i x>
K ®!
\ s*
*i PRO^
Regulatory Impact Analysis for the Proposed
Revised Cross-State Air Pollution Rule (CSAPR)
Update for the 2008 Ozone NAAQS
-------
ERRATA SHEET
After completion of the RIA, EPA received revised production cost projections for the proposed
rule IPM run, which reduced the projected cost of the proposed rule. This Errata presents these
technical corrections. The first table presents the changes in the text and is followed by sets of
tables each showing the current table and corrected table.
Page numbers
Current Value
Corrected Value
(Highlighted in yellow)
ES-15
The estimated social costs to
implement the proposal, as
described in this document,
are approximately $21
million in 2021 and $6
million in 2025
(2016$).
The estimated social costs to
implement the proposal, as
described in this document,
are approximately $20
million in 2021 and $1
million in 2025
(2016$).
ES-16
The annual net benefits of the
proposal in 2021 (in 2016$)
are approximately -$21
million using a 3 percent
discount rate and a 7 percent
real discount rate. The annual
net benefits of the proposal in
2025 are approximately $27
million using a 3 percent real
discount rate and
approximately -$0.9 million
using a 7 percent real
discount rate.
The annual net benefits of the
proposal in 2021 (in 2016$)
are approximately -$20
million using a 3 percent
discount rate and a 7 percent
real discount rate. The annual
net benefits of the proposal in
2025 are approximately $31
million using a 3 percent real
discount rate and
approximately $4 million
using a 7 percent real
discount rate.
ES-17
The present value
(PV) of the net benefits, in
2016$ and discounted to
2021, is -$68 million when
using a 7 percent
discount rate and $14 million
when using a 3 percent
discount rate. The equivalent
annualized
value (EAV), an estimate of
the annualized value of the
net benefits consistent with
the present
value, is -$17 million per year
when using a 7 percent
The present value
(PV) of the net benefits, in
2016$ and discounted to
2021, is -$59 million when
using a 7 percent
discount rate and $23 million
when using a 3 percent
discount rate. The equivalent
annualized
value (EAV), an estimate of
the annualized value of the
net benefits consistent with
the present
value, is -$14 million per year
when using a 7 percent
-------
Page numbers
Current Value
Corrected Value
(Highlighted in yellow)
discount rate and $3 million
when using a
3 percent discount rate.
discount rate and $5 million
when using a
3 percent discount rate.
7-2
As shown in
Chapter 4, the estimated
annual compliance costs to
implement the proposal, as
described in this
document, are approximately
$21 million in 2021 and $6
million in 2025 (2016$).
As shown in
Chapter 4, the estimated
annual compliance costs to
implement the proposal, as
described in this
document, are approximately
$20 million in 2021 and $1
million in 2025 (2016$).
7-3
The annual net benefits of the
proposal in 2021 (in 2016$)
are approximately -$21
million using both a 3 percent
and 7 percent
real discount rate for the
climate benefits. The annual
net benefits of the proposal in
2025 are
approximately $27 using a 3
percent real discount rate and
-$0.9 million using a 7
percent real
discount rate.
The annual net benefits of the
proposal in 2021 (in 2016$)
are approximately -$20
million using both a 3 percent
and 7 percent
real discount rate for the
climate benefits. The annual
net benefits of the proposal in
2025 are
approximately $31 using a 3
percent real discount rate and
$4 million using a 7 percent
real
discount rate.
7-5
The present value
(PV) of the net benefits, in
2016$ and discounted to
2021, is -$68 million when
using a 7 percent
discount rate and $14 million
when using a 3 percent
discount rate. The equivalent
annualized
value (EAV), an estimate of
the annualized value of the
net benefits consistent with
the present
value, is -$17 million per year
when using a 7 percent
discount rate and $3 million
when using a
3 percent discount rate.
The present value
(PV) of the net benefits, in
2016$ and discounted to
2021, is -$59 million when
using a 7 percent
discount rate and $23 million
when using a 3 percent
discount rate. The equivalent
annualized
value (EAV), an estimate of
the annualized value of the
net benefits consistent with
the present
value, is -$14 million per year
when using a 7 percent
discount rate and $5 million
when using a
3 percent discount rate.
-------
Location: Page ES-10
Current Table:
Table ES-1. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
19.4
80.6
1.6
2021 (Annual)
20.9
37.2
3.8
2025 (Annual)
6.3
132.2
-12.0
The 2021-2025 (Annualized) row reflects total estimated annual compliance costs levelized over
the period 2021 through 2025, discounted using a 4.25 real discount rate. The 2021 (Annual) and
2025 (Annual) rows reflect annual estimates in each of those years.
Corrected Table (Corrections highlighted in yellow):
Table ES-2. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
17.4
80.6
1.6
2021 (Annual)
20.5
37.2
3.8
2025 (Annual)
1.5
132.2
-12.0
The 2021-2025 (Annualized) row reflects total estimated annual compliance costs levelized over
the period 2021 through 2025, discounted using a 4.25 real discount rate. The 2021 (Annual) and
2025 (Annual) rows reflect annual estimates in each of those years.
-------
Location: Page ES-16
Current Table:
Table ES-7. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
21
-21 +B
7%
0.05 +B
-21 +B
More Stringent
Alternative
3%
0.80 +B
37
-36+B
7%
0.12+B
-37+B
Less Stringent
Alternative
3%
0.17+B
4
-4+B
7% 0.03 +B -4 +B
Corrected Table (Corrections highlighted in yellow):
Table ES-7. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
20
-20+B
7%
0.05 +B
-20+B
More Stringent
Alternative
3%
0.80 +B
37
-36+B
7%
0.12+B
-37+B
Less Stringent
Alternative
3%
0.17+B
4
-4+B
7%
0.03 +B
-4+B
-------
Location: Pages ES16-17
Current Table:
Table ES-8. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
33 +B
6
27+B
7%
5.4+B
-0.9 +B
More Stringent
Alternative
3%
71.5 +B
132
-61 +B
7%
11.7+B
-120 +B
Less Stringent
Alternative
3%
25 +B
-12
37+B
7% 4.2 +B 16 +B
Corrected Table (Corrections highlighted in yellow):
Table ES-8. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
33 +B
1
31 +B
7%
5.4+B
4+B
More Stringent
Alternative
3%
71.5 +B
132
-61 +B
7%
11.7+B
-120 +B
Less Stringent
Alternative
3%
25 +B
-12
37+B
7%
4.2+B
16+B
-------
Location: Page ES-18
Current Table:
Table ES-9. Summary of Present Values and Equivalent Annualized Values for the 2021-
2025 Timeframe for Estimated Compliance Costs, Climate Benefits, and Net
Benefits for the Proposed Rule (millions of 2016$, discounted to 2021)
3% Discount Rate
7% Discount Rate
Present Value
Benefitscd
ioi+p
15+P
Climate Benefits0
101
15
Compliance Costse
87
83
Net Benefits
14+p
-68+p
Equivalent
Annualized Value
Benefits
22+b
4+b
Climate Benefits
22
4
Compliance Costs
19
20
Net Benefits
3+b
-17+b
Corrected Table (Corrections highlighted in yellow):
3% Discount Rate
7% Discount Rate
Present Value
Benefitscd
ioi+p
15+P
Climate Benefits0
101
15
Compliance Costse
78
74
Net Benefits
23+p
-59+p
Equivalent
Annualized Value
Benefits
22+b
4+b
Climate Benefits
22
4
Compliance Costs
17
18
Net Benefits
5+b
-14+b
-------
Location: Page 4-20
Current Table:
Table 4-6. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Revised
CSAPR
Update
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
19.4
80.6
1.6
2021 (Annual)
20.9
37.2
3.8
2022 (Annual)
29.7
49.2
12.8
2023 (Annual)
27.8
47.3
12.8
2024 (Annual)
6.3
132.2
-12.0
2025 (Annual)
6.3
132.2
-12.0
Corrected Table (Corrections highlighted in yellow):
Table 4-6. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Revised
CSAPR
Update
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
17.4
80.6
1.6
2021 (Annual)
20.5
37.2
3.8
2022 (Annual)
29.6
49.2
12.8
2023 (Annual)
27.7
47.3
12.8
2024 (Annual)
1.5
132.2
-12.0
2025 (Annual)
1.5
132.2
-12.0
-------
Location: Page 7-3
Current Table:
Table 7-1. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
21
-21 + B
7%
0.05 + B
-21 + B
More Stringent
Alternative
3%
0.80+ B
37
-36+ B
7%
0.12 + B
-37+ B
Less Stringent
Alternative
3%
0.17 + B
4
-4 + B
7%
0.03 +B
-4 + B
Corrected Table (Corrections highlighted in yellow):
Table 7-1. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
20
-20+ B
7%
0.05 + B
-20+ B
More Stringent
Alternative
3%
0.80+ B
37
-36+ B
7%
0.12 + B
-37+ B
Less Stringent
Alternative
3%
0.17 + B
4
-4 + B
7%
0.03 +B
-4 + B
-------
Location: Page 7-3
Current Table:
Table 7-2. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
33 +B
6
27+ B
7%
5.4+ B
-0.9+ B
More Stringent
Alternative
3%
71.5 + B
132
-61 + B
7%
11.7 + B
-120+ B
Less Stringent
Alternative
3%
25 + B
-12
37 + B
7%
4.2+ B
16 + B
Corrected Table (Corrections highlighted in yellow):
Table 7-2. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
33 +B
1
31 +B
7%
5.4+ B
4+B
More Stringent
Alternative
3%
71.5 + B
132
-61 + B
7%
11.7 + B
-120+ B
Less Stringent
Alternative
3%
25 + B
-12
37 + B
7%
4.2+ B
16 + B
-------
Location: Page 7-4
Current Table:
Table 7-3. Summary of Present Values and Equivalent Annualized Values for the 2021-
2025 Timeframe for Estimated Compliance Costs, Climate Benefits, and Net
Benefits for the Proposed Rule (millions of 2016$, discounted to 2021)
3% Discount Rate
7% Discount Rate
Present Value
Climate Benefitscd
ioi+p
15+P
Compliance Costse
87
83
Net Benefits
14+p
-68+p
Equivalent
Annualized Value
Climate Benefits
22+b
4+b
Compliance Costs
19
20
Net Benefits
3+b
-17+b
Corrected Table (Corrections highlighted in yellow):
Table 7-3. Summary of Present Values and Equivalent Annualized Values for the 2021-
2025 Timeframe for Estimated Compliance Costs, Climate Benefits, and Net
Benefits for the Proposed Rule (millions of 2016$, discounted to 2021)
3% Discount Rate
7% Discount Rate
Present Value
Climate Benefitscd
ioi+p
15+P
Compliance Costse
78
74
Net Benefits
23+p
-59+p
Equivalent
Annualized Value
Climate Benefits
22+b
4+b
Compliance Costs
17
18
Net Benefits
5+b
-14+b
-------
ii
-------
EPA-452/P-20-003
October 2020
Regulatory Impact Analysis for the Proposed Revised Cross-State Air Pollution Rule (CSAPR)
Update for the 2008 Ozone NAAQS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, NC
in
-------
CONTACT INFORMATION
This document has been prepared by personnel from the Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency. Questions related to this document should be
addressed to Robin Langdon, U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, C439-02, Research Triangle Park, North Carolina 27711 (email:
langdon.robin@epa.gov).
ACKNOWLEDGEMENTS
In addition to EPA staff from the Office of Air Quality Planning and Standards, personnel from
the Office of Atmospheric Programs and the Office of Policy's National Center for
Environmental Economics contributed data and analysis to this document.
iv
-------
TABLE OF CONTENTS
LIST OF TABLES ix
LIST OF FIGURES xiii
EXECUTIVE SUMMARY ES-1
Overview ES-1
ES. 1 Identifying Needed Emissions Reductions and Description of the Remedy ES-2
ES.2 Baseline and Analysis Years ES-5
ES.3 Emissions and Air Quality Modeling ES-6
ES.4 Control Strategies and Emissions Reductions ES-7
ES.5 Cost Impacts ES-9
ES.6 Benefits ES-10
ES.6.1 Climate Benefits Estimates ES-12
ES.6.2 Unquantified Health and Welfare Benefits Categories ES-13
ES.6.3 Approach for Updating Health Effects from PM2.5 and Ozone ES-14
ES.7 Results of Benefit-Cost Analysis ES-15
CHAPTER 1: INTRODUCTION AND BACKGROUND 1-1
Overview 1-1
1.1 Background 1-3
1.1.1 Role of Executive Orders in the Regulatory Impact Analysis 1-4
1.1.2 Alternatives Analyzed 1-5
1.1.3 The Need for Air Quality or Emissions Regulation 1-5
1.2 Overview and Design of the RIA 1-6
1.2.1 Methodology for Identifying Needed Reductions 1-6
1.2.2 States Covered by the Proposed Rule 1-8
1.2.3 Regulated Entities 1-9
1.2.4 Baseline and Analysis Years 1-9
1.2.5 Emissions Controls, Emissions, and Cost Analysis Approach 1-11
1.2.6 B enefits Analy si s Approach 1-11
1.3 Organization of the Regulatory Impact Analysis 1-11
CHAPTER 2: ELECTRIC POWER SECTOR PROFILE 2-1
Overview 2-1
2.1 Background 2-1
2.2 Power Sector Overview 2-2
2.2.1 Generation 2-2
v
-------
2.2.2 Transmission 2-9
2.2.3 Distribution 2-10
2.3 Sales, Expenses, and Prices 2-10
2.3.1 Electricity Prices 2-11
2.3.2 Prices of Fossil Fuels Used for Generating Electricity 2-15
2.3.3 Changes in Electricity Intensity of the U.S. Economy from 2014 to 2018 2-15
2.4 Deregulation and Restructuring 2-17
CHAPTER 3: EMISSIONS AND AIR QUALITY IMPACTS 3-1
Overview 3-1
3.1 ACE Air Quality Modeling Platform 3-1
3.2. Applying Modeling Outputs to Create Spatial Fields 3-3
3.3 Application of ACE Approach for the Revised CSAPR Update 3-7
3.4 Spatial Distribution of Air Quality Impacts 3-9
3.5 Uncertainties and Limitations of ACE Approach 3-14
3.6 References 3-16
APPENDIX 3A: METHODOLOGY FOR DEVELOPING AIR QUALITY SURFACES
3A-1
3 A. 1 Air Quality Modeling Platform for the ACE Rule 3 A-l
3A. 1.1 Air Quality Model, Meteorology and Boundary Conditions 3A-l
3 A. 1.2 2011 and 2023 Emissions 3A-3
3 A. 1.3 2011 Model Evaluation for Ozone and PM2.5 3 A-7
3A.2 Source Apportionment Tags 3A-l 1
3A.3 Applying Source Apportionment Contributions to Create Air Quality Fields 3A-13
3A.3.2 Scaling Ratio Applied to Source Apportionment Tags 3A-13
3A.4 Creating Fused Fields Based on Observations and Model Surfaces 3A-16
3 A. 5 References 3 A-19
CHAPTER 4: COST, EMISSIONS, AND ENERGY IMPACTS 4-1
Overview 4-1
4.1 Regulatory Control Alternatives 4-1
4.1.2 Regulatory Control Alternatives Analyzed 4-2
4.2 Power Sector Modeling Framework 4-4
4.3 EPA's Power Sector Modeling of the Base Case and Three Regulatory Control
Alternatives 4-7
4.3.1 EPA'sIPM Base Case v.6 4-7
4.3.2. Methodology for Evaluating the Regulatory Control Alternatives 4-8
vi
-------
4.3.3 Methodology for Estimating Compliance Costs 4-13
4.4 Estimated Impacts of the Regulatory Control Alternatives 4-15
4.4.1 Emission Reduction Assessment 4-15
4.4.2 Impact of Emissions Reductions on Maintenance and Nonattainment Monitors 4-19
4.4.3 Compliance Cost Assessment 4-20
4.4.4 Impacts on Fuel Use, Prices and Generation Mix 4-22
4.5 Social Costs 4-28
4.6 Limitations 4-29
4.7 References 4-30
CHAPTER 5: BENEFITS 5-1
Overview 5-1
5.1 Estimated Climate Benefits from Reducing CO2 5-1
5.2 Unquantified Benefits 5-6
5.2.1 Ozone Health Benefits 5-15
5.2.2 PM2.5 Health Benefits 5-16
5.2.3 NO; Health Benefits 5-16
5.2.4 Ozone Welfare Benefits 5-17
5.2.5 NO2 Welfare Benefits 5-17
5.2.6 Visibility Impairment Benefits 5-18
5.3 References 5-19
APPENDIX 5A: UNCERTAINTY ASSOCIATED WITH ESTIMATING THE SOCIAL
COST OF CARBON 5A-1
Overview 5A-1
5A. 1 Overview of 2009 Endangerment Finding and Climate Science Assessments 5A-1
5A.2 Overview of Methodology Used to Develop Interim Domestic SC-C02 Estimates. 5A-3
5A.3 Treatment of Uncertainty in Interim Domestic SC-CO2 Estimates 5A-4
5A.4 Global Climate Benefits 5A-9
5A. 5 References 5A-10
APPENDIX 5B: AIR POLLUTION-RELATED HUMAN HEALTH BENEFITS
ESTIMATED USING PREVIOUS METHODS 5B-1
Overview 5B-1
5B. 1 Estimated Human Health Benefits 5B-1
5B. 1.1 Health Impact Assessment for Ozone and PM2.5 5B-3
5B. 1.1.1 Selecting Air Pollution Health Endpoints to Quantify 5B-3
vii
-------
5B. 1.1.2 Calculating Counts of Air Pollution Effects Using the Health Impact Function
5B-6
5B. 1.1.3 Quantifying Cases of Ozone-Attributable Premature Death 5B-8
5B. 1.1.4 Quantifying Cases of PM2.5-Attributable Premature Death 5B-9
5B. 1.2 Economic Valuation Methodology for Health Benefits 5B-10
5B.1.3 Characterizing Uncertainty in the Estimated Benefits 5B-12
5B. 1.4 Estimated Number and Economic Value of Health Benefits 5B-16
5B.2 References 5B-25
CHAPTER 6: STATUTORY AND EXECUTIVE ORDER REVIEWS 6-1
Overview 6-1
6.1 Executive Order 12866: Regulatory Planning and Review 6-1
6.2 Executive Order 13771 6-1
6.3 Paperwork Reduction Act 6-1
6.4 Regulatory Flexibility Act 6-1
6.4.1 Identification of Small Entities 6-3
6.4.2 Overview of Analysis and Results 6-7
6.4.2.1 Methodology for Estimating Impacts of the Revised CSAPR Update proposal on
Small Entities 6-7
6.4.2.2 Results 6-9
6.4.3 Summary of Small Entity Impacts 6-11
6.5 Unfunded Mandates Reform Act 6-11
6.6 Executive Order 13132: Federalism 6-12
6.7 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments 6-12
6.8 Executive Order 13045: Protection of Children from Environmental Health & Safety
Risks 6-13
6.9 Executive Order 13211: Actions that Significantly Affect Energy Supply, Distribution,
or Use 6-13
6.10 National Technology Transfer and Advancement Act 6-14
6.11 Executive Order 12898: Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations 6-14
CHAPTER 7: COMPARISON OF BENEFITS AND COSTS 7-1
Overview 7-1
7.1 Results 7-2
viii
-------
LIST OF TABLES
Table ES-1. NOx Ozone Season Emission Budgets (Tons) Evaluated ES-3
Table ES-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives ES-8
Table ES-3. Estimated 2021 and 2025 EGU Emissions Reductions in the 12 States of NOx, SO2,
and CO2 and More and Less Stringent Alternatives (Tons) ES-8
Table ES-4. National Compliance Cost Estimates (millions of 2016$) for the Regulatory Control
Alternatives ES-10
Table ES-5. Estimated Domestic Climate Benefits from Changes in CO2 Emissions for Selected
Years (Millions of 2016$) ES-12
Table ES-6. Estimated Total Annualized Domestic Climate Benefits (2021-25) from Changes in
CO2 Emissions (Millions of 2016$) ES-13
Table ES-7. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$) ES-16
Table ES-8. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$) ES-16
Table ES-9. Summary of Present Values and Equivalent Annualized Values for the 2021-2025
Timeframe for Estimated Compliance Costs, Climate Benefits, and Net Benefits for the
Proposed Rule (millions of 2016$, discounted to 2021) ES-18
Table 2-1. Total Net Summer Electricity Generating Capacity by Energy Source, 2014 and 2018
2-3
Table 2-2. Net Generation in 2014 and 2018 (Trillion kWh = TWh) 2-6
Table 2-3. Coal and Natural Gas Generating Units, by Size, Age, Capacity, and Average Heat
Rate in 2018 2-7
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2014 and 2018 (billion kWh) 2-11
Table 3A.1. Model Performance Statistics by Region for PM2.5 3A-10
Table 3A.2. Model Performance Statistics by Region for Ozone on Days Above 60 ppb (May-
Sep) 3 A-11
Table 3A.3. Source Apportionment Tags 3A-12
Table 4-1. NOx Ozone Season Emission Budgets (Tons) Evaluated 4-3
IX
-------
Table 4-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory Control
Alternatives 4-11
Table 4-3. Summary of Methodology for Calculating Compliance Costs Estimated Outside of
IPM for Revised CSAPR Update Proposal, 2021 (2016$) 4-14
Table 4-4. EGU Ozone Season NOx Emissions and Emissions Changes (thousand tons) for the
Base Case and the Regulatory Control Alternatives 4-16
Table 4-5. EGU Annual Emissions and Emissions Changes for NOx, SO2, PM2.5, and CO2 for
the Regulatory Control Alternatives 4-17
Table 4-6. National Compliance Cost Estimates (millions of 2016$) for the Regulatory Control
Alternatives 4-20
Table 4-7. 2021 Projected U.S. Power Sector Coal Use for the Base Case and the Regulatory
Control Alternatives 4-23
Table 4-8. 2021 Projected U.S. Power Sector Natural Gas Use for the Base Case and the
Regulatory Control Alternatives 4-23
Table 4-9. 2021 Projected Minemouth and Power Sector Delivered Coal Price for the Base Case
and the Regulatory Control Alternatives 4-23
Table 4-10. 2021 Projected Henry Hub and Power Sector Delivered Natural Gas Price for the
Base Case and the Regulatory Control Alternatives 4-24
Table 4-11. 2021 Projected U.S. Generation by Fuel Type for the Base Case and the Regulatory
Control Alternatives 4-24
Table 4-12. 2021 Projected U.S. Capacity by Fuel Type for the Base Case and the Regulatory
Control Alternatives 4-25
Table 4-13. Average Retail Electricity Price by Region for the Base Case and the Regulatory
Control Alternatives, 2021 4-26
Table 4-14. Average Retail Electricity Price by Region for the Base Case and the Regulatory
Control Alternatives, 2025 4-27
Table 5-1. Interim Domestic Social Cost of Carbon Values (2016$/Metric Tonne CO2) 5-3
Table 5-2. Estimated Domestic Climate Benefits from Changes in CO2 Emissions 2021 - 2025
(Millions of 2016$) 5-5
Table 5-3. Estimated Total Annualized Domestic Climate Benefits (2021-25) from Changes in
CO2 Emissions (Millions of 2016$) 5-6
Table 5-4. Unquantified Health and Welfare Benefits Categories 5-8
x
-------
Table 5-5. Estimated Summary of Causality Determination for each Ozone-Related Endpoint
5-14
Table 5-6. Summary of Causality Determination for each PM2.5-Related Endpoint 5-15
Table 5A-1. Climate Effects 5A-2
Table 5B-1. Health Effects of Ambient Ozone and PM2.5 5B-6
Table 5B-2. Estimated Avoided Ozone-Related Premature Deaths and Illnesses for the Proposal
and More and Less Stringent Alternatives for 2021 (95% Confidence Interval) 5B-18
Table 5B-3. Estimated Avoided PM2.5 and Ozone-Related Premature Deaths and Illnesses for the
Proposal and More and Less Stringent Alternatives for 2025 (95% Confidence Interval)
5B-19
Table 5B-4. Estimated Value of Avoided Ozone-Related Premature Deaths and Illnesses for the
Proposal and More and Less Stringent Alternatives for 2021 (95% Confidence Interval;
millions of 2016$fb 5B-20
Table 5B-5. Estimated Value of Avoided PM2.5 and Ozone-Related Premature Deaths and
Illnesses for the Proposal and More and Less Stringent Alternatives for 2025 (95%
Confidence Interval; millions of 2016$) 5B-21
Table 5B-6. Estimated Avoided PM-Related Premature Deaths Using Alternative Approaches
Using Two Approaches to Quantifying Avoided PM-Attributable Deaths (95%
Confidence Interval) in 2025 5B-22
Table 5B-7. Estimated Economic Value of Ozone-Attributable Deaths and Illnesses for the
Proposed Policy Scenarios in 2021 (95% Confidence Interval; millions of 2016$)
5B-23
Table 5B-8. Estimated Economic Value of Avoided Ozone and PM2.5-Attributable Deaths and
Illnesses for the Proposed Policy Scenario Using Alternative Approaches to Represent
PM2.5 Mortality Risk Effects in 2025 (95% Confidence Interval; millions of 2016$)
5B-24
Table 5B-9. Estimated Percent of Avoided PM2.5-related Premature Deaths Above and Below
PM2.5 Concentration Cut Points in 2025 5B-25
Table 6-1. SBA Size Standards by NAICS Code 6-6
Table 6-2. Projected Impact of the Revised CSAPR Update Proposal on Small Entities in 2021
6-10
Table 6-3. Incremental Annual Costs under the Revised CSAPR Update Proposal Summarized
by Ownership Group and Cost Category in 2021 (2016$ millions) 6-11
XI
-------
Table 7-1. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$) 7-3
Table 7-2. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$) 7-3
Table 7-3. Summary of Present Values and Equivalent Annualized Values for the 2021-2025
Timeframe for Estimated Compliance Costs, Climate Benefits, and Net Benefits for the
Proposed Rule (millions of 2016$, discounted to 2021) 7-4
xii
-------
LIST OF FIGURES
Figure 2-1. National New Build and Retired Capacity (MW) by Fuel Type, 2014-2018 2-4
Figure 2-2. Regional Differences in Generating Capacity (MW), 2018 2-5
Figure 2-3. Cumulative Distribution in 2018 of Coal and Natural Gas Electricity Capacity and
Generation, by Age 2-8
Figure 2-4. Fossil Fuel-Fired Electricity Generating Facilities, by Size 2-9
Figure 2-5. Real National Average Electricity Prices (including taxes) for Three Major End-Use
Categories 2-12
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for Major End-
Use Categories (including taxes), With Inflation Indices 2-13
Figure 2-7. Real National Average Electricity Prices for Three Major End-Use Categories
(including taxes), 1960-2018 (2018$) 2-14
Figure 2-8. Relative Change in Real National Average Electricity Prices (2018$) for Three Major
End-Use Categories (including taxes) 2-14
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in National
Average Real Price per MMBtu Delivered to EGU 2-15
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since 2014
2-16
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation Intensity
Since 2014 2-17
Figure 2-12. Status of State Electricity Industry Restructuring Activities 2-18
Figures 2-13. and 2-14. Capacity and Generation Mix by Ownership Type, 2014 & 2018 2-20
Figure 3-1. Air Quality Modeling Domain 3-2
Figure 3-2. Map of change in May-September MDA8 ozone (ppb): 2021 baseline - less
stringent regulatory alternative (scale: + 0.10 ppb) 3-10
Figure 3-3. Map of change in May-September MDA8 ozone (ppb): 2021 baseline - proposal
(scale: + 0.50 ppb) 3-10
Figure 3-4. Map of change in May-September MDA8 ozone (ppb): 2021 baseline - more
stringent regulatory alternative (scale: + 0.50 ppb) 3-11
Figure 3-5. Map of change in May-September MDA8 ozone (ppb): 2025 baseline - less
stringent regulatory alternative (scale: + 0.10 ppb) 3-11
Xlll
-------
Figure 3-6. Map of change in May-September MDA8 ozone (ppb): 2025 baseline - proposal
(scale: + 0.50 ppb) 3-12
Figure 3-7. Map of change in May-September MDA8 ozone (ppb): 2025 baseline - more
stringent regulatory alternative (scale: + 0.50 ppb) 3-12
Figure 3-8. Map of change in annual mean PM2.5 (]ag/m3): 2025 baseline - less stringent
regulatory alternative (scale: + 0.01 iag/m3) 3-13
Figure 3-9. Map of change in annual mean PM2.5 (iag/m3): 2025 baseline - proposal (scale: +
0.01 :;g/m3) 3-13
Figure 3-10. Map of change in annual mean PM2.5 (]ig/m3): 2025 baseline - more stringent
regulatory alternative (scale: + 0.01 iag/m3) 3-14
Figure 3A-1. Air Quality Modeling Domain 3A-2
Figure 3 A-2. NOAA Climate Regions 3 A-9
Figure 4-1. Electricity Market Module Regions 4-28
Figure 5A-1. Frequency Distribution of Interim Domestic SC-CO2 Estimates for 2030 (in 2016$
per Metric Ton CO2) 5A-7
Figure 5B-1. Stylized Relationship between the PM2.5 Concentrations Considered in
Epidemiology Studies and our Confidence in the Estimated PM-related Premature Deaths
5B-14
Figure 5B-2. Estimated Percentage of PM2.5-Related Deaths and Number of Individuals Exposed
by Annual Mean PM2.5 Level in 2025 5B-16
xiv
-------
EXECUTIVE SUMMARY
Overview
This proposed action is taken in response to the United States Court of Appeals for the
District of Columbia Circuit's (D.C. Circuit) September 13, 2019 remand of the Cross-State Air
Pollution Rule (CSAPR) Update. The CSAPR Update finalized Federal Implementation Plans
(FIPs) for 22 states to address their interstate pollution-transport obligations under the Clean Air
Act (CAA) for the 2008 ozone National Ambient Air Quality Standards (NAAQS).1 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. This proposed
rule, if finalized, will resolve 21 states' outstanding interstate ozone transport obligations with
respect to the 2008 ozone NAAQS. 2
This action proposes to find that for 9 of the 21 states with remanded FIPs (Alabama,
Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Texas, and Wisconsin), their
projected 2021 ozone season nitrogen oxides (NOx) emissions do not significantly contribute to
a continuing downwind nonattainment and/or maintenance problem; therefore the CSAPR
Update fully addresses their interstate ozone transport obligations with respect to the 2008 ozone
NAAQS. This action also proposes to find 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 NOx emissions significantly
contribute to downwind states' nonattainment and/or maintenance problems for the 2008 ozone
NAAQS.
EPA is proposing the creation of an additional geographic group and ozone season
trading program comprised of these 12 upwind states with remaining linkages to downwind air
1 The 2008 ozone NAAQS is an 8-hour standard that was set at 75 parts per billion (ppb). See 73 FR 16436 (March
27, 2008).
2 In the CSAPR Update, EPA found that the finalized Tennessee emissions budget fully addressed Tennessee's good
neighbor obligation with respect to the 2008 ozone NAAQS. As such, Tennessee is not considered in this proposal,
and the number of states included is reduced from 22 to 21 states.
ES-1
-------
quality problems in 2021. This new group, Group 3, will be covered by a new CSAPRNOx
Ozone Season (May 1 - September 30) Group 3 trading program and will no longer be subject to
Group 2 budgets. Aside from the removal of the 12 covered states from the current Group 2
program, this proposal leaves unchanged the budget stringency and geography of the existing
CSAPRNOx Ozone Season Group 1 and Group 2 trading programs. The electric generating
units (EGUs) covered by the FIPs and subject to the budget are all fossil-fired EGUs with >25
megawatt (MW) capacity.
ES.l Identifying Needed Emissions Reductions and Description of the Remedy
To reduce interstate emission transport under the authority provided in CAA section
110(a)(2)(D)(i)(I), this rule proposes to further limit ozone season NOx emissions from EGUs in
12 states using the same framework used by EPA in developing the CSAPR (the interstate
transport framework). The interstate transport framework provides a 4-step process to address
the requirements of the good neighbor provision for ground-level ozone and fine particulate
matter (PM2.5) 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 1 percent contribution threshold); (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 emissions reductions through enforceable
measures. In this proposed action, EPA applies this 4-step interstate transport framework to
respond to the D.C. Circuit's remand and revise the CSAPR Update with respect to the 2008
ozone NAAQS.
The remedy that emerges from the 4-step interstate transport framework is state
emissions budgets implemented as a cap-and-trade program. This RIA evaluates how the EGUs
covered by the proposed rule are expected to reduce their emissions in response to the
requirements and flexibilities provided by the remedy implemented by the proposed Revised
CSAPR Update and the benefits, costs, and impacts of doing so. The proposed rule sets EGU
ozone season NOx emissions budgets (allowable emission levels) for 2021 and future years. EPA
ES-2
-------
proposes to implement these reductions through FIPs in any state that does not have an approved
good neighbor SIP by the date this proposal is finalized. Furthermore, under the FIPs, affected
EGUs would participate in the CSAPRNOx ozone-season allowance trading program. The
allowance trading program essentially converts the EGU NOx emissions budget for each of the
12 states subject to the FIP into a limited number of NOx ozone-season allowances that, on a ton
basis, equal the state's ozone season emissions budget. Starting in 2021, emissions from affected
EGUs in the 12 states cannot exceed the sum of emissions budgets but for the ability to use
banked allowances from previous years for compliance. No further reductions in budgets occur
after 2025, and budgets remain in place for future years. Furthermore, emissions from affected
EGUs in a particular state are subject to the CSAPR assurance provisions, which require
additional allowance surrender penalties (a total of 3 allowances per ton of emissions) on
emissions that exceed a state's CSAPRNOx ozone season assurance level, or 121 percent of the
states' emissions budget. Similar to the approach taken in the CSAPR Update, EPA is proposing
a one-time conversion of banked Group 2 allowances according to a formula that ensures that
emissions in the Group 3 trading program region in the first year of the program do not exceed a
specified level (defined as emissions up to the sum of the states' seasonal emissions budgets and
variability limits) as a result of the use of banked allowances from the Group 2 trading program.
For the proposed Revised CSAPR Update, the EGU ozone season NOx budgets for each
state reflect EGU NOx reduction strategies that are widely available at a uniform cost of $1,600
per ton (2016$) of NOx for affected EGUs.3 Specifically, this uniform cost reflects turning on
idled SCR and installing state-of-the-art combustion controls. Furthermore, this RIA analyzes
regulatory control alternatives based on more and less stringent state emissions budgets based on
uniform NOx control costs of $9,600 per ton and $500 per ton, respectively. Table ES-1 shows
the EGU NOx ozone season emission budgets that are evaluated in this RIA.
Table ES-1. NOx Ozone Season Emission Budgets (Tons) Evaluated
Revised CSAPR Update Proposal
State 2021 2022 2023 2024 2025
Illinois 9,444 9,415 8,397 8,397 8,397
Indiana 12,500 11,998 11,998 9,447 9,447
3 For details, please see EGU NOx Mitigation Strategies Proposed Rule TSD available in the docket for this
proposed rule.
ES-3
-------
Kentucky
14,384
11,936
11,936
11,936
11,936
Louisiana
15,402
14,871
14,871
14,871
14,871
Maryland
1,522
1,498
1,498
1,498
1,498
Michigan
12,727
11,767
9,803
9,614
9,614
New Jersey
1,253
1,253
1,253
1,253
1,253
New York
3,137
3,137
3,137
3,119
3,119
Ohio
9,605
9,676
9,676
9,676
9,676
Pennsylvania
8,076
8,076
8,076
8,076
8,076
Virginia
4,544
3,656
3,656
3,395
3,395
West
Virginia
13,686
12,813
11,810
11,810
11,810
Total
106,280
100,096
96,111
93,092
93,092
Less-Stringent Alternative
State
2021
2022
2023
2024
2025
Illinois
9,667
9,632
8,579
8,599
8,579
Indiana
15,677
15,206
15,206
12,755
12,603
Kentucky
15,606
15,606
15,606
15,588
15,606
Louisiana
15,442
15,442
15,442
15,488
15,442
Maryland
1,565
1,565
1,565
1,565
1,565
Michigan
13,120
13,120
10,313
10,841
10,116
New Jersey
1,346
1,346
1,346
1,346
1,346
New York
3,182
3,182
3,182
3,169
3,163
Ohio
15,490
15,560
15,560
15,917
15,560
Pennsylvania
11,487
11,487
11,487
11,570
11,487
Virginia
4,588
4,172
4,172
3,912
3,908
West
Virginia
15,017
15,017
13,272
13,407
13,272
Total
122,187
121,334
115,730
114,156
112,647
More-Stringent Alternative
State
2021
2022
2023
2024
2025
Illinois
9,444
9,415
8,397
7,142
7,142
Indiana
12,500
11,998
11,998
8,264
8,264
Kentucky
14,384
11,936
11,936
8,852
8,852
Louisiana
15,402
14,871
14,871
12,636
12,636
Maryland
1,522
1,498
1,498
1,239
1,239
Michigan
12,727
11,767
9,803
7,315
7,315
New Jersey
1,253
1,253
1,253
1,257
1,257
New York
3,137
3,137
3,137
3,020
3,020
Ohio
9,605
9,676
9,676
9,126
9,126
Pennsylvania
8,076
8,076
8,076
7,578
7,578
ES-4
-------
Virginia
West
Virginia
4,544 3,656 3,656 3,022 3,022
13,686 12,813 11,810 9,569 9,569
Total
106,280 100,096 96,111 79,020 79,020
ES.2 Baseline and Analysis Years
The proposal sets forth the requirements for states to reduce states' significant
contribution to downwind nonattainment or interference with maintenance of the 2008 ozone
NAAQS. To develop and evaluate control strategies for addressing these obligations, it is
important to first establish a baseline projection of air quality and electricity sector and related
fuel market conditions in the analysis year of 2021, taking into account currently on-the-books
Federal regulations, substantial Federal regulatory proposals, enforcement actions, state
regulations, population, expected electricity demand growth, and where possible, economic
growth.4 Establishing this baseline for the analysis then allows us to estimate the incremental
costs and benefits of the additional emissions reductions that will be achieved by the proposal.
The analysis in this RIA focuses on benefits, costs and certain impacts in both 2021 and
2025. We focus on 2021 because it is by the 2021 ozone season, corresponding with the 2021
Serious area attainment date, that significant contribution from upwind states' must be
eliminated to the extent possible. It is also the first year in which some EGU NOx mitigation
technologies are available. In addition, impacts for 2023 to 2025 are important as these years
reflect the next model years in which additional NOx mitigation technologies are first available.
Presenting benefits, costs, and certain impacts in 2025 reflects the time needed to make
these retrofits on a regional scale and reflects full implementation of the proposed policy.
Additional benefits and costs are expected to occur after 2025 as EGUs subject to this proposal
4 The technical support document (TSD) for the 2016vl emissions modeling platform titled Preparation of
Emissions Inventories for 2016vl North American Emissions Modeling Platform is included in the docket for this
proposed rule. The TSD includes additional discussion on mobile source rules included in the baseline. The future
year onroad emission factors account for changes in activity data and the impact of on-the-books rules that are
implemented into MOVES2014b. These rules include the Light Duty Vehicle GHG Rule for Model-Year 2017-2025
and the Tier 3 Motor Vehicle Emission and Fuel Standards Rule. Local inspection and maintenance (I/M) and other
onroad mobile programs are included, such as California LEVIII, the National Low Emissions Vehicle (LEV) and
Ozone Transport Commission (OTC) LEV regulations, local fuel programs, and Stage II refueling control programs.
Regulations finalized after the year 2014 are not included, such as the Safer Affordable Fuel Efficient (SAFE)
Vehicles Final Rule for Model Years 2021-2026 and the Final Rule for Phase 2 Greenhouse Gas Emissions
Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles.
ES-5
-------
continue to comply with the tighter allowance budget, which is below their baseline emissions.
Because EPA did not estimate costs and benefits beyond 2025, the full costs and benefits of the
proposed policy may be understated in this RIA.
ES.3 Emissions and Air Quality Modeling
The air quality spatial fields for this proposal were constructed using the method and air
quality modeling data developed to support the regulatory impact analysis (RIA) for the Repeal
of the Clean Power Plan, and the Emission Guidelines for Greenhouse Gas Emissions from
Existing Electric Utility Generating Units (U.S. EPA 2019), also referred to the Affordable Clean
Energy (ACE) rule.5 The foundational data from the ACE approach includes the ozone
contributions from EGU emissions in each state based on the 2023 ACE EGU state-sector sector
contribution modeling and the 2023 emissions for coal and non-coal fired EGUs that were input
to that modeling.6
The air quality modeling used in the ACE analysis included annual model simulations for
a 2011 base year and a 2023 future year to provide hourly concentrations of ozone and primary
and secondarily formed PM2.5 component species (e.g., sulfate, nitrate, ammonium, elemental
carbon (EC), organic aerosol (OA), and crustal material7) for both years nationwide. The
photochemical modeling results for 2011 and 2023, in conjunction with modeling to characterize
the air quality impacts from groups of emissions sources (i.e., source apportionment modeling)
and emissions data for the baseline and regulatory control alternatives, were used to construct the
air quality spatial fields that reflect the influence of emissions changes between the baseline and
the regulatory control alternatives.
The air quality model simulations (i.e., model runs) were performed using the
Comprehensive Air Quality Model with Extensions (CAMx) (Ramboll Environ 2016). Our
CAMx nationwide modeling domain (i.e., the geographic area included in the modeling) covers
5 Additional details on the ACE modeling and methodology for developing spatial fields of air quality for EGU
control strategies are provided in Appendix 3A.
6 The 2023 emissions used for the ACE modeling were derived from the 2011-based emissions platform whereas the
emissions used in the air quality modeling to project ozone design values and contributions for this proposed rule
were based on the more recent 2016 platform.
7 Crustal material refers to metals that are commonly found in the earth's crust such as Aluminum, Calcium, Iron,
Magnesium, Manganese, Potassium, Silicon, Titanium and the associated oxygen atoms.
ES-6
-------
all lower 48 states plus adjacent portions of Canada and Mexico using a horizontal grid
resolution of 12 x 12 km.
To potentially calculate ozone-related benefits in 2021 and 2025 EPA applied the ACE
approach using as input the ozone season EGU NOx emissions (tons) for the 2021 and 2025
baseline along with emissions for the proposal and each of the two other regulatory control
alternatives. These emissions were applied using the ACE approach and source apportionment
data to produce spatial fields of the May-September seasonal average MDA8 ozone and the
April-October seasonal average MDA1 ozone concentrations as described in Chapter 3.8 The
emissions of SO2 and directly emitted PM2.5 in 2021 and 2025 for each of the regulatory
alternatives do not change from the 2025 baseline.
ES.4 Control Strategies and Emissions Reductions
Before undertaking power sector analysis to evaluate compliance with the regulatory
control alternatives, EPA first considered available EGU NOx mitigation strategies that could be
implemented for the upcoming ozone season (i.e., the 2021 ozone season). EPA considered all
widely-used EGU NOx control strategies: optimizing NOx removal by existing, operational
selective catalytic reduction (SCRs) and turning on and optimizing existing idled SCRs;9 turning
on existing idled selective non-catalytic reduction (SNCRs); installation of (or upgrading to)
state-of-the-art NOx combustion controls; shifting generation to units with lower NOx emission
rates; and installing new SCRs and SNCRs. Similarly, as proposed, EPA determined that the
power sector could implement all of these NOx mitigation strategies, except installation of new
SCRs or SNCRs, for the 2021 ozone-season.
The EGU NOx mitigation strategies that are assumed to operate or are available to reduce
NOx in order to comply with each of the regulatory control alternatives are shown in Table ES-2.
8 MDA8 is defined as maximum daily 8-hour average ozone concentration, and MDA1 is defined as the maximum
daily 1-hour ozone concentration.
9 Units may choose to idle SCRs in order to avoid fixed operation and maintenance (FOM) and variable operation
and maintenance (VOM) costs such as auxiliary fan power, catalyst costs, and additional administrative costs
(labor), depending on the prevailing CSAPR allowance price for those units otherwise not required to attain a NOx
emission rate that would require operating their SCRs more intensively.
ES-7
-------
Table ES-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives
Regulatory Control
Alternative
NOx Controls Implemented
Less Stringent Alternative (1) Shift generation to minimize costs (costs estimated within IPM)
(All controls above)
(2) Fully operating existing SCRs to achieve 0.08 lb/MMBtu NOx emission rate
(costs estimated outside IPM)
(3) Turn on idled SCRs (costs estimated outside IPM) and fully operate akin to
(2)
(4) Install state of the art combustion controls.
Revised CSAPR Update
Proposed Rule
More Stringent Alternative
(All controls above)
(5) In 2025, turn on idled SNCRs (costs estimated outside IPM)
(6) In 2025, install new SCRs (costs estimated outside IPM)
For the NOx controls identified in Table ES-2, under the proposed rule and the more
stringent alternative, 60 units are projected to fully operate existing SCRs and 4 units are
projected to turn on idled SCRs. Under the less stringent alternative, no units are projected to
either fully operate existing SCRs or turn on idled SCRs. Under the proposed rule and the more
stringent alternative, 27 units are projected to install state-of-the-art combustion controls, and
under the less stringent alternative no units are projected to install state-of-the-art combustion
controls. The book-life of the controls is assumed to be 15 years. Under the proposed rule and
the less stringent alternative, no units are projected to install new SCRs, and under the more
stringent alternative, 48 units are projected to install new SCRs. The book-life of the new SCRs
is assumed to be 15 years. For the final rule, EPA will provide analytic results for the years 2021
through 2040. In addition, EPA will provide information on the stream of costs and, as feasible,
information on the stream of benefits for these analytical years for all scenarios that are both
discounted and undiscounted to provide a more complete picture of the effects estimated to take
place. For additional details, see the EGUNOx Mitigation Strategies Proposed Rule TSD.
Table ES-3 shows the emissions reductions expected from the proposal in 2021 and 2025,
as well as the more and less stringent alternatives analyzed.
Table ES-3. Estimated 2021 and 2025a EGU Emissions Reductions in the 12 States of NOx,
SO2, and CO2 and More and Less Stringent Alternatives (Tons)b'c
2021
Proposal
More Stringent
Alternative
Less Stringent
Alternative
NOx (annual)
17,000
17,000
2,000
NOx (ozone season)
17,000
17,000
2,000
SO2 (annual)
--
--
--
ES-8
-------
2021
Proposal
More Stringent
Alternative
Less Stringent
Alternative
CO2 (annual, thousand metric)
—
—
—
2025
NOx (annual)
"" "(nil i
41,000
"" ()()()
NOx (ozone season)
21,000
35,000
2,000
SO2 (annual)
—
—
—
CO2 (annual, thousand metric)
4,000
10,000
3,000
a The 2021 emissions reductions estimates are based on IPM projections for 2021 and engineering analysis. For
more information, see the Ozone Transport Policy Analysis TSD.
b NOx emissions are reported in English (short) tons; CO2 is reported in metric tons.
0 In addition to no annual SO2 emissions reductions as shown in the table above, there are no annual direct PM2 5
emissions reductions.
The results of EPA's analysis show that, with respect to compliance with the EGU NOx
emission budgets in 2021, maximizing the use of existing operating SCRs provides the largest
amount of ozone season NOx emission reductions (52 percent, affecting 60 units), and turning on
idled SCRs produces an additional 34 percent (affecting 4 units) of the total ozone season NOx
reductions. Generation shifting primarily from coal to gas generation (14 percent) makes up the
remainder of the ozone season NOx reductions.
ES.5 Cost Impacts
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 ES-4. Since the rule does not result in any additional recordkeeping,
monitoring or reporting requirements, the costs associated with compliance with monitoring,
recordkeeping, and reporting requirements are not included within the estimates in this table and
can be found in preamble section VIII.C.6.
There are several notable aspects of the results presented in Table ES-4. The most notable
result is that the estimated annual compliance cost for the less stringent alternative is negative
(i.e., a cost reduction) in 2025, although this regulatory control alternative reduces NOx
emissions by over 2,000 tons as shown in Table ES-3. While seemingly counterintuitive,
estimating negative compliance costs in a single year is possible given the assumption of perfect
foresight. IPM's objective function is to minimize the discounted net present value (NPV) of a
ES-9
-------
stream of annual total cost of generation over a multi-decadal time period. For example, with the
assumption of perfect foresight it is possible that on a national basis within the model the least-
cost compliance strategy may be to delay a new investment or economic retirement that was
projected to occur sooner in the base case. Such a delay could result in a lowering of annual cost
in an early time period and increase it in later time periods. Since the less-stringent alternative is
designed to include only generation shifting, it does not necessitate full operation of existing
controls, or installation of new controls, leading to a negative cost point estimate in 2025.
Table ES-4. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
19.4
80.6
1.6
2021 (Annual)
20.9
37.2
3.8
2025 (Annual)
6.3
132.2
-12.0
The 2021-2025 (Annualized) row reflects total estimated annual compliance costs levelized over the period 2021
through 2025, discounted using a 4.25 real discount rate.10 The 2021 (Annual) and 2025 (Annual) rows reflect
annual estimates in each of those years.
Under the Revised CSAPR Update proposed rule, fully operating existing SCR controls
provides a large share of the total emissions reductions. These options are selected in 2021, while
upgrading to state-of-the-art combustion controls is assumed to begin in 2022. Generation
shifting costs are positive in 2021, but negative in 2025. The result is that the costs in 2021 are
higher than costs in 2025.
ES.6 Benefits
The proposed Revised CSAPR Update is expected to reduce concentrations of ground-
level ozone, PM2.5, and CO2 in the atmosphere (see Chapter 3 for discussion). EPA historically
has used evidence reported in the Integrated Science Assessment (ISA) for the most recent
NAAQS review to inform its approach for quantifying air pollution-attributable health, welfare,
and environmental impacts associated with that pollutant. The ISA synthesizes the
10 This table reports compliance costs consistent with expected electricity sector economic conditions. An NPV of
costs was calculated using a 4.25% real discount rate consistent with the rate used in IPM's objective function for
cost-minimization. The NPV of costs was then used to calculate the levelized annual value over a 5-year period
(2021-2025) using the 4.25% rate as well. Tables ES-9 and 7-3 report the NPV of the annual stream of costs from
2021-2025 using 3% and 7% consistent with OMB guidance.
ES-10
-------
epidemiologic, controlled human exposure and experimental evidence .useful in indicating the
kind and extent of identifiable effects on public health or welfare which may be expected from
the presence of [a] pollutant in ambient air."
The ISA uses a weight of evidence approach to assess the extent to which each criteria
pollutant causes a given health outcome. EPA generally estimates the number and economic
value of the effects for which the ISA identifies the pollutant as having "causal" or "likely to be
causal" relationship. The endpoints for which the 2020 final Ozone ISA and the 2019 final PM
ISA identified as being causal or likely causal differed in some cases from the endpoints for
which those pollutants were identified as being causal or likely causal in the Ozone and PM ISAs
completed for the previous NAAQS reviews (see Chapter 5, Tables 5-5 and 5-6). In addition to
statements of causality, each new ISA identifies an extensive number of epidemiologic studies
that may be suitable for supporting a PM or ozone benefits analysis.11
When updating its approach for quantifying the benefits of changes in PM2.5 and ozone,
the Agency will incorporate evidence reported in these two ISAs and account for forthcoming
recommendations from the Science Advisory Board on this issue. When updating the evidence
for a given endpoint, EPA will consider the extent to which there is a causal relationship,
whether suitable epidemiologic evidence exists to quantify the effect and whether the economic
value of the effect may be estimated. Carefully and systematically reviewing the full breadth of
this information requires significant time and resources. EPA intends to conduct the necessary
updates in time to report the number and economic value of PM2.5 and ozone health effects
resulting from this proposed rulemaking in the final Revised CSAPR Update RIA. However, to
provide perspective regarding the scope of the estimated benefits, Appendix 5B illustrates the
potential health effects associated with the change in PM2.5 and ozone concentrations as
11 In particular, the 2020 Ozone ISA concludes that the currently available evidence for cardiovascular effects and
total mortality is suggestive of, but not sufficient to infer, a causal relationship with short-term (as well as long-term)
ozone exposures (U.S. EPA, 2020b, sections IS.4.3.4 and IS.4.3.5). As such, EPA is in the process of recalibrating
its benefits estimates to quantify only premature mortality from respiratory causes (i.e., non-respiratory causes of
premature mortality associated with ozone exposure would no longer be estimated). Similarly, the 2019 PM ISA
concludes that the currently available evidence for nervous system effects and cancer is likely to be a causal
relationship with long term PM2 5 exposure. EPA is in the process of evaluating nervous system effects from long
term PM2 5 exposure and evaluating the relationship between long term PM2 5 exposure and cancer. Furthermore, the
ISA references a variety of additional studies for consideration in quantifying the health implications of changes in
PM2.5 and ozone exposure. EPA is updating the estimates for several other health endpoints to account for this new
scientific literature.
