Air Plan Approval; Arkansas; Arkansas Regional Haze and Visibility Transport State Implementation Plan Revisions

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Final rule.

SUMMARY:

Pursuant to the Clean Air Act (CAA or the Act), the Environmental Protection Agency (EPA) is finalizing approval of a revision to the Arkansas State Implementation Plan (SIP) submitted by the State of Arkansas through the Arkansas Department of Energy and Environment, Division of Environmental Quality (DEQ) on August 13, 2019. The SIP submittal addresses requirements of the Act and the Regional Haze Rule for visibility protection in mandatory Class I Federal areas (Class I areas) for the first implementation period. The EPA is approving an alternative measure to best available retrofit technology (BART) at the Domtar Ashdown Mill for sulfur dioxide (SO₂), particulate matter (PM), and nitrogen oxide (NO_X); and elements of the SIP submittal that relate to these BART requirements at this facility. In addition, we are approving the withdrawal from the SIP of the previously approved PM₁₀ BART limit for Power Boiler No. 1. The EPA is also concurrently approving Arkansas' interstate visibility transport provisions from the August 8, 2018, regional haze SIP submittal as supplemented by the visibility transport provisions in the October 4, 2019, interstate transport SIP submittal, which covers the following national ambient air quality standards (NAAQS): The 2006 24-hour fine particulate matter (PM_2.5) NAAQS; the 2012 annual PM_2.5 NAAQS; the 2008 and 2015 eight-hour ozone (O₃) NAAQS; the 2010 one-hour nitrogen dioxide (NO₂) NAAQS; and the 2010 one-hour SO₂ NAAQS. In conjunction with our final approval of these SIP revisions, we are finalizing in a separate rulemaking, published elsewhere in this issue of the Federal Register, our withdrawal of the Federal implementation plan (FIP) provisions for the Domtar Ashdown Mill.

DATES:

This rule is effective on April 21, 2021.

ADDRESSES:

The EPA has established a docket of all documents for this action at https://www.regulations.gov under Docket ID No. EPA-R06-OAR-2015-0189. Although listed in the index, some information is not publicly available, e.g., Confidential Business Information or other information whose disclosure is restricted by statute. Certain other material, such as copyrighted material, is not placed on the internet. Publicly available docket materials are available electronically through https://www.regulations.gov.

FOR FURTHER INFORMATION CONTACT:

James E. Grady, EPA Region 6 Office, Regional Haze and SO₂ Section, 1201 Elm Street, Suite 500, Dallas TX 72570, 214-665-6745; grady.james@epa.gov. Please call or email Mr. Grady or Mr. Bill Deese at 214-665-7253 if you need alternative access to material indexed but not provided in the docket.

SUPPLEMENTARY INFORMATION:

Throughout this document “we,” “us,” and “our” mean the EPA.

Table of Contents

I. Background

A. Regional Haze Principles

B. Requirements of the CAA and the EPA's Regional Haze Rule

C. BART Requirements

D. BART Alternative Requirements

E. Long-Term Strategy and Reasonable Progress Requirements

F. Previous Actions on Arkansas Regional Haze

G. Arkansas Regional Haze Phase III SIP Submittal

H. Arkansas Visibility Transport

II. Summary of Proposed Action and Our Final Decisions

III. Public Comments and EPA Responses

A. Demonstration That the BART Alternative Is Better-Than-BART

B. Monitoring, Recordkeeping and Reporting Requirements

C. Requirements for Emission Reductions To Occur During the First Implementation Period and for a Compliance Schedule

D. The CAA Section 110(l) Anti-Backsliding Provision

E. Interstate Visibility Transport and Regional Haze Reasonable Progress Requirements

F. Comments From Domtar

IV. Final Action

A. Arkansas Regional Haze Phase III SIP Submittal

B. Arkansas Visibility Transport

C. CAA Section 110(l)

V. Incorporation by Reference

VI. Statutory and Executive Order Reviews

I. Background

A. Regional Haze Principles

Regional haze is visibility impairment that is produced by a multitude of sources and activities that are located across a broad geographic area and emit fine particulates (PM_2.5) into the air. Fine particulates which cause haze are sulfates (SO₄ ^2-), nitrates (NO₃ ^-), organic carbon (OC), elemental carbon (EC), and soil dust. PM_2.5 precursors consist of SO₂, NO_X, volatile organic compounds (VOCs), and in some cases, ammonia (NH₃). Airborne PM_2.5 can scatter and absorb the incident light and, therefore, lead to atmospheric opacity and horizontal visibility degradation. Regional haze limits visual distance and reduces color, clarity, and contrast of view. PM_2.5 can cause serious adverse health effects and mortality in humans. It also contributes to environmental effects such as acid deposition and eutrophication. Emissions that affect visibility include a wide variety of natural and man-made sources. Natural sources can include windblown dust and soot from wildfires. Man-made sources can include major and minor stationary sources, mobile sources, and area sources. Reducing PM_2.5 and its precursor gases in the atmosphere is an effective method of improving visibility.

Data from the existing visibility monitoring network, “Interagency Monitoring of Protected Visual Environments” (IMPROVE), shows that visibility impairment caused by air pollution occurs virtually all of the time at most national parks and wilderness areas. In 1999, the average visual range in many mandatory Class I Federal areas in the western United States was 100-150 kilometers (km), or about one-half to two-thirds of the visual range that would exist under estimated natural conditions. In most of the eastern Class I areas of the United States, the average visual range was less than 30 km, or about one-fifth of the visual range that would exist under estimated natural conditions. Since the promulgation of the original Regional Haze Rule in 1999, CAA programs have reduced emissions of haze-causing pollution, lessening visibility impairment and resulting in improved average visual ranges.

B. Requirements of the CAA and the EPA's Regional Haze Rule

In section 169A, enacted as part of the 1977 CAA Amendments, Congress created a program for protecting visibility in the nation's national parks and wilderness areas. This section of the CAA establishes as a national goal the prevention of any future, and the remedying of any existing, visibility impairment in mandatory Class I Federal areas where impairment results from manmade air pollution. Congress added section 169B to the CAA in 1990, which strengthened the visibility protection program of the Act, and the EPA promulgated final regulations addressing regional haze as part of the 1999 Regional Haze Rule, which was most recently updated in 2017. The Regional Haze Rule revised the existing 1980 visibility regulations and established a more comprehensive visibility protection program for Class I areas. The requirements for regional haze, found at 40 CFR 51.308 and 51.309, are included in the EPA's broader visibility protection regulations at 40 CFR 51.300-309. The regional haze regulations require states to demonstrate reasonable progress toward meeting the national goal of restoring natural visibility conditions for Class I areas by 2064. The CAA requirement in section 169A(b)(2) to submit a regional haze SIP applies to all fifty states, the District of Columbia, and the Virgin Islands. States were required to submit the first implementation plan addressing visibility impairment caused by regional haze no later than December 17, 2007.

C. BART Requirements

Section 169A(b)(2)(A) of the CAA directs states to evaluate the use of BART controls at certain categories of existing major stationary sources built between 1962 and 1977. Under 40 CFR 51.308(e)(1)(ii), any BART-eligible source that is reasonably anticipated to cause or contribute to visibility impairment in a Class I area is classified as subject-to-BART. States are directed to conduct BART determinations to address visibility impacts for each source classified as subject-to-BART. These large, often under-controlled, older stationary sources are then required to procure, install, and operate the BART controls established in these determinations to reduce visibility impairment. The determinations must be based on an analysis of the best system of continuous emission control technology available and associated emission reductions achievable. States are required to identify the level of control representing BART after considering the five statutory factors set out in CAA section 169A(g)(2) for the potential BART controls. States must establish emission limits, a schedule of compliance, and other measures consistent with the BART determination process for each source subject-to-BART.

D. BART Alternative Requirements

A State may opt to implement or require participation in an emissions trading program or other alternative measure rather than require sources subject-to-BART to install, operate, and maintain BART. Such an emissions trading program or other alternative measure must achieve greater reasonable progress than would be achieved through the installation and operation of BART. In order to demonstrate that the alternative program achieves greater reasonable progress than source-specific BART, a state must demonstrate that its SIP meets the requirements in 40 CFR 51.308(e)(2)(i) to (iv). Among other things, the state must conduct an analysis of BART and the associated reductions for each source subject-to-BART covered by the alternative program, and compare the reductions and visibility improvements of the alternative program to what would have been achieved by BART.

Pursuant to 40 CFR 51.308(e)(2)(i)E), the state must provide a determination under 40 CFR 51.308(e)(3) or otherwise based on the “clear weight of evidence” that the alternative measure achieves greater reasonable progress than BART. 40 CFR 51.308(e)(3) provides two specific tests applicable under specific circumstances for determining whether the alternative measure achieves greater reasonable progress than BART. Under the first test, if the distribution of emissions is not substantially different than under BART, and the alternative measure results in greater emission reductions, then the alternative measure may be deemed to achieve greater reasonable progress. Under the second test, if the distribution of emissions is significantly different, then the State must conduct dispersion modeling to determine the difference in visibility between BART and the alternative measure for each impacted Class I area, for the twenty percent best and worst days. The modeling would demonstrate greater reasonable progress if both of the following two criteria are met: (i) Visibility does not decline in any Class I area, and (ii) there is an overall improvement in visibility, determined by comparing the average difference between BART and the alternative over all affected Class I areas.

Alternatively, under 40 CFR 51.308(e)(2)(i)(E), states may show based on the “clear weight of evidence” that the alternative achieves greater reasonable progress than would be achieved through the installation and operation of BART at the covered sources. As stated in the EPA's revisions to the Regional Haze Rule governing alternatives to source-specific BART determinations, weight of evidence demonstrations attempt to make use of all available information and data which can inform a decision while recognizing the relative strengths and weaknesses of that information in arriving at the soundest decision possible. This array of information and other relevant data must be of sufficient quality to inform the comparison of visibility impacts between BART and the alternative. A weight of evidence comparison may be warranted when there is confidence that the difference in visibility impacts between BART and the alternative scenarios are expected to be large enough to show that an alternative is better than BART. The EPA will carefully consider this evidence in evaluating any SIPs submitted by States employing such an approach.

Finally, under 40 CFR 51.308(e)(2)(iii) and (iv), all emission reductions for the alternative program must take place during the period of the first long-term strategy for regional haze, and all the emission reductions resulting from the alternative program must be surplus to those reductions resulting from measures adopted to meet requirements of the CAA as of the baseline date of the SIP.

E. Long-Term Strategy and Reasonable Progress Requirements

In addition to BART requirements, 40 CFR 51.308(d)(3)(i) to (iv) requires each state to include in its SIP a long-term strategy for the planning period that addresses regional haze visibility impairment for each Class I area located within the state and outside the state that may be affected by emissions generated from within the state. The long-term strategy is the vehicle for ensuring continuing reasonable progress toward achieving natural visibility conditions. It is a compilation of all control measures in the SIP that a state will use during the implementation period to meet the applicable reasonable progress goals (RPGs) established under 40 CFR 51.308(d)(1) for each Class I area. The RPGs established by the State provide an assessment of the visibility improvement anticipated to result for that planning period. Section 51.308(d)(3)(v) requires that a state consider certain minimum factors (the long-term strategy factors) in developing its long-term strategy for each Class I area. States have significant flexibility in establishing RPGs during the first planning period and must determine whether additional measures beyond BART are needed for reasonable progress. Under CAA section 169A(g)(1), once a set of potential control measures have been identified for a selected source, the State must collect data on and apply the four statutory factors that will be considered in selecting the measure(s) for that source that are necessary to make reasonable progress. The four statutory factors used to characterize potential emission controls are as follows: (1) The costs of compliance; (2) the time necessary for compliance; (3) the energy and non-air quality environmental impacts of compliance; and (4) the remaining useful life of any potentially affected sources. A state planning to consider visibility benefits will also need to characterize those benefits (often referred to as the 5th factor). States must demonstrate in their regional haze SIPs how these factors are considered when selecting the controls for their long-term strategies and provide an assessment of the visibility improvement anticipated to establish RPGs for each applicable Class I area. This is commonly referred to this as the “reasonable progress analysis” or “four-factor analysis.”

F. Previous Actions on Arkansas Regional Haze

The State of Arkansas submitted a regional haze SIP on September 9, 2008, intended to address the requirements of the first regional haze implementation period. On August 3, 2010, the State submitted a SIP revision with mostly non-substantive changes that addressed Arkansas Pollution Control and Ecology Commission (APCEC) Regulation 19, Chapter 15. On September 27, 2011, the State submitted a supplemental letter that clarified several aspects of the 2008 submittal. The EPA collectively refers to the original 2008 submittal, the supplemental letter, and the 2010 revision together as the 2008 Arkansas Regional Haze SIP. On March 12, 2012, the EPA partially approved and partially disapproved the 2008 Arkansas Regional Haze SIP. Specifically, the EPA disapproved certain BART compliance dates; the State's identification of certain BART-eligible sources and subject-to-BART sources; certain BART determinations for NO_X, SO₂, and PM₁₀; the State's reasonable progress analysis; and a portion of the State's long-term strategy. The remaining provisions of the 2008 Arkansas Regional Haze SIP were approved. The final partial disapproval started a two-year FIP clock that obligated the EPA to either approve a SIP revision and/or promulgate a FIP to address the disapproved portions of the SIP. Because a SIP revision addressing the deficiencies was not approved and the FIP clock expired in April 2014, the EPA promulgated a FIP (the Arkansas Regional Haze FIP) on September 27, 2016, to address the disapproved portions of the 2008 Arkansas Regional Haze SIP. Among other things, the FIP established SO₂, NO_X, and PM₁₀ emission limits under the BART requirements for nine units at six facilities: Arkansas Electric Cooperative Corporation (AECC) Carl E. Bailey Plant Unit 1 Boiler; AECC John L. McClellan Plant Unit 1 Boiler; American Electric Power/Southwestern Electric Power Company (AEP/SWEPCO) Flint Creek Plant Boiler No. 1; Entergy22 Lake Catherine Plant Unit 4 Boiler; Entergy White Bluff Plant Units 1 and 2 Boilers and the Auxiliary Boiler; and the Domtar Ashdown Mill Power Boilers No. 1 and 2. The FIP also established SO₂ and NO_X emission limits under the reasonable progress requirements for the Entergy Independence Plant Units 1 and 2.

Following petitions for reconsideration and administrative stay submitted by the State, industry, and ratepayers, on April 14, 2017, the EPA announced our decision to reconsider several elements of the FIP and on April 25, 2017, the EPA issued a partial administrative stay of the effectiveness of the FIP for ninety days. During that period, Arkansas started to address the disapproved portions of its regional haze SIP through several phases of SIP revisions. On July 12, 2017, the State submitted its Phase I SIP submittal (the Arkansas Regional Haze NO_X SIP revision) to address NO_X BART requirements for all electric generating units (EGUs) and the reasonable progress requirements with respect to NO_X. These NO_X provisions were previously disapproved by the EPA in our 2012 final action on the 2008 Arkansas Regional Haze SIP. The Arkansas Regional Haze NO_X SIP submittal replaced all source-specific NO_X BART determinations for EGUs established in the FIP with reliance upon the Cross-State Air Pollution Rule (CSAPR) emissions trading program for O₃ season NO_X as an alternative to NO_X BART. The SIP submittal addressed the NO_X BART requirements for Bailey Unit 1, McClellan Unit 1, Flint Creek Boiler No. 1, Lake Catherine Unit 4; White Bluff Units 1 and 2, and the Auxiliary Boiler. The revision did not address NO_X BART for Domtar Ashdown Mill Power Boilers No. 1 and 2. On February 12, 2018, we took final action to approve the Arkansas Regional Haze NO_X SIP revision and to withdraw the corresponding NO_X provisions of the FIP.

The State submitted its Phase II SIP revision (the Arkansas Regional Haze SO₂ and PM SIP revision) on August 8, 2018, that addressed most of the remaining parts of the 2008 Arkansas Regional Haze SIP that were disapproved in the March 12, 2012, action. The August 8, 2018, SIP submittal was intended to replace the federal SO₂ and PM₁₀ BART determinations as well as the reasonable progress determinations established in the FIP with the State's own determinations. Specifically, the SIP revision addressed the applicable SO₂ and PM₁₀ BART requirements for Bailey Unit 1; SO₂ and PM₁₀ BART requirements for McClellan Unit 1; SO₂ BART requirements for Flint Creek Boiler No. 1; SO₂ BART requirements for White Bluff Units 1 and 2; SO₂, NO_X, and PM₁₀ BART requirements for the White Bluff Auxiliary Boiler; and included a requirement that Lake Catherine Unit 4 not burn fuel oil until SO₂ and PM BART determinations for the fuel oil firing scenario are approved into the SIP by the EPA. The submittal addressed the reasonable progress requirements with respect to SO₂ and PM₁₀ emissions for Independence Units 1 and 2 and all other sources in Arkansas. In addition, it established revised RPGs for Arkansas' two Class I areas and revised the State's long-term strategy provisions. The submittal did not address BART and associated long-term strategy requirements for Domtar Ashdown Mill Power Boilers No. 1 and 2. On September 27, 2019, we took final action to approve a portion of the Arkansas Regional Haze SO₂ and PM SIP revision and to withdraw the corresponding parts of the FIP. The August 8, 2018, SIP also contained a discussion of the interstate visibility transport provisions, as discussed in more detail in Section I.H of this final action.

