Rule2024-29508
Partial Approval and Disapproval of Air Quality Implementation Plans; Arizona; Regional Haze State Implementation Plan for the Second Implementation Period and Prong 4 (Visibility) for the 2015 Ozone and 2012 Particulate Matter Standards
Primary source
Metadata and text below are from the Federal Register, a public-domain U.S. government work. Always verify the official published version before relying on it for any legal matter.
Published
December 18, 2024
Effective
January 17, 2025
Issuing agencies
Environmental Protection Agency
Abstract
The Environmental Protection Agency (EPA) is partially approving and partially disapproving the regional haze state implementation plan (SIP) revision submitted by Arizona on August 15, 2022 ("2022 Arizona Regional Haze Plan"), under the Clean Air Act (CAA) and the EPA's Regional Haze Rule (RHR) for the program's second implementation period. Arizona's SIP submission was developed to address the requirement that states must periodically revise their long-term strategies for making reasonable progress towards the national goal of preventing any future, and remedying any existing, anthropogenic impairment of visibility, including regional haze, in mandatory Class I Federal areas. The SIP submission also addresses other applicable requirements for the second implementation period of the regional haze program. Within this action, the EPA is also disapproving the visibility transport prong of Arizona's infrastructure SIP submittals for the 2012 annual fine particulate matter (PM<INF>2.5</INF>) and 2015 ozone National Ambient Air Quality Standards (NAAQS). The EPA is taking this action pursuant to CAA sections 110 and 169A.
Full Text
<html>
<head>
<title>Federal Register, Volume 89 Issue 243 (Wednesday, December 18, 2024)</title>
</head>
<body><pre>
[Federal Register Volume 89, Number 243 (Wednesday, December 18, 2024)]
[Rules and Regulations]
[Pages 102744-102773]
From the Federal Register Online via the Government Publishing Office [<a href="http://www.gpo.gov">www.gpo.gov</a>]
[FR Doc No: 2024-29508]
=======================================================================
-----------------------------------------------------------------------
ENVIRONMENTAL PROTECTION AGENCY
40 CFR Part 52
[EPA-R09-OAR-2024-0005; FRL-11919-02-R9]
Partial Approval and Disapproval of Air Quality Implementation
Plans; Arizona; Regional Haze State Implementation Plan for the Second
Implementation Period and Prong 4 (Visibility) for the 2015 Ozone and
2012 Particulate Matter Standards
AGENCY: Environmental Protection Agency (EPA).
ACTION: Final rule.
-----------------------------------------------------------------------
SUMMARY: The Environmental Protection Agency (EPA) is partially
approving and partially disapproving the regional haze state
implementation plan (SIP) revision submitted by Arizona on August 15,
2022 (``2022 Arizona Regional Haze Plan''), under the Clean Air Act
(CAA) and the EPA's Regional Haze Rule (RHR) for the program's second
implementation period. Arizona's SIP submission was developed to
address the requirement that states must periodically revise their
long-term strategies for making reasonable progress towards the
national goal of preventing any future, and remedying any existing,
anthropogenic impairment of visibility, including regional haze, in
mandatory Class I Federal areas. The SIP submission also addresses
other applicable requirements for the second implementation period of
the regional haze program. Within this action, the EPA is also
disapproving the visibility transport prong of Arizona's infrastructure
SIP submittals for the 2012 annual fine particulate matter
(PM<INF>2.5</INF>) and 2015 ozone National Ambient Air Quality
Standards (NAAQS). The EPA is taking this action pursuant to CAA
sections 110 and 169A.
DATES: This rule is effective on January 17, 2025.
ADDRESSES: The EPA has established a docket for this action under
Docket ID No. EPA-R09-OAR-2024-0005. All documents in the docket are
listed on the <a href="https://www.regulations.gov">https://www.regulations.gov</a> website. Although listed in
the index, some information is not publicly
[[Page 102745]]
available, e.g., Confidential Business Information (CBI) or other
information whose disclosure is restricted by statute. Certain other
material, such as copyrighted material, is not placed on the internet
and will be publicly available only in hard copy form. Publicly
available docket materials are available through <a href="https://www.regulations.gov">https://www.regulations.gov</a>, or please contact the person identified in the FOR
FURTHER INFORMATION CONTACT section for additional availability
information. If you need assistance in a language other than English or
if you are a person with a disability who needs a reasonable
accommodation at no cost to you, please contact the person identified
in the FOR FURTHER INFORMATION CONTACT section.
FOR FURTHER INFORMATION CONTACT: Michael Dorantes, Geographic
Strategies & Modeling Section (AIR-2-2), Planning & Analysis Branch,
Air and Radiation Division, EPA Region IX, 75 Hawthorne Street, San
Francisco, CA 94105; phone: (415) 972-3934; email:
<a href="/cdn-cgi/l/email-protection#e5818a97848b918096cb888c868d848089a5809584cb828a93"><span class="__cf_email__" data-cfemail="98fcf7eaf9f6ecfdebb6f5f1fbf0f9fdf4d8fde8f9b6fff7ee">[email protected]</span></a>.
SUPPLEMENTARY INFORMATION: Throughout this document, ``we,'' ``us'' and
``our'' refer to the EPA.
Table of Contents
I. Background
A. Regional Haze Plan for the Second Implementation Period
B. Prong 4 (Visibility) of the 2012 PM<INF>2.5</INF> NAAQS and
2015 Ozone NAAQS Infrastructure SIPs
II. Public Comments and EPA Responses
A. Comment Letter From Tri-State
B. Comment Letter From ADEQ
C. Comment Letter From the Chamber and AMC
D. Comment Letter From TEP
E. Comment Letter From SRP
F. Community Sign-On Letter
G. Comment Letter From NPCA et al.
III. Final Action
IV. Statutory and Executive Order Reviews
I. Background
A. Regional Haze Plan for the Second Implementation Period
On August 15, 2022,\1\ the Arizona Department of Environmental
Quality (ADEQ) submitted the 2022 Arizona Regional Haze Plan. ADEQ
supplemented its SIP revision on August 25, 2023, with nonpoint source
rules (``2023 Arizona Regional Haze Rules Supplement'').\2\ ADEQ made
these SIP submissions to address requirements of the CAA's regional
haze program pursuant to CAA sections 169A and 169B and 40 CFR 51.308.
---------------------------------------------------------------------------
\1\ Letter dated August 15, 2022, from Daniel Czecholinski,
Director, Arizona Department of Environmental Quality Air Quality
Division, to Martha Guzman, Regional Administrator, EPA Region IX
(submitted electronically August 15, 2022). On August 16, 2022, the
EPA determined that the SIP submittal met the completeness criteria
outlined in 40 CFR part 51, Appendix V. Letter dated August 16,
2022, from Elizabeth Adams, Director, Air and Radiation Division,
EPA Region IX, to Daniel Czecholinski, Director, Arizona Department
of Environmental Quality Air Quality Division.
\2\ Letter dated August 21, 2023, from Daniel Czecholinski,
Director, Arizona Department of Environmental Quality Air Quality
Division, to Martha Guzman, Regional Administrator, EPA Region IX
(submitted electronically August 25, 2023).
---------------------------------------------------------------------------
On May 31, 2024, the EPA published a notice of proposed rulemaking
proposing partial approval and partial disapproval of the 2022 Arizona
Regional Haze Plan SIP submission as partially satisfying the regional
haze requirements for the second implementation period contained in the
CAA and 40 CFR 51.308.\3\ We did not propose to act on the 2023 Arizona
Regional Haze Rules Supplement.
---------------------------------------------------------------------------
\3\ 89 FR 47398.
---------------------------------------------------------------------------
The EPA is now approving the elements of the 2022 Arizona Regional
Haze Plan related to requirements contained in 40 CFR 51.308(f)(1),
(f)(4)-(6), and (g)(1)-(5). The EPA is disapproving the elements of the
2022 Arizona Regional Haze Plan related to requirements contained in 40
CFR 51.308(f)(2), (f)(3), and (i)(2)-(4). Our proposed action and our
responses to comments in section II of this document contain more
information on the basis for this rulemaking and on our evaluation of
the submittal.
B. Prong 4 (Visibility) of the 2012 PM2.5 NAAQS and 2015 Ozone NAAQS
Infrastructure SIPs
Arizona submitted its infrastructure SIP submission for the 2012
PM<INF>2.5</INF> NAAQS on December 11, 2015 (``2015 PM<INF>2.5</INF> I-
SIP submittal'').\4\ Arizona also submitted its infrastructure SIP
submission for the 2015 ozone NAAQS on September 24, 2018 (``2018 Ozone
I-SIP submittal'').\5\
---------------------------------------------------------------------------
\4\ Letter dated December 11, 2015, from Eric Massey, Director,
Air Quality Division, ADEQ, to Jared Blumenfeld, Regional
Administrator, EPA Region IX.
\5\ Letter dated September 24, 2018, from Timothy S. Franquist,
Director, Air Quality Division, ADEQ, to Michael Stoker, Regional
Administrator, EPA Region IX (submitted electronically September 24,
2018).
---------------------------------------------------------------------------
Our May 31, 2024 proposed rulemaking action proposed to disapprove
the prong 4 portions of Arizona's 2018 Ozone I-SIP submittal and 2015
PM<INF>2.5</INF> I-SIP submittal. The EPA is now disapproving the Prong
4 elements of Arizona's 2018 Ozone I-SIP submittal and 2015
PM<INF>2.5</INF> I-SIP submittal. Our proposed action contains more
information on the basis for this rulemaking and on our evaluation of
the submittals.
II. Public Comments and EPA Responses
The EPA's May 31, 2024 proposed rulemaking action provided a 30-day
public comment period that would have ended on July 1, 2024. We
received four comments requesting an extension of the comment period.
On June 26, 2024,\6\ the EPA extended the comment period for the
proposed rulemaking action by 14 days in response to requests from
commenters. This action extended the close of the comment period to
July 15, 2024.
---------------------------------------------------------------------------
\6\ 89 FR 53372.
---------------------------------------------------------------------------
The EPA received an additional ten unique comments, including one
anonymous comment,\7\ two comments from private individuals,\8\ and
comment letters from Tri-State Generation and Transmission Association,
Inc. (``Tri-State''),\9\ ADEQ,\10\ the Arizona Chamber of Commerce and
Industry and the Arizona Manufacturers Council (``the Chamber and
AMC''),\11\ Tuscon Electric Power (TEP),\12\ the Salt River
Agricultural Improvement and Power District (SRP),\13\ 12 community
organizations (``Community Sign-on Letter''),\14\ and Earthjustice on
behalf of the National Parks Conservation Association, Sierra Club, and
the Coalition to Protect America's National Parks (``NPCA et
al.'').\15\ The anonymous comment and the comments from the private
individuals were unrelated to our proposed rulemaking. These three
comments do not require a response. We respond to the issues raised in
the seven remaining comment
[[Page 102746]]
letters received on our proposed rulemaking in this action.
---------------------------------------------------------------------------
\7\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0014">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0014</a>.
\8\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0015">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0015</a> and <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0019">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0019</a>.
\9\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0016">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0016</a>.
\10\ ADEQ submitted its comment letter twice. The letter is
available at both <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0017">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0017</a> and <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0018">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0018</a>.
\11\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0020">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0020</a>.
\12\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0021">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0021</a>.
\13\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0023">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0023</a>.
\14\ Available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0022">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0022</a>.
\15\ The comment letter and all Exhibits except for Exhibits 24
and 60 are available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0024">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0024</a>. Exhibits 24 and 60, including an emailed copy of
the NPCA et al.'s comment letter, are available at <a href="https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0025">https://www.regulations.gov/comment/EPA-R09-OAR-2024-0005-0025</a>.
---------------------------------------------------------------------------
A. Comment Letter From Tri-State
Sections I (``Introduction''), II (``Background Information on Tri-
State''), and III (``Tri-State Supports the Provisions of the Arizona
Department of Environmental Quality's State Implementation Plan
Submittal for which EPA has Proposed Approval'') of Tri-State's comment
letter either provide background information or are supportive of the
EPA's proposal and therefore do not require a response in our final
action. We respond to sections IV (``The Partial Disapproval is Overly
Vague and Should Be Reproposed with a Fulsome and Specific Explanation
of What EPA Finds Inconsistent with the Clean Air Act and the
Implementing Regulations''), V (``Additional Emission Reductions Are
Inappropriate for Springerville Unit 3''), and VI (``EPA's Reliance on
the July 8, 2021, Clarifications Memorandum Is Inappropriate Because It
Was Issued Only Days Prior to the Regional Haze State Implementation
Plans Being Due'') of Tri-State's comment letter below.
Comment A.1. Tri-State comments that the proposed rule,
specifically the partial disapproval, is overly vague and does not lend
itself to the general public being able to adequately understand what
the EPA finds objectionable regarding the SIP submission. Tri-State
asserts that the EPA made broad statements about divergences from the
Control Cost Manual and inadequacies in four-factor analyses and that
``it [is] impossible to understand with certainty the sources to which
these overly broad statements apply.'' The commenter also notes that
the proposed rulemaking action only gives limited examples of what the
EPA is referring to and asserts that the EPA needs to clearly provide
where it has issues with the SIP revision as it was submitted not just
examples of what it is concerned about. Specifically, Tri-State states
that the proposed rulemaking action makes general comments about ADEQ
conducting analyses for well-controlled sources to further reduce
emissions but makes no reference to which sources that the EPA finds to
be lacking in this area. Tri-State also comments that the vagueness of
the proposal is not consistent with the cooperative nature of the CAA
in the EPA working with states.
Response A.1. We do not agree that the proposal was overly vague or
that the basis for our proposed partial disapproval was unclear. The
proposal provided a detailed summary of the 2022 Arizona Regional Haze
Plan and the EPA's evaluation of the Plan with regard to each of the
applicable requirements of the CAA and the RHR. With respect to the
EPA's partial disapproval specifically, the proposal laid out multiple
reasons for the EPA's determination that the long-term strategy did not
fully meet the requirements of 40 CFR 51.308(f)(2), related to source
selection, four-factor analyses (specifically, controlled emissions
rates and deviations from the Control Cost Manual), and control
determinations (specifically, the application of cost thresholds, the
use of visibility as a factor to avoid controls, and the mass-based
emissions caps at Springerville Generating Station (SGS)).\16\ In each
instance, we provided at least one example of a unit to which this
reason applied.\17\ We also explained why the disapproval of the long-
term strategy also necessitated disapproval of the RPG and FLM
consultation elements.\18\
---------------------------------------------------------------------------
\16\ 89 FR 47398, 47428-47432.
\17\ Id.
\18\ Id. at 47432-47433 and 47435-47436.
---------------------------------------------------------------------------
We disagree with the commenter's suggestion that it was necessary
for the EPA to specify the extent to which each of the flaws identified
by the EPA applied to each unit considered by ADEQ. The EPA's role in
reviewing SIPs is to determine whether they meet all of the applicable
CAA requirements.\19\ In evaluating whether a SIP revision (or a
portion thereof) meets all of the applicable requirements, the EPA is
not required to separately evaluate and discuss each of the thousands
of pieces of information, analyses and determinations comprising the
SIP submission. Rather, the EPA may focus on those specific elements of
the SIP revision that form the basis for our determination that certain
applicable requirements are met and certain applicable requirements are
not met. Therefore, in this instance, it was reasonable for the EPA to
summarize our evaluation and cite to examples of where the State's
documentation, analyses, and determinations did not meet CAA
requirements, rather than separately evaluating and discussing every
such instance throughout the SIP revision.
---------------------------------------------------------------------------
\19\ See CAA section 110(k)(3) (``the Administrator shall
approve [a SIP] submittal as a whole if it meets all of the
applicable requirements of [the CAA]. If a portion of the plan
revision meets all the applicable requirements of [the CAA], the
Administrator may approve the plan revision in part and disapprove
the plan revision in part'').
---------------------------------------------------------------------------
For example, regarding source selection, our proposal stated that:
ADEQ did not provide an adequate justification for screening out
certain sources and units from conducting a four-factor analysis on
the basis that they are `effectively controlled' as part of its
source selection process. Specifically, in some cases, ADEQ did not
identify the controls for each pollutant at each unit or process,
the associated limits, or where the controls/limits currently exist
in the Arizona SIP. In other cases, ADEQ listed the controls, but
did not clearly explain why it is reasonable to assume, without
conducting a four-factor analysis, that no additional controls would
be reasonable.
We then provided examples of specific sources to which these
concerns applied. In addition, table 3 of the proposal listed all of
the units that ADEQ screened out as ``effectively controlled'' and
includes the entirety of ADEQ's rationale for each unit.\20\ This table
clearly shows that for many of the units, ADEQ did not list controls
for one or more of the three relevant pollutants (NO<INF>X</INF>,
SO<INF>2</INF>, and PM). Moreover, even for those units where ADEQ did
list this information, it did not explain why it is reasonable to
assume, without conducting a four-factor analysis, that no additional
controls would be reasonable. Furthermore, our proposed rulemaking
action clearly stated that ADEQ did not adequately explain whether any
of the existing controls for facilities evaluated within the SIP
submittal were necessary for reasonable progress and therefore a part
of the state's long-term strategy. Therefore, it was not necessary to
specifically identify each source that was deficient in this respect
because the deficiency applied to every source determined to be
effectively controlled.
---------------------------------------------------------------------------
\20\ Consistent with their labeling in the Plan, Appendix C.1,
Table 147, these rationales appear under the heading of
``Comments.''
---------------------------------------------------------------------------
Finally, regarding the cooperative nature of the CAA, we note that
the EPA worked extensively with ADEQ during SIP development. EPA and
ADEQ staff met on a monthly basis beginning in 2019 and continuing
through 2021 to discuss the development of the Plan. The EPA also
provided informal written feedback on various elements of the Plan
between 2019 and 2022. In these communications, the EPA identified many
of the flaws that are the basis for the partial disapproval, so ADEQ
was aware of the EPA's concerns prior to the EPA's proposal.
In conclusion, based on the findings discussed in our proposal and
elsewhere in this document, we find that the long-term strategy in the
2022 Arizona Regional Haze Plan does not meet the requirements of
51.308(f)(2) and we are disapproving the Plan with respect to this
requirement. Therefore, pursuant to CAA section 110(c), the EPA will be
required to develop a new
[[Page 102747]]
long-term strategy as part of a FIP, unless the EPA approves a
subsequent SIP submission that fully meets these requirements. That
long-term strategy would necessarily include updated source selection,
four factor analyses, and control determinations that address the
deficiencies we identified in the Plan. We are available to work with
ADEQ following this final action to develop a SIP revision, including
these elements.
Comment A.2. Tri-State asserted that additional emissions
reductions are ``inappropriate'' for Springerville Generating Station
(SGS) Unit 3. Tri-State made a few arguments in support of this
contention.
First, Tri-State noted that ADEQ evaluated the currently installed
NO<INF>X</INF> emissions controls against technically feasible
emissions controls and concluded that the current NO<INF>X</INF>
emissions controls constitutes best available control technology (BACT)
for coal-fired electric generating units (EGUs). For SO<INF>2</INF>,
Tri-State indicated that ADEQ's analysis for SGS Unit 3 clearly
demonstrates that Unit 3's emissions ranged from 0.069 to 0.090 lb/
MMBtu on an annual basis and has continuously complied with the Mercury
and Air Toxics Standard (MATS) SO<INF>2</INF> emissions standard of
0.20 lb/MMBtu, and therefore does not warrant further emissions
controls. Tri-State also noted language in the EPA's ``Guidance on
Regional Haze State Implementation Plans for the Second Implementation
Period'' (``2019 Guidance'') \21\ that describes scenarios in which the
EPA believes it may be reasonable for a state not to select a
particular source for further analysis for EGUs that have add-on flue
gas desulfurization (FGD) and that meets the applicable alternative
SO<INF>2</INF> emissions limit of the MATS rule for power plants.
---------------------------------------------------------------------------
\21\ Guidance on Regional Haze State Implementation Plans for
the Second Implementation Period. <a href="https://www.epa.gov/visibility/guidance-regional-haze-state-implementation-plans-second-implementation-period">https://www.epa.gov/visibility/guidance-regional-haze-state-implementation-plans-second-implementation-period</a>. The EPA Office of Air Quality Planning and
Standards, Research Triangle Park (August 20, 2019).
---------------------------------------------------------------------------
Second, Tri-State noted the visibility improvement at the Mount
Baldy Wilderness Area in the first planning period, and that Arizona
will have four additional 10-year planning periods to achieve the 1.3
deciview improvement needed to achieve natural conditions.
Finally, Tri-State stated that its intent is to retire SGS Unit 3
by September 15, 2031, and requested that the EPA include in the final
rule a provision to allow sources to work with their state regulatory
agencies to adopt an enforceable commitment to retire a unit, such as
through a permit condition, to allow the source to forego any further
emissions control or reduction requirements if a unit is retiring
within ten years of the Regional Haze SIP approval.
