Regulatory Impact Analysis for the Proposed
Cross-State Air Pollution Rule (CSAPR)
Update for the 2008 Ozone National Ambient
Air Quality Standards (NAAQS)
-------
EPA-452/R-15-009
November 2015
Regulatory Impact Analysis for the Proposed Cross-State Air Pollution Rule
(CSAPR) Update for the 2008 Ozone
National Ambient Air Quality Standards (NAAQS)
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
11
-------
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, the Office of Transportation and Air Quality, and the
Office of Policy's National Center for Environmental Economics contributed data and analysis to
this document.
in
-------
GLOSSARY OF TERMS AND ABBREVIATIONS
The following
CAA or Act
CAIR
CAMx
CBI
CEM
CSAPR
EGU
FIP
FR
EPA
GHG
GW
ICR
IPM
km
Ib/mmBtu
LNB
NAAQS
NBP
NEI
NOx
NSPS
OFA
PM2.5
ppb
RIA
SCR
SIP
are abbreviations of terms used in this Regulatory Impact Analysis.
Clean Air Act
Clean Air Interstate Rule
Comprehensive Air Quality Model with Extensions
Confidential Business Information
Continuous Emissions Monitoring
Cross-State Air Pollution Rule
Electric Generating Unit
Federal Implementation Plan
Federal Register
U.S. Environmental Protection Agency
Greenhouse Gas
Gigawatts
Information Collection Request
Integrated Planning Model
Kilometer
Pounds per Million British Thermal Unit
Low-NOx Burners
National Ambient Air Quality Standard
NOx Budget Trading Program
National Emission Inventory
Nitrogen Oxides
New Source Performance Standard
Overfire Air
Fine Particulate Matter
Parts Per Billion
Regulatory Impact Analysis
Selective Catalytic Reduction
State Implementation Plan
IV
-------
SMOKE Sparse Matrix Operator Kernel Emissions
SNCR Selective Non-catalytic Reduction
SO2 Sulfur Dioxide
TIP Tribal Implementation Plan
TPY Tons Per Year
TSD Technical Support Document
-------
-------
-------
TABLE OF CONTENTS
page
LIST OF TABLES ii
LIST OF FIGURES ii
EXECUTIVE SUMMARY ES-1
Overview ES-1
ES.l Identifying Required Emissions Reductions ES-1
ES.2 Baseline and Analysis Year ES-4
ES.3 Control Strategies and Emissions Reductions ES-6
ES.4 Costs ES-7
ES.5 Benefits to Human Health and Welfare ES-8
ES.5.1 Human Health Benefits and Climate Co-benefits ES-9
ES.5.2 Combined Health Benefits and Climate Co-Benefits Estimates ES-11
ES.5.3 Unquantified Co-Benefits ES-12
ES.5 Results of Benefit-Cost Analysis ES-15
E.S.6 References ES-16
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 1-2
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 Proposed Rule 1-5
1.2.3 Regulated Entities 1-6
1.2.4 Baseline and Analysis Year 1-6
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
CHAPTER2: ELECTRIC POWER SECTOR PROFILE 2-1
Overview 2-1
2.1 Background 2-1
2.2 Power Sector Overview 2-2
2.2.1 Generation 2-2
2.2.2 Transmission 2-10
2.2.3 Distribution 2-11
2.3 Sales, Expenses, and Prices 2-12
2.3.1 Electricity Prices 2-12
IX
-------
2.3.2 Prices of Fossil Fuels Used for Generating Electricity 2-17
2.3.3 Changes in Electricity Intensity of the U.S. Economy from 2000 to 2013 ..2-18
2.4 Deregulation and Restructuring 2-19
CHAPTERS: EMISSIONS AND AIR QUALITY MODELING IMP ACTS 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 Control Case Emissions 3-9
3.3 Post-Processing of Air Quality Modeling for Benefits Calculations 3-10
3.3.1 Converting CAMx Ozone Outputs to Benefits Inputs 3-10
3.3.2 Converting CAMx PM2.5 Outputs to Benefits Inputs 3-11
3.4 References 3-13
CHAPTER4: REGULATORY CONTROL Alternatives 4-1
Introduction 4-1
4.1 Background 4-2
4.2 Regulatory Control Alternatives Considered 4-3
4.3 Rationale for Regulatory Control Alternatives Chosen 4-6
CHAPTER 5: COST, ECONOMIC, ENERGY, AND EMPLOYMENT IMPACTS 5-1
Overview 5-1
5.1 Power Sector Modeling Framework 5-1
5.2 EPA's Power Sector Modeling Base Case for the Proposal to Update CSAPR 5-4
5.2.1 EPA'sIPMBaseCasev.5.15 5-5
5.2.2 The IPM Base Case Used to Analyze the Proposal to Update CSAPR 5-6
5.3 Evaluating the Regulatory Control Alternatives 5-7
5.3.1 Emission Reduction Assessment 5-10
5.3.2 Compliance Cost Assessment 5-12
5.3.3 Impacts on Fuel Use, Prices and Generation Mix 5-17
5.3.4 Effect of Emission Reductions on Downwind Receptors 5-20
5.4 Employment Impacts 5-22
5.4.1 Economic Theory and Employment 5-23
5.4.1.2 Current State of Knowledge Based on the Peer-Reviewed Literature .. 5-26
5.4.1.3 Regulated Sector 5-26
5.4.1.4 Economy-Wide 5-27
5.4.1.5 Labor Supply Impacts 5-27
5.4.1.6 Conclusion 5-28
5.4.2 Recent Employment Trends 5-28
IX
-------
5.4.2.1 Electric Power Generation 5-28
5.4.2.2 Fossil Fuel Extraction 5-29
5.4.3 Power and Fuels Sector Direct Employment Impacts 5-31
5.5 Social Costs 5-36
5.6 Secondary Economic Impacts 5-37
5.6.1 Methods 5-37
5.6.2 Summary of Secondary Market Impacts of Energy Price Changes 5-39
5.6.3 Share of Total Production Costs 5-40
5.6.4 Ability to Substitute between Inputs to the Production Process 5-40
5.6.5 Availability of Substitute Goods and Services 5-41
5.6.6 Effect of Changes in Input Demand from Electricity Sector 5-41
5.6.7 Conclusions 5-41
5.7 References 5-42
Chapter 6: Estimated Human Health Benefits and Climate Co-benefits 6-1
6.1 Introduction 6-1
6.2 Estimated Human Health Benefits 6-2
6.2.1 Health Impact Assessment for Ozone andPM2.s 6-4
6.2.1.1 Mortality Effect Coefficients for Short-term Ozone Exposure 6-7
6.2.1.2 PM2.5 Mortality Effect Coefficients for Adults and Infants 6-9
6.2.2 Economic Valuation for Health Benefits 6-14
6.2.3 Benefit-per-ton Estimates for Ozone 6-16
6.2.4 Benefit-per-ton Estimates for PM2.5 6-16
6.2.5 Estimated Health Benefits Results 6-17
6.3 Estimated Climate Co-Benefits from CO2 6-26
6.3.1 Climate Change Impacts 6-28
6.4 Combined Health Benefits and Climate Co-Benefits Estimates 6-35
6.5 Unqualified Co-benefits 6-37
6.5.1 HAP Impacts 6-39
6.5.2 Additional NO2 Health Co-Benefits 6-39
6.5.3 Additional SO2 Health Co-Benefits 6-40
6.5.4 Additional NO2 and SO2 Welfare Co-Benefits 6-41
6.5.5 Ozone Welfare Co-Benefits 6-42
6.5.6 Visibility Impairment Co-Benefits 6-42
6.6 References 6-43
6A.1 Overview of Benefit-per-Ton Estimates 6A-1
6A.2 Air Quality Modeling for the Proposed Transport Rule 6A-2
6A.3 National PM2.5 Benefit-per-Ton Estimates for EGUs Derived from Air
Quality Modeling of the Proposed Transport Rule 6A-3
6A.4 Regional Ozone Benefit-per-Ton Estimates 6A-6
6A.5 References 6A-7
CHAPTER 7: STATUTORY AND EXECUTIVE ORDER REVIEWS 7-1
Overview 7-1
7.1 Executive Order 12866: Regulatory Planning and Review 7-1
IX
-------
7.2 Paperwork Reduction Act 7-1
7.3 Regulatory Flexibility Act 7-4
7.3.1 Identification of Small Entities 7-5
7.3.2 Overview of Analysis and Results 7-9
7.3.2.1 Methodology for Estimating Impacts of the Proposed CSAPR update
on Small Entities 7-9
7.3.2.2 Results 7-11
7.3.3 Summary of Small Entity Impacts 7-14
7.4 Unfunded Mandates Reform Act 7-15
7.4.1 Identification of Government-Owned Entities 7-16
7.4.2 Overview of Analysis and Results 7-17
7.4.2.1 Methodology for Estimating Impacts of the proposed CSAPR update
on Government Entities 7-17
7.4.2.2 Results 7-19
7.4.3 Summary of Government Entity Impacts 7-22
7.5 Executive Order 13132: Federalism 7-22
7.6 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments 7-23
7.7 Executive Order 13045: Protection of Children from Environmental Health &
Safety Risks 7-24
7.8 Executive Order 13211: Actions that Significantly Affect Energy Supply,
Distribution, or Use 7-24
7.9 National Technology Transfer and Advancement Act 7-25
7.10 Executive Order 12898: Federal Actions to Address Environmental Justice
in Minority Populations and Low-Income Populations 7-25
CHAPTER 8: COMPARISON OF BENEFITS AND COSTS 8-1
Overview 8-1
8.1 Results 8-1
8.2 Net Present Value of a Stream of Costs and Benefits... .. 8-3
IX
-------
LIST OF TABLES
Table ES-1. Projected 2017* EGU Emissions Reductions of NOx, SO2, and CO2 with the
Proposed NOxx Emissions Budgets and More and Less Stringent Alternatives
(Tons)" ES-7
Table ES-2 Cost estimates (2011$) for proposal and more and less stringent alternatives .ES-8
Table ES-3. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the Proposal for 2017* ES-10
Table ES-4. Combined Health Benefits and Climate Co-Benefits for the Proposed
CSAPR EGU NOx Ozone Season Emissions Budgets and for More and Less
Stringent Alternatives for 2017 (millions of 2011$)* ES-12
Table ES-5. Unquantified Health and Welfare Co-benefits Categories ES-13
Table ES-6. Total Costs, Total Monetized Benefits, and Net Benefits of the Proposal in
2017 for U.S. (millions of 2011 $)a'b'c ES-16
Table 2-1. Existing National Electricity Generating Capacity by Energy Source, 2000
and 2013 2-3
Table 2-2. Net Generation in 2000 and 2013 (Trillion kWh = TWh) 2-6
Table 2-3. Coal and Natural Gas Generating Units, by Size, Age, Capacity, and
Thermal Efficiency in 2013 (Heat Rate) 2-7
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2000 and 2013 (billion
kWh) 2-12
Table 3-1 2011 Base Year and 2017 Baseline NOx and VOC Emissions by Sector
(thousand tons) 3-9
Table 3-2 Illustrative Control Case NOX Emissions and Changes from the 2017
Baseline 3-Error! Bookmark not defined.
Table 4-1 Ozone-Season NOx Emissions Budgets (Tons) for Proposed, More Stringent
and Less Stringent Regulatory Control Alternatives in 2017 and Later 4-5
Table 5-1 NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives 5-10
Table 5-2 EGU Ozone Season NOx Emission Reductions (tons) for the Proposal and
More and Less Stringent Alternatives 5-11
Table 5-3 EGU Annual Emission Reductions (tons) for SO2 and CO2 for the Proposal
and More and Less Stringent Alternatives 5-12
xn
-------
Table 5-4 Cost Estimates (millions of 2011$) for Proposal and More and Less
Stringent Alternatives 5-14
Table 5-5 Percent Changes* in Coal and Natural Gas Usage by EGUs for the Proposal
and More and Less Stringent Alternatives for 2017 5-18
Table 5-6 Percent Changes* in Generation by Major Type for the Proposal and More
and Less Stringent Alternatives for 2017 5-18
Table 5-7 Changes* in Generating Capacity by Major Type for the Proposal and More
and Less Stringent Alternatives in 2017 5-18
Table 5-8 Percent Changes* in Retail Electricity Price for the Proposal and More and
Less Stringent Alternatives for 2017 & 2020 5-20
Table 5-9. Estimated Percentage Changes in Average Energy Prices by Energy Type
for the Proposed Alternative* 5-39
Table 6-1. Human Health Effects of Ambient Ozone andPM2.s 6-5
Table 6-2. Summary of Ozone and PM2.5 Benefit-per-Ton Estimates Based on Air
Quality Modeling from The Illustrative Control Case in 2017 (2011 $)* 6-18
Table 6-3. Emission Reductions of Criteria Pollutants for the Proposal and More and
Less Stringent Alternatives in 2017 (thousands of short tons)* 6-18
Table 6-4. Summary of Estimated Monetized Health Benefits for the Proposal and
More and Less Stringent Alternatives Regulatory Control Alternatives for 2017
(millions of 2011$) * 6-18
Table 6-5. Summary of Avoided Health Incidences from Ozone-Related and PM2.5-
Related Benefits for the Proposal and More and Less Stringent Alternatives for
2017* 19
Table 6-8. Climate Effects 6-27
Table 6-9. Social Cost of CO2, 2015-2050 (in 2011$ per metric ton)* 6-34
Table 6-10. Estimated Global Climate Co-benefits of CO2 Reductions for the Proposal
and More and Less Stringent Alternatives for 2017 (millions of 2011$)* 6-34
Table 6-11. Combined Health Benefits and Climate Co-Benefits for the Proposal and
More and Less Stringent Alternatives for 2017 (millions of 2011$)* 6-37
Table 6-12. Unquantified Health and Welfare Co-benefits Categories 6-38
Table 7-1. SB A Size Standards by NAICS Code 7-8
Table 7-2. Projected Impact of the Proposed CSAPR Update on Small Entities in 2017 7-12
xn
-------
Table 7-3. Summary of Distribution of Economic Impacts of the Proposed CSAPR Update
on Small Entities in 2017 7-13
Table 7-4. Incremental Annual Costs under the Proposed CSAPR Update Summarized by
Ownership Group and Cost Category in 2017 (2011$ millions) 7-14
Table 7-5. Summary of Potential Impacts on Government Entities under the Proposed
CSAPR Update in 2017 7-20
Table 7-6. Incremental Annual Costs under the Proposed CSAPR Update Summarized by
Ownership Group and Cost Category (2011$ millions) in 2017 7-22
Table 8-1. Total Costs, Total Monetized Benefits, and Net Benefits of the Proposal and
More or Less Stringent Alternatives for 2017 for U.S. (millions of 2011$)a'b'c 8-2
Table 8-2. Projected 2017* Reductions in Emissions of NOx, SO2, and CO2 with the
proposed NOx Emissions Budgets and More or Less Stringent Alternatives (Tons) 8-3
xn
-------
LIST OF FIGURES
Figure ES-1. Nonattainment and Maintenance Receptors Identified for this Proposal and
Upwind States Linked to these Downwind Air Quality Problems with Respect to the
2008 Ozone NAAQS ES-3
Figure 2-1 National New Build and Retired Capacity (MW) by Fuel Type, 2000-2013.... 2-5
Figure 2-2 Regional Differences in Generating Capacity (MW), 2013 2-6
Figure 2-3. Cumulative Distribution in 2013 of Coal and Natural Gas Electricity
Capacity and Generation, by Age 2-9
Figure 2-4. Fossil Fuel-Fired Electricity Generating Facilities, by Size 2-10
Figure 2-5. Real National Average Electricity Prices for Three Major End-Use
Categories 2-14
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for Major
End-Use Categories, With Inflation Indices 2-15
Figure 2-7. Real National Average Electricity Prices (2011$) for Three Major End-Use
Categories 2-16
Figure 2-8. Relative Change in Real National Average Electricity Prices (2011$) for
Three Major End-Use Categories 2-17
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in
National Average Real Price per MBtu Delivered to EGU 2-17
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since
2000 18
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation
Intensity Since 2000 2-19
Figure 2-12. Status of State Electricity Industry Restructuring Activities 2-21
Figures 2-13 & 2-14. Capacity and Generation Mix by Ownership Type, 2002 & 2012 2-22
Figure 3-1. National air quality modeling domain 3-3
Figure 5-1 Time series of annual costs for the Proposal and More and Less Stringent
Alternatives 5-15
Figure 5-2 Time series of annual costs and annualized costs for the Proposal and More
and Less Stringent Alternatives 5-17
Figure 5-3 Electric Power Industry Employment 5-29
xiv
-------
Figure 5-4 Coal Production Employment 5-30
Figure 5-5 Oil and Gas Extraction Employment 5-31
Figure 6-1. Monetized Health Benefits of Proposal for 2017 6-20
Figure 6-2. Percentage of Adult Population (age 30+) by Annual Mean PM2.5 Exposure
in the Illustrative Modeling (used to generate the benefit-per-ton estimates) 6-26
Figure 6-3. Cumulative Distribution of Adult Population (age 30+) by Annual Mean
PM2.5 Exposure in the Illustrative Modeling (used to generate the benefit-per-ton
estimates) 6-26
xiv
-------
EXECUTIVE SUMMARY
Overview
The primary purpose of the proposed rule is to address interstate air quality impacts with
respect to the 2008 ozone National Ambient Air Quality Standards (NAAQS). The EPA
promulgated the Cross-State Air Pollution Rule (CSAPR) on July 6, 2011,1 to address interstate
transport of ozone pollution under the 1997 ozone NAAQS.2 The proposed rule would update
CSAPR to address interstate emissions transport of nitrogen oxides (NOx) that contribute
significantly to nonattainment or interfere with maintenance of the 2008 ozone NAAQS in
downwind states. The proposed rule also responds to the D.C. Circuit's July 28, 2015 remand of
certain CSAPR ozone season NOx emissions budgets to the EPA for reconsideration. This
Regulatory Impact Analysis (RIA) presents the health and welfare benefits and climate co-
benefits of the proposal to update CSAPR, and compares the benefits to the estimated costs of
implementing the proposed rule for the 2017 analysis year. This RIA also reports certain impacts
of the proposed rule such as its effect on employment and energy prices. This executive
summary both explains the analytic approach taken in the RIA and summarizes the RIA results.
ES.l Identifying Required Emissions Reductions
As described in the preamble for the proposal, CSAPR provides a 4-step process to
address the requirements of CAA section 110(a)(2)(D)(i)(I) (sometimes called the "good
neighbor provision") for ozone or fine particulate matter (PIVh.s) 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 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 by
1 See 76 FR 48208 (July 6, 2011)
2 CSAPR also addressed interstate transport of fine particulate matter (PM2 5) under the 1997 and 2006 PM2f
NAAQS.
ES-1
-------
quantifying available upwind emission reductions and apportioning upwind responsibility among
linked states; 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 emissions via regional emissions allowance trading programs. This action proposes to
apply this 4-step process to update CSAPR with respect to the 2008 ozone NAAQS. The
reductions required by the proposed rule would be achieved through a federal implementation
plan (FIP) for any state that does not have an approved state implementation plan (SIP)
addressing its contribution by the date this rule is finalized. Furthermore, under the FIPs, affected
EGUs would participate in the CSAPR NOx ozone-season allowance trading program. More
details on the methods and results of applying this process can be found in the preamble for this
proposal, and in Chapter 4 of this RIA.
Application of the first two steps of this process with respect to the 2008 ozone NAAQS
provides the analytic basis for proposing that ozone season emissions in 23 eastern states3 affect
the ability of downwind states to attain and maintain the 2008 ozone NAAQS. Figure ES-1
shows the affected states.
3 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia,
West Virginia, and Wisconsin.
ES-2
-------
I Geography for the proposal to update CSAPR for the 2008 ozone NAAQS (23 states)
Nonartainment receptor identified for this proposal
Maintenance receptor identified for this proposal
Figure ES-1. Nonattainment and Maintenance Receptors Identified for this Proposal and
Upwind States Linked to these Downwind Air Quality Problems with Respect to
the 2008 Ozone NAAQS
Applying Step 3 of this process, this action proposes to quantify electric generating unit
(EGU) NOx reductions in these 23 eastern states and to update CSAPR ozone season NOx
emissions budgets. A state's updated CSAPR ozone season NOx emissions budget is the quantity
of EGU NOx emissions that would remain after reducing significant contribution to
nonattainment and interference with maintenance of the 2008 ozone NAAQS in an average
year.4 These updated CSAPR NOx budgets were developed considering EGU NOx reductions
that are achievable for the 2017 ozone season.5 The EPA applied a multi-factor test in step 3 to
evaluate EGU NOx reduction potential for 2017 and is proposing to quantify EGU NOx ozone-
4 For example, assuming no abnormal variation in electricity supply due to events such as abnormal meteorology.
5 Non-EGU NOx emission control measures and reductions are not included in this proposal. More information on
non-EGU NOx control measures and the potential for emission reductions can be found in the Non-EGU TSD on
Emissions and Potential Reductions in 2017, which can be found in the docket for this proposal: EPA-HQ-OAR-
2015-0500.
ES-3
-------
season emissions budgets reflecting EGU NOx reduction strategies that are widely available at a
uniform annualized cost of $1,300 per ton (2011$). This assessment revealed that there is
significant EGU NOx reduction potential that can be achieved by 2017, which would make
meaningful and timely improvements in ozone air quality. Applying Step 4 of this process, the
EPA is proposing to implement these EGU NOx emissions budgets through the CSAPR NOx
ozone-season allowance trading program.
For the RIA, in order to implement the OMB Circular A-4 requirement to assess one less
stringent and one more stringent alternative to the proposal, the EPA is also analyzing EGU NOx
ozone season emissions budgets reflecting NOx reduction strategies that are widely available at a
uniform cost of $500 per ton (2011$) and strategies that are widely available at a uniform cost of
$3,400 per ton (2011$). The EPA applies these alternative uniform costs for quantifying EGU
NOx ozone-season emissions budgets in Step 3 of the 4-step process.
ES.2 Baseline and Analysis Year
The proposal sets forth the requirements for 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 year of 2017, taking into account
currently on-the-books Federal regulations, substantial Federal regulatory proposals,
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 proposal.
Below is a list of some of the national rules reflected in the baseline. 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).6 For a list of those regulations reflected in the compliance and cost modeling
6 U.S. Environmental Protection Agency, 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/2011v6/2011v6_2_2017_2025_EmisMod_TSD_aug2015.pdf).
ES-4
-------
of the electricity sector, please see "EPA Base Case v.5.15 Using IPM Incremental
Documentation" August, 2015.7
Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility
Generating Units (Final Rule) (U.S. EPA, 2015)
Standards of Performance for Greenhouse Gas Emissions from New, Modified, and
Reconstructed Stationary Sources: Electric Utility Generating Units (Final Rule) (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)
Cross State Air Pollution Rule (CSAPR) (U.S. EPA, 2011)8
Mercury and Air Toxics Standards (MATS) (U.S. EPA, 201 la)9
Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and
Heavy-Duty Engines and Vehicles (U.S. EPA, 201 lb)10
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, 201 Ob)
Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate Average Fuel
Economy Standards; Final Rule for Model-Year 2012-2016 (U.S. EPA, 2010c)
7 http://www.epa.gov/powersectormodeling/
8 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 II). The court largely
upheld CSAPR, but remanded to EPA without vacate certain states' phase 2 emissions budgets for reconsideration.
Compliance with these remanded emissions budgets was included in the baseline modeling for this proposal, which
was already underway when this opinion was issued.
9 On June 29, 2015, the United States Supreme Court reversed the D.C. Circuit opinion affirming the MATS. The
EPA is reviewing the decision and will determine any appropriate next steps once the review is complete. MATS is
included in the baseline for this analysis, and the EPA does not believe including MATS substantially alters the
results of this analysis, because MATS was remanded, not vacated.
10 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-5
-------
Hospital/Medical/Infectious Waste Incinerators: New Source Performance Standards and
Emission Guidelines: Final Rule 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)
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)
ES.3 Control Strategies and Emissions Reductions
The proposal requires EGUs in 23 eastern states to reduce interstate transport of NOx
emissions that significantly contribute to nonattainment or interfere with maintenance of the
2008 ozone NAAQS. The proposal sets EGU ozone season NOx emissions budgets (allowable
emission levels) for 2017 and future years. The proposal would implement these reductions
through FIPs in any state that does not have an approved good neighbor SIP by the date this rule
is finalized. Furthermore, under the FIPs, affected EGUs would 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 emissions reductions required by the proposed rule. The
allowance trading program essentially converts the EGU NOx emissions budget for each of the
23 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 emissions budget.
The EGU ozone season NOx budgets for each state reflect EGU NOx reduction strategies
that are widely available at a uniform cost of $1,300 per ton of NOx for affected EGUs.
Specifically, this uniform cost reflects turning on idled selective catalytic reduction (SCR),
selective non-catalytic reduction (SNCR) and upgrading combustion controls. Furthermore, this
RIA analyzes regulatory control alternatives based on more and less stringent state emissions
budgets based on uniform NOx control costs of $3,400 per ton and $500 per ton, respectively.
Table ES-1 show the emission reductions expected from the proposal and the more and
less stringent alternatives analyzed. Included in the table are annual and seasonal NOx, sulfur
dioxide (802), and carbon dioxide (CO2) reductions over the contiguous U.S.
ES-6
-------
Table ES-1. Projected 2017* ECU Emissions Reductions of NOx, SOi, and COi with the
Proposed NOx Emissions Budgets and More and Less Stringent Alternatives
(Tons)"
NOx (annual)
NOx (ozone season)
SO2 (annual)
CO2 (annual)
Proposal
90,000
85,000
1,200
660,000
More Stringent
Alternative
93,000
87,000
1,200
710,000
Less Stringent
Alternative
24,000
24,000
1,100
770,000
* The forecast of annual reductions of co-pollutants in 2017 is based on 2018 IPM direct model outputs.
" NOX and SO2 emissions are reported in English (short) tons; CO2 is reported in metric tons. All estimates rounded
to two significant figures.