ES-11
-------
calculated using methods developed prior to the 2019 PM ISA and 2020 Ozone ISA. The values
of these estimated benefits are not reflected in the estimated net benefits reported below.
The proposal is expected to reduce emissions of ozone season NOx. Reducing NOx
emissions generally reduces human exposure to ozone and the incidence of ozone-related health
effects, though the degree to which ozone is reduced will depend in part on local levels of VOCs
as discussed in Chapter 3. The proposal would also reduce emissions of NOx throughout the
year. Because NOx is also a precursor to formation of ambient PM2.5, reducing these emissions
would reduce human exposure to ambient PM2.5 throughout the year and would reduce the
incidence of PIvfc.s-attributable health effects.12 Reducing emissions of NOx would also reduce
ambient exposure to NO2 and its associated health effects.
ES. 6.1 Climate Benefits Estimates
We estimate the climate benefits for this proposed rulemaking using a measure of the
domestic social cost of carbon (SC-CO2). The SC-CO2 is a metric that estimates the monetary
value of projected impacts associated with marginal changes in CO2 emissions in a specific year.
The SC-CO2 includes a wide range of anticipated climate impacts, such as net changes in
agricultural productivity and human health, property damage from increased flood risk, and
changes in energy system costs, including reduced costs for heating and increased costs for air
conditioning. The metric is typically used to assess the avoided damages as a result of regulatory
actions (i.e., benefits of rulemakings that lead to an incremental reduction in cumulative global
CO2 emissions). The CO2 estimates presented in this RIA focus on the projected impacts of
climate change that are anticipated to directly occur within U.S. borders. Table ES-5 shows the
estimated monetary value of the estimated changes in CO2 emissions in 2021 and 2025 for the
Revised CSAPR Update proposal, the more stringent alternative, and the less stringent
alternative.
Table ES-5. Estimated Domestic Climate Benefits from Changes in CO2 Emissions for
Selected Years (Millions of 2016$)
Regulatory Option
Y ear 3 % Di scount Rate
7% Discount Rate
Proposal
2021 0.3
0.0
12 This RIA does not quantify PM2 5-related benefits associated with SO2 emission reductions. As discussed in
Chapter 4, EPA does not estimate significant SO2 emission reductions as a result of this proposal. Additionally, this
RIA does not estimate changes in emissions of directly emitted particles.
ES-12
-------
Table ES-5. Estimated Domestic Climate Benefits from Changes in CO2 Emissions for
Selected Years (Millions of 2016$)
2025
32.9
5.4
More Stringent
2021
0.8
0.1
Alternative
2025
71.5
11.7
Less Stringent
2021
0.2
0.0
Alternative
2025
25.5
4.2
Table ES-6 shows the total annualized monetary values associated with changes in CO2
emissions for the three regulatory options. The annualized values for the proposed Revised
CSAPR Update are $22 million and $3.6 million, using discount rates of 3 and 7 percent,
respectively.
Table ES-6. Estimated Total Annualized Domestic Climate Benefits (2021-25) from
Changes in CO2 Emissions (Millions of 2016$)
Regulatory Option
3% Discount Rate
7% Discount Rate
Proposal
22.1
3.6
More Stringent Alternative
38.9
6.3
Less Stringent Alternative
15.3
2.5
ES. 6.2 Unquantified Health and Welfare Benefits Categories
The monetized benefits estimated in this RIA reflect a subset of benefits attributable to
the climate benefits from reductions associated with CO2. The proposal is also expected to
reduce emissions of ozone season NOx. In the presence of sunlight, NOx and volatile organic
compounds (VOCs) can undergo a chemical reaction in the atmosphere to form ozone. Reducing
NOx emissions generally reduces human exposure to ozone and the incidence of ozone-related
health effects, though the degree to which ozone is reduced will depend in part on local levels of
VOCs. The proposal would also reduce emissions of NOx throughout the year. Because NOx is
also a precursor to formation of ambient PM2.5, reducing these emissions would reduce human
exposure to ambient PM2.5 throughout the year and would reduce the incidence of PM2.5-
attributable health effects. Reducing emissions of NOx would also reduce ambient exposure to
NO2 and its associated health effects.
Data, time, and resource limitations prevented EPA from quantifying the estimated
impacts or monetizing estimated benefits , including benefits associated with exposure to ozone,
PM2.5, and NO2 (independent of the role NO2 plays as precursors to PM2.5), as well as ecosystem
ES-13
-------
effects, and visibility impairment due to the absence of air quality modeling data for these
pollutants in this analysis. In Chapter 5, (Table 5-4), we provide a qualitative description of these
benefits.
ES.6.3 Approach for Updating Health Effects from PM2.sand Ozone
EPA is reviewing this evidence and is following a five-step approach as it updates its
methods for quantifying and monetizing ozone and PM2.5 attributable health endpoints:
1. Identify Ozone- and PJvfe.s-attributable health effects for which the ISA reports the
strongest evidence. EPA will consider the ISA-reported evidence for each endpoint,
including the extent to which the ISA identifies that endpoint as either causally, or likely-
to-be-causally, related to each pollutant.
2. Identify health outcomes that may be quantified in a benefits assessment. We would
select among clinically significant outcomes (e.g. premature mortality and hospital
admissions) for which endpoint-specific baseline incidence data are available.
3. Choose concentration-response parameters characteristic of the literature reviewed in the
ISA. We would weigh criteria including study design, location, population
characteristics, and other attributes. In some cases we will need to identify and select new
rates of baseline disease to quantify these effects.
4. Choose economic unit values. For each health endpoint we would identify a
corresponding willingness-to-pay or cost-of-illness measure to express the economic
value of the adverse effect.
5. Develop methods for characterizing uncertainty associated with quantified benefits
estimates. Building on EPA's current methods for characterizing uncertainty, these
approaches will include, among others, reporting confidence intervals calculated from
concentration-response parameter estimates and separate quantification using multiple
studies and concentration response parameters for particularly influential endpoints (e.g.,
mortality risk), and potentially approaches for aggregating and representing the results of
multiple studies evaluating a particular health endpoint.
At each of the four stages above, the Agency would report a Preferred Reporting Item for System
Reviews (PRISMA) diagram, detailing for each endpoint, study and concentration-response
ES-14
-------
(effect coefficients), which are included and excluded and the rationale for applying or excluding
this information.
ES.7 Results of Benefit-Cost Analysis
In applying the multi-factor test, EPA evaluated whether reductions resulting from
emitting at the level of the proposed emissions budgets for EGUs in 2021 and 2022 would
resolve any downwind nonattainment or maintenance problems. The assessment showed that the
emission budgets reflecting $1,600 per ton would change the status of one of the two
nonattainment receptors (first shifting the Stratford, Connecticut 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's assessment
shows that none of the 11 states are solely linked to the Stratford receptor that is resolved at the
$1,600 per ton level of control stringency in 2022. In addition, reductions resulting from the
$1,600 per ton emission budgets 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.
Below in Table ES-7 and Table ES-8, we present the annual costs and benefits estimates
for 2021 and 2025, respectively. This analysis uses annual compliance costs reported above as a
proxy for social costs. The net benefits of the proposal and more and less stringent alternatives
reflect the climate benefits of implementing EGU emissions reductions strategies for the affected
12 states via the proposed FIPs minus the costs of those emissions reductions. We represent the
present annual value of non-monetized benefits from reductions in ozone, PM2.5 and NO2
exposure as a B. The annual value of B will differ across discount rates, year of analysis, and the
regulatory alternatives analyzed. The estimated social costs to implement the proposal, as
described in this document, are approximately $21 million in 2021 and $6 million in 2025
(2016$).
The estimated climate benefits from implementation of the proposal are approximately
$0.31 million and $0.05 million in 2021 (2016$, based on a real discount rate of 3 percent and 7
percent for climate benefits). For 2025, the estimated climate benefits from implementation of
ES-15
-------
the proposal are approximately $33 million and $5.4 million (2016$, based on a real discount
rate of 3 percent and 7 percent for climate benefits). As discussed in Chapter 5, the monetized
benefits presented in this proposal RIA are those for climate (from CO2 emissions reductions).
The non-monetized benefits for ozone and PM2.5 are discussed qualitatively in Chapter 5.
EPA calculates the net benefits of the proposal by subtracting the estimated social costs
from the estimated benefits in both 2021 and 2025. The annual net benefits of the proposal in
2021 (in 2016$) are approximately -$21 million using a 3 percent discount rate and a 7 percent
real discount rate. The annual net benefits of the proposal in 2025 are approximately $27 million
using a 3 percent real discount rate and approximately -$0.9 million using a 7 percent real
discount rate. Table ES-7 presents a summary of the climate benefits, costs, and net benefits of
the proposal and the more and less stringent alternatives for 2021. Table ES-8 presents a
summary of these impacts for the proposal and the more and less stringent alternatives for 2025.
Table ES-7. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)a'b'C d
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
21
-21 +B
7%
0.05 +B
-21 +B
More Stringent
Alternative
3%
0.80 +B
37
-36+B
7%
0.12+B
-37+B
Less Stringent
Alternative
3%
0.17+B
4
-4+B
7%
0.03 +B
-4+B
a We focus results to provide a snapshot of costs and benefits in 2021, using the best available information to
approximate social costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in C02 emissions. The costs presented in this table are 2021 annual
estimates for each alternative analyzed.
0 All costs and benefits are rounded to two significant figures; rows may not appear to add correctly.
d B is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual value of B will differ across discount
rates, year of analysis, and the regulatory alternatives analyzed. While EPA did not estimate these benefits in this
RIA, Appendix 5B presents PM2 5 and ozone estimates quantified using methods consistent with the previously
published ISAs to provide information regarding the potential magnitude of the benefits of this proposed rule.
Table ES-8. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)a'b'c'd
Discount Rate
Benefits
Costs
Net Benefits
Proposal
ES-16
-------
3%
33 +B
6
27+B
7%
5.4+B
-0.9 +B
More Stringent
Alternative
3%
71.5+B
132
-61 +B
7%
11.7+B
-120 +B
Less Stringent
Alternative
3%
25 +B
-12
37+B
7%
4.2+B
16+B
" We focus results to provide a snapshot of costs and benefits in 2025, using the best available information to
approximate social costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in C02 emissions. The costs presented in this table are 2025 annual
estimates for each alternative analyzed.
0 All costs and benefits are rounded to two significant figures; rows may not appear to add correctly.
d B is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual value of B will differ across discount
rates, year of analysis, and the regulatory alternatives analyzed. While EPA did not estimate these benefits in this
RIA, Appendix 5B presents PM2 5 and ozone estimates quantified using methods consistent with the previously
published ISAs to provide information regarding the potential magnitude of the benefits of this proposed rule.
Also, as part of fulfilling analytical guidance with respect to E.O. 12866, EPA presents
estimates of the present value of the benefits and costs over the five-year period of 2021 to 2025,
which is the analytical period for this proposal. To calculate the present value of the social net-
benefits of the proposed Revised CSAPR Update, annual benefits and costs are discounted to
2021 at 3 percent and 7 discount rates as directed by OMB's Circular A-4. The present value
(PV) of the net benefits, in 2016$ and discounted to 2021, is -$68 million when using a 7 percent
discount rate and $14 million when using a 3 percent discount rate.13 The equivalent annualized
value (EAV), an estimate of the annualized value of the net benefits consistent with the present
value, is -$17 million per year when using a 7 percent discount rate and $3 million when using a
3 percent discount rate. The EAV represents a flow of constant annual values that, had they
occurred in each year from 2021 to 2025, would yield a sum equivalent to the PV. The EAV
represents the value of a typical cost or benefit for each year of the analysis, in contrast to the
year-specific estimates mentioned earlier in the RIA for the analysis years 2021 and 2025. The
comparison of benefits and costs in PV and EAV terms for the proposal can be found in Table
ES-9. Estimates in the table are presented as rounded values. The table represents the present
13 In annualizing compliance costs using social discount rates, this analysis treats the annual compliance costs as
reflecting the use of real resources in a particular year. In practice, annual costs from IPM and costs of NOx controls
estimated outside of IPM (e.g., capital costs of combustion controls) reflect annual payments for financed capital
and not solely the change in the use of real resources in a particular year (i.e., the opportunity cost of those
resources).
ES-17
-------
value of non-monetized benefits from ozone, PM2.5 and NO2 reductions as a P, while b represents
the equivalent annualized value of these non-monetized benefits. These values will differ across
the discount rates and depend on the values of the B's in the previous tables.
Table ES-9. Summary of Present Values and Equivalent Annualized Values for the 2021-
2025 Timeframe for Estimated Compliance Costs, Climate Benefits, and Net
Benefits for the Proposed Rule (millions of 2016$, discounted to 2021)a'b
3 % Discount Rate 7% Discount Rate
Present Value Benefitsc d 101+p 15+(3
Climate Benefits0 101 15
Compliance Costs6 87 83
N^Benefiis 14+0 -68+0
Equivalent Annualized
Value Benefits 22+b 4+b
Climate Benefits 22 4
Compliance Costs 19 20
Net Benefits 3+b -17+b
a All estimates in this table are rounded to two significant figures, so numbers may not sum due to independent
rounding.
b The annualized present value of costs and benefits are calculated over a 5 year period from 2021 to 2025.
0 Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in CO2 emissions.
d (3 and b is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual values of (3 and b will differ
across discount rates. While EPA did not estimate these benefits in this RIA, Appendix 5B presents PM2 5 and ozone
estimates quantified using methods consistent with the previously published ISAs to provide information regarding
the potential magnitude of the benefits of this proposed rule.
e The costs presented in this table reflect annualized present value compliance costs calculated over a 5 year period
from 2021 to 2025.
ES-18
-------
CHAPTER 1: INTRODUCTION AND BACKGROUND
Overview
EPA originally published the Cross-State Air Pollution Rule (CSAPR) on August 8, 2011,
to address interstate transport of ozone pollution under the 1997 ozone National Ambient Air
Quality Standards (NAAQS).1 On October 26, 2016, EPA published the CSAPR Update, which
finalized Federal Implementation Plans (FIPs) for 22 states that EPA found failed to submit a
complete good neighbor State Implementation Plan (SIP) (15 states)2 or for which EPA issued a
final rule disapproving their good neighbor SIP (7 states).3 The FIPs promulgated for these states
included new electric generating unit (EGU) oxides of nitrogen (NOx) ozone season emission
budgets to reduce interstate transport for the 2008 ozone NAAQS.4 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.5
This proposed action is taken in response to the United States Court of Appeals for the
District of Columbia Circuit's (D.C. Circuit) September 13, 2019 remand of the CSAPR
Update.6 The D.C. Circuit found that the CSAPR Update, which was a partial remedy, 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. This proposed rule, if finalized, will
1 CSAPR also addressed interstate transport of fine particulate matter (PM2 5) under the 1997 and 2006 PM2 5
NAAQS.
2 Alabama, Arkansas, Illinois, Iowa, Kansas, Maryland, Michigan, Mississippi, Missouri, New Jersey, Oklahoma,
Pennsylvania, Tennessee, Virginia, and West Virginia.
3 Indiana, Kentucky, Louisiana, New York, Ohio, Texas, and Wisconsin.
4 The 2008 ozone NAAQS is an 8-hour standard that was set at 75 parts per billion (ppb). See 73 FR 16436 (March
27, 2008).
5 In the CSAPR Update, EPA found that the finalized Tennessee emission budget fully addressed Tennessee's good
neighbor obligation with respect to the 2008 ozone NAAQS. As such, Tennessee is not considered in this proposal,
and the number of states included is reduced from 22 to 21 states.
6 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). The court 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.
1-1
-------
resolve 21 states' outstanding interstate ozone transport obligations with respect to the 2008
ozone NAAQS.
This action, the Revised CSAPR Update proposal, finds that for 9 of the 21 states with
remanded FIPs (Alabama, Arkansas, Iowa, Kansas, Mississippi, Missouri, Oklahoma, Texas, and
Wisconsin), their projected 2021 ozone season nitrogen oxides (NOx) emissions do not
significantly contribute to a continuing downwind nonattainment and/or maintenance problem;
therefore the CSAPR Update fully addresses their interstate ozone transport obligations with
respect to the 2008 ozone NAAQS. This action also proposes to find 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 NOx
emissions significantly contribute to downwind states' nonattainment and/or maintenance
problems for the 2008 ozone NAAQS. For these 12 states, EPA proposes to amend their FIPs to
revise the existing CSAPR NOx Ozone Season Group 2 emissions budgets for EGUs and
implement the revised budgets via a new CSAPR NOx Ozone Season Group 3 Trading
Program.7 EPA is proposing implementation of the revised emission budgets starting with the
2021 ozone season (May 1 - September 30), as outlined in section VIII of the preamble.
These emission budgets represent the remaining EGU emissions after reducing those
amounts of each state's emissions that significantly contribute to downwind nonattainment or
interfere with maintenance of the 2008 ozone NAAQS in downwind states, as required under
Clean Air Act (CAA) section 110(a)(2)(D)(i)(I). The allowance trading program is the proposed
remedy in the FIPs that achieves the ozone season NOx emission reductions proposed by the
rule. The allowance trading program essentially converts the EGU NOx emission budget for each
of the 12 states into a limited number of NOx allowances that, on a tonnage basis, equal the
state's ozone season NOx emission budget. EGUs covered by the FIPs can trade NOx ozone
season allowances among EGUs within their state and across state boundaries, with emissions
and the use of allowances subject to certain limits. The EGUs covered by the FIPs and subject to
the budget are all fossil-fired EGUs with >25MW capacity. The 12 Group 3 states may not use
7 The CSAPR Update established a second NOx ozone season trading program for the 22 states determined to have
good neighbor obligations with respect to the 2008 ozone NAAQS - the CSAPR NOx Ozone Season Group 2
trading program.
1-2
-------
allowances allocated under the CSAPR Update for compliance in 2021 and later.8 Also,
allowances allocated under the Revised CSAPR Update may not be used for compliance in the
10 Group 2 states that remain subject to the budgets established in the CSAPR Update.
Consistent with OMB Circular A-4 and EPA's Guidelines for Preparing Economic
Analyses (2010), this Regulatory Impact Analysis (RIA) presents the benefits and costs of the
proposal and compares the benefits and the costs of the proposed rule in 2021 and 2025. The
estimated benefits are those health benefits expected to arise from reduced air pollution and the
estimated costs are the increased costs of producing electricity and any state reporting
requirements as a result of this rule. Unquantified benefits and costs are described qualitatively.
The RIA also provides (i) estimates of other impacts of the proposed rule including its effect on
retail electricity prices and fuel production and (ii) an assessment of how expected compliance
with the proposed rule would affect concentrations at nonattainment and maintenance receptors.
This chapter contains background information relevant to the rule and an outline of the chapters
of this RIA.
1.1 Background
Clean Air Act (CAA or the 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. The statute vests states with the primary responsibility to address interstate
emission transport 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 FIPs
where states fail to submit good neighbor SIPs or EPA disapproves a submitted good neighbor
SIP.
8 EGUs can still use converted banked allowances from the CSAPR Update to comply with this proposed rule.
1-3
-------
As described in the preamble for the proposal, to reduce interstate emission transport
under the authority provided in CAA section 110(a)(2)(D)(i)(I), this rule proposes to further limit
ozone season (May 1 through September 30) NOx emissions from EGUs in 12 states using the
same framework used by EPA in developing the original CSAPR (the Interstate Transport
Framework). The Interstate Transport Framework provides a 4-step process to address the
requirements of the good neighbor provision for ground-level ozone and fine particulate matter
(PM2.5) 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 1 percent contribution threshold); (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 emissions reductions through enforceable
measures. Details on the methods and results of applying this process can be found in the
preamble for this proposal.
1.1.1 Role of Executive Orders in the Regulatory Impact Analysis
Several statutes and executive orders apply to any public document. The analyses
required by these statutes, along with a brief discussion of several executive orders, are presented
in Chapter 6. Below we briefly discuss the requirements of Executive Orders 12866, 13563, and
13771 and the guidelines of the Office of Management and Budget (OMB) Circular A-4 (U.S.
OMB, 2003).
Executive Order 13771 directs all federal agencies to repeal at least two existing
regulations for each new regulation issued in fiscal year (FY) 2017 and thereafter. It further
directs agencies that the "total incremental costs of all regulations should be no greater than
zero" in FY 2017. For FY 2018 and beyond, the director of the OMB is to provide agencies with
a total amount of incremental costs that will be allowed.
In accordance with Executive Orders 12866 and 13563 and the guidelines of OMB
Circular A-4, the RIA analyzes the benefits and costs associated with emissions reductions for
1-4
-------
compliance with the Revised CSAPR Update proposal. OMB Circular A-4 requires analysis of
one potential regulatory control alternative more stringent than the proposal and one less
stringent than the proposal. This RIA evaluates the benefits, costs, and certain impacts of a more
and a less stringent alternative to the primary alternative in this proposal.
1.1.2 Alternatives Analyzed
EPA proposes to amend FIPs for 12 states to revise the existing CSAPR NOx Ozone
Season Group 2 emissions budgets for EGUs and implement the revised budgets via a new
CSAPR NOx Ozone Season Group 3 Trading Program. Note that EGUs have flexibility in
determining how they will comply with the allowance trading program. EPA is proposing
implementation of the revised emission budgets starting with the 2021 ozone season.
In response to OMB Circular A-4, this RIA analyzes the Revised CSAPR Update proposed
emission budgets as well as a more and a less stringent alternative to the Revised CSAPR Update
proposal. The more and less stringent alternatives differ from the Revised CSAPR Update
proposal in that they set different EGU NOx ozone season emission budgets for the affected
EGUs. The less-stringent scenario uses emission budgets that were developed using uniform
control stringency represented by $500 per ton (2016$). The more-stringent scenario uses
emission budgets that were developed using uniform control stringency represented by $9,600
per ton (2016$). See section VIII of the preamble, and the EGU NOx Mitigation Strategies
Proposed Rule TSD, in the docket for this proposed rule9 for further details of these emission
budgets.
1.1.3 The Needfor Air Quality or Emissions Regulation
OMB Circular A-4 indicates that one of the reasons a regulation may be issued is to
address a market failure. The major types of market failure include externalities, market power,
and inadequate or asymmetric information. Correcting market failures is one reason for
regulation; it is not the only reason. Other possible justifications include improving the function
of government, correcting distributional unfairness, or securing privacy or personal freedom.
9 Docket ID No. EPA-HQ-OAR-2020-0272
1-5
-------
Environmental problems are classic examples of externalities - uncompensated benefits
or costs imposed on another party as a result of one's actions. For example, the smoke from a
factory may adversely affect the health of local residents and soil the property in nearby
neighborhoods. Pollution emitted in one state may be transported across state lines and affect air
quality in a neighboring state. If bargaining were costless and all property rights were well
defined, people would eliminate externalities through bargaining without the need for
government regulation.
From an economics perspective, achieving emissions reductions (i.e., by establishing the
EGUNOx ozone-season emissions budgets in this proposal) through a market-based mechanism
is a straightforward and cost-effective remedy to address an externality in which firms emit
pollutants, resulting in health and environmental problems without compensation for those
incurring the problems. Capping emissions through allowance allocations incentivizes those who
emit the pollutants to reduce their emissions, which lessens the impact on those who suffer the
health and environmental problems from higher levels of pollution.
1.2 Overview and Design of the RIA
1.2.1 Methodology for Identifying Needed Reductions
In order to apply the first and second steps of the CSAPR 4-step Interstate Transport
Framework to interstate transport for the 2008 ozone NAAQS, EPA first performed air quality
modeling coupled with ambient measurements in an interpolation technique to project ozone
concentrations at air quality monitoring sites in 2021. 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. In this analysis, downwind air quality problems are defined by receptors
that are projected to be unable to attain (i.e., nonattainment receptor) or maintain (i.e.,
maintenance receptor) the 2008 ozone NAAQS.
To apply the second step of the Interstate Transport Framework, EPA used air quality
modeling to quantify the contributions from upwind states to ozone concentrations in 2021 at
downwind receptors. Once quantified, EPA then evaluated these contributions relative to a
screening threshold of 1 percent of the NAAQS. States with contributions that equal or exceed 1
1-6
-------
percent of the NAAQS are identified as warranting further analysis for significant contribution to
nonattainment or interference with maintenance.10 States with contributions below 1 percent of
the NAAQS are considered to not significantly contribute to nonattainment or interfere with
maintenance of the NAAQS in downwind states.
To apply the third step of the Interstate Transport Framework, EPA applied a multi-factor
test to evaluate 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. EPA used this multi-
factor assessment to gauge the extent to which emission reductions are needed, and to ensure any
required reductions do not result in over-control.
Using the multi-factor test, EPA identified a control strategy for EGUs at a stringency
level that maximizes cost-effective emission reductions.11 This control strategy reflects the
optimization of existing selective catalytic reduction (SCR) controls and installation of state-of-
the-art NOx combustion controls, with an estimated marginal cost of $1,600 per ton (2016$).12 It
is at this control stringency where incremental EGUNOx reduction potential and corresponding
downwind ozone air quality improvements are maximized relative to the alternative options
analyzed. This strategy maximizes the ratio of emission reductions to marginal cost and the ratio
of ozone improvements to marginal cost. 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 VII.D.l of
the preamble. Further, this evaluation shows that emission budgets reflecting the $1,600 per ton
cost threshold do not over-control upwind states' emissions relative to either the downwind air
10 EPA assessed the magnitude of the maximum projected design value for 2021 at each receptor in relation to the
2008 ozone NAAQS. Where the value exceeds the NAAQS, EPA determined that receptor to be a maintenance
receptor for purposes of defining interference with maintenance. That is, monitoring sites with a maximum design
value that exceeds the NAAQS are projected to have a maintenance problem in 2021.
11 EPA's Guidelines for Preparing Economic Analysis states "[a] policy is cost-effective if it meets a given goal at
least cost, but cost-effectiveness does not encompass an evaluation of whether that goal has been set appropriately
to maximize social welfare. ... A policy is considered cost-effective when marginal abatement costs are equal across
all polluters. In other words, for any level of total abatement, each polluter has the same cost for their last unit
abated." (USEPA 2010, p 4-2). That is not the sense in which the term "cost-effective" is used in this paragraph. For
the sense of what this term means, and in particular what "maximize cost-effective reductions" means in the context
of this proposed rulemaking, see Section VII.D. 1 of the preamble.
12 EGU NOx Mitigation Strategies Proposed Rule TSD, in the docket for this proposed rule (Docket ID No. EPA-
HQ-OAR-2020-0272).
1-7
-------
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 Interstate Transport Framework for the 2008
ozone NAAQS.
In applying the multi-factor test, EPA evaluated whether reductions resulting from the
proposed emissions budgets for EGUs in 2021 and 2022 would resolve any downwind
nonattainment or maintenance problems. The assessment showed that the emission budgets
reflecting $1,600 per ton would change the status of one of the two nonattainment receptors (first
shifting the Stratford, Connecticut monitor to a maintenance-only receptor in 2021, then shifting
that receptor to attainment in 2022); however, no other nonattainment or maintenance problems
would be resolved in 2021 or 2022. EPA's assessment shows that none of the 11 states are solely
linked to the Stratford receptor that is resolved at the $1,600 per ton level of control stringency in
2022. In addition, reductions resulting from the $1,600 per ton emission budgets 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.
1.2.2 States Covered by the Proposed Rule
This rule proposes to find that the following 12 states require further ozone season NOx
emission reductions to address the good neighbor provision as to the 2008 ozone NAAQS:
Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio,
Pennsylvania, Virginia, and West Virginia.13 As such, EPA proposes to promulgate 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. EPA also proposes to adjust
states' emission budgets for each ozone season thereafter to incentivize ongoing operation of
13 This action proposes to find that for 9 of the 21 states with remanded FIPs (Alabama, Arkansas, Iowa, Kansas,
Mississippi, Missouri, Oklahoma, Texas, and Wisconsin), their projected 2021 ozone season NOx emissions do not
significantly contribute to a continuing downwind nonattainment and/or maintenance problem; therefore the CSAPR
Update fully addresses their interstate ozone transport obligations with respect to the 2008 ozone NAAQS. In
addition, in the CSAPR Update EPA found that the finalized Tennessee emission budget fully addressed
Tennessee's good neighbor obligation with respect to the 2008 ozone NAAQS, and Tennessee is also not considered
in this proposal. Allowances allocated under the Revised CSAPR Update may not be used for compliance in these
10 Group 2 states that remain subject to the budgets established in the CSAPR Update.
1-8
-------
identified emission controls to address significant contribution, until such time that our air
quality projections demonstrate anticipated resolution of the downwind nonattainment and/or
maintenance problems for the 2008 ozone NAAQS.
1.2.3 Regulated Entities
The proposed rule affects EGUs in these 12 states and regulates utilities (electric, natural
gas, other systems) classified as code 221112 by the North American Industry Classification
System (NAICS) and have a nameplate capacity of greater than 25 megawatts (MWe).
1.2.4 Baseline and Analysis Years
As described in the preamble, EPA proposes to align implementation of this rule with
relevant attainment dates for the 2008 ozone NAAQS. EPA's final 2008 Ozone NAAQS SIP
Requirements Rule established the attainment deadline of July 20, 2021 for ozone nonattainment
areas currently designated as Serious, and EPA proposes to establish emission budgets and
implementation of these emission budgets starting with the 2021 ozone season.
To develop and evaluate control strategies for addressing these obligations, it is important
to first establish a baseline projection of air quality in the analysis year of 2021, taking into
account currently on-the-books Federal regulations, substantial Federal regulatory proposals,
enforcement actions, state regulations, population, and where possible, economic growth.14
Establishing this baseline for the analysis then allows us to estimate the incremental costs and
benefits of the additional emissions reductions that will be achieved by the proposed transport
rule.
The baseline for this analysis does not assume states will adopt any emissions reduction
methods in and around the Air Quality Control Regions where the nonattainment and
14 The technical support document (TSD) for the 2016vl emissions modeling platform titled Preparation of
Emissions Inventories for 2016vl North American Emissions Modeling Platform is included in the docket for this
proposed rule. The TSD includes additional discussion on mobile source rules included in the baseline. The future
year onroad emission factors account for changes in activity data and the impact of on-the-books rules that are
implemented into MOVES2014b. These rules include the Light Duty Vehicle GHG Rule for Model-Year 2017-2025
and the Tier 3 Motor Vehicle Emission and Fuel Standards Rule. Local inspection and maintenance (I/M) and other
onroad mobile programs are included, such as California LEVIII, the National Low Emissions Vehicle (LEV) and
Ozone Transport Commission (OTC) LEV regulations, local fuel programs, and Stage II refueling control programs.
Regulations finalized after the year 2014 are not included, such as the Safer Affordable Fuel Efficient (SAFE)
Vehicles Final Rule for Model Years 2021-2026 and the Final Rule for Phase 2 Greenhouse Gas Emissions
Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles.
1-9
-------
maintenance receptors are located to reduce ozone other than those already taken into account. In
these areas that do not meet the NAAQS in the baseline that see decreased concentrations of
ozone, the states where these receptors are located may be able to avoid applying other measures
to assure NAAQS attainment. As a result, there would be benefits from avoided compliance
costs in these areas and the ozone and PM2.5 concentrations changes, and their associated health
and ecological benefits, would likely be lower relative to the projections in this RIA. The
baseline in this RIA respects that reductions are required of upwind states in order to improve air
quality at the nonattainment and maintenance receptors.
The analysis in this RIA focuses on benefits, costs and certain impacts in both 2021 and
2025. We focus on 2021 because it is by the 2021 ozone season, corresponding with the 2021
Serious area attainment date, that significant contribution from upwind states' must be
eliminated to the extent possible. In addition, impacts for 2023 to 2025 are important because it
is in this period that additional NOx control technologies could potentially be installed. EPA's
analysis for the third step of the Interstate Transport Framework indicates that by 2023 the
remaining ozone receptors in the two downwind states (Connecticut and Texas) are expected to
shift from nonattainment or maintenance status to meeting the NAAQS with application of
certain EGU controls beginning in 2021, except for one receptor in Westport, Connecticut.15 This
receptor is estimated to shift from nonattainment status to meeting the NAAQS in 2025 with the
application of additional EGU controls. Presenting benefits, costs and certain impacts in 2025
reflects the time needed to make these retrofits on a regional scale and reflects full
implementation of the proposed policy. Additional benefits and costs are expected to occur after
2025 as EGUs subject to this proposal continue to comply with the tighter allowance budget,
which is below their baseline emissions.16 Because EPA did not estimate costs and benefits
beyond 2025, the full costs and benefits of the proposed policy may be understated in this RIA.
15 This RIA also provides an assessment of how expected compliance with the proposed rule would affect
concentrations at nonattainment and maintenance receptors.
16 EPA designed the analysis for the RIA to be consistent with the methodology adopted in the CSAPR Update rule.
As such, the analytical timeframe chosen was 2021-25, which is the period between when the proposed rule would
begin to take effect and the date by which all linked receptors are projected to come clean.
1-10
-------
1.2.5 Emissions Controls, Emissions, and Cost Analysis Approach
EPA estimated the control strategies and compliance costs of the proposed rule using the
Integrated Planning Model (IPM) as well as certain costs that are estimated outside the model but
use IPM inputs for their estimation. These cost estimates reflect costs incurred by the power
sector and include (but are not limited to) the costs of purchasing, installing, and operating NOx
control technology, changes in fuel costs, and changes in the generation mix.17 A description of
the methodologies used to estimate the costs and economic impacts to the power sector is
contained in Chapter 4 of this RIA. This analysis also provides estimates of NOx emissions
changes during ozone season and year-round, as well as emissions changes in carbon dioxide
(CO2) due to changes in power sector operation.
1.2.6 Benefits Analysis Approach
Implementing the Revised CSAPR Update proposed rule is expected to reduce emissions
of NOx and provide ozone reductions, as well as consequent reductions in PM2.5 concentrations
and CO2 emissions. Data, resource, and methodological limitations prevent EPA from
monetizing health benefits from reducing concentrations of ozone and PM2.5, as well as the
benefits of reducing direct exposure to NO2, ecosystem effects and visibility impairment as well
as benefits from reductions in other pollutants, such as hazardous air pollutants (HAP). For more
details on these limitations and a qualitative discussion of the unquantified benefits, see Chapter
5. EPA estimated the climate benefits of the proposal, and a description of the methodologies
used to estimate the climate benefits is also contained in Chapter 5.
1.3 Organization of the Regulatory Impact Analysis
This RIA is organized into the following remaining chapters:
• Chapter 2: Electric Power Sector Profile. This chapter describes the electric power sector
in detail.
• Chapter 3: Emissions and Air Quality Modeling Impacts. The data, tools, and
methodology used for the air quality modeling are described in this chapter, as well as the
17 Under the proposed rule and the more stringent alternative, 27 units are projected to install state-of-the-art
combustion controls; under the less stringent alternative, no units are projected to install state-of-the-art combustion
controls. Under the more stringent alternative, 48 units are projected to install new SCRs; under the proposed rule
and the less stringent alternative, no units are projected to install new SCRs.
1-11
-------
post-processing techniques used to produce a number of air quality metrics for input into
the analysis of benefits and costs.
• Chapter 4: Cost, Emissions, and Energy Impacts. The chapter summarizes the data
sources and methodology used to estimate the costs and other impacts incurred by the
power sector.
• Chapter 5: Benefits. The chapter qualitatively discusses the health-related benefits of the
ozone-related air quality improvements associated with the three regulatory control
alternatives analyzed.
• Chapter 6: Statutory and Executive Order Impact Analyses. The chapter summarizes the
Statutory and Executive Order impact analyses.
• Chapter 7: Comparison of Benefits and Costs. The chapter compares estimates of the
total benefits with total costs and summarizes the net benefits of the three alternative
regulatory control scenarios analyzed.
1-12
-------
CHAPTER 2: ELECTRIC POWER SECTOR PROFILE
Overview
This chapter discusses important aspects of the power sector that relate to the Revised
CSAPR Update proposal with respect to the interstate transport of emissions of nitrogen oxides
(NOx) that contribute significantly to nonattainment or interfere with maintenance of the 2008
ozone NAAQS in downwind states. This chapter describes types of existing power-sector
sources affected by the proposed regulation1 and provides background on the power sector and
electricity generating units (EGUs). In addition, this chapter provides some historical
background on recent trends in the power sector, as well as about existing EPA regulation of the
power sector.
2.1 Background
In the past decade there have been significant structural changes in both the mix of
generating capacity and in the share of electricity generation supplied by different types of
generation. These changes are the result of multiple factors in the power sector, including normal
replacements of older generating units with new units, changes in the electricity intensity of the
U.S. economy, growth and regional changes in the U.S. population, technological improvements
in electricity generation from both existing and new units, changes in the prices and availability
of different fuels, and substantial growth in electricity generation by renewable and
unconventional methods. Many of these trends will continue to contribute to the evolution of the
power sector. The evolving economics of the power sector, specifically the increased natural gas
supply and subsequent relatively low natural gas prices, have resulted in more natural gas being
used as base load energy in addition to supplying electricity during peak load. This chapter
presents data on the evolution of the power sector from 2014 through 2018. Projections of future
power sector behavior and the impact of this proposed rule are discussed in more detail in
Chapter 4 of this RIA.
1 Only coal-fired EGUs will be directly affected (i.e., have to reduce NOx emissions) by this proposal.
2-1
-------
2.2 Power Sector Overview
The production and delivery of electricity to customers consists of three distinct segments:
generation, transmission, and distribution.
2.2.1 Generation
Electricity generation is the first process in the delivery of electricity to consumers. There
are two important aspects of electricity generation; capacity and net generation. Generating
Capacity refers to the maximum amount of production an EGU is capable of producing in a
typical hour, typically measured in megawatts (MW) for individual units, or gigawatts (1 GW =
1,000 MW) for multiple EGUs. Electricity Generation refers to the amount of electricity actually
produced by an EGU over some period of time, measured in kilowatt-hours (kWh) or gigawatt-
hours (GWh = 1 million kWh). Net Generation is the amount of electricity that is available to the
grid from the EGU (i.e., excluding the amount of electricity generated but used within the
generating station for operations). Electricity generation is most often reported as the total annual
generation (or some other period, such as seasonal). In addition to producing electricity for sale
to the grid, EGUs perform other services important to reliable electricity supply, such as
providing backup generating capacity in the event of unexpected changes in demand or
unexpected changes in the availability of other generators. Other important services provided by
generators include facilitating the regulation of the voltage of supplied generation.
Individual EGUs are not used to generate electricity 100 percent of the time. Individual
EGUs are periodically not needed to meet the regular daily and seasonal fluctuations of
electricity demand. Furthermore, EGUs relying on renewable resources such as wind, sunlight
and surface water to generate electricity are routinely constrained by the availability of adequate
wind, sunlight, or water at different times of the day and season. Units are also unavailable
during routine and unanticipated outages for maintenance. These factors result in the mix of
generating capacity types available (e.g., the share of capacity of each type of EGU) being
substantially different than the mix of the share of total electricity produced by each type of EGU
in a given season or year.
Most of the existing capacity generates electricity by creating heat to create high pressure
steam that is released to rotate turbines which, in turn, create electricity. Natural gas combined
2-2
-------
cycle (NGCC) units have two generating components operating from a single source of heat. The
first cycle is a gas-fired turbine, which generates electricity directly from the heat of burning
natural gas. The second cycle reuses the waste heat from the first cycle to generate steam, which
is then used to generate electricity from a steam turbine. Other EGUs generate electricity by
using water or wind to rotate turbines, and a variety of other methods including direct
photovoltaic generation also make up a small, but growing, share of the overall electricity
supply. The generating capacity includes fossil-fuel-fired units, nuclear units, and hydroelectric
and other renewable sources (see Table 2-1). Table 2-1 also shows the comparison between the
generating capacity in 2014 and 2018.
In 2018 the power sector consisted of over 22,000 generating units with a total capacity2 of
1,095 GW, an increase of 26 GW (or 2 percent) from the capacity in 2014 (1,068 GW). The 26
GW increase consisted primarily of natural gas fired EGUs (38 GW), and wind (30 GW) and
solar generators (22 GW), and the retirement/re-rating of 56 GW of coal capacity. Substantially
smaller net increases and decreases in other types of generating units also occurred.
Table 2-1. Total Net Summer Electricity Generating Capacity by Energy Source, 2014
and 2018
2014
2018
Change Between '14 and '18
Energy Source
Net
Summer
Capacity
(MW)
% Total
Capacity
Net
Summer
Capacity
(MW)
% Total
Capacity
%
Increase
Capacity
Change
(MW)
%of
Total
Capacity
Increase
Coal
299,094
28%
242,786
22%
-19%
-56,309
-214%
Natural Gas
432,150
40%
470,237
43%
9%
38,087
145%
Nuclear
98,569
9%
99,433
9%
0.9%
864
3.3%
Hydro
102,162
9.56%
102,702
9.38%
0.5%
540
2.1%
Petroleum
41,135
3.85%
32,218
2.94%
-22%
-8,917
-34%
Wind
64,232
6.01%
94,418
8.62%
47%
30,186
115%
Solar
10,323
0.97%
31,878
2.91%
209%
21,555
82%
Other Renewable
16,049
2%
16,178
1%
1%
129
0%
Misc
4,707
0.44%
4,891
0.45%
4%
184
1%
Total
1,068,422
100%
1,094,740
100%
2%
26,318
100%
2 This includes generating capacity at EGUs primarily operated to supply electricity to the grid and combined heat
and power facilities classified as Independent Power Producers (IPP) and excludes generating capacity at
commercial and industrial facilities that does not operate primarily as an EGU. Natural Gas information in this
chapter (unless otherwise stated) reflects data for all generating units using natural gas as the primary fossil heat
source. This includes Combined Cycle Combustion Turbine, Gas Turbine, steam, and miscellaneous (< 1 percent).
2-3
-------
Note: This table presents generation capacity. Actual net generation is presented in Table 2-2.
Source: EIA. Electric Power Annual 2014 and 2018, Table 4.3
The 2 percent increase in generating capacity is the net impact of newly built generating
units, retirements of generating units, and a variety of increases and decreases to the nameplate
capacity of individual existing units due to changes in operating equipment, changes in emission
controls, etc. During the period 2014 to 2018, a total of 98 GW of new generating capacity was
built and brought online, and 74 GW of existing units were retired. The net effect of the re-rating
of existing units reduced the total capacity by 9.4 GW. The overall net change in capacity was an
increase of 26 GW, as shown in Table 2-1.
The newly built generating capacity was primarily natural gas (44 GW), which was
partially offset by gas retirements (24 GW). Wind capacity was the second largest type of new
builds (30 GW), augmented by solar (21 GW). The largest decline was from coal retirements and
re-rating, which amounted to 56 GW over this period. The overall mix of newly built and retired
capacity, along with the net effect, is shown on Figure 2-1. The data for Figure 2-1 is from Form
EIA-860. Figure 2-1 does not show wind and solar retirements of 568 MW.
130,000
110,000
90,000
70,000
50,000
30,000
10,000
(10,000)
(30,000)
(50,000)
New Build
Retirement
i Other
i Coal
i Wind & Solar
Gas
Figure 2-1. National New Build and Retired Capacity (MW) by Fuel Type, 2014-2018
2-4
-------
The information in Table 2-1 and Figure 2-1 present information about the generating
capacity in the entire U.S. The CSAPR Update, however, directly affected EGUs in 23 eastern
states (i.e., the CSAPR 2008 Ozone Region. The share of generating capacity from each major
type of generation differs between the CSAPR 2008 Ozone Region and the rest of the U.S. (non-
region). Figure 2-2 shows the mix of generating capacity for each region. In 2018, the overall
capacity in the CSAPR 2008 Ozone Region is 59 percent of the national total, reflecting the
larger total population in the region. The mix of capacity is noticeably different in the two
regions. In the CSAPR 2008 Ozone Region in 2014, coal makes up a significantly larger share of
total capacity (26 percent) than it does in the rest of the country (17 percent). The share of
natural gas in the CSAPR 2008 Ozone Region is 45 percent as compared to 40 percent in the rest
of the country. The difference in the share of coal's capacity is primarily balanced by relatively
more hydro, wind, and solar capacity in the rest of country compared to the CSAPR 2008 Ozone
Region.
700,000
600,000
500,000
400,000
300,000
200,000
100,000
0
I Other
i Wind & Solar
I Hydro
I Nuclear
I Gas
I Coal
In Region
Non-Region
Figure 2-2. Regional Differences in Generating Capacity (MW), 2018
Source: FormEIA-860. Note: "Other" includes petroleum, geothermal, other renewable, waste materials and
miscellaneous.
In 2018, electric generating sources produced a net 4,204 TWh to meet national electricity
demand, a 2 percent increase from 2014. As presented in Table 2-2, 62 percent of electricity in
2018 was produced through the combustion of fossil fuels, primarily coal and natural gas, with
natural gas accounting for the largest single share. Although the share of the total generation
2-5
-------
from fossil fuels in 2018 (62 percent) was only modestly smaller than the total fossil share in
2014 (66 percent), the mix of fossil fuel generation changed substantially during that period.
Coal generation declined by 28 percent and petroleum generation by 17 percent, while natural
gas generation increased by 30 percent. This reflects both the increase in natural gas capacity
during that period as well as an increase in the utilization of new and existing gas EGUs during
that period. Wind and solar generation also grew from 5 percent of the mix in 2014 to 8 percent
in 2018.
Table 2-2. Net Generation in 2014 and 2018 (Trillion kWh = TWh)
2014
2018
Change Between '14
and '18
Net
Generation
(TWh)
Fuel Source
Share
Net
Generation
(TWh)
Fuel Source
Share
Net
Generati
on
Change
(TWh)
% Change
in Net
Generation
Coal
1,582
39%
1,146
27%
-436
-440%
Natural Gas
1,127
27%
1,469
35%
342
345%
Nuclear
797
19%
807
19%
10
10%
Hydro
253
6%
287
7%
33
34%
Petroleum
30
1%
25
1%
-5
-5%
Wind
182
4%
273
6%
91
92%
Solar
18
0%
64
2%
46
47%
Other Renewable
91
2%
107
3%
16
16%
Misc
25
1%
26
1%
1
1%
Total
4,105
100%
4,204
100%
99
100%
Source: EIA 2014 and 2018 Electric Power Annual, Tables 3.2 and 3.3.
Coal-fired and nuclear generating units have historically supplied "base load" electricity,
the portion of electricity loads that are continually present and typically operate throughout all
hours of the year. The coal units meet the part of demand that is relatively constant. Although
much of the coal fleet operates as base load, there can be notable differences across various
facilities (see Table 2-3). For example, coal-fired units less than 100 megawatts (MW) in size
compose 18 percent of the total number of coal-fired units, but only 2 percent of total coal-fired
capacity. Gas-fired generation is better able to vary output and is the primary option used to meet
the variable portion of the electricity load and has historically supplied "peak" and
"intermediate" power, when there is increased demand for electricity (for example, when
businesses operate throughout the day or when people return home from work and run appliances
2-6
-------
and heating/air-conditioning), versus late at night or very early in the morning, when demand for
electricity is reduced.
Table 2-3 also shows comparable data for the capacity and age distribution of natural gas
units. Compared with the fleet of coal EGUs, the natural gas fleet of EGUs is generally smaller
and newer. While 66 percent of the coal EGU fleet capacity is over 500 MW per unit, 82 percent
of the gas fleet is between 50 and 500 MW per unit. Many of the largest gas units are gas-fired
steam-generating EGUs.
Table 2-3. Coal and Natural Gas Generating Units, by Size, Age, Capacity, and Average
Heat Rate in 2018
Avg. Net
Total Net
Unit Size
Summer
Summer
Avg. Heat
Grouping
% of All
Capacity
Capacity
% Total
Rate
(MW)
No. Units
Units
Avg. Age
(MW)
(MW)
Capacity
(Btu/kWh)
COAL
0-24
37
7%
50
12
427
0%
11,948
25-49
39
7%
34
36
1,404
1%
12,386
50-99
26
5%
39
76
1,987
1%
12,027
100 - 149
39
7%
48
122
4,757
2%
11,223
150 - 249
73
13%
50
192
14,040
7%
10,882
250 - 499
142
25%
41
364
51,748
24%
10,659
500 - 749
143
26%
39
608
87,005
40%
10,310
750 - 999
49
9%
35
827
40,521
19%
10,057
1000 - 1500
11
2%
41
1,257
13,831
6%
9,802
Total Coal
559
100%
41
386
215,720
100%
10,838
NATURAL GAS
0-24
3,910
51%
32
5
20,540
4%
14,015
25-49
931
12%
26
41
37,792
8%
11,999
50-99
1,032
14%
26
71
73,129
15%
12,315
100 - 149
418
5%
22
127
52,927
11%
9,442
150 - 249
1,018
13%
16
179
181,772
38%
8,192
250 - 499
247
3%
22
332
82,114
17%
8,296
500 - 749
38
0%
39
577
21,910
5%
10,583
750 - 1000
9
0%
44
834
7,510
2%
11,625
Total Gas
7,603
100%
28
63
477,693
100%
12,301
Source: National Electric Energy Data System (NEEDS) v.6
Note: The average heat rate reported is the mean of the heat rate of the units in each size category (as opposed to a
generation-weighted or capacity-weighted average heat rate.) A lower heat rate indicates a higher level of fuel
efficiency. Table is limited to coal-steam units in operation in 2018 or earlier and excludes those units in NEEDS
with planned retirements in 2019 or 2020.
In terms of the age of the generating units, almost 50 percent of the total coal generating
capacity has been in service for more than 40 years, while nearly 50 percent of the natural gas
capacity has been in service less than 15 years. Figure 2-3 presents the cumulative age
distributions of the coal and gas fleets, highlighting the pronounced differences in the ages of the
2-7
-------
fleets of these two types of fossil-fuel generating capacity. Figure 2-3 also includes the
distribution of generation, which is similar to the distribution of capacity.
100%
90%
80%
C
| 70%
-------
Figure 2-4. Fossil Fuel-Fired Electricity Generating Facilities, by Size
Source: National Electric Energy Data System (NEEDS) v.6
Note: This map displays fossil capacity at facilities in the NEEDS v.6 IPM frame. NEEDS v.6 reflects generating
capacity expected to be on-line at the end of 2021. This includes planned new builds already under construction and
planned retirements. In areas with a dense concentration of facilities, some facilities may be obscured.
2.2.2 Transmission
Transmission is the term used to describe the bulk transfer of electricity over a network of
high voltage lines, from electric generators to substations where power is stepped down for local
distribution. In the U.S. and Canada, there are three separate interconnected networks of high
voltage transmission lines,3 each operating synchronously. Within each of these transmission
networks, there are multiple areas where the operation of power plants is monitored and
controlled by regional organizations to ensure that electricity generation and load are kept in
balance. In some areas, the operation of the transmission system is under the control of a single
3 These three network interconnections are the Western Interconnection, comprising the western parts of both the US
and Canada (approximately the area to the west of the Rocky Mountains), the Eastern Interconnection, comprising
the eastern parts of both the US and Canada (except those part of eastern Canada that are in the Quebec
Interconnection), and the Texas Interconnection (which encompasses the portion of the Texas electricity system
commonly known as the Electric Reliability Council of Texas (ERCOT)). See map of all NERC interconnections at
https://www.nerc.com/AboutNERC/keyplayers/PublisliingImages/NERC%20Interconnections.pdf.