G. Arkansas Regional Haze Phase III SIP Submittal

On August 13, 2019, DEQ submitted the Arkansas Regional Haze Phase III SIP revision (Phase III SIP revision), which we are finalizing approval of in this action. This submittal contains an alternative measure to address BART and the associated long-term strategy requirements for two subject-to-BART sources (Power Boilers No. 1 and 2) at the Domtar Ashdown paper mill located in Ashdown, Arkansas. Power Boiler No. 1 was first installed in 1967-1968. At the time of SIP submittal and our proposed approval, the unit was permitted to burn only natural gas. It was capable of burning a variety of other fuels too, including bark, wood waste, tire-derived fuel (TDF), municipal yard waste, pelletized paper fuel, fuel-oil, and reprocessed fuel-oil, but was not authorized to do so. It was equipped with a wet electrostatic precipitator (WESP) but the requirements to operate the WESP were removed when the permit was modified to combust natural gas only. In 2020, DEQ received a disconnection notice for Power Boiler No. 1 and it is now permanently retired. Power Boiler No. 1 has a design heat input rating of 580 million British Thermal units per hour (MMBtu/hr) and an average steam generation rate of approximately 120,000 pounds per hour (pph). Power Boiler No. 2 was installed in 1975 and is authorized to burn a variety of fuels including coal, petroleum coke, TDF, natural gas, wood waste, clean cellulosic biomass (e.g. bark, wood residuals, and other woody biomass materials), and wood chips used to absorb oil spills. It is equipped with a traveling grate; a combustion air system that includes over-fire air; multi-clones for PM₁₀ removal; and two venturi scrubbers in parallel for removal of SO₂ and remaining particulates. Power Boiler No. 2 has a heat input rating of 820 MMBtu/hr and an average steam generation rate of approximately 600,000 pph.

DEQ's original BART analyses and determinations (dated October 2006 and March 2007) for Power Boilers No. 1 and 2 were included in the 2008 Arkansas Regional Haze SIP. In our 2012 partial approval/partial disapproval action, we approved DEQ's identification of these two units as BART-eligible; DEQ's determination that these units are subject-to-BART; and DEQ's PM₁₀ BART determination for Power Boiler No. 1. In that action, we also disapproved the SO₂ and NO_X BART determinations for Power Boiler No. 1; and the SO₂, NO_X, and PM₁₀ BART determinations for Power Boiler No. 2. In the 2016 Arkansas Regional Haze FIP and its associated technical support document (TSD), the EPA promulgated SO₂, NO_X, and PM₁₀ emission limits for these boilers. The FIP BART limits were based on consideration of the 2006 and 2007 BART analyses, a revised BART analysis (dated May 2014), and additional information provided by Domtar for the disapproved BART determinations. On March 20, 2018, Domtar provided DEQ with a proposed BART alternative based on changing boiler operations as part of the company's planned re-purposing and mill transformation from paper production to fluff pulp production. On September 5, 2018, Domtar further revised its BART alternative approach in response to additional boiler operation changes planned at the Ashdown Mill. In October 2018, DEQ proposed a SIP revision that included Domtar's BART alternative approach to address the BART requirements for Power Boilers 1 and 2 at the Ashdown Mill. The October 2018 proposal included an administrative order as the enforceable mechanism for the emission limits established under the BART alternative; and the order also contained monitoring, reporting, and recordkeeping requirements for the boilers. During the State's public comment period, Domtar submitted comments stating that while it agrees with the BART alternative approach and with the emission limits themselves, it does not agree with the use of the administrative order as the enforceable mechanism of the proposed SIP revision. Domtar requested that the portion of its New Source Review (NSR) permit containing the regional haze requirements be included in the proposed SIP revision as the enforceable mechanism instead of the administrative order. DEQ addressed Domtar's request in April 2019 by proposing a supplemental SIP revision to the October 2018 proposal. The supplemental SIP revision proposal replaced the administrative order with the incorporation of certain provisions of Domtar's revised NSR permit into the SIP as the enforceable mechanism for Domtar's regional haze requirements. On August 1, 2019, DEQ issued a final minor permit modification letter to Domtar, which included enforceable emission limitations and compliance schedules for the BART alternative.

DEQ submitted its third corrective regional haze SIP submittal to the EPA on August 13, 2019, which is the subject of this final action (the Arkansas Regional Haze Phase III SIP revision). The Phase III SIP revision includes Domtar's BART alternative approach and revises all of the prior BART determinations for Power Boilers No. 1 and 2 at the Ashdown Mill. The Phase III SIP submittal also incorporates plantwide provisions from the August 1, 2019, permit including emission limits and conditions for implementing the BART alternative. With final approval of the Arkansas Regional Haze Phase III SIP revision in this action, DEQ now has a fully-approved regional haze SIP for the first implementation period. The Arkansas Regional Haze NO_X SIP revision (Phase I SIP), the Arkansas Regional Haze SO₂ and PM SIP revision (Phase II SIP), and the Arkansas Regional Haze Phase III SIP revision together fully address all deficiencies of the 2008 Arkansas Regional Haze SIP that EPA previously identified in the March 12, 2012 partial approval/partial disapproval action.

H. Arkansas Visibility Transport

We are also addressing the interstate visibility transport element required under CAA section 110(a)(2)(D)(i)(II) in this final action from multiple SIP revisions for several NAAQS. Sections 110(a)(1) and (2) of the CAA direct each state to develop and submit to the EPA a SIP that provides for the implementation, maintenance, and enforcement of a new or revised NAAQS. This type of SIP submission is referred to as an infrastructure SIP. Section 110(a)(1) provides the timing and procedural requirements for infrastructure SIPs. Specifically, each state is required to make a new SIP submission within three years after promulgation of a new or revised primary or secondary NAAQS. Section 110(a)(2) lists the substantive elements that states must address for infrastructure SIPs to be approved by the EPA. Section 110(a)(2)(D)(i) includes four distinct elements related to interstate transport of air pollution, commonly referred to as prongs, that must be addressed in infrastructure SIP submissions. The first two prongs are codified in section 110(a)(2)(D)(i)(I) and the third and fourth prongs are codified in section 110(a)(2)(D)(i)(II). These four prongs prohibit any source or type of emission activities in one state from:

• Contributing significantly to nonattainment of the NAAQS in another state (prong 1);
• Interfering with maintenance of the NAAQS in another state (prong 2);
• Interfering with measures that prevent significant deterioration of air quality in another state (prong 3); and
• Interfering with measures that protect visibility in another state (prong 4 or “visibility transport”).

We are only addressing the prong 4 element in this final approval. The prong 4 element is consistent with the requirements in the regional haze program, which explicitly require each state to address its share of emission reductions needed to meet the RPGs for surrounding Class I areas. The EPA most recently issued guidance that addressed prong 4 on September 13, 2013. The 2013 guidance indicates that a state can satisfy prong 4 requirements with a fully-approved regional haze SIP that meets 40 CFR 51.308 or 309. Alternatively, in the absence of a fully-approved regional haze SIP, a state may meet the prong 4 requirements through a demonstration showing that emissions within its jurisdiction do not interfere with another air agency's plans to protect visibility. Lastly, the guidance states that prong 4 is pollutant-specific, so infrastructure SIPs only need to address the particular pollutant (including precursors) for which there is a new or revised NAAQS for which the SIP is being submitted that is interfering with visibility protection.

On March 24, 2017, the State submitted a SIP revision that addressed all four infrastructure prongs from section 110(a)(2)(D)(i) for the 2008 lead (Pb) NAAQS, the 2006 and 2012 PM_2.5 NAAQS, the 2008 O₃ NAAQS, the 2010 SO₂ NAAQS, and the 2010 NO₂ NAAQS. We deferred taking action on the 110(a)(2)(D)(i)(II) prong 4 portion of that infrastructure SIP for a future rulemaking with the exception of the 2008 Pb NAAQS. On August 8, 2018, the State also included a discussion on visibility transport in its regional haze Phase II SIP revision, but we deferred taking action on the visibility transport requirements in that submittal too. In the Phase II SIP revision, the State considered all Class I areas in Arkansas and also considered those in Missouri, which is the only State that was determined to potentially be impacted by sources from within Arkansas for the first implementation period. Missouri is currently not relying on emission reductions from Domtar in its regional haze plan. DEQ concluded that Missouri is on track to achieve its visibility goals; that observed visibility progress from Arkansas sources are not interfering with Missouri's RPG achievements for Hercules-Glades Wilderness and Mingo National Wildlife Refuge; and that no additional controls on Arkansas sources are necessary to ensure that other states' Class I areas meet their visibility goals for the first planning period. On October 4, 2019, the State submitted the Arkansas 2015 O₃ NAAQS Interstate Transport SIP revision to meet the requirements of CAA section 110(a)(2)(D) regarding interstate transport for the 2015 O₃ NAAQS. In that SIP submittal, Arkansas also addressed the 2006 and 2012 PM_2.5 NAAQS, the 2008 O₃ NAAQS, the 2010 SO₂ NAAQS, and the 2010 NO₂ NAAQS prong 4 visibility transport obligations in 110(a)(2)(D)(i)(II), and we are finalizing approval of those prong 4 requirements in this action. The State's prong 4 visibility transport analysis in the October 4, 2019 submittal supersedes the prong 4 visibility transport portion of the March 24, 2017, infrastructure SIP submittal and supplements the August 8, 2018, Phase II Arkansas Regional Haze SO₂ and PM SIP revision for the 2006 and 2012 PM_2.5 NAAQS, the 2008 and 2015 O₃ NAAQS, the 2010 SO₂ NAAQS, and the 2010 NO₂ NAAQS. All other applicable infrastructure SIP requirements in the October 4, 2019, SIP submission have been or will be addressed in separate rulemakings.

II. Summary of Proposed Action and Our Final Decisions

On March 16, 2020, we published a Notice of Proposed Rulemaking (NPRM) proposing to approve the Arkansas Regional Haze Phase III SIP revision submitted by DEQ on August 13, 2019. The SIP submittal addressed requirements of the Act and the Regional Haze Rule for visibility protection in mandatory Federal Class I areas for the first implementation period. The EPA proposed to approve an alternative measure to BART for SO₂, PM, and NO_X at the Domtar Ashdown Mill and elements of the SIP submittal that relate to these BART requirements at this facility. We are finalizing our determination in the NPRM that the Arkansas Regional Haze Phase III SIP revision meets all of the applicable regional haze BART alternative provisions set forth in 40 CFR 51.308(e)(2)(i) to (iv) for the Domtar Ashdown Mill. We are also finalizing our approval of specific plantwide permit provisions as the enforceable mechanism for the BART alternative emission limits and conditions for implementing the BART alternative. We are finalizing our approval of the reasonable progress components under 40 CFR 51.308(d) relating to Domtar Power Boilers No. 1 and 2. With the final approval of the BART alternative requirements for the Domtar Ashdown Mill in this action, DEQ has satisfied all long-term strategy requirements under section 40 CFR 51.308(d)(3). We also proposed to approve Arkansas' consultation with FLMs and Missouri and our determination that the SIP submittal satisfies the consultation requirements under 40 CFR 51.308(i)(2) and 40 CFR 51.308(d)(3)(i). We also agreed with DEQ's determination that the revised 2018 RPGs in the Phase II action do not need to be further revised. We proposed to approve Arkansas' request to withdrawal from the approved SIP the previously approved PM₁₀ BART limit for Power Boiler No. 1. and the regional haze FIP provisions for the Domtar Ashdown Mill, and we are finalizing the withdrawal of those provisions in a separate rulemaking published elsewhere in this issue of the Federal Register.

The EPA also proposed to approve in its NPRM Arkansas' interstate visibility transport provisions from the August 8, 2018, regional haze Phase II SIP submittal as supplemented by the visibility transport provisions in the October 4, 2019, interstate transport SIP submittal, which cover the following six NAAQS: The 2006 24-hour PM_2.5 NAAQS; the 2012 annual PM_2.5 NAAQS; the 2008 and 2015 eight-hour O₃ NAAQS; the 2010 one-hour NO₂ NAAQS; and the 2010 one-hour SO₂ NAAQS. We are finalizing our approval of the prong 4 portions of these SIP submittals addressing CAA section 110(a)(2)(D)(i)(II) for these NAAQS on the basis that with our approval of the Arkansas Regional Haze Phase III SIP revision in this notice, Arkansas has a fully-approved regional haze SIP. The Arkansas Regional Haze NO_X SIP revision, the Arkansas Regional Haze SO₂ and PM SIP revision, and the Arkansas Regional Haze Phase III SIP revision together fully address the deficiencies of the 2008 Arkansas Regional Haze SIP that were identified in the March 12, 2012, partial approval/partial disapproval action. As an alternative basis for approval of the State's CAA section 110(a)(2)(D)(i)(II) prong 4 submittals for these NAAQS, we are finalizing our determination that Arkansas has provided an adequate demonstration in the October 4, 2019 submittal that emissions within its jurisdiction do not interfere with other air agencies' plans to protect visibility.

The public comment period for the NPRM closed on April 15, 2020. We received two sets of public comments concerning our proposed action. The comments are included in the publicly posted docket associated with this action at https://www.regulations.gov. We received a comment letter with adverse comments dated April 15, 2020, submitted on behalf of the National Parks Conservation Association, the Sierra Club, and Earthjustice regarding our proposed approval. We also received another comment letter dated April 15, 2020, from Domtar that was largely in support of our proposed approval. Below we provide a summary of the comments with our detailed responses. The complete comments can be found in the docket associated with this final rulemaking. After careful consideration of the public comments received, we have decided to finalize our action with no changes from the proposed action. For our complete, comprehensive evaluation of the Arkansas Regional Haze Phase III SIP revision, please refer to the proposed approval (See 85 FR 14847). Our final actions regarding the NPRM are summarized in section IV of this notice.

III. Public Comments and EPA Responses

A. Demonstration That the BART Alternative Is Better-Than-BART

Comment A.1: The BART alternative measure submitted by the State fails to demonstrate that the BART alternative achieves greater reasonable progress than BART. Rather than submit a revised BART analysis determination, DEQ's Arkansas Regional Haze Phase III SIP includes what it asserts are approvable SIP measures in a BART alternative for two subject-to-BART sources (Power Boilers No. 1 and 2) at the Domtar Ashdown paper mill located in Ashdown, Arkansas. Compared to BART, the BART alternative results in an overall (Power Boilers No. 1 and 2) increase in sulfur dioxide (SO₂) emissions and decrease in NO_X emissions. While DEQ claims that the NO_X decrease mitigates the SO₂ increase, the SIP fails to demonstrate the BART alternative achieves greater reasonable progress than BART.

Response: We disagree with the commenter's assertion that the BART alternative measure submitted by the State fails to demonstrate that the BART alternative achieves greater reasonable progress than BART.

As explained in the proposed action, the BART alternative would result in an overall decrease in SO₂, NO_X, and particulate matter (PM₁₀) emissions from the baseline for both power boilers at Domtar Ashdown paper mill. The BART alternative results in greater emission reductions of NO_X and PM₁₀ than the BART controls in the FIP. The BART alternative controls would reduce NO_X and PM₁₀ emissions by 1,096 and 111 tons per year (tpy), respectively, from the baseline. The BART alternative results in a smaller reduction in SO₂ emissions compared to the BART controls (BART achieves 3,051 tpy SO₂ reduction) but still achieves a decrease of 1,637 tpy SO₂ from the baseline. Despite a smaller reduction in SO₂ emissions than BART (a 1,414 tpy SO₂ difference), the BART alternative results in 300 tpy fewer NO_X emissions and 157 tpy fewer PM₁₀ emissions compared to BART. Model results show that the additional reduction in NO_X emissions under the BART alternative controls results in more overall modeled visibility improvement across the impacted Class I areas than BART even with the smaller reduction in SO₂ emissions.

We explained in our proposed action that greater visibility improvement occurs because Domtar's baseline NO_X emissions are the primary driver of visibility impacts from the source and contribute more to visibility impairment across the four-affected Class I areas in Arkansas and Missouri for Power Boiler No. 1, and also contribute more at Caney Creek for Power Boiler No. 2 than other pollutants emitted by the source. DEQ first included an analysis utilizing method 1 that shows that the BART alternative controls achieve greater overall cumulative reductions in visibility impairment (as expressed by the change in deciviews or Δdv) from the baseline across the four Class I areas when compared to BART (0.549 Δdv for the alternative versus 0.473 Δdv for BART). DEQ then determined that the BART alternative controls reduce the overall visibility impairment from the baseline by 0.520 Δdv under its method 2 evaluation and is greater than the overall visibility improvement modeled under BART, which is 0.516 Δdv. The DEQ noted that the most impacted Class I area, Caney Creek (1.137 dv baseline impairment), improved the greatest (0.384 Δdv) with the BART alternative under method 2, and would experience greater visibility improvement under the BART alternative scenario than under the BART scenario, which improves by 0.361 Δdv.