ES.4 Costs
The EPA analyzed ozone-season NOx emissions reductions, as well as the associated
compliance costs of the power sector of implementing the EGU NOx ozone-season emissions
budgets in each of the 23 states, using the Integrated Planning Model (IPM) and its underlying
data and inputs. All EGU emissions reduction estimates are taken directly from IPM modeling
results. We use the compliance cost estimate from IPM as a proxy for social costs. As currently
configured, IPM directly estimates the costs for two of the NOx reduction strategies: turning on
idled existing SCRs and SNCRs, and shifting generation to lower-NOx emitting EGUs. The
costs of other mitigation measures, which are optimizing existing operating SCR or SNCR and
installing or upgrading NOx combustion controls, are estimated outside of IPM but are
calculated using IPM inputs.
One additional aspect of the analysis conducted using IPM is that the existing but idle
SCRs or SNCRs in the model baseline are required to be turned on and fully utilized in the
model for compliance with the emissions budgets. For the proposal, the EPA estimates that,
given the characteristics of these units and the average cost per ton of NOx reduced to comply
with the proposed emissions budgets, turning on the existing but idle SCRs is a reasonable and
cost-effective compliance method to evaluate for this RIA.
The estimate of the cost of this proposal, therefore, is the combination of NOX costs
estimated by IPM and additional costs estimated outside of IPM. The cost estimates for the
ES-7
-------
proposal and more and less stringent alternatives are presented in Table ES-2.11 All costs are in
2011 dollars.
Table ES-2. Cost Estimates (2011$) for Proposal and More and Less Stringent
Alternatives
Policy Annualized*
Proposal $93,000,000
More Stringent Alternative $96,000,000
Less Stringent Alternative $4,700,000
*Costs are annualized over the period 2016 through 2040 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.
ES.5 Benefits to Human Health and Welfare
Implementing this proposal to update CSAPR 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 also depends partly on local levels of
volatile organic compounds (VOCs). In addition, implementing the proposal would 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 PIVh.s throughout
the year and would reduce the incidence of PIVh.s-related health effects. Finally, these emissions
reductions would lower ozone and PIVh.s concentrations in regions beyond those subject to this
proposal, and the RIA accounts for those additional benefits.
Based on the IPM modeling results, the EPA does not expect this proposal to
significantly change annual power sector emissions of SO2, which is also a PM2.5 precursor.
Accordingly, this RIA does not quantify SO2-related PIVh.s co-benefits. 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.
11 It should be noted that the costs associated with implementation of monitoring, recordkeeping, and reports
requirements are not included within the estimates in this table. Such costs, which are estimated separately from the
approaches listed above and are found in Chapter 7, are actually negative on net but are very small compared to the
magnitude of the costs in Table ES-2.
ES-8
-------
In this section, we provide an overview of the monetized ozone benefits and
related co-benefits estimated from NOx reductions for compliance with the proposed CSAPR
EGU NOx ozone season emissions 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, 2015).
ES.5.1 Human Health Benefits and Climate Co-benefits
We follow 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. In addition to transferring information from other contexts to the
context of this regulation, we also use a "benefit-per-ton" approach to estimate the ozone and
PM2.5 benefits in this RIA. Benefit-per-ton approaches apply an average benefit-per-ton derived
from modeling of benefits of specific air quality scenarios to estimates of emissions reductions
for scenarios where no air quality modeling is available. More information on these approaches
is available in Chapter 6 of the RIA. 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 emissions reductions from other air quality modeling.
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.
The Health Impact Assessment (HIA) for ozone and PM2.5, discussed further in Chapter 6
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
ES-9
-------
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 (Abt Associates, 2012). 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 proposed CSAPR EGU NOX ozone-season emissions
budgets and for more and less stringent alternatives for 2017.
Table ES-3. Summary of Avoided Health Incidences from Ozone-Related and PMi.s-
Related Benefits from NOx reductions for the Proposal for 2017*
Ozone-related Health Effects
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided Premature Mortality
Smith et al. (2009) (all ages)
Zanobetti and Schwartz (2008) (all ages)
48
81
50
83
14
23
Avoided Morbidity
Hospital admissions respiratory causes (ages > 65)
Emergency room visits for asthma (all ages)
Asthma exacerbation (ages 6-18)
Minor restricted-activity days (ages 18-65)
School loss days (ages 5-17)
79
320
93,000
240,000
77,000
81
330
95,000
240,000
79,000
22
90
26,000
67,000
22,000
PM2.s-related Health Effects
Avoided Premature Mortality
Krewski et al. (2009) (adult)
Lepeule etal. (2012) (adult)
Woodruff et al. (1997) (infant)
21
48
<1
22
50
<1
6
13
<1
Avoided Morbidity
Emergency department visits for asthma (all ages)
Acute bronchitis (age 8-12)
Lower respiratory symptoms (age 7-14)
Upper respiratory symptoms (asthmatics age 9-1 1)
Minor restricted-activity days (age 18-65)
Lost work days (age 18-65)
Asthma exacerbation (age 6-18)
Hospital admissions respiratory (all ages)
Hospital admissions cardiovascular (age > 18)
Non-Fatal Heart Attacks (age >18)
Peters etal. (2001)
Pooled estimate of 4 studies
12
31
390
560
16,000
2,700
580
6
8
25
3
12
32
400
580
16,000
2,700
600
7
8
26
3
3
8
100
150
4,200
700
150
2
2
7
1
* All estimates are rounded to whole numbers with two significant figures. Co-benefits for PM2 5 are based on
ozone season NOx emissions. Confidence intervals are unavailable for this analysis because of the incidence-per-ton
methodology. 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
etal. (2012).
ES-10
-------
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 jig/m3 for PIVh.s). These relative risks can be used to develop risk coefficients that
relate a unit reduction in ozone to changes in the incidence of a health effect. We refer the reader
to Chapter 6 of this RIA, as well as 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.
Co-benefits of the proposed rule come from reducing emissions of CO2. Chapter 6 of this
RIA provides a brief overview of the 2009 Endangerment Finding and climate science
assessments released since then. Chapter 6 also provides information regarding the economic
valuation of CO2 using the social cost of carbon (SC-CCh), 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 rates5 percent, 3 percent, 2.5
percentand the 95th percentile SC-CO2 at 3 percent as recommended in the current SC-CO2
TSD12; see Chapter 6 of this RIA for more details). In this analysis we were unable to monetize
the co-benefits associated with reducing exposure to SO2, and NO2, as well as ecosystem effects
12 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.
ES-11
-------
and visibility impairment. In addition, there are expected to be unquantified health and welfare
impacts associated with changes in hydrogen chloride.
Table ES-4 provides the combined health and climate benefits for the proposal 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.
Table ES-4. Combined Health Benefits and Climate Co-Benefits for the Proposed CSAPR
EGU NOx Ozone Season Emissions Budgets and for More and Less Stringent
Alternatives for 2017 (millions of 2011$)*
SC-CCh Discount Rate**
Health and Climate Benefits
(Discount Rate Applied to Health Co-Benefits)
3% 7%
Climate Co-
Benefits Only
Less Stringent Alternative
5%
3%
2.5%
3% (95th percentile)
$190 to $340
$2 10 to $360
$230 to $380
$260 to $4 10
$190 to $330
$210 to $350
$220 to $370
$260 to $400
$7.6
$27
$41
$78
Proposal
5%
3%
2.5%
3% (95th percentile)
$680 to $1200
$700 to $1200
$7 10 to $1300
$740 to $1300
$660 to $1200
$680 to $1200
$690 to $1200
$720 to $1200
$6.5
$23
$35
$66
More Stringent Alternative
5%
3%
2.5%
3% (95th percentile)
$700 to $1300
$720 to $1300
$730 to $1300
$760 to $1300
$680 to $1200
$700 to $1200
$7 10 to $1200
$740 to $1300
$6.5
$23
$35
$66
*A11 estimates are rounded to two significant figures. Co-benefits are based on benefit-per-ton estimates. Benefits
for ozone and PM2 5 co-benefits are based on ozone season NOx emissions. Ozone benefits and PM2 5 co-benefits
occur in the analysis year, so they are the same for all discount rates. 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)). Climate co-benefits are based on reductions in CO2 emissions. The monetized health co-
benefits do not include changes in health effects from direct exposure to NO2, SO2, ecosystem effects; or visibility
impairment.
**The SC-CO2 estimates are calculated with four different values of a one ton reduction. See RIA Chapter 6 for a
complete discussion.
ES. 5.3 Unquantified Co-Benefits
The monetized co-benefits estimated in this RIA reflect a subset of 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-
ES-12
-------
benefits from several important benefit categories, including co-benefits associated with reduced
exposure to 862 and NO2, as well as ecosystem effects, and visibility impairment from reduced
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 with changes in exposures to 862 and
NO2. These co-benefits are listed in Table ES-5 below, and discussed more fully in Chapter 6 of
this RIA.
Table ES-5.
Category
Unquantified Health
Specific Effect
and Welfare Co-benefits Categories
Effect Has Effect Has
Been Been More Information
Quantified Monetized
Improved
Human Health
Reduced
incidence of
morbidity from
exposure to
NO2
Reduced
incidence of
morbidity from
exposure to SO2
Asthma hospital admissions
(all ages)
Chronic lung disease hospital
admissions (age > 65)
Respiratory emergency
department visits (all ages)
Asthma exacerbation
(asthmatics age 4-18)
Acute respiratory symptoms
(age 7-14)
Premature mortality
Other respiratory effects
(e.g., airway
hyperresponsiveness and
inflammation, lung function,
other ages and populations)
Respiratory hospital
admissions (age > 65)
Asthma emergency
department visits (all ages)
Asthma exacerbation
(asthmatics age 4-12)
Acute respiratory symptoms
(age 7-14)
Premature mortality
Other respiratory effects
(e.g., airway
hyperresponsiveness and
inflammation, lung function,
other ages and populations)
NO2 ISA1
NO2 ISA1
N02 ISA1
N02 ISA1
N02 ISA1
N02 ISA1'2-3
NO2 ISA2-3
S02 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA1-2'3
S02 ISA1-2
ES-13
-------
Category
Specific Effect
Effect Has Effect Has
Been Been
Quantified Monetized
More Information
Improved
Environment
Reduced
visibility
impairment
Reduced effects
on materials
Reduced effects
from PM
deposition
(metals and
organic s)
Reduced
vegetation and
ecosystem
effects from
exposure to
ozone
Reduced effects
from acid
deposition
Reduced effects
from nutrient
enrichment
Visibility in Class 1 areas
Visibility in residential areas
Household soiling
Materials damage (e.g.,
corrosion, increased wear)
Effects on Individual
organisms and ecosystems
Visible foliar injury on
vegetation
Reduced vegetation growth
and reproduction
Yield and quality of
commercial forest products
and crops
Damage to urban ornamental
plants
Carbon sequestration in
terrestrial ecosystems
Recreational demand
associated with forest
aesthetics
Other non-use effects
Ecosystem functions (e.g.,
water cycling,
biogeochemical cycles, net
primary productivity, leaf-
gas exchange, community
composition)
Recreational fishing
Tree mortality and decline
Commercial fishing and
forestry effects
Recreational demand in
terrestrial and aquatic
ecosystems
Other non-use effects
Ecosystem functions (e.g.,
biogeochemical cycles)
Species composition and
biodiversity in terrestrial and
estuarine ecosystems
Coastal eutrophication
Recreational demand in
terrestrial and estuarine
ecosystems
Other non-use effects
Ecosystem functions (e.g.,
biogeochemical cycles, fire
regulation)
PMISA1
PMISA1
PMISA1'2
PM ISA2
PMISA2
Ozone ISA1
Ozone ISA1
Ozone ISA1
Ozone ISA2
Ozone ISA1
Ozone ISA2
Ozone ISA2
Ozone ISA2
NOxSOxISA1
NOxSOxISA2
NOx SOX ISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOx SOX ISA2
ES-14
-------
Effect Has Effect Has
Category Specific Effect Been Been More Information
Quantified Monetized
Reduced Injury to vegetation from
, , ~~
vegetation SO2 exposure
effects from
ambient Injury to vegetation from
, ~~ -I.T^
exposure to SCh NOX exposure
and NOx
1We 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.
ES.5 Results of Benefit-Cost Analysis
Below in Table ES-6, we present the primary costs and benefits estimates for 2017. The
net benefits of the proposal and more and less stringent alternatives reflect the benefits of
implementing EGU NOx emissions reductions for the affected 23 states via the proposed FIPs
minus the costs of 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. Ideally, streams of social costs and social
benefits over time would be estimated and the net present values of each would be compared to
determine net benefits of the illustrative control strategy. The three 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, and (iii) adjusting the cost of illness for non-fatal heart attacks to
adjust for lags in follow up costs - are all appropriate. We explain our discounting of benefits in
Chapter 6 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 5. Our estimates of net benefits are the
approximations of the net value (in 2017) of benefits attributable to emissions reductions needed
to implement the NOx emissions budgets for each state.
ES-15
-------
Table ES-6. Total Costs, Total Monetized Benefits, and Net Benefits of the Proposal in
2017 for U.S. (millions of 2011$)a'b'c
Proposal More Stringent AAA
Climate Co-Benefits
Air Quality Health Benefits
Total Benefits
Annualized Compliance
Net Benefits
Non-Monetized Benefits'1
$23
$670 to $1200
$700 to $1200
$93
$600 to $1100
Non-quantified climate benefits
$23
$690 to $1300
$720 to $1300
$96
$620 to $1200
Less Stringent
$27
$190 to $340
$2 10 to $360
$4.7
$2 10 to $360
Reductions in exposure to ambient NC>2 and SC>2
Ecosystem benefits assoc. with reductions in emissions of NOx, SC>2, and PM
Visibility impairment
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 6 for additional detail and explanation. The costs presented in this table reflect compliance costs annualized
at a 4.77 percent discount rate and include monitoring, recordkeeping, and reporting costs. See Chapter 5 for
additional detail and explanation.
0 All costs and benefits are rounded to two significant figures; columns may not appear to add correctly.
d Non-monetized benefits descriptions are for all three alternatives and are qualitative.
E.S.6 References
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, 2014. Tier 3 Motor Vehicle Emission and Fuel Standards,
http://www3 .epa.gov/otaq/tier3 .htm.
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, 201 Ib, 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.
ES-16
-------
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, 201 Ob. 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; Final Rule 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: Final Rule Amendments,
http ://www3. epa.gov/airtoxics/129/hmiwi/rihmiwi .html.
U.S. EPA, 2008a. Emissions Standards for Locomotives and Marine 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. EPA, 2005, NOX Emission Standard for 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-17
-------
CHAPTER 1: INTRODUCTION AND BACKGROUND
Introduction
The EPA is proposing to update the Cross-State Air Pollution Rule (CSAPR) to reduce
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 proposal (hereafter referred
to as the "proposal to update CSAPR", or simply "the proposal"), is to address interstate air
quality problems with respect to the 2008 ozone NAAQS. The proposal also responds to the D.C.
Circuit's July 28, 2015 remand of certain CSAPR ozone-season NOx emission budgets to the
EPA for reconsideration. This Regulatory Impact Analysis (RIA) presents the health and welfare
benefits of the proposal to update CSAPR, and compares the benefits of the rule to the estimated
costs of implementing the rule in 2017. This chapter contains background information relative to
the rule and an outline of the chapters of this RIA.
1.1 Background
Clean Air Act (CAA or the Act) section 110(a)(2)(D)(i)(I), 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
secondary NAAQS.13 The EPA promulgated CSAPR on July 6, 201114 to address interstate
transport for the 1997 ozone NAAQS and the 1997 and 2006 fine particulate matter (PIVh.s)
NAAQS.15 (See section IV of the preamble to the proposal to update CSAPR for a discussion of
CSAPR litigation and implementation.)
As described in the preamble for the proposal, CSAPR provides a 4-step process to
address the requirements of the good neighbor provision for ozone or PIVh.s standards: (1)
13 The EPA uses the term "states" to include the District of Columbia in this RIA.
14 See 76 FR 48208 (July 6, 2011)
15 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.
1-1
-------
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 by quantifying upwind NOx
reductions and apportioning upwind responsibility (this step establishes emissions budgets,
which are remaining allowable emissions after reducing significant contribution to nonattainment
and interference 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. In this action, the EPA proposes to
apply this 4-step process to update CSAPR with respect to the 2008 ozone NAAQS. The
reductions required by the proposed rule would be achieved through a FIP in any state that does
not have an approved SIP addressing its contribution by the date this rule is finalized.
Furthermore, under the FIPs, affected EGUs would participate in the CSAPR NOx ozone-season
allowance trading program. More details on the methods and results of applying this process can
be found in the preamble for this proposal and in Chapter 4 of this RIA. 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).
1.2.1 Role of Executive Orders in the Regulatory Impact Analysis
Several statutes and executive orders apply to any public document. The analyses
required by these statutes and executive orders are presented in detail in Chapter 7, and 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 emissions reductions for
compliance with the proposal to update CSAPR. OMB Circular A-4 requires analysis of one
potential alternative standard level more stringent than the proposal and one less stringent than
the proposal. This RIA evaluates the benefits, costs, and certain impacts of a more and a less
stringent alternative to the proposal.
1-2
-------
7.2.2 Illustrative Nature of this Analysis
The EPA proposes to implement the proposed EGU NOx emissions budgets via updating
the CSAPR regional NOx ozone-season allowance trading program. This implementation
approach provides utilities with the flexibility to determine their own compliance path. This RIA
develops and analyzes one possible scenario for compliance with the NOx budgets proposed by
this action and possible scenarios for EGU compliance with more and less stringent alternatives.
1.2.3 The Need for 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.
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 emissions budgets in this proposal) is a straightforward 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.
1-3
-------
1.2 Overview and Design of the RIA
7.2.7 Methodology for Identifying Required Reductions
Application of the first two steps of the CSAPR process (described above) with respect to
the 2008 ozone NAAQS provides the analytic basis for proposing that ozone season emissions in
23 eastern states affect the ability of downwind states to attain and maintain the 2008 ozone
NAAQS. Figure 1-1 shows the affected states.
I Geography for the proposal to update CSAPR for the 2008 ozone NAAQS (23 states)
Nonattainment receptor identified for this proposal
O Maintenance receptor identified for this proposal
Figure 1-1. Nonattainment and Maintenance Receptors Identified for this Proposal and
Upwind States Linked to these Downwind Air Quality Problems with Respect to
the 2008 Ozone NAAQS
Applying Step 3 of this process, this action proposes to quantify electric generating unit
(EGU) NOx reductions in these 23 eastern states and to update CSAPR ozone season NOx
emissions budgets. A state's CSAPR ozone season NOx emissions budget is the quantity of EGU
NOx emissions that would remain after reducing significant contribution to nonattainment and
1-4
-------
interference with maintenance of the 2008 ozone NAAQS in an average year.16 These NOx
budgets were developed considering EGU NOx reductions that are achievable for the 2017 ozone
season.17 The EPA applied a multi-factor test to evaluate EGU NOx reduction potential for 2017
and proposes to quantify EGU NOx ozone-season emissions budgets reflecting EGU NOx
reduction strategies that are widely available at a uniform cost of $1,300 per ton (2011$). This
assessment revealed that there is significant EGU NOx reduction potential that can be achieved
by 2017, which would make meaningful and timely improvements in ozone air quality. Applying
Step 4 of this process, the EPA is proposing to implement these EGU NOx emissions budgets
through the CSAPR NOx ozone-season allowance trading program.
For the RIA, in order to implement the OMB Circular A-4 requirement to assess one less
stringent and one more stringent alternative to the proposal, the EPA is also analyzing EGU NOx
ozone season emissions budgets reflecting NOx reduction strategies that are widely available at a
uniform cost of $500 per ton (2011$) and strategies that are widely available at a uniform cost of
$3,400 per ton (2011$).
7.2.2 States Covered by the Proposed Rule
For each state that would be affected by one of the proposed federal implementation
plans (FIPs)18 (as shown in Figure ES-1), and that is already included in the CSAPR NOx ozone-
season trading program to address interstate ozone transport for the 1997 NAAQS, the proposed
rule would lower EGU NOx ozone-season emissions budgets to reduce ozone transport for the
2008 ozone NAAQS. One state, Kansas, would have a new CSAPR ozone-season requirement
16 For example, assuming no abnormal variation in electricity supply due to events such as abnormal meteorology.
17 Non-EGU NOx emission control measures and reductions are not included in this proposal. More information on
non-EGU control measures and the potential for emission reductions can be found in the TSD on the Assessment of
Non-EGU NOx Emission Controls, Cost of Controls, and Time for Compliance in 2017, which can be found in the
docket for this proposal.
18 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia,
West Virginia, and Wisconsin.
1-5
-------
under this proposal. The remaining 22 states were included in the original CSAPR ozone-season
program as to the 1997 ozone NAAQS.
1.2.3 Regulated Entities
The proposed rule affects fossil fuel-fired EGUs in these 23 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).
1.2.4 Baseline and Analysis Year
As described in the preamble, the EPA proposes to align implementation of this rule with
relevant attainment dates for the 2008 ozone NAAQS, as required by the D.C. Circuit's decision
North Carolina v. EPA19 The EPA's final 2008 Ozone NAAQS SIP Requirements Rule revised
the attainment deadline for ozone nonattainment areas currently designated as moderate from
December 2018 to July 2018 in accordance with the D.C. Circuit's decision in NRDC v. EPA20
Because July 2018 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 2018
attainment date. We believe that North Carolina compels the EPA to identify upwind reductions
and implementation programs to achieve these reductions, to the extent possible, for the 2017
ozone season.
The proposal to update CSAPR 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
proposals, 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 transport
19 531 F.3d 896, 911-12 (D.C. Cir. 2008) (holding that EPA must coordinate interstate transport compliance
deadlines with downwind attainment deadlines).
20 777 F.3d 456, 469 (D.C. Cir. 2014).
1-6
-------
rule. 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 this proposal that are reported in
this RIA are limited to these two analysis years because important regulatory actions, including
the Clean Power Plan (CPP) and potentially the 2015 ozone NAAQS, are expected to have an
important and uncertain influence on the electricity sector in later years, as explained below. For
this reason, the EPA expects that most of the proposed CSAPR update's influence on emissions
reductions will occur between 2017 and 2020.
The first year that EGUs must comply with the CPP is 2022 and investments in
renewable generation and energy efficiency in 2020 and 2021 are expected to influence
generation patters in the electricity sector in those years, The CPP is not anticipated to have
significant interactions with the CPP and the near-term (i.e., starting in 2017) ozone-season EGU
NOx emission reduction requirements under this proposal. States must submit a state plan, or an
initial submittal with an extension request, by September 6, 2016 describing how they will
implement the emissions guidelines of the final CPP. See section VILE of the preamble for
further discussion.
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, as discussed in the memo
to EPA Regional Administrators, Implementing the 2015 Ozone National Ambient Air Quality
Standards, implementation of the 2015 ozone NAAQS may use the framework of the CSAPR.
Doing so may influence compliance with the budgets that are the subject of this proposal.
In addition to other on-the-books Federal regulations, the baseline for the analysis of the
benefits, costs and certain impacts of this proposal includes CSAPR phase II NOx ozone-season
emissions budgets. As discussed in section III.C of the preamble, in EME Homer City II21, the
D.C. Circuit declared invalid the CSAPR phase 2 ozone-season NOx emissions budgets of 11
states. Because the proposed rule modeling was performed prior to the D.C. Circuit's issuance of
21 EME Homer City Generation, L.P., v. EPA, No. 795 F.3d 118, 129-30, 138 (EME Homer City II).
1-7
-------
EME Homer City II, the modeling assumes that the baseline for all states includes the emission
reductions associated with the CSAPR phase 2 ozone-season budgets. Furthermore, as these
budgets were remanded, it is reasonable to treat them, for the purposes of the analysis in this
RIA, as on-the-books Federal regulations.22 As discussed in the preamble, this proposal
incorporates revised emissions budgets that would supplant and replace the budgets promulgated
in the original CSAPR rule including the remanded budgets. Furthermore, these proposed
budgets would be effective for the 2017 ozone season, the same period in which the phase 2
budgets that were invalidated by the court are currently scheduled to become effective. Of the 11
states for which the CSAPR phase 2 ozone-season NOx emissions budgets were remanded, 9 are
among the 23 states for which budgets are established in this proposal. They are: Maryland, New
Jersey, New York, North Carolina, Ohio, Pennsylvania, Texas, Virginia, and West Virginia. The
baseline and analysis year are discussed in more detail in Chapters 3 and 5 of this RIA.
In addition, on June 29, 2015, the United States Supreme Court reversed the D.C. Circuit
opinion affirming the MATS. The EPA is reviewing the decision and will determine any
appropriate next steps once the review is complete. MATS is included in the baseline for this
analysis, and the EPA does not believe including MATS substantially alters the results of this
analysis, because MATS was remanded, not vacated.
7.2.5 Emissions Controls and Cost Analysis Approach
The EPA estimated the control strategies and compliance costs of the proposed rule using
the Integrated Planning Model (IPM) as well as certain costs that are estimated outside the
model, but use IPM inputs for their estimation. These cost estimates reflect costs incurred by the
power sector, and include (but are not limited to) the costs of purchasing, installing, and
operating NOx control technology, changes in fuel costs, and changes in the generation mix. A
description of the methodologies used to estimate the costs and economic impacts to the power
sector is contained in chapter 5 of this RIA.
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 (PIVh.s). We estimated these benefits using benefit-per-ton estimates derived
from the BenMAP tool. The EPA also estimated the climate co-benefits of the proposal. A
description of the methodologies used to estimate the human health and climate benefits is
contained in chapter 6 of this RIA. In addition, chapter 6 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: Regulatory Control Scenarios. The chapter summarizes the rationale for the
three regulatory control alternatives analyzed and how the requirements of these
alternatives are represented in IPM.
Chapter 5: Cost, Economic, Employment, and Energy Impacts. The chapter summarizes
the data sources and methodology used to estimate the costs and other impacts incurred
by the power sector.
Chapter 6: Benefits Analysis Results. The chapter quantifies the health-related benefits of
the ozone-related air quality improvements associated with the three regulatory control
alternatives analyzed.
Chapter 7: Statutory and Executive Order Impact Analyses. The chapter summarizes the
Statutory and Executive Order impact analyses.
Chapter 8: 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 proposal
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 2013. Projections of future power
sector behavior and the impact of this rule are discussed in more detail in chapters 3 and 5 of this
RIA.
2-1
-------
2.2 Power Sector Overview
The production and delivery of electricity to customers consists of three distinct
segments: generation, transmission, and distribution.