2-9
-------
regional operator;4 in others, individual utilities5 coordinate the operations of their generation,
transmission, and distribution systems to balance the system across their respective service
territories.
2.2.3 Distribution
Distribution of electricity involves networks of lower voltage lines and substations that
take the higher voltage power from the transmission system and step it down to lower voltage
levels to match the needs of customers. The transmission and distribution system is the classic
example of a natural monopoly, in part because it is not practical to have more than one set of
lines running from the electricity generating sources to substations or from substations to
residences and businesses.
Over the last few decades, several jurisdictions in the United States began restructuring the
power industry to separate transmission and distribution from generation, ownership, and
operation. Historically, vertically integrated utilities established much of the existing
transmission infrastructure. However, as parts of the country have restructured the industry,
transmission infrastructure has also been developed by transmission utilities, electric
cooperatives, and merchant transmission companies, among others. Distribution, also historically
developed by vertically integrated utilities, is now often managed by a number of utilities that
purchase and sell electricity, but do not generate it. As discussed below, electricity restructuring
has focused primarily on efforts to reorganize the industry to encourage competition in the
generation segment of the industry, including ensuring open access of generation to the
transmission and distribution services needed to deliver power to consumers. In many states,
such efforts have also included separating generation assets from transmission and distribution
assets to form distinct economic entities. Transmission and distribution remain price-regulated
throughout the country based on the cost of service.
2.3 Sales, Expenses, and Prices
These electric generating sources provide electricity for ultimate commercial, industrial
and residential customers. Each of the three major ultimate categories consume roughly a quarter
4 For example, PMJ Interconnection, LLC, Western Area Power Administration (which comprises 4 sub-regions).
5 For example, Los Angeles Department of Power and Water, Florida Power and Light.
2-10
-------
to a third of the total electricity produced6 (see Table 2-4). Some of these uses are highly
variable, such as heating and air conditioning in residential and commercial buildings, while
others are relatively constant, such as industrial processes that operate 24 hours a day. The
distribution between the end use categories changed very little between 2014 and 2018.
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2014 and 2018 (billion kWh)
2014
2018
Sales/Direct
Sales/Direct
Use (Billion
Share of Total
Use (Billion
Share of Total
kWh)
End Use
kWh)
End Use
Residential
1,407
36%
1,469
37%
Sales
Commercial
1,352
35%
1,382
35%
Industrial
998
26%
1,001
25%
Transportation
8
0%
8
0%
Total
3,765
96%
3,859
96%
Direct Use
139
4%
144
4%
Total End Use
3,903
100%
4,003
100%
Source: Table 2.2, EIA Electric Power Annual, 2014 and 2018
Notes: Retail sales are not equal to net generation (Table 2-2) because net generation includes net imported
electricity and loss of electricity that occurs through transmission and distribution, along with data collection frame
differences and non-sampling error. Direct Use represents commercial and industrial facility use of onsite net
electricity generation; electricity sales or transfers to adjacent or co-located facilities; and barter transactions.
2.3.1 Electricity Prices
Electricity prices vary substantially across the United States, differing both between the
ultimate customer categories and by state and region of the country. Electricity prices are
typically highest for residential and commercial customers because of the relatively high costs of
distributing electricity to individual homes and commercial establishments. The higher prices for
residential and commercial customers are the result both of the necessary extensive distribution
network reaching to virtually every part of the country and every building, and also the fact that
generating stations are increasingly located relatively far from population centers (which
increases transmission costs). Industrial customers generally pay the lowest average prices,
reflecting both their proximity to generating stations and the fact that industrial customers
6 Transportation (primarily urban and regional electrical trains) is a fourth ultimate customer category which
accounts less than one percent of electricity consumption.
2-11
-------
receive electricity at higher voltages (which makes transmission more efficient and less
expensive). Industrial customers frequently pay variable prices for electricity, varying by the
season and time of day, while residential and commercial prices historically have been less
variable. Overall industrial customer prices are usually considerably closer to the wholesale
marginal cost of generating electricity than residential and commercial prices.
On a state-by-state basis, all retail electricity prices vary considerably. In 2018, the national
average retail electricity price (all sectors) was 10.53 cents/KWh, with a range from 7.71 cents
(Louisiana) to 29.18 (Hawaii).7
Average national retail electricity prices decreased between 2014 and 2018 by 5 percent
in real terms (2018$).8 The amount of decrease differed for the three major end use categories
(residential, commercial and industrial). National average industrial prices decreased the most (9
percent), and residential prices decreased the least (4 percent). The real year prices for 2014
through 2018 are shown in Figure 2-5.
W
00
c
<1)
Q_
>-
14
12
10
0
2014 2015 2016 2017 2018
Residential Commercial Industrial — —-Total
Figure 2-5. Real National Average Electricity Prices (including taxes) for Three Major
End-Use Categories
7 EIA State Electricity Profiles with Data for 2018 (http://www.eia.gov/electricity/state/)
8 All prices in this section are estimated as real 2018 prices adjusted using the GDP implicit price deflator unless
otherwise indicated.
2-12
-------
Source: EIA Monthly Energy Review (May 2020), Table 9.8.
Most of these electricity price decreases occurred between 2014 and 2015, when nominal
residential electricity prices followed inflation trends, while nominal commercial and industrial
electricity prices declined. The years 2016 and 2017 saw an increase in nominal commercial and
industrial electricity prices, while 2018 saw flattening of this growth. The increase in nominal
electricity prices for the major end use categories, as well as increases in the GDP price and CPI-
U indices for comparison, are shown in Figure 2-6.
Residential
Commercial
Industrial
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for Major
End-Use Categories (including taxes), With Inflation Indices
Source: EIA Monthly Energy Review (May 2020), Table 9.8.
For a longer-term perspective, Figure 2-7 shows real9 (2018$) electricity prices for the
three major customer categories since 1960, and Figure 2-8 shows the relative change in real
electricity prices relative to the prices since 1960. As can be seen in the figures, the price for
industrial customers has always been lower than for either residential or commercial customers,
but the industrial price has been more volatile. While the industrial real price of electricity in
9 All prices in this section are estimated as real 2018 prices adjusted using the GDP implicit price deflator unless
otherwise indicated.
2-13
-------
2018 was relatively unchanged from 1960 (5 percent lower), residential and commercial real
prices are 25 percent and 33 percent lower respectively than in 1960.
8
'= 2
.y j=
- 5
% ^
LU VI
20
18
16
14
12
10
8
6
4
2
0
1960
1970
• Residential
1980 1990
• Commercial
2000
Industrial
2010
— Total
Figure 2-7. Real National Average Electricity Prices for Three Major End-Use Categories
(including taxes), 1960-2018 (2018$)
Source: EIA Monthly Energy Review, May 2020, Table 9.8
~ -10%
.~2Tft0«*
"Residential
¦Commercial
Industrial
¦ Total
Figure 2-8. Relative Change in Real National Average Electricity Prices (2018$) for Three
Major End-Use Categories (including taxes)
Source: EIA Monthly Energy Review, May 2020, Table 9.8.
2-14
-------
2.3.2 Prices of Fossil Fuels Usedfor Generating Electricity
Another important factor in the changes in electricity prices are the changes in delivered
fuel prices10 for the three major fossil fuels used in electricity generation; coal, natural gas and
petroleum products. Relative to real prices in 2014, the national average real price (in 2018$) of
coal delivered to EGUs in 2018 had decreased by 18 percent, while the real price of natural gas
decreased by 33 percent. The real price of delivered petroleum products also decreased by 22
percent, but with petroleum products declining as an EGU fuel (in 2018 petroleum products
generated 1 percent of electricity) the higher delivered oil prices had little overall impact in the
electricity market. The combined real delivered price of all fossil fuels in 2014 decreased by 20
percent over 2014 prices. Figure 2-9 shows the relative changes in real price of all 3 fossil fuels
between 2014 and 2018.
2016
-60%
Coal Oil »Gas - —Average
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in
National Average Real Price per MMBtu Delivered to EGU
Source: EIA Monthly Energy Review, May 2020, Table 9.9.
2.3.3 Changes in Electricity Intensity of the U.S. Economy fi'om 2014 to 2018
An important aspect of the changes in electricity generation (i.e., electricity demand)
between 2014 and 2018 is that while total net generation increased by 2 percent over that period,
111 Fuel prices in this section are all presented in terms of price per MMBtu to make the prices comparable.
2-15
-------
the demand growth for generation was lower than both the population growth (3 percent) and
real GDP growth (10 percent). Figure 2-10 shows the growth of electricity generation,
population and real GDP during this period.
Population
Generation
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since
2014
Sources: Generation: U.S. EIA Monthly Energy Review, May 2020. Table 7.2a Electricity Net Generation: Total
(All Sectors). Population: U.S. Census. Real GDP: 2019 Economic Report of the President, Table B-3.
Because demand for electricity generation grew more slowly than both the population
and GDP, the relative electric intensity of the U.S. economy improved (i.e., less electricity used
per person and per real dollar of output) during 2014 to 2018. On a per capita basis, real GDP per
capita grew by 7 percent between 2014 and 2018. At the same time electricity generation per
capita decreased by 1 percent. The combined effect of these two changes improved the overall
electricity generation efficiency in the U.S. market economy. Electricity generation per dollar of
real GDP decreased 7 percent. These relative changes are shown in Figure 2-11.
2-16
-------
Real GDP/ Capita
Generation/ Capita
Generation/ Real GDP
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation Intensity
Since 2014
Sources: Generation: EIA Monthly Energy Review, May 2020. Table 7.2a Electricity Net Generation: Total (All
Sectors). Population: U.S. Census. Real GDP: 2019 Economic Report of the President, Table B-3.
2.4 Deregulation and Restructuring
The process of restructuring and deregulation of wholesale and retail electricity markets
has changed the structure of the electric power industry. In addition to reorganizing asset
management between companies, restructuring sought a functional unbundling of the generation,
transmission, distribution, and ancillary services the power sector has historically provided, with
the aim of enhancing competition in the generation segment of the industry.
Beginning in the 1970s, government policy shifted against traditional regulatory
approaches and in favor of deregulation for many important industries, including transportation
(notably commercial airlines), communications, and energy, which were all thought to be natural
monopolies (prior to 1970) that warranted governmental control of pricing. However,
deregulation efforts in the power sector were most active during the 1990s. Some of the primary
drivers for deregulation of electric power included the desire for more efficient investment
choices, the economic incentive to provide least-cost electric rates through market competition,
reduced costs of combustion turbine technology that opened the door for more companies to sell
power with smaller investments, and complexity of monitoring utilities' cost of service and
2-17
-------
establishing cost-based rates for various customer classes. Deregulation and market restructuring
in the power sector involved the divestiture of generation from utilities, the formation of
organized wholesale spot energy markets with economic mechanisms for the rationing of scarce
transmission resources during periods of peak demand, the introduction of retail choice
programs, and the establishment of new forms of market oversight and coordination.
The pace of restructuring in the electric power industry slowed significantly in response to
market volatility in California and financial turmoil associated with bankruptcy filings of key
energy companies. By the end of 2001, restructuring had either been delayed or suspended in
eight states that previously enacted legislation or issued regulatory orders for its implementation
(shown as "Suspended" in Figure 2-12). Eighteen other states that had seriously explored the
possibility of deregulation in 2000 reported no legislative or regulatory activity in 2001 (EIA,
2003) ("Not Active" in Figure 2-12). Currently, there are 15 states plus the District of Columbia
where price deregulation of generation (restructuring) has occurred ("Active" in Figure 2-12).
Power sector restructuring is more or less at a standstill; by 2010 there were no active proposals
under review by the Federal Energy Regulatory Commission (FERC) for actions aimed at wider
restructuring, and no additional states have begun retail deregulation activity since that time.
Electricity retail choice states, 2010
States with electricity retail choice programs
States without electricity retail choice programs
Figure 2-12. Status of State Electricity Industry Restructuring Activities
Source: EIA 2010. "Status of Electricity Restructuring by State." Available online at:
.
2-18
-------
One major effect of the restructuring and deregulation of the power sector was a
significant change in type of ownership of electricity generating units in the states that
deregulated prices. Throughout most of the 20th century electricity was supplied by vertically
integrated regulated utilities. The traditional integrated utilities provided generation, transmission
and distribution in their designated areas, and prices were set by cost of service regulations set by
state government agencies (e.g., Public Utility Commissions). Deregulation and restructuring
resulted in unbundling of the vertical integration structure. Transmission and distribution
continued to operate as monopolies with cost of service regulation, while generation shifted to a
mix of ownership affiliates of traditional utility ownership and some generation owned and
operated by competitive companies known as Independent Power Producers (IPPs). The
resulting generating sector differed by state or region, as the power sector adapted to the
restructuring and deregulation requirements in each state.
By the year 2000, the major impacts of adapting to changes brought about by
deregulation and restructuring during the 1990s were nearing completion. In 2014, traditional
utilities owned 61 percent of U.S. generating capacity (MW) while IPPs11 owned 39 percent of
U.S. generating capacity, respectively. The mix of electricity generated (MWh) was more
heavily weighted towards the utilities, with a distribution in 2014 of 61 percent, and 39 percent
for IPPs.
In 2018, the share of capacity (59 percent utility, 41 percent IPPs) and generation (58
percent utility, 42 percent IPP) has remained relatively stable relative to 2014 levels.
The mix of capacity and generation for each of the ownership types is shown in Figures
2-13 (capacity) and 2-14 (generation). The capacity and generation data for commercial and
industrial owners are not shown on these figures due to the small magnitude of those ownership
11 IPP data presented in this section include both combined and non-combined heat and power plants.
2-19
-------
Capacity Mix, 2014 & 2018
Generation Mix, 2014 & 2018
700
600
500
5
400
>•
o
(T!
300
Q.
ro
u
200
100
0
¦ I
ll..
2014 2018 2014 2018
Utility IPP
2,500
B
—
2,000
¦
Solar
Solar
¦ Other
¦ Wind
5
c
o
1,500
¦
¦
n
¦ Other
¦ Wind
¦ Hydro
re
i
QJ
1,000
1
1!
1!
¦ Hydro
¦ Nuclear
01
ej
¦
¦ Nuclear
¦ Gas
500
¦
¦ Gas
¦ Coal
0
¦
¦ Coal
2014 2018 2014 2018
Utility
ipp
Figures 2-13. and 2-14. Capacity and Generation Mix by Ownership Type, 2014 & 2018
types. A porti on of the shift of capacity and generation is due to sales and transfers of generation
assets from traditional utilities to IPPs, rather than strictly the result of newly built units.
2-20
-------
CHAPTER 3: EMISSIONS AND AIR QUALITY IMPACTS
Overview
This Chapter describes the methods for developing spatial fields of air quality
concentrations for the baseline and regulatory control alternatives in 2021 and 2025. These
spatial fields provide the air quality inputs to potentially calculate health benefits for the
proposed Revised CSAPR Update. The spatial fields for this proposal were constructed using the
method and air quality modeling developed to support the regulatory impact analysis (RIA) for
the Repeal of the Clean Power Plan, and the Emission Guidelines for Greenhouse Gas Emissions
from Existing Electric Utility Generating Units (U.S. EPA 2019), also referred to as the
Affordable Clean Energy (ACE) rule.1
In Section 3.1 we describe the ACE air quality modeling platform; in Section 3.2 we
describe the ACE approach for processing the air quality modeling outputs to create inputs for
estimating benefits; in Section 3.3 we describe how the ACE approach was applied in the
proposed Revised CSAPR Update, in Section 3.4 we present maps showing the impacts on ozone
and PM2.5 concentrations of each of the three regulatory control alternatives compared to the
corresponding baseline; and in Section 3.5 we identify uncertainties and limitations in the
application of the ACE approach for generating spatial fields of pollutant concentrations.
3.1 ACE Air Quality Modeling Platform
The air quality modeling for the ACE analysis utilized a 2011-based modeling platform
which included meteorology and base year emissions from 2011 and projected emissions for
2023. The air quality modeling included annual photochemical model simulations for a 2011
base year and a 2023 future year to provide hourly concentrations of ozone and primary and
secondarily formed PM2.5 component species (e.g., sulfate, nitrate, ammonium, elemental carbon
(EC), organic aerosol (OA), and crustal material2) for both years nationwide. In particular,
source apportionment modeling was performed for 2023 to quantify the contributions to ozone
and PM2.5 component species from coal-fired and non-coal EGUs on a state-by-state or multi-
1 Additional details on the ACE modeling and methodology for developing spatial fields of air quality for EGU
control strategies are provided in Appendix 3A.
2 Crustal material refers to metals that are commonly found in the earth's crust such as Aluminum, Calcium, Iron,
Magnesium, Manganese, Potassium, Silicon, Titanium and the associated oxygen atoms.
3-1
-------
state basis. As described below, the modeling results for 2011 and 2023, in conjunction with
emissions data for the baseline and regulatory control alternatives, were used to construct the air
quality spatial fields that reflect the influence of emissions changes between the baseline and the
regulatory control alternatives.
The air quality model simulations {i.e., model runs) were performed using the
Comprehensive Air Quality Model with Extensions (CAMx) (Ramboll Environ 2016). Our
CAMx nationwide modeling domain (i.e.., the geographic area included in the modeling) covers
all lower 48 states plus adjacent portions of Canada and Mexico using a horizontal grid
resolution of 12 x 12 km shown in Figure 3-1.
Figure 3-1. Air Quality Modeling Domain
The impact of specific emissions sources on ozone and PM2.5 in the 2023 modeled case
was tracked using a tool called "source apportionment." In general, source apportionment
modeling quantifies the air quality concentrations formed from individual, user-defined groups
of emissions sources or "tags". These source tags are tracked through the transport, dispersion,
chemical transformation, and deposition processes within the model to obtain hourly gridded3
contributions from the emissions in each individual tag to hourly modeled concentrations of
ozone and PM2.5 4 Thus, the source apportionment method provides an estimate of the effect of
3 Hourly contribution information is provided for each grid cell to provide spatial patterns of the contributions from
each tag.
4 Note that the sum of the contributions in a model grid cell from each tag for a pollutant equals the total
concentration of that pollutant in the grid cell.
3-2
-------
changes in emissions from each group of emissions sources {i.e., each tag) to changes in ozone
and PM2.5 concentrations. For this analysis we applied outputs from source apportionment
modeling for ozone and PM2.5 using the 2023 modeled case to obtain the contributions from
EGU emissions as well as other sources to ozone and to PM2.5 component species
concentrations.5 Ozone contributions were modeled using the Ozone Source Apportionment
Technique/Anthropogenic Precursor Culpability Assessment (OSAT/APCA) tool and PM2.5
component species contributions were modeled using the Particulate Source Apportionment
Technique (PSAT) tool.6 The source apportionment modeling, which was already available from
analysis performed to support the ACE rule RIA (U.S. EPA, 2019) was used to quantify the
contributions from EGU emissions on a state-by-state or, in some cases, on a multi-state basis.
For ozone, we modeled the contributions from the 2023 EGU sector emissions of NOx and VOC
to hourly ozone concentrations for the period April through October to provide data for
developing spatial fields for two seasonal ozone benefits metrics identified above (i.e., for the
May-September seasonal average of the maximum daily 8-hour average (MDA8) ozone and the
April-October seasonal average of the maximum daily 1-hour average (MDA1) ozone). For
PM2.5, we modeled the contributions from the 2023 EGU sector emissions of SO2, NOx, and
directly emitted PM2.5 for the entire year to inform the development of spatial fields of annual
mean PM2.5. For each state, or multi-state group, we separately tagged EGU emissions depending
on whether the emissions were from coal-fired units or non-coal units.7 In addition to tagging
coal-fired and non-coal EGU emissions we also tracked the ozone and PM2.5 contributions from
all other sources.
3.2. Applying Modeling Outputs to Create Spatial Fields
In this section we describe the ACE approach for creating spatial fields based on the 2011
and 2023 modeling performed for the ACE rule. The foundational data from ACE include the
ozone contributions from EGU emissions in each state based on the 2023 ACE EGU state-sector
sector contribution modeling and the 2023 emissions for coal and non-coal fired EGUs that were
5 In the source apportionment modeling for PM2 5 we tracked the source contributions from primary, but not
secondary organic aerosols (SOA). The method for treating SOA concentrations is described in U.S. EPA, 2019
chapter 8.
6 OSAT/APCA and PSAT tools are described in Ramboll Environ (2016).
7 For the purposes of this analysis non-coal fuels include emissions from natural gas, oil, biomass, municipal waste
combustion and waste coal EGUs.
3-3
-------
input to that modeling. These data are used to generate spatial fields based on ozone season EGU
NOx emissions (tons) and annual total EGU emissions of NOx, S02 and PM2.5. The inputs for
this method include emissions for each state with a breakout of emissions for coal-fired and non-
coal EGUs. The ozone season NOx emissions are used to prepare spatial fields of the May-
September seasonal average MDA8 ozone and the April-October seasonal average MDA1 ozone
concentration and the annual emissions are used to prepare spatial fields of annual PM2.5
concentrations. This method calculates the scaling ratios, described below, that are used to
prepare the air quality spatial fields.
To create the spatial fields for each future emissions scenario the 2023 state-sector source
apportionment modeling outputs from the ACE modeling described above are used in
combination with the EGU SO2, NOx, and PM2.5 emissions for each scenario. Contributions from
each state-sector contribution "tag" were scaled based on the ratio of emissions in the
year/scenario being evaluated to the emissions in the modeled ACE 2023 scenario. In this
approach, scaling ratios for PM2.5 components that are emitted directly from the source (OA, EC,
crustal) are based on relative changes in annual primary PM2.5 emissions between the modeled
ACE 2023 emissions scenario and the specific baseline or control scenario being analyzed. Also
the scaling ratios for components that are formed through chemical reactions in the atmosphere
were created as follows: scaling ratios for sulfate were based on relative changes in annual SO2
emissions; scaling ratios for nitrate were based on relative changes in annual NOx emissions; and
scaling ratios for ozone formed in NOx-limited regimes8 ("03N") were based on relative
changes in ozone season (May-September) NOx emissions. Tags representing sources other than
EGUs are held constant at 2023 ACE baseline levels for emissions scenarios analyzed by the
user. For each control scenario analyzed, the scaled contributions from all sources were summed
together to create a gridded surface of total modeled ozone or total modeled PM2.5. Finally,
spatial fields of ozone and PM2.5 are created based on "fusing" modeled data with measured
concentrations at air quality monitoring locations. The process is described in a step-by-step
manner below.
(1) The EGU annual SO2, NOx, and directly emitted PM2.5 emissions for the control scenario
of interest and the corresponding 2023 SO2, NOx, and directly emitted PM2.5 emissions
8 The CAMx model internally determines whether the ozone formation regime is NOx-limited or VOC-limited
depending on predicted ratios of indicator chemical species.
3-4
-------
used in the ACE modeling to calculate the ratio of control case emissions to the ACE
emissions for each of these pollutants for each EGU tag.
(2) The tag-specific 2025 to 2023 EGU emissions-based scaling ratios from step (1) are
multiplied by the corresponding 365 gridded daily 24-hour average PM2.5 component
species contributions from the 2023 contribution modeling. The emissions ratios for SO2
are applied to sulfate contributions; ratios for annual NOx are applied to nitrate
contributions; and ratios for directly emitted PM2.5 are applied to the EGU contributions
to primary OA, EC and crustal material. This step results in 365 adjusted gridded daily
PM2.5 component species contributions for each EGUs tag that reflects the emissions in
the control scenario.
(3) For each individual PM2.5 component species, the adjusted gridded contributions for each
EGU tag from step (2) are added together to produce a gridded daily EGU tag total.
(4) The daily total EGU contributions for each PM2.5 component species from step (3) are
then combined with the species contributions from source tags representing all other
sources of PM2.5. As part of this step we also add the total secondary organic aerosol
concentrations from the 2023 ACE modeling to the net EGU contributions of primary
OA. Note that the secondary organic aerosol concentration does not change between
scenarios. This step results in 24-hour average PM2.5 component species concentrations
for the control scenario in each model grid cell, nationwide for each day in the year.
(5) For each PM2.5 component species, the daily concentrations from step (4) are averaged
for each quarter of the year.
(6) The quarterly average PM2.5 component species concentrations from step (5)9 are divided
by the corresponding quarterly average species concentrations from the base period air
quality model run. This step provides a Relative Response Factor (i.e., RRF) between the
base period and the control scenario for each species in each model grid cell.
(7) The species-specific quarterly RRFs from step (6) are then multiplied by the
corresponding species-specific quarterly average concentrations from the base period
9 Ammonium concentrations are calculated assuming that the degree of neutralization of sulfate ions remains at
2011 levels (see Chapter 8 of U.S. EPA, 2019 for details).
3-5
-------
fused surfaces to produce quarterly average species concentrations for the control
scenario.
(8) The quarterly average species concentrations from step (7) are summed over the species
to produce total PM2.5 concentrations for each quarter. Finally, total PM2.5 concentrations
for the four quarters of the year are averaged to produce the spatial field of annual
average PM2.5 concentrations for the 2025 baseline.
To generate the spatial fields for each of the two ozone concentration metrics (i.e., April-
October MDA1 and May-September MDA8) we follow the steps similar to those above for
PM2.5.
(1) The EGU May through September (i.e., Ozone Season - OS) NOx for the control scenario
and the corresponding modeled 2023 OS NOx emissions are used to calculate the ratio of
control scenario emissions to 2023 ACE emissions for each EGU tag (i.e. an ozone-
season scaling factor for each tag).
(2) The source apportionment modeling provided separate ozone contributions for ozone
formed in VOC-limited chemical regimes (O3V) and ozone formed in NOx-limited
chemical regimes (O3N).10 The EGU OS NOx emissions for the control scenario and the
2023 ACE OS NOx baseline emissions are used to calculate the ratio of the control
scenario emissions to the 2023 ACE emission to create the EGU NOx emissions scaling
ratios. The emissions scaling ratios are multiplied by the corresponding O3N gridded
daily contributions to MDA1 and MDA8 concentrations. This step results in adjusted
gridded daily MDA1 and MDA8 contributions due to NOx changes for each EGUs tag
that reflect the emissions in the 2025 baseline.
(3) For MDA1 and MDA8, the adjusted contributions for each EGU tag from step (2) are
added together to produce a daily adjusted EGU tag total. Since IPM does not output
VOC from EGUs, there are no predicted changes in VOC emissions in these scenarios so
the O3V contributions remain unchanged. The contributions from the unaltered O3V tags
from the 2023 ACE modeling are added to the summed adjusted O3N EGU tags.
10 Information on the treatment of ozone contributions under NOx-limited and VOC-limited chemical regimes in the
CAMx APCA source apportionment technique can be found in the CAMx v6.40 User's Guide (Ramboll, 2016).
3-6
-------
(4) The daily total EGU contributions for MDA1 and MDA8 from step (3) are then
combined with the contributions to MDA1 and MDA8 from all other sources. This step
results in MDA1 and MDA8 concentrations for the control scenario in each model grid
cell, nationwide for each day in the ozone season.
(5) For MDA1, we average the daily concentrations from step (4) across all the days in the
period April 1 through October 31. For MDA8, we average the daily concentrations
across all days in the period May 1 through September 30.
(6) The seasonal mean concentrations from step (5) are divided by the corresponding
seasonal mean concentrations from the base period air quality model run. This step
provides a Relative Response Factor (i.e., RRF) between the base period and control
scenario for MDA1 and MDA8 in each model grid cell.
(7) Finally, the RRFs for the seasonal mean metrics from step (6) are then multiplied by the
corresponding seasonal mean concentrations from the base period MDA1 and MDA8
fused surfaces to produce seasonal mean concentrations for MDA1 and MDA8 for the
control scenario that are input to BenMAP-CE.
3.3 Application of ACE Approach for the Revised CSAPR Update
In this section we describe how we applied the ACE approach to generate spatial fields of
seasonal ozone and annual PM2.5 concentrations associated with the regulatory control
alternatives (i.e., the proposal and the less stringent and more stringent alternatives) in this
proposed rule RIA. The data for creating the Revised CSAPR Update spatial fields include EGU
emissions for the 2021 and 2025 baseline and the regulatory control alternatives. The EGU
emissions include OS NOx and annual NOx, S02, and PM2.5 for coal-fired and non-coal units in
each state in the continental U.S. These EGU emissions are taken from the electricity sector
analysis described in Chapter 4. In the case of the Revised CSAPR Update proposal analysis,
there are no impacts on SO2 or PM2.5 emissions in the regulatory control scenarios compared to
the 2025 baseline.
To potentially calculate ozone-related benefits in 2021 and 2025 we used the ozone season
EGU NOx emissions (tons) for the 2021 and 2025 baseline along with emissions for the
proposal, and each of the two other regulatory control alternatives. These emissions were applied
using the ACE approach and source apportionment data to produce spatial fields of the May-
3-7
-------
September seasonal average MDA8 ozone and the April-October seasonal average MDA1 ozone
concentrations as described in the previous section.
In 2021, the only control measure expected to be adopted for compliance in each of the
regulatory control alternatives is optimization of existing SCRs beginning in May of 2021, and
this measure will operate only during the ozone season. This is relevant because NOx reductions
in the ozone season provide minimal PM2.5 reductions 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 would be considered worthwhile to estimate PM2.5
benefits from NOx reductions in those months (Hand et al., 2012). In 2025, the presence of
additional control measures that operate year-round and other changes in market conditions as a
result of the proposed rule lead to notable NOx reductions in the winter months.
To create spatial fields for PM2.5 we pre-processed the 2025 coal and non-coal fired EGU
emissions in order to obtain annual emissions of NOx, S02, and directly emissions PM2.5 in a
manner that is appropriate for assessing the impacts on annual average PM2.5 concentrations.
This additional pre-processing was needed because the vast majority of the emissions reductions
are expected to occur during the ozone season but, as noted above, PM2.5 nitrate concentrations
are lowest during that time of year. In this regard, simply treating the summer emissions
reductions as if they were abated proportionately throughout the year would overstate the
impacts of the emissions reductions on PM2.5 and therefore overstate benefits associated with
reducing exposure to PM2.5.11 For those states in which there are NOx emissions reductions
during the ozone season only, we reset the annual NOx emission in the regulatory alternative to
be equivalent to the corresponding baseline emissions to avoid distributing the ozone season
reductions across the entire year. That is, we assumed that there would be no impact on PM2.5
nitrate concentrations of NOx reductions in the ozone season. For those states in which there are
NOx emissions changes between the baseline and regulatory control alternative outside of the
ozone season, we accounted for those reductions by "annualizing" the EGU emissions for the
11 The FAST-CE model described above essentially treats a ton of abated NOx emissions as if it were abated in
equal proportions per time (e.g., day) throughout the year when projecting PM2 5 fields. Therefore, when NOx
abatement is heavily concentrated in a particular time of the year, as in the proposed Revised CSAPR Update, the
inputs to the model need to be adjusted to avoid overestimating (as for this proposed rule) or underestimating (if
reductions were greater in the winter months) the change in annual PM2 5 concentrations and benefits from changes
in PM2 5 exposure.
3-8
-------
period outside the ozone season in the regulatory alternative as well as the corresponding
baseline. This method essentially applies the change in NOx tons outside the ozone season on a
daily basis to changes in NOx emissions tons within the ozone season.12 With this adjustment the
impact of the regulatory control alternative on annual average PM2.5 concentrations reflects the
emissions reductions that will occur outside the ozone season when PM2.5 nitrate concentrations
are highest. The emissions of SO2 and directly emitted PM2.5 in 2025 for each of the regulatory
alternatives do not change from the 2025 baseline. That is, the regulatory control alternatives
analyzed in this RIA reduce emissions of NOx, but do not impact emissions of SO2 and directly
emitted PM2.5.
3.4 Spatial Distribution of Air Quality Impacts
Below we present the estimated impacts on May-September MDA8 ozone13 between the
baseline and each of the regulatory control alternatives for 2021 and 2025 as well as the
estimated impacts on annual mean PM2.5 concentrations between the baseline and the regulatory
control alternatives in 2025 (Figure 3-2 through Figure 3-10). The data shown in these figures
are calculated as the baseline minus the regulatory control alternative concentrations (i.e.,
positive values indicate reductions in pollutant concentrations). The spatial patterns of the
impacts of emissions reductions are a result of (1) the spatial distribution of EGU sources that are
predicted to have changes in emissions and (2) the physical or chemical processing that the
model simulates in the atmosphere.
12 In all states the actual tons reduced in the ozone season is greater than or equal to the change outside the ozone
season between the baseline and the regulatory alternatives.
13 The estimated impacts on April-October 2021 and 2025 ozone for each scenario are not shown but are similar to
May-September impacts available in Figure 12-20.
3-9
-------
Min = O.OOE+O at (1,1), Max = 0.088 at (299,169)
Figure 3-2. Map of change in May-September MDA8 ozone (ppb):
2021 baseline - less stringent regulatory alternative (scale: + 0.10 ppb)
Min = 0.00E+0 at (1,1), Max = 1.234 at (308,134)
Figure 3-3. Map of change in May-September MDA8 ozone (ppb):
2021 baseline - proposal (scale: + 0.50 ppb)
3-10
-------
Delia
Figure 3-4. Map of change in May-September MDA8 ozone (ppb):
2021 baseline - more stringent regulatory alternative (scale: + 0.50 ppb)
Min = Q.00E+0 at (1,1), Max = 0 083 at (299,169)
Figure 3-5. Map of change in May-September MDA8 ozone (ppb):
2025 baseline - less stringent regulatory alternative (scale: + 0.10 ppb)
3-11
-------
1 80 159 239 318 397
Mi n = O.OOEtO at (1,1), Max = 1.303 at (308,134)
Figure 3-6. Map of change in May-September MDA8 ozone (ppb):
2025 baseline - proposal (scale: + 0.50 ppb)
Delta
Mm > O OOI<0 at (1,1). M». > 1 6SO M (187.1 JO)
Figure 3-7. Map of change in May-September MDA8 ozone (ppb):
2025 baseline - more stringent regulatory alternative (scale: + 0.50 ppb)
3-12
-------
Figure 3-8. Map of change in annual mean PM2.5 (pg/in3):
2025 baseline - less stringent regulatory alternative (scale: + 0.01 pg/m3)
Min = 0.00E+0 at (1,1), Max = 3.85E-3at (283,135)
Figure 3-9. Map of change in annual mean PM2.5 (pg/m3):
2025 baseline - proposal (scale: + 0.01 pg/m3)
3-13
-------
Min = O.OOE+O at (1,1), Max = 3.85E-3 at (283,135)
Figure 3-10. Map of change in annual mean PM2.5 (pg/m3):
2025 baseline - more stringent regulatory alternative (scale: + 0.01 pg/m3)
3.5 Uncertainties and Limitations of ACE Approach
One limitation of the scaling methodology for creating PM2.5 surfaces associated with the
baseline and regulatory alternatives described above is that it treats air quality changes from the
tagged sources as linear and additive. It therefore does not account for nonlinear atmospheric
chemistry and does not account for interactions between emissions of different pollutants and
between emissions from different tagged sources. This is consistent with how air quality
estimations have been treated in past regulatory analyses (U.S. EPA 2012; 2019; 2020b). We
note that air quality is calculated in the same manner for the baseline and the regulatory
alternatives, so any uncertainty associated with these assumptions is carried through both sets of
scenarios in the same manner and is thus not expected to impact the air quality differences
between scenarios. In addition, emissions changes between baseline and the regulatory
alternatives are relatively small compared to modeled 2023 emissions that form the basis of the
ACE source apportionment approach. Previous studies have shown that air pollutant
concentrations generally respond linearly to small emissions changes of up to 30 percent
(Dunker et al., 2002; Cohan et al., 2005; Napelenok et al., 2006; Koo et al., 2007; Zavala et al.,
2009; Cohan and Napelenok, 2011) and that linear scaling from source apportionment can do a
reasonable job of representing impacts of 100 percent of emissions from individual sources
(Baker and Kelly 2014). Therefore, while simplistic, it is reasonable to expect that the emissions
3-14
-------
concentration differences between the baseline and regulatory control alternatives can be
adequately represented using this methodology and any uncertainty should be weighed against
the speed in which this method may be used to account for spatial differences in the effect of
EGU emissions on ozone and PM2.5 concentrations.
A second limitation is that the source apportionment PM2.5 contributions represent the
spatial and temporal distribution of the emissions from each source tag as they occur in the 2023
modeled case. Thus, the contribution modeling results do not allow us to represent any changes
to "within tag" spatial distributions. As a result, the method does not account for any changes of
spatial patterns that would result from changes in the relative magnitude of sources within a
source tag in the scenarios investigated here. As described above, the EGU tags are generally by
state and by two EGU types; one for coal-fired units and one for non-coal units.
In addition, the 2023 CAMx-modeled concentrations themselves have some uncertainty.
While all models have some level of inherent uncertainty in their formulation and inputs, the
base-year 2011 model outputs have been evaluated elsewhere against ambient measurements
(U.S. EPA 2017; 2019) and have been shown to adequately reproduce spatially and temporally
varying ozone and PM2.5 concentrations.
The regulatory alternatives lead to decreased concentrations of ozone and PM2.5, the
extent to which varies by location, relative to the baseline. However, the analysis does not
account for how interaction with NAAQS compliance would affect the benefits and costs of the
regulatory alternatives, which introduces uncertainty in the benefits and costs of the alternatives.
To the extent the Revised CSAPR Update proposal will decrease NOx and consequentially ozone
and PM2.5, these changes may affect compliance with existing NAAQS standards and
subsequently affect the actual benefits and costs of the proposed rule. In areas not projected to
attain the 2015 ozone NAAQS without further emissions reductions from the baseline, states
may be able avoid applying some emissions control measures to reduce emissions from local
sources as a result of this proposed rule. If compliance behavior with the 2015 ozone NAAQS
were accounted for in the baseline in this RIA there may be additional benefits from reduced
compliance costs, while the level and spatial pattern of changes in ozone and PM2.5
concentrations, and their associated health and ecological benefits, would differ.
3-15
-------
Similarly, the regulatory alternatives may project decreases in ozone and PM2.5
concentrations in areas attaining the NAAQS in the baseline. In practice, these potential changes
in concentrations may influence NAAQS compliance plans in these areas, which in turn would
further influence concentrations and the cost of complying with the NAAQS. However, such
behavior will be mitigated by NAAQS requirements such as Prevention of Significant
Deterioration (PSD) requirements. This RIA does not account for how interaction with NAAQS
compliance would affect the benefits and costs of the regulatory alternatives.
3.6 References
Baker, Kirk R., and James T. Kelly. 2014. "Single Source Impacts Estimated with
Photochemical Model Source Sensitivity and Apportionment Approaches." Atmospheric
Environment 96 (October): 266-74. https://doi.Org/10.1016/j.atmosenv.2014.07.042.
Cohan Daniel S., Amir Hakami, Yongtao Hu, Armistead G. Russell. 2005. "Nonlinear response
of ozone to emissions: Source apportionment and sensitivity analysis." Environmental
Science & Technology 39:6739-6748
Cohan, Daniel, and Sergey Napelenok. 2011. "Air Quality Response Modeling for Decision
Support." Atmosphere 2 (December): 407-25. https://doi.org/10.3390/atmos2030407.
Ding, Dian, Yun Zhu, Carey Jang, Che-Jen Lin, Shuxiao Wang, Joshua Fu, Jian Gao, Shuang
Deng, Junping Xie, and Xuezhen Qiu. 2016. "Evaluation of Health Benefit Using
BenMAP-CE with an Integrated Scheme of Model and Monitor Data during Guangzhou
Asian Games." Journal of Environmental Sciences 42 (April): 9-18.
https://doi.Org/10.1016/j.jes.2015.06.003.
Dunker, Alan M., Greg Yarwood, Jerome P. Ortmann, and Gary M. Wilson. 2002. "The
Decoupled Direct Method for Sensitivity Analysis in a Three-Dimensional Air Quality
Model Implementation, Accuracy, and Efficiency." Environmental Science &
Technology 36 (13): 2965-76. https://doi.org/10.1021/es0112691.
Hand, J. L., B.A. Schichtel, M. Pitchford, W.C. Malm, andN.H. Frank. 2012. "Seasonal
Composition of Remote and Urban Fine Particulate Matter in the United States. Journal
of Geophysical Research, 117, D05209, doi: 10.1029/2011JD017122.
Koo, Bonyoung, Alan M. Dunker, and Greg Yarwood. 2007. "Implementing the Decoupled
Direct Method for Sensitivity Analysis in a Particulate Matter Air Quality Model."
Environmental Science & Technology 41 (8): 2847-54.
https://doi.org/10.1021/es0619962.
3-16
-------
Napelenok, Sergey L., Daniel S. Cohan, Yongtao Hu, and Armistead G. Russell. 2006.
"Decoupled Direct 3D Sensitivity Analysis for Particulate Matter (DDM-3D/PM)."
Atmospheric Environment A0 (32): 6112-21.
https://doi.Org/10.1016/j.atmosenv.2006.05.039.
Ramboll Environ. 2016. "Comprehensive Air Quality Model with Extensions Version 6.40."
User's Guide. Novato, CA: Ramboll Environ International Corporation.
http://www.camx.com/files/camxusersguide_v6-40.pdf.
US EPA, 2012. "Regulatory Impact Analysis for the Final Revisions to the National Ambient Air
Quality Standards for Particulate Matter." EPA-452/R-12-005. Research Triangle Park,
NC: U.S. Environmental Protection Agency.
https://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf.
US EPA, 2017. Documentation for the EPA's Preliminary 2028 Regional Haze Modeling.
Research Triangle Park, NC
(https://www3.epa.gov/ttn/scram/reports/2028_Regional_Haze_Modeling-TSD.pdf).
US EPA, 2019. "Regulatory Impact Analysis for the Repeal of the Clean Power Plan, and the
Emission Guidelines for Greenhouse Gas Emissions from Existing Electric Utility
Generating Units." EPA-452/R-19-003. Research Triangle Park, NC: U.S. Environmental
Protection Agency, https://www.epa.gov/sites/production/files/2019-
06/documents/utilities_ria_final_cpp_repeal_and_ace_2019-06.pdf.
US EPA, 2020a. "Regulatory Impact Analysis for Revisions to the Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source
Category". EPA-821-R-20-004. Washington, DC: U.S. Environmental Protection
Agency, https://www.epa.gov/sites/production/files/2020-
08/documents/steam_electric_elg_2020_final_reconsideration_rule_regulatory_impact_a
nalysis.pdf.
US EPA, 2020b. "Analysis of Potential Costs and Benefits for the "National Emission Standards
for Hazardous Air Pollutants: Coal- and Oil-Fired Electric Utility Steam Generating
Units - Subcategory of Certain Existing Electric Utility Steam Generating Units Firing
Easter". Memo to Docket for rulemaking: "National Emission Standards for Hazardous
Air Pollutants: Coal- and Oil-Fired Electric Utility Steam Generating Units -
Subcategory of Certain Existing Electric Utility Steam Generating Units Firing Eastern
Bituminous Coal Refuse for Emissions of Acid Gas Hazardous Air Pollutants" (EPA-
HQ-OAR-2018-0794), April 8, 2020. Available at:
https://www.epa.gOv/sites/production/files/2020-04/documents/mats_coal_refuse_cost-
benefit_memo.pdf
Zavala, M., Lei, W., Molina, M.J., Molina, L.T., 2009. Modeled and observed ozone sensitivity
to mobile-source emissions in Mexico City. Atmos. Chem. Phys. 9, 39-55.
3-17
-------
APPENDIX 3A: METHODOLOGY FOR DEVELOPING AIR QUALITY SURFACES
In this appendix we describe the air quality modeling platform and methodology that was
leveraged to prepare the air quality surfaces that could inform the calculation of health benefits
of the proposed Revised CSAPR Update. The modeling and methodology described here were
developed to support the Regulatory Impact Analysis for the Repeal of the Clean Power Plan,
and the Emission Guidelines for Greenhouse Gas Emissions from Existing Electric Utility
Generating Units (U.S. EPA 2019), also referred to the Affordable Clean Energy (ACE) rule.
The foundational data in the ACE approach include the 2023 ACE baseline EGU emissions and
the 2023 ACE EGU air quality contribution data described below. To generate spatial fields for
alternative EGU scenarios, such as the scenarios analyzed for the Revised CSPR Update
proposal, the user provides as input EGU emissions for coal-fired and non-coal units for each
state, separately. Ozone season EGUNOx emissions (tons) are used to prepare spatial fields of
the May-September seasonal average MDA8 ozone and the April-October seasonal average
MDA1 ozone concentrations and annual total EGU emissions of NOx, S02 and PM2.5 are used
to prepare spatial fields of annual PM2.5 concentrations. Emissions scaling ratios, described
below, that are used to prepare the air quality spatial fields.
3A.1 Air Quality Modeling Platform for the ACE Rule
As part of the ACE assessment we used existing air quality modeling for 2011 and 2023
to estimate PM2.5 and ozone concentrations in the future years analyzed for the ACE final rule.
The modeling platform consists of several components including the air quality model,
meteorology, estimates of international transport, and base year and future year emissions from
anthropogenic and natural sources. An overview of each of these platform comments is provided
in the subsections below.
3A. 1.1 Air Quality Model, Meteorology and Boundary Conditions
We used the Comprehensive Air Quality Model with Extensions (CAMx version 6.40)
with the Carbon Bond chemical mechanism CB6r4 for modeling base year and future year ozone
and PM2.5 concentrations (Ramboll, 2016). CAMx is a three-dimensional grid-based
photochemical air quality model designed to simulate the formation and fate of oxidant
precursors, primary and secondary particulate matter concentrations, and deposition over
3A-1
-------
national, regional and urban spatial scales. Consideration of the different processes (e.g.,
transport and deposition) that affect primary (directly emitted) and secondary (formed by
atmospheric processes) pollutants in different locations is fundamental to understanding and
assessing the effects of emissions on air quality concentrations.
The geographic extent of the modeling domain covers the 48 contiguous states along with
the southern portions of Canada and the northern portions of Mexico as shown in Figure 1. This
modeling domain contains 25 vertical layers with a top at about 17,550 meters1 and horizontal
grid resolution of 12 km x 12 km. The model simulations produce hourly air quality
concentrations for each 12-km grid cell across the modeling domain.
1
¦K VCl '
LTOt^n: UIMOOm H»a*"-Vn
tol JW row.it* t \ >»
fa
if
(
N ^
1 iJb,
Figure 3A-1. Air Quality Modeling Domain
The 2011 meteorological data for air quality modeling were derived from running
Version 3.4 of the Weather Research Forecasting Model (WRF) (Skamarock, et al., 2008). The
meteorological outputs from WRF include hourly-varying horizontal wind components (i.e.,
speed and direction), temperature, moisture, vertical diffusion rates, and rainfall rates for each
vertical layer in each grid cell. The 2011 meteorology was used for both the 2011 base year and
2023 future year air quality modeling. Details of the annual 2011 meteorological model
simulation and evaluation are provided in a separate technical support document (US EPA,
1 Since the model top is defined based on atmospheric pressure, the actual height of the model top varies somewhat
with time and location.
3A-2
-------
2014a) which can be obtained at:
http://www.epa.gov/ttn/scram/reports/MET_TSD_2011_final_ll-26-14.pdf
The lateral boundary and initial species condition concentrations are provided by a three-
dimensional global atmospheric chemistry model, GEOS-Chem (Yantosca, 2004) standard
version 8-03-02 with 8-02-01 chemistry. The global GEOS-Chem model simulates atmospheric
chemical and physical processes driven by assimilated meteorological observations from the
NASA's Goddard Earth Observing System (GEOS-5).2 GEOS-Chem was run for 2011 with a
grid resolution of 2.0 degrees x 2.5 degrees (latitude-longitude). The predictions were used to
provide one-way dynamic boundary condition concentrations at three-hour intervals and an
initial concentration field for the CAMx simulations. The 2011 boundary concentrations from
GEOS-Chem were used for both the 2011 and 2023 model simulations. The procedures for
translating GEOS-Chem predictions to initial and boundary concentrations are described
elsewhere (Henderson, 2014). More information about the GEOS-Chem model and other
applications using this tool is available at: http://www-as.harvard.edu/chemistry/trop/geos.
3A.1.2 2011 and 2023 Emissions
The purpose of the 2011 base year modeling is to represent the year 2011 in a manner
consistent with the methods used in the 2023 future year base case. The emissions data in this
platform are primarily based on the 2011 National Emissions Inventory (NEI) v2 for point
sources, nonpoint sources, commercial marine vessels, nonroad mobile sources and fires.3 The
onroad mobile source emissions are similar to those in the 2011 NEIv2, but were generated using
the 2014a version of the Motor Vehicle Emissions Simulator (MOVES2014a)
(http://www.epa.gov/otaq/models/moves/). The 2011 and 2023 emission inventories incorporate
revisions implemented based on comments received on the Notice of Data Availability (NODA)
2 Additional information is available at:
http://gmao.gsfc.nasa.gov/GEOS/ and http://wiki.seas.harvard.edu/geos-chem/index.php/GEOS-5).
3 Note that EPA used a more recent 2016-based emissions platform for air quality modeling to provide the
foundational data needed to identify receptors and interstate contributions for the proposed rule. The 2016-based
mobile emissions platform data were based on MOVES2014b. The 2016-based emissions platform is described in
the Emissions Modeling Technical Support Document available at: https://www.epa.gov/air-emissions-
modeling/2016vl-platform. Although the modeling data in the ACE approach are based on the 2011 platform (and
the 2011-based platform mobile emissions data were developed using MOVES2014a), the state-EGU contribution
modeling data, as described in this appendix, provide a means to develop spatial fields of air quality for the 2021
and 2025 baseline and the proposal and alternative control scenarios analyzed in this RIA.
3A-3
-------
issued in January 2017 "Preliminary Interstate Ozone Transport Modeling Data for the 2015
Ozone National Ambient Air Quality Standard" (82 FR 1733), along with revisions made from
prior notices and rulemakings on earlier versions of the 2011 platform. The preparation of the
emission inventories for air quality modeling is described in the Technical Support Document
(TSD) Additional Updates to Emissions Inventories for the Version 6.3, 2011 Emissions
Modeling Platform for the Year 2023 (US EPA, 2017a). Electronic copies of the emission
inventories and ancillary data used to produce the emissions inputs to the air quality model are
available from the 2011 en and 2023 en section of the EPA Air Emissions Modeling website for
the 2011v6.3 emissions modeling platform: https://www.epa.gov/air-emissions-modeling/2011-
version-63-platform.
The emission inventories for the 2023 ACE future year were developed using projection
methods that are specific to the type of emission source. Future emissions are projected from the
2011 current year either by running models to estimate future year emissions from specific types
of emission sources (e.g., EGUs, and onroad and nonroad mobile sources)4, or for other types of
sources by adjusting the base year emissions according to the best estimate of changes expected
to occur in the intervening years. For sectors which depend strongly on meteorology (such as
biogenic and fires), the same emissions are used in the base and future years to be consistent with
the 2011 meteorology used when modeling 2023. For the remaining sectors, rules and specific
legal obligations that go into effect in the intervening years, along with changes in activity for
the sector, are considered when possible. Emissions inventories for neighboring countries used in
our modeling are included in this platform, specifically 2011 and 2023 emissions inventories for
Mexico, and 2013 and 2025 emissions inventories for Canada. The meteorological data used to
create and temporalize emissions for the future year cases is held constant and represents the
year 2011. The same ancillary data files5 are used to prepare the future year emissions
inventories for air quality modeling as were used to prepare the 2011 base year inventories with
the exception of chemical speciation profiles for mobile sources and temporal profiles for EGUs.
4 California provided emissions for the modeling platform. As such, onroad mobile source emissions for California
were consistent with the emissions provided by the state.
5 Ancillary data files include temporal, spatial, and VOC/PM2 5 chemical speciation surrogates.
3A-4
-------
The projected EGU emissions reflect the emissions reductions expected due to the Final
Mercury and Air Toxics (MATS) rule announced on December 21, 2011, the Cross-State Air
Pollution Rule (CSAPR) issued July 6, 2011, and the CSAPR Update issued October 26, 2016.