The State's weight of evidence analysis of visibility improvement in the SIP was supported by our analysis of various metrics, which reinforced that the BART alternative achieves greater reasonable progress. We analyzed the pollutant species contribution to visibility impacts at the Class I areas from each power boiler. Specifically, for Power Boiler No. 1, baseline modeled nitrate (NO₃−) and nitrogen dioxide (NO₂) impacts had the highest contribution to visibility impairment at all Class I areas. For Power Boiler No. 2, baseline modeled NO₃− and NO₂ impacts are the primary driver for visibility impacts at Caney Creek, which is the Class I area impacted the most by the Domtar units. For Power Boiler No. 2, the visibility impacts resulting from NO_X at Caney Creek outweigh SO₄ ^{2 −} species contributions (from SO₂ precursors) to impacts at the other three Class I areas combined. In addition to pollutant species contributions to impacts, we also considered the ten highest impacted days. This analysis provided a broader look at those days with the highest impacts at each Class I area. The results were consistent with the State's analysis based on the 98th percentile day, which was selected as representative of the highest impact (the 8th highest day). The average results across the top ten highest impacted days also supported our position that it is appropriate to give greater weight to Caney Creek impacts (0.9819 dv baseline impairment) in our consideration of whether the BART alternative achieves greater reasonable progress than BART since they are much larger than impacts at the other Class I areas. The BART alternative resulted in more visibility improvement at Caney Creek and slightly less at the other Class I areas when compared to the BART limits, but the visibility improvement at Caney Creek outweighed the difference in visibility benefit at the other three Class I areas altogether. On average, the BART alternative controls achieved greater overall visibility improvement from the baseline compared to BART for the ten highest impacted days (0.439 Δdv for the alternative versus 0.423 Δdv for BART). Our analysis of the ten highest impacted days similarly supported the conclusion that the BART alternative provides for greater reasonable progress than BART. Finally, we complemented the State's analysis by evaluating the modeled number of days impacted by Domtar over 1.0 dv and 0.5 dv for each scenario at each Class I area. This compared the frequency and duration of higher visibility impacts between the two control scenarios. The BART FIP limits and the BART alternative both reduce the total modeled number of days with visibility impacts over 1.0 dv from fifteen days in the baseline to four days for each scenario. For the metric of days with modeled visibility impacts over 0.5 dv, the FIP limits and the BART alternative showed nearly identical reduction in the number of days, but very slightly favored the FIP limits over the BART alternative (from 82 to 36 days for the FIP limits compared to 37 days for the BART alternative). This single metric, however, on which BART performed slightly better than the BART alternative (days impacted over 0.5 dv) is not sufficient to outweigh the substantial evidence presented using the other metrics as to the relatively greater benefits of the BART alternative over BART. These different metrics reinforce the State's analysis in the SIP that greater reasonable progress was achieved by the BART alternative.

The State's weight of evidence analysis of emission reductions and visibility improvement (using the 98th percentile metric) as complemented by our analysis of different metrics, justify our approval of the State's determination that the BART alternative achieves greater reasonable progress than BART under 40 CFR 51.308(e)(2)(i)(E). The State followed the prescribed process for determining the level of control required for the BART alternative for the Domtar Ashdown Mill and adequately supported its determination with analysis that meets the requirements under section 40 CFR 51.308(e)(2).

Comment A.2: EPA proposes approving the Arkansas Regional Haze Phase III SIP and relaxing the BART emission limitations established in its 2016 FIP. The proposed facility-wide emission limitation would allow for fewer emission reductions from the Domtar Ashdown Mill. EPA's proposal reverses course on its FIP, failing to make reasonable progress on reducing visibility impairment in Class I areas in accordance with the CAA mandates and requirements.

Response: The BART alternative establishes pollutant-specific limits at each of the two BART sources at the Ashdown Mill. There is no “facility-wide emission limitation” as stated by the commenter. In addition, we disagree with the commenter that the EPA is reversing course on its FIP by relaxing BART limitations established in the FIP, and thus failing to make reasonable progress and reduce visibility impairment in Class I areas in accordance with the CAA and its mandates.

The BART alternative results in larger reductions in NO_X and PM emissions than required by the FIP, while SO₂ emissions are not reduced to the same extent as would be required under the FIP. As explained in our response to comment A.1 of this final action and also in section IV of our proposed action, our analysis of the State's weight of evidence conclusion as complemented by EPA's analysis, demonstrate that the State has met the BART and reasonable progress requirements for regional haze under the applicable provisions of the CAA and the Regional Haze Rule. Thus, the proposed withdrawal of the BART provisions in the FIP and replacement with the BART alternative requirements in the SIP will not result in a failure to meet the applicable requirements.

The Arkansas Regional Haze Phase III SIP revision and concurrent withdrawal of the corresponding parts of the FIP pertaining to Domtar will also not reverse course from the prior FIP with respect to the separate reasonable progress requirements for Arkansas. As mentioned in section IV of our proposed action, we determined in our September 27, 2019 Arkansas Regional Haze SO₂ and PM SIP revision that Arkansas had fully addressed the reasonable progress requirements under 40 CFR 51.308(d)(1) for the first implementation period in that final action. In that action, we also noted that the 2016 FIP BART determination requirements for Domtar were still in place but we agreed with the State that as long as those requirements continued to be addressed by the measures in the FIP, nothing further is needed to satisfy the reasonable progress requirements for the first implementation period. In the Arkansas Regional Haze Phase III SIP submittal, the State assessed whether changes would be needed with respect to the reasonable progress analysis, based on any differences between the SIP and FIP-based measures for Domtar. The BART alternative analysis performed for the Domtar power boilers was based, in part, on an assessment of the same factors that must be addressed in the reasonable progress analysis. The FIP BART determination analysis was compared to the proposed BART alternative controls in the Arkansas Regional Haze Phase III SIP submittal. The BART alternative measures for Domtar resulted in greater overall visibility improvement than the BART requirements in the FIP and the previously approved BART PM₁₀ limit for Power Boiler No. 1. As a result, nothing further is needed to satisfy the reasonable progress requirements for the first implementation period. For these reasons, approval of the Arkansas Regional Haze Phase III SIP revision and concurrent withdrawal of the corresponding parts of the FIP do not interfere or reverse course from the FIP with respect to the CAA requirements pertaining to BART or reasonable progress under 40 CFR 51.308(d) or (e).

Comment A.3: EPA's proposal cobbles together two pieces of information (a comparison of emission reductions and a modeling analysis) and fails to demonstrate that the BART alternative is clearly better than BART. The Regional Haze Rule provides different regulatory tests for a state to use to demonstrate that a BART alternative is better than BART. Arkansas claims that it used the “clear weight of evidence test,” but the information it provides falls under 40 CFR 51.308(e): An emission reduction comparison and modeling. The information Arkansas provides fails to meet the requirements in 40 CFR 51.308(e). Therefore, it is unreasonable for EPA to provide weight to the information.

Response: We disagree with the commenter's assertion that the information on which our approval of the State's SIP is based fails to provide an adequate clear weight of evidence analysis to meet the requirements in 40 CFR 51.308(e). The commenter is apparently alleging that the analysis provided by the State instead falls under 40 CFR 51.308(e)(3) rather than under 40 CFR 51.308(e)(2)(i)(E) because it is based on an emission reduction comparison and modeling. The argument that the kind of data and analysis to be used under the clear weight of evidence test must somehow be sufficiently different from what would be required under 40 CFR 51.308(e)(3) is not a reasonable interpretation of these regulations. EPA interprets 40 CFR 51.308(e)(2)(i)(E) as permitting data and analysis that may be relevant under 40 CFR 51.308(e)(3) analysis to be used in supporting a clear weight of evidence demonstration.

Pursuant to 40 CFR 51.308(e)(2)(i)(E), the state must provide a determination under 40 CFR 51.308(e)(3) or otherwise based on “clear weight of evidence” that the alternative measure achieves greater reasonable progress than BART. The State relied on a modeling analysis to determine if the BART alternative could be shown to make greater reasonable progress than BART, but that modeling was different than the modeling described under 40 CFR 51.308(e)(3). The State used an air quality modeling methodology approach using the maximum 98th percentile visibility impact of three modeled years using the CALPUFF model instead of modeled overall visibility conditions for the twenty percent best and worst days, as would be required under 40 CFR 51.308(e)(3). The State's approach could be considered a modified version of the two-part modeling test under 40 CFR 51.308(e)(3) and is more appropriate to classify under the weight of evidence analysis approach instead allowed under 40 CFR 51.308(e)(2)(i)(E).

The State's methodology and analysis under the clear weight of evidence test is reasonable. The State's CALPUFF modeling approach utilizing the 98th percentile visibility impacts is consistent with the approach recommended by the BART guidelines for comparing different control options at a single source when developing BART determinations relying on the 98th percentile visibility impact as the key metric. It is also consistent with the methodology followed in EPA's 2016 FIP BART determination for Domtar.

CALPUFF is a single source air quality model that is recommended in the BART Guidelines. Since CALPUFF was used for this BART alternative analysis, the modeling results were post-processed in a manner consistent with the BART guidelines. This approach is, therefore, acceptable and reasonable for the comparison of the proposed BART alternative to the FIP BART determination for Domtar since it is the same modeling used to determine BART in the FIP, and the BART alternative is focused on only the BART sources at Domtar. The State also considered two methods of modeling evaluation provided by Domtar for this approach of using the maximum 98th percentile visibility impact. Method 1 assessed visibility impairment on a per source per pollutant basis and method 2 allowed for interaction of the pollutants from both boilers. The State followed the same general CALPUFF modeling protocol and used the same meteorological data inputs for the BART alternative assessment as discussed in Appendix B to the FIP TSD. Only the modeled emission rates changed to represent the modeled scenarios for each method.

DEQ determined that the visibility benefits as measured under method 2 and the previous FIP BART determination formed an appropriate BART benchmark for the purposes of the evaluation of Domtar's BART alternative. We continue to agree with DEQ that because method 2 provides for the full chemical interaction of emissions from both power boilers, method 2 analysis results are a reliable assessment of the anticipated overall visibility improvement of controls utilizing the 98th percentile impact. Under the weight of evidence approach, we made use of all available information and data which could inform our decision while recognizing the relative strengths and weaknesses of that information in arriving at the soundest decision possible. This array of information and other relevant data was of sufficient quality to inform our comparison of visibility impacts between BART and the BART alternative. We carefully considered this evidence in evaluating the Arkansas Phase III SIP revision submitted by the State. Overall, the difference in visibility impacts between the BART and the BART alternative scenarios was large enough to show that the BART alternative achieves greater reasonable progress than BART based on the clear weight of the evidence.

As explained in response to comment A.1 in section III.A of this final action, we evaluated DEQ's analysis and additional model results (relying primarily on the analysis of the 98th percentile impacts), the analysis of emission reductions, and the analysis of Domtar's visibility impacts due to NO₃− compared to SO₄−. In addition, we also considered our analysis of the ten highest impacted days (8th to 17th highest) and our analysis of the number of days impacted over 0.5 dv and 1.0 dv. All of these metrics, except the number of days impacted over 0.5 dv (which only very slightly favored BART), provided substantial evidence and collectively supported the conclusion that the BART alternative provides for greater reasonable progress than BART. For these reasons, we are finalizing our approval of the State's weight of evidence analysis approach and the conclusions reached by the State. In the course of evaluating the SIP submittal, EPA developed some additional analysis that complements and supports the State's analysis. Taken as a whole, the record supports approval of the State's determination that the BART alternative achieves greater reasonable progress than BART under the clear weight of evidence pursuant to 40 CFR 51.308(e)(2)(i)(E).

Comment A.4: EPA fails to provide a basis to rely on a comparison of emissions. EPA merely presents the emission reductions under BART and the alternative, but fails to explain the strengths and weaknesses of this information and does not assign any weight to the emission comparison. A comparison of multiple pollutant species emission levels alone is not informative without visibility modeling. The pollutants' differing visibility impacts and complex interactions between them and in the atmosphere make it extremely difficult to discern their collective impacts without visibility modeling. EPA has consistently relied on modeling to assess the visibility impacts under these circumstances.

Response: We disagree with the commenter's assertion that EPA “merely presents the emission reductions under BART and the alternative.” In our proposed action, our basis for presenting the emission reduction information laid the foundation for describing the differences in visibility outcomes achieved between the FIP and the BART alternative, leading EPA to agree with the State that there was a need to support the BART alternative with visibility modeling. The State first showed reduced emissions from the baseline and then used the modeling to support a conclusion that the emission reduction differences between the FIP BART benchmark and BART alternative were acceptable because NO_X precursor emissions are the main driver contributing to the visibility impacts from this source. Thus, the State proceeded to conduct precisely the modeling analysis the commenter seems to assert is required, using CALPUFF. Indeed, recognizing the potential interaction between multiple species of visibility pollutants, the State used Method 2 in evaluating the visibility consequences of the BART alternative compared to the BART benchmark. EPA has relied on the modeling submitted by the State in reaching a conclusion that the SIP submittal is approvable. While EPA does not concede that modeling is required in all cases to conduct an approvable “clear weight of evidence” analysis under 51.308(e)(2)(i)(E), modeling was in fact done in this instance to support the analysis. This comment is thus premised on a misunderstanding of the record.

To the extent the commenter is asserting that the emissions comparisons alone cannot be used as even one part of a weight of evidence demonstration, the commenter is mistaken in how a “weight of evidence” analysis is conducted. The term “weight” connotes that multiple pieces of evidence are brought together and analyzed as a whole. Comparative emissions data is obviously a critical piece of that evidentiary record, and provides a foundation on which further analysis, such as modeling, may be conducted. To assert that EPA must ignore emissions comparisons—or any single piece of evidence—because it does not provide, on its own, a sufficient basis to make a “weight of evidence” determination is both illogical and a misreading of EPA's regulations. We also note that the regulations require an analysis of emission reductions under BART and the alternative, see 40 CFR 51.308(e)(2)(i)(C) and (D).

Comment A.5: EPA should not provide weight to modeling data of insufficient quality, which fails to meet the requirements of the regulations. It is disingenuous for EPA to suggest that the CALPUFF model is a “modified” version of the two-part modeling test. EPA has consistently interpreted the two-part dispersion modeling test under 40 CFR 51.308(e)(3) to mean the Comprehensive Air Quality Model with Extensions (CAMx) model, and not CALPUFF. EPA and states have consistently used CAMx to assess whether a BART alternative would result in “greater reasonable progress” under the two-prong test. CAMx and CALPUFF are vastly different models and 40 CFR 51.308(e)(3) requires a specific type of dispersion modeling. EPA's suggestion that use of CALPUFF is acceptable because it “is consistent with the approach recommended by the BART guidelines for comparing different control options at a single source when developing BART determinations relying on the 98th percentile visibility impact as the key metric” also fails. A comparison of control options at a single source compares changes in the emission reductions in one pollutant, but does not compare the complexities involved in analyzing interactions between multiple pollutants. It is also irrelevant that only the BART sources at Domtar are under consideration. While the FIP considered each pollutant separately, the alternative attempts to analyze and take credit for combined emission reductions from three pollutants as it fails to actually assess the effect of the alternative on visibility as compared to BART.

Response: We disagree with the comment that CAMx must be used for the two-part test under 40 CFR 51.308(e)(3) or that CALPUFF cannot be used to support the determination here, which is not under 40 CFR 51.308(e)(3) in any case. The first point is irrelevant because the State is not proceeding under 40 CFR 51.308(e)(3); however, it is worth noting that the regulatory text does not require the use of CAMx. CALPUFF is also an air dispersion model, and one that the Agency has recognized as available for use for BART alternatives under 40 CFR 51.308(e)(3).

Regarding the use of CALPUFF, we did not suggest that CALPUFF was replacing CAMx under 40 CFR 51.308(e)(3). We logically examined the two-part analysis under 40 CFR 51.308(e)(3) in the proposed action to show how the State arrived at classifying the approach as a weight of evidence approach. Our choice of using the term “modified” to describe the relationship of this analysis to the two-part test under 40 CFR 51.308(e)(3) was intended to describe how the State's approach was similar to 40 CFR51.308(e)(3) in considering distribution of emissions and visibility improvements using modeling, but different from 40 CFR 51.308(e)(3) because the analysis based on the CALPUFF modeling focused on the 98th percentile visibility impacts instead of the twenty percent best and worst days required by 40 CFR 51.308(e)(3). Therefore, the State's weight of evidence analysis is acceptable under 40 CFR 51.308(e)(2)(i)(E) and should not be judged according to 40 CFR 51.308(e)(3). The commenter's objection to 40 CFR 51.308(e)(3) not being met is immaterial since the weight of evidence approach followed in the SIP submittal does not fall under 40 CFR 51.308(e)(3) but under 40 CFR 51.308(e)(2)(i)(E).