2.2.7 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 =
1000 MW) for multiple EGUs. Electricity Generation refers to the amount of electricity actually
produced by an EGU over some period of time, measured in kilowatt-hours (kWh) or gigawatt-
hours (GWh = 1 million kWh). Net Generation is the amount of electricity that is available to the
grid from the EGU (i.e., excluding the amount of electricity generated but used within the
generating station for operations). Electricity generation is most often reported as the total annual
generation (or some other period, such as seasonal). In addition to producing electricity for sale
to the grid, EGUs perform other services important to reliable electricity supply, such as
providing backup generating capacity in the event of unexpected changes in demand or
unexpected changes in the availability of other generators. Other important services provided by
generators include facilitating the regulation of the voltage of supplied generation.
Individual EGUs are not used to generate electricity 100 percent of the time. Individual
EGUs are periodically not needed to meet the regular daily and seasonal fluctuations of
electricity demand. Furthermore, EGUs relying on renewable resources such as wind, sunlight
and surface water to generate electricity are routinely constrained by the availability of adequate
wind, sunlight or water at different times of the day and season. Units are also unavailable during
routine and unanticipated outages for maintenance. These factors result in the mix of generating
capacity types available (e.g., the share of capacity of each type of EGU) being substantially
different than the mix of the share of total electricity produced by each type of EGU in a given
season or year.
Most of the existing capacity generates electricity by creating heat to create high pressure
steam that is released to rotate turbines which, in turn, create electricity. Natural gas combined
2-2
-------
cycle (NGCC) units have two generating components operating from a single source of heat. The
first cycle is a gas-fired turbine, which generates electricity directly from the heat of burning
natural gas. The second cycle reuses the waste heat from the first cycle to generate steam, which
is then used to generate electricity from a steam turbine. Other EGUs generate electricity by
using water or wind to rotate turbines, and a variety of other methods including direct
photovoltaic generation also make up a small, but growing, share of the overall electricity
supply. The generating capacity includes fossil-fuel-fired units, nuclear units, and hydroelectric
and other renewable sources (see Table 2-1). Table 2-1 also shows the comparison between the
generating capacity in 2000 and 2013.
In 2013 the power sector consisted of over 19,000 generating units with a total capacity23
of 1,168 GW, an increase of 301 GW (or 35 percent) from the capacity in 2000 (867 GW). The
301 GW increase consisted primarily of natural gas fired EGUs (246 GW) and wind generators
(58 GW), with substantially smaller net increases and decreases in other types of generating
units.
Table 2-1. Existing National Electricity Generating Capacity by Energy Source,
2000 and 2013
Energy Source
Coal
Natural Gas
Nuclear
Hydro
2000
Generator
Nameplate
Capacity
(MW)
336,247
242,602
104,734
95,879
% Total
Capacity
39%
28%
12%
11%
2013
Generator
Nameplate
Capacity
(MW)
329,815
488,169
104,424
100,182
% Total
Capacity
29%
42%
9%
8%
Change Between '00
Nameplate
Capacity
% Change
Increase (MW)
-2% -6,433
101% 245,567
0% -311
4% 4,303
and '13
%of
Total
Capacity
Increase
-2%
82%
0%
1%
23 As with all data presented in this section, this includes generating capacity not only at EGUs primarily operated to
supply electricity to the grid, but also generating capacity at commercial and industrial facilities that produce both
electricity used onsite as well as dispatched to the grid. Unless otherwise indicated, capacity data presented in this
RIA is installed nameplate capacity (also known as nominal capacity), defined by EIA as "The maximum rated
output of a generator, prime mover, or other electric power production equipment under specific conditions
designated by the manufacturer." Nameplate capacity is consistently reported to regulatory authorities with a
common definition, where alternate measures of capacity (e.g., net summer capacity and net winter capacity) can
use a variety of definitions and specified conditions. 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, and miscellaneous (< 1 percent)
2-3
-------
Petroleum
Wind
Other
Renewable
Misc
Total
68,080
2,394
14,047
3,079
867,062
8%
0.3%
1.6%
0.4%
100%
49,794
60,712
25,748
5,180
1,167,995
5%
5.1%
1.8%
0.4%
100%
-27%
2436%
83.3%
68.2%
35%
-18,287
58,318
11,701
2,101
300,933
-6%
19%
3.9%
0.7%
100%
Note: This table presents generation capacity. Actual net generation is presented in Table 2-2.
Source: U.S. EIA. Downloaded from EIA Electricity Data Browser, Electric Power Plants
Generating Capacity By energy source, by producer, by state back to 2000 (annual data from
EIA Form 860). Available online at:
Accessed 8/7/15.
The 35 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 2013, a total of 408,022 MW of new generating capacity
was built and brought online, and 83,240 MW existing units were retired. The net effect of the
re-rating of existing units reduced the total capacity by 23,848 MW. The overall net change in
capacity was 300,933 MW, as shown in Table 2-1.
The newly built generating capacity was primarily natural gas (307,764 MW), which was
partially offset by gas retirements (36,876 MW). Wind capacity was the second largest type of
new builds (58,477 MW), augmented by solar (6,273 MW).24 The overall mix of newly built and
retired capacity, along with the net effect, is shown on Figure 2-1.
1 Partially offset by 86 MW retired older wind or solar capacity.
2-4
-------
i Coal Nat Gas Wind & Solar mO\\& Other
Figure 2-1 National New Build and Retired Capacity (MW) by Fuel Type, 2000-201325
The information in Table 2-1 and Figure 2-1 present information about the generating
capacity in the entire U.S. The proposal to update CSAPR, however, directly affects EGUs in 23
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
Region and the rest of the U.S. (non-region). Figure 2-2 shows the mix of generating capacity for
each region. The overall capacity and generation in the CSAPR 2008 Ozone Region is 60% of
the national total, reflecting the larger total population in the region. The mix of capacity and
generation are noticeably different in the two regions. In the CSAPR 2008 Ozone Region, coal
makes up a significantly larger share of total capacity (34 percent) than it does in the rest of the
country (19%). The shares of natural gas, however, are quite similar (41% in the region, and 43%
in the rest of country). The difference in the share of coal's capacity is balanced by relatively
more hydro, wind, solar and other capacity in the rest of country compared to the CSAPR 2008
Ozone Region.
25 Source: EIA Form 860. Not visible: wind and solar retirements = 87 MW, net change in coal
capacity = -4,186 MW
2-5
-------
800,000
700,000
I Coal
In Region
I Gas Nuclear
I Hydro
Non-Region
Wind & Solar Other
Figure 2-2 Regional Differences in Generating Capacity (MW), 2013.
Source: EIA Form 860 Note: "Other" includes petroleum, geothermal, other renewable, waste materials and
misc.
In 2013, electric generating sources produced a net 4,058 trillion kWh to meet national
electricity demand, a 7 percent increase from 2000. As presented in Table 2-2, almost 70 percent
of electricity in 2013 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 2013 (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 18 percent and petroleum generation by 76 percent, while
natural gas generation increased by 83 percent. This reflects both the increase in natural gas
capacity during that period as well as an increase in the utilization of new and existing gas EGUs
during that period. Wind generation also grew from a very small portion of the overall total in
2000 to 4.1 percent of the 2013 total.
Table 2-2.
Net Generation in 2000 and 2013 (Trillion kWh = TWh)
Coal
Natural Gas
2000
Net
Generation
(TWh)
1,966.3
615.0
Fuel
Source
Share
52%
16%
2013
Net
Generation
(TWh)
1,586.0
1,125.9
Fuel
Source
Share
39%
28%
Change Between '00 and '13
Net
Generation
Change
(TWh)
-380.3
510.9
% Change in
Net
Generation
-19.3%
83.1%
2-6
-------
Nuclear
Hydro
Petroleum
Wind
Other Renewable
Misc
Total
753.9
270.0
111.2
5.6
75.3
4.8
3,802
20%
7%
2.9%
0.1%
2.0%
0.1%
100%
789.0
264.7
26.9
167.7
85.7
12.4
4,058
19%
7%
0.7%
4.1%
2.1%
0.3%
100%
35.1
-5.3
-84.4
162.1
10.4
7.6
256
4.7%
-2.0%
-75.8%
-
13.7%
157.7%
7%
Source: U.S. EIA Monthly Energy Review, December 2014. Table 7.2a Electricity Net
Generation: Total (All Sectors). Available online at:
. Accessed 12/19/2014
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 37 percent of the total number of coal-fired units, but only 6 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.
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 55 percent of the coal EGU fleet is over 500 MW per unit, 77 percent of the
gas fleet is between 50 and 500 MW per unit. Many of the largest gas units are gas-fired steam-
generating EGUs.
Table 2-3. Coal and Natural Gas Generating Units, by Size, Age, Capacity, and
Thermal Efficiency in 2013 (Heat Rate)
Unit Size
Grouping
(MW)
No.
Units
% of All
Units
Avg.
Age
Avg. Net
Summer
Capacity
(MW)
Total Net
Summer
Capacity
(MW)
% Total
Capacity
Avg. Heat
Rate
(Btu/kWh)
COAL
2-7
-------
0-24
25-49
50-99
100 - 149
150-249
250 - 499
500 - 749
750 - 999
1000 - 1500
Total Coal
223
108
157
128
181
205
187
57
11
1257
18%
9%
12%
10%
14%
16%
15%
5%
1%
100%
40.7
44.2
49.0
50.6
48.7
38.4
35.4
31.4
35.7
42.6
11.4
36.7
74.1
122.7
190.4
356.2
604.6
823.9
1259.1
250.7
2,538
3,963
11,627
15,710
34,454
73,030
113,056
46,963
13,850
315,191
1%
1%
4%
5%
11%
23%
36%
15%
4%
100%
11,733
11,990
11,883
10,971
10,620
10,502
10,231
9,942
9,732
11,013
NATURAL GAS
0-24
25-49
50-99
100 - 149
150-249
250 - 499
500 - 749
750 - 1000
Total Gas
1992
410
962
802
167
982
37
14
5366
37%
8%
18%
15%
3%
18%
1%
0.3%
100%
37.6
21.8
15.6
23.4
28.7
24.6
40.0
35.9
27.7
7.0
125.0
174.2
39.9
342.4
71.1
588.8
820.9
79.2
13,863
51,247
167,536
31,982
57,179
69,788
21,785
11,492
424,872
3%
12%
39%
8%
13%
16%
5%
3%
100%
13,531
9,690
8,489
11,765
9,311
12,083
11,569
10,478
11,652
Source: National Electric Energy Data System (NEEDS) v.5.14
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, 50 percent of the total coal generating capacity
has been in service for more than 38 years, while 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.
2-8
-------
9O%
SOS
"~
_^x^^
30 40
Age of EGU (years)
- Gas Cap ...... Gas Gen
Figure 2-3. Cumulative Distribution in 2013 of Coal and Natural Gas Electricity
Capacity and Generation, by Age
Source: National Electric Energy Data System (NEEDS) v.5.15
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.
2-9
-------
^Facility Capacity (MW)
010100
10010500
500101.000
1.000102,000
V
2.000 to 3.700
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 2016. 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,26 each operating synchronously. Within each of these
transmission networks, there are multiple areas where the operation of power plants is monitored
and controlled by regional organizations to ensure that electricity generation and load are kept in
26 These three network interconnections are the Western Interconnection, comprising the western parts of both the
US and Canada (approximately the area to the west of the Rocky Mountains), the Eastern Interconnection,
comprising the eastern parts of both the US and Canada (except those part of eastern Canada that are in the Quebec
Interconnection), and the Texas Interconnection (which encompasses the portion of the Texas electricity system
commonly known as the Electric Reliability Council of Texas (ERCOT)). See map of all NERC interconnections at
http://www.nerc.co m/AboutNERC/keyplayers/Documents/NERC_Interconnections_Color_072512.jpg
2-10
-------
balance. In some areas, the operation of the transmission system is under the control of a single
regional operator;27 in others, individual utilities28 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.
' E.g., PMJ Interconnection, LLC, Western Area Power Administration (which comprises 4 sub-regions).
! E.g., Los Angeles Department of Power and Water, Florida Power and Light.
2-11
-------
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 produced29 (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 2013.
Table 2-4. Total U.S. Electric Power Industry Retail Sales, 2000 and 2013 (billion kWh)
Sales
Residential
Commercial
Industrial
Transportation
Other
Total
Direct Use
Total End Use
2000
Sales/Direct
Use (Billion
kWh)
1,192
1,055
1,064
NA
109
3,421
171
3,592
Share of Total
End Use
33%
29%
30%
3%
95%
5%
100%
2013
Sales/Direct
Use (Billion
kWh)
1,395
1,344
978
8
NA
3,725
143
3,869
Share of Total End
Use
36%
35%
25%
0.2%
96%
4%
100%
Source: Table 2.2, EIA Electric Power Annual, 2013 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
29 Transportation (primarily urban and regional electrical trains) is a fourth ultimate customer category which
accounts less than one percent of electricity consumption.
2-12
-------
distributing electricity to individual homes and commercial establishments. The higher prices for
residential and commercial customers are the result both of the necessary extensive distribution
network reaching to virtually every part of the country and every building, and also the fact that
generating stations are increasingly located relatively far from population centers (which
increases transmission costs). Industrial customers generally pay the lowest average prices,
reflecting both their proximity to generating stations and the fact that industrial customers
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 2013, the national
average retail electricity price (all sectors) was 10.12 cents/KWh, with a range from 7.09 cents
(Washington) to 33.26 (Hawaii).
Average national retail electricity prices increased between 2000 and 2013 by 7.1 percent
in real terms (2011$). The amount of increase differed for the three major end use categories
(residential, commercial and industrial). National average residential prices increased the most
(8.7 percent), and commercial prices increased the least (2.3 percent). The real year prices for
2000 through 2013 are shown in Figure 2-5.
2-13
-------
14.0
2000
2002
Residential
2004 2006
^^^Commercial
2008 2010
^Industrial
2012
Total
Figure 2-5. Real National Average Electricity Prices for Three Major End-Use
Categories
Source: EIA AEO 2013, Table 2.4
Most of these electricity price increases occurred between 2002 and 2008; since 2008
nominal electricity prices have been relatively stable while overall inflation continued to
increase. The increase in nominal electricity prices for the major end use categories, as well as
increases in the GDP price and CPI-U indices for comparison, are shown in Figure 2-6.
2-14
-------
2000 2002 2004
Residential ^^^ Commercial
2006 2008
^Industrial «
2010
CPI-U
2012
GDP Price
Figure 2-6. Relative Increases in Nominal National Average Electricity Prices for Major
End-Use Categories, With Inflation Indices
For a longer term perspective, Figure 2-7 shows real30 (2011$) 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 23 percent
and 28 percent lower respectively than in 1960.
30 All prices in this section are estimated as real 2011 prices adjusted using the GDP implicit price deflator unless
otherwise indicated.
2-15
-------
Real Electricity Prices, 1960-2014 (including taxes)
I960 1970
^^^ Residential
1980 1990 2000
Commercial ^^^ Industrial
2010
Figure 2-7. Real National Average Electricity Prices (2011$) for Three Major End-Use
Categories
Source: EIA Monthly Energy Review, April 2015, Table 9.8
Relative Change in Electricity Prices,
1960-2014 (including taxes)
-50%
1960
1970
Residential
1980
Commercial
1990
2000
Industrial
2010
Total
2-16
-------
Figure 2-8. Relative Change in Real National Average Electricity Prices (2011$) for
Three Major End-Use Categories
Source: EIA Monthly Energy Review, April 2015, Table 9.8
2.3.2 Prices of Fossil Fuels Used for Generating Electricity
Another important factor in the changes in electricity prices are the changes in fuel
prices31 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 2013 had increased by 50.5 percent, while the real price of natural gas decreased by
22.6 percent. The real price of oil increased by 112 percent, but with oil declining as an EGU
fuel (in 2013 oil generated only 1 percent of electricity) the doubling of oil prices had little
overall impact in the electricity market. The combined real delivered price of all fossil fuels in
2013 increased by 36.9 percent over 2000 prices. Figure 2-9 shows the relative changes in real
price of all 3 fossil fuels between 2000 and 2013.
2000
2002
2004
2006
Coal
Oil
2008
Gas
2010
Average
2012
Figure 2-9. Relative Real Prices of Fossil Fuels for Electricity Generation; Change in
National Average Real Price per MBtu Delivered to EGU
Source: EIA AEO 2015, Table 9.9
31 Fuel prices in this section are all presented in terms of price per Btu to make the prices comparable.
2-17
-------
2.3.3 Changes in Electricity Intensity of the U.S. Economy from 2000 to 2013
An important aspect of the changes in electricity generation (i.e., electricity demand)
between 2000 and 2013 is that while total net generation increased by 6.7 percent over that
period, the demand growth for generation was lower than both the population growth (12.2
percent) and real GDP growth (25.4 percent). Figure 2-10 shows the growth of electricity
generation, population and real GDP during this period.
Real GDP
Population
Generation
Figure 2-10. Relative Growth of Electricity Generation, Population and Real GDP Since
2000
Sources: U.S. EIA Monthly Energy Review, December 2014. Table 7.2a Electricity Net
Generation: Total (All Sectors). U.S. Census.
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 2013. On a per capita basis, real GDP per
capita grew by 11.8 percent, increasing from $44,500 (in 2011$) per person in 2000 to
$49,800/person in 2013. At the same time electricity generation per capita decreased by 4.8
percent, declining from 13.4 MWh/person in 2000 to 12.8 MWh/person in 2013. 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 14.9 percent, declining from
303 MWh per $1 million of GDP to 258 MWh/$l million GDP. These relative changes are
2-18
-------
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.
15%
10%
5%
0%
-5%
-10%
-15%
2000
2002 2004
-Real GDP/Capita
2006 2008 2010 2012
Generation/Capita Generation/ Real GDP
Figure 2-11. Relative Change of Real GDP, Population and Electricity Generation
Intensity Since 2000
Sources: U.S. EIA Monthly Energy Review, December 2014. Table 7.2a Electricity Net
Generation: Total (All Sectors). U.S. Census.
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
2-19
-------
(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
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.
2-20
-------
Electricity Restructuring by State
Figure 2-12. Status of State Electricity Industry Restructuring Activities
Source: EIA 2010. "Status of Electricity Restructuring by State." Available online at:
.
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
2-21
-------
utilities owned 79 percent of U.S. generating capacity (MW) while IPPs32 and commercial and
industrial producers owned 18 percent and 3 percent 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 81 percent, 15 percent and 4 percent for utilities, IPPs and
commercial/industrial, respectively.
Since 2000, IPPs have expanded faster than traditional utilities, substantially increasing
their share by 2013 of both capacity (59 percent utility, 38 percent IPPs, and 3 percent
commercial/industrial) and generation (58 percent, 38 percent and 4 percent).
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
Figures 2-13 & 2-14. Capacity and Generation Mix by Ownership Type, 2002 & 2012
Capacity Mix, 2002 & 2012
I Nuclear BCoal Gas Hydro Wind All Other
Generation Mix, 2002 & 2012
3,000,000
2,500,000
^2,000,000
(D
01,500,000
'4-»
(D
51,000,000
(D
500,000
I Nuclear Coal Gas Hydro Wind All Other
32IPP data presented in this section include both combined and non-combined heat and power plants.
2-22
-------
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.
2-23
-------
2A-1
-------
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 control case. 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 NOx on both ozone and fine particulate matter (PIVh.s) concentrations.33'34 These factors
were then used to estimate the benefits of the regulatory control alternatives, as described in
Chapter 6. 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 proposed 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
33 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.
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 5 and 6. See the discussion in section 3.2.2 of this chapter.
3-1
-------
modeling system includes (1) emissions for a 2011 base year and emissions for the 2017 baseline
and illustrative control case, (2) meteorological data inputs for the year 2011, 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 PIVh.s component species
concentrations.35 The model predictions for the 2011 base year, the 2017 baseline, and the 2017
illustrative control case were combined with ambient air quality observations to calculate
seasonal mean ozone air quality metrics and annual mean PIVh.s for the 2017 baseline and 2017
illustrative control case, 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 illustrative control case).
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.
35 The CAMx output data files are titled Air Quality Modeling Data Drives for the 2008 NAAQS CSAPR Proposal
and have been placed in the rulemaking docket EPA-HQ-OAR-2015-0500.
3-2
-------
Figure 3-1. National air quality modeling domain.
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 2017 illustrative control case. 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 National Emissions Inventory,
version 2 (201 !NEIv2) for point sources, nonpoint sources, commercial marine vessels (CMV),
3-3
-------
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 !NEIv2, but were generated using the official release
2014 version of the Motor Vehicle Emissions Simulator (MOVES2014)
(http://www3.epa.gov/otaq/models/moves/), while the 201 !NEIv2 emissions were generated
using a slightly earlier version of 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
Federal Register notices issued 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 3.6.5 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 6.2, 2011 Emissions Modeling Platform
(EPA, 2015) and can be found in the docket for this proposed rule.
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
biogenic, fire, offshore oil platforms, and Canadian emissions use the same emissions in the base
3-4
-------
and future years.36 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 December,
2014. 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.
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 5.14 base case.37 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. The 2017 baseline EGU emissions include impacts from the Final Mercury and Air
Toxics Standards announced on December 21, 201138 and the CSAPR issued on July 6, 2011.39
The EPA notes that because the modeling for the proposal was performed prior to the D.C.
Circuit's issuance of EMEHomer City //,40 that modeling assumed in its baseline for all states
36 The biogenic and fire emissions are normally held constant between base and future years. The offshore and
Canadian emissions were held constant due to the lack of detailed information available to adequately project those
emissions to future years.
37 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 5 of this RIA. The documentation for version 5.14 can be found on
EPA"s power sector modeling website: http://www.epa.gov/airmarkets/powersectormodeling.html
38 In Michigan v. EPA, the Supreme Court reversed on narrow grounds a portion of the D.C. Circuit decision
upholding the Mercury and Air Toxics Standards, 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 the Mercury and Air Toxics Standards currently remain in
place.
39 We note that the NOX emissions changes that result from the Mercury and Air Toxics Standards and in the
baseline do not substantially alter the results of the analysis. The NOX reductions based on EPA emissions modeling
are only 7 percent of the nationwide total for EGUs in 2016, as found by subtraction of emissions in Table 5 A-l 1
from Table 5A-9 of the Mercury and Air Toxics Standards RIA.
40 EME Homer City Generation, L.P., v. EPA, No. 795 F.3d 118, 129-30, 138 (EME Homer City IT)
3-5
-------
the emission reductions associated with the CSAPRNOx ozone-season phase 2 emissions
budgets.
The Clean Power Plan (CPP) is not included in the modeled IPM 5.14 2017 baseline. The
EPA notes that after the 2011 base year, the 2017 baseline, and 2017 illustrative control case air
quality modeling for this proposal were underway, the EPA released an updated IPM base case,
version 5.15, and the final CPP. In order to reflect all on-the-books policies as well as the most
current power sector modeling data, the EPA performed an assessment of benefits, costs, and
impacts of this proposal using the IPM 5.15 base case and including the final CPP, as described
further in Chapter 5 of this RIA.
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 79 FR 2437, 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
(NESHAP) are included. Reductions due to the New Source Performance Standards (NSPS)
3-6
-------
volatile organic compound (VOC) controls for oil and gas sources, and the NSPS 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 (NFb) 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 79 FR 2437.
3-7
-------
The MOVES2014-based 2017 onroad emissions41 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
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 included most of 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. The California
emissions do not reflect the impacts of the GHG/Smartway regulation, nor do they reflect state
GHG regulations for the projection of other emissions sectors because that information was not
included in the provided inventories.
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 (EVIO) 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 EVIO global NOX
and 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
41 The 2017 onroad mobile emissions were derived from a 2018 MOVES model run that was adjusted to account for
2017. See the emissions modeling TSD (US EPA, 2015)for more details.
3-8
-------
(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.
All modeled 2011 and 2017 emissions cases use 2010 Canada emissions data, which is
the latest year for which Environment Canada had provided data at the time the modeling was
performed. No accompanying future-year projected baseline inventories were provided in a form
suitable for this analysis. 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 2011 and 2017 NOx and VOC
emissions by sector. Additional details on the emissions by state are given in the emissions
modeling TSD (US EPA, 2015)42.
Table 3-1 2011 Base Year and 2017 Baseline NOx and VOC Emissions by Sector (thousand
tons)
Sector
EGU-point
NonEGU-point
Point oil and gas
Wildfires and Prescribed Fires
Nonpoint oil and gas
Residential wood combustion
Other nonpoint
Nonroad
Onroad
C3 Commercial marine vessel
(CMV)
Locomotive and C1/C2 CMV
Biogenics
TOTAL
2011 NOx
2,000
1,200
500
330
650
34
760
1,600
5,700
130
1,100
1,000
15,000
2017 NOx
1,500
1,200
410
330
690
35
730
1,100
3,200
130
910
1,000
11,200
2011 VOC
36
800
160
4,700
2,600
440
3,700
2,000
2,700
5
48
41,000
58,000
2017 VOC
40
810
170
4,700
3,200
440
3,500
1,400
1,500
6
35
41,000
57,000
3.2.3 2017 Illustrative Control Case Emissions
The EPA's approach to developing IPM v5.14-based emissions for the illustrative control
case is methodologically consistent with the EPA's approach to establishing the proposed EGU
42 Available in the rulemaking docket: EPA-HQ-OAR-2015-0500.
3-9
-------
NOx ozone-season emissions budgets to reduce interstate ozone transport for the 2008 ozone
NAAQS. Specifically, the EPA performed IPM modeling that applied a uniform cost of $1,300
per ton on EGU NOx emissions. Next, the EPA established illustrative EGU NOx ozone-season
emissions budgets for the 23 eastern states included in this proposal by multiplying the resulting
state-specific emissions rate for affected EGUs by the corresponding (i.e., affected EGU) 2014
heat input. 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 v5.14 to create the illustrative control case.
The emissions for the illustrative control 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 control case inventories
preserved the patterns from the 2017 baseline. This was accomplished by generating ratios of
2017 base hourly emissions to the total seasonal emissions by unit and then applying these ratios
to the total seasonal emissions from the 2017 illustrative control case. Thus, the same hourly
temporal patterns in the baseline are reflected in this control case, including any adjustments
made to constrain the hourly 2017 emissions below the maximum levels during the 2011-2014
period.
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
3-10
-------
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)43 to calculate benefit per ton factors
for each metric. Information on the calculation of the benefit per ton factors is provided in
Chapter 6.
3.3.2 Converting CAMx PM2.5 Outputs to Benefits Inputs
The CAMx model44 generates predictions of hourly PM2.5 species concentrations for every
grid. 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
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
44 ENVIRON, 2014. User's Guide Comprehensive Air Quality Model with Extensions version 6.11,
www.camx.com. ENVIRON International Corporation, Novato, CA.