The 2023 EGU projected inventory was developed using an engineering analysis approach. EPA
started with 2016 reported, seasonal, historical emissions for each unit. The emissions data for
NOx and SO2 for units that report data under either the Acid Rain Program (ARP) and/or the
CSAPR were aggregated to the summer/ozone season period (May-September) and winter/non-
ozone period (January-April and October-December).6 Adjustments to 2016 levels were made to
account for retirements, coal to gas conversion, retrofits, state-of-the-art combustion controls,
along with other unit-specific adjustments. Details and these adjustments, and information about
handling for units not reporting under Part 75 and pollutants other than NOx and SO2 are
described in the emissions modeling TSD (US EPA, 2017a).
The 2023 non-EGU stationary source emissions inventory includes impacts from
enforceable national rules and programs including the Reciprocating Internal Combustion
Engines (RICE) and cement manufacturing National Emissions Standards for Hazardous Air
Pollutants (NESHAPs) and Boiler Maximum Achievable Control Technology (MACT)
reconsideration reductions. Projection factors and percent reductions for non-EGU point sources
reflect comments received by EPA in response to the January 2017 NOD A, along with emissions
reductions due to national and local rules, control programs, plant closures, consent decrees and
settlements. Growth and control factors provided by states and by regional organizations on
behalf of states were applied. Reductions to criteria air pollutant (CAP) emissions from
stationary engines resulting as co-benefits to the Reciprocating Internal Combustion Engines
(RICE) National Emission Standard for Hazardous Air Pollutants (NESHAP) are included.
Reductions due to the New Source Performance Standards (NSPS) VOC controls for oil and gas
sources, and the NSPS for process heaters, internal combustion engines, and natural gas turbines
were also included.
6 EPA notes that historical state-level ozone season EGU NOx emission rates are publicly available and quality
assured data. They are monitored using continuous emissions monitors (CEMs) data and are reported to EPA
directly by power sector sources. They are reported under Part 75 of the CAA.
3A-5
-------
For point and nonpoint oil and gas sources, state projection factors were generated using
state-specific historical oil and gas production data available from EIA for 2011 to 2015 and
information from regional factors based AEO 2017 to project the emission to the year 2023. Co-
benefits of stationary engines CAP reductions (RICE NESHAP) and controls from New Source
Performance Standards (NSPS) are reflected for select source categories. Mid-Atlantic Regional
Air Management Association (MARAMA) factors for the year 2023 were used where applicable.
Projection factors for other nonpoint sources such as stationary source fuel combustion,
industrial processes, solvent utilization, and waste disposal, reflect emissions reductions due to
control programs along with comments on the growth and control of these sources as a result of
the January 2017 NOD A and information gathered from prior rulemakings and outreach to states
on emission inventories.
The MOVES2014a-based 2023 onroad mobile source emissions account for changes in
activity data and the impact of on-the-books national rules including: the Tier 3 Vehicle
Emission and Fuel Standards Program, the 2017 and Later Model Year Light-Duty Vehicle
Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards (LD GHG), the
Renewable Fuel Standard (RFS2), the Mobile Source Air Toxics Rule, the Light Duty Green
House Gas/Corporate Average Fuel Efficiency (CAFE) standards for 2012-2016, the Greenhouse
Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines
and Vehicles, the Light-Duty Vehicle Tier 2 Rule, and the Heavy-Duty Diesel Rule. The
MOVES-based emissions also include state rules related to the adoption of LEV standards,
inspection and maintenance programs, Stage II refueling controls, and local fuel restrictions.
The nonroad mobile 2023 emissions, including railroads and commercial marine vessel
emissions also include all national control programs. These control programs include the Clean
Air Nonroad Diesel Rule - Tier 4, the Nonroad Spark Ignition rules, and the Locomotive-Marine
Engine rule. For ocean-going vessels (Class 3 marine), the emissions data reflect the 2005
voluntary Vessel Speed Reduction (VSR) within 20 nautical miles, the 2007 and 2008 auxiliary
engine rules, the 40 nautical mile VSR program, the 2009 Low Sulfur Fuel regulation, the 2009-
2018 cold ironing regulation, the use of 1 percent sulfur fuel in the Emissions Control Area
(ECA) zone, the 2012-2015 Tier 2 NOx controls, the 2016 0.1 percent sulfur fuel regulation in
ECA zone, and the 2016 International Marine Organization (IMO) Tier 3 NOx controls. Non-
3A-6
-------
U.S. and U.S. category 3 commercial marine emissions were projected to 2025 using consistent
methods that incorporated controls based on ECA and IMO global NOx and SO2 controls.
3A.1.3 2011 Model Evaluation for Ozone andPM2.s
An operational model performance evaluation was conducted to examine the ability of
the 2011 base year model run to simulate the corresponding 2011 measured ozone and PM2.5
concentrations. This evaluation focused on four statistical metrics comparing model predictions
to the corresponding observations. The performance statistics include mean bias, mean error,
normalized mean bias, and normalized mean error. Mean bias (MB) is the sum of the difference
(predicted - observed) divided by the total number of replicates (n). Mean bias is given in units
of ppb and is defined as:
Where:
• Pis the model-predicted concentration;
• O is the observed concentrations; and
• n is the total number of observations
Mean error (ME) calculates the sum of the absolute value of the difference (predicted -
observed) divided by the total number of replicates (n). Mean error is given in units of ppb and is
defined as:
Normalized mean bias (NMB) is the sum of the difference (predicted - observed) over the
sum of observed values. NMB is a useful model performance indicator because it avoids over
inflating the observed range of values, especially at low concentrations. Normalized mean bias is
given in percentage units and is defined as:
MB= -Zi(P-O)
n
(Eq-l)
ME = ±Z?|P-0|
(Eq-2)
NMB = ^ 0) * 100
Zi(o)
(Eq-3)
3A-7
-------
Normalized mean error (NME) is the sum of the absolute value of the difference
(predicted - observed) divided by the sum of observed values. Normalized mean error is given in
percentage units and is defined as:
NME=sasf1*100
-------
U.S. Climate Regions
TX X LA ^
IIS
Figure 3A-2. NOAA Climate Regions
Model performance statistics for PM2.5 for each region are provided in Table 3 A. 1. These
data indicate that over the year as a whole, PM2.5 is over predicted in the Northeast, Ohio Valley,
Upper Midwest, Southeast, and Northwest regions and under predicted in the South and
Southwest regions. Normalized mean bias is within ±30 percent in all regions except the
Northwest which has somewhat larger model over-predictions. Model performance for PM2.5 for
the 2011 modeling platform is similar to the model performance results for other contemporary,
state of the science photochemical model applications (Simon et al., 2012). Additional details on
PM2.5 model performance for the 2011 base year model run can be found in the Technical
Support Document for EPA's preliminary regional haze modeling (US EPA, 2017b).
3A-9
-------
Table 3A.1. Model Performance Statistics by Region for PM2.5
Region
Network
No. of Obs
MB
frig/m3)
ME
frig/m3)
NMB
(%)
NME
(%)
Northeast
IMPROVE
1577
0.87
2.21
17.70
44.90
CSN
2788
0.97
4.04
9.70
40.40
Ohio Valley
IMPROVE
680
0.10
2.96
1.20
35.50
CSN
2475
0.13
3.85
1.10
32.80
Upper Midwest
IMPROVE
CSN
700
1343
0.83
1.37
2.37
3.66
14.20
13.60
40.40
36.30
Southeast
IMPROVE
1172
0.52
3.54
6.30
43.20
CSN
1813
0.19
3.92
1.70
34.20
South
IMPROVE
933
-0.47
2.69
-6.50
37.40
CSN
962
-0.08
4.48
-0.75
39.50
Southwest
IMPROVE
3695
-1.12
1.86
-28.00
46.30
CSN
746
-0.08
3.93
-1.00
47.10
N. Rockies/
IMPROVE
1952
0.07
1.39
2.40
44.90
Plains
CSN
275
-2.07
4.18
-21.80
43.90
Northwest
IMPROVE
1901
1.19
2.28
43.20
82.90
CSN
668
5.77
7.25
69.90
87.90
West
IMPROVE
1782
-1.08
2.08
-25.30
48.50
CSN
936
-2.92
5.08
-23.10
40.30
Model performance statistics for May through September MDA8 ozone concentrations for
each region are provided in Table 3A.2. Overall, measured ozone is under predicted in most
regions, except for the Northeast and Southeast where over prediction is found. Normalized
mean bias is within ±15 percent in all regions. Model performance for ozone for the 2011
modeling platform is similar to the model performance results for other contemporary, state of
the science photochemical model applications (Simon et al., 2012). Additional details on ozone
model performance for the 2011 base year model run can be found in the Air Quality Technical
Support Document for EPA's preliminary interstate ozone transport modeling for the 2015 ozone
National Ambient Air Quality Standard (US EPA, 2017c).
3 A-10
-------
Table 3A.2. Model Performance Statistics by Region for Ozone on Days Above 60 ppb
(May-Sep)
Region
No. of Obs
MB
(ppb)
ME
(ppb)
NMB
(%)
NME
(%)
Northeast
4085
1.20
7.30
1.80
10.70
Ohio Valley
6325
-0.60
7.50
-0.90
11.10
Upper Midwest
1162
-4.00
7.60
-5.90
11.10
Southeast
4840
2.30
6.80
3.40
10.20
South
5694
-5.30
8.40
-7.60
12.20
Southwest
6033
-6.20
8.50
-9.40
12.90
N. Rockies/Plains
380
-7.20
8.40
-11.40
13.40
Northwest
79
-5.60
9.00
-8.70
14.00
West
8655
-8.60
10.30
-12.20
14.50
Thus, the model performance results demonstrate the scientific credibility of our 2011
modeling platform for predicting PM2.5 and ozone concentrations. These results provide
confidence in the ability of the modeling platform to provide a reasonable projection of expected
future year ozone concentrations and contributions.
3A.2 Source Apportionment Tags
CAMx source apportionment modeling was used to track ozone and PM2.5 component
species impacts from pre-defined groups of emissions sources (source tags). Separate tags were
created for state-level EGUs split by fuel type (coal units versus non-coal units13). For some
states with low EGU emissions, EGUs are grouped with nearby states that also have low EGU
emissions. In addition, there are no coal EGUs operating in the 2023 emissions case for the
following states: Idaho, Oregon, and Washington. Therefore, there is no coal EGU tag for those
states. Similarly, there were no EGUs (coal or non-coal) in Washington D.C. in the 2023
emissions scenario, so there were no EGU tags for Washington D.C. There were also several
domain-wide tags for sources other than EGUs. Table 3 A.3 provides a full list of the emissions
group tags that were tracked in the source apportionment modeling.
13 For the purposes of this analysis non-coal fuels include emissions from natural gas, oil, biomass, and waste coal-
fired EGUs.
3 A-11
-------
Table 3A.3. Source Apportionment Tags
Coal-fired EGU tags
Non-coal EGU tags
Domain-wide tags
Alabama
Arizona
Arkansas
California
Colorado
Connecticut + Rhode Island
Delaware + New Jersey
Florida
Georgia
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine + Mass. + New Hamp. +
Vermont
Maryland
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Mexico
New York
North Carolina
North Dakota + South Dakota
Ohio
Oklahoma
Pennsylvania
South Carolina
Tennessee
Texas
Utah
Virginia
West Virginia
Wisconsin
Wyoming
Tribal Data*
Alabama
Arizona
Arkansas
California
Colorado
Connecticut + Rhode Island
Delaware + New Jersey
Florida
Georgia
Idaho + Oregon + Washington
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine + Mass. + New Hamp. +
Vermont
Maryland
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
New Mexico
New York
North Carolina
North Dakota + South Dakota
Ohio
Oklahoma
Pennsylvania
South Carolina
Tennessee
Texas
Utah
Virginia
West Virginia
Wisconsin
Wyoming
Tribal Data
EGU retirements
through 2025
EGU retirements
2026-2030
All U.S.
anthropogenic
emissions from
source sectors
other than EGUs
International
within-domain
emissions
(sources
occurring in
Canada, Mexico,
and from
offshore marine
vessels and
drilling
platforms)
Fires (wildfires
and prescribed
fires)
Biogenic sources
Boundary
conditions
14
14 EGUs operating on tribal lands were tracked together in a single tag. There are EGUs on tribal land in the
following states: Utah (coal), New Mexico (coal), Arizona (coal and non-coal), Idaho (non-coal). EGU emissions
occurring on tribal lands were not included in the state-level EGU source tags.
3A-12
-------
The contributions represent the spatial and temporal distribution of the emissions within
each source tag. Thus, the contribution modeling results do not allow us to represent any changes
to any "within tag" spatial distributions. For example, the location of coal-fired EGUs in
Michigan are held in place based on locations in the 2023 emissions. Additionally, the relative
magnitude of sources within a source tag do not change from what was modeled with the 2023
emissions inventory.
3A.3 Applying Source Apportionment Contributions to Create Air Quality Fields
We created air quality surfaces for the ACE future year baseline and illustrative policy
scenarios by scaling the EGU sector tagged contributions from the 2023 modeling based on
relative changes in EGU emissions associated with each tagged category between the 2023
emissions case and the ACE scenarios. Below, we provide equations used to apply these scaling
ratios along with tables of the ratios.
3A.3.2 Scaling Ratio Applied to Source Apportionment Tags
Scaling ratios for PM2.5 components that are emitted directly from the source (OA, EC,
crustal) were based on relative changes in annual primary PM2.5 emissions between the 2023
emissions case and the ACE baseline and the illustrative policy scenario. Scaling ratios for
components that are formed through chemical reactions in the atmosphere were created as
follows: scaling ratios for sulfate were based on relative changes in annual SO2 emissions;
scaling ratios for nitrate were based on relative changes annual NOx emissions; and scaling
ratios for ozone formed in NOx-limited regimes15 ("03N") were based on relative changes in
ozone season (May-September) NOx emissions. The scaling ratios that were determined based
on emissions provided for each scenario.
Scaling ratios were applied to create air quality surfaces for ozone using equation (9):
15 The CAMx model internally determines whether the ozone formation regime is NOx-limited or VOC-limited
depending on predicted ratios of indicator chemical species.
3 A-13
-------
Ozonerrlgdiy Cm,g,d,BC Cm,g,d,int ^m,g,d,bio ^m,g,d,fires
T
Cm,g,d,USanthro Cm,g,d,y,EGUret / CVOC,m,g,d,t
I
^ ' CNOx,m,g,d,t^t,i,y
t= 1
t= 1
T
(Eq-9)
where:
• Ozonemg d i y is the estimated ozone for metric, "m" (MDA8 or MDA1), grid-
cell, "g", day, "d", scenario, "i", and year, "y";
• Cm,g,d,Bc's the total ozone contribution from the modeled boundary inflow;
Cm,g,d,int is the total ozone contribution from international emissions within the
model domain;
• Cmg d bio is the total ozone contribution from biogenic emissions;
• Cm,g,d,fireS's the total ozone contribution from fires;
• Cm,g,d,usanthro's the total ozone contribution from U.S. anthropogenic sources
other than EGUs;
• Cm,g,d,y,EGUret's the total ozone contribution from retiring EGUs after year, "y"
(this term is equal to 0 in 2030 and 2035);
• Cvoc,m,g,d,t is the ozone contribution from EGU emissions of VOCs from tag, "t";
• CNOx,m,g,d,t is the ozone contribution from EGU emissions of NOx from tag, "t";
and
• St i y is the ozone scaling ratio for tag, "t", scenario, "i", and year, "y".
Scaling ratios were applied to create air quality surfaces for PM2.5 species using equation
(10) (for sulfate, nitrate, EC or crustal material) or using equation (11) (for OA):
3A-14
-------
s,g,d,USanthro
s,g,d,y,EGUret
(Eq-10)
OAgdiy — CpQA g age + CpoA,g,d,int C po A,g ,cl,bio CpOA,g,d,fires
CpOA,g,d,USanthro CpOA,g,d,y,EGUret SOAg Ci
T
(Eq-11)
^ ' CpOA,g,d,tSpri,t,i,y
t= 1
PMSig:d,i,y is the estimated concentration for species, "s" (sulfate, nitrate, EC, or crustal
material), grid-cell, "g", day, "d", scenario, "i", and year, "y";
Cs,g,d,BC is the species contribution from the modeled boundary inflow;
Cs,g,d,int is the species contribution from international emissions within the model
domain;
Cs,g,d,bio is the species contribution from biogenic emissions;
Cs,g,d,fires is the species contribution from fires;
Cs,g,d,usanthro is the species contribution from U.S. anthropogenic sources other than
EGUs;
Cs,g,d,y,EGUret is the species contribution from retiring EGUs after year, "y" (this term is
equal to 0 in 2030 and 2035);
Cs,g,d,t is the species contribution from EGU emissions from tag, "t"; and
Ss,t,i,y is the scaling ratio for species, "s", tag, "t", scenario, "i", and year, "y".
3 A-15
-------
Similarly, for Equation (11):
• 0Ag d i y is the estimated OA concentration for grid-cell, "g", day, "d", scenario, "i",
and year, "y";
• Each of the contribution terms refers to the contribution to primary OA (POA); and
• SOAg d represents the modeled secondary organic aerosol concentration for gird-
cell, "g", and day, "d", which does not change among scenarios
3A.4 Creating Fused Fields Based on Observations and Model Surfaces
In this section we describe steps taken to estimate PM2.5 and ozone gridded surfaces
associated with the baseline and the illustrative policy scenario for every year. For PM2.5, (daily
gridded PM2.5 species were processed into annual average surfaces which combine observed
values with model predictions using the enhanced Veronoi Neighbor Average (eVNA) method
(Gold et al., 1997; US EPA, 2007; Ding et al., 2015). These steps were performed using EPA's
software package, Software for the Modeled Attainment Test - Community Edition (SMAT-
CE)16 and have been previously documented both in the user's guide for the predecessor
software (Abt, 2014) and in EPA's modeling guidance document (U.S. EPA, 2014b). First, we
create a 2011 eVNA surface for each PM component species. To create the 2011 eVNA surface,
SMAT-CE first calculates quarterly average values (January-March; April-June; July-September;
October-December) for each PM2.5 component species at each monitoring site with available
measured data. For this calculation we used 3 years of monitoring data (2010-2012)17. SMAT-
CE then creates an interpolated field of the quarterly-average observed data for each PM2.5
component species using inverse distance squared weighting resulting in a separate 3-year
average interpolated observed field for each PM2.5 species and each quarter. The interpolated
observed fields are then adjusted to match the spatial gradients from the modeled data. These two
steps can be calculated using Equation (12):
16 Software download and documentation available at https://www.epa.gov/scram/photochemical-modeling-tools
17 Three years of ambient data is used to provide a more representative picture of air pollution concentrations.
3 A-16
-------
eVNAg,s,q,2011 = Z VKeightxMonitorxs q 2010_2012 ^!aA<*'2011 (Eq-12)
Moaetx,s,q,20ll
Where:
• eVNAg s q current is the gradient adjusted quarterly-average eVNA value at grid-
cell, g, for PM component species, s, during quarter, q for the year 2011;
• Weightx is the inverse distance weight for monitor x at the location of grid-cell,
g;
• Monitorxsq,2010-2012 is the 3-year (2010-2012) average of the quarterly
monitored concentration for species, s, at monitor, x, during quarter, q;
• Modelgsq 2011 is the 2011 modeled quarterly-average concentrations of species,
s, at grid cell, g, during quarter, q; and
• Modelx s q 2011 is the 2011 modeled quarterly-average concentration of species, s,
at the location of monitor, x, during quarter q.
The 2011 eVNA field serves as the starting point for future-year projections. To create a
gridded future-year eVNA surfaces for the baseline and ACE illustrative policy, we take the ratio
of the modeled future year18 quarterly average concentration to the modeled 2011 concentration
in each grid cell and multiply that by the corresponding 2011 eVNA quarterly PM2.5 component
species value in that grid cell (Equation 13).
eVNAgMmu„ = (eVNAgSiqml) x e(Eq-13)
lvIuutiLg,s,q, 2 011
This results in a gridded future-year projection which accounts for adjustments to match
observations in the 2011 modeled data.
Finally, particulate ammonium concentrations are impacted both by emissions of
precursor ammonia gas as well as ambient concentrations of particulate sulfate and nitrate.
18 In this analysis the "future year" modeled concentration is the result of Equations 9, 10 or 11 that represents either
the ACE scenarios.
3 A-17
-------
Because of uncertainties in ammonium speciation measurements combined with sparse
ammonium measurements in rural areas, the SMAT-CE default is to calculate ammonium values
using the degree of sulfate neutralization (i.e., the relative molar mass of ammonium to sulfate
with the assumption that all nitrate is fully neutralized). Degree of neutralization values are
mainly available in urban areas while sulfate measurements are available in both urban and rural
areas. Ammonium is thus calculated by multiplying the interpolated degree of neutralization
value by the interpolated sulfate value at each grid-cell location which allows the ammonium
fields to be informed by rural sulfate measurements in locations where no rural ammonium
measurements are available. The degree of neutralization is not permitted to exceed the
maximum theoretical molar ratio of 2:1 for ammonium:sulfate. When creating the future year
surface for particulate ammonium, we use the default SMAT-CE assumption that the degree of
neutralization for the aerosol remains at 2011 levels.
A similar method for creating future-year eVNA surfaces is followed for the two ozone
metrics with a few key differences. First, while PM2.5 is split into quarterly averages and then
averaged up to an annual value, we look at ozone as a summer-season average using definitions
that match metrics from epidemiology studies (May-Sep for MDA8 and Apr-Oct for MDA1).
The other main difference in the SMAT-CE calculation for ozone is that the spatial interpolation
of observations uses an inverse distance weighting rather than an inverse distance squared
weighting. This results in interpolated observational fields that better replicate the more gradual
spatial gradients observed in ozone compared to PM2.5.
3 A-18
-------
3A.5 References
82 FR 1733. Notice of Availability of the Environmental Protection Agency's Preliminary
Interstate Ozone Transport Modeling Data for the 2015 Ozone National Ambient Air
Quality Standard (NAAQS), (January 6, 2017).
Abt Associates, 2014. User's Guide: Modeled Attainment Test Software.
http://www.epa.gov/scram001/modelingapps_mats.htm
Cohan, D.S., Hakami, A., Hu, Y.T., Russell, A.G., 2005. Nonlinear response of ozone to
emissions: Source apportionment and sensitivity analysis. Environ. Sci. Technol. 39,
6739-6748.
Cohan, D.S., Napelenok, S.L., 2011. Air Quality Response Modeling for Decision Support.
Atmosphere. 2, 407-425.
Ding, D., Zhu, Y., Jang, C., Lin, C., Wang, S., Fu, J., Gao, J., Deng, S., Xie, J., Qui, X., 2015.
Evaluation of heath benefit using BenMAP-CE with an integrated scheme of model and
monitor data during Guangzhou Asian Games. Journal of Environmental Science. 29,
178-188.
Dunker, A.M., Yarwood, G., Ortmann, J.P., Wilson, G.M., 2002. The decoupled direct method
for sensitivity analysis in a three-dimensional air quality model—Implementation,
accuracy, and efficiency. Environ. Sci. Technol. 36, 2965-2976.
Gold C, Remmele P.R., Roos T., 1997. In: Algorithmic Foundation of Geographic Information
Systems. In: Lecture Notes in Computer Science, Vol. 1340 (van Kereveld M, Nievergelt
J, Roos T, Widmayer P, eds) Berlin, Germany: Springer-Verlag. Voronoi methods in
GIS. pp. 21-35.
Henderson, B.H., Akhtar, F., Pye, H.O.T., Napelenok, S.T., Hutzell, W.T., 2014. A Database and
Tool for Boundary Conditions for Regional Air Quality Modeling: Description and
Evaluations, Geoscientific Model Development. 7, 339-360.
Koo, B., Dunker, A.M., Yarwood, G., 2007. Implementing the decoupled direct method for
sensitivity analysis in a particulate matter air quality model. Environ. Sci. Technol. 41,
2847-2854.
Napelenok, S.L., Cohan, D.S., Hu, Y., Russell, A.G., 2006. Decoupled direct 3D sensitivity
analysis for particulate matter (DDM-3D/PM). Atmospheric Environment. 40, 6112-
6121.
Ramboll Environ, 2016. User's Guide Comprehensive Air Quality Model with Extensions
version 6.40. Ramboll Environ International Corporation, Novato, CA.
Simon, H., Baker, K. R., Phillips, S., 2012. Compilation and interpretation of photochemical
model performance statistics published between 2006 and 2012, Atmos. Environ. 61,
124-139.
3 A-19
-------
Skamarock, W.C., Klemp, J.B., Dudhia,J., Gill, D.O., Barker, D.M., Duda, M.G., Huang, X.-Y.,
Wang, W., Powers, J.G., 2008. A Description of the Advanced Research WRF Version 3.
NCAR Tech. Note NCAR/TN-475+STR.
(http://wwww.mmm.ucar.edu/wrf/users/docs/arw_v3.pdf).
US EPA, 2007, Technical Report on Ozone Exposure, Risk, and Impact Assessments for
Vegetation. EPA 452/R-07-002. Prepared by Abt Associates Inc. for U.S. Environmental
Protection Agency, Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division. Research Triangle Park, NC.
(https://www3.epa.gov/ttn/naaqs/standards/ozone/data/2007_01_environmental_tsd.pdf).
US EPA, 2014a. Meteorological Model Performance for Annual 2011 Simulation WRF v3.4,
Research Triangle Park, NC. (http://www.epa.gov/scram001/).
US EPA, 2014b, Modeling Guidance for Demonstrating Attainment of Air Quality Goals for
Ozone, PM2.5, and Regional Haze- December 2014 DRAFT, Research Triangle Park,
NC. (https://www3.epa.gov/ttn/scram/guidance/guide/Draft_03-PM-
RH_Modeling_Guidance-2014.pdf).
US EPA, 2015, Regulatory Impact Analysis of the Final Revisions to the National Ambient Air
Quality Standards for Ground-Level Ozone, EPA-452/R-15-07, Research Triangle Park,
NC. (https://www.epa.gov/sites/production/files/2016-02/documents/20151001ria.pdf).
US EPA, 2017a, Technical Support Document (TSD) Additional Updates to Emissions
Inventories for the Version 6.3, 2011 Emissions Modeling Platform for the Year 2023,
Research Triangle Park, NC. (https://www.epa.gov/sites/production/files/2017-
11/documents/ 201 Iv6.3_2023en_update_emismod_tsd_oct2017.pdf).
US EPA, 2017b. Documentation for the EPA's Preliminary 2028 Regional Haze Modeling.
Research Triangle Park, NC
(https://www3.epa.gov/ttn/scram/reports/2028_Regional_Haze_Modeling-TSD.pdf).
US EPA, 2017c. Air Quality Modeling Technical Support Document for the 2015 Ozone
NAAQS Preliminary Interstate Transport Assessment. Research Triangle Park, NC
(https://www.epa.gov/airmarkets/notice-data-availability-preliminary-interstate-ozone-
transport-modeling-data-2015 -ozone).
Yantosca, B. 2004. GEOS-CHEMv7-01-02 User's Guide, Atmospheric Chemistry Modeling
Group, Harvard University, Cambridge, MA.
Zavala, M., Lei, W., Molina, M.J., Molina, L.T., 2009. Modeled and observed ozone sensitivity
to mobile-source emissions in Mexico City. Atmos. Chem. Phys. 9, 39-55.
3A-20
-------
CHAPTER 4: COST, EMISSIONS, AND ENERGY IMPACTS
Overview
This chapter reports the compliance costs, emissions, and energy analyses performed for
the Revised CSAPR Update proposed rule. EPA used the Integrated Planning Model (IPM) to
conduct most of the analysis discussed in this chapter. As explained in detail below, this chapter
presents analysis for three regulatory control alternatives that differ in the level of electric
generating units (EGU) nitrogen oxides (NOx) ozone season emissions budgets in 12 states
subject to this action.1 These regulatory control alternatives impose different budget levels based
on alternative assumptions about the possible actions that EGUs may be able to pursue to reduce
their NOx emissions.
The chapter is organized as follows: following a summary of the regulatory control
alternatives analyzed and a summary of EPA's methodology, we present estimates of compliance
costs, as well as estimated impacts on emissions, generation, capacity, fuel use, fuel price, and
retail electricity price.
4.1 Regulatory Control Alternatives
Of the 22 states currently covered by the Cross-State Air Pollution Rule (CSAPR) NOx
Ozone Season Group 2 trading program, EPA is proposing to establish revised budgets for 12
states. Therefore, EPA is proposing the creation of an additional geographic group and ozone
season 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 and will no longer be subject to Group 2
budgets. Aside from the removal of the 12 covered states from the current Group 2 program, this
proposal leaves unchanged the budget stringency and geography of the existing CSAPR NOx
Ozone Season Group 1 and Group 2 trading programs. The EGUs covered by the FIPs and
subject to the budget are all fossil-fired EGUs with >25 megawatt (MW) capacity.
1 The 12 states for which EPA is proposing to promulgate FIPs to reduce interstate ozone transport for the 2008
ozone NAAQS are listed in Table I. A-2 of the preamble and include Illinois, Indiana, Kentucky, Louisiana,
Maryland, Michigan, New York, New Jersey, Ohio, Pennsylvania, Virginia, and West Virginia.
4-1
-------
This RIA evaluates the benefits, costs and certain impacts of compliance with three
regulatory control alternatives: the Revised CSAPR Update proposed rule, a less-stringent
alternative, and a more-stringent alternative. For details on the derivation of these budgets, please
see Section VII of the preamble. Aside from the difference in emission budgets, other key
regulatory features of the allowance trading program, such as the ability to bank allowances for
future use, are the same across all the three different sets of NOx emissions budgets analyzed.
4.1.2 Regulatory Control Alternatives Analyzed
In accordance with Executive Orders 12866 and 13563, the guidelines of OMB Circular A-
4, and EPA's Guidelines for Preparing Economic Analyses, this RIA analyzes the benefits and
costs associated with complying with the Revised CSAPR Update proposed rule. The Revised
CSAPR Update proposed emission budgets in this RIA represent EGU NOx ozone season
emission budgets for each state that were developed using uniform control stringency
represented by $1,600 per ton of NOx (2016$).2 This RIA analyzes the Revised CSAPR Update
proposed emission budgets, as well as a more and a less stringent alternative to the Revised
CSAPR Update proposal. The more and less stringent alternatives differ from the Revised
CSAPR Update proposal in that they set different NOx ozone season emission budgets for the
affected EGUs. The less-stringent scenario uses emission budgets that were developed using
uniform control stringency represented by $500 per ton of NOx (2016$). The more-stringent
scenario uses emission budgets that were developed using uniform control stringency
represented by $9,600 per ton of NOx (2016$). For details, please see EGU NOx Mitigation
Strategies Proposed Rule TSD, in the docket for this proposed rule.3
Table 4-1 reports the EGU NOx ozone season emission budgets that are evaluated in this
RIA. As described above, starting in 2021, emissions from affected EGUs in the 12 states cannot
exceed the sum of emissions budgets but for the ability to use banked allowances from previous
years for compliance. No further reductions in budgets occur after 2024, and budgets remain in
place for future years. Furthermore, emissions from affected EGUs in a particular state are
subject to the CSAPR assurance provisions, which require additional allowance surrender
2 The budget setting process is described in section VIII of the preamble and in detail in the Ozone Transport Policy
Analysis Technical Support Document (TSD).
3 Docket ID No. EPA-HQ-OAR-2020-0272
4-2
-------
penalties (a total of 3 allowances per ton of emissions) on emissions that exceed a state's CSAPR
NOx ozone season assurance level, or 121 percent of the emissions budget. Similar to the
approach taken in the CSAPR Update, EPA is proposing a one-time conversion of banked Group
2 allowances according to a formula. The size of the initial bank would be set at a level that
would ensure that the use of these converted allowances, in addition to the allowances provided
in the states' emissions budgets under the Group 3 trading program, would not authorize
emissions in the trading program region in the first year of the program to exceed the sum of the
states' budgets by more than the sum of the states' variability limits. The CSAPR NOx ozone
season allowance trading program is described in further detail in Section VIII of the preamble.
Table 4-1. NOx Ozone Season Emission Budgets (Tons) Evaluated
Revised CSAPR Update Proposal
State
2021
2022
2023
2024
2025
Illinois
9,444
9,415
8,397
8,397
8,397
Indiana
12,500
11,998
11,998
9,447
9,447
Kentucky
14,384
11,936
11,936
11,936
11,936
Louisiana
15,402
14,871
14,871
14,871
14,871
Maryland
1,522
1,498
1,498
1,498
1,498
Michigan
12,727
11,767
9,803
9,614
9,614
New Jersey
1,253
1,253
1,253
1,253
1,253
New York
3,137
3,137
3,137
3,119
3,119
Ohio
9,605
9,676
9,676
9,676
9,676
Pennsylvania
8,076
8,076
8,076
8,076
8,076
Virginia
4,544
3,656
3,656
3,395
3,395
West
Virginia
13,686
12,813
11,810
11,810
11,810
Total
106,280
100,096
96,111
93,092
93,092
Less-Stringent Alternative
State
2021
2022
2023
2024
2025
Illinois
9,667
9,632
8,579
8,599
8,579
Indiana
15,677
15,206
15,206
12,755
12,603
Kentucky
15,606
15,606
15,606
15,588
15,606
Louisiana
15,442
15,442
15,442
15,488
15,442
Maryland
1,565
1,565
1,565
1,565
1,565
Michigan
13,120
13,120
10,313
10,841
10,116
New Jersey
1,346
1,346
1,346
1,346
1,346
New York
3,182
3,182
3,182
3,169
3,163
4-3
-------
Ohio
15,490
15,560
15,560
15,917
15,560
Pennsylvania
11,487
11,487
11,487
11,570
11,487
Virginia
4,588
4,172
4,172
3,912
3,908
West
Virginia
15,017
15,017
13,272
13,407
13,272
Total
122,187
121,334
115,730
114,156
112,647
More-Stringent Alternative
State
2021
2022
2023
2024
2025
Illinois
9,444
9,415
8,397
7,142
7,142
Indiana
12,500
11,998
11,998
8,264
8,264
Kentucky
14,384
11,936
11,936
8,852
8,852
Louisiana
15,402
14,871
14,871
12,636
12,636
Maryland
1,522
1,498
1,498
1,239
1,239
Michigan
12,727
11,767
9,803
7,315
7,315
New Jersey
1,253
1,253
1,253
1,257
1,257
New York
3,137
3,137
3,137
3,020
3,020
Ohio
9,605
9,676
9,676
9,126
9,126
Pennsylvania
8,076
8,076
8,076
7,578
7,578
Virginia
4,544
3,656
3,656
3,022
3,022
West
Virginia
13,686
12,813
11,810
9,569
9,569
Total
106,280
100,096
96,111
79,020
79,020
Note that EGUs have flexibility in determining how they will comply with the allowance
trading program. As discussed below, the way that they comply may differ from the methods
forecast in the modeling for this RIA. See Section 4.3 for further discussion of the modeling
approach used in the analysis presented below.
4.2 Power Sector Modeling Framework
IPM is a state-of-the-art, peer-reviewed, dynamic linear programming model that can be
used to project power sector behavior under future business-as-usual conditions and to examine
prospective air pollution control policies throughout the contiguous United States for the entire
electric power system. EPA used IPM to project likely future electricity market conditions with
and without the Revised CSAPR Update proposal.
IPM, developed by ICF, is a multi-regional, dynamic, deterministic linear programming
model of the contiguous U.S. electric power sector. It provides estimates of least cost capacity
4-4
-------
expansion, electricity dispatch, and emissions control strategies while meeting energy demand
and environmental, transmission, dispatch, and reliability constraints.4 EPA has used IPM for
almost three decades to better understand power sector behavior under future business-as-usual
conditions and to evaluate the economic and emissions 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.5
The model incorporates a detailed representation of the fossil-fuel supply system that is
used to estimate equilibrium fuel prices. The model uses natural gas fuel supply curves and
regional gas delivery costs (basis differentials) to simulate the fuel price associated with a given
level of gas consumption within the system. These inputs are derived using ICF's Gas Market
Model (GMM), a supply/demand equilibrium model of the North American gas market.6
IPM also endogenously models the partial equilibrium of coal supply and EGU coal
demand levels throughout the contiguous U.S., taking into account assumed non-power sector
demand and imports/exports. IPM reflects 36 coal supply regions, 14 coal grades, and the coal
transport network, which consists of over four thousand linkages representing rail, barge, and
truck and conveyer linkages. The coal supply curves in IPM were developed during a thorough
bottom-up, mine-by-mine approach that depicts the coal choices and associated supply costs that
power plants would face if selecting that coal over the modeling time horizon. The IPM
4 Due to the compliance timing for the Revised CSAPR Update proposal, EPA does not allow IPM to build certain
new capital investments such as new, unplanned natural gas or renewable capacity or new SCR or SNCR through
2025. EPA's compliance modeling does allow for new combustion controls, which represent the most likely
potential capital expenditure in the 2021 analysis year.
5 Detailed information and documentation of EPA's Base Case using IPM (v6), including all the underlying
assumptions, data sources, and architecture parameters can be found on EPA's website at:
http://www.epa.gov/airmarkets/powersectormodeling.html.
6 See Chapter 8 of EPA's Base Case using IPM v6 documentation, available at:
https://www.epa.gov/airmarkets/power-sector-modeling-platform-v6-may-2019.
4-5
-------
documentation outlines the methods and data used to quantify the economically recoverable coal
reserves, characterize their cost, and build the 36 coal regions' supply curves.7
To estimate the annualized costs of additional capital investments in the power sector, 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. The
CRF is derived from estimates of the power sector's cost of capital (i.e., private discount rate),
the amount of insurance coverage required, local property taxes, and the life of capital.8 It is
important to note that there is no single CRF factor applied in the model; rather, the CRF varies
across technologies, book life of the capital investments, and regions in the model in order to
better simulate power sector decision-making.
EPA has used IPM extensively over the past three decades to analyze options for reducing
power sector emissions. Previously, the model has been used to estimate the costs, emission
changes, and power sector impacts for the Clean Air Interstate Rule (U.S. EPA, 2005), the
original Cross-State Air Pollution Rule (U.S. EPA, 2011), the Mercury and Air Toxics Standards
(MATS) (U.S. EPA, 201 la), the Clean Power Plan (CPP) for Existing Power Plants (U.S. EPA,
2015), the Carbon Pollution Standards for New Power Plants (U.S. EPA, 2015), the Affordable
Clean Energy Rule (U.S. EPA, 2019), and the Clean Power Plan Repeal (U.S. EPA, 2019). EPA
has also used IPM to estimate the air pollution reductions and power sector impacts of water and
waste regulations affecting EGUs, including Cooling Water Intakes (316(b)) Rule (U.S. EPA,
2014), Disposal of Coal Combustion Residuals from Electric Utilities (CCR) (U.S. EPA, 2015b)
and Steam Electric Effluent Limitation Guidelines (ELG) (U.S. EPA, 2015c).
The model and EPA's input assumptions undergo periodic formal peer review. The
rulemaking process also provides opportunity for expert review and comment by a variety of
stakeholders, including owners and operators of capacity in the electricity sector that is
represented by the model, public interest groups, and other developers of U.S. electricity sector
models. The feedback that the Agency receives provides a highly detailed review of key input
7 See Chapter 7 of the IPM v.6 documentation. The documentation for EPA's Base Case v.6 using IPM consists of a
comprehensive document for the November 2018 release of IPM v. 6, and incremental update documents for
subsequent releases: http://www.epa.gov/airmarkets/powersectormodeling.html.
8 See Chapter 10 of EPA's Base Case using IPM (v6) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
4-6
-------
assumptions, model representation, and modeling results. IPM has received extensive review by
energy and environmental modeling experts in a variety of contexts. For example, in October
2014 U.S. EPA commissioned a peer review9 of EPA Base Case version 5.13 using the
Integrated Planning Model. Additionally, and in the late 1990s, the Science Advisory Board
reviewed IPM as part of the CAA Amendments Section 812 prospective studies10 that are
periodically conducted. The Agency has also used the model in a number of comparative
modeling exercises sponsored by Stanford University's Energy Modeling Forum over the past 15
years. IPM has also been employed by states (e.g., for the Regional Greenhouse Gas Initiative,
the Western Regional Air Partnership, Ozone Transport Assessment Group), other Federal and
state agencies, environmental groups, and industry.
4.3 EPA's Power Sector Modeling of the Base Case and Three Regulatory Control
Alternatives
The IPM "base case" for any regulatory impact analysis is a business-as-usual scenario that
represents expected behavior in the electricity sector under market and regulatory conditions in
the absence of the proposed rule. As such, an IPM base case represents an element of the
baseline for this RIA.11 EPA frequently updates the IPM base case to reflect the latest available
electricity demand forecasts from the U.S. Energy Information Administration (EIA) as well as
expected costs and availability of new and existing generating resources, fuels, emission control
technologies, and regulatory requirements.
4.3.1 EPA 's IPM Base Case v. 6
For our analysis of the Revised CSAPR Update proposed rule, EPA used the January 2020
release of IPM version 6 to provide power sector emissions data for air quality modeling, as well
as a companion updated database of EGU units (the National Electricity Energy Data System, or
NEEDS v.6 rev: 1-8-202012) that is used in EPA's modeling applications of IPM. The IPM Base
9 See Response and Peer Review Report EPA Base Case Version 5.13 Using IPM, available at:
https://www.epa.gov/airmarkets/response-and-peer-review-report-epa-base-case-version-513-using-ipm.
10 http://www2.epa.gov/clean-air-act-overview/benefits-and-costs-clean-air-act
11 As described in Chapter 5 of EPA's Guidelines for Preparing Economic Analyses, the baseline "should
incorporate assumptions about exogenous changes in the economy that may affect relevant benefits and costs (e.g.,
changes in demographics, economic activity, consumer preferences, and technology), industry compliance rates,
other regulations promulgated by EPA or other government entities, and behavioral responses to the proposed rule
by firms and the public." (USEPA, 2010).
12 https://www.epa.gov/airmarkets/national-electric-energy-data-system-needs-v6
4-7
-------
Case includes the Affordable Clean Energy (ACE) Rule consistent with the RIA for the final rule
and includes both the CSAPR rule and CSAPR Update rule. The Base Case includes the 2015
Effluent Limitation Guidelines (ELG) and the 2015 Coal Combustion Residuals (CCR), but does
not include the recently finalized 2020 ELG and CCR rules.13 The analysis of cost and impacts
presented in this chapter is based on a single IPM base case, and represents incremental impacts
projected solely as a result of compliance with the emissions budgets presented in Table 4-1
above.
4.3.2. Methodology for Evaluating the Regulatory Control Alternatives
To estimate the costs, benefits, and economic and energy market impacts of the Revised
CSAPR Update proposal, EPA conducted quantitative analysis of the three regulatory control
alternatives: the Revised CSAPR Update proposed emission budgets and more and less stringent
alternatives. Details about these regulatory control alternatives, including state-specific EGU
NOx ozone-season emissions budgets for each alternative as analyzed in this RIA, are provided
above in Section 4.1.
Before undertaking power sector analysis to evaluate compliance with the regulatory
control alternatives, EPA first considered available EGU NOx mitigation strategies that could be
implemented for the upcoming ozone season (i.e., the 2021 ozone season). EPA considered all
widely-used EGU NOx control strategies: optimizing NOx removal by existing, operational
selective catalytic reduction (SCRs) and turning on and optimizing existing idled SCRs; turning
on existing idled selective non-catalytic reduction (SNCRs); installation of (or upgrading to)
state-of-the-art NOx combustion controls; shifting generation to units with lower NOx emission
rates; and installing new SCRs and SNCRs. EPA determined that affected EGUs within the 12
states could implement all of these NOx mitigation strategies, except installation of new SCRs or
SNCRs and state of the art combustion controls for the 2021 ozone season. After assessing the
available NOx mitigation methods for complying with the annual budgets, this RIA projects that
the system-wide least-cost strategies for compliance with the proposed Revised CSAPR Update
and the more and less stringent regulatory alternatives lead to the application of the same
13 For a Ml list of modeled policy parameters, please see:
https://www.epa.gov/sites/production/files/2020-
02/documents/incremental_documentation_for_epa_v6 January _2020_reference_case.pdf
4-8
-------
controls at the same sources as in the analysis used to calculate the budgets for these alternatives.
As a consequence, the sectoral analyses used to establish the budgets are the same analyses used
to estimate the compliance cost, benefits, and impacts of the proposed Revised CSAPR Update
and the more and less stringent alternatives. In the analysis of the proposed rule presented in this
RIA, in each year of the analysis period (2021-2025) and in each of the 12 states subject to
tighter seasonal NOx budgets, seasonal NOx emissions from the sources subject to the proposed
rule equal the seasonal NOx budget. For more details on these assessments, including the
assessment of EGU NOx mitigation costs and feasibility, please refer to the EGU NOx
Mitigation Strategies Proposed Rule TSD, in the docket for this proposed rule.14
These mitigation strategies are primarily captured within the model. However, due to
limitations on model size, IPM v.6 does not have the ability to endogenously determine whether
or not to operate existing EGU post-combustion NOx controls (i.e., SCR or SNCR) in response
to a regulatory emissions requirement.15 The operating status of existing post-combustion NOx
controls at a particular EGU in a model scenario is determined by the model user. In order to
evaluate compliance with the regulatory alternatives, EPA determined outside of IPM whether or
not operation of existing controls that are idle in the baseline would be reasonably expected for
compliance with each of the evaluated regulatory alternatives and for which model years they
can feasibly be applied. IPM includes optimization and perfect foresight in solving for least cost
dispatch. Given that the final rule will likely become effective either immediately prior to or
slightly after the start of the 2021 ozone season, to avoid overstating optimization and dispatch
decisions that are not possible in the short time frame, EPA complemented the 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 analysis forms the basis for the climate benefits
calculations presented in this RIA.
EPA considers a unit to have optimized use of an SCR if emissions rates are equal to (or
below) the "widely achievable" rate of 0.08 lbs/MMBtu.16 Within IPM, units with extant SCRs
are defined as SCR-equipped units with ozone season NOx emission rates less than 0.20 lbs/
14 Docket ID No. EPA-HQ-OAR-2020-0272
15 EGUs with idled SCR or SNCR in the base case represent a small percentage (less than 10 percent) of the EGU
fleet that is equipped with NOx post-combustion controls.
16 For details on the derivation of this standard, please see preamble Section VII.B. 1.
4-9
-------
MMBtu in the Base Case. These units had their emission rates lowered to the lower of their
mode 417 NOx rate in NEEDS and the "widely achievable" optimized emissions rate of 0.08 lbs/
MMBtu in the Revised CSAPR Update proposal. Units equipped with SCRs with an emissions
rate exceeding 0.20 lbs/ MMBtu were considered to have idled SCRs. These units had their
emission rates lowered to the lower of their mode 4 NOx rate in NEEDS and the "widely
achievable" optimized emissions rate of 0.08 lbs/ MMBtu in the Revised CSAPR Update
proposal. These control options are achievable in 2021 and were associated with a uniform
control cost of $800 per ton and $1,600 per ton respectively. No further adjustments were made
to the variable and fixed operating cost of these units, and their heat rates were also not adjusted
to reflect energy requirements from increasing SCR removal efficiency within IPM. Under the
proposed rule, 60 units are projected to fully run existing SCR controls, while 4 units are
projected to turn on idled SCR controls.
Finally, unit combustion control configurations listed in NEEDS were compared against
Table 3-11 in the Documentation for EPA Base Case v.5.13 Using the Integrated Planning
Model IPM v.6, which lists state-of-the-art combustion control configurations based on unit
firing type. This allowed EPA to identify units that would receive state-of-the-art combustion
control upgrades in IPM. EPA then followed the procedure in the EGU NOx Mitigation
Strategies Proposed Rule TSD to calculate each of these unit's new NOx emission rate. These
upgrades were assumed to occur in 2022 and were assigned a uniform control cost of $1,600 per
ton. No further adjustments were made to the variable and fixed operating cost of these units, and
their heat rates were also not adjusted to reflect increased energy input requirements at a given
load from the use of additional combustion controls, within IPM. Under the proposed rule, 27
units are projected to install state-of-the-art combustion controls.
The EGU NOx mitigation strategies that are assumed to operate or are available to reduce
NOx in order to comply with each of the regulatory control alternatives are shown in Table 4-2;
17 NEEDS includes four possible states of NOx control operations, designated Modes 1-4. For details, please see
Chapter 3.9.3 of IPM v6 documentation available at:
https://www.epa.gOv/sites/production/files/2018-08/documents/epa_platform_v6_documentation_-
_all_chapters_august_23_2018_updated_table_6-2.pdf.
4-10
-------
more information about the estimated costs of these controls can be found in the EGU NOx
Mitigation Strategies Proposed Rule TSD.
Table 4-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives
Regulatory Control
Alternative
NOx Controls Implemented
Less Stringent Alternative (1) Shift generation to minimize costs (costs estimated within IPM)
(All controls above)
(2) Fully operating existing SCRs to achieve 0.08 lb/MMBtu NOx emission rate
(costs estimated outside IPM)
(3) Turn on idled SCRs (costs estimated outside IPM) and fully operate akin to
(2)
(4) Install state of the art combustion controls.
Revised CSAPR Update
Proposed Rule
More Stringent Alternative
(All controls above)
(5) In 2025, turn on idled SNCRs (costs estimated outside IPM)
(6) In 2025, install new SCRs (costs estimated outside IPM)
For the NOx controls identified in Table 4-2, under the proposed rule and the more
stringent alternative, 60 units are projected to fully operate existing SCRs and 4 units are
projected to turn on idled SCRs. Under the less stringent alternative, no units are projected to
either fully operate existing SCRs or turn on idled SCRs. Under the proposed rule and the more
stringent alternative, 27 units are projected to install state-of-the-art combustion controls, and
under the less stringent alternative no units are projected to install state-of-the-art combustion
controls. The book-life of the controls is assumed to be 15 years. Under the proposed rule and
the less stringent alternative, no units are projected to install new SCRs, and under the more
stringent alternative, 48 units are projected to install new SCRs. The book-life of the new SCRs
is assumed to be 15 years.
In addition to the limitation on ozone season NOx emissions required by the EGU
emissions budgets for the 12 states, there are four important features of the allowance trading
program represented in the model that may influence the level and location of NOx emissions
from affected EGUs, including: the ability of affected EGUs to buy and sell NOx ozone season
allowances from one another for compliance purposes; the ability of affected EGUs to bank NOx
ozone season allowances for future use; the effect of limits on the total ozone season NOx
emissions from affected EGUs in each state required by the assurance provisions; and the
treatment of banked pre-2021 vintage NOx ozone season allowances issued under the CSAPR
4-11
-------
Update program now being revised under this proposal. Each of these features of the ozone
season allowance trading program is described below.
Affected EGUs are expected to choose the least-cost method of complying with the
requirements of the allowance trading program, and the distribution of ozone season NOx
emissions across affected EGUs is generally governed by this cost-minimizing behavior in the
analysis. The total ozone season NOx emissions from affected EGUs in this analysis are limited
to the amount allowed by the sum of the NOx budgets across the 12 states. Furthermore,
allowances may be banked for future use. The number of banked allowances is influenced by the
determination, outside the model, of whether (i) existing controls that are idle in the base case are
turned on and (ii) it is less costly to abate ozone season NOx emissions in a current ozone season
than to abate emissions in a later ozone season. Affected EGUs are expected to bank NOx ozone
season allowances in the 2021 ozone season for use in a later ozone season. The model starts
with an assumed bank level in 2021 and endogenously determines the bank in each subsequent
year. Based on observation, EPA believes that this is a reasonable compliance path for EGUs,
even though there may be other non-economic reasons, such as being prepared for future
variability in power sector operations, that can potentially influence this decision.