Response A.2. We partially agree with this comment. First, with
respect to NO<INF>X</INF> emissions at SGS Unit 3, ADEQ noted in the
Plan ``[t]he current controls ([low-NO<INF>X</INF> burners (LNB),
overfire air (OFA) with selective catalytic reduction (SCR)] represent
the most effective NO<INF>X</INF> control technologies for coal fired
EGUs and are estimated to achieve 85-95% removal efficiency.'' \22\ We
agree that the existing LNB, OFA and SCR constitute effective controls
for NO<INF>X</INF> at SGS Unit 3. As described in our proposal and in
response B.1 of this document, existing effective controls are
generally necessary to make reasonable progress and must be included in
the SIP, unless the state provides a weight-of-evidence demonstration
to justify that the existing effective controls are not necessary to
make reasonable progress. Because the Arizona SIP does not include a
NO<INF>X</INF> emissions limit corresponding to these controls for SGS
Unit 3, the State should have provided such demonstration.
Specifically, the State should have considered whether SGS Unit 3 is
subject to an enforceable NO<INF>X</INF> emissions limit that ensures
its NO<INF>X</INF> emissions rate will not increase. Without proper
justification that emissions of visibility impairing pollutants will
not increase, it is unclear how reasonable progress is being made
within the State's long-term strategy for the second planning
period.\23\
---------------------------------------------------------------------------
\22\ 2022 Arizona Regional Haze Plan, p. 219.
\23\ See CAA 169A(a)(1) ``Congress hereby declares as a national
goal the prevention of any future, and the remedying of any
existing, impairment of visibility in [Class I areas.]'' (emphasis
added).
---------------------------------------------------------------------------
Similarly, with respect to SO<INF>2</INF> emissions at SGS Unit 3,
we agree that, as described in the 2019 Guidance, an add-on FGD meeting
the appliable alternative SO<INF>2</INF> emissions standard under MATS
may constitute an effective control for SO<INF>2</INF>. However, as
noted in the previous paragraph, existing effective controls are
generally necessary to make reasonable progress and must be included in
the SIP, unless the state provides a weight-of-evidence demonstration
to justify that the existing effective controls are not necessary to
make reasonable progress. Because the Arizona SIP does not include a
SO<INF>2</INF> emissions limit corresponding to the existing
SO<INF>2</INF> controls at SGS Unit 3, the State should have provided
such a demonstration, including consideration of whether SGS Unit 3 is
subject to an enforceable SO<INF>2</INF> emissions limit that ensures
its SO<INF>2</INF> emissions rate will not increase.
Second, regarding the Mount Baldy Wilderness Area, although we
commend the efforts in Arizona that contributed to the noted visibility
improvement at the Mount Baldy Wilderness Area in the first planning
period, previous and ongoing measures are not automatically sufficient
to ensure ongoing reasonable progress. The regional haze requirements
of CAA sections 169A and 169B and 40 CFR 51.308 need to be
satisfied.\24\ In particular, the increment of progress that is
``reasonable progress'' for a given implementation period is determined
through the four statutory factors.\25\ While progress made in the
first implementation period, ongoing emissions trends, and anticipated
changes in emissions may inform a state's regional haze planning
process, these circumstances alone do not satisfy a state's obligation
to determine and include in its SIP the measures that are necessary to
make reasonable progress in the second planning period. We also note
that, while Mount Baldy is the closest Class I area to SGS and
therefore has the highest Q/d (emissions in tons per year divided by
distance to an affected Class I area in kilometers) value (339) with
respect to SGS,\26\ this does not mean that Mount Baldy is the only
Class I area affected by emissions from SGS. ADEQ did not specifically
identify all the Class I areas that may be affected by emissions from
each of the sources it evaluated. Nonetheless, given that SGS had a
2018 Q of 17,044 tons per year (tpy) \27\ and is located within 300 km
of 15 different Class I areas,\28\ it is likely to contribute to
visibility impairment at a number of Class I areas.
---------------------------------------------------------------------------
\24\ See 64 FR 35714 (July 1, 1999), 35721-35722 for additional
explanation as to the EPA's determination that emissions from all
States reasonably contribute to visibility impairment and thus are
subject to the regional haze regulations. Additionally, in the 2017
RHR, the EPA ``reiterat[ed] that the CAA requires States to consider
the four statutory factors . . . in each implementation period to
determine the rate of progress towards natural visibility conditions
that is reasonable for each Class I area.'' 82 FR 3078 (January 10,
2017), 3080.
\25\ 40 CFR 51.308(f)(2)(i).
\26\ Plan Appendix C, p. 21, Table 1. Q is calculated as the
total 2018 annual facility-wide NO<INF>X</INF>, SO<INF>2</INF>, and
PM<INF>10</INF> emissions in tpy, excluding processes determined by
ADEQ to be effectively controlled.
\27\ Id.
\28\ 79 FR 9318, 9360, Table 50 (February 18, 2014).
---------------------------------------------------------------------------
Finally, with respect to Tri-State's request regarding enforceable
shutdowns, we note that the EPA's role in acting on SIP submittals is
to evaluate whether they meet applicable CAA
[[Page 102748]]
requirement, not to establish new requirements. Tri-State may choose to
work with ADEQ to establish enforceable shutdowns as part of a
subsequent SIP revision. However, we note that even if there were an
enforceable shutdown in 2031, this would not automatically preclude the
unit from consideration under a four-factor analysis.\29\
---------------------------------------------------------------------------
\29\ See 2019 Guidance, pp. 20-21 (``It may be more challenging
for a state to reasonably use a shorter remaining useful life as the
basis for not selecting sources the further away the enforceable
shutdown date gets from 2028'').
---------------------------------------------------------------------------
Comment A.3. Tri-State comments that it was unreasonable for the
EPA to use the July 8, 2021 ``Clarifications Regarding Regional Haze
State Implementation Plans for the Second Implementation Period''
(``2021 Clarifications Memo'') as a basis for disapproval of the 2022
Arizona Regional Haze Plan. Tri-State indicated that although ADEQ
submitted its plan over a year late on August 22, 2022, ADEQ was over
three years into the process of developing the plan, working with the
Western Regional Air Partnership, the EPA, other states, Federal Land
Managers (FLMs) and members of the public, was nearly complete and
ready to go through the Arizona rulemaking process when the EPA
published the 2021 Clarifications Memo. Tri-State also noted that the
rulemaking process can take over a year due to various tasks required.
Response A.3. The EPA disagrees that it used the 2021
Clarifications Memo as a basis for disapproving portions of the 2022
Arizona Regional Haze Plan. Contrary to the commenter's suggestion, the
EPA's guidance, including the 2021 Clarifications Memo, is not the
basis for our disapproval. Rather, the partial disapproval is based on
the Plan's failure to satisfy the requirements of the relevant portions
of the RHR and CAA sections 169A and 169B. We did cite the guidance
documents because these documents provide helpful context explaining
the EPA's interpretations of the applicable statutory and regulatory
requirements against which we are required to evaluate SIP submittals.
Commenters are free to disagree and raise concerns with those
interpretations as part of the notice and comment process on individual
SIP actions. However, in this instance, the commenter does not appear
to object to any of the interpretations in the 2021 Clarifications
Memo, only to the fact that they were provided shortly before the due
date for the plans.
With regard to timing, we note that the 2021 Clarifications Memo
was developed in response to issues that EPA regions and other
stakeholders had raised regarding draft regional haze SIP revisions
that were already under development by states.\30\ The 2021
Clarifications Memo therefore necessarily came during the SIP
development process and, in comparison to the SIP-specific feedback
previously provided by the EPA prior to its issuance, was intended to
``offer feedback more broadly to help support SIP development,
submittal, review, and action for the second planning period.'' \31\
With regard to Arizona's Plan specifically, throughout the EPA's
collaboration with ADEQ during early engagement, EPA staff advised ADEQ
of many of the interpretations that would be expressed in the 2021
Clarifications Memo. Therefore, we do not agree that it was improper
for the EPA to cite to the 2021 Clarifications Memo as further
explanation for why portions of the 2022 Arizona Regional Haze Plan did
not comply with the applicable statutory and regulatory requirements.
---------------------------------------------------------------------------
\30\ 2021 Clarifications Memo, p. 1.
\31\ Id.
---------------------------------------------------------------------------
B. Comment Letter From ADEQ
Sections I (``Background'') and XI (``Conclusion'') of ADEQ's
comment letter are informational and therefore do not require a
response. We respond to sections II-X of ADEQ's comment letter below.
Comment B.1. ADEQ asserts that the EPA's changing guidance
increased the burden of ADEQ's planning efforts by introducing
uncertainty and rework. ADEQ noted delays between final publication of
the 2017 RHR and the 2019 Guidance, as well as later changes to the
EPA's interpretation of the RHR that came close to the plan submittal
deadline, including the 2021 Clarifications Memo. ADEQ asserts that the
state did not have the resources to undertake the evaluation of
existing control measures, as noted in the 2021 Clarifications Memo, a
process that ADEQ states was unreasonably broad-reaching and
duplicative.
Response B.1. We disagree with ADEQ's assertions about the EPA's
additional guidance increasing the burden of ADEQ's planning efforts by
introducing uncertainty and rework. First and foremost, as stated
previously and throughout this notice, the EPA did not rely on guidance
as the basis for its partial disapprovals. Rather, the 2019 Guidance
and the 2021 Clarifications Memo merely provide additional context to
the EPA's interpretations of the statutory and regulatory requirements.
Both the 2019 Guidance and the 2021 Clarifications Memo were drafted to
be used as tools by States in the development of their second planning
period regional haze plans. However, neither of these documents were
necessary for States to develop and submit their SIP revisions.
Regarding the contents of the guidance, we do not agree that the
EPA significantly changed its interpretations in either the 2019
Guidance or the 2021 Clarifications Memo. The commenter has not
provided any examples of interpretations that it believes were changed
under the 2019 Guidance and provided only a single example from the
2021 Clarifications Memo, relating to the section entitled,
``Determining When Existing Measures are Necessary for Reasonable
Progress.'' We do not agree that interpretations set forth in this
section of the 2021 Clarifications Memo represented a significant
change in interpretation. Rather, they were intended to clarify the
following statement in the 2019 Guidance:
If a state determines that an in-place emission control at a
source is a measure that is necessary to make reasonable progress
and there is not already an enforceable emission limit corresponding
to that control in the SIP, the state is required to adopt emission
limits based on those controls as part of its long-term strategy in
the SIP via the regional haze second planning period plan
submission.\32\
---------------------------------------------------------------------------
\32\ 2019 Guidance, p. 423.
Many states and other stakeholders raised questions about this
statement. In response, as part of the 2021 Clarifications Memo, the
EPA laid out in further detail our interpretation of the CAA and RHR
regarding how to make such a determination.
We also disagree, on multiple grounds, with the commenter's
assertion that the ``EPA's revised guidance requires an unreasonably
broad-reaching review of all existing control measures that are not
separately included in the regional haze plan to evaluate whether those
same measures should be duplicated in the regional plan to support
reasonable visibility progress.''
First, the 2021 Clarifications Memo did not establish any new
requirements. On the contrary, it clearly states that ``[t]his
memorandum does not change or substitute for provisions or requirements
of the CAA or RHR, nor does it create any new requirements. Rather,
this memorandum clarifies and provides further information on the
existing statutory and regulatory requirements.'' \33\ One of the key
requirements of the CAA and RHR is that all measures that are necessary
to make reasonable progress must be
[[Page 102749]]
included the SIP.\34\ However, neither the CAA, the RHR, or the 2019
Guidance explain how to determine whether an existing measure that
results from a four-factor analysis (or is used as the basis to avoid
such an analysis) is necessary to make reasonable progress. Based on
questions received on this subject during SIP development, it was clear
that further guidance on this question would be helpful. Accordingly,
the EPA provided this guidance in the 2021 Clarifications Memo.\35\
---------------------------------------------------------------------------
\33\ 2021 Clarifications Memo, p. 2.
\34\ CAA 169A(b)(2); 40 CFR 51.308(f)(2).
\35\ 2021 Clarifications Memo, pp. 8-9.
---------------------------------------------------------------------------
Regarding ``existing measures,'' the Memo explains:
When the outcome of a four-factor analysis is that no new
measures are reasonable for a source, the source's existing measures
are generally needed to prevent future visibility impairment (i.e.,
to prevent future emission increases) and thus necessary to make
reasonable progress. Measures that are necessary to make reasonable
progress must be included in the SIP.
However, there may be circumstances in which a source's existing
measures are not necessary to make reasonable progress.
Specifically, if a state can demonstrate that a source will continue
to implement its existing measures and will not increase its
emission rate, it may not be necessary to require those measures
under the regional haze program in order to prevent future emission
increases.
Similarly, with regard to existing ``effective controls'' used to
screen out sources from a four-factor analysis:
A decision to forgo a full four-factor analysis based on a
source's existing effective controls is equivalent to a
determination that no new measures are necessary to make reasonable
progress. In this scenario, existing effective controls are,
therefore, generally necessary to make reasonable progress and thus
must be adopted into the regulatory portion of the SIP. However, the
state may provide a weight-of-evidence demonstration as described in
Section 4.1 to justify that the existing effective control is not
necessary for reasonable progress.
Thus, the 2021 Clarifications Memo clarifies that, under the CAA
and the RHR, there is a general presumption that existing measures
resulting from a four-factor analysis (or relied upon to avoid such an
analysis) are necessary to prevent future visibility impairment and
therefore necessary to make reasonable progress. Accordingly, states
have the option to submit all such measures into the SIP (to the extent
they are not already approved into the SIP) without further evaluation
of whether the measures are necessary to make reasonable progress.
Alternatively, states may choose to provide a weight-of-evidence
demonstration that such measures are not necessary to make reasonable
progress. Importantly, however, such a demonstration is needed only
where an enforceable emissions limitation corresponding to an existing
measure has not already been approved into the SIP and is not being
submitted for SIP approval as part of the regional haze plan.
Therefore, we do not agree that a review of ``all existing control
measures that are not separately included in the regional haze plan''
is required.
Third, to the extent that a state chooses to undertake such a
demonstration that existing measures are not necessary to make
reasonable progress, we do not agree that it would be duplicative. On
the contrary, because such a demonstration is necessary only for
measures for which emissions limitations are not submitted into the
SIP, the state and the EPA need to evaluate relevant evidence
concerning whether the source will continue to implement its existing
measures and maintain its emissions rate in the absence of SIP-approved
requirements to do so, to ensure that visibility impairment does not
increase.
In sum, we disagree with ADEQ's characterization of the contents of
the 2021 Clarifications Memo and its role in our partial approval and
partial disapproval of the Plan.
Comment B.2. ADEQ comments that the EPA should not issue binding
decisions based on guidance alone where the bases for disapproval are
not in the rule or statute. ADEQ cites to statements in the 2019
Guidance and 2021 Clarifications Memo regarding screening out of
effectively controlled sources and determinations of whether existing
controls are necessary to make reasonable progress, as the examples of
guidance. Specifically, ADEQ asserts that ``[i]t was unreasonable for
the EPA's clarification memo to issue these additional specific
barriers to a determination that existing measures were effective at a
given source late in the development of second round regional haze
plans through guidance, and without additional notice and comment.''
Response B.2. We disagree that we issued our proposed decision
based on guidance. Rather, the proposal action clearly indicates that
the partial disapproval was based on failing to satisfy the
requirements of the relevant portions of the RHR and CAA sections 169A
and 169B, with citations to our guidance as further explanation. See
Response A.3 for more explanation.
We also note that ADEQ's comment appears to conflate two separate
questions: first, whether its justification for screening out a unit
based on existing effective emissions controls was sufficient, and
second, whether such existing effective control measures are necessary
to make reasonable progress. In our proposal, we discussed each of
these issues as two separate grounds for our proposed partial
disapproval with respect to 40 CFR 51.308(f)(2).\36\ While both
questions apply to sources that are screened out from a four-factor
analysis based on existing effective controls, they are distinct. The
first issue is further addressed in Response B.3 of this document,
while the second is addressed in Response B.1. Therefore, we do not
agree that the 2021 Clarifications Memo created any additional barriers
to a determination that existing measures were effective at a given
source.
---------------------------------------------------------------------------
\36\ See 89 FR 47398, 47428 (``ADEQ did not provide an adequate
justification for screening out certain sources and units from
conducting a four-factor analysis on the basis that they are
``effectively controlled'' as part of its source selection
process'') and 47431 (``ADEQ has not addressed whether any of the
existing measures relied upon in its four-factor analyses or its
`effective controls' determinations are necessary to make reasonable
progress and thus should be a part of the State's long-term strategy
for the second planning period.'').
---------------------------------------------------------------------------
Comment B.3. ADEQ disagreed with the EPA's proposed determination
that Arizona failed to provide adequate justification for deferring
certain emissions units from consideration. ADEQ indicated that the
state has the flexibility to reasonably select a set of sources for an
analysis of control measures, and that it did not exclude entire
facilities from consideration or exempt sources that had previously
adopted BART or reasonable progress controls, but rather excluded just
the emissions processes or units that recently installed highly
effective controls from the calculation of the Q/d value for that
facility. ADEQ also provided additional information regarding effective
controls in Attachment A of its letter.
Response B.3. We agree that states have flexibility to reasonably
select a set of sources for analysis of controls measures. However, as
described in our proposal, we find that ADEQ's approach to screening
out units from conducting a four-factor analysis on the basis that they
are ``effectively controlled'' was not adequately documented.\37\
Specifically, in some cases, ADEQ did not identify the controls for
each pollutant at each unit or process, the associated limits, or where
the controls and/or limits currently exist in the Arizona SIP. In other
cases, ADEQ listed the controls, but did not clearly explain why it is
[[Page 102750]]
reasonable to assume, without conducting a four-factor analysis, that
no additional controls would be reasonable.\38\ Accordingly, ADEQ
should have identified where the existing limits are found in the SIP
or FIP and clearly explained why no additional controls would likely be
reasonable under a four-factor reasonable progress analysis for the
second planning period. Without this analysis and explanation, it is
not clear what is a part of Arizona's long-term strategy for the second
planning period.
---------------------------------------------------------------------------
\37\ 89 FR 47398, 47428.
\38\ See 40 CFR 51.308(f)(2)(i) (``. . . The State must include
in its implementation plan a description of the criteria is used to
determine which sources or groups of sources it evaluated and how
the four factors were taken into consideration in selecting the
measures for inclusion in its long-term strategy''). See also 2021
Clarifications Memo, p. 5; 2019 Guidance, p. 23.
---------------------------------------------------------------------------
As noted in its comment, ADEQ conducted its effective controls
screening on a unit-specific basis. However, it did not do so on a
pollutant-specific basis. Rather, ADEQ screened out entire units from
further evaluation for NO<INF>X</INF>, SO<INF>2</INF>, or
PM<INF>10</INF> if the units met ADEQ's screening criteria for any one
of these pollutants. We find that this approach was unreasonable
because it resulted in the screening out of entire units without
consideration of whether the unit had effective controls for all three
of the pollutants covered in ADEQ's long-term strategy. For example,
ADEQ screened out AEPCO Apache Unit 3 and TEP Irvington Generating
Station (IGS) Unit 4 from any further analysis because these units were
converted from coal to natural gas under better-than-BART alternatives
during the first planning period.\39\ The EPA acknowledges that fuel
combustion units that are required to combust pipeline-quality natural
gas are generally considered to be effectively controlled for
SO<INF>2</INF> and PM.\40\ However, they are not necessarily
effectively controlled for NO<INF>X</INF>, based on burning natural gas
alone. Therefore, we find that ADEQ has not provided adequate
justification for screening these units out from an analysis of
NO<INF>X</INF> controls.
---------------------------------------------------------------------------
\39\ 2022 Arizona Regional Haze Plan, Appendix C, Table 147.
\40\ See 2019 Guidance, p. 24.
---------------------------------------------------------------------------
Additionally, we appreciate the documentation in Attachment A that
ADEQ provided in its attachment letter. However, this information would
need to be part of a SIP revision subject to review by the public and
FLMs in order for the EPA to consider it as part of the long-term
strategy. If ADEQ develops a new SIP revision intended to remedy the
deficiencies discussed in our proposed and final actions on the Plan,
it may be appropriate to include this information in that SIP revision.
Comment B.4. ADEQ asserts that, despite indicating that flawed
emissions rates were used for some of ADEQ's analyses, the proposed
action identified one example of differing achievable emissions rates
for selective catalytic reduction (SCR) and selective non-catalytic
reduction (SNCR) controls for SGS Units 1 and 2's four factor analyses.
In this example, ADEQ used 0.060 lb/MMBtu and 0.15 lb/MMBtu as
reasonable estimates of the achievable rates at TEP SGS Units 1 and 2
for SCR and SNCR, respectively.