The commenter states that EPA is wrong to consider CALPUFF as acceptable just because it “is consistent with the approach recommended by the BART guidelines for comparing different control options at a single source when developing BART determinations relying on the 98th percentile visibility impact as the key metric.” The commenter points out that a comparison of control options at a single source compares changes in the emission reductions in one pollutant, but does not compare the complexities involved in analyzing interactions between multiple pollutants. We disagree with this point in relation to the alternative analysis here. First, particularly for purposes of a BART alternative analysis for a single facility (with two BART units), EPA's regulations recognize CALPUFF to be an acceptable model, (explaining that CALPUFF is particularly suited for BART and BART alternative applications at a single source). Further, Method 2, incorporated by the State in its SIP submittal, is a full assessment method where all sources and pollutants are combined into a single CALPUFF modeling run per year for the baseline and each control scenario. Method 2 allows for interaction of the pollutants from both boilers, as emitted pollutants from each unit disperse and compete for the same reactants in the atmosphere, providing modeled overall impacts due to emissions from both units. It is because of this that method 2 analysis results are a more reliable assessment of the anticipated overall visibility improvement of controls under each scenario. Thus, this is an entirely suitable application of the CALPUFF model, and the commenter is incorrect to state that the CALPUFF modeling did not account for the interactive chemistry of visibility pollutants.

EPA recognizes that the CALPUFF model includes simplified chemistry to account for interactions between pollutants. The simplified chemistry tends to magnify the actual visibility effects of a single source; thus, it is appropriate to use the 98th percentile to avoid overprediction and not give undue weight to the extreme tail of the distribution. This approach will effectively capture the sources that contribute to visibility impairment in a Class I area, while minimizing the likelihood that the highest modeled visibility impacts might be caused by unusual meteorology or conservative assumptions in the model.

The EPA has previously recognized this approach of using CALPUFF as an acceptable approach in the past when analyzing BART alternatives that only include emission reductions at a single or small group of BART sources. Specifically, we approved this approach for the State of Arizona which established a BART alternative for Steam Units 2 and 3 at Arizona Electric Power Cooperative's Apache Generating Station. See also 70 FR 60616 (recognizing CALPUFF as particularly appropriate for single-source applications).

The commenter states that the FIP considered each pollutant separately, whereas the alternative attempts to analyze and take credit for combined emission reductions from three pollutants, which allegedly fails to assess the effect of the alternative on visibility as compared to BART. The commenter is incorrect in their premise. The CALPUFF modeling in the FIP evaluated each unit separately, but modeled the visibility impacts from all pollutants from that unit. For example, in evaluating the visibility benefit from NO_X controls on Power Boiler No. 1, the NO_X emissions varied between each control scenario modeled, while the SO₂ and PM emissions were included but held constant in these NO_X control scenarios. In evaluating the BART alternative, the State provided EPA with two separate methods of using the CALPUFF modeling to evaluate visibility impacts of the BART alternative as compared to BART, including Method 2 (described above) that modeled all pollutants from both BART units to assess the total visibility impact from these two units.

For these reasons, we disagree that the modeling data was of insufficient quality and failed to meet the requirements of the regulations.

Comment A.6: EPA lacks authority to give one Class I area more weight than others. EPA suggests that it is reasonable to give one of the Class I areas “greater weight” when considering visibility benefits and cherry-picks the Class I area with the greatest visibility improvement, which is closest to Domtar. Focusing on that Class I area serves to support a source's preferred control outcome. Showing that one Class I area will have greater visibility benefits does nothing to tip the weight of evidence scale in favor of the BART alternative. It merely shows one area will see more benefits. In addition, EPA fails to provide a basis for applying the 0.5 deciview threshold used by the State to determine if a source contributes to visibility impairment at a Class I area with the BART alternative analysis.

Response: We disagree with the commenter's assertion that EPA “cherry picks” the Class I areas with the greatest visibility improvement. We considered many metrics in analyzing the weight of evidence approach by the State, including the overall visibility improvement on average across the four impacted Class I areas. As a whole, these factors supported a conclusion that the BART alternative achieves greater reasonable progress than BART at the subject facility. One metric that we analyzed was the breakdown of pollutant speciation impacts across each Class I area due to modeled emissions from each power boiler. We highlighted impacts at Caney Creek specifically in this analysis because Domtar's Ashdown facility impacts this Class I area the greatest, and this is due to NO_X emissions from Power Boiler No. 2. We also found that NO_X emissions contributed more to visibility impairment across all four Class I areas for Power Boiler No. 1. The greater impact due to NO_X emissions is relevant because it demonstrates that the higher SO₂ emissions allowed under the BART alternative is offset by the larger reduction in NO_X emissions. This is just one factor among many that we considered in analyzing the State's weight of evidence approach as explained in the proposed approval and in preceding responses in this final approval. We took into account the visibility impacts at all impacted Class I areas (individually and on average) and did not solely focus on the benefits at the most impacted area.

We disagree with the assertion that we are supporting the source's preferred control outcome instead of addressing emissions cumulatively across all Class I areas. The commenter points out that the court in Nat'l Parks Conservation Ass'n v. EPA held that EPA's analysis in reviewing SIP submittals must take into account the visibility impacts at all impacted Class I areas rather than focusing solely on the benefits at the most impacted areas, 803 F.3d 151, 165 (3d Cir. 2015). However, the facts of Nat'l Parks Conservation Ass'n v. EPA, are not analogous to the facts surrounding our proposed approval. In Nat'l Parks Conservation Ass'n v. EPA, the court was reviewing EPA's approval of the state's assessment of the visibility-improvement factor within the five-factor BART analysis. The state calculated visibility improvement that could be achieved at Class I areas by implementing additional controls at BART-eligible sources. The state's calculations for each source, however, took into account only the potential impact such controls would have on the visibility in the Class I area most severely impacted by the source. The state did not consider “cumulative visibility impact,” which the EPA itself had conceded was improper under the visibility BART factor. The court in NPCA rejected that this flaw in the State's analysis could be dismissed as harmless error.

In this action, by contrast, both the State and EPA have evaluated the cumulative visibility impacts across all of the affected Class I areas. The State considered this with both of its methods of analysis, and EPA coupled those results with our own analysis of cumulative visibility improvement. DEQ first included an analysis utilizing method 1 that shows that the BART alternative controls achieve greater overall cumulative reductions in visibility impairment from the baseline cumulatively across the four Class I areas when compared to BART (0.549 Δdv for the alternative versus 0.473 Δdv for BART). DEQ also determined using method 2 that the BART alternative controls reduce the overall cumulative visibility impairment from the baseline by 0.520 Δdv, which is greater than the overall visibility improvement modeled under BART, which is 0.516 Δdv. We complemented the State's analysis by comparing the average visibility impact across the top ten highest impacted days at each Class I area (average 8th to 17th highest). This analysis provided a broader look at those days with the highest impacts at each Class I area. The results were consistent with the State's analysis based on the 98th percentile day, which was selected as representative of the highest impact (i.e., the 8th highest day). The BART alternative controls achieve greater overall visibility improvement from the baseline compared to BART for the ten highest impacted days (0.439 Δdv for the alternative versus 0.423 Δdv for BART). Thus, visibility benefits at each Class I area were considered and analyzed by multiple metrics that confirmed our proposed approval of the alternative.

The commenter argues that EPA “fails to provide a basis for applying the 0.5 deciview threshold used by the State to determine if a source contributes to visibility impairment at a Class I area with the BART alternative analysis,” noting that numerous BART determinations relied on lower deciview thresholds that resulted in significant emission reducing outcomes. The meaning of this comment is not clear. EPA did not apply a 0.5 deciview threshold to cut off its evaluation of other Class I areas. However, it is reasonable to provide additional analysis when one Class I area is much more heavily impacted by a source than others. In the case of Domtar, the baseline visibility impacts at Caney Creek are much larger than impacts at the other Class I areas, so it is reasonable to give greater weight to visibility benefits at Caney Creek resulting from the alternative as compared to BART. The level of visibility benefit from controls at the other three Class I areas are smaller than those at Caney Creek, and the baseline visibility impacts of the source at these areas was well below the 0.5 dv threshold used by the State to determine if a source contributes to visibility impairment at a Class I area. In making this observation, we do not categorically dismiss or ignore impacts to other Class I areas below 0.5 or any other threshold. We simply note that the changes in visibility at these other Class I areas were individually very small and collectively smaller than the comparative gain in visibility achieved by the BART alternative at Caney Creek.

The commenter mentioned that Congress provided no authority for EPA to treat one Class I area differently from others. As mentioned previously, we treated all Class I areas the same and measured the cumulative visibility impacts across all of them using multiple metrics. We specifically analyzed the effects at Caney Creek, since it is the Class I area impacted the most. But that analysis does not show favoritism and merely provides one metric for interpreting how impacts are correlated to overall emissions from the source at each Class I area.

B. Monitoring, Recordkeeping and Reporting Requirements

Comment B.1: EPA lacks authority to approve the State's SIP submission with respect to provisions pertaining to alternative test methods. EPA proposes to allow the State to authorize alternative sampling or monitoring methods (equivalent to methods in the permit) that EPA would concur on, outside the SIP process. Specifically, EPA proposes approving permit conditions 35 and 42 as a part of the SIP. Neither the State's SIP nor EPA's proposal explains what criteria and process EPA would use to approve an alternative method. Arkansas' alteration or elimination of SIP requirements can have no effect for purposes of federal law unless and until EPA ratifies that action with a SIP revision that is subject to the SIP requirements, including provisions for public notice and comment. Moreover, the monitoring, recordkeeping and reporting provisions in the State's SIP are not approvable and therefore, those methods cannot be used a basis for assessing whether an alternative method is approvable. Based on Arkansas' SIP provisions, there is no way for the public to assess whether an alternative method will comply with the Act. Therefore, EPA should not approve these provisions because they are inconsistent with the requirements of CAA section 110(i), 110(l) and 110(k)(3).

Response: We recognize that the commenter raises a concern that the State's ability to authorize (with EPA concurrence) alternative test methods in conditions 35 and 42 may be inconsistent with the Act insofar as “[n]either the State's SIP nor EPA's proposed approval explains what criteria and process EPA would use to approve an alternative method.” In general, EPA agrees that SIP provisions cannot authorize a State to make changes in the EPA-approved and federally enforceable SIP requirements applicable to sources without going through the statutorily required SIP-revision process. EPA refers to SIP provisions that purport to authorize States to make unilateral changes to existing SIP requirements as impermissible “director's discretion” provisions. However, EPA interprets the CAA to allow two types of such provisions: (i) Where the provision provides director's discretion for the State to make changes, but specifies that such changes have no effect for purposes of federal law or alter SIP requirements unless and until the EPA approves the changes through a SIP revision pursuant to CAA requirements; or (ii) where the provision provides director's discretion that is adequately bounded, such that at the time EPA approves the SIP provision the agency can evaluate it for compliance with applicable CAA requirements and evaluate the potential impacts of the State's exercise of that discretion. EPA interprets CAA section 110(i) to allow SIP provisions with director's discretion of either type. In the case of an adequately bounded provision, EPA considers such provisions consistent with section 110(i) because, at the time of initial approval into the SIP, the agency will already have evaluated the provision for compliance with applicable requirements and evaluated the potential impacts from exercise of the discretion. By their terms, conditions 35 and 42 do not specify that DEQ must seek a SIP revision to change the required monitoring at the source. Thus, to be approvable, EPA would have to determine that the State's discretion in these provisions is adequately bounded and assess the potential impacts from the exercise of that authority.

In response to the commenter's concerns, EPA has further evaluated conditions 35 and 42 to determine whether they provide adequate bounding, allowing EPA to assess the provisions for compliance with applicable requirements and the potential impacts that could result from DEQ's potential exercise of the discretion to authorize alternative monitoring. In support of EPA's proposed approval of plantwide conditions 35 and 42 into the Arkansas SIP, DEQ provided additional information in a letter (dated December 3, 2020) to EPA to clarify the process and standards that the State shall follow and apply to approve the use of any alternative method under plantwide conditions 35 and 42 of the Domtar permit. DEQ notes in the letter that DEQ has received a disconnection notice for Power Boiler No. 1 and that it is now permanently retired. In accordance with plantwide condition 34, Power Boiler No. 1 is in compliance with the BART alternative limits by virtue of being permanently retired and, therefore, not emitting any of the relevant visibility pollutants. The numerical emission limits will still apply, even though the unit has been taken out of service. As a result, the process to be used by DEQ in its approval of any request for an alternative sampling or monitoring method is only applicable to Power Boiler No. 2 under plantwide condition 42.

For Power Boiler No. 2, which currently relies on a continuous emissions monitoring system (CEMS) to monitor SO₂ and NO_X emissions, DEQ explained in its letter that it will use the criteria for alternate monitoring systems contained in 40 CFR part 75, subpart E in its evaluation of the approvability of any request for an alternative sampling or monitoring method for SO₂ and NO_X emissions. More specifically, the State explained that any request for approval of an alternative sampling or monitoring method under plantwide condition 42 shall meet the general demonstration requirements for alternative monitoring systems under 40 CFR 75.40 and require Domtar (or the current owner of the Ashdown Mill) to demonstrate adequately that the average hourly emission data for SO₂, NO_X, and/or volumetric flow in the proposed alternative sampling or monitoring has the same or better precision, reliability, accessibility, and timeliness as that provided by the currently applicable continuous emission monitoring system (see criteria in 40 CFR 75.41-75.46). Furthermore, DEQ will require all information in 40 CFR 75.48 of Domtar (or the current owner of Ashdown Mill) in the application for certification or recertification of the alternative monitoring system. DEQ notes that the requirements of 40 CFR part 75, subpart E shall be met by the alternative monitoring system when compared to a contemporaneously operating, fully certified continuous emission monitoring system or a contemporaneously operating reference method, where the appropriate reference methods are listed in 40 CFR 75.22.

With respect to any request for alternative sampling or monitoring methods for PM₁₀ under plantwide condition 42, we note that Power Boiler No. 2 is subject to 40 CFR part 63, subpart DDDDD and reference is made to those requirements for PM₁₀ compliance demonstrations in plantwide condition 41. Condition 41 clearly explains that the applicable PM₁₀ compliance demonstration requirements from 40 CFR part 63 subpart DDDDD shall be utilized by Domtar (or the current owner of Ashdown Mill). These requirements, which are at 40 CFR 63.7505—63.7541, do not cease and are ongoing. In response to comment B.8 in section III of this final action, we address the alternative option provided in the permit for monitoring emissions from Power Boiler #2 when that unit is combusting natural gas.

DEQ explained in its letter that it expects that Domtar will work with both DEQ and EPA in the development of equivalent testing protocols before seeking approval from DEQ (with EPA concurrence) and before performing the equivalency testing. The alternate sampling or monitoring protocol submittal to DEQ must contain EPA's official letter of documented recommendations and concurrence, as required for DEQ approval. Although not the same as EPA approval of an alternative sampling or monitoring requirement through a SIP revision, in the case of a valid director's discretion provision that is already adequately bounded, EPA considers the inclusion of consultation with EPA an extra measure of assurance that any such alternative will be appropriate. Given the process that DEQ will follow and standards that DEQ will apply in evaluating any potential alternative (and EPA's consultation in the process) EPA anticipates that DEQ's exercise of its well bounded discretion to authorize alternative sampling or monitoring will not result in adverse impacts, e.g., adverse impacts on regional haze requirements that are relevant to this SIP submission.

Based on the information contained in DEQ's December 3, 2020, letter which forms a critical part of the record basis for EPA's approval of this submittal, EPA has determined that conditions 35 and 42 as supplemented by the letter are adequately bounded director's discretion provisions. In particular, EPA agrees with DEQ that the criteria in 40 CFR part 75, subpart E for SO₂ and NO_X emissions and in 40 CFR part 63, subpart DDDDD for PM₁₀ emissions are appropriate to evaluate the approvability of any alternative sampling or monitoring methods and establish the proper bounds for DEQ's exercise of discretion and EPA approval for any future requests from the source to use alternative sampling and monitoring methods. Further, in determining whether it is appropriate for EPA to provide its concurrence to any future request for a change in sampling and monitoring methods under these conditions, EPA reserves the right to withhold its concurrence if EPA determines that the request falls outside the process and bounds specified in DEQ's letter. In such circumstances, the CAA would require that the State seek to make the change through the normal SIP revision process.

For these reasons, these permit provisions are consistent with the requirements of CAA sections 110(i), 110(l) and 110(k)(3).

Comment B.2: The Arkansas Regional Haze SIP for Domtar does not satisfy the requirement to provide for periodic testing of stationary sources and to use enforceable test methods for each emission limit specified in the plan, and should therefore be disapproved. For example, the SIP lacks specificity regarding test methods in permit conditions 38 and 40. Permit condition 38 refers to 40 CFR part 60, without identifying the specific rule provisions that apply. Similarly, permit condition 40 fails to identify the specific AP-42 emission factor.

