Regulatory Impact Analysis of the Cross-State
Air Pollution Rule (CSAPR) Update for the
2008 National Ambient Air Quality Standards
for Ground-Level Ozone

-------
EPA-452/R-16-004
September 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
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, NC 27711
ii

-------
CONTACT INFORMATION
This document has been prepared by staff from the Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency. Questions related to this document should be
addressed to Kathy Kaufman, U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, C439-02, Research Triangle Park, North Carolina 27711 (email:
kaufman.kathy@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.
111

-------
TABLE OF CONTENTS
LIST OF TABLES	viii
LIST OF FIGURES	xii
EXECUTIVE SUMMARY	ES-1
Overview	ES-1
ES. 1 Identifying Needed Emission Reductions	ES-1
ES.3 Control Strategies and Emissions Reductions	ES-7
ES.4 Costs	ES-9
ES.5 Benefits to Human Health and Welfare	ES-9
ES.5.1 Human Health Benefits and Climate Co-benefits	ES-10
ES.5.2 Combined Health Benefits and Climate Co-Benefits Estimates	ES-13
ES.5.3 Unquantified Health and Welfare Co-Benefits	ES-16
ES.5 Results of Benefit-Cost Analysis	ES-18
ES.6 Analytical Changes Subsequent to the Proposal	ES-19
ES.7 References	ES-21
CHAPTER 1: INTRODUCTION AND BACKGROUND	1-1
Introduction	1-1
1.1	Background	1-1
1.2.1	Role of Executive Orders in the Regulatory Impact Analysis	1-3
1.2.2	Illustrative Nature of this Analysis	1-3
1.2.3	The Need for Air Quality or Emissions Standards	1-3
1.2	Overview and Design of the RIA	1-4
1.2.1	Methodology for Identifying Required Reductions	1-4
1.2.2	States Covered by the CSAPR Update	1-6
1.2.3	Regulated Entities	1-6
1.2.4	Baseline and Analysis Year	1-7
1.2.5	Emissions Controls and Cost Analysis Approach	1-8
1.2.6	Benefits Analysis Approach	1-9
1.3	Organization of the Regulatory Impact Analysis	1-9
CHAPTER 2: ELECTRIC POWER SECTOR PROFILE	2-1
Overview	2-1
2.1	Background	2-1
2.2	Power Sector Overview	2-1
2.2.1	Generation	2-2
2.2.2	Transmission	2-9
2.2.3	Distribution	2-10
2.3	Sales, Expenses, and Prices	2-11
iv

-------
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 2000 to 2014 .. 2-16
2.4 Deregulation and Restructuring	2-18
CHAPTER 3: EMISSIONS AND AIR QUALITY MODELING IMPACTS	3-1
Overview	3-1
3.1	Air Quality Modeling Platform	3-1
3.1.1	Simulation Periods	3-2
3.1.2	Air Quality Modeling Domain	3-2
3.1.3	Air Quality Model Inputs	3-3
3.2	Development of Emissions Inventories	3-3
3.2.1	2011 Base Year Emissions	3-3
3.2.2	2017 Baseline Emissions	3-4
3.2.3	2017 Illustrative Emissions Case for the Final CSAPR Update Emissions
Budgets	3-8
3.2.4	Effect of Emissions Reductions on Downwind Receptors	3-9
3.3	Post-Processing of Air Quality Modeling for Benefits Calculations	3-11
3.3.1	Converting CAMx Ozone Outputs to Benefits Inputs	3-11
3.3.2	Converting CAMx PM2.5 Outputs to Benefits Inputs	3-12
3.4	Limitations	3-13
3.5	References	3-13
CHAPTER 4: COST, EMISSIONS, AND ENERGY IMPACTS	4-1
Overview	4-1
4.1	Regulatory Control Alternatives	4-1
4.2	Power Sector Modeling Framework	4-5
4.3	EPA's Power Sector Modeling of the Base Case and Three Regulatory Control
Alternatives	4-8
4.3.1 EPA's IPMv.5.15 Base Cases for the CSAPR Update	4-8
4.3.2. Methodology for Evaluating the Regulatory Control Alternatives	4-11
4.4	Estimated Impacts of the Regulatory Control Alternatives	4-16
4.4.1	Emission Reduction Assessment	4-16
4.4.2	Compliance Cost Assessment	4-19
4.4.3	Impacts on Fuel Use, Prices and Generation Mix	4-21
4.5	Social Costs	4-26
4.6	Limitations	4-27
4.7	References	4-28
APPENDIX 4A: COST, EMISSIONS, AND ENERGY IMPACTS OF FINAL CSAPR
UPDATE BUDGETS	4A-1
CHAPTER 5: ESTIMATED HUMAN HEALTH BENEFITS AND CLIMATE CO-
BENFITS	5-1
v

-------
5.1	Introducti on	5-1
5.2	Estimated Human Health Benefits	5-2
5.2.1	Health Impact Assessment for Ozone and PM2.5	5-3
5.2.1.1	Mortality Effect Coefficients for Short-term Ozone Exposure	5-7
5.2.1.2	PM2.5 Mortality Effect Coefficients for Adults and Infants	5-10
5.2.2	Economic Valuation for Health Benefits	5-14
5.2.3	Health Benefit Estimates for Ozone	5-16
5.2.4	Health Benefit Estimates for PM2.5	5-17
5.2.5	Updated Methodology in the Final RIA	5-18
5.2.6	Estimated Health Benefits Results	5-19
5.3	Estimated Climate Co-Benefits from CO2	5-29
5.3.1	Climate Change Impacts	5-30
5.4	Combined Health Benefits and Climate Co-Benefits Estimates	5-37
5.5	Unquantified Benefits and Co-benefits	5-39
5.5.2	Additional NO2 Health Co-Benefits	5-41
5.5.4	Additional NO2 Welfare Co-Benefits	5-42
5.5.5	Ozone Welfare Benefits	5-43
5.5.6	PM2.5 Visibility Impairment Co-Benefits	5-43
5.6	References	5-43
CHAPTER 6: ECONOMIC IMPACTS	6-1
Overview	6-1
6.1	Impacts on Small Entities	6-1
6.1.1	Identification of Small Entities	6-3
6.1.2	Overview of Analysis and Results	6-6
6.1.2.1	Methodology for Estimating Impacts of the CSAPR Update on Small
Entities	6-6
6.1.2.2	Results	6-8
6.1.3	Summary of Small Entity Impacts	6-11
6.2	Unfunded Mandates Reform Act	6-11
6.2.1	Identification of Government-Owned Entities	6-13
6.2.2	Overview of Analysis and Results	6-13
6.2.2.1	Methodology for Estimating Impacts of the CSAPR Update on
Government Entities	6-14
6.2.2.2	Results	6-16
6.2.3	Summary of Government Entity Impacts	6-18
6.3	Employment	6-18
6.3.1	Economic Theory and Employment	6-19
6.3.1.1	Current State of Knowledge Based on the Peer-Reviewed Literature .. 6-22
6.3.1.2	Regulated Sector	6-23
6.3.1.3	Economy-Wide	6-23
6.1.4 Labor Supply Impacts	6-23
6.3.1.5 Conclusion	6-24
6.3.2	Recent Employment Trends	6-24
6.3.2.1	Electric Power Generation	6-24
6.3.2.2	Fossil Fuel Extraction	6-25
vi

-------
6.3.3 Power and Fuels Sector Direct Employment Impacts	6-27
6.3.3.1	Methods Used to Estimate Changes in Employment in Electricity
Generation and Fuel Supply	6-30
6.3.3.2	Estimates of the Changes in Employment in Electricity Generation and Fuel
Supply	6-31
6.4 References	6-32
CHAPTER 7: COMPARISON OF BENEFITS AND COSTS	7-1
Overview	7-1
7.1 Results	7-1
vii

-------
LIST OF TABLES
Table ES-1. Projected 2017* EGU Emissions Reductions of NOxand CO2 with the
CSAPR Update NOx Emission Budgets and More and Less Stringent Alternatives
(Tons)"	ES-
Table ES-2. Cost Estimates (2011$) for CSAPR Update and More and Less Stringent
Alternatives	ES-9
Table ES-3. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the CSAPR Update and More and Less Stringent Alternatives
for 2017*	ES-12
Table ES-3. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for 2017
(millions of 2011$) *	ES-14
Table ES-4. Combined Health Benefits and Climate Co-Benefits for the CSAPR Update
and More and Less Stringent Alternatives for 2017 (millions of 2011$)*	ES-15
Table ES-5. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for 2017
(millions of 2011$) *	ES-16
Table ES-6. Unquantified Health and Welfare Co-benefits Categories	ES-16
Table ES-7. Total Costs, Total Monetized Benefits, and Net Benefits of the CSAPR
Update and More and Less Stringent Alternatives in 2017 for U.S. (millions of
201 l$)a'b'c'd	ES-19
Table 2-1. Total Net Summer Electricity Generating Capacity by Energy Source, 2000
and 2014	2-3
Table 2-2. Net Generation in 2000 and 2014 (Trillion kWh = TWh)	2-6
Table 2-3. Coal and Natural Gas Generating Units, by Size, Age, Capacity, and
Average Heat Rate in 2014	2-7
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2000 and 2014 (billion
kWh) 11
Table 3-1. 2011 Base Year and 2017 Baseline NOx and VOC Emissions by Sector
(thousand tons)	3-8
Table 4-1 Illustrative NOx Ozone Season Emission Budgets (Tons) Evaluated in this
RIA 4-4
Vlll

-------
Table 4-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives	4-13
Table 4-3. Summary of Methodology for Calculating Compliance Costs Estimated
Outside of IPVI for CSAPR Update, 2017 (2011$)	4-16
Table 4-4. EGU Ozone Season NOx Emissions and Emission Changes (thousand tons)
for the Base Case and the Regulatory Control Alternatives	4-17
Table 4-5. EGU Annual Emissions and Emissions Changes for NOx, SO2 and CO2 for
the Regulatory Control Alternatives	4-18
Table 4-6. Compliance Cost Estimates (millions of 2011$) for the Regulatory Control
Alternatives	4-19
Table 4-7. 2017 Projected Power Sector Coal Use for the Base Case and the Regulatory
Control Alternatives	4-21
Table 4-8. 2017 Projected Power Sector Natural Gas Use for the Base Case and the
Regulatory Control Alternatives	4-22
Table 4-9. 2017 Projected Minemouth and Power Sector Delivered Coal Price for the Base
Case and the Regulatory Control Alternatives	4-22
Table 4-10. 2017 Projected Henry Hub and Power Sector Delivered Natural Gas Price for
the Base Case and the Regulatory Control Alternatives	4-22
Table 4-11. 2017 Projected Generation by Fuel Type for the Base Case and the Regulatory
Control Alternatives	4-23
Table 4-12. 2020 Projected Capacity by Fuel Type for the Base Case and the Regulatory
Control Alternatives	4-23
Table 4-13. Average Retail Electricity Price by Region for the Base Case and the Regulatory
Control Alternatives, 2017	4-24
Table 4A-1 CSAPR Update NOx Ozone Season Emission Budgets (Tons)	4A-1
Table 4A-2. EGU Ozone Season NOx Emissions and Emission Changes (thousand tons)
for the Base Case and the CSAPR Update	4A-2
Table 4A-3. EGU Annual Emissions and Emissions Changes for NOx, SO2 and CO2 for
the CSAPR Update	4A-2
Table 4A-5. 2017 Projected Power Sector Coal Use for the Base Case and the CSAPR
Update 4A-3
Table 4A-6. 2017 Projected Power Sector Natural Gas Use for the Base Case and the
CSAPR Update	4A-3
ix

-------
Table 4A-9. 2017 Projected Generation by Fuel Type for the Base Case and the CSAPR
Update	4A-4
Table 4A-10. 2020 Projected Capacity by Fuel Type for the Base Case and the CSAPR
Update	4A-4
Table 4A-11. Average Retail Electricity Price by Region for the Base Case and the CSAPR
Update, 2017	4A-4
Table 5-1. Human Health Effects of Ambient Ozone and PM2.5	5-5
Table 5-2. Summary of Ozone and PM2.5 Benefit-per-Ton Estimates Based on Air
Quality Modeling in 2017 (2011$)*	5-19
Table 5-3. Emission Reductions of Criteria Pollutants in CSAPR Update States for the
CSAPR Update and More and Less Stringent Alternatives in 2017 (thousands of
short tons)*	5-20
Table 5-4. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for 2017
(millions of 2011$) *	5-20
Table 5-5. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the CSAPR Update and More and Less Stringent Alternatives
for 2017*	5-21
Table 5-7. Social Cost of CO2, 2015-2050 (in 2011$ per metric ton)*	5-36
Table 5-8. Estimated Global Climate Co-benefits of CO2 Reductions for the CSAPR
Update and More and Less Stringent Alternatives for 2017 (millions of 2011$)*	5-37
Table 5-9. Combined Health Benefits and Climate Co-Benefits for the CSAPR update
and More and Less Stringent Alternatives for 2017 (millions of 2011$)*	5-39
Table 5-10. Unquantified Health and Welfare Benefit and Co-benefit Categories	5-40
Table 6-1. SBA Size Standards by NAICS Code	6-5
Table 6-2. Projected Impact of the CSAPR Update on Small Entities in 2017	6-9
Table 6-3. Summary of Distribution of Economic Impacts of the CSAPR Update on Small
Entities in 2017	6-10
Table 6-4. Incremental Annual Costs under the CSAPR Update Summarized by Ownership
Group and Cost Category in 2017 (2011$ millions)	6-11
Table 6-5. Summary of Potential Impacts on Government Entities under the CSAPR Update
in 2017	6-16
x

-------
Table 6-6. Incremental Annual Costs under the CSAPR Update Summarized by Ownership
Group and Cost Category (2011$ millions) in 2017	6-17
Table 6-7. Annual Net Employment Impacts for Power and Fuels Sectors in 2017 & 2020.... 6-32
Table 7-1. Total Costs, Total Monetized Benefits, and Net Benefits of the CSAPR
Update and More and Less Stringent Alternatives in 2017 for U.S. (millions of
201 l$)a'b'c'd	7-2
Table 7-2. Projected 2017* Changes in Emissions of NOxand CO2 with the proposed
NOx Emissions Budgets and More or Less Stringent Alternatives (Tons)	7-3
XI

-------
LIST OF FIGURES	
Figure ES-1. States Covered by the Cross-State Air Pollution Rule Update	ES-3
Figure 1-1. States Covered by the Cross-State Air Pollution Rule Update	1-5
Figure 2-1 . National New Build and Retired Capacity (MW) by Fuel Type, 2000-2014... 2-4
Figure 2-2. Regional Differences in Generating Capacity (MW), 2014	2-5
Figure 2-3. Cumulative Distribution in 2012 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 for Three Major End-Use
Categories	2-13
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for Major
End-Use Categories, With Inflation Indices	2-13
Figure 2-7. Real National Average Electricity Prices for Three Major End-Use
Categories (including taxes), 1960-2014 (2011$)	2-14
Figure 2-8. Relative Change in Real National Average Electricity Prices (2011$) for
Three Major End-Use Categories	2-15
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in
National Average Real Price per MMBtu Delivered to EGU	2-16
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since
2000 17
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation
Intensity Since 2000	2-18
Figure 2-12. Status of State Electricity Industry Restructuring Activities	2-19
Figures 2-13 & 2-14. Capacity and Generation Mix by Ownership Type, 2000 & 2014	2-21
Figure 3-1. National air quality modeling domain	3-2
Figure 5-1. Monetized Health Benefits of CSAPR update for 2017 *	5-22
Figure 5-2. Percentage of Adult Population (age 30+) by Annual Mean PM2.5 Exposure
in the Baseline used for the Air Quality Analysis in Chapter 3	5-28
Figure 5-3. Cumulative Distribution of Adult Population (age 30+) by Annual Mean
PM2.5 Exposure in the Baseline used for the Air Quality Analysis in Chapter 3	5-29
xii

-------
Figure 6-1. Electric Power Industry Employment	6-25
Figure 6-2. Coal Production Employment	6-25
Figure 6-3 Oil and Gas Extraction Employment	6-27
xiii

-------
EXECUTIVE SUMMARY
Overview
The EPA promulgated the original Cross-State Air Pollution Rule (original CSAPR) on
August 8, 2011 (U.S. EPA, 2011), to address interstate transport of ozone pollution under the
1997 Ozone NAAQS.1 The primary purpose of this Cross-State Air Pollution Rule Update
(CSAPR Update) is to address interstate air quality impacts with respect to the 2008 Ozone
National Ambient Air Quality Standards (NAAQS). Specifically, this CSAPR Update will
reduce ozone season emissions of oxides of nitrogen (NOx) in 22 eastern states that can be
transported downwind as NOx or, after transformation in the atmosphere, as ozone and
contribute significantly to nonattainment or interfere with maintenance of the 2008 Ozone
NAAQS in downwind states. For the 22 eastern states affected by this rule, the EPA is issuing
Federal Implementation Plans (FIPs) that generally provide updated CSAPR NOx ozone season
emission budgets for electric generating units (EGUs) and is implementing these emission
budgets via modifications to the CSAPR NOx ozone season allowance trading program. The
CSAPR Update is also intended to respond to the D.C. Circuit's July 28, 2015, remand of certain
CSAPR NOx ozone season emission budgets to the EPA for reconsideration. This Regulatory
Impact Analysis (RIA) presents the health and welfare benefits and climate co-benefits of the
CSAPR Update, and compares the benefits to the estimated costs of implementing the CSAPR
Update for the 2017 analysis year. This RIA also reports certain other impacts of the CSAPR
Update, such as its effect on employment and energy prices. This executive summary explains
the analytic approach taken in the RIA and summarizes the RIA results.
ES.l Identifying Needed Emission Reductions
As described in the preamble for the CSAPR Update, CSAPR provides a 4-step
framework for addressing the requirements of CAA section 110(a)(2)(D)(i)(I) (sometimes called
the "good neighbor" provision) for ozone or fine particulate matter (PM2.5) standards: (1)
identifying downwind receptors that are expected to have problems attaining or maintaining
clean air standards (i.e., NAAQS); (2) determining which upwind states contribute to these
1 CSAPR also addressed interstate transport of fine particulate matter (PM2 5) under the 1997 and 2006 PM2 5
NAAQS.
ES-1

-------
identified problems in amounts sufficient to "link" them to the downwind air quality problems;
(3) for states linked to downwind air quality problems, identifying upwind emissions that
significantly contribute to downwind nonattainment or interfere with downwind maintenance of
a standard; and (4) reducing the identified upwind emissions via regional allowance trading
programs, for states that are found to have emissions that significantly contribute to
nonattainment or interfere with maintenance of the NAAQS downwind. The CSAPR Update
applies this 4-step framework to update CSAPR to address interstate emissions transport for the
2008 ozone NAAQS in the eastern United States.
Application of the first two steps of the 4-step framework with respect to the 2008 ozone
NAAQS provides the analytic basis for finding that ozone season emissions in 22 eastern states2
affect the ability of downwind states to attain and maintain the 2008 ozone NAAQS. Figure ES-1
shows these states, which are affected by this rule. More details on the methods and results of
applying this process can be found in the preamble for this CSAPR Update, and in Chapter 4 of
this RIA.
2 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia, West Virginia, and
Wisconsin.
ES-2

-------
| States covered by the Cross-State Air Pollution Update Rule (22 states)
Figure ES-1. States Covered by the Cross-State Air Pollution Rule Update
Applying Step 3 of the 4-step framework, the CSAPR Update quantifies EGU NOx
emission budgets for these 22 eastern states. A state's CSAPR Update NOx ozone season
emission budget represents the quantity of remaining EGU NOx emissions after reducing those
emissions that significantly contribute to downwind nonattainment or interfere with maintenance
of the 2008 Ozone NAAQS in an average year.3 These updated CSAPR NOx emissions budgets
were developed considering EGU NOx reductions that are achievable for the 2017 ozone
season.4 In calculating these budgets,the EPA applied the CSAPR 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. The EPA is finalizing EGU
3	For example, assuming no abnormal variation in electricity supply due to events such as abnormal meteorology.
4	Non-EGU NOx emission control measures and reductions are not included in this CSAPR Update.
ES-3

-------
NOx ozone season emission budgets developed using uniform control stringency represented by
$1,400 per ton control costs (2011$).5 Applying Step 4 of the 4-step framework, the EPA is
finalizing FIPs for each of the 22 states that require affected EGUs to participate in the CSAPR
NOx ozone season allowance trading program subject to the final emission budgets.
For this RIA, in order to implement the OMB Circular A-4 requirement to assess at least
one less stringent and one more stringent alternative to a rulemaking, the EPA is also analyzing
EGUNOx ozone season emission budgets developed using uniform control stringency
represented by $800 per ton (2011$) and emission budgets developed using uniform control
stringency represented by $3,400 per ton (2011$).6 The results of these analysis are summarized
in section ES.3 below.
ES.2 Baseline and Analysis Years
The CSAPR Update sets forth the requirements for 22 eastern states to reduce their
significant contribution to downwind nonattainment or interference with maintenance of the
2008 ozone NAAQS. To evaluate the benefits and costs of this regulation, it is important to first
establish a baseline projection of both emissions and air quality in the analysis years of 2017 and
2020, taking into account currently on-the-books Federal regulations,7 substantial Federal
regulatory updates, enforcement actions, state regulations,8 population, and where possible,
5	The basis for identifying this level of uniform control stringency is discussed in section VLB of the preamble to
the CSAPR Update rule and in the EGU NOx Mitigation Strategies Final Rule TSD. Further, the basis for finalizing
EGU NOx emission budgets developed using this level of uniform NOx control stringency is described in section
VI. C of the preamble to the CSAPR Update Rule.
6	The bases for identifying these levels of uniform control stringency are discussed in section VLB of the preamble
to the CSAPR Update rule.
7	The proposed CSAPR Update used an IPM base case that included the EPA's Clean Power Plan (CPP). Many
commenters requested that the agency not include the Clean Power Plan in the 2017 EGU projections. For the
reasons discussed in Section V.B of the preamble, we have excluded the CPP from the base case modeling for this
rule.
8	After the emissions and air quality modeling for the final rule were underway, Pennsylvania published a new
RACT rule that requires EGU and non-EGU NOx reductions starting on January 1, 2017. The EPA was unable to
explicitly include this final state rule in the baseline emission projections for the final CSAPR Update Rule.
However, the EPA recognizes that the implementation of this final state rule will precede the first control period for
the final CSAPR Update Rule. The agency quantifies costs and benefits of the CSAPR Update in this RIA that are
incremental to Pennsylvania's RACT rule.
ES-4

-------
economic growth. Establishing this baseline for the analysis then allows us to estimate the
incremental costs and benefits of the additional emission reductions that will be achieved by the
CSAPR Update 9
The analysis in this RIA focuses on benefits, costs and certain impacts in 2017. Certain
impacts in 2020, such as forecast emissions changes from the electricity sector, are also reported
in this RIA. The results from the analysis in support of the CSAPR Update that are reported in
this RIA are limited to these two analysis years. Other regulatory actions, including the 2015
ozone NAAQS, are expected to have a growing influence on the power sector in later years, as
explained below. For this reason, the EPA expects that most of the CSAPR Update's influence
on emissions reductions will occur between 2017 and 2020.
Below is a list of some of the national rules reflected in the baseline. Chapters 3 and 4
provide additional explanation about which rules are acccounted for in the baseline as well as
other details about how the baseline was constructed for this RIA. For a more complete list of the
rules reflected in the air quality modeling, please see the Technical Support Document:
Preparation of Emissions Inventories for the Version 6.2, 2011 Emissions Modeling Platform
(U.S. EPA, 2015). For a list of those regulations reflected in the compliance and cost modeling
of the electricity sector, please see "EPA Base Case v.5.15 Using IPM Incremental
Documentation" August, 2015.10
•	Standards of Performance for Greenhouse Gas Emissions from New, Modified, and
Reconstructed Stationary Sources: Electric Utility Generating Units (U.S. EPA, 2015a)
•	Tier 3 Motor Vehicle Emission and Fuel Standards (U.S. EPA, 2014)
•	2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate
Average Fuel Economy Standards (U.S. EPA, 2012)
9	Note that this modeling platform does not include the Regional Haze Plan for Texas and Oklahoma, published
January 5, 2016. The EPA does not believe this rule would substantially affect ozone season NOx emissions in
2017, and therefore budgets determined for this rule.
10	http://www.epa.gov/powersectormodeling/html
ES-5

-------
•	Cross State Air Pollution Rule (CSAPR) (U.S. EPA, 2011)11
•	Mercury and Air Toxics Standards (MATS) (U.S. EPA, 201 la)12
•	Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and
Heavy-Duty Engines and Vehicles (U.S. EPA, 201 lb)13
•	C3 Oceangoing Vessels (U.S. EPA, 2010)
•	Reciprocating Internal Combustion Engines (RICE) NESHAPs (U.S. EPA, 2010a)
•	Regulation of Fuels and Fuel Additives: Modifications to Renewable Fuel Standard
Program (RFS2) (U.S. EPA, 2010b)
•	Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel
Economy Standards; for Model-Year 2012-2016 (U.S. EPA, 2010c)
•	Hospital/Medical/Infectious Waste Incinerators: New Source Performance Standards and
Emission Guidelines: Amendments (U.S. EPA, 2009)
•	Emissions Standards for Locomotives and Marine Compression-Ignition Engines (U.S.
EPA, 2008a)
•	Control of Emissions for Nonroad Spark Ignition Engines and Equipment (U.S. EPA,
2008b)
11	On July 28, 2015, the D.C. Circuit issued its opinion regarding CSAPR on remand from the Supreme Court, EME
Homer City Generation, L.P., v. EPA, No. 795 F.3d 118, 129-30, 138 (EME Homer City IT). Unlike the modeling for
the proposed rule, which was conducted prior to the D.C. Circuit's issuance of EME Homer City II, this projected
base case accounts for compliance with the original CSAPR by including as constraints all original CSAPR
emission budgets with the exception of remanded phase 2 NOx ozone season emission budgets for 11 states and
phase 2 NOx ozone season emission budgets for four additional states that were finalized in the original CSAPR
supplemental rule. Specifically, to reflect original CSAPR ozone season NOx requirements, the modeling includes
as constraints the original CSAPR NOx ozone season emission budgets for 10 states ~ Alabama, Arkansas, Georgia,
Illinois, Indiana, Kentucky, Louisiana, Mississippi, Missouri, and Tennessee. For further discussion, see Chapter 4
of this RIA.
12	InMichigan v. EPA, the Supreme Court reversed on narrow grounds a portion of the D.C. Circuit decision
upholding the MATS rule, finding that the EPA erred by not considering cost when determining that regulation of
EGUs was "appropriate" pursuant to CAA section 112(n)(l). 135 S.Ct. 192 (2015). On remand, the D.C. Circuit left
the MATS rule in place pending the EPA's completion of its cost consideration in accordance with the Supreme
Court's decision. White Stallion Energy Ctr. v. EPA, No. 12-1100 (Dec. 15, 2015) (order remanding MATS rule
without vacatur). The EPA finalized its supplemental action responding to the Supreme Court's Michigan decision
on April 14, 2016. 81 FR 24420 (April 25, 2016). The MATS rule is currently in place.
13	This rule is Phase 1 of the Heavy Duty Greenhouse Gas Standards for New Vehicles and Engines (76 FR 57106,
September 15, 2011). Phase 2 of the Heavy Duty Greenhouse Gas Standards for New Vehicles and Engines (80 FR
40138, July 13, 2015) is not included because the rulemaking was not finalized in time to include in this analysis.
ES-6

-------
•	NOx Emission Standard for New Commercial Aircraft Engines (U.S. EPA, 2005)
•	Regional Haze Regulations and Guidelines for Best Available Retrofit Technology
Determinations (U.S. EPA, 2005a)
With regard to the increment of impacts attributable to the CSAPR Update and the original
CSAPR, the EPA does not believe that the costs and benefits for the original CSAPR and the
CSAPR Update are entirely additive. The EPA recognizes that the majority of the benefits of the
original CSAPR were derived from reductions in SO2 and annual NOx emissions, and the
benefits of the CSAPR Update are primarily based on ozone-season NOx emissions reductions.
However, five years have passed between promulgation of the original CSAPR and the CSAPR
Update, and the two rules have different baselines. In the intervening five years, changes in the
power sector that are independent of these rules, such as changes in fuel costs and electricity
markets as well as other federal and state level actions, which creates challenges when estimating
the sum of the costs and benefits of these two rules. In addition, implementation of the original
CSAPR was delayed such that its two phases were implemented as phase I - limits to be met by
2015, and phase II - limits to be met by 2017. The reductions estimated for the CSAPR Update
in 2017, given that it replaces remanded original CSAPR budgets, may overlap with reductions
that would have otherwise occurred for phase II. However, the benefits and costs of CSAPR are
still notable given the enduring original CSAPR ozone season NOx budgets, annual NOx
budgets, and SO2 budgets. While the EPA did remove the remanded ozone season NOx budgets
for three states, two of these states (North Carolina and South Carolina) remain subject to annual
NOx requirements. These original CSAPR budgets are all present in EPA's modeling of the
baseline and policy alternatives.
Also, EPA expects that most of the CSAPR Update's influence on emissions reductions
will occur between 2017 and 2020. We have excluded the CPP from the base case modeling for
this rule. The EPA does not anticipate significant interactions with the CPP and the near-term
ozone season EGU NOx emission reduction requirements under the CSAPR Update.
ES.3 Control Strategies and Emissions Reductions
The CSAPR Update requires EGUs in 22 eastern states to reduce interstate transport of
NOx emissions that significantly contribute to nonattainment or interfere with maintenance of
ES-7

-------
the 2008 ozone NAAQS. The CSAPR Update sets EGU NOx ozone season emission budgets
(allowable emission levels) for 2017 and future years. The CSAPR Update also finalizes FIPs
for each of the 22 states that require affected EGUs to participate in the CSAPR NOx ozone
season allowance trading program. The allowance trading program is the remedy in the FIP that
achieves the ozone season NOx emission reductions required by the CSAPR Update. The
allowance trading program essentially converts the EGU NOx emission budget for each of the 22
states subject to the FIP into a limited number of NOx ozone season allowances that, on a
tonnage basis, equal the state's ozone season emission budget.
The final CSAPR Update EGU NOx ozone season emission budgets for each state were
developed using uniform control stringency represented by $1,400 per ton of NOx reductions for
affected EGUs. Furthermore, this RIA analyzes regulatory control alternatives based on more
and less stringent state emission budgets developed using uniform control stringency represented
by $3,400 per ton and $800 per ton, respectively. As described in Chapter 4 the analysis in this
RIA uses illustrative budgets that differ somewhat from the finalized budgets for the CSAPR
Update, because the analysis for this RIA began before the budgets were finalized. Appendix 4A
reports the emissions reductions and costs of EPA's analysis of the CSAPR Update with the
finalized budgets.
The EPA analyzed ozone season NOx emission reductions from implementing the CSAPR
Update EGU NOx ozone season emission budgets using the Integrated Planning Model (IPM).
Table ES-1 shows the emission reductions expected from the CSAPR Update and the more and
less stringent alternatives analyzed. Included in the table are annual and seasonal NOx and
carbon dioxide (CO2) reductions over the contiguous U.S.
Table ES-1. Projected 2017* EGU Emissions Reductions of NOxand CO2 with the
CSAPR Update NOx Emission Budgets and More and Less Stringent
Alternatives (Tons)"

CSAPR Update
More Stringent
Alternative
Less Stringent
Alternative
NOx (annual)
75,000
79,000
27,000
NOx (ozone season)
61,000
66,000
27,000
CO2 (annual)
1,600,000
2,000,000
1,300,000
* The forecast of annual reductions of CO2 in 2017 is based on 2018 IPM direct model outputs.
" NOx emissions are reported in English (short) tons; CO2 is reported in metric tons. All estimates rounded to two
significant figures.
ES-8

-------
ES.4 Costs
In addition to emission reductions, the EPA estimated compliance costs associated with the
regulatory control alternatives. The compliance cost estimate represents the change in the cost of
supplying electricy under each regulatory control alternative. This change reflects both the
changes in electricity production costs resulting from application of NOx control strategies, as
well as differences in costs related to the small changes in the generation fuel mix projected to
occur as a result of compliance with the emissions budgets. The Agency uses the compliance
cost estimate from IPM as a proxy for social costs.
The estimate of the total cost of this CSAPR Update, therefore, is the combination of NOx
costs estimated by IPM and additional costs estimated outside of IPM. The cost estimates for the
CSAPR Update and more and less stringent alternatives are presented in Table ES-2. All costs
are in 2011 dollars.
Table ES-2. Cost Estimates (2011$) for CSAPR Update and More and Less Stringent
Alternatives
Alterantive
Annualized*
CSAPR Update
$68,000,000
More Stringent Alternative
$82,000,000
Less Stringent Alternative
$8,000,000
*Costs are annualized over the period 2017 through 2020 using the 4.77 percent discount rate used in IPM's
objective function for minimizing the net present value of the stream of total costs of electricity generation. An
explanation of the annualization of these costs can be found in Chapter 4 of this RIA. All estimates are rounded to
two significant figures.
ES.5 Benefits to Human Health and Welfare
Implementing this CSAPR Update is expected to reduce emissions of ozone season NOx.
In the presence of sunlight, NOx and VOCs can undergo a chemical reaction in the atmosphere
to form ozone. Reducing NOx emissions also reduces human exposure to ozone and the
incidence of ozone-related health effects, depending on local levels of volatile organic
compounds (VOCs). In addition, implementing the CSAPR Update is expected to reduce
emissions of NOx throughout the year. Because NOx is also a precursor to formation of ambient
PM2.5, reducing NOx emissions would also reduce human exposure to ambient PM2.5 throughout
the year and would reduce the incidence of PM2.5-related health effects. Finally, these emission
reductions would lower ozone and PM2.5 concentrations in regions beyond those subject to this
ES-9

-------
CSAPR Update, though this RIA does not account for benefits outside of the CSAPR Update 22-
state region. Additionally, although we do not have sufficient data to quantify these impacts in
this analysis, reducing emissions of NOx would also reduce ambient exposure to nitrogen
dioxide (NO2) and its associated health effects.
In this section, we provide an overview of the monetized ozone benefits and PM2.5-
related co-benefits estimated from NOx reductions for compliance with the CSAPR EGU NOx
ozone season emission budgets and for the more and less stringent alternatives. A full description
of the underlying data, studies, and assumptions is provided in the PM NAAQS RIA (U.S. EPA,
2012a) and Ozone NAAQS RIA (U.S. EPA, 2015b). The EPA does not view the projected
change in SO2 from IPM as a meaningful impact of the policy. Accordingly, this RIA does not
quantify S02-related PM2.5 co-benefits.
ES.5.1 Human Health Benefits and Climate Co-benefits
This analysis utilizes a "damage-function" approach in calculating benefits, which
estimates changes in individual health endpoints (specific effects that can be associated with
changes in air quality) and assigns values to those changes assuming independence of the values
for those individual endpoints. Because the EPA rarely has the time or resources to perform new
research to measure directly either health outcomes or their values for regulatory analyses, our
estimates are based on the best available methods of benefits transfer, which is the science and
art of adapting primary research from similar contexts to estimate benefits for the environmental
quality change under analysis. The benefit-per-ton approach we use in this RIA relies on
estimates of human health responses to exposure to ozone and PM obtained from the peer-
reviewed scientific literature. These estimates are used in conjunction with population data,
baseline health information, air quality data and economic valuation information to conduct
health impact and economic benefits assessments. These assessments form the key inputs to
calculating benefit-per-ton estimates. Thus, to develop estimates of benefits for this RIA, we are
transferring both the underlying health and economic information from previous studies and
information on air quality responses to emission reductions from other air quality modeling.
To perform the benefits transfer in this RIA we follow a "benefit-per-ton" approach to
estimating the ozone and PM2.5 benefits. Benefit-per-ton approaches apply an average benefit-
ES-10

-------
per-ton derived from modeling of benefits of specific air quality scenarios to estimates of
emission reductions for scenarios where no air quality modeling is available. The benefit-per-ton
values used in this RIA were estimating using air quality modeling conducted specifically for
this RIA. The baseline air quality modeling used to estimate the benefit-per-ton values does not
account for the Pennsylvania RACT, and the policy case is the CSAPR Update with the
illustrative budgets described in Chapter 4. More information on these approaches is available in
Chapter5 of the RIA.
The Health Impact Assessment (HIA) for ozone and PM2.5, discussed further in Chapter 5
of this RIA, quantifies the changes in the incidence of adverse health impacts resulting from
changes in human exposure to ozone and PM2.5. We use the environmental Benefits Mapping
and Analysis Program - Community Edition (BenMAP-CE) (version 1.1) to systematize health
impact analyses by applying a database of key input parameters, including population
projections, health impact functions, and valuation functions (US EPA, 2016). For this
assessment, the HIA is limited to those health effects that are directly linked to ambient ozone
and PM2.5 concentrations. Table ES-3 provides national summaries of the reductions in estimated
health incidences associated with the final CSAPR EGU NOx ozone season emission budgets
and for more and less stringent alternatives for 2017.
ES-11

-------
Table ES-3. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the CSAPR Update and More and Less Stringent
Alternatives for 2017*


More
Less

CSAPR
Stringent
Stringent
Ozone-related Health Effects
Update
Alternative
Alternative
Avoided Premature Mortality



Smith et al. (2009) (all ages)
21
23
9
Zanobetti and Schwartz (2008) (all ages)
60
65
26
Avoided Morbidity



Hospital admissions—respiratory causes (ages > 65)
59
64
26
Emergency room visits for asthma (all ages)
240
250
100
Asthma exacerbation (ages 6-18)
67,000
73,000
30,000
Minor restricted-activity days (ages 18-65)
170,000
180,000
75,000
School loss days (ages 5-17)
56,000
60,000
25,000
PlVhs-related Health Effects
Avoided Premature Mortality



Krewski et al. (2009) (adult)
10
11
3.7
Lepeule et al. (2012) (adult)
23
25
8.4
Woodruff et al. (1997) (infant)
<1
<1
<1
Avoided Morbidity



Emergency department visits for asthma (all ages)
6.1
6.5
2.2
Acute bronchitis (age 8-12)
15
15
5.2
Lower respiratory symptoms (age 7-14)
180
190
67
Upper respiratory symptoms (asthmatics age 9-11)
260
280
95
Minor restricted-activity days (age 18-65)
7,500
7,900
2,700
Lost work days (age 18-65)
1,300
1,300
450
Asthma exacerbation (age 6-18)
270
290
98
Hospital admissions—respiratory (all ages)
2.8
2.9
1.0
Hospital admissions—cardiovascular (age >18)
3.8
4.0
1.4
Non-Fatal Heart Attacks (age >18)



Peters et al. (2001)
12
13
4.3
Pooled estimate of 4 studies
1.3
1.4
0.46
* All estimates are rounded to whole numbers with two significant figures. Co-benefits for ozone are based on ozone
season NOx emissions. In general, the 95th percentile confidence interval for the health impact function alone ranges
from approximately ±30 percent for mortality incidence based on Krewski el al. (2009) and ±46 percent based on
Lepeule et al. (2012). The confidence intervals around the ozone mortality estimates are on the order of ± 60 percent
depending on the concentration-response function used.
There may be other indirect health impacts associated with reducing emissions, such as
occupational health exposures. We refer the reader to Chapter 5 of this RIA, as well as to the
Ozone NAAQS RIA (U.S. EPA, 2015b) and PM NAAQS RIA (U.S. EPA, 2012a) for more
information regarding the epidemiology studies and risk coefficients applied in this analysis.
ES-12