3-11
-------
particles. Future-year estimates of PIVh.s 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 PIVh.s species. The
modeled attainment test procedure for calculating future year PIVh.s 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 PIVh.s 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.s 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
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 6.
3-12
-------
3.4 References
Abt Associates, 2014. User's Guide: Modeled Attainment Test Software.
http://www3.epa.gov/scram001/modelingapps_mats.htm.
ERG, 2014. Develop Mexico Future Year Emissions Final Report. Available at
ftp://ftp.epa.gov/EmisInventory/20 Hv6/v2platform/2011emissions/Mexico_Emissions_WA
%204-09_fmal_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_O3-PM-
RH_Modeling_Guidance-2014.pdf)
U.S. Environmental Protection Agency, 2015. Preparation of Emissions Inventories for the
Version 6.2, 2011 Emissions Modeling Platform, Research Triangle Park, NC.
(http://www3 .epa.gov/ttn/chief/emch/2011 v6/2011 v6_2_2017_2025_EmisMod_TSD_aug20
15.pdf)
3-13
-------
3A-1
-------
CHAPTER 4: REGULATORY CONTROL ALTERNATIVES
Introduction
The primary purpose of this proposal is to address interstate air quality impacts with
respect to the 2008 ozone National Ambient Air Quality Standards (NAAQS). The EPA
promulgated the Cross-State Air Pollution Rule (CSAPR) on July 6, 2011,45 to address interstate
transport of ozone pollution under the 1997 ozone NAAQS.46 This action proposes to update
CSAPR to reduce interstate transport of electricity generating unit (EGU) ozone-season (May 1
through September 30) emissions of nitrogen oxides (NOx) that contribute significantly to
nonattainment or interfere with maintenance of the 2008 ozone NAAQS in downwind states. The
EPA proposes to implement the proposed EGU NOx reductions by setting emissions budgets
(limits on allowable emissions) that are implemented through the CSAPR NOx ozone-season
allowance trading program. As described in the preamble, the EPA is proposing transport
Federal Implementation Plans (FIPs) in this action. The EPA is proposing transport FIPs for 23
eastern states. The EPA would finalize a FIP for any state that does not have an approved State
Implementation Plan (SIP) addressing its good neighbor obligation by the date the rule is
finalized.
This Regulatory Impact Analysis (RIA) evaluates the benefits, costs and certain impacts
of three different illustrative regulatory control alternatives. The alternatives differ in the level of
the NOx emissions budget. One of the alternatives represents the proposed budgets which are
based on a uniform NOx control cost of $1,300 per ton (2011$), whereas the other two more and
less stringent alternatives represent budgets based on uniform NOx control costs of $3,400 per
ton and $500 per ton (2011$), respectively. The regulatory control alternatives also differ in the
size of the assurance limit, which equals 21 percent of the NOx ozone-season emissions budget,
for 2017 and following years for each state. The assurance provision limits the total ozone season
45 See 76 FR 48208 (July 6, 2011)
46 CSAPR also addressed interstate transport of fine paniculate matter (PM2s) under the 1997 and 2006 PM2s
NAAQS.
4-1
-------
NOx emissions from affected EGUs for each state as described below. Other key regulatory
features of the proposed allowance trading program, such as the ability to bank allowances for
future use, are the same across all three alternatives. This chapter describes the three alternatives
analyzed in the RIA.
4.1 Background
As described in detail in the preamble as well as in Chapter 1, the proposed rule requires
23 states in the eastern U.S. to reduce interstate emission transport that significantly contributes
to nonattainment, or interferes with maintenance, of the 2008 ozone NAAQS by reducing their
ozone season EGU NOx emissions in 2017 and future years. The rule proposes emissions
budgets, which are allowable emissions after reducing significant contribution to nonattainment
and interference with maintenance of the 2008 ozone NAAQS for the emissions of affected
EGUs in the 23 states. The reductions required by the proposed rule would be achieved through a
FIP in any state that does not have an approved SIP addressing its contribution by the date this
rule is finalized. Furthermore, under the FIPs, affected EGUs would 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 proposed rule.
The allowance trading program essentially converts the EGU NOx emissions budget for each of
the 23 states subject to the FIP into a limited number of NOx allowances that, on a tonnage basis,
equal the state's ozone season emissions budget. EGUs covered by the seasonal NOx allowance
trading program in the proposed FIPs are able to trade NOx ozone-season emission allowances
among EGUs within their state and across state boundaries, with emissions and the use of
allowances subject to certain limits. This RIA analyzes the benefits, costs, and certain impacts of
implementing the regulatory control alternative through the allowance trading program assuming
that all 23 states are subject to the proposed FIPs.
In accordance with Executive Orders 12866 and 13563, and the guidelines of OMB
Circular A-4, this RIA analyzes the benefits and costs associated with emissions controls to
comply with the proposal to update CSAPR. OMB Circular A-4 requires analysis of one
potential alternative standard level more stringent the proposed standard and one less stringent
than the revised standard. In response to this requirement, this RIA analyzes the proposed
remedy under the FIP as well as a more and a less stringent option (i.e., alternative) to the
4-2
-------
proposal under the FIP. The more and less stringent options differ from the proposed remedy in
that they set different EGU NOx ozone-season emissions budgets for the affected EGUs.
4.2 Regulatory Control Alternatives Considered
This RIA analyzes the benefits, costs and certain impacts of the proposed rule. In the
proposed rule, the EGU ozone-season NOx emissions budgets for each state are based on
applying a uniform cost of $1,300 per ton (2011$) to affected EGUs. This uniform cost of
reducing a ton of NOx emissions reflects the cost of NOx reduction technologies that are both
widely available and can be implemented in the near-term (i.e., by 2017). The budget setting
process is described in the preamble and in detail in the proposal's Ozone Transport Policy
Analysis Technical Support Document (TSD)47. Furthermore, this RIA analyzes regulatory
control alternatives with higher and with lower EGU emissions budgets. The EGU emissions
budgets in these more and less stringent regulatory control alternatives are based on uniform
NOx costs of $3,400 and $500 per ton (2011$), respectively. The rationale for choosing these
regulatory control alternatives for analysis in the RIA is described in Section 4.3.
As described in Chapter 5 of this RIA, the benefits, costs, and impacts of the proposed
rule are estimated, in part, by an economic model estimating how affected EGUs may comply
with the proposed rule. In addition to the limitation on ozone-season NOx emissions required by
the proposed EGU emissions budgets for the 23 states, there are four important features of the
allowance trading program that are 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 ozone-season NOx emissions allowances from one another for compliance purposes; the
ability of affected EGUs to bank ozone-season NOx 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 ozone-season NOx
allowances issued under 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.
Available in the rulemaking docket: EPA-HQ-OAR-2015-0500.
4-3
-------
Affected EGUs are expected to choose the least-cost method of complying with the
requirements of the allowance trading program. As described in Chapter 5, in the modeling of
EGU compliance, 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 are limited to the amount allowed by the sum of the NOx budgets
across the 23 states. Furthermore, as allowances may be banked for future use, if it is less costly
to abate ozone season NOx emissions in a current ozone season than to abate emissions in a later
ozone season, affected EGUs are expected to bank NOx ozone-season allowances in the current
ozone season for use in the later ozone season. Again, the ability of EGUs to bank allowances is
represented in the compliance modeling discussed in Chapter 5, and the number of banked
allowances is based on an outcome that minimizes the cost of complying with the NOx ozone-
season emissions budgets over time.
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 in the proposal, the
assurance provisions limit the amount of seasonal NOx emissions by affected EGUs in each of
the 23 states. The assurance level limits affected EGU emissions over an ozone season to the
state's NOx ozone-season emissions budget plus an increment equal to 21 percent of each state's
emissions budget. The increment is called the variability limit. See section VII.B.4 of the
preamble for a discussion of the purpose of the assurance provision and further detail about how
the variability and assurance limits are determined. If a state exceeds its assurance level in a
given year it is assessed a 2-to-l allowance surrender on the excess tons. Section VII.B.4 of the
preamble also explains how EPA then determines which EGUs are subject to this surrender
requirement. In the analysis in this RIA the assurance provisions are represented by a limit on the
total amount of ozone season NOx emissions that can be emitted by affected EGUs in each state.
That is, in the analysis, affected EGUs do not have the option of surrendering additional
allowances at the 2-to-l rate in the compliance modeling.
As described in section VII.B.5 of the preamble the rule proposes to allow 2015 and 2016
vintage NOx ozone-season allowances that were issued under CSAPR to address interstate ozone
transport for the 1997 ozone NAAQS to be used for compliance with this rule that reduces
interstate ozone transport for the 2008 ozone NAAQS. Specifically, 2015 and 2016 vintage
4-4
-------
CSAPR NOx ozone-season allowances may be used for compliance with this rule from 2017
forward.48 If an EGU in one of the 23 states affected by this proposed rule chooses to do so, it
must use four pre-2017 vintage allowances to cover one ton of NOx emitted. Based on EPA's
expectation of the size of the NOx allowance bank after the 2016 ozone season, the treatment of
these banked allowances is represented in the modeling as an additional 71,982 tons of NOx
allowances that may be used by affected EGUs during the 2017 ozone season or in later ozone
seasons.
Table 4.1 reports the seasonal budget for each of the 23 affected states. It also shows the
sum of all of the emissions budgets across all 23 states. As described above, in both the proposed
rule and the analysis in this RIA, emissions from affected EGUs in the entire region cannot
exceed this sum but for the ability to use banked allowances from previous years. Furthermore,
as described above, emissions from affected EGUs in a particular state may not exceed the
state's budget plus the assurance provision during the ozone season in any year (in 2017 and
later).
Table 4-1 Ozone-Season NOx Emissions Budgets (Tons) for Proposed, More Stringent and
Less Stringent Regulatory Control Alternatives in 2017 and Later
Alabama
Arkansas
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maryland
Michigan
Mississippi
Missouri
New Jersey
New York
North Carolina
Ohio
Oklahoma
Proposed Emissions
Budgets
Alternative
9,979
6,949
12,078
28,284
8,351
9,272
21,519
15,807
4,026
19,115
5,910
15,323
2,015
4,450
12,275
16,660
16,215
More Stringent
Control Alternative
9,931
6,101
11,992
27,585
8,118
9,259
20,945
15,378
4,026
18,624
5,487
15,240
2,011
4,391
10,705
16,637
16,215
Less Stringent Control
Alternative
11,886
7,038
12,144
33,483
8,614
9,278
32,783
15,861
4,026
22,022
6,083
15,380
2,016
4,607
12,278
20,194
16,215
48 Allowances were only issued under CSAPR for the 2015 and 2016 ozone seasons.
4-5
-------
Pennsylvania
Tennessee
Texas
Virginia
West Virginia
Wisconsin
TOTAL
Proposed Emissions
Budgets
Alternative
14,387
5,481
58,002
6,818
13,390
5,561
311,867
More Stringent
Control Alternative
14,358
5,449
55,864
5,834
12,367
5,511
302,028
Less Stringent Control
Alternative
38,270
5,520
58,492
6,955
22,932
5,588
371,665
These regulatory control alternatives are illustrative. For example, the EGUs have
flexibility in determining how they will comply with the allowance trading program. The way
that they comply may differ from the methods forecast in the modeling for this RIA.
4.3 Rationale for Regulatory Control Alternatives Chosen
As described in the preamble, the $1,300 per ton uniform cost used to set the proposed
EGU NOx ozone-season emissions budgets for the affected states is based on control
technologies that are both widely available and can be implemented in the near-term (i.e., by
2017). The uniform cost of $500 per ton was used to establish NOx ozone-season emissions
budgets for CSAPR and in this current rulemaking is the cost of fully optimizing post-
combustion controls that are already running. The uniform cost of $3,400/ton was used to
establish NOx ozone-season emissions budgets for the NOx SIP Call and reflects the cost of
turning on idled existing SNCRs. The proposal takes comment on budgets based on these two
uniform cost thresholds. Therefore evaluating the benefits and costs of complying with these
higher and lower budgets both responds to the requirement of Circular A-4 to evaluate a less and
a more stringent option (i.e., alternatives) and is informative to evaluating alternative control
stringencies on which the EPA is seeking comment. Achievable emission reductions from higher
uniform NOx cost levels were also evaluated for this proposal. Budgets are not being proposed
for this rule based on these higher uniform cost levels as they are based on the application of
NOx controls that could not be installed in the near term (i.e., by 2017). See section VI of the
preamble and the Ozone Transport Policy Analysis TSD for further explanation.
4-6
-------
4A-1
-------
CHAPTER 5: COST, ECONOMIC, ENERGY, AND EMPLOYMENT IMPACTS
Overview
This chapter reports the compliance cost, emissions, economic, energy and employment
impact analyses performed for the proposal to update CSAPR. The EPA used the Integrated
Planning Model (IPM), developed by ICF International, to conduct most of the analysis
discussed in this chapter. The EPA conducted additional analysis using the IPM estimates of
changes in the power and fuels sector as a basis for examining the potential direct labor impacts
in those sectors.
IPM is a dynamic linear programming model of the power sector that can be used to
examine air pollution control policies affecting NOx, CO2, 862, mercury (Hg), hydrochloric acid
(HC1), and other air pollutants throughout the contiguous United States for the entire power
system. Future-year electricity demand levels are based on projections from the Energy
Information Administration (EIA).
This chapter of the RIA presents analyses of the proposed rule that include assumptions
about the possible actions that EGUs may pursue as they reduce their NOx emissions to comply
with the proposed EGU NOx ozone-season emissions budgets in the 23-state region. Over the
past two decades, the EPA has used IPM and other analytical methods described in this chapter
to conduct extensive analyses of proposed and final federal environmental regulatory actions
affecting the power sector. These previous analytical efforts support the Agency's understanding
of key variables that influence the effects of a policy and provide the framework for how the
Agency estimates the costs and benefits associated with its actions.
5.1 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 proposed updates to CSAPR.
5-1
-------
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 emissions 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
modeling in IPM. The model documentation provides additional information on the assumptions
discussed here as well as all other model assumptions and inputs.49
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/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.50 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
49 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
50 See Chapter 10 of EPA's Base Case using IPM (v5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
5-2
-------
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.51
To estimate the annualized costs of additional capital investments 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.52 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 2011 Cross-State Air Pollution Rule (CSAPR) (U.S. EPA, 2005), 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). EPA has also used IPM to estimate the air pollution reductions and power sector impacts
of water and waste regulations affecting EGUs, including Cooling Water Intakes (316(b)) Rule
(U.S. EPA, 2014), Disposal of Coal Combustion Residuals from Electric Utilities (CCR) (U.S.
EPA, 2015b) and Steam Electric Effluent Limitation Guidelines (ELG) (U.S. EPA, 2015c).
The model and EPA's input assumptions undergo periodic formal peer review. The
rulemaking process also provides opportunity for expert review and comment by a variety of
stakeholders, including owners and operators of capacity in the electricity sector that is
51 See Chapter 9 of EPA's Base Case using IPM (v.5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
52 See Chapter 8 of EPA's Base Case using IPM (v5.15) documentation, available at:
http://www.epa.gov/airmarkets/powersectormodeling.html
5-3
-------
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
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 studies53 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.
5.2 EPA's Power Sector Modeling Base Case for the Proposal to Update CSAPR
The IPM "base case" for any regulatory impact analysis is a business-as-usual scenario
that would be expected under market and regulatory conditions in the absence of the rule. As
such, an IPM base case represents an element of the baseline for this RIA.54 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.
Our analysis of the proposal to update CSAPR involved two different IPM base cases. The
EPA used IPM version 5.14 (IPM v.5.14) to provide power sector emissions data for air quality
modeling. Specifically, IPM v.5.14 was used for the air quality modeling of a 2017 baseline and
2017 illustrative control case. This air quality modeling and IPM v.5.14 are detailed in Chapter
53
http://www2.epa.gov/clean-air-act-overview^eneflts-and-costs-clean-air-act
54 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).
5-4
-------
3.55 IPM v.5.14 was also used for the air quality modeling to quantify upwind state contributions,
described in Section V of the preamble for this proposal.
After these air quality modeling scenarios were underway, the EPA released an updated
IPM base case, version 5.15 (IPM v.5.15), and the final Clean Power Plan (CPP). The EPA used
IPM v.5.15 for developing the proposed state NOx emissions budgets discussed in Chapter 4 of
this RIA, and for analyzing the proposed rule's cost, benefits and impacts. The EPA relied on
IPM v.5.15 for these analyses so that the baseline for this RIA would reflect all on-the-books
policies, including the CPP, as well as the most current power sector modeling data. Using IPM
v.5.15 for these analyses provides EPA with the best information available to develop the
proposed rule and to provide the public with the most current information possible.
5.2.1 EPA 's IPM Base Case v. 5.15
When the EPA began using the current version of IPM (version 5), the EPA developed a
comprehensive updated base case (v.5.13), as well as a companion updated database of EGU
units (the National Electricity Energy Data System, or NEEDS v.5.13) that is used in EPA's
modeling applications of IPM. The EPA base case is updated periodically to reflect the annual
electricity demand forecasts from the EIA, among other data updates. EPA's IPM modeling
platform used to analyze this proposed rule (v.5.15) incorporates the version of the model used to
analyze the potential impacts of the CPP, which was finalized in August, 2015. As discussed
below in Section 5.2.2, the base case for the proposal to update CSAPR also includes certain
revisions to the v.5.15 base case used in the CPP analysis.
The updates to the base case between v.5.14 and v.5.15 include an update to the natural
gas supply as well as routine calibrations with the EIA's Annual Energy Outlook (AEO), such as
updating the electric demand forecast consistent with the AEO 2015.56 Additional updates, based
on the most up-to-date information and/or public comments received by the EPA, include unit-
55 As described in Chapter 3, the baseline used for the air quality modeling used to generate the benefit-per-ton
estimates used in Chapter 6, differs from the baseline used to estimate the benefits, costs and impacts of the
proposed rule and more and less stringent alternatives.
56 The naming conventions used for EPA's IPM modeling (such as v.5.15) reflect both the major overall version of
IPM (v.5) and which iteration of the AEO (currently AEO 2015) that supplies the electricity demand forecast used in
IPM.
5-5
-------
level specifications (e.g., pollution control configurations), planned power plant construction and
closures, and updated cost and performance for onshore wind and utility-scale solar technologies.
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 March 2015. This
update also includes two non-air federal rules affecting EGUs: Cooling Water Intakes (316(b))
Rule (U.S. EPA, 2014)..Combustion Residuals from Electric Utilities (CCR) (U.S. EPA,
2015b).. Additionally, all new capacity projected by the model is compliant with Clean Air Act
11 l(b) standards, including the final standards of performance for GHG emissions from new
sources57. For a detailed account of all updates made to the v.5.15 modeling platform, see the
Incremental Documentation for EPA Base Case v.5.15 Using IPM.58
EPA also updated the National Electric Energy Data System (NEEDS)59. This database
contains the unit-level data that is used to construct the "model" plants that represent existing and
committed units in EPA modeling applications of IPM. 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.
5.2.2 The IPM Base Case Used to Analyze the Proposal to Update CSAPR
As discussed above, IPM Base Case v.5.15 was used as the base case in EPA's analysis
of the final CPP, which was finalized in August, 2015. We are using a different IPM base case
for this analysis of the proposal to update CSAPR to reflect the fact that the Clean Power Plan is
now a finalized and promulgated federal regulation. The CPP is now a part of the business-as-
usual scenario that would be expected under market and existing regulatory conditions in the
absence of this proposal.
The base case used to analyze the proposal to update CSAPR is the IPM configuration
EPA used to analyze one of the final CPP illustrative policy options. Specifically, the base case
used in the current analysis is the IPM configuration for the CPP's "rate-based" illustrative plan
57 http://www2.epa.gov/cleanpowerplan/carbon-pollution-standards-new-modified-and-reconstructed-power-plants
58 Available at: http://www.epa.gov/airmarkets/powersectormodeling.html/
59 http://www2.epa.gov/airmarkets/power-sector-modeling-platform-v515
5-6
-------
approach. Using the CPP rate-based IPM configuration for the proposed CSAPR analysis assures
that the IPM results are the incremental impacts on the power sector of the proposal to update
CSAPR alone, and avoids "double counting" the impacts of the CPP as well as all other federal
and state regulations included the v.5.15 base case. Extensive information on the IPM
configuration and detailed results of the CPP rate-based analysis is available in the Clean Power
Plan Final Rule - Regulatory Impact Analysis60 for the CPP as well as at EPA's Power Sector
Modeling website.61
The 2017 baseline EGU emissions include impacts from the Final MATS announced on
December 21, 201162 and CSAPR issued on July 6, 2011. The EPA notes that because the
modeling for the proposal was performed prior to the D.C. Circuit's issuance of EME Homer
City II,63 that modeling assumed in its baseline for all states the emission reductions associated
with the CSAPR NOx ozone-season phase 2 emissions budgets.
5.3 Evaluating the Regulatory Control Alternatives
To estimate the costs, benefits, and economic and energy market impacts of the proposal to
update CSAPR, the EPA conducted quantitative analysis of three regulatory control alternatives:
the proposed EGU NOx ozone-season emissions budgets that reflect EGU NOx control strategies
represented by a uniform NOx cost of $1,300 per ton (2011$); and more and less stringent
alternative EGU NOx ozone-season emissions budgets that reflect EGU NOx control strategies
represented by uniform NOx costs of $3,400 per ton and $500 per ton (2011$), respectively.
Details about these regulatory control alternatives, including state-specific EGU NOx ozone-
season emissions budgets for each alternative, are provided in Chapter 4 of this RIA.
60 Available at: http://www2.epa.gov/cleanpowerplan/clean-power-plan-final-rule-regulatory-impact-analysis
61 Available at: http://www.epa.gov/airmarkets/powersectormodeling.html/
62 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.
63 EME Homer City Generation, L.P., v. EPA, No. 795 F.3d 118, 129-30, 138 (EME Homer City IT)
5-7
-------
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) assuming that this
rule is finalized in the summer of 2016. The EPA considered all widely used EGU NOx control
strategies: fully operating existing selective catalytic reduction (SCR) and selective non-catalytic
reduction (SNCR) - including optimizing NOx removal by existing, operational SCRs and
SNCRs as well as turning on and optimizing existing idled SCRs and SNCRs; installation of (or
upgrading to) state-of-the-art NOx combustion controls; shifting generation to units with lower
NOx emission rates within the same state; and installing new SCRs and SNCRs. 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, assuming this proposed rule is
finalized in the summer of 2016. The installation of new SCRs or SNCRs (the amount of time
from contract award through commissioning for retrofit) exceeds 18 and 12 months, respectively.
It would therefore not be feasible to retrofit new SCR or SNCR to achieve EGU NOx reductions
in the 2017 ozone season. For more details on these assessments, including the assessment of
EGU NOx mitigation costs and feasibility, please refer to the EGU NOx Mitigation Strategies
TSD, in the docket for this rule64.
The EPA notes that, due to limitations on model size, IPM v.5.15 does not have the
capacity to determine, within the model, whether or not to operate existing EGU post-
combustion NOx controls (i.e., SCR or SNCR) that are idle in the base case.65 In order to
evaluate compliance with the regulatory control alternatives (including the proposal and more
and less stringent alternatives), the EPA determined, outside the model, whether or not it would
be reasonably expected for these controls to operate in order to comply with each of the
evaluated regulatory control alternatives. After imposing the requirement to operate these control
systems, IPM then estimated the associated NOx reductions and costs.
64
Available at: EPA-HQ-OAR-2015-0500.
65 The EPA notes that EGUs with idled SCR or SNCR in the base represent a small percentage (less than 10 percent)
of the EGU fleet that is equipped with NOx post-combustion controls.
5-8
-------
The EGU NOx mitigation strategies that are assumed to operate or be available to reduce
NOx in order to comply with each of the regulatory control alternatives are shown in Table 5-1;
more information about the estimated costs of these controls can be found in the EGU NOx
Mitigation Strategies TSD.
5-9
-------
Table 5-1. NOx Mitigation Strategies Implemented for Compliance with the Regulatory
Control Alternatives
Regulatory Control , , T , , ,
. ,, , . NOx Controls Implemented
_ Alternative _ _ _
(1) Fully operating existing SCRs to achieve 0.075 Ib/MMBtu NOx emission
Less Stringent Alternative rate (outside IPM)
_ (2) Shift generation to minimize costs (within IPM) _
(All controls above)
Proposal (3) Turn on idled SCRs (within IPM)
_ (4) Install or upgrade combustion controls (outside IPM) _
More Stringent Alternative ... T (AU ^r * * >n
6 (5) Turn on idled SNCRs (outside IPM)
5.3.1 Emission Reduction Assessment
As discussed in Chapter 3 the EPA determined that NOx emissions in 23 eastern states
affect the ability of downwind states to attain and maintain the 2008 ozone NAAQS. For these
23 eastern states, the EPA is proposing to issue Federal Implementation Plans (FIPs) that
generally update the existing CSAPR NOx ozone-season emission budgets for EGUs and
implement these budgets via the CSAPR NOx ozone-season allowance trading program.
The EPA analyzed ozone-season NOx emission reductions from implementing each of the
three regulatory control alternatives. Specifically, the EPA conducted IPM modeling of each of
the regulatory control alternatives to evaluate the corresponding power sector emissions
reductions from complying with the EGU NOx ozone-season emissions budgets that are
provided in Chapter 4 of this RIA.
The NOx emissions reductions are presented for two time periods - 2017 (the principal
year of interest for the proposal to update CSAPR) and 2020. However, this version of IPM was
not structured to simulate 2017 directly. To evaluate the 2017 ozone season EGU NOx
reductions from compliance with the regulatory control alternatives, EPA developed estimates of
emissions for 2017 by adjusting IPM's direct estimates for 2018 to account for a number of
known differences. For example, these adjustments account for emissions from EGUs that are
expected to operate in 2017, but that have announced plans to retire by 2018 (and whose future-
year emissions are thus not otherwise represented in IPM's direct 2018 estimates). The EPA has
only made such adjustments for ozone season NOx emissions given the air quality objective of
5-10
-------
the proposal to address transported ozone pollution by the 2017 ozone season.66 For co-pollutant
EGU emissions (i.e., annual NOx, 862, and CCh), the EPA assumes that the 2018 IPM results
are representative of 2017 with no further adjustment. The EPA believes that this is reasonable
given that the overall magnitude of the adjustments to ozone season NOx emissions for 2017 is
relatively small and that the adjustments generally affect both the base case and the regulatory
control scenarios and therefore have an even smaller influence on the emissions difference
between the base case and the regulatory control scenarios.