Response: We disagree with the commenter that the SIP lacks specificity regarding test methods in permit conditions 38 and 40 for the boilers. The commenter states that permit condition 38 refers to 40 CFR part 60 regarding utilizing CEMS without identifying the specific rule provisions that apply. In permit condition 38, the State provided that “the permittee shall demonstrate compliance with the 30-boiler operating day rolling average SO₂ and NO_X limits utilizing a continuous emissions monitor (CEMS) subject to 40 CFR part 60.” Permit condition 38 identifies the source category type as being a boiler and the pollutants to be monitored by CEMS as SO₂ and NO_X. It is clear from the pollutant, fuel type, and the nature of the emission unit which of the tests would apply under 40 CFR 60 for demonstrating compliance. That is sufficient information to locate the performance specifications and quality assurance procedures for Power Boiler No. 2 to determine how to utilize CEMS to determine compliance with the SO₂ and NO_X limits of the Arkansas Regional Haze Phase III SIP revision. The State is being all-inclusive when referring to Part 60 to include all of the general provisions in Subpart A related to CEMS, such as 40 CFR 60.8 for performance tests, 40 CFR 60.13 pertaining to monitoring requirements, and Appendix B to Part 60, which includes performance specifications for CEMS. In addition, these permit conditions also implement APCEC Rule 19.703—Continuous Emission Monitoring, which is already part of the approved SIP, and applies to this source. Specific condition 54 of the permit provides additional information regarding CEMS requirements for Power Boiler No 2. Specifically, it says, “The permittee shall install, calibrate, maintain and operate continuous emissions monitoring systems for measuring SO₂ emissions, NO_X emissions, and either oxygen or carbon dioxide. The CEMS shall have readouts which demonstrate compliance with any of the applicable limits for the pollutant in question. The permittee shall comply with the DEQ CEMS conditions found in Appendix B. [Reg. 19.703, 40 CFR 52, Subpart E, and Ark. Code Ann. § 8-4-203 as referenced by Ark. Code Ann. §§ 8-4-304 and 8-4-311].” Appendix B sections II through IV of the permit lay out specific guidelines for CEMS operating conditions.

The commenter also states that permit condition 40 fails to identify the specific AP-42 emission factors. Condition 40 refers to “the applicable natural gas AP-42 emission factors” and provides an appropriate description because the applicable emission factors are based on the nature of the emissions unit, fuel, and pollutants in question. As explained in the proposed approval, if Power Boiler No. 2 switches to natural gas combustion, the applicable natural gas AP-42 emission factors of 0.6 lb SO₂/MMscf, 280 lb NO_X/MMscf, and 7.6 lb PM₁₀/MMscf in conjunction with natural gas fuel usage records shall be used to demonstrate compliance with the BART emission limits. Therefore, the boiler will operate under CEMs, and these AP-42 emissions factors would only be used for estimation of emissions if Power Boiler No. 2 burns natural gas. We note, just as we did in the FIP, for which these provisions are replacing, that burning only natural gas would very likely be sufficient in itself to demonstrate that the boiler is complying with the SO₂ emission limit. SO₂ emissions from combustion of natural gas are inherently very low and are virtually eliminated during the combustion process. Any SO₂ emissions will be in trace amounts well below the BART alternative emission limit so there should be no concern that the alternative limit for SO₂ will be met. NO_X and PM₁₀ emissions are also expected to be lower than the BART alternative emission limit for natural gas combustion. Using the most conservative NO_X, SO₂, and PM₁₀ AP-42 factors (highest factor) for boiler combustion indicates that the BART alternative emission limits will be met even when firing natural gas at full capacity. Based on this information, any ambiguity in the use of AP-42 factors for compliance using only natural gas is not of concern because of the characteristically lower emissions during natural gas combustion. When natural gas is used, the limits in the BART alternative demonstration will be met. DEQ has the State authority to enforce these emission factors to document compliance and EPA will have federal authority once this approval takes effect.

The State made clear in its SIP submittal that the BART alternative SIP requirements for this source would be implemented in conjunction with preexisting SIP requirements for monitoring, reporting, and recordkeeping, thus ensuring that the emissions limitations applicable to this source under the BART alternative are practically enforceable. See Aug. 2019 SIP Submittal at 2. These provisions of Arkansas's air regulations have been approved by EPA into Arkansas' federally enforceable SIP. In particular, APCEC Rule 19 Chapter 7—Sampling, Monitoring and Reporting Requirements, sets forth the powers of DEQ in requiring sampling, monitoring, and reporting requirements at stationary sources. Specifically, any stationary source is subject to air emission sampling (APCEC Rule 19.702); continuous emission monitoring (APCEC Rule 19.703); recordkeeping and reporting requirements (APCEC Rule 19.705); and Public Availability of Emissions Data (APCEC Rule 19.706). All of these requirements will become federally enforceable against Domtar with EPA's final approval of this SIP submittal. For these reasons, conditions 38 and 40 contain sufficient specificity regarding testing for compliance for Power Boiler No. 2.

Comment B.3: The provisions for recordkeeping are inadequate for permit conditions 36 and 43. In addition to failing to require that “owners and operators” are subject to these provisions, these provisions fail to specify necessary specifics to determine compliance. For example, these provisions lack requirements that records shall be maintained for CEMS data; quality assurance and quality control activities for emissions measuring systems; major maintenance activities conducted on emission units, control equipment, and CEMS; and any other records required by the underlying requirements.

Response: We disagree with the commenter's assertion that the provisions for recordkeeping are inadequate for conditions 36 and 43. The commenter cites CAA section 110(a)(2)(F), 40 CFR 51 Subpart K, and the BART guidelines in identifying the applicable recordkeeping and reporting requirements. However, these requirements do not mandate the level of specificity the commenter would like to see regarding recordkeeping, and the commenter cites no authority for the notion that that level of specificity is required. Nor did the commenter cite any examples from other BART alternative actions that would demonstrate that the level of specificity of the recordkeeping requirements here is inconsistent with what has been approved in other SIPs. Commenter's suggestions do not reflect how the regulations are worded regarding recordkeeping and reporting, therefore, we conclude that the commenter has failed to establish how the recordkeeping and reporting requirements in 40 CFR 51 Subpart K, and the BART guidelines are not met by conditions 36 and 43. Permit conditions 36 and 43 clearly require maintaining “all records” necessary to determine compliance “for at least 5 years.” This is sufficient under the regional haze regulations. Further, such broad terms encompass many if not all of the specific enumerated types of records the commenter claims should be retained. The recordkeeping provisions in conditions 36 and 43 are, therefore, not lacking and are sufficient enough on their own merit to meet 40 CFR 51 Subpart K and the BART-alternative requirements of subpart P. As mentioned in the previous response, Appendix B sections II through IV of the permit lay out specific guidelines for CEMS operating conditions. These CEMS conditions are reflected in and administered by the State under APCEC Rule 19.703—Continuous Emission Monitoring. The State applies APCEC Rule 19.705 —Record Keeping and Reporting Requirements to air pollution sources subject to the regulation. The State made clear in its August 2019 SIP Submittal, at page 2, that these provisions apply to the Domtar Ashdown Mill for purposes of implementing the BART alternative emission limitations at Power Boilers No. 1 and No. 2. These requirements will become federally enforceable against Domtar with final approval of this SIP submittal.

The commenter lastly mentioned that these conditions fail to require that “owners and operators” are subject to the provisions in them. We address this in response to comment B.5 in section III.B of this final action. As mentioned in that response, we recognize Domtar as both the permittee and the owner subject to the permit conditions. Further, because the permit conditions are being incorporated into the state's SIP, they are state- and federally-enforceable on any owner or operator of this facility regardless of any changes that may occur in ownership of the facility or in the permit itself. Therefore, Domtar and any future owner or operator is subject to the provisions being approved in this action, including conditions 36 and 43, and DEQ will continue to enforce these measures with EPA oversight.

Comment B.4: EPA's proposal suggests there are reporting requirements for Power Boiler No. 1 in conditions 33 to 36 and in conditions 38 to 43 for Power Boiler No. 2 but these provisions do not contain requirements for reporting. The SIP lacks any requirements for reporting and EPA must disapprove the SIP.

Response: The commenter asserts that conditions 33 to 36 for Power Boiler No. 1 and conditions 38 to 43 for Power Boiler No. 2 fail to contain reporting requirements as EPA suggests. However, permit conditions 36 and 43 state that all records “shall be made available to any agent of DEQ or EPA upon request.” Accordingly, the records will be provided upon request by DEQ or EPA. This is sufficient to satisfy periodic reporting of records in 40 CFR 51.211. The general BART alternative implementation requirements of 51.308(e)(2)(iii), which do not include a requirement of reporting on any specific time period, are also met. The commenter also suggests that the State is required to provide periodic reporting requirements as stated in 42 U.S.C. 7410(a)(2)(F)(ii) and the BART guidelines. However, section 110(a)(2)(F) requires EPA to “prescribe” its requirements, and thus this provision is implemented through the applicable regulations. The BART guidelines call for adequate reporting and recordkeeping so that air quality agency personnel can determine the compliance status of the source. Permit conditions 36 and 43 clearly require maintaining “all records” necessary to determine compliance “for at least 5 years” and permit conditions 36 and 43 state that all records “shall be made available to any agent of DEQ or EPA upon request” so determination of compliance can be made.

Further, other SIP-approved provisions of Arkansas' regulations also apply, ensuring the reporting obligations of 51.211 and the BART-alternative implementation measures of 51.308(e)(2)(iii) are satisfied. The commenter mentions that the SIP lacks any requirements for reporting, but that is not the case. APCEC Rule 19 Chapter 7—Sampling, Monitoring and Reporting Requirements, sets forth the powers of DEQ in requiring sampling, monitoring, and reporting requirements at stationary sources. As mentioned previously, the State made clear in its SIP submittal that the BART alternative SIP requirements for this source would be implemented in conjunction with preexisting SIP requirements for sampling, monitoring, and reporting requirements under APCEC Rule 19 Chapter 7, thus ensuring that the emissions limitations applicable to this source under the BART alternative are practically enforceable. Per APCEC Rule 19.705(C), Domtar must submit annual reports demonstrating compliance with applicable emission limitations. In addition, they must keep all records demonstrating compliance for at least five years (APCEC Rule 19.705(A)). Inspectors audit these records during site inspections. Therefore, Domtar does have a pre-existing annual reporting requirement, and, with the approval of the BART-alternative emission limits into the State's regional haze SIP, their compliance with these emission limits will also be a part of that annual report going forward. For these reasons, the SIP is not lacking reporting requirements, including any periodic reporting requirement as required under part 51, subpart K.

It is also worth noting that as a source subject to Title V requirements, it is subject to annual deviation reports under APCEC Rule 26.703(E)(3)(c). In addition, as a major source it is required to provide an annual emissions inventory. EPA finds that the reporting requirements applicable to Domtar under this SIP submittal are sufficient to meet the requirements of the BART alternative regulations and subpart K.

Comment B.5: The SIP fails to require that the source surveillance provisions apply to owners and operators. The source surveillance provisions must apply to owners and operators of the source instead of the Title V permittee in permit condition 32. This provision does not meet the requirements of subpart K. If the Title V permit were to expire, there would be no permittee to hold accountable. EPA must therefore disapprove this provision of the SIP because it fails to identify the appropriate liable entity. Similarly, permit condition 33 fails to specify the entity responsible for making the demonstration, and therefore, EPA must also disapprove this provision.

Response: The commenter stated that the SIP fails to require that the source surveillance provisions apply to owners and operators. EPA disagrees with this comment because the terms of the permit are incorporated into the SIP and are therefore applicable to both the permittee and any other owner or operator of this facility. Currently, those entities are one and the same: Domtar. Because conditions 32 and 33 in the permit both say “permittee” instead of “owner and operator,” the commenter asserts that nobody will be subject to the provisions in these conditions if the Title V permit were to expire. This is incorrect, and nothing in the State's SIP submittal or any other information before the EPA suggests that this is how these terms are to be interpreted. The terms “permittee” and “owner” are both used in the permit. Domtar is recognized as both the owner of the Ashdown mill who operates the boilers and the permittee of the Title V permit containing the revised conditions implementing the BART alternative. “The BART Alternative specific conditions” portion of the plantwide conditions section of the permit clarifies that the permittee is the one who is subject to these conditions.

In addition, these requirements would not cease to apply if Domtar were for any reason to cease to be the permittee. Although “permittee” is being used in the wording of the permit conditions, these conditions are being approved into the State's SIP and are state- and federally-enforceable by virtue of being in the SIP. As the State's SIP submittal explains, “For compliance with the CAA Regional Haze Program's requirements for the first planning period, the No. 1 and 2 Power Boilers are subject-to-BART alternative measures consistent with 40 CFR 51.308. The terms and conditions of the BART alternative measures are to be submitted to EPA for approval as part of the Arkansas SIP. Upon initial EPA approval of the permit into the SIP, the permittee shall continue to be subject to the conditions as approved into the SIP even if the conditions are revised as part of a permit amendment until such time as the EPA approves any revised conditions into the SIP. The permittee shall remain subject to both the initial SIP-approved conditions and the revised conditions, until EPA approves the revised conditions” (emphasis added). Because of this, should the Title V permit expire, be modified, or transferred, any person who owns or operates this facility, including the current permittee, will still be subject to these conditions as a result of their being incorporated into the federally enforceable SIP. We note in addition that permits are transferable due to changes in ownership of a source, given proper notification to the director including required disclosures. In terms of expiration, the Arkansas program is based on a one permit system meaning that a source contains a single document that contains both the Title I New Source Review (NSR) and Title V permit conditions/requirements. The conditions of the NSR permit do not ever expire. Title V permits do have a permit expiration date, but the expiration of the Title V permit does not impact the “status” of NSR permit requirements. These requirements live on unless modified/removed via an NSR permit action. Because NSR permit changes are automatically updated in the Title V permit there isn't any impact on operational status if the NSR permit was modified.

Therefore, the provisions in conditions 32 and 33 and in other provisions addressing ownership will continue to be enforceable requirements, regardless of who owns or operates this facility, and DEQ and EPA will continue to be able to enforce these measures. We, therefore, disagree that these conditions need to place requirements on the “owners and operators” rather than the “permittee” to be permanently enforceable.

Comment B.6: The SIP lacks enforceable provisions regarding permanent retirement. The SIP provides an option for permanent retirement of Power Boiler No. 1, but permit condition 34 lacks enforceable language. This permit condition and EPA's proposal lack the details necessary for enforcement. For example, it fails to explain what a “disconnection notice” is and what information is contained in the notice. Therefore, the public is unable to assess whether a “disconnection notice” is a permanent action that satisfies the BART requirements. EPA is prohibited from approving this additional BART alternative since the condition contains vague and unenforceable language.

Response: We disagree with the commenter that the SIP lacks enforceable provisions in condition 34 regarding permanent retirement. The term “disconnection notice” is self-defining in that it simply describes DEQ receiving communication in the form of a notice after Power Boiler No. 1 has already been taken out of service and is permanently retired. “Permanently retired” self-evidently means that once the power boiler is taken out of service it will never operate again. Indeed, this has already occurred. As indicated in a November 18, 2020, letter to DEQ from Domtar, the No. 1 Power Boiler was placed in standby mode and stopped operating in April 2016. That letter also documented that the unit was disconnected and permanently retired on August 6, 2018, with the removal of a section of boiler feedwater piping that prevents the boiler from producing steam. In addition, finalization of the permit amendment 0287-AOP-R23 removed authority for Domtar to operate No. 1 Power Boiler. As stated in an April 15, 2020, permit revision, “By request of the facility, this source has been retired and removed from the permit as a source in permit revision #23. The specific conditions have been marked, by request of the facility, as reserved in order to not change the numbering of the subsequent conditions. SN-03 is subject to the Regional Haze Program, specifically the BART Alternative. These conditions can be found starting with Plantwide Condition 32.” Because Domtar has requested that Power Boiler No. 1 be retired and removed as a source from the permit, the source specific permit provisions have been removed from the permit for Power Boiler No. 1 and they are not authorized to operate the unit. Power Boiler No. 1 is in compliance with the BART alternative limits by virtue of being permanently retired and therefore not emitting any of the relevant visibility pollutants. The numerical emission limits will apply, even though the unit has been taken out of service. DEQ has State authority established in its SIP, including APCEC Rule Chapter 7, for any other reporting requirements including documenting source retirement of this unit. For this reason, this condition does not lack enforceable provisions for retirement.

Comment B.7: The SIP neither specifies a compliance date nor requires compliance at all times. BART must reflect the best system of continuous emission reduction and the BART limits must apply at all times. EPA must clarify that the permit conditions proposed for approval in the SIP apply at all times. Furthermore, permit conditions 38 and 41 cross reference test methods found in other regulations that are inconsistent with the BART requirements since they do not require compliance at all times and exempt emissions during certain activities. These regulations and associated test methods are inconsistent with BART in that they do not require compliance at all times and exempt emissions during certain activities.