-------
Co-benefits of the CSAPR Update come from reducing emissions of CO2. Chapter 5 of this
RIA provides a brief overview of the 2009 Endangerment Finding and climate science
assessments released since then. Chapter 5 also provides information regarding the economic
valuation of CO2 using the social cost of carbon (SC-CO2), a metric that estimates the monetary
value of impacts associated with marginal changes in CO2 emissions in a given year.
ES.5.2 Combined Health Benefits and Climate Co-Benefits Estimates
In this analysis we were able to monetize the estimated benefits associated with the
reduced exposure to ozone and PM2.5 and co-benefits of decreased emissions of CO2.
Specifically, we estimated combinations of health benefits at discount rates of 3 percent and 7
percent (as recommended by the EPA's Guidelines for Preparing Economic Analyses [U.S.
EPA, 2014] and OMB's Circular A-4 [OMB, 2003]) and climate co-benefits using four SC-CO2
estimates (the average SC-CO2 at each of three discount rates—5 percent, 3 percent, 2.5
percent—and the 95th percentile SC-CO2 at 3 percent as recommended in the current SC-CO2
technical support document (TSD) [U.S. EPA, 2015c]; see Chapter 5 of this RIA for more
details). In this analysis we were unable to monetize the co-benefits associated with reducing
exposure to NO2, as well as ecosystem effects and visibility impairment associated with
reductions in NOx.
Table ES-3 reports the ozone and PM2.5-related benefits for the CSAPR Update and the
more and less stringent alternatives for the 2017 analysis year. ES-4 provides the combined
health and climate benefits for the CSAPR Update and for more and less stringent alternatives
for the 2017 analysis year. In the table, ranges within the total benefits rows reflect multiple
studies upon which the estimates of premature mortality were derived.
ES-13

-------
Table ES-3. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for
	2017 (millions of 2011$) *	
Pollutant

CSAPR Update
More Stringent
Alternative
Less Stringent
Alternative
NOx (as Ozone)

$370 to $610
$400 to $650
$160 to $270
NOx (as PM2.5)
3% Discount Rate
7% Discount Rate
$93 to $210
$83 to $190
$98 to $220
$88 to $200
$34 to $75
$30 to $67
Total
3% Discount Rate
7% Discount Rate
$460 to $810
$450 to $790
$500 to $870
$490 to $850
$200 to $340
$190 to $330
* All estimates are rounded to two significant figures so numbers may not sum down columns. The health benefits
range is based on adult mortality functions (e.g., from Krewski et al. (2009) with Smith et al. (2009) to Lepeule et al.
(2012) with Zanobetti and Schwartz (2008)). The estimated monetized co-benefits do not include reduced health
effects from direct exposure to NO2, ecosystem effects or visibility impairment. All fine particles are assumed to
have equivalent health effects. The CSAPR Update values, the more and less stringent alternatives were all
calculated using a benefits per ton approach. The monetized co-benefits incorporate the conversion from precursor
emissions to ambient fine particles and ozone. Benefits for ozone are based on ozone season NOx emissions. Ozone
benefits occur in analysis year, so they are the same for all discount rates. PM2 5 benefits are based on annual NOx
emissions and the nitrate-only fraction of PM2 5. In general, the confidence intervals around the ozone mortality
estimates are on the order of ± 60 percent depending on the concentration-response function used. The 95th
percentile confidence interval for monetized PM2 5 benefits ranges from approximately -90 percent to +180 percent
of the central estimates based on Krewski et al. (2009) and Lepeule et al. (2012)..
ES-14

-------
Table ES-4. Combined Health Benefits and Climate Co-Benefits for the CSAPR Update
and More and Less Stringent Alternatives for 2017 (millions of 2011$)*

Health and Climate Benefits
Climate Co-
SC-CO2 Discount Rate**
(Discount Rate Applied to Health Co-Benefits)
Benefits Only

3%
7%

CSAPR Update
5%
$480 to $830
$470 to $810
$19
3%
$530 to $880
$520 to $860
$66
2.5%
$560 to $910
$550 to $890
$100
3% (95th percentile)
$650 to $1,000
$640 to $980
$190
More Stringent Alternative
5%
$490 to $840
$480 to $820
$25
3%
$550 to $900
$540 to $880
$87
2.5%
$590 to $940
$580 to $920
$130
3% (95th percentile)
$710 to $1,100
$700 to $1,000
$250
Less Stringent Alternative
5%
$480 to $830
$470 to $810
$15
3%
$510 to $860
$500 to $840
$54
2.5%
$540 to $890
$530 to $870
$81
3% (95th percentile)
$610 to $960
$600 to $940
$150
*A11 estimates are rounded to two significant figures. Climate benefits are based on reductions in CO2 emissions.
Health benefits are based on benefit-per-ton estimates. Benefits for ozone are based on ozone season NOx
emissions. Ozone benefits occur in analysis year, so they are the same for all discount rates. The health benefits
reflect the sum of the ozone benefits and PM2 5 co-benefits and reflect the range based on adult mortality functions
(e.g., from Krewski et al. (2009) with Smith el al. (2009) to Lepeule el al. (2012) with Zanobetti and Schwartz
(2008)). The monetized health benefits do not include reduced health effects from direct exposure to NO2 as well as
ecosystem effects and visibility impairment associated with reductions in NOx. **As discussed in section 5.3, the
SC-CO2 estimates are calculated with four different values of a one metric ton reduction.
Table ES-5 summarizes the national monetized ozone-related and PM-related health
benefits estimated to occur for the CSAPR Update and two regulatory control alternatives for the
2017 analysis year using discount rates of 3 percent (non-fatal heart attacks quantified using
Peters et al. (2001)) and 7 percent (non-fatal heart attacks quantified using a pooled estimate that
includes Pope et al. (2006)).
ES-15

-------
Table ES-5. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for
	2017 (millions of 2011$) *	
Pollutant

CSAPR Update
More Stringent
Alternative
Less Stringent
Alternative
NOx (as Ozone)

$370 to $610
$400 to $650
$160 to $270
NOx (as PM2.5)
3% Discount Rate
7% Discount Rate
$93 to $210
$83 to $190
$98 to $220
$88 to $200
$34 to $75
$30 to $67
Total
3% Discount Rate
7% Discount Rate
$460 to $810
$450 to $790
$500 to $870
$490 to $850
$200 to $340
$190 to $330
* All estimates are rounded to two significant figures so numbers may not sum down columns. The health benefits
range is based on adult mortality functions (e.g., from Krewski et al. (2009) with Smith et al. (2009) to Lepeule et al.
(2012) with Zanobetti and Schwartz (2008)). The estimated monetized co-benefits do not include reduced health
effects from direct exposure to NO2, ecosystem effects, or visibility impairment. All fine particles are assumed to
have equivalent health effects, but the benefit-per-ton estimates vary depending on the location and magnitude of
their impact on PM2 5 levels, which drive population exposure. The CSAPR Update values, the more and less
stringent alternatives were all calculated using the benefits per ton approach based on the final modeling scenario.
The monetized co-benefits incorporate the conversion from precursor emissions to ambient fine particles and ozone.
Benefits for ozone are based on ozone season NOx emissions. Ozone co-benefits occur in analysis year, so they are
the same for all discount rates, and are based on annual NOx emissions and the nitrate-only fraction of PM2 5. In
general, the 95th percentile confidence interval for monetized PM2 5 benefits ranges from approximately -90 percent
to +180 percent of the central estimates based on Krewski et al. (2009) and Lepeule et al. (2012). The confidence
intervals around the ozone mortality estimates are on the order of ± 60 percent depending on the concentration-
response function used.
ES.5.3 Unquantified Health and Welfare Co-Benefits
The monetized health co-benefits estimated in this RIA reflect a subset of co-benefits
attributable to the health effect reductions associated with ambient fine particles. Data, time, and
resource limitations prevented the EPA from quantifying the impacts to, or monetizing the co-
benefits from several important benefit categories, including reduced exposure to NO2, as well as
ecosystem effects, and reduced visibility impairment from reduced NOx emissions. These
benefits were unable to be quantified due to the absence of air quality modeling data for these
pollutants. This does not imply that there are no co-benefits associated with changes in exposures
to NO2 or changes in ecosystem effects and visibility impairments from NOx reduction; the
identified co-benefits are listed in Table ES-6 below, and discussed more fully in Chapter 5 of
this RIA.
Table ES-6. Unquantified Health and Welfare Co-benefits Categories
Category
Specific Effect
Effect Has
Been
Quantified
Effect Has
Been
Monetized
More Information
Improved Human Health

Asthma hospital admissions (all ages)
—
—
NO2 ISA1
ES-16

-------


Effect Has
Effect Has

Category
Specific Effect
Been
Quantified
Been
Monetized
More Information

Chronic lung disease hospital admissions (age >
65)
—
—
NO2 ISA1

Respiratory emergency department visits (all
ages)
—
—
NO2 ISA1
Reduced incidence of
morbidity from exposure
to NO2
Asthma exacerbation (asthmatics age 4-18)
—
—
NO2 ISA1
Acute respiratory symptoms (age 7-14)
—
—
NO2 ISA1
Premature mortality
—
—
NO2 ISA1'2-3

Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
—
—
NO2 ISA2-3
Improved Environment
Reduced visibility
Visibility in Class 1 areas
—
—
PM ISA1
impairment
Visibility in residential areas
—
—
PM ISA1
Reduced effects on
materials
Household soiling
—
—
PM ISA1-2
Materials damage (e.g., corrosion, increased
wear)
—
—
PM ISA2
Reduced effects from PM




deposition (metals and
organic s)
Effects on Individual organisms and ecosystems
—
—
PM ISA2

Visible foliar injury on vegetation
—
—
Ozone ISA1

Reduced vegetation growth and reproduction
—
—
Ozone ISA1

Yield and quality of commercial forest products
and crops
—
—
Ozone ISA1
Reduced vegetation and
ecosystem effects from
Damage to urban ornamental plants
—
—
Ozone ISA2
Carbon sequestration in terrestrial ecosystems
—
—
Ozone ISA1
exposure to ozone
Recreational demand associated with forest
aesthetics
—
—
Ozone ISA2

Other non-use effects


Ozone ISA2

Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity,
leaf-gas exchange, community composition)
—
—
Ozone ISA2

Recreational fishing
—
—
NOxSOxISA1

Tree mortality and decline
—
—
NOxSOxISA2
Reduced effects from
acid deposition
Commercial fishing and forestry effects
—
—
NOxSOxISA2
Recreational demand in terrestrial and aquatic
ecosystems
—
—
NOxSOxISA2

Other non-use effects


NOxSOxISA2

Ecosystem functions (e.g., biogeochemical
cycles)
—
—
NOxSOxISA2

Species composition and biodiversity in terrestrial
and estuarine ecosystems
—
—
NOxSOxISA2

Coastal eutrophication
—
—
NOxSOxISA2
Reduced effects from
nutrient enrichment
Recreational demand in terrestrial and estuarine
ecosystems
—
—
NOxSOxISA2

Other non-use effects


NOxSOxISA2

Ecosystem functions (e.g., biogeochemical
cycles, fire regulation)
—
—
NOxSOxISA2
Reduced vegetation




effects from ambient
exposure to NOx
Injury to vegetation from NOx exposure
—
—
NOxSOxISA2
1 We assess these co-benefits qualitatively due to data and resource limitations for this RIA. More information is contained in the
integrated science assessments (ISAs) for the proposed or final NAAQS standards cited.
ES-17

-------
2We assess these co-benefits qualitatively because we do not have sufficient confidence in available data or methods.
3 We assess these co-benefits qualitatively because current evidence is only suggestive of causality or there are other significant
concerns over the strength of the association.
ES.5 Results of Benefit-Cost Analysis
Below in Table ES-7, we present the primary costs and benefits estimates for 2017. Net
benefits are also presented, reflecting the benefits of implementing the EGU NOx emission
budgets for the affected 22 states via the final FIPs, minus the costs of achieving those emissions
reductions.
The guidelines of OMB Circular A-4 require providing comparisons of social costs and
social benefits at discount rates of 3 and 7 percent. The four different uses of discounting in the
RIA - (i) construction of annualized costs, (ii) adjusting the value of mortality risk for lags in
mortality risk decreases, (iii) adjusting the cost of illness for non-fatal heart attacks to adjust for
lags in follow up costs, and (iv) discounting climate co-benefits - are all appropriate. We explain
our discounting of benefits in Chapter 5 of the RIA, specifically the application of discount rates
of 3 and 7 percent to PIVh.s-related co-benefits and 2.5, 3, and 5 percent to climate co-benefits;
we explain our discounting of costs, in which we use a single discount rate of 4.77 percent, in
Chapter 4. Our estimates of net benefits represent the net value (in 2017) of benefits attributable
to emission reductions needed to implement the NOx emission budgets for each state.
ES-18

-------
Table ES-7. Total Costs, Total Monetized Benefits, and Net Benefits of the CSAPR
Update and More and Less Stringent Alternatives in 2017 for U.S. (millions of
2011$)a'b'c'd
CSAPR Update
More Stringent
Alternative
Alternative
Less Stringent
Alternative
Climate Co-Benefits
Air Quality Health Benefits
Total Benefits
Annualized Compliance
Costs
Net Benefits
Non-Monetized Benefits®
$66
$460 to $810
$530 to $880
$87
$500 to $870
$580 to $960
$82
$500 to $880
$54
$200 to $340
$250 to $400
$460 to $810
Non-monetized climate benefits
Reductions in exposure to ambient NO2 and SO2
Ecosystem benefits assoc. with reductions in emissions of NOx
$240 to $390
a Estimating multiple years of costs and benefits is limited for this RIA by data and resource limitations. As a result,
we provide compliance costs and social benefits in 2017, using the best available information to approximate
compliance costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of health benefits and climate co-benefits at a discount rate of 3 percent. See
Chapter 5 for additional detail and explanation. The costs presented in this table reflect compliance costs annualized
at a 4.77 percent discount rate and do not include monitoring, recordkeeping, and reporting costs, which are reported
separately. See Chapter 4 for additional detail and explanation.
0 All costs and benefits are rounded to two significant figures; columns may not appear to add correctly.
d Ozone and PM2.5 benefits from NOx emission reductions are for the 22-state region only.
z Non-monetized benefits descriptions are for all three alternatives and are qualitative.
ES.6 Analytical Changes Subsequent to the Proposal
Costs
The EPA's IPM modeling platform used to analyze this rule (v.5.15) is similar to the
version used to analyze the CSAPR Update proposal, and incorporates minor updates made
primarily in response to comments received on an August 4, 2015 Notice of Data Availability
and the proposed rule.
Unlike the modeling for the proposed rule, which was conducted prior to the D.C. Circuit's
issuance of EME Homer City //,14 the base case for the final rule accounts for compliance with
14 In EME Homer City II, the D.C. Circuit declared invalid the CSAPR phase 2 NOx ozone season emission budgets
of 11 states: Florida, Maryland, New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Texas,
Virginia, and West Virginia. Id. 795 F.3d at 129-30, 138. The court remanded those budgets to the EPA for
reconsideration. Id. at 138. As a result, the EPA removed the original CSAPR phase 2 NOx ozone season emission
budgets as constraints for these 11 states in the 2017 IPM modeling.
ES-19

-------
the original CSAPR by including as constraints all original CSAPR emission budgets with the
exception of remanded phase 2 NOx ozone season emission budgets for 11 states and phase 2
NOx ozone season emission budgets for four additional states that were finalized in the original
CSAPR supplemental rule.15 Additionally, the Clean Power Plan (CPP) is not included in this
analysis. The base case results also reflect the recent Pennsylvania RACT, requires EGU NOx
reductions starting on January 1, 2017. For further discussion, see Chapter 4 of this RIA
Benefits
We modified our approach for estimating ozone and PIVh.s-related benefits between the
proposed and final rule. First, we calculated new ozone and PM2.5 benefit per ton estimates using
the results of an updated air quality modeling scenario. These air quality modeling predictions
more closely represent the selected policy option than the proposal modeling, but did not account
for either the final emissions budgets or the Pennsylvania RACT rule. Thus, the air quality
modeling scenario simulated a larger level of NOx emission reductions than the final policy
option implemented. Consequently, we applied ozone and PM2.5 benefit-per-ton values to
quantify the benefits of the final policy option and more and less stringent alternative options.
Second, when estimating the PIVh.s-related benefits for the final CSAPR rule we use a
benefit-per-ton value calculated using a nitrate-attributable PM2.5 benefit-per-ton estimate; the
proposal analysis used a total PM2.5 benefit per-ton-value. The EPA determined that,
considering the final CSAPR Update Rule illustrative emissions modeling results, using total
PM2.5 would incorrectly additionally account for the benefits of reduced sulfate and directly
emitted PM2.5 benefits, which the illustrative emissions modeling does not anticipate occurring.
Third, in this final rule the EPA estimated the benefits from the NOx emission reductions
only for the CSAPR states, whereas the proposed rule estimate national benefits from reductions
in NOx. The approach taken in the final rule likely underestimates total benefits to the extent that
15 The EPA acknowledges that the CSAPR NOx ozone season emission budgets for Iowa, Michigan, Oklahoma, and
Wisconsin ~ which were finalized in the original CSAPR Supplemental Rule (76 FR 80760, December 27, 2011) —
were linked to the same receptors that lead to the remand of other states' NOx ozone season emission budgets in
EME Homer City II.
ES-20

-------
downwind states in New England and certain Southeast states would likely improved air quality
from this rule.
ES.7 References
EME Homer City Generation, L.P., v. EPA, No. 795 F.3d 118, 129-30, 138 (EME Homer City
II)EME Homer City Generation, L.P., v. EPA, No. 11-1302, slip op. at 19, 36 (July 28,
2015).
U.S. EPA, 2016. Environmental Benefits Mapping and Analysis Program—Community Edition
vl.l. Research Triangle Park, NC. 
U.S. EPA, 2015. Preparation of Emissions Inventories for the Version 6.2, 2011 Emissions
Modeling Platform, Research Triangle Park, NC,
http://www.epa.gov/ttn/chief/emch/2011 v6/2011 v6_2_2017_2025_EmisMod_TSD_aug2015
.pdf.
U.S. EPA, 2015 a. Standards of Performance for Greenhouse Gas Emissions from New,
Modified, and Reconstructed Stationary Sources: Electric Utility Generating Units ,
http://www2.epa.gov/cleanpowerplan/carbon-pollution-standards-new-modified-and-
reconstructed-power-plants.
U.S. EPA, 2015b. Regulatory Impact Analysis of the Final Revisions to the National Ambient Air
Quality Standards for Ground-Level Ozone,
https://www3.epa.gov/ttn/ecas/docs/20151001ria.pdf.
U.S. EPA, 2015c. 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).
Available at: https://www.whitehouse.gov/omb/oira/social-cost-of-carbon.
U.S. EPA, 2014. Tier 3 Motor Vehicle Emission and Fuel Standards,
http://www3 .epa.gov/otaq/tier3 .htm.
U.S. Environmental Protection Agency (U.S. EPA). 2014d. Regulatory Impact Analysis of the
Proposed Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone.
EPA-452/P-14-006. Office of Air Quality Planning and Standards, Research Triangle Park,
NC. November. Available at .
Accessed June 4, 2015.
ES-21

-------
U.S. EPA, 2012. 2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and
Corporate Average Fuel Economy Standards, http://www3.epa.gov/otaq/climate/regs-light-
duty.htm#2017-2025.
U.S. EPA, 2011. Cross State Air Pollution Rule (CSAPR), http://www3.epa.gov/crossstaterule/.
U.S. EPA, 201 la . Mercury and Air Toxics Standards (MATS), http://www3.epa.gov/mats/.
U.S. EPA, 2011b, Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for
Medium- and Heavy-Duty Engines and Vehicles, http://www3.epa.gov/otaq/climate/regs-
heavy-duty.htm.
U.S. EPA, 2010. C3 Oceangoing Vessels, http://www3.epa.gov/otaq/oceanvessels.htm.
U.S. EPA, 2010a. Reciprocating Internal Combustion Engines (RICE) NESHAPs,
http ://www3. epa.gov/ttn/ atw/icengines/.
U.S. EPA, 2010b. Regulation of Fuels and Fuel Additives: Modifications to Renewable Fuel
Standard Program (RFS2), http://www2.epa.gov/renewable-fuel-standard-program.
U.S. EPA, 2010c. Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate
Average Fuel Economy Standards; for Model-Year 2012-2016,
http://www3.epa.gOv/otaq/climate/regs-light-duty.htm#2012-2016.
U.S. EPA, 2009. Hospital/Medical/Infectious Waste Incinerators: New Source Performance
Standards and Emission Guidelines: Amendments,
http://www3.epa.gov/airtoxics/129/hmiwi/rihmiwi.html.
U.S. Environmental Protection Agency (U.S. EPA). 2009. Integrated Science Assessment for
Particulate Matter (FinalReport). EPA-600-R-08-139F. National Center for Environmental
Assessment - RTP Division, Research Triangle Park, NC. December. Available at:
. Accessed June 4, 2015.
U.S. EPA, 2008a. Emissions Standards for Locomotives andMarine Compression-Ignition
Engines, http://www3.epa.gov/otaq/locomotives.htm.
U.S. EPA, 2008b. Control of Emissions for Nonroad Spark Ignition Engines and Equipment,
http ://www3 .epa.gov/nonroad/.
U.S. Environmental Protection Agency (U.S. EPA). 2008c. Integrated Science Assessment for
Sulfur Oxides—Health Criteria (FinalReport). National Center for Environmental
Assessment - RTP Division, Research Triangle Park, NC. September. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008d. Integrated Science Assessment for
Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
ES-22

-------
Assessment, Research Triangle Park, NC. July. Available at:
. Accessed June 4, 2015.
U.S. EPA, 2005, NOx Emission Standardfor New Commercial Aircraft Engines,
http://www3.epa.gov/otaq/aviation.htm.
U.S. EPA, 2005a. Regional Haze Regulations and Guidelines for Best Available Retrofit
Technology Determinations, http://www3.epa.gov/visibility/actions.html.
ES-23

-------
CHAPTER 1: INTRODUCTION AND BACKGROUND
Introduction
The EPA is finalizing this Cross-State Air Pollution Rule Update (CSAPR Update) to
address interstate transport of emissions of nitrogen oxides (NOx) that contribute significantly to
nonattainment or interfere with maintenance of the 2008 Ozone National Ambient Air Quality
Standard (NAAQS) in downwind states. The primary purpose of the CSAPR Update is to
address interstate air quality problems with respect to the 2008 ozone NAAQS. However, the
CSAPR Update is also intended to respond to the D.C. Circuit's July 28, 2015 remand of certain
CSAPR NOx ozone season emission budgets to the EPA for reconsideration. This Regulatory
Impact Analysis (RIA) presents the health and welfare benefits of the CSAPR Update, and
compares the benefits of the CSAPR Update to the estimated costs of implementing the rule in
2017. This RIA also reports certain other impacts of the CSAPR Update, such as its effect on
employment and energy prices. This chapter contains background information regarding the
CSAPR Update and an outline of the chapters of this RIA.
1.1 Background
The purpose of this rulemaking is to protect public health and welfare by reducing
interstate emission transport that significantly contributes to nonattainment, or interferes with
maintenance, of the 2008 ozone NAAQS in the eastern U.S. Ground-level ozone causes a variety
of negative effects on human health, vegetation, and ecosystems. In humans, acute and chronic
exposure to ozone is associated with premature mortality and a number of morbidity effects,
such as asthma exacerbation. Ozone exposure can also negatively impact ecosystems, for
example, by limiting tree growth. Studies have established that ozone occurs on a regional scale
(i.e., hundreds of miles) over much of the eastern U.S., with elevated concentrations occurring in
rural as well as metropolitan areas. 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).
Clean Air Act (CAA or the Act) section 110(a)(2)(D)(i)(I), sometimes called the "good
neighbor" provision, requires states to prohibit emissions that will contribute significantly to
nonattainment in, or interfere with maintenance by, any other state with respect to any primary or
1-1

-------
secondary NAAQS.16 The EPA promulgated the original Cross-State Air Pollution Rule (original
CSAPR) on August 8, 201117 to address interstate transport for the 1997 Ozone NAAQS and the
1997 and 2006 Fine Particulate matter (PM2.5) NAAQS.18 (See section III. A. 1.of the preamble to
the CSAPR Update for a discussion of CSAPR litigation and implementation.)
As described in the preamble for the CSAPR Update, CSAPR provides a 4-step
framework for addressing the requirements of the good neighbor provision for ozone or PM2.5
standards: (1) identifying downwind receptors that are expected to have problems attaining or
maintaining clean air standards (i.e., NAAQS); (2) determining which upwind states contribute
to these problems in amounts sufficient to "link" them to the downwind air quality problems; (3)
for states linked to downwind air quality problems, identifying upwind emissions that
significantly contribute to nonattainment or interfere with maintenance; and (4) for states that are
found to have emissions that significantly contribute to nonattainment or interfere with
maintenance of the NAAQS downwind, reducing the identified upwind NOx emissions via
regional allowance trading programs. In the CSAPR Update, the EPA applies this 4-step
framework to update CSAPR with respect to the 2008 ozone NAAQS. For 22 eastern states, this
CSAPR Update finalizes electric generating unit (EGU) NOx emission budgets representing the
quantity of remaining EGU NOx emissions after reducing those amounts that significantly
contribute to downwind nonattainment or interfere with maintenance of the 2008 ozone NAAQS
in an average year.19 The CSAPR Update finalizes FIPs for each of the 22 states that require
affected EGUs to participate in the CSAPR NOx ozone season allowance trading program
subject to these emission budgets. More details on the methods and results of applying this
framework can be found in the preamble for this CSAPR Update and in Chapter 4 of this RIA.
16	The EPA uses the term "states" to include the District of Columbia in this RIA.
17	See 76 FR 48208 (August 8, 2011)
18	CSAPR did not evaluate transport obligations for the 2008 ozone standard because the 2008 ozone NAAQS was
under reconsideration during the analytic work for CSAPR.
19	For example, assuming no abnormal variation in electricity supply due to events such as abnormal meteorology.
1-2

-------
1.2.1	Role of Executive Orders in the Regulatory Impact Analysis
Several statutes and executive orders apply to any public document. Certain analyses
required by these statutes and executive orders are presented in detail in Chapter 4, and all are
discussed in the preamble to the CSAPR Update. Below, we briefly discuss the requirements of
Executive Orders 12866 and 13563 and the guidelines of the Office of Management and Budget
(OMB) Circular A-4 (U.S. OMB, 2003).
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 emission reductions for
compliance with the CSAPR Update. OMB Circular A-4 requires analysis of at least one
potential alternative standard level more stringent than the CSAPR Update and one less stringent
than the CSAPR Update. This RIA evaluates the benefits, costs, and certain impacts of a more
and a less stringent alternative to the CSAPR Update.
1.2.2	Illustrative Nature of this Analysis
For the 22 CSAPR Update states, this rule finalizes EGU NOx emission budgets and
finalizes FIPs that require affected EGUs to participate in the CSAPR NOx ozone season
allowance trading program subject to these emission budgets. The EGU emission budgets
assessed in this RIA are illustrative of those that the EPA is finalizing. Further, implementation
via the CSAPR NOx ozone season allowance trading program provides utilities with the
flexibility to determine their own compliance path. This RIA develops and analyzes one possible
scenario for compliance with the illustrative EGU NOx emission budgets and possible scenarios
for EGU compliance with more and less stringent alternatives.
1.2.3	The Needfor Air Quality or Emissions Standards
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.
1-3

-------
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, setting an emissions standard (i.e., EGU NOx ozone
season emission budgets in this CSAPR Update) is a remedy to address an externality in which
firms emit pollutants, resulting in health and environmental problems without compensation for
those incurring the problems. Setting the emissions standard attempts to incentivize 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 Required Reductions
Application of the first two steps of the CSAPR framework (described above) with
respect to the 2008 ozone NAAQS provides the analytic basis for finding that ozone season
emissions in 22 eastern states20 affect the ability of downwind states to attain and maintain the
2008 ozone NAAQS. Figure 1-1 shows the covered states.
20 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia, West Virginia, and
Wisconsin.
1-4

-------
| 1 States covered by the Cross-State Air Pollution Update Rule (22 states)
Figure 1-1. States Covered by the Cross-State Air Pollution Rule Update
Applying Step 3 of the 4-step framework, the CSAPR Update quantifies EGU NOx
emission budgets for these 22 eastern states. A state's CSAPR Update NOx ozone season
emission budget represents the quantity of remaining EGU NOx emissions after reducing those
emissions that significantly contribute to downwind nonattainment or interfere with maintenance
of the 2008 Ozone NAAQS in an average year.21 These updated CSAPR NOx emissions budgets
were developed considering EGU NOx reductions that are achievable for the 2017 ozone
season.22 In calculating these budgets,the EPA applied the CSAPR 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. The EPA is finalizing EGU
21	For example, assuming no abnormal variation in electricity supply due to events such as abnormal meteorology.
22	Non-EGU NOx emission control measures and reductions are not included in this CSAPR Update.
1-5

-------
NOx ozone season emission budgets developed using uniform control stringency represented by
$1,400 per ton control costs (2011$). Applying Step 4 of the 4-step framework, the EPA is
finalizing FIPs for each of the 22 states that require affected EGUs to participate in the CSAPR
NOx ozone season allowance trading program subject to the final emission budgets.
For this RIA, in order to implement the OMB Circular A-4 requirement to assess at least
one less stringent and one more stringent alternative to a rulemaking, the EPA is also analyzing
EGUNOx ozone season emission budgets developed using uniform control stringency
represented by $800 per ton (2011$) and emission budgets developed using uniform control
stringency represented by $3,400 per ton (2011$).
1.2.2	States Covered by the CSAPR Update
For the 22 states affected by one of the FIPs finalized in the CSAPR Update, the EPA is
promulgating new FIPs with lower EGU NOx ozone season emission budgets to reduce
interstate transport for the 2008 ozone NAAQS. Of the 22 CSAPR Update states, 21 states23
have original CSAPR NOx ozone season FIP requirements with respect to the 1997 ozone
NAAQS. One state, Kansas, has newly added CSAPR NOx ozone season compliance
requirements under this CSAPR Update. One state for which the EPA proposed a FIP in the
proposed CSAPR Update rule, North Carolina, was found in the final air quality modeling not to
be linked to any downwind nonattainment or maintenance receptors. Therefore, the EPA is not
finalizing a FIP for North Carolina.
1.2.3	Regulated Entities
The CSAPR Update affects fossil fuel-fired EGUs in these 22 eastern states which are
classified as code 221112 by the North American Industry Classification System (NAICS) and
have a nameplate capacity of greater than 25 megawatts (MWe).
23 Alabama, Arkansas, Illinois, Indiana, Iowa, Kentucky, Louisiana, Maryland, Michigan, Mississippi, Missouri,
New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia, West Virginia, and Wisconsin.
1-6

-------
1.2.4 Baseline and Analysis Year
As described in the preamble, the EPA aligns implementation of the CSAPR Update with
relevant attainment dates for the 2008 ozone NAAQS, consistent with the D.C. Circuit's decision
North Carolina v. EPA24 The EPA's final 2008 Ozone NAAQS SIP Requirements Rule
established the attainment deadline of July 20, 2018, for ozone nonattainment areas currently
designated as Moderate.25 Because the attainment date falls during the 2018 ozone season, the
2017 ozone season will be the last full season from which data can be used to determine
attainment of the NAAQS by the July 20, 2018 attainment date. Therefore, the EPA has
identified achievable upwind emission reductions and aligned implementation of these
reductions, to the extent possible, for the 2017 ozone season.
The CSAPR Update sets forth the requirements for states to reduce their significant
contribution to downwind nonattainment and 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 in the analysis year of 2017, taking
into account currently on-the-books Federal regulations, substantial Federal regulatory CSAPR
updates, enforcement actions, state regulations, population, and where possible, economic
growth. 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 CSAPR
Update. Furthermore, the analysis in this RIA focuses on benefits, costs and certain impacts in
2017. Certain impacts in 2020, such as forecast emissions changes from the electricity sector, are
also reported in this RIA. The results from the analysis in support of the CSAPR Update that are
reported in this RIA are limited to these two analysis years. Other regulatory actions, including
the 2015 ozone NAAQS and the Clean Power Plan (CPP), are expected to have a growing
influence on the power sector in later years, as explained below. For this reason, the EPA expects
that most of the CSAPR Update's influence on emissions reductions will occur between 2017
and 2020.
24	5 31 F.3d 896, 911-12 (D.C. Cir. 2008) (holding that EPA should coordinate interstate transport compliance
deadlines with downwind attainment deadlines).
25	This deadline is in accordance with the D.C. Circuit's decision in NRDC v. EPA. Ill F.3d 456, 469 (D.C. Cir.
2014).
1-7

-------
EPA limits its analysis to this timeframe considering that on October 1, 2015, the EPA
strengthened the ground-level ozone NAAQS to 70 ppb. As discussed in the RIA for the final
2015 ozone NAAQS, it is assumed that potential nonattainment areas everywhere in the U.S.,
excluding California, will be designated such that they are required to attain the revised standard
by 2025. Furthermore, the EPA is mindful of the need to address ozone transport for the 2015
ozone NAAQS. As discussed in the memo to EPA Regional Administrators, Implementing the
2015 Ozone National Ambient Air Quality Standards, implementation of the good neighbor
provision for the 2015 ozone NAAQS may use the CSAPR framework. Given the statutory
implementation timeline of good neighbor requirements with respect to the 2015 ozone NAAQS,
the EPA anticipates that further actions to reduce interstate emission transport related to ozone
pollution could take place in the near future.26 Therefore, it is appropriate to evaluate the costs
of the regulatory control alternatives over the 2017-2020 timeframe.
For the reasons discussed in section V.B of the preamble, we have excluded the CPP
from the base case modeling for this rule. The EPA does not anticipate significant interactions
with the CPP and the near-term ozone season EGU NOx emission reduction requirements under
the CSAPR Update. See sections V.B and VII.F of the preamble for further discussion.
1.2.5 Emissions Controls and Cost Analysis Approach
The EPA estimated the control strategies and compliance costs of the CSAPR Update
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 turning on existing NOx control
technology, fully operating existing NOx control technology, purchasing, installing, and
operating NOx control technology, changes in fuel costs, and changes in the generation mix. 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.
26 See preamble section VII.
1-8

-------
1.2.6 Benefits Analysis Approach
The EPA estimated human health benefits (i.e., mortality and morbidity effects)
considering an array of health impacts attributable to changes in exposure to ozone and fine
particulate matter (PM2.5) from NOx reductions. We estimated these benefits using benefit-per-
ton estimates derived from the BenMAP tool. The EPA also estimated the climate co-benefits of
the CSAPR Update. A description of the methodologies used to estimate the human health and
climate benefits is contained in Chapter 5 of this RIA. In addition, Chapter 5 contains a
discussion of welfare co-benefits, such as ecosystem benefits from reduced nitrogen deposition.
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
post-processing techniques used to produce a number of air quality metrics for input into
the analysis of benefits and costs.
•	Chapter 4: Costs. The chapter summarizes the data sources and methodology used to
estimate the costs incurred by the power sector as well as changes in electricity and fuel
prices.
•	Chapter 5: Benefits. The chapter quantifies the health-related and climate benefits of the
ozone-related air quality improvements associated with the three regulatory control
alternatives analyzed.
•	Chapter 6: Economic Impacts. The chapter summarizes the data sources and
methodology used to estimate the economic impacts including employment impacts and
impacts on small entities.
•	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.
1-9

-------
CHAPTER 2: ELECTRIC POWER SECTOR PROFILE
Overview
This chapter discusses important aspects of the power sector that relate to today's final
action to update CSAPR 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 regulation, and provides background on the power sector and
electricity generating units (EGUs). In addition, this chapter provides some historical
background on trends in the past decade 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, in particular the increased natural gas
supply and subsequent relatively low natural gas prices, have resulted in more gas being utilized
as base load energy in addition to supplying electricity during peak load. This chapter presents
data on the evolution of the power sector from 2000 through 2014. Projections of future power
sector behavior and the impact of this rule are discussed in more detail in chapters 3 and 4 of this
RIA.
2.2	Power Sector Overview
The production and delivery of electricity to customers consists of three distinct
segments: generation, transmission, and distribution.
2-1

-------
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
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
2-2

-------
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 2000 and 2014.
In 2014 the power sector consisted of over 19,000 generating units with a total capacity27
of 1,038 GW, an increase of 255 GW (or 33 percent) from the capacity in 2000 (782 GW). The
255 GW increase consisted primarily of natural gas fired EGUs (211 GW) and wind generators
(62 GW), with substantially smaller net increases and decreases in other types of generating
units.
Table 2-1. Total Net Summer Electricity Generating Capacity by Energy Source,
2000 and 2014

2000
2014
Change Between '00 and '14
Energy Source
Net
Summer
Capacity
(MW)
% Total
Capacity
Net
Summer
Capacity
(MW)
% Total
Capacity
%
Increase
Capacity
Change
(MW)
%of
Total
Capacity
Increase
Coal
310,198
39%
295,906
29%
-5%
-14,293
-6%
Natural Gas
204,696
28%
415,592
40%
103%
210,896
83%
Nuclear
97,860
12%
98,569
10%
0.7%
709.3
0.3%
Hydro
97,769
11%
101,856
10%
4%
4,087
2%
Petroleum
60,710
8%
40,078
4%
-34%
-20,632
-8%
Wind
2,377
0.3%
64,156
6.2%
2599%
61,779
24%
Other
Renewable
8,190
1.6%
19,768
1.9%
141%
11,578
5%
Misc
331
0.4%
1,631
0.2%
393%
1,300
0.5%
Total
782,131
100%
1,037,556
100%
33%
255,425
100%
Note: This table presents generation capacity. Actual net generation is presented in Table 2-2.
Source: U.S. EIA. Electric Power Annual 2014, Table 4.3
27 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

-------
The 33 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 2000 to 2014, a total of 368 GW of new generating capacity was
built and brought online, and 80 GW existing units were retired. The overall net change in
capacity was an increase of 288 GW, as shown in Figure 2-1.
The newly built generating capacity was primarily natural gas (265 GW), which was
partially offset by gas retirements (35 GW). Wind capacity was the second largest type of new
builds (62 GW), augmented by solar (10 GW). The overall mix of newly built and retired
capacity, along with the net effect, is shown on Figure 2-1.
400,000
350,000
300,000
250,000
200,000
150,000
100,000
50,000
New Bui d
Net Change
Retirement
-50,000
Other
I Coal
Wind & Solar
Gas
Figure 2-1. National New Build and Retired Capacity (MW) by Fuel Type, 2000-201428
The information in Table 2-1 and Figure 2-1 present information about the generating
capacity in the entire U.S. The CSAPR Update Rule, however, directly affects EGUs in 22
eastern states (i.e., the CSAPR 2008 Ozone Region), as discussed in Chapter 1. The share of
generating capacity from each major type of generation differs between the CSAPR 2008 Ozone
28 Source: EIA Form 860. Not visible: wind and solar retirements = 87 MW, net change in coal capacity = -4,186
MW
2-4

-------
Region and the rest of the U.S. (non-region). Figure 2-2 shows the mix of generating capacity for
each region. In 2014, the overall capacity in the CSAPR 2008 Ozone Region is 59% 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 (34 percent) than it does in the rest of the country
(20%). The shares of natural gas, however, are quite similar (40% in the CSAPR 2008 Ozone
Region and 40% 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
Figure 2-2. Regional Differences in Generating Capacity (MW), 2014.
Source: 2014 EIA Form 860 Note: "Other" includes petroleum, geothermal, other renewable, waste materials
and misc."In-Region" refers to the 22 states within the CSAPR 2008 Ozone Region; "Non-Region" refers to all
other states in the contiguous U.S.
In 2014, electric generating sources produced a net 3,937 TWh to meet national electricity
demand, an 8 percent increase from 2000. As presented in Table 2-2, almost 70 percent of
electricity in 2014 was produced through the combustion of fossil fuels, primarily coal and
natural gas, with coal accounting for the largest single share. Although the share of the total
generation from fossil fuels in 2014 (67 percent) was only modestly smaller than the total fossil
share in 2000 (71 percent), the mix of fossil fuel generation changed substantially during that
period. Coal generation declined by 19 percent and petroleum generation by 73 percent, while
natural gas generation increased by 100 percent. This reflects both the increase in natural gas
Other
¦ Nuclear
2-5

-------
capacity during that period as well as an increase in the utilization of new and existing gas EGUs
during that period. Wind generation also grew from a very small portion of the overall total in
2000 to almost 5 percent of the 2014 total.
Table 2-2. Net Generation in 2000 and 2014 (Trillion kWh = TWh)