For SCR, ADEQ additionally states the study that the EPA cited for
its justification that SCR has been demonstrated to achieve 0.05 lb/
MMBtu (or up to 90 percent reduction) was published in 2005 and found
that the 20 SCR systems examined in 2003 achieved NO<INF>X</INF>
emissions rates between 0.04 and 0.07 lb/MMBtu.\41\ In addition, ADEQ
noted that EPA found 0.065 lb/MMBtu as a ``reasonable estimate of
average SCR performance'' in its 2016 Regional Haze FIP action for the
Salt River Project Coronado Generating Station Unit 1.
---------------------------------------------------------------------------
\41\ Ravi K. Srivastava, Robert E. Hall, Sikander Khan, Kevin
Culligan & Bruce W. Lani (2005) Nitrogen Oxides Emission Control
Options for Coal-Fired Electric Utility Boilers, Journal of the Air
& Waste Management Association, 55:9, 1367-1388, DOI: 10.1080/
10473289.2005.10464736. Available at: <a href="https://doi.org/10.1080/10473289.2005.10464736">https://doi.org/10.1080/10473289.2005.10464736</a>.
---------------------------------------------------------------------------
For SNCR, ADEQ additionally states that the EPA did not provide a
technical citation for disagreeing with ADEQ's use of 0.15 lb/MMBtu for
SNCR. ADEQ noted that the proposed rule indicated that ADEQ did not
demonstrate why source specific conditions would cause SNCR on these
units to achieve as little as a 15 percent reduction. ADEQ noted that
Srivastava et al. found that while smaller boilers (e.g., 76-78 MW
units) were able to achieve greater than 60 percent NO<INF>X</INF>
reductions, larger boilers (e.g., 500 MW units) ``may be capable of
achieving reductions of only ~30%.'' ADEQ indicates that SGS Units 1
and 2 units have nameplate ratings of 425 MW and would be expected to
achieve less reductions than smaller units. ADEQ also points to the
inlet concentration as another consideration for achievable emissions
rates with post combustion emissions control. Citing Srivastava et al.,
ADEQ notes that the study found that ``in the absence of reliable SCR
inlet NO<INF>X</INF> data, the SCR efficiencies are estimated using an
inlet NO<INF>X</INF> level of 0.5 lb/10\6\ Btu.'' However, in the case
of SGS Units 1 and 2, the NO<INF>X</INF> concentration in the exhaust
from these units is less than 100 parts per million by volume (ppmv)
with an assumed rate of 0.174 lb/MMBtu and 0.178 lb/MMBtu being used in
ADEQ's four factor analysis cost calculations for Unit 1 and Unit 2,
respectively. Given the already low NO<INF>X</INF> inlet concentration,
an achievable emissions rate of 0.15 lb/MMBtu was determined to be
reasonable. ADEQ further noted that additional information related to
achievable emissions rates for SNCR for SGS Unit 1 and Unit 2 can be
found in Appendix K, Section II(J), Comment 10 of the 2022 Arizona
Regional Haze Plan.
Response B.4. We acknowledge that we only highlighted a single
example of flawed emissions rates in our proposal--for NO<INF>X</INF>
at SGS Units 1 and 2. However, as explained in response A.1, in
evaluating whether a SIP revision (or a portion thereof) meets each of
these CAA requirements, the EPA is not required to separately evaluate
and discuss each of the thousands of pieces of information, analyses
and determinations comprising the SIP submission. Rather, the EPA may
focus on those specific elements of the SIP revision that form the
basis for our determination that certain applicable requirements are
met and certain applicable requirements are not met. In this instance,
we focused on these units because they are expected to have the highest
NO<INF>X</INF> emissions of any units in the State (2,099 and 2,283 tpy
respectively) by 2028, so it is important to carefully examine whether
additional NO<INF>X</INF> reductions from these units are necessary to
make reasonable progress. Given this context, and for the reasons
outlined below, we do not agree with ADEQ that it has adequately
documented the emissions rates assumed in this analysis, which are a
critical component of a four-factor analysis.
The emissions rate achievable by a unit equipped with SCR is
determined by several parameters and technological limitations. There
are periods of operation in which the SCR is not able to operate,
particularly during periods of startup and shutdown.\42\ The SCR
emissions rate (lb/MMBtu) achievable by a particular unit represents
the combination of two primary elements: (1) the controlled
NO<INF>X</INF> emissions rate during periods of normal unit operation
when the SCR is able to operate, and (2) the uncontrolled
NO<INF>X</INF> emissions rate
[[Page 102751]]
during periods of startup and shutdown when the SCR system cannot
operate, and result in unit emissions higher than the SCR controlled
emissions rate that increase the unit's overall emissions rate.
---------------------------------------------------------------------------
\42\ Control Cost Manual, Section 4, Chapter 2 Selective
Catalytic Reduction (June 2019), section, 2.2.1 Reduction Chemistry,
Reagents, and Catalyst, available at <a href="https://www.epa.gov/sites/default/files/2017-12/documents/scrcostmanualchapter7thedition_2016revisions2017.pdf">https://www.epa.gov/sites/default/files/2017-12/documents/scrcostmanualchapter7thedition_2016revisions2017.pdf</a>.
---------------------------------------------------------------------------
With regard to the first element, the 2005 study is not the sole
basis for our finding that an overall emissions rate of 0.050 lb/MMBtu
is achievable with SCR on an annual average basis. There are multiple
instances of coal-fired units installing SCR on a retrofit basis and
achieving 0.050 lb/MMBtu in practice on an annual average basis.\43\
These units are typically able to achieve this overall level of control
by being able to operate at NO<INF>X</INF> annual emissions rates at or
below 0.050 lb/MMBtu based upon periods of normal operation. Even
several of the units identified by ADEQ operating in the annual average
emissions rate range of 0.055 to 0.065 lb/MMBtu are still achieving
emissions rates of 0.050 lb/MMBtu and lower based upon periods of
normal operation.\44\ We consider this information sufficient to
establish that an 0.050 lb/MMBtu emissions rate warrants consideration
as technically feasible for coal fired units generally during periods
of normal operation, absent source specific factors affecting
feasibility. We are not aware of assertions by either ADEQ or TEP that
the Springerville units specifically cannot achieve 0.050 lb/MMBtu when
operating with SCR during periods of normal operation.\45\ Therefore,
we find that ADEQ should have considered a controlled NO<INF>X</INF>
emissions rate of 0.050 lb/MMBtu for SGS Units 1 and 2 when operating
with SCR during periods of normal operation.
---------------------------------------------------------------------------
\43\ See Docket Item F-16 ``SCR Tangentially fired Coal.xlsx.''
\44\ Id.
\45\ TEP has indicated that vendors have been unwilling to
provide manufacturer guarantees of 0.050 lb/MMBtu over the lifetime
of the SCR system. Because manufacturer guarantees include
contractual and financial considerations beyond technical
performance of the SCR system, we do not consider an inability to
secure a manufacturer guarantee to constitute a determination that
an emission rate of 0.050 lb/MMBtu is not technically feasible,
particularly with regard to periods of normal operation.
---------------------------------------------------------------------------
The majority of analysis performed by ADEQ is relevant to the
second element and is intended to support a position that, when the
annual emissions rate achievable during normal operations is combined
with emissions from the number of startup/shutdown cycles exhibited by
SGS Units 1 and 2, an annual average emissions rate of 0.06 lb/MMBtu is
what is reasonably achievable for these units. We consider it
appropriate to account for the effect of startup/shutdown emissions on
the emissions rate achievable by the unit, but disagree that the
analysis provided by ADEQ supports an annual average emissions rate of
0.060 lb/MMBtu. TEP's four factor analysis and ADEQ's SIP submittal did
not include startup/shutdown history to support the assertion that
Springerville has more startup/shutdown events than comparable
tangentially-fired coal fired boilers. A review of Clean Air Markets
Program Data (CAMPD) emissions and operating data over a 2021-2023
timeframe indicate that SGS Units 1 and 2 each experienced
approximately 9 to 14 startup events per year. While we acknowledge
that some portion of SGS baseline emissions consist of startup/shutdown
emissions that cannot be controlled by an SCR system, the substantial
majority of baseline emissions are attributable to emissions during
normal operation. We estimate that approximately 97-98% of baseline
emissions are attributable to normal operations that could be
controlled by SCR.\46\ Given that the majority of unit emissions can be
controlled by SCR to 0.050 lb/MMBtu or lower and that the remaining 2-3
percent of operations are characterized by low inlet SCR emissions
rates, we do not consider the historical startup/shutdown operating
profile to support deviating to an 0.06 lb/MMBtu emissions rate on an
annual average basis.
---------------------------------------------------------------------------
\46\ See Docket Item F-17 [SGS CAMPD 2021-23.xlsx]. To
illustrate SCR control potential during periods of normal operation,
this spreadsheet is based on the key assumption that emissions from
any day a unit did not operate a full 24 hours would be attributable
to startup/shutdown periods. An hourly analysis would provide a more
refined and precise assessment, though we consider this assumption
to overestimate the emissions attributable to startup/shutdown by
including all emissions from partial operating days towards startup/
shutdown.
---------------------------------------------------------------------------
ADEQ cites a limit of 0.065 lb/MMBtu established for SRP Coronado
Unit 1 in a 2016 Regional Haze FIP action as support for the use of an
annual average emissions rate of 0.060 lb/MMBtu. We wish to clarify
that the 0.065 lb/MMBtu value was not the annual average emissions rate
used in cost calculations, but was the emissions limit established on a
rolling 30-boiler operating day (BOD) average, for Coronado Unit 1. We
relied upon an 0.050 lb/MMBtu annual average emissions rate in
developing cost calculations for SCR,\47\ which is consistent with our
action here. In establishing a rolling 30-BOD limit for Coronado Unit 1
in that action, we acknowledged that upward revisions to the SCR design
rate achievable on an annual average basis would be appropriate in
order to accommodate the effect that multiple startup/shutdown events
would have to overall unit emissions rates on an averaging period that
could be as short as 30 days. Based upon startup/shutdown frequency and
projected controlled emissions rate information provided by SRP, we
finalized 0.065 lb/MMBtu as an appropriate emissions limit and
reasonable estimate of SCR performance over a short-term period.\48\
Given that the 0.065 lb/MMBtu limit reference here corresponds to a
different, shorter averaging period, and was itself based on an 0.05
lb/MMBtu annual average emissions rate, we do not consider it
supportive of an 0.060 lb/MMBtu annual average emissions rate.
---------------------------------------------------------------------------
\47\ See Docket Item EPA-R09-OAR-2012-0021-0204 for cost
calculation details. More information related to establishing a 30
BOD limit relative to an annual emissions rate can be found in our
March 31, 2015 (80 FR 17010) proposed reconsideration, including
Docket Items EPA-R09-OAR-2015-0165-0029 through -0033 for further
details. We note that SRP identified an SCR design target during
periods of normal operation as low as 0.03 lb/MMBtu for Coronado
Unit 1, though we acknowledge there are source specific differences
with the SGS units.
\48\ It is unclear to what extent a comparable 30-BOD limit may
be appropriate for the SGS units, but we note that the SGS units
appear to have historically had at least as many startup/shutdown
events as Coronado Unit 1.
---------------------------------------------------------------------------
For SNCR, we appreciate the additional analysis provided in the
comment and citation to Appendix K summarizing ADEQ's responses to
public comments. We acknowledge that low inlet NO<INF>X</INF>
concentrations are a general consideration in evaluating NO<INF>X</INF>
controls that can negatively impact control efficiencies and achievable
controlled emissions rates. Therefore, in order to further evaluate
whether a rate lower than 0.15 lb/MMBtu may be achievable with SNCR at
SGS Units 1 and 2 on an annual basis, we examined CAMPD emissions data
over a 2019-2023 time period for SNCR-equipped units comparable to SGS
Units 1 and 2, specifically filtering for tangentially-fired coal units
operating with SNCR on a retrofit basis.\49\ We identified four
currently operating SNCR-equipped units achieving NO<INF>X</INF>
emissions rates below 0.15 lb/MMBtu, ranging between 0.10 to 0.12 lb/
MMBtu.\50\ These values represent the highest performing SNCR-equipped
units, with the next best performing units operating at emissions rates
of 0.15 lb/MMBtu and higher. At least one of the four units we
identified has the capability to use natural gas,
[[Page 102752]]
which could have the effect of lowering emissions rates relative to
units that do not have this capability, such as SGS Units 1 and 2.\51\
Based on the three remaining units, each unit had relatively low pre-
SNCR emissions rates that are comparable to SGS Unit 1 and 2, and each
unit is able to achieve SNCR emissions rates below 0.15 lb/MMBtu and
control efficiencies better than 15 percent. Neither ADEQ or TEP has
provided documentation to support a claim that SGS Units 1 and 2
specifically cannot achieve an annual emissions rate lower than 0.15
lb/MMBtu with SNCR. Therefore, we find that ADEQ should have considered
a NO<INF>X</INF> emissions rate of lower than 0.15 lb/MMBtu for SNCR at
SGS Units 1 and 2.
---------------------------------------------------------------------------
\49\ See Docket Item F-21, ``SNCR Tangential Coal Units.xlsx.''
\50\ See Docket Item F-21. These units include Boswell Energy
Center (MN) Unit 4, Will County (IL) Unit 4, and Jeffrey Energy
Center (KS) Units 2 and 3.
\51\ Based on reported natural gas fuel usage, the Boswell
Energy Center appears to use natural gas primarily as a startup
fuel, but does periodically use quantities of natural gas that
suggest co-firing with coal for electricity generation purposes.
---------------------------------------------------------------------------
We also note that this was one of multiple flaws that formed the
basis of our determination that the State's long-term strategy did not
satisfy the requirements of 40 CFR 51.308(f)(2), including reasons
related to source selection and control determinations, as detailed in
our proposal and elsewhere in this document. Therefore, even assuming
that a control efficiency of 15 percent for SNCR at SGS Units 1 and 2
was reasonable, it would not have changed our determination that the
2022 Arizona Regional Haze Plan did not satisfy the requirements of 40
CFR 51.308(f)(2).
Comment B.5. ADEQ disagrees with the EPA's determination that
Arizona deviated from the Control Cost Manual without documentation as
part of its four factor analyses with regards to remaining useful life
calculations for the El Paso Natural Gas (EPNG) Williams facility and
the use of source specific interest rates without providing adequate
documentation in the control measure analyses for the EPNG Williams and
Willcox facilities.
ADEQ indicates that the EPA Control Cost Manual Section 4, Chapter
2, states that ``. . . a representative value of the equipment life for
SCR at power plants can be considered as 30 years . . . [f]or other
sources, the equipment life can be between 20 and 30 years.'' ADEQ
noted that while it erroneously omitted this citation from the EPNG
Williams Turbine analysis, the Control Cost Manual citation and
justification for use of 25 years is the midpoint between the 20-30
year range for non-EGU SCR systems and was included in the EPNG Willcox
SCR analysis for Turbines 1 and 2.\52\ ADEQ also noted that it received
cost calculation spreadsheets utilizing a 25-year useful life for SCR
for the EPNG Williams Turbine 1 from the US Forest Service.\53\
---------------------------------------------------------------------------
\52\ 2022 Arizona Regional Haze Plan, Appendix C, Section
C3.8.5.2, Page 146.
\53\ 2022 Arizona Regional Haze Plan, Appendix L, Section 4.2.4,
Comment 14.
---------------------------------------------------------------------------
ADEQ also explains that in its analysis of remaining useful life
for compressor engines at the EPNG Williams facility, ADEQ documented
the assumptions and basis for using 20 years to amortize NO<INF>X</INF>
controls in Appendix C, Section C3.7.6.5, which includes citations to
the Control Cost Manual and the EPA's 2016 technical support document
for the Cross State Air Pollution Rule for the 2008 Ozone NAAQS.
Response B.5. We appreciate the clarification regarding ADEQ's
reasoning for use of a 25-year remaining useful life for the EPNG
Williams turbines and 20 years for the EPNG Williams engines. While
there are instances of combustion turbines with operating lifetimes
beyond 25 years (with or without retrofit controls), we acknowledge
that EPA guidance such as the Control Cost Manual has not recommended a
value beyond the 20-30 year range. ADEQ's use of a 25-year useful life
represents the midpoint of Control Cost Manual recommendations, and
therefore we agree that it is consistent with the Control Cost Manual.
However, we note that the lack of documentation of remaining useful
life for the units at Williams Compressor Station was one of the many
flaws that we identified in the state's long-term strategy including
reasons related to source selection and control determinations, as
detailed in our proposal and elsewhere in this document. Therefore,
this clarification does not change our determination that the 2022
Arizona Regional Haze Plan did not satisfy the requirements of 40 CFR
51.308(f)(2).
Comment B.6. ADEQ disagrees with the EPA's determination that ADEQ
did not reasonably weigh the statutory factors in reaching its control
determinations with regards to application of cost thresholds. ADEQ
disagrees that Arizona's consideration of incremental cost
effectiveness in its four factor analyses were done in an unreasonable
manner. ADEQ cites to its explanation that the incremental cost of
requiring low-emission combustion 2 (LEC 2) as opposed to Air-Fuel
ratio adjustments is $5,034/ton, which ADEQ considered reasonable, and
therefore ADEQ found that LEC 2 is a more appropriate control for
Williams Reciprocating Engine 1 (RECIP-1).\54\
---------------------------------------------------------------------------
\54\ 2022 Arizona Regional Haze Plan, Appendix C, Section
C3.7.6.2, page 129.
---------------------------------------------------------------------------
ADEQ also asserts that it also analyzed other determinations from
the regional haze first implementation period besides the incremental
cost effectiveness value for the Nelson Lime Plant action, and ADEQ
provides this information in Table 1 of its comment letter.
Response B.6. We appreciate ADEQ's explanation about LEC 2 on
RECIP-1, but we note that ADEQ also rejected LEC 3 on Williams RECIP-1,
on the grounds that the incremental costs of these controls, relative
to less stringent controls, were excessive. In addition, ADEQ did not
provide or consider incremental cost effectiveness values for the same
controls for the other units at the same source (RECIP-2 or RECIP-5).
Although states may choose to consider incremental costs in a
reasonable manner,\55\ we find it was unreasonable for ADEQ to do so
only for specific units and controls, rather than in a consistent
manner across all units and controls. Such inconsistent treatment of
sources without explanation is the ``the hallmark of arbitrary
action.'' \56\
---------------------------------------------------------------------------
\55\ See, e.g., 2019 Guidance, p. 40.
\56\ Nat'l Parks Conservation Ass'n v. EPA, 788 F.3d 1134, 1145
(9th Cir. 2015).
---------------------------------------------------------------------------
We also appreciate the addition of Table 1 identifying other first
implementation period incremental cost effectiveness decisions.
However, this information was not included in the Plan and therefore
not subject to review by the public or FLMs. Accordingly, it cannot be
relied upon to meet the requirement of 40 CFR 51.308(f)(2)(iii) for
States to document the technical basis for their long-term strategy.
Moreover, even if the information had been included in the Plan, it
would not have justified ADEQ's inconsistent consideration of
incremental cost effectiveness, for the reasons described in our
proposal and the preceding paragraph.
Comment B.7. ADEQ states that, contrary to the EPA's contention,
ADEQ did not rely upon visibility benefits for its control
determinations, but rather visibility impacts were reported for some
sources to give reference to the reader as to the relative impact of
these sources or controls on visibility. ADEQ also asserts that nothing
in the CAA, RHR, or 2019 Guidance prevents the department from
considering visibility benefits as part of its analysis and stated that
``the ADEQ's labeling of the visibility benefits associated with
specific control scenarios as `small'
[[Page 102753]]
comports with similar language used by the EPA in their regional haze
actions.'' \57\
---------------------------------------------------------------------------
\57\ Citing e.g., ``relatively small visibility benefits'' in 79
FR 52419, 52439 (September 3, 2014).
---------------------------------------------------------------------------
Response B.7. We agree with ADEQ that in its Response to Comments,
ADEQ stated that ``[v]isibility impacts were reported for some sources
to give reference to the reader as to the relative impact of these
sources or controls considered on visibility. However, this information
was not considered in the Department's emissions control measure
determinations.'' \58\ However, this assertion is contradicted by the
language of some of the control determinations in the Plan. For
example, in the NO<INF>X</INF> four-factor analysis for SGS, ADEQ
stated that:
---------------------------------------------------------------------------
\58\ Id., Appendix K, p. 9.
ADEQ does find visibility impacts a useful consideration given
the goal of the regional haze program is to improve visibility in
Class I areas. As such, ADEQ reports modeled visibility impacts in
this documentation. The small modeled visibility benefits associated
with the modeled hypothetical NO<INF>X</INF> emission reduction
supports the determination that no additional NO<INF>X</INF>
controls are necessary to make reasonable progress towards natural
visibility at Class I areas during this implementation period.\59\
---------------------------------------------------------------------------
\59\ Plan, Appendix C, p. 221 (emphases added).