While there are no explicit limits on the exchange of allowances between affected EGUs
and on the banking of 2021 and future vintage NOx ozone season allowances, the assurance
provisions limit the amount of seasonal NOx emissions by affected EGUs in each of the 12
states. The assurance level limits affected EGU emissions over an ozone season to the state's
NOx ozone season emissions budget plus an increment equal to 21 percent of each state's
emissions budget. This increment is called the variability limit. See Section VIII.C.4 of the
preamble for a discussion of the purpose of the assurance provision and further detail about how
the variability limits and assurance levels are determined. If a state exceeds its assurance level in
a given year, sources within that state are assessed a total of 3-to-l allowance surrender on the
excess tons. Section VIII.C.4 of the preamble also explains how EPA then determines which
EGUs are subject to this surrender requirement. In the modeling, the assurance provisions are
represented by a limit on the total ozone season NOx emissions that may be emitted by affected
EGUs in each state, and thus the modeling does not permit affected EGUs to emit beyond the
assurance levels and thus incur penalties.
4-12
-------
As described in Section VIII.D.4 of the preamble, the rule allows pre-2021 vintage NOx
ozone season allowances (that had been issued under the CSAPR Update program now being
revised under this proposal) to be used for compliance with this proposed rule, following a one-
time conversion that reduces the overall quantity of banked allowances from that time period.
Based on EPA's expectation of the size of the NOx allowance bank after the one-time conversion
carried out pursuant to the terms of this proposed rule, the treatment of these banked allowances
is represented in the modeling as an additional 21,020 tons of NOx allowances, the equivalent of
one year of the variability limit associated with the emission budgets, that may be used by
affected EGUs during the 2021 ozone season or in later ozone seasons under the Revised
CSAPR Update rule. Under the more stringent and less stringent alternatives an additional
21,020 tons and 25,480 tons respectively may be used by affected EGUs during the 2021 ozone
season or in later ozone seasons.
4.3.3 Methodology for Estimating Compliance Costs
This section describes EPA's approach to quantify estimated compliance costs associated
with the three regulatory control alternatives. These compliance costs include estimates projected
directly by the model as well as calculations performed outside of the model that use IPM model
inputs and methods. The model projections capture the costs associated with shifting generation
to lower-NOx emitting EGUs. The costs of increasing the use and optimizing the performance of
existing and operating SCRs,18 and for installing or upgrading NOx combustion controls, were
estimated outside of the model. The costs for these two NOx mitigation strategies are calculated
based on engineering analytics emissions projections and use the same NOx control cost
equations used in IPM. Therefore, this estimate is consistent with modeled projections and
provides the best available quantification of the costs of these NOx mitigation strategies.
The following steps summarize EPA's methodology for estimating the component of
compliance costs that are calculated outside of the model for the Revised CSAPR Update
proposal alternative19:
18 This includes optimizing the performance of SCRs that were not operating.
19 For more information on the derivation of costs and useful life of combustion controls, please see EGU NOx
Mitigation Strategies Proposed Rule TSD.
4-13
-------
(1) In the model projections, identify all EGUs in the 12 states that can adopt the following
NOx mitigation strategies:
• Fully operating existing SCRs
• Installing state of the art combustion controls
(2) Estimate the total NOx reductions that are attributable to each of these strategies:20
• Fully operating existing SCRs (SCRs operating in base case): 9,154 tons
• Fully operating existing SCRs (SCRs not operating in base case): 5,870 tons
• Installing state-of-the-art combustion controls (not available in 2021): 0 tons
(3) Estimate the average cost associated with each of these strategies:21
• Fully operating existing SCRs (SCRs operating in base case): $800/ton
• Fully operating existing SCRs (SCRs not operating in base case): $l,600/ton
• Installing state-of-the-art combustion controls: $l,600/ton
(4) Multiply (2) by (3) to estimate the total cost associated with each of these strategies.
Table 4-3 summarizes the results of this methodology for the Revised CSAPR Update
proposal alternative in 2021.
Table 4-3. Summary of Methodology for Calculating Compliance Costs Estimated Outside
of IPM for Revised CSAPR Update Proposal, 2021 (2016$)
NOx Ozone Season
Emissions
Average Cost
Total Cost
NOx Mitigation Strategy
(Tons)
($/ton)
($MM)
Optimize existing SCRs
9,154
800
7
Operate existing SCRs
5,870
1,600
9
20 For more information on how NOx reductions were attributed to strategies, see the Ozone Transport Policy
Analysis TSD.
21 See NOx Mitigation Strategy TSD for derivation of cost-per-ton estimates for fully operating SCRs and upgrading
to state-of-the-art combustion controls.
4-14
-------
EPA exogenously updated the emissions rates for the identified EGUs within the 12 states
consistent with the set of controls determined for 2021-2025 within IPM. The model was
updated to incorporate the emissions budgets identified for each case, and the first-year bank
adjustment as outlined in Section 4.3.2. The Group 2 regional trading program was updated to
exclude the 12-state Group 3 regional trading program, and budgets for the remaining Group 2
states were left otherwise unchanged. The change in the reported power system production cost
between this model run and the base run was used to capture the cost of generation shifting. The
total costs of compliance with the regulatory control alternatives are estimated as the sum of the
costs that are modeled within IPM and the costs that are calculated outside the model.
4.4 Estimated Impacts of the Regulatory Control Alternatives
4.4.1 Emission Reduction Assessment
As discussed in Chapter 1, EPA determined that NOx emissions in 12 eastern states affect
the ability of downwind states to attain and maintain the 2008 ozone NAAQS. For these 12
eastern states, EPA is issuing Federal Implementation Plans (FIPs) that generally update the
existing CSAPR Update NOx ozone-season emission budgets for EGUs and implement these
budgets via the CSAPR NOx ozone-season allowance trading program.
As indicated in Chapter 1, the NOx emissions reductions are presented in this RIA for two
time periods: 2021 and 2025. The 2021 emissions estimates are based on IPM projections for
2021, and adjustments to account for historical data. For more information on these and other
adjustments, see the Ozone Transport Policy Analysis TSD.
Table 4-4 presents the estimated reduction in power sector NOx emissions resulting from
compliance with the evaluated regulatory control alternatives (i.e., emissions budgets) in the 12
states, as well as the impact on other states. The emission reductions follow an expected pattern:
the less stringent alternative produces substantially smaller emissions reductions than EPA's
proposed emissions budgets, and the more stringent alternative results in slightly more NOx
emissions reductions.
4-15
-------
Table 4-4. EGU Ozone Season NOx Emissions and Emissions Changes (thousand tons) for
the Base Case and the Regulatory Control Alternatives
Total Emissions
Ozone Season NOx
(thousand tons)
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
12 States
124
106
122
106
-17
-2
-17
2021
Other States
236
236
236
236
0
0
0
Total
359
342
358
342
-17
-2
-17
12 States
123
100
121
100
-23
-2
-23
2022
Other States
235
235
235
235
0
0
0
Total
358
335
356
335
-23
-2
-23
12 States
117
96
116
96
-21
-2
-21
2023
Other States
227
227
227
227
0
0
0
Total
345
324
343
324
-21
-2
-21
12 States
114
93
113
79
-21
-2
-35
2024
Other States
225
225
225
225
0
0
0
Total
340
319
338
304
-21
-2
-35
12 States
114
93
113
79
-21
-2
-35
2025
Other States
225
225
225
225
0
0
0
Total
340
319
338
304
-21
-2
-35
Change from Base Case
The results of EPA's analysis show that, with respect to compliance with the EGU NOx
emission budgets in 2021, maximizing the use of existing operating SCRs provides the largest
amount of ozone season NOx emission reductions (52 percent, affecting 60 units), and turning on
idled SCRs produces an additional 34 percent (affecting 4 units) of the total ozone season NOx
reductions. Generation shifting primarily from coal to gas generation (14 percent) makes up the
remainder of the ozone season NOx reductions. Based on this analysis of how EGUs are
expected to comply with the proposed Revised CSAPR Update, none of the Group 3 states are
projected to hit their variability limits, nor bank significant allowances during the analysis period
(2021-2025).22
22 As shown in Table 4-4, in 2021 and 2025 seasonal NOx emissions from affected EGUs in the Group 3 states are
projected to emit at levels equal to the seasonal budget, and therefore (i) will not bank additional allowances, or (ii)
on net, use any banked allowances available at the end of the previous year or, in the case of 2021, from the starting
bank.
4-16
-------
In addition to the ozone season NOx reductions, there will also be reductions of other air
emissions associated with EGUs burning fossil fuels (i.e., co-pollutants). These other emissions
include the annual total changes in emissions of NOx and CO2; there are no annual SO2 and
PM2.5 emissions changes. The emissions reductions are presented in Table 4-5. Consistent with
the limited impact of generation shifting, there were de minimis emissions changes of CO,
mercury, and HC1.
Table 4-5. EGU Annual Emissions and Emissions Changes for NOx, SO2, PM2.5, and CO2
for the Regulatory Control Alternatives
Total Emissions
Annual NOx
(thousand tons)
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
12 States
291
274
290
274
-17
-2
-17
2021
Other States
524
524
524
524
0
0
0
Total
815
797
813
797
-17
-2
-17
12 States
289
259
287
259
-30
-2
-30
2022
Other States
521
521
521
521
0
0
0
Total
809
780
808
780
-30
-2
-30
12 States
275
249
274
249
-27
-2
-27
2023
Other States
505
505
505
505
0
0
0
Total
780
753
778
753
-27
-2
-27
12 States
268
241
266
227
-27
-2
-41
2024
Other States
500
500
500
500
0
0
0
Total
768
741
766
727
-27
-2
-41
12 States
268
241
266
227
-27
-2
-41
2025
Other States
500
500
500
500
0
0
0
Total
768
741
766
727
-27
-2
-41
Change from Base Case
Total Emissions
Change from Base Case
Annual SO2
(thousand tons)
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
12 States
376
376
376
376
0
0
0
2021
Other States
556
556
556
556
0
0
0
Total
933
933
933
933
0
0
0
12 States
332
332
332
332
0
0
0
2022
Other States
492
492
492
492
0
0
0
Total
824
824
824
824
0
0
0
4-17
-------
12 States
302
302
302
302
0
0
0
2023
Other States
480
480
480
480
0
0
0
Total
781
781
781
781
0
0
0
12 States
338
338
338
338
0
0
0
2024
Other States
534
534
534
534
0
0
0
Total
872
872
872
872
0
0
0
12 States
338
338
338
338
0
0
0
2025
Other States
534
534
534
534
0
0
0
Total
872
872
872
872
0
0
0
Total Emissions
Change from Base Case
Annual PM2.5
(thousand tons)
Revised
CSAPR
Less-
More-
Revised
CSAPR
Less-
More-
Base
Update
Stringent
Stringent
Update
Stringent
Stringent
Case
Proposal
Alternative
Alternative
Proposal
Alternative
Alternative
12 States
50
50
50
50
0
0
0
2021
Other States
76
76
76
76
0
0
0
Total
126
126
126
126
0
0
0
12 States
50
50
50
50
0
0
0
2022
Other States
76
76
76
76
0
0
0
Total
125
125
125
125
0
0
0
12 States
48
48
48
48
0
0
0
2023
Other States
74
74
74
74
0
0
0
Total
122
122
122
122
0
0
0
12 States
47
47
47
47
0
0
0
2024
Other States
74
74
74
74
0
0
0
Total
121
121
121
121
0
0
0
12 States
47
47
47
47
0
0
0
2025
Other States
74
74
74
74
0
0
0
Total
121
121
121
121
0
0
0
Total Emissions
Change from Base Case
Annual CO2
(thousand tons)
Revised
CSAPR
Less-
More-
Revised
CSAPR
Less-
More-
Base
Update
Stringent
Stringent
Update
Stringent
Stringent
Case
Proposal
Alternative
Alternative
Proposal
Alternative
Alternative
12 States
478
478
478
478
0
0
0
2021
Other States
959
959
959
959
0
0
0
Total
1,437
1,437
1,437
1,437
0
0
0
12 States
507
505
506
504
-2
-1
-3
2022
Other States
985
985
985
985
0
0
0
Total
1493
1490
1491
1489
-2
-1
-3
2023
12 States
537
532
534
530
-5
-3
-6
4-18
-------
Other States
1011
1012
1012
1011
0
0
0
Total
1548
1544
1545
1542
-4
-3
-6
12 States
532
527
528
523
-4
-3
-8
2024
Other States
1004
1004
1004
1004
0
0
0
Total
1536
1531
1532
1527
-4
-3
-8
12 States
526
522
523
516
-4
-3
-10
2025
Other States
996
996
996
997
0
0
0
Total
1,523
1,518
1,519
1,513
-5
-4
-10
4.4.2 Impact of Emissions Reductions on Maintenance and Nonattainment Monitors
In 2021, there are two nonattainment receptors and two maintenance receptors (see section
VI.C of the preamble for additional discussion). EPA evaluated the air quality improvements at
the four receptors from projected compliance with the two regulatory alternatives at the highest
EGU cost threshold levels (i.e., $1,600 per ton and $3,900 per ton). EPA found that the average
air quality improvement at the four receptors relative to the engineering analytics baseline was
0.19 ppb at $1,600 per ton and 0.23 ppb at $3,900 per ton (see Table VII.D.1-1 in the preamble
for additional discussion). EPA found that the one of the receptors (Westport, Connecticut
receptor) remains nonattainment at all cost levels, another receptor the (Stratford, Connecticut
receptor) switches from nonattainment to maintenance at $1,600 per ton (i.e., its average design
value (DV)23 falls below the standard but its maximum DV remains above the NAAQS), while a
third receptor (Houston receptor) remains maintenance at all levels.24
EPA observes this $1,600 per ton level of stringency results in all downwind air quality
problems for the 2008 ozone NAAQS being resolved after 2024 (one year earlier than the base
case). There are also projected changes in receptor status (from projected nonattainment to
maintenance-only) for the Stratford and Westport receptors (the first in 2021, the second in
2024). In addition, the Houston receptor changes from maintenance to attainment in 2023.
23 The DV is calculated as the 3-year average of the annual 4th highest daily maximum 8-hour ozone concentration in
parts per billion, with decimals truncated. The D V is a metric compared to the standard level to determine whether a
monitor is violating the NAAQS.
24 The fourth receptor was clean in the engineering base case, which is the starting point for a Step 3 analysis.
4-19
-------
4.4.3 Compliance Cost Assessment
The estimates of the changes in the cost of supplying electricity for the regulatory control
alternatives are presented in Table 4-6. Since the rule does not result in any additional
recordkeeping, monitoring or reporting requirements, the costs associated with compliance,
monitoring, recordkeeping, and reporting requirements are not included within the estimates in
this table and can be found in preamble Section VIII.C.6.
Table 4-6. National Compliance Cost Estimates (millions of 2016$) for the Regulatory
Control Alternatives
Revised CSAPR
Update
Proposal
More-Stringent
Alternative
Less-Stringent
Alternative
2021-2025 (Annualized)
19.4
80.6
1.6
2021 (Annual)
20.9
37.2
3.8
2022 (Annual)
29.7
49.2
12.8
2023 (Annual)
27.8
47.3
12.8
2024 (Annual)
6.3
132.2
-12.0
2025 (Annual)
6.3
132.2
-12.0
"2021-2025 (Annualized)" reflects total estimated annual compliance costs levelized over the period 2021 through
2025, discounted using a 4.25 real discount rate.25 This does not include compliance costs beyond 2025. "2021
(Annual)" through "2025 (Annual)" costs reflect annual estimates in each of those years.
There are several notable aspects of the results presented in Table 4-6. The most notable
result in Table 4-6 is that the estimated annual compliance costs for the less stringent alternative
is negative (i.e., a cost reduction) in 2024 and 2025, although this regulatory control alternative
reduces NOx emissions by over 2,000 tons as shown in Table 4-5. While seemingly
counterintuitive, estimating negative compliance costs in a single year is possible given the
assumption of perfect foresight. IPM's objective function is to minimize the discounted net
present value (NPV) of a stream of annual total cost of generation over a multi-decadal time
period.26 For example, with the assumption of perfect foresight it is possible that on a national
basis within the model the least-cost compliance strategy may be to delay a new investment or
25 This table reports compliance costs consistent with expected electricity sector economic conditions. An NPV of
costs was calculated using a 4.25% real discount rate consistent with the rate used in IPM's objective function for
cost-minimization. The NPV of costs was then used to calculate the levelized annual value over a 5-year period
(2021-2025) using the 4.25% rate as well. Tables ES-9 and 7-3 report the NPV of the annual stream of costs from
2021-2025 using 3% and 7% consistent with OMB guidance.
26 For more information, please see Chapter 2 of the IPM documentation.
4-20
-------
retirement that was projected to occur sooner in the base case. Such a delay could result in a
lowering of annual cost in an early time period and increase it in later time periods. Since the
less-stringent alternative is designed to include only generation shifting, it does not necessitate
full operation of existing controls, nor installation of new controls, leading to a negative total
cost point estimate in 2025, reflecting the decision to delay retirements until later in the forecast
period. Under the Revised CSAPR Update proposed rule, fully operating existing SCR controls
provide a large share of the total emissions reductions. These options are selected in 2021, while
upgrading to state-of-the-art combustion controls is assumed to begin in 2022. Generation
shifting costs are positive in 2021 and 2023, but negative in 2025. The result is that the costs in
2021-23 are higher than costs in 2025.
Under the more stringent alternative, while 2021 includes the same set of controls as under
the Revised CSAPR Update proposed rule, a wider range of technologies is considered in
subsequent years. This, combined with a more stringent cap driving generation shifting costs
positive in every year, results in costs that grow over the 2021-25 period.
As part of the IPM model runs, the Group 2 regional trading program was updated to
exclude the 12-state Group 3 regional trading program, and budgets for the remaining Group 2
states were left otherwise unchanged. The Group 2 states did not exhibit significant changes in
projected allowance prices and level and location of Group 2 NOx emissions between the
baseline and regulatory alternatives as a result of this update.
In addition to evaluating annual compliance cost impacts, EPA believes that a full
understanding of these three regulatory control alternatives benefits from an evaluation of
annualized costs over the 2021-2025 timeframe. Starting with the estimated annual cost time
series, it is possible to estimate the net present value of that stream, and then estimate a levelized
annual cost associated with compliance with each regulatory control alternative.27 For this
analysis we first calculated the NPV of the stream of costs from 2021 through 202528 using a
4.25 percent discount rate. EPA typically uses a 3 and a 7 percent discount rate to discount future
27 The XNPV() function in Microsoft Excel 2013 was used to calculate the NPV of the variable stream of costs, and
the PMT() function in Microsoft Excel 2013 is used to calculate the level annualized cost from the estimated NPV.
28 Consistent with the relationship between IPM run years and calendar years, EPA assigned 2023 compliance cost
estimates to both 2022 and 2023 in the calculation of NPV, and 2025 compliance cost to 2024 and 2025. For more
information, see Chapter 7 of the IPM Documentation.
4-21
-------
year social benefits and social costs in regulatory impact analyses (USEPA, 2010). In this cost
annualization we use a 4.25 percent discount rate, which is consistent with the rate used in IPM's
objective function for minimizing the NPV of the stream of total costs of electricity generation.
This discount rate is meant to capture the observed equilibrium market rate at which investors
are willing to sacrifice present consumption for future consumption and is based on a Weighted
Average Cost of Capital (WACC).29 After calculating the NPV of the cost streams, the same 4.25
percent discount rate and 2021-2025 time period are used to calculate the levelized annual (i.e.,
annualized) cost estimates shown in Table 4-6.30
Additionally, note that the 2021-2025 equivalent annualized compliance cost estimates
have the expected relationship to each other; the annualized costs are lowest for the less stringent
alternative, and highest for the more stringent alternative.
4.4.4 Impacts on Fuel Use, Prices and Generation Mix
While the Revised CSAPR Update proposal is expected to result in significant NOx
emissions reductions, it is estimated to result in relatively modest impacts to the power sector.
While these impacts are relatively small in percentage terms, consideration of these potential
impacts is an important component of assessing the relative impact of the regulatory control
alternatives. In this section we discuss the estimated changes in fuel use, fuel prices, generation
by fuel type, capacity by fuel type, and retail electricity prices.
Table 4-7 and Table 4-8 present the percentage changes in national coal and natural gas
usage by EGUs in 2021. These fuel use estimates reflect a modest shift to natural gas from coal.
The projected impacts in 2025 are similarly small.
29 The IPM Base Case documentation (Section 10.4.1 Introduction to Discount Rate Calculations) states "The real
discount rate for all expenditures (capital, fuel, variable operations and maintenance, and fixed operations and
maintenance costs) in the EPA Platform v6 is 4.25%."
30 The PMT() function in Microsoft Excel 2013 is used to calculate the level annualized cost from the estimated
NPV.
4-22
-------
Table 4-7. 2021 Projected U.S. Power Sector Coal Use for the Base Case and the
Regulatory Control Alternatives
Million Tons
Percent Change from Base Case
Revised
Revised
CSAPR
Less-
CSAPR
Less-
More-
Base
Update
Stringent
More-
Update
Stringent
Stringent
Case
Proposal
Alt.
Stringent Alt.
Proposal
Alt.
Alt.
Appalachia
85
85
85
85
0.16%
0.11%
0.36%
Interior
115
115
115
115
0.00%
0.01%
0.05%
Waste Coal
0
0
0
0
0.00%
0.00%
0.00%
West
287
286
286
286
-0.08%
-0.06%
-0.20%
Total
487
487
487
487
-0.02%
-0.01%
-0.04%
Table 4-8. 2021 Projected U.S. Power Sector Natural Gas Use for the Base Case and the
Regulatory Control Alternatives
Trillion Cubic Feet
Percent Change from Base Case
Revised
Revised
CSAPR
More-
CSAPR
More-
Update Less-Stringent
Stringent
Update Less-Stringent
Stringent
Base Case Proposal Alternative
Alternative
Proposal Alternative
Alternative
11 11 11
11
0.00% 0.00%
0.00%
Table 4-9 and Table 4-10 present the projected coal and natural gas prices in 2021, as well
as the percent change from the base case projected as a result of the regulatory control
alternatives. These minor impacts in 2021 are consistent with the small changes in fuel use
summarized above. The projected impacts in 2025 are similarly very small.
Table 4-9. 2021 Projected Minemouth and Power Sector Delivered Coal Price for the Base
Case and the Regulatory Control Alternatives
$/MMBtu
Percent Change from Base Case
Revised
Revised
CSAPR
Less-
More-
CSAPR
Less-
More-
Base
Case
Update
Proposal
Stringent
Alternative
Stringent
Alternative
Update
Proposal
Stringent
Alternative
Stringent
Alternative
Minemouth
1.21
1.21
1.21
1.21
0.08%
0.06%
0.22%
Delivered
1.87
1.87
1.87
1.87
0.03%
0.03%
0.09%
4-23
-------
Table 4-10. 2021 Projected Henry Hub and Power Sector Delivered Natural Gas Price for
the Base Case and the Regulatory Control Alternatives
$/MMBtu
Percent Change from Base Case
Revised
Revised
CSAPR
Less-
More-
CSAPR
Less-
More-
Base
Case
Update
Proposal
Stringent
Alternative
Stringent
Alternative
Update
Proposal
Stringent
Alternative
Stringent
Alternative
Henry Hub
3.19
3.19
3.19
3.19
0.00%
0.00%
0.02%
Delivered
3.24
3.24
3.24
3.24
0.00%
0.00%
0.02%
Table 4-11 presents the projected percentage changes in the amount of electricity
generation in 2021 by fuel type. Consistent with the fuel use projections and emissions trends
above, EPA projects a small overall shift from coal to gas. The projected impact in 2025 is
similarly small.
Table 4-11. 2021 Projected U.S. Generation by Fuel Type for the Base Case and the
Regulatory Control Alternatives
Generation (TWh)
Percent
Change from Base Case
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Coal
797
797
797
797
-0.003%
-0.001%
0.001%
Natural Gas
1,582
1,582
1,582
1,582
0.001%
0.003%
0.002%
Nuclear
740
740
740
740
0.000%
0.000%
0.000%
Hydro
304
304
304
304
0.005%
-0.001%
-0.001%
Non-Hydro RE
536
536
536
536
0.000%
0.000%
0.000%
Oil\Gas Steam
58
58
58
58
-0.031%
-0.072%
-0.103%
Other
34
34
34
34
-0.043%
-0.042%
-0.020%
Total
4,051
4,051
4,051
4,051
-0.001%
0.000%
-0.001%
Note: In this table, "Non-Hydro RE" includes biomass, geothermal, landfill gas, solar, and wind.
Table 4-12 presents the projected percentage changes in the amount of generating capacity
in 2021 by primary fuel type. As explained above, none of the regulatory control alternatives are
expected to have a net impact on overall capacity by primary fuel type in 2021, and the model
was specified accordingly.
4-24
-------
Table 4-12. 2021 Projected U.S. Capacity by Fuel Type for the Base Case and the
Regulatory Control Alternatives
Capacity (GW)
Percent Change from Base Case
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Coal
216
216
216
216
0.0%
0.0%
0.0%
Natural Gas
421
421
421
421
0.0%
0.0%
0.0%
Nuclear
94
94
94
94
0.0%
0.0%
0.0%
Hydro
107
107
107
107
0.0%
0.0%
0.0%
Non-Hydro RE
184
184
184
184
0.0%
0.0%
0.0%
Oil\Gas Steam
74
74
74
74
0.0%
0.0%
0.0%
Other
8
8
8
8
0.0%
0.0%
0.0%
Total
1106
1106
1106
1106
0.0%
0.0%
0.0%
Note: In this table, "Non-Hydro RE" includes biomass, geothermal, landfill gas, solar, and wind
EPA estimated the change in the retail price of electricity (2016$) using the Retail Price
Model (RPM).31 The RPM was developed by ICF for EPA, and uses the IPM estimates of
changes in the cost of generating electricity to estimate the changes in average retail electricity
prices. The prices are average prices over consumer classes (i.e., consumer, commercial, and
industrial) and regions, weighted by the amount of electricity used by each class and in each
region. The RPM combines the IPM annual cost estimates in each of the 64 IPM regions with
EIA electricity market data for each of the 22 electricity supply regions (shown in Figure 4-1) in
the electricity market module of the National Energy Modeling System (NEMS).32
Table 4-13 and Table 4-14 present the projected percentage changes in the retail price of
electricity for the three regulatory control alternatives in 2021 and 2025, respectively. Consistent
with other projected impacts presented above, average retail electricity prices at both the national
and regional level are projected to be small. By 2025, EPA estimates that this rule will result in a
0.02 percent increase in national average retail electricity price, or by about 0.02 mills/kWh.
31 See documentation available at: https://www.epa.gov/airmarkets/retail-price-model
32 See documentation available at:
http://www.eia.gov/forecasts/aeo/nems/documentation/electricity /pdf/m068(2014).pdf
4-25
-------
Table 4-13. Average Retail Electricity Price by Region for the Base Case and the
Regulatory Control Alternatives, 2021
2021 Average Retail Electricity Price
(2016 mills/kWh)
Region
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
MROE
118
118
118
118
0%
0%
0%
NYCW
166
166
166
166
0%
0%
0%
NYLI
134
134
134
134
0%
0%
0%
NYUP
109
109
109
109
0%
0%
0%
RFCE
115
115
115
115
0%
0%
0%
RFCM
91
91
91
91
0%
0%
0%
RFCW
93
93
93
93
0%
0%
0%
SRDA
83
83
83
83
0%
0%
0%
SRGW
87
87
87
87
0%
0%
0%
SRCE
85
85
85
85
0%
0%
0%
SRVC
100
100
100
100
0%
0%
0%
SPSO
88
88
88
88
0%
0%
0%
NATIONAL
100
100
100
100
0%
0%
0%
Percent Change from Base Case
4-26
-------
Table 4-14. Average Retail Electricity Price by Region for the Base Case and the
Regulatory Control Alternatives, 2025
2025 Average Retail Electricity Price
(2016 mills/kWh)
Region
Base
Case
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
Revised
CSAPR
Update
Proposal
Less-
Stringent
Alternative
More-
Stringent
Alternative
MROE
116
116
116
116
0%
0%
0%
NYCW
198
198
198
198
0%
0%
0%
NYLI
159
159
159
159
0%
0%
0%
NYUP
135
135
134
134
0%
0%
0%
RFCE
132
132
132
132
0%
0%
0%
RFCM
105
105
105
105
0%
0%
1%
RFCW
104
104
104
104
0%
0%
0%
SRDA
83
83
83
83
0%
0%
0%
SRGW
98
97
97
97
0%
0%
0%
SRCE
83
83
83
83
0%
0%
0%
SRVC
101
101
101
101
0%
0%
0%
SPSO
93
93
93
93
0%
0%
0%
NATIONAL
107
107
107
107
0%
0%
0%
Percent Change from Base Case
4-27
-------
MEWf
NYUP
NYCW
RfCW
SfiGW
SRDA
Figure 4-1. Electricity Market Module Regions
Source: EIA (lit tp://www.eta.gov/forecasts/aeo/pdf/nercmap.pdf)
4.5 Social Costs
As discussed in EPA's Guidelines for Preparing Economic Analyses, social costs are the
total economic burden of a regulatory action (USEPA, 2010). This burden is the sum of all
opportunity costs incurred due to the regulatory action, where an opportunity cost is the value
lost to society of any goods and services that will not be produced and consumed as a result of
reallocating some resources towards pollution mitigation. Estimates of social costs may be
compared to the social benefits expected as a result of a regulation to assess its net impact on
society. The social costs of a regulatory action will not necessarily be equal to the expenditures
by the electricity sector to comply with the rule. Nonetheless, here we use compliance costs as a
proxy for social costs.
The compliance cost estimates for the proposed and more or less stringent regulatory
control alternatives presented in this chapter are the change in expenditures by the electricity
generating sector required by the power sector for compliance under each alternative. The
4-28
-------
change in the expenditures required by the power sector to maintain compliance reflect the
changes in electricity production costs resulting from application of NOx control strategies,
including changes in expenditures resulting from changes in the mix of fuels used for generation,
necessary to comply with the emissions budgets. Ultimately, part of the compliance costs may be
borne by electricity consumers through higher electricity prices. As discussed above, the
electricity and fossil fuel price impacts from this proposed rule are expected to be small.
4.6 Limitations
EPA's modeling is based on expert judgment of various input assumptions for variables
whose outcomes are uncertain. As a general matter, the Agency reviews the best available
information from engineering studies of air pollution controls and new capacity construction
costs to support a reasonable modeling framework for analyzing the cost, emission changes, and
other impacts of regulatory actions.
The IPM-projected annualized cost estimates of private compliance costs provided in this
analysis are meant to show the increase in production (generating) costs to the power sector in
response to the proposed rule. 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. The CRF is derived
from estimates of the cost of capital (private discount rate), the amount of insurance coverage
required, local property taxes, and the life of capital. The private compliance costs presented
earlier are EPA's best estimate of the direct private compliance costs of the proposed rule.
As discussed in section 4.3.2, IPM v6 does not have the capacity to endogenously
determine whether or not to maximize the use of existing EGU post-combustion NOx controls
(i.e., SCR), or install/upgrade combustion controls in response to a regulatory control
requirement. These decisions were imposed exogenously on the model, as documented in section
4.3.2 and Ozone Transport Policy Analysis TSD. While the emissions projections reflect
operation of these controls, the projected compliance costs were supplemented with exogenously
estimated costs of maximizing SCR operation and installing/upgrading combustion controls (see
section 4.3.3). As a result of this modeling approach, the dispatch decisions made within the
model do not take into consideration the additional operating costs associated with these two
4-29
-------
types compliance strategies (the operating costs of the units on which these strategies are
imposed do not reflect the additional costs of these strategies). The effect of changes in facility
and system-wide emissions from these changes in operating costs are also not accounted for in
the spatial fields for the regulatory alternatives described in Chapter 3. These additional costs
and their influence on projected changes in emissions and the level and location of ozone and
PM2.5 concentration patterns from the regulatory alternatives are relatively minor, and do not
have a significant impact on the overall finding that the economic impacts of this proposed rule
are minimal.
Additionally, the modeling includes two emission reduction strategies that are exogenously
imposed where applicable: turning on idled SCRs (Revised CSAPR Update proposal and more-
stringent alternative) and turning on idled SNCRs (mores stringent alternative only). While these
strategies are exogenously imposed, the operation of controls is imposed in IPM and the costs
and emissions reductions are estimated outside of IPM. Since the costs of these strategies are
accounted for within the model, they are able to influence the projected behavior of the EGUs
within the model.
The annualized cost of the final rule, as quantified here, is EPA's best assessment of the
cost of implementing the rule. These costs are generated from rigorous economic modeling of
changes in the power sector due to implementation of the Revised CSAPR Update proposal.
4.7 References
U.S. Energy Information Administration (EIA). 2014. The Electricity Market Module of the
National Energy Modeling System: Model Documentation 2014. Available at:
.
Accessed 9/17/2015.
U.S. EPA, 2015. Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric
Utility Generating Units (FinalRule), http://www2.epa.gov/cleanpowerplan/clean-
power-plan-existing-power-plants.
U.S. EPA, 2015 a. Standards of Performance for Greenhouse Gas Emissions from New,
Modified, and Reconstructed Stationary Sources: Electric Utility Generating Units (Final
Rule), http://www2.epa.gov/cleanpowerplan/carbon-pollution-standards-new-modified-
and-reconstructed-power-plants.
4-30
-------
U.S. EPA, 2015b. Disposal of Coal Combustion Residuals from Electric Utilities (Final Rule),
http://www2.epa.gov/coalash/coal-ash-rule.
U.S. EPA, 2015c. Steam Electric Power Generating Effluent Guidelines (Final Rule),
http://www2.epa.gov/eg/steam-electric-power-generating-effluent-guidelines-2015-final-
rule.
U.S. EPA, 2014. Final Rule for Existing Power Plants and Factories,
http://www2.epa.gov/cooling-water-intakes.
U.S. EPA, 2011. Cross-State Air Pollution Rule,
http://www3.epa.gov/airtransport/CSAPR/index.html
U.S. EPA, 201 la. Mercury and Air Toxics Standards (MATS), http://www3.epa.gov/mats/.
U.S. EPA. 2010. EPA Guidelines for Preparing Economic Analyses. Available at:
. Accessed 9/21/2015.
U.S. EPA. 2010a. Regulatory Impact Analysis for the Proposed Federal Transport Rule
Analyses. Available at: < http://www3.epa.gov/ttnecasl/ria.html>. Accessed 9/21/2015.
U.S. EPA, 2005. Clean Air Interstate Rule,
http://archive.epa.gov/airmarkets/programs/cair/web/html/index.html.
4-31
-------
CHAPTER 5: BENEFITS
Overview
This action proposes to revise the Cross-State Air Pollution Rule (CSAPR) Update to
reduce the emissions of nitrogen oxides (NOx) transported from states that contribute
significantly to nonattainment or interfere with maintenance of the 2008 ozone National Ambient
Air Quality Standard (NAAQS) in downwind states. Implementing the Revised CSAPR Update
proposed rule is expected to reduce emissions of NOx and provide ozone reductions, as well as
consequent reductions in fine particulate matter (PM2.5) concentrations and carbon dioxide (CO2)
emissions. This chapter describes the methods used to estimate the domestic climate benefits
from reductions of CO2 emissions. Data, resource, and methodological limitations prevent EPA
from monetizing health benefits from reducing concentrations of ozone and PM2.5, as well as the
benefits of reducing direct exposure to NO2, ecosystem effects and visibility impairment as well
as benefits from reductions in other pollutants, such as hazardous air pollutants (HAP). We
qualitatively discuss these unquantified benefits in this chapter. However, to provide perspective
regarding the scope of the estimated benefits, Appendix 5B illustrates the potential health effects
associated with the change in PM2.5 and ozone concentrations as calculated using methods
developed prior to the 2019 PM ISA and 2020 Ozone ISA. The values of these estimated
benefits are not reflected in the estimated net benefits reported in Tables 7-1 and 7-2.
This chapter reports the estimated monetized domestic climate benefits associated with
emission reductions for the three regulatory control alternatives across several discount rates.
5.1 Estimated Climate Benefits from Reducing CO2
We estimate the climate for this proposed rulemaking using a measure of the domestic
social cost of carbon (SC-CO2). The SC-CO2 is a metric that estimates the monetary value of
projected impacts associated with marginal changes in CO2 emissions in a given year. The SC-
CO2 includes a wide range of anticipated climate impacts, such as net changes in agricultural
productivity and human health, property damage from increased flood risk, and changes in
energy system costs, including reduced costs for heating and increased costs for air conditioning.
The metric is typically used to assess the avoided damages as a result of regulatory actions (i.e.,
benefits of rulemakings that lead to an incremental reduction in cumulative global CO2
5-1
-------
emissions). The SC-CO2 estimates presented in this RIA focus on the projected impacts of
climate change that are anticipated to directly occur within U.S. borders.
The SC-CO2 estimates used in this analysis are interim values developed under Executive
Order (EO) 13783 for use in regulatory analyses until an improved estimate of the impacts of
climate change to the U.S. can be developed based on the best available science and economics.
EO 13783 directed agencies to ensure that estimates of the social cost of greenhouse gases used
in regulatory analyses "are based on the best available science and economics" and are consistent
with the guidance contained in OMB Circular A-4, "including with respect to the consideration
of domestic versus international impacts and the consideration of appropriate discount rates" (EO
13783, Section 5(c)). In addition, EO 13783 withdrew the technical support documents (TSDs)
used in the benefits analysis of the 2016 CSAPR Update rule (U.S. EPA, 2016b) for describing
the global social cost of greenhouse gas estimates developed under the prior Administration as
no longer representative of government policy.
Regarding the two analytical considerations highlighted in EO 13783 - how best to
consider domestic versus international impacts and appropriate discount rates - current guidance
in OMB Circular A-4 is as follows. Circular A-4 states that analysis of economically significant
proposed and final regulations "should focus on benefits and costs that accrue to citizens and
residents of the United States." (OMB, 2003)1 EPA follows this guidance by adopting a domestic
perspective in our central analysis. Regarding discount rates, Circular A-4 states that regulatory
analyses "should provide estimates of net benefits using both 3 percent and 7 percent." (OMB,
2003) The 7 percent rate is intended to represent the average before-tax rate of return to private
capital in the U.S. economy. The 3 percent rate is intended to reflect the rate at which society
discounts future consumption, which is particularly relevant if a regulation is expected to affect
private consumption directly. EPA follows this guidance below by presenting estimates based on
both 3 and 7 percent discount rates in the main analysis. See Appendix 5A for a discussion of the
modeling steps involved in estimating the domestic SC-CO2 estimates based on these discount
rates. These SC-CO2 estimates developed under EO 13783 and presented below will be used in
1 Office of Management and Budget (OMB), 2003, Circular A-4,
http://www.whitehouse.gov/omb/circulars_a004_a-4 and OMB, 2011. Regulatory Impact Analysis: A Primer.
http://www.whitehouse.gov/sites/default/files/omb/inforeg/regpol/circular-a-4_regulatory-impact-analysis-
aprimer.pdf
5-2
-------
regulatory analysis until more comprehensive domestic estimates can be developed, which would
take into consideration recent recommendations from the National Academies of Sciences et al.
(2017) to further update the current methodology to ensure that the SC-CO2 estimates reflect the
best available science.
Table 5-1 presents the average domestic SC-CO2 estimate across all of the integrated
assessment model runs used to estimate the SC-CO2 for each discount rate for the years 2015 to
2050.2 As with the global SC-CO2 estimates, the domestic SC-CO2 increases over time because
future emissions are expected to produce larger incremental damages as economies grow and
physical and economic systems become more stressed in response to greater climate change.
EPA estimated the dollar value of the CCh-related effects for each analysis year between
2021 and 2025 by applying the SC-CO2 estimates, shown in Table 5-1, to the estimated changes
in CO2 emissions in the corresponding year under the regulatory options. EPA then calculated
the present value and annualized benefits from the perspective of 2020 by discounting each year-
specific value to the year 2020 using the same 3 percent and 7 percent discount rates.
Table 5-1. Interim Domestic Social Cost of Carbon Values (2016$/Metric
Tonne CO2)
Year
3% Discount Rate, Average
7% Discount Rate, Average
2020
$7
$1
2025
$7
$1
2030
$8
$1
2035
$9
$2
2040
$9
$2
2045
$10
$2
2050
$11
$2
Note: These SC-CO2 values are stated in $/metric tonne CO2 and rounded to the nearest dollar (1
metric tonne equals 1.102 short tons). The estimates vary depending on the year of CO2 emissions
and are defined in real terms, i.e., adjusted for inflation using the GDP implicit price deflator. EPA
interpolated annual values for intermediate years.
Source: U.S. EPA Analysis, 2020 based on U.S. EPA, 2019b
The limitations and uncertainties associated with the SC-CO2 analysis, which were
discussed in the 2016 CSAPR Update RIA (U.S. EPA, 2016b), likewise apply to the domestic
2 The SC-CO2 estimates rely on an ensemble of three integrated assessment models (IAMs): Dynamic Integrated
Climate and Economy (DICE) 2010; Climate Framework for Uncertainty, Negotiation, and Distribution (FUND)
3.8; and Policy Analysis of the Greenhouse Gas Effect (PAGE) 2009.
5-3
-------
SC-CO2 estimates presented in this chapter. Some uncertainties are captured within the analysis,
as discussed in Appendix 5A, while other areas of uncertainty have not yet been quantified in a
way that can be modeled. For example, limitations include the incomplete way in which the
integrated assessment models capture catastrophic and non-catastrophic impacts, their
incomplete treatment of adaptation and technological change, the incomplete way in which inter-
regional and intersectoral linkages are modeled, uncertainty in the extrapolation of damages to
high temperatures, and inadequate representation of the relationship between the discount rate
and uncertainty in economic growth over long time horizons. The science incorporated into these
models understandably lags behind the most recent research, and the limited amount of research
linking climate impacts to economic damages makes this comprehensive global modeling
exercise even more difficult. These individual limitations and uncertainties do not all work in the
same direction in terms of their influence on the SC-CO2 estimates. In accordance with guidance
in OMB Circular A-4 on the treatment of uncertainty, Appendix 5 A provides a detailed
discussion of the ways in which the modeling underlying the development of the SC-CO2
estimates used in this RIA addressed quantified sources of uncertainty and presents a sensitivity
analysis to show consideration of the uncertainty surrounding discount rates over long time
horizons.
Recognizing the limitations and uncertainties associated with estimating the SC-CO2, the
research community has continued to explore opportunities to improve SC-CO2 estimates.
Notably, the National Academies of Sciences, Engineering, and Medicine conducted a
multidiscipline, multi-year assessment to examine potential approaches, along with their relative
merits and challenges, for a comprehensive update to the current methodology. The task was to
ensure that the SC-CO2 estimates that are used in Federal analyses reflect the best available
science, focusing on issues related to the choice of models and damage functions, climate science
modeling assumptions, socioeconomic and emissions scenarios, presentation of uncertainty, and
discounting. In January 2017, the Academies released their final report, "Assessing Approaches
to Updating the Social Cost of Carbon," and recommended specific criteria for future updates to
the SC-CO2 estimates, a modeling framework to satisfy the specified criteria, and both near-term
updates and longer-term research needs pertaining to various components of the estimation
process (National Academies of Sciences et al., 2017).
5-4
-------
The Academies' 2017 report also discussed the challenges in developing domestic SC-
CO2 estimates, noting that current integrated assessment models do not model all relevant
regional interactions - i.e., how climate change impacts in other regions of the world could affect
the United States, through pathways such as global migration, economic destabilization, and
political destabilization. The Academies concluded that it "is important to consider what
constitutes a domestic impact in the case of a global pollutant that could have international
implications that impact the United States. More thoroughly estimating a domestic SC-CO2
would therefore need to consider the potential implications of climate impacts on, and actions by,
other countries, which also have impacts on the United States." (National Academies of Sciences
et al., 2017, pg. 12-13). In addition to requiring reporting of impacts at a domestic level, Circular
A-4 states that when an agency "evaluate[s] a regulation that is likely to have effects beyond the
borders of the United States, these effects should be reported separately" (OMB, 2003; page 15).
This guidance is relevant to the valuation of damages from C02 and other greenhouse gases
(GHGs), given that GHGs contribute to damages around the world independent of the country in
which they are emitted. Therefore, in accordance with this guidance in OMB Circular A-4,
Appendix 5A presents the global climate benefits from this proposed rule using global SC-C02
estimates based on both 3 and 7 percent discount rates. EPA did not quantitatively project the
full impact of the proposed rule on international trade and the location of production, so it is not
possible to present analogous estimates of international costs resulting from the regulatory
options. However, to the extent that the electricity market analysis endogenously models
international electricity and natural gas trade (see Chapter 4), and to the extent that affected firms
have some foreign ownership, some of the costs accruing to entities outside U.S. borders is
captured in the technology implementation costs presented in the RIA (U.S. EPA, 2020c).
Table 5-2 shows the estimated monetary value of the estimated changes in CO2 emissions
in 2021 and 2025 for the Revised CSAPR Update, the more-stringent alternative, and the less-
stringent alternative.
Table 5-2. Estimated Domestic Climate Benefits from Changes in CO2 Emissions 2021 -
2025 (Millions of 2016$)
Regulatory Option Year 3% Discount Rate 7% Discount Rate
Proposal 2021 0.3 0.0
5-5
-------
Table 5-2. Estimated Domestic Climate Benefits from Changes in CO2 Emissions 2021 -
2025 (Millions of 2016$)
2022
14.9
2.3
2023
30.0
4.8
2024
31.4
5.1
2025
32.9
5.4
More-Stringent
Alternative
2021
0.8
0.1
2022
21.6
3.4
2023
43.1
6.8
2024
57.1
9.2
2025
71.5
11.7
Less-Stringent
2021
0.2
0.0
Alternative
2022
9.4
1.5
2023
18.9
3.0
2024
22.1
3.6
2025
25.5
4.2
Table 5-3 shows the total annualized monetary values associated with changes in CO2
emissions for the three regulatory options. EPA annualized monetary value estimates to enable
consistent reporting across benefit categories (e.g., benefits from reduction in NOx emissions).
The annualized values for the Revised CSAPR update rule are $22.1 million and $3.6 million,
using discount rates of 3 and 7 percent, respectively.
Table 5-3. Estimated Total Annualized Domestic Climate Benefits (2021-25) from
Changes in CO2 Emissions (Millions of 2016$)
Regulatory Option
3% Discount Rate
7% Discount Rate
Proposal
22.1
3.6
More-Stringent Alternative
38.9
6.3
Less-Stringent Alternative
15.3
2.5
5.2 Unquantified Benefits
The monetized benefits estimated in this RIA reflect a subset of benefits attributable to
the climate benefits from reductions associated with CO2. The proposal is also expected to
reduce emissions of ozone season NOx. In the presence of sunlight, NOx and volatile organic
compounds (VOCs) can undergo a chemical reaction in the atmosphere to form ozone. Reducing
5-6
-------
NOx emissions generally reduces human exposure to ozone and the incidence of ozone-related
health effects, though the degree to which ozone is reduced will depend in part on local levels of
VOCs as discussed in Chapter 3. The proposal would also reduce emissions of NOx throughout
the year. Because NOx is also a precursor to formation of ambient PM2.5, reducing these
emissions would reduce human exposure to ambient PM2.5 throughout the year and would thus
reduce the incidence of PIvfc.s-attributable health effects.3 Reducing emissions of NOx would
also reduce ambient exposure to NO2 and its associated health effects.
Data, time, and resource limitations prevented EPA from quantifying the estimated
impacts or monetizing estimated benefits associated with exposure to ozone, PM2.5, and NO2
(independent of the role NO2 plays as precursors to PM2.5), as well as ecosystem effects, and
visibility impairment due to the absence of air quality modeling data for these pollutants in this
analysis. Lack of quantification does not imply that there are no benefits associated with
reductions in exposures to ozone, PM2.5, or NO2. In this section, we provide a qualitative
description of these benefits, which are listed in Table 5-4. However, to provide perspective
regarding the scope of the estimated benefits, Appendix 5B illustrates the potential health effects
associated with the change in PM2.5 and ozone concentrations as calculated using methods
developed prior to the 2019 PM ISA and 2020 Ozone ISA.
3 This RIA does not quantify PM2 5-related benefits associated with SO2 emission reductions. As discussed in
Chapter 4, EPA does not estimate significant SO2 emission reductions as a result of this proposal. Additionally, this
RIA does not estimate changes in emissions of directly emitted particles.
5-7
-------
Table 5-4. Unquantified Health and Welfare Benefits Categories
Category
Effect
Effect
Quantified
Effect
Monetized
More
Information
Premature mortality
Adult premature mortality
—
—
PMISA1-2
from exposure to PM2.5
Infant mortality
—
—
PMISA1-2
Non-fatal heart attacks
—
—
PMISA1-2
Hospital admissions—respiratory
—
—
PMISA1-2
Hospital admissions—cardiovascular
—
—
PMISA1-2
Emergency room visits for asthma
—
—
PMISA1-2
Acute bronchitis
—
—
PMISA1-2
Lower respiratory symptoms
—
—
PMISA1-2
Upper respiratory symptoms
—
—
PMISA1-2
Exacerbated asthma
—
—
PMISA1-2
Morbidity from
exposure to PM2.5
Lost work days
Minor restricted-activity days
—
—
PMISA1-2
PMISA1-2
Chronic Bronchitis
—
—
PMISA2
Emergency room visits for cardiovascular effects
—
—
PMISA2
Strokes and cerebrovascular disease
—
—
PMISA2
Other cardiovascular effects
—
—
PMISA3
Other respiratory effects (e.g., pulmonary function,
non-asthma ER visits, non-bronchitis chronic
PMISA3
diseases, other ages and populations)
Reproductive and developmental effects (e.g., low
birth weight, pre-term births, etc.)