2000
2014
Change Between '00 and
'14

Net
Generation
(TWh)
Fuel
Source
Share
Net
Generation
(TWh)
Fuel
Source
Share
Net
Generation
Change
(TWh)
% Change
in Net
Generation
Coal
1,943
52%
1,569
40%
-374
-19%
Natural Gas
517
16%
1,033
26%
516
100%
Nuclear
753
20%
797
20%
44
6%
Hydro
265
7%
252
6%
-13
-5%
Petroleum
105
3%
28
1%
-77
-73%
Wind
5
0%
181
5%
176
3530%
Other Renewable
43
2%
66
2%
23
53%
Misc
2
0%
11
0%
9
434%
Total
3,637
100%
3,937
100%
300
8%
Source: U.S. EIA 2014 Electric Power Annual, Tables 3.2 and 3.3. Columns may not sum to totals due to rounding.
Percent change based on rounded values
Coal-fired and nuclear generating units have historically supplied "base load" electricity,
the portion of electricity loads which 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 31 percent of the total number of coal-fired units, but only 4 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
and heating/air-conditioning), versus late at night or very early in the morning, when demand for
electricity is reduced.
2-6

-------
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 57 percent of the coal EGU fleet capacity is over 500 MW per unit, only 8
percent of the gas fleet capacity is greater than 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 2014
Unit Size



Avg. Net
Summer
Total Net

Avg. Heat
Grouping

% of All
Avg.
Capacity
Summer
% Total
Rate
(MW) No.
Units
Units
Age
(MW)
Capacity (MW)
Capacity
(Btu/kWh)
COAL
0-24
130
12%
47
14
1,772
1%
12,269
25-49
80
8%
40
36
2,919
1%
11,718
50-99
117
11%
48
73
8,545
3%
11,725
100 - 149
106
10%
52
123
13,052
4%
10,926
150-249
166
16%
48
190
31,531
11%
10,524
250 - 499
197
19%
40
356
70,150
23%
10,450
500 - 749
183
17%
37
606
110,952
37%
10,222
750 - 999
57
5%
33
824
46,981
16%
9,952
1000 - 1500
11
1%
38
1259
13,850
5%
9,644
Total Coal
1047
100%
43
286
299,753
100%
10,900
NATURAL GAS
0-24
1,990
36%
35
7
13,922
3%
13,212
25-49
837
15%
23
40
33,488
7%
11,712
50-99
1001
18%
23
71
71,185
16%
11,999
100 - 149
414
8%
21
125
51,753
11%
9,593
150-249
1024
19%
15
176
179,952
40%
8,368
250 - 499
192
3%
24
342
65,652
15%
8,935
500 - 749
41
1%
35
586
24,020
5%
10,808
750 - 1000
13
0.24%
38
851
11,062
2%
10,694
Total Gas
5512
100%
26
82
451,034
100%
11,419
Source: National Electric Energy Data System (NEEDS) v.5.15
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 2013 or earlier, and excludes those units in NEEDS
with planned retirements in 2014 or 2015.
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 2-2 presents the cumulative age
distributions of the coal and gas fleets, highlighting the pronounced differences in the ages of the
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.
2-7

-------
0%
0	10	20	30	40	50	60	70
Age of EGU (years)
Coal Cap 	Coal Gen	Gas Cap 	Gas Gen
Figure 2-3. Cumulative Distribution in 2012 of Coal and Natural Gas Electricity
Capacity and Generation, by Age
Source: eGRID 2012 (10-2015 release from EPA eGRID website). Figure presents data from generators that came
online between 1943 and 2012 (inclusive); a 70 year period. Full eGrid data includes generators that came online as
far back as 1915. Full data from 1915 onward is used in calculating cumulative distributions; figure truncation at 70
years is merely to improve visibility of diagram.
Not displayed: coal units (376 MW total, 1 percent of total) and gas units (62 MW, < .01 percent of total)) over 70
years old for clarity. Figure is limited to coal-steam units in NEEDS v5.13 in operation in 2013 or earlier (excludes
-2,100 MW of coal-fired IGCC and fossil waste capacity), and excludes those units in NEEDS with planned
retirements in 2014 or 2015.
The locations of existing fossil units in EPA's National Electric Energy Data System
(NEEDS) v.5.15 are shown in Figure 2-4. This map reflects generating capacity expected to be
on-line at the end of 2018, and includes planned new builds already under construction and
planned retirements. The size of each dot corresponds with the capacity of the facility it
represents.
2-8

-------
Figure 2-4. Fossil Fuel-Fired Electricity Generating Facilities, by Size
Source: National Electric Energy Data System (NEEDS) v.5.15
Note: This map displays fossil capacity at facilities in the NEEDS v.5.15 IPM frame. NEEDS v.5.15 reflects
generating capacity expected to be on-line at the end of 2018. 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,29 each operating synchronously. Within each of these
transmission networks, there are multiple areas where the operation of power plants is monitored
29 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 (w hich 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
http://www.nerc.com/AboutNERC/keyplayers/Documents/NERC_Interconnections_Color_072512.jpg
2-9

-------
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
regional operator;30 in others, individual utilities31 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.
30	E.g., PMJ Interconnection, LLC, Western Area Power Administration (which comprises 4 sub-regions).
31	E.g., Los Angeles Department of Power and Water, Florida Power and Light.
2-10

-------
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
to a third of the total electricity produced32 (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 2000 and 2014.
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2000 and 2014 (billion kWh)

2000
2014


Sales/Direct Use
Share of
Sales/Direct Use
Share of


(Billion kWh)
Total End Use
(Billion kWh)
Total End Use

Residential
1,192
33%
1,407
36%

Commercial
1,055
29%
1,352
35%
Sales
Industrial
1,064
30%
998
26%

Transportation
NA

8
0.2%

Other
109
3%
NA

Total
3,421
95%
3,765
96%
Direct Use
171
5%
139
4%
Total End Use
3,592
100%
3,903
100%
Source: Table 2.2, EIA Electric Power Annual, 2014 and 2010
Notes: Retail sales are not equal to net generation (Table 2-2) because net generation includes net exported
electricity and loss of electricity that occurs through transmission and distribution.
Direct Use represents commercial and industrial facility use of onsite net electricity generation; and
electricity sales or transfers to adjacent or co-located facilities for which revenue information is not
available.
2.3.1 Electricity Prices
Electricity prices vary substantially across the United States, differing both between the
ultimate customer categories and also 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
32 Transportation (primarily urban and regional electrical trains) is a fourth ultimate customer category which
accounts less than one percent of electricity consumption.
2-11

-------
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
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 2014, the national
average retail electricity price (all sectors) was 10.44 cents/KWh, with a range from 7.13 cents
(Washington) to 33.43 (Hawaii).33
Average national retail electricity prices increased between 2000 and 2014 by 15.5 percent
in real terms (2011$). The amount of increase differed for the three major end use categories
(residential, commercial and industrial). National average industrial prices increased the most
(15.3 percent), and commercial prices increased the least (8.9 percent). The real year prices for
2000 through 2014 are shown in Figure 2-5.
33 EIA State Electricity Profiles with Data for 2014 (http://www.eia.gov/electricity/state/)
2-12

-------

14


w

*—1

*—1
o
12
(N

-C
10
J*



V)
8
c
0J

u

OJ
6
u

o_

>
4


u

u
2
_0J


0
2000 2002 2004
Residential
2006 2008
Commercial
2010 2012
Industrial —
2014
Total
Figure 2-5. Real National Average Electricity Prices for Three Major End-Use
Categories
Source: EI A Monthly Energy Review, Table 9.8
Most of these electricity price increases occurred between 2002 and 2008; since 2008
nominal electricity prices have been relatively stable while overall inflation continued to
increase. 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.
2000 2002
^—Residential —
2004 2006
Commercial
2008
Industrial
2012 2014
GDP Price
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for IV
End-Use Categories, With Inflation Indices
ajor
2-13

-------
For a longer term perspective, Figure 2-7 shows real34 (2011$) electricity prices for the
three major customer categories since 1960, and Figure 2-8 shows the relative change in real
electricity prices relative to the prices since 1960. As can be seen in the figures, the price for
industrial customers has always been lower than for either residential or commercial customers,
but the industrial price has been more volatile. While the industrial real price of electricity in
2014 was relatively unchanged from 1960, residential and commercial real prices are 22 percent
and 28 percent lower respectively than in 1960.
18
u
Q-
>
u
aj
1960
1970
> Residential
—i	1—
1980	1990
Commercial
2000
Indistrial
2010
Total
Figure 2-7. Real National Average Electricity Prices for Three Major End-Use
Categories (including taxes), 1960-2014 (2011$)
Source: EIA Monthly Energy Review , May 2016, Table 9.8
34 All prices in this section are estimated as real 2011 prices adjusted using the GDP implicit price deflator unless
otherwise indicated.
2-14

-------
-40%	
-50%	
Residential	Commercial	Indistrial — — Total
Figure 2-8. Relative Change in Real National Average Electricity Prices (2011$) for
Three Major End-Use Categories
Source: EIA Monthly Energy Review, May 2016, Table 9.8
2.3.2 Prices of Fossil Fuels Usedfor Generating Electricity
Another important factor in the changes in electricity prices are the changes in delivered
fuel prices35 for the three major fossil fuels used in electricity generation; coal, natural gas and
oil. Relative to real prices in 2000, the national average real price (in 2011$) of coal delivered to
EGUs in 2014 had increased by 49 percent, while the real price of natural gas decreased by 12
percent. The real price of delivered oil increased by 109 percent, but with oil declining as an
EGU fuel (in 2014 oil generated only 1 percent of electricity) the doubling of delivered oil prices
had little overall impact in the electricity market. The combined real delivered price of all fossil
fuels in 2014 increased by 44 percent over 2000 prices. Figure 2-9 shows the relative changes in
real price of all 3 fossil fuels between 2000 and 2014.
35 Fuel prices in this section are all presented in terms of price per MMBtu to make the prices comparable.
2-15

-------
2000 2002 2004 2006 2008 2010 2012 2014
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in
National Average Real Price per MMBtu Delivered to EGU
Source: Monthly Energy Review, May 2016, Table 9.9
2.3.3 Changes in Electricity Intensity of the U.S. Economy from 2000 to 2014
An important aspect of the changes in electricity generation (i.e., electricity demand)
between 2000 and 2014 is that while total net generation increased by 8 percent over that period,
the demand growth for generation was lower than both the population growth (13 percent) and
real GDP growth (27 percent). Figure 2-10 shows the growth of electricity generation,
population and real GDP during this period.
2-16

-------
30%
25%
o
o
™ 20%
0J
u
15%
10%
u
+->
c
cu
u
k_
5%
cu
Q.
0%
20
2002
2004
2006
2008
2010
2012
2014
-5%
Real GDP	Population	Generation
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since
2000
Sources: Generation: U.S. EIA Monthly Energy Review, May 2016. Table 7.2a Electricity Net Generation: Total
(All Sectors). Population: U.S. Census. Real GDP: 2016 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 2000 to 2014. On a per capita basis, real GDP per
capita grew by 12 percent between 2000 and 2014. At the same time electricity generation per
capita decreased by 4 percent. The combined effect of these two changes improved the overall
electricity efficiency of the U.S. market economy. Electricity generation per dollar of real GDP
decreased 15 percent. These relative changes are shown in Figure 2-11. Figures 2-10 and 2-11
clearly show the effects of the 2007 - 2009 recession on both GDP and electricity generation, as
well as the effects of the subsequent economic recovery.
2-17

-------
15%
10%
0%
-5%
v -10%
-15%
-20%
2000
2002
Real GDP/Capita ^^—Generation/Capita Generation/Real GDP
2004
2006
2008
2010
2012
2014
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation
Intensity Since 2000
Sources: Generation: U.S. EIA Monthly Energy Review, May 2016. Table 7.2a Electricity Net Generation: Total
(All Sectors). Population: U.S. Census. Real GDP: 2016 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
2-18

-------
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 2-13). Currently, there are 15 states plus the District of Columbia
where price deregulation of generation (restructuring) has occurred ("Active" in Figure 2-13).
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 Restructuring by State
Not Active
Active
Figure 2-12. Status of State Electricity Industry Restructuring Activities
Source: EIA 2010. "Status of Electricity Restructuring by State." Available online at:
.
2-19

-------
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 2000, traditional
utilities owned 77 percent of U.S. generating capacity (MW) while IPPs36 owned 23 of U.S.
generating capacity, respectively. The mix of electricity generated (MWh) was more heavily
weighted towards the utilities, with a distribution in 2000 of 83 percent, and 17 percent for IPPs.
Since 2000, IPPs have expanded faster than traditional utilities, substantially increasing
their share by 2014 of both capacity (59 percent utility, 41 percent IPPs) and generation (60
percent utility, 40 percent IPP).
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
types. A portion 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.
36 IPP data presented in this section include both combined and non-combined heat and power plants.
2-20

-------
Capacity Mix, 2000 & 2014
Generation Mix, 2000 & 2014
700
13
500
g. 400
ns
u

ai
100
p
1—
1 1

1
3,500
Other
Other
Wind
Wind
3,000
.—2,500
Nuc ear
Nuc ear
2,000
21,500
wl,000
500

2000 2014 2000 2014
2000 2014 2000 2014
Utility
IPP
Utility
IPP
Figures 2-13 & 2-14. Capacity and Generation Mix by Ownership Type, 2000 & 2014
2-21

-------
CHAPTER 3: EMISSIONS AND AIR QUALITY MODELING IMPACTS
Overview
This Chapter describes the methods for estimating emissions and air quality for the 2017
baseline and 2017 illustrative final CSAPR Update emissions budgets described in Chapter 4. In
Section 3.1, we describe the air quality modeling platform, in Section 3.2 we describe the
development of emissions inventories used in the air quality modeling, and in Section 3.3 we
describe the methods for processing the air quality modeling outputs to create inputs for
estimating benefits. The 2017 baseline and illustrative control case air quality model predictions
were used to calculate "benefit per ton" factors of reduced nitrogen oxides (NOx) on both ozone
and fine particulate matter (PM2.5) concentrations.37'38 These factors were then used to estimate
the benefits of the regulatory control alternatives, as described in Chapter 5. Details on the air
quality modeling are provided in the Air Quality Modeling Technical Support Document, which
can be found in the docket for this rule.
3.1 Air Quality Modeling Platform
We use the emissions inputs described in Section 3.2 for national scale applications of the
Comprehensive Air Quality Model with Extensions (CAMx) modeling system to estimate ozone
and PM2.5 air quality in the contiguous U.S. CAMx is a three-dimensional grid-based Eulerian
photochemical model designed to estimate ozone and PM2.5 concentrations over seasonal and
annual time periods. Because it accounts for spatial and temporal variations as well as
differences in the reactivity of emissions, CAMx is useful for evaluating the impacts of the rule
on ozone and PM2.5 concentrations.
For this analysis we used CAMx to simulate air quality for every hour of every day of the
year. These model applications require a variety of input files that contain information
pertaining to the modeling domain and simulation period. In addition to the CAMx model, our
modeling system includes (1) emissions for a 2011 base year and 2017 emissions for the baseline
and the final CSAPR Update emissions budgets, (2) meteorological data inputs for the year 2011,
37 The 2017 baseline air quality model predictions were also used to inform the EPA's ozone transport policy
analysis by identifying which states significantly contribute to nonattainment or interfere with maintenance of
downwind receptors. See Ozone Transport Policy Analysis Proposed Rule Technical Support Document, which can
be found in the docket for this proposed rule.
TO
Note that the baseline underlying the air quality modeling does not reflect the updated IPM emissions baseline
used to develop costs and benefits in Chapters 4 and 5. See the discussion in section 3.2.2 of this chapter.
3-1

-------
and (3) estimates of intercontinental transport (i.e., boundary concentrations) from a global
photochemical model. Using these data, CAMx generates hourly predictions of ozone and PM2.5
component species concentrations. The model predictions for the 2011 base year, the baseline in
2017, and the final CSAPR Update emissions budgets were combined with ambient air quality
observations to calculate seasonal mean ozone air quality metrics and annual mean PM2.5 for the
baseline in 2017 and the final CSAPR Update emissions budgets, which were then used as input
for the benefits analysis.
3.1.1	Simulation Periods
For use in this benefits analysis, the simulation period modeled by CAMx included
separate full-year application for each of the three emissions scenarios (i.e., 2011 base year, 2017
baseline and 2017 final CSAPR Update emissions budgets).
3.1.2	Air Quality Modeling Domain
Figure 3-1 shows the geographic extent of the modeling domain that was used for air
quality modeling in this analysis. The domain covers the 48 contiguous states, along with the
southern portions of Canada and the northern portions of Mexico. This modeling domain
contains 25 vertical layers with a top at about 17,550 meters, or 50 millibars (mb), and horizontal
grid resolution of 12 km x 12 km. The model simulations produce hourly air quality
concentrations for each 12 km2 grid cell across the modeling domain.
12US2 domain *. c.
x,y origin: -2412000m, H620000r
col: 396 row:246 / V
Figure 3-1. National air quality modeling domain.
3-2

-------
3.1.3 Air Quality Model Inputs
CAMx requires a variety of input files that contain information pertaining to the
modeling domain and simulation period. These include gridded, hourly emissions estimates and
meteorological data, and initial and boundary conditions. Separate emissions inventories were
prepared for the 2011 base year, the 2017 baseline, and final CSAPR Update emissions budgets.
All other inputs were specified for the 2011 base year model application and remained
unchanged for each future-year modeling scenario.
CAMx requires detailed emissions inventories containing temporally allocated emissions
for each grid-cell in the modeling domain for each species being simulated, as described in
Section 3.2. The meteorological data model inputs for the 2011 base year were derived from
running Version 3.4 of the Weather Research Forecasting Model (WRF). 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 grid cell in
each vertical layer. The CAMx lateral boundary and initial species concentrations are provided
by a three-dimensional global atmospheric chemistry and transport model (GEOS-Chem). The
lateral boundary species concentrations varied with height and time (every 3 hours).
3.2 Development of Emissions Inventories
3.2.1 2011 Base Year Emissions
The 2011 emissions inventories are primarily based on the 2011 National Emissions
Inventory, version 2 (201 1NEIv2) for point sources, nonpoint sources, commercial marine
vessels (CMV), nonroad mobile sources and fires, although the inventories used for modeling
often have temporal resolution additional to what is available in the NEI. The onroad mobile
source emissions are similar to those in the 201 1NEIv2, but were generated using the official
release 2014a version of the Motor Vehicle Emissions Simulator (MOVES2014a)
(http://www3.epa.gov/otaq/models/moves/), while the 201 1NEIv2 emissions were generated
using MOVES2014. Biogenic emissions and emissions inventories for Canada and Mexico are
also included in the air quality modeling. The meteorological data used to develop and
temporally allocate emissions were consistent with the 2011 data used for the air quality
modeling.
The emissions inventories and modeling thereof incorporate comments received on the
Notice of Data Availability (NOD A) published in the Federal Register on August 4, 2015 (80 FR
3-3

-------
46271), and from comments on the earlier notices for the 2011 and 2018 emissions modeling
platforms: the Notice of Availability of the Environmental Protection Agency's 2011 Emissions
Modeling Platform issued November 27, 2013 (78 FR 70935) and the Notice of Availability of
the Environmental Protection Agency's 2018 Emissions Modeling Platform issued January 14,
2014 (79 FR 2437), respectively. The Sparse Matrix Operator Kernel Emissions (SMOKE)
modeling system (Houyoux et al., 2000) version 37 was used to prepare the emissions
inventories for CAMx. Details regarding the development of the emission inventories and
emissions modeling for the 2011 base year and the 2017 baseline are documented in the
Technical Support Document Preparation of Emissions Inventories for the Version 6.3, 2011
Emissions Modeling Platform (EPA, 2016) and can be found in the docket for the CSAPR
Update.
3.2.2 2017 Baseline Emissions
The emission inventories for the 2017 future baseline have been developed using
projection methods that are specific to emission source type. Future emissions are projected from
the 2011 base 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), 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 (e.g., non-EGU point and nonpoint sources). The
same emissions are used in the base and future years for biogenic, fire, and offshore oil platform
sources.39 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.
The modeled 2017 baseline emission inventories represent predicted emissions that account for
Federal and State measures promulgated or under reconsideration by February, 2016. With the
exception of speciation profiles for mobile sources and temporal profiles for EGUs, the same
ancillary data files are used to prepare the future year emissions inventories for air quality
modeling as were used to prepare the 2011 base year inventories. Details on the included
measures are provided the emissions modeling TSD (EPA, 2016) and in Chapter 4 .
39 The biogenic and fire emissions are normally held constant between base and future years. The offshore emissions
were held constant due to the lack of detailed information available to adequately project those emissions to future
years.
3-4

-------
The 2017 baseline inventory for EGUs represents demand growth, fuel resource
availability, generating technology cost and performance, and other economic factors affecting
power sector behavior. The EGU emissions for the air quality modeling were developed using
the IPM version 5.15 base case.40 The IPM base case reflects the expected emissions accounting
for the effects of environmental rules and regulations, consent decrees and settlements, plant
closures, units built, control devices installed, and forecast unit construction through the calendar
year 2017. Significant federal and state measures that area accounted for in the baseline EGU
emissions in 2017 are discussed in Chapter 4.
The 2018 emissions output from IPM were adjusted to reflect 2017 emissions levels as
described in "Calculating 2017 NOx Emissions" (see
http://www2.epa.gov/airmarkets/calculating-2017-nox-emissions). Temporal allocation was
used to process the seasonal emissions outputs from IPM to hourly emissions. To the extent
possible, this temporal allocation process preserved the emissions patterns from the base year
(2011), while keeping the maximum emissions below those that occurred in the period 2011-
2014.
Projections for most stationary emissions sources other than EGUs (i.e., non-EGUs)
were developed by using the EPA Control Strategy Tool (CoST) to create post-controls future
year inventories. CoST is described at http://www3.epa.gov/ttnecasl/cost.htm. The 2017 baseline
non-EGU stationary source emissions inventory includes all 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 80 FR 46271, along with emissions reductions due to national and local
rules, control programs, plant closures, consent decrees and settlements. Ancillary reductions to
criteria air pollutant (CAP) emissions from stationary engines as a result of the Reciprocating
Internal Combustion Engines (RICE) National Emission Standard for Hazardous Air Pollutants
40 IPM is a multiregional, dynamic, deterministic linear programming model of the U.S. electric power sector. This
model is described in more detail in Chapter 4 of this RIA. The documentation for version 5.15 can be found on
EPA's power sector modeling website: https://www.epa.gov/airmarkets/power-sector-modeling
3-5

-------
(NESHAP) are included. Reductions due to the New Source Performance Standards (NSPS)
volatile organic compound (VOC) controls for oil and gas sources, and the NSPS controls for
process heaters, internal combustion engines, and natural gas turbines are also included.
Regional projection factors for point and nonpoint oil and gas emissions were developed
using Annual Energy Outlook (AEO) 2014 (U.S. EIA, 2014) projections from year 2011 to year
2018. Projected emissions for corn ethanol, cellulosic ethanol and biodiesel plants, refineries and
upstream impacts represent the Energy Independence and Security Act (EISA) renewable fuel
standards mandate in the Renewable Fuel Standards Program (RFS2). Airport-specific terminal
area forecast (TAF) data were used for aircraft to account for projected changes in
landing/takeoff activity.
Projection factors for livestock are based on expected changes in animal population from
2005 Department of Agriculture data, updated according to EPA experts in July 2012; fertilizer
application ammonia (NH3) emissions projections include upstream impacts representing EISA.
Area fugitive dust projection factors for categories related to livestock estimates are based on
expected changes in animal population and upstream impacts from EISA. Fugitive dust for paved
and unpaved roads take growth in VMT and population into account. Residential Wood
Combustion (RWC) projection factors reflect assumed growth of wood burning appliances based
on sales data, equipment replacement rates and change outs. These changes include growth in
lower-emitting stoves and a reduction in higher emitting stoves. Impacts from the NSPS for
wood burning devices are also included.
Projection factors for the remaining nonpoint sources such as stationary source fuel
combustion, industrial processes, solvent utilization, and waste disposal, reflect comments
received on the projection of these sources as a result of rulemakings and outreach to states on
emission inventories, and they also include emission reductions due to control programs. Future
year portable fuel container (PFC) inventories reflect the impact of the final Mobile Source Air
Toxics (MSAT2) rule along with state comments received in response to 80 FR 46271.
The MOVES2014a-based 2017 onroad 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 Light-Duty Vehicle Tier 2 Rule, the Heavy Duty Diesel Rule, the Mobile
3-6

-------
Source Air Toxics Rule, the Renewable Fuel Standard (RFS2), the Light Duty Green House
Gas/Corporate Average Fuel Efficiency (CAFE) standards for 2012-2016, the Heavy-Duty
Vehicle Greenhouse Gas Rule, the 2017 and the Later Model Year Light-Duty Vehicle
Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards; Final Rule (LD
GHG). The 2017 onroad emissions also include state rules related to the adoption of low
emission vehicle (LEV) standards, inspection and maintenance programs, Stage II refueling
controls, and local fuel restrictions. For California, the baseline emissions were provided by the
California Air Resources Board and include most this state's on-the-books regulations, such as
those for idling of heavy-duty vehicles, chip reflash, public fleets, track trucks, drayage trucks,
and heavy duty trucks and buses (CARB, 2016).
The nonroad mobile source emissions for 2017, including those for 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% sulfur fuel in
the Emissions Control Area (ECA) zone, the 2012-2015 Tier 2 NOx controls, the 2016 0.1%
sulfur fuel regulation in ECA zone, and the 2016 International Marine Organization (IMO) Tier
3 NOx controls. Non-U.S. and U.S. category 3 commercial marine emissions were projected to
2017 using consistent methods that incorporated controls based on ECA and IMO global NOx
and sulfur dioxide (SO2) controls. For California, the 2017 emissions for these categories reflect
the state's Off-Road Construction Rule for "In-Use Diesel", cargo handling equipment rules in
place as of 2011 (see http://www.arb.ca.gov/ports/cargo/cargo.htm), and state rules through 2011
related to Transportation Refrigeration Units, the Spark-Ignition Marine Engine and Boat
Regulations adopted on July 24, 2008 for pleasure craft, and the 2007 and 2010 regulations to
reduce emissions from commercial harbor craft.
The modeled 2011 emission case uses 2010 Canada emissions data, which is the latest
year for which Environment Canada had provided data at the time the modeling was performed.
Although no accompanying future-year projected baseline inventories were provided in a form
3-7

-------
suitable for this analysis, for the 2017 emissions, known shutdowns to Canadian coal EGU units
in Ontario were incorporated. In addition, onroad and nonroad mobile source emissions were
scaled to represent average changes in U.S. emissions due to the similarities between U.S. and
Canadian mobile source regulations. For Mexico, emissions compiled from the Inventario
Nacional de Emisiones de Mexico, 2008 were used for 2011, as that was the latest complete
inventory available. For the 2017 baseline, projected emissions for the year 2018 based on the
2008 inventory were used (ERG, 2014). Table 3-1 shows the modeled national 2011 and 2017
NOx and VOC emissions by sector. Additional details on the base year and projected inventories
and on the emissions by state are given in the emissions modeling TSD (US EPA, 2016).
Table 3-1. 2011 Base Year and 2017 Baseline NOx and VOC Emissions by Sector
(thousand tons)
Sector
2011 NOx
2017 NOx
2011 VOC
2017 VOC
EGU-point
2,100
1,300
38
36
NonEGU-point
1,200
1,200
800
800
Point oil and gas
510
440
160
170
Fires
380
380
4,800
4,800
Nonpoint oil and gas
660
730
2,500
2,900
Residential wood combustion
34
36
440
440
Other nonpoint
720
730
3,700
3,500
Nonroad
1,600
1,100
2,000
1,400
Onroad
5,600
3,000
2,700
1,500
Commercial marine vessels
410
360
13
13
(CMV)
Locomotive
790
680
41
28
Biogenics
910
910
42.800
42.800
TOTAL
15,000
10,900
59,900
58,400
3.2.3 2017 Illustrative Emissions Case for the Final CSAPR Update Emissions Budgets
The EPA's approach to developing IPM v5.15-based emissions for the final CSAPR
Update emissions budgets is methodologically consistent with the EPA's approach to
establishing the final EGU NOx ozone-season emissions budgets to reduce interstate ozone
transport for the 2008 ozone NAAQS. These illustrative EGU NOx ozone-season emissions
budgets and their associated assurance levels, along with corresponding emission changes for
other pollutants as predicted by IPM, were modeled in IPM v5.15 to create the illustrative final
emissions case. As noted in Chapter 4, section 4.3.1, although IPM v5.15 was used for modeling
EGU emission for the baseline and the illustrative final emissions case, there were additional
3-8

-------
updates to EGU emissions that were included in the IPM run for the CSAPR Update illustrative
final emissions case that were not included in the baseline. See Chapter 4, Table 4-4 for the
illustrative final emissions.
The emissions for the illustrative final emissions case were processed for air quality
modeling in the same way as the 2017 baseline. The only difference in the emissions inventories
were the EGU emissions. The hourly temporal allocation for the illustrative final emissions case
inventories preserved the patterns from the 2017 baseline to the extent possible by maintaining
consistent unit-specific and regional, where appropriate, profiles in both cases. Thus, the same
hourly temporal patterns in the baseline are reflected in this final emissions case, including any
adjustments made to constrain the hourly 2017 emissions below the maximum levels during the
2011-2014 period.
3.2.4 Effect of Emissions Reductions on Downwind Receptors
As described in Sections V and VI of the preamble, and in the Ozone Transport Policy
Analysis Final Rule TSD, and summarized here, EPA evaluated the effect of the CSAPR Update
on nonattainment and maintenance receptors with respect to interstate transport for the 2008
ozone NAAQS. The 2008 ozone standard is 75 parts per billion (ppb), annual fourth-highest
daily maximum 8 hour concentration, averaged over 3 years. As described in Section V of the
preamble, the nonattainment and maintenance receptors with respect to interstate transport for
the 2008 ozone NAAQS in 2017 were identified using air quality modeling for 2011 and 2017
combined with measured design values41 for a base period encompassing 2009-2013. There are
19 receptors in 9 states identified as non-attainment and/or maintenance monitors for this
CSAPR Update.42 Six of these monitors are non-attainment monitors and 13 are maintenance
monitors. The average of the average design values of all 19 receptors is 75.9 ppb in 2017. The
average of the maximum design values of all 19 receptors is 78.1 ppb in 2017.
41	Ozone design value for a given monitoring site is the 3-year average (consecutive years) of the 4th highest 8-hour
daily maximum ozone concentrations at that site.
42	Section V.C of the preamble describes the approach for projecting future ozone design values.
3-9

-------
As described in the Ozone Transport Policy Analysis Final Rule TSD, these design
values were identified using an updated version of the EGU base case, the same one that was
used to establish emission budgets for the final CSAPR Update.43 Like the base case used to
estimate the costs and benefits of the CSAPR Update, this base case accounts for the
Pennsylvania NOx RACT final rule promulgated in April 2016. However, the 2017 EGU
emission levels in this base case also account for recent historical information about emissions,
which grounds the 2017 emission projections in historic data for the purpose of setting emission
budgets. To evaluate the effect of the CSAPR Update on the 19 nonattainment and/or
maintenance receptors, we assume that the affected source emissions under the CSAPR Update
equals the EGU NOx ozone season emission budgets.44 That is, that the difference in the affected
source emission levels from the updated base case and the final EGU NOx ozone season
emission budgets was used to estimate the change in average and maximum design values at the
19 receptors reported in this section of the RIA.
The ozone Air Quality Assessment Tool (AQAT) was used to estimate the impact of the
upwind states' EGU NOx reductions on downwind ozone pollution concentrations. Specifically,
AQAT was used to forecast both the average and maximum design values at the 19 receptors.
The AQAT was developed specifically for use in the CSAPR Update rule. This tool uses air
quality modeling outputs to calibrate the predicted change in ozone concentrations to reflect
changes in NOx emissions. See the Ozone Transport Policy Analysis Final Rule TSD for the air
quality estimates and for details on the construction of the AQAT. The effect of the CSAPR
Update on the 19 nonattainment and/or maintenance receptors is an average reduction in the
average and maximum ozone design values of 0.28 ppb and 0.29 ppb in 2017, respectively. The
emission reductions are expected to reduce the average and maximum design values below the
level of the NAAQS at three of the 19 receptors, therefore resolving their nonattainment and
maintenance issues, while bringing the other 16 receptors closer to attainment and maintenance.
Results for each of the 19 receptors are described in the Ozone Transport Policy Analysis Final
Rule TSD.
43	In the Ozone Transport Policy Analysis Final Rule TSD, this updated base case is referred to as the "$0/ton
emissions budget level with PA RACT."
44	In the Ozone Transport Policy Analysis Final Rule TSD, these budgets are referred to as the "Final $1400/ton
Emission Budgets."
3-10

-------
3.3 Post-Processing of Air Quality Modeling for Benefits Calculations
3.3.1 Converting CAMx Ozone Outputs to Benefits Inputs
The CAMx model generates predictions of hourly ozone concentrations for every grid
cell. Future-year estimates of ozone for each of three health benefits metrics for ozone were
calculated using model predictions. The modeled change in ozone between the 2011 base year
and the 2017 future baseline and illustrative control case were used to create relative reduction
factors (RRFs) which were then applied to 2011 ambient ozone concentrations, as described
below. The health benefits metrics for ozone are May through September seasonal average 8-
hour daily maximum ozone concentrations. The procedures for determining the ozone RRFs for
these metrics are similar to those described in EPA guidance for modeling attainment of the
ozone standard (EPA, 2014). This guidance recommends that model predictions be used in a
relative sense to estimate changes expected to occur in ozone concentrations for a future year
emissions case. The RRFs and future year ozone concentrations were calculated using EPA's
software Modeled Attainment Test Software (MATS) (Abt, 2014). EPA used MATS to estimate
the ozone impacts of the emissions reductions in the 2017 illustrative control case.
For the purposes of projecting future ozone concentrations for input to the benefits
calculations, we applied MATS using the base year 2011 modeling results and the results from
the 2017 baseline and 2017 illustrative control case scenarios. In our application of MATS for
ozone we used the ozone monitoring data centered about 2011 (2010-2012 ozone data) from the
Aerometric Information Retrieval System (AIRS) as the set of base-year measured
concentrations. The ambient ozone data and modeled ozone outputs were combined using the
MATS "eVNA" spatial fusion technique to generate gridded sets of spatial fields (interpolated
ozone metrics for each modeled 12km grid cell in the modeling domain) for each of the three
ozone metrics for the 2011 base year period. The ratio of the seasonal average model-predicted
future case ozone concentrations to the corresponding seasonal average model-predicted 2011
concentrations in each grid cell (RRF's) was calculated and then multiplied by the gridded
interpolated ozone concentrations for each metric to produce gridded ozone concentrations for
the 2017 baseline and 2017 illustrative control case. The resulting gridded files for the 2017
baseline and illustrative control cases were then input to the Benefits Mapping and Analysis
Program - Community Edition (BenMAP-CE) (version 1.1) (Abt, 2012) to calculate benefit per
3-11

-------
ton factors for each metric. Information on the calculation of the benefit per ton factors is
provided in Chapter 5.
3.3.2 Converting CAMx PM2.5 Outputs to Benefits Inputs
The CAMx model (ENVIRON, 2014) generates predictions of hourly PM2.5 species
concentrations for every grid cell. The species include a primary fraction and several secondary
PM2.5 species (e.g., sulfates, nitrates, and organics). PM2.5 is calculated as the sum of the primary
and the secondary formed particles. Future-year estimates of PM2.5 were calculated using RRFs
applied to 2010-2012 ambient PM2.5 and PM2.5 species concentrations, as described below.
The procedures for determining the RRFs are similar to those in EPA guidance for
modeling the PM2.5 NAAQS (EPA, 2014). This guidance recommends that model predictions be
used in a relative sense to estimate changes expected to occur in each PM2.5 species. The
modeled attainment test procedure for calculating future year PM2.5 values is described in the
modeling guidance and is codified in EPA's MATS. EPA used this procedure to estimate the
ambient impacts of the emissions reductions in the 2017 illustrative control case. For the
purposes of projecting future PM2.5 concentrations for input to the benefits calculations, we
applied the modeled attainment test procedure using the base year 2011 modeling results and the
results from the 2017 baseline and 2017 illustrative control case. In our application of MATS for
PM2.5 we used the PM2.5 monitoring data and speciated monitoring data centered about 2011
(2010-2012) from the state PM2.5 Federal Reference Method (FRM) network, the Chemical
Speciation Network (CSN) and Interagency Monitoring of Protected Visual Environments
(IMPROVE) network as the set of base-year measured concentrations. The ambient PM2.5 and
species data and modeled PM2.5 and species outputs were combined using the MATS "eVNA"
spatial fusion technique to generate gridded sets of spatial fields (interpolated annual average
PM2.5 and species concentrations for each modeled 12km grid cell in the modeling domain) for
the 2011 base year period. The ratio of the quarterly average model-predicted future case PM2.5
species concentrations to the corresponding quarterly average model-predicted 2011 species
concentrations in each grid cell (RRF's) were calculated and then multiplied by the gridded
interpolated PM2.5 species concentrations to produce gridded PM2.5 species concentrations for the
2017 baseline and 2017 illustrative control case. Output files from this process include both
3-12

-------
quarterly and annual mean PM2.5 mass concentrations and PM2.5 species concentrations which are
then processed to produce BenMAP input files containing annual mean PM2.5 mass
concentrations for the 2017 baseline and for the 2017 illustrative control case. These data files
were then input to BenMAP to calculate PM2.5 benefit per ton factors. Information on the
calculation of the benefit per ton factors is provided in Chapter 5.
3.4	Limitations
The air quality modeling for this analysis relied upon state-of-the-science tools, methods,
and data. Still, there are uncertainties associated with the projected baseline and illustrative
control case ozone concentrations that stem from limitations and uncertainties in the individual
components of the modeling process. These include (1) limitations in the emissions inventories
for specific source categories in terms of representing base year emissions and the methodologies
and economic assumptions associated with projecting emissions to a future year, (2)
uncertainties in the construct of the photochemical model that may affect the characterization of
physical properties and chemical reactions, (3) uncertainties in other model inputs such as
meteorology and international transport, and (4) uncertainties in the measured ozone
concentrations that are used as the basis for projecting future concentrations at individual
locations and the spatial fields used for benefits calculations. It is not clear that the net effect of
the limitations and uncertainties in the modeling process bias the analysis in either direction.
Rather, they should be viewed as considerations in interpreting the results.
3.5	References
Abt Associates, 2012. "BenMAP User's Manual Appendices," prepared for U.S. Research
Triangle Park, NC: U. S. Environmental Protection Agency, Office of Air Quality Planning
and Standards. Available at:
. Accessed June
6, 2015.
Abt Associates, 2014. User's Guide: Modeled Attainment Test Software.
http://www3.epa.gov/scram001/modelingapps_mats.htm.
California Air Resources Board, 2016. EMFAC2014 vl.0.7 technical documentation. Available
at http://www.arb.ca.gov/msei/downloads/emfac2014/emfac2014-vol3-technical-
documentation-052015 .pdf.
3-13

-------
ENVIRON, 2014. User's Guide Comprehensive Air Quality Model with Extensions version 6.11,
www.camx.com. ENVIRON International Corporation, Novato, CA.
ERG, 2014. Develop Mexico Future Year Emissions Final Report. Available at
ftp://ftp.epa.gov/EmisInventory/2011v6/v2platform/2011emissions/Mexico_Emissions_WA
%204-09_final_report_121814.pdf.
Houyoux, M.R., Vukovich, J.M., Coats, C.J., Wheeler, N.J.M., Kasibhatla, P.S. (2000),
Emissions inventory development and processing for the Seasonal Model for Regional Air
Quality (SMRAQ) project, Journal of Geophysical Research - Atmospheres, 105(D7), 9079-
9090.
U.S. Energy Information Administration, 2014. Annual Energy Outlook, 2014
(http ://www. eia. gov/forecasts/archive/ aeo 14/).
U.S. Environmental Protection Agency, 2014. Modeling Guidance for Demonstrating
Attainment of Air Quality Goals for Ozone, PM2.5, and Regional Haze, Research Triangle
Park, NC. (http://www3.epa.gov/ttn/scram/guidance/guide/Draft_03-PM-
RH_Modeling_Guidance-2014.pdf)
U.S. Environmental Protection Agency, 2016. Preparation of Emissions Inventories for the
Version 6.3, 2011 Emissions Modeling Platform, Research Triangle Park, NC. Available
from the 2011v6.3 platform section of https://www.epa.gov/air-emissions-modeling/2011-
version-6-air-emissions-modeling-platforms.
3-14