Table 5-2 presents the reduction in EGU NOx emissions resulting from compliance with
the regulatory control alternatives (i.e., emissions budgets) in the 23-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 proposed
emissions budgets, and the more stringent alternative results in modestly more NOx reductions.
Table 5-2. EGU Ozone Season NOx Emission Reductions (tons) for the Proposal and More
and Less Stringent Alternatives
Ozone Season NOx
2017
2020
23 -State Region
Non-Region
Total
23 -State Region
Non-Region
Total
More Stringent Less Stringent
Proposal Alternative Alternative
85,000
-400
85,000
83,000
-300
83,000
88,000
-500
87,000
84,000
-460
84,000
24,000
-300
24,000
24,000
-260
24,000
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 co-pollutant emission
reductions are presented in Table 5-3.
66 The 2018 emissions output from IPM were adjusted to reflect 2017 emissions levels as described in
http://www.epa.gov/airmarkets/documents/ipm/Adjusted2017.pdf
5-11
-------
Table 5-3. EGU Annual Emission Reductions (tons) for SOi and COi for the Proposal and
More and Less Stringent Alternatives
Annual NOx
23 -State Region
2017 Non-Region
Total
23 -State Region
2020 Non-Region
Total
Proposal
91,000
-560
90,000
92,000
-1,000
91,000
More Stringent
Alternative
93,000
-700
93,000
94,000
-1,700
92,000
Less Stringent
Alternative
24,000
-500
24,000
24,000
-700
24,000
Annual SCh
23 -State Region
2017 Non-Region
Total
23 -State Region
2020 Non-Region
Total
2,000
-820
1,200
1,700
-1,100
610
2,160
-920
1,200
1,600
-1,500
100
1,800
-710
1,100
1,300
-910
360
Annual CCh
23 -State Region
2017 Non-Region
Total
23 -State Region
2020 Non-Region
Total
1,300,000
-670,000
660,000
1,700,000
-1,400,000
270,000
1,600,000
-920,000
710,000
2,200,000
-2,400,000
-250,000
1,300,000
-520,000
770,000
1,400,000
-1,100,000
360,000,000
5.3.2 Compliance Cost Assessment
This section describes EPA's approach to quantify the costs for compliance with the
regulatory control alternatives. IPM directly estimated the costs for three of the NOx mitigation
strategies: turning on idled SCRs, turning on idled SNCRs, and shifting generation to lower-NOx
emitting EGUs. However, the costs of increasing the use and optimizing the performance of
existing and operating SCRs and SNCRs, and for installing or upgrading NOx combustion
controls, were estimated using an incremental analysis step that EPA developed to quantify the
additional costs of these controls. These methods and analysis relies on data and methods used
within IPM, including NOx control cost equations used in IPM. Therefore, this analysis is
consistent with IPM and provides the best available quantification of the costs of these NOx
mitigation strategies.
The following steps illustrate this analytical method and demonstrate the application of this
method to the EPA's cost estimate of compliance with the proposal:
5-12
-------
Analytical Method
Application to Estimate the Costs of Compliance
with the Proposal
Identify the IPM modeled EGU
NOx reductions that are
attributable to NOx mitigation
strategies with costs that are not
included in the model.
Maximizing the use of existing SCRs accounts
for approximately 33,000 tons of ozone season
EGU NOx reductions.
Installing or upgrading NOx combustion controls
accounts for approximately 7,800 tons of ozone
season EGU NOx reductions.
Estimate the costs associated with
these EGU NOx mitigation
strategies.
Maximizing the use of existing SCRs costs
approximately $500 per ton.
Installing or upgrading NOx combustion controls
costs approximately $1,200 per ton.
Calculate the cost of each EGU
NOx mitigation strategies as
applied to reduce emissions within
the model.
The cost of maximizing the use of existing SCRs
is approximately $17,000,000 annually for 2017.
The cost of installing or upgrading NOx
combustion controls is approximately
$9,400,000 annually for 2017.
The total costs of compliance with the regulatory control alternatives are estimated as the
sum of the costs that are modeled within IPM and the costs that are calculated outside the model.
The results of EPA's IPM analysis show that, with respect to compliance with the
proposed EGU NOx emissions budgets, turning on idle existing SCRs provides the largest
amount of ozone season NOx emission reductions (51 percent), and maximizing the use of
existing SCRs produces an additional 39 percent of the total ozone season NOx reductions.
Combustion controls (9 percent) and generation shifting (1 percent) make up the remainder of
the ozone season NOx reductions. In the more stringent alternative, compliance by turning on
idle existing SNCRs makes up 3 percent of the total reductions, while the shares attributed to the
other four mitigation measures are similar to the shares for compliance with the proposed EGU
NOx emissions budgets.
The estimates of the changes in the cost of supplying electricity for the regulatory control
alternatives are presented in Table 5-4. The costs associated with compliance with monitoring,
recordkeeping, and reports requirements are not included within the estimates in this table and
can be found in Chapter 7.
5-13
-------
Table 5-4. Cost Estimates (millions of 2011$) for Proposal and More and Less Stringent
Alternatives
Less Stringent
Proposal More Stringent Alternative Alternative
Annualized* $92.9 $95.7 $4.7
Annual 2017 $71.6 $63.4 -$47.6
Annual 2020 $48.6 $20.3 -$27.6
*Levelized annual costs over the period 2016 through 2040, discounted using the 4.77 discount rate used in IPM's
objective function of minimizing the net present value (NPV) of the stream of total costs of electricity generation.
There are several notable aspects of the results presented on Table 5-4. The most notable
result in Table 5-4 is that the estimated annual compliance costs for the less stringent alternative
are negative (i.e., a cost reduction) in 2017 and 2020, although this regulatory control alternative
reduces NOx emissions by over 25,000 tons as shown in Table 5-3. The estimate of the cost of
controls determined outside the model are positive. Therefore, the finding of net negative annual
compliance costs derives from the portion of the annual compliance cost estimate that comes
from IPM. While seemingly counter-intuitive, such an estimated cost reduction is quite possible
given the dynamic, perfect foresight linear programing structure of IPM. 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. For example, IPM's perfect foresight structure
makes it possible that the least cost way of complying may be to delay a new investment
otherwise expected to occur in the base case. Such a delay could result in a lowering of annual
cost in an early time period and increases in later time periods.
A better understanding of the cause of the negative costs for the less stringent alternative
can come from considering the entire time path of costs, rather than focusing on the 2017 and
2020 estimates in isolation. Because IPM only reports estimates for certain years, it is necessary
to estimate the cost values for the years between the years directly simulated. A simple but
straightforward method is to use a linear interpolation between the annual cost estimates for
directly simulated years to estimate a potential annual compliance cost time series across all
years in a given period. Figure 5-1 shows the annual cost estimates, which are the sum of the
annual compliance cost estimated by IPM and the annual compliance costs estimated outside the
5-14
-------
model, for each directly simulated year from 2016 through 2040. It also shows the linearly
interpolated annual cost between those years for all three of the regulatory control alternatives.67
$200
$150
$100
o
u
$50
$0
-$50
2016 2020 2024 2028
^^Less Stringent ^^Proposal
2032 2036
-More Stringent
2040
Figure 5-1. Time Series of Annual Costs for the Proposal and More and Less Stringent
Alternatives
Figure 5-1 shows that while two of the IPM year cost estimates (2017 and 2020) in the less
stringent alternative are negative, the rest of the annual costs are positive, resulting in the
majority of the years between 2016 and 2040 having positive costs. All years have positive costs
under compliance with the proposed EGU NOx ozone-season emissions budgets and the more
stringent alternative.
Using the estimated annual cost time series, it is possible to use a standard NPV
calculation to estimate an annualized cost (i.e., levelized annual cost) of the annual cost stream
associated with compliance with each regulatory control alternative.68 For this analysis we first
67 EPA estimated year-specific outside-the-model cost adjustments only for 2017 and 2020. For this analysis of the
time series of costs, the outside-the-model compliance cost adjustments estimated for 2020 are added to all years
between 2021 and 2040. While this approach is a relatively rougher estimate of the likely cost of the exogenously-
imposed NOx controls for those years, it is consistent with an assumption that once the CSAPR update is
implemented it will continue to have similar costs and emission reductions throughout the rest of the model's
forecast horizon.
68 The XNPV() function in Microsoft Excel 2013 was used to calculate the NPV of the variable stream of costs.
5-15
-------
calculated the NPV of the stream of costs from 2016 through 2040. We calculate the NPV in
2016 of each cost stream using a 4.77 percent discount rate.69 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 approach we use a 4.77 percent rate to be
consistent with the optimization approach used within IPM. IPM's cost minimization objective
function estimates the lowest possible NPV, and uses a 4.77 percent discount rate to calculate the
NPV. IPM v.5.15 uses a 4.77 percent discount rate as an estimate of the real opportunity cost of
capital in the power sector.70
After calculating the NPV of the cost streams, the same 4.77 percent discount rate and
2016-2040 time period is used to calculate the levelized annual (i.e., annualized) cost estimates
shown in Table 5-3.71
Figure 5-2 depicts the annualized cost estimates on top of the time stream of annual costs
previously presented on Figure 5-1.
69 While IPM also reports the costs for 2050, we do not include the 2050 estimates in this analysis. In this situation,
extending the present value calculations to include 2050 does not materially change the results.
70 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." Section 8.2 in that documentation describes the rationale for using a real opportunity cost of capital
specific to the power sector, and describes the data used to derive the 4.77 percent rate.
71 The PMT() function in Microsoft Excel 2013 is used to calculate the level annualized cost from the estimated
NPV.
5-16
-------
$200
o $150
o $100
o
u
c
c
$50
$0
-$50
2016
2020
Less Stringent
Annualized
2024
2028
Proposal
Annualized
2032
2036
2040
More Stringent
Annualized
Figure 5-2. Time Series of Annual Costs and Annualized Costs for the Proposal and More
and Less Stringent Alternatives
As can be seen in Table 5-3 and Figure 5-2, the annualized costs are positive for
compliance with each of the three regulatory control alternatives. Furthermore, the annualized
costs have the expected relationship; the annualized costs are lowest for the less stringent
alternatives, and highest for the more stringent alternative.
5.3.3 Impacts on Fuel Use, Prices and Generation Mix
The proposal to update CSAPR is estimated to have a variety of different impacts to the
power sector. While all the impacts are relatively small in percentage terms, considering the
potential impacts in addition to the cost and emissions estimates presented previously is an
important component of assessing the overall impact of the proposal. In this section of the RIA
we discuss the estimated changes in fuel use, prices of fuel and retail electricity, generation by
fuel type, and capacity by fuel type.
Table 5-5 presents the percentage changes in national coal and natural gas usage by EGUs
in 2017. Percent changes in fuel prices are also shown. As will be seen for the other measures of
impacts on the power sector, the fuel use estimates in Table 5-5 estimates reflect a modest shift
to natural gas from coal.
5-17
-------
Table 5-5. Percent Changes* in Coal and Natural Gas Usage by EGUs for the Proposal and
More and Less Stringent Alternatives for 2017
Proposal More Stringent Alternative
Less Stringent
Alternative
Change in Fuel Quantity (TBtu)
Coal
Natural Gas
-0.077%
0.105%
-0.072%
0.090%
-0.091%
0.134%
Change in Fuel Prices
Coal**
Natural Gas***
-0.211%
0.002%
-0.279%
-0.004%
-0.321%
0.004%
*Fuel use changes measured in TBtu
**Coal price changes are minemouth prices per MMBtu
***Natural gas price changes are delivered gas prices per MMBtu
Table 5-6 presents the projected percentage changes in the amount of electricity generation
nationally in 2017 by 3 major types of electricity generation: coal, gas, and
conventional hydro. The amount generated by nuclear energy is unchanged in
all three alternatives.Table 5-6. Percent Changes* in Generation by Major Fuel
Type for the Proposal and More and Less Stringent Alternatives for 2017
Proposal More Stringent Alternative
Coal
Natural Gas
Total Generation**
-0.087%
0.094%
0.009%
-0.079%
0.101%
0.008%
Less Stringent
Alternative
-0.103%
0.127%
0.009%
Detail by Region for Proposal
Coal
Natural Gas
Total Generation
In-Region
-0.185%
0.167%
-0.009%
Non-Region
0.174%
-0.009%
0.040%
Total
-0.087%
0.094%
0.009%
*Changes in GWh generated in 2017.
**Changes in total generation is from all types of generation nationally, including fossil, non-fossil and renewable
EGUs dispatching electricity to the grid.
Table 5-7 presents the projected percentage changes in the amount of generating capacity
in 2017 for coal, natural gas and total capacity.
Table 5-7. Changes* in Generating Capacity by Major Fuel Type for the Proposal and
More and Less Stringent Alternatives in 2017
Proposal More Stringent Alternative
Less Stringent
Alternative
Percent Change
Coal
Natural Gas
Total Capacity
-0.144%
-0.006%
-0.006%
-0.136%
-0.006%
-0.004%
-0.162%
-0.006%
-0.015%
Capacity Change (MW)
Coal
Natural Gas
-282
-24
-318
-95
-266
-22
5-18
-------
Less Stringent
Proposal More Stringent Alternative Alternative
Total Capacity -60 -146
*Changes in generating capacity available in 2017.
The EPA estimated the change in the retail price of electricity (2011$) using the Retail
Price Model (RPM).72 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 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).73
Table 5-8 presents the projected percentage changes in the retail price of electricity for
the proposal and more and less stringent alternatives. Estimates are presented for both 2017 and
2020. While all of the estimated changes in prices are quite small, Table 5-8 includes both
positive and negative retail electricity price changes. This is consistent with the pattern of
positive and negative annual cost estimates discussed above in Section 5.3.3, and also reflects
some influence from aggregating results for the 22 NEMS regions into separate estimates for the
23 -state region affected by the proposal to update CSAPR, and the rest of the country.
72 See documentation available at: http://www.epa.gov/powersectormodeling/
73 See documentation available at:
http://www.eia. gov/forecasts/aeo/nems/documentation/electricity/pdf/m068(2014).pdf
5-19
-------
Table 5-8. Retail Electricity Price for the Proposal and More and Less Stringent
Alternatives for 2017 & 2020
Base Case
Price
(cents/kWh)
Percent
Proposal
Change in Retail
More Stringent
Alternative
Electricity Price
Less Stringent
Alternative
2017
National Price
23 State Region
Non-Region
9.93
9.69
10.75
-0.069%
-0.161%
-0.154%
-0.099%
-0.371%
-0.371%
-0.025%
-0.047%
-0.027%
2020
National Price
23 State Region
Non-Region
10.33
10.26
11.11
0.031%
0.027%
0.018%
0.034%
0.028%
0.022%
0.009%
0.022%
-0.001%
5.3.4 Effect of Emission Reductions on Downwind Receptors
As described in Sections V and VI of the preamble, and in the Ozone Transport Policy
Analysis Proposed Rule TSD, and summarized here, EPA evaluated the effect of the proposed
rule on nonattainment and maintenance receptors with respect to interstate transport for the 2008
ozone NAAQS. The 2008 ozone standard is 0.075 parts per million (ppm), 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 the Comprehensive Air Quality Model
with Extensions (CAMx) modeling system. From this CAMx analysis there are 37 receptors in
10 states evaluated for this proposed update to CSAPR.74 All 37 have a maximum design value
of 0.076 ppm or higher in 2017. Fourteen of these receptors have an average design value of
0.076 ppm or higher in 2017.75 Furthermore, as described in the Ozone Transport Policy
74 The CAMx air quality modeling scenario used to identify these receptors was also used as the baseline in Chapter
3 to estimate benefit-per-ton values for NOx reductions.
75 As described in the preamble, section V.C, the approach for projecting future ozone design values involved the
projection of an average of up to 3 design value periods, which include the years 2009-2013 (design values for
2009-2011, 2010-2012, and 2011-2013). The average of the 3 design values creates a "5-year weighted average"
value. The 5-year weighted average values and the individual design values for 2009-2011, 2010-2012, and 2011-
2013 were projected to 2017. The highest of the three individual values is the "maximum" design value. We are
proposing to identify nonattainment receptors in this rulemaking as those sites that are violating the NAAQS based
on current measured air quality and also have projected average design values of 76 parts per billion (ppb) or
greater. Maintenance-only receptors, i.e., those that are at risk of not maintaining the NAAQS, therefore include
both (1) those sites with projected average design values above the NAAQS that are currently measuring clean data
and (2) those sites with projected average design values below the level of the NAAQS, but with projected
maximum design values of 76 ppb or greater. In addition to the maintenance-only receptors, the 2017 ozone
5-20
-------
Analysis Proposed Rule TSD, the forecast ambient concentration at these receptors was adjusted
to reflect a new IPM base case (IPM v.5.15 with CPP), different from the one that was
represented in the emissions inventory for the CAMx modeling of the ambient concentrations
(IPM v.5.14). The results described in this subsection, and in the rest of this RIA except where
otherwise noted, are from a baseline that includes IPM version 5.15 with CPP. After this
adjustment to reflect the updated IPM base case, 33 of the 37 receptors have a maximum design
values of 0.076 ppm or higher in 2017, while 12 have average design values of 0.076 ppm or
higher in 2017.
To evaluate the effect of the proposed rule on ambient ozone concentrations at the
receptors in 2017 relative to the baseline, the EPA first evaluated the response of the electricity
sector to the proposed rule using IPM. The IPM analysis used to estimate the compliance cost,
benefits, and impacts of the proposed rule is the same analysis used to identify the effect of the
proposed rule on the level and spatial pattern of ozone seasonal NOx emissions from EGUs.
The emissions outputs from IPM are then used to estimate the change in ambient ozone
concentrations at the 37 receptors. The ozone Air Quality Assessment Tool (AQAT) is used to
estimate the impact of the upwind states' EGU NOx reductions on downwind ozone pollution
concentrations. AQAT is used to forecast both the average and maximum design values at all
monitors, including the 37 receptors. The ozone AQAT was developed specifically for use in the
proposed rule's significant contribution analysis AQAT uses CAMx outputs to calibrate its
predicted change in ozone concentrations to changes in NOx emissions. See the Ozone Transport
Policy Analysis Proposed Rule TSD for the air quality estimates and for details on the
construction of ozone AQAT.
In summary, the proposed rule is expected to reduce the average design values at the 37
receptors by 0.61 parts per billion (ppb) on average, and to reduce the maximum design values at
these receptors by 0.63 ppb on average. At the 12 receptors with an average design values of
0.076 ppm or greater in the baseline, the proposed rule reduces the difference between the
baseline average ambient concentration and 0.076 ppm by 14% on average (an average reduction
nonattainment receptors are also maintenance receptors because the maximum design values for each of these sites
is always greater than or equal to the average design value.
5-21
-------
of 0.24 ppb). The proposed rule reduces the maximum design values below 0.076 ppm for 7 of
the 33 receptors. Of the remaining 26 receptors with maximum design values of 0.076 ppm or
greater, the proposed rule reduces the difference between the baseline maximum design value
and 0.076 ppm by 21% on average (an average of reduction of 0.37 ppb). The proposed rule is
therefore expected to provide notable improvements in ambient air quality at these monitors by
2017. Results for each of the 37 receptors, as well as for the more and less stringent alternatives,
are described in the Ozone Transport Policy Analysis Proposed Rule TSD.
5.4 Employment Impacts
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,76 we typically conduct
employment analyses for economically significant rules. This section discusses and projects
potential employment impacts related to today's proposal.77
Section 5.5.1 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 5.5.2 presents an overview of the peer-reviewed
literature relevant to evaluating the effect of environmental regulation on employment. Section
5.5.3 provides background regarding recent employment trends in the electricity generation, coal
and natural gas extraction sectors. Section 5.5.4 discusses the potential direct employment
impacts in these sectors.
76 Labor expenses do, however, contribute toward total costs in the EPA's standard benefit-cost analyses.
77 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).
5-22
-------
5.4.1 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 that response. 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.78 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.79-80 Berman and Bui (2001)
adapt this model to analyze how environmental regulations affect labor demand.81 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)
model two components that drive changes in firm-level labor demand: output effects and
substitution effects.82 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 demand, and resulting in a
decrease in production. The output effect describes how, holding labor intensity constant, a
78 See Layard and Walters (1978), a standard microeconomic theory textbook, for a discussion, in Chapter 9.
79 See Hamermesh (1993), Ch. 2, for a derivation of the firm's labor demand function from cost-minimization.
80 In this framework, labor demand is a function of quantity of output and prices (of both outputs and inputs).
81 Morgenstern, Pizer, and Shih (2002) develop a similar model.
82 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.
5-23
-------
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
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.83
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
1 See Greenstone (2002) p. 1212.
5-24
-------
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.84 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.85
In addition to changes to labor demand in the regulated industry, net employment impacts
encompass changes in other related sectors, for example, the environmental protection sector.
This proposal 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.86 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).87 Some workers may retrain or
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
84 See Ehrenberg & Smith, p. 108.
85 This discussion draws from Herman and Bui (2001), pp. 293.
86 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.
87 Arrow etal. 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.
5-25
-------
EPA's Mercury Air Toxics Standards 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.88 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.
5.4.1.2 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.
5.4.1.3 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
! E.g. Graff Zivin and Neidell (2012).
5-26
-------
(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.
5.4.1.4 Economy-Wide
Given the difficulty noted above 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.89 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.
5.4.1.5 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.90 Although this literature is limited by empirical challenges, researchers
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
89 For further information see:
http://yosemite.epa.gOv/sab/sabproduct.nsf/0/07E67CF77B54734285257BB0004F87ED7OpenDocument
90 For a recent review see Graff-Zivin and Neidell (2013).
5-27
-------
billion (ppb) decreases in ozone concentrations increases worker productivity by 5.5 percent."
(Graff Zivin andNeidell, 2012, p. 3654).91
5.4.1.6 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.
5.4.2 Recent Employment Trends
The U.S. electricity system includes employees that support electric power generation,
transmission and distribution; the extraction of fossil fuels; renewable energy generation; and
supply-side and demand-side energy efficiency. This section describes recent employment trends
in the electricity system.
5.4.2.1 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 (U.S. BLS, 2014) accounted for the largest share of workers (25 percent).
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
operations (7 percent) and management (7 percent). As shown in Figure 5-3, employment in the
electric power industry averaged about 420,000 workers 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.
91 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.
5-28
-------
Power Generation and Supply Employment
(NAICS = 2211, Annual Average, 1000s of Employees)
500
450
400
350
300
250
200
150
100
50
0
&
^
Figure 5-3. Electric Power Industry Employment
5.4.2.2 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 5-4). During the
2000 to 2014 period, coal mining employment peaked in 2011 at about 87,000 employees.
5-29
-------
Coal Mining Employment
(NAICS = 2121, Annual Average, 1000s of Employees)
100
90
80
70
60
50
40
30
20
10
D
Figure 5-4. Coal Production Employment
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 5-5 shows there has been a sharp
increase in employment in this sector over the past decade.
5-30
-------
9Rn
?nn
1 en
<
1 nn
en
Oil and Gas Extraction Employment
(NAICS = 211, Annual Average, 1000s of Employees)
^^~*~~*
-~*^
f^t *^\* *&\ f&t *^o *«fo f^\ *&y f&l k^ »\* kOv k'H k fet
^vO *vO r& r&) /vO ^v^?"^ /vO «0 /vO ^^y^ ^^V r^y f£y r^y
Figure 5-5 Oil and Gas Extraction Employment
Source: BLS (2014b)
5.4.3 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 net generation may shift from
higher NOx-emitting EGUs to units with lower NOx emission rates. All of these changes will
likely involve some amount of change in 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
direct employment effects in the power sector that could occur because of actions taken by the
2017 ozone season include:
5-31
-------
Optimizing NOx removal from existing and operational SCR and SNCR 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
emission rates.
In addition, there could be directly induced employment impacts (both positive and
negative) in the labor demand in the fossil fuels industry supplying fuels to the power sector.
Once implemented, both the potential increases in operating efficiency and NOx reductions, as
well as and shifting generation to lower NOx emitting assets, 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 this proposed 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 proposal in 2017, the estimated impacts relevant to changes in labor
demand include:
The overall total national cost of generation in 2017 increases by 0.046 percent (the
levelized annualized total cost of generation increases by 0.01 percent);
Total net generation increases by 0.009 percent (coal generation decreases by .084 percent,
and natural gas generation increases by 0.11 percent);
The power sector's total tons of coal used for electricity generation decreases by 0.04
percent (or 0.09 percent increase in BTUs);
Total natural gas increases by 0.11 percent.
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
5-32
-------
have been quantified for the proposed regulation for 2017 and 2020. The labor impacts for the
more and less stringent alternatives have not been quantified.
Affected EGUs may respond the proposed requirement for EGUs in 23 eastern states to
reduce NOx emissions during the ozone season by improving and optimizing existing NOx
emission control systems or to shift generation to generating sources with lower NOx emission
rates. Meeting the new EGU ozone season NOx budget limits will involve changes in the amount
of labor needed in different parts of the utility power sector. Installing and operating new
equipment, optimizing 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, increases in operating efficiency 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 generation-side employment analysis uses the cost projections from the engineering-
based IPM to project labor demand impacts of the final emission guidelines on affected EGUs in
the electricity power sector and the fuel production sector (coal and natural gas). These
projections include effects attributable to 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. These projections are based in part on the IPM estimates of the impact of the proposed
regulation previously discussed in this RIA.
5.4.3.1 Methods Used to Estimate Changes in Employment in Electricity Generation and Fuel
Supply
5-33
-------
The analytical approach used in this analysis is a bottom-up engineering method
combining the EPA's cost analysis of the NOx emission limits 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. This approach is different from the economy-wide types of economic analyses
discussed in section 5.4.1. 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., 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 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.
The methods the EPA uses to estimate the labor impacts are based on the analytical
methods used in many previous EPA regulatory analyses. The most relevant prior analysis was
the RIAs for CPP (2015), and the CSAPR (2011). The methods used in this analysis to estimate
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 CPP
and CSAPR RIAs (with updated data where available).