Similar language appears following the SO<INF>2</INF> four factor
analysis for SGS \60\ and in the summary of ADEQ's NO<INF>X</INF>
reasonable progress determination for IGS Unit 3.\61\ This language
indicates that ADEQ did consider visibility for these units, and
specifically, that it weighed the ``small modeled benefits'' of
controls in determining that no additional NO<INF>X</INF> controls were
warranted at SGS Units 1 and 2 and IGS Unit 3, and no more stringent
SO<INF>2</INF> controls were warranted at SGS Units 1 and 2.
---------------------------------------------------------------------------
\60\ Id. at 234 (``The small visibility benefits associated with
the modeled SO<INF>2</INF> controls supports the determination that
CDS and wet FGD control options are not necessary to make reasonable
progress towards natural visibility at Class I areas during this
implementation period.'').
\61\ Id. at 197 (``The small modeled visibility benefits
associated with additional controls support the determination that
no additional controls are necessary to make reasonable progress
towards natural visibility at Class I areas during this
implementation period.'').
---------------------------------------------------------------------------
While states have the option to consider visibility benefits, along
with the four statutory factors, in making control determinations, if
they choose to do so, they must do so ``in a reasonable way that does
not undermine or nullify the role of the four statutory factors in
determining what controls are necessary to make reasonable progress.''
\62\ In this case, we find it was not reasonable for ADEQ to consider
visibility benefits only for specific sources and without any
explanation of what would constitute a significant visibility benefit.
---------------------------------------------------------------------------
\62\ 2021 Clarifications Memo, p. 12 (quoting Response to
Comments on Protection of Visibility: Amendments to Requirements for
State Plans; Proposed Rule at 186).
---------------------------------------------------------------------------
In the action cited by ADEQ where the EPA considered ``relatively
small visibility benefits,'' \63\ we were comparing the relatively
small benefits of a control at one source to the relatively larger
visibility benefits expected to result from controls at other
sources.\64\ In contrast, in the 2022 Arizona Regional Haze Plan, ADEQ
did not find any visibility benefits at any source to be anything other
than small. Thus, as explained in our proposal regarding SGS Units 1
and 2, ``[i]n the absence of any opportunities for larger emissions
reductions and corresponding visibility benefits, we find that ADEQ's
reliance on `small' visibility benefits as an additional justification
for not adopting more stringent controls at these units is not
persuasive.'' \65\
---------------------------------------------------------------------------
\63\ 79 FR 52420, 52439.
\64\ See, e.g., id. at 52442 (referring to ``large visibility
benefits) and 52458 (``we consider this visibility benefit
sufficient to support installation of controls.'').
\65\ 89 FR 47398, 47430.
---------------------------------------------------------------------------
Finally, we note that, even assuming that ADEQ did not consider
visibility benefits as part of its control determinations for SGS Units
1 and 2, we would still conclude that the determinations were flawed
for other reasons. In particular, for NO<INF>X</INF>, ADEQ did not
adequately justify the control efficiency used for SCR and SNCR, as
discussed in the proposal and response B.4.\66\ For SO<INF>2</INF>,
ADEQ unreasonably rejected wet FGD on the basis of incremental cost,
and set mass-based caps that will not ensure implementation of the
emissions reduction measures that are necessary to make reasonable
progress at these units, as discussed in the proposal and response B.8.
Furthermore, the lack of clarity in the Plan regarding whether or not
ADEQ considered visibility benefits in making its control
determinations for SGS Units 3 and 4 and IGS Unit 3, indicates a lack
of reasoned decision making that also supports our disapproval of the
Plan's long-term strategy.
---------------------------------------------------------------------------
\66\ 89 FR 47398, 47428 (May 31, 2024).
---------------------------------------------------------------------------
Comment B.8. ADEQ disagrees with the EPA's determination that ADEQ
did not reasonably weigh the statutory factors in reaching its control
determinations with regards to three issues noted in the proposed rule
concerning the mass-based emissions caps at SGS and IGS.
First, ADEQ asserts that the EPA's rationale is arbitrary and
capricious in regard to the rejection of wet FGD for SGS Units 1 and 2.
ADEQ states that the use of emissions limits in lieu of codifying
specific control technologies is a flexibility that the EPA itself used
in its reasonable progress determination for Phoenix Cement Clarkdale
(PCC).\67\ ADEQ asserts that the EPA established an emissions limit for
PCC that did not require the installation of a particular control
technology but rather compliance through other means of meeting the
limit.
---------------------------------------------------------------------------
\67\ 79 FR 52420, 52460.
---------------------------------------------------------------------------
Second, regarding the EPA's contention that spray dry absorber
(SDA) upgrades may still be cost effective after the establishment of
the mass-based emissions caps, ADEQ comments that this manner of
analysis is not contemplated in the four-factor analysis as outlined in
the RHR or the 2019 Guidance. ADEQ asserts that the EPA has never
applied this standard whereby after the establishment of an emissions
limit based on the reductions achievable from a considered control
technology that a State must revisit and update the baseline emissions
of its four-factor analysis to reflect the new emissions limit. ADEQ
claims that for its analysis of SGS, ADEQ did not select a control
scenario in its four-factor analysis that included the imposition of
both emissions limits and the installation of SDA upgrades, and
therefore, the EPA should not substitute its judgment for ADEQ's
selection of SDA upgrades as the evaluated control measure for SGS
Units 1 and 2 or reject ADEQ's determination based on an arbitrary and
circular four factor analysis standard.
Third, ADEQ further disagrees with the EPA's assertion that the
mass-based emissions caps at SGS and IGS would not meaningfully
constrain the emissions from one unit during periods when the other
unit is not operating and argues that the rationale is arbitrary and
capricious. ADEQ notes that the EPA referenced TEP's 2023 Integrated
Resources Plan (IRP) and highlighted TEP's plans to retire SGS Unit 1
in 2027, but states that as the operating scenarios outlined in the IRP
are not federally enforceable conditions, ADEQ has no basis for the
consideration of these future scenarios as part of its control measure
analysis and the establishment of the mass-based emissions limits. ADEQ
states that the EPA should not rely upon unenforceable and hypothetical
operating scenarios to reject ADEQ's reasonable progress
determinations.
Lastly, ADEQ disagrees with the EPA's rationale that IGS Unit 3's
mass-
[[Page 102754]]
based emissions limits are not yet enforceable and therefore are not an
appropriate basis for modifying the baseline control scenario for a
four-factor analysis. ADEQ cites that for the EPA's BART determination
for Arizona Public Service (APS) Cholla generating station, the EPA
accepted a source specific permit revision for APS Cholla Unit 2 that
included a trigger that was conditional on the EPA's approval of the
SIP revision that altered the remaining useful life of the unit in
ADEQ's four factor analysis.\68\ Therefore, ADEQ concludes that the EPA
should approve ADEQ's reasonable progress determination for IGS Unit 3.
---------------------------------------------------------------------------
\68\ 81 FR 46852, 46860 (July 19, 2016).
---------------------------------------------------------------------------
Response B.8. We disagree with ADEQ's comments arguing that the
EPA's justification for disapproving the reasonable progress
determinations for SGS and IGS as it relates to the mass-based
emissions caps at SGS and IGS was improper.
First, we wish to clarify that we do not object to the use of
numeric emissions limitations as a means to implement control
determinations. Indeed, CAA section 169A(b)(2) specifically requires
the long-term strategy to include ``enforceable emissions limitations,
compliance schedules, and other measures that are necessary to make
reasonable progress.'' As explained in our proposal:
The amount of progress that is ``reasonable progress'' is based
on applying the four statutory factors in CAA section 169A(g)(1) in
an evaluation of potential control options for sources of visibility
impairing pollutants, which is referred to as a ``four-factor''
analysis. The outcome of that analysis is the emissions reduction
measures that a particular source or group of sources needs to
implement to make reasonable progress towards the national
visibility goal. . . . Such measures must be represented by
``enforceable emissions limitations, compliance schedules, and other
measures'' (i.e., any additional compliance tools) in a state's
long-term strategy in its SIP.\69\
---------------------------------------------------------------------------
\69\ 89 FR 47398, 47402-47403.
We find that the mass-based emissions caps set for SGS do not
represent the emissions reduction measures that were the outcome of the
state's four-factor analysis for the reasons described in our proposal
\70\ and herein. Therefore, these caps do not meet the requirements of
169A(b)(2) and (g)(1), or the corresponding provisions of the RHR.
---------------------------------------------------------------------------
\70\ 89 FR 47398, 47430.
---------------------------------------------------------------------------
Regarding PCC, we agree that in the EPA's reasonable progress
determination for PCC in the first implementation period, the EPA
established a mass-based emissions limitation for NO<INF>X</INF>.
However, the circumstances between PCC and SGS Units 1 and 2 differ in
important ways. The limit for PCC was set pursuant to 40 CFR
51.308(d)(3), which required the long-term strategy for the first
implementation period to ``include enforceable emissions limitations,
compliance schedules, and other measures as necessary to achieve the
reasonable progress goals established by States having mandatory Class
I Federal areas.'' \71\ In this instance, the Class I areas primarily
affected by emissions from PCC were in Arizona,\72\ and the emissions
reductions from PCC were reflected in the applicable RPGs for these
areas by scaling of visibility extinction components in proportion to
changes in total annual emissions.\73\ Under these circumstances, an
annual mass-based emissions limit corresponding to the level of annual
emissions reductions assumed in the RPG calculations was sufficient to
meet the applicable requirement for an emissions limit ``as necessary
to achieve'' the relevant RPGs.
---------------------------------------------------------------------------
\71\ 40 CFR 51.308(d)(3) (emphasis added).
\72\ See 79 FR 9318, 9354 Table 41 (showing impacts on Arizona I
class I areas).
\73\ 79 FR 52420, 52468-52469; FIP_RPG_estimates.xlsx.
---------------------------------------------------------------------------
In contrast, for the second planning period, the EPA clarified in
40 CFR 51.308(f)(2) that ``the long-term strategy must include the
enforceable emissions limitations, compliance schedules, and other
measures that are necessary to make reasonable progress, as determined
pursuant to (f)(2)(i) through (iv).'' 40 CFR 51.308(f)(i) in turn
requires the state to ``evaluate and determine the emission reduction
measures that are necessary to make reasonable progress by
considering'' the four statutory factors. Thus, the revised rule
clarifies that the long-term strategy must include emissions
limitations, compliance schedules, and other measures representing the
emissions reduction measures that the state determined to be necessary
to make reasonable progress, considering the four statutory factors.
The EPA provided recommendations on the appropriate form of such
emissions limitations and other measures in the 2019 Guidance and the
Clarifications Memo. Specifically, the 2019 Guidance recommends the use
of throughput-based limits, rather than mass-based limits (i.e.,
``caps'') for emissions limitations to implement measures necessary for
reasonable progress in most instances.\74\ The Guidance notes that
mass-based limits are allowed under the RHR, but explains that, ``[i]f
the state has determined, independent of the forecasted operating
level, that operation of the emission control equipment . . . is
necessary to make reasonable progress, a mass-based emission limit may
not be appropriate.'' \75\ The Clarifications Memo also reaffirms that
``whether for new or existing measures . . . an emission limit . . .
should be in the form of the emission rate achieved when implementing
those measures (e.g., pounds per million British thermal units or lbs/
MMBtu, pounds per hour or lbs/hr, or pounds per ton or lbs/ton of
produced material).'' \76\
---------------------------------------------------------------------------
\74\ 2019 Guidance p. 44.
\75\ Id. pp. 44-45.
\76\ 2021 Clarifications Memo, p. 11.
---------------------------------------------------------------------------
With regard to SGS 1 and 2, ADEQ concluded, based on a four-factor
analysis that, ``it is reasonable to require TEP to upgrade the current
SDA systems'' \77\ and did not indicate that this determination was
conditioned on a particular level of operation. Therefore, ADEQ should
have set limits in the form of the emissions rates achieved when
implementing SDA upgrades, e.g., lb/MMBtu limits, or should have
provided a rationale for why the mass-based limits, which could be met
without any control upgrades, nonetheless represent SDA upgrades.
---------------------------------------------------------------------------
\77\ Plan Appendix C, p. 232.
---------------------------------------------------------------------------
We also note that the mass-based limit at PCC was for a single
unit, meaning that it necessarily constrained the emissions from that
unit on an annual basis. In contrast, the mass-based limit at SGS Units
1 and 2 were set across two units. Accordingly, if one unit ceases to
operate, the limits would not meaningfully constrain emissions from the
other unit. In addition, ADEQ rejected more stringent controls at SGS
Units 1 and 2 based on its ``determination that another viable
reasonable control exists to reduce SO<INF>2</INF> emissions from Units
1 & 2 (upgraded SDA).'' In contrast, the EPA did not reject any more
stringent controls for PCC based on incremental cost effectiveness.
Furthermore, the cap at PCC was intended, in part, to address concerns
raised during consultation with the facility's owner, the Salt River
Pima Maricopa Indian Community (SRPMIC).\78\ No similar considerations
exist with respect to SGS Units 1 and 2. Therefore, we find that the
circumstances concerning the PCC
[[Page 102755]]
mass-based cap are distinguishable from those concerning SGS Units 1
and 2.
---------------------------------------------------------------------------
\78\ 79 FR 9318, 9366 (``An annual cap would allow SRPMIC to
delay installation of controls until the Plant's production returns
to pre-recession levels and would thus help to address the
Community's concerns about the budgetary impacts of control
requirements.'')
---------------------------------------------------------------------------
Second, regarding our finding that SDA upgrades would be cost
effective after the establishment of the mass-based emissions caps, we
disagree that we substituted our judgment for ADEQ's by arbitrarily
rejecting ADEQ's determination based on a circular four-factor analysis
standard. Contrary to ADEQ's suggestion, we did not state that ADEQ was
required to revisit and update the baseline emissions of its four-
factor analysis to reflect the new emissions limitation. However, we
note that ADEQ did use this approach for IGS Unit 3 and determined that
with the emissions reductions associated with the new Unit 3 emissions
caps, no additional controls are necessary to make reasonable progress.
ADEQ did not provide any rationale for why it used this approach for
IGS Unit 3, but not for SGS Units 1 and 2.
Furthermore, the fact that the SDA upgrades would still be cost
effective following implementation of the mass-based caps at SGS Units
1 and 2 is relevant to whether the use of caps (in lieu of throughput-
based limits) is reasonable and appropriate in this particular case.
The fact that throughput-based (e.g., lb/MMBtu) limits equivalent to
SDA would still be cost-effective following implementation of the caps
(i.e., if TEP complies with the caps by lowering its operating level)
indicates that throughput limits based on the emissions rates
achievable with SDA upgrades may be necessary to make reasonable
progress, regardless of the future operating level. As previously
noted, the 2019 Guidance explains that, under such circumstances, mass-
based caps are generally not appropriate. Therefore, we believe this
consideration is relevant to the evaluation of whether mass-based caps
were reasonable and consistent with the requirements of the RHR in this
case.
Third, we agree with ADEQ's statement that the operating scenarios
outlined in the IRP are not federally enforceable conditions. However,
ADEQ's statement in its comment letter that it has ``no basis for the
consideration of these future scenarios as part of its control measure
analysis and the establishment of the mass-based emission limits'' is
inconsistent with its stated rationale in the Plan for establishing
caps instead of mass-based limits. Specifically, ADEQ stated that:
As discussed in TEP's 2020 IRP, Units 1 will transition to
seasonal operation in 2023 and Unit 2 in 2024. TEP is planning to
retire Unit 1 in 2027 and Unit 2 in 2032. TEP will be very likely to
manage its operating level strategically instead of completing the
upgrades to the SDA systems for meeting the RP requirements.
Therefore, ADEQ determines that a mass-based emission limit is
reasonable.\79\
---------------------------------------------------------------------------
\79\ 2022 Arizona Regional Haze Plan, Appendix C, p. 236.
In other words, ADEQ elected to establish caps specifically because
it anticipated that TEP could comply with these caps by reducing its
operating level consistent with its then-current IRP. Therefore, we do
not agree that it was inappropriate for the EPA to consider information
contained in TEP's most recent (2023) IRP in evaluating whether the
mass-based caps were reasonable and consistent with the RHR.
Lastly, we agree that for APS Cholla Unit 2, the EPA accepted a
source-specific permit revision that included a trigger that was
conditional on the EPA's approval of the SIP revision that altered the
remaining useful life of the unit in ADEQ's four factor analysis.
However, the APS Cholla scenario was distinguishable because, as
explained in the proposed rule,\80\ the SIP revision for APS Cholla
replaced the FIP that was applicable to these units.\81\ Therefore, it
would not have been reasonable to subject them to two inconsistent
requirements, one State and one Federal, under the RHR at the same
time. Thus, under these circumstances it was appropriate for ADEQ to
make the effectiveness of the permit conditions contingent on EPA SIP
approval. In the current case, there is no existing FIP in place
applicable to IGS Unit 3, so no similar rationale exists for making the
cap contingent upon approval by the EPA.
---------------------------------------------------------------------------
\80\ 89 FR 47398, 47407, Table 1.
\81\ 77 FR 72511 (December 5, 2012).
---------------------------------------------------------------------------
Comment B.9. ADEQ asserts that the EPA should approve ADEQ's
reasonable progress goal (RPG) for the Sycamore Canyon Wilderness Area.
The state indicates that it provided the required ``robust
demonstration'' by including a detailed analysis of visibility data at
the Sycamore site to demonstrate that its slower rate of progress
results from significant increases in light extinction from coarse
mass.
Response B.9. As ADEQ notes in its comment, its analysis regarding
Sycamore Canyon focused on the substantial increase in coarse mass and
soil impairment at the Sycamore Canyon site. However, as explained in
our proposal, the rule requires a state with a Class I area whose RPG
is above the glidepath to demonstrate, based on the source selection
and four factor analyses required under 40 CFR 51.308(f)(2)(i), that
there are no additional emissions reduction measures for sources that
may reasonably be anticipated to contribute to visibility impairment in
the Class I area that would be reasonable to include in the long-term
strategy.\82\ Although ADEQ provided a detailed analysis of monitoring
data concerning Sycamore Canyon, it did not provide ``a robust
demonstration, including documenting the criteria used to determine
which sources or groups or sources were evaluated and how the four
factors required by paragraph (f)(2)(i) were taken into consideration
in selecting the measures for inclusion in its long-term strategy.''
\83\ Instead, it relied solely on monitoring data and the source
selection and four-factor analyses contained elsewhere in the Plan.\84\
For the reasons described in our proposal and elsewhere in this
document, we find that these analyses were inadequate to meet the
requirements of 40 CFR 51.308(f)(2). Therefore, the Plan also did not
meet the requirements of 40 CFR 51.308(f)(3)(ii)(A) with respect to
Sycamore Canyon.
---------------------------------------------------------------------------
\82\ 89 FR 47398, 47433.
\83\ 40 CFR 51.308(f)(ii)(A).
\84\ See Plan p. 106, footnote 112 (citing Plan Section 8 and
Appendix C).
---------------------------------------------------------------------------
Comment B.10. ADEQ indicates that the State intends to coordinate
with EPNG to provide supporting documentation for the interest rate.
ADEQ indicates that it intends to provide the interest rate
documentation as confidential business information (CBI) to the EPA for
review prior to publication of the final rule and requests that the EPA
approve the cost calculation for EPNG Willcox and Williams based on the
site-specific interest rate and supporting documentation.
Response B.10. Although ADEQ referenced the existence of
documentation and the possibility of sharing it with the EPA in its
response to comments in the Plan,\85\ no supporting documentation for
the interest rate was submitted as part of the Plan or prior to the
close the public comment period on the EPA's proposed partial approval
and partial disapproval. Therefore, the EPA is unable consider the
interest rate information as part of this action.
---------------------------------------------------------------------------
\85\ 2022 Arizona Regional Haze Plan, Appendix K.
---------------------------------------------------------------------------
Additionally, we note that the lack of documentation of EPNG's
firm-specific interest rate was one of several flaws that we identified
in the state's long-term strategy. Therefore, even if the appropriate
documentation had been submitted and within the proper time frame, that
alone would not have changed our determination that the
[[Page 102756]]
2022 Arizona Regional Haze Plan did not satisfy the long-term strategy
requirements of 40 CFR 51.308(f)(2).
Comment B.11. ADEQ voiced procedural concerns with the lack of
specificity in the EPA's proposed disapproval of the 2022 Arizona
Regional Haze Plan. ADEQ cites CAA 307(d)(3) as requiring a detailed
notice of rulemaking and cites Small Ref. Lead Phase-Down Task Force v.
EPA, 705 F.2d 506, 518-19 (D.C. Cir. 1983), for the proposition that
the notice provided in the proposal may be too general to be adequate
and that ``[a]gency notice must describe the range of alternatives
being considered with reasonable specificity.'' The commenter asserts
that the EPA's proposed rule does not provide ``detailed notice'' of
certain specific issues that form the basis for disapproval of entire
sections of the 2022 Arizona Regional Haze Plan, denying ADEQ the
opportunity to specifically address those bases for disapproval during
the comment period.
