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

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A2021-05362

## Record

- **Collection:** Federal Register
- **Document type:** Rule
- **Published:** March 22, 2021
- **Citation:** 86 FR 15104

## Text

ENVIRONMENTAL PROTECTION AGENCY
40 CFR Part 52
[EPA-R06-OAR-2015-0189; FRL-10019-63-Region 6]
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 (SO
2
), 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 PM
10
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 (O
3
) NAAQS; the 2010 one-hour nitrogen dioxide (NO
2
) NAAQS; and the 2010 one-hour SO
2
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 SO
2
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
)
1

into the air. Fine particulates which cause haze are sulfates (SO
4
2-
), nitrates (NO
3
-
), organic carbon (OC), elemental carbon (EC), and soil dust.
2

PM
2.5
precursors consist of SO
2
, NO
X
, volatile organic compounds (VOCs), and in some cases, ammonia (NH
3
). 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.

1
Fine particles are less than or equal to 2.5 microns (µm) in diameter and usually form secondary in nature indirectly from other sources. Particles less than or equal to 10 µm in diameter are referred to as PM
10
. Particles greater than PM
2.5
but less than PM
10
are referred to as coarse mass. Coarse mass can contribute to light extinction as well and is made up of primary particles directly emitted into the air. Fine particles tend to be man-made, while coarse particles tend to have a natural origin. Coarse mass settles out from the air more rapidly than fine particles and usually will be found relatively close to emission sources. Fine particles can be transported long distances by wind and can be found in the air thousands of miles from where they were formed.

2
Organic carbon can be emitted directly as particles or formed through reactions involving gaseous emissions. Elemental carbon, in contrast to organic carbon, is exclusively of primary origin and emitted by the incomplete combustion of carbon-based fuels. Elemental carbon particles are especially prevalent in diesel exhaust and smoke from wild and prescribed fires.

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
3

in many mandatory Class I Federal areas
4

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.
5

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.
6

3
Visual range is the greatest distance, in km or miles, at which a dark object can be viewed against the sky by a typical observer.

4
Mandatory Class I Federal areas consist of national parks exceeding 6,000 acres, wilderness areas and national memorial parks exceeding 5,000 acres, and all international parks that were in existence on August 7, 1977. The EPA, in consultation with the Department of Interior, promulgated a list of 156 areas where visibility was identified as an important value. The extent of a mandatory Class I area includes subsequent changes in boundaries, such as park expansions. Although states and tribes may designate additional areas as Class I, the requirements of the visibility program set forth in the CAA applies only to mandatory Class I Federal areas. Each mandatory Class I Federal area is the responsibility of a Federal Land Manager (FLM). When the term “Class I area” is used in this action, it means “mandatory Class I Federal areas.” See 44 FR 69122 (November 30, 1979) and CAA Sections 162(a), 169A, and 302(i).

5
64 FR 35714, 35715 (July 1, 1999).

6
An interactive story map depicting efforts and recent progress by the EPA and states to improve visibility at national parks and wilderness areas may be visited at:
http://arcg.is/29tAbS3.

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.
7

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.
8

7
See the July 1, 1999 Regional Haze Rule final action (64 FR 35714), as amended on July 6, 2005 (70 FR 39156), October 13, 2006 (71 FR 60631), June 7, 2012 (77 FR 33656) and on January 10, 2017 (82 FR 3079).

8
See 40 CFR 51.308(b). Also, under 40 CFR 51.308(f)-(i), the EPA requires subsequent updates to the regional haze SIPs for each implementation period. The next update for the second implementation period is due by July 31, 2021.

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.
9

Under 40 CFR 51.308(e)(1)(ii), any BART-eligible source
10

that is reasonably anticipated to cause or contribute to visibility impairment in a Class I area is classified as subject-to-BART.
11

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.
12

States must establish emission limits, a schedule of compliance, and other measures consistent with the BART determination process for each source subject-to-BART.

9
See 42 U.S.C. 7491(g)(7), which lists the 26 source categories of major stationary sources potentially subject-to-BART.

10
BART-eligible sources are those sources that fall within one of 26 source categories that began operation on or after August 7, 1962, and were in existence on August 7, 1977, with potential emissions greater than 250 tons per year (tpy). (See 40 CFR 51 Appendix Y, section II).