-------
CHAPTER 4: COST, EMISSIONS, AND ENERGY IMPACTS
Overview
This chapter reports the compliance costs, emissions, and energy analyses performed for
the final CSAPR Update. The EPA used the Integrated Planning Model (IPM), developed by ICF
International, to conduct most of the analysis discussed in this chapter.
As explained in detail below, this chapter presents analysis of three regulatory control
alternatives. These regulatory control alternatives include assumptions about the possible actions
that electric generating units (EGUs) may pursue to reduce their nitrogen oxides (NOx)
emissions in order to comply with the EGU NOx ozone season emission budgets in the 22-state
region.
The chapter is organized as follows: following a summary of the regulatory control
alternatives analyzed and a summary of the 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. Additional impacts are presented in subsequent chapters.
4.1 Regulatory Control Alternatives
The primary purpose of the CSAPR Update is to address interstate air quality impacts
with respect to the 2008 ozone National Ambient Air Quality Standards (NAAQS). The EPA
originally published CSAPR on August 8, 2011,45 to address interstate transport of ozone
pollution under the 1997 ozone NAAQS.46 The CSAPR Update will reduce ozone season (May 1
through September 30) NOx emissions in 22 eastern states that can be transported downwind as
NOx or, after transformation in the atmosphere, as ozone, and can negatively affect air quality
and public health in downwind areas. For these 22 eastern states, the EPA is issuing Federal
Implementation Plans (FIPs) that generally provide updated CSAPR NOx ozone season emission
budgets for EGUs. These emission budgets represent the remaining EGU emissions after
4-1

-------
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 CSAPR Update FIPs also
require affected EGUs to participate in the CSAPR NOx ozone season allowance trading
program subject to these emission budgets starting with the 2017 ozone season. The allowance
trading program is the remedy in the FIPs that achieves the ozone season NOx emission
reductions required by the rule. The allowance trading program essentially converts the EGU
NOx emission budget for each of the 22 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
are able to 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.
This RIA evaluates the benefits, costs and certain impacts of compliance with three
regulatory control alternatives. The CSAPR Update EGU NOx ozone season emission budgets
that the EPA is finalizing were developed using uniform control stringency represented by
$1,400 per ton (2011$), whereas the more and less stringent alternatives were developed using
uniform control stringency represented by $3,400 per ton and $800 per ton (2011$), respectively.
The EPA assesses compliance with these sets of emission budgets through implementation of the
CSAPR NOx ozone season allowance trading program. 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 emission
budgets analyzed. This chapter describes the EPA's analysis of the CSAPR Update and more and
less stringent alternatives. As described below, the emission budgets evaluated for the CSAPR
Update regulatory control alternatives in this RIA are illustrative because they differ somewhat
from the budgets finalized in this rule.
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 CSAPR Update. The final CSAPR Update emission
budgets in this RIA represents illustrative EGU NOx ozone season emission budgets for each
4-2

-------
state that were developed using uniform control stringency represented by $1,400 per ton
(2011$).47
Additionally, OMB Circular A-4 requires analysis of at least one potential alternative
standard level that is more stringent than the finalized standard and one that is less stringent than
the finalized standard. In response to this requirement, this RIA analyzes the final CSAPR
Update emission budgets as well as a more and a less stringent alternative to the CSAPR Update.
The more and less stringent alternatives differ from the CSAPR Update 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 $800
per ton (2011$). The more-stringent scenario uses emission budgets that were developed using
uniform control stringency represented by $3,400 per ton (2011$). These sets of emissions
budgets are analogous to those that the EPA explicitly took comment on in the CSAPR Update
proposal. We continue to analyze these scenarios alongside the finalized approach to evaluate
how economic and environmental information that has been updated since proposal affected
these non-finalized options. See section VI of the preamble for further details of these emission
budgets.
All three scenarios are illustrative in nature, and the budgets included in the "CSAPR
Update" scenario differ slightly from the budgets finalized in this rule. That is because
subsequent to completion of the analysis of these three scenarios, the EPA made minor updates
to budgets as well as to the modeling platform48. The EPA finds that the three illustrative
regulatory control alternatives presented in this RIA variously provide a reasonable
approximation of the impacts of the final rule, as well as an evaluation of the relative impacts of
two regulatory alternatives. This finding is supported by an analysis of the costs and impacts
(but not the benefits) of the final CSAPR Update emission budgets, as estimated using the
updated modeling platform. This analysis is provided in Appendix 4A.
47	The budget setting process is described in the preamble and in detail in the Ozone Transport Policy Analysis
Technical Support Document (TSD)
48	See EPA v.5.15 CSAPR Update Rule Base Cases Using IPM Incremental Documentation, available in the
docket.
4-3

-------
Table 4-1 reports the illustrative EGU NOx ozone season emission budgets that are
evaluated in this RIA. As described above, starting in 2017, emissions from affected EGUs
across this entire region cannot exceed the sum of emission budgets but for the ability to use
banked allowances from previous years for compliance. 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 CSAPR NOx ozone season assurance level, or 121 percent of the
emission budget. The CSAPR NOx ozone season allowance trading program is described in
further detail in Section VII of the preamble.
Table 4-1 Illustrative NOx Ozone Season Emission Budgets (Tons) Evaluated in this RIA

CSAPR Update (Not
Finalized Budgets)
More Stringent
Alternative
Less Stringent
Alternative
Alabama
12,599
11,406
13,548
Arkansas49
9,211
9,041
12,060
Delaware50
497
494
497
Illinois
14,588
14,464
14,632
Indiana
21,527
19,804
26,419
Iowa
11,272
11,065
11,477
Kansas
7,782
7,730
7,785
Kentucky
19,675
19,475
23,030
Louisiana
18,636
18,470
19,087
Maryland
3,457
2,838
3,795
Michigan
16,483
15,222
18,630
Mississippi
6,315
6,191
6,350
Missouri
15,085
14,604
16,628
New Jersey
2,057
2,061
2,063
New York
5,050
4,928
5,129
Ohio
18,763
18,599
22,372
Oklahoma
11,742
9,254
13,871
Pennsylvania
19,554
19,479
29,875
Tennessee
9,115
9,115
9,115
Texas
51,931
50,022
54,544
49	The EPA is finalizing CSAPR EGU NOx ozone season emission budgets for Arkansas of 12,048 tons for 2017
and 9,210 tons for 2018 and subsequent control periods. This RIA assessment assumes compliance with Arkansas'
illustrative emission budget for 2018 in 2017 and subsequent control periods.
50	Delaware is excluded from the final CSAPR Update policy, but was included in the illustrative policy modeling.
4-4

-------

CSAPR Update (Not
Finalized Budgets)
More Stringent
Alternative
Less Stringent
Alternative
Virginia
9,224
8,758
9,263
West Virginia
18,152
17,706
25,730
Wisconsin
7,862
7,791
7,922
TOTAL
310,577
298,515
353,821
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 CSAPR Update.
IPM is a multi-regional, dynamic, deterministic linear programming model of the
contiguous U.S. electric power sector. It provides estimates of least cost capacity expansion,
electricity dispatch, and emissions control strategies while meeting energy demand and
environmental, transmission, dispatch, and reliability constraints. The EPA has used IPM for
over two decades to better understand power sector behavior under future business-as-usual
conditions and to evaluate the economic and emission impacts of prospective environmental
policies. The model is designed to reflect electricity markets as accurately as possible. The 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
4-5

-------
modeling in IPM. The model documentation provides additional information on the assumptions
discussed here as well as all other model assumptions and inputs.51
The model incorporates a detailed representation of the fossil-fuel supply system that is
used to estimate equilibrium fuel prices. The model includes an endogenous representation of the
North American natural gas supply system through a natural gas module that reflects a partial
supply and demand equilibrium of the North American gas market, accounting for varying levels
of potential power sector and non-power sector gas demand and corresponding gas production
and price levels.52 This module consists of 118 supply, demand, and storage nodes and 15
liquefied natural gas re-gasification facility locations that are tied together by a series of linkages
(i.e., pipelines) that represent the North American natural gas transmission and distribution
network.
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
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.53
51	The documentation of EPA's Base Case using IPM (v5.15) contains detailed information, including all the
underlying assumptions, data sources, and architecture parameters. The documentation for EPA's Base Case v.5.15
using IPM consists of a comprehensive document for IPM v. 5.13, and an incremental update document for both
v.5.14 and v.5.15. All are available at available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
52	See Chapter 10 of EPA's Base Case using IPM (v5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
53	See Chapter 9 of EPA's Base Case using IPM (v.5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
4-6

-------
To estimate the annualized costs of additional capital investments54 in the power sector,
the 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.55 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.
The EPA has used IPM extensively over the past two 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), and the Carbon Pollution Standards for New Power Plants (U.S. EPA, 2015a).
The 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 the 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
assumptions, model representation, and modeling results. IPM has received extensive review by
54	Due to the near-term compliance timing for the CSAPR Update, the 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 in the near-
term. The EPA's illustrative compliance modeling does allow for new combustion controls, which represent the
most likely potential capital expenditure in the 2017 analysis year.
55	See Chapter 8 of EPA's Base Case using IPM (v5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
4-7

-------
energy and environmental modeling experts in a variety of contexts. For example, in the late
1990s, the Science Advisory Board reviewed IPM as part of the CAA Amendments Section 812
prospective studies56 that are periodically conducted. The model has also undergone considerable
interagency scrutiny when it was used to conduct over a dozen legislative analyses (performed at
Congressional request) over the past decade. 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 RGGI, 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 rule. As such, an IPM base case represents an element of the 2017 and 2020
baseline for this RIA.57 The EPA frequently updates the IPM base case to reflect the latest
available electricity demand forecasts from the U.S. Energy Information Agency (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 v.5.15 Base Cases for the CSAPR Update
EPA's IPM modeling platform used to analyze this rule (v.5.15) is similar to the version
used to analyze the CSAPR Update proposal, and incorporates minor updates made primarily in
response to comments received on an August 4, 2015 Notice of Data Availability and the
proposed rule.
56	http://www2.epa.gov/clean-air-act-overview/benefits-and-costs-clean-air-act
57	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).
4-8

-------
As with the CSAPR Update proposal, the IPM v.5.15 modeling platform incorporates
federal and most state laws and regulations whose provisions were either in effect or enacted and
clearly delineated by February 1, 2016.58 The base case includes the Final Mercury and Air
Toxics Standards (MATS),59 and two non-air federal rules affecting EGUs: Cooling Water
Intakes (316(b)) Rule (U.S. EPA, 2014), and Combustion Residuals from Electric Utilities
(CCR) (U.S. EPA, 2015b). Additionally, all new capacity projected by the model is compliant
with Clean Air Act 111(b) standards, including the final standards of performance for GHG
emissions from new sources. As described in section IV.B of the preamble, the Clean Power
Plan (CPP) is not included in this analysis.
Unlike the base case used in the analysis of the proposed rule, which was conducted prior
to the D.C. Circuit's issuance of EME Homer City II,60 the base case for the final rule accounts
for compliance with the original CSAPR by including as constraints all original CSAPR
emission budgets with the exception of remanded Phase 2 NOx ozone season emission budgets
for 11 states and Phase 2 NOx ozone season emission budgets for four additional states that were
finalized in the original CSAPR supplemental rule.61 For more information, see section V of the
preamble.
58	Note that this modeling platform does not include the Regional Haze Plan for Texas and Oklahoma, published
January 5, 2016. EPA does not believe this rule would substantially affect ozone season NOx emissions in 2017, and
therefore budgets determined for this rule.
59	In Michigan v. EPA, the Supreme Court reversed on narrow grounds a portion of the D.C. Circuit decision
upholding MATS, finding that EPA erred by not considering cost when determining that regulation of EGUs was
"appropriate" pursuant to CAA section 112(n)(l). 135 S.Ct. 192 (2015). The case was remanded to the D.C. Circuit
for further proceedings, and because MATS was remanded but not vacated, MATS currently remain in place.
60	In EME Homer City II, the D.C. Circuit declared invalid the CSAPR phase 2 NOx ozone season emission budgets
of 11 states: Florida, Maryland, New Jersey, New York, North Carolina, Ohio, Pennsylvania, South Carolina, Texas,
Virginia, and West Virginia. Id. 795 F.3d at 129-30, 138. The court remanded those budgets to the EPA for
reconsideration. Id. at 138. As a result, the EPA removed the original CSAPR phase 2 NOx ozone season emission
budgets as constraints for these 11 states in the 2017 IPM modeling.
61	The EPA acknowledges that the CSAPR NOx ozone season emission budgets for Iowa, Michigan, Oklahoma, and
Wisconsin ~ which were finalized in the original CSAPR Supplemental Rule (76 FR 80760, December 27, 2011) —
were linked to the same receptors that lead to the remand of other states' NOx ozone season emission budgets in
EME Homer City II.
4-9

-------
Additionally, after the IPM modeling for the final rule was underway, Pennsylvania
published a new RACT rule62 that would require EGU NOx reductions starting on January 1,
2017. The EPA was unable to explicitly include this final state rule in the IPM base case for the
final CSAPR Update. However, the EPA recognizes that the implementation of this final state
rule will precede the first control period for the final CSAPR Update. The agency believes that it
is reasonable to remove the impacts of the Pennsylvania RACT rule from the estimated impacts
of the CSAPR Update to appropriately reflect the emission reductions, costs, and benefits
attributable to the CSAPR Update. Therefore, the EPA evaluated the EGU emission reductions
expected to result from Pennsylvania's RACT rule exogenously and isolated these impacts from
the EPA's assessment of emission reductions, benefits, and costs estimated for the CSAPR
Update and the more and less stringent alternatives. For more information, see the Pennsylvania
Additional RACT Requirements for Major Sources of NOx and VOCs Memo to the Docket.
Other updates to the v.5.15 base cases used in this final rule include largely unit-level
specifications (e.g., pollution control configurations and emissions rates), and planned power
plant construction and closures that the EPA was aware of by February 1, 2016. In Maryland,
emission rates of units were updated to reflect compliance with the state's RACT rule.
Additionally, given the lead times for new combined cycle plants, EPA did not allow the model
to build additional capacity of that type in 2018 beyond announced new builds.63 Similarly, EPA
did not allow new renewable generation to be built before 2018, nor did EPA allow the model to
retire any units beyond announced retirements before 2020. Finally, NOx-specific pollution
control retrofits were limited to retrofits that occurred in the base case. For a detailed account of
all updates made to the v.5.15 modeling platform, see the EPA v.5.15 CSAPR Update Rule Base
Cases Using IPM Incremental Documentation , available in the docket..
EPA also updated the National Electric Energy Data System (NEEDS)64, based largely
on public comment received in response to an August 4, 2015 Notice of Data Availability and
62	Published April 23, 2017 (http://www.pabulletin.com/secure/data/vol46/46-17/694.html)
63	Additionally, note that no new coal-fired capacity was projected in 2016 or 2018 in any of the model runs
compelted for this analysis.
64	http://www2.epa.gov/airmarkets/power-sector-modeling-platform-v515
4-10

-------
the proposed rule. This database contains the unit-level data that is used to construct the "model"
plants that represent existing and planned-committed units in EPA modeling applications of
IPM.65 NEEDS includes detailed information on each individual EGU, including geographic,
operating, air emissions, and other data on every generating unit in the contiguous U.S.
While the EPA used the IPM v.5.15 platform throughout the development and analysis of
the final CSAPR Update, minor updates were made to this modeling platform over the course of
the final rule development. Subsequent to the initial base case projections that provided power
sector emissions data used for air quality modeling,66 the EPA made minor updates to the
modeling platform, which focus primarily on electricity generating unit-level input assumptions
regarding NOx rates. The EPA believes that these updates, while relatively minor in the context
of national emission projections, improve the model's ability to reflect the electric power system
in relation to the CSAPR Update, and enable the EPA to provide the best projections possible to
evaluate this rule. For more information, see the EPA v.5.15 CSAPR Update Rule Base Cases
Using IPM Incremental Documentation. .
The analysis of cost and impacts presented in this chapter is based on a single IPM base
case, the Illustrative Base Case, and represents incremental impacts projected solely as a result of
compliance with the illustrative emission budgets presented in Table 4-1 above. Note that
further analysis, which includes additional updates, is presented in Appendix 4A.
4.3.2. Methodology for Evaluating the Regulatory Control Alternatives
To estimate the costs, benefits, and economic and energy market impacts of the CSAPR
Update, the EPA conducted quantitative analysis of the three regulatory control alternatives: the
illustrative final CSAPR Update 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.
65	For more information, see Chapter 4 of the IPM Documentation.
66	The air quality modeling, used to quantify upwind state contributions, is described in Chapter 2 of this RIA and
Section V of the preamble.
4-11

-------
Before undertaking power sector analysis to evaluate compliance with the regulatory
control alternatives, the EPA first considered available EGU NOx mitigation strategies that could
be implemented for the first compliance period (i.e., the 2017 ozone season). The 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 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 2017 ozone-season. For more details on these assessments, including the assessment of EGU
NOx mitigation costs and feasibility, please refer to the Final EGU NOx Mitigation Strategies
TSD, in the docket for this rule.
These mitigation strategies are primarily captured within the model. However, due to
limitations on model size, IPM v.5.15 does not have the ability to determine, within the model,
whether or not to operate existing EGU post-combustion NOx controls (i.e., SCR or SNCR) in
response to a regulatory emission requirement.67 Whether or not an existing post-combustion
NOx control at a particular EGU is operating in a model scenario is determined by the model
user. In order to evaluate compliance with the regulatory control alternatives, the EPA
determined, outside the model, whether or not operation of existing controls that are idle in the
base case would be reasonably expected for compliance with each of the evaluated regulatory
control alternatives. After imposing the requirement to operate these controls, IPM estimated the
associated NOx reductions and impacts associated with each regulatory alternative.
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;
more information about the estimated costs of these controls can be found in the EGU NOx
Mitigation Strategies TSD.
67 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.
4-12

-------
Table 4-2. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives
Regulatory Control
Alternative
NOx Controls Implemented
Less Stringent Alternative
(1)	Fully operating existing SCRs to achieve 0.081 lb/MMBtu NOx emission
rate (costs estimated outside IPM)68
(2)	Shift generation to minimize costs (costs estimated within IPM)
CSAPR Update
(All controls above)
(3)	Turn on idled SCRs (costs estimated within IPM)
(4)	Install or upgrade combustion controls (costs estimated outside IPM)
More Stringent Alternative
(All controls above)
(5) Turn on idled SNCRs (costs estimated within IPM)
In addition to the limitation on ozone season NOx emissions required by the EGU
emissions budgets for the 22 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. They are: 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 2015 and 2016 vintage NOx ozone season allowances issued under the
original CSAPR to address interstate ozone transport for the 1997 ozone NAAQS. 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 22 states. Furthermore,
allowances may be banked for future use. The number of banked allowances is influenced by the
determination, outside the model, whether or not existing controls that are idle in the base case
are turned on and by if it is less costly to abate ozone season NOx emissions in a current ozone
68 For consistency with the budgets analyzed in this chapter, the illustrative policy cases assumes that fully operated
SCRs can achieve NOx emissions rates of up to 0.081 lbs/MMBtu. Note that the final budgets are based on an
assumed NOx rate of 0.10 lbs/MMBtu, which is the modeling assumption used in the analysis of the final budgets
presented in Appendix 4A.
4-13

-------
season than to abate emissions in a later ozone season. Affected EGUs are expected to bank NOx
ozone season allowances in the 2017 ozone season for use in the later ozone season. Based on
observation, the EPA believes that this is a reasonable illustrative compliance path for EGUs,
which may wish to bank allowances for future use under economic reasons or non-economic
reasons such as being prepared for future variability in power sector operations.
While there are no explicit limits on the exchange of allowances between affected EGUs
and on the banking of 2017 and future vintage NOx ozone season allowances, the assurance
provisions limit the amount of seasonal NOx emissions by affected EGUs in each of the 22
states. The assurance level limits affected EGU emissions over an ozone season to the state's
NOx ozone season emission budget plus an increment equal to 21 percent of each state's
emissions budget. This increment is called the variability limit. See section VII.E.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 VII.E.4 of the preamble also explains how the 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.
As described in section VII.C.2 of the preamble, the rule allows 2015 and 2016 vintage
NOx ozone season allowances (that had been issued under CSAPR to address interstate ozone
transport for the 1997 ozone NAAQS) to be used for compliance with this rule, following a one-
time conversion that reduces the overall quantity of banked allowances from that time period.
Based on EPA's expectation that its conversion of allowances will limit the use of banked
allowances to one year of states' aggregated variability limits,, the treatment of these banked
allowances is represented in the modeling as an additional 65,221 tons of NOx allowances, the
equivalent of one year of the variability limit associated with the illustrative emission budgets,
that may be used by affected EGUs during the 2017 ozone season or in later ozone seasons.
4-14

-------
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 three of
the NOx mitigation strategies: turning on idled SCRs, turning on idled SNCRs, and shifting
generation to lower-NOx emitting EGUs. The costs of increasing the use and optimizing the
performance of existing and operating SCRs,69 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 IPM emissions projections and utilize 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 the EPA's methodology for estimating the component of
compliance costs that are calculated outside of the model for the CSAPR Update scenario:
(1)	In the model projections, identify all model plants in the 22-state region that can adopt
the following NOx mitigation strategies:
•	Fully operating existing SCRs
•	Installing or upgrading NOx combustion controls
(2)	Estimate the total NOx reductions that are attributable to each of these strategies:70
•	Fully operating existing SCRs (SCRs operating in base case): 24,100 tons
•	Fully operating existing SCRs (SCRs not operating in base case): 4,500 tons
•	Installing or upgrading NOx combustion controls: 9,700 tons
69	This includes optimizing the performance of SCRs that were not operating.
70	For more information on how NOx reductions were attributed to strategies, see the excel files in the docket for
this rule entitled "Illustrative Cases Reduction Analysis 2018 and 2020 for RIA" and "Final Policy Case Reduction
Analysis 2018 and 2020 for RIA".
4-15

-------
(3)	Estimate the average cost associated with each of these strategies:71
•	Fully operating existing SCRs (SCRs operating in base case): $670/ton
•	Fully operating existing SCRs (SCRs not operating in base case): $l,000/ton
•	Installing or upgrading NOx combustion controls: $l,200/ton
(4)	Multiply (2) by (3) to estimate the total cost associate with each of these strategies.
Table 4-3 summarizes the results of this methodology for the illustrative CSAPR Update
scenario in 2017.
Table 4-3. Summary of Methodology for Calculating Compliance Costs Estimated Outside
of IPM for CSAPR Update, 2017 (2011$)
NOx Ozone Season
Emissions	Average Cost Total Cost
NOx Mitigation Strategy	(Tons)	($/ton)	($MM)
Maximizing the use of existing SCRs
(operating in Base Case) 24,100	670 16
Maximizing the use of existing SCRs
(not operating in Base Case) 4,500	1,000 4.5
Installing/upgrading NOx combustion
controls72 9,700	1,200 12
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 3, the EPA determined that NOx emissions in 22 eastern states
affect the ability of downwind states to attain and maintain the 2008 ozone NAAQS. For these
22 eastern states, the EPA is issuing Federal Implementation Plans (FIPs) that generally update
71	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.
72	This includes 3,030 tons of reductions from combustion control retrofits in Arkansas that are not expected until
2018.
4-16

-------
the existing CSAPR NOx ozone-season emission budgets for EGUs and implement these
budgets via the CSAPR NOx ozone-season allowance trading program.
The NOx emissions reductions are presented in this RIA for two time periods: 2017 (the
principal year of interest for the CSAPR update) and 2020. As with proposal RIA, the 2017
emissions estimates are based on IPM projections for 2018, and reflect exogenous adjustments to
account for known differences between 2017 and 2018 (e.g., planned closures, coal-to-gas
conversions, and planned SCR retrofits). For more information on these and other adjustments,
see 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 22-
state region, as well as the impact on states not in the region. The emission reductions follow an
expected pattern: the less stringent alternative produces substantially smaller emission reductions
than EPA's final emissions budgets, and the more stringent alternative results in slightly more
NOx reductions.
Table 4-4. EGU Ozone Season NOx Emissions and Emission Changes (thousand tons) for
the Base Case and the Regulatory Control Alternatives
Ozone Season NOx
(thousand tons)
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Region
369.5
308.3
342.7
303.2
-61.2
-26.8
-66.3
2017 Non-Region
205.4
205.3
205.4
205.5
-0.1
0.0
0.2
Total
574.8
513.5
548.1
508.8
-61.3
-26.8
-66.1
Region
374.6
302.8
347.7
297.8
-71.8
-26.9
-76.8
2020 Non-Region
181.6
181.5
181.6
181.8
-0.1
0.0
0.2
Total
556.2
484.3
529.3
479.6
-71.9
-26.9
-76.6
The results of EPA's IPM analysis show that, with respect to compliance with the
illustrative EGU NOx emission budgets, maximizing the use of existing operating SCRs
provides the largest amount of ozone season NOx emission reductions 42 percent), and turning
on idled SCRs produces an additional 32 percent of the total ozone season NOx reductions.
Combustion controls (16 percent) and generation shifting (10 percent) make up the remainder of
4-17

-------
the ozone season NOx reductions. In the more stringent alternative, compliance by turning on
idle existing SNCRs makes up 1 percent of the total reductions and generation shifting increases
to 16 percent, while the shares attributed to the other four mitigation measures are similar to, if
slightly smaller than, the shares for compliance with the finalized EGU NOx emissions budgets.
In the less stringent alternative, compliance by maximizing the use of existing operating SCRs
provides 85% of the total reductions, with the remainder attributable to generation shifting.
In addition to the ozone season NOx reductions, there will also be reductions of other air
emissions emitted by EGUs burning fossil fuels (i.e., co-pollutants). These other emissions
include the annual total changes in emissions of NOx, SO2 and CO2. The small SO2 emissions
increase is attributable primarily to a few model plants, for which the model projected a slightly
different 2016 MATS control strategy in the base case than with the CSAPR Update, resulting in
a small change in SO2 emissions. Since the MATS rule is currently effective, the EPA believes
that the MATS control strategies at these plants are currently in place, and not likely to change as
a result of the CSAPR Update. Therefore, the EPA does not view the projected SO2 increase as a
meaningful impact of the policy. The co-pollutant emission reductions are presented in Table 4-
5.
Table 4-5. EGU Annual Emissions and Emissions Changes for NOx, SO2 and CO2 for the
Regulatory Control Alternatives
Annual NOx
(thousand tons)
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Region
806.6
732.2
779.7
727.3
-74.5
-26.9
-79.3
2017 Non-Region
439.1
439.0
439.1
439.2
0.0
0.0
0.1
Total
1,245.7
1,171.2
1,218.8
1,166.5
-74.5
-26.9
-79.2
Region
820.2
735.5
793.2
730.6
-84.7
-27.0
-89.5
2020 Non-Region
415.3
415.3
415.3
415.4
0.0
0.0
0.1
Total
1,235.5
1,150.7
1,208.5
1,146.0
-84.8
-27.0
-89.4

Annual SO2
(thousand tons)
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Region
914.8
918.9
915.9
922.1
4.1
1.1
7.3
2017 Non-Region
324.1
322.1
323.7
321.7
-2.0
-0.4
-2.4
Total
1,238.9
1,241.0
1,239.6
1,243.8
2.2
0.7
5.0
4-18

-------
Region
914.8
918.9
915.9
922.1
4.1
1.1
7.3
2020 Non-Region
324.1
322.1
323.7
321.7
-2.0
-0.4
-2.4
Total
1,238.9
1,241.0
1,239.6
1,243.8
2.2
0.7
5.0

Annual CO2
(MM metric tonnes)
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Region
1,237.2
1,235.5
1,235.8
1,234.9
-1.7
-1.4
-2.3
2017 Non-Region
653.5
653.6
653.6
653.7
0.1
0.1
0.3
Total
1,890.7
1,889.1
1,889.4
1,888.6
-1.6
-1.3
-2.0
Region
1,237.2
1,235.5
1,235.8
1,234.9
-1.7
-1.4
-2.3
2020 Non-Region
653.5
653.6
653.6
653.7
0.1
0.1
0.3
Total
1,890.7
1,889.1
1,889.4
1,888.6
-1.6
-1.3
-2.0
4.4.2 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. The costs associated with compliance with monitoring,
recordkeeping, and reports requirements are not included within the estimates in this table and
can be found in preamble section X.B.
Table 4-6. Compliance Cost Estimates (millions of 2011$) for the Regulatory Control
Alternatives

CSAPR Update
Less-Stringent
Alternative
More-Stringent
Alternative
2017-2020 (Annualized)
68.4
8.0
82.0
2017 (Annual)
0.01
-55.7
-0.4
2020 (Annual)
136.9
77.1
164.6
"2017-2020 (Annualized)" reflects total estimated annual compliance costs levelized over the period 2017 through
2020, discounted using a 4.77 discount rate. "2017 (Annual)" and "2020 (Annual)" reflect point 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 and more stringent
alternatives are negative (i.e., a cost reduction) in 2017, although these regulatory control
alternatives reduce annual NOx emissions by approximately 27,000 and 79,000 tons respectively
4-19

-------
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.73 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 which 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.
In addition to evaluating annual compliance cost impacts, the EPA believes that a full
understanding of these three regulatory control alternatives benefits from an evaluation of
annualized costs over the 2017-2020 time frame. EPA limits its analysis to this timeframe
considering that on October 1, 2015, the EPA strengthened the ground-level ozone NAAQS to
70 ppb. The EPA is mindful of the need to address ozone transport for the 2015 ozone NAAQS.
Given the statutory implementation timeline of good neighbor requirements with respect to the
2015 ozone NAAQS, the EPA anticipates that further actions to reduce interstate emission
transport related to ozone pollution could take place in the near future.74 Therefore, it is
appropriate to evaluate the costs of the regulatory control alternatives over the 2017-2020
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.75 For this analysis we first calculated the
NPV of the stream of costs from 2017 through 202076 using a 4.77 percent discount rate. EPA
typically uses a 3 and a 7 percent discount rate to discount future year social benefits and social
costs in regulatory impact analyses (USEPA, 2010). In this cost annualization we use a 4.77
73	For more information, see Chapter 2 of the IPM Documentation.
74	See preamble section VII.
75	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.
76	Consistent with the relationship between IPM run years and calendar years, the EPA assigned 2020 compliance
cost estimates to both 2019 and 2020 in the calculation of NPV. For more information, see Chapter 7 of the IPM
Documentation.
4-20

-------
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.77
After calculating the NPV of the cost streams, the same 4.77 percent discount rate and
2017-2020 time period is used to calculate the levelized annual (i.e., annualized) cost estimates
shown in Table 4-6.78
Additionally, note that the 2017-2020 equivalent annualized compliance cost estimates
have the expected relationship to each other; the annualized costs are lowest for the less stringent
alternatives, and highest for the more stringent alternative.
4.4.3 Impacts on Fuel Use, Prices and Generation Mix
While the CSAPR Update 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.
Tables 4-7 and 4-8 present the percentage changes in national coal and natural gas usage
by EGUs in 2017 These fuel use estimates reflect a modest shift to natural gas from coal. The
projected impacts in 2020 are similarly very small.
Table 4-7.2017 Projected Power Sector Coal Use for the Base Case and the Regulatory
Control Alternatives
Million Tons
Percent
Change from Base Case



Less-
More-

Less-
More-

Base
CSAPR
Stringent
Stringent
CSAPR
Stringent
Stringent

Case
Update
Alternative
Alternative
Update
Alternative
Alternative
Appalachia
118
117
118
117
-0.2%
-0.1%
-0.4%
77	The IPM Base Case documentation (Section 8.2.1 Introduction to Discount Rate Calculations) states "The real
discount rate for expenditures (e.g., capital, fuel, variable operations and maintenance, and fixed operations and
maintenance costs) in the EPA Base Case v.5.13 is 4.77%. This serves as the default discount rate for all
expenditures."
78	The PMT() function in Microsoft Excel 2013 is used to calculate the level annualized cost from the estimated
NPV.
4-21

-------
Imports
0
0
0
0
N/A
N/A
N/A
Interior
227
227
227
227
0.0%
0.0%
0.0%
Waste Coal
6
6
6
6
0.0%
0.0%
0.0%
West
352
351
351
350
-0.4%
-0.4%
-0.5%
Total
703
701
701
700
-0.2%
-0.2%
-0.3%
Table 4-8. 2017 Projected Power Sector Natural Gas Use for the Base Case and the
Regulatory Control Alternatives
Trillion Cubic Feet
Percent Change from Base Case
Base Case
CSAPR Less-Stringent
Update Alternative
More-
Stringent
Alternative
CSAPR Less-Stringent
Update Alternative
More-
Stringent
Alternative
00
00
00
00
00
00
00
00
0.2% 0.1%
0.3%
Tables 4-9 and 4-10 present the projected coal and natural gas prices in 2017, as well as
the percent change from the base case projected as a result of the regulatory control alternatives.
These minor impacts in 2017 are consistent with the small changes in fuel use summarized
above. The projected impacts in 2020 are similarly very small.
Table 4-9. 2017 Projected Minemouth and Power Sector Delivered Coal Price for the Base
Case and the Regulatory Control Alternatives
$/MMBtu
Percent Change from Base Case

Base
Case
Less-
CSAPR Stringent
Update Alternative
More-
Stringent
Alternative
Less- More-
CSAPR Stringent Stringent
Update Alternative Alternative
Minemouth
1.51
1.51 1.51
1.51

-0.1% -0.2% -0.3%
Delivered
2.31
2.31 2.31
2.31

-0.1% -0.2% -0.2%
Table 4-10. 2017 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

Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Henry Hub
4.33
4.33
4.33
4.33
0.0%
0.0%
0.1%
Delivered
4.52
4.52
4.52
4.53
0.0%
0.0%
0.0%
4-22

-------
Table 4-11 presents the projected percentage changes in the amount of electricity
generation in 2017 by fuel type. Consistent with the fuel use projections and emissions trends
above, the EPA projects very small overall shift from coal to gas. The projected impact in 2020
is similarly very small.
Table 4-11. 2017 Projected Generation by Fuel Type for the Base Case and the Regulatory
Control Alternatives
Generation (MWh)
Percent Change from Base Case

Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Coal
1,388
1,386
1,387
1,385
-0.2%
-0.1%
-0.2%
Natural Gas
1,195
1,198
1,197
1,199
0.2%
0.1%
0.3%
Nuclear
787
787
787
787
0.0%
0.0%
0.0%
Hydro
281
281
281
281
0.0%
0.0%
0.0%
Non-Hydro RE
421
421
421
421
0.0%
0.0%
0.0%
Oil\Gas Steam
50
50
50
50
0.0%
0.0%
0.0%
Other
8
8
8
8
0.9%
-1.2%
0.4%
Total
4,131
4,131
4,131
4,131
0.0%
0.0%
0.0%
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 2020 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 2017, and the model
was specified accordingly.
Table 4-12. 2020 Projected Capacity by Fuel Type for the Base Case and the Regulatory
Control Alternatives
Capacity (GW)
Percent Change from Base Case

Base Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
Coal
209
209
209
209
-0.3%
-0.2%
-0.3%
Natural Gas
391
391
391
391
0.0%
0.0%
0.0%
Nuclear
101
101
101
102
0.3%
0.1%
0.3%
Hydro
107
107
107
107
0.0%
0.0%
0.0%
Non-Hydro RE
138
138
138
138
0.0%
0.0%
0.0%
Oil\Gas Steam
83
83
83
83
0.0%
0.0%
0.0%
Other
5
5
5
5
0.0%
0.0%
0.0%
Total
1,035
1,035
1,035
1,035
0.0%
0.0%
0.0%
Note: In this table, "Non-Hydro RE" includes biomass, geothermal, landfill gas, solar, and wind
4-23

-------
The EPA estimated the change in the retail price of electricity (2011$) using the Retail
Price Model (RPM).79 The RPM was developed by ICF International for the 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 in the
electricity market module of the National Energy Modeling System (NEMS).80
Tables 4-13 and 4-14 present the projected percentage changes in the retail price of
electricity for the three regulatory control alternatives in 2017 and 2020, 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 2020, the EPA estimates that this rule will result
in a 0.1% increase in national average retail electricity price, or by about 0.1 mills/kWh (about
0.01 cents/kWh).
Table 4-13. Average Retail Electricity Price by Region for the Base Case and the
Regulatory Control Alternatives, 2017
2017 Average Retail Electricity Price
(2011 mills/kWh)
Percent Change from Base Case
Region
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
ERCT
79.5
79.5
79.5
79.6
0.0%
0.0%
0.1%
FRCC
102.3
102.2
102.2
102.2
-0.1%
-0.1%
-0.1%
MROE
100.4
100.4
100.4
100.4
0.0%
0.0%
0.0%
MROW
87.6
87.5
87.6
87.5
-0.1%
-0.1%
-0.1%
NEWE
126.8
126.8
126.8
126.8
0.0%
0.0%
0.0%
NYCW
166.2
166.2
166.2
166.2
0.0%
0.0%
0.0%
NYLI
136.3
136.3
136.3
136.4
0.0%
0.0%
0.0%
NYUP
119.2
119.3
119.2
119.3
0.1%
0.0%
0.1%
79	See documentation available at: https://www.epa.gov/airmarkets/power-sector-modeling
80	See documentation available at:
http://www.eia. gov/forecasts/aeo/nems/documentation/electricity/pdf/m068(2014).pdf
4-24

-------
RFCE
103.1
103.0
103.5
103.1
-0.1%
0.4%
0.0%
RFCM
103.0
103.0
102.9
103.0
0.0%
-0.1%
0.0%
RFCW
88.6
88.7
88.6
88.7
0.1%
0.0%
0.1%
SRDA
82.5
82.5
82.4
82.5
0.0%
-0.1%
0.0%
SRGW
83.8
83.8
83.8
83.8
0.0%
0.0%
0.0%
SRSE
101.6
101.6
101.5
101.6
0.0%
-0.1%
0.0%
SRCE
79.7
79.7
79.6
79.6
0.0%
-0.1%
-0.1%
SRVC
98.3
98.3
98.3
98.3
0.0%
0.0%
0.0%
SPNO
102.2
102.2
102.2
102.1
0.0%
0.0%
-0.1%
SPSO
79.0
79.1
79.0
79.2
0.1%
0.1%
0.2%
AZNM
109.6
109.6
109.6
109.6
0.0%
0.0%
0.0%
CAMX
145.5
145.5
145.5
145.5
0.0%
0.0%
0.0%
NWPP
72.6
72.6
72.6
72.6
0.0%
0.0%
0.0%
RMPA
87.1
87.1
87.1
87.1
0.0%
0.0%
0.0%
NATIONAL
97.3
97.3
97.3
97.3
0.0%
0.0%
0.0%
Table 4-14. Average Retail Electricity Price by Region for the Base Case and the
Regulatory Control Alternatives, 2020
2020 Average Retail Electricity Price
(2011 mills/kWh)
Percent Change from Base Case
Region
Base
Case
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
CSAPR
Update
Less-
Stringent
Alternative
More-
Stringent
Alternative
ERCT
88.6
88.7
88.7
88.8
0.1%
0.1%
0.2%
FRCC
104.3
104.4
104.4
104.4
0.1%
0.1%
0.1%
MROE
99.1
99.1
99.1
99.1
0.0%
0.0%
0.0%
MROW
87.7
87.8
87.8
87.8
0.1%
0.1%
0.1%
NEWE
130.6
130.7
130.7
130.7
0.1%
0.0%
0.1%
NYCW
171.9
172.1
172.0
172.1
0.1%
0.1%
0.1%
NYLI
141.6
141.7
141.6
141.7
0.1%
0.0%
0.1%
NYUP
123.1
123.3
123.2
123.3
0.2%
0.1%
0.2%
RFCE
108.1
108.3
108.2
108.3
0.2%
0.1%
0.2%
RFCM
103.7
103.8
103.8
103.8
0.1%
0.0%
0.1%
RFCW
91.4
91.6
91.7
91.7
0.2%
0.3%
0.3%
SRDA
85.5
85.6
85.6
85.6
0.1%
0.1%
0.1%
SRGW
85.9
86.0
86.0
86.1
0.1%
0.2%
0.3%
SRSE
100.4
100.4
100.4
100.4
0.0%
0.0%
0.0%
SRCE
80.2
80.2
80.2
80.2
0.0%
0.0%
0.0%
SRVC
97.7
97.8
97.7
97.7
0.1%
0.0%
0.1%
SPNO
101.1
101.1
101.1
101.0
0.0%
0.0%
-0.1%
SPSO
81.7
81.9
81.8
82.0
0.2%
0.1%
0.3%
4-25