Response B.11. We disagree that the EPA's proposal provided
inadequate notice. The EPA's action on the 2022 Arizona Regional Haze
Plan is not subject to the requirements of CAA section 307(d). Those
requirements apply only to specific enumerated types of actions under
the CAA and to ``such other actions as the Administrator may
determine.'' \86\ Actions on SIPs are not one of the enumerated
actions, and the Administrator had not determined that this action is
subject to 307(d) pursuant to Section 307(d)(1)(V). Therefore, this
action is subject to the procedural requirements of the Administrative
Procedure Act (APA).
---------------------------------------------------------------------------
\86\ CAA Section 307(d)(1).
---------------------------------------------------------------------------
Accordingly, pursuant to 5 U.S.C. 553(b)(2) and (3), the EPA's
notice of proposed rulemaking regarding the Plan was required to
include ``reference to the legal authority under which the rule is
proposed'' and ``either the terms or substance of the proposed rule or
a description of the subjects and issues involved.'' The proposal
clearly met these requirements, as it stated the applicable legal
authorities and provided the EPA's review of the Plan in relation to
those requirements. The comment provides no basis to conclude that the
proposal failed to meet these requirements. Indeed, the opinion cited
by the commenter contrasts these more general APA requirements to the
more exacting requirements of CAA section 307(d) concerning the
contents of proposed rulemaking.\87\ While we agree with the commenter
that ``[a]gency notice must describe the range of alternatives being
considered with reasonable specificity,'' \88\ we find that our
proposal met this requirement, as it plainly stated that the EPA was
considering partially approving and partially disapproving the 2022
Arizona Regional Haze Plan and detailed the reason for this proposed
action. Please also see Responses A.1.
---------------------------------------------------------------------------
\87\ Small Ref. Lead Phase-Down Task Force v. EPA, 705 F.2d 506,
518-19 (D.C. Cir. 1983),
\88\ Id. at 549.
---------------------------------------------------------------------------
Comment B.12. ADEQ acknowledges that further FLM consultation is
required for a plan revision that will correct the deficiencies
identified in the proposed rulemaking action, but contends that the
inclusion of the nonpoint source selection analysis and selected
controls for nonpoint sources in the FLM review draft of the plan
provided FLMs adequate notice and review of Arizona's nonpoint source
rules that were codified after plan submission and submitted in the
2023 Arizona Regional Haze Rules Supplement.\89\
---------------------------------------------------------------------------
\89\ ADEQ SIP Revision: Nonpoint Rules to Supplement Arizona's
2022 Regional Haze SIP, ``2023 Arizona Regional Haze Rule
Supplement,'' August 22, 2023.
---------------------------------------------------------------------------
Response B.12.The EPA disagrees with ADEQ's contention that it
provided the FLMs with adequate notice and review of Arizona's nonpoint
source rules. The information about FLM consultation regarding the
rules specifically is not detailed in either the 2023 Arizona Regional
Haze Rules Supplement, or the 2022 Arizona Regional Haze Plan. The 2023
Arizona Regional Haze Rules Supplement indicated that further
information on how this SIP revision complied with 40 CFR 51.308(i)
requirements for federal land manager consultation is section 2.4 of
the 2022 Arizona Regional Haze Plan.\90\ However, section 2.4 of the
plan only describes the FLM review of the 2022 Arizona Regional Haze
Plan, and does not specify if ADEQ provided a draft of the three
nonpoint source rules to the FLMs for review. The rules were not
included in the 2022 Arizona Regional Haze Plan. Should Arizona choose
to submit a SIP revision, clarification of the FLM review process of
the nonpoint source rules that are listed in Table 1-1 of the 2023
Arizona Regional Haze Rules Supplement would be helpful for EPA review.
---------------------------------------------------------------------------
\90\ Id, at Chapter 4.
---------------------------------------------------------------------------
C. Comment Letter From the Chamber and AMC
Parts I (``The Chamber and AMC are supportive of the goals of the
Regional Haze Rule'') and III (``Conclusion'') of the Chamber and AMC's
comment letter are informational and therefore do not require a
response. We respond to part II, sections A-E of the Chamber and AMC's
comment letter below.
Comment C.1. The Chamber and AMC note that the EPA's delayed action
and partial disapproval of Arizona's Regional Haze Plan is an example
of erosion of cooperative federalism. The commenter contends that the
timeline of events leading up to the EPA's partial disapproval of
Arizona's Regional Haze Plan is problematic and that the EPA seems to
routinely miss statutory deadlines, only to get sued by third-party
entities for failure to act, resulting in agreements to deadline
extensions that delay action for years. In the meantime, the resources
and analysis invested by states depreciate in value, often requiring
states to reinvest in efforts to update an analysis with new
information. The commenter also asserted that ``[r]egularly, an even
worse scenario plays out in which EPA denies a SIP because the
information submitted in good faith by a state has since become dated
and stale.''
Response C.1. We do not agree that either the timing or substance
of the EPA's partial disapproval of Arizona's Regional Haze Plan is an
example of the erosion of cooperative federalism. We acknowledge that
the EPA did not act on the 2022 Arizona Regional Haze Plan within the
statutory deadline under CAA section 110(k)(3), and that we were
subsequently sued for failing to meet that deadline. This resulted in a
court-ordered deadline for the EPA to take action on the Plan by March
30, 2025.\91\ However, we do not agree that this resulted in a
``deadline extension'' of any sort, or an erosion of cooperative
federalism. On the contrary, in issuing a partial approval and partial
disapproval of the Plan, the EPA is fulfilling our statutorily-mandated
role to review SIPs for compliance with the requirements of the CAA and
the RHR, as further described in Response D.3.
---------------------------------------------------------------------------
\91\ Sierra Club v. EPA (D.D.C. Case No. 1:23-cv-01744-JDB),
Consent Decree entered July 12, 2024.
---------------------------------------------------------------------------
We also disagree with the suggestion that the EPA is disapproving
the 2022 Arizona Regional Haze Plan, partially or entirely because
information became ``dated or stale.'' The comment did not provide
examples of information becoming dated, resulting in disapproval; thus,
we cannot comment on any specific concerns the commenter has with the
information within the 2022 Arizona Regional Haze Plan.
Further, as explained in Responses A.1 EPA staff also discussed
with ADEQ many of the concerns that became bases for our disapproval
during the SIP development process.
[[Page 102757]]
Comment C.2. The Chamber and AMC state that guidance should not be
cited as grounds to disapprove the 2022 Arizona Regional Haze Plan. The
commenter asserts that guidance should be viewed only as a reference
and not a legal requirement to be used to approve or disapprove a
state's plan. The Chamber and AMC also claim that the EPA published
final guidance for Regional Haze Plans three weeks before the plans
were due, making compliance with the guidance practically impossible
before the deadline. The commenter concludes that relying on the
guidance to partially disapprove the Plan was therefore arbitrary and
capricious, and the EPA should withdraw all of the proposed
disapprovals based upon the 2021 Clarifications Memo.
Response C.2. The EPA disagrees that it relied on guidance,
including the 2021 Clarifications Memo, as the basis for our partial
disapproval. See responses A.5, B.1, and B.2.
Comment C.3. The Chamber and AMC assert that Arizona's source
selection methodology was reasonable and that the EPA should give
deference to the State on this matter. The commenter indicates that
sources that applied controls in the first round of Regional Haze had
recently made significant investments in the design, engineering,
procurement, construction and operation of those air pollution control
devices. They note that forcing facilities to consider improvement or
replacement of these air pollution control devices long before they
have depreciated is an unnecessary economic burden for the source and
the State. The Chamber and AMC note that ADEQ chose not to force
additional analysis from these sources and reasonably relied upon
reductions from other emissions sources for visibility improvement, a
method that resulted in all but one of Arizona's Class I areas either
meeting or exceeding the uniform rate of progress (URP) toward natural
conditions.
Response C.3. We disagree that it was reasonable for Arizona to
screen out sources solely because they applied controls during the
first planning period, for the reasons described in section IV.E.2.a.
of our proposal and in response B.3 of this document.
We also disagree that our disapproval will automatically force
sources to consider improving or replacing any recently installed air
pollution devices. Rather, ADEQ has the option to provide additional
documentation and justification for its effective control
determinations in a responsive SIP revision. We anticipate that for
many units that recently installed controls, ADEQ will be able to
provide an adequate demonstration of effective controls on a unit-
specific and pollutant-specific basis, if it chooses to do so, which
would preclude the need for a four-factor analysis for those units and
pollutants. Any affected units and pollutants for which ADEQ is unable
to make such demonstration would be subject to the four-factor analysis
requirement as required by 40 CFR 51.308(f)(2).
Finally, we note that all states are subject to the requirements at
40 CFR 51.308(f)(2) and (3) regardless of whether the 2028 RPGs for
Class I Areas they affect are above or below URP.
Comment C.4. The Chamber and AMC assert that the EPA should give
deference to Arizona's deviation from the EPA Control Cost Manual in
developing cost estimates. The commenter notes that the Cost Control
Manual is not accurate for all sources and cites examples such as
variable interest and emissions rates. They conclude that the use of
different interest rates and different control efficiencies for
different projects should be viewed as reasonable.
Response C.4. We do not agree that we should have deferred to
Arizona's deviation from the EPA Control Cost Manual in the absence of
adequate justification. As discussed in Response D.3, Congress charged
the EPA with independently evaluating and reviewing SIP submissions for
compliance with the applicable requirements under the CAA. 40 CFR
51.308(f)(2)(iii) requires states to ``document the technical basis,
including modeling, monitoring, cost, engineering, and emissions
information, on which the State is relying to determine the emission
reduction measures that are necessary to make reasonable progress in
each mandatory Class I Federal area it affects.'' The technical
documentation must include the modeling, monitoring, cost, engineering,
and emissions information on which the state relied to determine the
measures necessary to make reasonable progress. For the reasons noted
in section IV.E.2.b.ii (``Deviations from Cost Control Manual'') of our
proposed rule, we found that Arizona failed to adequately document the
technical basis that it relied upon to determine emissions reduction
measures, as required by 40 CFR 51.308(f)(2)(iii). Specifically, as
explained in our proposal, it is important to use consistent methods in
order to allow for reasoned comparisons between different sources
within a state, and cost analyses in other states.\92\ Therefore, while
our regulations allow for flexibility among various methodologies,
where a state deviates from these methods, it should explain how its
alternative approach is reasonable, appropriate, and consistent with
the regulations and the statutory requirement to make reasonable
progress towards the national goal. Arizona did not do so. We therefore
disagree that the EPA should give deference to Arizona's approach in
the Plan.
---------------------------------------------------------------------------
\92\ 89 FR 47398, 47428-47429.
---------------------------------------------------------------------------
Comment C.5. The Chamber and AMC indicate that Arizona's plan
should be viewed in the context that it results in all but one monitor
having an RPG that provides for a greater rate of visibility
improvement than the adjusted URP. The commenter asserts that Arizona
has created a plan that meets or exceeds the URP at all monitors except
for Sycamore Canyon, which was moved to an intersection of two dirt
roads in 2014. The commenter concludes that the fact that visibility at
Arizona's Class I areas is improving at a pace to reach natural
conditions prior to the RHR goal of 2064 is important context in
evaluating the source selection methodology and other decisions made by
Arizona.
Response C.5. We do not agree with the commenter's characterization
of the role of the URP. As explained in our proposal:
The URP is a planning metric used to gauge the amount of
progress made thus far and the amount left before reaching natural
visibility conditions. However, the URP is not based on
consideration of the four statutory factors and therefore cannot
answer the question of whether the amount of progress being made in
any particular implementation period is ``reasonable progress.''
\93\
---------------------------------------------------------------------------
\93\ 89 FR 47398, 47406.
Moreover, being on or below the URP does not exempt a state from
any of the requirements of the CAA or the Regional Haze Rule.\94\
---------------------------------------------------------------------------
\94\ See 82 FR 3078,3093 and 3099-3100.
---------------------------------------------------------------------------
It should also be noted that the URP represents the amount of
visibility improvement that would need to be achieved during each
implementation period to achieve natural visibility conditions by the
end of 2064. However, the 2064 date is used solely to calculate the URP
as a tracking metric. The CAA and RHR do not contain any end dates for
the regional haze program and do not have a ``goal'' or requirement to
achieve natural conditions by 2064 specifically.
Please also see Response B.9 for more information on the robust
[[Page 102758]]
demonstration required for Sycamore Canyon under 40 CFR
51.308(f)(3)(ii), which is missing from Arizona's submission.
D. Comment Letter From TEP
The ``Background'' section of TEP's comment letter is informational
and therefore does not require a response. We respond to the ``Comments
of Proposed Disapproval'' section of TEP's comment letter below.
Comment D.1. TEP claims that ADEQ's source selection methodology
was reasonable and the EPA should approve the determination. First, TEP
cites to the 2019 Guidance and 2021 Clarifications Memo as providing
states discretion for source selection and notes that ADEQ applied a Q/
d screening threshold of 10 for each process at a source. TEP further
describes ADEQ's screening out processes where the facility recently
adopted ``effective controls,'' which the Agency defined as controls
installed to meet the requirements of the PSD program (BACT), the first
regional haze planning period (BART), or other NAAQS requirements.
Second, TEP further suggests that the EPA is proposing to approve
``many aspects of ADEQ's source selection process,'' including ADEQ's
choice of screening threshold and focus on NO<INF>X,</INF>
SO<INF>2</INF>, and PM<INF>10</INF> in evaluating visibility impacts,
but simultaneously proposing to find that the State did not adequately
justify its determination of effective emissions reduction measures.
Response D.1. First, we disagree with the commenter that ADEQ
applied a Q/d screening threshold of 10 for each process at a source.
As noted in the 2022 Arizona Regional Haze Plan, the Q value was
calculated from facility-wide PM<INF>10,</INF> NO<INF>X</INF>, and
SO<INF>2</INF> annual emissions, not process-specific emissions.\95\
---------------------------------------------------------------------------
\95\ 2022 Arizona Regional Haze Plan, Appendix C2.2, p. 26.
---------------------------------------------------------------------------
Second, we clarify that the EPA did not propose to approve ``many
aspects of ADEQ's source selection process.'' We found that ADEQ
reasonably and adequately explained and documented many aspects of its
source selection process, such as its focus on sulfate, nitrate, and
coarse mass and its use of a Q/d value of 10 for point sources.
However, we are not separately approving or disapproving specific
elements of ADEQ's long-term strategy, including any elements of the
source-selection process. Rather, we are disapproving the long-term
strategy as a whole under 40 CFR 51.308(f)(2), for the reasons
described in our proposal and in this document.
Comment D.2. TEP asserts that ADEQ reasonably evaluated existing
controls at IGS Unit 4 and SGS Units 3 and 4.
First, the commenter states that ADEQ did not determine that
sources were effectively controlled based on BART controls alone, but
also evaluated additional emissions reduction measures at several units
that were controlled during the first regional haze planning period,
citing Table 8-2 in the 2022 Arizona Regional Haze Plan. For IGS Unit 4
specifically, the commenter states that ADEQ evaluated existing
emissions reduction measures at IGS Unit 4 from a source-specific
perspective, and determined that further analysis would be futile. TEP
cites to statements regarding post-combustion controls in the 2019
Guidance and states that, in a recent FIP action for Arizona, the EPA
determined that eliminating coal combustion at Unit 4 would control
emissions beyond the best available NO<INF>X</INF> and SO<INF>2</INF>
controls.\96\ TEP concludes that these emissions reduction measures
remain the best available controls at this unit in the second planning
period. TEP further notes that ADEQ's determination that IGS Unit 4 is
well-controlled is also consistent with recent decreasing emissions
trends across BART-eligible EGUs, including IGS Unit 4, and between
2014 and 2019. TEP points to the recent proposed rule for Georgia's
Regional Haze SIP,\97\ and comments that the EPA cited to similar
visibility-impairing emissions trends as support for Georgia's source-
selection methodology. Citing to the 2021 Clarifications Memo,\98\ TEP
also indicates that ADEQ was not required to consider emissions trends
and that the State has discretion to emphasize other considerations,
such as the EPA's prior FIP evaluation.
---------------------------------------------------------------------------
\96\ 79 FR 52420, 52422.
\97\ 89 FR 47481 (June 3, 2024).
\98\ Specifically, TEP quotes page 3 of the 2021 Clarifications
Memo (``[t]he [Regional Haze Rule] does not explicitly list factors
that states must or may not consider when selecting sources for
analysis,'') and cites page 5 of the 2021 Clarifications Memo as
supporting its assertion that ``EPA has recommended that states
consider projected and actual emissions in evaluating existing
emission reduction measures.''
---------------------------------------------------------------------------
Second, TEP states that the EPA's proposed disapproval fails to
engage with ADEQ's analysis for SGS Unit 3 and 4. TEP notes that ADEQ
considered potential additional control measures that could be used to
achieve emissions reductions at SGS Units 3 and 4 based on an initial
control analysis submitted by TEP.\99\ In this analysis, TEP provided
information on technically feasible control measures, as well as the
actual and projected emissions rates at each unit.
---------------------------------------------------------------------------
\99\ TEP, Identification and Evaluation of Emission Control
Measures for Units 3 and 4 at the Springerville Generating Station
for Purposes of the Regional Haze Second Planning Period Under 40
CFR 51.308(f)(2) (Mar. 2020), available at <a href="https://static.azdeq.gov/aqd/haze/tep_spr_4fa_u34.pdf">https://static.azdeq.gov/aqd/haze/tep_spr_4fa_u34.pdf</a>.
---------------------------------------------------------------------------
Response D.2. We disagree that ADEQ's evaluations regarding
effective controls at IGS Unit 4 and SGS Units 3 and 4 were reasonable
and justified. Contrary to TEP's claim, ADEQ did not evaluate
additional emissions reduction measures at several units that were
controlled during the first regional haze planning period. Table 8-2
cited by TEP to support its claim that ADEQ evaluated additional
emissions reduction measures at units that were controlled during the
first regional haze planning period only depicts annual, source-level
total emissions of NO<INF>X,</INF> SO<INF>2</INF> and PM<INF>10</INF>
and does not include any information regarding unit-specific or
pollutant specific emissions rates or controls. The commenter also
provides no citation for its assertion that ADEQ evaluated existing
emissions reduction measures at IGS Unit 4 from a source-specific
perspective. Therefore, based on the contents of the 2022 Arizona
Regional Haze Plan, it was reasonable for the EPA to determine that
ADEQ did not evaluate additional emissions reduction measures at units
that were controlled during the first regional haze planning period.
Similarly, the commenter's citation to the discussion of post-
combustion controls in the 2019 Guidance is misleading because no such
controls were installed at IGS Unit 4. Rather, the unit was converted
from coal to gas as part of a ``better-than-BART'' determination
pursuant to 40 CFR 51.308(e)(2).\100\ Therefore, the actual relevant
effective controls discussion in the 2019 Guidance is the discussion of
fuel combustion units that are required to burn pipeline quality
natural gas. Such units are generally considered to be effectively
controlled for SO<INF>2</INF> and PM.\101\ However, they are not
necessarily effectively controlled for NO<INF>X</INF>. Therefore, we
disagree with the commenter that, given TEP's recent conversion of IGS
Unit 4 from coal to natural gas, ADEQ's conclusion that further
analysis was not required was reasonable. As stated in the proposed
rule,\102\ ADEQ should have explained why it is reasonable to assume,
without conducting a four-factor analysis, that no additional
NO<INF>X</INF> controls would be reasonable.\103\
---------------------------------------------------------------------------
\100\ See 40 CFR 52.145(j)(4).
\101\ See 2019 Guidance, p. 24.
\102\ 89 FR 47398, 47428.
\103\ See 40 CFR 51.308(f)(2)(i) (``. . .The State must include
in its implementation plan a description of the criteria is used to
determine which sources or groups of sources it evaluated and how
the four factors were taken into consideration in selecting the
measures for inclusion in its long-term strategy.''). See also 2021
Clarifications Memo, p. 5; 2019 Guidance, p. 23.
---------------------------------------------------------------------------
[[Page 102759]]
The commenter's discussion of emissions trends is also misleading
on several grounds. First, the commenter mischaracterizes the EPA's
review of Georgia's regional haze source selection methodology. In the
section of the Georgia proposal cited by TEP, the EPA considered trends
in total measured visibility impairment at three Class I areas affected
by Georgia's sources as supporting the reasonableness of the state's
overall source selection methodology.\104\ The EPA did not, however,
consider trends in emissions from specific sources and did not indicate
that such trends would be relevant either to the reasonableness of a
state's overall source selection methodology or especially to the
question of a whether a particular source may be screened out on the
grounds that it is ``effectively controlled.'' Furthermore, the trends
cited by the commenter were for multiple BART-eligible EGUs, not just
IGS. Therefore, we do not agree that decreasing SO<INF>2</INF> and
NO<INF>X</INF> emissions at BART-eligible EGUs in Arizona between 2014
and 2019 are relevant to whether IGS Unit 4 is effectively controlled.