Response: We disagree with the commenter that the permit conditions do not apply at all times. There is no language in the proposed limits to suggest that they do not apply at all times. Conditions 32 and 37, which describe the emission rates for the power boilers, both say, “The permittee shall not exceed the emission rates set forth in the following table. The limits are based on a 30-day boiler operating day rolling average. 30 boiler operating day rolling average is defined as the arithmetic average of 30 consecutive daily values in which there is any hour of operation, and where each daily value is generated by summing the pounds of pollutant for that day and dividing the total by the sum of the hours the boiler was operating that day. A day is from 6 a.m. one calendar day to 6 a.m. the following calendar day. [Reg.19.304, 40 CFR 51.308(e)(2), and 40 CFR 52.173].” The language for permit conditions 38 and 41 describes ongoing compliance action into the future and does not indicate that the emission limits would cease or not apply continuously. Therefore, the BART alternative limits that we proposed to approve do indeed apply at all times.

The commenter argues that certain permit conditions cross-reference test methods in other regulations (i.e., the NESHAP, MACT and NSPS), which they allege are inconsistent with BART requirements since they do not require compliance at all times and exempt emissions during certain activities. The commenter specifically identifies this flaw in condition 38 pertaining to 40 CFR 60 and condition 41 pertaining to 40 CFR 63 subpart DDDDD. Programs like the NESHAP, MACT, and NSPS have different requirements, such as performance testing that is carried out over certain time frames that demonstrates compliance for particular pollutants. While those types of emission tests may have been designed to serve a different regulatory purpose, they are not in conflict with the BART requirements; nor do they override the BART alternative emission limits express set forth in the permit. There is no legal or regulatory barrier to incorporating performance testing requirements found in other regulatory programs as a means of implementing and ensuring compliance with a BART alternative. The commenter fails to demonstrate with reasonable specificity how the use of testing requirements that are intended to meet other criteria are in conflict or fail to meet the BART alternative requirements.

Further, the State made clear which test methods from those regulations are required for demonstrating compliance with these conditions. With respect to condition 38's reference to 40 CFR 60, the requirement to use CEMS to demonstrate compliance for SO₂ and NO_X is clear, unambiguous, and continuous. The State is being all-inclusive when referring to Part 60 to include all of the general provisions in Subpart A related to CEMS such as 40 CFR 60.8 for performance tests, 40 CFR 60.13 pertaining to monitoring requirements, and Appendix B to Part 60 that includes performance specifications. In addition, these permit conditions also implement APCEC Rule 19.703—Continuous Emission Monitoring, which is already part of the approved SIP, and applies to this source. Appendix B sections II through IV of the permit lay out specific guidelines for CEMS operating conditions. With respect to condition 41's reference to 40 CFR 63 subpart DDDDD, condition 41 clearly explains that the applicable PM₁₀ compliance demonstration requirements from 40 CFR part 63 subpart DDDDD shall be utilized. These requirements, which are at 40 CFR 63.7505-63.7541, do not cease and are ongoing. In response to comment B.8 in section III of this final action, we address the alternative option provided in the permit for monitoring emissions from Power Boiler #2 when that unit is combusting natural gas. Either method, however, provides for demonstration of continuous compliance with the BART alternative emission limits for PM₁₀. For these reasons, the test methods in conditions 38 and 41 are sufficient to provide continuous compliance and are not in conflict with the BART requirements.

The commenter particularly notes that because the permit conditions do not reference specific sections in these regulations, it is unclear whether the startup, shutdown, and malfunction emissions are included or exempt from monitoring. The commenter does not establish with reasonable specificity which of the performance testing or monitoring requirements from part 60 or part 63 would be affected here by provisions in those parts relating to “startup, shutdown, and malfunction.” Also, Table 10 to subpart DDDDD of Part 63 shows that SSM plan requirements and actions taken to minimize emissions during startup, shutdown, or malfunction are not required for subpart DDDDD.

The commenter lastly mentions that the State's SIP fails to include the schedule and timetable for compliance. We address comments regarding the schedule and timetable for compliance in response to comment C.1 in section III.C of this final action. These new BART alternative limits became enforceable by the State immediately upon issuance of a minor modification letter sent by the State to Domtar on February 28, 2019. The two Domtar power boilers have already been operating at emission levels below the proposed BART alternative emission limits since December 2016, three years before the limits became enforceable, continuing to do so through February 2019 and up to the present. The BART alternative limits and all associated permit conditions will become federally enforceable upon the effective date of this final action approving the SIP.

Comment B.8: The PM₁₀ test method for Power Boiler No. 2 permit is inappropriately conditioned on applicability under another regulation. The BART emission limits must have test methods that apply at all times. Permit condition 41 lacks enforceability in this regard. This permit condition is conditioned on when a National Emission Standards for Hazardous Air Pollutants (NESHAP) rule applies to this boiler. In other words, “while” the boiler “is subject to” the NESHAP, the requirements of the NESHAP rule are used to demonstrate compliance. In the event this boiler is no longer subject to the NESHAP, there would no longer be compliance demonstration requirements for the BART emission limits. This provision lacks specificity regarding the specific test methods in 40 CFR part 63 subpart DDDDD that apply and fails to identify what entity is required to meet these requirements.

Response: We disagree with the commenter that the PM₁₀ test method for Power Boiler No. 2 permit is inappropriately conditioned on applicability under another regulation. The commenter suggests that the word “while” in condition 41 is being used to allow avoidance of the BART alternative emission limit for PM₁₀. As we explained in our proposed action, “Since Power Boiler No. 2 is subject to 40 CFR part 63 subpart [DDDDD], the applicable PM₁₀ compliance demonstration requirements under the Boiler MACT shall be utilized to demonstrate compliance for PM₁₀ emissions (condition 41). If Power Boiler No. 2 switches to natural gas combustion, the applicable natural gas AP-42 emission factors of 0.6 lb SO₂/MMscf, 280 lb NO_X/MMscf, and 7.6 lb PM₁₀/MMscf in conjunction with natural gas fuel usage records (condition 40) shall be used to demonstrate compliance with the BART emission limits.” Therefore, “while” is used to draw a contrasting relationship between MACT, subpart DDDDD, and switching to natural gas combustion. If Power Boiler No. 2 switches to natural gas, fuel usage records will then apply for compliance demonstration. If the boiler does not burn natural gas only, then Power Boiler No. 2 is subject to 40 CFR 63 subpart DDDDD as an ongoing requirement for PM₁₀, and that requirement would not cease at any time.

The commenter also claims that permit condition 41 fails to identify which specific test methods found in 40 CFR 63 subpart DDDDD would apply. We disagree with this statement. Although the revised permit condition 41 does not spell out specific test methods, that does not mean it is not clear which test methods apply. In regard to 40 CFR 63 DDDDD, boiler MACT test methods are quite detailed and specific and are based on the source-specific unit type and pollutant emissions to be tested. It is clear from the pollutant, fuel type, and the nature of the emission unit here which of the tests would apply under DDDDD. Therefore, there is sufficient information to determine compliance. Table 10 to subpart DDDDD of Part 63 shows the applicable general provisions and includes performance testing requirements in 40 CFR 63.7. Continuous compliance is demonstrated for PM₁₀ under MACT, subpart DDDDD by maintaining the appropriate operating limit, depending on the control technology used (see Table 4 of subpart DDDDD). In this case, Power Boiler No. 2 uses venturi scrubbers so a site-specific minimum scrubber pressure drop and minimum flow rate operating limit according to 40 CFR 63.7530 would be used as the operating parameters. If no control device is used to demonstrate compliance with the PM₁₀ limit, the facility must monitor operating load (see item 8 of Table 4 and item 10 of Table 8) based on the operating limit set during the most recent PM₁₀ performance test (item 8 of Table 4 of subpart DDDDD), or by maintaining fuel records (40 CFR 63.7555(d)(1)) which is what will occur if Power Boiler No. 2 burns natural gas, as previously stated. Using the most conservative PM₁₀ AP-42 factor (highest factor) for boiler combustion indicates that the BART alternative emission limits will be met even when firing natural gas at full capacity.

Finally, the commenter mentions that this provision fails to identify what entity is required to meet these requirements (i.e., the owner or operator). The has been addressed previously in our response to comment B.5.

Comment B.9: The permit conditions appear to preclude the use of any credible evidence. EPA's proposal fails to explain whether the test procedures in the permit conditions are the “only” evidence that may be used to demonstrate compliance. EPA must disapprove the State's SIP submittal if approving these permit conditions were to preclude the use of any credible evidence.

Response: We disagree with the commenter that the permit conditions in any way preclude or appear to preclude the use of any credible evidence. The commenter does not identify anything in the permit or the Arkansas SIP that would preclude the use of other credible evidence. Both the SIP and the permit make clear that credible evidence can be used to determine compliance.

First, the SIP includes APCEC Regulation 19.701—Purpose, which states, “The purpose of this chapter is to generally define the powers of the Department in requiring sampling, monitoring, and reporting requirements at stationary sources. The Department shall enforce all properly incorporated and delegated federal testing requirements at a minimum. Any credible evidence based on sampling, monitoring, and reporting may be used to determine violations of applicable emission limitations” Similarly, general provision #27 of the Domtar permit provides that, “Any credible evidence based on sampling, monitoring, and reporting may be used to determine violations of applicable emission limitations. [Reg.18.1001, Reg.19.701, Ark. Code Ann. § 8-4-203 as referenced by Ark. Code Ann. §§ 8-4-304 and 8-4-311, and 40 CFR 52 Subpart E]” Lastly, the Credible Evidence Revisions rule revised 40 CFR parts 51, 52, 60, and 61 to permit the use of any credible evidence (i.e., both reference test data and comparable non-reference test data) to prove or disprove CAA violations in enforcement actions. In this regard, the preamble to the rule states: “These revisions make clear that enforcement authorities can prosecute actions based exclusively on any credible evidence, without the need to rely on any data from a particular reference test.” Therefore, although the permit does not specifically identify all types of evidence that may be used to determine compliance or non-compliance, neither the permit conditions nor the SIP preclude the use of any credible evidence. Furthermore, any attempt to specifically enumerate the types of evidence that may be used to determine compliance would undermine the purpose of the Credible Evidence Revisions rule. Thus, the requirement in subpart K, 40 CFR 51.212(c), is met.

Comment B.10: The proposal lacks an analysis and determination as to whether the monitoring requirements are met. Section 110(a)(2)(F)(i) covers monitoring emissions by owners and operators from stationary sources, and 40 CFR 51.214 contains explicit monitoring requirements. EPA's proposal fails to explain whether the permit conditions proposed for approving into the SIP meet these requirements.

Response: We disagree with the commenter's assertion that the proposal lacks an analysis and determination as to whether the permit conditions meet the monitoring requirements in CAA section 110(a)(2)(F)(i) and 40 CFR 51.214. The Arkansas Regional Haze Phase III SIP revision meets the applicable monitoring requirements under 40 CFR 51.214. In addition, it meets the applicable requirements found in 40 CFR 51.308(e)(2)(iii), which discusses rules for accounting and monitoring emissions, and procedures for enforcement of BART alternatives. This is established through our analysis of the monitoring regime discussed above in response to comments 2.B.3, 2.B.4, and 2.B.7. Commenter does not provide any further information with reasonable specificity as to how the applicable monitoring requirements in subparts K or P fail to be met. As discussed previously, the Arkansas SIP includes procedures in APCEC Regulation 19.703, including detailed information regarding CEMS, which DEQ has authority to administer. These procedures are already part of the State's plan requiring monitoring of this source's emissions. Because these monitoring provisions have already been adopted into the Arkansas SIP, the permit conditions pertaining to the BART alternative conditions will be administered under these existing approved provisions for monitoring. This is sufficient to meet the monitoring requirements in 40 CFR 51.214 and 40 CFR 51.308(e)(2)(iii). Therefore, the applicable monitoring requirements for this SIP revision are being met.

C. Requirements for Emissions Reductions To Occur During the First Implementation Period and a Compliance Schedule

Comment C.1: The SIP fails to demonstrate that emission reductions occurred during the first planning period by December 31, 2018 pursuant to 40 CFR 51.308(e)(2)(iii). EPA's proposal describes the emission reductions, but fails to explain whether the SIP contains the provisions necessary to satisfy regulatory requirements. For example, there are no compliance dates in the SIP that shows the emission limitations were enforceable in the first planning period. Furthermore, there is nothing in the SIP that demonstrates the monitoring, recordkeeping, and reporting requirements applied during the first planning period. Therefore, EPA lacks a basis to approve the SIP as meeting the element of the rule that the emission reductions occurred within the first planning period. Related to this issue, EPA's proposal suggests that the SIP included compliance schedules for Domtar, but the SIP fails to include any compliance schedules.

Response: We disagree with the commenter that the SIP fails to demonstrate that the required emission reductions occurred during the first planning period or that the SIP otherwise fails to meet the requirements of 40 CFR 51.308(e)(2)(iii). In our proposed approval, we explained that even though the BART alternative emission limits became enforceable by the State upon issuance of a minor modification letter sent by the State to Domtar on February 28, 2019, Domtar provided documentation demonstrating that Power Boilers No. 1 and 2 have been operating at emission levels below the BART alternative emission limits since December 2016. This shows that although the limits became enforceable shortly after the 2008 to 2018 planning period ended, Domtar had been in compliance with those limits for three years prior to the first planning period ending. Domtar's emission levels remained below the BART alternative levels up to the point at which the State's BART alternative emission limits and associated requirements became enforceable in February 2019. This is sufficient for the SIP submittal to meet the requirement of 40 CFR 51.308(e)(2)(iii).

The commenter argues that there is nothing in the SIP that demonstrates the monitoring, recordkeeping, and reporting requirements applied to the source during the first planning period. First, 40 CFR 51.308(e)(2)(iii) does not impose this requirement and neither does any other provision of the BART alternative regulations. Rather, in order to demonstrate that BART alternative emission limits are being achieved by the end of the first planning period, “the State must provide a detailed description of the emissions trading program or other alternative measure, including schedules for implementation, the emission reductions required by the program, all necessary administrative and technical procedures for implementing the program, rules for accounting and monitoring emissions, and procedures for enforcement.” EPA does not interpret this language as requiring that the monitoring, recordkeeping, and reporting requirements associated with a BART alternative must be in place and be state- or federally-enforceable before the end of the first planning period. The SIP must include such requirements, but with respect to demonstrating when they are applied to the source, it is reasonable that such requirements accompany the BART alternative. As discussed in the paragraph above, the reductions secured under the BART alternative have been documented to occur before the end of the first planning period, and the documentation further demonstrates that the requisite emission levels were maintained up until the point that the State imposed the enforceable BART-alternative emission limits and associated monitoring, recordkeeping, and reporting requirements on the source. This is sufficient to satisfy 40 CFR 51.308(e)(2)(iii).

In particular, the compliance documentation included a letter dated December 20, 2018, submitted to DEQ by Domtar, providing emissions data for Power Boilers No. 1 and 2 from December 2016 to November 2018. The letter noted that because Power Boiler No. 1 has been in standby mode, it has emitted zero emissions since early 2016. The letter also provided CEMS daily average and thirty-day rolling average emissions data for SO₂ and NO_X for Power Boiler No. 2 from December 1, 2016 through November 30, 2018. Based on that CEMS data, the highest thirty-day rolling averages for Power Boiler No. 2 were found to be 294 pph SO₂ and 179 pph NO_X, which are below the BART alternative emission limits of 435 pph SO₂ and 293 pph NO_X. The December 20, 2018 letter explained that compliance with the PM₁₀ BART alternative limit for Power Boiler No. 2 is demonstrated via compliance with the Boiler MACT. Based on previous compliance stack testing results conducted by Domtar in January 2016, PM₁₀ emissions for Power Boiler No. 2 are equal to 34 pph PM₁₀, which is below the BART alternative PM₁₀ emission limit of 81.6 pph PM₁₀. This demonstrates that Power Boilers No. 1 and No. 2 at the Ashdown Mill satisfy the timing requirements of 40 CFR 51.308(e)(2)(iii) and shows that the necessary emission reductions associated with the BART alternative occurred during the first long-term strategy period for regional haze.

In addition to being in compliance before the first implementation period ended, Domtar submitted additional letters to DEQ showing continued compliance for both power boilers. The letters contained CEMS emission data from January 2018 to April 2019. This CEMS data demonstrated continued compliance for Power Boiler No. 2 by showing emission levels below the BART alternative emission limits beyond 2018. Domtar noted that Power Boiler No. 1 continued to be in standby mode and that its emissions have been zero since early 2016.The Domtar letters also noted that the CEMS daily average and thirty-day rolling average emissions for SO₂ and NO_X were below the BART alternative limits for each month from January 2018 to April 2019. Additionally, based on the previous January 2016 Boiler MACT stack testing results, actual PM₁₀ emissions from Power Boiler No. 2 were conservatively estimated to be 48 pph PM₁₀, which is below the BART alternative emission limit of 81.6 pph PM₁₀ for Power Boiler No. 2.