-------
AZNM
110.6
110.6
110.6
110.6
0.0%
0.0%
0.0%
CAMX
144.4
144.4
144.4
144.4
0.0%
0.0%
0.0%
NWPP
69.4
69.4
69.4
69.4
0.0%
0.0%
0.0%
RYIPA
87.4
87.4
87.4
87.4
0.0%
0.1%
0.0%
NATIONAL
99.0
99.1
99.1
99.1
0.1%
0.1%
0.1%
MEWE
NWPP
NYUP
NYCW
SRGW
A2NM]
SRDA
Figure 4-1. Electricity Market Module Regions
Source: EI A (http ://www. eia. gov/forecasts/aeo/pdf/nerc_map.pdf)
4.5 Social Costs
As discussed in the EPA 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
4-26

-------
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 final and more or less stringent 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 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 final 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 in fact 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 final 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 final rule. These cost
estimates are based on rigorous power sector modeling using ICF's Integrated Planning Model.
IPM assumes "perfect foresight" of market conditions over the time horizon modeled; to the
4-27

-------
extent that utilities and/or energy regulators misjudge future conditions affecting the economics
of pollution control, costs may be understated.
As discussed in section 4.3.2, IPM v.5.15 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 Policy Analysis TSD. While the emission 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 types compliance
strategies (the operating costs of the units on which these strategies are imposed do not reflect
the additional costs of these strategies). These additional costs are relatively minor, and do not
have a significant impact on the overall finding that the economic impacts of this rule are
minimal.
Additionally, the modeling includes two emission reduction strategies that are exogenously
imposed where applicable: turning on idled SCRs (CSAPR Update and more-stringent
alternative) and turning on idled SNCRs (mores stringent alternative only). While these
strategies are exogenously imposed, the costs and emissions reductions are accounted for
endogenously. 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 CSAPR Update.
4.7 References
U.S. Energy Information Agency (EIA). 2014. The Electricity Market Module of the National
Energy Modeling System: Model Documentation 2014. Available at:
.
Accessed 9/17/2015.
4-28

-------
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-exi sting-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.
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/CSAPRyindex.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.
4-29

-------
APPENDIX 4A: COST, EMISSIONS, AND ENERGY IMPACTS OF FINAL CSAPR
UPDATE BUDGETS
This appendix reports the compliance costs, emissions, and energy analyses performed
for the final CSAPR Update NOx ozone season emission budgets. The tables below summarize
the analysis of the final emissions budgets, which differ slightly from the illustrative budgets
analyzed outside of this appendix. The differences between the results below and the results of
the illustrative budgets presented in this chapter are minor, consistent with the small differences
in NOx ozone season budgets and small updates to the modeling assumptions.81
Table 4A-1 CSAPR Update NOx Ozone Season Emission Budgets (Tons)
81 Consistent with the assumptions underlying the budgets, the final modeling assumes a NOx rate of 0.10
lbs/MMBtu for fully-operated SCRs (see Ozone Transport Policy Analysis Final Rule TSD). Additionally, as
discussed in the EPA v.5.15 CSAPR Update Rule Base Cases Using IPM Incremental Documentation, the NOx
rates of some units were updated to reflect recently observed performance.
State
CSAPR Update
Final Budgets
Alabama
Arkansas
Iowa
Illinois
Indiana
Kansas
Kentucky
Louisiana
Maryland
Michigan
Missouri
Mississippi
New Jersey
New York
Ohio
Oklahoma
Pennsylvania
Tennessee
Texas
Virginia
Wisconsin
West Virginia
13,211
9,210
11,272
14,601
23,303
8,027
21,115
18,639
3,828
16,545
15,780
6,315
2,062
5,135
19,522
11,641
17,952
7,736
52,301
9,223
7,915
17,815
4A-1

-------
TOTAL	313,148
Note: The budget displayed for Arkansas is its 2018 budget. In 2017, for all cases, Arkansas has a budget of 12,048.
Table 4A-2. EGU Ozone Season NOx Emissions and Emission Changes (thousand tons)
for the Base Case and the CSAPR Update
Ozone Season NOx



(thousand tons)
Base Case
CSAPR Update
Change
Region
371.7
319.8
-51.9
2017 Non-Region
206.4
206.4
0.0
Total
578.1
526.2
-51.9
Region
380.6
314.0
-66.6
2020 Non-Region
182.6
182.6
0.0
Total
563.2
496.6
-66.6
Table 4A-3. EGU Annual Emissions and Emissions Changes for NOx, SO2 and CO2 for
the CSAPR Update
Annual NOx
Base Case
CSAPR Update
Change
Region
812.4
750.3
-62.1
2017 Non-Region
441.1
441.1
0.0
Total
1,253.5
1,191.5
-62.1
Region
829.6
753.8
-75.8
2020 Non-Region
417.3
417.4
0.0
Total
1,246.9
1,171.2
-75.7
Annual SO2
(thousand tons) Base Case CSAPR Update Change
Region 909.4 919.4	10.0
2017 Non-Region 324.7 321.8	-2.9
	Total	1,234.1	1,241.2	7.0
Region 909.4 919.4	10.0
2020 Non-Region 324.7 321.8	-2.9
	Total	1,234.1	1,241.2	7.0
Annual CO2
(MM Metric Tonnes) Base Case CSAPR Update	Change
Region 1,235.9 1,235.4	-0.4
2017 Non-Region 653.4 653.6	0.1
	Total	1,889.3	1,889.0	-0.3
2020 Region 1,235.9 1,235.4	-0.4
4A-2

-------
Non-Region 653.4 653.6 0.1
Total	1,889.3	1,889.0 -0.3
Table 4A-4. Compliance Cost Estimates (millions of 2011$) for the CSAPR Update
CSAPR Update
2017-2020 (Annualized)
89.0
2017 (Annual)
-18.3
2020 (Annual)
198.2
"2017-2020 (Annualized)" reflects total estimated annual compliance costs levelized over the period 2017 through
2020, discounted using a 4.77 discount rate. "2017 (Annual)" and "2020 (Annual)" reflect point estimates in each of
those years. These costs do not include monitoring, reporting, and recordkeeping costs, which are estimated to be a
reduction of $1,347,291 peryear.
Table 4A-5. 2017 Projected Power Sector Coal Use for the Base Case and the CSAPR
Update
Percent
Change from
Base Case CSAPR Update	Base Case
116	117	1.1%
0	0	N/A
227	227	0.0%
6	6	0.0%
353	351	-0.6%
702	701	-0.1%
Appalachia
Imports
Interior
Waste Coal
West
Total
Table 4A-6. 2017 Projected Power Sector Natural Gas Use for the Base Case and the
CSAPR Update
Percent
Change from
	Base Case CSAPR Update	Base Case
Natural Gas Use	8.8	8.8	-0.01%
Table 4A-7. 2017 Projected Minemouth and Power Sector Delivered Coal Price for the
Base Case and the CSAPR Update



Percent



Change from

Base Case
CSAPR Update
Base Case
Minemouth
1.51
1.51
0.2%
Delivered
2.31
2.31
0.0%
4A-3

-------
Table 4A-8. 2017 Projected Henry Hub and Power Sector Delivered Natural Gas Price for
the Base Case and the CSAPR Update



Percent

Base Case
CSAPR Update
Change from
Base Case
Henry Hub
4.34
4.33
-0.1%
Delivered
4.53
4.52
-0.1%
Table 4A-9. 2017 Projected Generation by Fuel Type for the Base Case and the CSAPR
Update
Percent
Change from
Base Case CSAPR Update	Base Case
Coal
1,386
1,386
0.0%
Natural Gas
1,198
1,198
0.0%
Nuclear
787
787
0.0%
Hydro
281
281
0.2%
Non-Hydro RE
421
421
0.0%
Oil\Gas Steam
50
50
0.0%
Other
8
8
0.3%
Total
4,130
4,131
0.0%
Note: In this table, "Non-Hydro RE" includes biomass, geothermal, landfill gas, solar, and wind
Table 4A-10. 2020 Projected Capacity by Fuel Type for the Base Case and the CSAPR
Update
Percent
Change from
Base Case CSAPR Update	Base Case
Coal
209
209
-0.3%
Natural Gas
391
391
0.0%
Nuclear
101
101
0.3%
Hydro
107
107
0.0%
non-Hydro RE
138
138
0.0%
Oil\Gas Steam
83
83
0.0%
Other
5
5
0.0%
Total
1,035
1,035
0.0%
Note: In this table, "Non-Hydro RE" includes biomass, geothermal, landfill gas, solar, and wind
Table 4A-11. Average Retail Electricity Price by Region for the Base Case and the CSAPR
Update, 2017
Percent
CSAPR Change from
Region	Base Case Update Base Case
4A-4

-------
ERCT
79.6
79.5
0.0%
FRCC
102.3
102.2
-0.1%
MROE
100.4
100.4
0.0%
MROW
87.5
87.5
0.0%
NEWE
126.8
126.8
0.0%
NYCW
166.3
166.2
-0.1%
NYLI
136.4
136.4
0.0%
NYUP
119.3
119.3
0.0%
RFCE
103.2
103.1
-0.1%
RFCM
103.0
103.0
0.0%
RFCW
88.6
88.7
0.1%
SRDA
82.5
82.5
0.0%
SRGW
83.8
83.8
0.1%
SRSE
101.6
101.6
0.0%
SRCE
79.7
79.7
0.0%
SRVC
98.3
98.3
0.0%
SPNO
102.1
102.1
0.0%
SPSO
79.0
79.1
0.1%
AZNM
109.6
109.6
0.0%
CAMX
145.5
145.5
0.0%
NWPP
72.6
72.6
0.0%
RMPA
87.1
87.1
0.0%
NATIONAL
97.3
97.3
0.0%
Table 4A-12. Average Retail Electricity Price by Region for the Base Case and the CSAPR
Update, 2020
Percent
CSAPR Change from
Region
Base Case
Update
Base Case
ERCT
88.6
88.7
0.1%
FRCC
104.3
104.4
0.1%
MROE
99.1
99.1
0.0%
MROW
87.7
87.8
0.1%
NEWE
130.6
130.7
0.1%
NYCW
171.9
172.1
0.1%
NYLI
141.5
141.7
0.1%
NYUP
123.2
123.3
0.1%
RFCE
108.2
108.3
0.1%
RFCM
103.7
103.8
0.1%
RFCW
91.4
91.6
0.2%
SRDA
85.5
85.6
0.1%
SRGW
85.9
85.9
0.1%
4A-5

-------
SRSE
100.4
100.4
0.1%
SRCE
80.1
80.2
0.1%
SRVC
97.7
97.8
0.0%
SPNO
101.1
101.1
0.0%
SPSO
81.7
81.9
0.2%
AZNM
110.6
110.6
0.0%
CAMX
144.3
144.4
0.0%
NWPP
69.4
69.4
0.0%
RMPA
87.4
87.4
0.0%
NATIONAL
99.0
99.1
0.1%
4A-6

-------
CHAPTER 5: ESTIMATED HUMAN HEALTH BENEFITS AND CLIMATE CO-
BENFITS
5.1 Introduction
As discussed above, this final rule is an update of the Cross-State Air Pollution Rule
(CSAPR) to further reduce interstate transport of Electricity Generating Unit (EGU) ozone
season nitrogen oxides (NOx) emissions that contribute significantly to nonattainment or that
interfere with maintenance of the 2008 ozone National Ambient Air Quality Standard (NAAQS).
The EPA is implementing emission budgets for EGU NOx emissions through the CSAPR NOx
ozone season allowance trading program. Updating the CSAPR in this way will reduce emissions
of NOx during the summer ozone season and provide ancillary annual NOx and carbon dioxide
(CO2) benefits (i.e., co-benefits). This chapter describes the methods used to estimate the
monetized ozone-related air quality health benefits, the fine particulate matter (PM2.5)-related air
quality health co-benefits from reductions in NOx emissions, and climate co-benefits from
reductions of CO2 emissions. These health benefits are associated with reducing exposure to
ambient ozone and PM2.5 by reducing emissions of precursor pollutants (i.e., NOx). Data,
resource, and methodological limitations prevent the EPA from monetizing several important co-
benefits from reducing emissions of pollutants including SO2 and VOC as well as reduced
ecosystem effects and visibility impairment associated with reductions in NOx. We discuss these
and other unquantified benefits further in this chapter.
This chapter reports estimates of the monetized air quality health benefits and climate co-
benefits associated with emission reductions for the CSAPR Update and two regulatory control
alternatives across several discount rates. The estimated benefits associated with these emission
reductions are beyond those achieved by previous EPA air quality rules, including the original
CSAPR that affected cross-state transport of NOx and SO2.82
82 For reasons described in section IV.B of the preamble for the final CSAPR Update, the Clean Power Plan is not
included in the baseline for this analysis. Section 4.3.1 of this RIA discusses the treatment of remanded CSAPR
budgets.
5-1

-------
5.2 Estimated Human Health Benefits
The CSAPR update is expected to reduce emissions of ozone season NOx. In the presence
of sunlight, NOx and VOCs can undergo a chemical reaction in the atmosphere to form ozone.
Reducing NOx emissions also reduces human exposure to ozone and the incidence of ozone-
related health effects, though this depends partly on local levels of volatile organic compounds
(VOCs). The CSAPR update will also reduce emissions of NOx throughout the year. Because
NOx is also a precursor to formation of ambient PM2.5, reducing these emissions would also
reduce human exposure to ambient PM2.5 throughout the year and would reduce the incidence of
PM2.5-related health effects.83 This RIA does not quantify PIVh.s-related benefits associated with
SO2 emission changes. (For further explanation of the modeled SO2 emissions changes, see
Chapter 4, section 4.4.1).
The benefits estimates reported in this chapter are limited to those that would occur in the
22-state final CSAPR Update region. Reducing NOx may also reduce ozone and PM2.5
concentrations in areas outside the 22 states that are the subject of the CSAPR Update, though
the impact of reducing these pollutants in those areas are not assessed in this Chapter. Reducing
emissions of NOx would reduce ambient exposure to NO2 (which is a product of combustion)
and its associated health effects, though we do not quantify these effects because we lacked
sufficient data to quantify these effects. A full description of the epidemiological studies we use,
the methods we apply and the tools we employ to quantify the incidence of these effects may be
found in the PM NAAQS RIA (U.S. EPA, 2012a) and Ozone NAAQS RIA (U.S. EPA, 2015).
Implementing these updated CSAPR EGU NOx emissions budgets for the ozone season in
22 eastern states may reduce ambient ozone and PM2.5 concentrations below the National
Ambient Air Quality Standards (NAAQS) in some areas and assist other areas with attaining the
ozone and PM2.5 NAAQS. The NAAQS RIAs (U.S. EPA, 2008, 2012a, 2015) also calculated the
benefits of attaining alternate ozone and PM NAAQS, and so differences in the design and
analytical objectives of each RIA are worth noting here. The NAAQS RIAs illustrate the
potential costs and benefits of attaining a revised air quality standard nationwide based on an
array of emission reduction strategies for different sources reflecting the application of identified
83 Additionally, this RIA does not estimate changes in emissions of directly emitted particles.
5-2

-------
and unidentified controls, incremental to implementation of existing regulations and controls
needed to attain the NAAQS that currently is in effect. In short, NAAQS RIAs hypothesize, but
do not predict, the strategies that States may choose to enact when implementing a revised
NAAQS. Setting a NAAQS does not directly result in costs or benefits, and as such, the EPA's
NAAQS RIAs are illustrative. The estimated costs and benefits from NAAQS RIAs are not
intended to be added to the costs and benefits of other regulations that result in specific costs of
control and prescribe specific emission reductions. Indeed, some of the emissions reductions
estimated to result from implementing the CSAPR update may achieve some of the air quality
improvements that resulted from the hypothesized attainment strategies presented in the NAAQS
RIAs. The CSAPR Update is intended to achieve the air quality improvements identified in the
RIA for the 2008 NAAQS, with appropriate adjustments to baseline conditions between the
analysis in that RIA and the analysis presented in this RIA. Implementing this CSAPR Update
will assist downwind areas in attaining and maintaining the 2008 ozone NAAQS. The ambient
ozone reduced by this rule would also achieve some of the air quality improvements assumed in
the baseline for the 2015 ozone NAAQS RIA.84
As discussed in Chapter 4, the IPM modeling showing compliance with the CSAPR update
and two regulatory control alternatives for which emission reductions are estimated in this RIA is
one possible path for compliance with the CSAPR Update emissions budgets. However, the EPA
believes the magnitude and location of the air quality changes are well characterized because the
rule limits emissions from a specific sector. Emissions reduced by this rule will ultimately be
reflected in the baseline of future NAAQS analyses and would lower the additional emissions
reductions needed to attain revised future NAAQS. For more information on the relationship
between illustrative analyses, such as for the NAAQS and its associated implementation rules,
please see the Ozone NAAQS RIA (U.S. EPA, 2015).
5.2.1 Health Impact Assessment for Ozone and PM2.5
The Integrated Science Assessment for Ozone and Related Photochemical Oxidants
(Ozone ISA) (U.S. EPA, 2013b) identified the human health effects associated with chronic and
84 In other words, the 2015 ozone NAAQS RIA evaluated the costs and benefits of attaining the 2015 NAAQS,
starting from a baseline that included attainment of the 2008 NAAQS.
5-3

-------
acute ambient ozone exposure, which include premature death and a variety of morbidity effects.
Similarly, the Integrated Science Assessment for Particulate Matter (PM ISA) (U.S. EPA,
2009b) identified the human health effects associated with ambient PM2.5 exposure, which
include premature death and a variety of morbidity effects associated with acute and chronic
exposures. Table 5-1 identifies the quantified and monetized benefit and co-benefit categories
captured in the EPA's health benefits estimates for reduced exposure to ambient ozone and
PM2.5. Although the table below does not list unquantified health effects or welfare effects, such
as acidification and nutrient enrichment, these effects are described in detail in Chapters 5 and 6
of the PM NAAQS RIA (U.S. EPA, 2012a) and summarized later in this chapter. The list of
unquantified benefits categories is not exhaustive and effects may not have been quantified
completely.
5-4

-------
Table 5-1. Human Health Effects of Ambient Ozone and PM2.5


Effect Has
Effect Has
More
Information
Category
Specific Effect
Been
Quantified
Been
Monetized
Improved Human Health
Reduced incidence of
mortality from
exposure to ozone
Premature mortality based on short-term study
estimates (all ages)
~
~
Ozone ISA
Premature mortality based on long-term study
estimates (age 30-99)
—
—
Ozone ISA1

Hospital admissions—respiratory causes (age > 65)
~
~
Ozone ISA

Hospital admissions—respiratory causes (age <2)
~
~
Ozone ISA

Emergency department visits for asthma (all ages)
~
~
Ozone ISA
Reduced incidence of
morbidity from
exposure to ozone
Minor restricted-activity days (age 18-65)
~
~
Ozone ISA
School absence days (age 5-17)
~
~
Ozone ISA
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
Reduced incidence of
premature mortality
Adult premature mortality based on cohort study
estimates and expert elicitation estimates (age >25
~
~
PM ISA
from exposure to
PM2.5
or age >30)



Infant mortality (age <1)
~
~
PM ISA

Non-fatal heart attacks (age >18)
V
V
PM ISA

Hospital admissions—respiratory (all ages)
V
V
PM ISA

Hospital admissions—cardiovascular (age >20)
V
V
PM ISA

Emergency room visits for asthma (all ages)
V
V
PM ISA

Acute bronchitis (age 8-12)
V
V
PM ISA

Lower respiratory symptoms (age 7-14)
V
V
PM ISA

Upper respiratory symptoms (asthmatics age 9-11)
V
V
PM ISA

Asthma exacerbation (asthmatics age 6-18)
V
V
PM ISA
Reduced incidence of
morbidity from
exposure to PM2 5
Lost work days (age 18-65)
V
V
PM ISA
Minor restricted-activity days (age 18-65)
V
V
PM ISA
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
1	We assess these co-benefits qualitatively due to data and resource limitations for this analysis, but we have quantified them in
sensitivity analyses for other analyses.
2	We assess these co-benefits qualitatively because we do not have sufficient confidence in available data or methods.
3	We assess these co-benefits qualitatively because current evidence is only suggestive of causality or there are other significant
concerns over the strength of the association.
5-5

-------
We follow a "damage-function" approach in calculating benefits, which estimates changes
in individual health endpoints and assigns a dollar value to those changes. Because the EPA
rarely has the time or resources to perform new research to measure directly either health
outcomes or their values for regulatory analyses, our estimates are based on the best available
methods of benefits transfer, which is the science and art of adapting primary research from
similar contexts to estimate benefits for the environmental quality change under analysis. We use
two benefits transfer techniques to quantify the ozone and PIVh.s-attributable benefits. We first
perform a health impact assessment (HIA) to estimate the avoided deaths and illnesses resulting
from implementing the CSAPR Update. We next use a "benefit-per-ton" approach to estimate
the ozone and PM2.5 benefits of the CSAPR Update and the more and less stringent alternatives.
An HIA quantifies the changes in the incidence of adverse health impacts resulting from
changes in human exposure to ozone and PM2.5. We use the environmental Benefits Mapping
and Analysis Program - Community Edition (BenMAP-CE) (version 1.1) to calculate a health
impact function that combines information from the modeled air quality predictions for this rule
with a database of key input parameters, including population projections, health impact
functions, and valuation functions (EPA, 2014). For this assessment, the HIA is limited to those
health effects that are directly linked to ambient ozone and PM2.5 concentrations. There may be
other indirect health impacts associated with reducing emissions, such as occupational health
exposures. Epidemiological studies generally provide estimates of the relative risks of a
particular health effect for a given increment of air pollution (often per 10 ppb for ozone or per
10 |ig/m3 for PM2.5). These relative risks can be used to develop risk coefficients that relate a
unit reduction in pollution (e.g., ozone) to changes in the incidence of a health effect. We refer
the reader to the Ozone NAAQS RIA (U.S. EPA, 2015) and PM NAAQS RIA (U.S. EPA,
2012a) for more information regarding the epidemiology studies and risk coefficients applied in
this analysis.
The final air quality modeling simulation predicted changes in ozone and PM2.5 from a
baseline scenario that did not fully account for certain emission changes that are reflected in the
policy scenario. Chapter 4 describes in greater detail how the emissions baseline was
subsequently modified to account for the Pennsylvania RACT as well as other smaller-scale
changes to the estimated EGU-level emissions. Because we could not use these air quality
5-6

-------
predictions directly, we instead employed a benefit-per-ton approach. Using the BenMAP-CE
tool noted above, we first quantified the change in the number of ozone and PIVh.s-attributable
avoided deaths and illnesses, and the dollar value of these outcomes, estimated to result from the
modeled air quality scenario relative to the baseline. We divide these values by the change in
emissions to calculate an average benefit per ton. Thus, to develop estimates of benefits for this
RIA, we are transferring both the underlying health and economic information from a final air
quality modeling scenario to the illustrative policy emissions reductions, including more and less
stringent policy alternatives. Below, we describe in greater detail the data we used to calculate
these benefit per ton values.
Before describing our technique for calculating the benefit per ton estimates, we briefly
elaborate on the procedure for estimating the incidence of adult premature deaths in this RIA
below. The size of the mortality effect estimates from epidemiological studies, the serious nature
of the effect itself, and the high monetary value ascribed to reducing risks of premature death
make mortality risk reduction the most significant health endpoint quantified in this analysis.
5.2.1.1 Mortality Effect Coefficients for Short-term Ozone Exposure
The overall body of evidence indicates that there is likely to be a causal relationship
between short-term ozone exposure and premature death. The 2013 ozone Integrated Science
Assessment (ISA) concludes that the evidence suggests that ozone effects are independent of the
relationship between PM and mortality. (U.S. EPA, 2013a). However, the ISA notes that the
interpretation of the potential confounding effects of PM on ozone-mortality risk estimates
requires caution due to the PM sampling schedule (in most cities) which limits the overall
sample size available for evaluating potential confounding of the ozone effect by PM (U.S. EPA
2013a).
In 2006, the EPA requested a National Academies of Sciences (NAS) study to answer the
following four key questions regarding ozone-related mortality: (1) How did the epidemiological
literature to that point improve our understanding of the size of the ozone-related mortality
effect?; (2) How best can EPA quantify the level of ozone-related mortality impacts from short-
term exposure?; (3) How might EPA estimate the change in life expectancy?; and (4) What
5-7

-------
methods should EPA use to estimate the monetary value of changes in ozone-related mortality
risk and life expectancy?
In 2008, the NAS (NRC, 2008) issued a series of recommendations to the EPA regarding
the quantification and valuation of ozone-related short-term mortality. 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 multi-city and National Morbidity and Mortality Air Pollution Studies
(NMMAPS) studies without exclusion of the meta-analyses" (NRC, 2008). In addition, NAS
recommended that EPA "should give little or no weight to the assumption that there is no causal
association between estimated reductions in premature mortality and reduced ozone exposure"
(NRC, 2008). In 2010, the Health Effects Subcommittee of the Advisory Council on Clean Air
Compliance Analysis, while reviewing EPA's The Benefits and Costs of the Clean Air Act 1990
to 2020 (U.S. EPA, 201 la), also confirmed the NAS recommendation to include ozone mortality
benefits (U.S. EPA-SAB, 2010a).
In view of the findings of the ozone ISA, the NAS panel, the Science Advisory Board—
Health Effects Subcommittee (SAB-HES) panel, and the Clean Air Scientific Advisory
Committee (CASAC) panel, we estimate ozone-related premature mortality for short-term
exposure in the core health effects analysis using effect coefficients from the Smith et al. (2009)
NMMAPS analysis and the Zanobetti and Schwartz (2008) multi-city study with several
additional studies as sensitivity analyses. This emphasis on newer multi-city studies is consistent
with recommendations provided by the NAS in their ozone mortality report (NRC, 2008).
CASAC supported using the Smith et al. (2009) and Zanobetti and Schwartz (2008) studies for
the ozone Health Risk and Exposure Assessment (U.S. EPA-SAB, 2012, 2014), and these are
multi-city studies published more recently (as compared with other multi-city studies or meta-
analyses included in the sensitivity analyses - see discussion below).
Smith et al. (2009) reanalyzed the NMMAPS dataset, evaluating the relationship between
short-term ozone exposure and mortality. While this study reproduces the core national-scale
estimates presented in Bell et al. (2004), it also explored the sensitivity of the mortality effect to
5-8

-------
different model specifications including (a) regional versus national Bayes-based adjustment,85
(b) co-pollutant models considering PMio, (c) all-year versus ozone-season based estimates, and
(d) consideration of a range of ozone metrics, including the daily 8-hour max. In addition, the
Smith et al. (2009) study did not use the trimmed mean approach employed in the Bell et al.
(2004) study in preparing ozone monitor data.86 In selecting effect estimates from Smith et al.
(2009), we use an ozone-only estimate for non-accidental mortality using the 8-hour max metric
for the warmer ozone season. For the sensitivity analysis, we included a co-pollutant model
(ozone and PMio) from Smith et al. (2009) for all-cause mortality, using the 8-hour max ozone
metric for the ozone season. Using a single pollutant model for the core analysis and the co-
pollutant model in the sensitivity analysis reflects our concern that the reduced sampling
frequency for days with co-pollutant measurements (1/3 and 1/6) could affect the ability of the
study to characterize the ozone effect. This choice is consistent with the ozone ISA, which
concludes that ozone effects are likely to be independent of the relationship between PM and
mortality (U.S. EPA, 2013a).
The Zanobetti and Smith (2008) study evaluated the relationship between ozone exposure
(using an 8-hour mean metric for the warm season June-August) and all-cause mortality in 48
U.S. cities using data collected between 1989 and 2000. The study presented single pollutant C-
R functions based on shorter (0-3 day) and longer (0-20 day) lag structures, with the comparison
of effects based on these different lag structures being a central focus of the study. We used the
shorter day lag based C-R function since this had the strongest effect and tighter confidence
interval. We converted the effect estimate from an 8-hour mean metric to an equivalent effect
estimate based on an 8-hour max to account for the period of the day in which most individuals
oc
In Bayesian modeling, effect estimates are "updated" from an assumed prior value using observational data. In
the Smith et al. (2009) approach, the prior values are either a regional or national mean of the individual effect
estimates obtained for each individual city. The Bayesian adjusted city-specific effect estimates are then
calculated by updating the selected prior value based on the relative precision of each city-specific estimate and
the variation observed across all city-specific individual effect estimates. City-specific estimates are pulled
towards the prior value if they have low precision and/or if there is low overall variation across estimates. City-
specific estimates are given less adjustment if they are precisely estimated and/or there is greater overall variation
across estimates.
86 There are a number of concerns regarding the trimmed mean approach including (1) the potential loss of temporal
variation in the data when the approach is used (this could impact the size of the effect estimate), and (2) a lack of
complete documentation for the approach, which prevents a full reviewing or replication of the technique.
5-9

-------
are exposed to ozone. To do this, we used the ozone metric approach wherein the original effect
estimate (and standard error) is multiplied by the appropriate ozone metric adjustment ratio.
5.2.1.2 I'M2.5 Mortality Effect Coefficients for Adults and Infants
A substantial body of published scientific literature documents the association between
elevated PM2.5 concentrations and increased premature mortality (U.S. EPA, 2009b). This body
of literature reflects thousands of epidemiology, toxicology, and clinical studies. The PM ISA
completed as part of the most recent review of the PM NAAQS, which was twice reviewed by
the SAB-CASAC (U.S. 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 (U.S. EPA, 2009b). The size of the mortality effect estimates from
epidemiological studies, the serious nature of the effect itself, and the high monetary value
ascribed to prolonging life make mortality risk reduction the most significant health endpoint
quantified in this analysis.
Researchers have found statistically significant associations between PM2.5 and
premature mortality using different types of study designs. Time-series methods have been used
to relate short-term (often day-to-day) changes in PM2.5 concentrations and changes in daily
mortality rates up to several days after a period of exposure to elevated PM2.5 concentrations.
Cohort methods have been used to examine the potential relationship between community-level
PM2.5 exposures over multiple years (i.e., long-term exposures) and community-level annual
mortality rates that have been adjusted for individual level risk factors. When choosing between
using short-term studies or cohort studies for estimating mortality benefits, cohort analyses are
thought to capture more of the public health impact of exposure to air pollution over time
because they account for the effects of long-term exposures, as well as some fraction of short-
term exposures (Kunzli et al., 2001; NRC, 2002). The NRC stated that "it is essential to use the
cohort studies in benefits analysis to capture all important effects from air pollution exposure"
(NRC, 2002, p. 108). The NRC further noted that "the overall effect estimates may be a
combination of effects from long-term exposure plus some fraction from short-term exposure.
The amount of overlap is unknown" (NRC, 2002, p. 108-9). To avoid double counting, we focus
on applying the risk coefficients from the long-term cohort studies in estimating the mortality
impacts of reductions in PM2.5.
5-10

-------
Over the last three decades, several studies using "prospective cohort" designs have been
published that are consistent with the earlier body of literature. Two prospective cohort studies,
often referred to as the Harvard "Six Cities Study" (Dockery et al., 1993; Laden et al., 2006;
Lepeule et al., 2012) and the "American Cancer Society" or "ACS study" (Pope et al., 1995;
Pope et al., 2002; Pope et al., 2004; Krewski et al., 2009), provide the most extensive analyses of
ambient PM2.5 concentrations and mortality. These studies have found consistent relationships
between fine particle indicators and premature mortality across multiple locations in the United
States. The credibility of these two studies is further enhanced by the fact that the initial
published studies (Pope et al., 1995; Dockery et al., 1993) were subject to extensive
reexamination and reanalysis by an independent team of scientific experts commissioned by the
Health Effects Institute (HEI) and by a Special Panel of the HEI Health Review Committee
(Krewski et al., 2000). Publication of studies confirming and extending the findings of the 1993
Six Cities Study and the 1995 ACS study using more recent air quality data and a longer follow-
up period for the ACS cohort provides additional validation of the findings of these original
studies (Pope et al., 2002, 2004; Laden et al., 2006; Krewski et al., 2009; Lepeule et al., 2012).
The SAB-HES also supported using these two cohorts for analyses of the benefits of PM
reductions, and concluded, "the selection of these cohort studies as the underlying basis for PM
mortality benefit estimates [is] a good choice. These are widely cited, well studied and
extensively reviewed data sets" (U.S. EPA-SAB, 2010a). As both the ACS and Six Cities studies
have inherent strengths and weaknesses, we present benefits estimates using relative risk
estimates from the most recent extended reanalysis of these cohorts (Krewski et al., 2009;
Lepeule et al., 2012). Presenting results using both ACS and Six Cities is consistent with other
recent RIAs (e.g., U.S. EPA, 2010c, 201 la, 201 lc, 2015). The PM ISA concludes that the ACS
and Six Cities cohorts provide the strongest evidence of the association between long-term PM2.5
exposure and premature mortality with support from a number of additional cohort studies
(described below).
The extended analyses of the ACS cohort data (Krewski et al., 2009) refined the earlier
ACS studies by (a) extending the follow-up period by 2 years to the year 2000, for a total of 18
years; (b) incorporating almost double the number of urban areas; (c) addressing confounding by
spatial autocorrelation by incorporating ecological, or community-level, co-variates; and (d)
performing an extensive spatial analysis using land use regression modeling in two large urban
5-11

-------
areas. These enhancements make this analysis well-suited for the assessment of mortality risk
from long-term PM2.5 exposures for the EPA's benefits analyses.
In 2009, the SAB-HES again reviewed the choice of mortality risk coefficients for
benefits analysis, concluding that "[t]he Krewski et al. (2009) findings, while informative, have
not yet undergone the same degree of peer review as have the aforementioned studies. Thus, the
SAB-HES recommends that EPA not use the Krewski et al. (2009) findings for generating the
Primary Estimate" (U.S. EPA-SAB, 2010a). Since this time, the Krewski et al. (2009) has
undergone additional peer review, which we believe strengthens the support for including this
study in this RIA. For example, the PM ISA (U.S. EPA, 2009b) included this study among the
key mortality studies. In addition, the risk assessment supporting the PM NAAQS (U.S. EPA,
2010b) used risk coefficients drawn from the Krewski et al. (2009) study, the most recent
reanalysis of the ACS cohort data. The PM risk assessment cited a number of advantages that
informed the selection of the Krewski et al. (2009) study as the source of the core effect
estimates, including the extended period of observation, the rigorous examination of model
forms and effect estimates, the coverage for ecological variables, and the large dataset with over
1.2 million individuals and 156 MS As (U.S. EPA, 2010b). The CAS AC also provided extensive
peer review of the PM risk assessment and supported the use of effect estimates from this study
(U.S. EPA-SAB, 2009a, b, 2010b).
Consistent with the PM risk assessment (U.S. EPA, 2010b), which was reviewed by the
CASAC (U.S. EPA-SAB, 2009a, b), we use the all-cause premature mortality risk estimate based
on the random-effects Cox proportional hazard model that incorporates 44 individual and 7
ecological covariates (RR=1.06, 95% confidence intervals 1.04-1.08 per 10 |ig/m3 increase in
PM2.5). The relative risk estimate (1.06 per 10 |ig/m3 increase in PM2.5) is identical to the risk
estimate drawn from the earlier Pope et al. (2002) study, though the confidence interval around
the Krewski et al. (2009) risk estimate is narrower.
In the most recent Six Cities study, which was published after the last SAB-HES review,
Lepeule et al. (2012) evaluated the sensitivity of previous Six Cities results to model
specifications, lower exposures, and averaging time using eleven additional years of cohort
follow-up that incorporated recent lower exposures. The authors found significant associations
5-12

-------
between PM2.5 exposure and increased risk of premature all-cause, cardiovascular and lung
cancer mortality. The authors also concluded that the C-R relationship was linear down to PM2.5
concentrations of 8 |ig/m3 and that premature mortality rate ratios for PM2.5 fluctuated over time,
but without clear trends, despite a substantial drop in the sulfate fraction. We use the all-cause
mortality risk estimate based on a Cox proportional hazard model that incorporates 3 individual
covariates. (RR=1.14, 95% confidence intervals 1.07-1.22 per 10 |ig/m3 increase inPlVh.s). The
relative risk estimate is slightly smaller than the risk estimate drawn from Laden et al. (2006),
with relatively smaller confidence intervals.
Given that monetized benefits associated with PM2.5 are driven largely by reductions in
premature mortality, it is important to characterize the uncertainty in this endpoint. In order to do
so, we utilize the results of an expert elicitation sponsored by the EPA and completed in 2006
(Roman et al., 2008; IEc, 2006). The results of that expert elicitation can be used as a
characterization of uncertainty in the C-R functions.
In addition to the adult premature mortality studies described above, several studies show
an association between PM exposure and premature mortality in children under 5 years of age.87
The PM ISA states that less evidence is available regarding the potential impact of PM2.5
exposure on infant mortality than on adult mortality. Furthermore, the results of studies in
children under 5 from several countries include a range of findings with some finding significant
associations. Specifically, the PM ISA concluded that evidence exists for a stronger effect at the
post-neonatal period and for respiratory-related mortality, although this trend is not consistent
across all studies. In addition, compared to avoided premature deaths estimated for adult
mortality, avoided premature deaths for infants are significantly smaller because the number of
infants in the population is much smaller than the number of adults and the epidemiology studies
on infant mortality provide smaller risk coefficients associated with exposure to PM2.5.
In 2004, the SAB-HES noted the release of the WHO Global Burden of Disease Study
focusing on ambient air, which cites several recently published time-series studies relating daily
PM exposure to mortality in children (U.S. EPA-SAB, 2004). With regard to the cohort study
conducted by Woodruff et al. (1997), the SAB-HES noted several strengths of the study,
87 For the purposes of this analysis, we only calculate benefits for infants age 0-1, not all children under 5 years old.
5-13

-------
including the use of a larger cohort drawn from a large number of metropolitan areas and efforts
to control for a variety of individual risk factors in infants (e.g., maternal educational level,
maternal ethnicity, parental marital status, and maternal smoking status). Based on these
findings, the SAB-HES recommended that the EPA incorporate infant mortality into the primary
benefits estimate and that infant mortality be evaluated using an impact function developed from
the Woodruff et al. (1997) study (U.S. EPA-SAB, 2004).
In 2010, the SAB-HES again noted the increasing body of literature relating infant
mortality and PM exposure and supported the inclusion of infant mortality in the monetized
benefits (U.S. EPA-SAB, 2010a). The SAB-HES generally supported the approach of estimating
infant mortality based on Woodruff et al. (1997) but also noted that a more recent study by
Woodruff et al. (2006) continued to find associations between PM2.5 and infant mortality in
California. The SAB-HES also noted, "when PM10 results are scaled to estimate PM2.5 impacts,
the results yield similar risk estimates." Consistent with The Benefits and Costs of the Clean Air
Act 1990 to 2020 (U.S. EPA, 201 la), we continue to rely on the earlier 1997 study in part due to
the national-scale of the earlier study.
5.2.2 Economic Valuation for Health Benefits
After quantifying the change in adverse health impacts, we estimate the economic value of
these avoided impacts. Reductions in ambient concentrations of air pollution generally lower the
risk of future adverse health effects by a small amount for a large population. Therefore, the
appropriate economic measure is willingness to pay (WTP) for changes in risk of a health effect.
For some health effects, such as hospital admissions, WTP estimates are generally not 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,
2012a).
For this final rule avoided premature deaths account for over 90 percent of monetized
ozone-related benefits and 98 percent of monetized PM-related co-benefits. The economics
5-14