The bottom-up engineering-based labor analysis in the CPP RIA was primarily concerned
with the labor needs of improving CO2 emissions from operating of existing EGUs nationwide,
and changes in the expected retirements of existing coal-fired EGUs and avoiding the need for
new gas-fired EGUs. A central feature of the labor analysis for this RIA, however, involves the
labor needs of upgrading and optimizing NOx control systems on existing EGUs in the affected
23-state region. In addition to the changes at EGUs within the 23-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.
5-34
-------
The methods and data used to estimate the labor associated with upgrading combustion
control systems are assumed to be the same as those used to estimate heat rate improvements
(HRI, another type of combustion control system upgrade) in the CPP RIA. There are two
components of this assumption.
The mix of labor categories needed to implement the combustion control upgrades
in this NOx analysis (i.e., the share of the labor cost of the upgrades apportioned
general construction, boilermaker, engineering and management labor) is the same
as the HRI-related combustion control upgrades needed in the CPP CO2 analysis.
The fully loaded labor cost of each labor category is the same for NOx control
upgrades as for the HRI CO2 upgrades
5.4.3.2 Estimates of the Changes in Employment in Electricity Generation and Fuel Supply
The estimated labor impacts of the proposed revisions to the NOx budgets from EGUs in
the 23-state region are presented in Table 5-9. Given the methods the EPA uses to estimate labor
impacts, it is not possible to directly separate the labor impacts that occur within the 23-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 23-state region. The fuel supply labor impacts, however,
will occur both within the 23-state region as well as 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 23-state region and in other states. Second, the shifts in fossil-fired generation will also occur
both with within the 23-state region and in other states.
5-35
-------
Table 5-9. Annual Net Employment Impacts for Power and Fuels Sectors of the Proposed
Option in 2017 & 2020
2017 2020
Upgrades and Optimization
SCR 70 66
SNCR* 0
Combustion Control 67 66
Sub-Total 137 132
Plant Retirement
Coal -160 -95
Fuel Use Change
Coal -87 -42
Natural Gas 50 19
Subtotal -36 -22
Net Employment Impact -60 15
*SNCR optimization is only required 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). All job-year
estimates are full-time equivalent (FTE) jobs.
5.5 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
society. The social costs of a regulatory action will not necessarily be equivalent to the
expenditures associated with compliance. Nonetheless, here we use compliance costs as a proxy
for social costs.
The cost estimates for the proposed and more or less stringent alternatives presented in this
chapter are the change in expenditures 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 described above necessary to comply with the emissions budgets that are described in
Chapter 4. The changes in electricity production costs estimated by IPM as described earlier in
Chapter 5, and the change in expenditures required by the power sector to maintain compliance
are also described earlier in Chapter 5.
5-36
-------
5.6 Secondary Economic Impacts
The energy sector impacts presented earlier in Chapter 5 of this RIA include potential
changes in the prices for electricity, natural gas, and coal potentially resulting from the proposal.
This chapter addresses the impact of these potential changes on other markets outside of the
power sector and discusses some of the determinants of the magnitude of these impacts. We refer
to these changes as secondary market impacts.
The analysis of costs for this proposal includes strategies for affected states to reduce NOx
emissions from EGUs to comply with the proposed EGU NOx ozone-season emissions budgets
in 2017. Ultimately, given the flexibilities afforded EGUs and states in complying with the
proposed rule, the benefits, cost and economic impacts reported in this RIA may differ given the
compliance approaches affected EGUs adopt. The abatement strategies adopted by affected
EGUs, will ultimately drive the magnitude and timing of secondary impacts from changes in the
price of electricity, and the demand for inputs by the electricity sector, on other markets that use
and produce these inputs.
To estimate the costs, benefits, and impacts of implementing the proposal, the EPA
modeled a compliance approach for the proposal and more and less stringent alternatives.
Chapter 4 provides a description of the proposal and the more and less stringent alternatives
considered. This chapter provides a quantitative assessment of the energy price impacts for these
approaches and a qualitative assessment of the factors that will in part determine the timing and
magnitude of effects in other markets.
5.6.1 Methods
One potential quantitative approach to evaluating the secondary market impacts is to use a
computable general equilibrium (CGE) model. CGE models are able to provide aggregated
representations of the whole economy in equilibrium in the baseline and potentially with
regulation in place. As such, a CGE model may be able to capture interactions between
economic sectors and provide information on changes outside of the directly regulated sector. In
support of previous rulemakings, such as the 2008 Final Ozone NAAQS (U.S. EPA, 2008) and
the 2010 Transport Rule proposal (U.S. EPA, 2010a), the EPA used the Economic Model for
5-37
-------
Policy Analysis (BMPAX) CGE model to estimate the secondary market effects based on the
cost impacts projected by IPM for the directly regulated sector.
When considering the secondary market impacts of a regulation, the effects of the costs,
the benefits of improved air quality, and their interaction may be relevant. Therefore, in the
Second Prospective Analysis under Section 812 of the Clean Air Act Amendments, the EPA
incorporated a set of health benefits arising from air quality improvement into the EMPAX CGE
model when studying the economy-wide impacts of the Clean Air Act (U.S. EPA 2011). While
the external Council on Clean Air Compliance Analysis (COUNCIL) review of this study stated
that inclusion of benefits in an economy-wide model "represented] a significant step forward in
benefit-cost analysis" (Hammitt 2010), the EPA recognizes that serious technical challenges
remain when attempting to evaluate the benefits and costs of potential regulatory actions using
economy-wide models.
In light of these challenges, the EPA has established a Science Advisory Board (SAB)
panel on economy-wide modeling to consider the technical merits and challenges of using this
analytical tool to evaluate costs, benefits, and economic impacts in regulatory development. In
addition, EPA is asking the panel to identify potential paths forward for improvements that could
address the challenges posed when using economy-wide models to evaluate the effects of
regulations. The final panel membership was announced in March 2015 and the first of multiple
face-to-face meetings of the SAB panel was held October 22 and 23, 2015. The EPA will use the
recommendations and advice of this panel as an input into its process for improving benefit-cost
and economic impact analyses used to inform decision-making at the agency.
The advice from the SAB panel formed specifically to address the subject of economy-
wide modeling was not available in time for this proposed action. Given the ongoing SAB panel
on economy-wide modeling, and the ongoing challenges of accurately representing costs,
benefits, energy efficiency improvements in economy-wide modeling, this chapter considers the
energy impacts associated with the proposed alternatives analyzed and a qualitative assessment
of the factors that will, in part, determine the timing and magnitude of effects in other markets.
5-38
-------
5.6.2 Summary of Secondary Market Impacts of Energy Price Changes
Electricity, natural gas, and coal are important inputs to the production of other goods and
services. Therefore, changes in the price of these commodities will shift the production costs for
sectors that use electricity, natural gas, and coal in the production of other goods and services.
Changes in the types and levels of inputs used by producers in response to electricity and fuel
price changes may mitigate the production cost changes in these sectors. Such changes in
production costs may lead to changes in the quantities and/or prices of the goods or services
produced and changes in imports and exports.
The EPA used IPM to estimate electricity, natural gas, and coal price changes based on
compliance with the regulatory control alternatives for this rule. The RPM uses estimated
changes in wholesale prices to estimate changes in average retail prices. The prices are average
prices over consumer classes and regions weighted by the amount used. Table 5-10 shows these
estimated price changes for the proposal in 2017 and, for illustrative purposes only, 2020. For
other results generated by IPM and the RPM, please refer to earlier sections in Chapter 5.
There are many factors influencing the projected natural gas prices. IPM (and its integrated
gas resource and supply module) models natural gas natural gas supplies based on a multitude of
factors. Since the model simulates perfect foresight, it anticipates future demand for natural gas
and responds accordingly. In addition, IPM (and the natural gas module) are viewing a very long
time horizon (through 2050), such that the impacts in certain years may be responsive to other
modeling assumptions or drivers. The modeling framework is simultaneously solving for all of
these key market and policy parameters (both electric and natural gas), resulting in the impacts
discussed in previous sections and shown in Table 5-9.
Table 5-10. Estimated Percentage Changes in Average Energy Prices by Energy Type for
the Proposed Alternative*
Proposed Rule
Electricity Price Change
Delivered Natural Gas Price Change
Delivered Coal Price Change
2017
-0.1%
0.01%
-0.3%
2020
-0.2%
-0.1%
-0.4%
*A minus sign in front of the number denotes a negative change, or a price decrease.
5-39
-------
For years when the price of electricity, natural gas, or coal increased, one would expect decreases
in production and increases in market prices in sectors for which these commodities are inputs,
ceteris paribus. Conversely, for years when prices of these inputs decreased, one would expect
increases in production and decreases in market prices within these sectors. Smaller changes in
input price changes would lead to smaller impacts within secondary markets. For compliance
with the proposal and more and less stringent alternatives these price changes are quite small, or
even negative, in 2017. Thus, one could readily surmise that the impacts of this proposed rule
upon secondary markets should be minimal. As seen in Table 5-9, most of the price changes are
negative, thus showing an estimated decrease in energy prices, albeit very small. The variation
for these prices at the regional level is not substantially different from the national-level price
changes, and the variation of energy prices related to the proposal and the regional level can be
found among the IPM outputs in the public docket for this rulemaking. However, a number of
factors, in addition to the magnitude and sign of the energy price changes, influence the
magnitude of the impact on production and market prices for sectors using electricity, natural
gas, or coal as inputs to production. These factors are discussed below.
5.6.3 Share of Total Production Costs
The impact of energy price changes in a particular sector depends, in part, on the share of
total production costs attributable to those commodities. For sectors in which the directly
affected inputs are only a small portion of production costs, the impact will be smaller than for
sectors in which these inputs make up a greater portion of total production costs. Therefore, more
energy-intensive sectors would potentially experience greater cost increases when electricity,
natural gas, or coal prices increase, but would also experience greater reduced costs when these
input prices decrease.
5.6.4 Ability to Substitute between Inputs to the Production Process
The ease with which producers are able to substitute other inputs for electricity, natural
gas, or coal, or even amongst those commodities, influences the impact of price changes for
these inputs. Those sectors with a greater ability to substitute across energy inputs or to other
inputs will be able to, at least partially, offset the increased cost of these inputs resulting in
5-40
-------
smaller market impacts. Similarly, when prices for electricity, natural gas, or coal decrease, some
sectors may choose to use more of these inputs in place of other more costly substitutes.
5.6.5 Availability of Substitute Goods and Services
The ability of producers in sectors experiencing changes in their input prices to pass along
the increased costs to their customers in the form of higher prices for their products depends, in
part, on the availability of substitutes for the sectors' products. Substitutes may be either other
domestic products or foreign imports. If close substitutes exist, the demand for the product will
in general be more elastic and the producers will be less able to pass on the added cost through a
price increase.
Such substitution can also take place between foreign and domestic goods within the same
sector. Changes in the price of electricity, natural gas, and coal can influence the quantities of
goods imported or exported from sectors using these inputs. When the cost of domestic
production increases due to more expensive inputs, imports may increase as consumers substitute
towards relatively less costly foreign-produced goods, and vice versa.
5.6.6 Effect of Changes in Input Demand from Electricity Sector
Section 5.7.2 focuses on the effects of changes in energy prices, and possible responses to
those price changes, on sectors outside of the electricity sector. A change in demand for inputs in
the electricity sector, as well as changes in demand for energy efficiency services and products,
will also influence economic activity in other sectors of the economy. For example, EPA
estimates EGUs will increase use of ammonia and urea to reduce NOx emissions, which could
result in a price driven reduction in demand in other sectors (e.g., fertilizer manufacturing and
agriculture) using these industrial chemicals.
5.6.7 Conclusions
Changes in the price of electricity, natural gas, and coal can affect markets for goods and
services produced by sectors that use these energy inputs in the production process. The direction
and magnitude of these impacts are influenced by a number of factors. For example, the more
producers in these sectors are able to substitute away from the use of these energy inputs, the
smaller the effect of energy prices changes will be on their production cost. Changes in cost of
5-41
-------
production may lead to changes in price, quantity produced, and profitability of firms within
secondary markets. Furthermore, the demand inputs in the electricity sector will also affect
secondary markets. If regulation results in changes in domestic markets that lead to an increase
in imports, then domestic producers may experience less demand from their consumers, and vice
versa.
Modeling choices in IPM influence the estimated changes in electricity, natural gas, and
coal prices in this RIA. Actual market conditions will ultimately influence the price changes of
these energy inputs and consequent effects on secondary markets, as will the plan approaches
that states adopt.
5.7 References
Arrow, K. I; 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." Federal 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.
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
5-42
-------
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.
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.
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.
U.S. Energy Information Agency (EIA). 2014. The Electricity Market Module of the National
Energy Modeling System: Model Documentation 2014. Available at:
5-43
-------
.
Accessed 9/17/2015.
U.S. EPA, 2015. Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric
Utility Generating Units (FinalRule), http://www2.epa.gov/cleanpowerplan/clean-power-
plan-exi sting-power-plants.
U.S. EPA, 2015a. 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-fmal-rule.
U.S. EPA, 2014. Final Rule for Existing Power Plants andFactories,
http://www2.epa.gov/cooling-water-intakes.
U.S. EPA, 2011. Cross-State Air Pollution Rule,
http://www3.epa.gov/airtransport/CSAPR/index.html
U.S. EPA, 201 la. Mercury and Air Toxics Standards (MATS), http://www3.epa.gov/mats/.
U.S. EPA. 2010. EPA Guidelines for Preparing Economic Analyses. Available at:
. Accessed 9/21/2015.
U.S. EPA. 2010a. Regulatory Impact Analysis for the Proposed Federal Transport Rule
Analyses. Available at: < http://www3.epa.gov/ttnecasl/ria.html>. Accessed 9/21/2015.
U.S. EPA, 2005. Clean Air Inter state Rule,
http://archive.epa.gov/airmarkets/programs/cair/web/html/index.html.
5-44
-------
CHAPTER 6: ESTIMATED HUMAN HEALTH BENEFITS AND CLIMATE CO-
BENEFITS
6.1 Introduction
As discussed in Chapter 4, this action proposes to update CSAPR to reduce interstate
transport of EGU ozone season NOx emissions that contribute significantly to nonattainment or
interfere with maintenance of the 2008 ozone NAAQS. The EPA proposes to implement the
proposed EGU NOx reductions by setting emissions budgets that are implemented through the
CSAPR NOx ozone-season allowance trading program. Implementing this proposal to update
CSAPR is expected to 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,
fine particulate matter (PM2.s)-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 PIVh.s 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 several pollutants,
including those from reducing direct exposure to NO2, ecosystem effects and visibility
impairment. We qualitatively discuss these unquantified benefits in this chapter.
This chapter provides estimates of the monetized air quality health benefits and climate co-
benefits associated with emission reductions for the regulatory control alternatives across several
discount rates. The estimated benefits associated with these emission reductions are beyond
those achieved by previous EPA rulemakings, including the 2011 CSAPR92 and the finalized
Clean Power Plan (CPP).
92 As discussed in Chapters 1, 3 and 5, the 2017 baseline EGU emissions for this proposal include impacts from
CSAPR issued on July 6, 2011. As discussed and elaborated on in Chapter 1, because the modeling for the proposal
was performed prior to the D.C. Circuit's issuance of EME Homer City II, 11 that modeling assumed in its baseline
for all states the emission reductions associated with the CSAPR NOx ozone-season phase 2 emissions budgets.
5-1
-------
6.2 Estimated Human Health Benefits
The proposal to update CSAPR 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 proposal would also reduce emissions of NOx throughout the year.
Because NOx is also a precursor to formation of ambient PIVh.s, reducing these emissions would
also reduce human exposure to ambient PIVh.s throughout the year and would reduce the
incidence of PM2.5-related health effects.93 This RIA does not quantify PIVh.s-related benefits
associated with SO2 emission reductions. As discussed in Chapter 3, the EPA does not estimate
significant SO2 emission reductions as a result of this proposal. Furthermore, these reductions
would reduce ozone and PIVh.s concentrations in regions other than those with nonattainment
monitors that are the subject of this proposal, and the benefits of reducing these pollutants in
those regions are assessed in this Chapter. Reducing emissions of NOx would also reduce
ambient exposure to NO2 and its associated health effects, though we do not quantify these
effects because we have not developed a reduced-form technique for estimating NO2 impacts. In
this section, we provide an overview of the monetized ozone-related benefits and PIVh.s-related
co-benefits estimated for the proposed updated CSAPR EGU NOx ozone season emissions
budgets and for the more and less stringent alternatives. 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). The estimated benefits associated with these emissions reductions are additional to
those achieved by previous EPA rulemakings, including the finalized CPP.
Implementing these updated CSAPR EGU NOx emissions budgets for the ozone season in
23 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
93 Additionally, this RIA does not estimate changes in emissions of directly emitted particles.
6-2
-------
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
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. For example, some of the emissions
reductions estimated to result from implementing the proposal to update CSAPR may achieve
some of the air quality improvements that resulted from the hypothesized attainment strategies
presented in the NAAQS RIAs. The emissions reductions from implementing the proposal to
update CSAPR will decrease the remaining amount of emissions reductions needed in
nonattainment areas to, and reduce the costs to those areas to, meet the 2008 ozone NAAQS.
Specifically, the anticipated reductions in ozone concentrations from this rule would help areas
attain and maintain the 2008 ozone NAAQS and achieve some of the air quality improvements
from the hypothesized attainment strategy from the 2008 ozone NAAQS RIA. These ozone
improvements would similarly achieve some of the air quality improvements assumed in the
baseline (i.e., 2008 ozone NAAQS attainment projection) for the 2015 ozone NAAQS RIA.
As discussed in Chapter 4, the IPM modeling showing compliance with the regulatory
control alternatives for which emission reductions are estimated in this RIA is also illustrative in
nature. However, unlike the illustrative control strategies analyzed in NAAQS RIAs described
above, all of the emission reductions for the illustrative compliance modeling for this proposal
would occur in one well-characterized sector (i.e., the EGU sector). The EPA is more confident
in the magnitude and location of the emission reductions for this proposal because it imposes a
specific requirement that limits emissions from a specific sector. Emission reductions achieved
under this rule, if finalized, will ultimately be reflected in the baseline of future NAAQS
analyses and would lower the 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).
6-3
-------
6.2.1 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 ambient
ozone exposure, which also include premature mortality and a variety of morbidity effects
associated with acute and chronic exposures. Similarly, the Integrated Science Assessment for
Paniculate Matter (PM ISA) (U.S. EPA, 2009b) identified the human health effects associated
with ambient PIVh.s exposure, which include premature mortality and a variety of morbidity
effects associated with acute and chronic exposures. Table 6-1 identifies the quantified and
unquantified benefit and co-benefit categories captured in the EPA's health benefits estimates for
reduced exposure to ambient ozone and PIVh.s. 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. It is important to emphasize that the list of unquantified
benefits categories is not exhaustive, nor is quantification of each effect complete.
6-4
-------
Table 6-1. Human Health Effects of Ambient Ozone and PMi.s
Category
Effect Has
Specific Effect Been
Quantified
Effect Has
_ More
Been T .
, , ,. , Information
Monetized
Improved Human Health
Reduced incidence of
mortality from
exposure to ozone
Reduced incidence of
morbidity from
exposure to ozone
Reduced incidence of
premature mortality
from exposure to
PM25
Reduced incidence of
morbidity from
exposure to PM2 5
Premature mortality based on short-term study ,
estimates (all ages)
Premature mortality based on long-term study
estimates (age 30-99)
Hospital admissions respiratory causes (age > 65) S
Hospital admissions respiratory causes (age <2) S
Emergency department visits for asthma (all ages) S
Minor restricted-activity days (age 18-65) S
School absence days (age 5-17) S
Decreased outdoor worker productivity (age 1 8-65)
Other respiratory effects (e.g., premature aging of
lungs)
Cardiovascular and nervous system effects
Reproductive and developmental effects
Adult premature mortality based on cohort study
estimates and expert elicitation estimates (age >25 S
or age >30)
Infant mortality (age < 1 ) /
Non-fatal heart attacks (age > 1 8) V
Hospital admissions respiratory (all ages) V
Hospital admissions cardiovascular (age >20) V
Emergency room visits for asthma (all ages) V
Acute bronchitis (age 8-12) V
Lower respiratory symptoms (age 7-14) V
Upper respiratory symptoms (asthmatics age 9-11) /
Asthma exacerbation (asthmatics age 6-18) /
Lost work days (age 1 8-65) /
Minor restricted-activity days (age 1 8-65) /
Chronic Bronchitis (age >26)
Emergency room visits for cardiovascular effects
(all ages)
Strokes and cerebrovascular disease (age 50-79)
Other cardiovascular effects (e.g., other ages)
Other respiratory effects (e.g., pulmonary function,
non-asthma ER visits, non-bronchitis chronic
diseases, other ages and populations)
Reproductive and developmental effects (e.g., low
birth weight, pre-term births, etc.)
Cancer, mutagenicity, and genotoxicity effects
/ Ozone ISA
Ozone ISA1
/ Ozone ISA
/ Ozone ISA
/ Ozone ISA
/ Ozone ISA
/ Ozone ISA
Ozone ISA1
Ozone ISA2
Ozone ISA2
Ozone ISA2-3
-------
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. In addition to transferring information from other contexts to the context of this
regulation, we also use a "benefit-per-ton" approach to estimate the ozone and PM2.5 benefits in
this RIA. Benefit-per-ton approaches apply an average benefit per ton derived from modeling of
benefits of specific air quality scenarios to estimates of emissions reductions for scenarios where
no air quality modeling is available. 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 emissions reductions from other air quality modeling.
This section describes the underlying basis for the health and economic valuation estimates that
inform the benefit-per-ton estimates, and the subsequent section provides an overview of the
benefit-per-ton estimates, which are described in detail in the appendix to this chapter.
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. The
next sections provide an overview of the health impact assessment (HIA) methodology and
additional details on several key elements.
The HIA quantifies the changes in the incidence of adverse health impacts resulting from
changes in human exposure to ozone and PIVb.s. 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 (Abt Associates, 2012). 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
6-6
-------
estimates of the relative risks of a particular health effect for a given increment of air pollution
(often per 10 ppb for ozone or jig/m3 for PIVb.s). 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, and we briefly elaborate on adult premature mortality 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.
6.2.1.1 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 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 an 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 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 NMMAPS studies without exclusion of the meta-
6-7
-------
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 SAB-HES panel, and the
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
HREA (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
different model specifications including (a) regional versus national Bayes-based adjustment,94
(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.
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.
6-8
-------
(2004) study in preparing ozone monitor data.95 In selecting effect estimates from Smith et al.
(2009), we focused on 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
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.
6.2.1.2 PM2.5 Mortality Effect Coefficients for Adults and Infants
A substantial body of published scientific literature documents the association between
elevated PIVb.s 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
95 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.
6-9
-------
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 PIVb.s 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 National Research Council (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.
Over the last two 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
6-10
-------
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 Ic). 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
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
6-11
-------
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 MSAs (U.S. EPA, 2010b). The CASAC 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 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.s). The
relative risk estimate (1.06 per 10 |ig/m3 increase in PM2.s) 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 tighter.
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
between PM2.5 exposure and increased risk of 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 ug/m3 and that 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 in PM2.s). The
relative risk estimate is slightly smaller than the risk estimate drawn from Laden et al. (2006),
with relatively smaller confidence intervals.
6-12
-------
Given that monetized benefits associated with PIVh.s 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; ffic, 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 mortality studies described above, several studies show an
association between PM exposure and premature mortality in children under 5 years of age.96
The PM ISA states that less evidence is available regarding the potential impact of PIVh.s
exposure on infant mortality than on adult mortality and the results of studies in 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,
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
96 For the purposes of this analysis, we only calculate benefits for infants age 0-1, not all children under 5 years old.
6-13
-------
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 PMio results are scaled to estimate PIVb.s 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.
6.2.2 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 proposed 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
literature concerning the appropriate method for valuing reductions in premature mortality risk is
still developing. 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
community. Following the advice of the SAB's Environmental Economics Advisory Committee
(SAB-EEAC), the EPA currently uses the value of statistical life (VSL) approach in calculating
estimates of mortality benefits, because we believe this calculation provides the most reasonable
single estimate of an individual's willingness to trade off money for reductions in mortality risk
(U.S. EPA-SAB, 2000). The VSL approach is a summary measure for the value of small changes
in mortality risk experienced by a large number of people.
6-14
-------
The EPA continues work to update its guidance on valuing mortality risk reductions, and
the Agency consulted several times with the SAB-EEAC on this issue. Until updated guidance is
available, the Agency determined that a single, peer-reviewed estimate applied consistently, best
reflects the SAB-EEAC advice it has received. Therefore, the EPA has decided to apply the VSL
that was vetted and endorsed by the SAB in the Guidelines for Preparing Economic Analyses
(U.S. EPA, 2014)97 while the Agency continues its efforts to update its guidance on this issue.
This approach calculates a mean value across VSL estimates derived from 26 labor market and
contingent valuation studies published between 1974 and 1991. The mean VSL across these
studies is $6.3 million (2000$).98 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 $9.9 million for 2017.
The Agency is committed to using scientifically sound, appropriately reviewed evidence in
valuing mortality risk reductions and has made significant progress in responding to the SAB-
EEAC 's specific recommendations. In the process, the Agency has identified a number of
important issues to be considered in updating its mortality risk valuation estimates. These are
detailed in a white paper, "Valuing Mortality Risk Reductions in Environmental Policy" (U.S.
EPA, 2010c), which recently underwent review by the SAB-EEAC. A meeting with the SAB on
this paper was held on March 14, 2011 and formal recommendations were transmitted on
July 29, 2011 (U.S. EPA-SAB, 2011). The EPA is taking SAB's recommendations under
advisement.
In valuing PM2.s-related premature mortality, we discount the value of premature mortality
occurring in future years using rates of 3 percent and 7 percent (OMB, 2003). We assume that
there is a "cessation" lag between changes in PM exposures and the total realization of changes
in health effects. Although the structure of the lag is uncertain, the EPA follows the advice of the
SAB-HES to assume a segmented lag structure characterized by 30 percent of mortality
reductions in the first year, 50 percent over years 2 to 5, and 20 percent over the years 6 to 20
after the reduction in PM2.5 (U.S. EPA-SAB, 2004c). Changes in the cessation lag assumptions
97 In the updated Guidelines for Preparing Economic Analyses (U.S. EPA, 2010e), the EPA retained the VSL
endorsed by the SAB with the understanding that further updates to the mortality risk valuation guidance would be
forthcoming.