The commenter argues that there are no compliance dates in the SIP that show that the emission limitations were enforceable in the first planning period. This is not required by EPA's regulations, as explained above. In addition, there is no schedule for future compliance because the source is already complying with the emission limits which are already in place and enforceable through the State permit. Upon the effective date of this final action the emission limits (and associated requirements) will be federally enforceable. These provisions have never been administratively or judicially stayed, are currently in effect, and will remain in effect; the source has been compliant with those requirements. We note with respect to the SO₂ and NO_X BART limits promulgated by the FIP, which is now being withdrawn in this action, the compliance schedule did not require that these limits be in effect until October 27, 2021. Domtar has been in compliance with those schedules for both boilers for the past three years.

For these reasons, the State's BART alternative SIP revision for Domtar Ashdown Mill meets the provisions of 40 CFR 51.308(e)(2)(iii): It documents that the required reductions took place during the period of the first long-term strategy (i.e. before the end of 2018) and those reductions continued up until the point the enforceable BART alternative emission limits took effect at the state level. The BART alternative limits are now in effect, satisfying the implementation-schedule requirement of (e)(2)(iii), and the SIP establishes relevant monitoring, recordkeeping, and reporting requirements, as set forth in plantwide permit conditions 32 to 43 and the associated provisions of the State's SIP-approved monitoring and compliance regulations found at APCEC Rule 19, Chapter 7.

D. The CAA 110(l) Anti-Backsliding Provision

Comment D.1: The proposed rule violates the Clean Air Act's “anti-backsliding” requirement at 42 U.S.C. 7410(l) because compared to the existing federal plan, the State's plan would result in greater air pollution. EPA's proposal explains that “[b]ased on an assessment of current air quality in the areas most affected by this SIP revision, we are concluding that the less stringent SO₂ emission limits in the Phase III SIP will not interfere with attainment of the NAAQS.” EPA's proposal fails to explain and provide information regarding what areas it assessed and the basis for its assessment. Moreover, EPA's analysis only considers regional haze and the NAAQS, and not other CAA requirements such as PSD increments. Moreover, the increase in SO₂ emissions under the SIP relative to the FIP violates the Clean Air Act's section 110(l) anti-backsliding provision, which provides that “[t]he Administrator shall not approve a revision of a plan if the revision would interfere with any applicable requirement concerning attainment and reasonable further progress . . . or any other applicable requirement of this chapter.” Section 110(l) prohibits plan revisions that would interfere with any applicable requirement, including a BART determination. When determining whether a plan revision interferes with NAAQS attainment, EPA has interpreted section 110(l) as preventing plan revisions that would increase overall air pollution or worsen air quality. In Kentucky Resources Council, Inc. v. EPA, 467 F.3d 986 (6th Cir. 2006), EPA interpreted section 110(l) as allowing the agency to approve a plan revision that weakened some existing control measures while strengthening others, but only “[a]s long as actual emissions in the air are not increased.” The Eleventh Circuit and the Seventh Circuit have upheld EPA's section 110(l) interpretation as prohibiting plan revisions that would increase emissions or worsen air quality. In a discussion regarding a challenge to the Nevada regional haze plan, the Ninth Circuit also suggested that a haze plan that “weakens or removes any pollution controls” would violate section 110(l). Emissions under the Domtar BART alternative would increase, which is plainly at odds with CAA anti-backsliding requirements and the interpretation of these provisions in various circuit courts.

Response: We disagree with the commenter that “the proposed rule violates the CAA's anti-backsliding requirement due to an increase in SO₂ emissions under the SIP relative to the FIP.” For the reasons explained below, EPA concludes that CAA section 110(l) does not prohibit approval of this SIP.

Under CAA Section 110(l), the EPA cannot approve a plan revision “if the revision would interfere with any applicable requirement concerning attainment and reasonable further progress, or any other applicable requirement of this chapter.” Section 110(l) applies to all requirements of the CAA, and it applies to all areas of the country, whether attainment, nonattainment, unclassifiable, or maintenance for one or more of the six criteria pollutants. The EPA interprets section 110(l) as applying to all NAAQS that are in effect, including those for which SIP submissions have not been made. A section 110(l) demonstration should address all pollutants whose emissions and/or ambient concentrations may change as a result of a plan revision, even if the SIP provision was originally adopted only to address one particular NAAQS. In general, the level of rigor needed for any CAA section 110(l) demonstration will vary depending on the nature of the revision. Where available attainment demonstration or maintenance plans indicate that any change in emissions will not interfere with any applicable requirement concerning attainment and reasonable further progress, or any other applicable CAA requirement, EPA may rely on such plans to demonstrate that section 110(l) does not prohibit approval of the plan.

A state, instead of submitting an air quality analysis showing that the revision will not interfere with any applicable requirement, may substitute equivalent emissions reductions to compensate for any change to a plan to ensure actual emissions to the air are not increased and thus preserve status quo air quality. Equivalent emissions reductions are reductions that are equal to or greater than those reductions achieved by the control measure approved into the plan. To show that compensating emissions reductions are equivalent, adequate justification must be provided. The compensating, equivalent reductions should represent actual emissions reductions achieved in a contemporaneous time frame to the change of the existing control measure in order to preserve the status quo air quality. In addition to being contemporaneous, the equivalent emissions reductions should also be permanent, enforceable, quantifiable, and surplus. A showing that the substitute measures preserve status quo air quality is generally sufficient to demonstrate noninterference through this alternative approach.

As an initial matter, the commenter misstates the EPA's interpretation of CAA section 110(l). Neither EPA nor any court has concluded, as the commenter asserts, that plan revisions are permissible only if emissions to the air are not increased. The case cited by the commenter, Kentucky Resources Council, Inc. v. EPA, 467 F.3d 986 (6th Cir. 2006), involved a situation in which the state had opted to substitute equivalent emission reductions to compensate for emission changes associated with the plan revision, and the EPA concluded that the offsetting emission reductions were adequate to maintain the status quo air quality. Because no attainment demonstrations were available to guide an analysis of whether the revision would interfere with attainment of the NAAQS, the EPA had relied on its conclusion that status quo air quality would be maintained instead of conducting an air quality analysis evaluating the impact on attainment and maintenance of the NAAQS. The court upheld, as a reasonable reading of the statute entitled to deference, the EPA's conclusion that approval of the SIP revision was permissible in those circumstances. The court held that the use of substitute measures was permissible, not that such measures were required in every circumstance.

The Seventh Circuit decision mentioned by commenter—Indiana v. EPA, 796 F.3d 803 (7th Cir. 2015)—does not support commenter's argument. This case emphasizes that the EPA is required to determine whether the revision would, going forward, interfere with attainment. In Indiana, the court rejected arguments that the revised program could not be approved because it had led to a past O₃ NAAQS exceedance. The court also agreed that it was permissible for EPA to rely on the fact that the state demonstrated that substitute measures more than offset any increase associated with the plan revision. In the context of reviewing whether the substitute measures were sufficient, the court explained that “EPA can approve a SIP revision unless the agency finds it will make the air quality worse.” In doing so, however, the court did not hold that substitute measures are always required to demonstrate noninterference under CAA section 110(l) or that section 110(l) prohibits approval of any SIP revision which leads to an increase in emissions.

The Ninth Circuit decision commenters cite—WildEarth Guardians v. EPA, 759 F.3d 1064 (9th Cir. 2014)—also does not establish that EPA is prohibited from approving this SIP. In WildEarth Guardians, the Ninth Circuit rejected a challenge to an EPA action approving a haze plan and concluded that WildEarth had identified “nothing in the SIP that weakens or removes any pollution controls. And even if the SIP merely maintained the status quo, that would not interfere with the attainment or maintenance of the NAAQS.” For that reason, the court concluded that WildEarth had failed to show that EPA's approval of the SIP contravened CAA section 110(l). In brief, the court explained that a plan approval that does not weaken or remove pollution controls would not violate section 110(l). The court did not, however, suggest that any plan that weakens or removes pollution controls would necessarily violate CAA section 110(l). Several courts have deferred to EPA's interpretation of the phrase “would interfere” in CAA Section 110(l). In addition, determinations that are scientific in nature are entitled to the most deference on review. The county that Domtar is located in (Little River County) was previously designated as “Attainment/Unclassifiable.” for the 2010 SO₂ NAAQS. In addition, EPA has evaluated the air quality impact of the repeal of the FIP requiring BART controls and the approval of the BART alternative limits. As mentioned in the proposed approval, the BART alternative limits do not reduce SO₂ emissions as much as the BART controls in the FIP; however, all areas in Arkansas have been and are currently attaining all of the NAAQS, even though the SO₂ BART controls for Domtar have not been implemented. Those controls were not obligated to be in place until October 27, 2021, when the BART emission limits would have taken effect under the FIP. Therefore, even though the BART alternative will not achieve the same level of emission reductions for SO₂ as the BART FIP would have (in 2021), there is no reason to expect that this will negatively impact current air quality, which is already sufficient to attain the SO₂ NAAQS in Arkansas and (as discussed further below) any other areas that could be impacted by SO₂ emissions from this source. Further, the State of Missouri did not rely on reductions from Domtar for its regional haze plans, and the EPA is not aware of (nor has commenter identified) any other air quality analyses that rely on implementation of the BART requirements for Domtar in the FIP. The proposed withdrawal of the BART provisions in the FIP and replacement with the BART alternative requirements in the SIP will not cause air quality to become worse than current air quality or interfere with existing plans to attain and maintain the NAAQS.

The more stringent SO₂ emission limits for Domtar in the BART FIP did not go into effect before the SIP BART alternative replaced them. Given that current air quality is already sufficient to attain the SO₂ NAAQS in Arkansas and any other areas that could be impacted by SO₂ emissions from this source, there is no evidence that withdrawal of the SO₂ limits in the FIP for Domtar and the approval of the SO₂ emission limits in the Arkansas Regional Haze Phase III SIP revision will interfere with attainment of the 2010 one-hour SO₂ NAAQS or the 2006 24-hour or the 2012 annual PM_2.5 NAAQS (of which SO₂ is a precursor). In addition, Domtar provided documentation demonstrating that Power Boilers No. 1 and 2 have actually been operating at emission levels below the BART alternative emission limits since December 2016. At this time, all areas that would be potentially impacted by the increase in SO₂ emissions allowed under the SIP revision as compared to the FIP are attaining the 2010 one-hour SO₂ NAAQS, the 2006 24-hour PM_2.5 NAAQS, and the 2012 annual PM_2.5 NAAQS without the FIP-required controls being in operation. Based on this assessment of current air quality in the areas most affected by this SIP revision, we conclude that the less stringent SO₂ emission limits in the Phase III SIP will not interfere with attainment of these NAAQS.

The commenter states that EPA's proposal fails to explain and provide information regarding what areas it assessed and the basis for its assessment. With respect to regional haze requirements, we disagree with the commenter. We explained in the proposal that we considered all Class I areas in Arkansas and also considered those in Missouri, which is the only State that was determined to potentially be impacted by sources from within Arkansas for the first implementation period. Missouri is currently not relying on emission reductions from Domtar in its regional haze plan.

Further, there are no PM_2.5 or SO₂ nonattainment areas in any other state that could be impacted by the emissions from Domtar. Regarding PM nonattainment areas in other states, EPA previously approved Arkansas' interstate transport SIP submittals under CAA 110(a)(2)(D)(i)(I), which established that emissions from Arkansas do not significantly contribute to nonattainment or interfere with maintenance of the 2006 24-hour or 2012 annual PM_2.5 NAAQS in any other state. Concerning SO₂ nonattainment areas in other states, the nearest SO₂ nonattainment area to Domtar is within Titus County, Texas, approximately 100 km away. EPA designated part of Titus County, around the Monticello Power Plant, as nonattainment in Round 2 of the SO₂ designations process. Domtar is also not near any large SO₂ sources in other states. Large SO₂ sources greater than 100 tpy SO₂ in Oklahoma [IP Vaillant Paper Mill (100 km away) and Hugo Station (119 km away)] and Texas [Welsh Power Plant (95 km away)] are all approximately 100 km away from Domtar, which is too far for Domtar to contribute to air quality in those areas. 50 km is the useful distance to which AERMOD is considered accurate. Therefore, under the Data Requirements Rule (DRR), sources beyond 50 km were determined to not cause concentration gradient impacts within the area of analysis. The distance between Domtar and any of the large SO₂ sources in neighboring states makes it unlikely that SO₂ emissions from Arkansas interact with emissions from another state in such a way as to contribute to existing nonattainment of the 2010 one-hour SO₂ NAAQS. The DRR SO₂ monitor for the Welsh Power Plant (the closest large source to Domtar), showed attainment and characterized the air quality design value for 2017 to 2019 as 28 parts per billion (ppb) SO₂ which is below the 2010 one-hour SO₂ NAAQS of 75 ppb SO₂. For these reasons, we conclude that emissions from Domtar will not adversely impact air quality in PM_2.5 or SO₂ nonattainment areas in any other state.

The commenter argues that DEQ addressed the reasonable progress requirements under 40 CFR 51.308(d)(1) based on faulty analysis that the BART alternative for Domtar is approvable. We addressed objections to the BART alternative under 40 CFR 51.308(e) in section III.A of this final action and explained why the BART alternative provides greater reasonable progress for regional haze. We also explained how the reasonable progress requirements for regional haze under 40 CFR 51.308(d)(1) are being met, and found that reasonable progress was not impacted by the transition from the BART FIP requirements to the BART alternative at Domtar. Therefore, the BART alternative does not interfere with “reasonable progress” under the Regional Haze Rule as an “other CAA requirement” that could be affected under CAA 110(l).

The commenter mentioned that EPA's analysis only considers regional haze and the NAAQS, and not the other CAA requirements, for example, PSD increments. The commenter asserts that, for this reason also, EPA fails to demonstrate that withdrawing the FIP and approving the State's SIP complies with Section 110(l) of the Act. EPA did not evaluate PSD increments in the proposal for two reasons: (1) Both power boilers were in operation before the major source baseline trigger dates for all three pollutants with increments (SO₂, NO_X, and PM/PM₁₀/PM_2.5); and (2) both the FIP limits and alternative BART limits are less than past actual emissions (both on an annual tons per year basis and a short-term emission rate basis), so increment around the Domtar facility was being expanded, not consumed. We noted in our proposed approval that the BART alternative emission rates were 44 percent lower for SO₂ and 51 percent lower for NO₂ compared to previously permitted emission rates. Based on this and the knowledge that the power boilers historically have operated greater than 56 percent of their permitted rates on a short term and annual basis, it can be concluded that increment was being expanded by the BART alternative. The major source baseline trigger date for PM/PM₁₀/PM_2.5 and SO₂ increment was August 7, 1977. The major source baseline trigger date for NO_X was February 8, 1988. Both Power Boiler No. 1 and Power Boiler No. 2 are baseline increment sources since they received permits and/or were in operation before the major source baseline date for NO_X, SO₂ and PM/PM₁₀/PM_2.5 increments. PM/PM₁₀/PM_2.5, SO₂, and NO_X all have annual increment standards; SO₂ has a three-hour and a 24-hour increment standard, and PM/PM₁₀/PM_2.5 all have 24-hour Class II increment standards. The Air Quality Control Region (AQCR) that Domtar facility is located in is AQCR 22, and the minor source baseline date for AQCR 22 was triggered for PM/PM₁₀/PM_2.5 and SO₂ by a PSD permit modification (Domtar permit 287-AR-3) on May 31, 1983. The NO_X minor source baseline date was triggered for NO_X in AQCR 22 by a PSD permit modification (Domtar permit 946-A) on August 31, 1989.

The conversion of Power Boiler No. 1 to burn only natural gas was an increment expanding change. For the purpose of overall increment analysis, we evaluated the emissions of Power Boiler No. 1 prior to the conversion of only burning natural gas as these emissions were part of the pre-BART baseline. As can be seen in Table 1, the annual emission limits (tpy) for the Arkansas BART alternative are less than the Arkansas baseline actual emissions for SO₂, NO_X, and PM/PM₁₀/PM_2.5. Therefore, the Arkansas BART alternative results in annual increment expansion for all three pollutants.

As can be seen in Table 2, the short-term emission limits (pph) for the Arkansas BART alternative are less than the previously permitted limits, the Arkansas baseline (2001-2003 actual emissions), and the BART FIP baseline emissions (mixture of 2001-2003 and 2009-2011 actual emissions) for SO₂, NO_X, and PM/PM₁₀/PM_2.5. Therefore, the Arkansas BART alternative results in short-term increment expansion for SO₂ and PM/PM₁₀/PM_2.5 pollutants (there is no short term increment for NO_X). Therefore, removal of the FIP and approval of the Arkansas BART alternative would not interfere with PSD increments.