-------
literature concerning the appropriate method for valuing reductions in premature mortality risk is
still evolving. The adoption of a value for the projected reduction in the risk of premature
mortality is the subject of continuing discussion within the economics and public policy analysis
communities. Following the advice of the SAB's Environmental Economics Advisory
Committee (SAB-EEAC), the EPA uses the value of statistical life (VSL) approach in
calculating estimates of mortality benefits, because we believe this calculation provides the most
reasonable estimate of an individual's willingness to trade off wealth for reductions in mortality
risk (U.S. EPA-SAB, 2000). The VSL is a summary measure for the value of small changes in
mortality risk experienced by a large number of people.88
The EPA continues work to update its guidance on valuing mortality risk reductions, and,
in the process, has engaged the SAB-EEAC on different facets of this issue. Until updated
mortality risk valuation guidance is available, however, the Agency determined that applying a
single, peer-reviewed estimate in a consistent fashion best reflects the SAB-EEAC advice it has
received. Therefore, pending future revisions to its mortality risk valuation guidance, the EPA
continues to apply the VSL that was vetted and endorsed by the SAB in the Guidelines for
Preparing Economic Analyses (U.S. EPA, 2014).89 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 $6.3 million (2000$).90 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 VSL applied in this analysis in 2011$ after adjusting for income
growth is $9.9 million for 2017.
The Agency is committed to using scientifically sound, appropriately reviewed evidence in
valuing mortality risk reductions and has made significant progress in responding to recent SAB-
EEAC recommendations. In March 2016, the EPA presented to the SAB-EEAC a proposed
88	The SAB endorsed an EPA proposal to change the moniker and the units of the mortality risk valuation measure
applied in benefits analyses (US EPA 2011; Report #EPA-SAB-11-011) but encouraged EPA to explore
alternatives more formally before deciding on which to use. EPA plans to explore alternatives through focus groups
and other risk communication exercises.
89	In the updated Guidelines for Preparing Economic Analyses (U.S. EPA, 2010e), the EPA retained the VSL
endorsed by the SAB with the understanding that further updates to the mortality risk valuation guidance would be
forthcoming.
90	In 1990$, this base VSL is $4.8 million.
5-15

-------
methodology for updating Agency mortality risk valuation estimates based on a previous SAB
Advisory (US EPA 2016). The proposed methodology is currently under review, with formal
SAB recommendations anticipated later this year.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 (OMB, 2003). We assume that there is a "cessation" lag between changes in PM
exposures and the total realization of changes in health effects. Although the structure of the lag
is uncertain, the EPA follows the advice of the SAB-HES to assume a segmented lag structure
characterized by 30 percent of mortality reductions 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, 2004c).
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.
5.2.3 Health Benefit Estimates for Ozone
We performed an HIA in BenMAP-CE and then calculated benefit per ton values to
estimate the ozone benefits for the final CSAPR Update alternative and for the more and less
stringent alternatives in this RIA. The EPA has applied this approach in several previous RIAs
(e.g., U.S. EPA, 201 lb, 201 lc, 2012b, 2014a, 2015) to quantify the avoided number of deaths
and illnesses and the total monetized human health benefits (the sum of premature mortality and
premature morbidity) of reducing one ton of summer season NOx (an ozone precursor). We
generated benefit-per-ton estimates for ozone based on air quality modeling for the illustrative
CSAPR Update alternative described in Chapter 4 of this RIA. As described in Chapter 4 of this
RIA and further below, the air quality model runs for the baseline and CSAPR Update alternative
reflect different EGU NOx emission levels for reasons other than the abatement necessary to
comply with the CSAPR Update. For this reason, it was necessary to estimate a benefit-per-ton
value from these two air quality model runs which allows us to then value the benefits solely
attributable to NOx reductions associated with the CSAPR Update. We then applied that benefit-
per-ton value to the NOx emission reductions attributable to the CSAPR Update for the CSAPR
Update alternative, as well as for the more and less stringent alternatives. The BPT estimates
correspond to NOx emissions from U.S. EGUs during the ozone-season (May to September).
These estimates assume that EGU-attributable ozone formation at the regional-level is due to
5-16

-------
NOx alone. Because EGUs emit little VOC relative to NOx emissions, it is unlikely that VOCs
emitted by EGUs would contribute substantially to regional ozone formation.
When we characterize analytical uncertainty below we describe how the benefit-per-ton
estimates have certain limitations. Specifically, the benefit-per-ton estimates reflect the
geographic distribution of the modeled illustrative CSAPR Update. For this rule, the change in
EGU NOx emissions between the baseline and CSAPR Update alternative matches well the NOx
reductions solely attributable to the CSAPR Update, but not perfectly. For this reason, the
resulting ozone benefit per ton estimate may not reflect fully the size or geographic distribution
of emission reductions anticipated from the selected policy. In order to address this potential
limitation, we limited the benefits estimate for NOx reductions associated with ozone (and
PM2.5), to only those benefits that would occur in the 22-state region of the final CSAPR Update.
The benefit per ton estimates may also not reflect well the local variability in population density,
meteorology, exposure, baseline health incidence rates, or other local factors for any specific
location. Notwithstanding these limitations, we believe that this approach is reasonably able to
characterize the ozone-related benefits from the rule.
5.2.4 Health Benefit Estimates for PM2.5
We used a combination of an HIA and a "benefit-per-ton" approach to estimate the PM2.5
co-benefits for the final CSAPR Update alternative and for more and less stringent alternatives in
this RIA. These values represent the total monetized human health co-benefits (the sum of
valued avoided premature mortality and avoided premature morbidity), of reducing one ton of
nitrate-apportioned PM2.5 from EGU-attributable NOx. We generated benefit-per-ton estimates
for nitrate PM2.5 based on the same air quality modeling simulations used to generate the benefit-
per-ton estimate for ozone. To calculate nitrate-apportioned PM2.5 benefits we then multiplied
the benefit-per-ton estimates by the change in annual NOx emissions reductions attributable to
the CSAPR Update as well as the more and less stringent CSAPR Update alternatives. These
estimates correspond to the annual NOx emission reductions from U.S. EGUs. This nitrate
PM2.5 benefit-per-ton estimate shares the limitations of the ozone NOx benefit-per-ton estimate
noted above.
5-17

-------
5.2.5 Updated Methodology in the Final RIA
We modified our analytical approach between the proposal and this final RIA. For the final
RIA, an updated air quality modeling scenario was completed, which better reflected the selected
policy option than did the proposal air quality modeling, and therefore it is appropriate to use
updated benefit-per-ton values for the final rule. However, the final air quality model results
preceded final adjustments to the policy options. Furthermore, the Pennsylvania RACT was not
included in the base case IPM model scenario, and therefore is not reflected in the air quality
baseline. This omission accounts for the larger NOx emission reductions between the air quality
model runs than is seen between the IPM base case and the CSAPR Update alternative.
Consequently, the benefit-per-ton value for ozone and nitrate-attributed PM2.5 had to be applied
to the CSAPR Update alternative NOx emission reductions in addition to the more and less
stringent alternatives NOx emission reductions.
Unlike the CSAPR Update proposal RIA which provided national estimates of the benefits
of the proposed rule, for the final CSAPR Update we calculated benefits only for the 22 CSAPR
Update states. We applied the NOx emission reductions only from the CSAPR Update states in
order to provide a benefit-per-ton value for ozone and nitrate-attributed PM2.5 that captures the
benefits to the CSAPR states. We believed this approach was made necessary by the fact that the
air quality modeling simulation accounted for NOx emission reductions occurring outside of the
22 CSAPR state region that were not reflected in the final policy scenario. This approach to
calculating the benefit per ton values likely underestimates total benefits because it excludes
certain downwind states such as those in New England and in the southeast that would likely see
benefits from this rule.
When estimating PM2.5-attributable benefits we use benefit per ton values calculated
using a nitrate-attributable PM2.5 benefit per ton estimate; the proposal analysis used a total PM2.5
benefit per ton value. We determined that the controls in this rule would have a meaningful
influence on both NOx and PM2.5 formation from nitrate. The EPA determined that, considering
the final CSAPR Update Rule illustrative emissions modeling results, using total PM2.5 benefit
5-18

-------
per ton would incorrectly additionally account for the benefits of reduced sulfate and directly
emitted PM2.5 benefits, which the illustrative emissions modeling does not anticipate.91
5.2.6 Estimated Health Benefits Results
Table 5-2 provides the benefit-per-ton estimates for the analysis year 2017. Table 5-3
provides the emission reductions estimated to occur in the analysis year. Table 5-4 summarizes
the national monetized ozone-related and PM-related health benefits estimated to occur for the
CSAPR Update and two regulatory control alternatives for the 2017 analysis year using discount
rates of 3 percent (non-fatal heart attacks quantified using Peters etal. (2001)) and 7 percent
(non-fatal heart attacks quantified using a pooled estimate that includes Pope et al. (2006)).
Table 5-5 provides national summaries of the reduced counts of premature deaths and illnesses
associated with the CSAPR update and two more and less stringent alternatives for the 2017
analysis year.92 Figure 5-1 provides a visual representation of the range of estimated ozone and
PM2.5-related benefits using benefit-per-ton estimates based on concentration-response functions
from different studies and expert opinion for the CSAPR update evaluated for 2017.
Table 5-2. Summary of Ozone and PM2.5 Benefit-per-Ton Estimates Based on Air
	Quality Modeling in 2017 (2011$)*	
Pollutant	Discount Rate	National
NOx (as Ozone)	N/A
NOx (as PM2 5)	™
$6,000 to $9,900
$1,200 to $2,800
$1,100 to $2,500
91	This approach potentially excludes any impacts of NOx on changes in sulfate particles.
92	Incidence estimates were generated using the same "per ton" approach as used to generate the dollar benefit per
ton values.
5-19

-------
* The range of estimates reflects the range of epidemiology studies for avoided premature mortality for ozone and
PM2.5. All estimates are rounded to two significant figures. Benefit-per-ton estimates for ozone are based on the
modeled ozone season NOx emissions in the 22-state region (78,000 short tons) used in the air quality runs that
were used to estimate the benefit-per-ton value. Ozone co-benefits occur in the analysis year. The monetized co-
benefits do not include reduced health effects from direct exposure to NO2, or ecosystem effects or visibility
impairment from reduced NOx. All fine particles are assumed to have equivalent health effects, but the benefit-
per-ton estimates vary depending on the location and magnitude of their impact on PM2 5 concentrations, which
drive population exposure. The PM2.5 attributed to this rule only includes the nitrate fraction of PM2.5. Benefit-
per-ton estimates for PM are based on the annual modeled PM2 5 in the 22-state region (89,000 short tons) used in
the air quality runs that were used to estimate the benefit-per-ton value. The monetized benefits incorporate the
conversion from precursor emissions to ambient fine particles and ozone., so they are the same for all discount
rates. In general, the 95th percentile confidence interval for monetized PM2.5 benefits ranges from approximately -
90 percent to +180 percent of the central estimates based on Krewski et al. (2009) and Lepeule et al. (2012). The
confidence intervals around the ozone mortality estimates are on the order of ± 60 percent depending on the
concentration-response function used.
Table 5-3. Emission Reductions of Criteria Pollutants in CSAPR Update States for the
CSAPR Update and More and Less Stringent Alternatives in 2017 (thousands of
	short tons)*	



Less Stringent

CSAPR Update
More Stringent Alternative
Alternative
Ozone Season NOx
61,000
66,000
27,000
All Year NOx
75,000
79,000
27,000
* All emissions shown in the table are rounded.
Table 5-4. Summary of Estimated Monetized Health Benefits for the CSAPR Update
and More and Less Stringent Alternatives Regulatory Control Alternatives for
	2017 (millions of 2011$) *	
Pollutant

CSAPR Update
More Stringent
Alternative
Less Stringent
Alternative
NOx (as Ozone)

$370 to $610
$400 to $650
$160 to $270
NOx (as PM2.5)
3% Discount Rate
7% Discount Rate
$93 to $210
$83 to $190
$98 to $220
$88 to $200
$34 to $75
$30 to $67
Total
3% Discount Rate
7% Discount Rate
$460 to $810
$450 to $790
$500 to $870
$490 to $850
$200 to $340
$190 to $330
* All estimates are rounded to two significant figures so numbers may not sum down columns. The health benefits
range is based on adult mortality functions (e.g., from Krewski et al. (2009) with Smith et al. (2009) to Lepeule et al.
(2012) with Zanobetti and Schwartz (2008)). The estimated monetized co-benefits do not include reduced health
effects from direct exposure to NO2, ecosystem effects, or visibility impairment. All fine particles are assumed to
have equivalent health effects, but the benefit-per-ton estimates vary depending on the location and magnitude of
their impact on PM2 5 levels, which drive population exposure. The CSAPR Update values, the more and less
stringent alternatives were all calculated using the benefits per ton approach based on the final modeling scenario.
The monetized co-benefits incorporate the conversion from precursor emissions to ambient fine particles and ozone.
Benefits for ozone are based on ozone season NOx emissions. Ozone co-benefits occur in analysis year, so they are
the same for all discount rates, and are based on annual NOx emissions and the nitrate-only fraction of PM2 5. In
general, the 95th percentile confidence interval for monetized PM2.5 benefits ranges from approximately -90 percent
to +180 percent of the central estimates based on Krewski et al. (2009) and Lepeule et al. (2012). The confidence
5-20

-------
intervals around the ozone mortality estimates are on the order of ± 60 percent depending on the concentration-
response function used.
Table 5-5. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the CSAPR Update and More and Less Stringent
Alternatives for 2017*


More Stringent
Less Stringent
Ozone-related Health Effects
CSAPR Update
Alternative
Alternative
Avoided Premature Mortality



Smith et al. (2009) (all ages)
21
23
9
Zanobetti and Schwartz (2008) (all ages)
60
65
26
Avoided Morbidity



Hospital admissions—respiratory causes (ages > 65)
59
64
26
Emergency room visits for asthma (all ages)
240
250
100
Asthma exacerbation (ages 6-18)
67,000
73,000
30,000
Minor restricted-activity days (ages 18-65)
170,000
180,000
75,000
School loss days (ages 5-17)
56,000
60,000
25,000
PM2.5-related Health Effects
Avoided Premature Mortality



Krewski et al. (2009) (adult)
10
11
3.7
Lepeule et al. (2012) (adult)
23
25
8.4
Woodruff et al. (1997) (infant)
<1
<1
<1
Avoided Morbidity



Emergency department visits for asthma (all ages)
6.1
6.5
2.2
Acute bronchitis (age 8-12)
15
15
5.2
Lower respiratory symptoms (age 7-14)
180
190
67
Upper respiratory symptoms (asthmatics age 9-11)
260
280
95
Minor restricted-activity days (age 18-65)
7,500
7,900
2,700
Lost work days (age 18-65)
1,300
1,300
450
Asthma exacerbation (age 6-18)
270
290
98
Hospital admissions—respiratory (all ages)
2.8
2.9
1.0
Hospital admissions—cardiovascular (age >18)
3.8
4.0
1.4
Non-Fatal Heart Attacks (age >18)



Peters et al. (2001)
12
13
4.3
Pooled estimate of 4 studies
1.3
1.4
0.46
* All estimates are rounded to whole numbers with two significant figures. Co-benefits for ozone are based on ozone
season NOx emissions. In general, the 95th percentile confidence interval for the health impact function alone ranges
from approximately ±30 percent for mortality incidence based on Krewski et al. (2009) and ±46 percent based on
Lepeule et al. (2012). The confidence intervals around the ozone mortality estimates are on the order of ± 60 percent
depending on the concentration-response function used.
5-21

-------
&

1800
1500
1200
g g 900
° I
| 600
300



mill
l






Sr
:Ł>
C°


IV

J"

•*'
sy
* J*"
xF
" .Ł>	«S>

d>
¦v v
„A^
a;y
500
3% Discount
7% Discount
400
300
200
100
li II III

<
V
•>S

<
-------
5.2.7 Characterization of Uncertainty in the Estimated Health Benefits
In any complex analysis using estimated parameters and inputs from numerous models,
there are likely to be many sources of uncertainty. This analysis is no exception. This analysis
includes many data sources as inputs, including emission inventories, air quality data from
models (with their associated parameters and inputs), population data, population estimates,
health effect estimates from epidemiology studies, economic data for monetizing benefits, and
assumptions regarding the future state of the world (i.e., regulations, technology, and human
behavior). Each of these inputs may be uncertain and would affect the estimated benefits. When
the uncertainties from each stage of the analysis are compounded, even small uncertainties can
have large effects on the total quantified benefits. The use of the benefit-per-ton approach adds
additional uncertainties beyond those for analyses based directly on air quality modeling.
Therefore, the estimates of benefits should be viewed as illustrating the general magnitude of
benefits of the CSAPR update and regulatory control alternatives for the 2017 analysis year,
rather than the actual benefits anticipated from implementing the rule.
This RIA shares the same detailed uncertainty assessment found in the Ozone NAAQS
RIA (U.S. EPA, 2015) or the PM NAAQS RIA (U.S. EPA, 2012a) because of the air quality
modeling input data used to run the benefits model. The results of the quantitative and qualitative
uncertainty analyses presented in the Ozone NAAQS RIA and PM NAAQS RIA provide some
information regarding the uncertainty inherent in the estimated benefits results presented in this
analysis. For example, sensitivity analyses conducted for the PM NAAQS RIA indicate that
alternate cessation lag assumptions could change the estimated PM2.5-related mortality co-
benefits discounted at 3 percent by between 10 percent and -27 percent and that alternative
income growth adjustments could change the PM2.5-related mortality benefits by between 33
percent and -14 percent. Although we generally do not calculate confidence intervals for benefit-
per-ton estimates as they can provide an incomplete picture about the overall uncertainty in the
benefits estimates, the PM NAAQS RIA provides an indication of the random sampling error in
the health impact and economic valuation functions using Monte Carlo methods. In general, the
95th percentile confidence interval for monetized PM2.5 benefits ranges from approximately -90
percent to +180 percent of the central estimates based on Krewski et al. (2009) and Lepeule el al.
(2012). The 95th percentile confidence interval for the health impact function alone ranges from
5-23

-------
approximately ±30 percent for mortality incidence based on Krewski el al. (2009) and ±46 percent
based on Lepeule et al. (2012).
After determining the health impact assessment using the air quality modeling data, we
calculated and applied benefit-per-ton estimates, which reflect specific geographic patterns of
emissions reductions and specific air quality and benefits modeling assumptions. For example,
these estimates may not reflect local variability in population density, meteorology, exposure,
baseline health incidence rates, or other local factors that might lead to an over-estimate or
under-estimate of the actual co-benefits of controlling PM and ozone precursors. As such, it is
not feasible to estimate the proportion of co-benefits occurring in different locations. Use of
these benefit-per-ton values to estimate benefits may lead to higher or lower benefit estimates
than if benefits were calculated based on direct air quality modeling. Great care should be taken
in applying these estimates to emission reductions occurring in any specific location, as these are
all based on a broad emission reduction scenario and therefore represent average benefits-per-ton
over the entire region. The benefit-per-ton for emission reductions in specific locations may be
very different than the estimates presented here. To the extent that the geographic distribution of
the emissions reductions achieved by implementing the final rule relative to the baseline used to
estimate costs and emission reductions is different than the emissions reductions in the air quality
modeling of the illustrative budgets and the baseline described in Chapter 3, the benefits may be
underestimated or overestimated.
The benefits reported here reflect the reduction in NOx emissions among the 22 CSAPR
states alone. Excluding states outside of the 22-state region may under-estimates benefits
because it does not reflect the improved air quality that could occur among states downwind of
the 22-state region. However, for reasons noted above, the air quality modeling simulation for
this analysis did not account for the size and distribution of reduced NOx emissions in this rule.
The modeling used to estimate the BPT values simulated emission changes in certain states—
including North Carolina and Georgia—that were not attributable to the CSAPR Update. These
emissions changes are not reflected in the base case modeling run that was used to estimate the
BPT values. However, these emissions changes are reflected in the illustrative CSAPR Update
alternative modeling run that was ultimately used to estimate both the BPT values and the costs
5-24

-------
and benefits of the CSAPR Update.93 To avoid incorrectly accounting for ozone-related benefits
from reduced NOx emissions from such locations, we elected to calculate benefits only within
the 22-state region. Finally, by estimating ozone health impacts from May to September only, we
may have underestimated ozone related benefits in areas experiencing a longer ozone season.
Our estimate of the total monetized benefits is based on the EPA's interpretation of the
best available scientific literature and methods and supported by the SAB-HES and the National
Academies of Science (NRC, 20022.5-related premature mortality, which accounts for 98 percent
of the monetized PM2.5 health co-benefits.
1.	We assume that all fine particles, regardless of their chemical composition, are equally
potent in causing premature mortality. This is an important assumption, because PM2.5
varies considerably in composition across sources, but the scientific evidence is not yet
sufficient to allow differentiation of effect estimates by particle type. The PM ISA
concluded that "many constituents of PM2.5 can be linked with multiple health effects, and
the evidence is not yet sufficient to allow differentiation of those constituents or sources
that are more closely related to specific outcomes" (U.S. EPA, 2009b).
2.	We assume that the health impact function for fine particles is log-linear without a
threshold. Thus, the estimates include health co-benefits from reducing fine particles in
areas with varied concentrations of PM2.5, including both areas that do not meet the fine
particle standard and those areas that are in attainment, down to the lowest modeled
concentrations.
3.	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, 2004c), which affects the valuation of mortality co-benefits at different discount
rates.
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
93 These emissions updates were made to better represent subsequent baseline emissions from those EGUs. They are
also included in the base case used to estimate the cost and emissions changes from the CSAPR Update.
5-25

-------
data in these studies. Concentration benchmark analyses (e.g., lowest measured level [LML], one
standard deviation below the mean of the air quality data in the study, etc.) 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. In this analysis, we apply two concentration benchmark approaches (LML and
one standard deviation below the mean) that have been incorporated into recent RIAs and the
EPA's Policy Assessment for Particulate Matter (U.S. EPA, 201 Id). There are uncertainties
inherent in identifying any particular point at which our confidence in reported associations
becomes appreciably less, and the scientific evidence provides no clear dividing line. However,
the EPA does not view these concentration benchmarks as a concentration threshold below
which we would not quantify health benefits of air quality improvements.94 Rather, the co-
benefits estimates reported in this RIA are the best estimates because they reflect the full range
of air quality concentrations associated with the regulatory control alternatives. The PM ISA
concluded that the scientific evidence collectively is sufficient to conclude that the relationship
between long-term PM2.5 exposures and mortality is causal and that, overall, the studies support
the use of a no-threshold log-linear model to estimate PM-related long-term mortality (U.S. EPA,
2009b).
There is also a series of key assumptions associated with our analysis of ozone-related
effects which introduce uncertainty into our estimates:
• Key assumption and uncertainties related to modeling of ozone-related premature
mortality: Ozone-related short-term mortality represents a substantial proportion of total
monetized benefits (over 94% of the ozone-related-benefits), and these estimates have the
following key assumptions and uncertainties. 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, ozone-related premature deaths
94 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 PM2.5-related Mortality (U.S. EPA, 2010b).
5-26

-------
estimated at or below this level are subject to greater uncertainty, but we cannot judge
whether (and in what direction) these impacts might be biased.
• Avoided premature mortality according to baseline pollutant concentrations: We
recognize that, in estimating short-term ozone-related mortality, we are less confident in
specifying the shape of the C-R function at lower ambient ozone concentrations (at and
below 20 ppb, ozone ISA, section 2.5.4.4). Quantitative uncertainty analyses completed for
the Ozone NAAQS RIA (U.S. EPA, 2015) found almost 100% of mortality reductions
occurred above 20 ppb, where we are more confident in specifying the nature of the ozone-
mortality effect (ozone ISA, section 2.5.4.4). However, as discussed in section 6B.7 of that
RIA, care must be taken in interpreting these results since the ambient air metric used in
modeling this endpoint is the mean 8-hour max value in each grid cell (and not the full
distribution of 8-hour daily max values). Had the latter been used, then the distribution
would have likely been wider.
For this analysis, policy-specific air quality data are not available, and the control
scenarios are illustrative of what utilities may choose to do within the trading program. However,
we believe that it is still important to characterize the distribution of exposure to baseline
concentrations. As a surrogate measure of mortality impacts, we provide the percentage of the
population exposed at each PM2.5 concentration in the baseline of the air quality modeling used
to calculate the benefit-per-ton estimates for this RIA using 12 km grid cells across the
contiguous U.S.95 It is important to note that baseline exposure is only one parameter in the
health impact function, along with baseline incidence rates population and change in air quality.
In other words, the percentage of the population exposed to air pollution below the LML is not
the same as the percentage of the population experiencing health impacts as a result of a specific
emission reduction policy. The most important aspect, which we are unable to quantify without
rule-specific air quality modeling, is the shift in exposure anticipated by implementing the
CSAPR update. Therefore, caution is warranted when interpreting the LML assessment in this
95 As described in Chapter 3, the baseline for the air quality modeling used to calculate the benefit-per-ton values
differs from the baseline used to estimate the benefits, costs, and impacts of this rulemaking. See Chapter 3 for more
details about the differences between the two baselines.
5-27

-------
RIA because these results are not consistent with results from RIAs that had air quality
modeling.
Figure 5-3 shows a bar chart of the percentage of the population exposed to various air
quality levels, including the LML concentration benchmarks in the illustrative control case
modeling, and Figure 5-4 shows a cumulative distribution function of the same data. Both figures
identify the LML for each of the major cohort studies.
MOv.
f 20«n
— L" a
§> 10°<.
LML of Kiwski et al
(2009) study
LML of Lepeule et al
(2012) study
I 2 3 4 5 5.8 6 7 8 9 10 12 14 16 18 20
Baseline Annual Mean PM; < Level (jig'm?)
Among the populations exposed to PM2.5 in the baseline:
88% are exposed to PM; s levels at or above the LML of the Krewski et al. (2009) study
47% are exposed to PM2.5 levels at or above the LML of the Lepeule et al. (2012) study
Figure 5-2. Percentage of Adult Population (age 30+) by Annual Mean PM2.5 Exposure in
the Baseline used for the Air Quality Analysis in Chapter 3
5-28

-------
100%
90%
60%
70%
60%
50%
40%
30%
20%
10%
0%





/


f


/


T



LML ofKrewski etal
(.2009) study

LML ofL^peule et al,
(2012J study









1 2 3 4 5 5.8 6 7 S 9 10 12 14 16 18
Bastlijic Annual Mcaji PMj.j Level
Figure 5-3. Cumulative Distribution of Adult Population (age 30+) by Annual Mean
PM2.5 Exposure in the Baseline used for the Air Quality Analysis in Chapter 3
5.3 Estimated Climate Co-Benefits from CO2
A co-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 then. We also provide information regarding the economic valuation of CO2 using the
Social Cost of Carbon (SC-CO2), a metric that estimates the monetary value of impacts
associated with marginal changes in CO2 emissions in a given year.
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 PIVh.s-related health benefits would be affected by excluding
5-29

-------
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
modification of PM2.5 and ozone risks (Roberts 2004; Ren 2006a, 2006b, 2008a, 2008b). Third,
the estimated climate co-benefits reported in this RIA reflect global benefits, while the estimated
health benefits are calculated for the contiguous U.S. only. 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, we
do not estimate the climate co-benefits associated with reductions in PM and ozone precursors.
5.3.1 Climate Change Impacts
Through the implementation of CAA regulations, the 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 GHG
emissions endanger public health and welfare for current and future 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
5-30

-------
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.
5.3.2 Social Cost of Carbon
We estimate the global social benefits of CO2 emission reductions expected from the final
emission guidelines using the SC-CO2 estimates presented in the Technical Support Document:
Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866 (May 2013, Revised July 2015) ("current TSD").96 We refer to these estimates,
which were developed by the U.S. government, as "SC-CO2 estimates." The SC-CO2 is a metric
that estimates the monetary value of impacts associated with marginal changes in CO2 emissions
in a given year. It 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, such as reduced costs for heating and increased costs for air
conditioning. It 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 SC-CO2 estimates used in this analysis were developed over many years, using the
best science available, and with input from the public. Specifically, an interagency working
group (IWG) that included the EPA and other executive branch agencies and offices used three
integrated assessment models (IAMs) to develop the SC-CO2 estimates and recommended four
global values for use in regulatory analyses. The SC-CO2 estimates were first released in
February 2010 and updated in 2013 using new versions of each IAM. The 2013 update did not
96 Docket ID EPA-HQ-OAR-2013-0495, 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, 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 (May 2013, Revised July 2015).
Available at:  Accessed
7/11/2015.
5-31

-------
revisit the 2010 modeling decisions with regards to the discount rate, reference case
socioeconomic and emission scenarios, and equilibrium climate sensitivity distribution. Rather,
improvements in the way damages are modeled are confined to those that have been incorporated
into the latest versions of the models by the developers themselves and published in the peer-
reviewed literature. The 2010 SC-CO2 Technical Support Document (2010 SC-CO2 TSD)
provides a complete discussion of the methods used to develop these estimates and the current
SC-CO2 TSD presents and discusses the 2013 update (including recent minor technical
corrections to the estimates).97 One key methodological aspect discussed in the SC-C02 TSDs is
the global scope of the estimates. The SC-CO2 estimates represent global measures because of
the distinctive nature of climate change, which is highly unusual in at least three respects. First,
emissions of most GHGs contribute to damages around the world independent of the country in
which they are emitted. Second, the U.S. operates in a global and highly interconnected
economy, such that impacts on the other side of the world can affect our economy. This means
that the true costs of climate change to the U.S. are much larger than the direct impacts that
simply occur within the U.S. Third, climate change represents a classic public goods problem
because each country's greenhouse gas emissions reductions benefit everyone else and no
country can be excluded from enjoying the benefits of other countries' reductions, even if it
provides no reductions itself. In this situation, the only way to achieve an economically efficient
level of emissions reductions is for countries to cooperate in providing mutually beneficial
reductions beyond the level that would be justified only by their own domestic benefits. In
reference to the public good nature of mitigation and its role in foreign relations, thirteen
prominent academics noted that these "are compelling reasons to focus on a global SCC" in a
recent article on the SCC (Pizer et al., 2014). In addition, the IWG recently noted that there is no
bright line between domestic and global damages. Adverse impacts on other countries can have
spillover effects on the United States, particularly in the areas of national security, international
trade, public health and humanitarian concerns.98
97	Both the 2010 SC-CO2 TSD and the current SC-CO2 TSD are available at:
https://www.whitehouse.gov/omb/oira/social-cost-of-carbon
98	See Response to Comments: Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866,
Interagency Working Group on Social Cost of Carbon, July 2015, page 31.
https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf
5-32

-------
The 2010 TSD noted a number of limitations to the SC-CO2 analysis, including the
incomplete way in which the integrated assessment models capture catastrophic and non-
catastrophic impacts, their incomplete treatment of adaptation and technological change,
uncertainty in the extrapolation of damages to high temperatures, and assumptions regarding risk
aversion. Currently integrated assessment models do not assign value to all of the important
physical, ecological, and economic impacts of climate change recognized in the climate change
literature due to a lack of precise information on the nature of damages and because the science
incorporated into these models understandably lags behind the most recent research." The
limited amount of research linking climate impacts to economic damages makes the modeling
exercise even more difficult. These individual limitations do not all work in the same direction in
terms of their influence on the SC-CO2 estimates, though taken together they suggest that the
SC-CO2 estimates are likely conservative. In particular, the IPCC Fourth Assessment Report
(2007), which was the most current IPCC assessment available at the time of the IWG's 2009-
2010 review, concluded that "It is very likely that [SC-CO2 estimates] underestimate the damage
costs because they cannot include many non-quantifiable impacts." Since then, the peer-
reviewed literature has continued to support this conclusion. For example, the IPCC Fifth
Assessment report (2014) observed that SC-CO2 estimates continue to omit various impacts,
such as "the effects of the loss of biodiversity among pollinators and wild crops on agriculture."
Nonetheless, these estimates and the discussion of their limitations represent the best available
information about the social benefits of CO2 reductions to inform benefit-cost analysis. The new
versions of the models used to estimate the values presented below offer some improvements in
these areas, although further work is warranted.
The EPA and other agencies have continued to consider feedback on the SC-CO2 estimates
from stakeholders through a range of channels, including public comments on rulemakings that
use the SC-CO2 in supporting analyses and through regular interactions with stakeholders and
research analysts implementing the SC-CO2 methodology used by the interagency working
99 Climate change impacts and SCC modeling is an area of active research. For example, see: (1) Howard, Peter,
"Omitted Damages: What's Missing from the Social Cost of Carbon." March 13, 2014,
http://costofcarbon.org/files/Omitted_Damages_Whats_Missing_From_the_Social_Cost_of_Carbon.pdf; and (2)
Electric Power Research Institute, "Understanding the Social Cost of carbon: A Technical Assessment," October
2014, www.epri.com.
5-33

-------
group. In addition, OMB's Office of Information and Regulatory Affairs issued a separate
request for public comment on the approach used to develop the estimates.100 After careful
evaluation of the full range of comments submitted to OMB's Office of Information and
Regulatory Affairs, the IWG continues to recommend the use of these SC-CO2 estimates in
regulatory impact analysis. With the release of the response to comments101, the IWG announced
plans to obtain expert independent advice from the National Academies of Sciences,
Engineering, and Medicine (Academies) to ensure that the SC-CO2 estimates continue to reflect
the best available scientific and economic information on climate change.102 The Academies'
process will be informed by the public comments received and focuses on the technical merits
and challenges of potential approaches to improving the SC-CO2 estimates in future updates.103
Accordingly, EPA and other agencies continue to engage in research on modeling and
valuation of climate impacts with the goal to improve these estimates. The EPA and other
federal agencies also continue to consider feedback on the SC-CO2 estimates from stakeholders
through a range of channels, including public comments on Agency rulemakings that use the SC-
CO2 in supporting analyses and through regular interactions with stakeholders and research
analysts implementing the SC-CO2 methodology used by the IWG. In addition, OMB sought
public comment on the approach used to develop the SC-CO2 estimates through a separate
comment period and published a response to those comments in 2015.
After careful evaluation of the full range of comments submitted to OMB, the IWG
continues to recommend the use of the SC-CO2 estimates in regulatory impact analysis. With the
July 2015 release of the response to comments, the IWG announced plans to obtain expert
independent advice from the National Academies of Sciences, Engineering and Medicine to
100	See https://www.federalregister.gov/articles/2013/ll/26/2013-28242/technical-support-document-technical-
update-of-the-social-cost-of-carbon-for-regulatory-impact
101	See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf
102	See https://www.whitehouse.gov/blog/2015/07/02/estimating-benefits-carbon-dioxide-emissions-reductions.
103	See
http://sites.nationalacademies.org/DBASSE/BECS/CurrentProjects/DBASSE_167526?utm_source=All%20DBASS
E%20Newsletters&utm_campaign=e84cl3e8c4-
New_Project_the_Social_Cost_of_Carbon&utm_medium=email&utm_term=0_el6023964e-e84cl3e8c4-
267347161 for more information about the National Academies process and the status of the project.
5-34

-------
ensure that the SC-CO2 estimates continue to reflect the best available scientific and economic
information on climate change. The Academies then convened a committee, "Assessing
Approaches to Updating the Social Cost of Carbon," (Committee) which is reviewing the state of
the science on estimating the SC-CO2, and will provide expert, independent advice on the merits
of different technical approaches for modeling and highlight research priorities going forward.
EPA will evaluate its approach based upon any feedback received from the Academies' panel.
To date, the Committee has released an interim report, which recommended against doing
a near term update of the SC-CO2 estimates. For future revisions, the Committee recommended
the IWG move efforts towards a broader update of the climate system module consistent with the
most recent, best available science, and also offered recommendations for how to enhance the
discussion and presentation of uncertainty in the SC-CO2 estimates. Specifically, the Committee
recommended that "the IWG provide guidance in their technical support documents about how
[SC-CO2] uncertainty should be represented and discussed in individual regulatory impact
analyses that use the [SC-CO2]" and that the technical support document for each update of the
estimates present a section discussing the uncertainty in the overall approach, in the models used,
and uncertainty that may not be included in the estimates. At the time of this writing, the IWG is
reviewing the interim report and considering the recommendations. EPA looks forward to
working with the IWG to respond to the recommendations and will continue to follow IWG
guidance on SC-CO2.
The four SC-CO2 estimates are as follows: $12, $41, $63, and $120 per metric ton of CO2
emissions in the year 2017 (2011$).104 The first three values are based on the average SC-CO2
from the three IAMs, at discount rates of 5, 3, and 2.5 percent, respectively. SC-CO2 estimates
for several discount rates are included because the literature shows that the SC-CO2 is quite
sensitive to assumptions about the discount rate, and because no consensus exists on the
appropriate rate to use in an intergenerational context (where costs and benefits are incurred by
different generations). The fourth value is the 95th percentile of the SC-CO2 from all three
104 The current version of the TSD is available at: https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-
tsd-final-july-2015.pdf. The 2010 and 2013 TSDs present SC-CO2 in 2007$ per metric ton. The unrounded
estimates from the current TSD were adjusted to 2011$ using GDP Implicit Price Deflator (1.061374),
http://www.bea.gov/iTable/index_nipa.cfm. The estimates presented here have been rounded to two significant
digits.
5-35

-------
models at a 3 percent discount rate. It is included to represent lower probability but higher
impact outcomes from climate change, which are captured further out in the tail of the SC-CO2
distribution, and while less likely than those reflected by the average SC-CO2 estimates, would
be much more harmful to society and therefore, are relevant to policy makers.
Table 5-7 presents the global SC-CO2 estimates in metric tons for the years 2015 to 2050.
In order to calculate the dollar value for emission reductions, the SC-CO2 estimate for each
emissions year would be applied to changes in CO2 emissions for that year, and then discounted
back to the analysis year using the same discount rate used to estimate the SC-CO2.105'106 The
SC-CO2 increases over time because future emissions are expected to produce larger incremental
damages as physical and economic systems become more stressed in response to greater climate
change. Note that the interagency group estimated the growth rate of the SC-CO2 directly using
the three integrated assessment models rather than assuming a constant annual growth rate. This
helps to ensure that the estimates are internally consistent with other modeling assumptions.
Table 5-8 reports the incremental climate co-benefits from CO2 emission impacts estimated for
the final CSPAR update and more and less stringent alternatives for the 2017 analysis year.
Table 5-7.
Social Cost of CO2, 2015-2050 (in 2011$ per metric ton)*



Discount Rate and Statistic

Year
5% Average
3% Average
2.5% Average
3% (95th percentile)
2015
$12
$38
$59
$110
2017
$12
$41
$63
$120
2020
$13
$45
$66
$130
2025
$15
$49
$72
$150
2030
$17
$53
$77
$160
2035
$19
$58
$83
$180
2040
$22
$64
$89
$190
2045
$24
$68
$94
$210
2050
$28
$73
$100
$230
* These SC-CO2 values are stated in $/metric ton and rounded to two significant figures. 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.
105	CO2 emission impacts for this rulemaking are shown for the year 2017 and are calculated in metric tons.
106	This analysis considered the climate impacts of only CO2 emission change. As discussed below, the climate
impacts of other pollutants were not calculated for the final CSAPR Update. While CO2 is the dominant GHG
emitted by the sector, we recognize the representative facilities within these comparisons may also have different
emission rates for other climate forcers that will serve a minor role in determining the overall social cost of
generation.
5-36

-------
Table 5-8. Estimated Global Climate Co-benefits of CO2 Reductions for the CSAPR
	Update and More and Less Stringent Alternatives for 2017 (millions of 2011$)*
Discount rate and statistic
CSPAR Update
More Stringent
Alternative
Less Stringent
Alternative
Million metric tons of CO2 reduced
1.6
2.1
1.3
5% (average)
$19
$25
$15
3% (average)
$66
$87
$54
2.5% (average)
$100
$130
$81
3% (95th percentile)
$190
$250
$150
* The SC-CO2 values are dollar-year and emissions-year specific. SC-CO2 values represent only a partial accounting
of climate impacts.
It is important to note that the climate co-benefits presented above are associated with
changes in CO2 emissions only. Implementing the CSAPR update, however, will have an impact
on the emissions of other pollutants that would affect the climate. Both predicting reductions in
emissions and estimating the climate impacts of these other pollutants, however, is complex. The
climate impacts of these other pollutants have not been calculated for the rule.107
5.4 Combined Health Benefits and Climate Co-Benefits Estimates
In this analysis, we were able to monetize the estimated benefits associated with the
reduced exposure to ozone and PM2.5 and co-benefits of decreased emissions of CO2, but we
were unable to monetize the co-benefits associated with reducing exposure to mercury, carbon
monoxide, and NO2, as well as ecosystem effects and visibility impairment. In addition, there are
expected to be unquantified health and welfare impacts associated with changes in hydrogen
chloride. Specifically, we estimated combinations of health benefits at discount rates of 3 percent
and 7 percent (as recommended by the EPA's Guidelines for Preparing Economic Analyses
[U.S. EPA, 2014] and OMB's Circular A-4 [OMB, 2003]) and climate co-benefits at estimates
of the SC-CO2 (average SC-CO2 at each of three discount rates—5 percent, 3 percent, 2.5
percent—and the 95th percentile SC-CO2 at 3 percent) (as recommended by the IWG).
107 The SC-CO2 estimates used in this analysis are designed to assess the climate benefits associated with changes in
CO2 emissions only.
5-37