—
—
PMISA3-4
Cancer, mutagenicity, and genotoxicity effects
—
—
PMISA3-4
Premature mortality based on short-term study
Mortality from exposure estimates
—
—
Ozone ISA1-2
to ozone
Premature mortality based on long-term study
estimate
—
—
Ozone ISA1-2
Hospital admissions—respiratory causes
—
—
Ozone ISA1-2
Emergency department visits for asthma
—
—
Ozone ISA1-2
Exacerbated asthma
—
—
Ozone ISA1-2
Minor restricted-activity days
—
—
Ozone ISA1-2
Morbidity from
School absence davs
—
—
Ozone ISA1-2
exposure to ozone
Decreased outdoor worker productivity
—
—
Ozone ISA2
Other respiratory effects (e.g., premature aging of
lungs)
—
—
Ozone ISA3
Cardiovascular and nervous system effects
—
—
Ozone ISA3
Reproductive and developmental effects
—
—
Ozone ISA3-4
Improved Human Health
Asthma hospital admissions
—
—
NO2 ISA2
Chronic lung disease hospital admissions
—
—
NO2 ISA2
Respiratory emergency department visits
—
—
NO2 ISA2
Reduced incidence of
Asthma exacerbation
—
—
NO2 ISA2
morbidity from
exposure to NO2
Acute respiratory symptoms
—
—
NO2 ISA2
Premature mortality
—
—
NO2 ISA2-3-4
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
—
—
NO2 ISA3-4
Improved Environment
Visibility in Class 1 areas
—
—
PMISA2
5-8
-------
Reduced visibility
impairment
Visibility in residential areas
—
—
PM ISA2
Reduced effects on
Household soiling
—
—
PM ISA2,3
materials
Materials damage (e.g., corrosion, increased wear)
—
—
PM ISA3
Reduced effects from
PM deposition (metals
and organics)
Effects on Individual organisms and ecosystems
—
—
PM ISA3
Visible foliar injury on vegetation
—
—
Ozone ISA2
Reduced vegetation growth and reproduction
—
—
Ozone ISA2
Yield and quality of commercial forest products and
crops
—
—
Ozone ISA2
Reduced vegetation
Damage to urban ornamental plants
—
—
Ozone ISA3
and ecosystem effects
from exposure to
ozone
Carbon sequestration in terrestrial ecosystems
—
—
Ozone ISA2
Recreational demand associated with forest
aesthetics
—
—
Ozone ISA3
Other non-use effects
Ozone ISA3
Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity,
leaf-gas exchange, community composition)
—
—
Ozone ISA3
Recreational fishing
—
—
NOx SOxISA2
Tree mortality and decline
—
—
NOx SOxISA3
Reduced effects from
acid deposition
Commercial fishing and forestry effects
—
—
NOx SOxISA3
Recreational demand in terrestrial and aquatic
ecosystems
—
—
NOx SOxISA3
Other non-use effects
NOx SOxISA3
Ecosystem functions (e.g., biogeochemical cycles)
—
—
NOx SOxISA3
Species composition and biodiversity in terrestrial
and estuarine ecosystems
—
—
NOx SOxISA3
Coastal eutrophication
—
—
NOx SOxISA3
Reduced effects from
nutrient enrichment
Recreational demand in terrestrial and estuarine
ecosystems
—
—
NOx SOxISA3
Other non-use effects
NOx SOxISA3
Ecosystem functions (e.g., biogeochemical cycles,
fire regulation)
—
—
NOx SOxISA3
Reduced vegetation
Injury to vegetation from SO2 exposure
—
—
NOx SOxISA3
effects from ambient
exposure to SO2 and
NOx
Injury to vegetation from NOx exposure
—
—
NOx SOxISA3
5-9
-------
Improved Human Health
Asthma hospital admissions
—
—
NO2 ISA2
Chronic lung disease hospital admissions
—
—
NO2 ISA2
Respiratory emergency department visits
—
—
NO2 ISA2
Reduced incidence of
Asthma exacerbation
—
—
NO2 ISA2
morbidity from
exposure to NO2
Acute respiratory symptoms
—
—
NO2 ISA2
Premature mortality
—
—
NO2 ISA2'3-4
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
—
—
NO2 ISA3'4
Improved Environment
Reduced visibility
Visibility in Class 1 areas
—
—
PM ISA2
impairment
Visibility in residential areas
—
—
PM ISA2
Reduced effects on
Household soiling
—
—
PM ISA2'3
materials
Materials damage (e.g., corrosion, increased wear)
—
—
PM ISA3
Reduced effects from
PM deposition (metals
and organics)
Effects on Individual organisms and ecosystems
—
—
PM ISA3
Visible foliar injury on vegetation
—
—
Ozone ISA2
Reduced vegetation growth and reproduction
—
—
Ozone ISA2
Yield and quality of commercial forest products and
crops
—
—
Ozone ISA2
Reduced vegetation
Damage to urban ornamental plants
—
—
Ozone ISA3
and ecosystem effects
from exposure to
ozone
Carbon sequestration in terrestrial ecosystems
—
—
Ozone ISA2
Recreational demand associated with forest
aesthetics
—
—
Ozone ISA3
Other non-use effects
Ozone ISA3
Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity,
leaf-gas exchange, community composition)
—
—
Ozone ISA3
Recreational fishing
—
—
NOx SOxISA2
Tree mortality and decline
—
—
NOx SOxISA3
Reduced effects from
acid deposition
Commercial fishing and forestry effects
—
—
NOx SOxISA3
Recreational demand in terrestrial and aquatic
ecosystems
—
—
NOx SOxISA3
Other non-use effects
NOx SOxISA3
Ecosystem functions (e.g., biogeochemical cycles)
—
—
NOx SOxISA3
Species composition and biodiversity in terrestrial
and estuarine ecosystems
—
—
NOx SOxISA3
Coastal eutrophication
—
—
NOx SOxISA3
Reduced effects from
nutrient enrichment
Recreational demand in terrestrial and estuarine
ecosystems
—
—
NOx SOxISA3
Other non-use effects
NOx SOxISA3
Ecosystem functions (e.g., biogeochemical cycles,
fire regulation)
—
—
NOx SOxISA3
Reduced vegetation
Injury to vegetation from SO2 exposure
—
—
NOx SOxISA3
effects from ambient
exposure to SO2 and
NOx
Injury to vegetation from NOx exposure
—
—
NOx SOxISA3
1 These endpoints are generally quantified and monetized when EPA quantitatively characterizes the benefits of changes in PM2.5
and Ozone.
2 We assess these benefits qualitatively due to data and resource limitations for this RIA.
5-10
-------
3 We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
4 We assess these benefits qualitatively because current evidence is only suggestive of causality or there are other significant
concerns over the strength of the association.
The proposed Revised CSAPR Update is expected to reduce concentrations of both
ground-level ozone and fine particles (PM2.5) (see Chapter 3). EPA historically has used
conclusions of the most recent Integrated Science Assessment (ISA) to inform its approach for
quantifying air pollution-attributable health, welfare and environmental impacts associated with
that pollutant. There is a separate ISA for each of the criteria pollutants. The ISA synthesizes the
epidemiologic, controlled human exposure and experimental evidence ".. .useful in indicating the
kind and extent of identifiable effects on public health or welfare which may be expected from
the presence of [a] pollutant in ambient air."4
The ISA uses a weight of evidence approach to assess the extent to the evidence supports
conclusions about the likelihood that a given criteria pollutant causes a given health outcome.
EPA generally estimates the number and economic value of the effects for which the ISA
identifies the pollutant as having "causal" or "likely to be causal" relationship. The endpoints for
which the 2020 final Ozone ISA (U.S. EPA, 2020b) and the 2019 final PM ISA (U.S. EPA,
2019a) identified as being causal or likely causal differed in some cases the endpoints for which
those pollutants were identified as being causal or likely causal in the Ozone and PM ISAs
completed for the previous NAAQS reviews (Tables 5-5 and 5-6). EPA traditionally uses the
ISAs' characterizations of the health and ecological literature to identify individual studies that
may be of sufficient quality for use in supporting PM or ozone benefits analysis.
When updating its approach for quantifying the benefits of changes in PM2.5 and Ozone,
the Agency will incorporate evidence reported in these two recently completed ISAs and account
for forthcoming recommendations from the Science Advisory Board on this issue (U.S. EPA-
SAB 2020). When updating the evidence for a given endpoint, EPA will consider the extent to
which there is a causal relationship, whether suitable epidemiologic studies exist to allow
quantification of concentration response function, and whether there are robust economic
approaches for estimating the value of the impact of reducing human exposure to the pollutant.
Carefully and systematically reviewing the full breadth of this information requires significant
time and resources. This process is still underway and will not be completed in time for this
4 Section 108 of the Clean Air Act. 42 U.S.C. 7408
5-11
-------
proposal. For this reason, this RIA characterizes the potential benefits of reducing these two
pollutants in qualitative terms only.5 EPA intends to update its quantitative methods for
estimating the number and economic value of Ozone and PM2.5 health effects in time for
publication as part of the final Revised CSAPR Update RIA.6
EPA is reviewing this evidence and is following a five-step approach as it updates its
methods for quantifying and monetizing ozone and PM2.5 attributable health endpoints:
1. Identify Ozone- and PJvfe.s-attributable health effects for which the ISA reports the
strongest evidence. EPA will consider the ISA-reported evidence for each endpoint,
including the extent to which the ISA identifies that endpoint as either causally, or likely-
to-be-causally, related to each pollutant.
2. Identify health outcomes that may be quantified in a benefits assessment. We would
select among clinically significant outcomes (e.g. premature mortality and hospital
admissions) for which endpoint-specific baseline incidence data are available.
3. Choose concentration-response parameters characteristic of the literature reviewed in the
ISA. We would weigh criteria including study design, location, population
characteristics, and other attributes.7 In some cases we will need to identify and select
new rates of baseline disease to quantify these effects.
5 The RIA for the Effluent Limit Guidelines rule separately noted that".. .the 2020 Integrated Science Assessment
for Ozone concludes the currently available evidence for cardiovascular effects and total mortality is suggestive of,
but not sufficient to infer, a causal relationship with short-term (as well as long-term) 03 exposures (ISA, sections
IS.4.3.4 and IS.4.3.5)....Until a replacement method that only estimates the benefits associated with respiratory
causes of premature mortality has been developed, EPA will be removing the estimate of the impact of reduced
ozone exposure on premature mortality from its benefits estimates from subsequent rulemakings." (U.S. EPA,
2020a) Rather than selectively updating the evidence for individual endpoints, the Agency is instead systematically
updating its approach for quantifying all ozone and PM2 5 effects using evidence reported in the Final Ozone ISA
(2020) and the Final PM ISA (2019).
6 In particular, the 2020 Ozone ISA concludes that the currently available evidence for cardiovascular effects and
total mortality is suggestive of, but not sufficient to infer, a causal relationship with short-term (as well as long-term)
ozone exposures (U.S. EPA, 2020b, sections IS.4.3.4 and IS.4.3.5). As such, EPA is in the process of recalibrating
its benefits estimates to quantify only premature mortality from respiratory causes (i.e., non-respiratory causes of
premature mortality associated with ozone exposure would no longer be estimated). Similarly, the 2019 PM ISA
concludes that the currently available evidence for nervous system effects and cancer is likely to be a causal
relationship with long term PM2 5 exposure. EPA is in the process of evaluating nervous system effects from long
term PM2 5 exposure and evaluating the relationship between long term PM2 5 exposure and cancer. Furthermore, the
ISA references a variety of additional studies for consideration in quantifying the health implications of changes in
PM2 5 and ozone exposure. EPA is updating the estimates for several other health endpoints to account for this new
scientific literature.
7 In some cases, the ISA will identify whether there are more recent epidemiologic studies that are better suited than
the prior studies used for endpoints whose causality did not change between the prior ISA and the current ISA (e.g.
respiratory hospital admissions). In these cases, we may substitute this new epidemiologic evidence.
5-12
-------
4. Choose economic unit values. For each health endpoint we would identify a
corresponding willingness-to-pay or cost-of-illness measure to express the economic
value of the adverse effect.
5. Develop methods for characterizing uncertainty associated with quantified benefits
estimates. Building on EPA's current methods for characterizing uncertainty, these
approaches will include, among others, reporting confidence intervals calculated from
concentration-response parameter estimates and separate quantification using multiple
studies and concentration response parameters for particularly influential endpoints (e.g.,
mortality risk), and potentially approaches for aggregating and representing the results of
multiple studies evaluating a particular health endpoint.8
At each of the four stages above, the Agency would report a Preferred Reporting Items
for System Reviews (PRISMA) diagram, detailing for each endpoint, study and concentration-
response (effect coefficients), which are included and excluded and the rationale for applying or
excluding this information.9
8 Study quality, inter-study heterogeneity, and redundancy issues will be taken into consideration if epidemiologic
risk estimates are aggregated.
9 Additional information regarding the PRISMA can be found here:
https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed. 1000097
5-13
-------
Table 5-5. Estimated Summary of Causality Determination for each Ozone-Related
Endpoint
(endpoints for which EPA's causality determination has changed reported in bold italic)
Conclusion from the:
2013 Ozone ISA 2020 Ozone ISA
Health outcome
Respiratory effects
Causal relationship
Causal relationship
a
o
Cardiovascular effects
Likely to be causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
Metabolic effects
Not determined
Likely to be causal relationship
1
£
o
-s
Total mortality
Likely to be causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
Central nervous system
effects
Suggestive of a causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
Respiratory effects
Likely to be causal
relationship
Likely to be causal relationship
Cardiovascular effects
Suggestive of a causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
Suggestive of, but not
Metabolic effects
Not determined
sufficient to infer, a causal
relationship
s
to
o
Total mortality
Suggestive of a causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
5r
Effects on fertility and
S
reproduction: suggestive of,
but not sufficient to infer, a
o
<1
Reproductive effects
Suggestive of a causal
relationship
causal relationship
Effects on pregnancy and birth
outcomes: suggestive of, but
not sufficient to infer, a causal
relationship
Central nervous system
effects
Suggestive of a causal
relationship
Suggestive of, but not
sufficient to infer, a causal
relationship
Cancer
Inadequate to infer a causal
relationship
Inadequate to infer the
presence or absence of a
causal relationship
5-14
-------
Table 5-6. Summary of Causality Determination for each PlVh.s-Related Endpoint
(endpoints for which EPA's causality determination has changed reported in bold italic)
Conclusion from the:
2009 PM ISA 2019 PM ISA
Health outcome
Respiratory effects
Likely to be causal
relationship
Likely to be causal relationship
a
&5
O
Cardiovascular effects
Causal
Causal
1
Metabolic effects
Not determined
Suggestive of, but inadequate to
infer
o
Nervous system effects
Inadequate to infer
Suggestive of, but inadequate to
infer
Mortality
Causal
Causal
Respiratory effects
Likely to be causal
relationship
Likely to be causal relationship
Cardiovascular effects
Causal
Causal
Metabolic effects
Not determined
Suggestive of, but inadequate to
infer
Nervous system effects
Not determined
Likely to be causal relationship
&5
o
Cl
Effects on fertility and
Effects on fertility and
reproduction: suggestive of,
reproduction: suggestive of, but
but not sufficient to infer, a
not sufficient to infer, a causal
o
Reproductive effects
causal relationship
Effects on pregnancy and
relationship
Effects on pregnancy and birth
<1
birth outcomes: suggestive
of, but not sufficient to infer,
a causal relationship
outcomes: suggestive of, but not
sufficient to infer, a causal
relationship
Cancer
Suggestive of, but not
sufficient to infer
Likely to be causal
Pre-mature Mortality
Causal
Causal
5.2.1 Ozone Health Benefits
Following a comprehensive review of health evidence, the Integrated Science Assessment
for Ozone and Related Photochemical Oxidants (2020 ozone ISA) (U.S. EPA, 2020b) also made
several determinations of causal or likely causal impacts for long- and short-term ozone
exposures. Regarding long-term exposures, the ozone ISA found that evidence supports a likely
to be causal relationship with respiratory effects; and is suggestive, but inadequate to infer, a
causal relationship with cardiovascular effects, metabolic effects, total mortality, reproductive
5-15
-------
effects, and central nervous system effects. The 2020 ozone ISA found that short-term exposure
evidence supports a causal relationship with respiratory effects; a likely to be causal relationship
with metabolic effects; and is suggestive, but inadequate to infer, a causal relationship with
cardiovascular effects, total mortality, reproductive effects, and central nervous system effects.
While metabolic effects were not included in the 2013 ozone ISA, the determination for short-
term cardiovascular effects was increased and the determination for short-term total mortality
was decreased in the 2020 ozone ISA relative to the 2013 ozone ISA.
5.2.2 PM2.5 Health Benefits
Following a similar comprehensive review of health evidence from epidemiologic and
laboratory studies, the Integrated Science Assessment for Particulate Matter (2019 PM ISA)
(U.S. EPA, 2019a) made several determinations for long- and short-term PM2.5 exposures.
Regarding long-term exposures, the PM ISA found that evidence supports a causal relationship
with cardiovascular effects and total mortality; a likely to be causal relationship with respiratory
effects, nervous system effects, and cancer; and is suggestive, but inadequate to infer, a causal
relationship with metabolic effects and reproductive effects. The 2019 PM ISA found that short-
term exposure evidence supports a causal relationship with cardiovascular effects and total
mortality; a likely to be causal relationship with respiratory effects; and is suggestive, but
inadequate to infer, a causal relationship with metabolic effects and nervous system effects.
Metabolic effects and long-term nervous system effects were not included in the 2009 PM ISA,
and the determinations of causal or likely causal impacts for all health outcomes evaluated in that
ISA, other than cancer, which was increased, remained the same.
5.2.3 NO2 Health Benefits
In addition to being a precursor to PM2.5 and ozone, NOx emissions are also linked to a
variety of adverse health effects associated with direct exposure. We were unable to estimate the
health benefits associated with reduced NO2 exposure in this analysis. Following a
comprehensive review of health evidence from epidemiologic and laboratory studies, the
Integrated Science Assessment for Oxides of Nitrogen —Health Criteria (NOx ISA) (U.S. EPA,
2016c) concluded that there is a likely causal relationship between respiratory health effects and
short-term exposure to NO2. These epidemiologic and experimental studies encompass a number
5-16
-------
of endpoints including emergency department visits and hospitalizations, respiratory symptoms,
airway hyperresponsiveness, airway inflammation, and lung function. The NOx ISA also
concluded that the relationship between short-term NO2 exposure and premature mortality was
"suggestive but not sufficient to infer a causal relationship," because it is difficult to attribute the
mortality risk effects to NO2 alone. Although the NOx ISA stated that studies consistently
reported a relationship between NO2 exposure and mortality, the effect was generally smaller
than that for other pollutants such as PM.
5.2.4 Ozone Welfare Benefits
Exposure to ozone has been associated with a wide array of vegetation and ecosystem
effects in the published literature (U.S. EPA, 2013a). Sensitivity to ozone is highly variable
across species, with over 65 plant species identified as "ozone-sensitive", many of which occur
in state and national parks and forests. These effects include those that damage or impair the
intended use of the plant or ecosystem. Such effects can include reduced growth and/or biomass
production in sensitive plant species, including forest trees, reduced yield and quality of crops,
visible foliar injury, species composition shift, and changes in ecosystems and associated
ecosystem services.
5.2.5 NO2 Welfare Benefits
As described in the Integrated Science Assessment for Oxides of Nitrogen and Sulfur —
Ecological Criteria (NOx/SOx ISA) (U.S. EPA, 2008b), NOx emissions also contribute to a
variety of adverse welfare effects, including those associated with acidic deposition, visibility
impairment, and nutrient enrichment. Deposition of nitrogen causes acidification, which can
cause a loss of biodiversity of fishes, zooplankton, and macro invertebrates in aquatic
ecosystems, as well as a decline in sensitive tree species, such as red spruce (Picea rubens) and
sugar maple (Acer saccharum) in terrestrial ecosystems. In the northeastern U.S., the surface
waters affected by acidification are a source of food for some recreational and subsistence
fishermen and for other consumers and support several cultural services, including aesthetic and
educational services and recreational fishing. Biological effects of acidification in terrestrial
ecosystems are generally linked to aluminum toxicity, which can cause reduced root growth,
restricting the ability of the plant to take up water and nutrients. These direct effects can, in turn,
5-17
-------
increase the sensitivity of these plants to stresses, such as droughts, cold temperatures, insect
pests, and disease leading to increased mortality of canopy trees. Terrestrial acidification affects
several important ecological services, including declines in habitat for threatened and endangered
species (cultural), declines in forest aesthetics (cultural), declines in forest productivity
(provisioning), and increases in forest soil erosion and reductions in water retention (cultural and
regulating). (U.S. EPA, 2008b)
Deposition of nitrogen is also associated with aquatic and terrestrial nutrient enrichment.
In estuarine waters, excess nutrient enrichment can lead to eutrophication. Eutrophication of
estuaries can disrupt an important source of food production, particularly fish and shellfish
production, and a variety of cultural ecosystem services, including water-based recreational and
aesthetic services. Terrestrial nutrient enrichment is associated with changes in the types and
number of species and biodiversity in terrestrial systems. Excessive nitrogen deposition upsets
the balance between native and nonnative plants, changing the ability of an area to support
biodiversity. When the composition of species changes, then fire frequency and intensity can
also change, as nonnative grasses fuel more frequent and more intense wildfires. (U.S. EPA,
2008b)
5.2.6 Visibility Impairment Benefits
Reducing secondary formation of PM2.5 would improve levels of visibility in the U.S.
because suspended particles and gases degrade visibility by scattering and absorbing light (U.S.
EPA, 2009). Fine particles with significant light-extinction efficiencies include sulfates, nitrates,
organic carbon, elemental carbon, and soil (Sisler, 1996). Visibility has direct significance to
people's enjoyment of daily activities and their overall sense of wellbeing. Good visibility
increases the quality of life where individuals live and work, and where they engage in
recreational activities. Particulate sulfate is the dominant source of regional haze in the eastern
U.S. and particulate nitrate is an important contributor to light extinction in California and the
upper Midwestern U.S., particularly during winter (U.S. EPA, 2009). Previous analyses (U.S.
EPA, 201 la) show that visibility benefits can be a significant welfare benefit category. Without
air quality modeling, we are unable to estimate visibility-related benefits, and we are also unable
to determine whether the emission reductions associated with the final emission guidelines
would be likely to have a significant impact on visibility in urban areas or Class I areas.
5-18
-------
Reductions in emissions of NO2 will improve the level of visibility throughout the United
States because these gases (and the particles of nitrate and sulfate formed from these gases)
impair visibility by scattering and absorbing light (U.S. EPA, 2009). Visibility is also referred to
as visual air quality (VAQ), and it directly affects people's enjoyment of a variety of daily
activities (U.S. EPA, 2009). Good visibility increases quality of life where individuals live and
work, and where they travel for recreational activities, including sites of unique public value,
such as the Great Smoky Mountains National Park (U. S. EPA, 2009).
5.3 References
Amorim MIM, Mergler D, Bahia MO, Dubeau H, Miranda D, Lebel J, et al. 2000. Cytogenetic
damage related to low levels of methyl mercury contamination in the Brazilian Amazon.
An Acad Bras Cienc 72:497-507; doi:10.1590/S0001- 37652000000400004.
ATSDR. 1999. ATSDR - Toxicological Profile: Mercury.
Bell, M.L., A. McDermott, S.L. Zeger, J.M. Sarnet, and F. Dominici. 2004. "Ozone and Short-
Term Mortality in 95 U.S. Urban Communities, 1987-2000." Journal of the American
Medical Association. 292(19):2372-8.
Bell, M.L., F. Dominici, and J.M. Samet. 2005. "A Meta-Analysis of Time-Series Studies of
Ozone and Mortality with Comparison to the National Morbidity, Mortality, and Air
Pollution Study." Epidemiology. 16(4):436-45.
Fann N, Nolte CG, Dolwick P, Spero TL, Brown AC, Phillips S, et al. 2015. The geographic
distribution and economic value of climate change-related ozone health impacts in the
United States in 2030. J Air Waste Manag Assoc 65;
doi: 10.1080/10962247.2014.996270.
Gwinn, M.R., J. Craig, D.A. Axelrad, R. Cook, C. Dockins, N. Fann, R. Fegley, D.E. Guinnup,
G. Helfand, B. Hubbell, S.L. Mazur, T. Palma, R.L. Smith, J. Vandenberg, and
B. Sonawane. 2011. "Meeting report: Estimating the benefits of reducing hazardous air
pollutants—summary of 2009 workshop and future considerations." Environmental
Health Perspectives. 119(1): 125-30.
Huang Y, Dominici F, Bell M. 2004. Bayesian Hierarchical Distributed Lag Models for Summer
Ozone Exposure and Cardio-Respiratory Mortality. Johns Hopkins Univ Dept Biostat
Work Pap Ser.
IARC. 1994. Mercury Carcinogenicity.
Ito, K., S.F. De Leon, and M. Lippmann. 2005. "Associations Between Ozone and Daily
Mortality: Analysis and Meta-Analysis." Epidemiology. 16(4):446-57.
5-19
-------
Jerrett M, Burnett RT, Pope CA, Ito K, Thurston G, Krewski D, et al. 2009. Long-term ozone
exposure and mortality. NEngl J Med 360:1085-95; doi:10.1056/NEJMoa0803894.
Jhun I, Fann N, Zanobetti A, Hubbell B. 2014. Effect modification of ozone-related mortality
risks by temperature in 97 US cities. Environment International. 73:128-34.
Krewski D., M. Jerrett, R.T. Burnett, R. Ma, E. Hughes, Y. Shi, et al. 2009. Extended Follow-Up
and Spatial Analysis of the American Cancer Society Study Linking Particulate Air
Pollution and Mortality. HEI Research Report, 140, Health Effects Institute, Boston, MA.
Lepeule, J., F. Laden, D. Dockery, and J. Schwartz. 2012. "Chronic Exposure to Fine Particles
and Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to
2009." Environmental Health Perspectives. 120(7):965-70.
Levy, J.I., S.M. Chemerynski, and J.A. Sarnat. 2005. "Ozone Exposure and Mortality: An
Empiric Bayes Metaregression Analysis." Epidemiology. 16(4):458-68.
National Research Council (NRC). 2000. Toxicological Effects of Methylmercury: Committee
on the Toxicological Effects of Methylmercury." Board on Environmental Studies and
Toxicology. National Academies Press. Washington, DC.
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air
Pollution Regulations. National Academies Press. Washington, DC.
National Research Council (NRC). 2008. Estimating Mortality Risk Reduction and Economic
Benefits from Controlling Ozone Air Pollution. National Academies Press. Washington,
DC.
Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis.
Washington, DC. Available at: < http://www.whitehouse.gov/omb/circulars/a004/a-
4.html>.
Ren, C., G.M. Williams, K. Mengersen, L. Morawska, and S. Tong. 2008a. "Does Temperature
Modify Short-Term Effects of Ozone on Total Mortality in 60 Large Eastern U.S.
Communities? An Assessment Using the NMMAPS Data." Environment International.
34:451-458.
Ren, C., G.M. William, L. Morawska, K. Mengensen, and S. Tong. 2008b. "Ozone Modifies
Associations between Temperature and Cardiovascular Mortality: Analysis of the
NMMAPS Data." Occupational and Environmental Medicine. 65:255-260.
Sisler, J.F. 1996. Spatial and Seasonal Patterns and Long-Term Variability of the Composition of
the Haze in the United States: An analysis of data from the IMPROVE network. CIRA
Report, ISSN 0737-5352-32, Colorado State University.
Schwartz, J. 2005. "How Sensitive is the Association between Ozone and Daily Deaths to
Control for Temperature?" American Journal of Respiratory and Critical Care Medicine.
171(6): 627-31.
5-20
-------
Smith RL, Xu B, Switzer P. 2009. Reassessing the relationship between ozone and short-term
mortality in U.S. urban communities. Inhal Toxicol 21 Suppl 2:37-61;
doi: 10.1080/08958370903161612.
U.S. Environmental Protection Agency (U.S. EPA). 2002. Methylmercury (MeHg) CASRN
22967-92-6 | IRIS | US EPA, ORD
U.S. Environmental Protection Agency (U.S. EPA). 2008b. Integrated Science Assessment for
Oxides of Nitrogen and Sulfur-Ecological Criteria National (Final Report). National
Center for Environmental Assessment - RTP Division, Research Triangle Park, NC.
EPA/600/R-08/139. December. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment - RTP Division, Research Triangle Park, NC. December.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Integrated Science Assessment for
Carbon Monoxide. National Center for Environmental Assessment, Research Triangle
Park, NC. EPA/600/R-09/019F. January. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Technical Support Document:
Summary of Expert Opinions on the Existence of a Threshold in the Concentration-
Response Function for PM2.5-related Mortality. Research Triangle Park, NC. June.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Regulatory Impact Analysis (RIA)
for Existing Stationary Compression Ignition Engines NESHAP Final Draft.
U.S. Environmental Protection Agency (U.S. EPA). 2010d. Regulatory Impact Analysis for the
Proposed Federal Transport Rule.
U.S. Environmental Protection Agency (U.S. EPA). 201 la. Policy Assessment for the Review of
the Particulate Matter National Ambient Air Quality Standards. EPA-452/D-11-003.
Office of Air Quality Planning and Standards, Health and Environmental Impacts
Division. April. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 201 lb. The Benefits and Costs of the Clean
Air Act from 1990 to 2020. Office of Air and Radiation, Washington, DC. March.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 201 lc. Regulatory Impact Analysis for the
Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter
and Ozone in 27 States; Correction of SIP Approvals for 22 States.
5-21
-------
U.S. Environmental Protection Agency (U.S. EPA). 201 Id. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards.
U.S. Environmental Protection Agency (U.S. EPA). 2012. Regulatory Impact Analysis for the
Final Revisions to the National Ambient Air Quality Standards for Particulate Matter.
EPA-452/R-12-003. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division, Research Triangle Park, NC. Available at: <
http://www.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf>.
U.S. Environmental Protection Agency (U.S. EPA). 2013a. Integrated Science Assessment of
Ozone and Related Photochemical Oxidants (Final Report). EPA/600/R-10/076F.
National Center for Environmental Assessment - RTP Division, Research Triangle Park.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2013b. Regulatory Impact Analysis for the
Final Revisions to the National Ambient Air Quality Standards for Particulate Matter.
U.S. Environmental Protection Agency (U.S. EPA). 2014a. Regulatory Impact Analysis (RIA)
for Proposed Residential Wood Heaters NSPS Revision.
U.S. Environmental Protection Agency (U.S. EPA). 2014b. Regulatory Impact Analysis for the
Proposed Carbon Pollution Guidelines for Existing Power Plants and Emission Standards
for Modified and Reconstructed Power Plants.
U.S. Environmental Protection Agency (U.S. EPA). 2014c. Regulatory Impact Analysis of the
Proposed Revisions to the National Ambient Air Quality Standards for Ground-Level
Ozone.
U.S. Environmental Protection Agency (U.S. EPA). 2015a. Regulatory Impact Analysis for
Residential Wood Heaters NSPS Revision: Final Report.
U.S. Environmental Protection Agency (U.S. EPA). 2015b. Regulatory Impact Analysis for the
Clean Power Plan Final Rule.
U.S. Environmental Protection Agency (U.S. EPA). 2015c. Regulatory Impact Analysis for the
Proposed Cross-State Air Pollution Rule (CSAPR) Update for the 2008 Ozone National
Ambient Air Quality Standards (NAAQS).
U.S. Environmental Protection Agency (U.S. EPA). 2015d. Regulatory Impact Analysis for the
Proposed Federal Plan Requirements for Greenhouse Gas Emissions from Electric Utility
Generating Units Constructed on or Before January 8, 2014; Model Trading Rules;
Amendments to Framework Regulations. 4-52
U.S. Environmental Protection Agency (U.S. EPA). 2015e. Regulatory Impact Analysis of the
Final Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone,
EPA-452/R-15-07. EPA-452/R-12-003. Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division, Research Triangle Park, NC.
(https://www.epa.gOv/sites/production/files/2016-02/documents/20151001ria.pdf).
5-22
-------
U.S. Environmental Protection Agency (U.S. EPA). 2016a. Guidelines for Preparing Economic
Analyses.
U.S. Environmental Protection Agency (U.S. EPA). 2016b. Regulatory Impact Analysis of the
Cross-State Air Pollution Rule (CSAPR) Update for the 2008 National Ambient Air
Quality Standards for Ground-Level Ozone.
U.S. Environmental Protection Agency (U.S. EPA). 2016c. Integrated Science Assessment for
Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available at: <
https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310879>.
U.S. Environmental Protection Agency (U.S. EPA). 2017. Integrated Science Assessment for
Sulfur Oxides—Health Criteria (Final Report). National Center for Environmental
Assessment - RTP Division, Research Triangle Park, NC. September. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 2018. Environmental Benefits Mapping and
Analysis Program - Community Edition. User's Manual. Office of Air Quality Planning
and Standards, Health and Environmental Impacts Division, Research Triangle Park, NC.
Available at: < https://www.epa.gov/sites/production/files/2015-04/documents/benmap-
ce_user_manual_march_2015 ,pdf>
U.S. Environmental Protection Agency (U.S. EPA). 2019a. Integrated Science Assessment (ISA)
for Particulate Matter (Final Report, 2019). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-19/188, 2019.
U.S. Environmental Protection Agency (U.S. EPA). 2019b. Regulatory Impact Analysis for
Repeal of the Clean Power Plan; Emission Guidelines for Greenhouse Gas Emissions
from Existing Electric Utility Generating Units; Revisions to Emission Guidelines
Implementing Regulations. EPA-452/R19-003. June 2019
U.S. Environmental Protection Agency (U.S. EPA). 2020a. Benefit and Cost Analysis for
Revisions to the Effluent Limitations Guidelines and Standards for the Steam Electric
Power Generating Point Source Category. U.S. Environmental Protection Agency,
Washington, DC, EPA-821-R-20-003.
U.S. Environmental Protection Agency (U.S. EPA). 2020b. Integrated Science Assessment (ISA)
for Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-20/012, 2020.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2000. An
SAB Report on EPA's White Paper Valuing the Benefits of Fatal Cancer Risk Reduction.
EPA-SAB-EEAC-00-013. July. Available at:
.
5-23
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004.
Advisory Council on Clean Air Compliance Analysis Response to Agency Request on
Cessation Lag. EPA-COUNCIL-LTR-05-001. December. Available at:
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009b.
Review of Integrated Science Assessment for Particulate Matter (Second External
Review Draft, July 2009). EPA-CASAC-10-001. November. Available at:
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2010.
Review of EPA's Draft Health Benefits of the Second Section 812 Prospective Study of
the CAA
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2020.
Transmittal of the Science Advisory Board's Consideration of the Scientific and
Technical Basis of EPA's Proposed Rule titled "Increasing Consistency and
Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking
Process." Available at:
.
USGCRP. 2016. The Impacts of Climate Change on Human Health in the United States: A
Scientific Assessment.; doi:http://dx.doi.org/10.7930/J0R49NQX.
Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 1997. "The Relationship between Selected of
postneonatal infant mortality and particulate air pollution in the United States."
Environmental Health Perspectives. 105(6): 608-612
Zanobetti A, Schwartz J. 2008. Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med
177:184-9; doi: 10.1164/rccm.200706-8230C.
5-24
-------
APPENDIX 5A: UNCERTAINTY ASSOCIATED WITH ESTIMATING THE SOCIAL
COST OF CARBON
Overview
This appendix provides additional information on the climate benefits associated with CO2
emissions reductions. It first provides a brief overview of the 2009 Endangerment Finding and
climate science assessments released since 2009 and then provides greater detail about the
methodology used to estimate climate benefits due to changes in C02 emissions. The
methodology used to develop interim domestic SC-CO2 estimates and uncertainty associated
with the interim SC-CO2 values are the same as described in the RIA for the Affordable Clean
Energy (ACE) final rule (U.S. EPA, 2019). This appendix applies the methodology to the
analysis of the climate benefits of changes in CO2 emissions under the regulatory options
described in Chapter 4.
5A.1 Overview of 2009 Endangerment Finding and Climate Science Assessments
A benefit of this proposal is reducing emissions of CO2. In this section, we provide a brief
overview of the 2009 Endangerment Finding and climate science assessments released since
2009.
Through the implementation of Clean Air Act (CAA) regulations, EPA addresses the
negative externalities caused by air pollution. In 2009, the EPA Administrator found that
elevated concentrations of greenhouse gases in the atmosphere may reasonably be anticipated
both to endanger public health and to endanger public welfare. For health, these include the
increased likelihood of heat waves, negative impacts on air quality, more intense hurricanes,
more frequent and intense storms and heavy precipitation, and impacts on infectious and
waterborne diseases. For welfare, these include reduced water supplies in some regions,
increased water pollution, increased occurrences of floods and droughts, rising sea levels and
damage to coastal infrastructure, increased peak electricity demand, changes in ecosystems, and
impacts on indigenous communities.
Major scientific assessments released since the 2009 Endangerment Finding have
improved scientific understanding of the climate and provide even more evidence that
greenhouse gas (GHG) emissions endanger public health and welfare for current and future
5A-1
-------
generations. The National Climate Assessment (NCA), in particular, assessed the impacts of
climate change on human health in the United States, finding that Americans will be affected by
"increased extreme weather events, wildfire, decreased air quality, threats to mental health, and
illnesses transmitted by food, water, and disease-carriers such as mosquitoes and ticks." These
assessments also detail the risks to vulnerable groups such as children, the elderly and low-
income households. Furthermore, the assessments present an improved understanding of the
impacts of climate change on public welfare, higher projections of future sea level rise than had
been previously estimated, a better understanding of how the warmth in the next century may
reach levels that would be unprecedented relative to the preceding millions of years of history,
and new assessments of the impacts of climate change on permafrost and ocean acidification.
The impacts of GHG emissions will be realized worldwide, independent of their location of
origin, and impacts outside of the United States will produce consequences relevant to the United
States. Table 5A-1 summarizes the quantified and unquantified climate benefits in this analysis.
Table 5A-1. Climate Effects
Benefits Specific Effect
Effect Has Been
Effect Has Been
More
Category
Quantified
Monetized
Information
Improved Environment
Reduced Global climate impacts from C02
1
~2
SCC TSD
climate effects Climate impacts from ozone and black
—
—
Ozone ISA, PM
carbon (directly emitted PM)
ISA3
Other climate impacts (e.g., other GHGs
—
—
IPCC3
such as methane, aerosols, other impacts)
1 The global climate and related impacts of C02 emissions changes, such as sea level rise, are estimated within each
integrated assessment model as part of the calculation of the SC-CO2.
2 The monetized damages, which are relevant for conducting the benefit-cost analysis, are used in this RIA to
estimate the welfare effects of quantified changes in CO2 emissions. The SC-CO2 estimates used in the main
benefits analysis in this RIA focus on the projected impacts of climate change that are anticipated to directly occur
within U.S. borders. Monetized damages from global impacts are provided in Section 5A.4 of this appendix.
3 We assess these benefits qualitatively because we do not have sufficient confidence in available data or methods.
There are several important considerations in assessing the climate-related benefits for an
ozone air quality-focused rulemaking. First, the estimated health benefits do not account for any
climate-related air quality changes (e.g., increased ambient ozone associated with higher
temperatures). Excluding climate-related air quality changes may underestimate ozone-related
health benefits. It is unclear how PIvfc.s-related health benefits would be affected by excluding
climate-related air quality changes since the science is unclear as to how climate change may
affect PM2.5 exposure. Second, the estimated health benefits also do not consider temperature
5A-2
-------
modification of PM2.5 and ozone risks (Roberts 2004; Ren 2006a, 2006b, 2008a, 2008b).
Excluding temperature modification of air pollution risks and international air quality-related
health benefits likely leads to underestimation of quantified health benefits (Anenberg et al,
2009, Jhun et al, 2014). Fourth, as noted earlier, we do not estimate the climate benefits
associated with reductions in PM and O3 precursors.
5A.2 Overview of Methodology Used to Develop Interim Domestic SC-C02 Estimates
The methodology used to develop interim domestic SC-CO2 estimates and uncertainty
associated with the interim SC-CO2 values are the same as described in the RIA for the ACE
final rule (U.S. EPA, 2019). This section applies the methodology to the analysis of the climate
benefits of changes in CO2 emissions under the regulatory alternatives described in Chapter 4.
The domestic SC-CO2 estimates rely on the same ensemble of three integrated
assessment models (IAMs) that were used to develop the global SC-CO2 estimates (DICE 2010,
FUND 3.8, and PAGE 2009)1 used in the benefits analysis of the 2016 rule (U.S. EPA, 2016).
The three IAMs translate emissions into changes in atmospheric greenhouse concentrations,
atmospheric concentrations into changes in temperature, and changes in temperature into
economic damages. The emissions projections used in the models are based on specified
socioeconomic (GDP and population) pathways. These emissions are translated into atmospheric
concentrations, and concentrations are translated into warming based on each model's simplified
representation of the climate and a key parameter, equilibrium climate sensitivity. The effect of
the changes is estimated in terms of consumption-equivalent economic damages. As in the
estimation of SC-CO2 estimates used in the 2016 benefits analysis (U.S. EPA, 2016), three key
inputs were harmonized across the three models: a probability distribution for equilibrium
climate sensitivity; five scenarios for economic, population, and emissions growth; and discount
rates.2 All other model features were left unchanged. Future damages are discounted using
constant discount rates of both 3 and 7 percent, as recommended by OMB Circular A-4. The
domestic share of the global SC-CO2 - i.e., an approximation of the climate change impacts that
1 The full model names are as follows: Dynamic Integrated Climate and Economy (DICE); Climate Framework for
Uncertainty, Negotiation, and Distribution (FUND); and Policy Analysis of the Greenhouse Gas Effect (PAGE).
2 See the summary of the methodology in the 2015 Clean Power Plan docket, document ID number EPA-HQ-OAR-
2013-0602-37033, "Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866, Interagency Working Group on Social Cost of Carbon United States Government, 2015. See also
National Academies of Sciences et al., 2017 for a detailed discussion of each of these modeling assumptions.
5A-3
-------
occur within U.S. borders - are calculated directly in both FUND and PAGE. However, DICE
2010 generates only global SC-CO2 estimates. Therefore, EPA approximated U.S. damages as 10
percent of the global values from the DICE model runs, based on the results from a regionalized
version of the model (RICE 2010) reported in Table 2 of Nordhaus (2017).
The steps involved in estimating the social cost of CO2 are as follows. The three
integrated assessment models (FUND, DICE, and PAGE) are run using the harmonized
equilibrium climate sensitivity distribution, five socioeconomic and emissions scenarios, and
constant discount rates described above. Because the climate sensitivity parameter is modeled
probabilistically, and because PAGE and FUND incorporate uncertainty in other model
parameters, the final output from each model run is a distribution over the SC-CO2 in year t
based on a Monte Carlo simulation of 10,000 runs. For each of the IAMs, the basic
computational steps for calculating the social cost estimate in a particular year t are:
1. Calculate the temperature effects and (consumption-equivalent) damages in each year
resulting from the baseline path of emissions;
2. Adjust the model to reflect an additional unit of emissions in year /;
3. Recalculate the temperature effects and damages expected in all years beyond t
resulting from this adjusted path of emissions, as in step 1; and
4. Subtract the damages computed in step 1 from those in step 3 in each model period
and discount the resulting path of marginal damages back to the year of emissions. In
PAGE and FUND step 4 focuses on the damages attributed to the US region in the
models. As noted above, DICE does not explicitly include a separate US region in the
model and therefore, EPA approximates U.S. damages in step 4 as 10 percent of the
global values based on the results of Nordhaus (2017).
This exercise produces 30 separate distributions of the SC-CO2 for a given year, the
product of 3 models, 2 discount rates, and 5 socioeconomic scenarios. Following the approach
used by the IWG, the estimates are equally weighted across models and socioeconomic scenarios
in order to reduce the dimensionality of the results down to two separate distributions, one for
each discount rate.
5A.3 Treatment of Uncertainty in Interim Domestic SC-CO2 Estimates
There are various sources of uncertainty in the SC-CO2 estimates used in this BCA. Some
uncertainties pertain to aspects of the natural world, such as quantifying the physical effects of
5A-4
-------
greenhouse gas emissions on Earth systems. Other sources of uncertainty are associated with
current and future human behavior and well-being, such as population and economic growth,
GHG emissions, the translation of Earth system changes to economic damages, and the role of
adaptation. It is important to note that even in the presence of uncertainty, scientific and
economic analysis can provide valuable information to the public and decision makers, though
the uncertainty should be acknowledged and when possible taken into account in the analysis
(Institute of Medicine, 2013). OMB Circular A-4 also requires a thorough discussion of key
sources of uncertainty in the calculation of benefits and costs, including more rigorous
quantitative approaches for higher consequence rules. This section summarizes the sources of
uncertainty considered in a quantitative manner in the domestic SC-CO2 estimates.
The domestic SC-CO2 estimates consider various sources of uncertainty through a
combination of a multi-model ensemble, probabilistic analysis, and scenario analysis. EPA
provides a summary of this analysis here; more detailed discussion of each model and the
harmonized input assumptions can be found in the 2017 National Academies report. For
example, the three IAMs used collectively span a wide range of Earth system and economic
outcomes to help reflect the uncertainty in the literature and in the underlying dynamics being
modeled. The use of an ensemble of three different models at least partially addresses the fact
that no single model includes all of the quantified economic damages. It also helps to reflect
structural uncertainty across the models, which is uncertainty in the underlying relationships
between GHG emissions, Earth systems, and economic damages that are included in the models.
Bearing in mind the different limitations of each model and lacking an objective basis upon
which to differentially weight the models, the three integrated assessment models are given equal
weight in the analysis.
Monte Carlo techniques were used to run the IAMs a large number of times. In each
simulation the uncertain parameters are represented by random draws from their defined
probability distributions. In all three models the equilibrium climate sensitivity is treated
probabilistically based on the probability distribution from Roe and Baker (2007) calibrated to
the IPCC AR4 consensus statement about this key parameter.3 The equilibrium climate
3 Specifically, the Roe and Baker distribution for the climate sensitivity parameter was bounded between 0 and 10
with a median of 3 °C and a cumulative probability between 2 and 4.5 °C of two-thirds.
5A-5
-------
sensitivity is a key parameter in this analysis because it helps define the strength of the climate
response to increasing GHG concentrations in the atmosphere. In addition, the FUND and PAGE
models define many of their parameters with probability distributions instead of point estimates.
For these two models, the model developers' default probability distributions are maintained for
all parameters other than those superseded by the harmonized inputs (i.e., equilibrium climate
sensitivity, socioeconomic and emissions scenarios, and discount rates). More information on the
uncertain parameters in PAGE and FUND is available upon request.
For the socioeconomic and emissions scenarios, uncertainty is included in the analysis by
considering a range of scenarios selected from the Stanford Energy Modeling Forum exercise,
EMF-22. Given the dearth of information on the likelihood of a full range of future
socioeconomic pathways at the time the original modeling was conducted, and without a basis
for assigning differential weights to scenarios, the range of uncertainty was reflected by simply
weighting each of the five scenarios equally for the consolidated estimates. To better understand
how the results vary across scenarios, results of each model run are available in the docket for
the ACE final rule (Docket ID EPA-HQ-OAR-2017-0355).
The outcome of accounting for various sources of uncertainty using the approaches
described above is a frequency distribution of the SC-CO2 estimates for emissions occurring in a
given year for each discount rate. Unlike the approach taken for consolidating results across
models and socioeconomic and emissions scenarios, the SC-CO2 estimates are not pooled across
different discount rates because the range of discount rates reflects both uncertainty and, at least
in part, different policy or value judgements; uncertainty regarding this key assumption is
discussed in more detail below. The frequency distributions reflect the uncertainty around the
input parameters for which probability distributions were defined, as well as from the multi-
model ensemble and socioeconomic and emissions scenarios where probabilities were implied
by the equal weighting assumption. It is important to note that the set of SC-CO2 estimates
obtained from this analysis does not yield a probability distribution that fully characterizes
uncertainty about the SC-CO2 due to impact categories omitted from the models and sources of
uncertainty that have not been fully characterized due to data limitations.
5A-6
-------
Figure 5A-1 presents the frequency distribution of the domestic SC-CO2 estimates for
emissions in 2030 for each discount rate. Each distribution represents 150,000 estimates based
on 10,000 simulations for each combination of the three models and five socioeconomic and
emissions scenarios. In general, the distributions are skewed to the right and have long right tails,
which tend to be longer for lower discount rates. To highlight the difference between the impact
of the discount rate on the SC-CO2 and other quantified sources of uncertainty, the bars below
the frequency distributions provide a symmetric representation of quantified variability in the
SC-CO2 estimates conditioned on each discount rate. The full set of SC-CO2 results through
2050 is available in the docket for the ACE final rule (Docket ID EPA-HQ-OAR-2017-0355).
(/>
C
o
J?
E
CD
o
03
UO
o
0
CO
d
CM
O
o
0
7% Average = $1
Discount Rate
~ 7%
~ 3%
3% Average = $8
~dt
~ ~dl
_L
5,r - 95th Percentile
] J of Simulations
l l III
0 4
ill ill r~i i r~i i i i i i i i i i i
12 16 20 24 28 32 36 40
Interim U.S. Domestic Social Cost of Carbon in 2030 [2016$ / metric ton C02]
Figure 5A-1. Frequency Distribution of Interim Domestic SC-CO2 Estimates for 2030 (in
2016$ per Metric Ton CO2)
As illustrated by the frequency distributions in Figure 5A-1, the assumed discount rate
plays a critical role in the ultimate estimate of the social cost of carbon. This is because CO2
emissions today continue to impact society far out into the future, so with a higher discount rate,
costs that accrue to future generations are weighted less, resulting in a lower estimate. Circular
5A-7
-------
A-4 recommends that costs and benefits be discounted using the rates of 3 percent and 7 percent
to reflect the opportunity cost of consumption and capital, respectively. Circular A-4 also
recommends quantitative sensitivity analysis of key assumptions4, and offers guidance on what
sensitivity analysis can be conducted in cases where a rule will have important intergenerational
benefits or costs. To account for ethical considerations of future generations and potential
uncertainty in the discount rate over long time horizons, Circular A-4 suggests "further
sensitivity analysis using a lower but positive discount rate in addition to calculating net benefit
using discount rates of 3 and 7 percent" (page 36) and notes that research from the 1990s
suggests intergenerational rates "from 1 to 3 percent per annum" (OMB, 2003). EPA considers
the uncertainty in this key assumption by calculating the domestic SC-CO2 based on a
2.5 percent discount rate, in addition to the 3 and 7 percent used in the main analysis. Using a 2.5
percent discount rate, the average domestic SC-CO2 estimate across all the model runs for
emissions occurring over 2021-2025 $10 per metric ton of CO2 (in 2016$). In this case the
domestic climate benefits under the proposed alternative at a 2.5 percent discount rate in 2021
are $0.4 million (2016$) and in 2025 are $47 million.
In addition to the approach to accounting for the quantifiable uncertainty described
above, the scientific and economics literature has further explored known sources of uncertainty
related to estimates of the SC-CO2. For example, researchers have published papers that explore
the sensitivity of IAMs and the resulting SC-CO2 estimates to different assumptions embedded in
the models (e.g., Hope, 2013, Anthoff et al., 2013, and Nordhaus, 2014). However, there remain
additional sources of uncertainty that have not been fully characterized and explored due to
remaining data limitations. Additional research is needed in order to expand the quantification of
various sources of uncertainty in estimates of the SC-CO2 (e.g., developing explicit probability
distributions for more inputs pertaining to climate impacts and their valuation). On the issue of
intergenerational discounting, some experts have argued that a declining discount rate would be
appropriate to analyze impacts that occur far into the future (Arrow et al., 2013). However,
additional research and analysis is still needed to develop a methodology for implementing a
declining discount rate and to understand the implications of applying these theoretical lessons in
practice. The 2017 National Academies report also provides recommendations pertaining to
4 "If benefit or cost estimates depend heavily on certain assumptions, you should make those assumptions explicit
and carry out sensitivity analyses using plausible alternative assumptions." (OMB, 2003, page 42).
5A-8
-------
discounting, emphasizing the need to more explicitly model the uncertainty surrounding discount
rates over long time horizons, its connection to uncertainty in economic growth, and, in turn, to
climate damages using a Ramsey-like formula (National Academies of Sciences et al., 2017).
These and other research needs are discussed in detail in the 2017 National Academies'
recommendations for a comprehensive update to the current methodology, including a more
robust incorporation of uncertainty.
5A.4 Global Climate Benefits
In addition to requiring reporting of impacts at a domestic level, OMB Circular A-4 states
that when an agency "evaluate[s] a regulation that is likely to have effects beyond the borders of
the United States, these effects should be reported separately" (OMB, 2003; page 15).5 This
guidance is relevant to the valuation of damages from CO2 and other GHGs, given that GHGs
contribute to damages around the world independent of the country in which they are emitted.
Therefore, this section presents the global climate benefits in 2021 and 2025 from this proposed
rule using the global SC-CO2 estimates corresponding to the model runs that generated the
domestic SC-CO2 estimates used in the main analysis. The average global SC-CO2 estimate
across all the model runs for emissions occurring over 2021-2025 range from $5 to $6 per metric
ton of CO2 emissions (in 2016$) using a 7 percent discount rate, and $49 to $53 per metric ton of
CO2 emissions (in 2016$) using a 3 percent discount rate. In the 2021-2025 timeframe, the
domestic SC-CO2 estimates presented above are approximately 19 percent and 14 percent of the
global SC-CO2 estimates for the 7 percent and 3 percent discount rates, respectively. Applying
these estimates to the CO2 emission reductions results in estimated global climate benefits in
2021 of $0.2 million using a 7 percent discount rate and $2.0 million using a 3 percent discount
rate. By 2025, the estimated global climate benefits are $24.5 million using a 7 percent discount
rate and $222.6 million using a 3 percent discount rate. Under the sensitivity analysis considered
5 While Circular A-4 does not elaborate on this guidance, the basic argument for adopting a domestic only
perspective for the central benefit-cost analysis of domestic policies is based on the fact that the authority to regulate
only extends to a nation's own residents who have consented to adhere to the same set of rules and values for
collective decision-making, as well as the assumption that most domestic policies will have negligible effects on the
welfare of other countries' residents (U.S. EPA, 2010b; Kopp et al., 1997; Whittington et al., 1986). In the context
of policies that are expected to result in substantial effects outside of U.S. borders, an active literature has emerged
discussing how to appropriately treat these impacts for purposes of domestic policymaking (e.g., Gayer et al., 2016,
2017; Anthoff et al., 2010; Fraas et al., 2016; Revesz et al., 2017). This discourse has been primarily focused on the
regulation of GHGs, for which domestic policies may result in impacts outside of U.S. borders due to the global
nature of the pollutants.