---------------------------------------------------------------------------
\104\ 89 FR 47481, 47497-47498.
---------------------------------------------------------------------------
Second, the commenter mischaracterizes the contents of the 2021
Clarifications Memo as it relates to how to determine whether a source
is effectively controlled. The first section cited by the commenter,
regarding ``Factors to Consider for Source Selection'' \105\ relates to
a State's overall source selection methodology, which is generally
considered to be the first step of determining what measures are
necessary for reasonable progress.\106\ Whether and how to screen out
particular sources on the grounds that they are effectively controlled
is a subsequent step. As previously noted, we found that many aspects
of ADEQ's source selection process were reasonable and adequately
explained and documented,\107\ consistent with the statement in the
Clarifications Memo that, ``whatever choices states make should be
reasonably explained and produce a reasonable outcome.'' \108\ However,
once the sources were initially selected for evaluation of additional
control measures, we found that ADEQ did not provide an adequate
justification for subsequently screening out certain sources and units
from ultimately conducting a four-factor analysis on the basis that
they are effectively controlled.
---------------------------------------------------------------------------
\105\ 2021 Clarifications Memo, p. 3.
\106\ Id. (``Source selection is a critical step in states'
analytical processes. All subsequent determinations of what
constitutes reasonable progress flow from states' initial decisions
regarding the universe of pollutants and sources they will consider
for the second planning period.'')
\107\ 89 FR 47428.
\108\ 2021 Clarifications Memo, p. 3.
---------------------------------------------------------------------------
The second section of the 2021 Clarifications Memo mischaracterized
by the commenter, ``Sources that are Not Selected Based on Existing
Effective Controls,'' does address the screening out of particular
sources on the grounds that they are effectively controlled. In
particular, this section recommends that, ``States should first assess
whether the source in question already operates an `effective control'
as described in the August 2019 Guidance. They should further consider
information specific to the source, including recent actual and
projected emission rates, to determine if the source could reasonably
attain a lower rate.'' \109\ For the reasons detailed in our proposal
and elsewhere in this document, we find that ADEQ did not reasonably
explain and support its determination at the first step that IGS Unit 4
already operates effective controls, particularly for NO<INF>X</INF>.
Moreover, even if IGS Unit 4 does have effective controls, ADEQ should
have considered recent actual and projected emissions rates for this
particular unit, not for all BART-eligible units as a group, in order
to determine whether these controls are necessary to make reasonable
progress.
---------------------------------------------------------------------------
\109\ Id., p. 5 (emphasis added).
---------------------------------------------------------------------------
Finally, we disagree that our proposed disapproval fails to engage
with ADEQ's analysis for SGS Unit 3 and 4. Contrary to the commenter's
suggestion, ADEQ did not screen out SGS Units 3 and 4 at the source
selection stage,\110\ but instead conducted four-factor analyses for
these units.\111\ We summarized these analyses on pages 47422-47423 of
our proposal. We did not note any particular flaws in these analyses or
the resulting determinations that no additional controls were necessary
to make reasonable progress in our proposal. However, we found that
ADEQ had not addressed whether any of the existing measures relied upon
in these four-factor analyses were necessary to make reasonable
progress and thus should be a part of the State's long-term strategy
for the second planning period.\112\ We also noted that, as part of its
analysis of whether these existing measures are necessary to make
reasonable progress, the State should have considered whether the
relevant sources are subject to enforceable emissions limits that
ensure their emissions rates will not increase. Without this
information, it is not clear what measures are in the State's long-term
strategy for the second planning period and how controls on these units
result in each of the affected Class I areas making reasonable progress
towards the national goal.
---------------------------------------------------------------------------
\110\ 2022 Arizona Regional Haze Plan, Appendix C, Exhibit CI.
\111\ Id., Chapter C3.13.
\112\ 89 FR 47431.
---------------------------------------------------------------------------
Comment D.3. TEP asserts that ADEQ reasonably evaluated additional
control measures using a four-factor analysis for SGS Units 1 and 2.
Citing Oklahoma v. EPA, 723 F.3d 1201, 1209 (10th Cir. 2013), the
commenter asserts that while the statute identifies the factors that
must be considered, Congress left to states the determination as to how
these factors should be weighed.
Response D.3. While we agree that states have significant
discretion in how they consider and apply the four statutory factors as
part of a Regional Haze SIP, they do not have unlimited discretion. On
the contrary, the EPA has a crucial role in reviewing such SIP
submissions for compliance with the requirements of the CAA and the
RHR. Pursuant to CAA section 110, states must submit SIP revisions to
the EPA for review and the EPA must evaluate whether each SIP
submission meets all of the applicable requirements of the Act.\113\
The EPA must disapprove any SIP revision that ``would interfere with
any applicable requirement'' of the Act.\114\ CAA section 110(a)(2)(J)
specifically requires that SIPs ``meet the applicable requirements'' of
Part C of Title I of the CAA including the requirements for visibility
protection set forth in sections 169A and 169B.\115\ Pursuant to
section 169A(b), the EPA is required to promulgate visibility
protection regulations that apply to ``each applicable implementation
plan'' (i.e., each SIP or FIP) \116\ for each State containing one or
more Class I areas and each State ``emissions from which may reasonably
be anticipated to cause or contribute to any impairment of
[[Page 102760]]
visibility in any [Class I area].'' The CAA specifies that these
regulations (including the RHR) must require each such SIP or FIP to
``contain such emission limits, schedules of compliance and other
measures as may be necessary to make reasonable progress toward meeting
the national goal.'' \117\ Thus, the statute provides the EPA a key
oversight role in reviewing SIPs, including regional haze SIPs, and the
``EPA has substantive authority to assure that a state's proposals
comply with the Act, not simply the ministerial authority to assure
that the state has made some determination.'' \118\
---------------------------------------------------------------------------
\113\ CAA section 110(a)(1), (k)(3), 42 U.S.C. 7410(a)(1) and
(k)(3).
\114\ CAA section 110(l), 42 U.S.C. 7410(l).
\115\ CAA sections 110(a)(2)(J), 169A and 169B 42 U.S.C.
7410(a)(2)(J), 7491 and 7492.
\116\ Under the CAA, ``applicable implementation plan'' is
defined as ``the portion (or portions) of the implementation plan,
or most recent revision thereof, which has been approved under [CAA
section 110], or promulgated under [CAA section 110](c) . . . and
which implements the relevant requirements of [the CAA].'' CAA
section 302(q), 42 U.S.C. 7602(q). In other words, an ``applicable
implementation plan'' is an EPA-approved SIP or Tribal
Implementation Plan, or an EPA-promulgated FIP.
\117\ 42 U.S.C. 7491(b)(2).
\118\ Arizona ex. rel. Darwin v. EPA, 815 F.3d 519, 531 (9th
Cir. 2016).
---------------------------------------------------------------------------
Nothing in the Oklahoma case cited by the commenter undermines this
authority. On the contrary, the Oklahoma court upheld the EPA's
disapproval of BART determinations that were part of a regional haze
SIP, noting BART ``does not differ from other parts of the CAA--states
have the ability to create SIPs, but they are subject to EPA review.''
\119\ Likewise, all regional haze SIPs are subject to EPA review, as
described in the preceding paragraph.
---------------------------------------------------------------------------
\119\ Oklahoma v. EPA, 723 F.3d 1201, 1209 (10th Cir. 2013).
---------------------------------------------------------------------------
In our action on the Plan, we are exercising our substantive
authority to review the state's submittal for compliance with the
applicable requirements of the CAA and RHR. Based on our review of the
Plan, we find, among other things, that ADEQ had not reasonably
evaluated and weighed the four statutory factors for SGS Units 1 and 2
for the reasons noted in section IV.E.2 of our proposal and responses
B.4., B.6, B.7 and B.8 of this document.
Comment D.4. TEP states that ADEQ's choice of cost-effectiveness
threshold was reasonable, and that the EPA's proposed rule did not
afford appropriate deference to ADEQ's assessment of reasonable cost-
effectiveness values. TEP also cites a previous EPA action stating that
``Congress did not provide any direction as to how states should
consider `the costs of compliance' when determining reasonable
progress.'' \120\ TEP also provides the following reasons for why it
believes ADEQ's choice of cost-effectiveness threshold was reasonable.
---------------------------------------------------------------------------
\120\ 81 FR 296, 310 (January 5, 2016).
---------------------------------------------------------------------------
First, TEP states that ADEQ selected a threshold based on the
State's evaluation of the highest cost controls during the first
planning period. TEP notes that this threshold is nearly $1,000/ton
higher than the 98th percentile value for EGU boilers during the first
planning period and $1,500/ton higher than costs rejected by Georgia in
evaluating additional emissions reduction measures in its SIP
submission.\121\
---------------------------------------------------------------------------
\121\ 89 FR 47494.
---------------------------------------------------------------------------
Second, TEP claims that ADEQ's bright-line approach to analyzing
available controls above its cost-effectiveness threshold was also
reasonable. TEP disagrees with the EPA's concern about the average cost
of installing SNCR at SGS Unit 2 being $269/ton above ADEQ's $6,500/ton
threshold.\122\ TEP states that the 2019 Guidance emphasized that the
RHR does not prevent states from implementing ``bright line'' cost-
effectiveness thresholds when considering additional control measures,
consistent with the Ninth Circuit's decision in NPCA v. EPA.\123\ TEP
asserts that this threshold sets an amount above which a state would
reject control options as too expensive, and that for controls falling
below this threshold, it is reasonable for ADEQ to evaluate additional
factors, such as incremental costs, visibility impacts, and the other
statutory factors, in determining whether these controls are necessary
for reasonable progress.
---------------------------------------------------------------------------
\122\ 89 FR 47429.
\123\ 2019 Guidance at 38 (citing NPCA v. EPA, 788 F.3d 1134,
1142-43 (9th Cir. 2015)).
---------------------------------------------------------------------------
Response D.4. As TEP describes, the EPA noted in prior rulemakings
that ``Congress did not provide any direction as to how states should
consider `the costs of compliance' when determining reasonable
progress.'' However, consistent with our prescribed statutory role, as
described in Response D.3, the EPA is required to evaluate whether each
State exercised its flexibility and conducted the required analyses in
a reasonable way and in accordance with the applicable requirements. As
described throughout our proposal and this final action, we determined
that Arizona did not do so.
For example, contrary to TEP's claim, the EPA did not find that
ADEQ's average cost-effectiveness thresholds of $6,500/ton for point
sources and $5,000/ton for nonpoint sources were unreasonable. Rather,
as explained in our proposal, we found that the State inconsistently
and unreasonably applied these cost thresholds to the control
determinations.\124\
---------------------------------------------------------------------------
\124\ 89 FR 47398, 47429.
---------------------------------------------------------------------------
In particular, we found that ADEQ incorrectly characterized its
average cost effectiveness threshold for points sources as a bright-
line threshold,\125\ given that ADEQ, in fact, rejected controls that
were below this threshold. In these instances, ADEQ relied on
incremental cost effectiveness as a basis to find the cost of control
excessive, but, again, did not do so consistently, as described in
Response B.6 of this document. Thus, ADEQ did not consistently apply
either its chosen average cost effectiveness threshold or any defined
incremental cost effectiveness threshold as a basis to choose whether
to adopt or reject control measures, nor did it explain its reasoning
for these inconsistencies. The EPA finds this to be unreasonable. Based
on this and other flaws in the long-term strategy described in the
proposal and elsewhere in this document, we find that the long-term
strategy in the Plan did not meet the requirements of 40 CFR
51.308(f)(2).
---------------------------------------------------------------------------
\125\ See 2022 Arizona Regional Plan, Appendix K, p. 7.
---------------------------------------------------------------------------
We also note that, while our proposed rule recommended that ADEQ
revisit control determinations with controls that were slightly above
the cost thresholds, given the flaws in the cost-effectiveness
analyses, we did not indicate that, in the absence of such flaws, use
of a bright line threshold would have been improper, as long it was
appropriately justified and consistently applied. Finally, we note that
the fact that other states have applied lower or similar thresholds
does not automatically make Arizona's threshold reasonable.
Comment D.5. TEP states that ADEQ's evaluation of control costs is
conservative and results in projected costs that are lower than what
TEP would incur. TEP asserts that ADEQ selected baseline assumptions
consistent with the EPA's Control Cost Manual. In particular, TEP
indicates that ADEQ calculated control costs using an assumed interest
rate of 4.75 percent and a maximum equipment life of 30 years, but that
these assumptions resulted in unrealistically low control cost
estimates. TEP noted that the EPA previously used 20-year equipment
life and 7 percent interest rate for both NO<INF>X</INF> and
SO<INF>2</INF> candidate emissions reduction measures for IGS as part
of a previous regional haze FIP,\126\ and that these assumed values of
7 percent and 20 years better reflect actual control costs.
---------------------------------------------------------------------------
\126\ 79 FR 9318.
---------------------------------------------------------------------------
TEP also asserted that ADEQ is not required to treat the guidance
provided by the Control Cost Manual as dispositive, and, citing Wyoming
v. EPA, 78 F.4th 1171, 1180 (10th Cir. 2023), TEP asserted that the EPA
cannot treat nonbinding guidelines as mandatory in evaluating a SIP
submission.\88\
[[Page 102761]]
Response D.5 We do not agree that ADEQ's evaluation of costs was
conservative and results in projected costs that are lower than what
TEP would incur. Regarding the EPA's previous analyses for IGS, first,
we note that the EPA has revised several chapters of the Control Cost
Manual, since the 2014 regional haze FIP cited by TEP in the comment
letter. In particular, the chapter concerning SCR was updated in
2019.\127\ As part of that update, the remaining useful life for SCR
was revised from 20 years to 30 years.\128\ Consistent with this
change, ADEQ used a remaining useful life of 30 years for SCR, but used
20 years for SNCR.\129\
---------------------------------------------------------------------------
\127\ Control Cost Manual, Section 4, Chapter 2 Selective
Catalytic Reduction (June 2019), section, 2.4.2 Total Annual Costs,
Indirect costs, available at <a href="https://www.epa.gov/sites/default/files/2017-12/documents/scrcostmanualchapter7thedition_2016revisions2017.pdf">https://www.epa.gov/sites/default/files/2017-12/documents/scrcostmanualchapter7thedition_2016revisions2017.pdf</a>.
\128\ Id. at page 87 of 107 (PDF document pagination), (``a
representative value of the equipment life for SCR at power plants
can be considered as 30 years'').
\129\ Id., Appendix C, p. 219 (``The estimated life for SCR and
SNCR were set at 30 and 20 years respectively to match current EPA
guidance for these control technologies on utility boilers.'') The
commenter's citation to page 146 of the Plan appears to be in error,
as there is no such page in the main body of the Plan, and page 146
of Appendix C discusses the Willcox compressor station.
---------------------------------------------------------------------------
Second, as explained in the Control Cost Manual,\130\ interest
rates change with time due to changes in prices over time for all
relevant goods and services such as capital equipment, engineering
services, other materials and reagents used in the construction and
operation of control equipment. In the absence of source-specific
information, ADEQ relied on a 4.75 percent interest rate developed by
analyzing and averaging historical bank prime rate data. ADEQ looked at
3-year average bank prime rates for the periods of 2017-2019 (4.83
percent) and April 2018-March 2020 (4.78 percent). These dates were
chosen as they were the most recent data at the time of the
analysis.\131\ In contrast, ADEQ also explained in its SIP submittal
that the 7 percent interest rate from the first planning period FIP
cited by TEP was the 3-year average bank prime rate during 2005-
2007.\132\ Therefore, the 7 percent rate used by the EPA in our
previous FIP was not appropriate for the cost analysis for the Plan,
which was developed in 2020-2022.
---------------------------------------------------------------------------
\130\ Control Cost Manual, Section 4, Chapters 1 and 2,
available at <a href="https://www.epa.gov/economic-and-cost-analysis-air-pollution-regulations/cost-reports-and-guidance-air-pollution">https://www.epa.gov/economic-and-cost-analysis-air-pollution-regulations/cost-reports-and-guidance-air-pollution</a>.
\131\ 2022 Arizona Regional Haze Plan, Chapter 8.3.2.
\132\ Id.
---------------------------------------------------------------------------
Finally, we disagree with the commenter's suggestion that we
treated the Control Cost Manual, or any other guidance, as binding. As
discussed in Responses A.3 and C.4, the EPA's citations to guidance
documents were intended to provide further context on what is generally
considered to be a reasonable approach to fulfill the statutory and
regulatory requirements. We acknowledge that the suggestions in those
guidance documents are not binding, but are generally assumed to be
reasonable. States can deviate from the suggestions within EPA guidance
documents. However, they must do so in a reasonable way, accompanied by
sufficient justification. The Plan did not do so for the reasons
described in our proposal and elsewhere in this document.
Comment D.6. TEP asserts that ADEQ's analysis of SCR and SNCR
reasonably approximated emissions limits achievable at SGS Units 1 and
2. TEP explains that the emissions rate used in the analysis was unit-
specific, and that ADEQ considered baseline NO<INF>X</INF> emissions,
inlet concentrations, and the frequency of startup/shutdown cycles at
SGS Units 1 and 2. TEP reiterates that ADEQ explained that a higher
number of startup/shutdown cycles results in higher average
NO<INF>X</INF> emissions rates with SCR, and that SGS Units 1 and 2
experience a higher frequency of startup/shutdown cycles than average
EGUs.
TEP further disagrees that SGS Units 1 and 2 are able to achieve an
emissions rate of 0.05 lb/MMBtu. TEP notes that ADEQ further explained
in its response to public comments that it was inappropriate to use an
annual emissions limit of 0.05 lb/MMBtu because the CAMPD database only
contains three comparable units, out of fifty-eight units, which are
operating below this emissions limit in 2019-2021. The CAMPD database
shows that approximately 20% of the units achieved emissions below 0.06
lb/MMBtu in 2019-2021, which ADEQ believed was a reasonable estimate of
the rate achievable with SCR for SGS Units 1 and 2. TEP also notes that
based on their extensive industry experience, vendors will not
guarantee a rate of 0.05 lb/MMBtu due to concerns about degradation
over time.
TEP also states that even if ADEQ adopted the EPA's preferred
emissions rate of 0.05 lb/MMBtu, ADEQ's analysis of control measures
for SGS Units 1 and 2 would remain substantially the same. Table 3 of
TEP's comment letter provides the updated cost-effectiveness value
using a 0.05 lb/MMBtu rate for SGS Units 1 and 2, assuming a 4.75
percent interest rate and a remaining useful life of 30 years.
Response D.6. We do not agree that ADEQ's analysis of SCR and SNCR
reasonably approximated emissions limits achievable at SGS Units 1 and
2. Our proposed rule acknowledged the startup/shutdown considerations
noted by ADEQ.\133\ However, the proposed rule also explains that ADEQ
has not demonstrated why these startup/shutdown considerations would be
significant enough at SGS Units 1 and 2 on an annual average basis,
which is the averaging period used to calculate ton/year emissions
reductions for cost effectiveness calculations, to preclude them from
achieving this emissions reduction level with SCR. As discussed in more
detail in Response B.4, we would consider it appropriate for an
emissions limitation established on a shorter averaging period to have
a higher value to account for startup/shutdown emissions, which have a
greater effect on overall unit emissions rates over shorter averaging
periods.
---------------------------------------------------------------------------
\133\ 89 FR 47398, 47428.
---------------------------------------------------------------------------
We also disagree with TEP that ADEQ's review of emissions rates for
other tangentially fired coal units in CAMPD should be determinative of
the SCR emissions rate achievable for the SGS units. We do not
necessarily consider the inventory of unit emissions data in CAMPD to
be representative of what is achievable with SCR technology because
units whose emissions are included in CAMPD have been required to
install SCR as a result of a variety of regulatory programs. Not all of
these programs may have fully considered technological factors in
establishing emissions requirements or allowable emissions limits,
which would have the effect of elevating reported emissions rates. For
example, several units in the eastern U.S. have installed SCR but have
not been required to operate in a manner that fully accounts for
periods of startup operations,\134\ while other units may only operate
their SCR systems seasonally.\135\ Other SCR systems may have been
required by a Consent Decree, which involves a
[[Page 102762]]
negotiated settlement in which allowable emissions limits may be
established as part of injunctive relief, and may not necessarily be
representative of SCR technical capabilities. Therefore, we find that
ADEQ's analysis of other tangentially fired coal units in CAMPD is not
determinative of the SCR emissions rate achievable for the SGS units.
---------------------------------------------------------------------------
\134\ 87 FR 31798 (May 25, 2022) and Docket Item EPA-R03-OAR-
2022-0347-0059 for further details.