-------
Different discount rates are applied to SC-CO2 than to the health benefit estimates because
CO2 emissions are long-lived and subsequent damages occur over many years. Moreover, several
rates are applied to SC-CO2 because the literature shows that it is sensitive to assumptions about
discount rate and because no consensus exists on the appropriate rate to use in an
intergenerational context. The SC-CO2 interagency group centered its attention on the 3 percent
discount rate but emphasized the importance of considering all four SC-CO2 estimates.108 The
EPA has evaluated the range of potential impacts by combining all SC-CO2 values with health
benefits values at the 3 percent and 7 percent discount rates. Combining the 3 percent SC-CO2
values with the 3 percent health benefit values assumes that there is no difference in discount
rates
Table 5-9 provides the combined health and climate benefits for the CSAPR update and
more and less stringent alternatives for the 2017 analysis year.
108 See the 2010 SCC TSD. Docket ID EPA-HQ-OAR-2009-0472-114577 or
http://www.whitehouse.gov/sites/default/files/omb/inforeg/for-agencies/Social-Cost-of-Carbon-for-RIA.pdf for
details.
5-38

-------
Table 5-9. Combined Health Benefits and Climate Co-Benefits for the CSAPR update
and More and Less Stringent Alternatives for 2017 (millions of 2011$)*

Health and Climate Benefits
Climate Co-
SC-CO2 Discount Rate**
(Discount Rate Applied to Health Co-Benefits)
Benefits Only

3%
7%

CSAPR Update
5%
$480 to $830
$470 to $810
$19
3%
$530 to $880
$520 to $860
$66
2.5%
$560 to $910
$550 to $890
$100
3% (95th percentile)
$650 to $1,000
$640 to $980
$190
More Stringent Alternative
5%
$520 to $900
$510 to $870
$25
3%
$580 to $960
$570 to $940
$87
2.5%
$630 to $1,000
$620 to $980
$130
3% (95th percentile)
$750 to $1,100
$740 to $1,100
$250
Less Stringent Alternative
5%
$210 to $360
$210 to $350
$15
3%
$250 to $400
$250 to $390
$54
2.5%
$280 to $420
$270 to $420
$81
3% (95th percentile)
$350 to $500
$350 to $490
$150
*A11 estimates are rounded to two significant figures. Climate benefits are based on reductions in CO2 emissions.
Health benefits are based on benefit-per-ton estimates. Benefits for ozone are based on ozone season NOx
emissions. Ozone benefits occur in analysis year, so they are the same for all discount rates. The health benefits
reflect the sum of the ozone benefits and PM2 5 co-benefits and reflect the range based on adult mortality functions
(e.g., from Krewski el al. (2009) with Smith el al. (2009) to Lepeule el al. (2012) with Zanobetti and Schwartz
(2008)). The monetized health benefits do not include reduced health effects from direct exposure to NO2 as well as
ecosystem effects and visibility impairment associated with reductions in NOx.
**As discussed in section 5.3, the SC-CO2 estimates are calculated with four different values of a one metric ton
reduction.
5.5 Unquantified Benefits and Co-benefits
The monetized co-benefits estimated in this RIA reflect a subset of benefits and co-benefits
attributable to the health effect reductions associated with ambient ozone and fine particles. Data,
time, and resource limitations prevented the EPA from quantifying the impacts to, or monetizing
the co-benefits from several important benefit categories as well as ecosystem effects and
visibility impairment associated with reductions in NOx due to the absence of air quality
modeling data for these pollutants in this analysis. This does not imply that there are no co-
benefits associated reductions in exposures to NO2. In this section, we provide a qualitative
description of these benefits, which are listed in Table 5-10.
5-39

-------
Table 5-10. Unquantified Health and Welfare Benefit and Co-benefit Categories


Effect Has
Effect Has

Category
Specific Effect
Been
Quantified
Been
Monetized
More Information
Improved Human Health

Asthma hospital admissions (all ages)
—
—
NO2 ISA1

Chronic lung disease hospital admissions (age >
65)
—
—
NO2 ISA1
Reduced incidence of
Respiratory emergency department visits (all
ages)
—
—
NO2 ISA1
morbidity from exposure
Asthma exacerbation (asthmatics age 4-18)
—
—
NO2 ISA1
to NO2
Acute respiratory symptoms (age 7-14)
—
—
NO2 ISA1

Premature mortality
—
—
NO2 ISA1-2-3

Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
—
—
NO2 ISA2-3
Improved Environment
Reduced visibility
Visibility in Class 1 areas
—
—
PM ISA1
impairment
Visibility in residential areas
—
—
PM ISA1
Reduced effects on
materials
Household soiling
—
—
PM ISA1-2
Materials damage (e.g., corrosion, increased
wear)
—
—
PM ISA2
Reduced effects from PM




deposition (metals and
organic s)
Effects on individual organisms and ecosystems
—
—
PM ISA2

Visible foliar injury on vegetation
—
—
Ozone ISA1

Reduced vegetation growth and reproduction
—
—
Ozone ISA1

Yield and quality of commercial forest products
and crops
—
—
Ozone ISA1
Reduced vegetation and
ecosystem effects from
Damage to urban ornamental plants
—
—
Ozone ISA2
Carbon sequestration in terrestrial ecosystems
—
—
Ozone ISA1
exposure to ozone
Recreational demand associated with forest
aesthetics
—
—
Ozone ISA2

Other non-use effects


Ozone ISA2

Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity,
leaf-gas exchange, community composition)
—
—
Ozone ISA2

Recreational fishing
—
—
NOxSOxISA1

Tree mortality and decline
—
—
NOxSOxISA2
Reduced effects from
acid deposition
Commercial fishing and forestry effects
—
—
NOxSOxISA2
Recreational demand in terrestrial and aquatic
ecosystems
—
—
NOxSOxISA2

Other non-use effects


NOxSOxISA2

Ecosystem functions (e.g., biogeochemical
cycles)
—
—
NOxSOxISA2

Species composition and biodiversity in terrestrial
and estuarine ecosystems
—
—
NOxSOxISA2

Coastal eutrophication
—
—
NOxSOxISA2
Reduced effects from
nutrient enrichment
Recreational demand in terrestrial and estuarine
ecosystems
—
—
NOxSOxISA2

Other non-use effects


NOxSOxISA2

Ecosystem functions (e.g., biogeochemical
cycles, fire regulation)
—
—
NOxSOxISA2
5-40

-------


Effect Has
Effect Has

Category
Specific Effect
Been
Quantified
Been
Monetized
More Information
Reduced vegetation
effects from ambient
Injury to vegetation from NOx exposure
_
_
NOxSOxISA2
exposure to NOx




1 We assess these co-benefits qualitatively due to data and resource limitations for this RIA.
2We assess these co-benefits qualitatively because we do not have sufficient confidence in available data or methods.
3 We assess these co-benefits qualitatively because current evidence is only suggestive of causality or there are other significant
concerns over the strength of the association.
5.5.2 Additional NO2 Health Co-Benefits
NO and NO2 are often grouped together into their own group or family, which the
atmospheric sciences community refers to as NOx (U.S. EPA, 2016). In addition to being a
precursor to PM2.5 and ozone, NOx/NCh emissions—which emanate from a variety of sources
including EGU's—are also linked to a variety of adverse health effects associated with direct
exposure. We were unable to estimate the health co-benefits associated with reduced NO2
exposure in this analysis for two reasons. First, we lacked a reliable reduced-form approach for
quantifying NCh-attributable benefits. A second, and related reason, is that it is generally
necessary to perform air quality modeling that characterizes well the near-field gradient
associated with NO2 concentrations—particularly from the mobile sector (U.S. EPA, 2016); such
an analysis was not performed for this rule. Therefore, this analysis only quantified and
monetized the ozone benefits and PM2.5 co-benefits associated with the reductions in NO2
emissions.
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, 2016) concluded that there is a causal relationship between respiratory health effects
and short-term exposure to NO2. These epidemiologic and experimental studies encompass a
number 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.
5-41

-------
5.5.4 Additional NO2 Welfare Co-Benefits
As described in the Integrated Science Assessment for Oxides of Nitrogen and Sulfur —
Ecological Criteria (NOx/SOx ISA) (U.S. EPA, 2008d), 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,
increase the sensitivity of these plants to stresses, such as droughts, cold temperatures, insect
pests, and disease leading to increased mortality of canopy trees.
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,
2008d).
Reductions in emissions of NO2 will improve the level of visibility throughout the United
States because these gases (and the particles of nitrate formed from this gas as discussed below)
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
5-42

-------
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).
5.5.5	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, 2013b). 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.
5.5.6	PM2.5 Visibility Impairment Co-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, 2009b). 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, 2009b). Previous analyses (U.S.
EPA, 201 la) show that visibility co-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 CSAPR
Update would be likely to have a significant impact on visibility in urban areas or Class I areas.
5.6 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.
5-43

-------
Abt Associates, Inc. 2012. "BenMAP User's Manual Appendices," prepared for U.S. Research
Triangle Park, NC: U. S. Environmental Protection Agency, Office of Air Quality Planning
and Standards. Available at:
. Accessed June
6, 2015.
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.
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.
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.
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.
5-44

-------
Huang Y., F. Dominici, and M. Bell. 2005. "Bayesian Hierarchical Distributed Lag Models for
Summer Ozone Exposure and Cardio-Respiratory Mortality." Environmetrics. 16:547-562.
Industrial Economics, Incorporated (IEc). 2006. Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2.5 Exposure and Mortality. Prepared for:
Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC. September. Available at:
. Accessed June 6, 2015.
Industrial Economics, Incorporated (IEc). 2009. Section 812 Prospective Study of the Benefits
and Costs of the Clean Air Act: Air Toxics Case Study: Health Benefits of Benzene
Reductions in Houston, 1990-2020. Final Report, July 14, 2009. Available at:
. Accessed June 6, 2015.
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.
5-45

-------
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.
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.
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.
Karl, T., J. Melillo, and T. Peterson, Eds. 2009. Global Climate Change Impacts in the United
States. Cambridge, United Kingdom: Cambridge University Press.
Krewski D., M. Jerrett, R.T. Burnett, R. Ma, E. Hughes, Y. Shi, etal. 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.
Melillo, Jerry M., Terese (T.C.) Richmond, and Gary W. Yohe, Eds. 2014. Climate Change
Impacts in the United States: The Third National Climate Assessment. U.S. Global Change
Research Program. Available at 
National Research Council (NRC). 2000. ToxicologicalEffects 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.
National Research Council (NRC). 2010. Ocean Acidification: A National Strategy to Meet the
Challenges of a Changing Ocean. National Academies Press. Washington, DC.
National Research Council (NRC). 2011a. Climate Stabilization Targets: Emissions,
Concentrations, and Impacts over Decades to Millennia. National Academies Press,
Washington, DC.
National Research Council (NRC) 201 lb. National Security Implications of Climate Change for
U.S. Naval Forces. National Academies Press. Washington, DC.
5-46

-------
National Research Council (NRC, 201 lc). Understanding Earth's Deep Past: Lessons for Our
Climate Future. National Academies Press. Washington, DC
National Research Council (NRC). 2012a. Sea-Level Rise for the Coasts of California, Oregon,
and Washington: Past, Present, and Future. National Academies Press. Washington, DC.
National Research Council (NRC). 2013 a. Climate and Social Stress: Implications for Security
Analysis. National Academies Press. Washington, DC.
National Research Council (NRC). 2013b. Abrupt Impacts of Climate Change: Anticipating
Surprises. 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>.
Pizer, W., M. Adler, J. Aldy, D. Anthoff, M. Cropper, K. Gillingham, M. Greenstone, B. Murray,
R. Newell, R. Richels, A. Rowell, S. Waldhoff, J. Wiener. 2014. "Using and improving the
social cost of carbon." Science, Vol. 346, No. 6214, 12/05/14, pp 1189-1190.
Ren, C., G.M. William, L. Morawska, K. Mengensen, and S. Tong. 2008a. "Ozone Modifies
Associations between Temperature and Cardiovascular Mortality: Analysis of the NMMAPS
Data." Occupational and Environmental Medicine. 65:255-260.
Ren, C., G.M. Williams, K. Mengersen, L. Morawska, and S. Tong. 2008b. "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. and S. Tong. 2006b. "Temperature Modifies the Health Effects of Particulate Matter in
Brisbane, Australia " International Journal of Biometeorology. 51:87-96.
Ren. C., G.M. Williams, and S. Tong. 2006a. "Does Particulate Matter Modify the Association
between Temperature and Cardiorespiratory Diseases? Environmental Health Perspectives,
114:1690-1696.
Roberts, S. 2004. "Interactions between Particulate Air Pollution and Temperature in Air
Pollution Mortality Time Series Studies" Environmental Research. 96:328-337.
Roman, H., K. D. Walker, T.L. Walsh, L. Conner, H. M. Richmond, B.J. Hubbell, and P.L.
Kinney. 2008. "Expert Judgment Assessment of the Mortality Impact of Changes in Ambient
Fine Particulate Matter in the U.S." Environmental Science and Technology. 42(7):2268-
2274.
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.
5-47

-------
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.
U.S. Environmental Protection Agency (U.S. EPA). 1995. Regulatory Impact Analysis for the
Petroleum Refinery NESHAP. Revised Draft for Promulgation. Office of Air Quality
Planning and Standards, Research Triangle Park, N.C. Available on the Internet at <
http://yosemitel.epa.gov/ee/epa/ria.nsf/vwTD/9F39F2C26150BB21852564620063317F>.
Accessed June 6, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008a. Integrated Science Assessment for
Sulfur Oxides—Health Criteria (FinalReport). National Center for Environmental
Assessment - RTP Division, Research Triangle Park, NC. September. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008b. Final Ozone NAAQS Regulatory
Impact Analysis. EPA-452/R-08-003. Office of Air Quality Planning and Standards Health and
Environmental Impacts Division, Air Benefit and Cost Group Research Triangle Park, NC.
March. Available at: < http://www.epa.gov/ttnecasl/regdata/RIAs/6-ozoneriachapter6.pdf>.
Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008c. Integrated Science Assessment for
Oxides of Nitrogen - Health Criteria (Final Report). National Center for Environmental
Assessment, Research Triangle Park, NC. July. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008d. 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:
. Accessed June 4, 2015.
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:
. Accessed June 4, 2015.
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:
. Accessed June 4, 2015.
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:
. Accessed June 4, 2015.
5-48

-------
U.S. Environmental Protection Agency (U.S. EPA). 2010c. Valuing Mortality Risk Reductions
for Environmental Policy: A White Paper: SAB Review Draft. National Center for
Environmental Economics December. Available at:

Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2010d. Section 3: Re- analysis of the
Benefits of Attaining Alternative Ozone Standards to Incorporate Current Methods.
Available at: . Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2010e. Guidelines for Preparing Economic
Analyses.EPA 240-R-10-001. National Center for Environmental Economics, Office of
Policy Economics and Innovation.Washington, DC. December. Available at

U.S. Environmental Protection Agency (U.S. EPA). 201 la. The Benefits and Costs of the Clean
Air Act from 1990 to 2020. Office of Air and Radiation, Washington, DC. March. Available
at: . Accessed June 4,
2015.
U.S. Environmental Protection Agency (U.S. EPA). 2011b. Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards. EPA-452/R-11-011. December. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2011c. Regulatory Impact Analysis:
National Emission Standards for Hazardous Air Pollutants for Industrial, Commercial, and
Institutional Boilers and Process Heaters. February. Available at:
. Accessed June
4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 201 Id. 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: .
Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2012a. Regulatory Impact Analysis for the
Final Revisions to the National Ambient Air Quality Standards for Particulate Matter. EP A-
452/R-12-003. Office of Air Quality Planning and Standards, Health and Environmental
Impacts Division, Research Triangle Park, NC. December. Available at: <
http://www.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf>. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2012b. Regulatory Impact Analysis:
Petroleum Refineries New Source Performance Standards Ja. Office of Air Quality Planning
and Standards, Health and Environmental Impacts Division. June. Available at:
5-49

-------
. Accessed
June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2013a. Technical Support Document:
Estimating the Benefit per Ton of Reducing PM2.5 Precursors from 17 Sectors. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. February. Available at: <
http://www2.epa.gov/sites/production/files/2014-
10/documents/sourceapportionmentbpttsd.pdf >. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2013b. 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:
. Accessed June 4,
2015.
U.S. Environmental Protection Agency (U.S. EPA). 2014. Guidelines for Preparing Economic
Analyses. EPA 240-R-10-001. National Center for Environmental Economics, Office of the
Administrator. Washington, DC. Available at:
. Accessed June 6,
2015.
U.S. Environmental Protection Agency (U.S. EPA). 2014a. Regulatory Impact Analysis for the
Proposed Carbon Pollution Guidelines for Existing Power Plants and Emission Standards
for Modified and Reconstructed Power Plants. EPA-542/R-14-002. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. June. Available at
. Accessed June
4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2014b. Health Risk and Exposure
Assessment for Ozone: Final Report. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. EPA-452/R-14-004a. Available at:
. Accessed
June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2014c. Welfare Risk and Exposure
Assessment for Ozone: Final. EPA-452/R-14-005a. Office of Air Quality Planning and
Standards, Research Triangle Park, NC. August. Available at:
. Accessed
June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2014d. Regulatory Impact Analysis of the
Proposed Revisions to the National Ambient Air Quality Standards for Ground-Level Ozone.
EPA-452/P-14-006. Office of Air Quality Planning and Standards, Research Triangle Park,
NC. November. Available at .
Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2015. Regulatory Impact Analysis of the
Clean Power Plan. Office of Air Quality Planning and Standards, Research Triangle Park,
5-50

-------
NC. November. Available at< http://www2.epa.gov/cleanpowerplan/clean-power-plan-final-
rule-regulatory-impact-analysis>. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008c. 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>. Accessed July 11, 2016.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2002.
Workshop on the Benefits of Reductions in Exposure to Hazardous Air Pollutants:
Developing Best Estimates of Dose-Response Functions An SAB Workshop Report of an
EPA/SAB Workshop (Final Report). EPA-SAB-EC-WKSHP-02-001. January. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2008.
Characterizing Uncertainty in Particulate Matter Benefits Using Expert Elicitation. EPA-
COUNCIL-08-002. July. Available at:
. Accessed June 4, 2015.
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:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2004c.
Advisory Council on Clean Air Compliance Analysis Response to Agency Request on
Cessation Lag. EPA-COUNCIL-LTR-05-001. December. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2008.
Benefits of Reducing Benzene Emissions in Houston, 1990-2020. EPA-COUNCIL-08-001.
July. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009b.
Review of EPA 's Integrated Science Assessment for Particulate Matter (First External
Review Draft, December 2008). EPA-COUNCIL-09-008. May. Available at:
.
Accessed June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2009c.
Review of Integrated Science Assessment for Particulate Matter (Second External Review
5-51

-------
Draft, July 2009). EPA-CASAC-10-001. November. Available at:
. Accessed
June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2010a.
Review of EPA's DRAFT Health Benefits of the Second Section 812 Prospective Study of the
Clean Air Act. EPA-COUNCIL-10-001. June. Available at: <
http://yosemite.epa.gov/sab/sabproduct.nsf/9288428b8eeea4c885257242006935a3/72D4EFA
39E48CDB28525774500738776/$File/EPA-COUNCIL-10-001 -unsigned.pd^. Accessed
June 4, 2015.
U.S. Environmental Protection Agency—Science Advisory Board (U.S. EPA-SAB). 2011.
Review of Valuing Mortality Risk Reductions for Environmental Policy: A White Paper
(December 10, 2010). EPA-SAB-11-011 July. Available at:
. Accessed June 4, 2015.
Woodruff, T.J., J. Grillo, andK.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.
5-52

-------
CHAPTER 6: ECONOMIC IMPACTS
Overview
This chapter addresses economic impacts on small entities, other government entities, and
employment.
6.1 Impacts on Small Entities
The 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. The EPA has determined that 1 entity (of 11 small entities
identified as potentially affected) may experience an impact of greater than 3 percent of annual
revenues. 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 final rulemaking, it must prepare and make
available an final regulatory flexibility analysis, unless it certifies that the final rule 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.
The 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.5.15, used in this RIA.109
This analysis draws on the "parsed" unit-level estimates using IPM results for 2018,110 as well as
ownership, employment, and financial information for the potentially affected small entities
drawn from other resources described in more detail below.
109 Detailed documentation for IPM v.5.15 is available at:
http://www.epa.gov/airmarkets/powersectormodeling.html.
n° for this analysis, the 2018 parsed file is used to represent 2017 for the purposes of RIA analysis.
6-1

-------
The EPA identified the size of ultimate parent entities by using the Small Business
Administration (SBA) size threshold guidelines.111 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
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.112
•	Publicly-owned entities
o Publicly-owned entities include federal, state, municipal, and other political
subdivision entities.
o The federal and state governments were considered to be large. Municipalities
and other political units with population fewer than 50,000 were 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.
111	U.S. Small Business Administration (SBA). 2014. Small Business Size Standards. Effective as of July 14, 2014.
See: http://www.sba.gov/sites/default/files/Size_Standards_Table.pdf.
112	Certain affected EGUs are owned by ultimate parent entities whose primary business is not electric power
generation.
6-2

-------
6.1.1 Identification of Small Entities
In this analysis, the EPA considered EGUs which meet the following five criteria: 1)
EGU is represented in NEEDS v5.15; 2) EGU is fossil fuel-fired; 3) EGU is located in a state
covered by this rule; 4) EGU is neither a cogeneration unit nor solid waste incineration unit; 5)
EGU capacity is 25MW or larger. The EPA next refined this list of EGUs, narrowing it to those
that exhibit at least one of the following changes under the CSAPR Update scenario, in
comparison to the baseline.
•	Summer fuel use (BTUs) changes by +/- 1% or more
•	Summer generation (GWh) changes by +/- 1% or more
•	NOx summer emissions (tons) changes by +/- 1% or more
Based on these criteria, the EPA identified a total of 365 potentially affected EGUs
warranting examination in this RFA analysis. Next, we determined power plant ownership
information, including the name of associated owning entities, ownership shares, and each
entity's type of ownership. We primarily used data from SNL and Ventyx, supplemented by
limited research using publicly available data.113 Majority owners of power plants with affected
EGUs were categorized as one of the seven ownership types.114 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.
113	SNL Financial data covers the energy market and other industries, and includes detailed immediate and ultimate
ownership at the EGU level. For more information, see: www.snl.com. 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.
114	Throughout this analysis, the EPA refers to the owner with the largest ownership share as the "majority owner"
even when the ownership share is less than 51 percent.
6-3

-------
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, the EPA used both the Hoover's online database and the SNL database to identify
the ultimate owners of power plant owners identified in the SNL and Ventyx databases. This was
necessary, as many majority owners of power plants (listed in SNL or Ventyx) are themselves
owned by other ultimate parent entities (listed in Hoover's or SNL).115 In these cases, the
ultimate parent entity was identified via Hoover's or SNL, whether domestically or
internationally owned.
The 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 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 the EPA's IPM base case.
115 The Hoover's Inc. online platform includes company records that can contain NAICS codes, number of
employees, revenues, and assets. For more information, see: http://www.hoovers.com
6-4

-------
Table 6-1. SBA Size Standards by NAICS Code
NAICS
Code
NAICS Description
SBA Size Standard
221112
Fossil Fuel Electric Power Generation
750 employees
221118
Other Electric Power Generation
250 employees
221122
Electric Power Distribution
1,000 employees
221210
Natural Gas Distribution
1,000 employees
238210
Electrical Contractors and Other Wiring Installation Contractors
$15 million in revenue
324110
Petroleum Refineries
1,500 employees
325180
Other Basic Inorganic Chemical Manufacturing
1,000 employees
325320
Pesticide and Other Agricultural Chemical Manufacturing
1,000 employees
331313
Alumina Refining and Primary Aluminum Production
1,000 employees
333613
Mechanical Power Transmission Equipment Manufacturing
750 employees
424720
Petroleum and Petroleum Products Merchant Wholesalers (except
Bulk Stations and Terminals)
200 employees
486210
Pipeline Transportation of Natural Gas
$27.5 million in revenue
522110
Commercial Banking
$550 million in assets
522220
Sales Financing
$38.5 million in revenue
523120
Securities Brokerage
$38.5 million in revenue
523910
Miscellaneous Intermediation
$38.5 million in revenue
523930
Investment Advice
$38.5 million in revenue
524126
Direct Property and Casualty Insurance Carriers
1,500 employees
525120
Health and Welfare Funds
$32.5 million in revenue
525990
Other Financial Vehicles
$32.5 million in revenue
541611
Administrative Management and General Management Consulting
Services
$ 15 million in revenue
551112
Offices of Other Holding Companies
$20.5 million in revenue
Note: Based on size standards effective at the time the EPA conducted this analysis (SBA size standards, effective
February 26, 2016)
Source: SBA, 2016
The EPA compared the relevant entity size criterion for each ultimate parent entity to the
SBA threshold value 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: The EPA used the Hoover's database as the
primary source for information on ultimate parent entity employee numbers, revenue, and
assets.116 In parallel, the EPA also considered estimated revenues from affected EGUs
based on analysis of parsed-ftle estimates for the final rule. The EPA assumed that the
ultimate parent entity revenue was the larger of the two revenue estimates. In limited
116 Estimates of sales were used in lieu of revenue estimates when revenue data was unavailable.
6-5

-------
instances, supplemental research was also conducted to estimate an ultimate parent
entity's number of employees, revenue, or assets.
2. Population: Municipal entities are defined as small if they serve populations of less than
50,000. The EPA primarily relied on data from the Ventyx database and the U.S. Census
Bureau to inform this determination. Supplemental research of individual municipalities
was also conducted in some instances.
Ultimate parent entities for which the relevant measure is less than the SBA size criterion
were identified as small entities and carried forward in this analysis. In the case of one entity,
data limitations prevented the comparison of the entity against its appropriate SBA size standard.
For the purposes of this analysis, the EPA assumed that this entity is a small entity. Overall, the
EPA identified 30 potentially affected EGUs owned by 11 small entities included in the EPA's
Base Case.
6.1.2 Overview of Analysis and Results
This section presents the methodology and results for estimating the impact of the
CSAPR Update to small entities in 2017 based on the following endpoints:
•	annual economic impacts of the CSAPR Update on small entities, and
•	ratio of small entity impacts to revenues from electricity generation.
6.1.2.1	Methodology for Estimating Impacts of the CSAPR Update on Small Entities
An entity can comply with the CSAPR Update through some combination of the
following: optimizing existing SCR, turning on idled SCR or SNCR controls, upgrading to state
of the art combustion controls, using allocated allowances, purchasing allowances, or reducing
emissions through a reduction in generation or improved efficiency. 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, the EPA estimates compliance
costs as follows:
6-6

-------
Ccompliance A Coperating+Retrofit A CFuel A CAllowances A CTransaction A R
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
CSAPR Update in 2017. This analysis is based on the NOx budgets and modeling results
presented in Chapter 4.
In reality, 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 CSAPR
Update, 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 CSAPR Update. 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 rule on
small entities.
For this analysis, the 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 CSAPR Update relative to the base case. These individual
components of compliance cost were estimated as follows:
(1) Operating and retrofit costs: Using the IPM-parsed output for the base case and
the CSAPR Update, the EPA identified units that install control technology under
the policy, and what technology was installed. The equations for calculating
retrofit costs were adopted from the EPA's version of IPM. The model calculates
the capital cost (in $/MW); the fixed operation and maintenance (O&M) cost (in
$/MW-year); the variable O&M cost (in $/MWh); and the total annual cost for
6-7

-------
units projected to turn on existing idled SCR, fully operate existing SCR, or turn
on existing idled SNCR.
(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 $1,400 per ton. $1,400 per ton is the marginal cost of NOx
reductions used to set the budgets in the final rule. While this is a reasonable
approximation, it is possible that the actual allowance price could be lower. Units
were assumed to purchase or sell allowances to exactly cover their projected
emissions under the policy.
(3)	Fuel costs: The change in fuel expenditures under the policy was estimated by
taking the difference in projected fuel expenditures between the IPM estimates for
the CSAPR Update and the base case.
(4)	Value of electricity generated: To estimate the value of electricity generated, the
projected level of electricity generation is multiplied by the regional-adjusted
retail electricity price ($/MWh) estimate, for all entities except those categorized
as Private in Ventyx. For private entities, the EPA used segmental wholesale
electricity price instead 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 and 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, the EPA considered the primary
administrative cost to be transaction costs related to purchasing or selling
allowances. The 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
International.
6.1.2.2	Results
The potential impacts of the CSAPR Update on small entities are summarized in Table
6-2. All costs are presented in $2011. The EPA estimated the annual net compliance cost to small
entities to be approximately $23.9 million in 2017. At a plant level, the net compliance costs for
6-8

-------
all entities includes net savings at a number of plants in this analysis. These net savings are
driven by entities that are able to increase their revenues by increasing generation.
Table 6-2. Projected Impact of the CSAPR Update on Small Entities in 2017	
EGU
Ownership
Type
Number of
Potentially
Affected
Entities
Total Net
Compliance
Cost ($2011
millions)
Number of Small
Entities with
Compliance Costs
>1% of Generation
Revenues
Number of Small
Entities with
Compliance Costs
>3% of Generation
Revenues
Cooperative
3
24.1
1
1
Municipal
3
0.0
0
0
Private
5
-0.2
0
0
Total
11
23.9
1
1
Note: The total number of entities with costs greater than 1 percent or 3 percent of revenues includes only entities
experiencing positive costs. A negative cost value implies that the group of entities experiences a net
savings under the CSAPR Update.
Source: IPM analysis
The EPA assessed the economic and financial impacts of the 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 the 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. 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, the 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 11 small entities considered in this analysis, 1 entity may experience compliance
costs greater than 1 percent of generation revenues in 2017, and only 2 entities may experience
net positive compliance costs. The other 9 entities may experience negative net costs under the
CSAPR Update. The EPA has concluded that there is no significant economic impact on a
substantial number of small entities (No SISNOSE) for this rule. The number of entities with
compliance costs exceeding 3 percent of generation revenues is also included in Table 6-2.
6-9

-------
The distribution across entities of economic impacts as a share of base case revenue is
summarized in Table 6-3. Since there are few potentially-impacted small entities included in this
analysis, the distributions of economic impacts on each ownership type are in general fairly tight.
Table 6-3. Summary of Distribution of Economic Impacts of the CSAPR Update on Small
Entities in 2017
EGU Ownership
Type
Capacity-Weighted
Average Economic
Impacts as a % of
Generation Revenues
Min
Max
Cooperative
9.0%
-7.8%
9.0%
Municipal
-0.8%
-11.9%
0.2%
Private
-1.9%
-11.7%
-0.1%
Source: IPM analysis
The separate components of annual costs to small entities under the CSAPR Update are
summarized in Table 6-4. The most significant components of incremental cost to these entities
under the CSAPR Update are due to lower electricity revenues. The vast majority of the
decreased electricity revenue component is attributable to the single entity that may experience
compliance costs greater than 1 percent of generation revenues in 2017. Since this one entity
represents a large share of generation in this category, the projected reduction in generation at
this single entity relative to the base case leads to higher net costs for the entire category. The
fuel costs decreases are largely attributable to a few entities that are projected to decrease
generation relative to the base case, which translates to lower fuel costs for the whole group.
However, many of these entities are projected to increase generation relative to the base case and
thus counterbalance this overall impact. Additionally, increases in electricity generation
revenue, shown as cost savings or negative costs are experienced by cooperative, municipal, and
private entities. This is due largely to the projected increase in generation at these entities under
the CSAPR Update.
6-10

-------
Table 6-4. Incremental Annual Costs under the CSAPR Update Summarized by
Ownership Group and Cost Category in 2017 (2011$ millions)
EGU
Ownership
Type
Operating
Cost
Net Purchase
of Allowances
Fuel Cost
Lost
Electricity
Revenue
Administrative
Cost
Cooperative
-$1.8
$1.5
-$6.6
$31.0
$0.02
Municipal
$0.5
$0.3
$0.3
-$1.1
$0.00
Private
$0.4
-$0.6
-$0.1
$0.0
$0.01
Source: IPM analysis
6.1.3 Summary of Small Entity Impacts
The EPA examined the potential economic impacts to small entities associated with this
rulemaking based on assumptions of how the affected states will implement control measures to
meet their emissions. To summarize, of the 11 small entities potentially affected, 1 may
experience compliance costs in excess of 1 percent of revenues in 2017, based on assumptions of
how the affected states implement control measures to meet their emissions budgets as set forth
in this rulemaking. Potentially affected small entities experiencing compliance costs in excess of
1 percent of revenues have some potential for significant impact resulting from implementation
of the CSAPR Update.
The 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 22 states for which the EPA is finalzing FIPs. Additionally, the CSAPR Update
allows for the flexibility of trading, which greatly reduces compliance burden. For further
information, see the evaluation completed for the original CSAPR, available at 76 FR 48272-
48273 (August 8, 2011).
6.2 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
6-11

-------
governments and the private sector. Under Section 202 of the UMRA, 2 U.S.C. 1532, the 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
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).
Before promulgating an EPA rule for which a written statement is needed under Section
202 of the UMRA, Section 205, 2 U.S.C. 1535, of the UMRA generally requires the EPA to
identify and consider a reasonable number of regulatory alternatives and adopt the least costly,
most cost-effective, or least burdensome alternative that achieves the objectives of the rule.
Moreover, section 205 allows the EPA to adopt an alternative other than the least costly, most
cost-effective or least burdensome alternative if the Administrator publishes with the final rule
an explanation why that alternative was not adopted.
Furthermore, as the EPA stated in the preamble, the EPA is not directly establishing any
regulatory requirements that may significantly or uniquely affect small governments, including
Tribal governments. Thus, under the CSAPR Update, the EPA is not obligated to develop under
Section 203 of the UMRA a small government agency plan.
The EPA did analyze the economic impacts of the CSAPR Update on government
entities, however. This analysis does not examine potential indirect economic impacts associated
with the CSAPR Update, such as employment effects in industries providing fuel and pollution
control equipment, or the potential effects of electricity price increases on industries and
households.
6-12

-------
6.2.1	Identification of Government-Owned Entities
In this analysis, the EPA considered EGUs which meet the following five criteria: 1)
EGU is represented in NEEDS v5.15; 2) EGU is fossil-fuel fired; 3) EGU is located in a state
covered by this rule; 4) EGU is neither a cogeneration unit nor solid waste incineration unit; and
5) EGU capacity is 25 MW or larger.
The EPA next refined this list of EGUs, narrowing it to those that exhibit at least one of
the following changes under the final rule, in comparison to the base case.
•	Summer fuel use (BTUs) changes by +/- 1% or more
•	Summer generation (GWh) changes by +/- 1% or more
•	NOx summer emissions (tons) changes by +/- 1% or more
From the inventory of units meeting the criteria above, the EPA used Ventyx data to
identify state and municipality-owned utilities and subdivisions in the CSAPR Update region.
The EPA then used IPM-parsed output to associate these plants with individual generating units.
The EPA identified 12 municipality-owned utilities that are potentially affected by the CSAPR
Update.
6.2.2	Overview of Analysis and Results
After identifying potentially affected government entities, the EPA estimated the impact
of the CSAPR Update in 2017 based on the following:
•	total impacts of compliance on government entities; and
•	ratio of government entity impacts to revenues from electricity generation.
The financial burden to owners of EGUs under the CSAPR Update is composed of
compliance and administrative costs. This section outlines the compliance and administrative
costs for the 12 potentially affected government-owned units in the CSAPR Update region.
6-13

-------
6.2.2.1	Methodology for Estimating Impacts of the CSAPR Update on Government
Entities
An entity can comply with the CSAPR Update through any combination of the following:
optimizing existing SCR, turning on idled SCR or SNCR controls, upgrading to state of the art
combustion controls, using allocated allowances, purchasing allowances, or reducing emissions
through a reduction in generation or improved efficiency. Additionally, units with more
allowances than needed can sell these allowances on 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, the EPA estimates compliance
costs as follows:
CCompliance A Coperating+Retrofit A CFuel A CAllowances A CTransaction A R
where C represents a component of cost as labeled, and A R represents the retail value of
foregone electricity generation.
In reality, 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 CSAPR
Update, for example, some units will forgo some level of electricity generation (and thus
revenues) to comply, this impact will be lessened on these entities by the projected increase in
electricity prices under the policy, while those not reducing generation levels will see an increase
in electricity revenues. 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 rule on small entities.
In this analysis, the EPA used IPM-parsed output for the base case and the CSAPR
Update to estimate compliance cost at the unit level. These costs were then summed for each
entity, adjusting for ownership share. Compliance cost estimates were based on the following:
operating and retrofit costs, sale or purchase of allowances, and the change in fuel costs or
6-14

-------
electricity generation revenues under the CSAPR Update relative to the base case. These
components of compliance cost were estimated as follows:
(1)	Operating and retrofit costs: Using the IPM-parsed output for the base case and
the CSAPR Update, the EPA identified units that install control technology under
the policy and the technology installed. The equations for calculating retrofit costs
were adopted from the EPA's version of IPM. The model calculates the capital
cost (in $/MW); the fixed operation and maintenance (O&M) cost (in
$/MW-year); the variable O&M cost (in $/MWh); and the total annual cost for
units projected to turn on existing idled SCR, fully operate existing SCR, or turn
on existing idled SNCR.
(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 $1,400 per ton. $1,400 per ton is the marginal annualized
cost of NOx reductions used to set the budgets. While this is a reasonable
approximation, it is possible that the actual allowance price could be lower. Units
were assumed to purchase or sell allowances to exactly cover their projected
emissions under the CSAPR Update.
(3)	Fuel costs: The change in fuel expenditures under the policy was estimated by
taking the difference in projected fuel expenditures between the illustrative
CSAPR Update and the base case.
(4)	Value of electricity generated: To estimate the value of electricity generated, the
projected level of electricity generation is multiplied by the regional-adjusted
retail electricity price ($/MWh) estimate, for all entities except those categorized
as Private in Ventyx. For private entities, the EPA used wholesale electricity price
instead 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 and thus their revenue was estimated with
wholesale electricity prices.
6-15

-------
(5) Administrative costs: Because most affected units are already monitored as a
result of other regulatory requirements, the EPA considered the primary
administrative cost to be transaction costs related to purchasing or selling
allowances. The 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
International.
6.2.2.2	Results
A summary of economic impacts on government owned entities is presented in Table 6-5.
According to the EPA's analysis, the total net economic impact on government-owned entities
(state- and municipality-owned utilities and subdivisions) is expected to be $20.5 million in
2017.117
Table 6-5. Summary of Potential Impacts on Government Entities under the CSAPR
Update in 2017
EGU Ownership
Type
Potentially
Affected Entities
Projected
Annualized
Costs ($2011
millions)
Number of
Government
Entities with
Compliance
Costs >1% of
Generation
Revenues
Number of
Government
Entities with
Compliance
Costs >3% of
Generation
Revenues
Municipal
11
$14.7
2
2
State
1
$5.8
1
1
Total
12
$20.5
3
3
Note: The total number of entities with costs greater than 1 percent or 3 percent of revenues includes only entities
experiencing positive costs
As was done for the small entities analysis, the EPA further assessed the economic and
financial impacts of the rule using the ratio of compliance costs to the value of revenues from
electricity generation in the base case, also focusing specifically on entities for which this
117
All costs are reported in 2011 dollars.
6-16