98 In 1990$, this base VSL is $4.8 million.
6-15
-------
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 co-benefits are identical for all discount rates.
6.2.3 Benefit-per-ton Estimates for Ozone
We used a "benefit-per-ton" approach to estimate the ozone benefits in this RIA. The EPA
has applied this approach in several previous RIAs (e.g., U.S. EPA, 201 Ib, 201 Ic, 2012b, 2014a;
2015). These benefit-per-ton estimates provide the total monetized human health benefits (the
sum of premature mortality and premature morbidity) of reducing one ton of NOx (an ozone
precursor). We generated benefit-per-ton estimates for ozone based on air quality modeling for
the illustrative control case described in Chapter 3 of this RIA. The estimates correspond to NOx
emissions from U.S. EGUs during the ozone-season (May to September). Because we estimate
ozone health impacts from May to September only, this approach underestimates ozone benefits
in areas with a longer ozone season such as Texas. These estimates assume that EGU-attributable
ozone formation at the regional-level is due to 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. We provide more detailed information regarding the
generation of these estimates in the appendix to this chapter.
As noted below in the characterization of uncertainty, all benefit-per-ton estimates have
inherent limitations. Specifically, the benefit-per-ton estimates reflect the geographic distribution
of the modeled illustrative control case, which may not match the geographic distribution of
emission reductions anticipated for compliance with the regulatory control alternatives, and they
may not reflect local variability in population density, meteorology, exposure, baseline health
incidence rates, or other local factors for any specific location.
6.2.4 Benefit-per-ton Estimates for PM2.5
We used a "benefit-per-ton" approach to estimate the PM2.5 co-benefits in this RIA, which
represent the total monetized human health co-benefits (the sum of premature mortality and
premature morbidity), of reducing one ton of PM2.5 (or PM2.5 precursor such as NOx) from a
specified source. Specifically, in this analysis, we multiplied the benefit-per-ton estimates by the
corresponding emission reductions that were generated from air quality modeling of the
6-16
-------
illustrative control case. The appendix to this chapter provides additional detail regarding these
calculations. All benefit-per-ton estimates have inherent limitations and should be interpreted
with caution. Even though we assume that all fine particles have equivalent health effects, the
benefit-per-ton estimates vary between precursors depending on the location and magnitude of
their impact on PIVh.s levels, which drive population exposure.
6.2.5 Estimated Health Benefits Results
Table 6-2 provides the benefit-per-ton estimates for the analysis year 2017; because the
modeled change in SO2 levels was very small, we did not quantify the SCh-related PIVb.s
benefits. Table 6-3 provides the emission reductions estimated to occur in the analysis year.
Table 6-4 summarizes the national monetized ozone-related and PM-related health benefits
estimated to occur for regulatory control alternatives for the 2017 analysis year, by precursor
pollutant using discount rates of 3 percent and 7 percent. Table 6-5 provides national summaries
of the reductions in estimated health incidences associated with the proposal and more and less
stringent alternatives for the 2017 analysis year." Figure 6-1 provides a visual representation of
the range of estimated ozone and PIVb.s-related benefits using benefit-per-ton estimates based on
concentration-response functions from different studies and expert opinion for the proposal
evaluated for 2017.
99 Incidence estimates were generated using the same "per ton" approach as used to generate the dollar benefit per
ton values. See Appendix 6-A for details.
6-17
-------
Table 6-2. Summary of Ozone and PMi.s Benefit-per-Ton Estimates Based on Air
Quality Modeling from the Illustrative Control Case in 2017 (2011$)*
Pollutant Discount Rate National
NOx (as Ozone) N/A $5,700 to $9,400
3% $2,100 to $4,700
NOx(aSPM2'5) 7% $1,900 to $4,300
* The range of estimates reflects the range of epidemiology studies for avoided premature mortality for ozone and
PM25. All estimates are rounded to two significant figures. The monetized co-benefits do not include reduced
health effects from direct exposure to NO2, SO2, 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 PM25 concentrations, which drive population exposure. The monetized co-benefits
incorporate the conversion from precursor emissions to ambient fine particles and ozone. Benefit-per-ton
estimates for ozone are based on ozone season NOx emissions. Ozone co-benefits occur in the analysis year, so
they are the same for all discount rates. Confidence intervals are unavailable for this analysis because of the
benefit-per-ton methodology. 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).
Table 6-3. Emission Reductions of Criteria Pollutants for the Proposal and More and
Less Stringent Alternatives in 2017 (thousands of short tons)*
Ozone Season NOx
All Year NOx
All Year SO2
Proposal
85,000
90,000
380
More Stringent Alternative
87,000
93,000
430
Less Stringent
Alternative
24,000
24,000
300
* All emissions shown in the table are rounded, so regional emission reductions may appear to not sum to national
total.
Table 6-4. Summary of Estimated Monetized Health Benefits for the Proposal and
More and Less Stringent Alternatives Regulatory Control Alternatives for 2017
(millions of 201 IS)*
Pollutant
NOx (as Ozone)
NOx (as PM2 5)
Total
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
Proposal
$490 to $790
$190 to $430
$170 to $380
$670 to $1,200
$650 to $1,200
More Stringent
Alternative
$500 to $820
$190 to $440
$170 to $390
$690 to $1,300
$670 to $1,200
Less Stringent
Alternative
$140 to $220
$49 to $110
$45 to $100
$190 to $340
$180 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 climate benefits or
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 PM25 levels, which drive population exposure. The monetized co-benefits incorporate
the conversion from precursor emissions to ambient fine particles and ozone. Co-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.
Confidence intervals are unavailable for this analysis because of the benefit-per-ton methodology. 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
6-18
-------
around the ozone mortality estimates are on the order of +/- 60 percent depending on the concentration-response
function used.
Table 6-5. Summary of Avoided Health Incidences from Ozone-Related and PMi.s-
Related Benefits for the Proposal and More and Less Stringent Alternatives for
2017*
Ozone-related Health Effects
Proposal
More Stringent
Alternative
Less Stringent
Alternative
Avoided Premature Mortality
Smith et al. (2009) (all ages)
Zanobetti and Schwartz (2008) (all ages)
48
81
50
83
14
23
Avoided Morbidity
Hospital admissions respiratory causes (ages > 65)
Emergency room visits for asthma (all ages)
Asthma exacerbation (ages 6-18)
Minor restricted-activity days (ages 18-65)
School loss days (ages 5-17)
79
320
93,000
240,000
77,000
81
330
95,000
240,000
79,000
22
90
26,000
67,000
22,000
PM2.s-related Health Effects
Avoided Premature Mortality
Krewski et al. (2009) (adult)
Lepeule etal. (2012) (adult)
Woodruff et al. (1997) (infant)
21
48
<1
22
50
<1
6
13
<1
Avoided Morbidity
Emergency department visits for asthma (all ages)
Acute bronchitis (age 8-12)
Lower respiratory symptoms (age 7-14)
Upper respiratory symptoms (asthmatics age 9-11)
Minor restricted-activity days (age 18-65)
Lost work days (age 18-65)
Asthma exacerbation (age 6-18)
Hospital admissions respiratory (all ages)
Hospital admissions cardiovascular (age > 18)
Non-Fatal Heart Attacks (age >18)
Peters etal. (2001)
Pooled estimate of 4 studies
12
31
390
560
16,000
2,700
580
6
8
25
3
12
32
400
580
16,000
2,700
600
7
8
26
3
3
8
100
150
4,200
700
150
2
2
7
1
* All estimates are rounded to whole numbers with two significant figures.. Co-benefits for ozone are based on
ozone season NOx emissions. Confidence intervals are unavailable for this analysis because of the incidence-per-ton
methodology. 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.
6-19
-------
0)
c
o
M
O
$1,800
o? ,-ar ,<
«* , > ^
r,vr ^ r&
L w v v iCr
«1 ^' <*' c\b *
n»° d-
-------
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 estimate of benefits. When
the uncertainties from each stage of the analysis are compounded, even small uncertainties can
have large effects on the total quantified benefits. In addition, 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 representative of the general
magnitude of benefits of the regulatory control alternatives for the 2017 analysis year, rather than
the actual benefits anticipated from implement the proposal.
This RIA does not include the type of detailed uncertainty assessment found in the Ozone
NAAQS RIA (U.S. EPA, 2015) or the PM NAAQS RIA (U.S. EPA, 2012a) because we lack the
necessary air quality modeling input and/or monitoring data to run the benefits model. However,
the results of the quantitative and qualitative uncertainty analyses presented in the Ozone
NAAQS RIA and PM NAAQS RIA can 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.s-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.s-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 et al. (2012). 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).
6-21
-------
We 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 proposal is different than the emissions
in the air quality modeling of the illustrative control case, the benefits may be underestimated or
overestimated.
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, 2002). Below are key assumptions underlying the estimates for
PM2.s-related premature mortality, which accounts for 98 percent of the monetized PIVh.s 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 PIVh.s
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 PIVh.s 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 PIVh.s, including both areas that do not meet the fine
particle standard and those areas that are in attainment, down to the lowest modeled
concentrations.
6-22
-------
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 PIVh.s 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 PIVh.s concentrations that fall below the bulk of the observed
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 PIVh.s levels at or above different
concentrations, which provides some insight into the level of uncertainty in the estimated PIVh.s
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
EP'A''s Policy Assessment for P'articulate 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.100 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).
We report also the key assumptions associated with our analysis of ozone-related effects:
i oo por a summary Of the scientific review statements regarding the lack of a threshold in the PM2 s-mortality
relationship, see the TSD entitled Summary of Expert Opinions on the Existence of a Threshold in the
Concentration-Response Function for PM2.s-related Mortality (U.S. EPA, 2010b).
6-23
-------
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
estimated at or below this level are subject to greater uncertainty, but we cannot judge
whether (and in what direction) these impacts are 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 is not available, and the control scenarios
are illustrative of what states may choose to do. 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 PIVh.s
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.101 It is important to note
101 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.
6-24
-------
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 proposal. Therefore, caution is
warranted when interpreting the LML assessment in this RIA because these results are not
consistent with results from RIAs that had air quality modeling.
Figure 6-3 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 6-4 shows a cumulative distribution function of the same data. Both figures
identify the LML for each of the major cohort studies.
LML of Kiewski *t al
(2009) study
LML of Lepeule et al
(2012) study
I
'
5.8 6
8 9 10 12 14 16 18 20
Baseline Annual Mean PM: < Level
6-25
-------
Among the populations exposed to PlVh.5 in the baseline:
88% are exposed to PM2s levels at or above the LML of the Krewski et al. (2009) study
47% are exposed to PM2s levels at or above the LML of the Lepeule et al. (2012) study
Figure 6-2. Percentage of Adult Population (age 30+) by Annual Mean PMi.s Exposure in
the Baseline used for the Air Quality Analysis in Chapter 3
80H
70%
1 50%
i
LML of Krewski et al,
(2009J study
I OH
OH
LML of Lepeule «t al,
(2012) study
[ 2 3 4 5 5.S 6 8 9
Baseline Annual Mean PMi < Level
10 12 14 16 IS
Among the populations exposed to PM2.s in the baseline:
88% are exposed to PM2.s levels at or above the LML of the Krewski et al. (2009) study
47% are exposed to PM25 levels at or above the LML of the Lepeule et al. (2012) study
Figure 6-3. Cumulative Distribution of Adult Population (age 30+) by Annual Mean
PMi.s Exposure in the Baseline used for the Air Quality Analysis in Chapter 3
6.3 Estimated Climate Co-Benefits from COi
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-CCh), a metric that estimates the monetary value of impacts
6-26
-------
associated with marginal changes in CO2 emissions in a given year. Table 6-6 summarizes the
quantified and unquantified climate benefits in this analysis.
Table 6-6. Climate Effects
Benefits Specific Effect Effect Has Been Effect Has Been More
Category _ Quantified _ Monetized _ Information
Improved Environment
Reduced Global climate impacts from CO2 ~ ^ SCC TSD
climate effects Climate impacts from ozone and Ozone ISA, PM
black carbon (directly emitted PM) ISA2
Other climate impacts (e.g., other IPCC2
GHGs such as methane, aerosols,
_ other impacts)
The global climate and related impacts of COa emissions changes, such as sea level rise, are estimated within each
integrated assessment model as part of the calculation of the SC-CO2. The resulting monetized damages, which are
relevant for conducting the benefit-cost analysis, are used in this RIA to estimate the welfare effects of quantified
changes in CO2 emissions.
2 We assess these co-benefits qualitatively because we do not have sufficient confidence in available data or
methods.
There are several important considerations in assessing the climate-related benefits for an
ozone air quality-focused rulemaking. First, the estimated health benefits do not account for any
climate-related air quality changes (e.g., increased ambient ozone associated with higher
temperatures). Excluding climate-related air quality changes may underestimate ozone-related
health benefits. It is unclear how PM2.s-related health benefits would be affected by excluding
climate-related air quality changes since the science is unclear as to how climate change may
affect PM2.5 exposure. Second, the estimated health benefits also do not consider temperature
modification of PM2.s 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, as
noted earlier, we do not estimate the climate co-benefits associated with reductions in PM and
ozone precursors.
6-27
-------
6.3.1 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.
The preamble also summarizes new scientific assessments and recent climatic observations.
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 (NC A), in particular, assessed the impacts of climate change on human health in the
United States, finding that Americans will be affected by "increased extreme weather events,
wildfire, decreased air quality, threats to mental health, and illnesses transmitted by food, water,
and disease-carriers such as mosquitoes and ticks." These assessments also detail the risks to
vulnerable groups such as children, the elderly and low income households. Furthermore, the
assessments present an improved understanding of the impacts of climate change on public
welfare, higher projections of future sea level rise than had been previously estimated, a better
understanding of how the warmth in the next century may reach levels that would be
unprecedented relative to the preceding millions of years of history, and new assessments of the
impacts of climate change on permafrost and ocean acidification. The impacts of GHG emissions
will be realized worldwide, independent of their location of origin, and impacts outside of the
United States will produce consequences relevant to the United States.
6.3.2 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:
6-28
-------
Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866 (May 2013, Revised July 2015) ("current TSD").102 We refer to these estimates,
which were developed by the U.S. government, as "SC-CCh estimates." The SC-CCh 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 (lAMs) to develop the SC-CCh estimates and recommended four
global values for use in regulatory analyses. The SC-CCh estimates were first released in
February 2010 and updated in 2013 using new versions of each IAM. As discussed further
below, the IWG published two minor corrections to the SC-CO2 estimates in July 2015.
The SC-CO2 estimates were developed using an ensemble of the three most widely cited
integrated assessment models in the economics literature with the ability to estimate the SC-CCh.
A key objective of the IWG was to draw from the insights of the three models while respecting
the different approaches to linking GHG emissions and monetized damages taken by modelers in
the published literature. After conducting an extensive literature review, the interagency group
selected three sets of input parameters (climate sensitivity, socioeconomic and emissions
trajectories, and discount rates) to use consistently in each model. All other model features were
102 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.
6-29
-------
left unchanged, relying on the model developers' best estimates and judgments, as informed by
the literature. Specifically, a common probability distribution for the equilibrium climate
sensitivity parameter, which informs the strength of climate's response to atmospheric GHG
concentrations, was used across all three models. In addition, a common range of scenarios for
the socioeconomic parameters and emissions forecasts were used in all three models. Finally, the
marginal damage estimates from the three models were estimated using a consistent range of
discount rates, 2.5, 3.0, and 5.0 percent. See the 2010 TSD for a complete discussion of the
methods used to develop the estimates and the key uncertainties, and the current TSD for the
latest estimates.103
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.
The SC-CO2 must therefore incorporate the full (global) damages caused by GHG emissions to
address the global nature of the problem. 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 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 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, as noted in OMB's Response to Comments on the SCC, there is no bright line
between domestic and global damages. Adverse impacts on other countries can have spillover
See https://www.whitehouse.gov/omb/oira/social-cost-of-carbon for both TSDs.
6-30
-------
effects on the United States, particularly in the areas of national security, international trade,
public health and humanitarian concerns.104
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.105 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 observed that SC-CCh estimates continue to omit various impacts that would
likely increase damages. The 95th percentile estimate was included in the recommended range
for regulatory impact analysis to address these concerns.
The EPA and other agencies have continued to consider feedback on the SC-CCh 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
104 See Endangerment and Cause or Contribute Findings for Greenhouse Gases Under Section 202(a) of the Clean
Air Act, 74 Fed. Reg. 66,496, 66,535 (Dec. 15, 2009) and National Research Council 2013a.
105 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.
6-31
-------
research analysts implementing the SC-CO2 methodology used by the interagency working
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.106 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 comments107, 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.108 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.109
Concurrent with OMB's publication of the response to comments on SC-CO2 and
announcement of the Academies process, OMB posted a revised TSD that includes two minor
technical corrections to the current estimates. One technical correction addressed an inadvertent
omission of climate change damages in the last year of analysis (2300) in one model and the
second addressed a minor indexing error in another model. On average the revised SC-CO2
estimates are one dollar less than the mean SC-CO2 estimates reported in the November 2013
revision to the May 2013 TSD. The change in the estimates associated with the 95th percentile
estimates when using a 3% discount rate is slightly larger, as those estimates are heavily
influenced by the results from the model that was affected by the indexing error.
106 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
107 See https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-response-to-comments-final-july-2015.pdf
108 See https://www.whitehouse.gov/blog/2015/07/02/estimating-benefits-carbon-dioxide-emissions-reductions.
109 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.
6-32
-------
The four SC-CO2 estimates are as follows: $12, $41, $63, and $120 per metric ton of CO2
emissions in the year 2017 (2011$).110 The first three values are based on the average SC-CO2
from the three lAMs, 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
models at a 3 percent discount rate. It is included to represent higher-than-expected impacts from
temperature change further out in the tails of the SC-CO2 distribution (representing less likely,
but potentially catastrophic, outcomes).
Table 6-7 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. m'112 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 6-8 reports the incremental climate co-benefits from CO2 emission impacts estimated for
the proposal and more and less stringent alternatives for the 2017 analysis year.
110 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.
111 Emission impacts for this rulemaking are shown for the year 2017 and are calculated in short tons. Therefore, the
conversion factor 0.90718474 metric tons in a short ton was applied to the calculate the benefits in the year 2017.
112 This analysis considered the climate impacts of only CCh emission change. As discussed below, the climate
impacts of other pollutants were not calculated for the proposed guidelines. While 62 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.
6-33
-------
Table 6-7.
Year
2015
2017
2020
2025
2030
2035
2040
2045
2050
Social Cost
5% Average
$12
$12
$13
$15
$17
$19
$22
$24
$28
of COi, 2015-2050 (in
Discount
3% Average
$38
$41
$45
$49
$53
$58
$64
$68
$73
2011$ per metric ton)*
Rate and Statistic
2.5% Average
$59
$63
$66
$72
$77
$83
$89
$94
$100
3%(95thpercentile)
$110
$120
$130
$150
$160
$180
$190
$210
$230
* These SC-CCh values are stated in $/metric ton and rounded to two significant figures. The estimates vary
depending on the year of CCh emissions and are defined in real terms, i.e., adjusted for inflation using the GDP
implicit price deflator.
Table 6-8. Estimated Global Climate Co-benefits of COi Reductions for the Proposal
and More and Less Stringent Alternatives for 2017 (millions of 2011$)*
Discount rate and statistic
Thousands of short tons of CCh
reduced**
5% (average)
3% (average)
2.5% (average)
3% (95th percentile)
Proposal
610
$6.5
$23
$35
$66
More Stringent
Alternative
610
$6.5
$23
$35
$66
Less Stringent
Alternative
720
$7.6
$27
$41
$78
* The SC-CCh values are dollar-year and emissions-year specific. SC-CCh values represent only a partial accounting
of climate impacts.
**Emission impacts for this rulemaking are shown for the year 2017 and are calculated in short tons. The
conversion factor 0.90718474 metric tons in a short ton was applied to the calculate the benefits in the year 2017.
It is important to note that the climate co-benefits presented above are associated with
changes in CO2 emissions only. Implementing this proposal, 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 this proposal.113
113 The SC-COa estimates used in this analysis are designed to assess the climate benefits associated with changes in
CO2 emissions only.
6-34
-------
The other emissions potentially reduced as a result of the proposal to update CSAPR
include other greenhouse gases (such as methane), aerosols and aerosol precursors such as black
carbon, organic carbon, sulfur dioxide and nitrogen oxides, and ozone precursors such as
nitrogen oxides and volatile organic carbon compounds. Changes in emissions of these pollutants
(both increases and decreases) could directly result from changes in electricity generation,
upstream fossil fuel extraction and transport, and/or downstream secondary market impacts.
Reductions in black carbon or ozone precursors are projected to lead to further cooling, but
reductions in the other aerosol species and precursors are projected to lead to warming.
Therefore, changes in non-CCh pollutants could potentially augment or offset the climate co-
benefits calculated here. These pollutants can act in different ways and on different timescales
than carbon dioxide. For example, aerosols reflect (and in the case of black carbon, absorb)
incoming radiation, whereas greenhouse gases absorb outgoing infrared radiation. In addition,
these aerosols are thought to affect climate indirectly by altering properties of clouds. Black
carbon can also deposit on snow and ice, darkening these surfaces and accelerating melting. In
terms of lifetime, while carbon dioxide emissions can increase concentrations in the atmosphere
for hundreds or thousands of years, many of these other pollutants are short lived and remain in
the atmosphere for short periods of time ranging from days to weeks and can therefore exhibit
large spatial and temporal variability.
6.4 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-CCh at each of three discount rates5 percent, 3 percent, 2.5
percentand the 95th percentile SC-CCh at 3 percent) (as recommended by the IWG).
6-35
-------
Different discount rates are applied to SC-CCh 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-CCh 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.114 The
EPA has evaluated the range of potential impacts by combining all SC-CCh values with health
benefits values at the 3 percent and 7 percent discount rates. Combining the 3 percent SC-CCh
values with the 3 percent health benefit values assumes that there is no difference in discount
rates
Table 6-9 provides the combined health and climate benefits for the proposal and more and
less stringent alternatives for the 2017 analysis year.
114 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.pdffor
details.
6-36
-------
Table 6-9. Combined Health Benefits and Climate Co-Benefits for the Proposal and
More and Less Stringent Alternatives for 2017 (millions of 2011$)*
SC-CCh Discount Rate**
Health and Climate Benefits
(Discount Rate Applied to Health Co-Benefits)
3% 7%
Climate Co-
Benefits Only
Proposal
5%
3%
2.5%
3%(95thpercentile)
$680 to $1200
$700 to $1200
$7 10 to $1300
$740 to $1300
$660 to $1200
$680 to $1200
$690 to $1200
$720 to $1200
$6.5
$23
$35
$66
More Stringent Alternative
5%
3%
2.5%
3%(95thpercentile)
$700 to $1300
$720 to $1300
$730 to $1300
$760 to $1300
$680 to $1200
$700 to $1200
$7 10 to $1200
$740 to $1300
$6.5
$23
$35
$66
Less Stringent Alternative
5%
3%
2.5%
3%(95thpercentile)
$190 to $340
$2 10 to $360
$230 to $380
$260 to $4 10
$190 to $330
$210 to $350
$220 to $370
$260 to $400
$7.6
$27
$41
$78
*A11 estimates are rounded to two significant figures. Climate benefits are based on reductions in CO2 emissions.
Co-benefits are based on regional benefit-per-ton estimates. Co-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 PM2s and ozone co-benefits and reflect the range 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
monetized health co-benefits do not include reduced health effects from direct exposure to NC>2 and HAP;
ecosystem effects; or visibility impairment.
**As discussed in section 6.3, the SC-CCh estimates are calculated with four different values of a one ton reduction.
6.5 Unquantified Co-benefits
The monetized co-benefits estimated in this RIA reflect a subset of co-benefits attributable
to the health effect reductions associated with ambient fine particles and ozone. Data, time, and
resource limitations prevented the EPA from quantifying the impacts to, or monetizing the co-
benefits from several important benefit categories, including co-benefits associated with
exposure to several HAP and NO2, as well as ecosystem effects, and visibility impairment 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 6-10.
6-37
-------
Table 6-10. Unquantified Health and Welfare Co-benefits Categories
Category
Specific Effect
Effect Has Effect Has
Been Been More Information
Quantified Monetized
Improved Human Health
Reduced incidence of
morbidity from exposure
toNO2
Reduced incidence of
morbidity from exposure
toSO2
Asthma hospital admissions (all ages)
Chronic lung disease hospital admissions (age >
65)
Respiratory emergency department visits (all
ages)
Asthma exacerbation (asthmatics age 4-18)
Acute respiratory symptoms (age 7-14)
Premature mortality
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
Respiratory hospital admissions (age > 65)
Asthma emergency department visits (all ages)
Asthma exacerbation (asthmatics age 4-12)
Acute respiratory symptoms (age 7-14)
Premature mortality
Other respiratory effects (e.g., airway
hyperresponsiveness and inflammation, lung
function, other ages and populations)
NO2 ISA1
NO2 ISA1
N02 ISA1
N02 ISA1
N02 ISA1
N02 ISA1-2'3
N02 ISA2-3
SO2 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA1
SO2 ISA1-2-3
SO2 ISA1-2
Improved Environment
Reduced visibility
impairment
Reduced effects on
materials
Reduced effects from PM
deposition (metals and
organic s)
Reduced vegetation and
ecosystem effects from
exposure to ozone
Reduced effects from
acid deposition
Visibility in Class 1 areas
Visibility in residential areas
Household soiling
Materials damage (e.g., corrosion, increased
wear)
Effects on Individual organisms and ecosystems
Visible foliar injury on vegetation
Reduced vegetation growth and reproduction
Yield and quality of commercial forest products
and crops
Damage to urban ornamental plants
Carbon sequestration in terrestrial ecosystems
Recreational demand associated with forest
aesthetics
Other non-use effects
Ecosystem functions (e.g., water cycling,
biogeochemical cycles, net primary productivity,
leaf-gas exchange, community composition)
Recreational fishing
Tree mortality and decline
Commercial fishing and forestry effects
Recreational demand in terrestrial and aquatic
ecosystems
Other non-use effects
Ecosystem functions (e.g., biogeochemical
cycles)
PM ISA1
PM ISA1
PM ISA1-2
PM ISA2
PM ISA2
Ozone ISA1
Ozone ISA1
Ozone ISA1
Ozone ISA2
Ozone ISA1
Ozone ISA2
Ozone ISA2
Ozone ISA2
NOxSOxISA1
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
6-38
-------
Category
Reduced effects from
nutrient enrichment
Reduced vegetation
effects from ambient
exposure to SCh and NOX
Effect Has
Specific Effect Been
Quantified
Species composition and biodiversity in terrestrial
and estuarine ecosystems
Coastal eutrophication
Recreational demand in terrestrial and estuarine
ecosystems
Other non-use effects
Ecosystem functions (e.g., biogeochemical
cycles, fire regulation)
Injury to vegetation from SCh exposure
Injury to vegetation from NOX exposure
Effect Has
Been More Information
Monetized
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
NOxSOxISA2
1We 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.