\135\ Docket Items EPA-HQ-OAR-2021-0668-0115 and EPA-HQ-OAR-
2021-0668-0096 examine ozone season data from several mid-atlantic
states and identifies best performing months for several units. An
examination of operating data on a monthly or daily basis over the
course of a calendar year indicates that several units, such as
Pleasants Power Station 1 and 2 (WV), Conemaugh 2 (PA), and DB
Wilson W1 (KY) operate at much lower NO<INF>X</INF> emission rates
during ozone season months.
---------------------------------------------------------------------------
Finally, though we appreciate the summary provided in Table 3 of
the comment letter, the analysis is new information not included in the
Plan. This information would need to be part of a SIP revision subject
to review by the public and FLMs in order for the EPA to consider it as
part of the long-term strategy.
Comment D.7. TEP asserts that ADEQ reasonably determined that
SO<INF>2</INF> emissions limits were an appropriate and equivalent
control measure compared to SDA upgrades. TEP notes that while the EPA
suggests that TEP should also be required to install the SDA upgrades
in addition to the mass-based limits, the EPA does not explain why SDA
upgrades would achieve emissions reductions beyond the mass-based
limits adopted by TEP, which are equivalent to SDA upgrades.
Further, TEP states that the selection of emissions limits is
consistent with the CAA and is entitled to deference. First, TEP
reasons that the CAA does not require the installation of specific
control technologies, and that the EPA has recognized in the 2019
Guidance that mass-based limits may be a reasonable alternative to
specific control technologies, particularly where fixed capital costs
are high, so long as such limits do not enable a source to cease
operating an existing control technology. Second, TEP asserts that the
emissions limits are equally or more protective than control
technologies because the total SO<INF>2</INF> emitted cannot exceed the
caps, and whether one or both units is operating is irrelevant, and
because it provides the flexibility for greater total emissions
reductions to occur, should TEP ultimately retire SGS Unit 1 in 2027.
The commenter specifies that if TEP elects to achieve compliance with
the proposed caps by shutting down Unit 1, it is projected to reduce
SO<INF>2</INF> emissions by 2,982 tpy and achieve significant
reductions in NO<INF>X</INF> and PM emissions, whereas the installation
of SDA upgrades at both units is projected to reduce SO<INF>2</INF>
emissions by 2,122 tpy. Third, TEP asserts that the emissions limits
are consistent with EPA precedent. TEP listed previous examples of EPA-
approved mass-based emissions limits, such as the Hawaii Regional Haze
FIP for three Hawaiian Electric facilities \136\ and the Regional Haze
FIP for PCC.\137\
---------------------------------------------------------------------------
\136\ 77 FR 61478 (October 9, 2012).
\137\ 79 FR 52420.
---------------------------------------------------------------------------
Response D.7. We disagree with TEP's assertions that the selection
of mass-based SO<INF>2</INF> emissions limits for SGS Units 1 and 2 was
reasonable for the reasons described in Section IV.E.2.c.iii of our
proposal,\138\ and Response B.8 of this document.
---------------------------------------------------------------------------
\138\ 89 FR 47398, 47430-47431.
---------------------------------------------------------------------------
We also disagree with TEP's assertion that the EPA did not explain
why SDA upgrades would achieve emissions reductions beyond the mass-
based limits adopted by ADEQ. As stated in the proposed rule,\139\
because the limits are set across two units and the ton per day (tpd)
limit is based on a 30-calendar-day average (rather than a 30-day-
boiler-operating day average),\140\ they would not meaningfully
constrain the emissions from one unit during periods when the other
unit is not operating. In particular, the annual SO<INF>2</INF> cap of
3,739 tpy is significantly higher than ADEQ's projected 2028
SO<INF>2</INF> emissions for either Unit 1 or Unit 2 (2,869 and 2,982
tpy, respectively) \141\ and nearly double each unit's recent emissions
(1,980 and 1,988 tpy respectively on average 2021-2023).\142\
Accordingly, if TEP shuts down SGS Unit 1 by 2028, as it has stated it
intends to do,\143\ Unit 2 would be able to emit 3,739 tpy
SO<INF>2</INF> in 2028, nearly double what it emitted on average in
2021-2023 and significantly more than the 2,982 tpy it is projected to
emit in the absence of a cap and closure of Unit 1. In contrast, a lb/
MMBtu limit representing SDA upgrades on each unit would ensure
emissions from Unit 2 would be reduced by approximately \1/3\ from
recent emissions levels (i.e., a reduction of roughly 663 tpy) even if
Unit 1 ceases operation. Therefore, under the scenario projected by TEP
to occur in 2028 (i.e., closure of SGS Unit 1), implementation of SDA
upgrades at Unit 2 would achieve significant emissions reductions
beyond the mass-based limits adopted by ADEQ.
---------------------------------------------------------------------------
\139\ Id.
\140\ A limit based on boiler operating days would effectively
exclude days with zero emissions from the calculation of the 30-day
average whereas a limit based on calendar days does not.
\141\ Plan, appendix C, p. 213, Table 83.
\142\ Emissions information can be publicly accessed through the
EPA Clean Air Markets Program data, available at <a href="https://campd.epa.gov/">https://campd.epa.gov/</a>.
\143\ As part of its preferred alternative in its 2023
Integrated Resources Plan, p. 56, TEP states that ``Initially, the
units will alternate idling between spring and fall (both seasons
include the adjacent winter months). TEP plans to transition Unit 1
to summer-only operations prior to full retirement at the end of
2027.''
---------------------------------------------------------------------------
We also note that, if SGS Unit 1 does not close and both units
continue operation at roughly the same level as 2021-2023, a lb/MMBtu
limit representing SDA upgrades on each unit would ensure emissions
from both Unit 1 and Unit 2 would be reduced by approximately \1/3\
(i.e., a reduction of roughly 1,323 tpy based on 2021-2023 emissions),
whereas ADEQ's annual cap of 3,739 tpy would only ensure reductions of
229 tpy (i.e., compared to total 2021-2023 annual average emissions of
3,968). Therefore, we expect that lb/MMBtu limits representing SDA
upgrades on each unit would achieve significantly greater emissions
reductions than the two-unit mass-based limits adopted by ADEQ.
For similar reasons, we also do not agree with TEP's statements
that ``the proposed form of the limit is potentially more
environmentally protective than the installation of controls because it
provides the flexibility for greater total emission reductions to
occur, should TEP ultimately retire SGS Unit 1 in 2027'' and that
``[r]equiring the use of air pollution control technology at each unit
individually would foreclose a more environmentally beneficial
compliance option.'' Specifically, TEP incorrectly cites Table 83 as
supporting its assertion that if it ``elects to achieve compliance with
the proposed caps by shutting down Unit 1, it is projected to reduce
SO<INF>2</INF> emissions by 2,982 tpy and also to achieve significant
reductions in NO<INF>X</INF> and PM emissions.'' This mischaracterizes
the content of Table 83, which reflects ADEQ's baseline projection of
emissions at the TEP units, based on emissions and throughput data for
2016, 2018 and 2019.\144\ These values were the starting point for
ADEQ's four-factor analysis and do not represent the projected
emissions of these units following implementation of the emissions caps
and/or the closure of Unit 1. No such projection is included in the
Plan. However, as noted in the preceding paragraph, if TEP shuts down
SGS Unit 1 by 2028, as it has stated it intends to do, Unit 2 could
emit as much as 3,739 tpy SO<INF>2</INF> in 2028, nearly double what it
emitted on average in 2021-2023 and significantly more than the 2,982
tpy it is projected to emit in the absence of a cap and closure of Unit
1.
---------------------------------------------------------------------------
\144\ Plan, Appendix C, pp. 212.
---------------------------------------------------------------------------
Third, in response to TEP's claim about EPA precedent, both FIP
actions cited by TEP apply to first planning period requirements and
thus preceded the 2017 revisions to the RHR, which added 40 CFR
51.308(f). Please see Response B.8 for more information
[[Page 102763]]
about PCC. The other precedent cited by ADEQ, a cap on the Hawaiian
Electric sources, was not based solely on a determination that a
particular control measure was necessary to make reasonable progress
under the four statutory factors. Rather, the cap was intended
primarily to ensure that no degradation in visibility conditions would
occur at the affected Class I area during the first or subsequent
planning periods, as required under the RHR.\145\ This was based on
circumstances that were specific to Hawaii in the first planning
period, namely, that no photochemical modeling had been performed for
Hawaii's Class I areas and the EPA set the RPGs for these areas based
on island-specific emissions inventories.\146\ None of these
circumstances apply to Arizona generally, or to SGS Units 1 and 2
specifically, in the second regional haze planning period and thus are
inappropriate to rely upon.
---------------------------------------------------------------------------
\145\ 77 FR 31692, 31712 (May 29, 2012) (``without further
control, emissions of SO<INF>2</INF> on the Big Island are projected
to increase by nearly 4% between 2005 and 2018. Therefore,
additional, federally enforceable SO<INF>2</INF> reductions are
needed on the Big Island to ensure reasonable progress.'')
\146\ Id. at 31708.
---------------------------------------------------------------------------
Comment D.8. TEP comments that ADEQ reasonably rejected additional
control measures based on the four factors. For remaining useful life
specifically, TEP indicates that a 20-year remaining useful life is
more appropriate for SGS Units 1 and 2 because the shutdowns are
publicly documented in the 2023 IRP. TEP cites the 2021 Clarifications
Memo as stating, ``reasonable bases for projecting that future
emissions will be significantly different than past emissions are
enforceable requirements and energy efficiency, renewable energy, or
other similar programs, where there is a documented commitment to
participate and a verifiable basis for quantifying changes in future
emissions.'' \147\ TEP noted that consistent with this guidance, TEP's
planned retirement of these units significantly shortens the remaining
useful life of any controls and projected emissions reductions, and a
20-year remaining useful life is a reasonable, conservative basis to
analyze cost-effectiveness of additional controls.
---------------------------------------------------------------------------
\147\ 2021 Clarifications Memo, p. 12.
---------------------------------------------------------------------------
Response D.8. We disagree that 20 years is a reasonable remaining
useful life for SGS Units 1 and 2. The portion of the 2021
Clarifications Memo cited by TEP concerns setting emissions limits for
control measures found necessary to make reasonable progress, not
remaining useful life. Therefore, this citation does not support TEP's
assertion. Additionally, the 2019 Guidance and Clarifications Memo
clearly indicate that, under the RHR, where a shutdown date is used to
shorten a source's remaining useful life as part of a reasonable
progress determination, an enforceable requirement to shutdown must be
included in the SIP and/or be federally enforceable.\148\ The potential
shutdowns of SGS Units 1 and 2 are not federally enforceable.
Therefore, they cannot be relied upon to shorten the remaining useful
life of these units. In the absence of an enforceable requirement for
SGS Units 1 and 2 to retire, we do not agree that a 20-year remaining
useful life is reasonable.
---------------------------------------------------------------------------
\148\ 2019 Guidance, p. 34 (citing 40 CFR 51.308(f)(2));
Clarifications Memo, p. 10.
---------------------------------------------------------------------------
Comment D.9. TEP comments that ADEQ reasonably rejected additional
control measures in part by considering incremental costs. TEP noted
that ADEQ determined incremental costs for additional controls would
range from $9,400-13,500/ton, and that even the low end of this range
exceeded similar incremental costs that the EPA determined to be
excessive for sources in Arizona as part of the 2014 BART FIP.\149\
---------------------------------------------------------------------------
\149\ 79 FR 9318, 9341, Table 24 (rejecting DSI based on
incremental costs of $8,576/ton compared to lower sulfur fuel
blend).
---------------------------------------------------------------------------
TEP further states that ADEQ reasonably considered incremental
cost-effectiveness where the cost impacts were not clear due to
uncertainty in the remaining useful life of the additional controls.
Citing American Corn Growers Association vs. EPA, TEP states that it is
reasonable for a state to consider incremental visibility improvements
and other incremental metrics to inform its assessment of whether a
particular control is ``too costly . . . for a particular source.''
\150\ TEP notes that such an assessment does not give ``controlling
weight'' to the cost factor, but rather considers whether, on balance,
it would be unreasonable to require installation of a control,
consistent with the language of CAA section 169A.
---------------------------------------------------------------------------
\150\ American Corn Growers Association v. EPA, 291 F.3d 1, 6-7
(D.C. Cir. 2002).
---------------------------------------------------------------------------
Lastly, TEP asserts that ADEQ reasonably compared the costs of and
emissions reductions achievable with wet FGD and circulating dry
scrubbers (CDS) to the costs of and emissions reductions achievable
with SDA upgrades, even where such upgrades were not required to be
installed. The commenter argues that ADEQ reasonably used SDA upgrades
as a proxy for its mass-based emissions limits, since these limits were
developed based on equivalence with SDA upgrades and the EPA's approach
for Hawaiian Electric.
Response D.9. We disagree that ADEQ's use of incremental cost
effectiveness was reasonable for the reasons explained in Section
IV.E.2.C.IX of our proposal \151\ and Response B.6 of this document. We
also note that, while the commenter refers to ``similar incremental
costs that the EPA determined to be excessive for sources in Arizona as
part of the 2014 BART FIP,'' the accompanying citation refers to only
to a single source, i.e., Nelson Lime Kiln, which is the same source
cited by ADEQ in the Plan.
---------------------------------------------------------------------------
\151\ 89 FR 47398, 47429.
---------------------------------------------------------------------------
We also find that the commenter's citation to American Corn Growers
is inapposite. That decision concerned the EPA's interpretation of the
BART provisions of the CAA and turned on the fact that the CAA includes
the ``the degree in improvement in visibility that would be expected at
each Class I area as a result of imposing BART'' as one of five factors
to be considered in BART determinations.\152\ In contrast, the Act does
not mandate visibility improvement as one of the four factors to be
considered in determining reasonable progress.\153\ As discussed in
Response B.7, ADEQ indicated in its Response to Comments that it did
not consider this factor in its determinations.\154\ Furthermore, even
if ADEQ did consider visibility improvement in making control
determinations for SO<INF>2</INF> at SGS Units 1 and 2,\155\ it is
evident that ADEQ did not consider incremental visibility improvement
associated with Wet FGD compared to SDA upgrades. Rather, ADEQ
considered only ``a hypothetical SO<INF>2</INF> emission reduction of
3,236 tpy, which is approximately equivalent to 0.08 lb/MMBtu for SGS
Units 1 and 2.'' \156\ This level of reductions does not correspond
either to SDA upgrades or wet FGD, so ADEQ could not possibly have
considered the incremental visibility benefit between the levels of
control. Accordingly, American Corn Growers has no bearing on our
assessment of whether ADEQ's approach to its four factor analyses and
[[Page 102764]]
determinations for SGS Units 1 and 2 were reasonable.
---------------------------------------------------------------------------
\152\ Id. (interpreting CAA section 169A(g)(2)).
\153\ CAA 169A(g)(1).
\154\ 2022 Arizona Regional Haze Plan, Appendix K, p. 9.
\155\ See 2022 Arizona Regional Haze Plan, Appendix C, p. 234
(``small visibility benefits associated with the modeled
SO<INF>2</INF> controls supports the determination that CDS and wet
FGD control options are not necessary to make reasonable progress
towards natural visibility at Class I areas during this
implementation period'').
\156\ Id.
---------------------------------------------------------------------------
In addition, we also disagree that there is uncertainty in the
remaining useful life of SGS Units 1 and 2, because the potential
shutdowns at these units are not federally enforceable. See Response
D.8.
Finally, we disagree that ADEQ reasonably used SDA upgrades as a
proxy for its mass-based emissions limits. See Responses B.8 and D.7.
Comment D.10. TEP states that the EPA incorrectly implies that ADEQ
based its control determinations on visibility benefits, when ADEQ
stated otherwise in its SIP submission.\157\
---------------------------------------------------------------------------
\157\ 2022 Arizona Regional Haze Plan, Appendix K at 9
(explaining that visibility information ``was not considered in the
Department's emission control measure determination'').
---------------------------------------------------------------------------
Second, TEP notes that any consideration of visibility would be
consistent with recent EPA actions on regional haze and the text of the
statute. Citing the EPA's recent proposed action on the Georgia
Regional Haze SIP, TEP states that the EPA emphasized incremental
visibility improvements and that ``Georgia is also not contributing to
visibility impairment at any Class I areas above the URP.'' \158\ TEP
claims that because Arizona has similarly de minimis contributions to
visibility impairment at Class I areas,\159\ the EPA cannot apply
inconsistent criteria to its review of SIP submissions by different
states. Finally, citing Loper Bright Enterprises v. Raimondo, 144 S.
Ct. 2244, 2263 (2024), TEP argues that ``EPA's suggestion that
visibility should play little-to-no role in a state's assessment of
reasonable progress is unreasonable and counter to the text of the
statute.''
---------------------------------------------------------------------------
\158\ Citing 89 FR 47481, 47498.
\159\ 89 FR 47398, 47432, Table 21.
---------------------------------------------------------------------------
Response D.10. Regarding whether ADEQ considered visibility
benefits, in its control determinations, please see Response B.7.
We disagree that the EPA is applying inconsistent criteria to
review of SIP submissions by different states. As explained in Response
D.2, in the section of the Georgia action cited by the commenter, the
EPA considered overall trends in visibility impairment in evaluating
the reasonableness of Georgia's source selection methodology. This is
entirely different from weighing the potential visibility benefits of
specific controls at specific units, which is what the commenter
appears to be advocating for. Additionally, the quoted section of the
Georgia proposal simply states as a fact that Georgia is not
contributing to visibility impairment at any Class I areas above the
URP. That information is not used and should not be used as a ``safe
harbor'' argument to exclude contributing sources from a four-factor
analysis and potentially including cost-effective controls in the long-
term strategy.
Moreover, we do not agree that Arizona has de minimis contributions
to visibility impairment at Class I areas. In support of this
assertion, TEP cites Table 21 on page 47432 of the proposal. This table
shows Baseline Conditions, Adjusted URP and 2028 RPGs at Arizona's
Class I areas. However, the table provides no indication of which
sources contribute to visibility impairment at these Class I areas, or
whether these sources are located in Arizona or elsewhere. This table
also does not address any Class I areas outside of Arizona.
Furthermore, there is no other information in the Plan that suggests
that emissions from Arizona have de minimis contributions to visibility
impairment at all potentially affected Class I areas. Therefore, we do
not agree that the EPA is applying inconsistent criteria to review of
SIP submissions by different states.
Finally, we disagree that we suggested that ``visibility should
play little-to-no role in a state's assessment of reasonable
progress.'' Rather, we stated that ``[w]hether a particular visibility
impact is meaningful should be assessed in context and cannot be used
to undermine the four statutory factors that are to be analyzed in
order to determine what measures are necessary for reasonable
progress.'' \160\ Applying these considerations to ADEQ's discussion of
visibility benefits at SGS Units 1 and 2, we found that ``[i]n the
absence of any opportunities for larger emissions reductions and
corresponding visibility benefits, we find that ADEQ's reliance on
`small' visibility benefits as an additional justification for not
adopting more stringent controls at these units is not persuasive.''
\161\ Therefore, we do not agree with the commenter's characterization
of our proposal.
---------------------------------------------------------------------------
\160\ 89 FR 47398, 47430 (citing 40 CFR 51.308(f)(2)(i)).
\161\ Id.
---------------------------------------------------------------------------
Comment D.11. TEP asserts that ADEQ appropriately incorporated
``on-the-way'' measures in assessing baseline conditions at IGS Unit 3.
TEP cites the 2021 Clarifications Memo and states that the EPA has
explicitly recognized ``on-the-way'' measures that ``have not yet been
implemented and the associated emissions reductions have not yet
occurred as of the SIP submission date,'' and that the EPA indicated
that these measures may impact a state's choice of baseline for a four-
factor analysis at a particular source (``reasonable bases for
projecting that future emissions will be significantly different than
past emissions are enforceable requirements and energy efficiency,
renewable energy, or other similar programs, where there is a
documented commitment to participate and a verifiable basis for
quantifying changes in future emissions'').\162\ TEP claims that the
operational conditions at IGS Unit 3 have been incorporated in an
enforceable permit, and will become effective upon approval of
Arizona's SIP submission. TEP also claims that it has a documented
commitment to complying with these limits by requesting these permit
limits.
---------------------------------------------------------------------------
\162\ 2021 Clarifications Memo, pp. 10, 12.
---------------------------------------------------------------------------
Response D.11. TEP mischaracterizes the contents of the 2021
Clarifications Memo. The first section of the memo cited by the
commenter, entitled ``On-the-Way'' Measures and Shutdowns, states that
``on-the-way measures . . . are necessary to make reasonable progress
and must be included in a SIP.'' \163\ In this case, because the IGS
Unit 3 limits have not taken effect under State law, their inclusion as
part of the SIP revision is not meaningful, and they are not an
appropriate basis for modifying the baseline control scenario for a
four-factor analysis.
---------------------------------------------------------------------------
\163\ Id. p. 10.