5A-9
-------
above using a 2.5 percent discount rate, the average global SC-CO2 estimate across all the model
runs for emissions occurring over 2021-2025 ranges from $72 to $77 per metric ton of CO2
(2016 dollars); in this case the global climate benefits in 2021 are $3.1 million; by 2025, the
global benefits in this sensitivity case increase to $331.8 million.
5A.5 References
40 CFR Chapter I [EPA-HQ-OAR-2009-0171; FRL-9091-8] RIN 2060-ZA14,
"Endangerment and Cause or Contribute Findings for Greenhouse Gases under Section
202(a) of the Clean Air Act," Federal Register, Vol. 74, No. 239, Tuesday, December 15,
2009, Rules and Regulations.
Anenberg SC, West IJ, Fiore AM, Jaffe DA, Prather MJ, Bergmann D, Cuvelier K, Dentener FJ,
Duncan BN, Gauss M, Hess P, Jonson JE, Lupu A, Mackenzie IA, Marmer E, Park RJ,
Sanderson MG, Schultz M, Shindell DT, Szopa S, Vivanco MG, Wild O, Zeng G. 2009.
Intercontinental impacts of ozone pollution on human mortality. Environmental Science
and Technology. 43(17): 6482-7.
Docket ID EPA-HQ-OAR-2009-0472-114577, Technical Support Document: Social Cost of
Carbon for Regulatory Impact Analysis Under Executive Order 12866, Interagency
Working Group on Social Cost of Carbon, with participation by the Council of Economic
Advisers, Council on Environmental Quality, Department of Agriculture, Department of
Commerce, Department of Energy, Department of Transportation, Environmental
Protection Agency, National Economic Council, Office of Energy and Climate Change,
Office of Management and Budget, Office of Science and Technology Policy, and
Department of Treasury (February 2010). Available at:
.
Docket ID EPA-HQ-OAR-2013-0602, Technical Support Document: Technical Update of the
Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866,
Interagency Working Group on Social Cost of Carbon, with Participation by Council of
Economic Advisers, Council on Environmental Quality, Department of Agriculture,
Department of Commerce, Department of Energy, Department of Transportation,
Domestic Policy Council, Environmental Protection Agency, National Economic
Council, Office of Management and Budget, Office of Science and Technology Policy,
and Department of Treasury (May 2013, Revised July 2015). Also available at: <
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf>.
Accessed July 15, 2015.
5 A-10
-------
Interagency Working Group on Social Cost of Carbon, with participation by Council of
Economic Advisers, Council on Environmental Quality, Department of Agriculture,
Department of Commerce, Department of Energy, Department of Transportation,
Environmental Protection Agency, National Economic Council, Office of Management
and Budget, Office of Science and Technology Policy, and Department of Treasury.
Response to to Comments: Social Cost of Carbon for Regulatory Impact Analysis Under
Executive Order 12866. July 2015. Available at
Accessed July 15, 2015.
Intergovernmental Panel on Climate Change (IPCC). 2007. Climate Change 2007: Synthesis
Report Contribution of Working Groups I, II and III to the Fourth Assessment Report of
the IPCC. Available at:
. Accessed June 6, 2015.
Intergovernmental Panel on Climate Change (IPCC). 2012: Managing the Risks of Extreme
Events and Disasters to Advance Climate Change Adaptation. A Special Report of
Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge, UK, and New York, NY, USA.
Intergovernmental Panel on Climate Change (IPCC). 2013. Climate Change 2013: The Physical
Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA.
Intergovernmental Panel on Climate Change (IPCC). 2014a. Climate Change 2014: Impacts,
Adaptation, and Vulnerability. Contribution of Working Group II to the Fifth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.
Intergovernmental Panel on Climate Change (IPCC). 2014b. Climate Change 2014: Mitigation
of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of
the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.
Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis.
Washington, DC. Available at: < http://www.whitehouse.gov/omb/circulars/a004/a-
4.html>.
U.S. Environmental Protection Agency. 2016. Regulatory Impact Analysis of the Cross-State Air
Pollution Rule (CSAPR) Update for the 2008 National Ambient Air Quality Standards for
Ground-Level Ozone. Available at:
https://www3.epa.gOv/ttn/ecas/docs/ria/transport_ria_final-csapr-update_2016-09.pdf.
5 A-11
-------
U.S. Environmental Protection Agency. 2019. Regulatory Impact Analysis for the Repeal of the
Clean Power Plan, and the Emission Guidelines for Greenhouse Gas Emissions from
Existing Electric Utility Generating Units. Available at:
5 A-12
-------
APPENDIX 5B: AIR POLLUTION-RELATED HUMAN HEALTH BENEFITS
ESTIMATED USING PREVIOUS METHODS
Overview
This appendix reports the estimated number and economic value of reducing PM2.5 and
Ozone associated with the three regulatory control alternatives across several discount rates. We
estimate the incidence of air pollution-attributable premature deaths and illnesses using methods
first developed in 2009 for PM2.5 (U.S. EPA, 2009) and 2013 for Ozone (U.S. EPA, 2013a).
These methods have not yet been updated to reflect the new evidence reported in the most recent
PM and Ozone Integrated Science Assessments (U.S. EPA 2019a, 2020b). These limitations
notwithstanding, this appendix provides useful context for readers and sheds some light on the
potential magnitude of the benefits associated with the type of emissions reductions estimated in
this proposal.
As noted in the Executive Summary and Chapter 5, EPA intends to update its
methodology in time to quantify the number and economic value of Ozone and PM2.5 health
effects resulting from this rulemaking in the final Revised CSAPR Update RIA. When updating
its approach for quantifying the benefits of changes in PM2.5 and Ozone, the Agency will
incorporate evidence reported in these two recently published ISAs and account for forthcoming
recommendations from the EPA Science Advisory Board. When updating the evidence for each
endpoint, EPA will consider the extent to which there is a causal relationship, whether suitable
epidemiologic evidence exists to quantify the effect and whether the economic value of the effect
may be estimated. Carefully and systematically reviewing the full breadth of this information
requires significant time and resources that were unavailable to the Agency at the time of this
proposal.
5B.1 Estimated Human Health Benefits
The proposal is expected to reduce emissions of ozone season NOx. In the presence of
sunlight, NOx and volatile organic compounds (VOCs) can undergo a chemical reaction in the
atmosphere to form ozone. Reducing NOx emissions generally reduces human exposure to ozone
and the incidence of ozone-related health effects, though the degree to which ozone is reduced
will depend in part on local levels of VOCs. The proposal would also reduce emissions of NOx
5B-1
-------
throughout the year. Because NOxis also a precursor to formation of ambient PM2.5, reducing
these emissions would reduce human exposure to ambient PM2.5 throughout the year and would
reduce the incidence of PM2.5-attributable health effects.1 Reducing emissions of NOx would
also reduce ambient exposure to NO2 and its associated health effects, though we do not quantify
these effects due to lack of data.
The Regulatory Impact Analysis (RIA) for the Particulate Matter (PM) National Ambient
Air Quality Standards (NAAQS) (U.S. EPA 2012), the RIA for the Ozone NAAQS (U.S. EPA
2015e) and the user manual for the Benefits Mapping and Analysis Program—Community
Edition (BenMAP-CE) program (U.S. EPA 2018) each provide a full discussion of the Agency's
approach for quantifying the number and value of estimated air pollution-related impacts.2 In
these documents the reader can find the rationale for selecting health endpoints to quantify; the
demographic, health and economic data we apply within BenMAP-CE; modeling assumptions;
and our techniques for quantifying uncertainty. Additional information regarding our approach
for characterizing uncertainty in PM-attributable risk of premature mortality may be found in the
RIA for the Affordable Clean Energy Rule (U.S. EPA 2019b).
These estimated health benefits do not account for the influence of future changes in the
climate on ambient concentrations of pollutants (USGCRP 2016). For example, recent research
suggests that future changes to climate may create conditions more conducive to forming ozone;
the influence of changes in the climate on PM2.5 concentrations are less clear (Fann et al. 2015).
The estimated health benefits also do not consider the potential for climate-induced changes in
temperature to modify the relationship between ozone and the risk of premature death (Jhun et al.
2014; Ren et al. 2008a, 2008b).
Implementing the proposal will affect the distribution of ozone and PM2.5 concentrations
in much of the U.S.; this includes locations both meeting and exceeding the NAAQS for ozone
1 This RIA does not quantify PM2 5-related benefits associated with S02 emission reductions. As discussed in
Chapter 4, EPA does not estimate significant S02 emission reductions as a result of this proposal. Additionally, this
RIA does not estimate changes in emissions of directly emitted particles. As a result, quantified PM25-related
benefits are subject to uncertainty.
2 The Agency is evaluating the adequacy of the Ozone and PM NAAQS, see 85 FR 49830 and 85 FR 24094. Once
EPA promulgates final PM and Ozone NAAQS, the Agency will revisit its approach for estimating benefits for each
pollutant. Until that point, EPA will continue to apply methods for estimating benefits that are consistent with the
evidence supporting the 2012 PM and 2015 Ozone NAAQS.
5B-2
-------
and PM. This RIA estimates avoided ozone- and PIvfc.s-related health impacts that are distinct
from those reported in the RIAs for both NAAQS (U.S. EPA 2012, 2015e). The ozone and PM2.5
NAAQS RIAs hypothesize, but do not predict, the benefits and costs of strategies that States may
choose to enact when implementing a revised NAAQS; these costs and benefits are illustrative
and cannot be added to the costs and benefits of policies that prescribe specific emission control
measures. This RIA estimates the benefits (and costs) of specific emissions control measures.
We project levels of ozone and PM2.5 to increase and decrease over the U.S. compared to
the baseline. Some portion of the air quality and health benefits from the regulatory control
alternatives would occur in areas not attaining the Ozone or PM2.5 NAAQS. However, we do not
simulate how states would account for this rule when complying with the NAAQS; this affects
the estimated benefits (and costs) of the proposal and more and less stringent alternatives, which
introduces uncertainty in the estimated benefits (and costs).
5B. 1.1 Health Impact Assessment for Ozone and PM2.5
We estimate the quantity and economic value of air pollution-related effects using a
"damage-function" approach. This approach quantifies counts of air pollution-attributable cases
of adverse health outcomes and assigns dollar values to those counts, while assuming that each
outcome is independent of one another. We construct this damage function by adapting primary
research— specifically, air pollution epidemiology studies and economic value studies—from
similar contexts. This approach is sometimes referred to as "benefits transfer." Below we
describe the procedure we follow for: (1) selecting air pollution health endpoints to quantify; (2)
calculating counts of air pollution effects using a health impact function; (3) specifying the
health impact function with concentration-response parameters drawn from the epidemiological
literature.
5B. 1.1.1 Selecting Air Pollution Health Endpoints to Quantify
As a first step in quantifying ozone and PM2.5-related human health impacts, the Agency
consults the Integrated Science Assessment for Ozone and Related Photochemical Oxidants
(Ozone ISA) (U.S. EPA 2013a) and the Integrated Science Assessment for Particulate Matter
(PM ISA) (U.S. EPA 2009). These two documents synthesize the toxicological, clinical and
5B-3
-------
epidemiological evidence to determine whether each pollutant is causally related to an array of
adverse human health outcomes associated with either acute (i.e., hours or days-long) or chronic
(i.e. years-long) exposure; for each outcome, the ISA reports this relationship to be causal, likely
to be causal, suggestive of a causal relationship, inadequate to infer a causal relationship or not
likely to be a causal relationship. In brief, the ISA for ozone found acute exposure to ozone to be
causally related to respiratory effects, a likely-to-be-causal relationship with cardiovascular
effects and total mortality and a suggestive relationship for central nervous system effects.
Among chronic effects, the ISA reported a likely-to-be-causal relationship for respiratory
outcomes and respiratory mortality, and a suggestive relationship for cardiovascular effects,
reproductive and developmental effects, central nervous system effects, and total mortality. The
Agency estimates the incidence of air pollution effects for those health endpoints above where
the ISA classified as either causal or likely-to-be-causal. The PM ISA found acute exposure to
PM2.5 to be causally related to cardiovascular effects and mortality (i.e., premature death), and
respiratory effects as likely-to-be-causally related. The ISA identified cardiovascular effects and
total mortality as being causally related to long-term exposure to PM2.5 and respiratory effects as
likely-to-be-causal; and the evidence was suggestive of a causal relationship for reproductive and
developmental effects as well as cancer, mutagenicity and genotoxicity. Table ES-2 reports the
effects we quantified and those we did not quantify in this RIA.3 The list of benefit categories not
quantified is not exhaustive. And, among the effects quantified, it might not have been possible
to quantify completely either the full range of human health impacts or economic values. The
table below omits health effects associated with direct PM2.5, and NO2, and any welfare effects
such as acidification and nutrient enrichment; these effects are described in the Ozone and PM
NAAQS RIA (U.S. EPA 2015e, 2012) and summarized later in this appendix.
Consistent with economic theory, the willingness to pay (WTP) for reductions in
exposure to environmental hazard will depend on the expected impact of those reductions on
human health and welfare. All else equal, the WTP will be higher when there is strong evidence
of a causal relationship between exposure to the contaminant and changes in the endpoint of
interest. When there is some evidence of a relationship, but that evidence is insufficient to
definitively determine a causal relationship, individuals are still expected to have a positive WTP
3 Note that, as discussed in Chapter 5, EPA has not updated the endpoints in this table to reflect the evidence
reported in the most recent PM or Ozone IS As (U.S. EPA 2019a, 2020b).
5B-4
-------
for a reduction in exposure to that environmental hazard. However, the WTP for reductions in
exposure would be less than the case where the relationship can be determined to be causal.
Conversely, valuing expected changes in the risk of an endpoint as if the relationship was known
to be causal would overestimate the benefits of pollution reductions if there is uncertainty about
whether the relationship is indeed causal. EPA currently lacks a robust methodology to adjust
WTP estimates when the evidence is insufficient to conclude a causal relationship and therefore,
endpoints for which the association cannot be determined to be causal or likely causal are not
currently quantified or monetized.
5B-5
-------
Table 5B-2. Health Effects of Ambient Ozone and PM2.5
Category
Effect
Effect
Quantified
Effect
Monetized
More
Information
Premature mortality
from exposure to
Adult premature mortality based on cohort study
estimates and expert elicitation estimates (age >25
or age >30)
~
~
PM ISA
PM2.5
Infant mortality (age <1)
~
~
PM ISA
Non-fatal heart attacks (age >18)
~
~
PM ISA
Hospital admissions—respiratory (all ages)
~
~
PM ISA
Hospital admissions—cardiovascular (age >20)
~
~
PM ISA
Emergency room visits for asthma (all ages)
~
~
PM ISA
Acute bronchitis (age 8-12)
~
~
PM ISA
Lower respiratory symptoms (age 7-14)
~
~
PM ISA
Upper respiratory symptoms (asthmatics age 9-11)
~
~
PM ISA
Exacerbated asthma (asthmatics age 6-18)
~
~
PM ISA
Lost work days (age 18-65)
~
~
PM ISA
Morbidity from
Minor restricted-activity days (age 18-65)
~
~
PM ISA
exposure to PM2 5
Chronic Bronchitis (age >26)
—
—
PM ISA1
Emergency room visits for cardiovascular effects
(all ages)
—
—
PM ISA1
Strokes and cerebrovascular disease (age 50-79)
—
—
PM ISA1
Other cardiovascular effects (e.g., other ages)
—
—
PM ISA2
Other respiratory effects (e.g., pulmonary function,
non-asthma ER visits, non-bronchitis chronic
PM ISA2
diseases, other ages and populations)
Reproductive and developmental effects (e.g., low
birth weight, pre-term births, etc.)
—
—
PM ISA2'3
Cancer, mutagenicity, and genotoxicity effects
—
—
PM ISA2'3
Mortality from
exposure to ozone
Premature mortality based on sliori-ierm study
estimates (all ages)
Premature mortality based on long-term study
estimates (age 30-99)
~
~
~
~
Ozone ISA
Ozone ISA1
Hospital admissions—respiratory causes (age > 65)
~
~
Ozone ISA
Emergency department visits for asthma (all ages)
~
~
O/.onc ISA
Exacerbated asthma (asthmatics age 6-18)
~
~
O/.onc ISA
Minor restricted-activity days (age 18-65)
~
~
Ozone ISA
Morbidity from
School absence days (age 5-17)
~
~
Ozone ISA
exposure to ozone
Decreased outdoor worker productivity (age 18-65)
—
—
Ozone ISA1
Other respiratory effects (e.g., premature aging of
lungs)
—
—
Ozone ISA2
Cardiovascular and nervous system effects
—
—
Ozone ISA2
Reproductive and developmental effects
—
—
Ozone ISA2,3
1 Not quantified due to data and resource limitations for this analysis.
2 Not quantified because we do not have sufficient confidence in available data or methods.
3 Not quantified because current evidence is only suggestive of causality or there are other significant concerns over the strength
of the association.
5B. 1.1.2 Calculating Counts of Air Pollution Effects Using the Health Impact Function
We use BenMAP-CE to quantify counts of premature deaths and illnesses attributable to
photochemical modeled changes in summer season average ozone concentrations for the year
5B-6
-------
2021, and summer season average ozone concentrations and annual mean PM2.5 for the year 2025
using a health impact function. A health impact function combines information regarding the:
concentration-response relationship between air quality changes and the risk of a given adverse
outcome; population exposed to the air quality change; baseline rate of death or disease in that
population; and, air pollution concentration to which the population is exposed.
The following provides an example of a health impact function, in this case for PM2.5
mortality risk. We estimate counts of PIVfo.s-related total deaths (yij) during each year i (i=2025)
among adults aged 30 and older (a) in each county in the contiguous U.S. j (j=l,.. ,,J where J is
the total number of counties) as
yij 2a yija
yija = moija x(ep ACij-l) x Pija, Eq[l]
where moija is the baseline all-cause mortality rate for adults aged a=30-99 in county j in year i
stratified in 10-year age groups, P is the risk coefficient for all-cause mortality for adults
associated with annual average PM2.5 exposure, Cij is the annual mean PM2.5 concentration in
county j in year i, and Pija is the number of county adult residents aged a=30-99 in county j in
year i stratified into 5-year age groups.4
The BenMAP-CE tool is pre-loaded with projected population from the Woods & Poole
company; cause-specific and age-stratified death rates from the Centers for Disease Control and
Prevention, projected to future years; recent-year baseline rates of hospital admissions,
emergency department visits and other morbidity outcomes from the Healthcare Cost and
Utilization Program and other sources; concentration-response parameters from the published
epidemiologic literature cited in the Integrated Science Assessments for fine particles and
ground-level ozone; and, cost of illness and willingness to pay economic unit values for each
endpoint. Ozone and PM2.5 concentrations are taken from the air pollution spatial surfaces
described in Chapter 3.
4 In this illustrative example, the air quality is resolved at the county level. For this RIA, we simulate air quality
concentrations at 12km by 12km grids. The BenMAP-CE tool assigns the rates of baseline death and disease stored
at the county level to the 12km by 12km grid cells using an area-weighted algorithm. This approach is described in
greater detail in the appendices to the BenMAP-CE user manual.
5B-7
-------
This health impact assessment quantifies outcomes using a suite of concentration-
response parameters described in the PM NAAQS RIA (U.S. EPA 2012), Ozone NAAQS RIA
(U.S. EPA 2015) and the user manual for the BenMAP-CE program (U.S. EPA 2018). These
documents describe in detail our rationale for selecting air pollution-related health endpoints, the
source of the epidemiological evidence, the specific concentration-response parameters applied,
and our approach for pooling evidence across epidemiological studies. Given both the severity of
air pollution-related mortality and its large economic value, below we describe the source of the
concentration-response parameters for this endpoint.
5B. 1.1.3 Quantifying Cases of Ozone-A ttributable Premature Death
In 2008, the National Academies of Science (NRC 2008) issued a series of
recommendations to EPA regarding the procedure for quantifying and valuing ozone-related
mortality due to short-term exposures. Chief among these was that"... short-term exposure to
ambient ozone is likely to contribute to premature deaths" and the committee recommended that
"ozone-related mortality be included in future estimates of the health benefits of reducing ozone
exposures..The NAS also recommended that".. .the greatest emphasis be placed on the
multicity and [National Mortality and Morbidity Air Pollution Studies (NMMAPS)] ... studies
without exclusion of the meta-analyses" (NRC 2008). Prior to the 2015 Ozone NAAQS RIA, the
Agency estimated ozone-attributable premature deaths using an NMMAPS-based analysis of
total mortality (Bell et al. 2004), two multi-city studies of cardiopulmonary and total mortality
(Huang et al. 2004; Schwartz 2005) and effect estimates from three meta-analyses of non-
accidental mortality (Bell et al. 2005; Ito et al. 2005; Levy et al. 2005). Beginning with the 2015
Ozone NAAQS RIA, the Agency began quantifying ozone-attributable premature deaths using
two newer multi-city studies of non-accidental mortality (Smith et al. 2009; Zanobetti and
Schwartz 2008) and one long-term cohort study of respiratory mortality (Jerrett et al. 2009). We
report the ozone-attributable deaths in this RIA as a range reflecting the concentration-response
parameters from Smith et al. (2009) on the low end to Jerrett et al. (2009) on the high end.5
5 See section 6.6 of the RIA for the Ozone NAAQS (U.S. EPA 2015e) for further details.
5B-8
-------
5B. 1.1.4 Quantifying Cases of PM2.s-Attributable Premature Death
For adult PM-related mortality, we use the effect coefficients from two epidemiology
studies examining two large population cohorts: the American Cancer Society cohort (Krewski et
al. 2009) and the Harvard Six Cities cohort (Lepeule et al. 2012). The Integrated Science
Assessment for Particulate Matter (PM ISA) (U.S. EPA 2009) concluded that the analyses of the
ACS and Six Cities cohorts provide the strongest evidence of an association between long-term
PM2.5 exposure and premature mortality with support from additional cohort studies. The SAB's
Health Effects Subcommittee (SAB-HES) also supported using effect estimates from these two
analyses to estimate the benefits of PM reductions (U.S. EPA-SAB 2010). There are distinct
attributes of both the ACS and Six Cities cohort studies that make them well-suited to being used
in a PM benefits assessment and so here we present PM2.5 related effects derived using relative
risk estimates from both cohorts.
The PM ISA, which was twice reviewed by the Clean Air Scientific Advisory Committee
of EPA's Science Advisory Board (SAB-CASAC) (EPA-SAB 2009a, 2009b), concluded that
there is a causal relationship between mortality and both long-term and short-term exposure to
PM2.5 based on the entire body of scientific evidence. The PM ISA also concluded that the
scientific literature supports the use of a no-threshold log-linear model to portray the PM-
mortality concentration-response relationship while recognizing potential uncertainty about the
exact shape of the concentration-response function. The PM ISA, which informed the setting of
the 2012 PM NAAQS, reviewed available studies that examined the potential for a population-
level threshold to exist in the concentration-response relationship. Based on such studies, the ISA
concluded that the evidence supports the use of a "no-threshold" model and that "little evidence
was observed to suggest that a threshold exists" (U.S. EPA 2009) (pp. 2-25 to 2-26). Consistent
with this evidence, the Agency historically has estimated health impacts above and below the
prevailing NAAQS (U.S. EPA 2010c, 2010d, 2011c, 201 Id, 2012, 2013b, 2014a, 2014b, 2014c,
2015a, 2015b, 2015c, 2015d, 2015e, 2016b).
Following this approach, we report the estimated PM2.5-related benefits (in terms of both
health impacts and monetized values) calculated using a log-linear concentration-response
function that quantifies risk from the full range of simulated PM2.5 exposures (NRC 2002; U.S.
EPA 2009). When setting the 2012 PM NAAQS, the Administrator also acknowledged greater
5B-9
-------
uncertainty in specifying the "magnitude and significance" of PM-related health risks at PM
concentrations below the NAAQS. As noted in the preamble to the 2012 PM NAAQS final rule,
"EPA conclude[d] that it [was] not appropriate to place as much confidence in the magnitude and
significance of the associations over the lower percentiles of the distribution in each study as at
and around the long-term mean concentration." (78 FR 3154, 15 January 2013). The preamble
separately noted that "[a]s both the EPA and CASAC recognize, in the absence of a discernible
threshold, health effects may occur over the full range of concentrations observed in the
epidemiological studies." (78 FR 3149, 15 January 2013). In general, we are more confident in
the size of the risks we estimate from simulated PM2.5 concentrations that coincide with the bulk
of the observed PM concentrations in the epidemiological studies that are used to estimate the
benefits. Likewise, we are less confident in the risk we estimate from simulated PM2.5
concentrations that fall below the bulk of the observed data in these studies. To give insight to
the level of uncertainty in the estimated PM2.5 mortality benefits at lower ambient concentrations,
we report the PM benefits according to alternative concentration cut points. Below we further
describe our rationale for selecting these cut points. In addition to adult mortality discussed
above, we use effect coefficients from a multi-city study to estimate PM-related infant mortality
(Woodruff et al. 1997).
5B.1.2 Economic Valuation Methodology for Health Benefits
We next quantify the economic value of the ozone and PM2.5-related deaths and illnesses
estimated above. Changes in ambient concentrations of air pollution generally yield small
changes in the risk of future adverse health effects for a large number of people. Therefore, the
appropriate economic measure is WTP for changes in risk of a health effect. For some health
effects, such as hospital admissions, WTP estimates are not generally available, so we use the
cost of treating or mitigating the effect. These cost-of-illness (COI) estimates generally (although
not necessarily in every case) understate the true value of reductions in risk of a health effect.
They tend to reflect the direct expenditures related to treatment but not the value of avoided pain
and suffering from the health effect. The unit values applied in this analysis are provided in
Table 5-9 of the PM NAAQS RIA for each health endpoint (U.S. EPA 2012).
The value of avoided premature deaths account for over 95 percent of monetized ozone-
related benefits and over 98 percent of monetized PM2.5-related benefits. The economics
5B-10
-------
literature concerning the appropriate method for valuing reductions in premature mortality risk is
still developing. The value for the projected reduction in the risk of premature mortality is the
subject of continuing discussion within the economics and public policy analysis community.
Following the advice of the SAB's Environmental Economics Advisory Committee (SAB-
EEAC), EPA currently uses the value of statistical life (VSL) approach in calculating estimates
of mortality benefits, because we believe this calculation provides the most reasonable single
estimate of an individual's willingness to trade off money for changes in the risk of death (U.S.
EPA-SAB 2000). The VSL approach is a summary measure for the value of small changes in the
risk of death experienced by a large number of people.
EPA continues work to update its guidance on valuing mortality risk reductions, and the
Agency consulted several times with the SAB-EEAC on this issue. Until updated guidance is
available, the Agency determined that a single, peer-reviewed estimate applied consistently, best
reflects the SAB-EEAC advice it has received. Therefore, EPA applies the VSL that was vetted
and endorsed by the SAB in the Guidelines for Preparing Economic Analyses (U.S. EPA 2016a)
while the Agency continues its efforts to update its guidance on this issue. This approach
calculates a mean value across VSL estimates derived from 26 labor market and contingent
valuation studies published between 1974 and 1991. The mean VSL across these studies is $6.3
million (2000$). We then adjust this VSL to account for the currency year and to account for
income growth from 1990 to the analysis year. Specifically, the VSLs applied in this analysis in
2016$ after adjusting for income growth is $10.5 million for 2021 and $10.7 million for 2025.
The Agency is committed to using scientifically sound, appropriately reviewed evidence
in valuing changes in the risk of premature death and continues to engage with the SAB to
identify scientifically sound approaches to update its mortality risk valuation estimates. Most
recently, the Agency proposed new meta-analytic approaches for updating its estimates (U.S.
EPA 2010d), which were subsequently reviewed by the SAB-EEAC. EPA is taking the SAB's
formal recommendations under advisement (U.S. EPA 2017).
In valuing PM2.5-related premature mortality, we discount the value of premature
mortality occurring in future years using rates of 3 percent and 7 percent (U.S. Office of
Management and Budget 2003). We assume that there is a multi-year "cessation" lag between
5B-11
-------
changes in PM exposures and the total realization of changes in health effects. Although the
structure of the lag is uncertain, EPA follows the advice of the SAB-HES to use a segmented lag
structure that assumes 30 percent of premature deaths are reduced in the first year, 50 percent
over years 2 to 5, and 20 percent over the years 6 to 20 after the reduction in PM2.5 (U.S. EPA-
SAB 2004). Changes in the cessation lag assumptions do not change the total number of
estimated deaths but rather the timing of those deaths.
Because short-term ozone-related premature mortality occurs within the analysis year, the
estimated ozone-related benefits are identical for all discount rates. When valuing changes in
ozone-attributable deaths using the Jerrett et al. (2009) study, we follow advice provided by the
Health Effects Subcommittee of the SAB, which found that".. .there is no evidence in the
literature to support a different cessation lag between ozone and particulate matter. The HES
therefore recommends using the same cessation lag structure and assumptions as for particulate
matter when utilizing cohort mortality evidence for ozone" (U.S. EPA-SAB 2010).
5B. 1.3 Characterizing Uncertainty in the Estimated Benefits
This analysis includes many data sources as inputs that are each subject to uncertainty.
Input parameters include projected emission inventories, projected emissions from the electricity
planning model, air quality data from models (with their associated parameters and inputs),
population data, population estimates, health effect estimates from epidemiology studies,
economic data, and assumptions regarding the future state of the world (i.e., regulations,
technology, and human behavior). When compounded, even small uncertainties can greatly
influence the size of the total quantified benefits.
Our estimate of the total monetized PM2.5 and ozone-attributable benefits is based on
EPA's interpretation of the best available scientific literature and methods and supported by the
SAB-HES and the National Academies of Science (NRC 2002). Below are key assumptions
underlying the estimates for PM2.5-related premature mortality, followed by key uncertainties
associated with estimating the number and value of ozone-related premature deaths.
We assume that all fine particles, regardless of their chemical composition, are equally
potent in causing premature mortality. This is an important assumption, the PM ISA concluded
that "many constituents of PM2.5 can be linked with multiple health effects, and the evidence is
5B-12
-------
not yet sufficient to allow differentiation of those constituents or sources that are more closely
related to specific outcomes" (U.S. EPA 2009).
As noted above, we assume that the health impact function for fine particles is log-linear
without a threshold. Thus, the estimates include health benefits from reducing fine particles in
areas with different concentrations of PM2.5, including both areas that do not meet the fine
particle standard and those areas that are in attainment and reflect the full distribution of PM2.5
air quality simulated above.
Also, as noted above, we assume that there is a "cessation" lag between the change in PM
exposures and the total realization of changes in mortality effects. Specifically, we assume that
some of the incidences of premature mortality related to PM2.5 exposures occur in a distributed
fashion over the 20 years following exposure based on the advice of the SAB-HES (U.S. EPA-
SAB 2004), which affects the valuation of mortality benefits at different discount rates. The
above assumptions are subject to uncertainty.
In general, we are more confident in the magnitude of the risks we estimate from
simulated PM2.5 concentrations that coincide with the bulk of the observed PM concentrations in
the epidemiological studies that are used to estimate the benefits. Likewise, we are less confident
in the risk we estimate from simulated PM2.5 concentrations that fall below the bulk of the
observed data in these studies. There are uncertainties inherent in identifying any particular point
at which our confidence in reported associations decreases appreciably, and the scientific
evidence provides no clear dividing line. This relationship between the air quality data and our
confidence in the estimated risk is represented below in Figure 5B-2.
5B-13
-------
Less confident
More
confident
<
Below LML of PM25 data
in epidemiology study
(extrapolation)
I standard deviation below the
mean PM25 observed in
epidemiology study
Mean of PM2.5 data in
epidemiology study
Figure 5B-2. Stylized Relationship between the PM2.5 Concentrations Considered in
Epidemiology Studies and our Confidence in the Estimated PM-related
Premature Deaths
In this analysis, we build upon the concentration benchmark approach (also referred to as
the Lowest Measured Level (LML) analysis) that has been featured in recent RIAs and EPA's
Policy Assessment for Particulate Matter (U.S. EPA 2011) by reporting the estimated PM-
related deaths according to alternative concentration cut points.
Concentration benchmark analyses allow readers to determine the portion of population
exposed to annual mean PM2.5 levels at or above different concentrations, which provides some
insight into the level of uncertainty in the estimated PM2.5 mortality benefits. EPA does not view
these concentration benchmarks as concentration thresholds below which we would not quantify
health benefits of air quality improvements.6 Rather, the PM2.5-attributable benefits estimates
reported in this RIA are the most appropriate estimates because they reflect the full range of air
quality concentrations associated with the emission reduction strategies being evaluated in this
proposal. The PM ISA concluded that the scientific evidence collectively is sufficient to
conclude that there is a causal relationship between long-term PM2.5 exposures and mortality and
that overall the studies support the use of a no-threshold log-linear model to estimate mortality
attributed to long-term PM2.5 exposure (U.S. EPA 2009). Furthermore, while the tables below
show the benefits above the LML only, it is the benefits below those cut-points that are more
6 For a summary of the scientific review statements regarding the lack of a threshold in the PM2 5-mortality
relationship, see the TSD entitled Summary of Expert Opinions on the Existence of a Threshold in the
Concentration-Response Function for PAF.s-related Mortality (U.S. EPA, 2010b).
5B-14
-------
uncertain, and may be greater or smaller in magnitude when estimating PM2.5 benefits across the
full range of projected exposures.
Figure 5B-3 reports the percentage of the population, and number of PM-related deaths,
both above and below concentration benchmarks in the proposed policy modeling for the year
2025. The figure identifies the LML for each of the major cohort studies and the annual mean
PM2.5 NAAQS of 12 |ig/m3. For Krewski, the LML is 5.8 |ig/m3 and for Lepeule et al., the LML
is 8 |ig/m3. These results are sensitive to the annual mean PM2.5 concentration the air quality
model predicted in each 12km by 12km grid cell. The air quality modeling predicts PM2.5
concentrations to be at or below the PM2.5 NAAQS (12 |ig/m3) in nearly all locations. The
photochemical modeling we employ accounts for the suite of local, state and federal policies
expected to reduce PM2.5 and PM2.5 precursor emissions in future years, such that we project a
very small number of locations exceeding the annual standard. After presenting the full suite of
results below we stratify these estimated PM2.5 mortality deaths according to the concentration at
which they occurred: below the LML, between the LML and the NAAQS, and above the
NAAQS in future years across different policy scenarios. The results above should be viewed in
the context of the air quality modeling technique we used to estimate PM2.5 concentrations. We
are more confident in our ability to use the air quality modeling technique described above to
estimate changes in annual mean PM2.5 concentrations than we are in our ability to estimate
absolute PM2.5 concentrations.
5B-15
-------
LML LML
Krewski (2009) Lepeule (2012)
2012
PM NAAQS
0.4-
0.3-
C
CD
Q
£= 0.2-
la
cc
_Q
o
Mortality
Population Exposed
0.0-1
PM2.5 (ng/m )
Figure 5B-3. Estimated Percentage of PlVh.s-Related Deaths and Number of Individuals
Exposed by Annual Mean PM2.5 Level in 2025
The estimated number and value of avoided ozone-attributable deaths are also subject to
uncertainty. When estimating the economic value of avoided premature mortality from long-term
exposure to ozone, we use a 20-year segment lag (as used for PM2.5) as there is no alternative
empirical estimate of the cessation lag for long-term exposure to ozone. The 20-year segmented
lag accounts for the onset of cardiovascular related mortality, an outcome which is not relevant
to the long-term respiratory mortality estimated here. We utilize a log-linear impact function
without a threshold in modeling short-term ozone-related mortality. However, we acknowledge
reduced confidence in specifying the nature of the C-R function in the range of <20ppb and
below (ozone ISA, section 2.5.4.4). Thus, the estimates include health benefits from reducing
ozone in areas with varied concentrations of ozone down to the lowest modeled concentrations.
5B. 1.4 Estimated Number and Economic Value of Health Benefits
Below we report the estimated number of reduced premature deaths and illnesses in each
year relative to the baseline along with the 95% confidence interval (Table 5B-3and Table 5B-4).
The number of reduced estimated deaths and illnesses from the proposal and more and less
5B-16
-------
stringent alternatives are calculated from the sum of individual reduced mortality and illness risk
across the population. Table 5B-5 and Table 5B-6 report the estimated economic value of
avoided premature deaths and illness in each year relative to the baseline along with the 95%
confidence interval. The tables below are followed by the estimated number of avoided PM2.5-
related premature deaths calculated using different approaches to help the reader determine the
fraction of PM2.5 attributable deaths occuring at lower ambient concentrations. We summarize
the dollar value of these impacts for the proposal and more and less stringent alternatives across
all PM2.5 and ozone-related premature deaths and illnesses, using alternative approaches to
representing and quantifying PM mortality risk effects (and Table 5B-9). The alternative
approaches to quantifying and presenting mortality risk effects include both different means for
quantifying expected impacts using concentration-response functions over the entire domain of
exposure (i.e., the no-threshold model) along with different means of presenting impacts by
limiting consideration to only those impacts at exposures above the LML or above the NAAQS
(Table 5B-10).7
7 EPA continues to refine its approach for estimating and reporting PM-related effects at lower concentrations,
particularly at levels below those considered by the long-term exposure epidemiology studies used here to quantify
PM-related premature deaths. The Agency acknowledges the additional uncertainty associated with effects estimated
at these lower levels (particularly below the LML of the long-term exposure mortality studies) and seeks to develop
quantitative approaches for reflecting this uncertainty in the estimated PM benefits.
5B-17
-------
Table 5B-3. Estimated Avoided Ozone-Related Premature Deaths and Illnesses for the
Proposal and More and Less Stringent Alternatives for 2021 (95% Confidence
Interval) a,b
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided premature death among adults
Smith et al. (2009)
30
(-2.8 to 57)
30
(-2.8 to 57)
2.7
(-0.25 to 5)
Jerrett et al. (2009)
110
(36 to 180)
110
(36 to 180)
9.1
(3.1 to 15)
All other morbidity effects
Hospital admissions—respiratory
47
(-11 to 100)
47
(-11 to 100)
4.1
(-0.95 to 9)
ED visits for asthma
200
200
17
(19 to 490)
(19 to 490)
(1.7 to 42)
Exacerbated asthma
69,000
69,000
6,100
(-59,000 to 170,000)
(-59,000 to 170,000)
(-5,200 to 15,000)
Minor restricted-activity days
140,000
(59,000 to 230,000)
140,000
(59,000 to 230,000)
13,000
(5,200 to 20,000)
School absence days
43,000
(15,000 to 97,000)
43,000
(15,000 to 97,000)
3,800
(1,400 to 8,600)
a Values rounded to two significant figures.
b We estimated changes in annual mean PM25 and PM25 -related benefits in 2025, 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 beginning in May of 2021, 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 2025, 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.
5B-18
-------
Table 5B-4. Estimated Avoided PM2.5 and Ozone-Related Premature Deaths and Illnesses
for the Proposal and More and Less Stringent Alternatives for 2025 (95%
Confidence Interval) a
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided premature death among adults
0
(0 to 0)
S Krewski et al. (2009)
5
7
(4.8 to 9.3)
7
(4.8 to 9.3)
Ph
Lepeule et al. (2012)
16
(7.9 to 24)
16
(7.9 to 24)
0
(0 to 0)
S Smith et al. (2009)
0
38
(-3.5 to 70)
61
(-5.7 to 110)
2.6
(-0.25 to 4.9)
^ Jerrett et al. (2009)
130
220
9.1
(45 to 220)
(73 to 360)
(3.1 to 15)
PM2.5- related non-fatal heart attacks among adults
7.2
(1.8 to 13)
0
(0 to 0)
Peters et al. (2001)
7.2
(1.8 to 13)
Pooled estimate
0.77
0.77
0
(0.29 to 2.1)
(0.29 to 2.1)
(0 to 0)
All other morbidity effects
Hospital admissions—
cardiovascular (PM2 5)
1.8
(0.78 to 3.3)
1.8
(0.78 to 3.3)
0
(0 to 0)
Hospital admissions—
65
110
4.4
respiratory (PM2 5 & O3)
(-16 to 150)
(-25 to 240)
(-1 to 9.7)
ED visits for asthma
250
400
17
(PM2 5 & 03)
(22 to 600)
(37 to 960)
(1.6 to 41)
Exacerbated asthma
85,000
140,000
6,000
(PM2 5 & 03)
(-73,000 to 210,000)
(-120,000 to 340,000)
(-5,100 to 14,000)
Minor restricted-activity
170,000
280,000
12,000
days (PM25 & 03)
(74,000 to 270,000)
(120,000 to 440,000)
(4,900 to 19,000)
Acute bronchitis
8.4
8.4
0
(PM25)
(-2 to 19)
(-2 to 19)
(0 to 0)
Upper resp. symptoms
(PM25)
150
(27 to 270)
150
(27 to 270)
0
(0 to 0)
Lower resp. symptoms
(PM25)
110
(40 to 170)
110
(40 to 170)
0
(0 to 0)
Lost work days
740
740
0
(PM25)
(630 to 850)
(630 to 850)
(0 to 0)
School absence days
54,000
89,000
3,800
(03)
(19,000 to 120,000)
(31,000 to 200,000)
(1,300 to 8,400)
a Values rounded to two significant figures.
5B-19
-------
Table 5B-5. Estimated Value of Avoided Ozone-Related Premature Deaths and Illnesses for
the Proposal and More and Less Stringent Alternatives for 2021 (95%
Confidence Interval; millions of 2016$)a,b
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided premature death among adults
Smith et al. (2009)
$310
($-7.5 to $1,000)
$310
($-7.5 to $1,000)
$28
($-0.66 to $89)
Jerrett et al. (2009)°
$1,000
($81 to $3,000)
$1,000
($81 to $3,000)
$86
($6.9 to $260)
All other morbidity effects
Hospital admissions—respiratory
$1
($-0.24 to $2.3)
$1
($-0.24 to $2.3)
$0.09
($-0.02 to $0.2)
ED visits for asthma
$0.06
($0.01 to $0.15)
$0.06
($0.01 to $0.15)
$0.01
($0.0 to $0.01)
Exacerbated asthma
$3.2
$3.2
$0.28
($-2.7 to $9.5)
($-2.7 to $9.5)
($-0.24 to $0.84)
Minor restricted-activity days
$7.7
($2.9 to $14)
$7.7
($2.9 to $14)
$0.68
($0.25 to $1.3)
School absence days
$3.3
($1.2 to $7.4)
$3.3
($1.2 to $7.4)
$0.29
($0.1 to $0.66)
a Values rounded to two significant figures.
b We estimated changes in annual mean PM25 and PM2.5 -related benefits in 2025, but not 2021. As discussed in
Chapter 4, in 2021, the only control measure expected to be adopted for compliance in the regulatory control
alternatives is optimization of existing SCRs beginning in May of 2021, and this measure will operate only during
the ozone season. As discussed in Chapter 3, NOx reductions in the ozone season provide minimal PM2 5 benefits
since PM2 5 nitrate concentrations, which result from conversion of NOx emissions to nitrate, are minimal during the
warmer temperatures during the ozone season. Conversely, the conversion of nitrates to PM2 5 is much greater in
cooler (non-ozone season) months, and thus it becomes worthwhile to estimate PM2 5 benefits from NOx reductions
in those months. In 2025, the presence of additional control measures that operate year round and other changes in
market conditions as a result of the rule lead to notable NOx reductions in the winter months.
c Discounted at 3%. Summary tables below report mortality benefits discounted at 7%.
5B-20
-------
Table 5B-6. Estimated Value of Avoided PM2.5 and Ozone-Related Premature Deaths and
Illnesses for the Proposal and More and Less Stringent Alternatives for 2025
(95% Confidence Interval; millions of 2016$)"
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided premature death among adults
S Krewski et al. (2009)b
5
$67
($6.3 to $180)
$67
($6.3 to $180)
$0
($0 to $0)
Ph
Lepeule et al. (2012)b
$150
($14lo $430)
$150
($14 to $430)
$0
($0 to 0)
S Smith et al. (2009)
o
$400
($-9.5 to $1,300)
$640
($-15 to $2,100)
$28
($-0.67 to $89)
^ Jerrett et al. (2009)b
$1,300
($100 to $3,800)
$2,100
($170 to $6,200)
$88
($7.1 to $260)
PM2.5- related non-fatal heart attacks among adults
Peters et al. (200l)b
$1
($0.16 to $2.6)
$1
($0.16 to $2.6)
$0
($0 to $0)
Pooled estimate13
$0.11
($0,023 to $0.39)
$0.11
($0,023 to $0.39)
$0
($0 to $0)
All other morbidity effects
Hospital admissions—
$0,082
$0,082
$0
cardiovascular (PM2 5)
($0,036 to $0.15)
($0,036 to $0.15)
($0 to $0)
Hospital admissions—
$1.5
$2.4
$0,097
respiratory (PM2 5 & O3)
($-0.36 to $3.3)
($-0.57 to $5.3)
($-0,023 to $0.22)
ED visits for asthma
$0,074
$0.12
$0,005
(PM2 5 & 03)
($0.0059 to $0.19)
($0.01 to $0.31)
($0.0005 to $0,013)
Exacerbated asthma
$4
$6.5
$0.28
(PM2 5 & 03)
($-3.4 to $12)
($-5.6 to $19)
($-0.24 to $0.83)
Minor restricted-activity
$9.6
$15
$0.65
days (PM25 & O3)
($3.6 to $18)
($5.8 to $28)
($0.24 to $1.2)
Acute bronchitis
$0.0044
$0.0044
$0
(PM25)
($-0,001 to $0,013)
($-0,001 to $0,013)
($0 to $0)
Upper resp. symptoms
(PM25)
$0.0055
($0,001 to $0,014)
$0.0055
($0,001 to $0,014)
$0
($0 to $0)
Lower resp. symptoms
(PM25)
$0.0025
($0.0008 to $0,005)
$0.0025
($0.0008 to $0,005)
$0
($0 to $0)
Lost work days
$0.13
$0.13
$0
(PM25)
($0.11 to $0.15)
($0.11 to $0.15)
($0 to $0)
School absence days
$4.1
$6.8
$0.29
(03)
($1.5 to $9.2)
($2.4 to $15)
($0.1 to $0.64)
" Values rounded to two significant figures.
b Discounted at 3%. Summary tables below report mortality benefits discounted at 7%.
5B-21
-------
Table 5B-7. Estimated Avoided PM-Related Premature Deaths Using Alternative
Approaches Using Two Approaches to Quantifying Avoided PM-Attributable
Deaths (95% Confidence Interval) in 2025"
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Log-Linear no-threshold model
Krewski et al. (2009)
7
(4.8 to 9.3)
7
(4.8 to 9.3)
0
(0 to 0)
Lepeule et al. (2012)
16
(7.9 to 24)
16
(7.9 to 24)
0
(0 to 0)
Quantifying effect ofPM2.5 above the LML in each study and below the NAAQS
Krewski et al. (2009)
(LML= 5.8 ug/m3)
6.5
(4.4 to 8.6)
6.5
(4.4 to 8.6)
0
(0 to 0)
Lepeule et al. (2012)
(LML=8ug/m3)
4.4
(2.2 to 6.6)
4.4
(2.2 to 6.6)
0
(0 to 0)
Quantifying effect ofPM2.5 above the NAAQS
Krewski et al. (2009)
0.021
(0.014 to 0.028)
0.021
(0.014 to 0.028)
0
(0 to 0)
Lepeule et al. (2012)
0.048
(0.024 to 0.072)
0.048
(0.024 to 0.072)
0
(0 to 0)
a Values rounded to two significant figures.
5B-22
-------
Table 5B-8. Estimated Economic Value of Ozone-Attributable Deaths and Illnesses for the
Proposed Policy Scenarios in 2021 (95% Confidence Interval; millions of
2016$)a,b
Proposal
More Stringent Alternative
Less Stringent Alternative
3% Discount
Rate
$330
$1,000
$330
$1,000
| $29
$87
($-6.5 to
to
($82 to
($-6.5 to to
($82 to
| ($-0.57 to to
($7 to
$1,000)
$3,000)
$1,000)
$3,000)
1 $92)
$260)
7% Discount
Rate
$330
$930
$330
$930
| $29
$79
($-6.5 to
to
($75 to
($-6.5 to to
($75 to
| ($-0.57 to to
($6.4 to
$1,000)
$2,800)
$1,000)
$2,800)
I $92)
$240)
a Values rounded to two significant figures.
b We estimated changes in annual mean PM2 5 and PM2 5 -related benefits in 2025, but not 2021. As discussed in
Chapter 4, in 2021, the only control measure expected to be adopted for compliance in the regulatory control
alternatives is optimization of existing SCRs beginning in May of 2021, and this measure will operate only during
the ozone season. As discussed in Chapter 3, NOx reductions in the ozone season provide minimal PM2 5 benefits
since PM25 nitrate concentrations, which result from conversion of NOx emissions to nitrate, are minimal during the
warmer temperatures during the ozone season. Conversely, the conversion of nitrates to PM2 5 is much greater in
cooler (non-ozone season) months, and thus it becomes worthwhile to estimate PM2 5 benefits from NOx reductions
in those months. In 2025, the presence of additional control measures that operate year round and other changes in
market conditions as a result of the rule lead to notable NOx reductions in the winter months.
5B-23
-------
Table 5B-9. Estimated Economic Value of Avoided Ozone and PlVh.s-Attributable Deaths
and Illnesses for the Proposed Policy Scenario Using Alternative Approaches to
Represent PM2.5 Mortality Risk Effects in 2025 (95% Confidence Interval;
millions of 2016$)a
Proposal
More Stringent Alternative
Less Stringent
Alternative
Ozone
3% Discount Rate
$410
($-8.3 to
$1,300)
to
$1,300
($110 to
$3,900)
$670
($-14 to
$2,100)
to
$2,100
($170 to $
6,300)
$29
($-0.59
to $92)
to
$89
($7.2 to
$260)
7% Discount Rate
$410
($-8.3 to
$1,300)
to
$1,200
($96 to
$3,500)
$670
($-14 to
$2,100)
$1,900
to ($160 to $
5,700)
$29
($-0.59
to $92)
to
$81
($6.6 to
$240)
PMzs
No-threshold
q model
$69
($6.6 to
$190)
to
$150
($14 to
$440)
$69
($6.6 to
$190)
to
$150
($14 to $440)
$0
($0 to
$0)
to
$0
($0 to 0)
g Limited to
g above LML
$44
($4.2 to
$120)
to
$63
($6.1 to
$170)
$44
($4.2 to
$120)
to
$63
($6.1 to $170)
$0
($0 to
$0)
to
$0
($0 to 0)
Effects above
NAAQS
$1.3
($0.37 to
$3)
to
$2.5
($0.53 to
$6)
$1.3
($0.37 to
$3)
to
$2.5
($0.53 to $6)
$0
($0 to
$0)
to
$0
($0 to 0)
No-threshold
& model
$63
($6.1 to
$170)
to
$140
($13 to
$400)
$63
($6.1 to
$170)
to
$140
($13 to $400)
$0
($0 to
$0)
to
$0
($0 to 0)
Oh
s
g Limited to
'd above LML
Q
$41
($3.9 to
$110)
to
$58
($5.6 to
$160)
$41
($3.9 to
$110)
to
$58
($5.6 to $160)
$0
($0 to
$0)
to
$0
($0 to 0)
^ Effects above
NAAQS
$1.3
($0.36 to
$3)
to
$2.4
($0.51 to
$5.9)
$1.3
($0.36 to
$3)
to
$2.4
($0.51 to $5.9)
$0
($0 to
$0)
to
$0
($0 to 0)
1 Values rounded to two significant figures.
5B-24
-------
Table 5B-10. Estimated Percent of Avoided PlVh.s-related Premature Deaths Above and
Below PM2.5 Concentration Cut Points in 2025
Avoided PIVh.s-related premature deaths
reported by air quality cut point
Epidemiological
study
Total
mortality
Above
NAAQS
Below NAAQS and
Above LMLa
Below LMLa
Krewski
Lepeule
0.021
6.5
0.55
Proposal
/
16
(0%)
0.048
(0%)
(92%)
4.4
(27%)
(8%)
12
(72%)
More
Stringent
Alternative
Krewski
Lepeule
7
16
0.021
(0%)
0.048
(0%)
6.5
(92%)
4.4
(27%)
0.55
(8%)
12
(72%)
Less Stringent
Krewski
0
0
0
0
Alternative
Lepeule
0
0
0
0
a The LML of the Krewski study is 5.8 |ig/m3 and 8 |ig/m3 for Lepeule et al study.