11
Under the BART Guidelines, states may select a visibility impact threshold, measured in deciviews (dv), below which a BART-eligible source would not be expected to cause or contribute to visibility impairment in any Class I area. The State must document this threshold in the SIP and specify the basis for its selection of that value. Any source with visibility impacts that model above the threshold value would be subject to a BART determination review. The BART Guidelines acknowledge varying circumstances affecting different Class I areas. States should consider the number of emission sources affecting the Class I areas at issue and the magnitude of the individual sources' impacts. Any visibility impact threshold set by the state should not be higher than 0.5 dv. (See 40 CFR part 51, Appendix Y, section III.A.1).

12
The five statutory factors in determining BART controls are: (1) Costs of compliance, (2) the energy and non-air quality environmental impacts, (3) any existing control technology present at the source, (4) the remaining useful life of the source, and (5) the degree of improvement in visibility which may reasonably be anticipated to result from the use of such technology.

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.
13

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.

13
See 71 FR 60612, 60622 (October 13, 2006). Factors which can be used in a weight of evidence determination in this context may include, but not be limited to, future projected emissions levels under the alternative as compared to under BART; future projected visibility conditions under the two scenarios; the geographic distribution of sources likely to reduce or increase emissions under the alternative as compared to BART sources; monitoring data and emissions inventories; and sensitivity analyses of any models used.

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.
14

The RPGs established by the State provide an assessment of the visibility improvement anticipated to result for that planning period.
15

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.
16

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).
17

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.”

14
See 40 CFR 51.308(d)(3)(i) to (iv). For the first planning period, contributing and impacted states must develop coordinated emission management strategies. Impacted states must demonstrate that they have included all measures necessary in their SIPs to obtain their share of emission reductions needed to meet the RPGs for a Class I area. States must document the technical basis that they relied upon to determine the apportionment of emission reduction obligations necessary and identify the baseline emissions inventory on which their strategies are based. States must also identify all anthropogenic sources of visibility impairment considered in developing the strategy, such as major and minor stationary sources, mobile sources, and area sources.

15
The process for setting RPGs is as follows: (1) Identify sources that impact visibility; (2) evaluate potential controls based on consideration of the four reasonable progress factors; (3) project the visibility conditions based on implementation of on-the-books and additional selected controls; (4) compare the projected visibility conditions to the uniform rate of progress (URP) needed to attain natural visibility conditions by year 2064 for each Class I area; (5) determine an RPG for each Class I area based on this analysis that will improve the visibility at or beyond the URP on the most impaired days and ensure no degradation for the least impaired days. The Regional Haze Rule allows for the selection of an RPG at a given Class I area that provides for a slower rate of improvement than the URP for that area, but in that case a state must demonstrate that the URP is not reasonable and that the RPG selected is. See 40 CFR 51.308(d)(1)(ii).

16
These factors are: (1) Emission reductions due to ongoing air pollution control programs, including measures to address reasonably attributable visibility impairment (RAVI); (2) measures to mitigate the impacts of construction activities; (3) emissions limitations and schedules for compliance to achieve the reasonable progress goal; (4) source retirement and replacement schedules; (5) smoke management techniques for agricultural and forestry management purposes including plans as currently exist within the state for these purposes; (6) enforceability of emissions limitations and control measures; and (7) the anticipated net effect on visibility due to projected changes in point, area, and mobile source emissions over the period addressed by the long-term strategy.

17

Guidance for Setting Reasonable Progress Goals under the Regional Haze Program,
June 1, 2007, memorandum from William L. Wehrum, Acting Assistant Administrator for Air and Radiation, to the EPA Regional Administrators, EPA Regions 1-10 (pp. 4-2, 5-1).

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.
18

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.
19

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
, SO
2
, and PM
10
; 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.
20

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.
21

Among other things, the FIP established SO
2
, NO
X
, and PM
10
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 SO
2
and NO
X
emission limits under the reasonable progress requirements for the Entergy Independence Plant Units 1 and 2.

18
The September 9, 2008 SIP submittal included APCEC Regulation 19, Chapter 15, which is the state regulation that identified the BART-eligible and subject-to-BART sources in Arkansas and established BART emission limits for subject-to-BART sources. The August 3, 2010 SIP revision did not revise Arkansas' list of BART-eligible and subject-to-BART sources or revise any of the BART requirements for affected sources. Instead, it included mostly non-substantive revisions to the state regulation.

19
See the final action on (March 12, 2012) (77 FR 14604).

20
Under CAA section 110(c), the EPA is required to promulgate a FIP within two years of the effective date of a finding that a state has failed to make a required SIP submission or has made an incomplete submission, or of the effective date that the EPA disapproves a SIP in whole or in part. The FIP requirement is terminated only if a state submits a SIP, and the EPA approves that SIP as meeting applicable CAA requirements before promulgating a FIP.

21
See FIP final action on September 27, 2016 (81 FR 66332) as corrected on October 4, 2016 (81 FR 68319).

Following petitions for reconsideration and administrative stay submitted by the State, industry, and ratepayers, on April 14, 2017,
23

the EPA announced our decision to reconsider several elements of the FIP
24

and on April 25, 2017, the EPA issued a partial administrative stay of the effectiveness of the FIP for ninety days.
25

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 O
3
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.
26

23
Copies of the petitions for reconsideration and administrative stay submitted by the State of Arkansas; Entergy; Arkansas Electric Cooperative Corporation (AECC); and the Energy and Environmental Alliance of Arkansas (EEAA) are available in the docket of this action.

24
Letter from E. Scott Pruitt, Administrator, EPA, to Nicholas Jacob Bronni and Jamie Leigh Ewing, Arkansas Attorney General's Office (April 14, 2017). A copy of this letter is included in the docket,
https://www.regulations.gov/document?D=EPAR06-OAR-2015-0189-0240.

25
See 82 FR 18994.

26
See 82 FR 42627 (September 11, 2017) for the proposed approval. See also 83 FR 5915 and 83 FR 5927 (February 12, 2018) for the final action.

The State submitted its Phase II SIP revision (the Arkansas Regional Haze SO
2
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 SO
2
and PM
10
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 SO
2
and PM
10
BART requirements for Bailey Unit 1; SO
2
and PM
10
BART requirements for McClellan Unit 1; SO
2
BART requirements for Flint Creek Boiler No. 1; SO
2
BART requirements for White Bluff Units 1 and 2; SO
2
, NO
X
, and PM
10
BART requirements for the White Bluff Auxiliary Boiler;
27

and included a requirement that Lake Catherine Unit 4 not burn fuel oil until SO
2
and PM BART determinations for the fuel oil firing scenario are approved into the SIP by the EPA.
28

The submittal addressed the reasonable progress requirements with respect to SO
2
and PM
10
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 SO
2
and PM SIP revision and to withdraw the corresponding parts of the FIP.
29 30

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.

27
The Arkansas Regional Haze SO
2
and PM SIP revision established a new NO
X
emission limit of 32.2 pounds per hour (pph) for the Auxiliary Boiler to satisfy NO
X
BART and replaced the SIP determination that we previously approved in our final action on the Arkansas Regional Haze NO
X
SIP revision. In the Arkansas Regional Haze NO
X
SIP revision, DEQ incorrectly identified the Auxiliary Boiler as participating in the CSAPR trading program for O
3
season NO
X
to satisfy the NO
X
BART requirements. The new source-specific NO
X
BART emission limit that we approved in our final action on the Arkansas Regional Haze SO
2
and PM SIP revision corrected that error.

28
The 2012 action disapproved SO
2
, NO
X
, and PM BART for the fuel oil firing scenario for the Entergy Lake Catherine Plant Unit 4, but a FIP BART determination was not established. Instead, the FIP included a requirement that Entergy not burn fuel oil at Lake Catherine Unit 4 until final EPA approval of BART determinations for SO
2
, NO
X
, and PM. In the Arkansas Regional Haze NO
X
SIP revision, Arkansas relied on participation in CSAPR for O
3
season NO
X
to satisfy the NO
X
BART requirement for its subject-to-BART EGUs, including Lake Catherine Unit 4. When we took final action on the Arkansas Regional Haze NO
X
SIP revision, we also took final action to withdraw the FIP NO
X
emission limit for the natural gas firing scenario for Lake Catherine Unit 4. In the Arkansas Regional Haze SO
2
and PM SIP revision, Entergy committed to not burn fuel oil at Lake Catherine Unit 4 until final EPA approval of BART for SO
2
and PM. This commitment was made enforceable by the State through an Administrative Order that was adopted and incorporated in the Arkansas Regional Haze SO
2
and PM SIP revision.

29
See 83 FR 62204 (November 30, 2018) for proposed action and 84 FR 51033 (September 27, 2019) for final approval. The Arkansas Regional Haze SO
2
and PM SIP revision also addressed separate CAA requirements related to interstate visibility transport under CAA section 110(a)(2)(D)(i)(II), but we did not take action on that part of the submittal. We are acting on the interstate visibility transport portion of the Arkansas Regional Haze SO
2
and PM SIP revision in this final action.

30
See 84 FR 51056 (September 27, 2019) for the final withdrawal 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.
31

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)
32

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
33

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;
34

a combustion air system that includes over-fire air;
35

multi-clones for PM
10
removal;
36

and two venturi scrubbers in parallel for removal of SO
2
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.

31
Power Boiler No. 1 operates as natural gas only subject to the Gas 1 subcategory defined under 40 CFR 63.7575. See DEQ Air Permit No. 0287-AOP-R22 (page 64) in the docket of this action.

32
An electrostatic precipitator is an air pollution control device that functions by electrostatically charging particles in a gas stream that passes through collection plates with wires. The ionized particulate matter is attracted to and deposited on the plates as the cleaner air passes through. A wet electrostatic precipitator is designed to operate with water vapor saturated air streams to remove liquid droplets such as sulfuric acid.

33
See November 18, 2020 Disconnection Notice from Domtar for Power Boiler No. 1 (SN‐03) in the docket of this action.

34
A traveling grate is a moving grate used to feed fuel to the boiler for combustion.

35
Over-fire air typically recirculates a portion of the flue gas back to both the fuel-rich zone and the combustion zone to achieve complete burnout by encouraging the formation of nitrogen (N
2
) rather than NO
X
.

36
A cyclone separator is an air pollution control device shaped like a conical tube that creates an air vortex as air moves through it causing larger particles (PM
10
) to settle as the cleaner air passes through. Multi-clones are a sequence of cyclone separators in parallel used to treat a higher volume of air. In this particular case, the cleaner air travels to the venturi scrubbers to remove the smaller remaining particles like PM
2.5
and SO
2
.

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.
37

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 PM
10
BART determination for Power Boiler No. 1.
38

In that action, we also disapproved the SO
2
and NO
X
BART determinations for Power Boiler No. 1; and the SO
2
, NO
X
, and PM
10
BART determinations for Power Boiler No. 2. In the 2016 Arkansas Regional Haze FIP and its associated technical support document (TSD),
39

the EPA promulgated SO
2
, NO
X
, and PM
10
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),
40

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.
41

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.
42

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,
43

which included enforceable emission limitations and compliance schedules for the BART alternative.

37
See “Best Available Retrofit Technology Determination Domtar Industries Inc., Ashdown Mill (AFIN 41-00002),” originally dated October 31, 2006 and revised on March 26, 2007, prepared by Trinity Consultants Inc. This was included as part of the Phase III submittal and included in the docket of this action.

38
See the March 12, 2012 final action (77 FR 14604).

39
See final FIP action on September 27, 2016 (81 FR 66332) as corrected on October 4, 2016 (81 FR 68319) and the associated TSD, “AR020.0002-00 TSD for EPA's Proposed Action on the Arkansas Regional Haze FIP” in Docket No. EPA-R06-OAR-2015-0189 for the FIP BART analysis for SO
2
and NO
X
for Power Boiler No. 1; and SO
2
, NO
X
, and PM
10
for Power Boiler No. 2. This was included as part of the Phase III submittal and included in the docket of this action.

40
See “Supplemental BART Determination Information Domtar A.W. LLC, Ashdown Mill (AFIN 41-00002),” originally dated June 28, 2013 and revised on May 16, 2014, prepared by Trinity Consultants Inc. in conjunction with Domtar A.W. LLC. This was included as part of the Phase III SIP submittal and is included in the docket of this action.

41
See section III.B of the Arkansas Regional Haze Phase III submittal and the associated September 4, 2018, “Ashdown Mill BART Alternative TSD” in the docket of this action.

42
The proposed October 2018 SIP revision was intended to replace the portion of our FIP addressing Domtar and would also resolve the claims regarding Domtar in petitions for review of the FIP that are currently being held in abeyance,
State of Arkansas
v.
EPA,
No. 16-4270 (8th Cir.).

43
See DEQ Air permit #0287-AOP-R22 (effective August 1, 2019) included as part of the Phase III submittal and is included in the docket of this action.

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.
44

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),
45

the Arkansas Regional Haze SO
2
and PM SIP revision (Phase II SIP),
46

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.
47

44
See DEQ Air permit #0287-AOP-R22, Section VI, Plantwide Conditions #32 to #43. The “Regional Haze Program (BART Alternative) Specific Conditions” portion of the Plantwide Conditions section of the permit states the following: “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.”

45
See final action approved on February 12, 2018 (83 FR 5927).

46
See final action approved on September 27, 2019 (84 FR 51033) and the proposed approval on November 30, 2018 (83 FR 62204).

47
The Arkansas Regional Haze Phase III SIP submittal did not revise any aspects of the previous Phase I or II SIP revisions.

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.
48

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:

48
See the final rules promulgating the revised NAAQS: 71 FR 61144 (October 17, 2006); 77 FR 50033 (August 20, 2012); 80 FR 11573 (March 4, 2015); 80 FR 38419 (July 6, 2015); 78 FR 53269 (August 29, 2013); 73 FR 16436 (March 27, 2008). 81 FR 74504 (October 26, 2016); 75 FR 35520 (June 22, 2010); 75 FR 6474 (February 9, 2010); and 78 FR 3086 (January 15, 2013).

• 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.
49

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.

49
See “Guidance on Infrastructure State Implementation Plan (SIP) Elements under CAA sections 110(a)(1) and 110(a)(2)” by Stephen D. Page (Sept. 13, 2013), (pages 32-35).

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 O
3
NAAQS, the 2010 SO
2
NAAQS, and the 2010 NO
2
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.
50

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.
51

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 O
3
NAAQS Interstate Transport SIP revision to meet the requirements of CAA section 110(a)(2)(D) regarding interstate transport for the 2015 O
3
NAAQS. In that SIP submittal, Arkansas also addressed the 2006 and 2012 PM
2.5
NAAQS, the 2008 O
3
NAAQS, the 2010 SO
2
NAAQS, and the 2010 NO
2
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 SO
2
and PM SIP revision
52

for the 2006 and 2012 PM
2.5
NAAQS, the 2008 and 2015 O
3
NAAQS, the 2010 SO
2
NAAQS, and the 2010 NO
2
NAAQS. All other applicable infrastructure SIP requirements in the October 4, 2019, SIP submission have been or will be addressed in separate rulemakings.

50
The EPA approved the visibility transport requirement for the 2008 Pb NAAQS only in the February 2018 final action effective March 16, 2018 (see 83 FR 6470).

51
See 84 FR 51033, 51054 (September 27, 2019).

52
See 83 FR 62204 (November 30, 2018) for proposed approval and 84 FR 51033 (September 27, 2019) for final action. The Arkansas Regional Haze SO
2
and PM SIP revision addressed separate CAA requirements related to interstate visibility transport under CAA section 110(a)(2)(D)(i)(II), but we did not take action on that part of the submittal. We are acting on the prong 4 portion of the Arkansas Regional Haze SO
2
and PM SIP revision in this final action.

II. Summary of Proposed Action and Our Final Decisions

On March 16, 2020, we published a Notice of Proposed Rulemaking (NPRM)
53

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 SO
2
, 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 PM
10
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
.

53
See March 16, 2020 proposed approval (85 FR 14847).

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 O
3
NAAQS; the 2010 one-hour NO
2
NAAQS; and the 2010 one-hour SO
2
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,
54

the Arkansas Regional Haze SO
2
and PM SIP revision,
55

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.

54
Final action approved on February 12, 2018 (83 FR 5927).

55
See 83 FR 62204 (November 30, 2018) for proposed approval and 84 FR 51033 (September 27, 2019) for final approval.

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 (SO
2
) emissions and decrease in NO
X
emissions. While DEQ claims that the NO
X
decrease mitigates the SO
2
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 SO
2
, NO
X
, and particulate matter (PM
10
) 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 PM
10
than the BART controls in the FIP. The BART alternative controls would reduce NO
X
and PM
10
emissions by 1,096 and 111 tons per year (tpy), respectively, from the baseline. The BART alternative results in a smaller reduction in SO
2
emissions compared to the BART controls (BART achieves 3,051 tpy SO
2
reduction) but still achieves a decrease of 1,637 tpy SO
2
from the baseline. Despite a smaller reduction in SO
2
emissions than BART (a 1,414 tpy SO
2
difference), the BART alternative results in 300 tpy fewer NO
X
emissions and 157 tpy fewer PM
10
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 SO
2
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
56

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.

56
Method 1 assessed visibility impairment on a per source per pollutant basis and Method 2 allowed for interaction of the pollutants from both boilers. See descriptions of method 1 and 2 modeling evaluations in the March 16, 2020 proposed approval (85 FR 14847, 14857-14858).

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 (NO
3
−) and nitrogen dioxide (NO
2
) impacts had the highest contribution to visibility impairment at all Class I areas. For Power Boiler No. 2, baseline modeled NO
3
− and NO
2
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 SO
4
2
−
species contributions (from SO
2
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.
57

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.
58

57
The “ten highest impacted days” means the 8th to 17th highest days at each Class I area. The 98th percentile means that for a given distribution, it is equal to or higher than 98 percent of the rest of the distribution. The 98th percentile impact day means that only two percent of the 365 days in a calendar year, or 7.3 days (rounded up to 8 days) have higher impacts. The simplified chemistry in the CALPUFF model tends to magnify the actual visibility effects of that source so it is appropriate to use the 98th percentile, or 8th highest day, to 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. See 70 FR 39104, 39121 (July 6, 2005), Regional Haze Regulations and Guidelines for BART Determinations.

58
See discussion regarding the different metrics in the March 16, 2020 proposed approval (85 FR 14847, 14859-14860).

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 SO
2
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 SO
2
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 PM
10
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
59

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
60 61

for Domtar.

59

See
40 CFR 51 Appendix Y section III.A.3 and IV.D.5, “Guidelines for BART Determinations Under the Regional Haze Rule.”

60
See proposed FIP on April 8, 2015 (80 FR 18979).

61
See final FIP action on September 27, 2016 (81 FR 66332) as corrected on October 4, 2016 (81 FR 68319) and the associated TSD, “AR020.0002-00 TSD for EPA's Proposed Action on the Arkansas Regional Haze FIP” in Docket No. EPA-R06-OAR-2015-0189 for the FIP BART analysis for SO
2
and NO
X
for Power Boiler No. 1; and SO
2
, NO
X
, and PM
10
for Power Boiler No. 2. This was included as part of the Phase III submittal and included in the docket of this action.

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.
62

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.

62
See 71 FR 60612, 60622 (October 13, 2006). Factors which can be used in a weight of evidence determination in this context may include, but not be limited to, future projected emissions levels under the alternative as compared to under BART; future projected visibility conditions under the two scenarios; the geographic distribution of sources likely to reduce or increase emissions under the alternative as compared to BART sources; monitoring data and emissions inventories; and sensitivity analyses of any models used.

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),
63

the analysis of emission reductions,
64

and the analysis of Domtar's visibility impacts due to NO
3
− compared to SO
4
−.
65

In addition, we also considered our analysis of the ten highest impacted days (8th to 17th highest)
66

and our analysis of the number of days impacted over 0.5 dv and 1.0 dv.
67

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).

63
See Tables 7 and 8 of the proposed approval, 85 FR 14847, 14858.

64
See Tables 5 and 6 of the proposed approval, 85 FR 14847, 14856-14857.

65
See Appendix C “Supplemental BART Determination Information Domtar A.W. LLC, Ashdown Mill (AFIN 41-00002),” originally dated June 28, 2013 and revised on May 16, 2014, prepared by Trinity Consultants Inc. in conjunction with Domtar A.W. LLC.

66
See 85 FR 14847, 14859. This data is based on the CALPUFF modeling provided by Domtar and relied on by the State in the Phase III SIP. See “EPA-CALPUFF summary for Method 2.xlsx” for the EPA's summary of the modeling data, available in the docket for this action.

67
See 85 FR 14847, 14860. This data is based on the CALPUFF modeling provided by Domtar and relied on by the State in the Phase III SIP revision. See “EPA-CALPUFF summary for Method 2.xlsx” for the EPA's summary of the modeling data, available in the docket for this action.

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,
68

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.

68
85 FR 14847, 14857.

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.
69

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).

69
See 71 FR 60612, 60622 (October 13, 2006).

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).
70

70

See
71 FR 60612, 60616.

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).
71

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.

71
See 71 FR 60616.

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.
72

See also
70 FR 60616 (recognizing CALPUFF as particularly appropriate for single-source applications).

72
See Arizona's September 19, 2014 proposed approval (79 FR 56322) which was finalized on April 10, 2015 (80 FR 19220).

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 SO
2
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 SO
2
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.
73

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.
74

The court in
NPCA
rejected that this flaw in the State's analysis could be dismissed as harmless error.
75

73

Id.
at 164.

74

Id.
at 165.

75

Id.
at 167.

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.
76

DEQ notes in the letter that DEQ has received a disconnection notice
77

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.

76
See December 3, 2020 clarification letter to EPA from DEQ posted in the docket of this action.

77
See November 18, 2020 Disconnection Notice from Domtar for Power Boiler No. 1 (SN-03) in the docket of this action.

For Power Boiler No. 2, which currently relies on a continuous emissions monitoring system (CEMS) to monitor SO
2
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 SO
2
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 SO
2
, 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 PM
10
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 PM
10
compliance demonstrations in plantwide condition 41. Condition 41 clearly explains that the applicable PM
10
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 SO
2
and NO
X
emissions and in 40 CFR part 63, subpart DDDDD for PM
10
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 SO
2
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 SO
2
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 SO
2
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,
78

which is already part of the approved SIP, and applies to this source.
79

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 SO
2
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.

78
Under APCEC Rule 19.703—Continuous Emission Monitoring, any stationary source subject to this regulation shall, as required by federal law and upon request of the Department: (A) Install, calibrate, operate, and maintain equipment to continuously monitor or determine federally regulated air pollutant emissions in accordance with applicable performance specifications in 40 CFR part 60 Appendix B as of the effective date of the federal final rule published by EPA in the
Federal Register
on February 27, 2014 (79 FR 11271), and quality assurance procedures in 40 CFR part 60 Appendix F as of the effective date of the federal final rule published by EPA in the
Federal Register
on February 27, 2014 (79 FR 11274), and other methods and conditions that the Department, with the concurrence of the EPA, shall prescribe. Any source listed in a category in 40 CFR part 51 Appendix P as of the effective date of the federal final rule published by EPA in the
Federal Register
on November 7, 1986 (51 FR 40675), or in 40 CFR part 60 as of August 30, 1992, shall adhere to all continuous emissions monitoring or alternative continuous emission monitoring requirements stated therein, if applicable. (B) Report the data collected by the monitoring equipment to the Department at such intervals and on such forms as the Department shall prescribe, in accordance with 40 CFR part 51, Appendix P, Section 4.0 (Minimum Data Requirements) as of the effective date of the federal final rule published by EPA in the
Federal Register
on November 7, 1986 (51 FR 40675), and any other applicable reporting requirements promulgated by the EPA.

79
See 52.170(c) (table) for EPA-approved regulations in the Arkansas SIP.

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,
80

if Power Boiler No. 2 switches to natural gas combustion, the applicable natural gas AP-42 emission factors of 0.6 lb SO
2
/MMscf, 280 lb NO
X
/MMscf, and 7.6 lb PM
10
/MMscf in conjunction with natural gas fuel usage records shall be used to demonstrate compliance with the BART emission limits.
81

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,
82 83

that burning only natural gas would very likely be sufficient in itself to demonstrate that the boiler is complying with the SO
2
emission limit. SO
2
emissions from combustion of natural gas are inherently very low and are virtually eliminated during the combustion process. Any SO
2
emissions will be in trace amounts well below the BART alternative emission limit so there should be no concern that the alternative limit for SO
2
will be met. NO
X
and PM
10
emissions are also expected to be lower than the BART alternative emission limit for natural gas combustion.
84 85

Using the most conservative NO
X
, SO
2
, and PM
10
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.

80
See 85 FR 14847, 14862.

81
See AP 42, Fifth Edition Compilation of Air Pollutant Emissions Factors, Volume 1: Stationary Point and Area Sources, section 1.4, Tables 1.4-1 and 2 pertaining to natural gas combustion.

82
See 40 CFR 52.173(c)(8)(iv) and (v). However, the FIP regulations required burning only pipeline quality natural gas, and no such requirement to burn only pipeline quality natural gas can be located in the permit or the SIP for this unit. Nonetheless, there is no indication (nor has the commenter supplied any such information) that burning other types of natural gas would result in SO
2
emissions that would even approach the BART alternative emission limit.

83
Table 1.4-2 from
Fifth Edition Compilation of Air Pollutant Emissions Factors, Volume 1: Stationary Point and Area Sources, section 1.4
indicates that the AP-42 factor contemplates varying amounts of sulfur and the potential need to adjust the emission factor. The AP-42 factor for sulfur from natural gas (0.6 lb/10
6
scf) is based on 100% conversion of fuel sulfur to SO
2.
It assumes a sulfur content for natural gas of 2,000 grains/10
6
scf. The SO
2
emission factor in this table can be converted to other natural gas sulfur contents by multiplying the SO
2
emission factor by the ratio of the site-specific sulfur content (grains/10
6
scf) to 2,000 grains/10
6
scf. To convert the emission factors in the AP-42 tables on a volume basis (lb/10
6
scf) to an energy basis (lb/MMBtu) divide by a heating value of 1,020 MMBtu/10
6
scf. Then, multiply the result by the heat input capacity of the boiler (MMBtu/hr) to get a mass flow rate (lb/hr). Accordingly, an AP factor of 0.6 lb SO
2
/MMscf multiplied by Power Boiler No. 2 maximum heat input of 820 MMBtu/hr would result in 0.5 lb/hr SO
2
, showing that the sulfur emissions would be very low and almost negligible. It is also more conservative than the FIP (“pipeline quality natural gas” would result in 1.2 lb/hr SO
2
assuming pipeline natural gas contains 0.5 grains or less of total sulfur per 100 standard cubic feet). These results are well below the BART alternative limit for SO
2
of 435 lb/hr.

84
From Table 1.4-1 of
Fifth Edition Compilation of Air Pollutant Emissions Factors, Volume 1: Stationary Point and Area Sources, section 1.4
we can also appropriately select the most conservative NO
X
emission factor based on the design heat input capacity for Power Boiler No. 2 of 820 MMBtu/hr. From this, we can choose emission factors from the combustor type. The applicable AP-42 emission factor (280 lb NO
X
/MMscf) is consistent with what was used in the FIP for a large wall-fired boiler > 100 MMBtu/hr. This is the highest emission factor in the table for NO
X
and results in 225 lb/hr NO
X
(985 tpy NO
X
) which can be calculated from the heat input capacity of the boiler (820 MMBtu/hr) similarly as explained in previous footnote. The result is less than both the FIP NO
X
limit of 345 lb/hr (1,511 tpy) and the BART alternative NO
X
rate of 293 lb/hr (1,283 tpy).

85
From Table 1.4-2 of
Fifth Edition Compilation of Air Pollutant Emissions Factors, Volume 1: Stationary Point and Area Sources, section 1.4
an AP factor of 7.6 lb PM
10
/MMscf represents total PM and equates to 6.1 lb/hr PM applying a heat input capacity of 820 MMBtu/hr. This is less than the BART alternative rate of 81.6 lb/hr PM.

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.
86

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);
87

continuous emission monitoring (APCEC Rule 19.703); recordkeeping and reporting requirements (APCEC Rule 19.705);
88

and Public Availability of Emissions Data (APCEC Rule 19.706).
89

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.

86
See 40 CFR 52.170(c) (table) for EPA-approved regulations in the Arkansas SIP.

87
Under APCEC Rule 19.702—Air Emissions Sampling, any stationary source subject to this regulation shall be subject to the following requirements: (A) Sampling Ports To provide any sampling ports, at the request of the Department, required for federally regulated air pollutant emissions sampling, including safe and easy access to such ports. (B) Sampling To conduct federally regulated air pollutant emissions sampling, at the request of the Department, to determine the rate, opacity, composition, and/or contaminant concentration of the emissions. All compliance testing shall be done at the expense of the permittee by an independent firm, unless otherwise approved by the Department. Sampling shall not be required for those pollutants with continuous emissions monitors. (C) Averaging Times All compliance testing averaging times shall be consistent with the averaging times of the applicable federally regulated air pollutant emissions limitations stated in the applicable permit, which in no case shall be greater than the minimum averaging times of the applicable NAAQS. (D) Process Rates Unless otherwise approved by the Department, all federally regulated air pollutant emissions sampling shall be performed with the equipment being tested operating at least at ninety percent of its permitted capacity. Emissions results shall be extrapolated to correlate with 100 percent of permitted capacity to determine compliance.

88
Under APCEC Rule 19.705—Record Keeping and Reporting Requirements, any stationary source subject to this regulation shall, upon request by the Department: (A) Maintain records on the nature and amounts of federally regulated air pollutants emitted to the air by the equipment in question. All records, including compliance status reports and excess emissions measurements shall be retained for at least five (5) years, and shall be made available to any agent of the Department or EPA during regular business hours. (B) Supply the following information, correlated in units of the applicable emissions limitations, to the Department: (1) General process information related to the emissions of federally regulated air pollutants into the air. (2) Emissions data obtained through sampling or continuous emissions monitoring. (C) Information and data shall be submitted to the Department by a responsible official on such forms and at such time intervals as prescribed by applicable federal regulations or the Department. Reporting periods shall be a twelve-month period. (D) Each emission inventory is to be accompanied by a certifying statement, signed by the owner(s) or operator(s) and attesting that the information contained in the inventory is true and accurate to the best knowledge of the certifying official. The certification shall include the full name, title, signature, date of signature, and telephone number of the certifying official.

89
Emissions data obtained by the Department shall be correlated in units of applicable emissions limitations and be made available to the public at the Department's central offices during normal business hours.

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,
90

and the BART guidelines

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A2021-05362. Public record. Not legal advice.