---------------------------------------------------------------------------
The second portion of the 2021 Clarifications Memo cited by TEP
concerns setting emissions limits for control measures that have been
found necessary to make reasonable progress. It is not relevant to the
question of the extent to which existing measures may be considered as
part of the baseline for a control analysis. Regarding this latter
question, the 2019 Guidance states that:
Enforceable requirements are one reasonable basis for projecting
a change in operating parameters and thus emissions; energy
efficiency, renewable energy, or other such programs where there is
a documented commitment to participate and a verifiable basis for
quantifying any change in future emissions due to operational
changes may be another. A state considering using assumptions about
future operating parameters that are significantly different than
historical operating parameters should consult with its EPA Regional
office.\164\
---------------------------------------------------------------------------
\164\ 2019 Guidance p. 29 (emphases added).
Again, the emissions limits at IGS Unit 3 are not enforceable by
the State or the EPA unless and until the resulting reasonable progress
determination is approved into the SIP. Moreover, the permit conditions
that would implement the caps were adopted specifically to meet
regional
[[Page 102765]]
haze requirements and were not part of ``energy efficiency, renewable
energy, or other such programs.'' \165\ Therefore, we find that they
are not a reasonable basis for projecting a change in operating
parameters.
---------------------------------------------------------------------------
\165\ See, e.g., 2022 Arizona Regional Haze Plan, Appendix C, p.
197 (describing TEP's submittal of permit application for
NO<INF>X</INF> caps in response to ADEQ's initial regional haze
control determination); Appendix G, p. 58 (``The purpose of the
proposed SIP and significant permit revision is to support ADEQ's
forthcoming periodic comprehensive regional haze SIP submittal to
EPA.'').
---------------------------------------------------------------------------
Comment D.12. TEP asserts that ADEQ reasonably evaluated additional
control measures using a four-factor analysis for IGS Unit 3 based on
excessive costs. TEP explains that each of the NO<INF>X</INF> controls
available for IGS Unit 3 exceed Arizona's cost threshold of $6,500/ton,
and ADEQ therefore determined that no additional controls were
reasonable for the second planning period. TEP also noted that while
the average cost of installing combustion control retrofits at IGS Unit
3 was only $230/ton above ADEQ's $6,500/ton threshold, the RHR does not
prevent states from implementing ``bright line'' cost-effectiveness
thresholds when considering additional control measures.
Response D.12. We partly agree with this comment. We agree that the
NO<INF>X</INF> controls analyzed for IGS Unit 3 exceed the cost
threshold of $6,500, when the emissions limits in the permit (which are
not yet in effect) are considered as part of the baseline control
scenario. However, as described in section IV.E.2.c.iii of our
proposal, and Response D.11, we do not agree that these limits are an
appropriate basis for modifying the baseline control scenario for a
four-factor analysis. We also agree the RHR does not prevent states
from implementing ``bright line'' cost-effectiveness thresholds, but we
find that ADEQ did not do so in a reasonable or consistent manner in
the Plan, as discussed in Response D.4.
E. Comment Letter From SRP
Sections I (``Introduction''), II (``Background''), III (``SRP
Facilities Subject to the Proposed Rule'') and IV (``EPA Has
Appropriately Proposed Approval of Several Aspects of Arizona's
Regional Haze SIP'') of SRP's comment letter either provide background
information or are supportive and therefore do not require a response.
We respond to sections V-XI of SRP's comment letter below.
Comment E.1. SRP comments that the proposed rule is vague and
deprives the public of an adequate opportunity for comment. Citing
American Iron & Steel Institute v. EPA, SRP notes that the EPA is
obligated to provide the public with information sufficient to
facilitate public comment on the proposal, and that it is not clear
which grounds for the proposed partial disapproval of the SIP
submission apply to which sources. SRP provides a few examples of
allegedly vague language in the proposed rule, for example, noting
``some'' of Arizona's four-factor analyses were affected in a few
instances. SRP asserts that because it is not clear which aspects of
the EPA's analysis apply to which facilities, the public cannot
reasonably evaluate the grounds for the proposed disapproval actions,
and therefore the EPA needs to issue a new proposed rule that provides
additional necessary detail in support of the proposed actions.
Response E.1. We disagree with this comment. In American Iron &
Steel Institute,\166\ the court denied, in part, review of an EPA
interim final rule due to a lack of proper notice. However, as
described in Response A.1, we do not agree that the proposed rule
deprives the public of an adequate opportunity for comment.
---------------------------------------------------------------------------
\166\ See 568 F.2d 284 (3d Cir.1977), cert. denied, 435 U.S.
914, 98 S.Ct. 1467, 55 L.Ed.2d 505 (1978).
---------------------------------------------------------------------------
We also note that, in acting on the 2022 Arizona Regional Haze
Plan, the EPA is not issuing new requirements applicable to any
emissions sources. Rather, this final partial disapproval establishes a
two-year deadline for the EPA to promulgate a FIP to address the
relevant requirements under CAA section 110(c), unless the EPA approves
a subsequent SIP submission that meets these requirements. Accordingly,
Arizona can develop and submit a SIP revision addressing the
disapproved elements of the Plan. Both the state's adoption of that SIP
revision and the EPA's subsequent action on that SIP revision would be
subject to public notice and comment requirements.\167\ Similarly, if
the EPA does not fully approve a SIP submission addressing the
disapproved elements of the Plan, any FIP promulgated by the EPA would
be subject to public notice and comment. Accordingly, there will be
ample additional opportunities for the public, including potentially
regulated entities, to engage in the rulemaking process before any
additional requirements take effect.
---------------------------------------------------------------------------
\167\ See CAA section 110(l), 40 CFR 40 CFR 51.102, and 5 U.S.C.
553.
---------------------------------------------------------------------------
Comment E.2. SRP asserts that the EPA should approve Arizona's
determination that Coronado Generating Station and SGS Unit 4 are
effectively controlled because they are consistent with the law and EPA
guidance. SRP states that the 2019 Guidance indicates that sources that
have recently installed effective controls are the prime example of
sources that do not require evaluation during the current planning
period. Specifically, SRP notes that the 2019 Guidance states that
BART-eligible units that installed and began operating controls to meet
BART emissions limits for the first implementation period, including
sources that installed controls to comply with a better-than-BART
alternative, may be considered to be effectively controlled.
For Coronado, SRP states that the source is subject to a better-
then-BART alternative that the EPA approved in October 2017, and that
there have been no intervening changes in technology since that time.
SRP states that Coronado will also install new controls (SCR) to comply
with the BART alternative by January 2026.\168\ SRP noted that the
first planning period regional haze SIP submission for Coronado also
includes two additional requirements for SO<INF>2</INF>: an
SO<INF>2</INF> emissions limit of 0.060 lb/MMBtu, calculated on a 30-
boiler-operating-day rolling average and an annual plant-wide
SO<INF>2</INF> emissions cap of 1,970 tons per year.
---------------------------------------------------------------------------
\168\ SRP cited in the comment letter, but we note that the
correct citation is 82 FR 46903 (October 10, 2017).
---------------------------------------------------------------------------
SRP further asserts that SGS Unit 4 is similarly well-controlled.
ADEQ determined that Unit 4's existing controls of combustion controls
(LNB+OFA) and SCR is the most effective control technology available
for NO<INF>X</INF> for coal fired EGUs, and thus, no further analysis
for other control technologies was needed. For SO<INF>2</INF>, Unit 4
is equipped with SDA systems subject to the MATS rule and has been
achieving an SO<INF>2</INF> emissions rate over the most recent five
years ranging from 0.076 to 0.10 lb/MMBtu on an annual basis. For PM,
Unit 4 is equipped with a baghouse, which ADEQ found to have the
highest PM control efficiency of any PM control considered for Unit 4.
Response E.2. Regarding Coronado Generating Station, we agree that
the source was subject to a better-than BART alternative. However, we
note that the commenter mischaracterizes the contents of the 2019
Guidance regarding effective control determinations for BART-eligible
sources. The relevant portion of the Guidance includes as an example of
potentially effective controls, ``BART-eligible units that installed
and began operating controls to meet BART emission limits for the first
implementation period, on a
[[Page 102766]]
pollutant-specific basis.'' \169\ The Guidance further explains that:
---------------------------------------------------------------------------
\169\ See 2019 Guidance, p. 25 (emphasis added).
---------------------------------------------------------------------------
Although the Regional Haze Rule anticipates the re-assessment of
BART-eligible sources under the reasonable progress Rule provisions, if
a source installed and is currently operating controls to meet BART
emission limits, it may be unlikely that there will be further
available reasonable controls for such sources. However, states may not
categorically exclude all BART-eligible sources, or all sources that
installed BART controls, as candidates for selection for analysis of
control measures.
The associated footnote clarifies that this consideration is not
applicable to BART-subject units for which the BART requirement was met
in whole or in part by emissions reductions at other units as part of a
better-than-BART alternative or trading program.\170\
---------------------------------------------------------------------------
\170\ Id., n. 54.
---------------------------------------------------------------------------
As discussed in our proposal and Response B.3 of this document,
ADEQ excluded all units that installed BART (or better-than-BART)
controls between 2014 and 2028 for any pollutant. We therefore disagree
that Arizona followed the 2019 Guidance in evaluating effective
controls because it categorically excluded all such units without
considering whether the unit in question installed effective controls
for NO<INF>X</INF>, SO<INF>2</INF> and PM<INF>10</INF>, or whether the
BART requirement was met in part by reductions at other units as part
of a better-than-BART alternative.
Regarding SGS Unit 4, see Response D.2.
Comment E.3. SRP states that the EPA's assumed emissions rate for
SCR controls of 0.05 lb/MMBtu is not feasible at coal-fired EGUs. SRP
asserts that the EPA cannot assume all coal-fired units are capable of
the same efficiencies following pollution control installation or that
these controls incur the same costs for each unit nationwide,
irrespective of the local conditions and operations impacting
individual units. SRP further notes that the EPA's conclusion that all
SCR retrofitted units can uniformly meet a NO<INF>X</INF> emissions
limit of 0.05 lb/MMBtu without proper evaluation and consideration of
individual units is arbitrary and capricious and may unlawfully impose
limits on EGUs that are unachievable. SRP further points out that the
cited Srivastava et al. study notes units that achieved NO<INF>X</INF>
emissions rates between 0.04 and 0.07 lb/10\6\ Btu, and that this range
illustrates the variability associated with SCR-controlled coal-fired
unit NO<INF>X</INF> emissions rates and the importance of considering
unit-specific factors when identifying a controlled emissions rate.
Citing Appendix K of the Plan, SRP further asserts that ADEQ
specifically addressed unit-specific considerations when setting the
SGS Unit 1 and 2 NO<INF>X</INF> emissions rates by reviewing CAMPD
data. By assuming an SCR controlled NO<INF>X</INF> emissions rate of
0.06 lb/MMBtu, ADEQ is assuming that SGS Units 1 and 2 will achieve a
controlled emissions rate within the top 21 percent of tangentially-
fired EGUs. SRP indicates that without acknowledging ADEQ's evaluation,
the EPA states that ADEQ did not provide adequate justification and
unreasonably assumes an emissions rate within the top 5 percent of SCR
controlled tangentially-fired EGUs is appropriate for the SGS Unit 1
and 2. SRP concludes that while the state made reasonable conclusions
as to the emissions rates achievable by SCR at specific facilities, the
EPA has not.
Response E.3. We disagree that the EPA has assumed that all coal-
fired units are capable of the same efficiencies following pollution
control installation or that these controls incur the same costs for
each unit. As described in Responses B.4 and D.6, we have considered
unit-specific factors in evaluating the emissions rates achievable with
SCR at SGS Units 1 and 2, and we are not aware of any assertions that
SGS specifically cannot achieve 0.050 lb/MMBtu when operating with SCR
during periods of normal operation. Therefore, we find that ADEQ should
have considered a controlled NO<INF>X</INF> emissions rate of 0.050 lb/
MMBtu for SGS Units 1 and 2 when operating with SCR during periods of
normal operation. As further explained in Response B.4, this does not
mean that 0.050 lb/MMBtu would be an appropriate 30-BOD limit for these
units, but rather that it should be considered as annual emissions rate
for purposes of the control cost analysis.
Comment E.4. SRP comments that it is not necessary to include
control requirements for Coronado Generating Station in Arizona's
Regional Haze SIP. SRP states that under CAA section 169A, a state (or
the EPA) may only require a long-term strategy to include those control
measures that are found to be necessary to make reasonable progress
through the evaluation of the four statutory reasonable progress
factors. SRP claims that the EPA's position that in the absence of a
four-factor analysis supporting new controls, existing controls should
generally be deemed necessary to make reasonable progress and be
included in the regional haze SIP is not a valid reading of the CAA's
visibility provisions. Further, SRP claims that the CAA framework ties
reasonable progress controls to the four-factor analysis, and does not
leave room for the EPA's presumption that existing controls must be
included in a SIP even whenever a four-factor analysis failed to
identify new controls that should be implemented. Such existing
controls may only be deemed necessary components of a regional haze SIP
if a four-factor analysis independently identifies such controls as
necessary for reasonable progress.
SRP further cites both the 2021 Clarifications Memo (``There may be
other cases where, after having conducted robust source selection and
rigorous analysis of the four factors, states have not identified any
new measures that are reasonable to require for a source. In such
cases, states will have to address whether the source's existing
measures are necessary to make reasonable progress'') and 2019
Guidance: (``[i]f a state determines that an in-place emission control
at a source is a measure that is necessary to make reasonable progress
and there is not already an enforceable emission limit corresponding to
that control in the SIP, the state is required to adopt emission limits
based on those controls as part of its long-term strategy in the SIP
via the regional haze second planning period plan submission''). SRP
claims that the EPA therefore acknowledged that there should be no
presumption that existing measures are needed for reasonable progress.
Lastly, SRP states that Coronado Generating Station is already
subject to a source-specific SIP revision that was designed to
implement the better-than-BART alternative during the first planning
period of the regional haze program. SRP asserts that because these
requirements are already binding and enforceable, there is no need for
any additional action to address Coronado.
Response E.4. First, we disagree that existing controls being
necessary for reasonable progress does not have a basis in the statute
for two reasons. First, under CAA 169A(a)(1), the national visibility
goal is generally seperated into two parts: (1) the prevention of any
future, and (2) the remedying of any existing anthropogenic visibility
impairment. As noted in the proposed rule,\171\ and in response B.1,
continued implementation of the source's existing measures is generally
necessary to prevent future emissions increases and thus necessary
[[Page 102767]]
to make reasonable progress towards the national goal.
---------------------------------------------------------------------------
\171\ 89 FR 47398, 47404.
---------------------------------------------------------------------------
Second, control measures used to fulfill a CAA requirement must be
in the SIP.\172\ In this instance, in order to make reasonable progress
toward the national goal, the CAA requires every regional haze SIP to
contain ``such emission limits, schedules of compliance, and other
measures as may be necessary for reasonable progress.'' \173\ The CAA
also requires each regional haze SIP submission to include a long-term
strategy ``for making reasonable progress toward meeting the national
goal.'' \174\ Finally, reasonable progress is defined in the CAA as a
consideration of the four factors outlined in CAA 169A(g)(1). If the
State opts to avoid conducting the required consideration of the four
statutory factors on a source or group of sources based solely on the
source's existing measures, then, in order to fulfill its long-term
strategy requirements, those existing measures must also be in the SIP,
or else the state must demonstrate that the existing measures are not
necessary to make reasonable progress. This ensures that this source's
contribution to visibility impairment will not increase, and also
ensures that all measures being relied upon to fulfill the regional
haze requirements are in the SIP. Therefore, if Arizona is relying on
existing measures to avoid the statutorily required four factor
analysis, then those existing measures must be in the SIP and thus a
part of its long-term strategy for the second planning period, unless
the State demonstrates that they are not necessary to make reasonable
progress.
---------------------------------------------------------------------------
\172\ See Committee for a Better Arvin v. EPA, 786 F.3d 1169,
1175-77 (9th Cir. 2015).
\173\ CAA 169A(b)(2).
\174\ CAA 169A(b)(2)(B).
---------------------------------------------------------------------------
Additionally, we disagree that the EPA's guidance documents
acknowledged that there should be no presumption that existing measures
are needed for reasonable progress. The 2021 Clarifications Memo is
clear that, under the CAA and the RHR, ``when the outcome of a four-
factor analysis is that no new measures are reasonable for a source,
the source's existing measures are generally needed to prevent future
visibility impairment (i.e., to prevent future emission increases) and
thus necessary to make reasonable progress.'' \175\ ADEQ did not
provide this analysis of whether existing measures are necessary to
make reasonable progress.
---------------------------------------------------------------------------
\175\ 2021 Clarifications Memo, pp. 8-9 (emphasis added).
---------------------------------------------------------------------------
Finally, while we agree that Coronado Generating Station is already
subject to a source-specific SIP that was designed to implement a
better-than-BART alternative for the first implementation period, we do
not agree that this automatically means, without further justification,
that there is no need for any additional action to Coronado in the
second implementation period for the reasons described in E.2.
Comment E.5. SRP asserts that the EPA should approve Arizona's
control determinations and four-factor analyses with respect to cost-
effectiveness thresholds, incremental costs, compliance with the
Control Cost Manual, and consideration of visibility.
First, regarding cost-effectiveness thresholds, SRP indicates that
Arizona's cost-effectiveness threshold of $6,500/ton is reasonable and
should be approved, along with the control determinations the state
made in reliance on that threshold. SRP noted other examples of cost-
effectiveness thresholds in other states (Georgia and Arkansas) that
are lower than the values Arizona adopted. SRP also cites the EPA's
recent proposal for Missouri that cites a $6,060 to $7,600/ton
threshold from the Central Regional Air Planning Association and Texas
BART FIP threshold range of $5,300/ton to $6,500/ton that is consistent
with the $6,500/ton threshold adopted by Arizona. Therefore, SRP
concludes that ADEQ's threshold is reasonable and represents a
conservatively high threshold from a historical perspective.
Second, regarding incremental costs, SRP asserts that the EPA
provided no rationale for finding ADEQ's approach unreasonable, and
that it was appropriate to only consider incremental costs if a simple
dollar-per-ton analysis suggests a control might be cost-effective in
the absence of more nuanced information. SRP notes that suggesting that
states cannot reasonably take incremental costs into account to reject
control requirements far exceeds the EPA's statutory authority, citing
the court decision in American Corn Growers Association vs. EPA as
stating that ``[t]he Haze Rule calls for states to play the lead role
in designing and implementing regional haze program to clear the air in
national parks and wilderness areas.'' SRP further states that the EPA
states that Arizona considered only a single BART determination in
finding that incremental costs of $9,400-13,500/ton were excessive, but
this value is in line with past EPA actions finding incremental costs
excessive.\176\
---------------------------------------------------------------------------
\176\ See, e.g., 79 FR 9318, 9342 (rejecting controls based on
incremental costs of $8,803 and $8,576/ton); 76 FR 80754, 80756
(December 27, 2011) (rejecting controls based on incremental costs
of $5,367/ton).
---------------------------------------------------------------------------
Third, regarding compliance with the Control Cost Manual, SRP
states that the EPA did not provide enough information for the public
to tell what it considers to be the full range of the deviations from
the Cost Control Manual, except for some of the State's remaining
useful life values. SRP asserts that the EPA's rules specifically allow
for deviations from the Manual's methodology and sample calculations
and values whenever site-specific information is more accurate.
Lastly, regarding consideration of visibility, SRP asserts that the
EPA appears to place inappropriate limits on Arizona's consideration of
visibility impacts as part of its assessment of reasonable progress,
and that the State took visibility into account as additional
confirmation that controls were not reasonable. SRP also states that
the EPA has provided no context or analysis for stating that visibility
impacts must be assessed in context to determine if they are truly
meaningful and justify expensive control requirements. SRP points out
that Table 10-5 of the 2022 Arizona Regional Haze Plan shows that
natural visibility is projected to occur at all statewide sites between
2028 and 2056, well ahead 2064 natural conditions. SRP concludes that
Arizona appropriately determined that controls were not reasonable for
SGS when visibility is rapidly improving, costs are over or very near a
reasonable cost threshold, and where visibility improvements from
potential controls would be relatively small.
Response E.5. The EPA disagrees with this comment for a number of
reasons. First, with respect to cost-effectiveness thresholds, the EPA
clarifies that we are not disapproving Arizona's average cost-
effectiveness threshold specifically, but rather, finds that the State
inconsistently applied the threshold and did not adequately justify how
this approach resulted in a reasonable set of control measures in the
long-term strategy for the second planning period. We also note that
the fact that other states have applied lower or similar thresholds
does not automatically make Arizona's threshold reasonable. For further
explanation on this point, see Response D.4.
Second, with respect to incremental costs, contrary to the
commenter's assertion, our proposed rule did n
[…truncated; see source link]Indexed from Federal Register on December 18, 2024.
This is legal information, not legal advice. Laws vary by jurisdiction and change frequently. Always verify current law with official sources and consult a licensed attorney in your jurisdiction for advice on your specific situation.