The estimated number of deaths above and below the LML varies considerably according
to the epidemiology study used to estimate risk. Thus, for any year analyzed, we estimate a
substantially larger fraction of PM-related deaths above the LML of the Krewski et al. (2009)
study than we do the Lepeule et al. (2012) study as shown in Table 5B-10. Likewise, we estimate
a greater percentage of PM2.5-related deaths below the LML of the Lepeule et al. (2012) study
than we do the Krewski et al. (2009) study. Table 5B-10 also shows we estimate a very small
percentage of PM-related premature deaths occurring above the NAAQS in 2025 using either of
these two studies.
5B.2 References
Amorim MIM, MERGLER D, BAHIA MO, DUBEAU H, MIRANDA D, LEBEL J, et al. 2000.
Cytogenetic damage related to low levels of methyl mercury contamination in the
Brazilian Amazon. An Acad Bras Cienc 72:497-507; doi:10.1590/S0001-
37652000000400004.
ATSDR. 1999. ATSDR - Toxicological Profile: Mercury.
Bell, M.L., A. McDermott, S.L. Zeger, J.M. Sarnet, and F. Dominici. 2004. "Ozone and Short-
Term Mortality in 95 U.S. Urban Communities, 1987-2000." Journal of the American
Medical Association. 292(19):2372-8.
5B-25
-------
Bell, ML., F. Dominici, and J.M. Samet. 2005. "A Meta-Analysis of Time-Series Studies of
Ozone and Mortality with Comparison to the National Morbidity, Mortality, and Air
Pollution Study." Epidemiology. 16(4):436-45.
Fann N, Nolte CG, Dolwick P, Spero TL, Brown AC, Phillips S, et al. 2015. The geographic
distribution and economic value of climate change-related ozone health impacts in the
United States in 2030. J Air Waste Manag Assoc 65;
doi: 10.1080/10962247.2014.996270.
Gwinn, M.R., J. Craig, D.A. Axelrad, R. Cook, C. Dockins, N. Fann, R. Fegley, D.E. Guinnup,
G. Helfand, B. Hubbell, S.L. Mazur, T. Palma, R.L. Smith, J. Vandenberg, and
B. Sonawane. 2011. "Meeting report: Estimating the benefits of reducing hazardous air
pollutants—summary of 2009 workshop and future considerations." Environmental
Health Perspectives. 119(1): 125-30.
Huang Y, Dominici F, Bell M. 2004. Bayesian Hierarchical Distributed Lag Models for Summer
Ozone Exposure and Cardio-Respiratory Mortality. Johns Hopkins Univ Dept Biostat
Work Pap Ser.
IARC. 1994. Mercury Carcinogenicity.
Ito, K., S.F. De Leon, and M. Lippmann. 2005. "Associations Between Ozone and Daily
Mortality: Analysis and Meta-Analysis." Epidemiology. 16(4):446-57.
Jerrett M, Burnett RT, Pope CA, Ito K, Thurston G, Krewski D, et al. 2009. Long-term ozone
exposure and mortality. NEngl J Med 360:1085-95; doi:10.1056/NEJMoa0803894.
Jhun I, Fann N, Zanobetti A, Hubbell B. 2014. Effect modification of ozone-related mortality
risks by temperature in 97 US cities. Environment International. 73:128-34.
Krewski D., M. Jerrett, R.T. Burnett, R. Ma, E. Hughes, Y. Shi, et al. 2009. Extended Follow-Up
and Spatial Analysis of the American Cancer Society Study Linking Particulate Air
Pollution and Mortality. HEI Research Report, 140, Health Effects Institute, Boston, MA.
Lepeule, J., F. Laden, D. Dockery, and J. Schwartz. 2012. "Chronic Exposure to Fine Particles
and Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to
2009." Environmental Health Perspectives. 120(7):965-70.
Levy, J.I., S.M. Chemerynski, and J.A. Sarnat. 2005. "Ozone Exposure and Mortality: An
Empiric Bayes Metaregression Analysis." Epidemiology. 16(4):458-68.
National Research Council (NRC). 2000. Toxicological Effects of Methylmercury: Committee
on the Toxicological Effects of Methylmercury." Board on Environmental Studies and
Toxicology. National Academies Press. Washington, DC.
National Research Council (NRC). 2002. Estimating the Public Health Benefits of Proposed Air
Pollution Regulations. National Academies Press. Washington, DC.
5B-26
-------
National Research Council (NRC). 2008. Estimating Mortality Risk Reduction and Economic
Benefits from Controlling Ozone Air Pollution. National Academies Press. Washington,
DC.
Office of Management and Budget (OMB). 2003. Circular A-4: Regulatory Analysis.
Washington, DC. Available at: < http://www.whitehouse.gov/omb/circulars/a004/a-
4.html>.
Ren, C., G.M. Williams, K. Mengersen, L. Morawska, and S. Tong. 2008a. "Does Temperature
Modify Short-Term Effects of Ozone on Total Mortality in 60 Large Eastern U.S.
Communities? An Assessment Using the NMMAPS Data." Environment International.
34:451-458.
Ren, C., G.M. William, L. Morawska, K. Mengensen, and S. Tong. 2008b. "Ozone Modifies
Associations between Temperature and Cardiovascular Mortality: Analysis of the
NMMAPS Data." Occupational and Environmental Medicine. 65:255-260.
Sisler, J.F. 1996. Spatial and Seasonal Patterns and Long-Term Variability of the Composition of
the Haze in the United States: An analysis of data from the IMPROVE network. CIRA
Report, ISSN 0737-5352-32, Colorado State University.
Schwartz, J. 2005. "How Sensitive is the Association between Ozone and Daily Deaths to
Control for Temperature?" American Journal of Respiratory and Critical Care Medicine.
171(6): 627-31.
Smith RL, Xu B, Switzer P. 2009. Reassessing the relationship between ozone and short-term
mortality in U.S. urban communities. Inhal Toxicol 21 Suppl 2:37-61;
doi: 10.1080/08958370903161612.
U.S. Environmental Protection Agency (U.S. EPA). 2002. Methylmercury (MeHg) CASRN
22967-92-6 | IRIS | US EPA, ORD
U.S. Environmental Protection Agency (U.S. EPA). 2008b. Integrated Science Assessment for
Oxides of Nitrogen and Sulfur-Ecological Criteria National (Final Report). National
Center for Environmental Assessment - RTP Division, Research Triangle Park, NC.
EPA/600/R-08/139. December. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for
Particulate Matter (Final Report). EPA-600-R-08-139F. National Center for
Environmental Assessment - RTP Division, Research Triangle Park, NC. December.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2010a. Integrated Science Assessment for
Carbon Monoxide. National Center for Environmental Assessment, Research Triangle
Park, NC. EPA/600/R-09/019F. January. Available at:
.
5B-27
-------
U.S. Environmental Protection Agency (U.S. EPA). 2010b. Technical Support Document:
Summary of Expert Opinions on the Existence of a Threshold in the Concentration-
Response Function for PM2.5-related Mortality. Research Triangle Park, NC. June.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Regulatory Impact Analysis (RIA)
for Existing Stationary Compression Ignition Engines NESHAP Final Draft.
U.S. Environmental Protection Agency (U.S. EPA). 2010d. Regulatory Impact Analysis for the
Proposed Federal Transport Rule.
U.S. Environmental Protection Agency (U.S. EPA). 201 la. Policy Assessment for the Review of
the Particulate Matter National Ambient Air Quality Standards. EPA-452/D-11-003.
Office of Air Quality Planning and Standards, Health and Environmental Impacts
Division. April. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 201 lb. The Benefits and Costs of the Clean
Air Act from 1990 to 2020. Office of Air and Radiation, Washington, DC. March.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 201 lc. Regulatory Impact Analysis for the
Federal Implementation Plans to Reduce Interstate Transport of Fine Particulate Matter
and Ozone in 27 States; Correction of SIP Approvals for 22 States.
U.S. Environmental Protection Agency (U.S. EPA). 201 Id. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards.
U.S. Environmental Protection Agency (U.S. EPA). 2012. Regulatory Impact Analysis for the
Final Revisions to the National Ambient Air Quality Standards for Particulate Matter.
EPA-452/R-12-003. Office of Air Quality Planning and Standards, Health and
Environmental Impacts Division, Research Triangle Park, NC. Available at: <
http://www.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf>.
U.S. Environmental Protection Agency (U.S. EPA). 2013a. Integrated Science Assessment of
Ozone and Related Photochemical Oxidants (Final Report). EPA/600/R-10/076F.
National Center for Environmental Assessment - RTP Division, Research Triangle Park.
Available at: .
U.S. Environmental Protection Agency (U.S. EPA). 2013b. Regulatory Impact Analysis for the
Final Revisions to the National Ambient Air Quality Standards for Particulate Matter.
U.S. Environmental Protection Agency (U.S. EPA). 2014a. Regulatory Impact Analysis (RIA)
for Proposed Residential Wood Heaters NSPS Revision.
U.S. Environmental Protection Agency (U.S. EPA). 2014b. Regulatory Impact Analysis for the
Proposed Carbon Pollution Guidelines for Existing Power Plants and Emission Standards
for Modified and Reconstructed Power Plants.
5B-28
-------
U.S. Environmental Protection Agency (U.S. EPA). 2014c. Regulatory Impact Analysis of the
Proposed Revisions to the National Ambient Air Quality Standards for Ground-Level
Ozone.
U.S. Environmental Protection Agency (U.S. EPA). 2015a. Regulatory Impact Analysis for
Residential Wood Heaters NSPS Revision: Final Report.
U.S. Environmental Protection Agency (U.S. EPA). 2015b. Regulatory Impact Analysis for the
Clean Power Plan Final Rule.
U.S. Environmental Protection Agency (U.S. EPA). 2015c. Regulatory Impact Analysis for the
Proposed Cross-State Air Pollution Rule (CSAPR) Update for the 2008 Ozone National
Ambient Air Quality Standards (NAAQS).
U.S. Environmental Protection Agency (U.S. EPA). 2015d. Regulatory Impact Analysis for the
Proposed Federal Plan Requirements for Greenhouse Gas Emissions from Electric Utility
Generating Units Constructed on or Before January 8, 2014; Model Trading Rules;
Amendments to Framework Regulations. 4-52
U.S. Environmental Protection Agency (U.S. EPA). 2015e. Regulatory Impact Analysis of the
Final Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone,
EPA-452/R-15-07. EPA-452/R-12-003. Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division, Research Triangle Park, NC.
(https://www.epa.gOv/sites/production/files/2016-02/documents/20151001ria.pdf).
U.S. Environmental Protection Agency (U.S. EPA). 2016a. Guidelines for Preparing Economic
Analyses.
U.S. Environmental Protection Agency (U.S. EPA). 2016b. Regulatory Impact Analysis of the
Cross-State Air Pollution Rule (CSAPR) Update for the 2008 National Ambient Air
Quality Standards for Ground-Level Ozone.
U.S. Environmental Protection Agency (U.S. EPA). 2016c. Integrated Science Assessment for
Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available at: <
https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=310879>.
U.S. Environmental Protection Agency (U.S. EPA). 2017. Integrated Science Assessment for
Sulfur Oxides—Health Criteria (Final Report). National Center for Environmental
Assessment - RTP Division, Research Triangle Park, NC. September. Available at:
.
U.S. Environmental Protection Agency (U.S. EPA). 2018. Environmental Benefits Mapping and
Analysis Program - Community Edition. User's Manual. Office of Air Quality Planning
and Standards, Health and Environmental Impacts Division, Research Triangle Park, NC.
Available at: < https://www.epa.gov/sites/production/files/2015-04/documents/benmap-
ce_user_manual_march_2015 ,pdf>
5B-29
-------
U.S. Environmental Protection Agency (U.S. EPA). 2019a. Integrated Science Assessment (ISA)
for Particulate Matter (Final Report, 2019). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-19/188, 2019.
U.S. Environmental Protection Agency (U.S. EPA). 2019b. Regulatory Impact Analysis for
Repeal of the Clean Power Plan; Emission Guidelines for Greenhouse Gas Emissions
from Existing Electric Utility Generating Units; Revisions to Emission Guidelines
Implementing Regulations. EPA-452/R19-003. June 2019
U.S. Environmental Protection Agency (U.S. EPA). 2020a. Benefit and Cost Analysis for
Revisions to the Effluent Limitations Guidelines and Standards for the Steam Electric
Power Generating Point Source Category. U.S. Environmental Protection Agency,
Washington, DC, EPA-821-R-20-003.
U.S. Environmental Protection Agency (U.S. EPA). 2020b. Integrated Science Assessment (ISA)
for Ozone and Related Photochemical Oxidants (Final Report). U.S. Environmental
Protection Agency, Washington, DC, EPA/600/R-20/012, 2020.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2000. An
SAB Report on EPA's White Paper Valuing the Benefits of Fatal Cancer Risk Reduction.
EPA-SAB-EEAC-00-013. July. Available at:
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004.
Advisory Council on Clean Air Compliance Analysis Response to Agency Request on
Cessation Lag. EPA-COUNCIL-LTR-05-001. December. Available at:
.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009b.
Review of Integrated Science Assessment for Particulate Matter (Second External
Review Draft, July 2009). EPA-CASAC-10-001. November. Available at:
.
5B-30
-------
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2010.
Review of EPA's Draft Health Benefits of the Second Section 812 Prospective Study of
the CAA
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2020.
Transmittal of the Science Advisory Board's Consideration of the Scientific and
Technical Basis of EPA's Proposed Rule titled "Increasing Consistency and
Transparency in Considering Benefits and Costs in the Clean Air Act Rulemaking
Process." Available at:
.
USGCRP. 2016. The Impacts of Climate Change on Human Health in the United States: A
Scientific Assessment.; doi:http://dx.doi.org/10.7930/J0R49NQX.
Woodruff, T.J., J. Grillo, and K.C. Schoendorf. 1997. "The Relationship between Selected of
postneonatal infant mortality and particulate air pollution in the United States."
Environmental Health Perspectives. 105(6): 608-612
Zanobetti A, Schwartz J. 2008. Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med
177:184-9; doi: 10.1164/rccm.200706-8230C.
5B-31
-------
CHAPTER 6: STATUTORY AND EXECUTIVE ORDER REVIEWS
Overview
This chapter presents the statutory and executive orders applicable to EPA rules, and
discusses EPA's actions taken pursuant to these orders.
6.1 Executive Order 12866: Regulatory Planning and Review
This action is an economically significant regulatory action that was submitted to the
Office of Management and Budget (OMB) for review. EPA believes if the ozone and PM2.5-
related health benefits were quantified and monetized that the benefits of the proposed rule
would exceed $100 million in one of the analytic years. Any changes made in response to OMB
recommendations have been documented in the docket. EPA prepared an analysis of the
potential costs and benefits associated with this proposed action. This analysis is available in the
docket and is briefly summarized in Section IX of the preamble.
6.2 Executive Order 13771
This action is expected to be an Executive Order 13771 regulatory action.
6.3 Paperwork Reduction Act
This action does not impose any new information collection burden under the PRA. This
action would relocate certain existing information collection requirements for certain sources
from subpart EEEEE of 40 CFR part 97 to a new subpart GGGGG of 40 CFR part 97 but would
make no changes to any existing information collection requirements for any source. OMB has
previously approved the information collection activities contained in the existing regulations
and has assigned OMB control number 2060-0667.
6.4 Regulatory Flexibility Act
EPA certifies that this action will not have a significant economic impact on a substantial
number of small entities under the Regulatory Flexibility Act (RFA). The small entities subject
to the requirements of this action are small businesses, small organizations, and small
governmental jurisdictions. EPA has determined that no small entities potentially affected by the
6-1
-------
proposal will have compliance costs greater than 1 percent of annual revenues in 2021. Details of
this analysis are presented below.
The Regulatory Flexibility Act (5 U.S.C. 601 et seq.), as amended by the Small Business
Regulatory Enforcement Fairness Act (Public Law No. 104 121), provides that whenever an
agency is required to publish a general notice of proposed rulemaking, it must prepare and make
available an initial regulatory flexibility analysis, unless it certifies that the proposed rule, if
promulgated, will not have a significant economic impact on a substantial number of small
entities (5 U.S.C. 605[b]). Small entities include small businesses, small organizations, and
small governmental jurisdictions.
EPA conducted regulatory flexibility analysis at the ultimate (i.e., highest) level of
ownership, evaluating parent entities with the largest share of ownership in at least one
potentially-affected EGU included in EPA's base case using the IPM v.6, used in this RIA.l This
analysis draws on the "parsed" unit-level estimates using IPM results for 2021, as well as
ownership, employment, and financial information for the potentially affected small entities
drawn from other resources described in more detail below. This analysis is focused on
estimating impacts in 2021 because implementation of the proposed EGU controls occurs in the
2021 ozone season.
EPA identified the size of ultimate parent entities by using the Small Business
Administration (SBA) size standard guidelines.2 The criteria for size determination vary by the
organization/operation category of the ultimate parent entity, as follows:
• Privately-owned (non-government) entities (see Table 6-1)
o Privately-owned entities include investor-owned utilities, non-utility entities,
and entities with a primary business other than electric power generation.
o For entities with electric power generation as a primary business, small entities
are those with less than the threshold number of employees specified by SBA
1 Detailed documentation for IPM v.6 is available at: http://www.epa.gov/airmarkets/powersectormodeling.html.
2 U.S. Small Business Administration (SBA). 2019. Small Business Size Standards, effective as of August 19, 2019
and available at the following link: https://www.sba.gov/sites/default/files/2019-
08/SBA%20Table%20of%20Size%20Standards_Effective%20Aug%2019%2C%202019.pdf.
6-2
-------
for each of the relevant North American Industry Classification System
(NAICS) sectors (NAICS 2211).
o For entities with a primary business other than electric power generation, the
relevant size criteria are based on revenue, assets, or number of employees by
NAICS sector.3
• Publicly-owned entities
o Publicly-owned entities include federal, state, municipal, and other political
subdivision entities.
o The federal and state governments are considered to be large. Municipalities
and other political units with populations fewer than 50,000 ae considered to be
small.
• Rural Electric Cooperatives
o Small entities are those with fewer than the threshold level of employees or
revenue specified by SBA for each of the relevant NAICS sectors.
6.4.1 Identification of Small Entities
In this analysis, EPA considered EGUs that meet the following five criteria: 1) EGU is
represented in NEEDS v6; 2) EGU is fossil fuel-fired; 3) EGU is located in a state covered by
this proposed rule; 4) EGU is neither a cogeneration unit nor solid waste incineration unit; and 5)
EGU capacity is 25 MW or larger. EPA next refined this list of EGUs, narrowing it to those that
exhibit at least one of the following changes under the proposal, in comparison to the baseline.
• Summer fuel use (BTUs) changes by +/- 1 percent or more
• Summer generation (GWh) changes by +/- 1 percent or more
• NOx summer emissions (tons) changes by +/- 1 percent or more
Based on these criteria, EPA identified a total of 97 potentially affected EGUs warranting
examination in this RFA analysis. Next, we determined power plant ownership information,
3 Certain affected EGUs are owned by ultimate parent entities whose primary business is not electric power
generation.
6-3
-------
including the name of associated owning entities, ownership shares, and each entity's type of
ownership. We primarily used data from Ventyx, supplemented by limited research using
publicly available data.4 Majority owners of power plants with affected EGUs were categorized
as one of the seven ownership types.5 These ownership types are:
1. Investor-Owned Utility (IOU): Investor-owned assets (e.g., a marketer, independent
power producer, financial entity) and electric companies owned by stockholders, etc.
2. Cooperative (Co-Op): Non-profit, customer-owned electric companies that generate
and/or distribute electric power.
3. Municipal: A municipal utility, responsible for power supply and distribution in a small
region, such as a city.
4. Sub-division: Political subdivision utility is a county, municipality, school district,
hospital district, or any other political subdivision that is not classified as a municipality
under state law.
5. Private: Similar to an investor-owned utility, however, ownership shares are not openly
traded on the stock markets.
6. State: Utility owned by the state.
7. Federal: Utility owned by the federal government.
Next, EPA used both the D&B Hoover's online database and the Ventyx database to
identify the ultimate owners of power plant owners identified in the Ventyx database. This was
necessary, as many majority owners of power plants (listed in Ventyx) are themselves owned by
other ultimate parent entities (listed in D&B Hoover's).6 In these cases, the ultimate parent entity
was identified via D&B Hoover's, whether domestically or internationally owned.
4 The Ventyx Energy Velocity Suite database consists of detailed ownership and corporate affiliation information at
the EGU level. For more information, see: www.ventyx.com.
5 Throughout this analysis, EPA refers to the owner with the largest ownership share as the "majority owner" even
when the ownership share is less than 51 percent.
6 The D&B Hoover's online platform includes company records that can contain NAICS codes, number of
employees, revenues, and assets. For more information, see: https://www.dnb.com/products/marketing-sales/dnb-
hoovers.html.
6-4
-------
EPA followed SBA size standards to determine which non-government ultimate parent
entities should be considered small entities in this analysis. These SBA size standards are
specific to each industry, each having a threshold level of either employees, revenue, or assets
below which an entity is considered small. SBA guidelines list all industries, along with their
associated NAICS code and SBA size standard. Therefore, it was necessary to identify the
specific NAICS code associated with each ultimate parent entity in order to understand the
appropriate size standard to apply. Data from D&B Hoover's was used to identify the NAICS
codes for most of the ultimate parent entities. In many cases, an entity that is a majority owner of
a power plant is itself owned by an ultimate parent entity with a primary business other than
electric power generation. Therefore, it was necessary to consider SBA entity size guidelines for
the range of NAICS codes listed in Table 6-1. This table represents the range of NAICS codes
and areas of primary business of ultimate parent entities which are majority owners of potentially
affected EGUs in EPA's IPM base case.
6-5
-------
Table 6-1. SBA Size Standards by NAICS Code
Size
Size
Standards
Standards
NAICS
(millions of
(number of
Codes
NAICS U.S. Industry Title
dollars)
employees)
221111
Hydroelectric Power Generation
500
221112
Fossil Fuel Electric Power Generation
750
221113
Nuclear Electric Power Generation
750
221114
Solar Electric Power Generation
250
221115
Wind Electric Power Generation
250
221116
Geothermal Electric Power Generation
250
221117
Biomass Electric Power Generation
250
221118
Other Electric Power Generation
Electric Bulk Power Transmission and
250
221121
Control
500
221122
Electric Power Distribution
1000
221210
Natural Gas Distribution
1000
221310
Water Supply and Irrigation Systems
$30
221320
Sewage Treatment Facilities
$22
221330
Steam and Air-Conditioning Supply
$16
Note: Based on size standards effective at the time EPA conducted this analysis (SBA size standards, effective
August 19, 2019. Available at the following link: https://www.sba.gov/document/support--table-size-standards).
Source: SBA, 2019
EPA compared the relevant entity size criterion for each ultimate parent entity to the SBA
size standard noted in Table 6-1. We used the following data sources and methodology to
estimate the relevant size criterion values for each ultimate parent entity:
1. Employment, Revenue, and Assets: EPA used the D&B Hoover's database as the
primary source for information on ultimate parent entity employee numbers, revenue, and
assets.7 In parallel, EPA also considered estimated revenues from affected EGUs based
on analysis of parsed-ftle estimates for the proposal. EPA assumed that the ultimate
parent entity revenue was the larger of the two revenue estimates. In limited instances,
supplemental research was also conducted to estimate an ultimate parent entity's number
of employees, revenue, or assets.
7 Estimates of sales were used in lieu of revenue estimates when revenue data was unavailable.
6-6
-------
2. Population: Municipal entities are defined as small if they serve populations of less than
50,000. EPA primarily relied on data from the Ventyx database and the U.S. Census
Bureau to inform this determination.
Ultimate parent entities for which the relevant measure is less than the SBA size standard
were identified as small entities and carried forward in this analysis. In total EPA identified 97
potentially affected EGUs, owned by 16 entities. Of these, EPA identified 7 potentially affected
EGUs owned by 2 small entities8 included in EPA's Base Case.
6.4.2 Overview of Analysis and Results
This section presents the methodology and results for estimating the impact of the
Revised CSAPR Update proposal on small entities in 2021 based on the following endpoints:
• annual economic impacts of the Revised CSAPR Update proposal on small
entities, and
• ratio of small entity impacts to revenues from electricity generation.
6.4.2.1 Methodology for Estimating Impacts of the Revised CSAPR Update proposal on Small
Entities
An entity can comply with the Revised CSAPR Update proposal through some
combination of the following: optimizing existing SCRs, turning on idled SCR controls, using
allocated allowances, purchasing allowances, or reducing emissions through a reduction in
generation. Additionally, units with more allowances than needed can sell these allowances in
the market. The chosen compliance strategy will be primarily a function of the unit's marginal
control costs and its position relative to the marginal control costs of other units.
To attempt to account for each potential control strategy, EPA estimates compliance costs
as follows:
Ccompliance A Coperating+Retrofit A CFuel A CAllowances A Cfransaction ~ A R
8 Both of these small entities are inNAICS 221118, which is defined as establishments primarily engaged in
operating electric power generation facilities (except hydroelectric, fossil fuel, nuclear, solar, wind, geothermal,
biomass). These facilities convert other forms of energy, such as tidal power, into electric energy.
6-7
-------
where C represents a component of cost as labeled, and A R represents the value of foregone
electricity generation, calculated as the difference in revenues between the base case and the
Revised CSAPR Update proposal in 2021.
Realistically, compliance choices and market conditions can combine such that an entity
may actually experience a savings in any of the individual components of cost. Under the
Revised CSAPR Update proposed rule, some units will forgo some level of electricity generation
(and thus revenues) to comply and this impact will be lessened on these entities by the projected
increase in electricity prices under the Revised CSAPR Update proposed rule. On the other hand,
those increasing generation levels will see an increase in electricity revenues and as a result,
lower net compliance costs. If entities are able to increase revenue more than an increase in fuel
cost and other operating costs, ultimately they will have negative net compliance costs (or
savings). Overall, small entities are not projected to install relatively costly emissions control
retrofits but may choose to do so in some instances. Because this analysis evaluates the total
costs along each of the compliance strategies laid out above for each entity, it inevitably captures
savings or gains such as those described. As a result, what we describe as cost is really more of a
measure of the net economic impact of the proposal on small entities.
For this analysis, EPA used IPM-parsed output to estimate costs based on the parameters
above, at the unit level. These impacts were then summed for each small entity, adjusting for
ownership share. Net impact estimates were based on the following: operating and retrofit costs,
sale or purchase of allowances, and the change in fuel costs or electricity generation revenues
under the Revised CSAPR Update proposal relative to the base case. These individual
components of compliance cost were estimated as follows:
(1) Operating and retrofit costs: Using engineering analytics, EPA identified which
compliance option was selected by each EGU in 2021 (i.e., SCR optimization or
turning on existing SCR controls) and applied the appropriate cost to this choice.
EPA assumes that state of the art combustion controls may be installed in 2022
and are not part of the controls available in 2021.
(2) Sale or purchase of allowances: To estimate the value of allowances holdings,
allocated allowances were subtracted from projected emissions, and the difference
was then multiplied by $1,600 (2016$) per ton, which is the marginal cost of NOx
6-8
-------
reductions used to set the proposed budgets in the Revised CSAPR Update
proposal. While this is a reasonable approximation, it is possible that the actual
allowance price could be lower. Units were assumed to purchase or sell
allowances to exactly cover their projected emissions under the Revised CSAPR
Update proposal.
(3) Fuel costs: The change in fuel expenditures under the Revised CSAPR Update
proposal was estimated by taking the difference in projected fuel expenditures
between the IPM estimates for the Revised CSAPR Update proposed rule and the
base case.
(4) Value of electricity generated: To estimate the value of electricity generated, the
projected level of electricity generation is multiplied by the regional-adjusted
retail electricity price ($/MWh) estimate, for all entities except those categorized
as private in Ventyx. For private entities, EPA used the wholesale electricity price
instead of the retail electricity price because most of the private entities are
independent power producers (IPP). IPPs sell their electricity to wholesale
purchasers and do not own transmission facilities. Thus, their revenue was
estimated with wholesale electricity prices.
(5) Administrative costs: Because most affected units are already monitored as a
result of other regulatory requirements, EPA considered the primary
administrative cost to be transaction costs related to purchasing or selling
allowances. EPA assumed that transaction costs were equal to 1.5 percent of the
total absolute value of the difference between a unit's allocation and projected
NOx emissions. This assumption is based on market research by ICF.
6.4.2.2 Results
The potential impacts of the Revised CSAPR Update proposal on small entities are
summarized in Table 6-2. All costs are presented in 2016$. EPA estimated the annual net
compliance cost to small entities to be approximately $0.04 million in 2021.
6-9
-------
Table 6-2. Projected Impact of the Revised CSAPR Update Proposal on Small Entities in
2021
EGU
Ownership
Type
Number of
Potentially
Affected
Entities
Total Net
Compliance
Cost
($2016
millions)
Number of Small
Entities with
Compliance Costs
>1% of Generation
Revenues
Number of Small
Entities with
Compliance Costs
>3% of Generation
Revenues
Cooperative
1
0.04
0
0
Private
1
0.00
0
0
Total
2
0.04
0
0
Source: IPM analysis
EPA assessed the economic and financial impacts of the proposed rule using the ratio of
compliance costs to the value of revenues from electricity generation, focusing in particular on
entities for which this measure is greater than 1 percent. Although this metric is commonly used
in EPA impact analyses, it makes the most sense when as a general matter an analysis is looking
at small businesses that operate in competitive environments.9 However, small businesses in the
electric power industry often operate in a price-regulated environment where they are able to
recover expenses through rate increases. Given this, EPA considers the 1 percent measure in this
case a crude measure of the price increases these small entities will be asking of rate
commissions or making at publicly owned companies. Of the 2 small entities considered in this
analysis, neither is projected to experience compliance costs greater than 1 percent of generation
revenues in 2021. EPA has concluded that there is no significant economic impact on a
substantial number of small entities (no SISNOSE) for this rule.
The separate components of annual costs to small entities under the Revised CSAPR
Update proposal are summarized in Table 6-3. The most significant components of incremental
cost to the cooperative category under the Revised CSAPR Update proposal are due to higher
operating costs (reflecting the cost of controls). Among the private category, however, reduced
generation is the key driver. Total impacts to the private category are well below $10,000.
9 U.S. EPA. EPA's Action Development Process. Final Guidance for EPA Rulewriters: Regulatory Flexibility Act as
Amended by the Small Business Regulatory Enforcement Fairness Act. September 2006. Available at
https://www.epa.gov/sites/production/files/2015-06/documents/guidance-regflexact.pdf.
6-10
-------
Table 6-3. Incremental Annual Costs under the Revised CSAPR Update Proposal
EGU
Ownership
Type
Operating
Cost
Net Purchase
of Allowances
Fuel Cost
Lost
Electricity
Revenue
Administrative
Cost
Cooperative
0.06
0.00
-0.02
0.00
0.00
Private
0.00
0.00
0.00
0.00
0.00
Source: IPM analysis
6.4.3 Summary of Small Entity Impacts
EPA examined the potential economic impacts to small entities associated with this
proposal based on assumptions of how the affected states will implement control measures to
meet their emissions. To summarize, of the 2 small entities potentially affected, none are
projected to experience compliance costs in excess of 1 percent of revenues in 2021, based on
assumptions of how the affected states implement control measures to meet their emissions
budgets as set forth in this proposal.
EPA has lessened the impacts for small entities by excluding all units smaller than 25 MW.
This exclusion, in addition to the exemptions for cogeneration units and solid waste incineration
units, eliminates the burden of higher costs for a substantial number of small entities located in
the 12 states for which EPA is proposing FIPs.
6.5 Unfunded Mandates Reform Act
Title II of the UMRA of 1995 (Public Law 104-4) (UMRA) establishes requirements for
federal agencies to assess the effects of their regulatory actions on state, local, and Tribal
governments and the private sector. Under Section 202 of the UMRA, 2 U.S.C. 1532, EPA
generally must prepare a written statement, including a cost-benefit analysis, for any proposed or
final rule that includes any Federal mandate that may result in the expenditure by State, local,
and Tribal governments, in the aggregate, or by the private sector, of $100,000,000 or more in
any one year. A Federal mandate is defined under Section 421(6), 2 U.S.C. 658(6), to include a
Federal intergovernmental mandate and a Federal private sector mandate. A Federal
intergovernmental mandate, in turn, is defined to include a regulation that would impose an
6-11
-------
enforceable duty upon State, Local, or Tribal governments, Section 421(5)(A)(i), 2 U.S.C.
658(5)(A)(i), except for, among other things, a duty that is a condition of Federal assistance,
Section 421(5)(A)(i)(I). A Federal private sector mandate includes a regulation that would
impose an enforceable duty upon the private sector, with certain exceptions, Section 421(7)(A), 2
U.S.C. 658(7)(A).
As outlined in Section 4.4.2, EPA projects the total cost of compliance with the Revised
CSAPR Update proposal to be well below $100 million in every year. Furthermore, as EPA
stated in the proposal, EPA is not directly establishing any regulatory requirements that may
significantly or uniquely affect small governments, including Tribal governments. Thus, under
the Revised CSAPR Update proposal, EPA is not obligated to develop under Section 203 of the
UMRA a small government agency plan.
6.6 Executive Order 13132: Federalism
This proposed action does not have federalism implications. If finalized, this proposed
action will not have substantial direct effects on the states, on the relationship between the
national government and the states, or on the distribution of power and responsibilities among
the various levels of government.
6.7 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments
This proposed action has tribal implications. However, it would neither impose substantial
direct compliance costs on federally recognized tribal governments, nor preempt tribal law.
This action proposes to implement EGU NOx ozone season emissions reductions in 12
eastern states (Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New
York, Ohio, Pennsylvania, Virginia, and West Virginia). However, at this time, none of the
existing or planned EGUs affected by this rule are owned by tribes or located in Indian country.
This proposed action may have tribal implications if a new affected EGU is built in Indian
country. Additionally, tribes have a vested interest in how this proposed rule would affect air
quality.
6-12
-------
In developing the CSAPR, which was promulgated on July 6, 2011, to address interstate
transport of ozone pollution under the 1997 ozone NAAQS, EPA consulted with tribal officials
under EPA Policy on Consultation and Coordination with Indian Tribes early in the process of
developing that regulation to allow for meaningful and timely tribal input into its development.
A summary of that consultation is provided at 76 FR 48346.
EPA received comments from several tribal commenters regarding the availability of the
CSAPR allowance allocations to new units in Indian country. EPA responded to these comments
by instituting Indian country new unit set-asides in the final CSAPR. In order to protect tribal
sovereignty, these set-asides are managed and distributed by the federal government regardless
of whether the CSAPR in the adjoining or surrounding state is implemented through a FIP or
SIP. While there are no existing affected EGUs in Indian country covered by this proposal, the
Indian country set-asides will ensure that any future new units built in Indian country will be able
to obtain the necessary allowances. This proposal maintains the Indian country new unit set-aside
and adjusts the amounts of allowances in each set-aside according to the same methodology of
the CSAPR rule.
EPA informed tribes of our development of this proposal through a National Tribal Air
Association - EPA air policy conference call on June 25, 2020. EPA plans to further consult with
tribal officials under EPA Policy on Consultation and Coordination with Indian Tribes early in
the process of developing this proposed regulation to solicit meaningful and timely input into its
development. EPA will facilitate this consultation before finalizing this proposed rule.
6.8 Executive Order 13045: Protection of Children from Environmental Health &
Safety Risks
This proposed action is not subject to EO 13045 because EPA does not believe the
environmental health risks or safety risks addressed by this action present a disproportionate risk
to children. This action's health and risk assessments are discussed in Chapter 5.
6.9 Executive Order 13211: Actions that Significantly Affect Energy Supply,
Distribution, or Use
This action, which is a significant regulatory action under EO 12866, is likely to have a
significant effect on the supply, distribution, or use of energy. EPA has prepared a Statement of
6-13
-------
Energy Effects for the proposed regulatory control alternative as follows. We estimate a less than
1 percent change in retail electricity prices on average across the contiguous U.S. in 2021, and a
less than 1 percent change in coal-fired electricity generation in 2021 as a result of this proposed
rule. EPA projects that utility power sector delivered natural gas prices will change by less than 1
percent in 2021. For more information on the estimated energy effects, please see Chapter 4 of
this RIA.
6.10 National Technology Transfer and Advancement Act
The proposed rulemaking does not involve technical standards.
6.11 Executive Order 12898: Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations
EPA believes the human health or environmental risk addressed by this action will not
have potential disproportionately high and adverse human health or environmental effects on
minority, low-income, or indigenous populations.
EPA notes that this action proposes to revise the CSAPR Update to reduce interstate ozone
transport with respect to the 2008 ozone NAAQS. This rule uses EPA's authority in CAA section
110(a)(2)(d) (42 U.S.C. 7410(a)(2)(d)) to reduce NOx pollution that significantly contributes to
downwind ozone nonattainment or maintenance areas. As a result, the rule will reduce exposures
to ozone in the most-contaminated areas (i.e., areas that are not meeting the 2008 ozone
NAAQS). In addition, the proposed rule separately identifies both nonattainment areas and
maintenance areas. This requirement reduces the likelihood that areas close to the level of the
standard will exceed the current health-based standards in the future. EPA proposes to implement
these emission reductions using the CSAPR NOx Ozone Season Group 3 program with
assurance provisions.
EPA recognizes that many environmental justice communities have voiced concerns in the
past about emission trading and the potential for any emission increases in any location. The
CSAPR NOx Ozone Season Group 3 trading program in the proposed action is the result of
EPA's application of the 4-step framework to reduce interstate ozone pollution and implement
those reductions, similar to the emissions trading programs developed in the CSAPR (CSAPR
NOx Ozone Season Group 1 trading program) and modified in the CSAPR Update (CSAPR
6-14
-------
NOx Ozone Season Group 2 trading program), both of which also resulted from the application
of the 4-step framework. EPA believes that this approach used in the CSAPR and in the CSAPR
Update mitigated community concerns about emissions trading, and that this proposal, which
applies the same 4-step framework and proposes an emissions trading program similar to those
used in the CSAPR and the CSAPR Update, will also minimize community concerns. EPA seeks
comment from communities on this proposal.
Ozone pollution from power plants has both local and regional components: part of the
pollution in a given location—even in locations near emission sources—is due to emissions from
nearby sources and part is due to emissions that travel hundreds of miles and mix with emissions
from other sources. It is important to note that the section of the Clean Air Act providing
authority for this proposed rule, section 110(a)(2)(D) (42 U.S.C. 7410(a)(2)(D)), unlike some
other provisions, does not dictate levels of control for particular facilities. In this proposed
action, as in the CSAPR and the CSAPR Update, sources in the emissions trading program may
trade allowances with other sources in the same or different states, but any emissions shifting
that may occur is constrained by an effective ceiling on emissions in each state (the assurance
level). As in the CSAPR and the CSAPR Update, assurance provisions in the proposed rule
outline the allowance surrender penalties for failing to meet the assurance level (see section
VIII.C.2.); there are additional allowance for failing to hold an adequate number of allowances to
cover emissions.
This approach will reduce EGU emissions in each state that significantly contributes to
downwind nonattainment or maintenance areas with respect to the 2008 ozone NAAQS, while
allowing power companies to adjust generation as needed and ensure that the country's
electricity needs will continue to be met. As in the CSAPR and the CSAPR Update, EPA
believes that the existence of these assurance provisions in the emissions trading program,
including the penalties imposed when triggered, will ensure that emissions from states covered
by this proposal will stay below the level of the budget plus variability limit.
In addition, under this proposed rule all sources participating in the CSAPR NOx Ozone
Season Group 3 trading program must hold enough allowances to cover their emissions.
Therefore, if a source emits more than its allocation in a given year, either another source must
6-15
-------
have used less than its allocation and be willing to sell some of its excess allowances, or the
source itself had emitted less than its allocation in one or more previous years (i.e., banked
allowances for future use).
In summary, like the CSAPR and the CSAPR Update, this proposed rule minimizes
community concerns about localized hot spots and reduces ambient concentrations of pollution
where they are most needed by sensitive and vulnerable populations by: considering the science
of ozone transport to set strict state emissions budgets to reduce significant contributions to
ozone nonattainment and maintenance (i.e., the most polluted) areas; implementing air quality-
assured trading; requiring any emissions above the level of the allocations to be offset by
emission decreases; and imposing strict penalties for sources that contribute to a state's
exceedance of its budget plus variability limit. In addition, it is important to note that nothing in
this proposed rule allows sources to violate their title V permit or any other federal, state, or local
emissions or air quality requirements.
In addition, it is important to note that CAA section 110(a)(2)(D), which addresses
transport of criteria pollutants between states, is only one of many provisions of the CAA that
provide EPA, states, and local governments with authorities to reduce exposure to ozone in
communities. These legal authorities work together to reduce exposure to these pollutants in
communities, including for minority, low-income, and tribal populations, and provide substantial
health benefits to both the general public and sensitive sub-populations.
EPA has already taken steps to begin informing communities of our development of this
proposal through a National Tribal Air Association - EPA air policy conference call on June 25,
2020. EPA plans to further consult with communities early in the process of developing this
regulation to permit them to have meaningful and timely input into its development. EPA will
facilitate this engagement before finalizing this proposed rule.
6-16
-------
CHAPTER 7: COMPARISON OF BENEFITS AND COSTS
Overview
EPA performed an analysis to estimate the costs and climate benefits of compliance with
the proposed Revised CSAPR Update and more and less stringent alternatives. EPA is proposing
electric generating unit (EGU) oxides of nitrogen (NOx) ozone season emissions budgets for 12
states.1 This action proposes to find that for these states, their projected 2021 ozone season NOx
emissions significantly contribute to downwind states' nonattainment and/or maintenance
problems for the 2008 ozone national ambient air quality standards (NAAQS). For these 12
states, EPA proposes to amend their federal implementation plans (FIPs) to revise the existing
Cross-State Air Pollution Rule (CSAPR) NOx Ozone Season Group 2 emissions budgets for
EGUs and implement the revised budgets beginning in the 2021 ozone season (May 1, 2021 -
September 30, 2021) via a new CSAPR NOx Ozone Season Group 3 Trading Program.
The proposed Revised CSAPR Update state budgets reflect the optimization of existing
selective catalytic reduction (SCR) controls and installation of state-of-the-art NOx combustion
controls, with an estimated marginal cost of $1,600 per ton (2016$). For the RIA, in order to
implement the OMB Circular A-4 requirement for fulfilling Executive Order 12866 to assess one
less stringent and one more stringent alternative to the proposal, EPA is also analyzing EGU
NOx ozone season emissions budgets reflecting NOx reduction strategies that are widely
available at a uniform cost of $9,600 per ton (2016$) and strategies that are widely available at a
uniform cost of $500 per ton (2016$). These alternatives are used illustrate the monetized cost
and climate benefit impacts of varying program stringency. They are designed to show the
effects of more stringent and less stringent NOx reduction requirements in a regulatory structure
that is otherwise the same as the proposed NOx emissions budgets. We show the results for 2021
to reflect the year in which implementation of this proposal begins, and for 2025 to reflect full
implementation of the proposal. This RIA evaluates how the EGUs covered by the proposed rule
are expected to reduce their emissions in response to the requirements and flexibilities provided
1 The 12 states include Illinois, Indiana, Kentucky, Louisiana, Maryland, Michigan, New Jersey, New York, Ohio,
Pennsylvania, Virginia, and West Virginia.
7-1
-------
by the remedy implemented by the proposed Revised CSAPR Update and the benefit, cost and
impacts of their expected compliance behavior. This chapter summarizes these results.
7.1 Results
The proposal and regulatory control alternatives' compliance costs are estimated using the
IPM model and an evaluation of control technologies evaluated outside of IPM. As shown in
Chapter 4, the estimated annual compliance costs to implement the proposal, as described in this
document, are approximately $21 million in 2021 and $6 million in 2025 (2016$). As described
in Section 4.5, this RIA uses compliance costs as a proxy for social costs. As shown in Chapter
5, the estimated monetized climate benefits from implementation of the proposal are
approximately $0.31 million and $0.05 million in 2021 (2016$, based on a real discount rate of 3
percent and 7 percent, respectively). For 2025, the estimated monetized climate benefits from
implementation of the proposal are approximately $33 million and $5.4 million (2016$, based on
a real discount rate of 3 percent and 7 percent, respectively). As discussed in Chapter 5, the
monetized benefits presented in this proposal RIA are those for climate (from CO2 emissions
reductions). The non-monetized benefits for ozone and PM2.5 are discussed qualitatively in
Chapter 5.
EPA calculates the net benefits of the proposal by subtracting the estimated compliance
costs from the estimated climate benefits in both 2021 and 2025. The annual net benefits of the
proposal in 2021 (in 2016$) are approximately -$21 million using both a 3 percent and 7 percent
real discount rate for the climate benefits. The annual net benefits of the proposal in 2025 are
approximately $27 using a 3 percent real discount rate and -$0.9 million using a 7 percent real
discount rate. Table 7-1 presents a summary of the climate benefits, costs, and net benefits of the
proposal and the more and less stringent alternatives for 2021. Table 7-2 presents a summary of
these impacts for the proposal and the more and less stringent alternatives for 2025. The tables
represent the present annual value of non-monetized benefits from ozone, PM2.5 and NO2
reductions as a B. The annual value of B will differ across discount rates, year of analysis, and
the regulatory alternatives analyzed. At a 3 and 7 percent real discount rate the least stringent
alternative has the greatest annual monetized net-benefits in the two analytic years. The
monetized net-benefit estimates exclude important benefits from reductions in ozone and PM2.5
concentrations.
7-2
-------
Table 7-1. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2021 for the U.S. (millions of 2016$)a'b,c,d
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
0.31 +B
21
-21 + B
7%
0.05+B
-21 + B
More Stringent
Alternative
3%
0.80+ B
37
-36+B
7%
0.12 + B
-37+B
Less Stringent
Alternative
3%
0.17 + B
4
-4 + B
7%
0.03 +B
-4 + B
a We focus results to provide a snapshot of costs and benefits in 2021, using the best available information to
approximate social costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in C02 emissions. The costs presented in this table are 2021 annual
estimates for each alternative analyzed.
0 All costs and benefits are rounded to two significant figures; rows may not appear to add correctly.
d B is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual value of B will differ across discount
rates, year of analysis, and the regulatory alternatives analyzed. While EPA did not estimate these benefits in this
RIA, Appendix 5B presents PM2 5 and ozone estimates quantified using methods consistent with the previously
published ISAs to provide information regarding the potential magnitude of the benefits of this proposed rule.
Table 7-2. Benefits, Costs, and Net Benefits of the Proposal and More and Less Stringent
Alternatives for 2025 for the U.S. (millions of 2016$)a,b,c,d
Discount Rate
Benefits
Costs
Net Benefits
Proposal
3%
33 +B
6
27+B
7%
5.4+B
-0.9+ B
More Stringent
Alternative
3%
71.5 + B
132
-61 + B
7%
11.7 + B
-120+ B
Less Stringent
Alternative
3%
25+B
-12
37+B
7%
4.2+B
16+B
" We focus results to provide a snapshot of costs and benefits in 2025, using the best available information to
approximate social costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in CO2 emissions. The costs presented in this table are 2025annual
estimates for each alternative analyzed.
0 All costs and benefits are rounded to two significant figures; rows may not appear to add correctly.
d B is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual value of B will differ across discount
rates, year of analysis, and the regulatory alternatives analyzed. While EPA did not estimate these benefits in this
RIA, Appendix 5B presents PM2 5 and ozone estimates quantified using methods consistent with the previously
published ISAs to provide information regarding the potential magnitude of the benefits of this proposed rule.
7-3
-------
As part of fulfilling analytical guidance with respect to E.O. 12866, EPA presents
estimates of the present value of the benefits and costs over the five-year period of 2021 to 2025,
which is the analytical period for this proposal. To calculate the present value of the social net-
benefits of the proposed Revised CSAPR Update, annual benefits and costs are discounted to
2021 at 3 percent and 7 discount rates as directed by OMB's circular A-4. The present value
(PV) of the net benefits, in 2016$ and discounted to 2021, is -$68 million when using a 7 percent
discount rate and $14 million when using a 3 percent discount rate.2 The equivalent annualized
value (EAV), an estimate of the annualized value of the net benefits consistent with the present
value, is -$17 million per year when using a 7 percent discount rate and $3 million when using a
3 percent discount rate. The EAV represents a flow of constant annual values that, had they
occurred in each year from 2021 to 2025, would yield a sum equivalent to the PV. The EAV
represents the value of a typical cost or benefit for each year of the analysis, in contrast to the
year-specific estimates mentioned earlier in the RIA for the analysis years 2021 and 2025. The
comparison of benefits and costs in PV and EAV terms for the proposal can be found in Table 7-
3. Estimates in the table are presented as rounded values. The table represents the present value
of non-monetized benefits from ozone, PM2.5 and NO2 reductions as a P, while b represents the
equivalent annualized value of these non-monetized benefits. These values will differ across the
discount rates and depend on the value of the B's in the previous tables.
Table 7-3. Summary of Present Values and Equivalent Annualized Values for the 2021-
2025 Timeframe for Estimated Compliance Costs, Climate Benefits, and Net
Benefits for the Proposed Rule (millions of 2016$, discounted to 2021)a'b
3% Discount Rate
7% Discount Rate
Present Value
Benefitscd
101+p
15+p
Climate Benefits0
101
15
Compliance Costs6
87
83
Net Benefits
14+P
-68+P
Equivalent Annualized
value
Benefits
22+b
4+b
Climate Benefits
22
4
Compliance Costs
19
20
2 In annualizing compliance costs using social discount rates, this analysis treats the annual compliance costs as
reflecting the use of real resources in a particular year. In practice, annual costs from IPM and costs of NOx controls
estimated outside of IPM (e.g., capital costs of combustion controls) reflect annual payments for financed capital
and not solely the change in the use of real resources in a particular year (i.e., the opportunity cost of those
resources).
7-4
-------
Net Benefits 3+b -17+b
a All estimates in this table are rounded to two significant figures, so numbers may not sum due to independent
rounding.
b The annualized present value of costs and benefits are calculated over a 5 year period from 2021 to 2025.
0 Benefits ranges represent discounting of climate benefits at a real discount rate of 3 percent and 7 percent. Climate
benefits are based on changes (reductions) in CO2 emissions.
d (3 and b is the sum of all unqualified ozone, PM2 5, and NO2 benefits. The annual values of (3 and b will differ
across discount rates. While EPA did not estimate these benefits in this RIA, Appendix 5B presents PM2 5 and ozone
estimates quantified using methods consistent with the previously published ISAs to provide information regarding
the potential magnitude of the benefits of this proposed rule.
e The costs presented in this table reflect annualized present value compliance costs calculated over a 5 year period
from 2021 to 2025.
7-5
-------
United States Office of Air Quality Planning and Standards Publication No. EPA-452/P-20-003
Environmental Protection Health and Environmental Impacts Division October 2020
Agency Research Triangle Park, NC
------- |