As discussed above, EPA's technical documentation shows that approval of the Arkansas SIP submittal is not prohibited under CAA section 110(l). As also explained above, CAA section 110(l) does not prohibit states from submitting a SIP less stringent than a FIP or replacing a SIP with a less stringent SIP. Even though the requirements adopted in the SIP revision here do not match the emissions limitations in the FIP, there is no expectation that approval of the SIP will interfere with attainment or maintenance of the NAAQS or any other requirements under the Act.

E. Interstate Visibility Transport and Regional Haze Reasonable Progress Requirements

Comment E.1: A state can satisfy prong 4 interstate transport requirements with a fully approved regional haze SIP. EPA's proposal contains numerous fatal flaws and EPA cannot approve the State's SIP submittal for Domtar Ashdown Mill. Therefore, EPA similarly cannot approve prong 4 since the State does not have a fully approvable regional haze SIP. Similarly, EPA cannot determine the State's SIP meets the reasonable progress requirements under 40 CFR 51.308(d)(1) since the State's BART alternative fails to comply with the Act and regulations.

Response: We disagree with the commenter's assertion that EPA is prohibited from approving the Arkansas SIP submission regarding interstate visibility transport requirements and regional haze reasonable progress requirements. As explained in our proposed rule, a state can demonstrate compliance with CAA section 110(a)(2)(D)(i)(II) prong 4 by either having a fully-approved regional haze SIP or by demonstrating that emissions within its jurisdiction do not interfere with another air agency's plans to protect visibility. The State addressed interstate visibility transport requirements in its 2018 Phase II SIP revision, as supplemented by the Arkansas 2015 O₃ NAAQS Interstate Transport SIP revision (submitted October 4, 2019), for the following NAAQS: the 2006 24-hour PM_2.5 NAAQS; the 2012 annual PM_2.5 NAAQS; the 2008 and 2015 eight-hour O₃ NAAQS; the 2010 one-hour NO₂ NAAQS; and the 2010 one-hour SO₂ NAAQS. The State's analysis in the Arkansas 2015 O₃ NAAQS Interstate Transport SIP supersedes the interstate visibility transport portion of the 2017 infrastructure SIP.

As to the first basis for approval, the Arkansas Regional Haze NO_X SIP revision (Phase I), the Arkansas Regional Haze SO₂ and PM SIP revision (Phase II), and this action (the Arkansas Regional Haze Phase III SIP revision) together fully address the State's outstanding regional haze requirements for the first planning period and address the deficiencies of the 2008 Arkansas Regional Haze SIP that were identified in EPA's March 12, 2012, action. Thus, Arkansas now has a fully-approved regional haze SIP for the first planning period. This is sufficient under EPA's 2013 infrastructure SIP guidance to determine that Arkansas has also adequately addressed interstate visibility transport under “prong 4” for the above-listed NAAQS.

As an alternative basis for approval of Arkansas' CAA section 110(a)(2)(D)(i)(II) prong 4 SIP submittals, EPA finds that DEQ provided an adequate demonstration that it is not interfering with other states' visibility programs in the Arkansas 2015 O₃ NAAQS Interstate Transport SIP submittal, which addressed the prong 4 requirements for the six NAAQS previously mentioned. Arkansas documented its apportionment of emission reduction obligations needed at the affected Class I areas in other states and provided a demonstration that the SIP includes approved federally enforceable measures that contribute to achieving the 2018 RPGs set for those areas. The demonstration showed that emissions within Arkansas' jurisdiction do not interfere with other air agencies' plans to protect visibility, as expressed via the 2018 reasonable progress goals for Class I areas in other states. In particular, Arkansas' SIP submittals demonstrated that the RPGs for the only two Class I areas outside Arkansas potentially impacted by Arkansas emissions, Hercules-Glades Wilderness and Mingo National Wildlife Refuge, in Missouri, were achieving the visibility goals that were determined through interstate consultation. Further, the emissions from certain EGU sources in Arkansas are demonstrated to be below the levels Arkansas had agreed to in the interstate consultation process.

For these reasons, Arkansas has fulfilled its prong 4 visibility transport requirements for the 2006 24-hour PM_2.5 NAAQS; the 2012 annual PM_2.5 NAAQS; the 2008 and 2015 eight-hour O₃ NAAQS; the 2010 one-hour NO₂ NAAQS; and the 2010 one-hour SO₂ NAAQS in accordance with EPA's 2013 infrastructure SIP guidance. This alternative basis for approving these SIP submittals is not dependent on Arkansas having a fully approved Regional Haze SIP for the first planning period, and it is not dependent on the emission reductions achieved by the BART alternative for the two BART sources at Domtar Ashdown Mill. Thus, this basis for these prong 4 approvals is independent and severable from any other aspect of this action. Such approvals, on this basis, would not be affected by any administrative or judicial action altering, modifying, vacating, remanding, staying, or enjoining any other aspect of this action.

The commenter's objections to EPA approving reasonable progress requirements have been addressed in previous responses in this document.

F. Comments From Domtar

Comment F.1: Overall the commenter agrees with EPA's summary of ADEQ's BART Alternative for the Ashdown Mill, and further agrees that the BART Alternative, by the clear weight of evidence, achieves greater reasonable progress than the FIP. Commenter supports EPA's determination that the BART Alternative meets the applicable Regional Haze requirements and supports approving DEQ's Regional Haze Phase III SIP submittal. Commenter also agrees and supports EPA's determination that with this submittal ADEQ has satisfied all of the regional haze first planning period SIP requirements for Domtar.

Response: We appreciate the commenter's support of our proposed approval.

Comment F.2: The commenter believes a sufficient demonstration was made to grant an exemption under 40 CFR 51.303. However, for purposes of these comments, the commenter supports EPA's proposal with the reservation that it reserves the right to raise challenges to EPA's modeling approach in any effort to impose further reductions on the Ashdown Mill emissions in any subsequent Regional Haze SIP proceedings that may involve the Ashdown Mill.

Response: We appreciate the commenter's support of our proposed approval. An exemption under 40 CFR 51.303 is outside the scope of this action.

Comment F.3: Two nonsubstantive corrections were suggested for consideration to make the proposed action record factually correct, but do not affect the BART alternative limits or conditions:

• At 14851, middle column about two-thirds of the way down, referring to Power Boiler 1: `It is equipped with a wet electrostatic precipitator. . . .' It should be stated `It was. . . .' The wet electrostatic precipitator is no longer needed after the boiler was converted to burn natural gas.
• At 14855, middle column just above Table 2, referring to the FIP's nitrogen oxide (NO_X) BART determination for Power Boiler 2: `. . . achieved by the installation and operation of low NO_X burners.' The reference to low NO_X burners needs to be removed.”

Response: The EPA agrees with commenter's non substantive textual edits and the proposed SIP approval should read as follows:

At 14851, “It is equipped with a wet electrostatic precipitator” should be changed to read:

“It was equipped with a wet electrostatic precipitator.” With the conversion and permit modification to burn only natural gas, the wet electrostatic precipitator is no longer needed to control PM emissions from Power Boiler 1.

At 14855, “The NO_X Best Available Retrofit Technology (BART) determination for Power Boiler No. 2 is an emission limit of 345 pph on a thirty boiler-operating-day rolling average, achieved by the installation and operation of low NO_X burners” should be changed to read: “The NO_X BART determination for Power Boiler No. 2 is an emission limit of 345 pph on a thirty boiler-operating-day rolling average consistent with the installation and operation of low NO_X burners.” (see 81 FR 66332, 66348). A BART determination is an emission limit based on the determination of a particular control strategy considering the BART factors, rather than a requirement to undertake the selected control.

These non-substantive textual edits do not impact our analysis and our final decision regarding approval of the BART alternative for Power Boilers No. 1 and 2.

IV. Final Action

A. Arkansas Regional Haze Phase III SIP Submittal

We finalize approval of the Arkansas Regional Haze Phase III SIP revision (submitted August 13, 2019) as meeting the applicable regional haze BART alternative provisions set forth in 40 CFR 51.308(e)(2) for the Domtar Ashdown Mill. Specifically, we finalize approval of the regional haze program-specific plantwide conditions 32 to 43 from section VI of permit revision #0287-AOP-R22 (effective August 1, 2019) into the SIP for implementing the Domtar BART alternative. These plantwide conditions of permit #0287-AOP-R22 include SO_2, NO_X, and PM₁₀ emission limits and associated conditions for implementing these BART alternative limits for Power Boiler No. 1 and Power Boiler No. 2.

We finalize approval of the reasonable progress components under 40 CFR 51.308(d)(1), to the extent they relate to Domtar Power Boilers No. 1 and 2. With the approved Phase I and II SIP revision requirements and the Arkansas Regional Haze Phase III BART alternative requirements being approved in this final action, Arkansas has addressed all reasonable progress requirements under 40 CFR51.308(d)(1) with a fully-approved regional haze SIP. We, therefore, finalize approval of the emission limits and schedules of compliance long-term strategy element under 40 CFR 51.308(d)(3)(v)(3) pertaining to the Domtar Ashdown Mill in the August 13, 2019, submittal. With the final approval of the BART alternative requirements for the Domtar Ashdown Mill being addressed in this action, DEQ has satisfied all long-term strategy requirements under 40 CFR 51.308(d)(3), as pertains to the first planning period for regional haze. We agree with DEQ's determination that the revised 2018 RPGs in the Phase II action do not need to be revised further. We finalize approval of the State's withdrawal of the current PM₁₀ BART determination of 0.07 lb/MMBtu for Power Boiler No. 1 in the 2008 Arkansas Regional Haze SIP, and approve its replacement with the PM₁₀ BART alternative limit in the Arkansas Regional Haze Phase III SIP submittal. We finalize approval of Arkansas' consultation with FLMs and Missouri and finalize our determination that the SIP submittal satisfies the consultation requirements under 40 CFR 51.308(i)(2) and 40 CFR 51.308(d)(3)(i).

B. Arkansas Visibility Transport

We finalize approval of the portion of the Arkansas 2015 O₃ NAAQS Interstate Transport SIP revision (submitted October 4, 2019) addressing CAA section 110(a)(2)(D)(i)(II) prong 4 visibility transport for the following six NAAQS: 2006 24-hour PM_2.5 NAAQS; the 2012 annual PM_2.5 NAAQS; the 2008 and 2015 eight-hour O₃ NAAQS; the 2010 one-hour NO₂ NAAQS; and the 2010 one-hour SO₂ NAAQS. We also finalize approval of the visibility transport portion of the 2018 Phase II SIP revision, as supplemented by the Arkansas 2015 O₃ NAAQS Interstate Transport SIP revision. The State's analysis in the Arkansas 2015 O₃ NAAQS Interstate Transport SIP supersedes the visibility transport portion of the 2017 infrastructure SIP. We finalize approval of the prong 4 portions of these SIP submittals on the basis that Arkansas has a fully-approved regional haze SIP with our final approval of the Arkansas Regional Haze Phase III SIP submittal. The Arkansas Regional Haze NO_X SIP revision, the Arkansas Regional Haze SO₂ and PM SIP revision, and the Arkansas Regional Haze Phase III SIP revision together fully address the deficiencies of the 2008 Arkansas Regional Haze SIP that were identified in the March 12, 2012 partial approval/partial disapproval action. Arkansas has a fully-approved regional haze SIP comprised of the portion of the 2008 Arkansas Regional Haze SIP approved in our 2012 final action, the Arkansas Regional Haze NO_X SIP revision, the Arkansas Regional Haze SO₂ and PM SIP revision, and the Arkansas Regional Haze Phase III SIP revision. A fully-approved regional haze plan ensures that emissions from Arkansas sources do not interfere with measures required to be included in another air agencies' plans to protect visibility. As an alternative basis for approval of CAA section 110(a)(2)(D)(i)(II) prong 4 for these NAAQS, we finalize a determination that Arkansas has provided an adequate demonstration in the October 4, 2019 submittal showing that emissions within its jurisdiction do not interfere with other air agencies' plans to protect visibility.

C. CAA Section 110(l)

We finalize our determination that approval of the Arkansas Regional Haze Phase III SIP revision and concurrent withdrawal of the corresponding parts of the FIP meet the provisions of CAA section 110(l).

V. Incorporation by Reference

In this rule, the EPA is finalizing regulatory text that includes incorporation by reference. In accordance with requirements of 1 CFR 51.5, the EPA is finalizing the incorporation by reference of revisions to the Arkansas source specific requirements as described in the Final Action section above. The EPA has made, and will continue to make, these materials generally available through www.regulations.gov a (please contact the person identified in the FOR FURTHER INFORMATION CONTACT section of this preamble for more information). Therefore, these materials have been approved by EPA for inclusion in the SIP, have been incorporated by reference by EPA into that plan, are fully federally enforceable under sections 110 and 113 of the CAA as of the effective date of the final rulemaking of EPA's approval, and will be incorporated by reference in the next update to the SIP compilation.

VI. Statutory and Executive Order Reviews

Under the Clean Air Act, the Administrator is required to approve a SIP submission that complies with the provisions of the Act and applicable Federal regulations. 42 U.S.C. 7410(k); 40 CFR 52.02(a). Thus, in reviewing SIP submissions, the EPA's role is to approve state choices, provided that they meet the requirements of the Clean Air Act. Accordingly, this action merely approves state law as meeting Federal requirements and does not impose additional requirements beyond those imposed by state law. For that reason, this action:

• Is not a “significant regulatory action” subject to review by the Office of Management and Budget under Executive Orders 12866 (58 FR 51735, October 4, 1993) and 13563 (76 FR 3821, January 21, 2011);
• Does not impose an information collection burden under the provisions of the Paperwork Reduction Act (44 U.S.C. 3501 et seq.);
• Is certified as not having a significant economic impact on a substantial number of small entities under the Regulatory Flexibility Act (5 U.S.C. 601 et seq.);
• Does not contain any unfunded mandate or significantly or uniquely affect small governments, described in the Unfunded Mandates Reform Act of 1995 (Pub. L. 104-4);
• Does not have federalism implications as specified in Executive Order 13132 (64 FR 43255, August 10, 1999);
• Is not an economically significant regulatory action based on health or safety risks subject to Executive Order 13045 (62 FR 19885, April 23, 1997);
• Is not a significant regulatory action subject to Executive Order 13211 (66 FR 28355, May 22, 2001);
• Is not subject to requirements of section 12(d) of the National Technology Transfer and Advancement Act of 1995 (15 U.S.C. 272 note) because application of those requirements would be inconsistent with the CAA; and
• Does not provide EPA with the discretionary authority to address, as appropriate, disproportionate human health or environmental effects, using practicable and legally permissible methods, under Executive Order 12898 (59 FR 7629, February 16, 1994).

In addition, the SIP is not approved to apply on any Indian reservation land or in any other area where EPA or an Indian tribe has demonstrated that a tribe has jurisdiction. In those areas of Indian country, the rule does not have tribal implications and will not impose substantial direct costs on tribal governments or preempt tribal law as specified by Executive Order 13175 (65 FR 67249, November 9, 2000).

The Congressional Review Act, 5 U.S.C. 801 et seq., as added by the Small Business Regulatory Enforcement Fairness Act of 1996, generally provides that before a rule may take effect, the agency promulgating the rule must submit a rule report, which includes a copy of the rule, to each House of the Congress and to the Comptroller General of the United States. EPA will submit a report containing this action and other required information to the U.S. Senate, the U.S. House of Representatives, and the Comptroller General of the United States prior to publication of the rule in the Federal Register. A major rule cannot take effect until 60 days after it is published in the Federal Register. This action is not a “major rule” as defined by 5 U.S.C. 804(2).

Under section 307(b)(1) of the Clean Air Act, petitions for judicial review of this action must be filed in the United States Court of Appeals for the appropriate circuit by May 21, 2021. Filing a petition for reconsideration by the Administrator of this final rule does not affect the finality of this action for the purposes of judicial review nor does it extend the time within which a petition for judicial review may be filed, and shall not postpone the effectiveness of such rule or action. This action may not be challenged later in proceedings to enforce its requirements. (See section 307(b)(2).)

List of Subjects in 40 CFR Part 52

• Environmental protection
• Air pollution control
• Best available retrofit technology
• Incorporation by reference
• Intergovernmental relations
• Interstate transport of pollution
• Nitrogen dioxide
• Ozone
• Particulate matter
• Regional haze
• Reporting and recordkeeping requirements
• Sulfur oxides
• Visibility

Title 40, chapter I, of the Code of Federal Regulations is amended as follows:

PART 52—APPROVAL AND PROMULGATION OF IMPLEMENTATION PLANS

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

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

Subpart E—Arkansas

2. In § 52.170:

a. The table in paragraph (d), entitled “EPA-Approved Arkansas Source-Specific Requirements” is amended by adding an entry for “Domtar Ashdown Mill” at the end of the table.

b. In paragraph (e), the third table titled “EPA-Approved Non-Regulatory Provisions and Quasi-Regulatory Measures in the Arkansas SIP” is amended by adding an entry for “Arkansas Regional Haze Phase III SIP Revision” at the end of the table.

The additions read as follows:

3. In § 52.173, add paragraphs (h) and (i) to read as follows: