Skip to Content

We invite you to try out our new beta eCFR site at https://ecfr.federalregister.gov. We’ve made big changes to make the eCFR easier to use. Be sure to leave feedback using the 'Feedback' button on the bottom right of each page!

Rule

Revised Cross-State Air Pollution Rule Update for the 2008 Ozone NAAQS

Document Details

Information about this document as published in the Federal Register.

Document Statistics
Document page views are updated periodically throughout the day and are cumulative counts for this document. Counts are subject to sampling, reprocessing and revision (up or down) throughout the day.
Enhanced Content

Relevant information about this document from Regulations.gov provides additional context. This information is not part of the official Federal Register document.

Published Document

This document has been published in the Federal Register. Use the PDF linked in the document sidebar for the official electronic format.

Start Preamble Start Printed Page 23054

AGENCY:

Environmental Protection Agency (EPA).

ACTION:

Final rule.

SUMMARY:

The U.S. Environmental Protection Agency (EPA) is taking this action to address interstate transport of ozone pollution under the “good neighbor provision” of the Clean Air Act (CAA). This final action is taken in response to the United States Court of Appeals for the District of Columbia Circuit's (D.C. Circuit) remand of the Cross-State Air Pollution Rule (CSAPR) Update in Wisconsin v. EPA on September 13, 2019. The CSAPR Update finalized Federal Implementation Plans (FIPs) for 22 states to address their good neighbor obligations for the 2008 ozone National Ambient Air Quality Standards (NAAQS). The D.C. Circuit found that the CSAPR Update, which was published on October 26, 2016 as a partial remedy to address upwind states' obligations prior to the 2018 Moderate area attainment date under the 2008 ozone NAAQS, was unlawful to the extent it allowed those states to continue their significant contributions to downwind ozone problems beyond the statutory dates by which downwind states must demonstrate their attainment of the air quality standards. On the same grounds, the D.C. Circuit also vacated the CSAPR Close-Out in New York v. EPA on October 1, 2019. This final rule resolves 21 states' outstanding interstate ozone transport obligations with respect to the 2008 ozone NAAQS.

This action finds that for 9 of the 21 states for which the CSAPR Update was found to be only a partial remedy (Alabama, Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Texas, and Wisconsin), their projected ozone precursor emissions in the 2021 ozone season and thereafter do not significantly contribute to a continuing downwind nonattainment and/or maintenance problem, and therefore the states' CSAPR Update FIPs (or the SIPs subsequently approved to replace certain states' CSAPR Update FIPs) fully address their interstate ozone transport obligations with respect to the 2008 ozone NAAQS. This action also finds that for the 12 remaining states (Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West Virginia), their projected 2021 ozone season nitrogen oxides (NOX) emissions significantly contribute to downwind states' nonattainment and/or maintenance problems for the 2008 ozone NAAQS. In this final action, EPA is issuing new or amended FIPs for these 12 states to replace their existing CSAPR NOX Ozone Season Group 2 emissions budgets for electricity generating units (EGUs) with revised budgets via a new CSAPR NOX Ozone Season Group 3 Trading Program. EPA is requiring implementation of the revised emission budgets beginning with the 2021 ozone season. Based on EPA's assessment of remaining air quality issues and additional emission control strategies for EGUs and other emissions sources in other industry sectors (non-EGUs), EPA is further determining that these NOX emission reductions fully eliminate these states' significant contributions to downwind air quality problems for the 2008 ozone NAAQS. In this action, EPA is also finalizing an error correction of its June 2018 approval of Kentucky's good neighbor SIP.

DATES:

This final rule is effective on June 29, 2021.

ADDRESSES:

EPA has established a docket for this action under Docket ID No. EPA-HQ-OAR-2020-0272. All documents in the docket are listed on the www.regulations.gov website. 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 and will be publicly available only in hard copy form. Publicly available docket materials are available electronically through www.regulations.gov.

Start Further Info

FOR FURTHER INFORMATION CONTACT:

Mr. Daniel Hooper, Clean Air Markets Division, Office of Atmospheric Programs (Mail Code 6204M), Environmental Protection Agency, 1200 Pennsylvania Avenue NW, Washington, DC 20460; telephone number: (202) 343-9167; email address: Hooper.Daniel@epa.gov.

End Further Info End Preamble Start Supplemental Information

SUPPLEMENTARY INFORMATION:

Preamble Glossary of Terms and Abbreviations

The following are abbreviations of terms used in the preamble.

4-step good neighbor framework 4-step framework

AEO Annual Energy Outlook

AQAT Air Quality Assessment Tool

AQM TSD Air Quality Modeling Technical Support Document

CAA or Act Clean Air Act

CAIR Clean Air Interstate Rule

CAMx Comprehensive Air Quality Model with Extensions

CBI Confidential Business Information

CEMS Continuous Emission Monitoring System(s)

CFR Code of Federal Regulations

CMDb Control Measures Database

CMV Commercial Marine Vehicle

CoST Control Strategy Tool

CRA Congressional Review Act

CSAPR Cross-State Air Pollution Rule

EGU Electric Generating Unit

EISA Energy Independence and Security Act

EPA U.S. Environmental Protection Agency

FIP Federal Implementation Plan

FR Federal Register

HDGHG Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles

IC Internal Combustion

ICI Industrial, Commercial, and Institutional

ICR Information Collection Request

IPM Integrated Planning Model

iSIP Infrastructure State Implementation Plan

km Kilometer

lb/mmBtu Pounds per Million British Thermal Units

LEC Low Emission Combustion

LNB Low-NOX Burners

MJO Multi-Jurisdictional Organizations

mmBtu Million British Thermal Units

MOVES Motor Vehicle Emission Simulator

MSAT2 Mobile Source Air Toxic Rule

NAAQS National Ambient Air Quality Standard

NEI National Emission Inventory

NESHAP National Emission Standards for Hazardous Air Pollutants

NOX Nitrogen Oxides

NODA Notice of Data Availability

Non-EGU Non-electric Generating Unit

NSPS New Source Performance Standard

NUSA New Unit Set-Aside

OSAT/APCA Ozone Source Apportionment Technology/Anthropogenic Precursor Culpability Analysis

OMB Office of Management and Budget

OTC Ozone Transport Commission

OTR Ozone Transport Region

PEMS Predictive Emissions Monitoring System

PM2.5 Fine Particulate Matter

ppb Parts Per Billion

RACT Reasonably Available Control Technology

RIA Regulatory Impact Analysis

RICE Reciprocating Internal Combustion Engines

RRF Relative Response Factor

RTC Document Response to Comment Document

SCR Selective Catalytic Reduction

SIP State Implementation Plan

SMOKE Sparse Matrix Operator Kernel Emissions

SNCR Selective Non-catalytic ReductionStart Printed Page 23055

SO2 Sulfur Dioxide

TIP Tribal Implementation Plan

TSD Technical Support Document

tpy Ton Per Year

ULNB Ultra-low NOX Burner

VOC Volatile Organic Compound

WRF Weather Research and Forecasting Model

Table of Contents

I. Executive Summary

A. Purpose of Regulatory Action

B. Summary of the Major Provisions of the Regulatory Action

C. Costs and Benefits

II. General Information

A. Does this action apply to me?

III. EPA's Legal Authority for the Final Rule

A. Statutory Authority

B. Prior Good Neighbor Rulemakings Addressing Regional Ozone

IV. Air Quality Issues Addressed and Overall Approach for the Final Rule

A. The Interstate Ozone Transport Challenge

1. Nature of Ozone and the Ozone NAAQS

2. Ozone Transport

3. Health and Environmental Effects

B. Relationship Between This Regulatory Action and the 2015 Ozone NAAQS

C. Approach To Address the Remanded Transport Obligations for the 2008 Ozone NAAQS

1. Events Affecting Application of the Good Neighbor Provision for the 2008 Ozone NAAQS

2. FIP Authority for Each State Covered by the Final Rule

3. The 4-Step Good Neighbor Framework

V. Analyzing Downwind Air Quality and Upwind-State Contributions

A. Overview of Air Quality Modeling Platform

B. Emission Inventories

1. Foundation Emission Inventory Data Sets

2. Development of Emission Inventories for EGUs

3. Development of Emission Inventories for Non-EGU Point Sources

4. Development of Emission Inventories for Onroad Mobile Sources

5. Development of Emission Inventories for Commercial Marine Vessels

6. Development of Emission Inventories for Other Nonroad Mobile Sources

7. Development of Emission Inventories for Nonpoint Sources

C. Air Quality Modeling To Identify Nonattainment and Maintenance Receptors

D. Pollutant Transport From Upwind States

1. Air Quality Modeling To Quantify Upwind State Contributions

2. Application of Screening Threshold

VI. Quantifying Upwind-State NOX Reduction Potential To Reduce Interstate Ozone Transport for the 2008 NAAQS

A. The Multi-Factor Test

B. Identifying Levels of Control Stringency

1. EGU NOX Mitigation Strategies

2. Non-EGU NOX Mitigation Strategies

3. Mobile Source NOX Mitigation Strategies

C. Emission Reduction Potential of Control Stringencies

1. EGU Emission Reduction Potential

2. Non-EGU Emission Reduction Potential

D. Assessing Cost, EGU and Non-EGU NOX Reductions, and Air Quality

1. EGU Assessment

2. Non-EGU Assessment

3. Overcontrol Analysis

VII. Implementation of Emission Reductions

A. Regulatory Requirements for EGUs

B. Quantifying State Emissions Budgets

C. Elements of New Trading Program

1. Applicability

2. State Budgets, Variability Limits, Assurance Levels, and Penalties

3. Unit-Level Allocations of Emission Allowances

4. Transitioning From Existing CSAPR NOX Ozone Season Group 2 Trading Program

5. Compliance Deadlines

6. Monitoring and Reporting

7. Recordation of Allowances

8. Conforming Revisions to Regulations for Existing Trading Programs

D. Submitting a SIP

1. SIP Option To Modify 2022 Allocations

2. SIP Option To Modify Allocations in 2023 and Beyond

3. SIP Revisions That Do Not Use the New Group 3 Trading Program

4. No SIP Option for Additional States To Participate in the New Trading Program

E. Title V Permitting

F. Relationship to Other Emission Trading and Ozone Transport Programs

1. Existing Trading Programs

2. Title IV Interactions

3. NOX SIP Call Interactions

VIII. Costs, Benefits, and Other Impacts of the Final Rule

IX. Summary of Changes to the Regulatory Text for the Federal Implementation Plans and Trading Programs

A. Amended CSAPR Update FIP Provisions

B. New CSAPR NOX Ozone Season Group 3 Trading Program Provisions

C. Transitional Provisions

D. Conforming Revisions, Corrections, and Clarifications to Existing Regulations

X. Statutory and Executive Order Reviews

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

B. Paperwork Reduction Act (PRA)

C. Regulatory Flexibility Act (RFA)

D. Unfunded Mandates Reform Act (UMRA)

E. Executive Order 13132: Federalism

F. Executive Order 13175: Consultation and Coordination With Indian Tribal Governments

G. Executive Order 13045: Protection of Children From Environmental Health Risks and Safety Risks

H. Executive Order 13211: Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution or Use

I. National Technology Transfer and Advancement Act (NTTAA)

J. Executive Order 12898: Federal Actions To Address Environmental Justice in Minority Populations and Low-Income Populations

K. Congressional Review Act

L. Determinations Under CAA Section 307(b)(1) and (d)

I. Executive Summary

This final rule resolves the interstate transport obligations of 21 states under the good neighbor provision of the Clean Air Act (CAA or the Act), CAA section 110(a)(2)(D)(i)(I), for the 2008 ozone National Ambient Air Quality Standards (NAAQS). The 2008 ozone NAAQS is an 8-hour standard that was set at 75 parts per billion (ppb).[1] The U.S. Environmental Protection Agency (EPA or the Agency) published the Cross-State Air Pollution Rule (CSAPR) Update on October 26, 2016, which, among other things, partially addressed the interstate transport of emissions from 21 states with respect to the 2008 ozone NAAQS.[2] See 81 FR 74504. On December 21, 2018, EPA published the CSAPR Close-Out Rule, which found that the CSAPR Update was a complete remedy for 20 of those states based on air quality analysis of the year 2023.[3]

On September 13, 2019, the United States Court of Appeals for the District of Columbia Circuit (D.C. Circuit) remanded the CSAPR Update, concluding that it was invalid in one respect because it unlawfully allowed upwind states to continue their significant contributions to downwind air quality problems beyond the statutory dates by which downwind States must demonstrate their attainment of ozone air quality standards. Wisconsin v. EPA, 938 F.3d 303, 318-20 (D.C. Cir. 2019) (Wisconsin) (per curiam); see also id. 336-37 (concluding that remand without vacatur was appropriate). Subsequently, on October 1, 2019, in a judgment order, the D.C. Circuit vacated the CSAPR Close-Out on the same grounds on which it had remanded without vacatur the CSAPR Update in Wisconsin. New York v. EPA, 781 Fed. App'x 4, 7 (D.C. Cir. 2019) (New York). The court found the CSAPR Close-Out inconsistent with the Wisconsin holding because the rule analyzed the year 2023 rather than 2021 and failed to demonstrate that it was an impossibility to address significant contribution by the 2021 Serious area attainment date (“the next applicable attainment date”). To address the Wisconsin and New York decisions, EPA proposed this rule in the Federal Start Printed Page 23056Register on October 30, 2020 to revise the CSAPR Update (85 FR 68964).[4]

In this final rule, in accordance with Wisconsin and New York, EPA has aligned its analysis and the implementation of emission reductions required to address significant contribution with the 2021 ozone season, which corresponds to the July 20, 2021 Serious area attainment date for the 2008 ozone NAAQS. EPA has further determined which emission reductions are impossible to achieve by the 2021 attainment date and whether any such additional emission reductions should be required beyond that date. See Wisconsin, 938 F.3d at 320; New York, 781 Fed. App'x at 7.

In this action on remand, EPA is not reopening any determinations, findings, or statutory or regulatory interpretations that are not required to address the Wisconsin remand, unless the Agency has explicitly so stated. This final action addressing the remand of the CSAPR Update in Wisconsin also has the effect of addressing the outstanding obligations that resulted from the D.C. Circuit's vacatur of the CSAPR Close-Out in New York. See New York, 781 Fed. App'x at 7.

A. Purpose of the Regulatory Action

The purpose of this rulemaking is to protect public health and welfare by eliminating emissions in certain upwind states that significantly contribute to nonattainment, or interfere with maintenance, of the 2008 ozone NAAQS in the U.S. Ground-level ozone causes a variety of negative effects on human health, vegetation, and ecosystems. In humans, acute and chronic exposure to ozone is associated with premature mortality and a number of morbidity effects, such as asthma exacerbation. Ozone exposure can also negatively impact ecosystems, for example, by limiting tree growth. Studies have established that ozone transport occurs on a regional scale (i.e., hundreds of miles) over much of the eastern U.S., with elevated concentrations occurring in rural as well as metropolitan areas.[5 6] As discussed in more detail in section IV.A.1, assessments of ozone control approaches have concluded that nitrogen oxides (NOX) control strategies are effective to reduce regional-scale ozone transport.[7]

Clean Air Act section 110(a)(2)(D)(i)(I), which is also known as the “good neighbor provision,” requires states to prohibit emissions that will contribute significantly to nonattainment or interfere with maintenance in any other state with respect to any primary or secondary NAAQS.[8] The statute vests states with the primary responsibility to address this “interstate transport” of air pollutants through the development of good neighbor State Implementation Plans (SIPs), which are one component of larger SIP submittals typically required three years after EPA promulgates a new or revised NAAQS. These larger SIPs are often referred to as “infrastructure” SIPs or iSIPs. See CAA section 110(a)(1) and (2). EPA supports state efforts to submit good neighbor SIPs for the 2008 ozone NAAQS and has shared information with states to facilitate such SIP submittals. However, the CAA also requires EPA to fill a backstop role by issuing Federal Implementation Plans (FIPs) where states fail to submit good neighbor SIPs or EPA disapproves a submitted good neighbor SIP. See generally CAA section 110(k) and 110(c).

On October 26, 2016, EPA published the CSAPR Update, which finalized FIPs for 22 states that EPA found failed to submit a complete good neighbor SIP (15 states) [9] or for which EPA issued a final rule disapproving their good neighbor SIP (7 states).[10] The FIPs promulgated for these states included new NOX ozone season emission budgets for electric generating units (EGUs) to reduce interstate transport for the 2008 ozone NAAQS. These emission budgets took effect in 2017 in order to assist downwind states with attainment of the 2008 ozone NAAQS by the 2018 Moderate area attainment date. EPA acknowledged at the time that the FIPs promulgated for 21 of the 22 states only partially addressed good neighbor obligations under the 2008 ozone NAAQS. The 22 states for which EPA promulgated FIPs to reduce interstate ozone transport as to the 2008 ozone NAAQS are listed in Table I.A-1.

Table I.A-1—List of 22 Covered States for the 2008 8-Hour Ozone NAAQS in the CSAPR Update

State
        Alabama
        Arkansas
        Illinois
        Indiana
        Iowa
        Kansas
        Kentucky
        Louisiana
        Maryland
        Michigan
        Mississippi
        Missouri
        New Jersey
        New York
        Ohio
        Oklahoma
        Pennsylvania
        Tennessee
        Texas
        Virginia
        West Virginia
        Wisconsin

In response to the D.C. Circuit's remand of the CSAPR Update in Wisconsin and the court's vacatur of the CSAPR Close-Out in New York, this rule finds that 12 of the 22 states listed in Table I.A-1 require further ozone season NOX emission reductions to address the good neighbor provision as to the 2008 ozone NAAQS. As such, EPA is promulgating new or revised FIPs for these states that include new EGU NOX ozone season emission budgets, with implementation of these emission budgets beginning with the 2021 ozone season.[11] The 12 states for which EPA is promulgating new or revised FIPs to reduce interstate ozone transport as to the 2008 ozone NAAQS in this rulemaking are listed in Table I.A-2.

Table I.A-2—List of 12 Covered States for the 2008 8-Hour Ozone NAAQS

State
        Illinois
        Indiana
        Kentucky
        Louisiana
        Maryland
        Michigan
Start Printed Page 23057
        New Jersey
        New York
        Ohio
        Pennsylvania
        Virginia
        West Virginia

The enhanced control stringency represented by the new EGU NOx ozone season emission budgets for these states will take effect 60 days after publication in the Federal Register, which corresponds to the effective date of the rule as a whole.[12] This date will fall before the July 20, 2021, Serious area attainment date for the 2008 ozone NAAQS. EPA has determined that it is feasible for the EGUs subject to this rule to comply with the enhanced stringency of the budgets and that there is sufficient time before the effective date to prepare to meet these budgets by either undertaking the emission control measures EPA has identified in this action, or by taking advantage of compliance flexibilities available through the new interstate emissions trading program EPA is establishing.[13] As explained in greater detail below, due to timing considerations, one aspect of EPA's selected EGU control stringency—installation of state-of-the-art combustion controls—will not take effect until the 2022 ozone season, and this is accounted for in EPA's budget-setting process.

EPA is further adjusting these states' emission budgets for each ozone season from 2022 to 2024 to incentivize ongoing operation of identified emission controls to address significant contribution, until such time that air quality projections demonstrate resolution of the downwind nonattainment and/or maintenance problems for the 2008 ozone NAAQS. No further budget adjustments will be made after that time (i.e., after the 2024 ozone season). EPA is implementing the new state-level ozone season emission budgets through a new CSAPR NOX Ozone Season Group 3 Trading Program. Based on EPA's assessment of remaining air quality issues and additional emission controls, EPA is further determining that these NOX emission reductions fully eliminate these states' significant contribution to nonattainment and interference with maintenance of the 2008 ozone NAAQS in other states.

As discussed in more detail in section IV.C.2.b below, for one state, Kentucky, EPA is making an error correction under CAA section 110(k)(6) of its June 2018 approval of the Commonwealth's SIP, which had concluded that the CSAPR Update was a complete remedy based on modeling of the 2023 analytic year. EPA finds that the basis for that conclusion was invalidated by the decisions in Wisconsin and New York. With finalization of this error correction and disapproval of Kentucky's SIP, Kentucky's good neighbor obligations are outstanding. In light of the Wisconsin remand of Kentucky's FIP and EPA's error correction, the Agency has the necessary authority to amend the CSAPR Update FIP for Kentucky.

For the nine remaining states with FIPs promulgated under the CSAPR Update that EPA previously found partially addressed good neighbor obligations for the 2008 ozone NAAQS (Alabama, Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Texas, and Wisconsin), EPA's updated air quality and contributions analysis shows that these states are not linked to any downwind air quality problems in 2021.[14] Therefore, EPA finds that the existing CSAPR Update FIPs (or the SIP revisions later approved to replace the CSAPR Update FIPs) for these states satisfy their good neighbor obligations for the 2008 ozone NAAQS.[15] Consequently, EPA is not requiring additional emission reductions from sources in these states in this final rule.

B. Summary of the Major Provisions of the Regulatory Action

To reduce interstate ozone transport under the authority provided in CAA section 110(a)(2)(D)(i)(I), this rule further limits ozone season (May 1 through September 30) NOX emissions from EGUs in 12 states using the same framework EPA used in the CSAPR and other good neighbor rules (the 4-step good neighbor framework or 4-step framework). The 4-step good neighbor framework provides a process to address the requirements of the good neighbor provision for ground-level ozone NAAQS: (1) Identifying downwind receptors that are expected to have problems attaining or maintaining the NAAQS; (2) determining which upwind states contribute to these identified problems in amounts sufficient to “link” them to the downwind air quality problems (i.e., here, a contribution threshold equal to or greater than 1 percent of the NAAQS); (3) for states linked to downwind air quality problems, identifying upwind emissions that significantly contribute to downwind nonattainment or interfere with downwind maintenance of the NAAQS; and (4) for states that are found to have emissions that significantly contribute to nonattainment or interfere with maintenance of the NAAQS downwind, implementing the necessary emission reductions through enforceable measures. In this final rule, EPA applies this 4-step framework to respond to the D.C. Circuit's remand in Wisconsin and to revise the CSAPR Update with respect to the 2008 ozone NAAQS.

In order to apply the first step of the 4-step framework to the 2008 ozone NAAQS, EPA performed air quality modeling coupled with ambient measurements in an interpolation technique to project ozone concentrations at air quality monitoring sites in 2021.[16] (“Interpolation” is a Start Printed Page 23058numerical method for constructing new data points within the range of a discrete set of known data points, in this case the known data are the 2016 measured-based and 2023 modeling-based ozone concentrations.) EPA evaluated 2021 projected ozone concentrations at individual monitoring sites and considered current ozone monitoring data at these sites to identify receptors that are anticipated to have problems attaining or maintaining the 2008 ozone NAAQS. Such monitoring sites are referred to as nonattainment and/or maintenance receptors. Based on EPA's analysis, the Agency identified four nonattainment and/or maintenance receptors in 2021 (i.e., three receptors in Connecticut and one in Texas). EPA received comments on its approach to identify nonattainment and/or maintenance receptors in 2021. A summary of these comments, as well as EPA's responses, can be found in section V and in the Response to Comments (RTC) document for this final rule.

To apply the second step of the framework, EPA used an air quality modeling-based technique to quantify the contributions in 2021 from upwind states to ozone concentrations at individual monitoring sites, as described in section V. Once quantified, EPA then evaluated these contributions relative to a screening threshold of 1 percent of the NAAQS (i.e., 0.75 ppb) for those monitoring sites identified as nonattainment and/or maintenance receptors in step 1. States with contributions that equal or exceed 1 percent of the NAAQS were identified as warranting further analysis for significant contribution to nonattainment or interference with maintenance. States with contributions below 1 percent of the NAAQS were considered to not significantly contribute to nonattainment or interfere with maintenance of the NAAQS in downwind states. Based on EPA's updated air quality and contribution analysis using 2021 as the analytic year, EPA is determining that the following 12 states have contributions that equal or exceed 1 percent of the 2008 ozone NAAQS, and thereby warrant further analysis for significant contribution to nonattainment or interference with maintenance: Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West Virginia. EPA received comments on its approach to quantify interstate contributions and the use of a 1 percent of the NAAQS screening threshold. A summary of these comments, as well as EPA's responses, can be found in section V and in the RTC document for this final rule.

At the third step of the 4-step framework, EPA applied the multi-factor test used in the CSAPR Update, which evaluates cost, available emission reductions, and downwind air quality impacts to determine the amount of linked upwind states' emissions that “significantly” contribute to downwind nonattainment or maintenance receptors. In this action, EPA applied the multi-factor test to both EGU and non-EGU source categories and assessed potential emission reductions in all years for which there is a potential remaining interstate ozone transport problem (i.e., through 2025), in order to ensure a full remedy in accordance with the Wisconsin decision.

In the proposed rule, EPA identified a control stringency that reflects the optimization of existing selective catalytic reduction (SCR) controls and installation of state-of-the-art NOX combustion controls at EGUs, represented by a cost of $1,600 per ton of NOX reduced. In this final rule, EPA is determining that optimization of existing selective non-catalytic reduction (SNCR) controls should also be included in EPA's identified EGU control stringency. As discussed in further detail in Section VI, EPA adjusted its representative cost for optimizing existing SNCR controls to $1,800 per ton in response to comments received on the proposed rule, as well as further EPA review of available information. EPA views $1,600 per ton for optimization of existing SCR controls and installation of state-of-the-art NOX combustion controls and $1,800 per ton for optimization of existing SNCRs as comparable for policy purposes. In addition, other considerations beyond marginal cost and air quality improvement, as outlined in the section VI.D discussion of the multi-factor test, support inclusion of emission reduction potential from optimization of existing SNCR controls in EPA's identified EGU control stringency in this rule.

At the selected control stringency in this final rule, downwind ozone air quality improvements continue to be maximized relative to a representative marginal cost. That is, the ratio of emission reductions to marginal cost and the ratio of ozone improvements to marginal cost are maximized relative to the other control stringency levels evaluated. EPA finds that these cost-effective EGU NOX reductions will make meaningful and timely improvements in downwind ozone air quality to address interstate ozone transport for the 2008 ozone NAAQS, as discussed in section VI.D.1 below. Further, this evaluation shows that emission budgets reflecting the optimization of existing SCRs and SNCRs, and installation of state-of-the-art NOX combustion controls at EGUs do not over-control upwind states' emissions relative to either the downwind air quality problems to which they are linked at step 1 or the 1 percent contribution threshold that triggers further evaluation at step 2 of the 4-step framework.

EPA notes that two of these EGU emission controls (optimization of existing SCR controls and installation of state-of-the-art NOX combustion controls) were also selected in the CSAPR Update for the 2017 ozone season, and which at that time EPA characterized as only a partial remedy. For this rule, EPA extends its evaluation of the reduction potential from these emission controls to years beyond 2017 in order to assess a full remedy. EPA's updated analysis, as discussed in more detail in section VI, leads the Agency to find that these emission controls can provide additional cost-effective emission reductions for the 2021 through 2024 ozone seasons. While EPA's analysis indicates that the majority of EGUs implemented these emission controls in response to the CSAPR Update, changes in the power sector since the 2017 ozone season and updated air quality and contribution analysis show that there is a demonstrated need to update the emission budgets for these 12 states to incentivize ongoing operation of identified emission controls to fully eliminate significant contribution and interference with maintenance. Likewise, EPA finds that many EGUs are already operating their existing SNCR controls to some extent but that additional cost-effective emission reductions for the 2021 through 2024 ozone seasons are available. Taken together, the emission budgets established in this final rule reflect EPA's identified EGU control stringency of optimization of all existing post-combustion controls (SCRs and SNCRs) by the 2021 ozone season, and the installation of state-of-the-art NOX combustion controls by the 2022 ozone season.

For non-EGU industry sectors and emissions sources, EPA applied the step 3 multi-factor test to determine whether any emission reductions should be required from non-EGU sources to address significant contribution under the 2008 ozone NAAQS. EPA acknowledged in the proposed rule that its current datasets with information on emissions, existing controls on Start Printed Page 23059emissions sources, emission-reduction potential, and air quality impacts for these sources are not as well developed as the datasets it has for EGUs. Nonetheless, using the best information currently available to the Agency, including some additional analysis conducted between the proposed rule and this final action, EPA is concluding that there are relatively fewer emission reductions available at a cost threshold comparable to the cost threshold selected for EGUs. In EPA's reasoned judgment, the Agency concludes such reductions are estimated to have a much smaller effect on any downwind receptor in the year by which EPA finds such controls could be installed. For these reasons, EPA is finding that limits on ozone season NOX emissions from non-EGU sources are not required to eliminate significant contribution or interference with maintenance under the 2008 ozone NAAQS (see section VI.D.2).

Based on EPA's analysis at step 3, the Agency is promulgating EGU NOX ozone season emission budgets developed using a uniform control stringency of optimization of existing SCRs and SNCRs, and installation of state-of-the-art NOX combustion controls. EPA is determining that with implementation of this control stringency, the 12 states in Table I.A-2 will have fully addressed significant contribution under the good neighbor provision for the 2008 ozone NAAQS. EPA is aligning implementation of emission budgets with relevant attainment dates for the 2008 ozone NAAQS, consistent with CAA requirements and the D.C. Circuit's decision in Wisconsin v. EPA.[17] As EPA's final 2008 Ozone NAAQS SIP Requirements Rule [18] established the attainment date of July 20, 2021, for ozone nonattainment areas currently designated as Serious, EPA is establishing emission budgets and implementation of these emission budgets starting with the 2021 ozone season as shown in Table I.B-1.[19]

Table I.B-1—EGU NOX Ozone Season Emission Budgets

[Ozone Season NOX Tons] *

State2021 Budget2022 Budget2023 Budget2024 Budget
Illinois9,1029,1028,1798,059
Indiana13,05112,58212,5539,564
Kentucky15,30014,05114,05114,051
Louisiana14,81814,81814,81814,818
Maryland1,4991,2661,2661,348
Michigan12,72712,2909,9759,786
New Jersey1,2531,2531,2531,253
New York3,4163,4163,4213,403
Ohio9,6909,7739,7739,773
Pennsylvania8,3798,3738,3738,373
Virginia4,5163,8973,9803,663
West Virginia13,33412,88412,88412,884
Total107,085103,705100,52696,975
* Note—The 2022 and beyond budgets incorporate the installation of state-of-the-art NOX combustion controls, whereas the 2021 budgets do not. Additionally, the 2024 emissions budget applies to 2024 and each year thereafter.

EPA further determined which emission reductions are impossible to achieve by the 2021 attainment date and whether any such additional emission reductions should be required beyond that date.[20] See Wisconsin, 938 F.3d at 320. EPA estimates that one part of the selected control stringency—installation of state-of-the-art NOX combustion controls—requires approximately one to six months depending on the unit. Recognizing that the final rule will become effective slightly after the start of the 2021 ozone season, EPA determined it is not possible to install state-of-the-art NOX combustion controls on a regional scale by the 2021 ozone season. Therefore, the 2021 ozone season emission budgets reflect only the optimization of existing SCR and SNCR controls at the affected EGUs, but the emission budgets for the 2022 ozone season and beyond reflect both the continued optimization of existing SCR and SNCR controls and installation of state-of-the-art NOX combustion controls. Detailed installation-timing information for this technology is available in section VI.B and the EGU NOX Mitigation Strategies Final Rule TSD.

As discussed in section VI.D.1, EPA's air quality projections anticipate that with the implementation of the identified control stringency for EGUs, downwind nonattainment and maintenance problems for the 2008 ozone NAAQS will persist through the 2024 ozone season. Therefore, EPA is adjusting emission budgets for upwind states that remain linked to downwind nonattainment and maintenance problems through the 2024 ozone season to incentivize the continued optimization of existing SCR and SNCR controls, and installation of state-of-the-art NOX combustion controls. The 2024 emission budgets will then continue to apply in each year thereafter.

To apply the fourth step of the 4-step framework (i.e., implementation), EPA is including enforceable measures in the promulgated FIPs to achieve the required emission reductions in each of the 12 states. Specifically, the FIPs require power plants in the 12 states to participate in a new CSAPR NOX Ozone Season Group 3 Trading Program that largely replicates the existing CSAPR NOX Ozone Season Group 2 Trading Start Printed Page 23060Program with the main differences being the geography and budget stringency. This final rule leaves unchanged the budget stringency of the existing CSAPR NOX Ozone Season Group 1 and Group 2 trading programs for the states that remain covered by those programs.

EPA is finalizing the proposed feature of the budget-setting process in which budgets are adjusted in 2022, 2023, and 2024 to account for future unit retirements and construction of new units that are known with sufficient certainty as of this final action. As discussed in section VII.C.3.b, in response to comments, EPA has made the methodology for allocating allowances to existing units in this final rule more consistent with the budget-setting process by eliminating allocations to units following their retirements in instances where the future retirements were scheduled in advance with sufficient certainty to be taken into account in the budget-setting process.

As proposed, to promote compliance flexibility without relaxing the program stringency identified as appropriate to address states' obligations under CAA section 110(a)(2)(D)(i)(I), EPA is creating a limited initial bank of allowances for use in the new Group 3 trading program by converting allowances banked in 2017-2020 under the existing Group 2 trading program at a formula-based conversion ratio. The target bank amount is based on the sum of the states' “variability limits”—that is, the amounts by which emissions from a given state's units can exceed the state's emission budget before incurring a penalty surrender ratio. As discussed in section VII.C.4.b, in response to comments requesting greater certainty, in the final rule EPA has modified the proposed conversion ratio formula so as to yield an expected fixed conversion ratio of 8:1 (i.e., eight Group 2 allowances must be exchanged for each Group 3 allowance). Participation in the conversion process is mandatory for the sources in states covered by the Group 3 trading program and, if the Group 3 sources' accounts collectively do not hold enough Group 2 allowances to exchange for the entire target bank amount, for holders of Group 2 allowances in non-source accounts as well.[21]

As discussed in section VII.C.4.c, the final rule also provides a second opportunity for sources to create an additional limited number of Group 3 allowances through the voluntary conversion of additional Group 2 allowances at an 18:1 conversion ratio (known as a “safety valve”). Any 2017-2020 Group 2 allowances that have not already been exchanged for Group 3 allowances through the process of creating the initial bank may be used to obtain additional Group 3 allowances through the safety valve mechanism. The availability of the starting bank and any additional allowances converted using this “safety valve” ensures that compliance with the rule is feasible and addresses any market liquidity concerns raised by commenters.

The remainder of this preamble is organized as follows: section III describes EPA's legal authority for this final action; section IV describes the human health and environmental context, as well as EPA's approach for addressing interstate transport for the 2008 ozone NAAQS; section V describes EPA's assessment of downwind receptors of concern and upwind state ozone contributions to those receptors, including the air quality modeling platform and emission inventories that EPA used; section VI describes EPA's application of the multifactor test at step 3 of the 4-step framework to EGU and non-EGU sources, quantification of upwind state obligations in the form of final EGU NOX emission budgets, and assessment of overcontrol; section VII details the implementation requirements including key elements of the CSAPR NOX Ozone Season Group 3 Trading Program and deadlines for compliance; section VIII describes the expected costs, benefits, and other impacts of this final rule; section IX discusses changes to the existing regulatory text; and section X discusses the statutes and executive orders affecting this final rule. Each section includes a summary of the principal comments received with respect to that topic, as well as EPA's responses. The Revised Cross State Air Pollution Update Rule—Response to Comment document (RTC), which includes a compilation of all comments received and EPA's responses, has been included in the docket for this action.

C. Costs and Benefits

A summary of the key results of the cost-benefit analysis that was prepared for this final rule is presented in Table I.C-1. Table I.C-1 presents estimates of the present values (PV) and equivalent annualized values (EAV), calculated using discount rates of 3 and 7 percent as directed by OMB's Circular A-4, of the health benefits, climate benefits, compliance costs, and net benefits of the final rule, in 2016 dollars, discounted to 2021. The estimated net benefits are the estimated benefits minus the estimated costs of the final rule.

Table I.C-1—Estimated Health Benefits, Climte Benefits, Compliance Costs, and Net Benefits of the Final Rule, 2021 Through 2040

[Millions 2016$, discounted to 2021]

3% Discount rate7% Discount rate
Present Value:
Health Benefits b$4,800 and $37,000$3,200 and $25,000.
Climate Benefits b$4,400$4,400.
Compliance Costs c$370$260.
Net Benefits$8,800 and $41,000$7,300 and $29,000.
Equivalent Annualized Value:
Health Benefits$320 and $2,500$300 and $2,400.
Climate Benefits$290$290.
Compliance Costs$25$25.
Net Benefits$590 and $2,800$570 and $2,700.
a Numbers may not sum due to independent rounding.Start Printed Page 23061
b The health benefits are associated with several point estimates and are presented at real discount rates of 3 and 7 percent. The two benefits estimates are separated by the word “and” to signify that they are two separate estimates. The estimates do not represent lower- and upper-bound estimates and should not be summed. Climate benefits are based on changes (reductions) in CO2 emissions and are calculated using four different estimates of the social cost of carbon (SC-CO2) (model average at 2.5 percent, 3 percent, and 5 percent discount rates; 95th percentile at 3 percent discount rate). For the presentational purposes of this table, we show the climate benefits associated with the average SC-CO2 at a 3 percent discount rate, but the Agency does not have a single central SC-CO2 point estimate. We emphasize the importance and value of considering the benefits calculated using all four SC-CO2 estimates; the additional benefit estimates are presented in Table VIII.5 in Section VIII. As discussed in Chapter 5 of the Regulatory Impact Analysis for the Final Revised Cross-State Air Pollution Rule Update for the 2008 Ozone NAAQS, a consideration of climate benefits calculated using discount rates below 3 percent, including 2 percent and lower, are also warranted when discounting intergenerational impacts.
c To estimate these annualized costs, EPA uses a conventional and widely accepted approach that applies a capital recovery factor (CRF) multiplier to capital investments and adds that to the annual incremental operating expenses. Annual costs were calculated using a 4.25% real discount rate consistent with the rate used in IPM's objective function for cost-minimization.

As shown in Table I.C-1, the PV of the health benefits of this final rule, discounted at a 3-percent discount rate, is estimated to be about $4,800 million and $37,000 million, with an EAV of about $320 million and $2,500 million. At a 7-percent discount rate, the PV of the health benefits is estimated to be $3,200 million and $25,000 million, with an EAV of about $300 million and $2,400 million. The two health benefits estimates for each discount rate reflect alternative ozone and PM2.5 mortality risk estimates. The PV of the climate benefits of this final rule, discounted at a 3-percent rate, is estimated to be about $4,400 million, with an EAV of about $290 million. The PV of the compliance costs, discounted at a 3-percent rate, is estimated to be about $370 million, with an EAV of about $25 million. At a 7-percent discount rate, the PV of the compliance costs is estimated to be about $260 million, with an EAV of about $25 million.

II. General Information

A. Does this action apply to me?

This final rule affects EGUs, and regulates the groups identified in Table II.A-1:

Table II.A-1—Regulated Groups

Industry groupNAICS *
Fossil fuel-fired electric power generation221112
* North American Industry Classification System.

This table is not intended to be exhaustive, but rather provides a guide for readers regarding entities likely to be regulated by this action. This table lists the types of entities that EPA is now aware are regulated by this action. Other types of entities not listed in the table could also be regulated. To determine whether your EGU entity is regulated by this action, you should carefully examine the applicability criteria found in 40 CFR 97.1004, as promulgated in this final action. If you have questions regarding the applicability of this action to a particular entity, consult the person listed in the FOR FURTHER INFORMATION CONTACT section.

III. EPA's Legal Authority for the Final Rule

A. Statutory Authority

The statutory authority for this final action is provided by the CAA as amended (42 U.S.C. 7401 et seq.). Specifically, sections 110 and 301 of the CAA provide the primary statutory underpinnings for this action. The most relevant portions of CAA section 110 are subsections 110(a)(1), 110(a)(2) (including 110(a)(2)(D)(i)(I)), 110(c)(1), and 110(k)(6).

CAA section 110(a)(1) provides that states must make SIP submissions “within 3 years (or such shorter period as the Administrator may prescribe) after the promulgation of a national primary ambient air quality standard (or any revision thereof),” and that these SIP submissions are to provide for the “implementation, maintenance, and enforcement” of such NAAQS.[22] The statute directly imposes on states the duty to make these SIP submissions, and the requirement to make the submissions is not conditioned upon EPA taking any action other than promulgating a new or revised NAAQS.[23]

EPA has historically referred to SIP submissions made for the purpose of satisfying the applicable requirements of CAA sections 110(a)(1) and 110(a)(2) as “infrastructure SIP” or “iSIP” submissions. CAA section 110(a)(1) addresses the timing and general requirements for iSIP submissions, and CAA section 110(a)(2) provides more details concerning the required content of these submissions.[24] It includes a list of specific elements that “[e]ach such plan” submission must address.[25]

CAA section 110(c)(1) requires the Administrator to promulgate a FIP at any time within two years after the Administrator: (1) Finds that a state has failed to make a required SIP submission; (2) finds a SIP submission to be incomplete pursuant to CAA section 110(k)(1)(C); or (3) disapproves a SIP submission. This obligation applies unless the state corrects the deficiency through a SIP revision that the Administrator approves before the FIP is promulgated.[26]

CAA section 110(a)(2)(D)(i)(I), also known as the “good neighbor” provision, provides the primary basis for this final action.[27] It requires that each state SIP include provisions sufficient to “prohibit[ ], consistent with the provisions of this subchapter, any source or other type of emissions activity within the State from emitting any air pollutant in amounts which will—(I) contribute significantly to nonattainment in, or interfere with maintenance by, any other State with respect to any [NAAQS].” [28] EPA often refers to the emission reduction requirements under this provision as “good neighbor obligations” and submissions addressing these requirements as “good neighbor SIPs.”

Once EPA promulgates a NAAQS, EPA must designate areas as being in “attainment” or “nonattainment” of the NAAQS, or “unclassifiable.” CAA section 107(d).[29] For ozone, nonattainment is further split into five classifications based on the severity of the violation—Marginal, Moderate, Serious, Severe, or Extreme. Higher classifications provide states with progressively more time to attain while imposing progressively more stringent control requirements. See CAA sections 181, 182.[30] In general, states with nonattainment areas classified as Moderate or higher must submit plans to EPA to bring these areas into Start Printed Page 23062attainment according to the statutory schedule. CAA section 182.[31] If an area fails to attain the NAAQS by the attainment date associated with its classification, it is “bumped up” to the next classification. CAA section 181(b).[32]

Section 301(a)(1) of the CAA also gives the Administrator the general authority to prescribe such regulations as are necessary to carry out functions under the Act.[33] Pursuant to this section, EPA has authority to clarify the applicability of CAA requirements and undertake other rulemaking action as necessary to implement CAA requirements. In this final rule, among other things, EPA is clarifying the applicability of CAA section 110(a)(2)(D)(i)(I) with respect to the 2008 ozone NAAQS. In particular, EPA is using its authority under CAA sections 110 and 301 to issue new or amended FIPs to revise NOX ozone season emission budgets for 12 states to eliminate their significant contribution to nonattainment or interference with maintenance of the 2008 ozone NAAQS in another state, and EPA is making findings as to 9 additional states that the CSAPR Update FIPs (or SIP revisions later approved to replace those FIPs) are a complete remedy and need no further revision.[34] In addition, EPA is addressing its obligation to respond to the D.C. Circuit's remand of the CSAPR Update in Wisconsin v. EPA, 938 F.3d 303, with respect to the 21 states for which the FIPs created by that rule were found to be only a partial remedy. This final rule wholly resolves the Agency's obligations on remand. Finally, CAA section 301 [35] affords the Agency any additional authority that may be needed in order to make certain other changes to its regulations under 40 CFR parts 51, 52, 78, and 97, in order to effectuate the purposes of the Act. Such changes are discussed in section VII of this preamble.

B. Prior Good Neighbor Rulemakings Addressing Regional Ozone

EPA has issued several rules interpreting and clarifying the requirements of CAA section 110(a)(2)(D)(i)(I) with respect to the regional transport of ozone for states in the eastern United States. These rules, and the associated court decisions addressing these rules, summarized here, provide important direction regarding the requirements of CAA section 110(a)(2)(D)(i)(I).

The NOX SIP Call, promulgated in 1998, addressed the good neighbor provision for the 1979 1-hour ozone NAAQS.[36] The rule required 22 states and the District of Columbia to amend their SIPs to reduce NOX emissions that contribute to ozone nonattainment in downwind states. EPA set ozone season NOX budgets for each state, and the states were given the option to participate in a regional trading program, known as the NOX Budget Trading Program.[37] The D.C. Circuit largely upheld the NOX SIP Call in Michigan v. EPA, 213 F.3d 663 (D.C. Cir. 2000), cert. denied, 532 U.S. 904 (2001).

EPA's next rule addressing the good neighbor provision, the Clean Air Interstate Rule (CAIR), was promulgated in 2005 and addressed both the 1997 fine particulate matter (PM2.5) NAAQS and 1997 ozone NAAQS.[38] CAIR required SIP revisions in 28 states and the District of Columbia to reduce emissions of sulfur dioxide (SO2) and/or NOX—important precursors of regionally transported PM2.5 (SO2 and annual NOX) and ozone (summer-time NOX). As in the NOX SIP Call, states were given the option to participate in regional trading programs to achieve the reductions. When EPA promulgated the final CAIR in 2005, EPA also issued findings that states nationwide had failed to submit SIPs to address the requirements of CAA section 110(a)(2)(D)(i) with respect to the 1997 PM2.5 and 1997 ozone NAAQS.[39] On March 15, 2006, EPA promulgated FIPs to implement the emission reductions required by CAIR.[40] CAIR was remanded to EPA by the D.C. Circuit in North Carolina v. EPA, 531 F.3d 896 (D.C. Cir. 2008), modified on reh'g, 550 F.3d 1176. For more information on the legal issues underlying CAIR and the D.C. Circuit's holding in North Carolina, refer to the preamble of the CSAPR rule.[41]

In 2011, EPA promulgated the CSAPR to address the issues raised by the remand of CAIR. The CSAPR addressed the two NAAQS at issue in CAIR and additionally addressed the good neighbor provision for the 2006 PM2.5 NAAQS.[42] The CSAPR required 28 states to reduce SO2 emissions, annual NOX emissions, and/or ozone season NOX emissions that significantly contribute to other states' nonattainment or interfere with other states' abilities to maintain these air quality standards.[43] To align implementation with the applicable attainment deadlines, EPA promulgated FIPs for each of the 28 states covered by the CSAPR. The FIPs require EGUs in the covered states to participate in regional trading programs to achieve the necessary emission reductions. Each state can submit a good neighbor SIP at any time that, if approved by EPA, would replace the CSAPR FIP for that state.

The CSAPR was the subject of an adverse decision by the D.C. Circuit in August 2012.[44] However, this decision was reversed in April 2014 by the Supreme Court, which largely upheld the rule, including EPA's approach to addressing interstate transport in the CSAPR. EPA v. EME Homer City Generation, L.P., 572 U.S. 489 (2014) (EME Homer City I). The rule was remanded to the D.C. Circuit to consider claims not addressed by the Supreme Court. Id. In July 2015 the D.C. Circuit generally affirmed EPA's interpretation of various statutory provisions and EPA's technical decisions. EME Homer City Generation, L.P. v. EPA, 795 F.3d 118 (2015) (EME Homer City II). However, the court remanded the rule without vacatur for reconsideration of EPA's emissions budgets for certain states, which the court found may have over-controlled those states' emissions with respect to the downwind air quality problems to which the states Start Printed Page 23063were linked. Id. at 129-30, 138. For more information on the legal issues associated with the CSAPR and the Supreme Court's and D.C. Circuit's decisions in the EME Homer City litigation, refer to the preamble of the CSAPR Update.[45]

In 2016, EPA promulgated the CSAPR Update to address interstate transport of ozone pollution with respect to the 2008 ozone NAAQS.[46] The final rule updated the CSAPR ozone season NOX emissions budgets for 22 states to achieve cost-effective and immediately feasible NOX emission reductions from EGUs within those states.[47] EPA aligned the analysis and implementation of the CSAPR Update with the 2017 ozone season in order to assist downwind states with timely attainment of the 2008 ozone NAAQS.[48] The CSAPR Update implemented the budgets through FIPs requiring sources to participate in a revised CSAPR NOX ozone season trading program beginning with the 2017 ozone season. As under the CSAPR, each state could submit a good neighbor SIP at any time that, if approved by EPA, would replace the CSAPR Update FIP for that state. The final CSAPR Update also addressed the remand by the D.C. Circuit of certain states' CSAPR phase 2 ozone season NOX emissions budgets in EME Homer City II. Further details regarding the CSAPR Update are discussed in sections IV.C.1.a and IV.C.1.b below.

In December 2018, EPA promulgated the CSAPR “Close-Out,” which determined that no further enforceable reductions in emissions of NOX were required with respect to the 2008 ozone NAAQS for 20 of the 22 eastern states covered by the CSAPR Update, and reflected that determination in revisions to the existing state-specific sections of the CSAPR Update regulations for those states.[49] Further details on the CSAPR Close-Out are discussed in section IV.C.1.c below.

The CSAPR Update and the CSAPR Close-Out were both subject to legal challenges in the D.C. Circuit. Wisconsin v. EPA, 938 F.3d 303 (D.C. Cir. 2019) (Wisconsin); New York v. EPA, 781 Fed. App'x 4 (D.C. Cir. 2019) (New York). As discussed in greater detail in section IV.C.1.d below, in September 2019, the D.C. Circuit upheld the CSAPR Update in virtually all respects, but remanded the rule because it was partial in nature and did not fully eliminate upwind states' significant contribution to nonattainment or interference with maintenance of the 2008 ozone NAAQS by “the relevant downwind attainment deadlines” in the CAA. Wisconsin, 938 F.3d at 313-15. In October 2019, the D.C. Circuit vacated the CSAPR Close-Out on the same grounds that it remanded the CSAPR Update in Wisconsin, specifically that the Close-Out rule did not address good neighbor obligations by “the next applicable attainment date” of downwind states. New York, 781 Fed. App'x at 7.

IV. Air Quality Issues Addressed and Overall Approach for the Final Rule

A. The Interstate Ozone Transport Challenge

Interstate transport of NOX emissions poses significant challenges with respect to the 2008 ozone NAAQS in the eastern U.S. and thus presents a threat to public health and welfare.

1. Nature of Ozone and the Ozone NAAQS

Ground-level ozone is not emitted directly into the air but is created by chemical reactions between NOX and volatile organic compounds (VOC) in the presence of sunlight. Emissions from electric utilities and industrial facilities, motor vehicles, gasoline vapors, and chemical solvents are some of the major sources of NOX and VOC.

Because ground-level ozone formation increases with temperature and sunlight, ozone levels are generally higher during the summer. Increased temperature also increases emissions of volatile man-made and biogenic organics and can indirectly increase NOX emissions as well (e.g., increased electricity generation for air conditioning).

The 2008 primary and secondary ozone standards are both 75 ppb as an 8-hour level.[50] Specifically, the standards require that the 3-year average of the fourth highest 24-hour maximum 8-hour average ozone concentration may not exceed 75 ppb as a truncated value (i.e., digits to right of decimal removed).[51] In general, areas that exceed the ozone standard are designated as nonattainment areas, pursuant to the designations process under CAA section 107 and are subject to heightened planning requirements depending on the degree of severity of their nonattainment classification, see CAA sections 181, 182.

2. Ozone Transport

Studies have established that ozone formation, atmospheric residence, and transport occur on a regional scale (i.e., thousands of kilometers) over much of the eastern U.S.[52] While substantial progress has been made in reducing ozone in many areas, interstate ozone transport is still an important component of peak ozone concentrations during the summer ozone season.

EPA has previously concluded in the NOX SIP Call, CAIR, and the CSAPR that, for reducing regional-scale ozone transport, a NOX control strategy would be most effective. NOX emissions can be transported downwind as NOX or, after transformation in the atmosphere, as ozone. As a result of ozone transport, in any given location, ozone pollution levels are impacted by a combination of local emissions and emissions from upwind sources. The transport of ozone pollution across state borders compounds the difficulty for downwind states in meeting health-based air quality standards (i.e., NAAQS). Assessments of ozone, for example those conducted for the October 2015 Regulatory Impact Analysis of the Final Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone (EPA-452/R-15-007), continue to show the importance of NOX emissions for ozone transport. This analysis is in the docket for this final rule and can be also found at EPA's website at: https://www.epa.gov/​ttnecas1/​docs/​20151001ria.pdf.

Further, studies have found that EGU NOX emission reductions can be effective in reducing individual 8-hour peak ozone concentrations and in Start Printed Page 23064reducing 8-hour peak ozone concentrations averaged across the ozone season. For example, a study that evaluates the effectiveness on ozone concentrations of EGU NOX reductions achieved under the NOX Budget Trading Program (i.e., the NOX SIP Call) shows that regulating NOX emissions in that program was highly effective in reducing both ozone and dry-NO3 concentrations during the ozone season. Further, this study indicates that EGU emissions, which are generally released higher in the air column through tall stacks and are significant in quantity, may disproportionately contribute to long-range transport of ozone pollution on a per-ton basis.[53]

Previous regional ozone transport efforts, including the NOX SIP Call, CAIR, and the CSAPR, required ozone season NOX reductions from EGUs to address interstate transport of ozone. EPA took comment on regulating EGU NOX emissions to address interstate ozone transport in the notice-and-comment process for these rulemakings. EPA received some comments suggesting it modify its pollutant focus to either include VOCs in addition to NOX, or apply a more granular time scale. However, EPA did not modify its proposed approach in this final rule. These comments, as well as EPA's responses, are addressed in section VI.A and VII.B.

As described in section VI, EPA's analysis finds that the power sector continues to be capable of making NOX reductions at reasonable cost that reduce interstate transport with respect to ground-level ozone. EGU NOX emission reductions can be made in the near-term under this final rule by fully operating existing EGU NOX post-combustion controls (i.e., SCRs and SNCRs)—including optimizing NOX removal by existing operational controls and turning on and optimizing existing idled controls; installation of (or upgrading to) state-of-the-art NOX combustion controls; and shifting generation to units with lower NOX emission rates. Further, additional assessment reveals that these available EGU NOX reductions would make meaningful and timely improvements in ozone air quality.

EPA also observes that significant emission reduction potential from EGUs is available through post-combustion control retrofits (e.g., new SCRs and new SNCRs). These controls reduce emissions and can have a meaningful air quality impact, but, in contrast to the controls discussed above, they are only available on a longer time frame (reflecting the time required to develop, construct, and install the technology) that exceeds the expected downwind nonattainment and maintenance problems for the 2008 ozone NAAQS and are estimated to have a higher cost.

3. Health and Environmental Effects

Exposure to ambient ozone causes a variety of negative effects on human health, vegetation, and ecosystems. In humans, acute and chronic exposure to ozone is associated with premature mortality and a number of morbidity effects, such as asthma exacerbation. In ecosystems, ozone exposure causes visible foliar injury, decreases plant growth, and affects ecosystem community composition. See EPA's October 2020 Regulatory Impact Analysis for the Proposed Revised Cross-State Air Pollution Rule (CSAPR) Update for the 2008 Ozone NAAQS (EPA-452/P-20-003), in the docket for this rule and available on EPA's website at: https://www.epa.gov/​sites/​production/​files/​2020-10/​documents/​revised_​csapr_​update_​ria_​proposal.pdf, for more information on the human health and welfare and ecosystem effects associated with ambient ozone exposure.

B. Relationship Between This Regulatory Action and the 2015 Ozone NAAQS

On October 1, 2015, EPA strengthened the ground-level ozone NAAQS to 70 ppb on an eight-hour averaging time.[54] While reductions achieved by this rule may have the effect of aiding in attainment and maintenance of the 2015 standard, this action is taken solely with respect to EPA's authority to address remaining CAA good neighbor obligations under the 2008 ozone NAAQS. EPA and states are working outside of this final action to address the CAA good neighbor provision for the 2015 ozone NAAQS, including consideration of any necessary control requirements for EGU and non-EGU sources.

EPA received several comments regarding the relationship of this rule to the 2015 ozone NAAQS and the schedule for implementation of good neighbor obligations related to that NAAQS. These comments are out of the scope of this action, which considers states' obligations under 2008 ozone NAAQS in response to the Wisconsin remand and the New York vacatur. Wisconsin v. EPA, 938 F.3d 303 (D.C. Cir. 2019). New York v. EPA, 781 F. App'x 4 (D.C. Cir. 2019). This action does not address any state's obligations under the 2015 ozone NAAQS. Nonetheless, the emission reductions and associated improvement in ozone levels achieved by this action are beneficial toward reducing ozone for purposes of the 2015 ozone NAAQS and its associated attainment planning and good neighbor requirements. In some cases, the reductions necessary to address significant contribution or interference with maintenance at receptors identified in this action for purposes of the 2008 ozone NAAQS will have the effect of incidentally improving ozone levels at potential receptors under the 2015 ozone NAAQS.

C. Approach To Address the Remanded Transport Obligations for the 2008 Ozone NAAQS

1. Events Affecting Application of the Good Neighbor Provision for the 2008 Ozone NAAQS

EPA is taking this action to address the remand of the CSAPR Update in Wisconsin v. EPA, 938 F.3d 303 (D.C. Cir. 2019). This section will discuss the key, relevant aspects of the CSAPR Update, the related CSAPR Close-Out, and the D.C. Circuit's decisions in Wisconsin and New York v. EPA, 781 Fed. App'x 4 (D.C. Cir. 2019) (the latter of which vacated the Close-out Rule based on the same reasoning as the Wisconsin decision remanding the Update). The basis for EPA's authority under CAA section 110(c) (42 U.S.C. 7410(c)) to promulgate good neighbor FIPs for the 21 states subject to this action on remand is discussed in sections III and IV.C.2.

a. The CSAPR Update

On October 26, 2016, the CSAPR Update was published in the Federal Register. 81 FR 74504. The purpose of the CSAPR Update was to address the good neighbor provision for the 2008 ozone NAAQS, as well as address the remanded CSAPR obligations for the 1997 ozone NAAQS. The CSAPR Update required EGUs in 22 states to reduce ozone season NOX emissions that significantly contribute to other states' nonattainment or interfere with other states' abilities to maintain the 2008 ozone NAAQS.

To establish and implement the CSAPR Update emissions budgets, EPA followed the same 4-step analytic process that it used in the CSAPR, an Start Printed Page 23065approach which reflects the evolution of the Agency's prior regional interstate transport rulemakings related to ozone NAAQS. The 4-step framework is described in more detail in sections IV.C.3 and VI.A.

In the CSAPR Update, to evaluate the scope of the interstate ozone transport problem at step 1, EPA identified downwind areas that were expected to have problems attaining and maintaining the 2008 ozone NAAQS using modeling that projected air quality to a future compliance year. See 81 FR 74517. EPA aligned the analysis and implementation of the CSAPR Update with the 2017 ozone season (May 1-September 30) in order to assist downwind states with attainment of the 2008 ozone NAAQS by the 2018 Moderate area attainment date. Id. at 74516. (EPA's final 2008 Ozone NAAQS SIP Requirements Rule established the attainment deadline of July 20, 2018, for ozone nonattainment areas classified as Moderate.[55] ) Because the attainment date fell during the 2018 ozone season, the 2017 ozone season was the last full season from which data could be used to determine attainment of the NAAQS by that date.

At step 2, EPA identified upwind states that collectively contribute to these identified downwind areas. In the CSAPR Update, EPA used a screening threshold of 1 percent of the NAAQS to identify states “linked” to downwind ozone problems sufficient for further evaluation for significant contribution to nonattainment or interference with maintenance of the NAAQS under the good neighbor provision. 81 FR 74518. This same threshold for analysis was used in the CSAPR as to the 1997 ozone NAAQS. See 76 FR at 48237-38.

At step 3, EPA quantified emissions from upwind states that would significantly contribute to nonattainment or interfere with maintenance by first evaluating various levels of uniform NOX control stringency, each represented by an estimated representative marginal cost per ton of NOX reduced. EPA then applied the same multi-factor test that was used in the CSAPR to evaluate cost, available emission reductions, and downwind air quality impacts to determine the appropriate level of uniform NOX control stringency that addressed the impacts of interstate transport on downwind nonattainment or maintenance receptors. EPA used this multi-factor assessment to gauge the extent to which emission reductions could be implemented in the future compliance year (i.e., 2017) and to evaluate the potential for over- and under-control of upwind state emissions.

Within the multi-factor test, EPA identified a “knee in the curve,” i.e., a point at which the cost-effectiveness of the emission reductions was maximized, so named for the discernable turning point observable in a multi-factor (i.e., multi-variable) curve. See 81 FR 74550. EPA concluded that this was at the point where emissions budgets reflected a uniform NOX control stringency represented by an estimated marginal cost of $1,400 per ton (2011$) of NOX reduced. This cost threshold in turn represented a control strategy of installing or upgrading combustion controls and optimizing existing SCR controls. In light of this multi-factor test, EPA determined this level of stringency in emissions budgets represented the level at which incremental EGU NOX reduction potential and corresponding downwind ozone air quality improvements were maximized—relative to other control stringencies evaluated—with respect to marginal cost. That is, the ratio of emission reductions to marginal cost and the ratio of ozone improvements to marginal cost were maximized relative to the other levels of control stringency evaluated. EPA found that feasible and cost-effective EGU NOX reductions were available to make meaningful and timely improvements in downwind ozone air quality to address interstate ozone transport for the 2008 ozone NAAQS for the 2017 ozone season. Id. at 74508. Further, the Agency's evaluation showed that emissions budgets reflecting the $1,400 per ton cost threshold did not over-control upwind states' emissions relative to either the downwind air quality problems to which they were linked or the 1 percent contribution threshold in step 2 that triggered their further evaluation in step 3. Id. at 74551-52.

At step 4, EPA finalized EGU ozone season NOX emissions budgets developed using uniform control stringency represented by $1,400 per ton. These budgets represented emissions remaining in each state after elimination of the amounts of emissions that EPA identified would significantly contribute to nonattainment or interfere with maintenance of the 2008 ozone NAAQS in downwind states. EPA promulgated FIPs requiring the covered power plants in the 22 covered states to participate in the CSAPR NOX Ozone Season Group 2 Trading Program starting in 2017.[56]

b. Partial Nature of the CSAPR Update

At the time it promulgated the CSAPR Update, EPA considered the FIPs to be “partial” and that the rule “may not be sufficient to fully address these states' good neighbor obligations” for the 2008 ozone NAAQS for 21 of the 22 states included in that rule. 81 FR 74508, 74521 (Oct. 26, 2016). Based on information available at the time of the rule's promulgation, EPA was unable to conclude that the CSAPR Update fully addressed most of the covered states' good neighbor obligations for the 2008 ozone NAAQS. Id. at 74521. Information available at the time indicated that, even with the CSAPR Update implementation, several downwind receptors were expected to continue having problems attaining and maintaining this NAAQS and that emissions from upwind states were expected to continue to contribute greater than or equal to 1 percent of the NAAQS to these areas during the 2017 ozone season. Id. at 74551-52. Further, EPA could not conclude at that time whether additional EGU and non-EGU reductions implemented on a longer timeframe than 2017 would be needed to address states' good neighbor obligations for this NAAQS.

Additionally, EPA determined it was not feasible to complete an emissions control analysis that may otherwise have been necessary to evaluate full elimination of each state's significant contribution to nonattainment or interference with maintenance and also ensure that emission reductions already quantified in the rule would be achieved by 2017. Id. at 74522. EPA was unable to fully consider both non-EGU ozone season NOX reductions and further EGU reductions that may have been achievable after 2017. Id. at 74521. See section IV.D.3 below.

Thus, EPA also could not make an emission reduction-based conclusion that the CSAPR Update would fully resolve states' good neighbor obligations with respect to the 2008 ozone NAAQS because the reductions evaluated and required by the CSAPR Update were limited in scope (both by technology and sector). As a result of the remaining air quality problems and the limitations Start Printed Page 23066on EPA's analysis, for all but one of the 22 affected states, EPA did not determine in the CSAPR Update that the rule fully addressed those states' downwind air quality impacts under the good neighbor provision for the 2008 ozone NAAQS. Id. at 74521. For one state, Tennessee, EPA determined in the final CSAPR Update that Tennessee's emissions budget fully eliminated the state's significant contribution to downwind nonattainment and interference with maintenance of the 2008 ozone NAAQS because the downwind air quality problems to which the state was linked were projected to be resolved with implementation of the CSAPR Update. Id. at 74552.

c. The CSAPR Close-Out

Following implementation of the CSAPR Update and the approval of Kentucky's SIP (under a court-ordered deadline),[57] on December 21, 2018, EPA issued the CSAPR “Close-Out” to address any good neighbor obligations that remained for the 2008 ozone NAAQS for the 20 remaining states in the CSAPR Update region. See 83 FR 65878 (Dec. 21, 2018). The CSAPR Close-Out made a determination that, based on additional information and analysis, the CSAPR Update fully addressed the remaining 20 affected states' good neighbor obligations for the 2008 ozone NAAQS. In particular, EPA determined that 2023 was an appropriate future analytic year considering relevant attainment dates and the time EPA estimated to be necessary to implement new NOX control technologies at EGUs. Based on EPA's analysis of projected air quality in that year, EPA determined that, for the purposes of addressing good neighbor obligations for the 2008 ozone NAAQS, there would be no remaining nonattainment or maintenance receptors in the eastern U.S. As a result of this determination, EPA found that, with continued implementation of the CSAPR Update, these 20 states would no longer contribute significantly to nonattainment in, or interfere with maintenance by, any other state with respect to the 2008 ozone NAAQS. Id.

d. D.C. Circuit Decisions in Wisconsin v. EPA and New York v. EPA

The CSAPR Update was subject to petitions for judicial review, and the D.C. Circuit issued its opinion in Wisconsin v. EPA on September 13, 2019. 938 F.3d 303. The D.C. Circuit upheld the CSAPR Update in all respects save one: The court concluded that the CSAPR Update was inconsistent with the CAA to the extent that it was partial in nature and did not fully eliminate upwind states' significant contribution to nonattainment or interference with maintenance of the 2008 ozone NAAQS by the downwind states' 2018 Moderate attainment date. Id. at 313.

The court identified three bases for this holding: (1) The D.C. Circuit's prior opinion in North Carolina v. EPA, 531 F.3d 896 (2008), which held, in the context of CAIR, that the good neighbor provision requires states to eliminate significant contribution “consistent with the provisions” of Title I of the CAA, including the attainment dates applicable in downwind areas, 938 F.3d at 314 (citing 531 F.3d at 912); (2) the unreasonableness of EPA's interpretation of the phrase “consistent with the provisions [of Title I]” in the good neighbor provision as allowing for variation from the attainment schedule in CAA section 181 because it would enable significant contribution from upwind states to continue beyond that statutory timeframe, 938 F.3d at 315-18; and (3) the court's finding that the practical obstacles EPA identified regarding why it needed more time to implement a full remedy did not rise to the level of an “impossibility,” id. at 318-20. With respect to the third basis, the court also found EPA must make a higher showing of uncertainty regarding non-EGU point-source NOX mitigation potential before declining to regulate such sources. Id. at 318-20.

However, the court identified flexibilities that EPA retains in administering the good neighbor provision, acknowledging that EPA has latitude in defining which upwind contribution “amounts” count as significant and thus must be abated, permitting EPA to consider, among other things, the magnitude of upwind states' contributions and the cost associated with eliminating them. 938 F.3d at 320. The court further noted that, in certain circumstances, EPA can grant extensions of the attainment deadlines under the Act; for instance, the court cited CAA section 181(a)(5), which allows EPA to grant one-year extensions from attainment dates under certain circumstances. Id. Finally, the court noted that EPA can attempt to show “impossibility.” Id. The court also recognized that the statutory command that compliance with the good neighbor provision must be achieved consistent with Title I might be read, upon a sufficient showing of necessity, to allow some deviation from downwind deadlines, so long as it is rooted in Title I's framework and provides a sufficient level of protection to downwind States. Id.

The court in Wisconsin remanded but did not vacate the CSAPR Update, finding that vacatur of the rule could cause harm to public health and the environment or disrupt the trading program EPA had established and that the obligations imposed by the rule may be appropriate and sustained on remand. Id. at 336. The court also rejected petitioners' request to place EPA on a six-month schedule to address the remand, noting the availability of “mandamus” relief before the D.C. Circuit should EPA fail to “modify the rule in a manner consistent with our opinion.” Id. at 336-37.

On October 1, 2019, in a judgment order, the D.C. Circuit vacated the CSAPR Close-Out on the same grounds that it remanded the Update in Wisconsin. New York v. EPA, 781 Fed. App'x 4 (D.C. Cir. 2019). Because the Close-Out analyzed the year 2023 rather than 2021 (“the next applicable attainment date”) and failed to demonstrate that it was impossible to address significant contribution by the 2021 attainment date, the court found the rule ran afoul of the Wisconsin holding. Id. at 7. “As the EPA acknowledges, the Close-Out Rule `relied upon the same statutory interpretation of the Good Neighbor Provision' that we rejected in Wisconsin. Thus, the Agency's defense of the Close-Out Rule in these cases is foreclosed.” Id. at 6-7 (internal citation omitted). The court left open the possibility that the flexibilities identified in Wisconsin, 938 F.3d at 320, and outlined above, may be available to EPA on remand. Id.

Following Wisconsin and New York, EPA on remand must address good neighbor obligations for the 21 states within the CSAPR Update region for which the Update was only a partial remedy. As explained in the following section, EPA already retains FIP authority as to 20 of these states. In addition, EPA is taking action pursuant to CAA section 110(k)(6) (42 U.S.C. 7410(k)(6)) to find that Kentucky's SIP was approved in error and is thus promulgating a FIP for Kentucky consistent with the obligations for the other remaining CSAPR Update region states.

2. FIP Authority for Each State Covered by the Final Rule

On March 12, 2008, EPA promulgated a revision to the ozone NAAQS, lowering both the primary and Start Printed Page 23067secondary standards to 75 ppb. See National Ambient Air Quality Standards for Ozone, Final Rule, 73 FR 16436 (March 27, 2008). Specifically, the standards require that an area may not exceed 0.075 parts per million (75 ppb) using the 3-year average of the fourth highest 24-hour maximum 8-hour rolling average ozone concentration. These revisions of the NAAQS, in turn, triggered a 3-year deadline for states to submit SIP revisions addressing infrastructure requirements under CAA sections 110(a)(1) and 110(a)(2), including the good neighbor provision. Several events affected the timely application of the good neighbor provision for the 2008 ozone NAAQS, including reconsideration of the 2008 ozone NAAQS and legal developments pertaining to the CSAPR, which created uncertainty surrounding EPA's statutory interpretation and implementation of the good neighbor provision.[58] Notwithstanding these events, EPA ultimately affirmed that states' good neighbor SIPs were due on March 12, 2011.

a. FIP Authority for the CSAPR Update States

EPA subsequently took several actions that triggered EPA's obligation under CAA section 110(c) to promulgate FIPs addressing the good neighbor provision for several states.[59] First, on July 13, 2015, EPA published a rule finding that 24 states failed to make complete submissions that address the requirements of section 110(a)(2)(D)(i)(I) related to the interstate transport of pollution as to the 2008 ozone NAAQS. See 80 FR 39961 (effective August 12, 2015). This finding triggered a two-year deadline for EPA to issue FIPs to address the good neighbor provision for these states by August 12, 2017. The CSAPR Update finalized FIPs for 13 of these states (Alabama, Arkansas, Illinois, Iowa, Kansas, Michigan, Mississippi, Missouri, Oklahoma, Pennsylvania, Tennessee, Virginia, and West Virginia), requiring their participation in a NOX trading program. EPA also determined in the CSAPR Update that the Agency had no further FIP obligation as to nine additional states identified in the finding of failure to submit because these states did not contribute significantly to nonattainment in, or interfere with maintenance by, any other state with respect to the 2008 ozone NAAQS. See 81 FR 74506.[60] 61 On June 15, 2016, and July 20, 2016, EPA published additional rules finding that Maryland and New Jersey, respectively, also failed to submit transport SIPs for the 2008 ozone NAAQS. See 81 FR 38963 (June 15, 2016) (New Jersey, effective July 15, 2016); 81 FR 47040 (July 20, 2016) (Maryland, effective August 19, 2016). The finding actions triggered two-year deadlines for EPA to issue FIPs to address the good neighbor provision for Maryland by August 19, 2018, and for New Jersey by July 15, 2018. The CSAPR Update also finalized FIPs for these two states.

In addition to these findings, EPA finalized disapproval or partial disapproval actions for good neighbor SIPs submitted by Indiana, Kentucky, Louisiana, New York, Ohio, Texas, and Wisconsin.[62] These disapprovals triggered EPA's obligation to promulgate FIPs to implement the requirements of the good neighbor provision for those states within two years of the effective date of each disapproval or, in the case of Kentucky, within two years of the issuance of the judgment in a subsequent Supreme Court decision.[63] EPA promulgated FIPs in the CSAPR Update for each of these states.

As discussed in more detail above in section IV.C.1, in issuing the CSAPR Update, EPA could not determine that it had entirely addressed EPA's outstanding CAA obligations to implement the good neighbor provision with respect to the 2008 ozone NAAQS for 21 of 22 states covered by that rule. Accordingly, the CSAPR Update did not fully satisfy EPA's obligation under CAA section 110(c) to address the good neighbor provision requirements for those states by approving SIPs, issuing FIPs, or some combination of those two actions. EPA found that the CSAPR Update FIPs fully addressed the good neighbor provision for the 2008 ozone NAAQS only with respect to Tennessee.

b. Correction of EPA's Determination Regarding Kentucky's SIP Revision and Its Impact on EPA's FIP Authority for Kentucky

After promulgating the CSAPR Update and before promulgating the CSAPR Close-Out, EPA approved a SIP submission from Kentucky resolving the Commonwealth's good neighbor obligations for the 2008 ozone NAAQS based on a demonstration that no further emission reductions were needed from Kentucky with the CSAPR Update FIP for Kentucky in place. See 83 FR 33730 (July 17, 2018). The action was separate from the CSAPR Close-Out because it was taken in response to a May 23, 2017 order from the U.S. District Court for the Northern District of California requiring EPA to take a final action fully addressing the good neighbor obligation for the 2008 ozone NAAQS for Kentucky by June 30, 2018.[64] EPA was obligated to address the outstanding obligation by either approving a SIP revision submitted by Kentucky or promulgating a FIP to address any remaining obligation.[65]

On May 10, 2018, Kentucky submitted a final SIP revision to EPA, on which the Agency finalized approval consistent with the court-ordered deadline. See 83 FR 33730. The Kentucky SIP revision that EPA approved relied on the reductions from the CSAPR Update FIP for Kentucky and provided a technical analysis, including emission projections and air quality modeling for 2023, showing that Start Printed Page 23068with the CSAPR Update level of reductions, the receptors to which Kentucky was linked were attaining and maintaining the 2008 ozone NAAQS in 2023. This allowed EPA to conclude that Kentucky did not have any further obligation for the 2008 ozone NAAQS, and EPA approved the SIP revision. The SIP revision from Kentucky was an analytical demonstration only, and it did not replace the CSAPR Update FIP; rather, the CSAPR Update FIP was left in place for Kentucky and was relied on in the state's demonstration.

The approval relied on the same rationale and technical analysis—including the use of a 2023 analytic year—that was eventually used for the other CSAPR Update FIP states in the CSAPR Close-Out. EPA's approval stated:

“no additional emission reductions are necessary to address the good neighbor provision for the 2008 ozone NAAQS beyond those required by the Cross-State Air Pollution Rule Update (CSAPR Update) federal implementation plan (FIP). Accordingly, EPA is approving Kentucky's submission because it partially addresses the requirements of the good neighbor provision for the 2008 ozone NAAQS, and it resolves any obligation remaining under the good neighbor provision after promulgation of the CSAPR Update FIP. The approval of Kentucky's SIP submission and the CSAPR Update FIP, together, fully address the requirements of the good neighbor provision for the 2008 ozone NAAQS for Kentucky.”

83 FR 33730.

Subsequent to EPA's approval of the Kentucky SIP submission, EPA issued the CSAPR Close-Out, which concluded that, based on essentially the same analysis used for Kentucky, none of the other 20 CSAPR Update states had further good neighbor obligations to address the 2008 8-hour ozone NAAQS. In the Fall of 2019, the D.C. Circuit issued the Wisconsin and New York decisions remanding the CSAPR Update Rule and vacating the CSAPR Close-Out (see section IV.C.1.d.).

Kentucky's CSAPR Update FIP, which Kentucky relied on (and did not replace) in its SIP revision, is part of the CSAPR Update remand, and EPA must address it in this action. Further, the D.C. Circuit's review of the CSAPR Close-Out found fault with, and vacated, the same rationale for other states that EPA had used to approve Kentucky's SIP submission in June 2018.

Therefore, in light of the remand of Kentucky's CSAPR Update FIP in Wisconsin and vacatur of the CSAPR Close-Out in New York, EPA is determining in this final action that its approval of Kentucky's SIP revision as fully resolving the state's 2008 ozone NAAQS good neighbor obligations was in error. Section 110(k)(6) of the CAA (42 U.S.C. 7410(k)(6)) gives the Administrator authority, without any further submission from a state, to revise certain prior actions, including actions to approve SIPs, upon determining that those actions were in error. The court's remand of the partial FIP for Kentucky in Wisconsin and the vacatur of EPA's conclusions for states identically situated to Kentucky in the CSAPR Close-Out means that EPA's approval of Kentucky's SIP was in error. EPA is compelled on remand to act consistently with the court's opinion and has reassessed Kentucky's good neighbor obligations under the 2008 ozone NAAQS here. In doing so, EPA's analysis identifies an additional emission reduction obligation for Kentucky. Therefore, EPA is correcting the error in Kentucky's SIP approval through this final rulemaking, as allowed by the CAA when a prior SIP approval was in error. This error correction under CAA section 110(k)(6) revises the approval of Kentucky's SIP to a disapproval and rescinds any statements that the SIP submission fully addresses the requirements of the good neighbor provision for the 2008 ozone NAAQS for Kentucky. The Kentucky approval relied on the same analysis that the D.C. Circuit determined to be unlawful in the CSAPR Close-Out: It only addressed conditions in 2023, ignoring the 2021 attainment date without a showing of impossibility or necessity in doing so. Kentucky's remanded partial FIP has been reassessed in this action, consistent with EPA's methodology to address the other 20 states with remanded CSAPR Update FIPs, and consistent with the D.C. Circuit's direction in Wisconsin and New York. As discussed in greater detail in the sections that follow, EPA is determining that there are additional emission reductions that are required for Kentucky to fully satisfy its good neighbor obligations for the 2008 ozone NAAQS. The analysis on which EPA reaches this conclusion for Kentucky is the same, regionally consistent analytical framework on which the Agency is taking action for all of the other CSAPR Update states with remanded FIPs.

Comment: The Agency received several comments regarding its error correction for Kentucky from the state and from sources in Kentucky. The commenters generally disagreed with EPA's use of CAA section 110(k)(6) to correct the error in the SIP approval based on the Wisconsin and New York decisions. Commenters did not agree that the court decisions are applicable to the Kentucky action or that EPA had any other basis to determine that Kentucky has outstanding good neighbor obligations under the 2008 ozone NAAQS. Two commenters also argued that EPA overestimated emissions from Kentucky in the modeling released with the proposed rule of this action.

Response: EPA disagrees that there is no basis to correct its error in approving Kentucky's SIP revision or to find that Kentucky has outstanding good neighbor obligations under the 2008 ozone NAAQS. Wisconsin and New York require the state or EPA to analyze the interstate impacts of Kentucky's emissions by the 2021 Serious attainment date. The Kentucky SIP approval is based on analysis of the 2023 ozone season. Further, the Kentucky SIP approval relies on reductions achieved from Kentucky's CSAPR Update FIP, which was remanded by Wisconsin. The information provided by commenters on emissions from Kentucky was already reflected in EPA's modeling and did not present information with regard to Kentucky that changed EPA's 2021 analysis, which shows Kentucky has further good neighbor obligations under the 2008 ozone NAAQS. Comments related to EPA's technical basis for concluding that Kentucky has further obligations, including comments regarding alleged additional emission reductions achieved by Kentucky sources, are addressed in the RTC document.

c. CSAPR Update SIP Revisions That Do Not Affect FIP Authority

Subsequent to the promulgation of the CSAPR Update, EPA approved SIPs fully replacing the CSAPR Update FIPs for Alabama, Indiana, and Missouri.[66] In those SIP approvals and consistent with the conclusions of the CSAPR Update, EPA found that the SIPs partially satisfy Alabama's, Indiana's, and Missouri's good neighbor obligations for the 2008 ozone NAAQS. Thus, EPA continues to have an obligation to fully address good neighbor requirements for the 2008 ozone NAAQS with respect to Alabama and Missouri, stemming from the July 13, 2015, findings of failure to submit, and Indiana, due to the June 15, 2016, disapproval of the state's good neighbor SIP. See 80 FR 39961; 81 FR 38957. Other states have also submitted 2008 ozone NAAQS good neighbor SIPs or SIPs to replace their CSAPR FIPs, some of which EPA has approved and some Start Printed Page 23069of which still remain pending. These circumstances do not affect the scope or basis for this rulemaking.

d. Summary of Authority for FIPs for This Action

Table IV.C-1 summarizes the statutory deadline for EPA to address its FIP obligation under CAA section 110(c) and the event that activated EPA's obligation for each of the 21 CSAPR Update states that are the subject of this final action. For more information regarding the actions triggering EPA's FIP obligation and EPA's action on SIPs addressing the good neighbor provision for the 2008 ozone NAAQS, see the memorandum, “Final Action, Status of 110(a)(2)(D)(i)(I) SIPs for the 2008 Ozone NAAQS,” in the docket for this action.

Table IV.C-1—Actions That Activated EPA's Statutory FIP Deadlines

StateType of action (Federal Register citation, publication date)Statutory FIP deadline 
AlabamaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
ArkansasFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
IllinoisFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
IndianaSIP disapproval (81 FR 38957, 6/15/2016)7/15/2018
IowaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
KansasFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
KentuckySIP disapproval (78 FR 14681, 3/7/2013)6/2/2016
LouisianaSIP disapproval (81 FR 53308, 8/12/2016)9/12/2018
MarylandFinding of Failure to Submit (81 FR 47040, 7/20/2016)8/19/2018
MichiganFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
MississippiFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
MissouriFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
New JerseyFinding of Failure to Submit (81 FR 38963, 6/15/2016)7/15/2018
New YorkSIP disapproval (81 FR 58849, 8/26/2016)9/26/2018
OhioSIP disapproval (81 FR 38957, 6/15/2016)7/15/2018
OklahomaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
PennsylvaniaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
TexasSIP disapproval (81 FR 53284, 8/12/2016)9/12/2018
VirginiaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
West VirginiaFinding of Failure to Submit (80 FR 39961, 7/13/2015)8/12/2017
WisconsinPartial SIP disapproval as to prong 2 (81 FR 53309, 8/12/2016)9/12/2018
 For states other than Kentucky, the FIP deadline is two years from the effective date of the SIP disapproval or Finding of Failure to Submit, which generally trails the publication date by 30 days. For Kentucky, the FIP deadline is two years after the issuance of the Supreme Court's judgment in EPA v. EME Homer City Generation, L.P., 572 U.S. 489 (2014). See supra note 63.

3. The 4-Step Good Neighbor Framework

The CSAPR and the subsequent CSAPR Update, building on EPA's prior methodologies in the NOX SIP Call and CAIR, established a 4-step process to address the requirements of the good neighbor provision.[67] In this final action to address the remand of the CSAPR Update, EPA follows the same steps. These steps are: (1) Identifying downwind receptors that are expected to have problems attaining or maintaining the NAAQS; (2) determining which upwind states contribute to these identified problems in amounts sufficient to “link” them to the downwind air quality problems; (3) for states linked to downwind air quality problems, identifying upwind emissions that significantly contribute to downwind nonattainment or interfere with downwind maintenance of the NAAQS; and (4) for states that are found to have emissions that significantly contribute to nonattainment or interfere with maintenance of the NAAQS downwind, implementing the necessary emission reductions through enforceable measures.

Step 1—In the CSAPR, downwind air quality problems were assessed using modeled future air quality concentrations for a year aligned with attainment deadlines for the NAAQS considered in that rulemaking. The assessment of future air quality conditions generally accounts for on-the-books emission reductions and the most up-to-date forecast of future emissions in the absence of the transport policy being evaluated (i.e., base case conditions). The locations of downwind air quality problems are identified as those with receptors that are projected to be unable to attain (i.e., nonattainment receptor) or maintain (i.e., maintenance receptor) the NAAQS. In the CSAPR Update, EPA also considered current monitored air quality data to further inform the projected identification of downwind air quality problems. These same considerations are included for this final rule. EPA is not reopening the definition of nonattainment and maintenance receptors promulgated in the CSAPR Update. Further details and application of step 1 for this rule are described in section V.

Step 2—The CSAPR and the CSAPR Update used a screening threshold of 1 percent of the NAAQS to identify upwind states that were “linked” to downwind air pollution problems. States with contributions greater than or equal to the threshold for at least one downwind problem receptor (i.e., nonattainment or maintenance receptor identified in step 1) were identified as needing further evaluation for actions to address transport if their air quality was impacted.[68] EPA evaluated a given state's contribution based on the average relative downwind impact calculated over multiple days.[69] States whose air Start Printed Page 23070quality impacts to all downwind problem receptors were below this threshold did not require further evaluation for actions to address transport—that is, these states were determined to not contribute to downwind air quality problems and therefore had no emission reduction obligations under the good neighbor provision. EPA has used this threshold because a notable portion of the transport problem in the eastern half of the United States can result from relatively small contributions from a number of upwind states. Use of the 1 percent threshold for the CSAPR is discussed in the preambles to the proposed and final CSAPR rules. See 75 FR 45237 (Aug. 2, 2010); 76 FR 48238 (Aug. 8, 2011). The same metric is discussed in the CSAPR Update Rule. See 81 FR 74538. While EPA has updated its air quality data for determining contributions, the Agency is not reopening the use of the 1 percent threshold in this action to address the remand of the CSAPR Update. Application of step 2 for this rule is described in section V.

Step 3—For states that are linked in step 2 to downwind air quality problems, the CSAPR and the CSAPR Update evaluated NOX reductions that were available in upwind states by applying a uniform control stringency (represented by a cost per ton of NOX reduced) to entities in these states. EPA evaluated multiple factors, including NOX reduction potential, cost, and downwind air quality improvements available at several control stringencies in the multi-factor test. This evaluation quantified the magnitude of emissions that significantly contribute to nonattainment or interfere with maintenance of a NAAQS downwind and apportioned upwind responsibility among linked states, an approach upheld by the U.S. Supreme Court in EPA v. EME Homer City.[70] In this action, EPA applies this same approach to identify NOX emission reductions necessary to address significant contribution for the 2008 ozone NAAQS.

In EME Homer City, the Supreme Court held that “EPA cannot require a State to reduce its output of pollution by more than is necessary to achieve attainment in every downwind State or at odds with the one-percent threshold the Agency has set.” 572 U.S. at 521. The Court acknowledged that “instances of `over-control' in particular downwind locations may be incidental to reductions necessary to ensure attainment elsewhere.” Id. at 492.

“Because individual upwind States often `contribute significantly' to nonattainment in multiple downwind locations, the emissions reductions required to bring one linked downwind State into attainment may well be large enough to push other linked downwind States over the attainment line. As the Good Neighbor Provision seeks attainment in every downwind State, however, exceeding attainment in one State cannot rank as `over-control' unless unnecessary to achieving attainment in any downwind State. Only reductions unnecessary to downwind attainment anywhere fall outside the Agency's statutory authority.”

Id. at 522 (footnotes excluded).

The Court further explained that “while EPA has a statutory duty to avoid over-control, the Agency also has a statutory obligation to avoid `under-control,' i.e., to maximize achievement of attainment downwind.” Id. at 523. Therefore, in the CSAPR Update, EPA evaluated possible over-control by considering whether an upwind state is linked solely to downwind air quality problems that can be resolved at a lower cost threshold, or if upwind states would reduce their emissions at a lower cost threshold to the extent that they would no longer meet or exceed the 1 percent air quality contribution threshold. See 81 FR at 74551-52. This evaluation of cost, NOX reductions, and air quality improvements, including consideration of potential over-control, results in EPA's determination of upwind emissions that significantly contribute to nonattainment or interfere with maintenance of the NAAQS downwind and should therefore be eliminated. This allows EPA to then determine an enforceable emissions limit (often embodied in the form of an emissions budget) for the covered sources. Emissions budgets are the remaining allowable emissions after the elimination of emissions identified as significantly contributing to nonattainment or interfering with maintenance of the standard downwind.

In both the CSAPR and the CSAPR Update, EPA focused its step 3 analysis on EGUs. In the CSAPR Update, EPA did not quantify non-EGU stationary source emission reductions to address interstate ozone transport for the 2008 ozone NAAQS for two reasons. First, EPA explained that there was greater uncertainty in EPA's assessment of non-EGU NOX mitigation potential, and that more time would be required for states and EPA to improve non-EGU point source data and pollution control assumptions before it could develop emission reduction obligations based on that data. See 81 FR 74542. Second, EPA explained that it did not believe that significant, certain, and meaningful non-EGU NOX reduction was in fact feasible for the 2017 ozone season. Id. In Wisconsin, the D.C. Circuit found that the practical obstacles EPA identified with respect to its evaluation of non-EGUs did not rise to the level of an “impossibility,” 938 F.3d at 318-20. The court also found that EPA must make a higher showing of uncertainty regarding non-EGU point-source NOX mitigation potential before declining to regulate such sources on such a basis, id. Therefore, as discussed in more detail in section VI, in this final action on remand from Wisconsin, EPA has included all major stationary source sectors in the linked upwind states in its “significant contribution” analysis at step 3 of the 4-step framework.

Step 4—the CSAPR and the CSAPR Update established interstate trading programs to implement the necessary emission reductions. Each state subject to the program is assigned an emissions budget for the covered sources. Emissions allowances are allocated to units covered by the trading program, and the covered units then surrender allowances after the close of each control period in an amount equal to their ozone season EGU NOX emissions. Emissions allowances are allocated to units covered by the respective trading program, and the covered units then surrender allowances after the close of each control period in an amount equal to their ozone season EGU NOX emissions.

All of EPA's trading programs established under the good neighbor provision allow for interstate trading. However, in order to ensure that each state achieves reductions proportional to the level of their significant contribution, beginning with the CSAPR, EPA established “assurance levels” set as percentage of each state's budget (e.g., 121 percent) above which emissions from sources in that state become subject to a higher “penalty” surrender ratio. These assurance levels are designed to allow for a certain level of year-to-year variability within power sector emissions to account for fluctuations in demand and EGU Start Printed Page 23071operations. The levels are therefore set by determining a “variability limit,” calculated based on an analysis of the historical level of variability in EGU operations.

Thus, both the CSAPR and the CSAPR Update set assurance levels equal to the sum of each state's emissions budget plus its variability limit. The CSAPR and the CSAPR Update included assurance provisions to limit state emissions to levels below 121 percent of the state's ozone season NOx emissions budget by requiring additional allowance surrenders in the instance that emissions in the state exceed this level. This limit on the degree to which a state's emissions can exceed its budget is responsive to previous court decisions (see discussion in section VII.C.2 of this preamble) and was not part of the CSAPR Update aspects remanded to EPA in Wisconsin. EPA is applying the same variability limits and assurance provisions in this rule.[71] Implementation using a trading program is further described in section VII.

EPA received several comments related to its overall approach in this rulemaking. These comments related to the following topics: (1) Whether this rule remains only a partial remedy in terms of both the amount of emission reductions achieved and the timing of implementation; (2) whether any additional EGU emission reductions relative to the CSAPR Update are permissible in light of the CSAPR Update record and the scope of the D.C. Circuit's decision in Wisconsin; and (3) EPA's use of cost to define significant contribution. Other comments on EPA's overall approach in this action are addressed in the RTC document.

Comment: Numerous commenters asserted that despite EPA purporting to fully address the covered states' good neighbor obligations, the rule remains only a partial solution, and allows upwind states' significant contribution to nonattainment and interference with maintenance of the 2008 ozone NAAQS to continue past the next attainment date. One commenter asserts that this rule will “hinder” attainment of the 2015 ozone NAAQS in downwind states. Many commenters claim that the rule is insufficient to ensure downwind attainment of the NAAQS. The commenters question EPA's application of the 4-step framework and disagree with the Agency's conclusions drawn from that analysis, particularly with respect to the EPA's determinations at step 3 and the emissions controls adopted at step 4. Some commenters also challenge the legal basis for the selection of the 2021 analytic year, as opposed to 2020, and whether EPA has met the requirement to obtain reductions “as expeditiously as practicable” or otherwise complied with the holdings in Wisconsin and New York to eliminate significant contribution on par with the relevant downwind attainment deadlines. See, e.g. Wisconsin, 938 F.3d at 315. Some argue that EPA depends on claims of technical infeasibility or scientific uncertainty and flawed cost effectiveness considerations in not requiring more emission reductions on a shorter timeframe. Others believe the implementation timeframe of this rule to be a phased plan in direct conflict with Wisconsin and New York. One commenter concludes there is a “mismatch” between EPA's 4-step framework's multi-factor test at step 3 and the implementation timeframes in this rule. They also argue that EPA should consider the cost of RACT in downwind states when analyzing the maximized cost effectiveness of controls in upwind states. Several commenters also brought attention to the length of time between when 2008 ozone NAAQS good neighbor SIPs were initially due and the proposed rule in October 2020.

Response: This rule is a full remedy for the good neighbor provision for the covered upwind states for the 2008 ozone NAAQS based on EPA's analysis. The good neighbor provision does not obligate upwind states to fully resolve a downwind nonattainment or maintenance problem. CAA section 110(a)(2)(D)(i)(I) only requires that upwind states prohibit those emissions that “contribute significantly to nonattainment” or “interfere with maintenance of the NAAQS.” As such, the objective of the good neighbor provision is the elimination of upwind significant contribution or interference with maintenance. It does not require that the upwind states bear the full burden of bringing downwind states into attainment. Ultimate achievement of the NAAQS downwind is accomplished through the larger framework of the CAA, including under sections 110, 181, 182 and other provisions to attain the NAAQS. Thus, in this action, EPA must determine what amount of upwind contribution is significant (or interferes with maintenance) and require elimination of that significant contribution while avoiding overcontrol or undercontrol. EPA v. EME Homer City Generation, L.P., 572 U.S. 489, 521-23 (2014).

Further, it is not correct to say that good neighbor obligations can only be found to be fully addressed when there is no longer any remaining air quality problem at the downwind receptors. Indeed, the Supreme Court recognized in EME Homer City, 572 U.S. at 521-22, that under the framework EPA has adopted, EPA could not require a state to further reduce its emissions once it is at or below the 1 percent contribution threshold at all receptors. The aim of the good neighbor provision is to eliminate significant contribution to nonattainment and interference with maintenance, not to achieve final attainment at the downwind receptor. Further, in upholding EPA's approach to defining and allocating upwind responsibility in the CSAPR, the Court in EME Homer City recognized the discretion EPA has in defining what constitutes “significant” contribution, and did not hold that obligations on upwind states must be imposed to “maximize” downwind attainment without consideration of any other factors. Accord Wisconsin, 938 F.3d at 320 (recognizing EPA's discretion to interpret “significant contribution”).

The comments do not establish a basis for asserting that EPA's approach to defining significant contribution or interference with maintenance is unlawful or unreasonable. They do not explain what is meant by “excessive amounts of ozone pollution,” “excessive upwind contributions,” “sufficient emission reductions,” or “sufficient upwind reductions.” These comments do not inform how EPA should define significant contribution nor do they recognize that EPA has discretion to define significant contribution. The D.C. Circuit first upheld the validity of using cost as part of the method for determining “significance” in Michigan v. EPA, 213 F.3d 663, 675-79 (D.C. Cir. 2000). The Supreme Court upheld that same approach in EPA v. EME Homer City Generation, L.P., 572 U.S. 489, 512-20 (2014) (“Eliminating those amounts that can cost‐effectively be reduced is an efficient and equitable solution to the allocation problem the Good Neighbor Provision requires the Agency to address.”). EPA applied this approach again in the CSAPR Update, its first action to address good neighbor obligations under the 2008 ozone NAAQS. And while that action only provided a partial remedy, no party in Wisconsin challenged as a general matter EPA's ability to use cost-effectiveness in determining and allocating upwind responsibility. Wisconsin and New York recognized EPA's discretion to define significant Start Printed Page 23072contribution. Wisconsin v. EPA, 938 F.3d 303, 319-20 (D.C. Cir. 2019) (“EPA, though, possesses a measure of latitude in defining which upwind contribution `amounts' count as `significant[ ]' and thus must be abated.”); New York v. Envtl. Prot. Agency, 781 F. App'x 4, 7 (D.C. Cir. 2019) (“[I]n determining what constitutes a significant contribution to downwind nonattainment, the agency can consider the amount of upwind states' contributions and the cost of abating them.”).

With respect to the timing of when such reductions must be achieved, EPA agrees that “as expeditiously as practicable” is the first-order statutory directive. See CAA section 181(a)(1); Wisconsin, 938 F.3d at 313. EPA's approach in this rule, after determining significant contribution, implements all reductions that EPA identified as possible by the 2021 attainment date, and requires additional reductions of EGUs in later ozone seasons to the extent not possible by that date to fully eliminate significant contribution. In this case, implementing reductions any faster than the 2021 ozone season is impossible because 2020 is in the past. Commenters are incorrect to assert that EPA has unlawfully failed to require all necessary reductions by the 2021 attainment date. EPA has required those reductions that it has determined are possible by that date; EPA has also made a determination that additional reductions that are only possible after that date are nonetheless necessary to eliminate significant contribution or interference with maintenance, as EPA has interpreted those terms, and is requiring those later reductions as expeditiously as practicable. Achieving necessary reductions past the next attainment date when EPA finds it is impossible to do so beforehand is consistent with the statute and prior caselaw. Wisconsin and New York recognized these flexibilities available to EPA in acknowledging that for reasons of necessity or impossibility, EPA may deviate from the attainment schedule for downwind areas established in the Act. Wisconsin v. EPA, 938 F.3d 303, 320 (D.C. Cir. 2019). New York v. EPA, 781 F. App'x 4, 7 (D.C. Cir. 2019). Indeed, these commenters are not asserting that EPA lacks authority to require reductions beyond the attainment date, only that EPA should have required the reductions by that date. But these comments fail to establish a technical or evidentiary basis to overturn EPA's judgment that such additional reductions are not in fact possible by the 2021 attainment date.

EPA disagrees that Wisconsin held that it must address good neighbor obligations by the full ozone season prior to the attainment date (i.e., here, 2020). The decision recognized that the agency must fully address good neighbor obligations (to the extent EPA determines possible) by the attainment date itself. 938 F.3d at 315. EPA's practice of addressing obligations by the full ozone season prior to the attainment date, while not mandated by statute or caselaw, continues to make good policy sense, because it assists downwind areas with improved three-year design values [72] used in determining whether attainment has been achieved. However, in this instance, as one commenter correctly notes, reductions in 2020 are not possible since this rule was not proposed until after the 2020 ozone season. EPA nonetheless can still meet the legal mandate to achieve those reductions that are possible by the 2021 attainment date.

Further, EPA is not relying on “scientific uncertainty” as a justification for not requiring reductions earlier. As explained elsewhere in this record, EPA has determined the amount of time needed for installation and operation of various control strategies. With respect to the optimization of existing SNCR controls, EPA notes that it is requiring that strategy as reflected in the final budgets by the 2021 attainment date, as explained in sections VI.B.1, C.1, and D.1.

EPA defined significant contribution in this rule based on an assessment of control alternatives under the 4-step good neighbor framework's step 3 multi-factor test. EPA's determination of what controls to require and when they can first be implemented are based on EPA's technical evaluation and application of the third step multi-factor analysis in the 4-step framework. The only “mismatch” that one commenter identified at the third step is no mismatch at all; it is simply the reality that some of the controls that EPA is requiring in this rule cannot be installed before the 2021 ozone season, and some controls that EPA assessed cannot be installed and operational before air quality problems are projected to resolve under the 2008 ozone NAAQS (i.e., by the 2025 ozone season). These comments have not explained how EPA's evaluation of control options under that test was arbitrary or capricious.

The reasons for alleged past delays in implementing ozone transport obligations is out of the scope of this action on remand. However, EPA notes that the time it has taken to get reductions in place to address interstate ozone transport is due to multiple factors, including past judicial stays of major transport rules such as the NOX SIP Call and the CSAPR. In addition, EPA had made a determination in the CSAPR Close-out that it had fully addressed good neighbor obligations; it was not until the D.C. Circuit ruled in Wisconsin that the basis for this conclusion was revealed to be insufficient. The CSAPR Update has and continues to achieve upwind reductions for the 2008 ozone NAAQS. As explained elsewhere in the preamble to this action, EPA now finds it to be a full remedy for nine upwind states.

Comment: Several commenters said that the CSAPR Update was already a complete remedy with regard to the EGU sector. One commenter described EPA's response to the remand as “unreasonable” and its re-application of the 4-step framework as “erroneous.” Other commenters opined that EPA has no legal basis to require short-term EGU controls under the Wisconsin remand. In their opinion, Wisconsin found that the CSAPR Update fully eliminated significant contribution from EGUs, which they supported by quoting portions of the decision. They asserted that Wisconsin only authorized EPA to search for emission reductions from non-EGUs and to narrowly reconsider the CSAPR Update in terms of the statutory downwind attainment dates.

Response: The commenters are incorrect that EPA lacks a legal basis to re-assess and fully address good neighbor obligations for the covered states under the Wisconsin remand. As an initial matter, the CSAPR Update was, by EPA's own admissions, a partial rule. See 81 FR at 74521-22. The court's analysis upholding the portions of the rule in Wisconsin cited by these commenters was against a backdrop that the rule was only partial in nature. See, e.g., 938 F.3d at 327. Wisconsin required EPA to provide a complete remedy by the next applicable attainment date. This was confirmed in the New York decision vacating the CSAPR Close-out. The D.C. Circuit found that rule violated the holding in Wisconsin by failing to analyze the 2021 analytic year without a sufficient showing of impossibility or necessity. To the extent that EPA had attempted to fully address the relevant obligations in the CSAPR Close-Out Rule, that action has been vacated. Therefore, on remand, EPA not only needs to use a different analytic year to Start Printed Page 23073inform its analysis under the 4-step framework, but it also needs to apply that framework in order to determine what, if any, obligations must be addressed, and what emission reductions must be required.

EPA disagrees that Wisconsin prevents requiring additional necessary controls on EGUs. As stated in the preamble to the CSAPR Update, EPA did not view the CSAPR Update as necessarily fully eliminating significant contribution from EGUs. See 81 FR 74522. Wisconsin recognized that EPA anticipated “further EGU reductions that are achievable after 2017” may be necessary to completely eliminate significant contribution or interference with maintenance for the 2008 ozone NAAQS. Wisconsin v. EPA, 938 F.3d 303 (D.C. Cir. 2019) (quoting 81 FR 74522). In the present action, evaluation of a full remedy in accordance with Wisconsin under the 4-step framework, and particularly the step 3 multi-factor test, establishes that additional reductions from EGUs should be required in 12 of the states currently subject to the CSAPR Update. For nine other states, their continued obligations under the CSAPR Update satisfy their good neighbor obligations for the 2008 ozone NAAQS. That same analysis shows that reductions from non-EGUs are not justified under the same test.

Comment: Some commenters argued that EPA's use of cost in defining significant contribution has no statutory basis and is contrary to NAAQS attainment planning caselaw and the Supreme Court's holding in EPA v. EME Homer City Generation, L.P., 572 U.S. 489 (2014) (“EME Homer City”), because it does not result in sufficient emission reduction for attainment and maintenance of the NAAQS. The commenter also said that even if EPA could use cost as a basis for defining significant contribution for non-attainment, the Agency could not do so for interference with maintenance. Another commenter described EPA's proposed cost threshold of $1,600 per ton as “arbitrary” and inconsistent with the CAA and EME Homer City, as this cost threshold is insufficient to enable downwind states reach attainment or maintenance. Further, commenters argued, EPA's use of cost-effectiveness as a metric at step 3 fails to identify what the ultimate goal should be, as cost-effectiveness can only be used to evaluate which way to best achieve a goal. One commenter argued that EPA should require upwind reductions so long as the downwind benefit of such reductions continues to outweigh their cost.

Response: The approach used here is materially the same approach the Agency applied in the NOX SIP Call, the CSAPR, and in the CSAPR Update. These comments essentially seek to relitigate EME Homer City, as well as the D.C. Circuit's prior opinion in Michigan v. EPA, 213 F.3d 663 (D.C. Cir. 2000). Contrary to the commenters' interpretation, EME Homer City allowed the use of cost both to define and to allocate upwind state responsibility. 572 U.S. 489, 518-520 (2014) (“The Agency, tasked with choosing which among equal “amounts” to eliminate, has chosen sensibly to reduce the amount easier, i.e., less costly, to eradicate.”). Notably, in the CSAPR rulemaking, EPA used cost as part of a multi-factor effectiveness metric in the multi-factor test to determine the “amount” of upwind contribution that is “significant” in a very similar manner as EPA did in the CSAPR Update and now here in this action on remand. See 76 FR 48208, 48248-51 (Aug. 8, 2011). In the NOX SIP Call, EPA took a similar approach. See 213 F.3d at 675 (“Although the dividing line was a very low threshold of contribution, in the end EPA's rule called for termination of only a subset of each state's contribution. EPA decided that the 23 `significant contributors' need only reduce their ozone by the amount achievable with `highly cost-effective controls.' ”) (emphasis added) (citing 63 FR at 57403).

Commenters fail to identify why an alternative method for determining “contribution” is compelled by the statute, or that EPA's approach is unlawful, arbitrary, or capricious. Contrary to these commenters' assertion, the good neighbor provision does not contemplate that an upwind state's obligation can only ever be resolved once a downwind receptor is fully in attainment. The Supreme Court recognized in EME Homer City that the 1 percent contribution threshold used at step 2 must necessarily be a stopping point in EPA's analysis because a state that contributed less than that would not be assessed for reductions at step 3 in the first place. 572 U.S. at 521. The Supreme Court in EME Homer City recognized that the problem of defining “significant contribution” in the context of a regional pollutant like ozone is inherently extremely complex. Id. at 514. The Court found that using cost (and specifically, a uniform cost-effectiveness threshold) to allocate the reduction obligation was both equitable and efficient. Id. at 519.

Further, the case law on barring use of cost considerations in the attainment planning context cited by one commenter is inapplicable. EPA has discretion to interpret significant contribution, as recognized by Wisconsin and New York. Wisconsin v. EPA, 938 F.3d 303, 319-20 (D.C. Cir. 2019) (“EPA, though, possesses a measure of latitude in defining which upwind contribution `amounts' count as `significant[ ]' and thus must be abated.”). New York v. Envtl. Prot. Agency, 781 F. App'x 4, 7 (D.C. Cir. 2019) (“[I]n determining what constitutes a significant contribution to downwind nonattainment, the agency can consider the amount of upwind states' contributions and the cost of abating them.”).

The comment that cost effectiveness does not provide an adequate basis for EPA to select the correct level of stringency misapprehends the full scope of the step 3 multi-factor analysis EPA applies in the 4-step framework. EPA's analysis at step 3 additionally considers the total amount of reductions to be achieved by a control stringency as well as the effect on air quality at downwind receptors. EPA also must take into consideration the minimum amount of time needed for controls to be installed and operational, because if an air quality problem is no longer present by the time controls could be operational, then there is no need for those controls to be required. See 572 U.S. at 521. Thus, it is not just the relative cost effectiveness of a control stringency but its ultimate effect on a downwind problem that informs EPA's determination of “significance.” [73]

Start Printed Page 23074

The uniform control stringency selected in this rule for EGUs compares favorably with prior transport rulemakings in terms of cost-effectiveness, overall cost, total reductions, and downwind benefits. By contrast, when EPA analyzed the best available current data on non-EGUs for potential control, EPA's analysis showed that at a comparable cost level ($2,000/ton—on a weighted average basis, rather than the 90th percentile value used as a representative marginal cost used for EGU SCR optimization, far fewer NOX emission reductions were available and their corresponding effect on downwind receptors was much smaller, on the order of a few hundredths of a ppb.

Regarding the comment that EPA has failed to give independent effect to the requirement to prohibit emissions that interfere with maintenance of the NAAQS in other states (i.e., prong 2): EPA gives effect to prong 2 through identifying receptors that may have trouble attaining the NAAQS under varying air quality and meteorological conditions. EME Homer City upheld EPA's approach to using cost to determine “amounts” with respect to both prong 1 and 2, and this is settled law. EPA v. EME Homer City Generation, 572 U.S. at 518-520. EPA's use of the term “significant contribution” in its analysis at the third step of the 4-step framework is applied for both prongs 1 and 2. This approach to giving effect to the “interfere with maintenance” prong has been upheld twice by the D.C. Circuit. See EME Homer City, 795 F.3d at 136; Wisconsin, 938 F.3d at 325-27. In effect, EPA's determination of what level of upwind contribution constitutes “interference” with a maintenance receptor is the same determination as what constitutes “significant contribution” for a nonattainment receptor. Nonetheless, this continues to give independent effect to prong 2 because EPA applies a broader definition for identifying maintenance receptors, which accounts for the possibility of problems maintaining the NAAQS under realistic potential future conditions. While EPA and others may occasionally use the language of “significance” as a shorthand for determinations at the third step under both prongs 1 and 2, this does not detract from the fact that EPA gives prong 2 independent effect under the 4-step framework.

EPA has explained elsewhere in the record for this action why the selected control stringency selected in this rule is appropriate in light of EPA's application of the step 3 multi-factor test of the 4-step framework. To the extent commenters argue that EPA should have selected a higher cost threshold or required more reductions based on the technical data, those issues are addressed elsewhere in the record.

V. Analyzing Downwind Air Quality and Upwind-State Contributions

In this section, EPA describes the air quality modeling and analyses performed to identify nonattainment and/or maintenance receptors and evaluate interstate contributions to these receptors from individual upwind states for the 2021 analytic year. Although the air quality modeling was performed using an air quality modeling platform that covers the contiguous 48 states, the analysis to identify receptors and evaluate contributions focuses on the 21 upwind states that are the subject of this rule with respect to the 2008 ozone NAAQS. In this action, EPA is not addressing the good neighbor obligations of any other state, nor is it addressing the obligations of any state, including the 21 covered by this action, with respect to the 2015 ozone NAAQS.

The year 2021 was selected as the appropriate future analytic year for this rule because it coincides with the July 20, 2021, Serious area attainment date under the 2008 ozone NAAQS. In the CSAPR Update, EPA had aligned its analysis and implementation of emission reductions with the 2017 ozone season (ozone seasons run each year from May 1-September 30) in order to assist downwind states with timely attainment of the 2008 ozone NAAQS by the Moderate area attainment date of July 20, 2018. See 81 FR 74516. In order to demonstrate attainment by this deadline, states were required to rely on design values calculated using ozone season data from 2015 through 2017, since the July 20, 2018, deadline did not afford enough time for measured data of the full 2018 ozone season. Similarly, for the Serious area attainment date in 2021, states will rely on design values calculated using ozone season data from 2018 through 2020. However, it is not possible to impose emission reductions on upwind states in the 2020 ozone season, which has already passed. Reductions in the 2021 ozone season will nonetheless occur in time for the 2021 attainment date and therefore assist downwind states in achieving attainment by the July 20, 2021, attainment date, in compliance with the Wisconsin holding. See Wisconsin, 938 F.3d at 309 (the CSAPR Update was unlawful to the extent it allowed upwind states to “continue their significant contributions to downwind air quality problems beyond the statutory deadlines by which downwind States must demonstrate their attainment of air quality standards”) (emphasis added). Further, EPA continues to interpret the good neighbor provision as forward-looking, based on Congress's use of the future-tense “will” in section 110(a)(2)(D)(i), an interpretation upheld in Wisconsin, 938 F.3d at 322. It would be “anomalous,” id., for EPA to impose good neighbor obligations in 2021 and future years based solely on finding that “significant contribution” had existed at some time in the past.

EPA has also conducted additional analysis of remaining air quality receptors and contribution in years beyond 2021, in order to ensure a complete step 3 analysis. EPA has analyzed these later years to determine whether any additional emission reductions that are impossible to obtain by the 2021 attainment date may yet be necessary in order to fully address significant contribution. This comports with the D.C. Circuit's direction in Wisconsin that implementing good neighbor obligations beyond the dates established for attainment may be justified on a proper showing of impossibility and/or necessity. See 938 F.3d at 320. However, for purposes of EPA's initial analysis of air quality at step 1 of the 4-step framework, in accordance with Wisconsin, EPA has selected the 2021 ozone season, corresponding with the 2021 Serious area attainment date.

The remainder of this section includes information on: (1) The air quality modeling platform used in support of this final rule with a focus on the base year and future year base case emission inventories, (2) the method for projecting design values in 2021, and (3) the approach for calculating ozone contributions from upwind states.[74] The Agency also provides the design values for nonattainment and maintenance receptors and the predicted interstate contributions that are at or above the 1 percent of the NAAQS screening threshold. The 2016 base period and 2021, 2023, and 2028 future design values and contributions for all ozone monitoring sites are provided in the docket for this rule. The Air Quality Modeling Technical Support Document Start Printed Page 23075(AQM TSD) in the docket for this rule contains more detailed information on the air quality modeling aspects of this rule.

A. Overview of Air Quality Modeling Platform

EPA used the 2016-based modeling platform for the air quality modeling for this final rule. This modeling platform includes 2016 base year emissions from anthropogenic and natural sources and 2016 meteorology. The platform also includes anthropogenic emission projections for 2023 and 2028. The emissions data contained in this platform were developed by EPA, Multi-Jurisdictional Organizations (MJOs), and state and local air agencies as part of the Emissions Inventory Collaborative Process. This process resulted in a common-use set of emissions data for a 2016 base year and 2023 and 2028 that can be leveraged by EPA and states for regulatory air quality modeling.[75] The air quality modeling was performed for a modeling region (i.e., modeling domain) that covers the contiguous 48 states using a horizontal resolution of 12 x 12 km. EPA used the CAMx version 7beta6 for air quality modeling for both the proposed rule and this final rule.[76] Additional information on the 2016-based air quality modeling platform can be found in the AQM TSD.

B. Emission Inventories

EPA developed emission inventories for the proposed rule, including emission estimates for EGUs, non-EGU point sources, stationary nonpoint sources, onroad mobile sources, nonroad mobile sources, wildfires, prescribed fires, and biogenic emissions that are not the result of human activities. EPA's air quality modeling relies on this comprehensive set of emission inventories because emissions from multiple source categories are needed to model ambient air quality and to facilitate comparison of model outputs with ambient measurements. To prepare the emission inventories for air quality modeling, EPA processed the emission inventories using the Sparse Matrix Operator Kernel Emissions (SMOKE) Modeling System version 4.7 to produce the gridded, hourly, speciated, model-ready emissions for input to the air quality model. Additional information on the development of the emission inventories and on data sets used during the emissions modeling process are provided in the Technical Support Document (TSD) “Preparation of Emissions Inventories for the 2016v1 North American Emissions Modeling Platform,” hereafter known as the “Emissions Modeling TSD.” This TSD is available in the docket for this rule and at https://www.epa.gov/​air-emissions-modeling/​2016v1-platform.

1. Foundation Emission Inventory Data Sets

Emissions data were developed that represented the year 2016 to support air quality modeling of a base year from which future air quality could be forecasted. As noted above, EPA used the Inventory Collaborative 2016 version 1 (2016v1) Emissions Modeling Platform, released in October 2019, as the primary basis for the inventories supporting the air quality modeling. This platform was developed through a national collaborative effort between EPA and state and local agencies along with MJOs. The original starting point for the U.S. portions of the 2016 inventory was the 2014 National Emissions Inventory (NEI), version 2 (2014NEIv2), although all of the inventory sectors were updated to better represent the year 2016 through the incorporation of 2016-specific state and local data along with nationally applied adjustment methods. The future base case inventories developed for 2023 and 2028 represent projected changes in activity data and predicted emission reductions from on-the-books actions, planned emission control installations, and promulgated federal measures that affect anthropogenic emissions.[77]

2. Development of Emission Inventories for EGUs

Annual NOX and SO2 emissions for EGUs in the 2016 base year inventory are based primarily on data from continuous emission monitoring systems (CEMS) and other monitoring systems allowed for use by qualifying units under 40 CFR part 75, with other EGU pollutants estimated using emission factors and annual heat input data reported to EPA. For EGUs not reporting under part 75, EPA used the most recent data submitted to the NEI by the states. Emissions data for sources that did not have data provided for the year 2016 were pulled forward from data submitted for 2014. The Air Emissions Reporting Rule, (80 FR 8787 February 19, 2015), requires that Type A point sources large enough to meet or exceed specific thresholds for emissions be reported to EPA every year, while the smaller Type B point sources must only be reported to EPA every three years. For more information on how the 2016 EGU emissions data were developed and prepared for air quality modeling, see the Emissions Modeling TSD.

EPA projected future 2023 and 2028 baseline EGU emissions using the version 6—January 2020 reference case of the Integrated Planning Model (IPM).[78] [79] IPM, developed by ICF Consulting, is a state-of-the-art, peer-reviewed, multi-regional, dynamic, deterministic linear programming model of the contiguous U.S. electric power sector. It provides forecasts of least cost capacity expansion, electricity dispatch, and emission control strategies while meeting energy demand and environmental, transmission, dispatch, and reliability constraints. EPA has used IPM for over two decades to better understand power sector behavior under future business-as-usual conditions and to evaluate the economic and emission impacts of prospective environmental policies. The model is designed to reflect electricity markets as accurately as possible. EPA uses the best available information from utilities, industry experts, gas and coal market experts, financial institutions, and government statistics as the basis for the detailed power sector modeling in IPM. The model documentation provides additional information on the assumptions discussed here as well as all other model assumptions and inputs.[80]

The IPM version 6—January 2020 reference base case accounts for updated federal and state environmental regulations, committed EGU retirements and new builds, and technology cost and performance assumptions as of late 2019. This projected base case accounts for the effects of the finalized Mercury and Air Toxics Standards rule, the CSAPR and the CSAPR Update, New Source Review settlements, and other on-the-books federal and state rules through 2019 [81] impacting SO2, NOX, Start Printed Page 23076directly emitted particulate matter, and CO2, and final actions EPA has taken to implement the Regional Haze Rule.

Additional 2021 EGU emissions baseline levels were developed through engineering analytics as an alternative approach that did not involve IPM. EPA developed this inventory for use in step 3 of this final rule, where it determines emission reduction potential and corresponding emission budgets. IPM includes optimization and perfect foresight in solving for least cost dispatch. Given that this final rule will likely become effective either immediately prior to or slightly after the start of the 2021 ozone season, EPA adopted a similar approach to the CSAPR Update where it relied on IPM in a relative way in step 3 to avoid overstating optimization and dispatch decisions that were not possible in the short time frame. EPA does this by using the difference in emission rate observed between IPM runs with and without the cost threshold applied, rather than using absolute values. In both the CSAPR Update and in this rule at step 3, EPA complemented that projected IPM EGU outlook with historical (e.g., engineering analytics) perspective based on historical data that only factors in known changes to the fleet. This 2021 engineering analytics data set is described in more detail in the Ozone Transport Policy Analysis Final Rule TSD.

3. Development of Emission Inventories for Non-EGU Point Sources

The non-EGU point source emissions in the 2016 base case inventory match those in the 2016v1 platform. Some non-EGU point source emissions were based on data submitted for 2016, others were projected from 2014 to 2016, and the emissions for remaining small sources were kept at 2014 levels. Prior to air quality modeling, the emission inventories were processed into a format that is appropriate for the air quality model to use. Projection factors and percent reductions in this final rule reflect comments received as a result of the Inventory Collaborative development process, along with emission reductions due to national and local rules, control programs, plant closures, consent decrees, and settlements. Reductions from several Maximum Achievable Control Technology and National Emission Standards for Hazardous Air Pollutants (NESHAP) standards are included. Projection approaches for corn ethanol and biodiesel plants, refineries and upstream impacts represent requirements pursuant to the Energy Independence and Security Act of 2007 (EISA). Details on the development and processing of the non-EGU emissions inventories for 2016, 2023, and 2028 are available in the Emissions Modeling TSD.

For aircraft emissions at airports, the emissions used were based on adjustments to emissions in the 2017 NEI (see https://www.epa.gov/​air-emissions-inventories/​2017-national-emissions-inventory-nei-data for data and a TSD). EPA developed and applied factors to adjust the 2017 emissions to 2016, 2023, and 2028 based on activity growth projected by the Federal Aviation Administration Terminal Area Forecast [82] system, published in 2018.

Emissions at rail yards were represented as non-EGU point sources. The 2016 rail yard emissions are largely consistent with the 2017 NEI rail yard emissions. The 2016, 2023, and 2028 rail yard emissions were developed through the Inventory Collaborative process. The rail yard emissions were interpolated from the 2016 and 2023 emissions. Class I rail yard emissions were projected using the Energy Information Administration's 2019 Annual Energy Outlook (AEO) freight rail energy use growth rate projections for 2016, 2023, and 2028 with the fleet mix assumed to be constant throughout the period.

Point source oil and gas emissions for 2016 were based on the 2016v1 point inventory, while nonpoint oil and gas emissions were primarily based on a run of EPA Oil and Gas Tool for the year 2016. The 2016 oil and gas inventories were projected to 2023 and 2028 using regional projection factors by product type based on AEO 2019 projections. NOX and VOC reductions that are co-benefits to the NESHAP and New Source Performance Standards (NSPS) for Stationary Reciprocating Internal Combustion Engines (RICE) are reflected for select source categories. In addition, Natural Gas Turbines and Process Heaters NSPS NOX controls and NSPS Oil and Gas VOC controls are reflected for select source categories. Additional information on the development and modeling of the oil and gas emission inventories can be found in the Emissions Modeling TSD.

4. Development of Emission Inventories for Onroad Mobile Sources

Onroad mobile sources include exhaust, evaporative, and brake and tire wear emissions from vehicles that drive on roads, parked vehicles, and vehicle refueling. Emissions from vehicles using regular gasoline, high ethanol gasoline, diesel fuel, and electric vehicles were represented, along with buses that used compressed natural gas. EPA developed the onroad mobile source emissions for states other than California using EPA's Motor Vehicle Emissions Simulator (MOVES) 2014b. MOVES2014b was used with inputs provided by state and local agencies, where available, in combination with nationally available data sets. Onroad emissions for the platform were developed based on emissions factors output from MOVES2014b run for the year 2016, coupled with activity data (e.g., vehicle miles traveled and vehicle populations) representing the year 2016. The 2016 activity data were provided by some state and local agencies, and the remaining activity data were derived from the 2014NEIv2. The onroad emissions were computed within SMOKE by multiplying emission factors developed using MOVES with the appropriate activity data. Onroad mobile source emissions for California were consistent with the emissions provided by the state.

The future-year emissions for onroad mobile sources represent all national control programs known at the time of modeling except for the Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles (HDGHG)—Phase 2 [83] and the Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule.[84] Finalized rules incorporated into the onroad mobile source emissions include: Tier 3 Standards (March 2014), the Light-Duty Greenhouse Gas Rule (March 2013), Heavy (and Medium)-Duty Greenhouse Gas Rule (August 2011), the Renewable Fuel Standard (February 2010), the Light Duty Greenhouse Gas Rule (April 2010), the Corporate-Average Fuel Economy standards for 2008-2011 (April 2010), the 2007 Onroad Heavy-Duty Rule (February 2009), and the Final Mobile Source Air Toxics Rule (MSAT2) (February 2007). Estimates of the impacts of rules that were in effect in 2016 are included in the 2016 base year Start Printed Page 23077emissions at a level that corresponds to the extent to which each rule had penetrated into the fleet and fuel supply by the year 2016. Local control programs such as the California LEV III program are included in the onroad mobile source emissions. The future year onroad emissions reflect projected changes to fuel properties and usage. MOVES was run for the years 2023 and 2028 to generate the emissions factors relevant to those years. Future year activity data for onroad mobile sources were provided by some state and local agencies, and otherwise were projected to 2023 and 2028 using AEO 2019-based factors. The future year emissions were computed within SMOKE by multiplying the future year emission factors developed using MOVES with the year-specific activity data. Additional information on the approach for generating the onroad mobile source emissions is available in the Emissions Modeling TSD.

5. Development of Emission Inventories for Commercial Marine Vessels

The commercial marine vessel (CMV) emissions in the 2016 base case emission inventory for this rule were based on those in the 2017 NEI. Factors were then applied to adjust the 2017 NEI emissions backward to represent emissions for the year 2016. The CMV emissions reflect reductions associated with the Emissions Control Area proposal to the International Maritime Organization control strategy (EPA-420-F-10-041, August 2010); reductions of NOX, VOC, and CO emissions for new C3 engines that went into effect in 2011; and fuel sulfur limits that went into effect prior to 2016. The cumulative impacts of these rules through 2023 and 2028 were incorporated into the projected emissions for CMV sources. The CMV emissions were split into emissions inventories from the larger category 3 (C3) engines, and those from the smaller category 1 and 2 (C1C2) engines. Some minor adjustments to the CMV emissions were implemented following the October 2019 2016v1 release. These updated CMV inventories were released publicly by February, 2020.[85]

6. Development of Emission Inventories for Other Nonroad Mobile Sources

Nonroad mobile source emission inventories (other than CMV, locomotive, and aircraft emissions) were developed from monthly, county, and process level emissions output from MOVES2014b. MOVES2014b included important updates to nonroad engine population growth rates. Types of nonroad equipment include recreational vehicles, pleasure craft, and construction, agricultural, mining, and lawn and garden equipment. State-submitted emissions data for nonroad sources were used for California.

EPA also ran MOVES2014b for 2023 and 2028 to prepare nonroad mobile emissions inventories for future years. The nonroad mobile emission control programs include reductions to locomotives, diesel engines, and recreational marine engines, along with standards for fuel sulfur content and evaporative emissions. A comprehensive list of control programs included for mobile sources is available in the Emissions Modeling TSD.

Line haul locomotives are also considered a type of nonroad mobile source but the emissions inventories for locomotives were not developed using MOVES2014b. Year 2016 locomotive emissions were developed through the Inventory Collaborative and are mostly consistent with those in the 2017 NEI. The projected locomotive emissions for 2023 and 2028 were developed by applying factors to the base year emissions using activity data based on 2018 AEO freight rail energy use growth rate projections and emission rates adjusted to account for recent historical trends.

7. Development of Emission Inventories for Nonpoint Sources

The emissions for stationary nonpoint sources in our 2016 base case emission inventory are largely consistent with those in the 2014NEIv2, although some were adjusted to more closely reflect year 2016 using factors based on changes to human population from 2014 to 2016. Stationary nonpoint sources include evaporative sources, consumer products, fuel combustion that is not captured by point sources, agricultural livestock, agricultural fertilizer, residential wood combustion, fugitive dust, and oil and gas sources. For more information on the nonpoint sources in the 2016 base case inventory, see the Emissions Modeling TSD and the 2014NEIv2 TSD.

Where states provided the Inventory Collaborative information about projected control measures or changes in nonpoint source emissions, those inputs were incorporated into the projected inventories for 2023 and 2028. Adjustments for state fuel sulfur content rules for fuel oil in the Northeast were included. Projected emissions for portable fuel containers reflect the impact of projection factors required by the final MSAT2 rule and the EISA, including updates to cellulosic ethanol plants, ethanol transport working losses, and ethanol distribution vapor losses.

For 2016, nonpoint oil and gas emissions inventories were developed based on a run of EPA Oil and Gas Tool for 2016. To develop the future year inventories, regional projection factors for nonpoint oil and gas sources were developed by product type based on AEO 2019 projections to 2023 and 2028. Estimates of criteria air pollutant (CAP) co-benefit reductions resulting from the NESHAP for RICE and NSPS rules and Oil and Gas NSPS VOC controls for select source categories were included. Additional details on the application of these rules and projections for nonpoint sources are available in the Emissions Modeling TSD. EPA received comments on the emissions inventories used in the proposed rule. These comments and EPA's responses are provided below and in the RTC.

Comment: EPA received comments that contend that the Agency did not include emission reductions from all “on the books” control programs in certain states. These commenters say that monitoring sites that were identified as nonattainment and/or maintenance receptors might not be receptors if the Agency had accounted for the impacts of all control programs.[86]

Response: The emissions inventories used for the step 1 and step 2 air quality modeling of 2023 and 2028 were developed through a collaborative process through which input from state and local agencies and multijurisdictional organizations was solicited and accepted. For point sources, the 2016 inventories were derived from state and local submissions to the 2016 NEI as required by the Air Emissions Reporting Rule see 80 FR 8787 (February 19, 2015). Any rules promulgated by 2016 that would have impacted emissions in the year 2016 would be included in those inventories. EPA then accounted for known changes in those inventories that would occur by 2023 and 2028 using EPA projection methods along with stakeholder-developed information. The Midatlantic Regional Air Management Association (MARAMA) worked with their member states and Ozone Transport Commission (OTC) states to develop projection and control factors for the years 2023 and 2028. These factors were provided to EPA in May Start Printed Page 230782019 and reflect rules impacting nonpoint sources that were promulgated prior to 2019. Through the Inventory Collaborative process, the inventories used for modeling included the “on the books” control programs that were identified by EPA and the state and multijurisdictional organization (MJO) partners such as MARAMA that provided inputs to the collaborative inventories. Rules related to emissions for sources other than EGUs promulgated in 2019 or later following the completion of the inventories for those sources are not included in the modeling for this rule.

The commenter has listed multiple pages of various state-level NOX and VOC control programs and regulations, promulgated over multiple decades. The commenter did not provide quantitative information or data to support their claim that EPA failed to include the control programs cited by the commenter in the emissions inventories used to support the proposed rule, what the effect would be had they been included or characterized differently, and whether the effect would have changed any of the regulatory outcomes in EPA's analysis. This comment is further addressed in the RTC.

Comment: EPA received comment suggesting changes to its EGU emissions inventory used in its step 1 and step 2 evaluations based on more recent data.

Response: EPA is not changing the emissions inventory derived from its IPM modeling that incorporated the latest data at the time of execution in January of 2020 used at step 1 and step 2 of the 4-step framework. However, both in the proposed rule and at final, EPA reaffirmed its step 1 and step 2 findings using an updated/alternative EGU emissions inventory from the engineering analytics tool used in step 3 and discussed in the Ozone Transport Policy Analysis Final Rule TSD. This tool reflects known changes (e.g., retirements and new builds) applied to historical data to estimate future year EGU emissions. It represents alternative EGU emissions inventory perspective as it does not factor in model-projected changes. Moreover, it incorporates the latest available data and commenter input regarding any fleet changes. EPA, in the proposed and final rule, uses this alternative inventory in conjunction with its air quality assessment tool (AQAT) to estimate air quality impacts and upwind state contributions. Both in the proposed rule and final, this alternative emissions inventory and subsequent AQAT sensitivity analysis led to the same step 1 and step 2 findings as the IPM-based EGU emissions inventory and related CAMx modeling results. That is, EPA has examined a range of EGU inventories using different future year projections and incorporating the latest available data and commenter input. Across this range of EGU emission inventory estimates, EPA reaches the same conclusion for step 1 and step 2 downwind receptors and upwind linkages. Therefore, EPA's EGU emission inventories and corresponding step 1 and step 2 analytic findings have been robustly examined, tested across a range of assumptions, and are robust to a variety of assumptions, including the unit updates suggested by the commenter. For a complete unit-by-unit inventory of all EGUs included in the future year baseline for the engineering analytic tool, see the Ozone Transport Policy Analysis Final Rule TSD; Appendix A. The data in this Appendix reflect future unit level operating status taking into account retirement and new build announcements from both commenter input and the latest EIA Form 860 monthly (October 2020) available.

C. Air Quality Modeling and Analyses To Identify Nonattainment and Maintenance Receptors

In this section the Agency describes the air quality modeling and analyses performed in Step 1 to identify locations where the Agency expects there to be nonattainment or maintenance receptors for the 2008 8-hour ozone NAAQS in the 2021 analytic future year. Where EPA's analysis shows that an area or site does not fall under the definition of a nonattainment or maintenance receptor in 2021, that site is excluded from further analysis under EPA's good neighbor framework.

In this final rule, EPA is not reopening the approach used in the CSAPR Update to identify nonattainment and maintenance receptors. Wisconsin upheld EPA's approach to identifying nonattainment and maintenance receptors against specific challenges. See 938 F.3d at 325-27.[87] As this action is taken in response to the Wisconsin remand and to complete the good neighbor obligations that were partially addressed in the CSAPR Update, it is entirely appropriate to continue to apply the same approach to identifying receptors to fully address the outstanding obligations as EPA took in partially addressing them. Indeed, to do otherwise would be anomalous and could lead to inconsistent treatment of states under the 4-step framework for purposes of the 2008 ozone NAAQS. However, as an aid to understanding EPA's approach to identifying receptors, a summary of this approach follows.

EPA's approach gives independent effect to both the “contribute significantly to nonattainment” and the “interfere with maintenance” prongs of section 110(a)(2)(D)(i)(I), consistent with the D.C. Circuit's direction in North Carolina.[88] Further, in its decision on the remand of the CSAPR from the Supreme Court in the EME Homer City case, the D.C. Circuit confirmed that EPA's approach to identifying maintenance receptors in the CSAPR comported with the court's prior instruction to give independent meaning to the “interfere with maintenance” prong in the good neighbor provision. EME Homer City II, 795 F.3d at 136.

In the CSAPR Update, EPA identified nonattainment receptors as those monitoring sites that are projected to have average design values that exceed the NAAQS and that are also measuring nonattainment based on the most recent monitored design values. This approach is consistent with prior transport rulemakings, such as the NOX SIP Call and CAIR, where EPA defined nonattainment receptors as those areas that both currently monitor nonattainment and that EPA projects will be in nonattainment in the future compliance year.[89]

The Agency explained in the NOX SIP Call and CAIR and then reaffirmed in the CSAPR Update that EPA has the most confidence in our projections of nonattainment for those counties that also measure nonattainment for the most recent period of available ambient data. EPA separately identified maintenance receptors as those receptors that would have difficulty maintaining the relevant NAAQS in a scenario that takes into account historical variability in air quality at that receptor. The variability in air quality was determined by evaluating the “maximum” future design value at each receptor based on a projection of the maximum measured design value Start Printed Page 23079over the relevant period. EPA interprets the projected maximum future design value to be a potential future air quality outcome consistent with the meteorology that yielded maximum measured concentrations in the ambient data set analyzed for that receptor (i.e., ozone conducive meteorology). EPA also recognizes that previously experienced meteorological conditions (e.g., dominant wind direction, temperatures, air mass patterns) promoting ozone formation that led to maximum concentrations in the measured data may reoccur in the future. The maximum design value gives a reasonable projection of future air quality at the receptor under a scenario in which such conditions do, in fact, reoccur. The projected maximum design value is used to identify upwind emissions that, under those circumstances, could interfere with the downwind area's ability to maintain the NAAQS.

Therefore, applying this methodology in this final rule, EPA assessed the magnitude of the maximum projected design value for 2021 at each receptor in relation to the 2008 ozone NAAQS and, where such a value exceeds the NAAQS, EPA determined that receptor to be a “maintenance” receptor for purposes of defining interference with maintenance, consistent with the method used in the CSAPR and upheld by the DC Circuit in EME Homer City II.[90] That is, monitoring sites with a maximum design value that exceeds the NAAQS are projected to have a maintenance problem in 2021.

Recognizing that nonattainment receptors are also, by definition, maintenance receptors, EPA often uses the term “maintenance-only” to refer to receptors that are not also nonattainment receptors. Consistent with the methodology described above, monitoring sites with a projected maximum design value that exceeds the NAAQS, but with a projected average design value that is below the NAAQS, are identified as maintenance-only receptors. In addition, those sites that are currently measuring ozone concentrations below the level of the applicable NAAQS, but are projected to be nonattainment based on the average design value and that, by definition, are projected to have a maximum design value above the standard are also identified as maintenance-only receptors.

As described above in section V.B., EPA is using the 2016 and 2023 base case emissions developed under the EPA/MJO/state collaborative project as the primary source for base year and 2023 future year emissions data for this final rule. Because this platform does not include emissions for 2021, EPA developed an interpolation technique based on modeling for 2023 and measured ozone data to determine ozone concentrations for 2021. To estimate average and maximum design values for 2021, EPA first performed air quality modeling for 2016 and 2023 to obtain design values in 2023. The 2023 design values were then coupled with the corresponding 2016 measured design values to estimate design values in 2021 using the interpolation technique described below.

Consistent with EPA's modeling guidance, the 2016 and 2023 air quality modeling results were used in a “relative” sense to project design values for 2023. That is, the ratios of future year model predictions to base year model predictions are used to adjust ambient ozone design values up or down depending on the relative (percent) change in model predictions for each location. The modeling guidance recommends using measured ozone concentrations for the 5-year period centered on the base year as the air quality data starting point for future year projections. This average design value is used to dampen the effects of inter-annual variability in meteorology on ozone concentrations and to provide a reasonable projection of future air quality at the receptor under “average” conditions. In addition, the Agency calculated maximum design values from within the 5-year base period to represent conditions when meteorology is more favorable than average for ozone formation. Because the base year for the air quality modeling used in this final rule is 2016, the base period 2014-2018 ambient ozone design value data was used in order to project average and maximum design values in 2023.

The ozone predictions from the 2016 and 2023 air quality model simulations were used to project 2014-2018 average and maximum ozone design values to 2023 using an approach similar to the approach in EPA's guidance for attainment demonstration modeling. This guidance recommends using model predictions from the “3 x 3” array of grid cells [91] surrounding the location of the monitoring site to calculate a Relative Response Factor (RRF) for that site.[92] The 2014-2018 average and maximum design values were multiplied by the RRF to project each of these design values to 2023. In this manner, the projected design values are grounded in monitored data, and not the absolute model-predicted 2023 concentrations. In light of comments on the Notice of Data Availability (82 FR 1733; January 6, 2017) and other analyses, EPA also projected 2023 design values based on a modified version of the “3 x 3” approach for those monitoring sites located in coastal areas. In this alternative approach, EPA eliminated from the RRF calculations the modeling data in those grid cells that are dominated by water (i.e., more than 50 percent of the area in the grid cell is water) and that do not contain a monitoring site (i.e., if a grid cell is more than 50 percent water but contains an air quality monitor, that cell would remain in the calculation). The choice of more than 50 percent of the grid cell area as water as the criteria for identifying overwater grid cells is based on the treatment of land use in the Weather Research and Forecasting model (WRF).[93] Specifically, in the WRF meteorological model those grid cells that are greater than 50 percent overwater are treated as being 100 percent overwater. In such cases the meteorological conditions in the entire grid cell reflect the vertical mixing and winds over water, even if part of the grid cell also happens to be over land with land-based emissions, as can often be the case for coastal areas. Overlaying land-based emissions with overwater meteorology may be representative of conditions at coastal monitors during times of on-shore flow associated with synoptic conditions and/or sea-breeze or lake-breeze wind flows. But there may be other times, particularly with off-shore wind flow when vertical mixing of land-based emissions may be too limited due to the presence of overwater meteorology. Thus, for our modeling EPA calculated 2023 projected average and maximum design values at individual monitoring sites based on both the “3 x 3” approach as well as the alternative approach that eliminates overwater cells in the RRF calculation Start Printed Page 23080for near-coastal areas (i.e., “no water” approach).

The 2023 average and maximum design values for both the “3 x 3” and “no water” approaches were then paired with the corresponding base period measured design values at each ozone monitoring site. Design values for 2021 for both approaches were calculated by linearly interpolating between the 2016 base period and 2023 projected values.[94] The steps in the interpolation process for estimating 2021 average and maximum design values are as follows:

(1) Calculate the ppb change in design values between the 2016 base period and 2023;

(2) Divide the ppb change by 7 to calculate the ppb change per year over the 7-year period between 2016 and 2023;

(3) Multiply the ppb per year value by 5 to calculate the ppb change in design values over the 5-year period between 2016 and 2021;

(4) Subtract the ppb change between 2016 to 2021 from the 2016 design values to produce the design values for 2021.

The projected 2021 and 2023 design values using both the “3 x 3” and “no-water” approaches are provided in the AQM TSD.[95] For this final rule, EPA is relying upon design values based on the “no water” approach for identifying nonattainment and maintenance receptors.

Consistent with the truncation and rounding procedures for the 8-hour ozone NAAQS, the projected design values are truncated to integers in units of ppb.[96] Therefore, projected design values that are greater than or equal to 76 ppb are considered to be violating the 2008 ozone NAAQS. For those sites that are projected to be violating the NAAQS based on the average design values in 2021, the Agency examined the design values for 2019, which are the most recent certified measured ozone design values at the time of this action. As noted above, the Agency identified nonattainment receptors in this rulemaking as those sites that are violating the NAAQS based on current measured air quality and also have projected average design values of 76 ppb or greater. Maintenance-only receptors include both (1) those sites with projected average design values above the NAAQS that are currently measuring clean data and (2) those sites with projected average design values below the level of the NAAQS, but with projected maximum design values of 76 ppb or greater. In addition to the maintenance-only receptors, the 2021 ozone nonattainment receptors are also maintenance receptors because the maximum design values for each of these sites is always greater than or equal to the average design value. The monitoring sites that the Agency projects to be nonattainment and maintenance receptors for the ozone NAAQS in the 2021 base case are used for assessing the contribution of emissions in upwind states to downwind nonattainment and maintenance of ozone NAAQS as part of this action.

Table V.C-1 contains the 2014-2018 base period average and maximum 8-hour ozone design values, the 2021 base case average and maximum design values,[97] and the 2019 design values for the two sites that are projected to be nonattainment receptors in 2021 and the two sites that are projected to be maintenance-only receptors in 2021.[98] The design values for all monitoring sites in the U.S. are provided in the docket for this rule. Additional details on the approach for projecting average and maximum design values are provided in the AQM TSD.

Table V.C-1—Average and Maximum 2014-2018 and 2021 Base Case 8-Hour Ozone Design Values and 2019 Preliminary Design Values (ppb) at Projected Nonattainment and Maintenance-Only Sites

Monitor IDStateSiteAverage design value 2014-2018Maximum design value 2014-2018Average design value 2021Maximum design value 20212019 Design value
Nonattainment Receptors
090013007CTStratford82.08376.577.482
090019003CTWestport82.78378.578.882
Maintenance-Only Receptors
090099002CTMadison79.78273.976.182
482010024TXHouston79.38175.577.181

Comment: Some commenters said that EPA's interpolation method for determining design values in 2021 is flawed because (1) the method incorrectly assumes that ozone precursor emissions in all source sectors in all states change at an equal rate between 2016 and 2023, (2) linearly interpolated EGU emissions for 2021 overstate EPA's IPM-predicted EGU emissions for 2021, and (3) the method does not account for the non-linear Start Printed Page 23081response of ozone to emissions changes. These commenters say that EPA should have developed a 2021 specific emissions inventory or at a minimum developed an interpolated 2021 emission inventory and then rerun the photochemical model to account for the reactivity of ozone formation from the distribution of ozone precursor emissions. The commenters contend that failing to take this step, EPA has introduced significant uncertainty into the air quality projections of the proposed rule and potentially subjected multiple upwind states to unnecessary additional control requirements.

Response: As an initial matter, there is no legal obligation for EPA to directly model the selected analytic year, here 2021, in order to make regulatory determinations within the 4-step good neighbor framework. Given the limited amount of time EPA had to complete this rulemaking in order to meet the court-ordered March 15 deadline, EPA reasonably chose to use existing air quality modeling and contribution information to derive an appropriately reliable projection of air quality conditions and contributions in 2021. The Supreme Court recognized in EME Homer City that it is not possible to perfectly account for all factors that will affect downwind air quality problems in a future year. Regulators, the Court noted, “must account for the vagaries of the wind” and in assigning upwind responsibility face a “thorny causation problem.” 572 U.S. 489, 497, 514. EPA's ultimate task is not to achieve a perfect understanding of atmospheric conditions in some future year, but “to quantify the amount of upwind gases . . . that must be reduced to enable downwind states to keep their levels of ozone . . . in check. Id. 497. See also EME Homer City Generation, L.P. v. EPA, 795 F.3d 118, 135-36 (“We will not invalidate EPA's predictions solely because there might be discrepancies between those predictions and the real world. . . . [A] model is meant to simplify reality in order to make it tractable.”).

EPA continues to view the interpolation analysis presented at proposal as sufficiently reliable for purposes of the regulatory determinations made in this rulemaking. Commenters assert that it is possible EPA may have found certain upwind state linkages not to exist had EPA taken a different approach to developing its projections for 2021. But no commenter has established an actual instance of overcontrol, which the courts have held must be clearly established through as-applied challenges. See Wisconsin, 938 F.3d at 325 (“ `[T]he Supreme Court has made clear . . . that the way to contest instances of over-control is not through generalized claims that EPA's methodology would lead to over-control, but rather through a “particularized, as-applied challenge.” ' ”) (quoting EME Homer City, 795 F.3d at 137).

Nonetheless, in consideration of these comments, EPA has performed additional analysis, which confirms the regulatory determinations EPA proposed and is now finalizing. EPA was able to construct an emissions inventory for 2021, using available data and the same approach as EPA used to develop projection inventories for 2023 and 2028. Details on the construct of the 2021 emissions are provided in the Emissions Modeling TSD. There was, however, insufficient time to perform air quality modeling using this newly constructed 2021 inventory. Instead EPA used the Air Quality Assessment Tool (AQAT) to perform a sensitivity analysis to determine whether there would be any change in the outcome of this rule if the projection of 2021 air quality were based on projected 2021 emissions rather than EPA's interpolation method, as described above. In brief, AQAT uses the results of existing base year and future year air quality modeling as part of an interpolation technique to estimate ozone design values and contributions for analytic years that are not modeled as well as to analyze the air quality impacts of control scenarios in step 3 of the 4-step transport framework. AQAT is calibrated using model simulations to account for the non-linearity response of ozone to emissions changes. As noted by the commenter, EPA's interpolation approach inherently assumes that the relative change in emissions between 2016 and 2023 is the same across all states. Because this application of AQAT considered 2021 state level emissions on a state-by-state basis, the analysis accounted for any state-to-state differences in the change in emissions between 2016 and 2023. As part of this sensitivity analysis EPA coupled the 2021 emissions and 2023 model-predicted ozone design values and contributions to estimate design values and contributions in 2021. EPA also used the 2021 emissions in AQAT to create a more-refined interpolated 2022 emission inventory. EPA then used the AQAT to examine the effects of this refined 2022 emission inventory on ozone design values and contributions. The results indicate that any changes in the nonattainment or maintenance status of individual receptors using 2021 and 2022 projected emissions would not affect which upwind states significantly contribute to nonattainment and/or interfere with maintenance of the 2008 NAAQS in another state.[99] Details on AQAT and this sensitivity analysis can be found in the Ozone Policy Analysis Final Rule TSD.

Comment: Other commenters claim that there is a disconnect between EPA's projected 2021 design values and current ozone monitoring data. These commenters said that EPA should give priority to monitored data over modeled data when evaluating which areas need transport obligations resolved. Specifically, one commenter performed an analysis to estimate 2021 design values by first estimating a fourth high maximum daily average 8-hour (MDA8) ozone concentration in 2021 based on the four-year average of the measured fourth high values during the period 2017 through 2020 and second, calculating the 2021 design value as the average of the measured fourth high value in 2019, the preliminary fourth high value in 2020 and the estimated fourth high value in 2021.[100] Another commenter performed a statistical linear regression analysis of the fourth highest measured values for each of three time periods: 2012 through 2020, 2014 through 2020, and 2016 through 2020 to estimate fourth highest values in 2021 that would result in nonattainment in 2021 at individual monitoring sites. This commenter said that an assessment of actual ambient monitor data, such as the analysis performed by this commenter, should be given as much weight, if not more, in identifying receptors in 2021 as the modeling-based analysis performed by EPA. Both commenters said that the results of their analyses support EPA's finding that the four monitoring sites identified in Table V.C-1, above will be receptors in 2021. However, both commenters claim that the Madison, Connecticut monitoring site 090099002 will be a nonattainment receptor, whereas EPA projects this site to be a maintenance-only receptor in 2021. Also, both commenters claim that there will be an additional 2021 nonattainment receptor at the Greenwich, Connecticut monitoring site 090010017. One commenter noted that identifying the Madison monitoring site as nonattainment instead of maintenance-only and the Greenwich Start Printed Page 23082monitoring site as a receptor will not alter the outcome of EPA's determination of which upwind states are linked to downwind receptors at step 2 of the 4-step transport framework.

In addition to the 2021 receptors in Connecticut, one commenter said that there will be two additional monitoring sites in the eastern U.S. that each have a chance of being a nonattainment or maintenance receptor in 2021. These monitoring sites are Houston-Deer Parksite 492011039 and Dallas-Grapevine site 48439007). The other commenter said that their analysis shows that there will be up to four additional nonattainment receptors in 2021 in the eastern U.S. outside of Connecticut. These monitoring sites include the Chicago-Northbrook, Illinois monitoring site 170314201, the Michigan City, Indiana monitoring site 180910005, the El Paso, Texas monitoring site 481410037, and the Dallas-Eagle Mountain Rock monitoring site 484390075.

Response: EPA agrees with these commenters that the four monitoring sites identified by EPA as receptors in Table V.C-1 will be receptors in 2021. EPA also agrees that there would be no change in the upwind states covered by this rule if the Madison, Connecticut maintenance-only receptor is a nonattainment receptor rather than maintenance-only receptor. As described above, a maintenance-only receptor is a monitoring site that is at risk of being in nonattainment under meteorological conditions that are more conducive than average for ozone formation. Also, upwind states that are linked to maintenance-only receptors are evaluated by EPA using the same approach as those upwind states linked to nonattainment receptors in EPA's analysis of significant contribution in step 3 of the 4-step transport framework. Regarding the Greenwich, Connecticut monitoring site, EPA's contribution data, as provided in the docket for this rule, shows that there would be no additional upwind states covered by this rule if this monitoring site was included as a receptor in 2021. That is, all the upwind states that are linked to this monitoring site, using a 1 percent of the NAAQS threshold, are also linked to one or more of the other 2021 nonattainment and/or maintenance receptors in Connecticut that are identified in Table V.C-1.

EPA disagrees with the commenters that the six additional monitoring sites (i.e., Chicago/Northbrook, Dallas/Eagle Mountain Rock, Dallas/Grapevine, El Paso, Houston/Deer Park, and Michigan City) will be nonattainment or maintenance receptors in 2021. First, as explained in the Air Quality TSD, these sites are not identified in the methodology EPA uses to identify nonattainment and maintenance receptors. These conclusions are bolstered by EPA's review of measured design values for the period 2012 through 2019 at each of these six monitoring sites (see Table V.C-2). These data show that each of these sites, except for the site in Michigan City, is not measuring nonattainment based on their 2019 design value, which are the most recent official design values based on state-certified data. Moreover, the monitoring site in El Paso has not measured a violation during this entire eight-year time period; the Houston/Deer Park site has not measured a violation in the most recent 6 years; the Dallas/Eagle Mountain Lake site has not measured a violation in the most recent 4 years; the Chicago/Northbrook site has measured only 1 violation in the most recent 6 years; and the Dallas/Grapevine site has measured only one violation in the most recent 4 years. At the Michigan City site, there are no official measured design values in 2016, 2017, and 2018 because there was no valid fourth high MDA8 ozone concentration in 2016. As a result, the data at this site did not meet the criteria in EPA's modeling guidance for calculating valid future year design values. As such, EPA has not calculated projected design values nor any contributions for this site.

Table V.C-2—Ozone Design Values at Monitoring Sites Identified as Receptors by Commenters

Site IDStateCountySite name20122013201420152016201720182019
170314201ILCookChicago/Northbrook7877746871727774
180910005INLaPorteMichigan City8383796876
481410037TXEl PasoEl Paso7272727170717375
482011039TXHarrisHouston/Deer Park8479726967687175
484390075TXTarrantDallas/Eagle Mountain Lake8281797672717073
484393009TXTarrantDallas/Grapevine8686807875757675

Comment: In the proposed rule EPA requested comment on applying the “3 x 3” approach and the “no water cell” approach, described above, to identify modeled-grid cells for use in projecting ozone design values to a future year. One commenter said that both the “3 x 3” and “no water cell” approaches are acceptable, a second commenter supported the use of the “no water cell” approach, while a third commenter suggested that EPA modify the “no water cell” approach to exclude from the calculation of projected design values any data from the grid cell containing the monitoring site, if the monitor grid cell is also dominated by water.

Response: EPA has considered these comments and will continue to rely upon the “no water cell” approach used for the proposed rule to calculate projected design values at monitoring sites in coastal areas. The alternative suggested by one commenter to exclude model data from the grid cell containing the monitoring site, if that grid cells is classified as a “water” grid cell, ignores the modeling data for the location of the monitoring state which is contrary to EPA's air quality modeling guidance. This guidance recommends that the calculation of ozone relative response factors, which are used in projecting future year design values, include the modeled data in grid cells immediately surrounding the monitoring site along with the grid cell in which the monitor is located. For coastal monitoring sites, the grid cell in which the monitor is located is more likely to be representative of the monitor locations, than adjacent over-water grid cells. In this regard, the approach suggested by the commenter is too restrictive in that modeling data in the grid cell containing the monitoring site would never be used in projecting design values for that monitor.

D. Pollutant Transport From Upwind States

1. Air Quality Modeling To Quantify Upwind State Contributions

This section documents the procedures EPA used to quantify the impact of emissions from specific upwind states on 2021 8-hour design values for the identified downwind nonattainment and maintenance Start Printed Page 23083receptors. EPA used CAMx photochemical source apportionment modeling to quantify the impact of emissions in specific upwind states on downwind nonattainment and maintenance receptors for 8-hour ozone. CAMx employs enhanced source apportionment techniques that track the formation and transport of ozone from specific emissions sources and calculates the contribution of sources and precursors to ozone for individual receptor locations. The strength of the photochemical model source apportionment technique is that all modeled ozone at a given receptor location in the modeling domain is tracked back to specific sources of emissions and boundary conditions to fully characterize culpable sources.

EPA performed nationwide, state-level ozone source apportionment modeling using the CAMx Ozone Source Apportionment Technology/Anthropogenic Precursor Culpability Analysis (OSAT/APCA) technique [101] to quantify the contribution of 2023 base case NOX and VOC emissions from all sources in each state to projected 2023 ozone design values at air quality monitoring sites. The CAMx OSAT/APCA model run was performed for the period May 1 through September 30 using the projected 2023 base case emissions and 2016 meteorology for this time period. As described below, in the source apportionment modeling the Agency tracked (i.e., tagged) the amount of ozone formed from anthropogenic emissions in each state individually as well as the contributions from other sources (e.g., natural emissions).

To determine upwind contributions in 2021 the Agency applied the contributions from the 2023 modeling in a relative manner to the 2021 ozone design values. The analytic steps in the process are as follows:

(1) Calculate the 8-hour average contribution from each source tag to each monitoring site for the time period of the 8-hour daily maximum modeled concentrations in 2023;

(2) Average the contributions and concentrations for each of the top 10 modeled ozone concentration days in 2023 [102] and then divide the average contribution by the corresponding concentration to obtain a Relative Contribution Factor (RCF) for each monitoring site; and

(3) Multiply the 2021 design values by the 2023 RCF at each site to produce the average contribution metric values in 2021.[103] The resulting 2021 contributions from each tag to each monitoring site in the U.S. along with additional details on the source apportionment modeling and the procedures for calculating contributions can be found in the AQM TSD.

In the source apportionment model run, EPA tracked the ozone formed from each of the following tags:

  • States—anthropogenic NOX and VOC emissions from each state tracked individually (emissions from all anthropogenic sectors in a given state were combined);
  • Biogenics—biogenic NOX and VOC emissions domain-wide (i.e., not by state);
  • Boundary Concentrations—concentrations transported into the modeling domain;
  • Tribes—the emissions from those tribal lands for which the Agency has point source inventory data in the 2016v1 emissions modeling platform (EPA did not model the contributions from individual tribes);
  • Canada and Mexico—anthropogenic emissions from sources in the portions of Canada and Mexico included in the modeling domain (EPA did not model the contributions from Canada and Mexico separately);
  • Fires—combined emissions from wild and prescribed fires domain-wide (i.e., not by state); and
  • Offshore—combined emissions from offshore marine vessels and offshore drilling platforms.

The contribution modeling provided contributions to ozone from anthropogenic NOX and VOC emissions in each state, individually. The contributions to ozone from chemical reactions between biogenic NOX and VOC emissions were modeled and assigned to the “biogenic” category. The contributions from wildfire and prescribed fire NOX and VOC emissions were modeled and assigned to the “fires” category. That is, the contributions from the “biogenic” and “fires” categories are not assigned to individual states nor are they included in the state contributions.

The average contribution metric is intended to provide a reasonable representation of the contribution from individual states to the projected 2021 design value, based on modeled transport patterns and other meteorological conditions generally associated with modeled high ozone concentrations at the receptor. An average contribution metric constructed in this manner is beneficial since the magnitude of the contributions is directly related to the magnitude of the design value at each site.

The largest contribution from each state that is the subject of this rule to 8-hour ozone nonattainment and maintenance receptors in downwind states in 2021 is provided in Table V.D-1.

Table V.D-1—Largest Contribution to Downwind 8-Hour Ozone Nonattainment and Maintenance Receptors in 2021

Upwind stateLargest downwind contribution to nonattainment receptors for ozone (ppb)Largest downwind contribution to maintenance-only receptors for ozone (ppb)
Alabama0.110.27
Start Printed Page 23084
Arkansas0.180.15
Illinois0.810.80
Indiana1.261.08
Iowa0.170.22
Kansas0.130.11
Kentucky0.870.79
Louisiana0.274.68
Maryland1.211.56
Michigan1.711.62
Mississippi0.100.37
Missouri0.360.33
New Jersey8.625.71
New York14.4412.54
Ohio2.552.35
Oklahoma0.200.14
Pennsylvania6.865.64
Texas0.590.36
Virginia1.301.69
West Virginia1.491.55
Wisconsin0.230.23

Comment: One commenter said that the future year average contribution metric should be calculated using the modeled contributions on the same days that were used to calculate the RRFs for projecting future ozone design values.

Response: EPA believes that its approach, as described above, for calculating the future year average contribution metric provides a more technically reliable estimate of contributions than the method suggested by the commenter. In calculating the average contribution metric, EPA uses modeled contributions on the 10 days in the future year with the highest model-predicted concentrations.[104] In part because the formation of ozone from precursor emissions can be highly nonlinear and dependent on meteorological conditions, the response of ozone to emission reductions can vary from day to day. In this regard, the days with the highest model-predicted ozone concentrations in the 2016 base year that are used for projecting ozone design values may not be among the highest ozone days in the future analytic year. In this situation, the calculation of the contribution metric could exclude days with higher concentrations in the future year in favor of lower future-concentration days that happened to correspond to the highest days in 2016. The problems with basing the calculation of future year average contributions on the days that were used to project design values are illustrated in Table V.D-2. Table V.D-2 includes the data for all the days that were either used to project design values and/or to calculate the average contribution values from each upwind state to a particular receptor. The data in the “2016 Modeled” column are the 2016 base year MDA8 ozone concentrations and the data in the “2023 Modeled” column are the MDA8 ozone concentrations in 2023. The data in the table are ranked based on the magnitude of the 2016 MDA8 concentrations.[105] Comparing the 2023 MDA8 ozone concentrations to the corresponding 2016 values shows that the days with the highest MDA8 ozone concentrations in 2016 are not the same days as the highest MDA8 ozone concentrations in 2023. Of importance, the top 10 days based on 2016 model predictions includes five days with 2023 MDA8 ozone concentrations below 60 ppb. In calculating the average contribution metric EPA excludes from the calculation all days with future year modeled MDA8 concentrations below 60 ppb. Thus, using EPA's approach the average contribution metric in this example would be calculated based on daily contribution data for the top 6 MDA8 concentration days in 2023, because the remaining top 10 future year days are below 60 ppb (i.e., 05/06, 05/13, 06/08, 09/12, and 09/28). Moreover, even though the concentration on the sixth-highest day in 2023 is 60 ppb, the contribution data on this day would be excluded from the calculations because this day is not among the top 10 days used to project design values.

Table V.D-2—MDA8 Ozone Concentrations in 2016 Used To Project Design Values and the 2023 Modeled MDA8 Concentrations on the Same Days (ppb)

Date2016 Rank2016 Modeled2023 Rank2023 Modeled
07/01179.4369.1
06/27279.4174.5
05/12376.4269.7
Start Printed Page 23085
06/08471.9759.5
09/12569.41351.8
09/28668.51056.3
08/09768.5561.0
05/13867.8957.1
09/19967.5461.3
05/061067.1858.1
08/081165.81254.4
07/211265.21155.9
06/301364.81450.0
05/101463.4660.0

It is obviously impossible for EPA, or anyone, to predict which exact days in a future year will have high ozone levels, nor does it make sense to analyze contribution on modeled days of low ozone concentration. EPA's methodology is reasonable in projecting where ozone problems are likely to recur in a future year and analyzing who is contributing to those problems under the conditions for high ozone formation in those locations.

Comment: One commenter said that EPA should base the calculation of the future year contribution metric on days with measured exceedances of the NAAQS. Specifically, the comment asked EPA to examine the 2016 measured concentrations at receptors in Connecticut to ensure that the contribution from Illinois to these receptors was calculated on days when the monitor measured exceedances.

Response: EPA continues to believe that the future year contribution metric should be based on the highest ozone concentration days in the future year. However, as a sensitivity analysis EPA recalculated the average contribution from Illinois to the three receptors in Connecticut using the daily contributions on days with measured exceedances of the NAAQS, after applying the 60 ppb screening criteria to eliminate from the calculations those days with future year model-predicted MDA8 ozone concentrations below 60 ppb. The results of this sensitivity analysis, as provided in Table V.D-3, show that Illinois would contribute above the 1 percent of the NAAQS screening threshold to each of the three Connecticut receptors using the approach suggested by the commenter.

Table V.D-3—Contributions From Illinois (ppb) to Receptors in Connecticut

ReceptorContribution based on EPA's methodContribution based on measured exceedance days
Stratford0.690.98
Westport0.810.76
Madison0.801.03

2. Application of Screening Threshold

EPA evaluated the magnitude of the contributions from each upwind state to downwind nonattainment and maintenance receptors. In step 2 of the good neighbor framework, EPA uses an air quality screening threshold to identify upwind states that contribute to downwind ozone concentrations in amounts sufficient to “link” them to these to downwind nonattainment and maintenance receptors. The contributions from each of the CSAPR Update states to each downwind nonattainment and/or maintenance receptor that were used for the step 2 evaluation can be found in the AQM TSD.

As discussed above in section IV, EPA is not reopening the air quality screening threshold of 1 percent of the NAAQS used in the CSAPR Update. Therefore, as in the CSAPR Update, EPA uses an 8-hour ozone value for this air quality threshold of 0.75 ppb as the quantification of 1 percent of the 2008 ozone NAAQS.

Comment: Several commenters said that EPA's 1 percent of the NAAQS threshold is too low and that, instead, a threshold of 1 ppb or 2 ppb should be used as the contribution screening threshold in step 2.

Response: As noted above, the Agency is not reopening the use of the 1 percent threshold in this action to address the remand of the CSAPR Update. This action is taken in response to the Wisconsin remand and to complete the good neighbor obligations that were partially addressed in the CSAPR Update. It is entirely appropriate to continue to apply the same screening threshold to identifying receptors to fully address the outstanding obligations as EPA took in initially addressing them. Indeed, to do otherwise would be anomalous and pose a risk of inconsistent requirements for different states. While the Agency is not reopening the application of the 1 percent threshold in this action on remand, explanation for how this value was originally derived is available in the CSAPR rulemaking in 2011. See 76 FR 48208, 48237-38. Further, in the CSAPR Update, EPA re-analyzed the threshold for purposes of the 2008 ozone NAAQS and determined it was appropriate to continue to apply this threshold. EPA compared the 1 percent threshold to a 0.5 percent of NAAQS threshold and a 5 percent of NAAQS threshold. EPA found that the lower threshold did not capture appreciably more upwind state contribution compared to the 1 percent threshold, while the 5 percent threshold allowed too much upwind state contribution to drop out from further analysis.[106] EPA therefore determined the 1 percent threshold was appropriate for purposes of good neighbor obligations under the 2008 ozone NAAQS. This determination was not challenged in the Wisconsin case. Thus, EPA is applying the 1 percent threshold at step 2, consistent with its initial analysis of obligations in the CSAPR Update and without reopening its prior determination on this issue in that rule.

a. States That Contribute Below the Screening Threshold

Of the 21 states that are the subject of this final rule, EPA has determined that the contributions from each of the Start Printed Page 23086following states to nonattainment and/or maintenance-only receptors in the 2021 analytic year are below the threshold: Alabama, Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Texas, and Wisconsin. Because these states are considered not to contribute to projected downwind air quality problems, EPA is determining that the CSAPR Update FIPs for these states (or, in the case of Alabama and Missouri, the SIP revisions later approved to replace the states' CSAPR Update FIPs) are a complete remedy to address their significant contribution under the good neighbor provision for the 2008 ozone NAAQS. These states remain subject to the ozone season NOX emission budgets established in the CSAPR Update, and EPA is not reopening the determinations in the CSAPR Update regarding these states.[107]

b. States That Contribute at or Above the Screening Threshold

In this final rule, states with remanded emission budgets under the CSAPR Update that contribute to a specific receptor in an amount at or above the screening threshold in 2021 are considered linked to that receptor. The ozone contributions and emissions (and available emission reductions) for these states are analyzed further at step 3, as described in section VI, to determine whether and to what extent emission reductions might be required from each state.

Based on the maximum downwind contributions in Table V.D-1, the step 2 analysis identifies that the following 11 states contribute at or above the 0.75 ppb threshold to downwind nonattainment receptors: Illinois, Indiana, Kentucky, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West Virginia. Based on the maximum downwind contributions in Table V.D-1, the following 12 states contribute at or above the 0.75 ppb threshold to downwind maintenance-only receptors: Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West Virginia. The levels of contribution between each of these linked upwind state and downwind nonattainment receptors and maintenance-only receptors are provided in Table V.D-2 and Table V.D-3, respectively.

Table V.D-2—Contribution (ppb) From Each Linked Upwind State to Downwind Nonattainment Receptors in 2021

Upwind stateNonattainment receptors
Stratford, CTWestport, CT
Illinois0.690.81
Indiana0.991.26
Kentucky0.780.87
Louisiana0.270.27
Maryland1.211.20
Michigan1.161.71
New Jersey7.708.62
New York14.4214.44
Ohio2.342.55
Pennsylvania6.726.86
Virginia1.291.30
West Virginia1.451.49

Table V.D-3—Contribution (ppb) From Each Linked Upwind State to Downwind Maintenance-Only Receptors in 2021

Upwind stateMaintenance-only receptors
Madison, CTHouston, TX
Illinois0.800.02
Indiana1.080.02
Kentucky0.790.02
Louisiana0.154.68
Maryland1.560.00
Michigan1.620.00
New Jersey5.710.00
New York12.540.00
Ohio2.350.00
Pennsylvania5.640.00
Virginia1.690.00
West Virginia1.550.00

In conclusion, as described above, states with contributions that equal or exceed 1 percent of the NAAQS to either nonattainment or maintenance receptors are identified as “linked” at step 2 of the good neighbor framework and warrant further analysis for significant contribution to nonattainment or interference with maintenance under step 3. EPA is determining that the following 12 States are linked at step 2: Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio, Pennsylvania, Virginia, and West Virginia.

VI. Quantifying Upwind-State NOX Reduction Potential To Reduce Interstate Ozone Transport for the 2008 Ozone NAAQS

A. The Multi-Factor Test

This section describes EPA's methodology at step 3 of the 4-step framework for identifying upwind emissions that constitute “significant” contribution for the states subject to this final rule. This analysis focuses on the 12 states linked at steps 1 and 2 of the framework, as identified in the sections above. Following the existing framework as applied in the CSAPR Update, EPA's assessment of linked upwind state emissions reflects analysis of uniform NOX emission control stringency. The analysis has been extended to include assessment of non-EGU sources in addition to EGU sources in the linked upwind states.

Each level of uniform NOX control stringency is characterized by a set of pollution control measures. EPA applies a multi-factor test—the same multi-factor test that was used in the CSAPR and the CSAPR Update [108] —to evaluate increasing levels of uniform NOX control stringency. The multi-factor test, which is central to EPA's step 3 quantification of significant contribution, considers cost, available emission reductions, and downwind air quality impacts to determine the appropriate level of uniform NOX control stringency that addresses the impacts of interstate transport on downwind nonattainment or maintenance receptors. The uniform NOX emission control stringency, represented by marginal cost (or a weighted average cost in the case of EPA's non-EGU analysis), also serves to apportion the reduction responsibility among collectively contributing upwind states. This approach to quantifying upwind state emission-reduction obligations using uniform cost was reviewed by the Supreme Court in EME Homer City Generation, which held that using such an approach to apportion emission reduction responsibilities among upwind states that are collectively responsible for downwind air quality impacts “is an efficient and equitable solution to the allocation problem the Good Neighbor Provision requires the Agency to address.” 572 U.S. at 519.

There are four stages in developing the multi-factor test: (1) Identify levels of uniform NOX control stringency; (2) evaluate potential NOX emission reductions associated with each identified level of uniform control stringency; (3) assess air quality improvements at downwind receptors for each level of uniform control stringency; and (4) select a level of control stringency considering the identified cost, available NOX emission reductions, and downwind air quality impacts, while also ensuring that emission reductions do not unnecessarily over-control relative to Start Printed Page 23087the contribution threshold or downwind air quality.

Comment: Some commenters suggested EPA also consider regulating volatile organic compounds (VOCs) as it represents another precursor to ozone formation. They assert EPA's failure to reduce significant contributions to downwind nonattainment/maintenance by reducing upwind VOC emissions disproportionately harms communities of color, low-income communities, and children, perpetuating environmental injustice.

Response: EPA agrees that VOCs are a precursor along with NOX in forming ground-level ozone and that ozone formation chemistry can be “NOX-limited”, where ozone production is primarily determined by the amount of NOX emissions or “VOC-limited”, where ozone production is primarily determined by the amount of VOC emissions.[109] EPA also acknowledges that VOCs can contain toxic chemicals that affect public health. EPA's obligation in this action is to complete the elimination of significant contribution to nonattainment or interference with maintenance of NAAQS in other states for 12 states in the East to meet the requirements of section 110(a)(2)(D)(i)(I) of the Act. Provisions for local NAAQS attainment and exposure to toxic pollutant concentrations are addressed by other sections of the statute. EPA and others have long regarded NOX to be the more significant ozone precursor in the context of interstate ozone transport.[110] In response to this comment, EPA examined the results of the contribution modeling performed for this rule to identify the portion of the ozone contribution attributable to anthropogenic NOX emissions versus VOC emissions from each linked upwind state to each downwind receptor. Table VI.A provides the ozone contribution from each upwind state linked to the receptors in Connecticut along with the percent (in parenthesis) of the contribution that is formed under “NOX-limited” photochemistry. The data show that NOX is the determinative precursor for over 80 percent of the total contribution from each upwind state to each of these receptors. In addition to the Connecticut receptors, ozone primarily formed from NOX emissions is 95 percent of the 4.58 ppb contribution from Louisiana to the receptor in Harris County, Texas. Therefore, EPA's review of the data leads to the finding that, as proposed, a focus on NOX emission reductions is appropriate for the purpose of addressing interstate ozone transport.

Table VI.A.—Contribution (ppb) From Each Linked Upwind State to Receptors in Connecticut and the Percent of the Contribution From NOX

StateReceptorILINKYMDMINJNYOHPAVAWV
CTStratfordNot Linked0.96 (95%)0.76 (96%)1.18 (90%)1.13 (95%)7.48 (83%)14.01 (81%)2.27 (95%)6.53 (93%)1.25 (93%)1.41 (97%)
CTWestport0.79 (94%)1.23 (95%)0.85 (96%)1.18 (89%)1.67 (94%)8.44 (83%)14.14 (81%)2.50 (95%)6.72 (92%)1.27 (92%)1.45 (96%)
CTMadison0.78 (95%)1.04 (96%)0.77 (96%)1.51 (91%)1.57 (95%)5.53 (84%)12.15 (86%)2.27 (95%)5.47 (92%)1.63 (93%)1.51 (96%)

For both EGUs and non-EGUs, section VI.B describes the available NOX emission controls considered and their associated cost levels (in 2016$). Section VI.C discusses EPA's application of that information to assess emission reduction potential of the identified control stringencies. Finally, section VI.D describes EPA's assessment of associated air quality impacts and EPA's subsequent identification of appropriate control stringencies considering the relevant factors (cost, available emission reductions, and downwind air quality impacts). As discussed in greater detail in section VI.D, the multi-factor test informed EPA's determination of appropriate EGU NOX ozone season emission budgets necessary to reduce emissions that significantly contribute to nonattainment or interfere with maintenance of the 2008 ozone NAAQS for the 2021 ozone season and subsequent control periods.

This multi-factor approach is consistent with EPA's approach in the prior CSAPR and CSAPR Update actions. In addition, as was done in the CSAPR Update, EPA evaluated possible over-control by determining if an upwind state is linked solely to downwind air quality problems that could have been resolved at a lower representative cost threshold, or if upwind states could reduce their emissions below the 1 percent air quality contribution threshold at a lower representative cost threshold. This analysis is described in section VI.D.

In the proposed rule, EPA identified a control stringency that reflects the optimization of existing SCR controls and installation of state-of-the-art NOX combustion controls at EGUs, with an estimated marginal cost of $1,600 per ton. As explained in greater detail in section VI.D, EPA is finalizing an EGU control stringency that also includes optimizing existing SNCR controls. Application of the multi-factor test to non-EGU sources has led EPA to conclude, as the Agency proposed, that emission reductions from non-EGU sources are not necessary to address significant contribution or interference with maintenance under the 2008 ozone NAAQS.

B. Identifying Levels of Control Stringency

1. EGU NOX Mitigation Strategies

In identifying levels of uniform control stringency for EGUs, EPA reassessed the same NOX emission controls that it had analyzed in the CSAPR Update, all of which are considered to be widely available in this sector: (1) Fully operating existing SCR, including both optimizing NOX removal by existing operational SCRs and turning on and optimizing existing idled SCRs; (2) installing state-of-the-art NOX combustion controls; (3) fully operating existing SNCRs, including both optimizing NOX removal by existing operational SNCRs and turning on and optimizing existing idled SNCRs; (4) installing new SNCRs; and (5) installing new SCRs. For the reasons explained in the EGU NOX Mitigation Strategies Final Rule TSD included in the docket for this final rule, EPA determined that for the regional, multi-state scale of this rulemaking, only EGU NOX emission controls 1 and 3 are possible for the 2021 ozone season (fully operating existing SCRs and SNCRs). As discussed Start Printed Page 23088in section VI.B.1.b, EPA finds that it is not possible to install state-of-the-art NOX combustion controls by the 2021 ozone season on a regional scale. EPA determined state-of-the-art NOX combustion controls at EGUs are available by the beginning of the 2022 ozone season.

a. Optimizing Existing SCRs

Optimizing (i.e., turning on idled or improving operation of partially operating) existing SCRs can substantially reduce EGU NOX emissions quickly using investments that have already been made in pollution control technologies. With the promulgation of the CSAPR Update, most operators improved their SCR performance and have continued to maintain that level of improved operation. However, this SCR performance is not universal and some drop has been observed as the CSAPR Update ozone-season allowance price has declined steadily since 2017. For example, recent power sector data from 2019 reveal that, in some cases, operating units have SCR controls that have been idled or are operating partially, and therefore suggest that there remains reduction potential through optimization.[111] EPA determined that optimizing all of these remaining SCRs in the 12 linked states is a readily available approach for EGUs to reduce NOX emissions.

EPA estimates a representative cost of optimizing SCR controls to be approximately $1,600 per ton. EPA's analysis of this emission control is informed by comment on the CSAPR Update proposed rule and updated information on operation and industrial-input costs that have become available since the CSAPR Update.[112] While the costs of optimizing existing, operational SCRs include only variable costs, the cost of optimizing SCR units that are currently idled back into service considers both variable and fixed costs. Variable and fixed costs include labor, maintenance and repair, parasitic load, and ammonia or urea for use as a NOX reduction reagent in SCR systems. EPA performed an in-depth cost assessment for all coal-fired units with SCRs. More information about this analysis is available in the EGU NOX Mitigation Strategies Final Rule TSD, which is found in the docket for this rule. The TSD notes that, for the subset of SCRs that are already partially operating, the cost of optimizing is often much lower than the $1,600 per ton marginal cost and often under $800 per ton.

EPA is using the same methodology to identify SCR performance as it did in the CSAPR Update. To estimate EGU NOX reduction potential from optimizing, EPA considers the difference between the non-optimized NOX emission rates and an achievable operating and optimized SCR NOX emission rate. To determine this rate in the CSAPR Update, EPA evaluated nationwide coal-fired EGU NOX ozone season emissions data from 2009 through 2015 and calculated an average NOX ozone season emission rate across the fleet of coal-fired EGUs with SCR for each of these seven years. EPA found it prudent to not consider the lowest or second-lowest ozone season NOX emission rates, which may reflect new SCR systems that have all new components (e.g., new layers of catalyst). Data from these new systems are not representative of ongoing achievable NOX emission rates considering broken-in components and routine maintenance schedules. To identify the potential reductions from SCR optimization in this final action, EPA followed the same methodology and incorporated the latest reported coal-fired EGU NOX ozone season emissions data. EPA updated the timeframe to include the most recent and best available operational data (i.e., 2009 through 2019). Considering the emissions data over the full time period of available data results in a third-best rate of 0.08 pounds per million British thermal units (lb/mmBtu). EPA notes that over half of the SCR-controlled EGUs achieved a NOX emission rate of 0.068 lbs/mmBtu or less over their third-best entire ozone season. Moreover, for the SCR-controlled coal units that EPA identified as having a 2019 emission rate greater than 0.08 lb/mmBtu, EPA verified that in prior years, the majority (approximately 95 percent) of these same units had demonstrated and achieved a NOX emission rate of 0.08 lb/mmBtu or less on a seasonal and/or monthly basis. This further supports EPA's determination that 0.08 lb/mmBtu reflects a reasonable emission rate for representing SCR optimization in quantifying state emission budgets as discussed in section VII.B. This fleet-level emission rate assumption of 0.08 lb/mmBtu for non-optimized units reflects, on average, what those units would achieve when optimized. Some of these units may achieve rates that are lower than 0.08 lb/mmBtu, and some units may operate above that rate based on unit-specific configuration and dispatch patterns.

EPA evaluated the feasibility of optimizing idled SCRs for the 2021 ozone season. Based on industry past practice, EPA determined that idled controls can be restored to operation quickly (less than two months). This timeframe is informed by many electric utilities' previous long-standing practice of utilizing SCRs to reduce EGU NOX emission during the ozone season while putting the systems into protective lay-up during the non-ozone season months. For example, this was the long-standing practice of many EGUs that used SCR systems for compliance with the NOX Budget Trading Program. It was quite typical for SCRs to be turned off following the September 30 end of the ozone season control period. These controls would then be put into protective lay-up for several months of non-use before being returned to operation by May 1 of the following ozone season.[113] Therefore, EPA believes that optimization of existing SCRs is possible for the portion of the 2021 ozone season covered under this final rule.

The vast majority of SCR controlled units (nationwide and in the 12 linked states) are already partially operating these controls during the ozone season based on historical 2019 emissions rates. EPA believes that this widely demonstrated seasonal behavior of turning on idled SCRs also supports the Agency's determination that optimizing existing SCR systems currently being operated to some degree within the ozone season, which would necessitate fewer changes to SCR operation relative to restarting idled systems, is also feasible for the 2021 ozone season. Full operation of existing SCRs that are already operating to some extent involves increasing reagent (i.e., ammonia or urea) flow rate, and maintaining and replacing catalyst to sustain higher NOX removal rate operations. Increasing NOX removal by SCR controls that are already operating can be implemented by procuring more reagent and catalyst. EGUs with SCR routinely procure reagent and catalyst as Start Printed Page 23089part of ongoing operation and maintenance of the SCR system. In many cases, where EPA has identified EGUs that are operating their SCR at non-optimized NOX removal efficiencies, EGU data indicate that these units historically have achieved more efficient NOX removal rates. Therefore, EPA determined that optimizing existing SCRs currently being operated could generally be done by reverting back to previous operation and maintenance plans. Regarding full operation activities, existing SCRs that are only operating at partial capacity still provide functioning, maintained systems that may only require increased chemical reagent feed rate up to their design potential and catalyst maintenance for mitigating NOX emissions. Units must have adequate inventory of chemical reagent and catalyst deliveries to sustain operations. Considering that units have procurement programs in place for operating SCRs, this may only require updating the frequency of deliveries. This may be accomplished within a few weeks.

Comment: EPA received comments supporting the 0.08 lb/mmBtu emission rate as achievable and, according to some commenters, conservative. Some of these commenters went on to provide their own analysis demonstrating that the 0.08 lb/mmBtu was achievable not only on average for the non-optimized fleet, but also for these individual units and that the resulting state emission budgets were likewise achievable. Some commenters suggested that the rate should be lower and premised on EPA using a longer historical baseline (e.g., extending baseline back to year 2006) and relying on the first- or second-best year instead of the third best year of SCR performance. In addition to supporting the 0.08 lb/mmBtu optimization rate as viable for 2021, these same commenters noted the 2021 attainment data and suggested implementation by 2021 was not only achievable, but necessary under Clean Air Act requirements and the Wisconsin directive.

Response: As explained above, EPA chose 2009 for the start of its baseline period of SCR performance examination because that is the first year of annual compliance under the CAIR NOX program. The analysis focuses on the third best ozone season average rate because EPA believes that the first or second best rate, as discussed in the CSAPR Update final rule, could continue to capture disproportionately new SCR components and/or the onset of new regulatory programs and does not necessarily reflect achievable ongoing NOX emission rates. Therefore, EPA is finalizing analysis using the third best rate starting from 2009—consistent with its approach in the CSAPR Update.

Comment: Other commenters suggested that EPA should apply a higher emission rate than 0.08 lb/mmBtu premised on considerations such as: A generally reduced average capacity factor for coal units in recent years, the age of the boiler, coal rank (bituminous or subbituminous), or other unit-specific considerations that make the 0.08 lb/mmBtu rate unattainable for a specific unit. They also suggested that EPA's determination of the rate should be premised on EPA using a shorter historical baseline (e.g., shortening the baseline to year 2013).

Response: EPA did not find sufficient justification to apply a higher average emission rate than 0.08 lb/mmBtu or for shortening the baseline to exclude representative operational data starting in 2009. EPA found that some commenters were misunderstanding or misconstruing both EPA's assumption and implementation mechanism as a unit-level requirement for every SCR-controlled unit instead of a reflection of a fleet-wide average based on a third-best rate. The commenters' observation—that 0.08 lb/mmBtu may be difficult for some units to achieve or may not be a preferred compliance strategy for a given unit given its dispatch levels—does not contradict EPA's assumption, but rather supports its methodology and assumptions. As EPA pointed out in the proposed rule, “this fleet-level emission rate assumption of 0.08 lb/mmBtu for non-optimized units reflects, on average, what those units would achieve when optimized. Some of these units may achieve rates that are lower than 0.08 lb/mmBtu, and some units may operate above that rate based on unit-specific configuration and dispatch patterns.” [114] In other words, EPA is using this assumption as the average performance of a unit that optimizes its SCR, recognizing that heterogeneity within the fleet will likely lead some units to overperform and others to underperform this rate. Moreover, a review of unit-specific historical data indicates that this is a reasonable assumption: Not only has the group of units with SCR optimization potential demonstrated they can perform at or better than the 0.08 lb/mmBtu rate on average, but 95 percent of the individual units in this group have met this rate on a seasonal and/or monthly basis based on their reported historical data.[115] Additionally, EPA's examination of units with the largest emission reduction potential based on SCR optimization levels of 0.08 lb/mmBtu indicates the ability of units to improve emission rate performance. As an example, Miami Fort Unit 7 had considerably more hours operating at a 70 to 79 percent capacity factor in 2019 compared to previous years. However, Miami Fort Unit 7's ozone-season NOX emission rate substantially increased in 2019 compared to previous years. This runs counter to the notion that an increase in emission rates is purely driven by reduced capacity factor, as suggested by commenters. This substantial deterioration in the median emission rate performance is observable even when comparing specific hours in 2019 to specific hours in prior years when the unit operated in the same 70 to 79 percent capacity factor range. In fact, in 2019 the unit experienced notable emission rate increases from prior years across multiple capacity factor ranges as low as 40 percent to as high as 80 percent. This type of data indicates instances where the increase in emission rate (and emissions) is not necessitated by load changes but is more likely due to the erosion of the existing incentive to optimize controls (i.e., the ozone-season NOx allowance price has fallen so low that unit operators find it more economic to surrender additional allowances instead of continuing to operate pollution controls at an optimized level). This type of decline in emission rate performance at some SCR-controlled units is what EPA disincentivizes with the full remedy nature of this action.

Start Printed Page 23090

EPA observed this pattern in other units identified in this rulemaking as having significant SCR optimization emission reduction potential. In the accompanying Emissions Data TSD for the supplemental notice that EPA recently released in a proceeding to address a recommendation submitted to EPA by the Ozone Transport Commission under CAA section 184(c), EPA noted, “In their years with the lowest average ozone season NOX emission rates in this analysis, these EGUs had relatively low NOX emission rates at mid- and high-operating levels; moreover, there was little variability in NOX emission rates at these operating levels. However, during the 2019 ozone season, these EGUs had higher NOX emission rates and greater variability in NOX emission rates across operating levels than in the past, particularly at mid-operating levels.” [116] That hourly data analysis, included in this docket, controls for operating level changes and still finds there to be instances across multiple SCR-controlled units in the 12-state region where hourly emission rates are increasing even when compared to the same load levels in previous years.

To the extent commenters have alleged that in recent years coal-fired EGUs have declined in capacity factor and that SCR performance declines at those lower operating levels, EPA notes that this does not necessarily result in a compliance feasibility challenge. First, as explained elsewhere in this section, EPA believes the 0.08 rate assumption is achievable on a fleetwide average basis. Second, the implementation mechanism of a mass-based emission trading program eliminates any compliance feasibility concern. Even if reduced operation of a unit were to affect the rate-based performance of a unit, it would also lower emissions-producing generation from that unit, which in turn reduces the number of allowances the unit operator must hold for compliance under this emission trading program. Commenters have failed to establish that compliance with the mass-based implementation mechanism of this rule is actually unachievable. Further, hourly data indicate that maintaining consistent SCR performance at lower capacity factors is possible. For example, the unit-level performance data in the graph below show the emission rate at a plant staying relatively low (consistent with our optimization assumption of 0.08 lb/mmBtu) and stable across a wide range of capacity factors.[117]

Start Printed Page 23091

Comment: EPA received comment suggesting that EPA subcategorize its SCR optimization rate assumption by coal rank (i.e., bituminous or subbituminous) as the difference between the two would imply that the 0.08 lb/mmBtu rate is not appropriate.

Response: EPA reviewed historical data for SCR operation by coal rank and assessed it against its 0.08 lb/mmBtu fleet-wide average assumption and did not find any change necessary or appropriate. EPA found many instances of both SCR-controlled coal units combusting subbituminous coal and SCR-controlled coal units combusting bituminous coal (including instances in earlier years where these very same units that EPA is identifying as having optimization potential relative to their 2019 levels) operating at emission rate levels at or below the 0.08 lb/mmBtu rate. In other words, although these units may not be operating at this emission rate in 2019, it is not due to coal rank as they have—in the vast majority of cases—met that rate in some period prior to 2019. In this case, the use of the average rate and the third best year accommodates any heterogeneity in emission rate that may stem from a unit's coal choice and makes 0.08 lb/mmBtu a reasonable average performance rate regardless of coal rank. Moreover, EPA notes that the covered fleet with the identified SCR optimization potential identified in this rule is composed of sources who have purchased and consumed both subbituminous and bituminous coal. The presence of both types of coal burning units within the region coupled with this observation that some units have utilized both types of coal, further support the use of a single fleet-wide average for purposes of estimating reduction potential and implementing state emission budgets—consistent with the CSAPR Update. This use of an average value, instead of two separate values is also consistent with EPA's approach in the CSAPR Update. EPA further examines and addresses this comment in the EGU NOX Mitigation Strategies Final Rule TSD.

Comment: EPA also received comment suggesting it should deviate from its approach in the CSAPR Update of using a nationwide set of data to establish a third best year, and instead use an average from just the 12 covered states.

Response: EPA reviewed the data and its methodology and evaluated it against its intention to identify a technology-specific representative emission rate for SCR optimization. In doing so, EPA did not identify any need to make the suggested change. EPA is interested in the performance potential of a technology, and a larger dataset provides a superior indication of that potential as opposed to a smaller, state-limited dataset. In both the CSAPR Update and in this rule, EPA appropriately relied on the largest dataset possible (i.e., nationwide) to derive technology performance averages that it then applied respectively to the CSAPR Update 22-state region and this rule's 12-state region. Finally, as noted above, in affirming the reasonableness of this approach, EPA examined the historical reported data (pre-2019) for the units in the 12 states with SCR optimization potential and found the nationwide derived average appropriate and consistent with demonstrated capability and performance of units within those states. That is, the vast majority of units for which this resulting emission rate assumption was being applied had demonstrated the ability to achieve this rate in some prior time period. This information is discussed further in the EGU NOx Mitigation Strategies Final Rule TSD in the docket.

In the proposed rule, EPA relied on the same SCR optimization timing assumptions it utilized in the CSAPR Update. EPA received comments on the feasibility of implementing SCR optimization mitigation measures by the start of the 2021 ozone season.

Comment: While many commenters supported the feasibility of 2021 ozone-season implementation by noting the “immediate availability” of SCR optimization, those that did not focused on two concerns: (1) That the engineering, procurement, and other steps required for SCR optimization were not feasible given the anticipated 1.5 months between rule finalization and the start of the 2021 ozone season and (2) that the short implementation time frame may not allow enough time Start Printed Page 23092for allowance trading to occur, and thus jeopardize allowance market liquidity and the overall that the implementation mechanism of a trading program.

Response: EPA disagrees that these concerns justify a change in approach, as explained below, and is finalizing the same SCR optimization timing assumptions it proposed.

As an initial matter, sources will have more than two months between the date of signature on this final action and the rule's effective date when the enhanced control stringency being adopted in this rule will take effect.[118] Further, EPA has determined that this implementation schedule is achievable and necessary in order to address good neighbor obligations by the July 20, 2021 Serious area attainment date for certain downwind receptors, in accordance with the Wisconsin decision of the D.C. Circuit.[119] While EPA observes that implementation of this control stringency is viable during the 2021 ozone season at the unit level as described below, it also notes that the flexible implementation mechanism of a trading program, starting bank, and safety valve (as discussed in VII.C.4) obviate any unit-specific compliance challenges raised by commenters.

As indicated in the discussion and graphics above, data in the EGU NOX Mitigation Strategies Final Rule TSD, and in the CSAPR Update, there is ample evidence of units restoring their optimal performance within a two-month timeframe. Not only do units reactivate SCR performance level at the start of an ozone-season when tighter emission limits begin, but unit-level data also shows instances where sources have demonstrated the ability to quickly alter their emission rate within an ozone-season and even within the same day in some cases. Moreover, this emission control is familiar to sources and was analyzed and included in the CSAPR Update emission budgets finalized in 2016. With this experience, and notice through the October 2020 proposed rule, as well as over two months from final rule to effective date, the viability of this emission control for the 2021 ozone season is entirely consistent with the 2-week to 2-month timeframe that EPA identified as reasonable in both the CSAPR Update and the proposed rule. Similar to prior rules, commenters provide some unit-level examples where it has taken longer. Also similar to those prior rules, EPA does not find those unit-level examples compelling in the context of its fleet average assumptions and in the implementation context of a trading program which provides compliance alternatives in the event a specific unit prefers more time to implement the control stringency. As noted in Wisconsin, “. . . all those anecdotes show is that installation can drag on when companies are unconstrained by the ticking clock of the law.” 938 F.3d at 330. Commenters also provide logistical details for certain engineering steps (e.g., procuring catalyst replacement) that will not be necessary in many instances to improve performance at existing SCRs. The majority of emission reductions from units with SCRs would be available within hours (from turning on and fully operating those existing control devices) even in the absence of catalyst that is not as optimally configured or with reagent sprayers that have not been recently tuned as commenters suggest they must be. And as noted previously, a prudent EGU operator has had since at least the publication date of the proposed rule in October 2020 to take steps to prepare for compliance, such as planning for the necessary products to run their controls.

EPA further disagrees with commenters' assertions that the 2021 emission budgets are not feasible. Claiming that ultimate compliance with the emissions trading program is infeasible ignores the flexibilities of EPA's trading program implementation mechanism, including the starting allowance bank and the “safety valve” mechanism for accessing even more allowances. EPA uses a fleet-wide average assumption that non-optimized units with SCR will optimize to 0.08 lb/mmBtu on average by 2021. EPA uses this average assumption in its derivation of state-emission budgets, but then implements the reductions through a trading program that provides sources the flexibility to operate at different emission rates, as they need only hold allowances adequate to cover their emissions for the relevant control period. Not every unit need implement this emission control or meet this rate in order to comply with the state budget under the trading program. For some units, the timing and rate performance will likely be easier to meet than estimated by EPA, creating space for other sources to achieve different rates on different schedules while collectively complying with the state emission budget. Additionally, while given the large amount of historical data demonstrating that units can operate their controls (often within hours of startup), unit operation (i.e., seasonal capacity factor) is another variable that operators can utilize to reduce seasonal emissions. In short, because compliance is based on seasonal emission totals, variation in emission rates is not on its own a barrier to meeting a seasonal total state emission target. In short, commenters concerned about 2021 implementation viability largely neglected these critical aspects of the trading program and did not provide any comprehensive state or system modeling showing the 2021 implementation of the state budgets was not achievable when factoring in the program's trading program. Instead of performing this critical evaluation step, commenters most often limited their arguments to a hypothetical unit-specific rate requirement evaluation, ignoring the broader mechanisms of EPA's quantification and implementation of good neighbor obligations.

EPA notes that historical emission data and program experience support its assumption regarding timing of these emission controls. Similar arguments regarding next-season implementation challenges were made against the CSAPR Update but were not borne out in the data as both unit-level and state-level emissions adjusted consistent with EPA's assumptions for that first season of implementation (the emission rate at SCR controlled units dropped by nearly half in the 2017 ozone season, the first ozone-season of the CSAPR Update implementation) when EPA examined this challenge in the context of EPA's Response to CAA section 126(b) petitions from Maryland and Delaware.[120] Moreover, the future modeling data, as well as some commenters' own analysis, supported the viability of EPA's 2021 implementation. Finally, some utilities with a significant footprint in this region even have their own near-term and medium-term emission reduction goals, which, if realized, reflect even Start Printed Page 23093more fleet alignment with emission reductions.[121 122] For all of these reasons, EPA determined it was not necessary to change its emission control implementation timing assumptions from those utilized in the CSAPR Update.

With regard to market liquidity concerns, EPA notes that those same concerns have been voiced in the lead-up to past trading programs but ultimately did not materialize. For example, a functioning allowance market formed and resulted in 100 percent compliance with the allowance holding requirements during the first year of implementation. See more discussion on this issue in section VII.C.3. EPA notes that the date by which sources must hold allowances to cover their emissions for the first control period under this final rule is June 1, 2022—more than 14 months after the date of signature of the rule. Moreover, shortly after the final rule's effective date and well before the end of the 2021 control period, the allowances allocated to most sources from both the state emission budgets and from the initial Group 3 bank will be recorded in sources' accounts and available for trading. Finally, as an additional measure promoting market liquidity, EPA will allow the use of Group 2 allowances at an 18:1 trade-in ratio to provide additional assurance to sources that allowances will be available, but ensuring that the cost of this compliance option is such that entities will take it only in the very unlikely event that access to such additional allowances proves to be necessary. The safety valve is described further in section VII.C.4.c., The presence of the safety valve, combined with the recordation of allowances from the state budgets and the starting bank shortly after the rule's effective date, should obviate any market liquidity concerns, as the number of allowances available for trading in the market for the first control period well in advance of the compliance deadline will accommodate a variety of compliance pathways and unit operational decisions.

b. Installing State-of-the-Art NOX Combustion Controls

EPA estimates that the representative cost of installing state-of-the-art combustion controls is comparable to, if not notably less than, the estimated cost of optimizing existing SCR (represented by $1,600 per ton). State-of-the-art combustion controls such as low-NOX burners (LNB) and over-fire air (OFA) can be installed and/or updated quickly and can substantially reduce EGU NOX emissions. In the 12 states linked to downwind receptors in this final rule, approximately 99 percent of coal-fired EGU capacity is equipped with some form of combustion control; however, the control configuration and/or corresponding emission rates at a few units indicate they do not currently have state-of-the-art combustion control technology. As discussed in EPA's response to comments below, the Agency has updated its NOX emission rates for upgrading existing combustion controls to state-of-the-art combustion control from the proposed rule, where EPA estimated a range of 0.139 to 0.155 lbs/mmBtu. In this final rule, EPA is determining that NOX emission rates of 0.146 to 0.199 lbs/mmBtu can be achieved on average depending on the unit's boiler configuration,[123] and, once installed, reduce NOX emissions at all times of EGU operation.

The feasibility of installing combustion controls was examined by EPA in the CSAPR where industry demonstrated the ability to install state-of-the-art LNB controls on a large unit (800 MW) in under six months when including the pre-installation phases (design, order placement, fabrication, and delivery).[124] In the proposed rule, EPA discussed comments it had received on the CSAPR Update regarding installation of combustion controls from the Institute of Clean Air Companies.[125] Those comments provided information on the equipment and typical installation time frame for new combustion controls, accounting for all steps, and noted it generally takes between 6-8 months on a typical boiler—covering the time through bid evaluation through start-up of the technology. The deployment schedule was described as:

  • 4-8 weeks—bid evaluation and negotiation
  • 4-6 weeks—engineering and completion of engineering drawings
  • 2 weeks—drawing review and approval from user
  • 10-12 weeks—fabrication of equipment and shipping to end user site
  • 2-3 weeks—installation at end user site
  • 1 week—commissioning and start-up of technology

Given the above timeframe of approximately 6 to 8 months to complete combustion control installation in the region, EPA is determining that the installation of state-of-the-art combustion controls is a readily available approach for EGUs to reduce NOX emissions by the start of the 2022 ozone season. More details on these analyses can be found in the EGU NOX Mitigation Strategies Final Rule TSD.

The cost of installing state-of-the-art combustion controls per ton of NOX reduced is dependent on the combustion control type and unit type. EPA estimates the cost per ton of state-of-the-art combustion controls to be $400 per ton to $1,200 per ton of NOX removed using a representative capacity factor of 70 percent. See the NOX Mitigation Strategies Final Rule TSD for additional details.

Comment: EPA received comment on the proposed timing, cost, and performance rate of combustion controls.

Response: EPA is finalizing its proposed assumptions on the cost and timing for upgrading combustion controls. These assumptions are consistent with the CSAPR Update. They are described above and further discussed in the RTC document and in the EGU NOx Mitigation Strategies Final Rule TSD. EPA is updating its assumed performance rate for state-of-the-art combustion controls from the proposed rule based on two factors. First, as commenters pointed out, EPA was in the process of updating these assumptions based on the latest representative-year data and an updated inventory of units with like controls. This update and corresponding emission rates were in the October 2020 NEEDS file placed in the docket for the proposed rule, but the data were not available in time to be included in EPA's proposed rule analysis. This adjustment raised the average emission rate assumption to 0.199 lb/mmBtu for combustion controls on dry bottom wall fired units and 0.146 lb/mmBtu for tangentially fired units. Additionally, commenters provided detailed analysis of how other unit considerations, such as coal rank, can result in large deviations from what has been historically demonstrated with this combustion control technology. Based Start Printed Page 23094on these comments and EPA's review of historical performance data for tangentially-fired units by coal rank with state-of-the-art combustion controls, EPA determined it was appropriate to use the 0.199 lb/mmBtu rate for both tangentially and wall-fired units in this final rule. As noted by commenters, many of the likely impacted units burn bituminous coal, and the 0.146 lb/mmBtu nationwide average for tangentially-fired (inclusive of subbituminous units) appeared to be below the demonstrated emission rate of state-of-the-art combustion controls for bituminous coal units of this boiler type. EPA notes that its analysis of illustrative units indicates the costs are often lower than the $1,600 per ton level EPA assumes in this rule. Similarly, the pervasiveness of this technology (i.e., 99 percent of units have some form of combustion controls) in response to previous EPA actions indicates the wide spread cost-effectiveness of this control and therefore its inclusion in the final EGU NOx emission budgets beginning in the 2022 ozone season (noting that the trading program gives units flexibility in compliance options to accommodate their specific circumstances).

c. Optimizing Already Operating SNCRs or Turning on Idled Existing SNCRs

Optimizing already operating SNCRs or turning on idled existing SNCRs can also reduce EGU NOX emissions quickly, using investments in pollution control technologies that have already been made. Compared to no post-combustion controls on a unit, SNCRs can achieve a 25 percent reduction on average in EGU NOX emissions (with sufficient reagent). They are less capital intensive but less efficient at NOx removal than SCRs. These controls are in use to some degree across the U.S. power sector. In the 12 states identified in this final rule, approximately 14 percent of coal-fired EGU capacity is equipped with SNCR. Recent power sector data suggest that, in some cases, SNCR controls have been operating less in 2019 relative to performance in prior years.[126]

In the proposed rule, EPA determined that optimizing already operating SNCRs or turning on idled SNCRs is an available approach for EGUs to reduce NOX emissions, has similar implementation timing to restarting idled SCR controls (less than two months for a given unit), and therefore could be done in time for the 2021 ozone season. EPA is finalizing its proposed determination that this emission control technology can be implemented in the 2021 ozone season. As explained in section VI.D.1 below, EPA is including optimization of existing SNCRs in its selected EGU control stringency. Thus, EPA provides further discussion here confirming the implementation timing of this emission control technology.

First, as noted with respect to SCR optimization, this rule will have an effective date over two months from the date of signature. In light of EPA's timing estimates of roughly 0.5 to 2 months for EGU operators to optimize their controls, this timing provides sufficient advance notice for operators of SNCR-equipped units to undertake any preparatory activities that may be needed prior to the effective date of the rule, and the onset of the increased stringency represented by the new emission budgets. Furthermore, because the emission reduction obligation is implemented through a mass-based trading program, these sources (and all others in the newly established Group 3 trading program) have abundant flexibility to choose other means of complying with their emission budget. Finally, as explained in section VII.C.4.d, EPA is providing a safety valve allowing access to additional allowances usable in the Group 3 trading program (through exchange of banked 2017-2020 Group 2 allowances at an 18:1 conversion ratio). As the amount of additional Group 3 allowances made available through the safety valve mechanism exceeds the effect on the emission budgets of including the optimization of existing SNCR controls several times over, there is no basis to believe that there will be compliance difficulty for any covered units.

In the proposed rule, EPA estimated a representative cost of approximately $3,900 per ton for turning on and fully operating idled SNCRs. For existing SNCRs that have been idled, unit operators may need to restart payment of some fixed and variable operating costs associated with these controls. Fixed and variable costs include labor, maintenance and repair, parasitic load, and ammonia or urea. The majority of the total fixed and variable operating costs for SNCR is related to the cost of the reagent used (e.g., ammonia or urea) and the resulting cost per ton of NOX reduction is sensitive to the NOX rate of the unit prior to SNCR operation. EPA is finalizing its adjusted representative cost of $1,800 per ton as described in the response to comments below, but applies the same performance, and timing assumptions for SNCRs that are idled as in the proposed rule.

Comment: Commenters observed that many SNCRs are already operating over the past several years (in an environment with an allowance price signal much lower than the $3,900 per ton threshold that EPA proposed represented turning on and optimizing idled controls). This observation suggests that the representative cost for this technology to optimize is likely less than estimated by EPA in the proposed rule when these operating patterns are accounted for.

Response: First, EPA examined the portion of the fleet with SNCR optimization potential and determined that the majority of units were already partially operating their controls. Therefore, EPA revisited the cost for SNCR optimization for units that are partially operating their controls. At proposal, EPA had noted a representative cost of $1,800 per ton for SNCR-controlled unit to optimize their controls if that control was already on and partially operating reflecting the cost of adding more reagent. This is similar to its analysis for SCR optimization that revealed an $800 per ton cost for SCR optimization at units with partially operating controls (as opposed to $1,600 per ton at units with idled SCR controls). EPA revisited this assessment of SNCR optimization cost at units with partially operating controls and found $1,800 per ton to still be a representative cost.[127] Therefore, given the majority of the SNCR-controlled fleet with identified optimization potential was already partially operating their controls based on 2019 historical data, EPA determined that $1,800 per ton (as opposed to the $3,900 per ton cost estimated in the proposed rule for turning on idled SNCRs) was a more representative cost for the mitigation strategy in this rulemakng. The representative cost of optimizing SNCR that is already partially operating excludes the fixed operating and maintenance (FOM) cost associated with starting up an idled SNCR control. For more details on this assessment, refer to the EGU NOX Mitigation Strategies Final Rule TSD in the docket for this rule. This adjustment in the expected cost of implementing this emission control has factored into EPA's determination to include optimization of existing SNCRs in its selected control stringency as Start Printed Page 23095discussed in more detail in section VI.D.1.

d. Installing New SNCRs

EPA is finalizing its determination not to include installation of new SNCRs in its selected control stringency in this rule. The amount of time needed to retrofit an EGU with new SNCR extends beyond the 2021 Serious area attainment date. However, similar to SCR retrofits discussed in section VI.B.1.e, and consistent with the Wisconsin decision, EPA evaluated potential emission reductions and associated costs from this emission control technology, and assessed the impacts and need for this emission control at the earliest point in time when post combustion control installation could be achieved. SNCR installations, while generally having shorter project timeframes (i.e., as little as 16 months (including pre-contract award steps) for an individual power plant installing controls on more than one boiler), share similar implementation steps with and also need to account for the same regional factors as SCR installations.[128] One recent example of installation timing took over a year—SNCR installation at the Jeffrey power plant (Kansas) was in the planning phase in 2013 but not in service until 2015.[129] Therefore, EPA is determining that at least 16 months would be needed to complete all necessary steps of SNCR development and installation at the EGUs not currently equipped with SNCRs in the 12 states linked to downwind receptors in this final rule. EPA discusses the timing of SNCR and SCR post-combustion retrofits together and in more detail in section VI.C.1.

SNCR technology provides owners a relatively less capital-intensive option for reducing NOX emissions compared to SCR technology, albeit at the expense of higher operating costs on a per-ton basis and less total emission reduction potential. EPA examined the remaining nationwide coal-fired fleet that lack SNCR or other NOX post-combustion control to estimate a representative cost of SNCR installation on a dollar per ton basis. Costs were estimated using the operating and unit characteristics specific to this fleet. As described in the EGU NOX Mitigation Strategies Final Rule TSD, EPA estimated that $5,800 per ton reflects a representative cost level at which they are available for a majority of the uncontrolled fleet.

Comment: EPA received some comments on timing and performance assumptions of this technology that largely focused on the decision to couple timing considerations for reduction evaluation purposes of SCR and SNCR retrofits together.

Response: EPA used the same cost, performance, and timing assumptions for this technology as it used in the proposed rule. EPA evaluates new retrofit technologies (i.e., SCR and SNCR) timing in tandem at step 3, and therefore it addresses this timing component in section VI.C.1. Remaining comments on SNCR performance potential are addressed in the RTC Document and in the EGU NOx Mitigation Strategies Final Rule TSD.

e. Installing New SCRs

The amount of time needed to retrofit an EGU with new SCR extends beyond the 2021 Serious area attainment date. However, similar to SNCR retrofits discussed above, and consistent with the Wisconsin decision, EPA evaluated potential emission reductions and associated costs from this control technology, as well as the impacts and need for this emissions control strategy, at the earliest point in time when their installation could be achieved. The amount of time to retrofit EGUs with new SCR varies between approximately 2 and 4 years depending on site-specific engineering considerations and on the number of installations being considered. In prior actions, EPA has noted 39-48 months as appropriate for regionwide actions when EPA is evaluating multiple installations at multiple locations.[130]

The Agency examined the cost for retrofitting a unit with new SCR technology, which typically attains controlled NOX rates of 0.07 lbs/mmBtu or less. Based on the characteristics of the remaining nationwide coal fleet that does not have a post-combustion control retrofit, EPA estimated that for unit and performance characteristics representative of that subgroup, $9,600 per ton reflects a representative cost level at which the SCR retrofit technology was typically available for the majority of these sources.

Comment: EPA received comments on the cost and performance of this technology, as well as comment on its timing assumption (as part of the collective timing assumptions in step 3).

Response: For this final rule's analyses, EPA used the same cost, performance, and timing assumptions that it used for this technology in the proposed rule. For more details on this assessment, refer to the EGU NOX Mitigation Strategies Final Rule TSD in the docket for this final rule and the RTC Document. Section VI.C.1 presents comments and EPA responses on the timing assumptions for installation of new SCRs.

f. Generation Shifting.

Finally, EPA evaluates emission reduction potential from generation shifting across the representative dollar per ton levels estimated for the other emission controls considered above. Shifting generation to lower NOX-emitting or zero-emitting EGUs occurs in response to economic factors (including regulatory signals such as pollution control costs). As the cost of emitting NOX increases, it becomes increasingly cost-effective for units with lower NOX rates to increase generation, while units with higher NOX rates reduce generation. Because the cost of generation is unit-specific, this generation shifting occurs incrementally on a continuum. Consequently, there is more generation shifting at higher cost NOX-control levels. It is reasonable for EPA to quantify and include the emission reduction potential from generation shifting at cost levels that are representative of the emission control technologies evaluated in the multi-factor analysis. Including emission reductions from generation shifting is important, ensuring that other cost-effective reductions (e.g., fully operating controls) can be expected to occur in a competitive electricity marketplace where generation shifting will inevitably occur in response to pollution control requirements. Generation shifting treatment and results are discussed in greater detail in the EGU NOX Mitigation Strategies Final Rule TSD.

In general, when EPA estimates emission reduction potential from generation shifting, EPA finds small amounts of generation shifting to existing lower NOX-emitting or zero-emitting units could occur consistent with the near-term implementation timing for this final rule. As a proxy for Start Printed Page 23096limiting the amount of generation shifting that is feasible for the near-term ozone seasons, EPA limits its assessment to shifting generation to other EGUs within the same state. EPA believes that limiting its evaluation of shifting generation (which EPA sometimes refers to as re-dispatch) to the amount that could occur within the state represents a conservatively small amount of generation-shifting because it does not capture further potential emission reductions that would occur if generation was shifted more broadly among units in different states within the interconnected electricity grid.

Comment: Commenters suggested that EPA should have included additional reductions from generation shifting beyond those levels that are commensurate with the emission controls identified. Commenters note that the statutory command is to eliminate significant contribution to downwind nonattainment or maintenance problems, 42 U.S.C. 7410(a)(2)(D)(i)(I), not merely to create a strong enough incentive that sources will likely install certain control technology. Because generation shifting is an independent measure that EGUs have widely deployed to reduce NOX emissions, EPA has no basis for evaluating only the emission reductions that result from a NOX price that matches—but goes no further than—the estimated representative NOX control costs of other emission control technologies assessed.

Response: EPA is finalizing the same approach to generation shifting that it proposed and that it included in the CSAPR Update. This rule's approach to capturing emission reduction potential from generation shifting in the state's emission budgets focuses on preserving the incentive for combustion and post-combustion controls to operate. Factoring generation shifting into the state emissions budgets helps promote an allowance price that will incentivize these controls to operate.

EPA recognizes that looking at higher levels of reductions purely through generation shifting is possible, assuming the availability for dispatch of lower or zero emitting generation assets that could substitute for the higher emitting EGUs. Shifting to such generators that are already in existence and operating at capacity factors that allow for some increase in their generation is the most economically efficient form of generation shifting, assuming other considerations such as availability, cost, reliability, and other factors are accounted for. Even greater shifting of generation to lower or zero emitting assets may be considered with the construction of new assets, although cost, timing, and economic considerations are generally of a greater magnitude and complexity in this context. Sophisticated power sector modeling tools, such as EPA's Integrated Planning Model (IPM) platform, can provide realistic and reliable assessments of the degree of generation shifting that may be accomplished at different cost levels. Indeed, in the Regulatory Impact Analysis for the proposed rule and for this final rule, EPA assessed a less-stringent control alternative for EGUs at the $500 per ton level, which was based solely on generation shifting rather than any at-the-source control technology. In general, EPA continues to stand by its discussion of its legal authority for and the technical viability of generation shifting as a method of emission reduction under the good neighbor provision, as set forth in the final CSAPR Update rule. See especially 81 FR 74504, 74545-47; see also CSAPR Update Response to Comment Document at 546-550 (legal authority); id. 528-533 (technical feasibility). (EPA had no occasion and did not reopen this portion of the CSAPR Update in this action on remand.)

Nonetheless, while generation shifting as a stand-alone strategy for emission reductions is available for both states' and EPA's consideration in the context of good neighbor SIPs or FIPs, EPA maintains the position discussed in the proposed rule for this action that further generation shifting than is captured by the methodology of the proposed rulemaking is unnecessary in the context of the resolution of good neighbor obligations for the 2008 ozone NAAQS in this action. The remaining timeframe for addressing upwind contribution to downwind nonattainment and maintenance receptors is through the 2024 ozone season, as downwind air quality problems for the 2008 ozone NAAQS are projected to be resolved by the 2025 ozone season. In EPA's judgment, the capital intensive nature of new builds and the likely multi-year timeframe necessary for the permitting and construction of new units make generation shifting to new generating resources, beyond those already planned and included in the baseline, not possible before downwind receptors are already resolved. With respect to generation shifting to existing generation resources with excess capacity, again, this rule already incorporates a certain amount of such generation shifting at cost levels representative of the other control technologies selected to quantify the state emission budgets in this rule. EPA believes that this degree of emission reduction through generation shifting is appropriate to include under the step 3 multi-factor analysis for the circumstances and compliance timetable currently presented by the 2008 ozone NAAQS, particularly the finding that downwind receptors will be resolved under this NAAQS by the 2025 ozone season.

Comment: Other commenters suggest that EPA should not factor in any generation shifting based reductions into state emission budgets, noting that EPA rejected the use of generation shifting in rescinding the Clean Power Plan and should do the same here in establishing emission reduction obligations under the good neighbor provision of section 110 of the Clean Air Act. According to these commenters, the emission budgets should be based on cost-effective emission reduction strategies that reflect technologies that can be implemented within the affected source's fence line.

Response: EPA notes again that its treatment of generation shifting here is consistent with both the CSAPR Update and the CSAPR, and the statute. Moreover, this comment incorrectly conflates the question of statutory authority under section 111 of the Act, the authority at issue in the Clean Power Plan and its repeal and subsequent litigation, with the question of statutory authority under section 110. As EPA explained in the CSAPR Update:

The good neighbor provision requires state and federal plans implementing its requirements to “prohibit[ ] . . . any source or other type of emissions activity within the State from emitting any air pollutant in amounts which will” significantly contribute to nonattainment or interfere with maintenance of the NAAQS in any other state. CAA section 110(a)(2)(D)(i)(I) (emphasis added). The EPA's consideration of the potential for generation shifting in developing state budgets is consistent with this statutory requirement. First, contrary to the commenters' contention, the statute does not limit the EPA's authority under the good neighbor provision to basing regulation only to control strategies for individual sources. The statute authorizes the state or EPA in promulgating a plan to prohibit emissions from “any source or other type of emissions activity within the State” that contributes (as determined by EPA) to the interstate transport problem with respect to a particular NAAQS. This broad statutory language shows that Congress was directing the states and the EPA to address a wide range of entities and activities that may be responsible for downwind emissions. However, this provision is silent as to the type of emission reduction measures that the states and the EPA may consider in establishing emission Start Printed Page 23097reduction requirements, and it does not limit those measures to individual source controls. The EPA reasonably interprets this provision to authorize consideration of a wide range of measures to reduce emissions from sources, which is consistent with the broad scope of this provision, as noted immediately above.

81 FR 74545.[131]

Finally, EPA notes that its interpretation of section 111 of the Act as unambiguously precluding the use of generation shifting as a “best system of emission reduction” under that provision was recently rejected by the D.C. Circuit. American Lung Association v. EPA, No. 19-1140 (D.C. Cir. Jan. 19, 2021). The court there also rejected arguments that generation shifting in the Clean Power Plan runs afoul of the federalism doctrine, slip op. 92 (“Interstate air pollution is not an area of traditional state regulation. And federalism concerns do not bar the United States government from addressing areas of Federal concern just because its actions have incidental effects on areas of state power.”) (emphasis in original) (citing FERC v. EPSA, 136 S. Ct. 760, 775-778 (2016)), or conflicts with FERC's authority, id. 95 n.12 (“The effects of environmental regulations on the power grid do not amount to power regulation statutorily reserved to FERC.”). In this rule, as in prior transport rules, EPA has established emission budgets that capture a certain degree of generation shifting that is modeled to occur as an economical response by the power sector to a particular cost threshold associated with at-the-plant control technologies. EPA has not mandated or ordered any particular degree of generation shifting to occur or that it occurs in a particular way. Further, this action is related solely to air pollution, in this case NOX as an ozone-precursor, and does not affect or purport to regulate any particular type of generation or achieve any type of generation mix, except as related to those NOX emissions. Cf. id. 88 (“The Clean Power Plan was aimed not at regulating the grid, but squarely and solely at controlling air pollution—a task at the heart of the EPA's mandate.”). The budgets here simply reflect an expectation that the power sector can and will take advantage of the compliance flexibility of a mass-based emission trading program to shift generation when it is economical to do so in response to an environmental mandate.

Finally, EPA solicited comment on whether other ozone-season NOX mitigation technologies should be considered. EPA invited comments on the cost and performance of the above listed technologies and any other potential mitigation technologies. For example, in January of 2020 the New York Department of Environmental Conservation adopted a rule to limit emissions from combustion turbines that operate as peaking units. EPA has not historically considered NOX mitigation technologies for these sources in its rulemakings, such as the CSAPR and the CSAPR Update, but invited comment on their appropriateness for this rulemaking. Separately, location and high emission rates of grid-connected municipal solid waste combustors, generally not covered under EPA's transport rules given their small size and differing purpose, have also led some stakeholders to suggest mitigation measures be considered for those sources.

Comment: EPA received comments calling on the Agency to reduce NOX from peaking units and municipal waste combustors and claimed that the agency's focus in its proposed rule on the suite of EGU emission controls above failed to address large sources of NOX emissions that are relatively close to the Connecticut receptors. Some of these commenters go one step further and say not only should EPA regulate these sources, but that EPA should only require emission reductions from local sources in place of reductions from larger emitting sources upwind.

Response: EPA is finalizing its evaluation of the same suite of emission controls as in the proposed rule. EPA notes that several states close to, or that have, nonattainment or maintenance receptors are already taking some of these measures. For example, New York finalized the state regulation mentioned above and New Jersey notes in their comment that the measures documented in New Jersey's Good Neighbor SIP include controls for sources such as behind-the-meter distributed generation/demand response (DG/DR) electric generators and municipal waste combustors. Even with these local measures, nonattainment and maintenance receptors persist in the region with demonstrable upwind state contribution, and thus the presence of these initiatives does not absolve upwind states and sources from the responsibility of addressing their significant contribution.[132] In the proposed rule, EPA inquired whether these additional emission controls should be considered in addition to, not in place of, the other proposed controls. EPA did not receive determinative evidence that (1) there were meaningful, upwind reductions from these emission controls that are not already being addressed by state rules, or (2) that any further reductions could be implemented in a timeframe consistent with the remaining nonattainment and maintenance receptors that resolve after 2024. EPA notes the New York rule referenced above was finalized in early 2020, but its control measures will phase in during the 2023-2025 period. Therefore, EPA is not finalizing any additional reductions from new control measures at these sources in this final rule, but, pending further analysis, doing so may be appropriate in a future context (e.g., under a different NAAQS). Finally, EPA notes to the extent that any of the sources meet the applicability requirements and are covered in the Group 3 trading program under this rulemaking, they would have an incentive to reduce emissions consistent with the ozone NOX allowance price. Moreover, as identified in the discussion the EGU NOX Mitigation Final Rule TSD, a significant number of units with this technology are located in states with rules addressing those sources.

2. Non-EGU NOX Mitigation Strategies

EPA has not regulated emissions from non-EGU sources as part of its regional transport rulemakings since the 1998 NOX SIP Call. In Wisconsin, the DC Circuit held that EPA must, on remand, implement a full remedy by the next attainment date (2021 for this final rule), or as soon as possible thereafter on a showing of impossibility, to achieve necessary reductions by that date. 938 F.3d at 320. The court also directed the Agency to address non-EGU sources, unless “the scientific uncertainty is so profound that it precludes EPA from making a reasoned judgment.” Id. at 318-20 (quoting Massachusetts v. EPA, 549 U.S. 497, 534 (2007)). The DC Circuit found that the practical obstacles EPA identified with respect to its evaluation of non-EGUs in the CSAPR Update did not rise to the level of an “impossibility,” id. The court also found that EPA must make a higher Start Printed Page 23098showing of uncertainty regarding non-EGU point-source NOX mitigation potential before declining to regulate such sources on the basis of “uncertainty.” Id. Thus, in the proposed rule, EPA extended its analysis to include all major stationary source sectors in the linked upwind states, including non-EGU emissions sources in various industry sectors. As discussed in section V, of the 22 states originally included in the CSAPR Update, EPA has determined that 12 states warrant analysis at step 3 for significant contribution to downwind nonattainment and/or maintenance receptors for the 2008 ozone NAAQS. Therefore, the Agency focused its step 3 assessment on non-EGU sources in these 12 states. For these sources, EPA retained its focus on NOX as the most effective precursor pollutant for addressing interstate ozone transport at a regional scale. See 82 FR 51238, 51248 (Nov. 3, 2017) (citing 76 FR 48222) and 63 FR 57381.

The remainder of this section summarizes the analysis EPA conducted in the proposed rule. EPA is finalizing this analysis using the best available current data, largely as proposed, and determines on the basis of this analysis that emission reductions from non-EGU sources/units in the 12 states are not needed to eliminate their significant contribution to nonattainment or interference with maintenance in any other state. EPA made some minor updates to its analysis of non-EGU emission reduction potential, and these changes did not affect its overall conclusion that reductions are not warranted under the step 3 multi-factor test. EPA responds to significant comments on its assessment of non-EGU emission reduction potential at the end of the relevant section below, and addresses remaining comments on potential non-EGU emission reductions in the RTC document located in the docket for this action.

For non-EGU sources, there are many types of emissions sources or units that emit NOX and many control technologies or combinations of control technologies for these sources or units. As such, there are many approaches to assessing emission reduction potential from non-EGU emission sources or units. In this final rule, EPA applied the multi-factor test used for EGUs in an effort to determine an appropriate stringency level for non-EGU sources/units in linked upwind states. EPA identified available control technologies and estimated their costs and potential emission reductions. The Agency considered the information it has regarding control technology implementation timeframes, including information on such timeframes provided by commenters on the proposed rule, to determine potential air quality impacts in relevant future years.

To identify levels of control for non-EGU sources/units, EPA used the Control Strategy Tool (CoST),[133] the Control Measures Database (CMDb), and the projected 2023 inventory from the 2016v1 modeling platform. EPA assessed potential emission reductions associated with applying controls to emissions units with 150 tons per year (tpy) or more of pre-control NOX emissions in 2023, which is an emissions threshold that represents a comparable unit size to 25 MW for EGUs used in prior interstate transport rulemakings. To derive this emissions threshold, EPA used emissions expected from an average 25 MW EGU unit operating at a median heat rate, emission rate, and capacity factor for a coal-fired unit.[134] In CoST, the Agency used the maximum emission reduction strategy[135] to estimate the largest quantity of potential emission reductions from each emissions source or unit located in the 12 upwind states linked to downwind receptors in this final rule. Eleven of the 12 upwind states had sources/units with 150 tpy or more of pre-control NOX emissions in 2023; the projected 2023 emissions inventory did not include non-EGU point sources/units in New Jersey with pre-control NOx emissions greater than 150 tpy for which CoST had applicable control measures.[136]

For the 12 linked states, EPA categorized the CoST results for control technologies that comprise approximately 92 percent of the total estimated potential emission reductions from the non-EGU sources/units with 150 tpy or more of NOX emissions in these states; [137] those technologies and related emissions sources/units are summarized in Table VI.B.2-1 below. In tranche one before further refinement and verification, the number of emissions units CoST applied SCR to was 51 and the number of emissions units CoST applied SNCR to was 23. The estimated emission reductions from those control applications were 12,724 ozone season tons. In tranche two, before further refinement and verification, the number of emissions units to which CoST applied layered combustion (a type of combustion control technology) was 49, the number of emissions units to which CoST applied NSCR [138] or layered combustion was 65, and the number of emissions units to which CoST applied ultra-low NOx burner and SCR was 56. The estimated emission reductions from those control applications were 17,283 ozone season tons. EPA then calculated a weighted average cost per ton (in 2016$) for estimated potential reductions associated with each control technology and plotted the weighted average cost per ton values. From the resulting curve, EPA identified a clear break point that defined two tranches of potential emission reduction, as shown in Table VI.B.2-1. For additional details on the curve and the potential emission reductions in tranches one and two, please see the memorandum titled Assessing Non-EGU Emission Reduction Potential, available in the docket for this rule.

Start Printed Page 23099

Table VI.B.2-1—Details on Tranches One and Two of Potential Emission Reductions

TrancheTechnologies/industry sectors or source groupsWeighted average cost (2016$ per ton)Cost range (2016$ per ton)
Tranche OneSCR/Glass Manufacturing, IC Engines SNCR/Cement Manufacturing2,000139 64-5,700
Tranche TwoLayered Combustion/Lean Burn IC Engines NSCR or Layered Combustion/Industrial Rich Burn Natural Gas IC Engines Ultra-low NOX Burner and SCR/Industrial Boilers5,000-6,6001,400-9,700

Given the large number of emissions units in one or more industry sectors that could require control installation, EPA does not have detailed information on the time needed to install all of the control technologies identified in Table VI.B.2-1. Any installation timing estimates would need to reflect the time needed to install controls across a potentially large number of sources, the time needed to have appropriate NOX monitoring installed, the time needed to raise the necessary financing, and other steps in the permitting, construction and procurement processes. EPA previously examined the time necessary to install some of the controls indicated in Table VI.B.2-1 for different industries in the 2016 Final Technical Support Document (TSD) for the Final Cross-State Air Pollution Rule for the 2008 Ozone NAAQS, Assessment of Non-EGU NOXEmission Controls, Cost of Controls, and Time for Compliance Final TSD (“CSAPR Update Non-EGU TSD”), which is discussed in section VI.C.2. EPA expects that the controls for glass furnaces and cement kilns would take at least 2 years to install on a sector-wide basis across the 12-state region affected by this final rule. Information available to the Agency, including information provided by commenters, does not establish that implementation of NOX control technologies for non-EGU emission sources/units could take place in less than 2 years. Therefore, EPA has determined that the 2023 ozone season is the earliest ozone season by which these non-EGU controls could be installed. EPA thus concludes that no NOX controls for non-EGUs included in this cost analysis can be installed by the 2021 ozone season. Additional information on installation times for non-EGU NOX controls can be found in section VI.C.

3. Mobile Source NOX Mitigation Strategies

Under a variety of CAA programs, EPA has established federal emissions and fuel quality standards that reduce emissions from cars, trucks, buses, nonroad engines and equipment, locomotives, marine vessels, and aircraft (i.e., “mobile sources”). Because states are generally preempted from regulating new vehicles and engines with certain exceptions (see generally CAA sections 209, 177), mobile source emissions are primarily controlled through EPA's federal programs. EPA has been regulating mobile source emissions since it was established as a federal agency in 1970, and all mobile source sectors are currently subject to NOX emissions standards. EPA factors these standards and associated emission reductions into its baseline air quality assessment in good neighbor rulemaking, including in this action. Such reductions are an important reason for the historical and long-running trend of improving air quality in the United States. These trends help explain why the overall number of receptors and severity of ozone nonattainment problems under the 2008 ozone NAAQS continues to decline. Such data are factored into EPA's analysis at steps 1 and 2 of the 4-step framework. As a result of this long history, NOX emissions from onroad and nonroad mobile sources have substantially decreased (73 percent and 57 percent since 2002, for onroad and nonroad, respectively) [140] and are predicted to continue to decrease into the future as newer vehicles and engines that are subject to the most recent, stringent standards replace older vehicles and engines.[141]

For example, in 2014 EPA promulgated new, more stringent emissions and fuel standards for light-duty passenger cars and trucks.[142] The fuel standards took effect in 2017, and the vehicle standards are phasing in between 2017 and 2025. Other EPA actions that are continuing to reduce NOX emissions include the Heavy-Duty Engine and Vehicle Standards and Highway Diesel Fuel Sulfur Control Requirements (66 FR 5002; January 18, 2001); the Clean Air Nonroad Diesel Rule (69 FR 38957; June 29, 2004); the Locomotive and Marine Rule (73 FR 25098; May 6, 2008); the Marine Spark-Ignition and Small Spark-Ignition Engine Rule (73 FR 59034; October 8, 2008); the New Marine Compression-Ignition Engines at or Above 30 Liters per Cylinder Rule (75 FR 22895; April 30, 2010); and the Aircraft and Aircraft Engine Emissions Standards (77 FR 36342; June 18, 2012).

EPA is currently developing a new regulatory effort to reduce NOX and other pollution from heavy-duty trucks (known as the Cleaner Trucks Initiative), as described in the January 21, 2020, Advance Notice of Proposed Rulemaking (85 FR 3306). Heavy-duty vehicles are the largest contributor to mobile source emissions of NOX and will be one of the largest mobile source contributors to ozone in 2025.[143] Reducing heavy-duty vehicle emissions nationally would improve air quality where the trucks are operating as well as downwind. As required by CAA section 202(a)(3)(A) of the Act, EPA will be proposing NOX emission standards that “reflect the greatest degree of emission reduction achievable through the application of technology which the Administrator determines will be available for the model year to which such standards apply, giving appropriate consideration to cost, energy, and safety factors associated with the application of such technology.” Section 202(a)(3)(C) Start Printed Page 23100requires that standards apply for no less than 3 model years and apply no earlier than 4 years after promulgation.

Given these requirements, EPA is considering implementation of new heavy-duty NOX emission standards beginning in model year 2027. In addition, any new rulemaking process for other mobile source sectors would not achieve actual NOX emission reductions before 2025, given the lead time necessary for EPA and for manufacturers.

However, EPA's existing regulatory program will continue to reduce NOX emissions into the future, and EPA is currently taking active steps to ensure that these NOX reductions occur. The CAA prohibits tampering with emissions controls, as well as manufacturing, selling, and installing aftermarket devices intended to defeat those controls. EPA currently has a National Compliance Initiative called “Stopping Aftermarket Defeat Devices for Vehicles and Engines,” which focuses on stopping the manufacture, sale, and installation of hardware and software specifically designed to defeat required emissions controls on onroad and nonroad vehicles and engines.

C. Emission Reduction Potential of Control Stringencies

1. EGU Emission Reduction Potential

For EGUs, as discussed in section VI.A, the multi-factor test considers increasing levels of uniform control stringency in combination with consideration of total NOX reduction potential and corresponding air quality improvements. EPA evaluated EGU NOX emission controls that are widely available (described previously in section VI.B.1), that were assessed in previous rules to address ozone transport, and that have been incorporated into state requirements to address ozone nonattainment.

The tables below summarize the emission reduction potentials (in absolute ozone season tons) from these emission controls across the 12-state region. Table VI.C.1-2 focuses on near-term mitigation emission controls while Table VI.C.1-3 includes emission controls with extended time frames for implementation.

Table VI.C.1-2—EGU Ozone-Season Emission Reduction Potential—2021

StateBaseline 2021 OS NOXReduction potential (tons) for varying levels of technology inclusion
SCR optimizationSCR optimization + LNB upgradeSCR/SNCR optimization + LNB upgrade
Illinois9,368171171267
Indiana15,8562,7712,7712,805
Kentucky15,5882821,5311,538
Louisiana15,4768787658
Maryland1,501111
Michigan13,8981,1661,2841,288
New Jersey1,346929292
New York3,469535353
Ohio15,8296,1406,1406,140
Pennsylvania11,8963,5173,5173,517
Virginia4,66450320380
West Virginia15,1651,4791,9602,281
Total124,05715,80917,92719,021
* EPA shows reduction potential from state-of-the-art LNB upgrade as a near-term reduction emission control but explains in sections VI.B and VI.D that this reduction potential would not be implemented until 2022. Sum of state values may vary slightly from total due to rounding.

Table VI.C.1-3—EGU Ozone-Season Emission Reduction Potential—2025

StateBaseline 2025 OS NOXReduction potential (tons) for varying levels of technology inclusion*
SCR optimization + LNB upgradeSCR/SNCR optimization + LNB upgradeSCR/SNCR optimization + LNB upgrade + SNCR retrofit + generation shiftingSCR/SNCR optimization + LNB upgrade + SCR retrofit + generation shifting
Illinois8,2811382331,0531,401
Indiana12,2322,6482,6683,3093,802
Kentucky14,5511,1991,2052,7555,022
Louisiana15,476876591,0982,854
Maryland1,35022181181
Michigan11,0091,2051,2092,3313,656
New Jersey1,34692928989
New York3,4565353159159
Ohio15,9276,1556,1556,2846,706
Pennsylvania11,8963,5233,5233,9754,045
Virginia4,162323367417850
West Virginia15,1651,9602,2812,3284,597
Start Printed Page 23101
Total114,85017,38418,44823,97833,363
* Both tables VI.C.1-2 and VI.C.1-3 include limited generation shifting (reflecting that which would occur at the price level consistent with control operation). It does not factor in generation shifting reduction potential that may be attributable to incremental new builds or incremental retirements. Sum of state values may vary slightly from total due to rounding.

Comment: Some commenters suggested that the emission reduction estimates for an identified technology needed to be updated based on new or updated data.

Response: EPA has updated the total emission reduction potential for each technology based on information provided by commenters. Further details are provided in the RTC Document included in the docket, the EGU NOx Mitigation Strategies Final Rule TSD, and in the Ozone Transport Policy Analysis Final Rule TSD. In summary, comments containing new data, information, or analysis that resulted in changes to the values in the tables above included information on (1) shared stack emissions apportionment, (2) updated information and data on retirements and new builds, (3) updated information and data on combustion control performance, and (4) and updated information on SNCR optimization cost. In the first three instances, the resulting impact was a change in the inventory of units with identified emission reduction potential (and therefore overall emission reduction potential from that category). For instance, multiple commenters provided EPA with data on shared stack emissions apportionment not readily available in unit-level data reported to EPA. In some cases where stack data are measured and reported, and that stack is shared by two units (one with an SCR and one without), the apportionment method of those reported stack emissions for reporting purposes is heat input-based and therefore may not reflect the unit-level operation of the control at that unit, even when that control is operating. In other words, it may have apportioned those stack emissions (e.g., 10 tons) as 5 tons to each unit, while the actual operation is 9 tons from the uncontrolled unit and 1 ton from the controlled unit. This can give the appearance of a controlled unit emitting above the optimized rate, when in fact it is already operating below the 0.08 lb/mmBtu threshold. Similar to the CSAPR Update, EPA incorporated the information from this comment and new data into this final rule, and EPA has adjusted the Agency's inventory of units that may have SCR reduction potential accordingly. Likewise, EPA received some updated information on unit-level retirement status such as changes to the retirement status of the Colver Power Plant in Pennsylvania and the Pleasants Power Station in West Virginia. As these units are no longer retiring, their retirement is not factored into the step 3 baseline or resulting state emission budgets. Similarly, EPA also incorporated comments and new data regarding new units expected to come online and retiring units expected to go offline after 2019 but prior to 2024 ozone season.[144] Also, as noted above, EPA updated its performance rate assumption for LNB controls based on updated data and comments, resulting in less emission reduction potential from this technology category. Finally, the emission reduction levels associated with SNCR optimization were updated to be consistent with the representative cost (and commensurate generation shifting-based reductions) adjustments discussed above.

Comment: Some commenters asserted that EPA should change its timing assumptions for post-combustion control retrofits by parsing out different timing assumptions for SNCR and SCR retrofits. They claim that doing so would result in more emission reductions available starting in earlier years (e.g., 2023) given that SNCR retrofit technology could be installed by that year.

Response: EPA is finalizing the same timing assumptions that it proposed for the installation of post-combustion controls. As discussed in section VI.B.1.e and noted in prior actions, EPA generally views 39-48 months as an appropriate implementation timeframe for regionwide installation of new post-combustion control technologies when EPA is evaluating multiple installations at multiple locations. As discussed further below, this is primarily based on SCR retrofit rather than SNCR. The period from finalization of this rule until the start of the 2024 ozone-season would allow less than 39 months for post combustion controls to be regionally installed and operating. The 2025 ozone season represents a period approximately 48 months after finalization of this rule and reflects a more demonstrably possible window for making retrofits on a regional scale. Therefore, EPA finds that 2025 is the earliest ozone season by which new SNCR or SCR may be installed across multiple EGUs on a regional basis.

Installing new SCR or SNCR controls for EGUs generally involves the following steps: Conducting an engineering review of the facility to determine suitability and project scope; advertising and awarding a procurement contract; obtaining a construction permit; installing the control technology; testing the control technology; and obtaining or modifying an operating permit. These timeframes are intended to accommodate a plant's need to conduct an engineering assessment of the possible NOX mitigation technologies necessary to then develop and send a bid request to potential suppliers. Control specifications are variable based on individual plant configuration and operating details (e.g., operating temperatures, location restrictions, and ash loads). Before making potential large capital investments, plants need to complete these careful reviews of their system to inform and develop the control design they request. They then need to solicit bids, review bid submissions, and award a procurement Start Printed Page 23102contract—all before construction can begin.

Scheduled curtailment, or planned outage, for pollution control installation would also be necessary to complete SCR or SNCR projects on a regional scale. Given that peak demand and rule compliance would both fall in the ozone season, sources would likely need to schedule installation projects for the “shoulder” seasons (i.e., the spring and/or fall seasons), when electricity demand is lower than in the summer, reserves are higher, and ozone season compliance requirements are not in effect. If multiple units were under the same timeline to complete the retrofit projects as soon as feasible from an engineering perspective, this could lead to bottlenecks of scheduled outages as each unit attempts to start and finish its installation in roughly the same compressed time period. Thus, any compliance timeframe that would assume installation of new SCR or SNCR controls should be developed to reasonably encompass multiple shoulder seasons to accommodate scheduling of curtailment for control installation purposes and better accommodate the regional nature of the program.[145]

Finally, the time lag observed between the planning phase and in-service date of SCR operations in certain cases also illustrates that site-specific conditions can lead to installation times of four years or longer—even for individual power plants. For instance, SCR projects for units at the Ottumwa power plant (Iowa), Columbia power plant (Wisconsin), and Oakley power plant (California) were all in the planning phase in 2014. By 2016, these projects were under construction with estimated in-service dates of 2018.[146] Further, large-scale projects also illustrate that timelines can extend beyond the general estimate for a single power plant when the project is part of a larger, multifaceted air pollution reduction goal. For instance, the Big Bend power plant in Florida completed a multifaceted project that involved adding SCRs to all four units as well as converting furnaces, over-fire air changes, and making windbox modifications, during which a decade elapsed between the initial planning stages and completion.[147]

EPA notes that differences between these control technologies exist with respect to the potential viability of achieving cost-effective, regional NOX reductions from EGUs. SCR controls generally achieve greater EGU NOX reduction efficiency (up to 90 percent) than SNCR controls (25 percent). EPA observes that for the remaining uncontrolled coal fleet in the 12 states, SCRs are, on average, more expensive on a cost per ton basis. However, the analysis in the EGU NOX Mitigation Strategies Final Rule TSD notes that the cost range varies widely for units depending on inlet NOX rate and capacity factor. Therefore, for some units, it is possible that SCR retrofit costs are lower than SNCR costs on a cost per ton basis. Moreover, there are a host of other market and policy drivers that may lead a specific unit to prefer an SCR retrofit over an SNCR retrofit. As a result, EPA finds it is reasonable to allow sufficient time for EGU operators to select installation of SCR in response to a multi-state emission control program whose emission budgets would reflect emission reductions from new post-combustion controls. Therefore, EPA is using an SCR-inclusive planning and installation schedule to represent new post-combustion retrofit potential on a regional basis (be it SNCR or SCR as determined by individual EGU owners under our flexible market-based emission trading program).

Furthermore, SNCR installation at an individual source would render later installation of an SCR less cost-effective, because such a unit would have already expended some unrecoverable capital on the less-effective pollution control technology. As a result, it would be counterproductive to assume EGUs should install the less effective SNCR control technology to address a short-run air quality concern under an older and less stringent NAAQS when it may later prove necessary to require the more effective SCR control technology to address longer-run air quality concerns under a more stringent NAAQS for the same pollutant. Considering these factors, EPA finds it is appropriate to give particular weight to the timeframe required for implementation of SCR across the region as compared to SNCR to allow sources the flexibility to make the most efficient post-combustion control investment. Historically, units have chosen to retrofit with higher performing SCR at a much greater rate than they have chosen SNCR. For SCR, the total time associated with project development is estimated to be up to 39 months for an individual power plant installing controls on more than one boiler. However, more time is needed when considering installation timing for new SCR controls regionally. EPA has previously determined that a minimum of 48 months (four years) is a reasonable time period to allow to complete all necessary steps of SCR projects at EGUs on a regional scale. This timeframe would allow for regional implementation of these controls (i.e., at multiple power plants with multiple boilers) considering the necessary stages of post-combustion control project planning, shepherding of labor and material supply, installation, coordination of outages, testing, and operation.[148]

In addition to its engineering assessment, EPA looked at historical data to validate this 39-48 month installation timeframe. EPA observed over 12 GW of uncontrolled coal capacity in the linked states covered in this rule. For comparison, EPA looked at the last 15 years of data to see if a similar amount of capacity had come online in a shorter time frame. It observed that it had not. Most notably, the CAIR was finalized in March of 2005 covering much of the Eastern U.S. and drove significant SCR retrofit activity, with incentives for early installation and reductions. From this date, 39-48 months would have placed the SCRs online in the mid 2008 to 2009 time frame. The graphic below illustrates an uptick in coal-fired capacity retrofitted with SCRs in response to the rule (Figure VI.D.2). Most of this capacity comes online in 2009 and 2010. Although EPA's data on when sources started planning these controls and whether it was driven purely by CAIR or other factors are not perfect, the Agency finds the chart below consistent with its determination that a 39-48 month time frame is reasonable for SCR retrofit possibility on a regional level.

Start Printed Page 23103

Comment: EPA received comment on the timing assumptions regarding SCR and SNCR retrofit. Commenters noted that EPA should require SNCR installation as it can be installed in as little as 16 months, and that EPA's reliance on SCR timing to justify not considering SNCR is not reasonable given that EPA is not considering SCR installation. Moreover, the commenter also suggested that if these controls are not available on a region-wide level by the start of 2024, that EPA should still include them for a limited number of units (e.g., 30 percent of the unretrofitted fleet) as the Clean Air Act requires that upwind states limit emissions “as expeditiously as practicable.”

Response: EPA believes its proposed collective timing assumptions for post-combustion control retrofit are practicable given that the preferable capital-intensive investment retrofit decision would be highly unit-specific and subject to a unit's compliance strategy choices with respect to multiple regulatory requirements. For the reasons described above, EPA believes that separating the post-combustion retrofit timing consideration would create a framework that potentially inhibits greater emission reductions from technologies like SCR that may be both preferable to the unit's operator and beneficial to overall emission reductions. While the commenter observed that SCR installation is not included as part of EPA's proposed control stringency, states and EPA may consider requiring this emission control technology to address good neighbor obligations or other attainment planning requirements for the 2015 ozone NAAQS or other CAA programs. Therefore, while the commenter suggests that the exclusion of new SCR installations from the control stringency selected for this rule should result in the decoupling of SNCR and SCR for timing considerations, EPA observes that the broader regulatory context potentially presents situations where a better performing emission reduction technology is the preferred retrofit choice. If EPA were to ignore the observation that this post-combustion retrofit technology decision is a binary choice, as these technologies substitute for rather than complement one another, it would potentially eliminate or make more costly the eventual decision to implement a better performing SCR technology by implementing on a schedule that did not allow for that compliance strategy.

With regard to the suggestion that, if it is not possible to require all non-retrofitted units to install new controls, EPA should at least require some units to retrofit with SNCR and SCR, EPA observed that doing so would result in making selective choices about which linked upwind states should face more stringent requirements and would upset the uniform control stringency scheme allowing for “equitable and efficient” implementation of good neighbor obligations. EME Homer City, 572 U.S. at 519. In addition, it would necessitate far greater unit-level analysis, which would likely have prevented EPA from finalizing a rule in time to implement reductions for the 2021 ozone season.

2. Non-EGU Emission Reduction Potential

EPA performed a similar analysis of reduction potential for the non-EGU mitigation technologies identified, as discussed in section VI.B.2 of this notice. EPA's assessment of emission reduction potential from the controls in the tranches reflects ongoing uncertainty resulting from the quality of the current information available to the Agency. This uncertainty has been addressed to some extent through further research conducted since the proposed rule. Because information for existing controls on non-EGU emissions sources is missing in the 2016 base year inventory for some states and incomplete for some sources, EPA went through a process in the proposed rule to further verify existing control information and refine the NOX emission reduction potential estimated by CoST, the CMDb, and the 2023 projected inventory. In the proposed rule EPA focused its verification and refinement efforts on those upwind states with the largest estimated potential air quality impacts from potential non-EGU emission reductions. Since the proposed rule, EPA extended its verification and refinement efforts to several additional linked states.

In the proposed rule, EPA identified two tranches of controls for non-EGU emissions sources/units associated with two levels of weighted average cost per ton. EPA assumed that the potential reductions in tranche one were likely cost-effective because tranche one's weighted average cost of $2,000 per ton is similar to the identified control stringency for EGUs represented by Start Printed Page 23104$1,800 per ton (see section VI.D.1). The additional steps EPA took, discussed in more detail below, included:

  • Looked at potential emission reductions in tranche one that were estimated to cost less than $2,000 per ton; and
  • For those potential reductions in tranche one that were estimated to cost less than $2,000 per ton, reviewed online facility permits and industrial trade literature to verify and determine if the estimated emission reductions may be actual, achievable emission reductions or if the estimated emission reductions are associated with emissions units that are already controlled.

EPA focused its verification and refinement efforts on those upwind states with the largest estimated potential air quality impacts from potential non-EGU emission reductions. Specifically, EPA used an estimate of 0.02 ppb as a threshold for air quality improvement that may be obtained from reductions from non-EGU emissions sources in each state to better target its efforts to verify and refine the potential estimated non-EGU NOX emission reductions. The Agency explained that it was not applying a 0.02 ppb impact threshold as a step in the step 3 multi-factor test. Rather, the threshold allowed the Agency to better target its efforts toward the potentially effective states for non-EGU NOX emission reductions. Based on this, the states for which the Agency verified existing control information and refined the NOX emission reduction estimates in the proposed rule included: Indiana, New York, Ohio, Pennsylvania, and West Virginia. For additional discussion on the air quality impacts by state, see the section titled Air Quality Impacts from Potential Non-EGU Emissions Reductions in the technical memorandum titled Assessing Non-EGU Emission Reduction Potential in the docket for this rule. In this final rule, EPA extended its verification process to additional linked states, including Maryland, Michigan, and Virginia.[149]

As noted above to focus the set of non-EGU emissions sources/units in the linked upwind states (Indiana, Maryland, Michigan, New York, Ohio, Pennsylvania, Virginia, and West Virginia) for which EPA could verify existing control information and refine the NOX emission reduction estimates, the Agency assumed that the potential reductions in tranche one were likely cost-effective because tranche one's weighted average cost of $2,000 per ton is similar to the identified control stringency for EGUs represented by up to $1,800 per ton (see section VI.D.1).

In the proposed rule, EPA found in Indiana, New York, Ohio, and Pennsylvania, that the estimated emission reductions in tranche one that cost less than $2,000 per ton were 6,346 ozone season tons. Note that no potential emission reductions at a cost of less than $2,000 per ton were identified in West Virginia because CoST originally estimated control costs for two IC engines in West Virginia inappropriately, and CoST did not identify likely cost-effective controls for any other non-EGU emissions units in the state. EPA removed the two IC engines in West Virginia from further consideration because the corrected potential cost was greater than $2,000 per ton.

In reviewing the potential controls in tranche one that were estimated to cost less than $2,000 per ton for Indiana, New York, Pennsylvania, and Ohio, EPA found that these reductions were from SCR applied to glass furnaces and SNCR applied to cement kilns. In addition in this final rule, EPA found in Maryland, Michigan, and Virginia the estimated emission reductions in tranche one that cost less than $2,000 per ton are 664 ozone season tons. These estimated reductions were also from glass furnaces and cement kilns. The total estimated emission reductions in tranche one in Indiana, Maryland, Michigan, New York, Pennsylvania, Ohio, and Virginia that cost less than $2,000 per ton are 7,010 ozone season tons.

Next, to verify the information on the application of these controls and estimated emission reductions, EPA reviewed facilities' online title V permits and industrial trade literature for the likely cost-effective emission reductions associated with SCR applied to glass furnaces and SNCR applied to cement kilns. In the proposed rule, EPA determined that of the 20 emissions units in Indiana, New York, Pennsylvania, and Ohio included in the cost analysis, source permits identified that 10 units (i) already have controls and monitors (primarily CEMS), (ii) are installing controls and CEMS or consolidating operations in the next few years as a result of recent consent decrees issued as part of EPA's New Source Review Air Enforcement Initiative, (iii) have shut down, or (iv) are planning to shut down by 2023. These 10 units account for approximately 34 percent of estimated potential emissions reductions from Pennsylvania, New York, Ohio, and Indiana in tranche 1 that cost <$2,000 per ton. The results of the online permit review and review of industrial trade literature, summarized in Table VI.C.2-1 below, suggested that approximately 14 percent of the CoST-estimated potential emission reductions in these four states may be possible to achieve.

Table VI.C.2-1—Status of Potential Emission Reductions

Number of emissions unitsOS tonsPercent of total
Shutdowns482413
Lehigh Cement—Kiln Replacements33666
NEI Discrepancy/Uncertain 15013,28651
Start Printed Page 23105
Already Controlled/Uncertain596715
Possible Emission Reductions790314
Total206,346

In EPA's analysis for this final rule, the online permit review for Maryland, Michigan, and Virginia identified approximately 62 ozone season tons out of the estimated 664 ozone season tons that are from sources/units already controlled, leaving an estimated 602 ozone season tons of likely cost-effective emission reductions from these states. For additional details on the review of online permits and industrial trade literature, please see the memorandum titled Assessing Non-EGU Emission Reduction Potential, available in the docket for this rule.

EPA previously examined the time necessary to install the controls indicated in the table above (with details on the technology tranches) for different industries. The 2016 CSAPR Update Non-EGU TSD provided preliminary estimates of installation times for a variety of NOX control technologies applied to a large number of sources in non-EGU industry sectors.[151] For virtually all NOX controls applied to cement manufacturing and glass manufacturing, information on installation times was not available to provide an estimate, and the installation time for these controls was “uncertain.” There was an exception for SNCR applied to cement kilns; however, the installation time estimate of 42-51 weeks listed in the CSAPR Update Non-EGU TSD does not account for implementation across multiple sources, the time needed to have NOX monitoring installed, and other steps in the permitting and construction processes.

To improve upon information from the CSAPR Update Non-EGU TSD on installation times for SCR on glass furnaces and SNCR on cement kilns, EPA reviewed information from permitting actions and a consent decree. For two glass manufacturing facilities that installed SCR on glass furnaces, from the time of permit application to the time of SCR operation was approximately 19 months for one facility and is currently at least 20 months for another facility.[152] These installation times do not reflect time needed for pre-construction design and engineering, financing, and factors associated with scaling up construction services for multiple installations at several emissions units. With respect to cement kilns, an April 2013 consent decree between EPA and CEMEX, Inc. required installation of SNCR at a kiln within 450 days, or approximately 15 months, of the effective date of the consent decree. Similarly, this installation time does not reflect time associated with scaling up construction services for multiple control installations at several emissions units.

This information and EPA's general experience indicate that a two-year installation timeframe for a rule requiring installation of new control technologies across a variety of emissions sources in several industry sectors on a regional basis is a relatively fast installation timeframe. A shorter installation timeframe of approximately one year (i.e., in time for the 2022 ozone season) would raise significant challenges for sources, suppliers, contractors, and other economic actors, potentially including customers relying on the products or services supplied by the regulated sources.[153]

Thus, for this rule, EPA estimates that these controls for glass furnaces and cement kilns would take at least 2 years to install on a sector-wide basis across the 12-state region. Therefore, based on the information currently available, EPA in its reasoned judgment finds that the 2023 ozone season is the earliest ozone season by which these non-EGU controls could be installed.

D. Assessing Cost, EGU and Non-EGU NOX Reductions, and Air Quality

To determine the emissions that are significantly contributing to nonattainment or interfering with maintenance, EPA applied the multi-factor test to EGUs and non-EGUs separately, considering for each the relationship of cost, available emission reductions, and downwind air quality impacts. Specifically, EPA determined the appropriate level of uniform NOX control stringency that addresses the impacts of interstate transport on downwind nonattainment or maintenance receptors. EPA also evaluated possible over-control by determining if an upwind state is linked solely to downwind air quality problems that could have been resolved at a lower cost threshold, or if an upwind state could have reduced its emissions below the 1 percent air quality contribution threshold at a lower cost threshold.

1. EGU Assessment

For EGUs, EPA examined the emission reduction potential associated with each EGU emission control technology (presented in section VI.C.1) and its impact on the air quality at downwind receptors. Specifically, EPA identified the projected air quality improvement relative to the base case, as well as whether the air quality improvements are sufficient to shift the status of receptors from nonattainment to maintenance or from maintenance to clean. Combining these air quality factors, cost, and emission reductions, EPA identified a control stringency for EGUs that maximizes the air quality improvement from emission controls available in the timeframe for which air quality problems at downwind receptors Start Printed Page 23106persist. This control stringency reflects the optimization of existing SCR controls and installation of state-of-the-art NOX combustion controls, which are widely available at a representative marginal cost of $1,600 per ton. It also includes the optimization of existing SNCR controls at sources that are already partially operating these controls, which becomes widely available as a mitigation technology at $1,800 per ton. EPA's evaluation also shows that emission budgets reflecting the operation of these existing post combustion controls and combustion control upgrades do not over-control upwind states' emissions relative to either the downwind air quality problems to which they are linked at step 1 or the 1 percent contribution threshold that triggers further evaluation at step 2 of the 4-step framework for the 2008 ozone NAAQS. To assess downwind air quality impacts for each nonattainment and maintenance receptor identified in section V.C, EPA evaluated the air quality change at that receptor expected from the progressively more stringent upwind EGU control stringencies that were available for that time period. This assessment provides the downwind ozone improvements for consideration and provides air quality data that is used to evaluate potential over-control.

To assess the air quality impacts of the various control stringencies at downwind receptors, EPA evaluated changes resulting from the emission reductions associated with the identified emission controls in each of the upwind states, as well as corresponding reductions of similar stringency in the downwind state containing the receptor they are linked to. By applying these emission controls to the state containing the receptor, EPA assumes that the downwind state will implement (if it has not already) an emissions control stringency for its sources that is comparable to the upwind control stringency identified here. Consequently, EPA explicitly ensures that it is accounting for the downwind state's share of a nonattainment or maintenance problem (which is a part of the overcontrol evaluation).[154]

For states that were not linked to that receptor, the air quality change at that receptor was evaluated assuming emissions equal to the engineering analytics base case emission level. This method holds each upwind state responsible for its share of the specific downwind problems to which it is linked. For states that are not linked to that receptor (even if they are linked to a different receptor), EPA assumes that they are not making emission reductions beyond those in the base case to that receptor. In practice, because these states, by definition, do not impact such receptors above the contribution threshold, the changes in emissions have little to no effect on the non-linked receptor. Furthermore, if EPA were to explicitly consider these reductions within the framework, it would introduce interdependency into the solution for significant contribution. The state-and-receptor-specific definition of significant contribution would devolve into a simultaneous regional action, where particular states would have to either “go first” or where non-linked states would shoulder burdens to receptors to which they are not linked while other linked states would do less. In any case, EPA has verified that even if it were to account for non-linked state reductions under the selected control stringency, the changes in concentrations at the receptors are so small that they do not affect the attainment or maintenance status of any receptor.

For this assessment, EPA used an ozone air quality assessment tool (ozone AQAT) to estimate downwind changes in ozone concentrations related to upwind changes in emission levels. EPA used this tool to analyze the years for which downwind nonattainment and maintenance problems persist for the 2008 ozone NAAQS. Under the base case, EPA projects that such air quality problems persist through 2025. Therefore, EPA focused its assessment on the years 2021 through 2025.

This tool is similar to the AQAT tool used in the CSAPR Update to evaluate changes in ozone concentrations. The ozone AQAT uses simplifying assumptions regarding the relationship between each state's change in NOX emissions and the corresponding change in ozone concentrations at nonattainment and maintenance receptors to which that state is linked. This method is calibrated using two CAMx air quality modeling scenarios that fully account for the non-linear relationship between emissions and air quality associated with atmospheric chemistry. The two CAMx modeling scenarios are the 2016fh1 base year and the 2023fh1 future year scenarios for the 2021 time period. For the 2024 and 2025 AQAT simulations, the two CAMx modeling scenarios are the 2023fh1 future year and the 2028fh1 scenario. See the Ozone Transport Policy Analysis Final Rule TSD for additional details.

For each EGU emission control technology, EPA first evaluated the magnitude of the change in ozone concentrations at the nonattainment and maintenance receptors for each relevant year. EPA next evaluated whether the estimated change in concentration would resolve the receptor's nonattainment or maintenance concern by lowering the average or maximum design values below 76 ppb, respectively. For a complete set of estimates, see the Ozone Transport Policy Analysis Final Rule TSD or the ozone AQAT excel file.

In 2021, there are two nonattainment receptors and two maintenance receptors (see section V.C for details). Table VI.D.1-1 summarizes the results of EPA's evaluation of air quality improvements in 2021 at these receptors using AQAT. EPA evaluated the air quality improvements at the four receptors for the three EGU emission control technologies that are available in the near-term. EPA determined that the average air quality improvement at the four receptors relative to the engineering analytics base case was 0.16 ppb for optimization of existing SCRs and LNB upgrades, and 0.17 ppb when also including optimization of existing SNCRs. EPA determined that the Westport receptor (090019003) remains nonattainment and the Houston receptor (482010024) remains maintenance across these control stringencies, while the Stratford receptor (090013007) switches from nonattainment to maintenance with the optimization of existing SCRs and LNB upgrades (i.e., its average DV becomes clean but its maximum DV remains above the NAAQS). Lastly, the New Haven receptor has all nonattainment and maintenance resolved in the engineering analytics base case. For more information about how this assessment was performed and the results of the analysis for each receptor, refer to the Ozone Transport Policy Analysis Final Rule TSD and to the Ozone AQAT included in the docket for this rule.Start Printed Page 23107

Table VI.D.1-1—Air Quality Improvements at the Four Receptors in 2021 From Near-Term Emission Control Technologies

Monitor ID #StateCountyBaselineSCR optimization + LNB upgradeSCR/SNCR optimization + LNB upgradeBaselineSCR optimization + LNB upgradeSCR/SNCR optimization + LNB upgrade
Average DV (ppb)Average DV (ppb)Average DV (ppb)Max DV (ppb)Max DV (ppb)Max DV (ppb)
90013007ConnecticutFairfield76.1375.9375.9377.0576.8576.85
90019003ConnecticutFairfield78.2778.1278.1278.5878.4378.42
90099002ConnecticutNew Haven73.5973.3873.3775.7475.5375.52
482010024TexasHarris75.6275.5175.5077.2577.1577.13
Average AQ Improvement Relative to Base (ppb)0.000.160.17

Figure 1 illustrates the air quality improvement relative to the estimated representative cost associated with the previously identified near-term emission control technologies. This graph shows improving air quality at the downwind receptors as emission control technologies are assumed to be implemented. In this final rule, EPA has adjusted this graph to reflect a revised estimated representative cost of $1,800 per ton for optimization of already operating SNCRs (which, as explained in section VI.B.1, EPA has adjusted from a value of $3,900 per ton in the proposed rule, which reflected turning on idled SNCRs). In the proposed rule, the SNCR cost extended the right terminus of the solid line out to $3,900 per ton and showed a “knee-in-the-curve” pattern. As noted by commenters, a “knee-in-the-curve” is not on its own a justification for not requiring reductions beyond that point in the cost curve. Even though EPA did not solely rely on this factor in the proposed rule, it notes that this inflection point is greatly diminished and there a less discernable knee when the SNCR optimization cost is updated to reflect $1,800 per ton. In fact, as explained below, EPA does not view the now very slight difference in cost thresholds between $1,600 per ton and $1,800 per ton as significant, and together, EPA views them as comparable in terms of the relationship of available emission reductions to air quality improvement.

The graph in Figure 1 to Section VI.D.1 highlights that the majority of emission reduction potential and air quality improvement occurs from optimization of existing SCRs, with some additional reductions from installation of state-of-the-art combustion control at the same cost threshold. At the slightly higher cost threshold of $1,800 per ton, there is some additional air quality improvement from optimization of existing SNCRs. Taken together, this level of control stringency in emission budgets represents the level at which incremental EGU NOX reduction potential and corresponding downwind ozone air quality improvements are maximized with respect to identified near-term emission control technologies. While the more stringent emission budget levels (e.g., emission budgets reflecting the inclusion of optimization of existing SNCRs) yield a relatively small amount of incremental emission reductions and fewer air quality improvements, they still demonstrate meaningful air quality improvement. Further, after consideration of comments and examining cost data, EPA has identified additional compelling reasons favoring the inclusion of optimization of SNCR controls in the context of this full remedy rulemaking, discussed below. This evaluation shows that EGU NOX reductions for each of the near-term emission control technologies are available at reasonable cost and that these reductions can provide improvements in downwind ozone concentrations at the identified nonattainment and maintenance receptors.

Start Printed Page 23108

EPA finds that the control stringency that reflects optimization of existing SCRs and SNCRs, and installation of state-of-the-art combustion controls results in a substantial number of emission reductions totaling nearly 19,000 tons (approximately 16 percent of the baseline level), resulting in all downwind air quality problems for the 2008 ozone NAAQS being resolved after 2024 (one year earlier than the base case).[155] There are also projected changes in receptor status (from projected nonattainment to maintenance-only) for the Stratford and Westport receptors (the first in 2021, the second in 2024). In addition, the Houston receptor changes from maintenance to attainment in 2023. In 2021, the average level of improvement in ozone concentrations at all four of the receptors is 0.17 ppb.

Including optimization of existing SNCRs yields incremental emission reductions of approximately 1,200 tons as there are fewer sources with this emission control technology. As noted in the proposed rule, a small portion of the coal fleet had this technology in place (14 percent), and of that small portion, many units with these SNCR controls had emission rates of 0.13 lb/mmBtu or less (many operating less than 0.1 lb/mmBtu), suggesting they were already optimizing their SNCRs. Analysis using the AQAT tool suggests that optimization of existing SNCRs has an average air quality improvement of 0.01 ppb. While having no further impact on receptors' classification status, it does deliver additional improvement at the problematic receptors.

Given the small portion of the EGU fleet with existing SNCRs in the 12 linked states, the limited number of additional reductions, and the relatively higher cost for this emission control technology, EPA had proposed to determine that the potential emission reductions associated with optimizing existing SNCRs not be required to eliminate significant contribution from the 12 linked states under the 2008 ozone NAAQS. Based on comments EPA received and outlined below, along with subsequent review of cost data and additional considerations, EPA is including emission reduction potential from this emission control technology in the state emission budgets for this final rule.

Comment: Commenters suggested that reductions from optimizing existing SNCRs should be included in the final rule consistent with Clean Air Act requirements and the full remedy nature of this action. These commenters noted that EPA's touchstone metric in the step 3 multi-factor analysis of “maximizing” air quality improvement relative to representative marginal cost was not a sufficient reason to exclude these reductions. They suggest it is eminently “reasonable” to require EGUs to operate all existing controls, for which they have already made significant capital expenditures to purchase and install. These commenters argued that the reductions, even if small, still delivered air quality improvement in a meaningful timeframe at downwind receptors linked to upwind contribution. The same commenters also noted that these emission control technologies may cost less than EPA suggested in the proposed rule because most of the SNCRs are already operating to some degree at a much lower allowance price incentive.Start Printed Page 23109

Response: EPA is including SNCR optimization in its EGU control stringency in this final rule. EPA examined both its methodology and its cost assumptions and determined it was appropriate to include SNCR optimization-driven reductions in identifying significant contribution. EPA concludes that most of these units are already operating and, in most cases, would not incur the additional FOM cost associated with $3,900 per ton estimate included in the proposed rule, and reductions were likely significantly less expensive, consistent with the commenters' observation that a broad set of units appeared to be incentivized to operate these controls under the CSAPR NOX Ozone Season Group 2 Trading Program (which applied control stringency levels with a representative cost of $1,400 per ton). This technology inclusion was further supported by the observation that most SNCR-controlled units already appear to be operating at lower rates compared with their higher historical emission rates (indicating partial operation) even with the current allowance price substantially under $1,800 per ton.

There are additional considerations unique to EGUs with existing SNCRs that EPA has determined support including their optimization as part of EPA's identified control stringency, such as:

  • These controls are already installed and available for operation on these units;
  • they are on average already partially operating, but not necessarily optimized;
  • the reductions are available in the near-term (during ozone seasons when the problematic receptors are projected to persist), including by the 2021 Serious area attainment date;
  • these sources are already covered under the existing CSAPR NOX Ozone Season Group 2 Trading Program and thus have the monitoring, reporting, recordkeeping, and all other necessary elements of compliance with the trading program already in place;
  • the overall compliance burden and total cost is relatively low, and the incremental cost of operating the technology is not capital intensive.

Indeed, when comparing units of similar size and operation, the absolute annual cost of operating SNCR controls in total dollar terms is often comparable to or less than the cost of operating SCR controls. However, the significantly lower NOX removal efficiency for existing SNCRs (20 to 25 percent) compared to existing SCRs (60 to 90 percent) results in a higher cost-per-ton estimate.

Another consideration that weighs heavily in favor of including the optimization of existing SNCRs as part of EPA's identified control stringency is that emission budgets are set using historical data as a starting point, thus capturing the emission reductions achieved by the EGUs already optimizing their SNCR controls. In other words, state emission budgets assume these units are to continue optimizing those controls. At the same time, EPA's proposed approach would have implicitly allowed EGUs not fully operating their SNCRs to continue to not do so, avoiding the associated cost, and reaping a competitive advantage over those EGUs who, in fully operating their controls, are acting in a more environmentally responsible manner. EPA views this treatment of higher emitting units to be problematic, when the number of EGUs already optimizing their SNCR controls underscores the cost-effectiveness and feasibility of this control measure. Further, as proposed, EPA is including optimization of existing SCRs in its identified control stringency. SCRs are more capital-intensive investments with much better environmental performance. If EPA failed to include optimization of existing SNCRs in its identified control stringency in this action, EGUs that chose SNCRs, which is a less effective form of emission control, would be allowed to continue not operating that control. Considerations of effective and equitable environmental policy strongly weigh against allowing such a result and the perverse incentives it would tend to foster.

These factors, coupled with EPA's final rule cost evaluation, leads the Agency to include optimization of existing SNCRs as part of its identified control stringency. As such, EPA is determining that the full operation of all existing post-combustion controls (both SCRs and SNCRs) and state-of-the-art combustion control upgrades from units constitute the Agency's identified control stringency for EGUs and the associated emision reductions are reflected in the new emission budgets in this final rule. This determination for EGUs is the result of the assessment of the multiple factors and considerations listed above rather than any single factor.

Finally, EPA is determining to not incorporate any additional generation shifting associated with optimization of existing SNCRs, as its updated costs are commensurate with levels of generation shifting already associated with the optimization of existing SCRs reflected in the new state emission budgets. In the proposed rule, EPA identified 1,700 tons of emission reductions from generation shifting associated with optimization of existing SNCRs at a representative cost of $3,900 per ton. Because EPA is determining that $3,900 per ton is not the cost associated with optimizing these partially operating SNCR controls, the Agency is not including that 1,700 tons of generation shifting reduction potential in the state emission budgets in this final rule. Therefore, the emission reductions associated with optimization of existing SNCRs are approximately 1,200 tons for the 12-state region.

Comment: Some commenters suggest that EPA examine higher cost thresholds consistent with downwind state RACT requirements.

Response: EPA first notes that it is including all identified EGU emission controls that are possible to implement during the period during which the upwind state remains linked to a downwind nonattainment or maintenance receptor. While EPA believes the stringency of downwind emission requirements can be useful information in evaluating which control stringencies should be considered upwind, it is—on its own—not dispositive of what that upwind stringency should be. As demonstrated through EPA's air quality modeling, the air quality impact (generally expressed in ppb of ambient ozone concentration at a downwind receptor) of a ton of emissions reduced varies by geography, with areas where the receptor is located generally having a much higher ppb per ton of emissions impact. Therefore, the home state where a receptor is located may generate much greater environmental and public health benefit from a ton of emissions reduced in that state than in an upwind state. In many cases, that may merit a different level of stringency for the home state. However, EPA does view the EGU control stringency it is implementing in this final rule as largely consistent with those EGU emission controls covered by RACT requirements in downwind states (e.g., optimize existing controls and upgrade to state-of-the-art combustion controls). While installation of new post-combustion controls (SCR or SNCR) may also qualify for RACT, EPA's analysis is that such controls could not be operational on a fleetwide scale before all downwind receptors are projected to resolve. Controls associated with the selected EGU control stringency are implementable by the 2021 ozone season (or in the case of upgraded or new combustion controls, by the 2022 ozone season; see the discussion in section VI.C and in the EGU NOX Mitigation Strategies Final Start Printed Page 23110Rule TSD for details). Thus, as to the 2021 and 2022 ozone seasons these are the only emision controls for EGUs that EPA is assessing for this timeframe because they are the only ones that are possible. See Wisconsin, 938 F.3d at 320.

As discussed above in section VI.C, EPA estimates that the time necessary to install new SNCR or new SCR controls (represented by $5,800 per ton and $9,600 per ton cost) on a regional basis across multiple EGUs is approximately 39 to 48 months. While a single new SNCR may be installed within 16 months, for the reasons explained in section VI.C.1, a time frame that encompasses the ability for a unit to make a unit-specific choice of what post-combustion control (SCR or SNCR) is best for its configuration and future operating plans is appropriate. Therefore, EPA considers the timing estimates for SNCR and SCR together, and the 39-48 month time frame for SCR installation (with its superior NOx control efficiency) is the most appropriate time period to use for assessing post-combustion controls. Assuming a final rule in the spring of 2021, this means that these controls could not be operational by the 2024 ozone season, and therefore the reduction potential is not available until the 2025 ozone season. According to EPA's air quality assessment, there are no remaining air quality receptors in 2025 assuming the control stringency identified in this final rule for EGUs is already in place in the 12 linked states. It is not necessary to require emission controls that can only be operational at a point in time when EPA's projections demonstrate there is no remaining interstate transport problem for the 2008 Ozone NAAQS.

EPA requested comment on its proposed determination that new post-combustion controls (SCR or SNCR) are not possible to implement on a regional basis by the start of the 2024 ozone season), and if evidence established such controls were possible, how EPA might apply its step 3 multi-factor analysis in that circumstances. EPA received comments on this topic and addresses the timing assumptions in Section VI.C.1. Moreover, the Appendix to the Ozone Transport Policy Analysis Final Rule TSD further discusses how, even if the controls were available on an earlier time scale, the multi-factor assessment would not necessarily indicate their inclusion in this rule.

2. Non-EGU Assessment

The Agency used CoST and the 2023 projected emissions inventory to identify uncontrolled emissions sources or units and applied controls to emissions units with 150 tpy or more of pre-control NOX emissions, which is an emissions threshold that represents a comparable unit size to 25 MW for EGUs. EPA categorized the CoST results by the control technologies, calculated a weighted average cost per ton (in 2016$) for emission reductions associated with each technology, and identified two tranches of potential reductions based on estimated cost effectiveness (for details see section VI.B.2). EPA took a series of steps to further verify and refine the NOX emission reduction potential estimated by CoST, the CMDb, and the 2023 projected inventory and found that the cost-effective emission reductions in tranche one were from SCR applied to glass furnaces and SNCR applied to cement kilns. These controls could likely take 2-4 years to install. Therefore, at the time of this final rule, EPA is concluding that the 2023 ozone season is the earliest ozone season by which these non-EGU controls could be installed (for details see section VI.C.2).

Using 2023 as the potential earliest date by which controls for glass furnaces and cement kilns can be installed, EPA assessed whether these emission controls should be required at step 3 under its multi-factor test.[156] EPA estimated that across the 11 states linked to the remaining receptor in Connecticut in 2023 (Westport), the available emission reductions from tranche one at less than $2,000 per ton are 1,505 ozone season tons.[157]

Using AQAT, EPA assessed whether this level of emission reductions would have a meaningful effect on the Connecticut receptor. EPA determined that the improvement in air quality at this receptor from these emission reductions is 0.03 ppb. This potential air quality improvement is about an order of magnitude less than the air quality improvement EPA expects to obtain from the emission controls identified in its selected control stringency for EGUs in 2023, which, at a representative cost of $1,800 per ton,[158] is estimated to improve air quality at the remaining Connecticut receptor by 0.28 ppb. These air quality improvements and representative costs support the Agency's position, consistent with its proposed rule, that requiring these non-EGU controls is not warranted under EPA's step 3 multi-factor analysis.

Based on this assessment, the Agency determines under the multi-factor test that even the likely most cost-effective reductions from non-EGU sources (i.e., those below $2,000 per ton in tranche one) do not rise to the level of “significance” that would justify mandating them under the good neighbor provision for the 2008 ozone NAAQS.[159] In the proposed rule, EPA encouraged stakeholder comments on the analysis with respect to the tranche one non-EGU control strategies.

Comment: One commenter suggested that EPA should consider SCR as a control technology for cement plants. The commenter stated that SCR has been used at cement kilns across the globe and that a cement plant in Joppa, Illinois has successfully demonstrated its use with a reported 80 percent removal rate for NOX, while a plant in Midlothian, Texas, has obtained a permit to install SCR units on its kilns.

Response: The Agency appreciates the information from the commenter about SCR controls on cement kilns. However, what the comment does not consider is the time it has taken to install controls at the two plants cited. The SCR installation at the Joppa, IL plant took approximately 6 years to install. The SCR at the Midlothian, TX cement plant is currently not operating, to the best of EPA's knowledge. Cost and testing Start Printed Page 23111concerns have led to slow acceptance of SCR at cement kilns in the United States. The examples provided suggest the time to install these SCRs is much longer than downwind air quality problems are projected to persist for the 2008 ozone NAAQS.

EPA estimates that the 2023 ozone season is the earliest ozone season by which the 111 identified non-EGU emissions units in tranche two could be retrofitted or have controls installed. In tranche two, the weighted average cost of the estimated emission reductions from non-EGU emissions sources ranges from $5,000 to $6,600 per ton. Across the 11 states linked to the remaining receptor in Connecticut in 2023 (Westport), the Agency identified approximately 11,100 tons of potential ozone season emission reductions by applying layered combustion, NSCR (non-selective catalytic reduction) or layered combustion, and ultra-low NOX burners in combination with SCR to 111 emissions units in the oil and gas industry and several manufacturing industries. Since the proposed rule, EPA verified existing control information and refined the NOX emission reduction estimates for emissions sources/units in tranche two. Of the approximately 11,100 tons of potential ozone season emission reductions, EPA determined that approximately 10 percent of those estimated reductions are from sources/units already controlled. In the proposed rule EPA sought comment on the feasibility of further controlling NOX from IC engines and large ICI boilers, including optimizing combustion and installing ultra-low NOX burners.

EPA's assessment is that, with implementation of the new emission budgets for EGUs reflecting the Agency's selected control stringency (see section VI.D.1.), there will no longer be any downwind receptors in 2025 with respect to the 2008 ozone NAAQS. Focusing then on whether there are any non-EGU NOX emission reductions available to address significant contribution under the step 3 multi-factor test in either the 2023 or 2024 ozone seasons, based on its assessment EPA is concluding that any such potentially available reductions would not be justified. EPA's assessment determined that there is a relatively smaller quantity of NOX reductions that may be available from the non-EGU control strategies in tranches one and two in these years, across the 11 states linked to the remaining receptor. These control strategies are estimated to have a limited impact on further improving air quality at this receptor for this rulemaking. As shown in the Ozone Policy Analysis Final Rule TSD, the incremental effects of emission reductions from non-EGUs do not affect the status of any of the four receptors in any of the relevant years compared with EPA's EGU control stringency. For more information, refer to the Ozone Transport Policy Analysis Final Rule TSD. EPA therefore is concluding that no emission reductions from non-EGU sources are necessary to eliminate significant contribution under the good neighbor provision for the 2008 ozone NAAQS.

EPA solicited comment on its analysis, and whether, based on updated or more complete information, there may be grounds to find non-EGU emission reductions are necessary to address significant contribution for the 2008 ozone NAAQS.

Comment: EPA received several comments in response to this request. Some commenters tended to agree that with more complete information, further analysis would not find it necessary to further control emissions from non-EGU sources in this rule. A group of industry trade associations stated that without highly cost-effective options to reduce emissions from non-EGU emissions sources/units, the estimated reductions did not rise to the level of significance to mandate controls. Another commenter stated that the most appropriate mechanisms to consider whether further limits on NOX emissions from industry boilers, furnaces and other emission sources are cost-effective are the existing NSR/PSD, NSPS, and RACT programs. This commenter stated that there is no need to apply additional programs on top of existing programs, or to circumvent existing programs, that are designed to address the issue of cost-effective emissions controls.

Another commenter stated that EPA should direct states to submit revisions to their SIPs because the SIP planning process is the best platform for the identification of potential NOX emission reductions at the local level that may be necessary in non-EGU industry sectors. State and local air pollution control agencies have access to the detailed emissions inventory data from sources and emissions units in non-EGU industry sectors. With this data, states can assess whether additional emission reductions are necessary at the local level from non-EGU industry sectors.

Lastly, other commenters disagreed and stated that EPA lacks statutory authority to exclude non-EGU emissions sources from the coverage of the good neighbor provision, which extends to “any source or other type of emissions activity” that significantly contributes to downwind nonattainment or interferes with downwind maintenance.

Response: EPA stated in the proposed rule that it understands the methodology employed in its assessment was one approach to assessing emission reduction potential from non-EGU emissions sources or units and to determining an appropriate control stringency level for non-EGU sources. EPA also provided details on determining the 150 tpy emissions threshold in the section titled Background for Determining Source Size/Threshold for Non-EGU Emissions Sources in the memorandum titled Assessing Non-EGU Emission Reduction Potential. Based on EPA's analysis for this final rule and considering comments received, EPA determined that its analysis presents a credible analytical foundation on which to conclude that new emission controls on non-EGU sources are not required from the linked upwind states in order to address significant contribution or interference with maintenance of the 2008 ozone NAAQS.

Comment: The Agency received a number of comments on its step 3 analysis to determine whether any emission reductions should be required from non-EGU sources/units to address significant contribution under the 2008 ozone NAAQS. These comments covered a variety of issues related to the assessment of emission reduction potential from non-EGU sources/units. Environmental organizations and downwind states submitted comments that focused on the Agency's determination that further emission reductions would not need to be required from non-EGU sources/units. These comments emphasized that the assessment of non-EGU emission reductions was improperly limited to (a) Controls that would cost $2,000 per ton of emission reductions and (b) a narrow set of potential source types or emissions units. There were also several comments on EPA's decision to analyze emissions units of 150 tpy and larger for the non-EGU analysis. Commenters stated that previous transport rule makings analyzed emission units of 100 tpy and greater.

There were also several comments on the legal requirements to evaluate and include emission reductions from non-EGU emissions sources/units in the rule. The comments emphasized both the impossibility threshold from recent court decisions and data availability. One commenter said that a refusal to include non-EGU emission reductions in the rule would represent an abdication of statutory responsibility. Several comments expressed frustration Start Printed Page 23112that the Agency has claimed data uncertainty issues in interstate transport rulemakings for years and that should no longer be a viable reason to exclude non-EGU emissions sources/units. The commenters stated that this is backed up by the decision in Wisconsin.

Finally, a number of stakeholders from industry associations and upwind states submitted comments stating they agreed with the proposed decision not to include emission reductions from non-EGU emissions sources/units in this rule. The commenters recognized the data limitations faced by the Agency, saying that additional emission reductions from this sector are not necessary to meet obligations under the good neighbor provisions. All of these groups provided limited additional information beyond what the Agency possessed and came to the same conclusions with regard to emission reductions from non-EGU sources/units. A point made in several comments was that emission reductions would not be able to be achieved before the 2023 ozone season due to the timing it would take to install and make operational the emission control devices.

Response: EPA disagrees that the assessment of non-EGU emission reduction potential was unnecessarily limited by carving out large numbers of potential sources, controls, and locations. Using the best information currently available to the Agency, EPA extended its emission reduction and air quality analyses beyond EGUs to include many major stationary source sectors in the linked upwind states, including non-EGU emissions sources in various industry sectors (see Table 2 in the September 1, 2020 document titled Assessing Non-EGU Emission Reduction Potential for a summary). In the analyses, we determined that emissions reductions from non-EGU sources will have a relatively small effect on any downwind receptor in the year by which such controls could likely be installed and do not rise to the level of “significance” that would justify mandating them under the good neighbor provision for the 2008 ozone NAAQS. Further, in the September 1, 2020 memorandum, EPA included a discussion of the assessment for determining an appropriate emissions size threshold comparable to those EGUs included in this and previous transport rulemakings.

In addition, EPA disagrees that the use of a $1,600/ton EGU threshold as a roughly equivalent threshold to assess non-EGU controls is inappropriate. We note that the $2,000/ton threshold value used for assessing non-EGU controls provides a rough equivalence with the threshold value and analysis for EGUs. The $2,000/ton threshold value is a weighted average of control costs, while EPA's cost threshold for EGUs is based on a 90th percentile metric. A 90th percentile metric provides a higher cost threshold for assessing potential controls than a weighted average cost. In other words, the $2,000/ton threshold EPA used for evaluating non-EGU emissions sources/units represents a relatively lower cost/ton threshold value for considering potential controls.

EPA also believes that its determination with respect to emissions reductions from non-EGU sources in this action is not premised on “uncertainty,” or lack of information, but rather a finding based on the analysis of tranche 1 and tranche 2 controls that those non-EGU emission controls that could be potentially available at a cost-effectiveness comparable to EGU controls do not produce sufficient total emission reductions or downwind air quality impacts to be justified under EPA's step 3 multi-factor analysis. The emissions control strategy EPA assessed for non-EGU emissions sources across all twelve states did not generate sufficient air quality improvements to justify requiring.

Additional responses to these comments are provided in the RTC Document included in the docket.

EPA also requested comment on a number of questions related to specific control technologies on non-EGU emissions sources the Agency evaluated, and in particular sought feedback and data from stakeholders with relevant expertise or knowledge. Recognizing the limitations and uncertainties in the existing data EPA used in the assessment of non-EGU emission reductions in the proposed rule, EPA requested comment to assist in substantiating whether the assessment is fully supportable based on additional information and analyses not currently available to the Agency.

Comment: One industry association (National Lime Association) prepared a cost estimate using publicly available information from the EPA Control Cost Manual Worksheet and generic public emission factors from EPA Standard AP-42. The industry-specific report demonstrated: (a) The industry could not possibly achieve any meaningful reductions in NOx emissions by the 2021 ozone season to eliminate “significant” contribution under the 2008 ozone NAAQS; and (b) even considering the most favorable application of retrofit SNCR control in the industry, installation of such controls could not be considered “cost-effective” in the context of this rule. Another trade association stated that obtaining information on NOx emissions units, much less sector-specific information on NOx emission units for purposes of the multi-factor test, would be exceedingly challenging based on available state and local air authority emission inventories and potentially proprietary technology and site-specific cost information.

Another commenter provided unit-specific information prepared for four-factor analyses for recent Regional Haze SIPs for several units in the iron and steel industry. Lastly, another commenter stated that developing a more complete non-EGU inventory is an essential task for EPA. EPA should continue to develop its non-EGU inventory for two purposes: (i) If the New York metropolitan area does not attain the 2008 ozone NAAQS in 2024, as EPA projects, additional emission reductions throughout the region may be necessary, and (ii) EPA is statutorily mandated to act on states' Good Neighbor SIPs for the more stringent 2015 ozone NAAQS now or in the coming months. The commenter concluded that EPA may ultimately need to issue FIPs in instances of SIP disapprovals and emission reductions beyond the EGU sector will likely be required for the New York metropolitan area to attain the 2015 ozone NAAQS. The commenter concluded that because of the complexity of non-EGU operations and control options, EPA should engage with states and affected industries to ensure an accurate inventory and control analysis.

Response: EPA agrees that securing sufficient, detailed sector- and unit-specific information for NOx emission units and related costs to use for the multi-factor test has been difficult.

In the proposed rule, to help inform further technical review and comments, the following Excel workbooks were available in the docket and referenced in the memorandum titled Assessing Non-EGU Emission Reduction Potential: (i) For a summary of the CoST run results CoST Control Strategy—Max Reduction $10k 150 tpy cutoff 12 States Updated Modeling—No Replace—07-23-2020, and (ii) for summaries of emission reductions by control technologies, Control Summary—Max Reduction $10k 150 tpy cutoff 12 States Updated Modeling—No Replace—05-18-2020. Note that the CoST Control Strategy—Max Reduction $10k 150 tpy cutoff 12 States Updated Modeling—No Replace—07-23-2020 Excel workbook includes a READ ME worksheet that provides details on the parameters used for the CoST run.Start Printed Page 23113

To improve the underlying data used in an assessment of emission reduction potential from non-EGU sources, EPA requested comments on: (i) The existing assessment of emission reduction potential from glass furnaces and cement kilns; and (ii) emission reduction potential from other control strategies or measures on a variety of emissions sources in several industry sectors.

Comment: EPA received limited comments on the existing assessment of emission reduction potential from glass furnaces and cement kilns. A commenter noted that EPA incorrectly identified two cement kilns as eligible for SNCR installation in its analysis. Through a 2017 consent decree with EPA and the Department of Justice, SNCR was not feasible for one of the kilns because of the current configuration of the equipment. For the second kiln, SNCR was already installed because a different configuration allowed for the control installation.

Response: EPA appreciates the submittal of this information.

Comment: EPA received several comments regarding emission reduction potential from other control strategies or measures on a variety of emissions sources in several industry sectors. A few commenters indicated that a 2017 OTC paper titled White Paper on Control Technologies and OTC State Regulations for Nitrogen Oxides (NO) Emissions from Eight Source Categories reflects appropriate control strategies, identifies emission limits and regulations for eight source categories, and details information for four of the 12 states identified as significantly contributing to downwind areas with attainment or maintenance issues for the 2008 ozone NAAQS.

Other commenters cited a 2009 OTC paper that analyzed the cost of installing NOX controls on ICI boilers. The paper concluded that key variables that impact cost analyses include boiler type, boiler firing type, type of fuel combusted, type of emission control, uncontrolled emission rate, controlled emission rate, capital cost of control equipment, financial costs, unit capacity factor (hours/year), flue gas flow rates and temperatures, and commodity prices. The analysis found that NOX control costs for non-EGU emissions sources are highly variable and site-specific and the cost per ton of NOX removed from several control technologies reviewed was significantly above the proposed rule representative cost of the selected EGU control stringency ($1,600 per ton).

Response: EPA appreciates the references provided regarding the assessment of non-EGU emissions sources/units. Non-EGU emissions sources/units are diverse, making them challenging to analyze. Nonetheless, EPA's determination with respect to emission reductions from non-EGU sources in this action is not premised on “uncertainty,” or lack of information. Rather, EPA's finding is based on the analysis of tranche one and tranche two controls. EPA determined that those non-EGU emission controls that could be potentially available at a cost-effectiveness comparable to EGU controls do not produce sufficient total emission reductions or downwind air quality impacts that would justify requiring them under EPA's step 3 multi-factor analysis.

EPA requested comment on the aspects of the assessment presented above of emission reduction potential from the glass and cement manufacturing sectors. The Agency did not receive any comments directly addressing this that were independent of the other comments.

In addition, EPA requested comment on the following:

  • Other than glass and cement manufacturing, are there other sectors or sources that could achieve potentially cost-effective emission reductions? What are those sectors or sources? What control technologies achieve the reductions? What are cost estimates and installation times for those control technologies?
  • Are there other sectors where cost effective emission reductions could be obtained by, in lieu of installing controls, replacing older, higher emitting equipment with newer equipment?
  • Are there sectors or sources where cost effective emission reductions could be obtained by switching from coal-fired units to natural gas-fired units?
  • For non-EGU sources without emissions monitors, what would CEMS cost to install and operate? How long would CEMS take to program and install? Are monitoring techniques other than CEMS, such as predictive emissions monitoring systems (PEMS), sufficient for certain non-EGU facilities that would not be brought into a trading program? If so, for what types of non-EGU facilities, and under what circumstances, would PEMS be sufficient? What would be the cost to install and operate monitoring techniques other than CEMS?

Comment: EPA received several comments in response to this set of questions. Two industry association commenters indicated that where feasible, facilities have already largely replaced or repowered boilers to comply with several other EPA rules (e.g., boiler MACT, Regional Haze Rule, and 1-hour SO2 NAAQS). With respect to fuel switching for boilers, one industry association stated that many of their members undertook fuel switching as a compliance strategy for the boiler MACT. Another commenter cautioned that EPA should consider other factors when evaluating the time necessary to retrofit add-on controls, including the availability of the specialized trades that are needed to complete the retrofit installation of low NOx burners and Clean Air Act permitting obligations, which increase the time needed for a retrofit.

With respect to installation timing and the cost of CEMS, three trade associations provided the following estimates:

  • Installation Timing
  • 28 weeks (7 months)—delivery time for a CEMS shelter with pre-installed analyzers and other equipment is about 24 weeks; installation time and programming would take about another 4 weeks.
  • 16-24 weeks (4-6 months)—CEMS installation would likely take 4 to 6 months if a facility was currently ready to start. However, this timeline does not take into account the time required to obtain capital approval, issue an RFP, engage a consultant, and make any necessary structural modifications to the stack if it cannot accommodate CEMS.
  • Cost
  • $500,000-cost will depend on whether the stack is designed to accommodate a CEMS. If a stack is designed to support a NOX CEMS, the cost to install, program, and certify the NOX CEMS could be $500,000. Ongoing operation and maintenance costs are likely around $150,000 per year.
  • $300,000—$400,000—capital cost for the equipment (assuming a single boiler installation) is approximately $300,000 to $400,000 (2016$). Actual costs at a given facility will vary and will depend on factors including the availability of space and the location of the CEMS air-conditioned shelter.

Additionally, one commenter stated that a rigorous PEMS, if a feasible alternative, would be more expensive than a CEMS. While another commenter stated that PEMS have proven to be very reliable and significantly less expensive to operate and maintain than CEMS. The commenter observed that PEMS minimize the up-front capital costs, as well as the on-going cost of operation, maintenance, and quality assurance.

Response: EPA thanks for the commenters for this information.

EPA requested comments on the feasibility of further controlling NOXStart Printed Page 23114from large ICI boilers and IC engines, including optimizing combustion and installing low NOX burners.

Comment: EPA received several comments in response to this request. One commenter stated that EPA should pursue requiring additional NOX controls on IC engines and large ICI boilers, including optimizing combustion and installing ultra-low NOX burners and offered no specific supporting information. An industry association stated that most of their members' boilers are already equipped with low NOX burners. The members' experiences with the retrofit installation of low NOX burners on existing boilers are that the reductions achieved vary from boiler to boiler as a function of: (a) The existing configuration of the boiler, (b) the boiler fuel, and (c) the day-to-day operation of the boiler to meet the demands for thermal energy from the end-use processes or customers.

Another industry association noted that recent Regional Haze Rule-related analyses for forest products industry boilers indicated that the cost of installing additional controls (LNB/FGR, SNCR, or SCR retrofits) is generally more than $5,000/ton, based on representative actual emissions. The commenter stated that if EPA were to determine that NOX controls on ICI boilers should be required, no new controls could be implemented by the 2021 ozone season and it would be difficult to implement controls before 2024. Facilities would need a minimum of four years to implement controls after promulgation of any requirement to do so because the process to undertake a retrofitting project is complex, involving design, engineering, permitting, procurement, and installation. The commenter stated that since the start of the COVID-19 pandemic, the time necessary to implement construction projects has increased considerably.

Additionally, two energy companies offered their experiences with modifying IC engines. One energy company indicated that after the modification it took three to five years to get engine performance back to previous levels. The same energy company stated that as operations evolve, where feasible, they will install newer engines, or turbines, at natural gas compressor and storage sites. A second energy company has already replaced some older uncontrolled IC engines with new, state-of-the art low NOX compressor engines and/or combustion turbines within its fleet, intending to operate the newer IC engines preferentially over the older units. They stated that regulating IC engines at compressor stations will not result in significant reductions in actual NOX emissions, and they do not believe it is cost-effective.

Response: EPA thanks the commenters for this information.

EPA requested comment on whether EPA should require that large non-EGU boilers and turbines—as defined in the NOX SIP call as boilers and turbines with heat inputs greater than 250 Million British Thermal Units (mmBtu) per hour or with NOX emissions greater than 1 ton per ozone season day [160] —within the 12 states employ controls that achieve emission reductions greater than or equal to what can be achieved through the installation of low NOX burners.

Comment: EPA received a few comments in response to this request. One industry association stated that there is no justification for a requirement for large industrial boilers within the 12 states covered by this rule to employ controls that achieve emission reductions greater than or equal to what can be achieved through the installation of low NOX burners. Such a requirement could be infeasible for certain types of boilers without a significant capital investment and could increase CO emissions above allowable levels. The commenter suggested that these types of requirements are better implemented through the New Source Review (NSR) permitting process where a site-specific analysis is required. Another commenter stated that such a requirement could require very significant capital investment for retrofitting certain types of existing boilers and may not be feasible for certain types of boilers.

Response: EPA thanks for the commenters for this information.

EPA requested comment on (i) the magnitude of the emission reductions that could be achieved by requiring that large non-EGU boilers and turbines install controls that achieve emission reductions greater than or equal to what could be achieved through the installation of low NOX burners, (ii) the prevalence of these or better NOX controls already in place on this equipment in these 12 states, and (iii) the time it typically takes to install such controls. EPA did not receive any comments in direct response to this comment solicitation.

As mentioned in the discussion above on emission reductions from the EGU sector, EPA understands that it is generally possible to install LNB on EGU boilers fairly quickly and that these burners can significantly reduce NOX emissions. EPA notes that in the original interstate transport rule, the NOX SIP call, the Agency concluded that controls on large, non-EGU boilers and turbines were cost effective and allowed states to include those emissions sources in their budgets as a means of providing additional opportunities to reduce state-wide NOX emissions in a cost-effective manner.[161]

Also, five of the 12 states that are subject to this rulemaking are within the Ozone Transport Region (OTR)—Maryland, New Jersey, New York, Pennsylvania, and Virginia. As member states of the OTR, these five states are required to implement reasonably available control technology (RACT) state-wide on major sources of emissions.[162] It is likely that NOX controls, such as low NOX burners, are already in wide-spread use on large non-EGU boilers and turbines within these five states. However, such controls may not be as widely used in states outside of the OTR. Therefore, the Agency also solicited comments on the following:

  • How effective are ultra-low NOX burners or low NOX burners in controlling NOX emissions from ICI boilers?
  • Are they generally considered part of the process or add-on controls? If they are part of a process, how could EPA estimate the cost associated with changing the process to accommodate ultra-low NOX burners and low NOX burners?
  • What are the costs (capital and annual) for these as add-on control technologies on ICI boilers?
  • What are the earliest possible installation times for these control technologies on ICI boilers? EPA believes it is generally possible to install low NOX burners on EGU boilers relatively quickly and that low NOX burners can significantly reduce NOX emissions. EPA solicited comment on whether this is also true for large non-EGU ICI boilers.
  • Do some of the emissions units included in the summary already have either add-on controls or controls that are part of a process? If so, what control is on the unit and what is the control device (or removal) efficiency?
  • Natural gas compressor stations are the largest NOX-emitting non-EGU Start Printed Page 23115sector [163] affecting the 12 states that are the subject of this final rule, and many of these facilities are powered by decades-old, uncontrolled IC engines. Should emission reductions be sought from the IC engines at these stations, either through installing controls, upgrading equipment, or other means?
  • How effective is low emission combustion in controlling NOX from IC engines?
  • What is the cost (capital and annual) for low emission combustion on IC engines?
  • What is the earliest possible installation time for low emission combustion on IC engines? In lieu of installing controls, is replacing older, higher emitting equipment with newer equipment a cost-effective way to reduce emissions from IC engines?
  • Do some of the emissions units included in the summary already have either add-on controls or controls that are part of a process? If so, what control is on the unit and what is the control device (or removal) efficiency?

The Agency encouraged stakeholders with particular expertise, such as source owners and operators, state agencies, trade associations, and knowledgeable non-governmental organizations, to evaluate the information available in the docket and presented above and provide updates, corrections, and other information as may assist in improving EPA's ability to more accurately assess non-EGU emission control strategies relevant to addressing interstate ozone transport.

Comment: EPA received relatively few comments directly in response to this request. One NGO cited EPA's 2016 Final Technical Support Document (TSD) for the Final Cross-State Air Pollution Rule for the 2008 Ozone NAAQS, Assessment of Non-EGU NOXEmission Controls, Cost of Controls, and Time for Compliance Final TSD with information on controls and costs for IC engines. Another comment encouraged the Agency to pursue controlling NOX from ICI boilers and IC engines, including optimizing combustion and installing low NOX burners.

Response: EPA notes that the 2016 Final Technical Support Document (TSD) for the Final Cross-State Air Pollution Rule for the 2008 Ozone NAAQS, Assessment of Non-EGU NOXEmission Controls, Cost of Controls, and Time for Compliance Final TSD was prepared for the purpose of presenting and seeking comment on the then currently available information on emissions and control measures for sources of NOX other than EGUs; it was not prepared for use in conducting a rigorous regulatory analysis under the step 3 multi-factor test, nor for establishing specific emissions limits.

3. Overcontrol Analysis

As part of the air quality analysis using the Ozone AQAT, EPA evaluated potential over-control with respect to whether (1) the expected ozone improvements would be greater than necessary to resolve the downwind ozone pollution problem (i.e., beyond what is necessary to resolve all nonattainment and maintenance problems to which an upwind state is linked) or (2) the expected ozone improvements would reduce the upwind state's ozone contributions below the screening threshold (i.e., 1 percent of the NAAQS; 0.75 ppb).

In EME Homer City, the Supreme Court held that EPA cannot “require[ ] an upwind State to reduce emissions by more than the amount necessary to achieve attainment in every downwind State to which it is linked.” 572 U.S. at 521. On remand from the Supreme Court, the D.C. Circuit held that this means that EPA might overstep its authority “when those downwind locations would achieve attainment even if less stringent emissions limits were imposed on the upwind States linked to those locations.” EME Homer City II, 795 F.3d at 127. The D.C. Circuit qualified this statement by noting that this “does not mean that every such upwind State would then be entitled to less stringent emission limits. Some of those upwind States may still be subject to the more stringent emissions limits so as not to cause other downwind locations to which those States are linked to fall into nonattainment.” Id. at 14-15. As the Supreme Court explained, “while EPA has a statutory duty to avoid over-control, the Agency also has a statutory obligation to avoid `under-control,' i.e., to maximize achievement of attainment downwind.” 572 U.S. at 523. The Court noted that “a degree of imprecision is inevitable in tackling the problem of interstate air pollution” and that incidental over-control may be unavoidable. Id. “Required to balance the possibilities of under-control and over-control, EPA must have leeway in fulfilling its statutory mandate.” Id.[164]

Consistent with these instructions from the Supreme Court and the D.C. Circuit, EPA first evaluated whether reductions resulting from the emission budgets for EGUs in 2021 and 2022 can be anticipated to resolve any downwind nonattainment or maintenance problems. As discussed in Section VI.D.1, the proposed control stringency (represented by a $1,600 per ton cost threshold) was adjusted to a control stringency that includes optimization of existing SNCRs (represented by a $1,800 per ton cost threshold) in this final rule. This assessment shows that the emission budgets reflecting $1,800 per ton would change the status of one of the two nonattainment receptors (first shifting the Stratford monitor to a maintenance-only receptor in 2021 and then shifting that monitor to attainment in 2022). However, no other nonattainment or maintenance problems would be resolved in 2021 or 2022. EPA determined that none of the 11 states are solely linked to the Stratford receptor that is resolved at the $1,800 per ton level of control stringency in 2022.

Reductions resulting from the $1,800 per ton emission budgets for EGUs would shift the Houston receptor in Harris County, Texas, from maintenance to attainment in 2023. These emission reductions would also shift the last remaining nonattainment receptor (the Westport receptor in Fairfield, Connecticut) to a maintenance-only receptor in 2024. No nonattainment or maintenance receptors would remain after 2024.

Next, EPA evaluated the potential for over-control with respect to the 1 percent of the NAAQS threshold applied in this final rulemaking at step 2 of the good neighbor framework for the $1,800 per ton cost threshold level for each year downwind nonattainment and maintenance problems persist (i.e., 2021 through 2024). Specifically, EPA evaluated whether the emission levels would reduce upwind EGU emissions to a level where the contribution from any of the 12 upwind states would be below the 1 percent threshold that linked the upwind state to the downwind receptors. EPA finds that under the $1,800 per ton EGU cost threshold level for 2021 to 2024 emission levels, all 12 states that contributed greater than or equal to the 1 percent threshold in the base case continued to contribute greater than or equal to 1 percent of the NAAQS to at least one remaining downwind nonattainment or maintenance receptor for as long as that receptor remained in nonattainment or maintenance. For more information about this assessment, refer to the Start Printed Page 23116Ozone Transport Policy Analysis Final Rule TSD and the Ozone AQAT.

Since emission reductions resulting from the $1,800 per ton emission budgets for EGUs are not projected to result in the expected ozone improvements: (1) Being greater than necessary to resolve the downwind ozone pollution problem (i.e., beyond what is necessary to resolve all nonattainment and maintenance problems to which an upwind state is linked) or (2) reducing the upwind state's ozone contributions below the screening threshold (i.e., 1 percent of the NAAQS; 0.75 ppb), EPA finds that the $1,800 control strategy does not result in overcontrol.

Based on the multi-factor test applied to both EGU and non-EGU sources and subsequent assessment of overcontrol, EPA finds that the emission reductions associated with the $1,800 per ton control stringency for EGUs constitute elimination of significant contribution and interference with maintenance without overcontrol from the 12 linked upwind states. Therefore, as discussed in section VII, EPA is establishing emission budgets for EGUs in the 12 linked states that reflect the remaining allowable emissions after the emission reductions associated with the $1,800 per ton control stringency have been achieved. For additional comments and responses and details about the test and the overcontrol analysis, see the RTC and Ozone Transport Policy Analysis Final Rule TSD.

VII. Implementation of Emission Reductions

A. Regulatory Requirements for EGUs

The CSAPR established a seasonal NOX trading program for states determined in that rulemaking to have good neighbor obligations with respect to the 1997 ozone NAAQS. The CSAPR Update established a new seasonal NOX trading program for 22 states determined to have good neighbor obligations with respect to the 2008 ozone NAAQS—the CSAPR NOX Ozone Season Group 2 Trading Program—and renamed the seasonal NOX trading program established in the CSAPR, which now covers only Georgia, the CSAPR NOX Ozone Season Group 1 Trading Program.[165] Each of these trading programs for seasonal NOX emissions established state-level budgets for EGUs and allowed affected sources within each state to use, trade, or bank allowances within the same trading group for compliance. In the CSAPR NOX Ozone Season Group 1 and Group 2 trading programs, sources are required to retire one Group 1 or Group 2 allowance, respectively, for each ton of NOX emitted during a given ozone season. EPA is using the same regional trading approach, with modifications to reflect updated budgets, trading groups, and certain additional revisions, as the compliance remedy implemented through the FIPs to address interstate transport for the states having further good neighbor obligations with respect to the 2008 ozone NAAQS in this rule.

Of the 22 states currently covered by the CSAPR NOX Ozone Season Group 2 Trading Program, EPA is establishing revised budgets for 12 states, as explained below. Therefore, EPA is creating an additional geographic group and trading program comprised of these 12 upwind states with remaining linkages to downwind air quality problems in 2021. This new group, Group 3, will be covered by a new CSAPR NOX Ozone Season Group 3 Trading Program. Aside from the removal of the 12 covered states from the current Group 2 trading program, this rule leaves unchanged the budget stringency and geography of the existing CSAPR NOX Ozone Season Group 1 and Group 2 trading programs.

EPA is using the existing CSAPR NOX ozone season allowance trading system framework, established in the CSAPR for Group 1 and used again in the CSAPR Update for Group 2, to implement the emission reductions identified and quantified in the FIPs for this rule. The new Group 3 trading program is being codified at 40 CFR part 97, subpart GGGGG. As with the existing CSAPR trading programs, emissions monitoring and reporting will be performed according to the provisions of 40 CFR part 75, and decisions of the Administrator under the program will be subject to the administrative appeal procedures in 40 CFR part 78.

Comment: EPA received several comments suggesting that Louisiana not be included in the Group 3 trading program. Commenters suggested that EPA has no basis for including Louisiana in the Group 3 trading program because its linkage geography (i.e., to a receptor in Texas) is separate from the 11 remaining Group 3 states which have linkages to receptors in Connecticut. Several commenters also raised the possibility of under-control in the 12-state trading program should EPA allow trading of emission allowances between Louisiana and the remaining 11 states.

Response: EPA disagrees with comments that Louisiana should not be included in the Group 3 trading program. All covered states in the Group 3 trading program, regardless of the downwind monitors to which they are linked, are subject to emission budgets established based on the same set of emission control measures applied at the same levels of stringency. In similar circumstances in earlier rulemakings to address the good neighbor provision, EPA has routinely included states in a common trading program based on a uniform level of control stringency, not based on whether the states were all found to be linked to the same downwind receptors. For example, the states required to participate in the Group 2 trading program under the CSAPR Update included one state linked only to downwind receptors in Connecticut, two states linked only to downwind receptors in Michigan, and two states linked only to downwind receptors in Texas, as well as other states linked to downwind receptors in multiple states. See 81 FR 74538 tbls. V.E-2 and V.E-3.

Moreover, all states subject to the new Group 3 trading program will be required to comply with the assurance provisions in this final action. The assurance provisions ensure that emissions within a covered state do not exceed that state's emission reduction obligations (see section VII.C.2.). The assurance provisions, and associated variability limits, impose an additional allowance surrender requirement when a state's emissions exceed its budget for a given control period by 21 percent. The additional allowance surrender requirement associated with the assurance provisions provides an incentive for sources within a state to comply with the emission budgets for a given control period, while accounting for the inherent variability in operations and emissions from one year to the next. By limiting the degree to which any state's emissions exceed that state's emissions budget, the assurance provisions reduce concerns that a state covered by the new Group 3 trading program would be able to routinely rely on surplus allowances purchased from another state in the trading program in a different geographic region (or in the same geographic region) instead of reducing emissions within the state. Establishing assurance levels with compliance penalties responds to and complies with the D.C. Circuit's holding in North Carolina requiring EPA to ensure that sources in each state meet their good neighbor obligations while Start Printed Page 23117still taking advantage of the benefits of an interstate trading program. See 531 F.3d at 908. See also 81 FR 74566-67.

Comment: Some commenters asserted that implementation of emission reductions through a state-level, seasonal emissions budget program with trading flexibilities is not sufficient to ensure that reductions are realized on high ozone days when they are most needed. These commenters suggested that EPA replace or supplement its emission trading program with unit-specific emission rate requirements applied on a shorter time scale (e.g., daily). Commenters assert that existing controls must be maintained and operated in accordance with good pollution control practices whenever feasible. Commenters assert that shorter-term NOX emission rate limits must ensure that SCRs are operated in accordance with good pollution control practices at all times the units are operating. They suggest that short-term limits are necessary to prevent units from turning controls off intermittently on days with high ozone in order to harvest additional power that would otherwise be used for control operation.

Response: EPA is finalizing the implementation of required emission reductions through the same ozone season trading program structure successfully used in prior CSAPR rules, CAIR, and the NOX Budget Trading Program associated with the 1998 NOX SIP Call. These trading programs have been demonstrated to be highly effective at achieving emission reductions. For instance, as discussed in greater detail below, EPA has previously demonstrated that in the first CSAPR Update compliance period (i.e., the 2017 ozone season), the budget drove sources, nearly uniformly, to operate their controls for that control period.[166] EPA acknowledges that without adjustments in budget stringency to ensure continued operation of the selected control strategy (or equivalent reductions), this analysis may not hold in later years of a trading program should a sufficient bank of allowances develop that the price signal for continued control operation is weakened. However, EPA has addressed that concern in this rule by making downward adjustments in the budgets to account for known fleet changes. Early in the implementation of the CSAPR Update in 2017, when emission budgets were binding and allowance prices were higher, EPA conducted an analysis on how effectively units were operating their SCRs (1) in response to a trading program implementation measure and (2) on High Electricity Demand Days (HEDD). This analysis was done in the context of responding to petitions from Maryland and Delaware under CAA section 126(b) petition.[167] With this rule in place as of 2021, the situation will be comparable and the analysis of 2017 data provides a good indication of how EPA anticipates sources with post-combustion controls will respond to a trading program implementation measure designed to be a full remedy. Moreover, EPA performed the same analysis using 2019 data and continues to find that units operate their SCRs on HEDD as described below.

In the Maryland/Delaware CAA section 126(b) action, EPA examined the complete set of 2017 ozone-season data and did not find evidence of sources regularly idling controls on high ozone days when subject to a sufficiently stringent budget.[168] EPA found that, based on 2017 emissions data reflecting implementation of the CSAPR Update, 261 of 274 units had ozone-season emission rates below 0.20 lb/mmBtu, indicating they were likely operating their post-combustion controls through most of the ozone season. On average, the 274 units were operating at an average emission rate of approximately 0.088 lb/mmBtu.[169] Consequently, EPA found that on average, SCR-controlled units were operating their SCRs throughout the season and that the petitioner's assertion of the likelihood of trading programs leading to widespread idling of controls was not borne out in the most recently available data. In years following 2017, EPA has seen the seasonal emission rates of some SCR-controlled units increase, while the vast majority continue to operate and optimize their controls. As noted above, this is attributable to the partial nature of the CSAPR Update and consequently that program not being configured to account for fleet changes after 2017. Nonetheless, EPA's analysis of 2017 data shows that the CSAPR Update regional trading program and other EPA regional trading programs have driven significant reductions and can provide continued incentive for control operation in a full-remedy context, so long as the budget is sufficiently stringent.

EPA has revisited the aforementioned examinations of SCR performance rates using 2019 hourly NOX emissions data in place of 2017 data. While there was an increased frequency and number of units turning off their controls in 2019, EPA again found that this did not happen during the hours with the highest generation.[170] As was shown in the analysis conducted for the Maryland/Delaware action, and confirmed based on 2019 analysis, SCR-controlled units generally operated with lower emission rates during high generation hours, suggesting SCRs generally were in better operating condition—not worse, let alone idling—during those days/hours. In other words, EPA compared NOX rates for EGUs for hours with high energy demand and compared them with seasonal average NOX rates and found very little difference, just as it had observed in the 2017 data. Thus, the data do not support the notion of wide-spread reduction of SCR operation on high demand days. Moreover, the auxiliary power used for control operation is small—typically less than one percent of the generation at the facility—and it is, therefore, unlikely that sources would cease operation of controls for such a limited energy savings. Instead, the previous analysis indicated that increases in total emissions on days with high generation are generally the result of additional units that do not normally operate coming online to satisfy increased energy demand and units that do regularly operate increasing hourly utilization, rather than reduced functioning of control equipment. In this action, the Agency concludes that while short-term limits and a regional trading budgets are not necessarily mutually exclusive and could complement each other (and do in fact complement each other since many states already have established emission rate requirements for their EGUs through other control programs such as RACT), in this specific instance, where the Agency is addressing regional air quality issues with regionally uniform levels of control through the flexibilities afforded by a mass-based trading program, specific unit-level control requirements, particularly short-term emissions limits, are not necessary, so long as the mass-based budget is sufficiently stringent. This rule addresses the need for sufficiently stringent budgets through budget adjustments in each year through 2024 Start Printed Page 23118to ensure that stringency levels account for known future changes in the fleet.

Further, EPA finds there to be environmental benefits associated with a mass-based trading program that controls units' total amounts of emissions. This creates an incentive structure resulting in lower-emitting sources tending to operate more than dirtier units. Moreover, EPA's implementation program provides—through an allowance price—an incentive to optimize emissions performance as much as possible. This approach not only encourages units to achieve the rates assumed in the budget-setting process, but to perform at even better rates where better performance can be achieved at a cost lower than the allowance price. By contrast, an implementation mechanism that provides a unit-specific emission rate would not incentivize the unit to perform better than its rate requirement. Thus, the trading program encourages controls to not only operate on high electric demand days, but it could provide a unit additional incentive (through its allowance price) to outperform an equivalent ermission rate assumption implemented through a unit-specific rate requirement.

Finally, as other commenters pointed out, unit-specific short-term emission rates pose significant implementation and rulemaking challenges, because there are more unit-specific characteristics that must be taken into account to arrive at unit-specific rate requirements. In establishing a trading program, EPA is better able to rely with confidence on fleet averages used for calculating state budgets. Were EPA to choose to implement a unit-specific emissions rate regime for implementation, the compliance flexibility afforded by emissions trading would not be available and it would not be possible to rely on fleet average information to the same extent for purposes of establishing appropriate levels of control stringency. EPA would likely be unable to establish such requirements or mandate them in time to meet the 2021 Serious area attainment date.

B. Quantifying State Emissions Budgets

EPA is quantifying state emission budgets consistent with the approach used in the CSAPR Update. However, given Wisconsin's direction to implement a full remedy, EPA must address upwind emission reduction potential beyond the initial year for which it is establishing emission budgets. Whereas in the partial-remedy context of the CSAPR Update, EPA established budgets based only on its assessment of the 2017 analytic year and noted it would revisit future years at a later date, in this action EPA is simultaneously looking at budgets for all relevant future years to comply with the full-remedy directive. Consequently, for the Group 3 states EPA is quantifying specific budgets in each year to ensure that EGUs continue to be incentivized to implement the full extent of EPA's selected control stringency while linkages to downwind nonattainment and maintenance receptors remain unresolved. In effect, by doing this, EPA is accounting for scheduled fleet turnover after the first-year budget. For instance, if State X's budget was 100 tons in 2021, but there are 10 tons of emissions from a unit scheduled to retire at the end of the year and 5 tons expected from a new unit coming online, then the state emission budget for 2022 will reflect these scheduled changes by establishing a budget of 100 tons—(10 tons −5 tons) = 95 tons for the subsequent year. This adjustment in methodology reflects the need to anticipate and respond to scheduled fleet turnover in the power sector in ensuring that the control stringency selected to eliminate significant contribution remains incentivized. Based on the Agency's experience implementing prior good neighbor trading programs, setting emissions budgets that do not account for planned retirements in subsequent years can lead to an erosion in the allowance price signal and hence a reduced incentive to take the mitigation measures identified in EPA's significant contribution determination (e.g., optimize SCRs). EPA's air quality projections demonstrate that even with a $1,800 per ton EGU control stringency, the Group 3 states continue to contribute above the 1 percent of the NAAQS threshold to at least one receptor whose nonattainment and maintenance concerns persist through the 2024 ozone season (with the exception of Louisiana, as discussed in more detail below). As such, and in order to implement a full remedy as required under the Wisconsin decision, EPA is determining that it is necessary to design a step 4 implementation framework that effectively ensures the continued optimization of existing SCR and SNCR controls and the incentive to install or upgrade combustion controls for so long as downwind nonattainment and maintenance concerns persist. Therefore, for all Group 3 states except Louisiana, the emission budget setting process described below applies to each year from 2021 through 2024, with the budgets held constant from 2024 onwards. For Louisiana, the emission budget setting process applies to 2021 and 2022 only, with the budget held constant from 2022 onwards, as the Houston receptor to which Louisiana is linked is projected to be resolved by the 2023 ozone season.

EPA is not increasing the stringency of the program over these years in the sense of requiring any further emission reductions than the control stringency represented by $1,800 per ton achieves. Rather, these budget adjustments account for pre-existing, on-going changes in the EGU sector, which if not accounted for, could significantly weaken the incentive to optimize existing SCR and SNCR controls and install or upgrade combustion controls. By determining emissions budgets for a given emissions control across a range of years (e.g., 2021-2024), EPA is able to best reflect the realization of that technology in any given year. For instance, a unit may be scheduled to retire (independent of any environmental regulation) in 2023. Therefore, the same $1,800 per ton uniform control stringency (i.e., SCR and SNCR optimization, and combustion control installation or upgrade) will produce a different state emissions level (i.e., budget) in 2021 and 2024 due to this change in fleet composition. Having the emissions estimated for each year allows EPA to best ensure the reductions available from the identified control stringency continue to be achieved to eliminate that state's significant contribution. This type of phased implementation preserves the intended control stringency of the rule and is consistent with the direction under the Wisconsin decision to promulgate a full-remedy rule. In prior trading programs, commenters observed that the program's static emission budgets quickly fell behind the rapid pace of change in the power sector fleet. As this occurs, a large allowance bank builds and the price of allowances falls below the price in the initial years. For example, the price of CSAPR Update Group 2 allowances started out at levels near $800 per ton in 2017 and provided a strong signal for the mitigation technology identified in the significant contribution determination. However, in subsequent years as the fleet of covered EGUs changed, the price of those allowances declined to less than $70 per ton in July 2020.[171] Stakeholders have pointed out that these low prices could allow for some backsliding of the emission control technologies (e.g., reduced incentive to operate SCR Start Printed Page 23119controls) that were initially determined to be cost-effective and required to eliminate significant contribution. At the same time that the incentive for EPA's selected control stringency weakens, EPA's data show that downwind air quality receptors continue to persist at step 1, and the overall level of anthropogenic emissions from an upwind state continues to contribute to those receptors above the contribution threshold at step 2. Under these conditions, a legal basis exists within EPA's 4-step framework to undertake measures that ensure EGUs continue to implement EPA's selected control stringency. Stated differently, EPA is confident that it is well within its statutory authority under CAA section 110(a)(2)(D)(i)(I) to impose on each covered EGU in a linked upwind state an emission limit that is enforceable and permanent, reflective of the control stringency EPA has determined is needed to eliminate significant contribution from that state. EPA's approach in this rule better incentivizes the selected control stringency while retaining the flexible compliance benefits of an interstate-trading approach to implementation.[172]

In summary, in order to implement a full remedy, EPA is implementing ozone season budgets for each year that reflect ongoing incentivization of the emission reduction measures identified in this rule, with a final budget being implemented in 2024 (the last year EPA projects downwind receptors to remain unresolved) and then held constant for each year thereafter. EPA requested comment on this approach and is finalizing the same approach that it proposed.

Comment: EPA received comment noting some stakeholders' strong support for the issuance of NOX emissions budgets that were updated each ozone season to account for fleet changes. Commenters also claimed that failing to do so would raise concerns that, as the cost of allowances falls, units would be incentivized to buy cheaper allowances rather than optimize controls. They note this dynamic would undercut the purpose of the trading program, and EPA's efforts to address this issue by adjusting the NOX emissions budgets each ozone season in response to fleet changes are necessary to avoid such an outcome. They conclude it is a fair and equitable practice that ensures continued optimization of emissions controls. EPA also received comment opposing this methodology, generally for the stated reasons that (1) the methodology differs from past EPA methodology, (2) EPA's budget methodology should allow for other existing sources to replace the retiring generation by assuming a corresponding replacement or even increase in emissions, (3) some of the scheduled future retirements are uncertain, and (4) reducing budgets based on retirements but continuing to allocate allowances to those retiring units penalizes the non-retiring units by reducing their allocation in a manner disproportionate to their needs.

Response: EPA determined that in order to fulfill the Wisconsin directive to implement a full remedy, these phased budgets are necessary to ensure an incentive for existing controls to continue to operate. Not including such a mechanism in a full-remedy approach would lead to the possibility highlighted in EPA's proposed rule and some comments, and supported by historical data, where the incentive to operate controls decreases over time with fleet turnover, even though upwind states remain linked to downwind receptors. If EPA did not include such a phase-down mechanism in budgets accounting for fleet turnover, then the other alternative to ensure a full remedy would be unit-specific emission rate requirements (as the only alternative to continue to incentivize existing controls to operate). EPA notes that the some of the commenters who oppose the phase-down mechanism which preserves the trading program's effectiveness across time also support EPA's trading program as the preferred implementation mechanism relative to unit-specific emission rate requirements and even explicitly oppose unit-specific emission rate requirements in some cases. However, the continued reliance on a trading program for full-remedy policy solutions requires this mechanism to ensure the program's effectiveness remains robust in the context of scheduled fleet turnover.

With regard to comments that this approach is different than EPA's past approaches, EPA notes that this approach is not unprecedented or inconsistent with past EPA programs. In the first CSAPR rule, EPA implemented phase 1 and phase 2 NOX budgets for states, which tightened over time even as the rule stringency remained constant for that pollutant. In the CSAPR Update, EPA examined only 2017 for its partial remedy and noted it would revisit future years to see if additional reductions were necessary when implementing a full remedy. This rule achieves that full remedy.

Comment: Some commenters suggested that EPA should assume increased generation from existing units (beyond recent historical data and beyond baseline levels) as some of these units retire, thus offsetting some of the emission reductions.

Response: EPA first notes that it does include emissions and additional generation from additional new sources that are under construction and/or that have received their permit approvals. This new-unit generation offsets the amount of retiring generation in EPA's baseline at the regional level. Second, EPA notes that in both the proposed and the final rule it evaluated the assumed fossil generation from covered sources within its future year baseline (after factoring in retiring fossil generation) relative to historical trends and continues to find that its assumed future level of fossil fuel-fired generation is well within the trend observed over the past four years. In other words, whereas fossil generation from the covered fleet in these 12 states has been declining at approximately 2 percent on average over the past four years, EPA's future year baseline contains fossil generation well within this historical trend (i.e., continued decline at less than 2 percent). Moreover, EPA's assumption that existing, higher-emitting sources will, on average, not raise their generation levels in the future is consistent not only with historical trends, but also with both modeling outlooks for future generation from these EGUs as well as announced plans to replace retiring fossil generation with non-fossil sources. For many of these scheduled retirements, utilities not only have broad plans stating their intention to replace higher-emitting fossil sources with lower emitting sources, but already have those plans for replacement generation, such as renewable technologies, underway.[173 174]

Comment: Some stakeholders note the uncertainty of some scheduled retirements, and the potential for them Start Printed Page 23120to be possibly altered pending information from regulatory entities.

Response: With regard to commenters noting that some retirements are uncertain and therefore should not be factored into EPA future baseline and budget estimates, EPA notes it is using the best available data at the time of the final rule and that no retirement plans included in the final rule were contradicted by commenter data submitted on the proposed rule. EPA relies on a compilation of data from DOE EIA Form 860 where facilities report their future retirement plans and on the information included in its NEEDS database. This information is considered to be highly reliable, real-world information that provides EPA with the high confidence that such retirements will in fact occur. Indeed, in response to commenters' suggestions to factor in yet additional potential retirements, EPA has declined to do so where the intention to retire a unit is not abundantly supported by utility-reported information. Despite this conservative approach to identifying known fleet changes, if a unit's future retirement status ultimately does not materialize on the scheduled date, EPA observes that such an unexpected departure from the currently available evidence would still not contradict its future state-level and region-level estimates. EPA's approach of using historical data and incorporation only of announced fleet changes in estimating its future baseline means that its future year baseline generation and retirement outlook for higher emitting sources is likely conservative, as EPA does not assume any retirements beyond those that are announced. In other words, there are more likely to be additional future EGU retirements that materialize post-rule signature that impact the 2021-2024 timeframe than there are to be announced retirement plans that are subsequently unwound. The analytic tools and information resources used in any estimation of state and regional future EGU emission totals inherently have some discrepancies between what is projected for the future and how the future unfolds—particularly at the unit level. But those potential unit-level discrepancies, inherent in the enterprise of prediction, would at most impact emissions both ways and do not, on their own, undermine EPA's aggregate state and regional estimates. Additionally, as noted elsewhere, EPA's use of a market-based program, a starting bank of converted allowances, availability of additional converted allowances through the “safety valve” mechanism, and variability limits are all features that will readily accommodate whatever small discrepancies there may be between EPA's projection of the EGU fleet and actual fleet conditions in any of the relevant future years. Therefore, EPA's resulting state emission budgets are robust to the inherent uncertainty in future year baseline conditions.

Finally, with regard to comments concerning the impacts of the successive year emissions budget changes' on unit-level allocations for non-retiring units, EPA considers this not to be a budget-setting issue, but rather a question of how to allocate allowances within the budget. Thus, this topic is addressed in section VII.C.3.

EPA's emissions budget methodology and formula for establishing Group 3 budgets are described in detail in the Ozone Transport Policy Analysis Final Rule TSD and summarized below.

For determining emission budgets, EPA generally used historical ozone season data from the 2019 ozone season, the most recent data whose representativeness was not called into question by the unusual circumstances of the Covid-19 pandemic. This is similar to its approach in the CSAPR Update where EPA began with 2015 data (the most recent year at the time). As in the CSAPR Update, EPA combined historical data with IPM data to determine emission budgets. The budget setting process has three primary steps:

(1) Determine a future year baseline—Start with the latest reported historical unit-level data (e.g., 2019), and adjust any unit data where a retirement or new build is known to occur by the baseline year. This results in a future year (e.g., 2021) baseline for emissions budget purposes.[175]

(2) Factor in additional emission controls for the selected control stringency (e.g., $1,800 per ton)—For the unit-level emission control technologies identified in this control stringency, adjust the baseline unit-level emissions and emission rates. For example, if an SCR-controlled unit had a baseline greater than 0.08 lb/mmBtu, its rate and corresponding emissions would be adjusted down to levels reflecting its operation at 0.08 lb/mmBtu.

(3) Incorporate generation shifting—Use IPM in a relative way to capture the reductions expected from generation shifting at a given $ per ton level that reflects control optimization (constrained to within-state shifting).

By using historical unit and state-level NOX emission rates, heat input, and emissions data at step 1 of the budget setting process, EPA is grounding its budgets in the most recent representative historical operation for the covered units.[176] This data set is a reasonable starting point for the budget setting process as it reflects the latest data reported by affected facilities under 40 CFR part 75. The reporting requirements include quality control measures, verification measures, and instrumentation to best record and report the data. In addition, the designated representatives of EGU sources are required to attest to the accuracy and completeness of the data. In step 1 of the budget setting process, EPA first adjusted the 2019 ozone-season data to reflect committed fleet changes under a baseline scenario (i.e., announced and confirmed retirements, new builds, and retrofits that have already occurred). For example, if a unit emitted in 2019, but retired in 2020, its 2019 emissions would not be included in the 2021 estimate. For units that had no known changes, the 2021 emissions assumption was the actual reported data from 2019 at this first step of adjusting the baseline. EPA also included known new units and scheduled retrofits in this manner. Using this method, EPA arrived at a baseline emission, heat input, and emission rate estimate for each unit for a future year (e.g., 2021), and then was able to aggregate those unit-level estimates to state-level totals. These state-level totals constituted the state's baseline from an engineering analytics perspective. The ozone-season state-level emissions, heat input, and emissions rates for covered sources under a baseline scenario were determined for each future year examined (2021 through 2024). Because 2024 is the last ozone season for which EPA projects continued contribution to any downwind receptors, 2024 is the last year for which EPA is making an adjustment to emission budgets.

For step 2 of the emissions budget setting process, EPA examined how the baseline emissions and emission rates would change under different control stringencies for EGUs. For instance, under the SCR optimization scenario, if a unit was not operating its SCR at 0.08 lb/mmBtu or lower in the baseline, EPA Start Printed Page 23121lowered that unit's assumed emission rate to 0.08 lb/mmBtu and calculated the impact on the unit's and state's emission rate and emissions. Note, the heat input is held constant for the unit in the process, reflecting the same level of unit operation compared to historical 2019 data. An improved emission rate is then applied to this heat input, reflecting control optimization. In this manner, the state-level baseline totals from step 1 reflecting known baseline changes were adjusted to reflect the additional application of the assumed control technology at a given control stringency.

Finally, at step 3 of the emissions budget setting process, EPA used IPM to capture any generation shifting at a given control stringency necessary for the majority of the respective emission control technology to operate. EPA explains how it accounts for generation shifting in more detail in in section VI.B and in the Ozone Transport Policy Analysis Final Rule TSD. In this rule, as a proxy for the near-term reductions required by 2021, EPA has constrained generation shifting to occur only within-state.

EPA requested comment on the approach described above, as well as alternatives discussed in the budget-setting TSD. Specifically, EPA requested comment on its consideration of using 2020 data in place of 2019 data as the most recent historical data set to inform final rule budgets. Although the reduction potential associated with the selected control stringency described in section VI would likely not change substantially with that data set, the baseline values calculated in step 1 of the emissions budget setting process may change significantly and possibly result in lower or higher state-level emission budgets.

Comment: EPA received comment highlighting the unique impact of the Covid-19 pandemic on 2020 emissions and generation data due to changes in market conditions that may not be representative in subsequent years (e.g., changes in net generation, time-of-day impacts on demand, and natural gas prices). Commenters cautioned against relying on 2020 data for informing step 3 analysis in this rule.

Response: EPA is finalizing, as proposed and consistent with these comments, the continued use of 2019 EGU data as the latest, most representative historical year for informing the Agency's step 3 analysis. EPA examined the unique Covid-related impacts on the power sector and energy market data. It observed significant changes for some variables where the change appeared to be specific to the 2020 dataset and pandemic-related conditions, and therefore not representative of future power sector operations or market conditions. These included changes in natural gas prices, the demand profiles for electricity (which influence what units generate at different parts of the day), and overall electricity demand. This was further borne out by comparing quarterly year-over-year data which revealed that changes in Q4 2020 data relative to Q4 2019 data were not as pronounced as changes in Q2 2020 data relative to Q2 2019 data, indicating the temporary status of some of changes observed in the 2020 ozone season. For instance, Q2 2020 NOX emissions were down 20 percent year-over-year, but Q4 2020 NOX emissions were down only 9 percent year-over-year. EPA provides additional detail in the RTC document on its consideration of 2019 and 2020 data as the most recent historical representative year of the power sector. Had EPA utilized 2020 data as the starting point for its future year baseline in Engineering Analytics, it likely would have been incorporating some 2020-fleet operational changes (and corresponding emission levels) unique to the pandemic year instead of fleet changes expected to endure into post-2020 years. As also explained in the RTC document, while EPA did continue to use 2019 as the starting historical data set, it recognized commenters' observations that New York and Virginia were differently situated in that their emissions were higher in 2020 than 2019 (whereas all other states were lower, at least partially attributable to Covid impacts). Additionally, reflecting the 2020 fleet dynamics in the future year baseline for New York helps capture some of the dynamics related to the retirement of one unit at the Indian Point nuclear facility as pointed out by the commenter. To account for these atypical circumstances, EPA incorporated upward adjustments to its future year baseline values for New York and Virginia that reflected the incremental changes in heat input, generation, and emissions for 2020 relative to 2019.

Comment: Some commenters suggested EPA use a multi-year historical baseline for its step 3 analysis on the theory that this would provide a more robust set of historical data and a more representative baseline for the power sector.

Response: EPA is finalizing use of the same single-year historical baseline approach it used in the proposed rule. This approach is similar to the CSAPR Update, where EPA also relied on a single-year historical baseline to inform its step 3 approach. EPA's interest in a historical data set to inform this part of the analysis is to capture the current status of the power sector (i.e., incorporating the latest new builds, retirements, and unit operation in response to current regulations and market conditions). Incorporating prior years through a multi-year historical baseline would dilute, rather than strengthen, the methodology's ability to capture the most representative perspective of the current power sector. It would in effect include units that no longer exist, market conditions that have since evolved, and a regulatory landscape that has likewise since changed. It would diminish the effect of newer generation resources that have come online which reflect the impacts of the latest changes in technology performance and cost levels. EPA finds that, particularly at the state and regional level, the most recent year data is a better representation and basis for future year baselines rather than incorporating older data. In other applications, where the purpose is not forward looking, but rather distribution-based and unit-level focused, lengthier historical baselines have more value. See additional response to this comment in the State Emission Budgets section of the RTC document.

C. Elements of New Trading Program

To implement the updated emissions budgets developed according to the process described in section VII.B, EPA is requiring EGUs in each of the 12 covered states to participate in a new CSAPR NOX Ozone Season Group 3 Trading Program. The provisions of the new “Group 3” trading program are largely identical to the provisions of the “Group 2” trading program in which affected EGUs in the 12 covered states participated from 2017 through 2020. The principal differences between the Group 2 and Group 3 trading programs are the differences in state budgets and geography established in this rule to address the covered states' remaining obligations under CAA section 110(a)(2)(D)(i)(I) with respect to the 2008 ozone NAAQS. One other difference, which EPA is adopting in response to comments, concerns the determination of which units are eligible to receive allocations of allowances for use in the new Group 3 trading program as “existing units” under EPA's default allocation methodology. Specifically, certain units with scheduled future retirement dates will not receive allocations as existing units for use in the Group 3 trading program starting with the first control period for which the units' scheduled Start Printed Page 23122retirements are reflected in adjustments to the state emission budgets. This aspect of implementation of the Group 3 trading program is discussed in section VII.C.3.b.

The proposed rule included several provisions designed to address the transition from the Group 2 trading program to the Group 3 trading program. The provisions for allocation of supplemental allowances to ensure that the enhanced control stringency established in this action applies only after the rule's effective date are finalized as proposed. The provisions concerning creation of an initial bank of Group 3 allowances in exchange for banked 2017-2020 Group 2 allowances at a formula-based conversion ratio and the provisions concerning the recall of certain previously recorded 2021-2024 Group 2 allowances are finalized with certain modifications adopted after consideration of comments. Also, in response to comments, the final rule includes transitional provisions establishing a “safety valve” mechanism under which sources may obtain additional Group 3 allowances in exchange for additional 2017-2020 Group 2 allowances at a higher conversion ratio. All of these transitional provisions are discussed in section VII.C.4.

The only other differences between the new Group 3 trading program regulations and the Group 2 trading program regulations that applied for emissions through the 2020 control periods are a small number of corrections and administrative simplifications that have no effect on program stringency; EPA is eliminating these differences by making the same corrections and simplifications to the regulations for the Group 2 trading program and the other existing CSAPR trading programs starting with the 2021 control periods.[177] In this section, the Agency discusses major elements of the new Group 3 trading program, with emphasis on the elements that differ from the previous provisions of the Group 2 trading program as well as the provisions specifically designed to address the transition from the Group 2 trading program to the Group 3 trading program.

1. Applicability

In this rule, EPA is using the same EGU applicability provisions in the new Group 3 trading program as in the existing Group 2 trading program and the other CSAPR trading programs, without change. Under the general CSAPR applicability provisions, a covered unit is any stationary fossil-fuel-fired boiler or combustion turbine serving at any time on or after January 1, 2005, a generator with nameplate capacity exceeding 25 MW, which is producing electricity for sale, with the exception of certain cogeneration units and solid waste incineration units.

2. State Budgets, Variability Limits, Assurance Levels, and Penalties

EPA is establishing revised state budgets for EGU emissions of ozone season NOX for the 12 “Group 3” states subject to new or amended FIPs in this final rule in order to fully address these states' significant contribution with respect to the 2008 ozone NAAQS. The budgets have been established according to the process described in section VII.B. As discussed in that section, for each of the covered states, separate budgets are established for the three individual years 2021, 2022, and 2023, and then for 2024 and beyond.[178] Portions of the updated NOX ozone season emission budgets are reserved as updated new unit set-asides and Indian country new unit set-asides for the same control periods, as further described in section VII.C.3.a. The amounts of the state emissions budgets for 2021, 2022, 2023, and 2024 and beyond are shown in tables VII.C.2-1, VII.C.2-2, VII.C.2-3, and VII.C.2-4.

Similar to the previous requirements to hold Group 2 allowances sufficient to cover their NOX emissions in each control period from 2017 through 2020, sources in states covered by the new Group 3 trading program will be required to hold new Group 3 allowances sufficient to cover their NOX emissions in each control period in 2021 and thereafter. For Group 3 states that were found in the CSAPR Update to still have good neighbor obligations with respect to the 1997 ozone NAAQS, EPA is determining that participation of the state's EGUs in the more stringent Group 3 trading program will satisfy those obligations.[179]

In the CSAPR and the CSAPR Update, EPA developed assurance provisions, including variability limits and assurance levels (with associated compliance penalties), to ensure that each state will meet its pollution control and emission reduction obligations and to accommodate inherent year-to-year variability in state-level EGU operations. Establishing assurance levels with compliance penalties responds to the D.C. Circuit's holding in North Carolina requiring EPA to ensure within the context of an interstate trading program that sources in each state are required to eliminate emissions that significantly contribute to nonattainment or interfere with maintenance of the NAAQS in another state.[180]

Like the emission budgets promulgated in the CSAPR and the CSAPR Update, the revised emission budgets promulgated in this rule reflect EGU operations in an “average year.” However, year-to-year variability in EGU operations occurs due to the interconnected nature of the power sector, changing weather patterns, changes in electricity demand, or disruptions in electricity supply from other units or from the transmission grid. Recognizing this, the trading program provisions finalized in the CSAPR and CSAPR Update rulemakings include variability limits, which define the amount by which an individual state's emissions may exceed the level of its budget in a given year to account for variability in EGU operations. A state's budget plus its variability limit equals the state's assurance level, which acts as a cap on the state's NOX emissions during a given control period (in this rulemaking, the relevant control period is the May-September ozone season). The new CSAPR NOX Ozone Season Group 3 Trading Program provisions established for affected sources in the 12 states subject to the new trading program under this final rule contain equivalent assurance provisions to the prior CSAPR and CSAPR Update trading programs.

The variability limits ensure that the trading program can accommodate the inherent variability in the power sector while ensuring that each state eliminates the amount of emissions within the state, in a given control period, that must be eliminated to meet Start Printed Page 23123the statutory mandate of CAA section 110(a)(2)(D)(i)(I). Moreover, the structure of the trading program, which achieves required emission reductions through limits on the total numbers of allowances allocated, assurance provisions, and penalty mechanisms, ensures that the variability limits only allow the amount of temporal and geographic shifting of emissions that is likely to result from the inherent variability in power generation, and not from decisions to avoid or delay the optimization or installation of necessary controls.

To establish the variability limits in the CSAPR, EPA analyzed historical state-level heat input variability as a proxy for emissions variability, assuming constant emission rates. See 76 FR 48265. The variability limits for ozone season NOX in both the CSAPR and the CSAPR Update were calculated as 21 percent of each state's budget, and these variability limits for the NOX ozone season trading programs were then codified in 40 CFR 97.510 and 40 CFR 97.810, along with the respective state budgets. For this final rule, EPA is retaining variability limits for the 12 Group 3 states covered by this rule calculated as 21 percent of each state's revised budget.[181]

Table VII.C.2-1—CSAPR NOX Ozone Season Group 3 State Budgets, Variability Limits, and Assurance Levels for 2021 182

StateEmissions budget (tons)Variability limit (tons)Assurance level (tons)
Illinois9,1021,91111,013
Indiana13,0512,74115,792
Kentucky15,3003,21318,513
Louisiana14,8183,11217,930
Maryland1,4993151,814
Michigan12,7272,67315,400
New Jersey1,2532631,516
New York3,4167174,133
Ohio9,6902,03511,725
Pennsylvania8,3791,76010,139
Virginia4,5169485,464
West Virginia13,3342,80016,134

Table VII.C.2-2—CSAPR NOX Ozone Season Group 3 State Budgets, Variability Limits, and Assurance Levels for 2022

StateEmissions budget (tons)Variability limit (tons)Assurance level (tons)
Illinois9,1021,91111,013
Indiana12,5822,64215,224
Kentucky14,0512,95117,002
Louisiana14,8183,11217,930
Maryland1,2662661,532
Michigan12,2902,58114,871
New Jersey1,2532631,516
New York3,4167174,133
Ohio9,7732,05211,825
Pennsylvania8,3731,75810,131
Virginia3,8978184,715
West Virginia12,8842,70615,590

Table VII.C.2-3—CSAPR NOX Ozone Season Group 3 State Budgets, Variability Limits, and Assurance Levels for 2023

StateEmissions budget (tons)Variability limit (tons)Assurance level (tons)
Illinois8,1791,7189,897
Indiana12,5532,63615,189
Kentucky14,0512,95117,002
Louisiana14,8183,11217,930
Start Printed Page 23124
Maryland1,2662661,532
Michigan9,9752,09512,070
New Jersey1,2532631,516
New York3,4217184,139
Ohio9,7732,05211,825
Pennsylvania8,3731,75810,131
Virginia3,9808364,816
West Virginia12,8842,70615,590

Table VII.C.2-4—CSAPR NOX Ozone Season Group 3 State Budgets, Variability Limits, and Assurance Levels for 2024 and Beyond

StateEmissions budget (tons)Variability limit (tons)Assurance level (tons)
Illinois8,0591,6929,751
Indiana9,5642,00811,572
Kentucky14,0512,95117,002
Louisiana14,8183,11217,930
Maryland1,3482831,631
Michigan9,7862,05511,841
New Jersey1,2532631,516
New York3,4037154,118
Ohio9,7732,05211,825
Pennsylvania8,3731,75810,131
Virginia3,6637694,432
West Virginia12,8842,70615,590

The assurance provisions include penalties that are triggered in the event that the covered sources' emissions in a given state, as a whole, exceed the state's assurance level. The CSAPR and the CSAPR Update provided that, when the emissions from EGUs in a state exceed that state's assurance level in a given year, particular sources within that state will be assessed a 3-to-1 allowance surrender on emissions exceeding the assurance level. Specifically, each excess ton above a given state's assurance level must be met with one allowance, per standard compliance, and two additional allowances to satisfy the penalty. The penalty was designed to deter state-level emissions from exceeding assurance levels. In both the CSAPR and the CSAPR Update, the assurance provisions were designed to account for variability in the electricity sector while ensuring that the necessary emission reductions occur within each covered state, consistent with the court's holding in North Carolina, 531 F.3d at 908. If EGU emissions in a given state do not exceed that state's assurance level, no penalties are incurred by any source.

To assess the penalty under the assurance provisions, EPA is following the same methodology finalized in the CSAPR Update. See 81 FR 74567. In that methodology, EPA evaluates whether any state's total EGU emissions in a control period exceeded the state's assurance level, and if so, EPA then determines which groups of units in the state represented by a “common designated representative” emitted in excess of the common designated representative's share of the state assurance level and, therefore, will be subject to the allowance surrender requirement described above. Thus, penalties under the assurance provisions are triggered for the group of sources represented by a common designated representative when two conditions are met: (1) The group of sources and units with a common designated representative are located in a state where the total state EGU emissions for a control period exceed the state assurance level; and (2) that group with the common designated representative had emissions exceeding the respective common designated representative's share of the state assurance level. EPA is establishing assurance provisions for the CSAPR NOX Ozone Season Group 3 Trading Program that are equivalent to the assurance provisions in the CSAPR NOX Ozone Season Group 2 Trading Program.

In this final rule, EPA is simplifying the procedures for administering the assurance provisions, as compared to the analogous provisions included in the existing CSAPR trading programs for control periods before 2021.[183] The simplifications are made possible by the revisions to the process for allocating allowances from the new unit set-asides that are discussed in section VII.C.3.c. The same simplifications are also being implemented in the existing CSAPR trading programs, as discussed in section VII.C.8. These simplifications concern the procedures for determining the portion of the state's assurance level to be assigned to each common designated representative. Specifically, certain provisions of these procedures as previously implemented in the existing CSAPR trading programs were designed to address circumstances where a new unit operates but has no allowance allocation determined for it. Administration of these provisions Start Printed Page 23125required EPA to issue a notice to collect information needed solely for this purpose that is not otherwise required to be reported to EPA. Because the revised new unit set-aside (“NUSA”) allocation procedures eliminate the possibility that a new unit would not have an allowance allocation determined for it, EPA is eliminating the provisions for issuance of the related extra notice. EPA also is extending the date as of which a common designated representative is determined under both the new Group 3 trading program and the existing CSAPR trading programs from April 1 of the year following the control period to July 1 so as to preserve the relationship of those dates to the allowance transfer deadline, which is being extended from March 1 of the year following the control period to June 1.[184] Further discussion of these changes from the current provisions in the existing trading programs is provided in section VII.C.8.

Comment: EPA received several comments concerning the achievability of state emissions budgets in 2021 that highlighted the quick implementation timeframe and suggesting that such a timeframe would not allow enough times for a liquid allowance market to form and thus inhibit sources' ability to obtain the allowances that they need for compliance.

Response: As an initial matter, EPA observes that in 25 years of promulgating and administering trading programs for NOX and SO2 as mechanisms to address acid precipitation or interstate transport of air pollution, the Agency has never encountered a single instance where a source was unable to comply with the requirements of any of these trading programs because of an inability to find allowances available for purchase. Almost all of the sources that will participate in the trading program established under this final rule have previously participated in some of these other trading programs and therefore are, or should be, fully aware that under every such trading program, a functioning allowance market has developed. Nevertheless, some commenters assert that in the trading program established under this specific final rule—where the emission reductions required for the first control period are set at levels designed to be achievable without installation of any new controls by any source—for the first time ever, either no industry participants will be willing to take actions that would create surplus allowances or all industry participants will refuse to sell such surplus allowances at any price. The comments fly in the face of 25 years of evidence and common experience, not to mention principles of economics and market participants' self-interest. EPA views the comments as unsupported and speculative to the point of irrationality.

EPA first addresses the viability of 2021 implementation for the emission reductions required under this rule in detail in section VI. B above. With regard to the specific market liquidity concerns expressed here, EPA notes that those same concerns have been voiced in the lead-up to past CSAPR trading programs and have never materialized. Instead, a functioning allowance market has always formed and resulted in 100 percent compliance with the allowance holding requirements for the first control period (and subsequent control periods). As described in this section, under the new trading program sources are not required to hold (and subsequently surrender) any allowances for compliance purposes until June 1, 2022—well after the end of the 2021 ozone season. In the current CSAPR Update ozone season programs, EPA observes that most trades occur near or after end of the ozone season. Therefore, the approximately two months between final rule promulgation and the start of the compliance period is in no way a limit on the time sources have to buy and sell allowances for that compliance period. Rather, sources will have eight months after the end of the control period in which to engage in any necessary or desired allowance market transactions. The total quantity of allowances usable for the 2021 control periods from state emission budgets and from the initial Group 3 bank (discussed in section VII.C.4.b) will be known before the start of the 2021 control period, and EPA expects that almost all such allowances will be recorded in sources' compliance accounts well before the end of the 2021 control period, ensuring that there will be no logistical impediments to such transactions. Moreover, in many cases, units that may have the need to procure allowances will also have associated units under common ownership elsewhere in the fleet that hold a surplus of allowances In this case, it is only a matter of intra-owner allowance movement needed to align allowancing holdings with allowance surrender obligations, and the need for accessing a broader allowance market is mooted.

Further, the level of the budgets, in addition to the initial Group 3 bank, should obviate any market liquidity concerns as the number of allowances on the market for the first year will accommodate a variety of compliance pathways and unit operational decisions. Moreover, the experience of the CSAPR programs reveals that the allowance price is highest in the first compliance period, creating an incentive for all sources to implement achievable emission reductions and for sources with surplus of allowances to sell them while allowance prices are highest, generating the conditions for a robust market to form—further promoting market liquidity. While EPA strongly disagrees, based on previous program implementation and forward-looking analysis, that there is any risk of market illiquidity, the Agency is creating an additional “safety valve” in this final rule due to the near-term implementation timetable. Consistent with commenters' suggestions, EPA will allow the one-time conversion of Group 2 allowances at an 18:1 ratio to provide additional assurance to sources that allowances will be available, but ensuring that the cost of this compliance option is such that entities will utilize it only in the very unlikely event that access to such additional allowances proves to be necessary. The safety valve is described further in section VII.C.4.c.

Comment: EPA received several comments concerning the proposed variability limits and associated assurance levels for the states in the Group 3 trading program. Some commenters suggested that EPA should eliminate or tighten variability limits for the Group 3 trading program. One commenter justified these changes by observing that the Group 2 trading program established under the CSAPR Update had excess availability of allowances and low allowance prices.

One commenter suggested that EPA eliminate variability limits for the Group 3 trading program on the basis that the variability limits and associated assurance levels as proposed do not result in the elimination of downwind non-attainment by the end of 2021. This commenter stated that EPA failed to provide a full explanation in the proposed action as to why the 21 percent variability limit used in the trading programs for ozone season NOX established in the CSAPR and the CSAPR Update was still applicable in the new Group 3 trading program. The commenter stated there is no justification for EPA to increase the budget amounts due to variability in Start Printed Page 23126EGU fleet operation if EPA is correct in its assessment that the proposed NOX mass emission budget levels are representative of near-term achievable NOX emission control obligations based on historical EGU fleet operation. The commenter stated that increasing budgets by 21 percent to arrive at an assurance level permits an upwind state's EGU fleet to emit NOX mass emissions more than the levels necessary to meet the given state's obligation to downwind areas. The commenter further states a belief that EPA is misapplying the concept of EGU fleet operational variability to permit a state's EGU fleet to emit NOX mass emissions at levels that may negatively impact the health and welfare of downwind populations.

Response: EPA disagrees with the commenters and is retaining the variability limits and associated assurance levels as reflected in both the CSAPR and the CSAPR Update. EPA believes a variability limit of 21 percent continues to be appropriate for states in the Group 3 trading program. The assertion that state budgets are increased by 21 percent in response to the variability limit is incorrect. Rather, as described in the CSAPR, the CSAPR Update, and reiterated in this final action, the variability limits reflect expected year-to-year or season-to-season variability in demand for electricity, and therefore, variability in the use of fuel and in emissions. While a given state may emit up to the assurance level (i.e., that state's budget plus the 21 percent variability limit) during years with adverse meteorology and atypical levels of electricity demand, allowances banked from prior control periods may then be used for compliance obligations. However, the total number of allowances issued for each control period in the Group 3 trading program is equal to the sum of the Group 3 states' emission budgets, not the sum of the Group 3 states' assurance levels. Although EPA is also creating an initial bank of allowances in an amount equal to the sum of the states' variability limits for the 2022 control period (see section VII.C.4.b), creation of the bank is a one-time event and does not represent a 21 percent increase in the state emission budgets established for each control period.

With regard to the comment that EPA has not sufficiently justified reusing in the Group 3 trading program the same 21 percent variability limits used in the trading programs for ozone season NOX established in the CSAPR and the CSAPR Update, EPA disagrees that updating these limits is necessary. The original variability analysis performed in the CSAPR rulemaking considered data for 26 states (including all 12 Group 3 states) and reflected over a decade of operational variability (from 2000 through 2010), producing relatively robust standard deviation estimates. EPA would not necessarily view changes of a few percent above or below the previously identified variability level of 21 percent from an updated analysis as significant enough to require establishment of different variability limits in the Group 3 trading program. Nevertheless, in response to the comment, EPA has performed an updated variability analysis for the 12 Group 3 states reflecting data for all control periods from 2000 to 2019. The updated analysis again results in a variability estimate of 21 percent. EPA also considered shorter time periods for the updated analysis and found that the resulting variability estimates are not especially sensitive to the particular time period analyzed. Accordingly, EPA concludes that it is reasonable to finalize the variability limits for the Group 3 trading program at the level of 21 percent as proposed.[185]

3. Unit-Level Allocations of Emissions Allowances

For states participating in the CSAPR Group 3 trading program, EPA is issuing CSAPR NOX Ozone Season Group 3 allowances to be used for compliance beginning with the 2021 ozone season. This section explains the default process by which EPA is allocating total amounts of these allowances equal to each state's budget amount existing units and new units in the state. Section VII.C.3.a describes the determination of the portions of each state's budget that will be set aside for potential allocation to new units in the state and in any Indian country within the state's borders. Section VII.C.3.b discusses the methodology used to allocate shares of each state's budget not reserved in a set-aside to the existing units in the state, including in some cases to units that have ceased operations. Sections VII.C.3.c and VII.C.3.d discuss the process for allocating the allowances in the new unit set-asides and Indian country new unit set-asides, respectively, to individual units.

As under both the CSAPR and the CSAPR Update, states have several options under this final rulemaking to submit SIP revisions which, if approved, may result in the replacement of EPA's default allocations with state-determined allocations for the 2022 control period and beyond. The provisions described in this section do not prevent any state from employing an alternative allocation methodology for control periods after 2021 through a SIP submission. See section VII.D. for details on the development of approvable SIP submissions.

a. Set-Asides of Portions of State Budgets for New Units

As part of the default allocation process that will apply where a state does not employ an alternative allocation process pursuant to an approved SIP revision, EPA is promulgating allocations to a new unit set-aside for each state equal to a minimum of 2 percent of the total state budget, plus the projected amount of emissions from planned units in that state. For example, if planned units in a state are projected to emit 3 percent of the state's NOX ozone season emission budget, then the new unit set-aside for the state would be set at 5 percent, which is the sum of the minimum 2 percent set-aside plus an additional 3 percent for planned units. As further discussed in section VII.C.3.d., for the three Group 3 states with Indian country within their borders (Louisiana, Michigan, and New York), EPA is reserving 5 percent of the minimum 2 percent new unit set-aside, or 0.1 percent of the total state budget, for any new units in Indian country within the borders of state,[186] with no additional amount to address planned units in Indian country.[187] This is the same approach previously used to establish the amounts of new unit set-asides and Indian country new unit set-asides for all the CSAPR and CSAPR Update trading programs. See, e.g., 76 FR 48292 (August 8, 2011). Note that New York has set its NUSA percentage within its approved SIP for the existing Group 2 trading program to 5 percent of the state emission budget without consideration of planned units; therefore, this NUSA percentage is used Start Printed Page 23127for New York. As described in greater detail in sections VII.C.3.c and VII.C.3.d, new units are eligible to receive allocations from a new unit set-aside or Indian country new unit set-aside starting with the first year they are subject to the allowance-holding requirements of this rule. If the allowances in the NUSA for a state or the Indian country NUSA for Indian country within the borders of a state are not allocated to new units, the allowances are redistributed to existing units in the state before each compliance deadline.

The process described above for determining the portions of each state budget that will be set aside for potential allocation to new units is unchanged from the process described in the proposed rule. EPA received no comments concerning the portions of the emission budgets established under the new Group 3 trading program that would be set aside for this purpose. One commenter suggested that the amounts of the new unit set-asides should be increased by adding allowances from the existing Group 2 trading program that would have been allocated to retired units under that program. EPA is not implementing this suggestion and responds more fully to the comment in section VII.C.4.b.

Because the budgets under the Group 3 trading program vary across control periods, the amounts of the default new unit set-asides and Indian country new unit set-asides also vary. The amounts for each state for 2021 through 2023 and for 2024 and beyond are set forth in tables VII.C.3-1 through VII.C.3-4.[188]

Table VII.C.3-1—CSAPR NOX Ozone Season Group 3 New Unit Set-Aside (NUSA) Amounts for 2021

StateEmission budgets (tons)New unit set-aside amount (percent)Total new unit set-aside amount for new units (tons)New unit set-aside amount for new units not in Indian country (tons)Indian country new unit set-aside amount (tons)
Illinois9,1023265265
Indiana13,0512262262
Kentucky15,3002309309
Louisiana14,818344543015
Maryland1,4999135135
Michigan12,727451350013
New Jersey1,25322727
New York3,41651711683
Ohio9,6903291291
Pennsylvania8,3794335335
Virginia4,5164185185
West Virginia13,3342266266

Table VII.C.3-2—CSAPR NOX Ozone Season Group 3 New Unit Set-Aside (NUSA) Amounts for 2022

StateEmission budgets (tons)New unit set-aside amount (percent)Total new unit set-aside amount for new units (tons)New unit set-aside amount for new units not in Indian country (tons)Indian country new unit set-aside amount (tons)
Illinois9,1023265265
Indiana12,5822254254
Kentucky14,0512283283
Louisiana14,818344543015
Maryland1,2669115115
Michigan12,290449448212
New Jersey1,25322727
New York3,41651711683
Ohio9,7733290290
Pennsylvania8,3734339339
Virginia3,8974161161
West Virginia12,8842261261
Start Printed Page 23128

Table VII.C.3-3—CSAPR NOX Ozone Season Group 3 New Unit Set-Aside (NUSA) Amounts for 2023

StateEmission budgets (tons)New unit set-aside amount (percent)Total new unit set-aside amount for new units (tons)New unit set-aside amount for new units not in Indian country (tons)Indian country new unit set-aside amount (tons)
Illinois8,1793248248
Indiana12,5532249249
Kentucky14,0512283283
Louisiana14,818344543015
Maryland1,2669115115
Michigan9,975439838810
New Jersey1,25322727
New York3,42151711683
Ohio9,7733290290
Pennsylvania8,3734339339
Virginia3,9804166166
West Virginia12,8842261261

Table VII.C.3-4—CSAPR NOX Ozone Season Group 3 New Unit Set-Aside (NUSA) Amounts for 2024 and Beyond

StateEmission budgets (tons)New unit set-aside amount (percent)Total new unit set-aside amount for new units (tons)New unit set-aside amount for new units not in Indian country (tons)Indian country new unit set-aside amount (tons)
Illinois8,0593244244
Indiana9,5642190190
Kentucky14,0512283283
Louisiana14,818344543015
Maryland1,3489122122
Michigan9,786439238210
New Jersey1,25322727
New York3,40351701673
Ohio9,7733290290
Pennsylvania8,3734339339
Virginia3,6634150150
West Virginia12,8842261261

b. Allocations to Existing Units, Including Units That Cease Operation

The portion of a state budget remaining after the portions reserved for new units have been set aside is allocated among the existing units in the state. EPA in this action is generally allocating allowances to existing units in the Group 3 states following the same methodology for allowance allocation that was used in the CSAPR Update, which relies on historical heat input data and historical emissions data for each eligible existing unit in the state. See 81 FR 74564-65. For the new Group 3 trading program, EPA is applying this methodology using historical data through 2019. In response to comments, EPA is also making one change to the approach used to determine which existing units are eligible to receive allocations for a given control period, specifically by excluding certain units with scheduled future retirements from receiving allocations for control periods after the years in which the scheduled retirements are reflected in adjustments to the respective states' emission budgets. For the amounts of the allocations to existing units, see the TSD “Unit Level Allocations and Underlying Data for the Revised CSAPR Update for the 2008 Ozone NAAQS,” in the docket for this final rule. Note that this final rule addresses allocations of only the newly created CSAPR NOX Ozone Season Group 3 allowances issued under and used for compliance in the Group 3 trading program. EPA is not changing allocations of allowances used in the CSAPR NOX Ozone Season Group 1 or Group 2, NOX Annual, or SO2 Group 1 or Group 2 trading programs and is not reopening the previously established default allocations under these programs.

For the purpose of allocations, the CSAPR considered an “existing unit” to be a unit that commenced commercial operation prior to January 1, 2010, and the CSAPR Update considered an “existing unit” to be a unit that commenced commercial operation prior to January 1, 2015. For the 12 states subject to new or amended FIPs in this rulemaking, EPA is considering an “existing unit” for purposes of the Group 3 trading program to be a unit that commenced commercial operation prior to January 1, 2019 (although only existing units that did not cease operation before January 1, 2021 will be eligible to receive allocations of Group 3 allowances as existing units). This change will allow units commencing commercial operation between 2015 and 2019 to be directly allocated allowances from each state's budget as existing units and will allow the full amounts of the new unit set-asides and Indian country new unit set-asides to be available for any future new units Start Printed Page 23129locating in covered states or Indian country. Using data available at the time of the proposed rule's development, EPA identified which units in the proposed Group 3 states that currently submit quarterly emissions reports to EPA appear to be eligible or ineligible to receive allowance allocations as existing units; [189] for this final rule, EPA has updated the lists of units with the most recent data. EPA is not reconsidering which units are “existing units” for purposes of any other CSAPR trading program. Sources in most of the Group 3 states also participate in the CSAPR NOX Annual and SO2 Group 1 trading programs, for which an “existing unit” is a unit that commenced commercial operation before January 1, 2010. Thus, a unit that is located in one of these states and that commenced commercial operation between January 1, 2010, and January 1, 2019, would be considered an “existing unit” for purposes of default allowance allocations under the Group 3 trading program but would continue to be considered a “new unit” for purposes of default allowance allocations under the CSAPR NOX Annual and SO2 Group 1 trading programs.

As noted earlier in this section, in response to comments EPA is finalizing a change from the allocation methodology used in the existing CSAPR trading programs with respect to which existing units are eligible to receive allocations from the budget for a given control period following retirement. Specifically, in cases where, before finalization of this rule, a unit was scheduled to retire with sufficient certainty for the retirement to be taken into account in EPA's process in this rule for setting the emission budgets for the state where the unit is located, EPA is not providing allocations of allowances to the unit as an existing unit from the budget for any control period starting with the first control period for which the state's emission budget has been adjusted to reflect the unit's scheduled retirement. This approach to determining eligibility to receive allocations as an existing unit does not apply to other units that may cease operations but whose upcoming retirements were not scheduled as of finalization of this action with sufficient certainty to be reflected in the process for setting the emission budgets. These other units would continue to receive allowance allocations as existing units for five control periods of non-operation, consistent with the allocation methodology used in the existing CSAPR trading programs. EPA provides additional discussion of these aspects of the allocation methodology in the responses to comments at the end of this section. The criteria that EPA has applied to determine whether a unit's scheduled retirement is sufficiently certain to serve as a basis for adjusting emission budgets and unit-level allocations are discussed in section VII.B and in the Ozone Transport Policy Analysis Final Rule TSD.

EPA is applying the default methodology finalized in the CSAPR Update for allocating emission allowances to existing units, updated to use more recent historical data. This methodology allocates allowances to each unit based on the unit's share of the state's heat input, limited by the unit's maximum historical emissions. As discussed in the CSAPR Update, see 81 FR 74563-65, EPA finds this allowance allocation approach to be fuel-neutral, control-neutral, transparent, based on reliable data, and similar to allocation methodologies previously used in the CSAPR, the NOX SIP Call, and the Acid Rain Program.[190] EPA is therefore continuing the application of this default methodology for allocating allowances to existing sources in this final rule.

This final rule uses the average of the three highest years of heat input data out of the most recent five-year period that is considered representative to establish the heat input baseline for each unit.[191] These heat input data are used to calculate each unit's proportion of state-level heat input (the average of the unit's three highest non-zero years of heat input divided by the total of such averages for all eligible units within the given state). In general, EPA applies this proportion to the total amount of existing unit allowances to be allocated to quantify unit-level allocations. However, EPA also constrains each unit's allocation so as not to exceed the unit's maximum historical baseline emissions, calculated as the highest year of emissions out of the most recent eight-year period that is considered representative.[192] In other words, if the allocation that a unit would receive from the emission budget for its state based solely on consideration of the unit's share of the state-level heat input exceeds that unit's maximum historical baseline emissions, the unit's allocation is capped at its maximum historical baseline emissions and the excess allowances are instead allocated to other units in the state whose allocations do not exceed their respective maximum historical baseline emissions, again in proportion to those other units' shares of the state-level heat input. Like the proposed rule, this final rulemaking uses 2015-2019 heat input data and 2012-2019 emissions data for purposes of computing unit-level allocations. Although EPA proposed to update the data used in this action to include 2016-2020 heat input data and 2013-2020 emission data, most comments received on this topic opposed the use of 2020 data as potentially unrepresentative because of changes in economic conditions related to the COVID-19 pandemic. EPA is persuaded that in the unusual circumstance of the pandemic, 2020 data have the potential to be less representative for at least some units than data from earlier control periods, and accordingly EPA is not updating the periods of the data used in the allocation calculations for the final rule to include 2020 data.

Under the CSAPR Update, if, at the time the rule was finalized, a state had already submitted a SIP revision addressing the allocation of CSAPR NOX ozone season allowances among the units in the state, and if the SIP submission's allocation provisions could be applied to an updated budget, EPA applied the state's preferred allocation methodology to determine the allocation of allowances among that state's units under the final CSAPR Update. Two of the Group 3 states (Indiana and New York) have such methodologies for allocating CSAPR NOX Ozone Season Group 2 allowances among their units. As under the CSAPR Update, in this final rule EPA is carrying out the intent of these SIPs by establishing initial allowance allocations to existing units under the FIPs for these two states using the allocation methodologies already adopted by the states. EPA received no comments opposing this approach to establishing the default allocation methodologies for these states. EPA Start Printed Page 23130notes that, consistent with the approach taken for other states, when applying these states' allocation methodologies, the set of units deemed eligible to receive allocations for each control period as existing units is updated to eliminate any units whose scheduled retirements were considered sufficiently certain to be reflected in the budget-setting process with respect to that control period.

Comment: EPA received a comment suggesting modifications to the proposed methodology for calculating allowance allocations in the event that a unit has fewer than three years of operating history for use in calculating allocations. The commenter suggests that EPA either revise the allocation methodology for existing units so as to recompute existing unit-level allocations for each ozone season through 2024 to take account of additional years of heat input data for units in this situation or else allow units in this situation to receive allocations from new unit set-asides to the extent that their allocations as existing units are less than their actual emissions.

Response: EPA disagrees that the revisions to the allocation approach suggested by the commenter are needed in order to effectuate a reasonable allocation of allowances among all of the units in a given state. The suggested revisions would require promulgating new allocation methodologies for either all existing units or all new units that would differ from the allocation methodologies used in all of the existing CSAPR trading programs and that would change the allocation to the commenter's unit by at most one allowance for the 2021 control period. EPA notes that any state may submit SIP revisions to replace EPA's default allocations with state-determined allocations if the state would prefer that allowances be allocated differently among the state's units.

Comment: Commenters expressed diverse views on questions concerning allocations of allowances to units with unscheduled future retirements, ranging from recommendations that EPA end such allocations immediately upon a unit's retirement to recommendations that EPA continue allocations to retired units indefinitely.

Response: With respect to units with unscheduled future retirements, EPA proposed to provide allocations of allowances according to the methodology used in the existing CSAPR trading programs. Under this methodology, when a unit ceases operation for two consecutive control periods, it continues to receive allocations for five control periods starting with the first control period of non-operation. After the fifth control period, allowances that would otherwise have been allocated to the unit for future control periods are instead directed to the state's new unit set-aside for each control period, and if the unit happens to resume operation in a future control period, it is eligible to receive allocations only as a “new” unit from the new unit set-aside for that control period. The rationale for continuing to make allocations to sources that have ceased operations for five control periods, rather than ending allocations immediately or extending the allocations indefinitely, is to balance two concerns. The first concern, which tends to suggest reallocating allowances more quickly, is to ensure that allowances are available for new units as the generating fleet evolves. The second concern, which tends to suggest reallocating allowances less quickly, is to ensure that the program's allowance allocation provisions do not distort a unit owner's incentives to pursue what would otherwise be the most economic compliance strategy. Such distortion could occur if a predicted immediate loss of allowance allocations upon closure would give the owner of an otherwise uneconomic unit an incentive to keep the unit in operation just to receive allowance allocations.

None of the comments recommending a change to the approach for allocating allowances to units with unscheduled future retirements advocate allocating the allowances to other units instead. Rather, all of these comments appear to either explicitly or implicitly incorporate an assumption that the recommended change in allocations to the units with unscheduled future retirements would be accompanied by a corresponding change in the total number of allowances made available collectively to all units in the state under the Group 3 trading program. In other words, the comments recommending earlier discontinuation of allocations to retired units are actually advocating for reduced emission budgets implemented through the mechanism of reduced allocations to retired units, while the comments recommending more extended allocations to retired units are actually advocating for increased emission budgets implemented through the mechanism of increased allocations to retired units. EPA was unable to identify any comments advocating for changes in the methodology establishing the allocations to units with unscheduled retirements that were not effectively comments advocating for changes in the amounts of the emission budgets, which EPA considers a different issue. Comments on the amounts of the emission budgets are addressed elsewhere.

Comment: EPA received several comments regarding the question of when to treat a unit with an unscheduled future retirement as retired for purposes of triggering the count of the five control periods for which the unit would continue to receive allocations. One comment suggested that EPA begin counting the five control periods following the unit's announced retirement date rather than when the unit has ceased operating for two consecutive control periods. Other comments suggested that EPA evaluate non-operation on the basis of full calendar years rather than on the basis of control periods (i.e., the ozone season portions of calendar years).

Response: EPA disagrees with these comments. With respect to the suggestion to wait for a retirement announcement even if a unit has ceased operation for two control periods, EPA sees no reason to also wait for a retirement announcement before taking observed information about a unit's actual non-operation into account for purposes of determining allowance allocations. With respect to the suggestion that EPA evaluate non-operation over full calendar years instead of control periods, EPA has followed the approach of considering a unit's non-operation during the relevant control period for each trading program—in other words, the calendar year for annual programs and the ozone season for seasonal NOX programs—in all of the existing CSAPR trading programs since 2015 without encountering any problems. Commenters have not identified any new issues with the existing procedures that would justify establishing different procedures for the new Group 3 trading program. The example cited by commenters where a unit might operate during the non-ozone season portion but not the ozone season portion of two consecutive calendar years is neither a new issue nor a significant problem. If such a unit loses its allocation as an existing unit but then operates in a subsequent ozone season, under the Group 3 trading program (as under the other seasonal NOX trading programs) the unit becomes eligible to receive allocations from the new unit set-aside.

Comment: EPA received several comments concerning units that have already retired. Some commenters recommended that these units should receive allocations under the Group 3 trading program at least until the units Start Printed Page 23131had received allocations for a total of five years of operation under the current Group 2 trading program and the new Group 3 trading program, generally citing the rationale described above for why EPA continues to provide allocations for a limited amount of time in the case of units with unscheduled future retirements.

Response: EPA disagrees with these comments. With respect to units that permanently retired before January 1, 2021, EPA will not provide allocations of Group 3 allowances. As noted above, the reason that the existing CSAPR trading programs have provided allowances to units for a period of time following their retirement is to avoid a distortion that immediate discontinuation of allocations otherwise could cause to the owner's incentives in making decisions about the unit's future. Where a unit's owners have already retired a unit, this reason for continuing allowance allocations for some period after retirement no longer applies. Thus, including a provision in the new Group 3 trading program that would allocate allowances to past retirements would simply redistribute allowances from operating units to retired units for no useful purpose. EPA again notes that any state may submit SIP revisions to replace EPA's default allocations with state-determined allocations if the state would prefer that allowances be allocated differently among the state's units.

Comment: Another commenter on units that have already retired recommended not only that the units continue to receive allocations for a combined five-year period under the Group 2 and Group 3 trading programs but also that after the end of the five-year period EPA should add an equivalent quantity of allowances to the states' new unit set-asides to ensure that sufficient allowances are available for new units.

Response: EPA disagrees with this additional recommendation. As discussed above with respect to the comments received on the methodology for allocating allowances to units with unscheduled future retirements, the recommendation to add allowances to the new unit set-asides is effectively a comment on the amounts of the emission budgets rather than on the allocation methodology, and comments on the amounts of the emission budgets are addressed elsewhere. EPA notes that the process for setting the emission budgets already includes a procedure to ensure that the emission budgets account for estimated emissions from planned new units.

Comment: With respect to units with scheduled future retirements, in the proposed rule EPA proposed to treat the units identically to units with unscheduled future retirements, allocating allowances to these units for five years starting with their first year of non-operation. Commenters observed that EPA was proposing to account for units with scheduled future retirements by reducing emission budgets in the control periods immediately following the retirements and suggested that it would be inconsistent for EPA to treat these units differently for purposes of the budget-setting process and the unit-level allocation process. The commenters implied that allowance markets might not be sufficiently liquid to facilitate the transfer of allowances from retired units to units that continue to operate, and suggested that the reductions in allocations to the operating units caused by the reductions in the overall emission budgets would constitute an increase in program stringency for the operating units.

Response: EPA disagrees with commenters' implicit assumption that allowance markets will be illiquid and with the suggestion that changes in how allowances are allocated among operating and retired units, as opposed to changes in state emission budgets, represent changes in overall program stringency. However, EPA agrees that it is reasonable to treat the units with scheduled future retirements more consistently across the budget-setting and unit-level allocation processes. Accordingly, in the final action, EPA is accounting for units with scheduled future retirements not only by reducing emission budgets in the control periods immediately following the retirements but also by ending allocations to those retired units in the control periods immediately following the retirements. Just as units that have already retired before 2021 are not included in the set of existing units to which allowances are allocated for 2021 under EPA's default methodology, in the final rule units with scheduled retirements before 2022, 2023, and 2024 are not included in the sets of existing units to which allowances are allocated for 2022, 2023, and 2024, respectively. As with EPA's rationale for not allocating allowances to units that have already retired, allocating allowances to a unit for control periods after the control period in which the unit is already scheduled to retire serves no useful purpose, because in such circumstances there is no potential distortion of economic incentives that needs to be considered. The effect of this change in the final rule is to ensure that allocations to a state's units that continue to operate do not change from control period to control period based solely on the retirements of other units in the state, where those retirements are known and already factored into the budgets. EPA notes that in the unlikely event that one of these retiring units operates in a control period after the control period in which it was scheduled to retire, it would be eligible to receive an allocation of allowances as a “new” unit from the state's new unit set-aside for the control period.

Finally, EPA notes that because this change addresses scheduled future retirements occurring in 2021, 2022, and 2023 that are first reflected in the state emission budgets for the 2022, 2023, and 2024 control periods, respectively, the change first affects unit-level allocations as of the 2022 control period. Under this final rule, every Group 3 state has the ability to establish state-determined unit-level allocations to replace EPA's default unit-level allocations through SIP revisions for any control period after 2021. Thus, any state that that would prefer to allocate allowances for control periods after 2021 to units with scheduled future retirements has the ability to do so through SIP revisions.

c. Allocations to New Units

Consistent with the updates to which units are considered to be “existing units” described in section VII.C.3.b, for purposes of this final rule a “new unit” that is eligible to receive allocations from the new unit set-aside (NUSA) for a state includes any covered unit that commences commercial operation on or after January 1, 2019, as well as a unit that becomes covered by meeting applicability criteria subsequent to January 1, 2019; a unit that relocates to a different state covered by a FIP promulgated by this rule; and an “existing” covered unit that loses its allocation as an existing unit due to a scheduled retirement or by otherwise ceasing operation but that resumes operation at some point thereafter. The amounts of allowances initially placed in each new unit set-aside for potential allocation to new units are determined as described in section VII.C.3.a. In addition, any allowances that would otherwise have been allocated to a unit with an unscheduled future retirement that is no longer eligible to receive allocations as an existing unit are redirected to the new unit set-aside for the state in which the unit is located. Units qualifying to receive allocations from a new unit set-aside may receive such allocations starting with the first Start Printed Page 23132year they are subject to the allowance-holding requirements of the rule. If the allowances in the NUSA for a given state are not allocated to new units, the allowances are redistributed to the existing units in the state before each compliance deadline.

In the final rule, under the new Group 3 trading program EPA will allocate allowances from each new unit set-aside using a one-round approach that will be carried out after the end of the control period at issue. Under the one-round approach, any eligible units in the state that operated during the control period will be allocated allowances in proportion to their respective emissions during the control period, up to the amounts of those emissions if the NUSA contains sufficient allowances, and not exceeding those emissions. Any allowances remaining in a new unit set-aside after the allocations to new units will be reallocated to the existing units in the state. EPA will issue a notice of data availability concerning the proposed allocations by March 1 following the control period, provide an opportunity for submission of objections, and issue a final notice of data availability and record the allocations by May 1 following the control period, one month before the June 1 compliance deadline. EPA believes this one-round approach for allocating allowances from each state's NUSA to eligible units is both simpler and more equitable that the two-round approach that EPA historically used in all the previous CSAPR trading programs. The existing CSAPR trading programs are being amended to also adopt the one-round approach starting with the 2021 control periods. The differences between the two-round and one-round procedures and reasons for adopting the revisions are discussed in section VII.C.8.b.[193]

Comment: EPA received comments concerning allocation of the portions of new unit set-asides composed of allowances redirected to the new unit set-asides from existing units that have retired and lost their allocations. Some commenters suggested that while EPA should make these allowances available to new units, EPA should not reallocate these allowances to existing units after the completion of allocations to eligible new units, or should reallocate allowances only where the existing units demonstrated emission rates at or below the levels EPA used in setting the state budgets. Other commenters supported the proposed approach of reallocating the allowances to existing units, as provided under the existing CSAPR trading programs.

Response: After consideration of the comments, EPA is finalizing the provisions that allow all allowances in the new unit set-asides, including allowances redirected from retired units, to be reallocated to existing units. As with many of the comments received concerning allocations to existing units, these comments are effectively advocating for reductions in the overall emission budgets through the mechanism of reduced allocations to certain units. In the final rule, this issue applies only the allowances no longer allocated to units with unscheduled future retirements, because the allowances formerly allocated to units with scheduled future retirements will be removed from the budgets for control periods after the scheduled retirements instead of being added to the new unit set-asides for the future control periods. EPA has not included a mechanism in this rule to adjust the emission budgets over time to account for either units with unscheduled future retirements or the construction of unplanned new units and is not prepared at this time to reduce the budgets for units with unscheduled future retirements without consideration of whether and how to increase the budgets for the construction of unplanned new units. Although EPA has determined that it is reasonable in this rule to reduce the emission budgets over time to account for units with scheduled future retirements, this is in part because EPA's budget-setting process also accounts for the construction of planned new units over time.

d. Allocations to New Units in Indian Country

Clean Air Act programs in Indian reservations and other areas of Indian country over which a tribe or EPA has demonstrated that a tribe has jurisdiction generally may be implemented either by a tribe through an EPA-approved tribal implementation plan (TIP) or by EPA through a FIP. Tribes may, but are not required to, submit TIPs. Under EPA's Tribal Authority Rule (TAR), 40 CFR 49.1-49.11, EPA is authorized to promulgate FIPs for sources in Indian country as necessary or appropriate to protect air quality if a tribe does not submit and receive EPA approval of a TIP. See 40 CFR 49.11(a); see also 42 U.S.C. 7601(d)(4). To date, no tribes have sought approval of a TIP implementing the good neighbor provision at CAA section 110(a)(2)(D)(i)(I) with respect to the 2008 ozone NAAQS. EPA has therefore determined that it is necessary and appropriate for EPA to implement the FIPs in any affected Indian reservations or other areas of Indian country over which a tribe has jurisdiction. However, there are no existing units that would qualify as “covered units” in Indian country located in the Group 3 states under this final rule.

EPA is generally applying the CSAPR Update approach for allocating allowances to any new units located in Indian country, with parallel modifications to those described above with respect to unit-level allocations from the new unit set-asides for units not in Indian country. Under this approach, allowances to possible future new units located in Indian Country will be allocated by EPA from an Indian country new unit set-aside established for the Indian country (if any) within each state's borders. The amounts of allowances initially placed in each Indian country new unit set-aside for potential allocation to new units are determined as described in section VII.C.3.a. Because states generally have no SIP authority in these areas, EPA will continue to administer the allocation of allowances to any sources that locate in such areas of Indian country within a state's borders over which a tribe or EPA has demonstrated that a tribe has jurisdiction, even if the state submits a SIP to replace the applicable FIP for the sources in the state.

EPA will allocate allowances from each Indian country new unit set-aside using a one-round approach that will be carried out after the end of the control period at issue. Under the one-round approach, any eligible units in the area of Indian country that operated during the control period will be allocated allowances in proportion to their respective emissions during the control period, up to the amounts of those emissions if the Indian country NUSA contains sufficient allowances, and not exceeding those emissions. Unallocated allowances from the Indian country new unit set-aside for Indian country within a particular state's borders will be returned to the state's new unit set-aside and allocated according to the methodology for that new unit set-aside. EPA believes this one-round approach for allocating allowances from each Indian country NUSA to eligible units is both simpler and more equitable than the two-round approach that EPA historically used in all the previous CSAPR trading programs. The existing CSAPR trading programs are being Start Printed Page 23133amended to also adopt the one-round approach starting with the 2021 control periods. The differences between the two-round and one-round procedures and reasons for adopting the revisions are discussed in section VII.C.8.b.[194]

Comment: Some commenters suggested alternatives to the provisions concerning the Indian country new unit set-asides. These commenters suggested that EPA should consolidate all allowances held back from all state budgets for potential new sources in Indian country into a single newly-created new unit set-aside for all Indian country within the borders of all states covered by the new Group 3 trading program in order to provide a clearer separation between state and tribal jurisdictions. The commenters further suggested that if any allowances in the consolidated Indian country set-aside are not allocated to new units, the allowances either should be retired or should be sold, with the sale proceeds dedicated to tribes through grant programs.

Response: EPA disagrees with these comments. There are no existing EGUs in Indian country within the borders of any state covered by the Group 3 trading program. All of the allowances being held back for potential allocation to new units in Indian country are being held back from state emission budgets. These budgets were determined based on the projected emissions of the existing units in the states after accounting for emission reductions achievable through implementation of the selected control strategy (with adjustments for known changes to the fleet of units such as scheduled future retirements of existing units and construction of planned new units). Because the allowances added to the Indian country new unit set-asides are being held back from the overall state budget amounts that would otherwise be allocated among each state's existing units, EPA believes the most reasonable disposition for these allowances if they are not allocated to potential new units in Indian country is to return them to the states from whose emission budgets the allowances were held back, after which the allowances can be redistributed in accordance with the procedures used to allocate the remainder of those states' budgets.

With respect to maintaining separation between state and tribal jurisdictions, EPA believes that the regulations for the new Group 3 trading program already maintain a clear separation between the new unit set-asides for each state and the Indian country new unit set-asides for Indian country within the borders of certain states, with the consequence that no greater clarity of separation would be achieved by consolidating the various Indian country new unit set-asides established under the new trading program into a single new unit set-aside for all Indian country within the borders of all states covered by the Group 3 trading program. Further, EPA believes such an approach would be likely to cause confusion because it would differ from the established approach already being implemented in all the existing CSAPR trading programs, where a separate Indian country new unit set-aside is established for any Indian country within the borders of any individual state covered by the trading program.

4. Transitioning From Existing CSAPR NOX Ozone Season Group 2 Trading Program

This section discusses four sets of provisions that EPA is implementing in order to address the transition of sources from the Group 2 trading program to the Group 3 trading program. First, to address the fact that the effective date for the final action in this rulemaking will fall after the start of the ozone season on May 1, 2021, and to ensure that under these circumstances the Group 3 trading program can be implemented for the full May-September ozone season in 2021 without imposing retroactive emission reduction requirements, EPA will allocate additional allowances, and make corresponding adjustments to states' 2021 assurance levels, so as to offset the otherwise applicable emission reduction requirements under this rulemaking for the portion of the 2021 ozone season occurring before the final rule's effective date. Second, in order to facilitate the continued use of market-based trading programs as the compliance mechanism for sources covered by this action while ensuring an appropriate level of stringency in the Group 3 trading program, EPA is implementing a process by which an initial bank of CSAPR NOX Ozone Season Group 3 allowances will be created through the conversion of certain banked CSAPR NOX Ozone Season Group 2 allowances allocated for the control periods in 2017 through 2020. Third, to provide additional compliance flexibility in extreme circumstances, EPA is creating a safety valve mechanism that would allow Group 3 sources to exchange additional 2017-2020 Group 2 allowances for Group 3 allowances at a higher conversion ratio. Finally, to maintain the previously established levels of stringency of the Group 2 trading program for the states and sources that remain subject to that program, CSAPR NOX Ozone Season Group 2 allowances equivalent in amount and usability to the vintage year 2021-2024 CSAPR NOX Ozone Season Group 2 allowances previously recorded in the compliance accounts for sources in the new Group 3 region are being recalled.

a. Supplemental Allowance Allocations To Avoid Retroactive Emission Reduction Requirements

Although EPA anticipates that this final rule will be published in the Federal Register by early April 2021, before the start of the 2021 ozone season on May 1, 2021, the effective date of the rule will fall after May 1, 2021 because of the requirements of the Congressional Review Act (CRA), 5 U.S.C. 801-808. Under CRA section 801(a)(3), a “major rule,” as defined under the CRA, generally may not take effect sooner than 60 days after the date of publication in the Federal Register (or, if later, 60 days after the date on which Congress receives a report on the final rule from EPA). Under CRA section 804(2), a “major rule” includes any rule that the Office of Management and Budget (OMB) finds is “economically significant” under Executive Order 12866, that is, a rule likely to result in an annual effect on the economy of $100 million or more. Because this rule is projected to result in annualized benefits greater than $100 million per year, as discussed in section VIII of the preamble, OMB has found that the rule is “economically significant.” It is thus a “major rule” for CRA purposes, with the result that the rule's effective date will occur after the start of the 2021 ozone season.

EPA finds that, notwithstanding that the final rule's effective date will be after May 1, 2021, it will nevertheless serve the public interest and greatly aid in administrative efficiency for most elements of the Group 3 trading program—specifically, all elements of the trading program other than the elements designed to establish more stringent emissions limitations for the sources in Group 3 states—to start on May 1, 2021. This will facilitate implementation of the Group 3 trading program in an orderly manner for the entire 2021 ozone season and reduce compliance burdens and potential confusion. Each of the CSAPR trading programs for ozone season NOX is Start Printed Page 23134designed to be implemented over an entire ozone season. Implementing the transition from the Group 2 trading program to the Group 3 trading program in a manner that required the covered sources to participate in the Group 2 trading program for part of the 2021 ozone season and the Group 3 trading program for the remainder of that ozone season would be complex and burdensome for sources. Attempting to address the issue by splitting the Group 2 and Group 3 requirements into separate years is not a viable approach, because EPA has no legal basis for releasing the Group 3 sources from the emission reduction requirements found to be necessary in the CSAPR Update for a portion of the 2021 ozone season, and EPA similarly has no legal basis for deferring implementation of the 2021 emission reduction requirements found to be necessary under this rule until 2022. Moreover, the requirements of the Group 2 trading program and the Group 3 trading program are substantively identical as to almost all provisions, such that with respect to those provisions, a source will not need to alter its operations in any manner or face different compliance obligations as a consequence of a transition from the Group 2 trading program to the Group 3 trading program. Thus, EPA believes that no substantive concerns regarding retroactivity arise from implementing the Group 3 trading program starting on May 1, 2021, so long as those aspects of the Group 3 trading program that do meaningfully differ from the analogous aspects of the Group 2 trading program—that is, the relative stringencies of the two trading programs, as reflected in the emissions budgets and associated assurance levels—are applied only as of the effective date of the final rule.

Thus, with respect to two aspects of the final rule, EPA is making the following adjustments in 2021 ozone season obligations in order to ensure that no new requirements are imposed on any regulated parties prior to the effective date of the final rule.

To cause the more stringent budgets of the Group 3 trading program to apply only after the effective date of the final rule, EPA will make supplemental allocations of Group 3 allowances to Group 3 sources for the portion of the 2021 ozone season occurring before the effective date of the final rule. The total amounts of the supplemental allowances available for allocation to the sources in each state will be calculated by multiplying the difference between the state's Group 2 and Group 3 budgets by the fraction of the 2021 ozone season, measured in days, occurring before the final rule's effective date. The state's total amount of supplemental allowances will then be allocated among the state's existing units as if the supplemental allowances had been included in the state's 2021 emissions budget for the Group 3 trading program. The allocations of supplemental allowances will be recorded at the same time as the allocations from the budget.

To cause the more stringent assurance levels of the Group 3 trading program to apply only after the effective date of the final rule, EPA will include an increment in each state's assurance level for 2021 in addition to the state's emissions budget and variability limit for 2021. The amount of the increment will be computed as 1.21 times the total amount of supplemental allowances determined for the state as described above, where 1.21 is the ratio of the Group 2 state assurance levels to the Group 2 state budgets and is also the ratio of the Group 3 state assurance levels to the Group 3 state budgets. In the event of an exceedance of a state's assurance level, the allocations of supplemental allowances and the increment to the state's variability limit will also be taken into account for purposes of the calculations used to apportion responsibility for any exceedance of a state's assurance level among the owners and operators of the state's sources.

In all respects other than the allocation of supplemental Group 3 allowances and the addition of an increment to the states' assurance levels, EPA is implementing the Group 3 trading program for the 2021 control period exactly as the program would be implemented for any other control period. Thus, allocations of Group 3 allowances from each state's emissions budget to existing and new units are being made for the entire 2021 ozone season (i.e., May 1, 2021, through September 30, 2021), emissions will be monitored and reported for the entire 2021 ozone season, and as of the allowance transfer deadline for the 2021 control period (i.e., June 1, 2022) each source will be required to hold in its compliance account vintage-year 2021 Group 3 allowances not less than the source's emissions of NOX during the entire 2021 ozone season. Because of the supplemental allowances allocated for the portion of the 2021 ozone season before the rule's effective date, EPA finds that implementing the program in this manner will substantively apply the final rule's emission reduction requirements only from the rule's effective date. Similarly, because of the increment to the states' assurance levels for 2021, EPA finds that implementing the trading program in this manner will substantively apply the final rule's more stringent assurance levels only from the rule's effective date. Moreover, any efforts undertaken by a source to reduce its emissions during the portion of the 2021 ozone season before the effective date of the rule will aid the source's compliance by reducing the amount of Group 3 allowances that the source will need to hold in its compliance account as of the allowance transfer deadline, increasing the range of options available to the source for meeting its compliance obligations under the Group 3 trading program.

EPA requested comment on this approach for implementing the Group 3 trading program in a manner that would apply the substantive increases in stringency established under the final rule on and after, but not before, the final rule's effective date. No commenters opposed this approach.

b. Creation of Initial Group 3 Allowance Bank

For this rulemaking, EPA is creating a limited initial bank of allowances that can be used for compliance in the CSAPR NOX Ozone Season Group 3 Trading Program by converting certain allowances banked in 2017-2020 under the CSAPR NOX Ozone Season Group 2 Trading Program at a conversion ratio determined by a formula. Any such conversion of banked allowances from the Group 2 trading program for use in the Group 3 trading program must ensure that implementation of the Group 3 trading program will result in NOX emission reductions sufficient to address significant contribution in the 12 linked Group 3 states, while also providing industry certainty (and obtaining an environmental benefit) through continued recognition of the value of saving allowances through early reductions in emissions. EPA's approach to balancing these concerns in the CSAPR Update through the conversion of banked allowances from the CSAPR trading program for ozone season NOX emissions was upheld in Wisconsin v. EPA, see 938 F.3d at 321.

Similar to the approach taken in the CSAPR update, EPA is creating the initial bank of allowances for the Group 3 trading program through a one-time conversion of banked Group 2 allowances. The allowances in the initial Group 3 bank will be allocated for the 2021 control period and will therefore be useable in that control period or any subsequent control period. Because the purpose of an initial bank is to assist in compliance flexibility without relaxing the program Start Printed Page 23135stringency identified as appropriate to address states' obligations under the good neighbor provision, EPA's objective is to set the target amount for the initial bank at a level high enough to accommodate year-to-year variability in operations and emissions, as reflected in states' variability limits, but not high enough to allow sources collectively to plan to emit in excess of the collective state budgets. For this rulemaking, as proposed, EPA is determining that an initial bank amount approximately equal to the sum of the states' variability limits is a reasonable level to accomplish this objective, given the expectation that sources would generally seek to carry a bank of roughly that amount forward from year to year in order to retain a comparable degree of compliance flexibility in subsequent control periods.[195] Further, because emission reductions from some of the emission controls that EPA has identified as appropriate to use in setting budgets are first reflected in the 2022 state budgets rather than the 2021 state budgets, EPA is basing the initial bank target amount on the sum of the states' 2022 variability limits rather than the 2021 variability limits. This approach results in an initial bank target amount of 21,777 allowances, computed on the basis of a full ozone season.

As discussed in section VII.C.4.a, the effective date of this rule will occur after the start of the 2021 ozone season, and adjustments are being made to ensure that the increased stringency of this rule's state budgets and state assurance levels (i.e., the sums of the budgets and variability limits) takes effect only after the rule's effective date. Consistent with these other adjustments, and as proposed, the initial bank target amount will be similarly prorated. For example, if the effective date of the final rule is June 1, 2021, which would be after the first 31 days of the 153-day ozone season have passed, the full-season initial bank target amount of 21,777 allowances would be prorated to an initial bank target amount of 17,365 allowances.[196] EPA notes that prorating the bank amount in this manner will not reduce sources' compliance flexibility for the 2021 ozone season, because the amounts of Group 3 allowances that sources will receive for the portion of the 2021 ozone season before the rule's effective date will be based on the existing Group 2 trading program budgets. The Group 2 budgets exceed the sources' collective 2019 emissions by over 38,000 tons (and exceed the sources' 2020 emissions by almost 60,000 tons), indicating potentially surplus allowances well above the full-season initial bank target amount of 21,777 allowances. Thus, although the prorating procedure will reduce the amount of Group 3 allowances that sources will receive in the form of an initial bank, the reduction in the quantity of these allowances will be more than offset by the supplemental Group 3 allowances that will be allocated in excess of sources' recent historical emission levels for the portion of the ozone season before this final rule's effective date.

Taking the same approach as was followed in the CSAPR Update, EPA will allocate the new Group 3 allowances constituting the initial bank through a conversion process in which Group 2 allowances allocated for the 2017 through 2020 control periods and banked under the existing Group 2 trading program will be exchanged for Group 3 allowances allocated for the 2021 control period at a uniform conversion ratio determined by a formula. The conversions will be carried out at the level of individual sources and general accounts, in each case using the same uniform conversion ratio. By creating the new Group 3 allowances through the conversion of previously banked Group 2 allowances, the bank creation mechanism rewards holders of banked Group 2 allowances for conducting emission reduction activities that contributed to the creation of those banked allowances as well as for financially supporting emission reductions activities at other sources through allowance purchases. Creating the new Group 3 allowances through conversion of previously banked Group 2 allowances also helps preserve the stringency of the Group 2 trading program for the states that remain covered by that trading program at levels consistent with the stringency found to be appropriate to address those states' good neighbor obligations with respect to the 2008 ozone NAAQS in the CSAPR Update.

Under EPA's proposed approach for creation of the bank, the conversion formula would have used the total quantity of 2017-2020 Group 2 allowances being converted as the numerator and the total quantity of 2021 Group 3 allowances being created as the denominator. EPA also proposed to give holders of 2017-2020 Group 2 allowances complete flexibility to choose how many of those allowances they wanted to include in the conversion process, making the formula numerator entirely dependent on those choices. An unavoidable consequence of this proposed flexibility was that EPA would have been unable to predict the conversion ratio until holders finalized their choices shortly before the conversion date. In the proposed rule, the formula denominator was also uncertain to a lesser degree because of the then-unknown magnitude of the prorating adjustment affecting the quantity of 2021 Group 3 allowances being created, although this uncertainty will be resolved as of publication of the final rule in the Federal Register. Commenters requested that EPA provide greater certainty concerning the conversion process, as discussed later in this section; further, commenters submitted no comments asking EPA to finalize the proposed flexibility for Group 2 allowance holders.

After consideration of comments, EPA is not finalizing the proposed flexibility for Group 2 allowance holders to decide how many Group 2 allowances to include in the conversion process and is instead finalizing a formula for the conversion ratio based on an alternative offered for comment that provides greater certainty. Under the alternative being finalized, the formula numerator is the portion of the total existing bank of 2017-2020 Group 2 allowances attributable to the Group 3 states, which is more specifically defined as: (1) The sum of the budgets of the Group 3 states under the Group 2 trading program for the 2017-2020 control periods, plus (2) the portion of the initial Group 2 bank target amount attributable to the Group 3 states, minus (3) the emissions of sources in the Group 3 states for the 2017-2020 control periods.[197] The formula denominator in the final rule continues to be based on the initial Start Printed Page 23136Group 3 bank target amount, but reflects the full-season target amount instead of the target amount after the prorating adjustment. The final rule also provides for the value computed from the formula to be rounded down to the nearest whole number. Using data as of January 2021, the formula numerator will be 186,014 allowances [198] and the formula denominator will be 21,777 allowances, yielding a rounded conversion ratio of 8:1. In other words, the result of applying the formula in the final rule is that eight 2017-2020 Group 2 allowances will be exchanged for each 2021 Group 3 allowance created in the initial bank.[199] Continuing the previous example, if the rule's effective date is June 1, 2021 and the initial Group 3 bank target amount after prorating is therefore 17,365 allowances, then 138,920 Group 2 allowances [200] would be removed from the accounts where those allowances are held and 17,365 Group 3 allowances would be recorded in the same accounts.

In addition to requesting greater certainty about the conversion process, commenters also indicated an interest in receiving the allowances in the initial Group 3 bank more quickly than would have occurred under the proposed rule. In response to these comments, EPA has advanced the conversion process schedule such that the conversions will be completed more than two months earlier than proposed and shortly after recordation of Group 3 allowance allocations from the state budgets for the 2021 control period for most sources. Approximately 45 days after the rule's effective date, EPA will temporarily suspend acceptance of transfers of Group 2 allowances. Before resuming acceptance of such transfers, EPA will allocate Group 3 allowances up to the initial Group 3 bank target amount to Group 3 sources at the 8:1 conversion ratio in proportion to the amounts of 2017-2020 Group 2 allowances held in each such source's compliance account immediately before the conversion.[201] If the Group 3 sources' compliance accounts do not collectively hold enough Group 2 allowances to exchange for the full target amount of the initial Group 3 bank at the 8:1 conversion ratio, EPA will allocate Group 3 allowances up to the remainder of the initial Group 3 bank target amount to general accounts at the same 8:1 conversion ratio in proportion to each such general account's holdings of 2017-2020 Group 2 allowances immediately before the conversion.[202] For each Group 3 allowance allocated ard recorded in a given account, EPA will deduct from the same account eight 2017-2020 Group 2 allowances on a first-in, first-out basis. After recording the Group 3 allowance allocations and the 2017-2020 Group 2 allowance deductions, EPA will resume acceptance of transfers of Group 2 allowances.

Comment: Some commenters stated that EPA should not create an initial bank of Group 3 allowances because of a consequent reduction in stringency of the rule.

Response: EPA disagrees with the comment that it should not create an initial bank of Group 3 allowances under the new trading program. EPA believes creating an initial bank of Group 3 allowances will provide Group 3 sources flexibility to comply with the stringency of the new trading program in light of year-to-year variability in unit operations and emissions. Creating the initial bank of Group 3 allowances through the conversion of banked 2017-2020 Group 2 allowances also provides sources within the Group 3 states with an opportunity to benefit under the Group 3 trading program from their efforts to bank allowances under the Group 2 trading program. Failure to establish an initial bank could reduce the incentive to achieve early reductions and bank allowances in the future by signaling to market participants that banked allowances accrued under existing trading programs will hold no value in any future new or modified trading program. EPA's approach of establishing an initial Group 3 bank in an amount equal to the sum of the new trading program's aggregate variability limits is similar to the methodology followed in the CSAPR Update that was upheld against challenge in the Wisconsin decision.

Comment: Some commenters stated that EPA should create a larger bank, possibly by allowing some or all banked Group 2 allowances to be used for compliance in the Group 3 trading program on a 1-for-1 basis instead of being converted to Group 3 allowances at a conversion ratio greater than 1:1.

Response: EPA disagrees with these comments. Creating an overly large initial bank of Group 3 allowances, regardless of the conversion ratio used, would dilute the intended control stringency and emission budgets established in this rule to address Group 3 states' obligations under the good neighbor provision with respect to the 2008 ozone NAAQS. Certainly, given the large existing bank of 2017-2020 Group 2 allowances, allowing these Group 2 allowances to be used for compliance in the Group 3 trading program at a 1:1 ratio would unacceptably dilute the control stringency and emission budgets established by EPA in this rulemaking. As explained earlier in this section, EPA believes that creating an initial bank of Group 3 allowances in an amount not exceeding the sum of the Group 3 state's variability limits, and doing so through conversion of Group 2 allowances at an 8:1 ratio, is consistent with both achieving the requisite level of stringency and encouraging continued use of banking.

Comment: A commenter suggested that EPA should base the initial Group 3 bank target amount not on the sum of Group 3 states' variability limits for 2022, but instead on the sum of Group 3 states' variability limits for 2021.

Response: EPA disagrees with this comment. The initial Group 3 allowance bank is intended to accommodate year-to-year variability in operations and emissions, and EPA expects that on average, sources collectively will aim to carry forward the bank from year to year so that in each subsequent control period, sources will continue to have the flexibility needed to accommodate year-to-year variability in operations and emissions. Unlike the 2022 state emission budgets, the 2021 state emission budgets do not reflect emission reductions achievable from application of the full control stringency Start Printed Page 23137that EPA is finding necessary to resolve the Group 3 states' obligations under the good neighbor provision with regard to the 2008 ozone NAAQS. Because the Group 3 bank is intended to be an element of the Group 3 trading program on an ongoing basis, not just in 2021, as a matter of program design EPA considers it appropriate for the amount of the initial Group 3 bank to represent the full control stringency found to be necessary under this rule. For this reason, it is appropriate to base the initial Group 3 bank target amount on the sum of states' variability limits for the 2022 control period rather than the 2021 control period.

EPA also views creation of the larger initial Group 3 bank suggested by the commenter as unnecessary to ensure compliance is achievable. After consideration of the prorating adjustment discussed earlier in this section, using the 2021 variability limits instead of the 2022 variability limits as the basis for determining the size of the initial Group 3 bank would increase the size of the bank by less than 600 allowances. In the very unlikely event that Group 3 sources are unable to reduce their emissions in the 2021 control period sufficiently to meet their compliance obligations by holding the Group 3 allowances allocated from the state emission budgets and from the initial Group 3 bank, sources would be able to obtain well over 600 additional Group 3 allowances for 2021 compliance by electing to use the safety valve mechanism discussed in section VII.C.4.c.

Comment: Some commenters stated that under the proposed conversion procedures, sources would not know the quantities of Group 3 allowances they would receive in sufficient time to rely on that information for purposes of planning their compliance activities.

Response: EPA acknowledges that the large degree of flexibility offered to holders of Group 2 allowances under the proposed rule created uncertainty regarding one aspect of the process for creating the initial Group 3 bank—specifically, the conversion ratio that would be used to create the initial Group 3 bank. As discussed above, the final rule modifies the formula for the conversion ratio so that the value of the ratio (i.e., 8:1 based on 2017-2020 data) is knowable as of the date of this final rule. In the final rule EPA has also advanced the schedule for carrying out the conversion process so that the allowances in the initial Group 3 bank will be recorded in accounts by 120 days after publication of the final rule in the Federal Register, or roughly two months before the end of the 2021 ozone season.

However, EPA disagrees that uncertainty regarding either the conversion ratio or the amount of Group 3 allowances received by any individual source from the initial bank has any meaningful effect on sources' ability to plan their compliance activities. As an initial matter, under a trading program, the most cost-effective compliance strategy for a source is generally to reduce its emissions if it believes it can do so for less than the market price of an allowance and then to purchase allowances as needed to cover its remaining emissions or to sell surplus allowances to other sources that cannot reduce emissions as cheaply. Because allowance prices under any trading program are uncertain, sources must always make these compliance planning decisions based on their best allowance price projections while recognizing the existence of price uncertainty. For purposes of forecasting future allowance prices under the Group 3 trading program, the only relevant question concerning the initial Group 3 bank is the total quantity of allowances that will be created in that bank, because that total amount will factor into the market balance between the overall supply of allowances and the overall demand for allowances. EPA's proposed rule provided essentially complete information about the total quantity of allowances that would be created in the initial Group 3 bank, and the final rule closely follows the proposed rule on this point. In short, nothing about EPA's proposed or final approach to creation of the initial Group 3 bank increased the uncertainty about future Group 3 allowance prices beyond the degree of uncertainty that is inherent in trading program-based approaches to environmental regulation.

Further, even if a particular source decides to not to avail itself of the flexibility provided by a trading program and instead chooses to plan its compliance strategy based on the number of allowances it expects to receive as zero-cost allocations, the quantity of allowances that a source might receive from the initial Group 3 bank would necessarily play a relatively modest role in such a strategy. Of the total allowances available for 2021 compliance that will be allocated to sources from the state emission budgets and from the initial Group 3 bank, more than 80 percent will come from the state emission budgets, and for subsequent control periods the proportion that will come from the state emission budgets will be 100 percent. In the proposed rule, EPA included extensive information on the proposed unit-level allocations from the proposed state emission budgets, including both a complete description of the allocation methodology and spreadsheets showing the allocations to each individual unit that would result from applying that methodology to the proposed state emission budgets. In the final rule, the only change to the allocation methodology is that, because certain units with scheduled future retirements will no longer receive allocations starting with the 2022, 2023, or 2024 control period when their scheduled retirements are taken into account for budget-setting purposes, the remaining units in those states that continue to operate will receive larger shares of the respective state budgets in those later control periods. It was clear from the proposed rule that any allocations of allowances from the initial Group 3 bank would be considerably smaller and therefore less relevant for an allocation-based compliance planning process than the allocations of allowances from the state emission budgets. This is consistent with the intended purpose of the initial Group 3 bank, which is to accommodate year-to-year variability in operations and emissions but not to allow for collective planned emissions to exceed the state emission budgets.

c. Opportunity To Obtain Additional Group 3 Allowances Through Further Conversion of Group 2 Allowances (“Safety Valve” Mechanism)

As discussed in section VI.B.1, in order to further ensure allowance market liquidity and compliance flexibility, in this final rule EPA is creating a “safety valve” mechanism that will allow Group 3 sources to access additional Group 3 allowances for the 2021 control period. The new Group 3 allowances would be created in exchange for banked 2017-2020 Group 2 allowances that have not already been exchanged for Group 3 allowances as part of the process of creating the initial Group 3 allowance bank described in section VII.C.4.b. The safety valve mechanism will be available for the month of February 2022, which falls approximately midway between October 30, 2021 (the deadline for reporting of emissions for the last three months of the 2021 control period under the Group 3 trading program) and June 1, 2022 (the deadline by which Group 3 sources must hold Group 3 allowances in their compliance accounts sufficient to cover their emissions during the 2021 control period). The conversion ratio used in the safety valve mechanism will be 18:1—in other words, 18 banked 2017-2020 Group 2 allowances would have to Start Printed Page 23138be surrendered in exchange for each Group 3 allowance issued through this mechanism. No Group 2 allowances will be exchanged for Group 3 allowances under the safety valve mechanism except as specifically requested by the designated representative for a Group 3 source.

EPA is establishing the safety valve mechanism and has designed its features to be responsive to comments on the proposed rule. Even without the safety valve mechanism, EPA considers it extremely unlikely that any source would be unable to achieve compliance with the Group 3 trading program's requirements. Sources have a flexible combination of options to achieve compliance, including reducing emission at the source, using the allowances allocated to the source from the applicable state's emissions budget and from the initial Group 3 bank, and purchasing allowances from other sources that have made emission reductions. However, given the short time-frame before implementation of the Group 3 trading program in the 2021 ozone season, inclusion of a safety valve mechanism will increase sources' confidence in their ability to comply. EPA views this as an enhancement to the trading program consistent with mechanisms included in other emissions trading programs, as long as the mechanism is implemented in a manner that does not weaken the ability of the trading program to address Group 3 states' obligations under the good neighbor provision with respect to the 2008 ozone NAAQS. The assurance that the safety valve mechanism does not undermine the purpose of the trading program is provided by the use of the 18:1 conversion ratio.

As discussed in section VII.C.4.b., EPA received comments expressing widely varying perspectives concerning whether and, if so, what quantities of Group 3 allowances should be made available for compliance flexibility beyond the allowances allocated from state budgets. Some industry commenters advocated for a mechanism that would allow them to purchase additional allowances at a price of $1,600 per allowance, consistent with EPA's estimate of the cost per ton of emission reductions achievable through optimization of installed SCR controls. In contrast, some commenters from downwind states advocated for no issuance of any Group 3 allowances beyond the state emission budgets, but one of these commenters also suggested that if any such Group 3 allowances were issued through the exchange of banked Group 2 allowances, the conversion ratio should reflect the relation of the estimated cost of the control strategy reflected in the Group 3 budgets to the market price of Group 2 allowances. The commenter suggested a conversion ratio of 11:1 based on the ratio of the same estimated $1,600 cost per ton of emission reductions available from SCR optimization to an estimated average market price for Group 2 allowances of $137 per Group 2 allowance during 2019 and 2020.[203] After consideration of these comments, EPA is setting the conversion ratio for the safety valve mechanism at 18:1, consistent with the principles underlying the recommendations of the commenters but using updated data. For the numerator of the conversion ratio, EPA is using $1,800 per ton, based on the estimated cost of the emission reductions available from SNCR optimization that are reflected in the final state emissions budgets. For the denominator of the updated ratio, EPA is using $100 per ton, reflecting an estimated average market price over the period from March 2020 through January 2021 for Group 2 allowances allocated for the 2020 control period.[204] EPA finds, first, that this conversion ratio is high enough to avoid interfering with incentives for sources to reduce emissions through the use of the control technologies identified as appropriate for establishing states' emissions budgets in this action, and second, that it is low enough to provide additional flexibility that, in extreme circumstances, could facilitate compliance by some sources. Based on the total quantity of banked 2017-2020 Group 2 allowances expected to remain after completion of the deductions necessary for Group 2 trading program compliance for the 2020 control period and the deductions made in the process of creating the initial Group 3 bank, EPA estimates that the maximum quantity of Group 3 allowances that could be created through the safety valve mechanism will be in the range of 7,000 to 9,000 Group 3 allowances. This degree of conversion of Group 2 allowances would be highly unlikely to occur, and indeed, EPA considers it more likely that no source will need to make use of the safety valve mechanism.

Under the final regulations, any use of the safety valve mechanism will be at the initiative of the designated representatives of Group 3 sources. Throughout the month of February 2022, EPA will accept requests from designated representatives for allocations of additional Group 3 allowances under the safety valve mechanism. It will be the responsibility of the Group 3 designated representatives to obtain any Group 2 allowances needed for this purpose, either by using any 2017-2020 Group 2 allowances remaining in the Group 3 source's compliance account after the initial Group 3 bank conversion process, transferring 2017-2020 Group 2 allowances held in the account of a Group 2 source under the control of the same owners and operators, or purchasing 2017-2020 Group 2 allowances from third parties. Holders of Group 2 allowances are not obligated to sell or transfer their allowances to effectuate such conversions if they prefer to retain such allowances for use in the Group 2 trading program.[205] As soon as practicable on or after March 1, 2022, if a request was received from the designated representative for a particular Group 3 source, EPA will deduct 2017-2020 Group 2 allowances in sets of 18 from the source's compliance account on a first-in, first-out basis up to the maximum number of sets of 2017-2020 Group 2 allowances available in the account. For each set of 2017-2020 Group 2 allowances deducted from a Group 3 source's compliance account, EPA will record one Group 3 allowance in the account.

Start Printed Page 23139

d. Recall of Group 2 Allowances Allocated for Control Periods After 2020

To maintain the previously established levels of stringency of the Group 2 trading program for the states and sources that remain subject to that program under this action, EPA is recalling CSAPR NOX Ozone Season Group 2 allowances equivalent in amount and usability to all vintage year 2021-2024 CSAPR NOX Ozone Season Group 2 allowances previously allocated to sources in Group 3 states and recorded in the sources' compliance accounts. Consistent with the proposed rule, the recall provisions established in this final rule apply to all sources in Group 3 states in whose compliance accounts CSAPR NOX Ozone Season Group 2 allowances for a control period from 2021 through 2024 were recorded, including sources where some or all units have permanently retired or where the previously recorded 2021-2024 allowances have been transferred out of the compliance account.[206] However, in response to comments discussed at the end of this section, and as further detailed below, the final rule provides a more flexible compliance schedule intended to accommodate any sources that have already transferred the previously recorded 2021-2024 allowances out of their compliance accounts and provides greater flexibility as to the vintage years of Group 2 allowances that sources may surrender to achieve compliance. As requested in comments, the final rule also clarifies how the recall provisions apply in instances where a source and its allowances have been transferred to different parties and adds more specificity regarding the procedures that EPA will follow to implement the recall.

Under the Group 2 trading program regulations, each Group 2 allowance is a “limited authorization to emit one ton of NOX during the control period in one year,” where the relevant limitations include the EPA Administrator's authority “to terminate or limit the use and duration of such authorization to the extent the Administrator determines is necessary or appropriate to implement any provision of the Clean Air Act.” 40 CFR 97.806(c)(6)(ii). In this action, the Administrator is determining that, in order to effectively implement the Group 2 trading program as a compliance mechanism through which states not subject to the Group 3 trading program may continue to meet their obligations under CAA section 110(a)(2)(D)(i)(I) with regard to the 2008 ozone NAAQS, it is necessary to limit the use of Group 2 allowances equivalent in quantity and usability to all Group 2 allowances previously allocated for the 2021-2024 control periods and recorded in the compliance accounts of sources in Group 3 states. The Group 2 allowances that have already been allocated to sources in Group 3 states for the 2021-2024 control periods and recorded in the sources' compliance accounts represent more than half of the total quantity of Group 2 allowances that have been allocated and recorded for the 2021-2024 control periods. Because allowances can be freely traded, if the use of the 2021-2024 Group 2 allowances previously recorded in Group 3 sources' compliance accounts (or equivalent Group 2 allowances) were not limited, the effect would be the same as if EPA had issued to sources in the states that will remain covered by the Group 2 trading program a quantity of allowances available for compliance under the 2021-2024 control periods more than double the levels that EPA determined to be appropriate emissions budgets for these states in the CSAPR Update. Through the use of banked allowances, the excess Group 2 allowances would affect compliance under the Group 2 trading program in control periods after 2024 as well. Continued implementation of the Group 2 trading program at levels of stringency consistent with the levels contemplated under the CSAPR Update therefore requires that EPA limit the use of the excess allowances, as EPA is doing in this final rule.

In the recall provisions finalized in this action, limitations on the use of the excess 2021-2024 Group 2 allowances are being implemented through requirements to surrender, for each 2021-2024 Group 2 allowance recorded in a Group 3 source's compliance account, one Group 2 allowance of equivalent usability under the Group 2 trading program. The surrender requirements apply to the owners and operators of the Group 3 sources in whose compliance account the excess 2021-2024 Group 2 allowances were initially recorded. In general, each source's current owners and operators will be required to comply with the surrender requirements for the source by ensuring that sufficient allowances to complete the deductions are available in the source's compliance account by one of two possible deadlines discussed below. However, an exception is provided if it is demonstrated to EPA's satisfaction that a source's current owners and operators obtained ownership and operational control of the source in a transaction that did not include rights to direct the use and transfer of some or all of the 2021-2024 Group 2 allowances allocated and recorded (either before or after that transaction) in the source's compliance account. The final rule provides that in such a circumstance, with respect to the 2021-2024 Group 2 allowances for which rights were not included in the transaction, the surrender requirements apply to the most recent former owners and operators of the source for which such a demonstration is not made. Because in this situation a source's former owners and operators might lack the ability to access the source's compliance account for purposes of complying with the surrender requirements, the former owners and operators will instead be allowed to meet the surrender requirements with Group 2 allowances held in a general account.[207]

To provide as much flexibility as possible consistent with the need to limit the use of the excess Group 2 allowances, for each 2021-2024 Group 2 allowance recorded in a Group 3 source's compliance account, EPA will accept the surrender of either the same specific 2021-2024 Group 2 allowance or any other Group 2 allowance with equivalent (or greater) usability under the Group 2 trading program. Thus, a surrender requirement with regard to a Group 2 allowance allocated for the 2021 control period may be met through the surrender of any Group 2 allowance allocated for the 2021 control period or the control period in any earlier year—in other words, any 2017-2021 Group 2 allowance.[208] Similarly, the surrender requirement with regard to a 2022 Group 2 allowance, a 2023 Group 2 allowance, or a 2024 Group 2 allowance may be met through the surrender of any 2017-2022 Group 2 allowance, any 2017-2023 Group 2 allowance, or any 2017-2024 Group 2 allowance, respectively.

Owners and operators subject to the surrender requirements can choose from two possible deadlines for meeting the requirements. The first deadline will be July 14, 2021. As soon as practicable or Start Printed Page 23140after this date, EPA will make a first attempt to complete the deductions of Group 2 allowances required for each Group 3 source from the source's compliance account. EPA will deduct Group 2 allowances first to address any surrender requirements for the 2021 control period, then to address any surrender requirements for the 2022, 2023, and 2024 control periods in turn. When deducting Group 2 allowances to address the surrender requirements for each control period, EPA will first deduct allowances allocated for that control period and then will deduct allowances allocated for each successively earlier control period. This order of deductions is intended to ensure that whatever Group 2 allowances are available in the account are applied to the surrender requirements in a manner that both maximizes the extent to which all of the source's surrender requirements will be met and also ensures that any Group 2 allowances left in the source's compliance account after completion of all required deductions will be the earliest allocated, and therefore most useful, Group 2 allowances possible. Among the Group 2 allowances allocated for a given control period, EPA will first deduct allowances that were initially recorded in that account, in the order of recordation, and will then deduct allowances that were transferred into that account after having been initially recorded in some other account, in the order of recordation.

Following the first attempt to deduct Group 2 allowances to address Group 3 sources' surrender requirements, EPA will send a notification to the designated representative for each such source (as well as any alternate designated representative) indicating whether all required deductions were completed and, if not, the additional amounts of Group 2 allowances usable in the 2021, 2022, 2023, and/or 2024 control periods that must be held in the appropriate account by the second surrender deadline of September 15, 2021. Each notification will be sent to the email addresses most recently provided to EPA for the recipients and will include information on how to contact EPA with any questions. Consistent with the proposed rule, the final rule provides that no allocations of Group 3 allowances will be recorded in a source's compliance account until all the source's surrender requirements with regard to 2021-2024 Group 2 allowances have been met. For this reason, the principal consequence to a source of failure to fully comply with the surrender requirements by the July 14, 2021 surrender deadline is that any Group 3 allowances allocated to the units at the source for the 2021 and 2022 control periods that would otherwise have been recorded in the source's compliance account by July 29, 2021 will not be recorded as of that recordation date.

If all surrender requirements of 2021-2024 Group 2 allowances for a source have not been met in EPA's first attempt, EPA will make a second attempt to complete the required deductions from the source's compliance account (or from a specified general account, in the limited circumstance noted above) as soon as practicable on or after September 15, 2021. The order in which Group 2 allowances will be deducted will be the same as described above for the first attempt.

If the second attempt to deduct Group 2 allowances to meet the surrender requirements through deductions from the source's compliance account (or from a specified general account) is unsuccessful for a given source, the final regulations provide that as soon as practicable on or after November 15, 2021, to the extent necessary to address the unsatisfied surrender requirements for the source, EPA will deduct the 2021-2024 Group 2 allowances that were initially recorded in the source's compliance account from whatever accounts the allowances are held in as of the date of the deduction, except for any allowances where, as of January 31, 2021, no person with an ownership interest in the allowances was an owner or operator of the source, was a direct or indirect parent or subsidiary of an owner or operator of the source, or was directly or indirectly under common ownership with an owner or operator of the source. Although this consequence of a source's failure to hold the allowances necessary to comply with the surrender requirements in the source's compliance account (or a specified general account) by the surrender deadline was not expressly stated in the recall provisions in the proposed rule, the provision merely makes explicit a remedy for a source's noncompliance that is inherent in EPA's existing authority under 40 CFR 97.806(c)(6)(ii) to limit the use of any Group 2 allowance as necessary or appropriate to address the requirements of CAA section 110(a)(2)(d)(i)(I). Before making any deduction under this provision, EPA will send a notification to the authorized account representative for the account in which the allowance is held and will provide an opportunity for submission of objections concerning the data upon which EPA is relying. In EPA's view, this provision does not unduly interfere with the legitimate expectations of participants in the allowance markets because the provision would not be invoked in the case of any allowance that was transferred to an independent party in an arms-length transaction before EPA's intent to recall 2021-2024 Group 2 allowances became widely known. The provision would apply only to a Group 2 allowance that, as of January 31, 2021, was still controlled either by the owners and operators of the source in whose compliance account it was initially recorded or by an entity affiliated with such an owner or operator. EPA believes that by January 31, 2021, which was three months after publication of the proposed rule for this rulemaking in the Federal Register, all market participants had ample opportunity to become informed of the proposed rule provisions to recall 2021-2024 Group 2 allowances recorded in Group 3 sources' compliance accounts.[209]

The final rule includes the proposed provision under which failure of a source's owners and operators to comply with the surrender requirements is subject to enforcement as a violation of the Clean Air Act, with each allowance and each day of the control period constituting a separate violation.

To eliminate any possible uncertainty regarding the amounts of Group 2 allowances allocated for the 2021-2024 control periods (or earlier control periods) that the owners and operators of each Group 3 source must surrender under the final rule's recall provisions, EPA has prepared a list of the sources in Group 3 states in whose compliance accounts allocations of 2021-2024 Group 2 allowances were recorded with the amounts of the allocations recorded in each such compliance account for each control period from 2021 through 2024. An additional list shows, for each Group 3 source, the specific Group 2 allowances (batched by serial number) allocated for each control period and recorded in the source's compliance account and indicates whether, as of January 31, 2021, that batch of allowances was held in the source's compliance account, in an account believed to be partially or fully Start Printed Page 23141controlled by a related party (i.e., an owner or operator of the source or an affiliate of an owner or operator of the source), or in an account believed to be fully controlled by independent parties. The lists are in a spreadsheet entitled “Recall of CSAPR NOX Ozone Season Group 2 Allowances”, available in the docket for this action. After the first and second surrender deadlines, EPA intends to update the lists to indicate for each Group 3 source whether or not the surrender requirements for the source under the recall provisions have been fully satisfied. EPA will post the updated lists on a publicly accessible website to ensure that all market participants have the ability to determine which specific 2021-2024 Group 2 allowances initially recorded in any given Group 3 source's compliance account do or do not remain subject to potential deduction to address the source's surrender requirements under the recall provisions.

Comment: One commenter requested that EPA provide greater flexibility for complying with the recall provisions in the case of sources that may have already sold the 2021-2024 Group 2 allowances previously recorded in the sources' compliance accounts. The commenter suggested that such sources might have difficulty acquiring the Group 2 allowances needed to comply with the surrender requirements by the proposed surrender date, which would have been 60 days after publication of the final rule in the Federal Register.

Response: In the final regulations, EPA has modified the recall provisions to provide two deadlines for compliance with the surrender requirements: July 14, 2021 and September 15, 2021. The final provisions also provide greater flexibility than the proposed rule by allowing the surrender requirements to be satisfied not only with Group 2 allowances allocated for the same control periods as the excess 2021-2024 Group 2 allowances, but also with Group 2 allowances allocated for earlier control periods. Any source may miss the first surrender deadline with no consequence except that any Group 3 allowances allocated to the units at the source will not be recorded in the source's compliance account by the otherwise applicable recordation date of July 29, 2021, but instead will be recorded after the source has fully complied with the surrender requirements. The second surrender deadline is expected to be more than five months after the publication date—and six months after the signature date—of this final rule. EPA believes that the second deadline provides sufficient time for any source that has sold the 2021-2024 Group 2 allowances initially recorded in the source's compliance account to acquire replacement Group 2 allowances for purposes of complying with the recall provisions. Further, because at the time of the proposed rule in this action, the large majority of Group 3 sources subject to the recall provisions still held all 2021-2024 Group 2 allowances initially recorded in their compliance accounts, EPA expects that most sources will be able to easily comply with the first deadline simply by not transferring those Group 2 allowances to another account before that deadline. Inclusion of the first deadline thus ensures that EPA will be able to record most Group 3 allowances within 30 days after the effective date of this final rule.

Comment: One commenter suggested that retired sources that have already sold the 2021-2024 Group 2 allowances recorded in their compliance accounts should not be subject to the recall provisions on the grounds that “requiring already-retired units to purchase allowances would be antithetical to the goal of a trading program, in which allowances can be freely traded.” The commenter also suggested that there might be no willing sellers of Group 2 allowances from whom the retired sources could purchase replacement Group 2 allowances to comply with the recall provisions. The commenter further asserted that the proposed rule did not provide adequate notice that the recall provisions would apply to retired sources because the proposed regulatory text included a cross-reference to an existing rule section that addresses retired units.

Response: EPA disagrees with this comment. As explained earlier in this section, recall of Group 2 allowances equivalent in quantity and usability to the 2021-2024 Group 2 allowances allocated to Group 3 sources is necessary to maintain the previously established levels of stringency of the Group 2 trading program for the states and sources that remain subject to that program, because not recalling the excess allowances would be equivalent to increasing the budgets for the remaining Group 2 states, contrary to the stringency of the requirements established for those states in the CSAPR Update. The necessity of recalling the excess Group 2 allowances exists regardless of whether the sources in whose compliance accounts the excess allowances were initially recorded continue to operate or have retired.

The commenter provides no support for the assertion that requiring retired sources to comply with the recall provisions is somehow inconsistent with a trading program, and EPA sees no basis for the assertion. EPA has made clear that the recall provisions apply to the owners and operators of the sources in whose compliance accounts the excess Group 2 allowances were initially recorded (and who paid nothing for those Group 2 allowances), not to persons who may have purchased the excess Group 2 allowances in arms-length transactions before EPA provided general notice of the proposed recall. By honoring arms-length market transactions for Group 2 allowances, EPA is executing the recall in a manner that is entirely consistent with the normal freedom to trade allowances under EPA's trading programs.

The commenter's suggestion that there might be no willing sellers of Group 2 allowances is speculative and contrary to EPA's experience in administering every trading program implemented by the Agency over the course of the last 25 years, starting with the Acid Rain Program. The commenter's statement that “some Group members are already finding that Group 2 allowances are not readily available because companies are holding onto them” is vague and insufficient to counter EPA's reasonable expectation, supported by decades of experience, that Group 2 allowances will be available for purchase in the six-month period following finalization of this action. To the extent that public notice of proposed changes to regulatory requirements may have temporarily affected activity in the market for Group 2 allowances, any such temporary effects would indicate only that, as intended, public notice made market participants aware of the possibility of changed regulatory requirements. The fact that some market participants may view waiting for the additional information contained in the final regulatory requirements as sensible does not serve as a reasonable basis for assertions that allowance markets will be illiquid when those final regulatory requirements are made public.

The commenter's assertion that the proposed rule did not provide adequate notice that the recall provisions would apply to retired sources is belied by the fact that the commenter, as well as other commenters, understood that the recall provisions were proposed to apply to retired sources and submitted comments on that aspect of the proposed rule. Moreover, the commenter offers no basis to support the notion that any person reviewing the proposed rule would reasonably have believed that the proposed recall did not apply to retired Start Printed Page 23142sources. The section of the preamble to the proposed rule that discusses the recall provisions states that the recall was proposed to apply with respect to “all” 2021-2024 Group 2 allowances allocated not only to sources in Group 3 states but also to non-source entities in Group 3 states. There is no language indicating that any source, retired or not, would be treated differently under the provisions than any other source, and the inclusion of non-source entities left no room for an interpretation that continued production of electricity and emissions was a prerequisite for applicability of the recall. The fact that, in order to be as clear as possible that the recall applied to sources with retired units, the proposed regulatory text included a cross-reference to an existing regulatory text provision identifying permanently retired units, but did not use the exact words “permanently retired units,” does not somehow manufacture a lack of notice. The use of cross-references is common and appropriate in regulatory text. Further, the proposed regulatory text would have encompassed 2021-2024 Group 2 allowances allocated to retired units even without the clarifying cross-reference.

Comment: One commenter requested that the regulations lay out in greater detail the specific procedures EPA would follow to administer the recall. The commenter sought clarification specifically as to how the recall provisions would apply in instances where a source or its allowances had been sold, potentially to different purchasers.

Response: As described earlier in this section, the final regulations include more detailed provisions concerning the procedures EPA will follow to deduct Group 2 allowances to implement the surrender requirements. Consistent with the proposed rule, the final surrender requirements apply with respect to all Group 3 sources in whose compliance accounts 2021-2024 Group 2 allowances were recorded, regardless of whether some or all units at the source may have retired or whether the source or its allowances may have been sold. However, in response to the comment, the final regulations provide that if it is demonstrated to EPA's satisfaction that the current owners and operators of a source obtained ownership and operational control of the source in a transaction that did not include rights to direct the use and transfer of some or all of the 2021-2024 Group 2 allowances allocated and recorded (either before or after that transaction) in the source's compliance account, then with regard to the 2021-2024 Group 2 allowances for which such rights were not included in the transaction, the surrender requirements apply to the most recent former owners and operators of the source for which such a demonstration is not made. EPA believes that this provision identifies the appropriate parties to whom the surrender requirements should apply in the situation identified by the commenter, consistent with EPA's intent expressed in the proposed rule for the requirements to apply to the owners and operators of the Group 3 source in whose compliance account the excess 2021-2024 Group 2 allowances were initially recorded.

Comment: One commenter requested that where a Group 3 source has purchased additional 2021-2024 Group 2 allowances beyond those initially recorded in the source's compliance account, the additional 2021-2024 Group 2 allowances should not be subject to the recall provisions but should remain available for transfer to a Group 2 source for future use in the Group 2 trading program. Alternatively, the commenter requested that EPA convert the additional 2021-2024 Group 2 allowances to Group 3 allowances that could be used in the Group 3 trading program.

Response: Under the final procedures for implementing the recall provisions, where a Group 3 source continues to hold the 2021-2024 Group 2 allowances initially recorded in the source's compliance account and also holds additional 2021-2024 Group 2 allowances purchased in an arms-length transaction before January 31, 2021, surrender of the initially recorded 2021-2024 Group 2 allowances would satisfy the recall provisions and the purchased 2021-2024 Group 2 allowances would remain available for transfer to a Group 2 source for future use in the Group 2 trading program. The purchased 2021-2024 Group 2 allowances would not be available for use in the Group 3 trading program, either through conversion to Group 3 allowances or otherwise.

5. Compliance Deadlines

As discussed in section IV.C. of this preamble, the final rule requires sources to comply with the revised respective NOX emission budgets for the ozone seasons (May 1 through September 30 of each year) in 2021 and subsequent years in order to ensure that these necessary NOX emission reductions are implemented to assist in downwind states' attainment and maintenance of the 2008 ozone NAAQS. The increased stringency of Group 3 budgets for the 2021 ozone season will take effect on the effective date of this action, which will be part of the way into the 2021 ozone season, but before the July 20, 2021 Serious area attainment date. Thus, under the new CSAPR NOX Ozone Season Group 3 Trading Program established in this rulemaking, the first affected control period is the 2021 ozone season (i.e. May 1, 2021, through September 30, 2021).

Under all CSAPR trading programs, compliance at the source level is achieved by each source surrendering by a compliance deadline—defined in the new Group 3 trading program regulations at 40 CFR 97.1002 as the “allowance transfer deadline”—a number of allowances equal to the source's total emissions for the preceding ozone-season control period.

For previous control periods under the existing CSAPR trading programs, the allowance transfer deadline was March 1 of the year following the control period. In this action, EPA is establishing the allowance transfer deadline for the Group 3 trading program—and for all the other CSAPR trading programs starting with the 2021 control periods [210] —to be June 1 of the year after the control period.[211] For example, under this coordinated deadline, June 1, 2022, is the date by which Group 3 sources will be required to hold Group 3 allowances for the 2021 control period. The reason for the change from earlier practice is to accommodate the change in the methodology and schedule for allocating allowances to units from the new unit set-asides that will start with the 2021 control periods. Under that revised methodology, allowances from the new unit set-asides will be recorded in sources' compliance accounts by May 1 of the year following the control period, and some additional period after that date is needed to allow for allowance purchases in case a source receives fewer allowances from the new unit set-aside than anticipated. Under the previous regulations at 40 CFR 97.812, the deadline for recording second-round allocations from the new unit set-asides was February 15, two weeks before the March 1 allowance Start Printed Page 23143transfer deadline. EPA believes sources will have greater trading flexibility if this interval is extended to a full month, resulting in the allowance transfer deadline of June 1. Extension of the allowance transfer deadline is not expected to have any impact on the achievement of the CSAPR trading programs' environmental objectives because it will not affect the quantities of allowances that sources will be required to hold as of the deadline or the total quantities of allowances that will be made available for compliance in advance of the deadline. Further discussion is provided in sections VII.C.3.c and VII.C.8.b.

EPA received no comments on the Group 3 trading program compliance deadlines for holding allowances after the end of each control period. Comments concerning the implementation of emission budgets that require emission reductions as of the 2021 ozone season instead of a later ozone season are addressed in sections VI.B.1 and VI.C.1.

6. Monitoring and Reporting

Monitoring and reporting in accordance with the provisions of 40 CFR part 75 are required for all units subject to all the CSAPR trading programs, which includes all units covered under this final rule. Consistent with these existing requirements, the monitoring system certification deadline by which monitors are installed and certified for compliance use under the CSAPR NOX Ozone Season Group 3 Trading Program generally will be May 1, 2021, the beginning of the first control period in this final rule, with potentially later deadlines for units that commence commercial operation less than 180 days before that date. Units already in compliance with monitoring system certification requirements for the Group 2 trading program will not have to undertake any additional activities to certify their monitoring systems for the Group 3 trading program. Similarly, Group 3 units will not have to undertake additional activities to update any facility account demographic information. All account demographic information of current Group 2 facility accounts will be transitioned to the Group 3 trading program, including an account's designated representative, alternate designated representative, and any agents. The first period in which emission reporting is required under the new Group 3 trading program will be the quarter that includes May 1, 2021, (i.e., the second quarter of the year that covers April, May, and June). These monitoring and reporting requirements and deadlines are analogous to the current deadlines under the CSAPR NOX Ozone Season Group 2 Trading Program.

Under 40 CFR part 75, a unit has several options for monitoring and reporting, including the use of a CEMS; an excepted monitoring methodology based in part on fuel-flow metering for certain gas- or oil-fired peaking units; low-mass emissions monitoring for certain non-coal-fired, low emitting units; or an alternative monitoring system approved by the Administrator through a petition process. In addition, sources can submit petitions to the Administrator for alternatives to individual monitoring, recordkeeping, and reporting requirements specified in 40 CFR part 75. Each CEMS must undergo rigorous initial certification testing and periodic quality assurance testing thereafter, including the use of relative accuracy test audits and 24-hour calibrations. In addition, when a monitoring system is not operating properly, standard substitute data procedures are applied and result in a conservative estimate of emissions for the period involved.

Further, 40 CFR part 75 requires electronic submission of quarterly emissions reports to the Administrator, in a format prescribed by the Administrator. The reports will contain all of the data required concerning ozone season NOX emissions.

Units currently subject to the CSAPR NOX Ozone Season Group 2 Trading Program are required to monitor and report NOX emissions in accordance with 40 CFR part 75, so covered sources in the Group 3 trading program will simply continue the same monitoring and reporting practices as required by 40 CFR part 75 under the Group 2 trading program.

7. Recordation of Allowances

EPA is establishing a schedule for recording allocations of vintage-year 2021 CSAPR NOX Ozone Season Group 3 allowances to ensure that affected sources are allocated vintage year 2021 allowances as soon as practicable and well before the 2021 ozone season compliance deadline (June 1, 2022). EPA is also establishing a schedule for recording allocations of vintage-year 2022 CSAPR NOX Ozone Season Group 3 allowances that accommodates sources' expectation to receive these allowance allocations soon after the publication of this final rule while also ensuring that states have the opportunity to develop and submit to EPA SIP revisions concerning allocations of allowances for vintage year 2022 and later.

Specifically, allocations to existing units for the first control period outlined in this final rule (i.e. the 2021 ozone season) will be recorded by July 29, 2021. EPA will also record allocation of vintage year 2022 allowances by this deadline for all units except those in states that provided to EPA, by June 29, 2021, a letter indicating an intent to submit a SIP revision that, if approved, would substitute state-determined allocations for the default allocations determined by EPA for the 2022 control period. The deadline for states to submit to EPA such SIP revisions will be September 1, 2021. If a state that notified EPA of its intent to submit a SIP revision fails to submit such a SIP by the SIP submission deadline, EPA will record vintage year 2022 FIP allocations to the sources in the state no later than September 15, 2021. No later than March 1, 2022, EPA will record the SIP allocations of vintage year 2022 Group 3 allowances for states with approved SIP revisions. By this same deadline, EPA will record the FIP allocations of vintage year 2022 Group 3 allowances for states whose SIP revisions are not approved by EPA.

The recordation deadline for vintage year 2021 allowances to existing units is anticipated to be approximately 11 months before the date by which sources are required to hold allowances sufficient to cover their emissions for that first control period (June 1, 2022, as discussed above). This schedule allows sources ample time to engage in allowance trading activities consistent with their preferred compliance strategies. EPA will record vintage year 2023 and 2024 Group 3 allowance allocations to existing units by July 1, 2022, and vintage year 2025 and 2026 Group 3 allowance allocations by July 1, 2023. By July 1 of each year after 2023, EPA will record Group 3 allowance allocations to existing units for the control period in the third year after the year of recordation. The recordation deadlines will apply to recordation of both allocations based on the default allocation provisions and allocations provided by states pursuant to approved SIP revisions.

As an exception to all of the recordation deadlines that would otherwise apply, EPA will not record any allocations of Group 3 allowances in a source's compliance account unless that source has complied with the requirements to surrender previously allocated 2021-2024 Group 2 allowances. The surrender requirements are necessary to maintain the previously established levels of stringency of the Group 2 trading program for the states and sources that remain subject to that Start Printed Page 23144program under this final rule. EPA finds that it is reasonable to condition the recordation of Group 3 allowances on compliance with the surrender requirements because the condition will spur compliance and will not impose an inappropriate burden on sources. EPA considers establishment of this condition, which will facilitate the continued functioning of the Group 2 trading program, to be an appropriate exercise of the Agency's authority under CAA section 301 (42 U.S.C. 7601) to prescribe such regulations as are necessary to carry out its functions under the Act.

EPA notes that recording allocations to existing units generally three years in advance under the new Group 3 trading program represents a change from the historical recordation schedules for allocations to existing units under the other CSAPR trading programs, which have generally provided for such allocations to be recorded four years in advance. In this action, EPA is revising the recordation schedules under the other CSAPR trading programs, as well as the similarly structured Texas SO2 Trading Program, so as to generally record allocations to existing units three years in advance. This change will take effect with allocations for the 2025 control periods, which will be recorded by July 1, 2022, instead of by July 1, 2021. The reason for the change is the discovery of a timing conflict in all the CSAPR trading programs between the requirement to record four years in advance and the separate provisions governing allocations to existing units that have ceased operations. Under those separate provisions, EPA is unable to determine whether some existing units are entitled to continue to receive their allowance allocations more than three years in advance, and thus EPA does not have the information necessary to record all the allocations four years in advance. Further discussion of this revision to the schedule for recording allocations to existing units is provided in section VII.C.8.a.

With respect to allocations of allowances from the new unit set-asides and Indian country new unit set-asides, in previous control periods under the existing CSAPR trading programs, EPA has recorded these allocations in two rounds, by August 1 of the control period and by February 15 of the year following the control period. In this action, EPA is adopting a new one-round process for determining allocations from the new unit set-asides and Indian country new unit set-asides, and consistent with that revised allocation process, starting with allocations for the 2021 control periods,[212] EPA will record all allocations from these set-asides as of May 1 in the year following the control period, in both the Group 3 trading program and the existing CSAPR trading programs, and both where the allocations are determined by EPA and where the allocations are provided by states pursuant to approved SIP revisions. Further discussion is provided in sections VII.C.3.c and VII.C.8.b.

8. Conforming Revisions to Regulations for Existing Trading Programs

As discussed elsewhere in this preamble, in most respects, but not in every respect, the provisions of the CSAPR NOX Ozone Season Group 3 Trading Program at subpart GGGGG of 40 CFR part 97 parallel the provisions that have applied for control periods through 2020 under the other CSAPR trading programs [213] at subparts AAAAA through EEEEE of part 97 established in the CSAPR rulemaking and the CSAPR Update and, to a somewhat lesser extent, the provisions of the similarly structured Texas SO2 Trading Program established at subpart FFFFF of part 97. This section discusses the provisions of the new Group 3 trading program that differ from the provisions that have applied under the existing CSAPR trading programs, beyond the provisions discussed in section VII.C.4. addressing the transition to the new Group 3 trading program. This section also discusses various minor corrections and clarifications to the regulations.

To clarify and facilitate administration of the regulations for all of EPA's trading programs in 40 CFR part 97, and to maintain their parallel nature to the extent possible, EPA is amending the regulations for the existing trading programs to reflect certain revisions as noted in the sections of this preamble describing the new Group 3 trading program. Section VII.C.8.a. addresses the revisions discussed in section VII.C.7. to address a timing conflict in the current regulations for all of the existing programs. Section VII.C.8.b. addresses the revisions discussed in sections VII.C.3.c and VII.C.3.d to simplify and improve the process for allocating allowances from the new unit set-asides and Indian country new unit set-asides under the existing CSAPR programs. Section VII.C.8.c. addresses additional minor revisions and corrections. EPA received no adverse comment regarding any of these conforming revisions or corrections.

In this action, EPA did not reopen or request comment on the regulations for any of the existing trading programs in 40 CFR part 97, subparts AAAAA through FFFFF, except with respect to specific revisions to these subparts identified in this section, as well as the revisions to the regulations for the Group 2 trading program discussed in section VII.C.4. that address the transition from the Group 2 trading program to the Group 3 trading program.

a. Resolution of Timing Conflict Between Certain Existing Provisions

Consistent with the provisions of the new CSAPR trading program finalized in this action, EPA is amending the regulations for the existing CSAPR trading programs and the Texas SO2 Trading Program to resolve a timing conflict between the provisions that set deadlines for recordation of allowances allocated to existing units and the provisions that govern allocations of allowances to units that have ceased operation for the control periods in at least two consecutive years. The recordation provisions in all of the trading programs generally have required EPA to record allocations of allowances to existing units four years in advance of the control periods for which the allowances are being allocated. For example, on July 1, 2020, EPA recorded allocations to most existing units of allowances for use in the 2024 control periods for all the existing trading programs. However, other provisions of all the trading programs require EPA not to record allocations to existing units that do not operate for two consecutive control periods, starting with the fifth control period after the first control period in which the unit did not operate. For example, if a unit that would otherwise receive allocations as an existing unit does not operate in the 2019 and 2020 control periods, the unit will continue to receive allocations for the control periods in 2019 through 2023 but will no longer be entitled to receive allocations for control periods in 2024 and thereafter. These two sets of timing requirements are in conflict, as demonstrated by the examples just presented: as of the July 1, 2020, Start Printed Page 23145deadline to record allocations for the 2024 control periods, EPA could not yet know whether any units that did not operate in 2019 might resume operation later in 2020, and EPA therefore could not yet know whether all such units would lose their eligibility to receive allocations for the 2024 control periods or not.[214]

To address the timing conflict described above, EPA is amending the regulations for each of the CSAPR trading programs and the Texas SO2 Trading Program to generally require recordation of allowances allocated to existing units to take place three years rather than four years in advance of the control period for which allowances are being allocated. Returning to the examples above, if these amendments had been in effect with respect to allocations for the control periods in 2024, EPA would not have been required to record allocations for the 2024 control period until July 1, 2021, by which time complete information on all units' operations in 2019 and 2020 will be available. Relatedly, for states that determine allocations of allowances to their sources under approved SIP revisions, EPA is amending the deadlines by which the states must submit the allocations to EPA for recordation. Under the amended deadlines, the states' submissions are due three years instead of four years before the applicable control period.[215]

The amended recordation and submission schedules will be effective beginning with recordation of allocations for control periods in 2025 and will apply to EPA's schedule for recording not only the allocations determined by EPA under the federal CSAPR trading programs but also the allocations determined by states or EPA under state CSAPR trading programs that are similarly recorded by EPA. EPA believes these amendments address the timing conflict in the existing trading program regulations in a manner that is as consistent as possible with the other provisions of the regulations, because while the amendments alter the point in time at which trading program participants receive allowances, the amendments will not alter the quantities of allowances received by any participant in any of the existing trading programs. In contrast, the only two other simple options for resolving the timing conflict—either shortening the period of non-operation that would cause a unit to lose its allocation from two years to one year or lengthening the period for which non-operating units would retain their allowance allocations from five years to six years—would cause changes in the amounts of allowances received by some trading program participants, and some stakeholders might view those changes as inequitable or undesirable for other policy reasons.

Further details on the specific regulatory provisions that are affected by the revisions are provided in section IX.D. of the preamble.

b. Modifications to NUSA Provisions

Consistent with the provisions of the new CSAPR trading program in this action for ozone season emissions of NOX from sources in Group 3 states, EPA is amending the regulations for the existing CSAPR trading programs governing allocations of allowances to units from NUSAs and Indian country NUSAs to reduce the potential for inequitable outcomes and to clarify and simplify the regulations. In order to ensure maximum consistency across all participants in the trading programs, the amendments will govern EPA's administration of the integrated trading programs not only under FIPs but also under approved SIPs where the NUSA allocation procedures are specified in provisions of the federal CSAPR trading programs in 40 CFR part 97 that have been incorporated into the SIP by reference.

The regulations applicable to control periods through 2020 under the existing CSAPR trading programs have provided for a two-round allocation process. For purposes of the first round, a unit was generally eligible to receive allocations from the NUSA for its state regardless of when it commenced commercial operation, as long as either no allocation of allowances to the unit as an existing unit was previously determined [216] or the unit was no longer entitled to receive its previously determined allocation as an existing unit. The first-round allocations were calculated during the control period at issue and were proportional to the eligible units' emissions during the preceding control period, up to the amount of allowances available in the NUSA. EPA performed preliminary calculations and published a notice by June 1, provided an opportunity for objections, and then adjusted the calculations as necessary, issued a final notice, and recorded the allocations by August 1 of the control period.

If any allowances remained in the NUSA for a given state after the first round, EPA carried out a second round, for which eligibility was limited to units that commenced commercial operation in the year of the control period at issue or the preceding year. The second-round allocations were calculated early in the year after the year of the control period at issue (very shortly after the January 30 deadline for submission of emissions data for October through December) and were proportional to the positive differences, if any, between the eligible units' emissions during the control period at issue and the amounts of any allocations the units received in the first round, up to the remaining amount of allowances available in the NUSA. Any allowances remaining after the second round were allocated to existing units in the state in proportion to their previous allocations. EPA made a preliminary identification of eligible units and published a notice by December 15, provided an opportunity for objections, and then performed the calculations, issued a final notice, and recorded the allocations by February 15 following the year of the control period, two weeks before the then-applicable March 1 allowance transfer deadline.

As indicated in the description above, the previous procedures had the potential to produce inequitable results, where some units could receive allowances in the first round (based on their emissions in the preceding control period) that exceeded the amounts needed to cover their emissions during the control period at issue, while other units that commenced operation more recently might not receive any allowances in either the first round (because the units had no covered emissions in the preceding control period) or the second round (because the NUSA may have been exhausted in the first round). Further, based on the experience of administering the two-round NUSA allocation process since 2015, EPA believes the previous procedures were unnecessarily complex and caused confusion for some market participants.Start Printed Page 23146

To simplify the NUSA allocation process and eliminate the potential inequities noted, EPA is amending the regulations for the existing CSAPR programs to replace the previous two-round NUSA allocation process with a one-round process that will allocate allowances to all eligible units in proportion to their emissions in the control period at issue. The amended provisions will be effective beginning with NUSA allocations for the control periods in 2021. Under the procedures, which apply to both NUSAs and Indian country NUSAs, EPA will perform preliminary calculations and issue a notice by March 1 of the year after the control period at issue, one month after the January 30 deadline for submission of the required emission data. After providing an opportunity for objections, EPA will make any necessary adjustments, issue a final notice, and record the allowances by May 1. To accommodate this process, the amendments also extend the allowance transfer deadline (i.e., the date by which all covered sources must hold allowances in their compliance accounts sufficient to cover their emissions during the preceding control period) by three months, from March 1 of the year following the control period to June 1. In coordination with the revised recordation deadlines, EPA is also extending the deadline for states to submit to EPA their state-determined allocations for new units from July 1 in the year of the control period to April 1 in the year following the control period. Finally, although the Texas SO2 Trading Program does not have NUSA provisions, in order to minimize unnecessary differences between the deadlines for analogous provisions in that program and the CSAPR programs, EPA is also revising the Supplemental Allowance Pool recordation deadline and the allowance transfer deadline under the Texas SO2 Trading Program to May 1 and June 1, respectively, of the year after the control period. Like the amendments to the NUSA provisions, the amendments to the deadlines described in this paragraph would apply for purposes of EPA's administration of the integrated trading programs under both FIPs and approved SIPs.

The revisions to the NUSA allocation procedures also allow for related simplification of the CSAPR trading programs' assurance provisions. Under the assurance provisions that have applied for control periods through 2020, when emissions in a state for a given control period exceed the state's assurance level, if there are any units in the state that operated during the control period but that did not receive an actual allowance allocation either as an existing unit or from the NUSA, the regulations require EPA to publish a notice calling for the owners and operators of such units to submit certain information which EPA would use to determine imputed allowance allocations for the units. EPA then would use the imputed allowance allocations for these units, together with the actual allowance allocations for other units, to apportion responsibility for the assurance level exceedance among the owners and operators of all the state's units. Under the amendments to the NUSA allocation process, all units that have covered emissions during any control period will receive allocations either as an existing unit or from the NUSA, making the procedures for determining imputed allocations unnecessary. Accordingly, EPA is simplifying the assurance provisions for all of the existing CSAPR trading programs by removing the requirement for EPA to issue the additional notice just discussed, starting with the 2021 control periods.[217] EPA is also revising the date as of which the “common designated representative” for a group of sources is determined for purposes of the assurance provisions from April 1 to July 1 of the year following the control period, preserving that date's current position of being one month after the allowance transfer deadline. This revision maintains the existing coordination between these two regulatory deadlines and applies to all the existing CSAPR trading programs, whether administered under FIPs or approved SIPs, as well as the Texas SO2 Trading Program.

EPA is making the changes to the NUSA allocation provisions, assurance provisions, and related deadlines effective as of the 2021 control period. EPA proposed to make the changes effective as of the 2023 control period, which is the first control period by which it would have been possible for states to fully replace the FIP requirements established in this action with a SIP revision. However, EPA also specifically requested comment on implementing the changes as of the 2021 control period. Having received no comment opposing the substance of the proposed revisions and no comment favoring implementation as of the 2023 control period, EPA is finalizing the amendments as of the 2021 control period in order to simplify the programs and clarify the regulations to the greatest extent possible.

Further details on the specific regulatory provisions that are affected by the revisions are provided in section IX.D. of the preamble.

c. Minor Corrections and Clarifications to Existing Regulations

EPA is implementing a small number of additional minor corrections and clarifications to the NUSA provisions in the existing CSAPR trading programs. First, EPA is amending the provisions that address the disposition of allowances that are determined to have been allocated incorrectly and that consequently are recalled and added to the NUSA for reallocation. The regulations that have applied through the 2020 control periods provided for the recalled allowances to be reallocated through the NUSA allocation process for the same control period for which the allowances were originally allocated incorrectly. Because some corrections may occur after the NUSA allocation process for a control period has already been completed, EPA is revising these provisions to also allow the recalled allowances to be reallocated as part of the NUSA allocation process for a subsequent control period.

Second, EPA is correcting the specific numbers of allowances identified as the NUSA amounts for several states under the existing CSAPR programs established in the CSAPR rulemaking.[218] Following the promulgation of the CSAPR regulations in August 2011, EPA issued two rules revising the amounts of the emissions budgets, NUSAs, and Indian country NUSAs for several states.[219] Subsequent to these rule revisions, EPA recalculated the allocations to individual existing units and published a notice of data availability establishing the new allocations.[220] However, because of rounding differences, in certain instances the sum of the recalculated allocations to the individual units in a state plus the amounts identified in the regulations for the NUSA and Indian country NUSA for the state does not exactly equal the state budget.[221] In this Start Printed Page 23147final action, EPA is adjusting the amounts of the NUSAs identified in the regulations for control periods in future years up or down by the amount needed to eliminate the rounding differences. The sizes of the NUSA adjustments range from 1 to 17 allowances. These revisions do not affect the amounts of any state emissions budgets.

Third, EPA is adding provisions to the regulations for each of the existing CSAPR trading programs addressing the disposition of allowances held in the compliance accounts of sources in states that are no longer covered by those programs. Under the added provisions, EPA would identify or, if necessary, establish a general account controlled by each such source's owners and operators and would transfer any such allowances to that general account. The added provisions parallel analogous provisions that were proposed and are being finalized in this action to address the disposition of any CSAPR NOX Ozone Season Group 2 allowances that may remain in the compliance accounts of sources in states covered by the new CSAPR NOX Ozone Season Group 3 trading program after the various procedures governing conversion or recall of such allowances have been carried out.

Finally, EPA is making non-substantive revisions to the sections of the existing CSAPR trading program regulations that set forth the amounts of the budgets, new unit set-asides, and variability limits. The revisions clarify the regulations by indicating the specific control periods when such amounts no longer apply to the sources in a given state because the state's sources are no longer required to participate in that trading program.

Further details on the specific regulatory provisions that would be affected by the revisions are provided in section IX.D. of the preamble.

D. Submitting a SIP

States may replace a FIP with a SIP under the Clean Air Act at any time if the SIP is approved by EPA, see CAA section 110(c)(1)(B). EPA has established certain specialized provisions for replacing FIPs with SIPs within all of the CSAPR trading programs, including the use of so-called “abbreviated SIPs” and “full SIPs,” see 40 CFR 52.38(a)(4) and (5) and (b)(4), (5), (8), and (9); 40 CFR 52.39(e), (f), (h), and (i). Under the new or amended FIPs for the 12 states whose sources are required to participate in the new CSAPR NOX Ozone Season Group 3 Trading Program, “abbreviated” and “full” SIP options continue to be available. An “abbreviated SIP” allows a state to submit a SIP revision that would establish state-determined allowance allocation provisions replacing the default FIP allocation provisions but leaves the remaining FIP provisions in place. A “full SIP” allows a state to adopt a trading program meeting certain requirements that would allow sources in the state to continue to use the EPA-administered trading program through an approved SIP revision, rather than a FIP. In addition, as under the CSAPR and the CSAPR Update, EPA is providing states with an opportunity to adopt state-determined allowance allocations for existing units for the second control period under this rule—in this case, the 2022 control period—through streamlined SIP revisions. See 76 FR 48326-48332 for additional discussion on full and abbreviated SIP options and 40 CFR 52.38(b).

1. SIP Option To Modify 2022 Allocations

As under the CSAPR and the CSAPR Update, EPA is allowing a state to submit a SIP revision establishing allowance allocations for existing units in the state for the second control period of the new requirements, in 2022, to replace the EPA-determined default allocations. This process is the same as the process used at the start of other CSAPR trading programs but with updated deadlines, i.e., a state must submit a letter to EPA by June 29, 2021 indicating its intent to submit a complete SIP revision by September 1, 2021. The SIP would provide in an EPA-prescribed format a list of existing units and their allocations for the 2022 control period. If a state does not submit a letter of intent to submit a SIP revision, the EPA-determined default allocations will be recorded by July 29, 2021. If a state submits a timely letter of intent but fails to submit a SIP revision, the EPA-determined default allocations will be recorded by September 15, 2021. If a state submits a timely letter of intent followed by a timely SIP revision that is approved, the approved SIP allocations will be recorded by March 1, 2022.

2. SIP Option To Modify Allocations in 2023 and Beyond

For the 2023 control period and later, states in the CSAPR NOX Ozone Season Group 3 Trading Program can modify the EPA-determined default allocations with an approved SIP revision. The SIP submittal deadline is December 1, 2021. The deadline for states to submit state-determined allocations beginning with the 2023 and 2024 control periods under an approved SIP would be June 1, 2022, and the deadline for EPA to record those allocations would be July 1, 2022. Similarly, a state can submit a SIP revision beginning with control periods in 2025 and beyond by December 1, 2022, with state allocations for the 2025 and 2026 control periods due June 1, 2023, and EPA recordation of the allocations by July 1, 2023. For the 2023 control period and later, SIPs can be full or abbreviated SIPs. As discussed in section VII.F.3. below, states will also have the option to expand applicability to include EGUs between 15 MWe and 25 MWe or, in the case of states subject to the NOX SIP Call, large non-EGU boilers and combustion turbines. Inclusion of the large non-EGUs would serve as a mechanism to address the state's outstanding regulatory obligations under the NOX SIP Call with respect to those sources, and the state would be allowed to allocate a defined quantity of additional Group 3 allowances because of the expanded set of sources. See above and 76 FR 48326-48332 for additional discussion on full and abbreviated SIP options and 40 CFR 52.38(b).

3. SIP Revisions That Do Not Use the New Group 3 Trading Program

States can submit SIP revisions to replace the FIP that achieve the necessary emission reductions but do not use the CSAPR NOX Ozone Season Group 3 Trading Program. For a transport SIP revision that does not use the CSAPR NOX Ozone Season Group 3 Trading Program, EPA would evaluate the transport SIP based on the particular control strategies selected and whether the strategies as a whole provide adequate and enforceable provisions ensuring that the necessary emission reductions (i.e., reductions equal to or greater than what the Group 3 trading program will achieve) will be achieved. In order to best ensure its approvability, the SIP revision should include the following general elements: (1) A comprehensive baseline 2021 statewide NOX emission inventory (which includes existing control requirements), which should be consistent with the 2021 emission inventory that EPA used to calculate the required state budget in this final action (unless the state can explain the discrepancy); (2) a list and description of control measures to satisfy the state emission reduction obligation and a demonstration showing when each measure would be in place to meet the 2021 and successive control periods; (3) fully-adopted state rules providing for such NOX controls during Start Printed Page 23148the ozone season; (4) for EGUs greater than 25 MWe, monitoring and reporting under 40 CFR part 75, and for other units, monitoring and reporting procedures sufficient to demonstrate that sources are complying with the SIP (see 40 CFR part 51 subpart K (“source surveillance” requirements)); and (5) a projected inventory demonstrating that state measures along with federal measures will achieve the necessary emission reductions in time to meet the 2021 compliance deadline. The SIPs must meet procedural requirements under the Act, such as the requirements for public hearing, be adopted by the appropriate state board or authority, and establish by a practically enforceable regulation or permit a schedule and date for each affected source or source category to achieve compliance. Once the state has made a SIP submission, EPA will evaluate the submission(s) for completeness. EPA's criteria for determining completeness of a SIP submission are codified at 40 CFR part 51 appendix V.

For further information on replacing a FIP with a SIP, see the discussion in the final CSAPR rulemaking (76 FR 48326).

4. No SIP Option for Additional States To Participate in the New Trading Program

EPA is not finalizing the proposed option that would have allowed EPA to approve a SIP submitted by a state whose sources are required to participate in the CSAPR NOX Ozone Season Group 1 Trading Program (i.e., Georgia) or a state whose sources are required to continue to participate in the CSAPR NOX Ozone Season Group 2 Trading Program (Alabama, Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Tennessee, Texas, and Wisconsin) requiring its sources to participate instead in the new Group 3 trading program. No comments were received indicating interest in such an option, and elimination of the option facilitates simplification and clarification of several areas of the regulations. A similar option was made available to Georgia in the CSAPR Update (with respect to the Group 2 trading program) to address possible concerns expressed by some commenters in the CSAPR Update rulemaking that if sources in Georgia were not allowed to trade with sources in other states, the allowances issued to the sources in Georgia would otherwise be of limited use. See 81 FR 74504, 74588 (former 40 CFR 52.38(b)(6)). Because EPA has already approved a SIP revision under which Georgia adopted a state program requiring its sources to participate in the Group 1 trading program, EPA in this action is simplifying and clarifying the regulations by removing the option for Georgia to instead adopt a SIP instead requiring its sources to participate in the Group 2 trading program. Relatedly, EPA is removing the provisions in the Group 2 trading program regulations setting forth the amounts of the emissions budget, new unit set-aside, and variability limit that would have applied if EPA had approved a SIP revision from Georgia's requiring the state's sources to participate in that program as well as the provisions in the Group 1 trading program regulations that would have converted all remaining Group 1 allowances into amounts of Group 2 allowances.

E. Title V Permitting

This final rule, like the CSAPR and the CSAPR Update, does not establish any permitting requirements independent of those under Title V of the CAA and the regulations implementing Title V, 40 CFR parts 70 and 71.[222] All major stationary sources of air pollution and certain other sources are required to apply for title V operating permits that include emission limitations and other conditions as necessary to ensure compliance with the applicable requirements of the CAA, including the requirements of the applicable SIP. CAA sections 502(a) and 504(a), 42 U.S.C. 7661a(a) and 7661c(a). The “applicable requirements” that must be addressed in title V permits are defined in the title V regulations (40 CFR 70.2 and 71.2 (definition of “applicable requirement”)).

EPA anticipates that, given the nature of the units subject to this final rule and given that all of the units covered here are already subject to the CSAPR Update, most if not all of the sources at which the units are located are already subject to title V permitting requirements. For sources subject to title V, the interstate transport requirements for the 2008 ozone NAAQS that are applicable to them under the new or amended FIPs would be “applicable requirements” under title V and therefore must be addressed in the title V permits. For example, requirements concerning designated representatives, monitoring, reporting, and recordkeeping, the requirement to hold allowances covering emissions, the assurance provisions, and liability are “applicable requirements” that must be addressed in the permits.

Title V of the CAA establishes the basic requirements for state title V permitting programs, including, among other things, provisions governing permit applications, permit content, and permit revisions that address applicable requirements under final FIPs in a manner that provides the flexibility necessary to implement market-based programs such as the trading programs established by the CSAPR and the CSAPR Update and this final rule. 42 U.S.C. 7661a(b); 40 CFR 70.6(a)(8) & (10); 40 CFR 71.6(a)(8) & (10).

In the CSAPR and the CSAPR Update, EPA established standard requirements governing how sources covered by that rule would comply with title V and its regulations.[223] 40 CFR 97.506(d) and 97.806(d). For any new or existing sources under this final rule establishing the Group 3 trading program, identical title V compliance provisions would apply, just as they would have in the CSAPR NOX Ozone Season Group 2 Trading Program. For example, the title V regulations provide that a permit issued under title V must include “[a] provision stating that no permit revision shall be required under any approved . . . emissions trading and other similar programs or processes for changes that are provided for in the permit.” 40 CFR 70.6(a)(8) and 71.6(a)(8). Consistent with these provisions in the title V regulations, in the CSAPR and the CSAPR Update, EPA included a provision stating that no permit revision is necessary for the allocation, holding, deduction, or transfer of allowances. 40 CFR 97.506(d)(1) and 97.806(d)(1). This provision is also included in each title V permit for an affected source. This final rule maintains the approach taken under the CSAPR and the CSAPR Update that allows allowances to be traded (or allocated, held, or deducted) without a revision to the title V permit of any of the sources involved.

Similarly, this final rule would also continue to support the means by which a source in a CSAPR trading program can use the title V minor modification procedure to change its approach for monitoring and reporting emissions, in certain circumstances. Specifically, sources may use the minor modification procedure so long as the new monitoring and reporting approach is one of the prior-approved approaches under the CSAPR and the CSAPR Update (i.e., approaches using a Start Printed Page 23149continuous emission monitoring system under subparts B and H of Part 75, an excepted monitoring system under appendices D and E to Part 75, a low mass emissions excepted monitoring methodology under 40 CFR 75.19, or an alternative monitoring system under subpart E of part 75), and the permit already includes a description of the new monitoring and reporting approach to be used. See 40 CFR 97.506(d)(2) and 97.806(d)(2); 40 CFR 70.7(e)(2)(i)(B) and 40 CFR 71.7(e)(1)(i)(B). As described in EPA's 2015 guidance, the Agency suggests in its template that sources may comply with this requirement by including a table of all of the approved monitoring and reporting approaches under the CSAPR and the CSAPR Update trading programs in which the source is required to participate, and the applicable requirements governing each of those approaches. Inclusion of the table in a source's title V permit therefore allows a covered unit that seeks to change or add to its chosen monitoring and recordkeeping approach to easily comply with the regulations governing the use of the title V minor modification procedure.

Under the CSAPR and the CSAPR Update, in order to employ a monitoring or reporting approach different from the prior-approved approaches discussed previously, unit owners and operators must submit monitoring system certification applications to EPA establishing the monitoring and reporting approach actually to be used by the unit, or, if the owners and operators choose to employ an alternative monitoring system, to submit petitions for that alternative to EPA. These applications and petitions are subject to EPA review and approval to ensure consistency in monitoring and reporting among all trading program participants. EPA's responses to any petitions for alternative monitoring systems or for alternatives to specific monitoring or reporting requirements are posted on EPA's website.[224] EPA maintains the same approach in this final rule.

Consistent with EPA's approach under the CSAPR and the CSAPR Update, the applicable requirements resulting from the new and amended FIPs generally will have to be incorporated into affected sources' existing title V permits either pursuant to the provisions for reopening for cause (40 CFR 70.7(f) and 71.7(f)) or the standard permit renewal provisions (40 CFR 70.7(c) and 71.7(c)).[225] For sources newly subject to title V that are affected sources under the FIPs, the initial title V permit issued pursuant to 40 CFR 70.7(a) should address the final FIP requirements.

As was the case in the CSAPR and the CSAPR Update, the new and amended FIPs impose no independent permitting requirements and the title V permitting process will impose no additional burden on sources already required to be permitted under title V and on permitting authorities.

F. Relationship to Other Emission Trading and Ozone Transport Programs

1. Existing Trading Programs

This final rule ends the requirements for sources in certain states to participate in the existing CSAPR NOX Ozone Season Group 2 Trading Program with respect to emissions occurring after 2020 and requires those same sources instead to participate in a new CSAPR NOX Ozone Season Group 3 Trading Program with more stringent emissions budgets with respect to those emissions.[226] As discussed in section VII.C.4. above, the final rule lays out certain requirements associated with this transition, including provisions to accommodate an effective date sometime after the start of the 2021 ozone season, two mechanisms for the creation of limited quantities of Group 3 allowances available for use in the new Group 3 trading program in exchange for certain banked 2017-2020 Group 2 allowances, and the recall of 2021-2024 Group 2 allowances previously allocated to the sources in Group 3 states. In addition, in section VII.C.8. of this document, EPA describes certain features of the new Group 3 trading program that differ from the current features of the other CSAPR trading programs and that EPA is adopting as revisions to the other CSAPR trading programs, as well as a subset of those new features adopted as revisions to the similarly structured Texas SO2 Trading Program. Beyond these items, nothing else in this rule affects any requirements for any source under the CSAPR NOX Annual, SO2 Group 1 or Group 2, or NOX Ozone Season Group 1 or Group 2 trading programs or the Texas SO2 Trading Program. These trading programs all remain in place and will continue to be administered by EPA.

2. Title IV Interactions

This final rule does not affect any Acid Rain Program requirements. Acid Rain Program SO2 and NOX requirements are established independently in Title IV of the Clean Air Act and will continue to apply independently of this final rule's provisions. Acid Rain sources will still be required to comply with Title IV requirements, including the requirement to hold Title IV allowances to cover SO2 emissions after the end of each annual control period. EPA notes that the deadline by which sources affected under the Acid Rain Program must hold Title IV allowances is not affected by this final action and will continue to be 60 days after the end of the control period (i.e., February 29 or March 1 of the following year). Thus, starting with the compliance deadlines in 2022 for the control periods in 2021, the Acid Rain Program deadline will be approximately three months earlier than the corresponding deadline by which sources affected under all the CSAPR trading programs and the Texas SO2 Trading Program must hold allowances available for compliance under those programs, which will be June 1 of the year following the year of the control period, as discussed in sections VII.C.5 and VII.C.8.b.

3. NOX SIP Call Interactions

States affected by both the NOX SIP Call and this action will be required to comply with the requirements of both rules. This final rule requires NOX ozone season emission reductions from EGUs larger than 25 MWe in many NOX SIP Call states and at greater stringency than required by the NOX SIP Call. Therefore, this final rule will achieve emission reductions sufficient to address the emission reduction requirements of the NOX SIP Call for these large EGUs.

The NOX SIP Call states used the NOX Budget Trading Program to comply with the NOX SIP Call requirements both for EGUs serving generators with a nameplate capacity greater than 25 MWe and for large non-EGU boilers and combustion turbines with a maximum design heat input greater than 250 mmBtu/hr. (In some states, EGUs Start Printed Page 23150serving a generator with a nameplate capacity equal to or smaller than 25 MWe were also part of the NOX Budget Trading Program as a carryover from the Ozone Transport Commission NOX Budget Program.) However, EPA discontinued the NOX Budget Trading Program after 2008 when implementation of the CAIR NOX Ozone Season Trading Program began. Since that time, states have had to find appropriate alternative ways to continue to show compliance with the NOX SIP Call, particularly for large non-EGUs. As one option, EPA has allowed states to modify the applicability provisions of the NOX ozone season trading programs established under CAIR and later the CSAPR Update (although not the CSAPR) to include all NOX Budget Trading Program units as a way to continue to meet the requirements of the NOX SIP Call for these sources.

In this action, as under CAIR and the CSAPR Update, EPA is again allowing any NOX SIP Call state affected by this final rule to voluntarily submit a SIP revision to expand the applicability of the CSAPR NOX Ozone Season Group 3 Trading Program to include all NOX Budget Trading Program units. As part of such a SIP revision, the state would be allowed to issue additional emission allowances capped at a level intended to preserve the stringency of the Group 3 trading program. Analysis shows that the NOX Budget Trading Program units (mainly large non-EGU boilers, combustion turbines, and combined cycle units with a maximum design heat input greater than 250 mmBtu/hr) continue to emit well below their portions of the NOX SIP Call state budgets. In order to ensure that the necessary amounts of EGU emission reductions occur for this final rule, the corresponding state ozone-season emissions budget amount can be increased by the lesser of: (1) The relevant non-EGU budget under the NOX SIP Call or (2) the highest emissions of the relevant set of non-EGUs in the most recent 3 years. EPA believes that the environmental impact would be neutral using this approach, and hourly reporting of emissions under 40 CFR part 75 would continue. This option will address requests by states for help in determining an appropriate way to address the continuing NOX SIP Call requirement for large boilers and turbines. If a state elects to bring its NOX SIP Call non-EGUs into the Group 3 trading program, the program's assurance provisions continue to apply only to the EGUs covered by the program, and the amounts of the variability limits and assurance levels established for EGUs will remain unchanged.

The NOX SIP Call generally requires that states choosing to rely on large EGUs and large non-EGU boilers and turbines for meeting NOX SIP Call emission reduction requirements must establish a NOX mass emissions cap on each source and require 40 CFR part 75, subpart H monitoring or alternative monitoring. As an alternative to source-by-source NOX mass emission caps, a state may impose NOX emission rate limits on each source and use maximum operating capacity for estimating NOX mass emissions or may rely on other requirements that the state demonstrates to be equivalent to either the NOX mass emission caps or the NOX emission rate limits that assume maximum operating capacity. Collectively, the caps or their alternatives cannot exceed the portion of the state budget for those sources. See 40 CFR 51.121(f)(2) and (i)(1). If a state submits and EPA approves a SIP expanding the applicability to include all of the state's NOX Budget Trading Program units in the CSAPR NOX Ozone Season Group 3 Trading Program, the cap requirement would be met through the new budget and the monitoring requirement would be met through the trading program provisions, which require part 75 monitoring. Whether states choose to include NOX Budget Trading Program units in the CSAPR NOX Ozone Season Group 3 Trading Program through SIPs or not, EPA will work with states to ensure that NOX SIP Call obligations continue to be met.

Comment: One commenter questioned the need to allow states to include large non-EGUs of the types that participated in the NOX Budget Trading Program in the Group 3 trading program since current ozone season NOX emissions from the large non-EGUs are a small fraction of historical emissions because many units have retired and the remaining ones have moved away from coal as the main fuel and are now largely natural gas-fired.

Response: EPA is not requiring states to include non-EGUs of the types that participated in the NOX Budget Trading program in the Group 3 trading program. EPA continues to believe that allowing states to include these sources in the Group 3 trading program (or for some states, the Group 2 trading program) provides states a potentially useful option for continued compliance with ongoing NOX SIP Call requirements.

Comment: A commenter questioned the methodology EPA would use to ensure that including these sources in the Group 3 trading program could be done in an environmentally neutral way. The commenter did not feel EPA had explained how that would work in practice.

Response: EPA has allowed the voluntary inclusion of these sources in the CSAPR programs for a number of years. The methodology for determining the amount of allowances to provide for the additional sources is capped at the lesser of recent historical actual emissions and the allocations from the original NOX SIP Call program. This methodology accounts for the kinds of overall emission reductions that have occurred as cited by the commenter and holds emissions at actual levels, thus not allowing emissions increases from a decision by a state to voluntarily include these sources in the new CSAPR trading program.

VIII. Costs, Benefits, and Other Impacts of the Final Rule

In the Regulatory Impact Analysis for the Final Revised Cross-State Air Pollution Rule Update for the 2008 Ozone NAAQS (RIA), EPA estimated the health and climate benefits, compliance costs, and emissions changes that may result from the final rule for the analysis period 2021 to 2040. The estimated health and climate benefits and compliance costs are presented in detail in the RIA accompanying this final action. EPA notes that the estimated health and climate benefits and compliance costs are directly associated with optimizing NOX removal by turning on and optimizing existing idled SCRs; optimizing existing idled selective non-catalytic reduction (SNCRs); and installing state-of-the-art combustion controls. The estimated health and climate benefits and compliance costs also result from a small amount of generation shifting as the power system adjusts to the regulatory requirements.

EPA analyzed this final action's emission budgets, using a uniform control stringency represented by $1,800 per ton of NOX (2016$), as well as a more and a less stringent alternative. The more and less stringent alternatives differ in that they set different NOX ozone season emission budgets for the affected EGUs. The less stringent alternative uses emission budgets that were developed using uniform control stringency represented by $500 per ton of NOX (2016$). The more stringent alternative uses emission budgets that were developed using uniform control stringency represented by $9,600 per ton of NOX (2016$). Table VIII.1 provides the projected 2021, 2025, 2030, 2035, and 2040 EGU emission reductions for the evaluated regulatory Start Printed Page 23151control alternatives.[227] For additional information on emissions changes, see Table 4.5 in Chapter 4 of the RIA.

Table VIII.1—Estimated 2021, 2025, 2030, 2035, and 2040 a EGU Emissions Reductions in the 12 States of NOX, SO2, and CO2 and More and Less Stringent Alternatives

[Tons] b c

Final ruleMore stringent alternativeLess stringent alternative
2021:
NOX (annual)16,00016,0002,000
NOX (ozone season)16,00016,0002,000
SO2 (annual) *(*)(*)(*)
CO2 (annual, thousand metric)
2025:
NOX (annual)21,00037,0002,000
NOX (ozone season)19,00034,0002,000
SO2 (annual) *(*)(*)(*)
CO2 (annual, thousand metric)5,00014,0004,000
2030:
NOX (annual)16,00027,0002,000
NOX (ozone season)13,00025,0002,000
SO2 (annual) *(*)(*)(*)
CO2 (annual, thousand metric)8,00019,0006,000
2035:
NOX (annual)15,00026,0002,000
NOX (ozone season)13,00025,0002,000
SO2 (annual) *(*)(*)(*)
CO2 (annual, thousand metric)8,00019,0006,000
2040:
NOX (annual)14,00025,0002,000
NOX (ozone season)13,00024,0002,000
SO2 (annual) *(*)(*)(*)
CO2 (annual, thousand metric)4,00013,0003,000
a The 2021-2040 emissions reductions estimates are based on IPM projections for CO2 and engineering analysis for annual and ozone season NOX. SO2 and PM2.5 emissions were only partially analyzed. IPM was run for the following years: 2021, 2023, 2025, 2030, 2035, 2040, 2045 and 2050. For more information, see Chapter 4 and the Ozone Transport Policy Analysis Final Rule TSD.
b NOX emissions are reported in English (short) tons; CO2 is reported in metric tons.
c In addition to no annual SO2 emissions reductions as shown in the table above, there are no annual direct PM2.5 emissions changes.
* There are no annual SO2 and PM2.5 emissions reductions that come from turning on SCRs and SNCRs assuming that nothing else changes, but EPA did not analyze the effects on SO2 and direct PM that may come from shifting power generation, for example from coal-fired power plants to gas-fired or other types of power plants. EPA does expect some changes in SO2 and PM2.5 emissions due to shifting of power generation.

EPA analyzed ozone-season NOX emission reductions and the associated costs to the power sector of implementing the EGU NOX ozone-season emissions budgets in each of the 12 states using the Integrated Planning Model (IPM) and its underlying data and inputs. The estimates of the changes in the cost of supplying electricity for the regulatory control alternatives are presented in Table VIII.2. Total costs continue to change in later IPM run years as the modeled system responds to projected demand growth and shifts in the power sector under the illustrative scenarios. For a detailed description of these cost trends, please see Chapter 4, Section 4.4.3 of the RIA.

Table VIII.2—National Compliance Cost Estimates (Millions of 2016$) for the Regulatory Control Alternatives

Final ruleMore-stringent alternativeLess-stringent alternative
2021-2025 (Annualized)$10.0$41.4$(2.9)
2021-2040 (Annualized)24.828.519.6
2021 (Annual)5.15.21.6
2022 (Annual)19.261.55.9
2023 (Annual)19.261.55.9
2024 (Annual)2.14.5(14.9)
2025 (Annual)1.64.0(14.9)
2030 (Annual)63.632.367.0
Start Printed Page 23152
2035 (Annual)18.241.214.3
2040 (Annual)8.8134.018.9
“2021-2025 (Annualized)” reflects total estimated annual compliance costs levelized over the period 2021 through 2025 and discounted using a 4.25 real discount rate.228 This does not include compliance costs beyond 2025. “2021-2040 (Annualized)” reflects total estimated annual compliance costs levelized over the period 2021 through 2040 and discounted using a 4.25 real discount rate. This does not include compliance costs beyond 2040. “2021 (Annual)” through “2040 (Annual)” costs reflect annual estimates in each of those years.

Tables VIII.3 and VIII.4 report the estimated economic value of avoided premature deaths and illness in each year relative to the baseline along with the 95% confidence interval. In each of these tables, for each discount rate and regulatory control alternative, multiple benefits estimates are presented reflecting alternative ozone and PM2.5 mortality risk estimates. For additional information on health benefits, see Chapter 5 of the RIA.

Table VIII.3—Table VIII.3. Estimated Discounted Economic Value of Ozone-Attributable Premature Mortality and Illnesses for the Final Policy Scenarios in 2021

[95% Confidence Interval; millions of 2016$] a b

Final ruleMore stringent alternativeLess stringent alternative
3% Discount Rate$230 ($58 to $480) c and $1,900 ($210 to $5,000) d$260 ($88 to $520) c and $1,900 ($210 to $5,000) d$22 ($6 to $47) c and $190 ($20 to $490) d
7% Discount Rate$200 ($38 to $460) c and $1,700 ($170 to $4,500) d$200 ($38 to $460) c and $1,700 ($170 to $4,500) d$20 ($4 to $45) c and $170 ($17 to $440) d
a Values rounded to two significant figures. The two benefits estimates are separated by the word “and” to signify that they are two separate estimates. The estimates do not represent lower- and upper-bound estimates and should not be summed.
b We estimated changes in annual mean PM2.5 and PM2.5 -related benefits in 2024, but not 2021. As discussed in Chapter 4, in 2021, the only control measure expected to be adopted for compliance in the regulatory control alternatives is optimization of existing SCRs, and this measure will operate only during the ozone season. As discussed in Chapter 3, NOX reductions in the ozone season provide minimal PM2.5 benefits since PM2.5 nitrate concentrations, which result from conversion of NOX emissions to nitrate, are minimal during the warmer temperatures during the ozone season. Conversely, the conversion of nitrates to PM2.5 is much greater in cooler (non-ozone season) months, and thus it becomes worthwhile to estimate PM2.5 benefits from NOX reductions in those months. In 2024, the presence of additional control measures that operate year-round and other changes in market conditions as a result of the rule lead to notable NOX reductions in the winter months.
c Sum of ozone mortality estimated using the pooled Katsouyanni et al. (2009) and Zanobetti and Schwartz (2008) short-term risk estimate and the Di et al. (2017) long-term mortality risk estimate. As PM-related mortality quantified using risk estimates from the Di et al. (2017) and Turner et al. (2016) are within 5% of one another, in the interest of clarity and simplicity, we present the results estimated using the risk estimate from Di et al. (2017) alone.
d Sum of ozone mortality estimated using the long-term risk estimate and the Di et al. (2017) long-term mortality risk estimate. As PM-related mortality quantified using risk estimates from the Di et al. (2017) and Turner et al. (2016) are within 5% of one another, in the interest of clarity and simplicity, we present the results estimated using the risk estimate from Di et al. (2017) alone.

Table VIII.4—Estimated Discounted Economic Value of Avoided Ozone and PM2.5-Attributable Premature Mortality and Illnesses for the Final Policy Scenario in 2024

[95% Confidence Interval; millions of 2016$] a b

Final ruleMore stringent alternativeLess stringent alternative b
3% Discount Rate$310 ($72 to $680) c and $2,400 ($250 to $6,200) d$530 ($130 to $1,100) c and $4,200 ($450 to $11,000) d$22 ($6 to $47) c and $190 ($20 to $490). d
7% Discount Rate$280 ($48 to $640) c and $2,100 ($210 to $5,600) d$470 ($84 to $1,100) c and $3,800 ($370 to $9,900) d$20 ($4 to $45) c and $170 ($17 to $440). d
a Values rounded to two significant figures. The two benefits estimates are separated by the word “and” to signify that they are two separate estimates. The estimates do not represent lower- and upper-bound estimates and should not be summed.
b No PM-attributable benefits accrue for this scenario.
c Sum of ozone mortality estimated using the pooled Katsouyanni et al. (2009) and Zanobetti and Schwartz (2008) short-term risk estimate and the Di et al. (2017) long-term mortality risk estimate. As PM-related mortality quantified using risk estimates from the Di et al. (2017) and Turner et al. (2016) are within 5% of one another, in the interest of clarity and simplicity, we present the results estimated using the risk estimate from Di et al. (2017).
d Sum of ozone mortality estimated using the long-term risk estimate and the Di et al. (2017) long-term mortality risk estimate. PM-related mortality quantified using risk estimates from the Di et al. (2017) and Turner et al. (2016) are within 5% of one another. In the interest of clarity and simplicity, we present the results estimated using the risk estimate from Di et al. (2017) alone.
Start Printed Page 23153

Table VIII.5 shows the estimated monetary value of the estimated changes in CO2 emissions expected to occur over 2021-2040 for the final rule, the more-stringent alternative