-------
measure is greater than 1 percent.118 The EPA projects that 3 government entities may have
compliance costs greater than 1 percent of revenues from electricity generation in 2017. The
majority of the units that have higher costs are not expected to make operational changes as a
result of this rule (e.g., turn on controls). Their increased costs are largely due to a change in
generation level, which results in a decrease in electricity revenue. This approach is more
indicative of a significant impact when an analysis is looking at entities operating in a
competitive market environment. Government-owned entities do not operate in a competitive
market environment and therefore will be able to recover expenses under the CSAPR Update
through rate increases. Given this, the EPA considers the 1 percent measure in this case a crude
measure of the extent to which rate increases will be made at publicly owned companies.
For municipality- and state-owned entities, the capacity-weighted average economic
impact as a share of base case revenue is slightly less than zero percent. This average reflects the
fact that 6 of the 12 entities are projected to experience a negative economic impact as a share of
base case revenue, which implies that this group of 5 entities experiences a net savings under the
CSAPR Update.
The various components of annual incremental cost under the CSAPR Update to
government entities are summarized in Table 6-6. In 2017, state and municipal entities are a net
purchaser of allowances, and experience both a decrease in fuel expenditures and a decrease in
electricity revenue under the CSAPR Update. Incremental fuel costs are negative because most
of these entities are projected to decrease generation
Table 6-6. Incremental Annual Costs under the CSAPR Update Summarized by
Ownership Group and Cost Category (2011$ millions) in 2017
EGU
Ownership
Type
Retrofit +
Operating
Cost
Net Purchase
of Allowances
Fuel Cost
Lost
Electricity
Revenue
Administrative
Cost
Municipal
-$0.8
$1.0
-$5.8
-$20.2
$0.1
State
-$1.4
$0.9
-$5.2
-$11.6
$0.0
118 Neither the costs nor the revenues of units that retire under the illustrative CSAPR Update are included in this
portion of the analysis. Because these units are better off retiring under the policy than continuing operation, the true
cost of the rule on these units is not represented by our modeling. The true cost of the policy for these units is the
differential between their costs in the base case and the costs of meeting their customers' demand under the rule.
6-17

-------
Source: IPM analysis
6.2.3 Summary of Government Entity Impacts
The EPA examined the potential economic impacts on government-owned entities
associated with this rulemaking based on assumptions of how the affected states will implement
control measures to meet their emissions. According to the EPA's analysis, the total net
economic impact on government-owned entities is expected to be $20.5 million in 2017. This
does not mean that each government entity will experience net cost as the overall net savings is
driven by some entities garnering savings. Of the 12 government entities considered in this
analysis, three may experience compliance costs in excess of 1 percent of revenues in 2017,
based on our assumptions of how the affected states implement control measures to meet their
emissions budgets as set forth in this rulemaking.
Government entities projected to experience compliance costs in excess of 1 percent of
revenues have some potential for significant impact resulting from implementation of the
CSAPR Update. However, as noted above, it is the EPA's position that because these
government entities can pass on their costs of compliance to rate-payers, they will not be
significantly affected.
6.3 Employment
Executive Order 13563 directs federal agencies to consider regulatory impacts on job
creation and employment. According to the Executive Order, "our regulatory system must
protect public health, welfare, safety, and our environment while promoting economic growth,
innovation, competitiveness, and job creation. It must be based on the best available science"
(Executive Order 13563, 2011). Although standard benefit-cost analyses have not typically
included a separate analysis of regulation-induced employment impacts,119 we typically conduct
employment analyses for economically significant rules. This section discusses and projects
potential employment impacts related to today's final rule.120
119	Labor expenses do, however, contribute toward total costs in the EPA's standard benefit-cost analyses.
120	The employment analysis in this RIA is part of EPA's ongoing effort to "conduct continuing evaluations of
potential loss or shifts of employment which may result from the administration or enforcement of [the Act]"
pursuant to CAA section 321(a).
6-18

-------
Section 6.3.1 describes the theoretical framework used to analyze regulation-induced
employment impacts, discussing how economic theory alone cannot predict whether such
impacts are positive or negative. Section 6.3.2 presents an overview of the peer-reviewed
literature relevant to evaluating the effect of environmental regulation on employment. Section
6.3.3 provides background regarding recent employment trends in the electricity generation, coal
and natural gas extraction sectors. Section 6.3.4 discusses the potential direct employment
impacts in these sectors.
6.3.1 Economic Theory and Employment
Regulatory employment impacts are difficult to disentangle from other economic changes
affecting employment decisions over time and across regions and industries. Labor market
responses to regulation are complex. They depend on labor demand and supply elasticities and
possible labor market imperfections (e.g., wage stickiness, long-term unemployment, etc.). The
unit of measurement (e.g., number of jobs, types of jobs, hours worked, and earnings) may affect
observability of those responses. Net employment impacts are composed of a mix of potential
declines and gains in different areas of the economy (e.g., the directly regulated sector, the
environmental protection sector, upstream and downstream sectors, etc.) over time. In light of
these difficulties, economic theory provides a constructive framework for analysis.
Microeconomic theory describes how firms adjust their use of inputs in response to
changes in economic conditions.121 Labor is one of many inputs to production, along with capital,
energy, and materials. In competitive markets, firms choose inputs and outputs to maximize
profit as a function of market prices and technological constraints.122 123 Berman and Bui (2001)
adapt this model to analyze how environmental regulations affect labor demand.124 They model
environmental regulation as effectively requiring certain factors of production, such as pollution
abatement capital, at levels that firms would not otherwise choose. Berman and Bui (2001)
121	See Layard and Walters (1978), a standard microeconomic theory textbook, Chapter 9, for a discussion.
122	See Hamermesh (1993), Chapter 2, for a derivation of the firm's labor demand function from cost-minimization.
123	In this framework, labor demand is a function of quantity of output and prices (of both outputs and inputs).
124	Morgenstern, Pizer, and Shih (2002) develop a similar model.
6-19

-------
model two components that drive changes in firm-level labor demand: output effects and
substitution effects.125 Regulation affects the profit-maximizing quantity of output by changing
the marginal cost of production. If a regulation causes marginal production cost to increase, it
will place upward pressure on output prices, leading to a decrease in quantity demanded, and
resulting in a decrease in production. The output effect describes how, holding labor intensity
constant, a decrease in production causes a decrease in labor demand. As noted by Berman and
Bui, although many assume that regulations must increase marginal cost, in some cases they may
decrease it. A regulation could induce a firm to upgrade to less polluting and more efficient
equipment that lowers the marginal cost of production. In such a case, output could increase after
firms comply with the regulation. An unregulated profit-maximizing firm may not have chosen
to install such an efficiency-improving technology if the return on investment were too low, but
once the technology is in place it lowers marginal production costs.
The substitution effect describes how, holding output constant, regulation affects the labor-
intensity of production. Although increased environmental regulation may increase use of
pollution control equipment and energy to operate that equipment, the impact on labor demand is
ambiguous. For example, equipment inspection requirements, specialized waste handling,
completing required paperwork, or pollution technologies that alter the production process may
affect the number of workers necessary to produce a unit of output. Berman and Bui (2001)
model the substitution effect as the effect of regulation on pollution control equipment and
expenditures required by the regulation and the corresponding change in the labor-intensity of
production.
In summary, as output and substitution effects may be positive or negative, economic
theory alone cannot predict the direction of the net effect of regulation on labor demand. In
addition, the empirical literature illustrates difficulties with estimation of net employment
impacts. The most commonly used empirical methods, for example, Greenstone (2002), likely
overstate employment impacts because they rely on relative comparisons between more
125 The authors also discuss a third component, the impact of regulation on factor prices, but conclude that this effect
is unlikely to be important for large competitive factor markets, such as labor and capital. Morgenstern, Pizer and
Shih (2002) use a very similar model, but they break the employment effect into three parts: 1) a demand effect; 2) a
cost effect; and 3) a factor-shift effect.
6-20

-------
regulated and less regulated counties, which can lead to "double counting" of impacts when
production and employment shift from more regulated towards less regulated areas. Thus these
empirical methods cannot be used to estimate net employment effects.126
The conceptual framework described thus far focused on regulatory effects on plant-level
decisions within a regulated industry, but employment impacts at an individual plant do not
necessarily represent impacts for the sector as a whole. At the industry-level, labor demand is
more responsive if: (1) the price elasticity of demand for the product is high, (2) other factors of
production can be easily substituted for labor, (3) the supply of other factors is highly elastic, or
(4) labor costs are a large share of total production costs.127 For example, if all firms in an
industry are faced with the same regulatory compliance costs and product demand is inelastic,
then industry output may not change much, and output of individual firms may change slightly.128
In addition to changes to labor demand in the regulated industry, net employment impacts
encompass changes in other related sectors such as the environmental protection sector. This
final rule may increase demand for the nitrogenous reagent (typically ammonia or urea) used in
SCRs and SNCRs to reduce NOx, which may increase revenue and employment in the firms
providing these chemicals.
If the U.S. economy is at full employment, even a large-scale environmental regulation is
unlikely to have a noticeable impact on aggregate net employment.129 Instead, labor in affected
sectors would primarily be reallocated from one productive use to another (e.g., from producing
electricity to manufacturing, installing, or operating and maintaining pollution-abatement
equipment), and net national employment effects from environmental regulation would be small
and transitory (e.g., as workers move from one job to another).130 Some workers may retrain or
126	See Greenstone (2002) p. 1212.
127	See Ehrenberg & Smith (2000), p. 108.
128	This example is from Berman and Bui (2001), pp. 293.
129	Full employment is a conceptual target for the economy where everyone who wants to work and is available to
do so at prevailing wages is actively employed. The unemployment rate at full employment is not zero.
130	Arrow el al. (1996); see discussion on bottom of p. 8. In practice, distributional impacts on individual workers
can be important, as discussed in later paragraphs of this section.
6-21

-------
relocate in anticipation of new requirements or require time to search for new jobs, while
shortages in some sectors or regions could bid up wages to attract workers. These adjustment
costs can lead to local labor disruptions.
If, on the other hand, the economy is operating at less than full employment, economic
theory does not clearly indicate the direction or magnitude of the net impact of environmental
regulation on employment; it could cause either a short-run net increase or short-run net decrease
(Schmalansee and Stavins, 2011). For example, the Congressional Budget Office considered
EPA's MATS and regulations for industrial boilers and process heaters as potentially leading to
short-run net increases in economic growth and employment, driven by capital investments for
compliance with the regulations (Congressional Budget Office, 2011). Environmental regulation
may also affect labor supply and productivity. In particular, reducing pollution and other
environmental risks may improve labor productivity or employees' ability to work.131 While the
theoretical framework for analyzing labor supply effects is analogous to that for labor demand, it
is more difficult to study empirically. There is a small emerging literature that uses detailed labor
and environmental data to assess these impacts.
To summarize, economic theory provides a framework for analyzing the impacts of
environmental regulation on employment. The net employment effect incorporates expected
employment changes (both positive and negative) in the regulated sector and other related
sectors including the environmental protection sector. Labor demand impacts for regulated firms,
and also for the regulated industry, can be decomposed into output and substitution effects which
may be either negative or positive. Estimation of net employment effects for regulated sectors is
possible when data of sufficient detail and quality are available. Finally, economic theory
suggests that labor supply effects are also possible. In the next section, we discuss the empirical
literature.
6.3.1.1 Current State of Knowledge Based on the Peer-Reviewed Literature
The peer-reviewed empirical literature specifically estimating employment effects of
environmental regulations is limited but growing. We summarize it briefly in this section.
131 E.g. Graff Zivin and Neidell (2012).
6-22

-------
6.3.1.2	Regulated Sector
Several empirical studies, including Berman and Bui (2001) and Ferris, Shadbegian, and
Wolverton (2014), suggest that regulation-induced net employment impacts may be zero or
slightly positive, but small in the regulated sector. Gray et al (2014) find that pulp mills that had
to comply with both the air and water regulations in EPA's 1998 "Cluster Rule" experienced
relatively small, and not always statistically significant, decreases in employment. Other research
on regulated sectors suggests that employment growth may be lower in more regulated areas
(Greenstone 2002, Walker 2011, 2013). However, since these latter studies compare more
regulated to less regulated counties, this methodological approach likely overstates employment
impacts to the extent that regulation causes plants to locate in one area of the country rather than
another, which would lead to "double counting" of the employment impacts. List et al. (2003)
find some evidence that this type of geographic relocation may be occurring.
6.3.1.3	Economy-Wide
Given the difficulty with estimating national impacts of regulations, EPA has not generally
estimated economy-wide employment impacts of its regulations in its benefit-cost analyses.
However, in its continuing effort to advance the evaluation of costs, benefits, and economic
impacts associated with environmental regulation, EPA has formed a panel of experts as part of
EPA's Science Advisory Board (SAB) to advise EPA on the technical merits and challenges of
using economy-wide economic models to evaluate the impacts of its regulations, including the
impact on net national employment.132 Once EPA receives guidance from this panel, it will
carefully consider this input and then decide if and how to proceed on economy-wide modeling
of employment impacts of its regulations.
6.1.4 Labor Supply Impacts
The empirical literature on environmental regulatory employment impacts focuses
primarily on labor demand. However, there is a nascent literature focusing on regulation-induced
effects on labor supply.133 Although this literature is limited by empirical challenges, researchers
132	For further information see:
http://yosemite.epa.gOv/sab/sabproduct.nsf/0/07E67CF77B54734285257BB0004F87ED70penDocument
133	For a recent review see Graff-Zivin and Neidell (2013).
6-23

-------
have found that air quality improvements lead to reductions in lost work days (e.g., Ostro, 1987).
Limited evidence suggests worker productivity may also improve when pollution is reduced.
Graff Zivin and Neidell (2012) used detailed worker-level productivity data from 2009 and 2010,
paired with local ozone air quality monitoring data for one large California farm growing
multiple crops, with a piece-rate payment structure. Their quasi-experimental structure identifies
an effect of daily variation in monitored ozone levels on productivity. They find "ozone levels
well below federal air quality standards have a significant impact on productivity: a 10 parts per
billion (ppb) decreases in ozone concentrations increases worker productivity by 5.5 percent."
(Graff Zivin and Neidell, 2012, p. 3654).134
6.3.1.5 Conclusion
This section has outlined the challenges associated with estimating regulatory effects on
both labor demand and supply for specific sectors. These challenges make it difficult to estimate
net national employment estimates that would appropriately capture the way in which costs,
compliance spending, and environmental improvements propagate through the macro-economy.
6.3.2 Recent Employment Trends
The U.S. electricity system includes employees that support electric power generation,
transmission and distribution; the extraction of fossil fuels; and supply-side and demand-side
energy efficiency. This section describes recent employment trends in the electricity system.
6.3.2.1 Electric Power Generation
In 2014, the electric power generation, transmission and distribution sector (NAICS 2211)
employed about 390,000 workers (U.S. BLS, 2015) in the U.S. Installation, maintenance, and
repair occupations accounted for the largest share of workers (25 percent) (U.S. BLS, 2014).
These categories include inspection, testing, repairing and maintaining of electrical equipment
and/or installation and repair of cables used in electrical power and distribution systems. Other
major occupation categories include office and administrative support (18 percent), production
occupations (16 percent), architecture and engineering (10 percent), business and financial
134 The EPA is not quantifying productivity impacts of reduced pollution in this rulemaking using this study. In light
of this recent research, however, the EPA is considering how best to incorporate possible productivity effects in the
future.
6-24

-------
operations (7 percent) and management (7 percent). Asd shown in Figure 6-1, employment in the
electric power industry averaged about 420,000 workers from 2000 to 2005, declining to an
average of about 400,000 workers for the rest of the decade, and then declining to about 390,000
workers in 2014.
Figure 6-1. Electric Power Industry Employment
Power Generation and Supply Employment
(NAICS = 2211, Annual Average, 1000s of Employees)
500
350
300
250
200
150
100
50
0
6.3.2.2 Fossil Fuel Extraction
Coal Mining. The coal mining sector (NAICS 2121) is primarily engaged in coal mining and
coal mine site development, excluding metal ore mining and nonmetallic mineral mining and
quarrying. In 2014, BLS reported about 74,000 coal mining employees (Figure 6-2). During the
2000 to 2014 period, coal mining employment peaked in 2011 at about 87,000 employees.
Figure 6-2. Coal Production Employment
6-25

-------
Coal Mining Employment
(NAICS = 2121, Annual Average, 1000s of Employees)
100
90
80
70
60
50
40
30
20
10
0
Source: BLS (2014a)
Oil and Gas Extraction. In 2014, there were close to 200,000 employees in the oil and gas
extraction sector (NAICS 211). This sector includes production of crude petroleum, oil from oil
shale and oil sands, production of natural gas, sulfur recovery from natural gas, and recovery of
hydrocarbon liquids. Activities include the development of gas and oil fields, exploration
activities for crude petroleum and natural gas, drilling, completing, and equipping wells, and
other production activities. In contrast with coal, Figure 6-3 shows there has been a sharp
increase in employment in this sector over the past decade.
6-26

-------
Figure 6-3 Oil and Gas Extraction Employment
Oil and Gas Extraction Employment
(NAICS = 211, Annual Average, 1000s of Employees)
250
100
50
0
Source: BLS (2014b)
6.3.3 Power and Fuels Sector Direct Employment Impacts
As described above, affected EGUs may respond to the CSAPR Update by upgrading or
improving performance of existing combustion controls, or by upgrading, improving, or utilizing
post-combustion NOx systems already in place. In addition, some generation may shift from
higher NOx-emitting EGUs to units with lower or zero NOx emission rates. All of these NOx-
related changes will likely result in changes in the amount of the various types of amount of
labor needed in different parts of the fuels and utility power sectors. There also may be other
labor impacts in sectors that provide products and materials used in reducing NOx emissions at
EGUs, such as catalysts used in SCR control systems. These direct labor impacts will likely
include both increased demand for certain types of labor in some portions of the affected sectors,
and reduced demand for labor in other portions of the affected sectors.
Installing and operating new equipment could change labor demand in the electricity
generating sector itself, as well as associated equipment and services sectors. Specifically, the
6-27

-------
direct employment effects in the power sector that could occur because of actions taken by the
2017 ozone season include:
•	Optimizing NOx removal from existing and operational SCR systems;
•	Turning on and optimizing idled SCR and SNCR systems;
•	Installing, optimizing or upgrading combustion-side improvements resulting in reduced
NOx emissions;
•	Shifting generation from units with higher NOx emission rates to units with lower or zero
emission rates.
In addition, there could be directly induced employment impacts (both positive and
negative) in the labor demand in the industries supplying fossil fuels to the power sector and
industries supplying materials used by the NOx reduction systems. Once implemented, both the
potential increases in operating efficiency of NOx reductions, as well as shifting generation to
lower NOx-emitting or zero-emitting EGUs, could impact the utility power sector's demand for
fossil fuels, and hence the demand for labor needed in the coal mining and gas extraction sectors.
The direct net employment impacts of the final rule, in terms of the power sector and fuels
sector, however, are anticipated to be relatively small. This is consistent with the relatively small
estimated changes in the power sector's overall cost of generation, as well as relatively small
changes in generation, fuel use, capacity, and the percent of total generation produced by each
type of fuel.
For example, for the final rule in 2017, the estimated impacts relevant to changes in labor
demand include:
•	The overall total national cost of generation in 2017 decreases by 0.01 percent;
•	Total net generation increases by 0.001 percent (coal generation decreases by 0.17 percent,
and natural gas generation increases by 0.18 percent);
•	The power sector's total tons of coal used for electricity generation decreases by 0.25
percent (or 0.19 percent decrease in BTUs);
•	Total natural gas use increases by 0.20 percent.
6-28

-------
The results of the power sector modeling suggest that because of the very small changes in
the power and fuels sector, the direction and magnitude of the potential labor impacts are very
small in all three regulatory alternatives analyzed. To illustrate this point, the direct labor impacts
are quantified for the final regulation for 2017 and 2020. The labor impacts for the more and less
stringent alternatives have not been quantified.
Affected EGUs may respond to the requirement for EGUs in 22 eastern states to reduce
NOx emissions during the ozone season by improving and optimizing existing NOx emission
control systems or to shift generation to lower NOx-emitting or zero-emitting EGUs. Meeting the
new EGU ozone season NOx budget limits will result in changes in the amount of labor needed
in different parts of the utility power sector. Installing and operating new equipment, upgrading
combustion control operations to reduce NOx emissions, and shifting generation to other sources
could affect labor demand in the electricity generating sector itself, as well as associated
equipment and services sectors. Specifically, the direct employment effects of initiatives at
existing fossil EGUs would include increases in labor demand during the implementation phase
for manufacturing, installing, and operating the NOx emissions controls at existing fossil units.
Once implemented, reductions in NOx emissions from existing EGUs and shifting generation to
existing generation resources will impact the utility power sector's demand for fossil fuels and
potentially plans for EGU retirement.
The employment analysis uses the cost projections from IPM to project labor demand
impacts of the final CSAPR Update on affected EGUs in the electricity power sector and the fuel
production sector (coal and natural gas). These projections include effects attributable to
installing and improving the NOx control performance of combustion control systems,
optimizing the operation of post-combustion NOx control systems, generation shifts, and
changes in fuel use. The following section presents the EPA's quantitative projections of
potential employment impacts in the electricity generation sector, as well as the impacts in the
coal and natural gas fuel sectors.
6-29

-------
6.3.3.1 Methods Used to Estimate Changes in Employment in Electricity Generation and Fuel
Supply
The analytical approach used in this analysis is a bottom-up engineering method
combining the EPA's cost analysis of compliance with the NOx emissions budgets with data on
labor productivity, engineering estimates of the amount and types of labor needed to
manufacture, construct, and operate different types of NOx control systems, and prevailing wage
rates for skilled and general labor categories. Lacking robust peer-reviewed methods to estimate
economy-wide impacts, the engineering-based analysis focuses on the supply-side direct impact
on labor demand in industries closely involved with electricity generation. The engineering
approach projects labor changes measured as the change in each analysis year in job-years
employed in the utility power sector and directly related sectors (e.g., emission control
equipment manufacturing and fuel supply). Some of the quantified employment impacts in this
analysis are one-time impacts, such as changes associated with upgrading the combustion
controls. Other labor impacts will continue, such as changes associated with operating and
maintaining generating units that will be constructed or retired, shifting generation to lower
emitting generating units, and changes in the demand for labor providing the fuels supplied to the
affected fossil-fired EGUs. All of these continuing labor impacts are estimated as annual impacts
on employment.
The methods the EPA uses to estimate the labor impacts are based on the analytical
methods used in many previous EPA regulatory analyses. The methods used in this analysis to
estimate many of the labor impacts (e.g., labor associated with changes in operating and
maintaining generating units, as well as labor needed to mine coal and natural gas) are the same
as we used in the Clean Power Plan (CPP) (U.S. EPA, 2015) and CSAPR (U.S. EPA, 2011), with
updated data where available. In addition, a central feature of the labor analysis for this RIA,
involves the labor needs of upgrading and optimizing NOx control systems on existing EGUs in
the affected 22-state region. In addition to the changes at EGUs within the 22-state region, there
are also estimated changes in the utilization of existing generating units in other states, as well as
changes in the gas and coal supply sectors.
The methods and data used to estimate the labor associated with upgrading combustion
control systems to reduce NOx emissions rely on three critical components:
6-30

-------
•	The mix of labor categories needed to implement the NOx combustion control
upgrades (i.e., the share of the labor cost of the upgrades apportioned to general
construction, boilermaker, engineering and management labor) is the same as was
used for heat rate improvement combustion control upgrades needed in the final
CPP RIA analysis.
•	The fully loaded labor cost of each labor category is the same as was used for the
NOx control upgrades and is the same labor cost assumed for heat rate
improvements in the CPP final RIA.
•	The amount of labor needed to implement the NOx combustion control upgrades is
derived from the total costs of the NOx combustion upgrades estimated by IPM.
The labor analysis relies on an estimate (McAdams et al., 2001) that the labor
needed to install the combustion upgrades accounts for 30% of the total cost, and
the remaining 70% of the total cost is for capital expenditures on equipment.135
6.3.3.2 Estimates of the Changes in Employment in Electricity Generation and Fuel Supply
The estimated labor impacts of the revisions to the NOx budgets from EGUs in the 22-state
region are presented in Table 6-7. Given the methods the EPA uses to estimate labor impacts, it
is not possible to directly separate the labor impacts that occur within the 22-state region from
the labor impacts in the states not in the region. However, all the labor changes associated with
combustion control upgrades, and optimization of existing post-combustion NOx control
systems, will occur within the 22-state region. The fuel supply labor impacts, however, will
occur both within the 22-state region and in other states. This occurs for two reasons. First, coal
and natural gas used at EGUs throughout the United States are both extracted within the 22-state
region and in other states. Second, the shifts in fossil-fired generation will also occur both within
the 22-state region and in other states.
135 In the RIA for the proposed CASPR Update, labor was assumed to account for 40% of the total cost. The 40%
estimate was consistent with the labor share of cost for heat rate improvements used in the CPP RIA. The 30% labor
share estimate used in this final CSAPR Update analysis comes from a published article (McAdams et al., 2001),
which specifically examined the labor and capital costs of improving NOx emission rates at industrial boilers by
retrofitting flue gas recirculation systems and upgraded low-NOx burners.
6-31

-------
Table 6-7. Annual Net Employment Impacts for Power and Fuels Sectors in 2017 & 2020

2017
2020
Upgrades and Optimization
SCR
11
14
SNCR*
0
0
Combustion Control
55
66
Upgrades & Optimization Sub-
Total
65
80
Plant Retirement
Coal
0
-366
Fuel Use Change
Coal
-95
-339
Natural Gas
87
128
Fuel Use Sub-Total
-8
-211
Net Employment Impact
58
-497
* All results in this table are those for the CSAPR Update alternative only. Turning on idled SNCR takes place only
in the more stringent alternative. Job-year estimates are derived from IPM investment and upgrade estimates, as well
as IPM fuel use estimates (tons coals or MMBtu gas). Employment impacts in the upgrades and optimization
category includes both employment on-site (e.g., installing improved combustion control systems) and employment
involved in manufacturing the improved combustion control systems. All job-year estimates are full-time equivalent
(FTE) jobs.
6.4 References
Arrow, K. J.; M. L. Cropper; G. C. Eads; R. W. Hahn; L. B. Lave; R. G. Noll; Paul R. Portney;
M. Russell; R. Schmalensee; V. K. Smith and R. N. Stavins. 1996. "Benefit-Cost Analysis in
Environmental, Health, and Safety Regulation: A Statement of Principles." American
Enterprise Institute, the Annapolis Center, and Resources for the Future; AEI Press.
Available at: . Accessed June 5, 2015.
Berman, E. and L. T. M. Bui. 2001. "Environmental Regulation and Labor Demand: Evidence
from the South Coast Air Basin." Journal of Public Economics. 79(2): 265-295.
Congressional Budget Office (2011), Statement of Douglas W. Elmendorf, Director, before the
Senate Budget Committee, "Policies for Increasing Economic Growth and Employment in
2012 and 2013" (November 15)
Ehrenberg, R. G. and R. S. Smith. 2000. Modern Labor Economics: Theory and Public Policy.
Addison Wesley Longman, Inc., Chapter 4.
Executive Order 13563 (January 21, 2011). "Improving Regulation and Regulatory Review.
Section 1. General Principles of Regulation." Inderal Register 76(14): 3821-3823.
Ferris, A. E., R. J. Shadbegian, A. Wolverton. 2014. "The Effect of Environmental Regulation on
Power Sector Employment: Phase I of the Title IV SO2 Trading Program." Journal of the
Association of Environmental and Resource Economists. 1(4): 521-553.
6-32

-------
Graff Zivin J. and M. Neidell. 2012. "The Impact of Pollution on Worker Productivity."
American Economic Review. 102(7):3652-73.
Gray, W., R. J. Shadbegian, C. Wang and M. Meral. 2014 "Do EPA Regulations Affect Labor
Demand? Evidence from the Pulp and Paper Industry", Journal of Environmental Economics
and Management. 68: 188-202.
Greenstone, M. 2002. "The Impacts of Environmental Regulations on Industrial Activity:
Evidence from the 1970 and 1977 Clean Air Act Amendments and the Census of
Manufactures." Journal of Political Economy. 110(6): 1175-1219.
Hamermesh, D. S. 1993. Labor Demand. Princeton, NJ: Princeton University Press. Chapter 2.
Layard, P.R.G. and A. A. Walters. 1978. Microeconomic Theory. McGraw-Hill, Inc. Chapter 9.
List, J. A.; D. L. Millimet; P. G. Fredriksson and W. W. McHone. 2003. "Effects of
Environmental Regulations on Manufacturing Plant Births: Evidence from a Propensity
Score Matching Estimator." The Review of Economics and Statistics. 55(4): 944-952.
McAdams, J. D., S. D. Reed and D.C. Itse. 2001. "Minimize NOx Emissions Cost-Effectively "
Hydrocarbon Processing. June, 2001. Pgs. 51-58. Available at:
. Accessed
June 14, 2016.
Ostro, B.D. 1987. "Air Pollution and Morbidity Revisited: A Specification Test." Journal of
Environmental Economics Management. 14:87-98.
Schmalansee, R. and R. Stavins (2011). "A Guide to Economic and Policy Analysis for the
Transport Rule." White Paper. Boston, MA. Exelon Corp.
Walker, W. R. 2011."Environmental Regulation and Labor Reallocation." American Economic
Review. 101(2): 442-47.
Walker, W. R. 2013."The Transitional Costs of Sectoral Reallocation: Evidence From the Clean
Air Act and the Workforce." The Quarterly Journal of Economics 128 (4): 1787-1835.
U.S. Bureau of Labor Statistics (BLS). 2014. "Occupational Employment Statistics, May 2014
National Industry-Specific Occupational Employment and Wage Estimates, Electric Power
Generation, Transmission, and Distribution (NAICS 2211)". Available at:
. Accessed June 9, 2015.
U.S. Bureau of Labor Statistics (BLS). 2014a. "May 2014 National Industry-Specific
Occupational Employment and Wage Estimates: NAICS 212100 - Coal Mining". Available
at: < http://www.bls.gOv/oes/current/naics4_212100.htm#00-0000>. Accessed June 9, 2015.
U.S. Bureau of Labor Statistics (BLS). 2014b. "May 2014 National Industry-Specific
Occupational Employment and Wage Estimates: NAICS 212100 - Oil and Gas Extraction".
Available at: < http://www.bls.gov/oes/current/naics4_212100.htm>. Accessed June 9, 2015.
6-33

-------
U.S. Bureau of Labor Statistics (BLS). 2015. "Current Employment Survey Seasonally Adjusted
Employment for Electric Power Generation (national)" and "Current Employment Survey
Seasonally Adjusted Employment for Transmission, and Distribution (national)." Series ID:
CES4422110001. Available at . Accessed June 9, 2015.
6-34

-------
CHAPTER 7: COMPARISON OF BENEFITS AND COSTS
Overview
The EPA performed an illustrative analysis to estimate the costs, human health benefits,
and climate co-benefits of compliance with the proposed and more and less stringent alternatives
and is finalizing EGU NOx ozone season emissions budgets for 22 states.136 The emissions
reductions evaluated in the CSAPR update reflect EGU NOx reduction strategies that are
achievable for the 2017 ozone season. The EPA has quantified EGU NOx ozone-season
emissions budgets reflecting EGU NOx reduction strategies that are widely available at a
uniform annualized cost of $1,400 per ton (2011$). For the RIA, in order to implement the OMB
Circular A-4 requirement to assess at least one less stringent and one more stringent alternative
to the CSAPR update, the EPA has also analyzed EGU NOx ozone season emissions budgets
reflecting NOx reduction strategies that are widely available at a uniform annualized cost of
$800 per ton (2011$) and strategies that are widely available at a uniform annualized cost of
$3,400 per ton (2011$). This chapter summarizes these results.
7.1 Results
As shown in Chapter 4, the estimated annualized costs to implement the CSAPR update,
are approximately $68 million (2011 dollars, rounded to two significant figures). As shown in
Chapter 5, the total estimated combined benefits from implementation of the CSAPR update are
approximately $530 million to $880 million in 2017 (2011 dollars, rounded to two significant
figures). EPA can thus calculate the net benefits of the CSAPR update by subtracting the
estimated annualized costs from the estimated benefits in 2017. The net benefits of the CSAPR
update are approximately $460 to $810 million (based on air quality benefits discounted at 3
percent, the central estimate of CO2 co-benefits, and annualized cost estimates). Therefore, the
EPA expects that implementation of this rule, based solely on economic efficiency criteria, will
provide society with a significant net gain in social welfare, notwithstanding the expansive set of
health and environmental effects we were unable to quantify. Further quantification of directly
emitted PM2.5-, mercury-, acidification-, and eutrophication-related impacts would increase the
136 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia, West Virginia, and
Wisconsin.
7-1

-------
estimated net benefits of the rule. Table 7-1 presents a summary of the benefits, costs, and net
benefits of the CSAPR update and also the more and less stringent alternatives.
Table 7-1. Total Costs, Total Monetized Benefits, and Net Benefits of the CSAPR Update
and More and Less Stringent Alternatives in 2017 for U.S. (millions of
2011$)a'b'c'd
CSAPR Update
More Stringent
Alternative
Alternative
Less Stringent
Alternative
Climate Co-Benefits
Air Quality Health Benefits
Total Benefits
Annualized Compliance
Costs
Net Benefits
Non-Monetized Benefits®
$66
$450 to $790
$520 to $860
$87
$490 to $850
$580 to $940
$82
$490 to $850
$54
$190 to $330
$240 to $390
$240 to $380
$450 to $790
Non-monetized climate benefits
Reductions in exposure to ambient NO2
Ecosystem benefits and visibility improvments assoc. with reductions in
emissions of NOx
a Estimating multiple years of costs and benefits is limited for this RIA by data and resource limitations. As a result,
we provide compliance costs and social benefits in 2017, using the best available information to approximate
compliance costs and social benefits recognizing uncertainties and limitations in those estimates.
b Benefits ranges represent discounting of health benefits and climate co-benefits at a discount rate of 7 percent. See
Chapter 5 for additional detail and explanation. The costs presented in this table reflect compliance costs annualized
at a 4.77 percent discount rate and do not include monitoring, recordkeeping, and reporting costs, which are reported
separately. See Chapter 4 for additional detail and explanation.
0 All costs and benefits are rounded to two significant figures; columns may not appear to add correctly.
d Ozone and PM2 5 benefits from NOx emission reductions are for the 22-state region only.
e Non-monetized benefits descriptions are for all three alternatives and are qualitative.
In accordance with Circular A-4 Guidance (OMB, 2003), the EPA also analyzed the costs
and benefits of two regulatory control alternatives that impose relatively more stringent and
relatively less stringent EGU NOx emissions budgets, compared to the CSAPR Update. 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 final NOx emissions budgets. Table 7-2
presents the projected emissions reductions for ozone season NOx, as well as reductions in co-
pollutant annual NOx, annual SO2, and annual CO2, in 2017 under the CSAPR update and the
more and less stringent alternatives.
7-2

-------
Table 7-2. Projected 2017* Changes in Emissions of NOxand CO2 with the proposed
NOx Emissions Budgets and More or Less Stringent Alternatives (Tons)

CSAPR update
More Stringent
Alternative
Less Stringent
Alternative
NOx (annual)
-75,000
-79,000
-27,000
NOx (ozone season)
-61,000
-66,000
-27,000
CO2 (annual short tons)
-1,600,000
-2,000,000
-1,300,000
* Annual reductions are based on 2018 IPM direct model outputs relied upon in this RIA to represent 2017 co-
pollutant reductions
In this RIA, we quantify an array of adverse health impacts attributable to ozone and
PM2.5. The Integrated Science Assessment for Ozone and Related Photochemical Oxidants
("Ozone ISA") (U.S. EPA, 2013a) identifies the human health effects associated with ozone
exposure, which include premature death and a variety of illnesses associated with acute (days-
long) and chronic (months to years-long) exposures. Similarly, the Integrated Science
Assessment for Particulate Matter ("PM ISA") (U.S. EPA, 2009) identifies the human health
effects associated with ambient particles, which include premature death and a variety of
illnesses associated with acute and chronic exposures.
The EPA believes that providing comparisons of social costs and social benefits at
discount rates of 3 and 7 percent is appropriate to the extent this is possible given available
models and techniques. The four different uses of discounting in the RIA - (i) construction of
annualized costs, (ii) adjusting the value of mortality risk for lags in mortality risk decreases, (iii)
adjusting the cost of illness for non-fatal heart attacks to adjust for lags in follow up costs, and
(iv) discounting climate co-benefits — are all appropriate. We explain our discounting of
benefits in Chapter 5 of the RIA, specifically the application of 3 and 7 percent to air quality
benefits and 2.5, 3, and 5 percent to climate co-benefits; we explain our discounting of costs, in
which we use a single discount rate of 4.77 percent, in Chapter 4. Our estimates of net benefits
are the approximations of the net value (in 2017) of benefits attributable to emissions reductions
needed to attain just for the year 2017.
The EPA presents annualized costs and benefits in a single year for comparison in this RIA
because there are a number of methodological complexities associated with calculating the net
present value (NPV) of a stream of costs and benefits for a NAAQS. While NPV analysis allows
evaluation of alternatives by summing the present value of all future costs and benefits, insights
into how costs will occur over time, necessary for a NPV calculation, are limited by underlying
7-3

-------
assumptions and data. Calculating a present value (PV) of the stream of future benefits also
poses special challenges, which we describe below. In addition, calculating NPV requires
definition of the length of the future time period considered, which is not straightforward for this
analysis and subject to uncertainty. We provide annualized costs of compliance instead of using
NPV or alternatives in this RIA, and our explanation for this is in Chapter 4.
The theoretically appropriate approach for characterizing the PV of benefits is the life table
approach. The life table, or dynamic population, approach explicitly models the year-to-year
influence of air pollution on baseline mortality risk, population growth and the birth rate—
typically for each year over the course of a 50-to-100 year period (U.S. EPA SAB, 2010; Miller,
2003). In contrast to the pulse approach137, a life table models these variables endogenously by
following a population cohort over time. For example, a life table will "pass" the air pollution-
modified baseline death rate and population from year to year; impacts estimated in year 50 will
account for the influence of air pollution on death rates and population growth in the preceding
49 years.
Calculating year-to-year changes in mortality risk in a life table requires some estimate of
the annual change in air quality levels. It is both impractical and challenging to model air quality
levels for each year and to account for changes in federal, state and local policies that will affect
the annual level and distribution of pollutants. For each of these reasons, the EPA has not
generally reported the PV of benefits for air rules but has instead pursued a pulse approach.
While we agree that providing the NPV of a stream of costs and benefits could be informative,
based on the challenges with calculating NPV outlined above, we are not able to provide the
NPV of a stream of costs and benefits in this RIA.
Finally, with regard to the increment of impacts attributable to the CSAPR Update and
the original CSAPR, the EPA does not believe that the costs and benefits for the original CSAPR
and the CSAPR Update are entirely additive. The EPA recognizes that the majority of the
benefits of the original CSAPR were derived from reductions in SO2 and annual NOx emissions,
and the benefits of the CSAPR Update are primarily based on ozone-season NOx emissions
137 The pulse approach assumes changes in air pollution in a single year and affects mortality estimates over a 20-
year period.
7-4

-------
reductions. However, five years have passed between promulgation of the original CSAPR and
the CSAPR Update, and the two rules have different baselines. In the intervening five years,
changes in the power sector that are independent of these rules, such as changes in fuel costs and
electricity markets as well as other federal and state level actions, which creates challenges when
estimating the sum of the costs and benefits of these two rules. In addition, implementation of
the original CSAPR was delayed such that its two phases were implemented as phase I - limits
to be met by 2015, and phase II - limits to be met by 2017. The reductions estimated for the
CSAPR Update in 2017, given that it replaces remanded original CSAPR budgets, may overlap
with reductions that would have otherwise occurred for phase II. However, the benefits and
costs of CSAPR are still notable given the enduring original CSAPR ozone season NOx budgets,
annual NOx budgets, and SO2 budgets. While the EPA did remove the remanded ozone season
NOx budgets for three states, two of these states (North Carolina and South Carolina) remain
subject to annual NOx requirements. These original CSAPR budgets are all present in EPA's
modeling of the baseline and policy alternatives.
7-5

-------
United States	Office of Air Quality Planning and Standards	Publication No. EPA-452/R-16-004
Environmental Protection	Health and Environmental Impacts Division	September, 2016
Agency	Research Triangle Park, NC

-------