6.5.1 HAP Impacts
Due to methodology and resource limitations, we were unable to estimate the impacts
associated with changes in emissions of the hazardous air pollutants in this analysis. The EPA's
SAB-HES concluded that "the challenges for assessing progress in health improvement as a
result of reductions in emissions of hazardous air pollutants (HAPs) are daunting...due to a lack
of exposure-response functions, uncertainties in emissions inventories and background levels, the
difficulty of extrapolating risk estimates to low doses and the challenges of tracking health
progress for diseases, such as cancer, that have long latency periods" (U.S. EPA-SAB, 2008). In
2009, the EPA convened a workshop to address the inherent complexities, limitations, and
uncertainties in current methods to quantify the benefits of reducing HAP. Recommendations
from this workshop included identifying research priorities, focusing on susceptible and
vulnerable populations, and improving dose-response relationships (Gwinn etal., 2011).
6.5.2 Additional NO2 Health Co-Benefits
In addition to being a precursor to PlVb.s and ozone, NOx emissions are also linked to a
variety of adverse health effects associated with direct exposure. We were unable to estimate the
health co-benefits associated with reduced NO2 exposure in this analysis. Therefore, this analysis
only quantified and monetized the PlVb.s and ozone co-benefits associated with the reductions in
NO2 emissions.
6-39
-------
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, 2008c) concluded that there is a likely causal relationship between respiratory health
effects and short-term exposure to NO2. These epidemiologic and experimental studies
encompass a number 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.
6.5.3 Additional SO2 Health Co-Benefits
In addition to being a precursor to PIVb.s, SO2 emissions 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 SO2 in this analysis because we do not have air quality
modeling data available. Therefore, this analysis only quantifies and monetizes the PIVh.s co-
benefits associated with the reductions in SO2 emissions.
Following an extensive evaluation of health evidence from epidemiologic and laboratory
studies, the Integrated Science Assessment for Oxides of Sulfur Health Criteria (SO2 ISA)
concluded that there is a causal relationship between respiratory health effects and short-term
exposure to SO2 (U.S. EPA, 2008a). The immediate effect of SO2 on the respiratory system in
humans is bronchoconstriction. Asthmatics are more sensitive to the effects of SO2 likely
resulting from preexisting inflammation associated with this disease. A clear concentration-
response relationship has been demonstrated in laboratory studies following exposures to SO2 at
concentrations between 20 and 100 ppb, both in terms of increasing severity of effect and
percentage of asthmatics adversely affected. Based on our review of this information, we
identified three short-term morbidity endpoints that the SO2 ISA identified as a "causal
relationship": asthma exacerbation, respiratory-related emergency department visits, and
respiratory-related hospitalizations. The differing evidence and associated strength of the
evidence for these different effects is described in detail in the SO2 ISA. The SO2 ISA also
6-40
-------
concluded that the relationship between short-term 862 exposure and premature mortality was
"suggestive of a causal relationship" because it is difficult to attribute the mortality risk effects to
862 alone. Although the 862 ISA stated that studies are generally consistent in reporting a
relationship between 862 exposure and mortality, there was a lack of robustness of the observed
associations to adjustment for other pollutants. We did not quantify these co-benefits due to data
constraints.
6.5.4 Additional NO2 and SO2 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), SO2 and 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. Terrestrial acidification affects
several important ecological services, including declines in habitat for threatened and endangered
species (cultural), declines in forest aesthetics (cultural), declines in forest productivity
(provisioning), and increases in forest soil erosion and reductions in water retention (cultural and
regulating). (U.S. EPA, 2008d)
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
6-41
-------
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 and SO2 will improve the level of visibility throughout the
United States because these gases (and the particles of nitrate and sulfate formed from these
gases) impair visibility by scattering and absorbing light (U.S. EPA, 2009). Visibility is also
referred to as visual air quality (VAQ), and it directly affects people's enjoyment of a variety of
daily activities (U.S. EPA, 2009). Good visibility increases quality of life where individuals live
and work, and where they travel for recreational activities, including sites of unique public value,
such as the Great Smoky Mountains National Park (U. S. EPA, 2009).
6.5.5 Ozone Welfare Co-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.
6.5.6 Visibility Impairment Co-Benefits
Reducing secondary formation of PIVb.s 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
6-42
-------
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 emission
guidelines would be likely to have a significant impact on visibility in urban areas or Class I
areas.
6.6 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.
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, Marnier 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): 23 72-8.
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
6-43
-------
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-fmal-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.
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 (ffic). 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.
6-44
-------
Industrial Economics, Incorporated (ffic). 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.
6-45
-------
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, et al. 2009. Extended Follow-Up
and Spatial Analysis of the American Cancer Society Study Linking Particulate Air Pollution
and Mortality. HEI Research Report, 140, Health Effects Institute, Boston, MA.
Lepeule, J., F. Laden, D. Dockery, and J. Schwartz. 2012. "Chronic Exposure to Fine Particles
and Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009."
Environmental Health Perspectives. 120(7):965-70.
Levy, J.I., S.M. Chemerynski, and J.A. Sarnat. 2005. "Ozone Exposure and Mortality: An
Empiric Bayes Metaregression Analysis." Epidemiology. 16(4):458-68.
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 ofMethylmercury: Committee on
the Toxicological Effects ofMethylmercury.'" 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.
6-46
-------
National Research Council (NRC). 201 la. Climate Stabilization Targets: Emissions,
Concentrations, and Impacts over Decades to Millennia. National Academies Press,
Washington, DC.
National Research Council (NRC) 201 Ib. National Security Implications of Climate Change for
U.S. Naval Forces. National Academies Press. Washington, DC.
National Research Council (NRC, 201 Ic). 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." InternationalJournalofBiometeorology. 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.
6-47
-------
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, BJ. 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.
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 OxidesHealth Criteria (Final Report). National Center for Environmental
Assessment - RTF Division, Research Triangle Park, NC. September. Available at:
. Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 2008b. Final Ozone NAAQSRegulatory
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.
6-48
-------
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 - RTF 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 (FinalReport). EPA-600-R-08-139F. National Center for Environmental
Assessment - RTF 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 PIVh.s-related Mortality. Research Triangle Park, NC. June. Available at:
. Accessed June 4, 2015.
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
6-49
-------
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). 201 Ib. 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). 201 Ic. 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 Paniculate 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 P articulate Matter. EPA-
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:
. 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: <
6-50
-------
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 - RTF 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:
-------
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,
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 AgencyScience 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 AgencyScience 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 AgencyScience 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 AgencyScience 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 AgencyScience Advisory Board (U.S. EPA-SAB). 2008.
Benefits of Reducing Benzene Emissions in Houston, 1990-2020. EPA-COUNCIL-08-001.
July. Available at:
6-52
-------
. Accessed June 4, 2015.
U.S. Environmental Protection AgencyScience Advisory Board (U.S. EPA-SAB). 2009b.
Review of EPA 's Integrated Science Assessment for P articulate Matter (First External
Review Draft, December 2008). EPA-COUNCIL-09-008. May. Available at:
.
Accessed June 4, 2015.
U.S. Environmental Protection AgencyScience Advisory Board (U.S. EPA-SAB). 2009c.
Review of Integrated Science Assessment for P articulate Matter (Second External Review
Draft, July 2009). EPA-C AS AC-10-001. November. Available at:
. Accessed
June 4, 2015.
U.S. Environmental Protection AgencyScience 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.pdf>. Accessed
June 4, 2015.
U.S. Environmental Protection AgencyScience 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-Oil July. Available at:
. Accessed June 4, 2015.
Woodruff, T.J., J. Grille, 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
6-53
-------
6-1
-------
APPENDIX 6A: ESTIMATING THE BENEFIT-PER-TON OF REDUCING PMi.s AND OZONE
PRECURSORS
This appendix reports additional detail regarding our approach for estimating the benefit-
per-ton (BPT) of reducing PlVh.s and ozone precursor emissions. These BPT were applied in
Chapter 6 of this Regulatory Impact Analysis (RIA). Specifically, this appendix describes the
methods for estimating these values for the contiguous U.S. for PlVh.s and ozone precursor NOX
emissions from by the electrical generating unit (EGU) sector in the proposed CSAPR Update
for the 2008 Ozone NAAQS Rule.
6A.1 Overview of Benefit-per-Ton Estimates
As described in the Technical Support Document: Estimating the Benefit per Ton of
Reducing PM2.5 Precursors from 17 Sectors (U.S. EPA, 2013), the general procedure for
calculating average benefit-per-ton coefficients generally follows three steps. As an example, in
order to calculate average benefit-per-ton estimates for the key precursor pollutant emitted from
EGU sources, we:
1. Use air quality modeling to predict changes in ambient concentrations of primary PlVh.s,
nitrate, sulfate, and ozone at a 12km2 grid resolution across the contiguous U.S. that are
attributable to the illustrative control strategy.
2. For each grid cell, estimate the health impacts, and the economic value of these impacts,
associated with the attributable ambient concentrations using the environmental Benefits
Mapping and Analysis Program - Community Edition (BenMAP-CE vl.l).115'116
Aggregate those impacts and economic values to the United States.
3. Divide the national health impacts attributable to each precursor, and the national
monetary value of these impacts, by the amount of associated national precursor
emissions. For example, ozone benefits are divided by ozone-season NOx emissions.
115 When estimating these impacts we apply effect coefficients that relate changes in total PM mass to the risk of adverse health outcomes; we do not apply effect coefficients
that are differentiated by PM species.
116 Previous RIAs have used earlier versions of the BenMAP software. BenMAP-CE vl.l provides results consistent
with earlier versions of BenMAP and is available for download at http://www.epa.gov/auYbenmap/.
6A-1
-------
6A.2 Air Quality Modeling for the Proposed Transport Rule
The EPA ran the Comprehensive Model with Extensions (CAMx) photochemical model
(ENVIRON, 2014) to predict ozone and PIVh.s concentrations. The model predictions for the
2011 base year, the 2017 baseline, and the 2017 illustrative control case were combined with
ambient air quality observations to calculate seasonal mean ozone air quality metrics and annual
mean PM2.5 for the 2017 baseline and 2017 illustrative control case, which were then used as
input for the benefits analysis.
Each of the CAMx model simulations was performed for a nationwide modeling domain117
using a full year of meteorological conditions for 2011. The modeling for 2011 was used as the
anchor point for projecting ozone and annual PM2.5 concentration values for the 2017 base case
and for the 2017 illustrative control scenario using methodologies consistent with the EPA's air
quality modeling guidance (U.S. EPA, 2007). The air quality modeling results for the 2017 base
case served as the baseline for gauging the future year impacts on ozone and annual PM2.5 of the
illustrative control case scenario. The 2017 base case reflects emissions reductions between 2011
and 2017 that are expected to result from regional and national rules including the Cross-State
Air Pollution Rule, the Mercury and Air Toxics Standards (MATS), mobile source rules up
through Tier-3, and various state emissions control programs and consent decrees. The methods
for estimating the EGU emissions for the proposal are described in Chapter 3 of this RIA. The
data indicate that, overall nationwide, EGU emissions with the illustrative control case scenario
would be about 14% lower than the 2017 base case.
As indicated above, the air quality modeling was used to project gridded ozone and
annual PM2.5 concentrations at the 12km by 12km resolution for the 2017 base case and the
illustrative control case scenario modeled for this analysis. The air quality modeling results were
combined with monitored ozone and PM2.5 data to create projected spatial fields of annual PM2.5
and seasonal mean (May through September) 8-hour daily maximum ozone for the 2017 base
case and for the illustrative control scenario. These spatial fields were then used as inputs to
estimate the health co-benefits of the proposed rule as described below.
117 The modeling domain (i.e., region modeled) includes all of the lower 48 states plus adjacent portions of Canada
and Mexico) at a spatial resolution of 12 km.
6A-2
-------
6A.3 National PMi.s Benefit-per-Ton Estimates for EGUs Derived from Air Quality
Modeling of the Proposed Transport Rule
After estimating the 12km by 12km resolution PIVh.s benefits for each of the analysis years
applied in this RIA (i.e., 2017), we aggregated the benefits results nationally.118 In order to
calculate the benefit-per-ton estimates, we divided the regional benefits estimates by the
corresponding emissions, as shown in Table 4A-1. Lastly, we adjusted the benefit-per-ton
estimates for a currency year of 201 IS.119
This method provides estimates of the national average benefit-per-ton for a subset of the
major PIVh.s precursors emitted from EGU sources. For precursor emissions of NOx, there is
generally a non-linear relationship between emissions and formation of PIVh.s. This means that
each ton of NOx reduced would have a different impact on ambient PIVh.s depending on the
initial level of emissions and potentially on the levels of emissions of other pollutants. In
contrast, SO2 is generally linear in forming PIVh.s. For precursors like NOx which form PM2.5
non-linearly, a marginal benefit-per-ton approach would better approximate the specific benefits
associated with an emissions reduction scenario for a given set of base case emissions, because it
would allow the benefit-per-ton to vary depending on the level of emissions reductions and the
baseline emissions levels. However, we do not have sufficient air quality modeling data to
calculate marginal benefit-per-ton estimates for the EGU sector. Therefore, using an average
benefit-per-ton estimate for NOx adds uncertainty to the co-benefits estimated in this RIA.
In this RIA, we estimate emission reductions from EGUs using IPM.120 IPM outputs
provide endogenously projected unit level emissions of SO2, NOx, CO2, Hg, hydrogen chloride
(HC1) from EGUs, but carbon monoxide, volatile organic compounds, ammonia and total
directly emitted PIVh.s and PMio emissions are post-calculated.121 In addition, directly emitted
particle emissions calculated from IPM outputs do not include speciation, i.e. they are only the
total emissions. In order to conduct air quality modeling, directly emitted PIVh.s from EGUs is
118 This aggregation is identified as the shapefile "Report Regions" in BenMAP's grid definitions.
119 Currently, BenMAP does not have an inflation adjustment to 2011$. We ran BenMAP for a currency year of 2010$ and calculated the benefit-per-ton estimates in 2010$. We
then adjusted the resulting benefit-per-ton estimates to 2011 $ using the Consumer Price Index.
120 See Chapter 3 of this PJA for additional information regarding+Ug JntGffnited Planning rVTodel f^^-
121 Detailed documentation of this post-processing is available at
http://www.epa.gov/powersectormodeling/docs/v513MalFile_Methodology.pdf
6A-3
-------
speciated into components during the emissions modeling process based on emission profiles for
EGUs by source classification code. Even though these speciation profiles are not unit-specific,
an emission profile based on the source classification code reflects the fuel and the unit
configuration. Model-predicted concentrations of nitrate and sulfate include both the directly
emitted nitrate and sulfate from speciated PM2.5 and secondarily formed nitrate and sulfate from
emissions of NOx and SO2, respectively. Because this proposed rule is not expected to greatly
affect the levels of directly emitted PM2.5 or SO2, we do not quantify a benefit per-ton value for
either pollutant.
Although it is possible to calculate 95th percentile confidence intervals using the approach
described in this appendix (e.g., U.S. EPA, 201 Ib), we generally do not calculate confidence
intervals for benefit-per-ton estimates because of the additional unquantified uncertainties that
result from the benefit transfer methods, including those related to the transfer of air quality
modeling information. Instead, we refer the reader to Chapter 5 of PM NAAQS RIA (U.S. EPA,
2012a) for an indication of the combined random sampling error in the health impact and
economic valuation functions using Monte Carlo methods. In general, the 95th percentile
confidence interval for the total monetized PM2.5 benefits ranges from approximately -90% to
±180% of the central estimates based on concentration-response functions from Krewski et al.
(2009) and Lepeule et al. (2012). The 95th percentile confidence interval for the health impact
function alone ranges from approximately ±30% for mortality incidence based on Krewski et al.
(2009) and ±46% based on Lepeule et al. (2012). These confidence intervals do not reflect other
sources of uncertainty inherent within the estimates, such as baseline incidence rates, populations
exposed, and transferability of the effect estimate to diverse locations. As a result, the reported
confidence intervals and range of estimates give an incomplete picture about the overall
uncertainty in the benefits estimates.
Table 6A-1 reports the national benefit-per-ton estimates for the EGU sector at discount
rates of 3% and 7% in 2017. Tables 6A-2 reports the incidence per ton estimates (which follows
the same general methodology as for the benefit-per-ton calculations) for the EGU sector in 2017
for the set of health endpoints used to calculate the benefit-per-ton estimates.
6A-4
-------
Table 6A-1. Summary of Regional PMi.s Benefit-per-Ton Estimates Based on Air Quality
Modeling from the Illustrative Control Case in 2017 (2011$)*
Pollutant Discount Rate Benefit per ton
NOX (as PM2.5) 3% $2,100 to $4,700
7% $1,900 to $4,300
* The range of estimates reflects the range of epidemiology studies for avoided premature mortality for PM2.5. All
estimates are rounded to two significant figures. 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 PM25 levels, which
drive population exposure. The monetized benefits incorporate the conversion from precursor emissions to ambient
fine particles. The estimates do not include reduced health effects from direct exposure to ozone, NCh, SCh,
ecosystem effects, or visibility impairment.
Table 6A-2. Summary of Regional PMi.s Incidence-per-Ton Estimates Based on Air
Quality Modeling from the Illustrative Control Case in 2017*
Health Endpoint NOx Incidence per ton
Premature Mortality
Krewski et al. (2009) - adult 0.00024
Lepeule et al. (2012) - adult 0.00054
Woodruff et al. (1997) - infants 0.00001
Morbidity
Emergency department visits for asthma 0.00013
Acute bronchitis 0.00034
Lower respiratory symptoms 0.00436
Upper respiratory symptoms 0.00623
Minor restricted-activity days 0.17575
Lost work days 0.02947
Asthma exacerbation 0.00642
Hospital Admissions, Respiratory 0.00007
Hospital Admissions, Cardiovascular 0.00009
Non-fatal Heart Attacks (age>18)
Peters etal (2001) 0.00028
Pooled estimate of 4 studies 0.00003
* All estimates are rounded to two significant figures. All fine particles are assumed to have equivalent health
effects, but the incidence-per-ton estimates vary depending on the location and magnitude of their impact on PM25
levels, which drive population exposure. The incidence benefit-per-ton estimates incorporate the conversion from
precursor emissions to ambient fine particles.
6A-5
-------
6A.4 Regional Ozone Benefit-per-Ton Estimates
The process for generating the ozone benefit-per-ton estimates is consistent with the
process for PIVh.s. Ozone is not directly emitted, and is a non-linear function of NOx and VOC
emissions. For the purpose of estimating benefit-per-ton for this RIA, we assume that all of the
ozone impacts from EGUs are attributable to NOx emissions. VOC emissions, which are also a
precursor to ambient ozone formation, are insignificant from the EGU sector relative to both
NOx emissions from EGUs and the total VOC emissions inventory. Therefore, we believe that
our assumption that EGU-attributable ozone formation at the regional-level is due to NOx alone
is reasonable.
Similar to PIVh.s, this method provides estimates of the regional average benefit-per-ton.
Due to the non-linear chemistry between NOx emissions and ambient ozone, using an average
benefit-per-ton estimate for NOx adds uncertainty to the ozone co-benefits estimated for the
proposed guidelines.
In the ozone co-benefits estimated in this RIA, we apply the benefit-per-ton estimates
calculated using NOx emissions derived from modeling the the Illustrative Control Case during
the ozone-season only (May to September). Because we estimate ozone health impacts from May
to September only, this approach underestimates ozone co-benefits in areas with longer ozone
seasons such as southern California and Texas. When the underestimated benefit-per-ton
estimate is multiplied by ozone-season only NOx emission reductions, this results in an
underestimate of the monetized ozone co-benefits. Table 6A-3 reports the ozone benefit-per-ton
estimates using ozone-season only NOx for 2017. Table 6A-4 reports the ozone season
incidence-per-ton estimates for 2017.
Table 6A-3. Summary of National Ozone Benefit-per-Ton Estimates Based on Air
Quality Modeling from the Illustrative Control Case in 2017 (2011$)*
Ozone precursor Pollutant Benefit per ton
Ozone season NOX $4,300 to $20,000
* The range of estimates reflects the range of epidemiology studies for avoided premature mortality for ozone. All
estimates are rounded to two significant figures. The monetized benefits incorporate the conversion from NOx
precursor emissions to ambient ozone.
6A-6
-------
Table 6A-4. Summary of National Ozone Incidence-per-Ton Estimates Based on Air
Quality Modeling from Proposed Transport Rule in 2017*
Health Endpoint Incidence per ton
Premature Mortality - adult
Smith et al. (2009) 0.000569
Zanobetti & Schwartz (2008) 0.000956
Morbidity
Hospital Admissions, Respiratory (ages < 2 & > 65) 0.000927
Emergency Room Visits, Respiratory 0.003771
Acute Respiratory Symptoms 2.781551
School Loss Days 0.903847
* All estimates are rounded to two significant figures. The incidence benefit-per-ton estimates incorporate the
conversion from NOx precursor emissions to ambient ozone. These estimates reflect ozone-season NOx emissions.
6A.5 References
Abt Associates, Inc. 2010. "User's Guide: Modeled Attainment Test Software." Available at:
. Accessed June 6, 2015.
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.
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): 23 72-8.
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.
ENVIRON International Corporation. 2014. User's Guide: Comprehensive Air Quality Model
with Extensions, Version 6.1, Novato, CA. April. Available at .
Accessed June 6, 2015.
Fann, N., C.M. Fulcher, and B.J. Hubbell. 2009. "The Influence of Location, Source, and
Emission Type in Estimates of the Human Health Benefits of Reducing a Ton of Air
Pollution." Air Quality and Atmospheric Health. 2:169-176.
6A-7
-------
Fann, N., K.R. Baker, and C.M. Fulcher. 2012. "Characterizing the PM2.5-Related Health
Benefits of Emission Reductions for 17 Industrial, Area and Mobile Emission Sectors Across
the U.S." Environment International. 49:41-151.
Fann, N., K.R. Baker, and C.M. Fulcher. 2013. "The Recent and Future Health Burden of Air
Pollution Apportioned Across 23 U.S. Sectors." Environmental Scientific Technology. 47(8):
3580-3589.
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.
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.
Krewski D., M. Jerrett, R.T. Burnett, R. Ma, E. Hughes, Y. Shi, et al. 2009. Extended Follow-Up
and Spatial Analysis of the American Cancer Society Study Linking Paniculate 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.
Roman, H.A., 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 Scientific 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.
U.S. Environmental Protection Agency (U.S. EPA). 2007. Guidance on the Use of Models and
Other Analyses for Demonstrating Attainment of Air Quality Goals for Ozone, PM2.5, and
Regional Haze. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
Available at .
Accessed June 4, 2015.
U.S. Environmental Protection Agency (U.S. EPA). 201 Ib Regulatory Impact Analysis for the
Final Mercury and Air Toxics Standards. Research Triangle Park, NC: Office of Air Quality
Planning and Standards, Health and Environmental Impacts Division. (EPA document
6A-8
-------
number EPA-452/R-11-011, December). 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. EPA-
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). 2013. 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). 2015. Regulatory Impact Analysis for the
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 <
http://www3 .epa.gov/airquality/cpp/cpp-fmal-rule-ria.pdf >.
6A-9
-------
6B-1
-------
CHAPTER 7: STATUTORY AND EXECUTIVE ORDER REVIEWS
Overview
This section explains the statutory and executive orders applicable to EPA rules, and
discusses EPA's actions taken pursuant to these orders.
7.1 Executive Order 12866: Regulatory Planning and Review
This action is an economically significant regulatory action that was submitted to the
Office of Management and Budget (OMB) for review. Any changes made in response to OMB
recommendations have been documented in the docket. This RIA estimates the costs and
monetized human health and welfare benefits of the proposal. Consistent with Executive Order
(EO) 12866 and EO 13563, the EPA estimated the costs and benefits for three regulatory control
alternatives: the proposed EGUNOx ozone-season emissions budgets and more and less
stringent alternatives. The proposed action would reduce ozone-season NOx emissions from
EGUs in 23 eastern states. Actions taken to comply with the proposed EGU NOx ozone-season
emissions budgets would also reduce emissions of other criteria air pollution and hazardous air
pollution emissions, including annual NOx, and CO2. The benefits associated with these co-
pollutant reductions are referred to as co-benefits, as these reductions are not the primary
objective of the rule.
7.2 Paperwork Reduction Act
The information collection activities in this proposed rule have been submitted for
approval to the Office of Management and Budget (OMB) under the Paperwork Reduction Act
(PRA), 44 U.S.C. 3501 et seq. The Information Collection Request (ICR) document that the EPA
prepared has been assigned EPA ICR number 2527.01. You can find a copy of the ICR in the
docket for the proposed rule, and it is briefly summarized here.
The information generated by information collection activities under CSAPR is used by
the EPA to ensure that affected facilities comply with the emission limits and other requirements.
Records and reports are necessary to enable EPA or states to identify affected facilities that may
not be in compliance with the requirements. The recordkeeping requirements require only the
specific information needed to determine compliance. These recordkeeping and reporting
requirements are established pursuant to CAA sections 110(a)(2)(D) and (c) and 301 (a) (42
7-1
-------
U.S.C. 7410(a)(2)(D) and (c) and 7601 (a)) and are specifically authorized by CAA section 114
(42 U.S.C. 7414). Reported data may also be used for other regulatory and programmatic
purposes. All information submitted to the EPA for which a claim of confidentiality is made will
be safeguarded according to EPA policies in 40 CFR part 2, subpart B, Confidentiality of
Business Information.
All of the EGUs that would be subject to changed information collection requirements
under this proposed rule are already subject to information collection requirements under
CSAPR. Most of these EGUs also are already subject to information collection requirements
under the Acid Rain Program (ARP) established under Title IV of the 1990 Clean Air Act
Amendments. Both CSAPR and the ARP have existing approved ICRs or pending renewals:
EPA ICR Number 2391.03/OMB Control Number 2060-0667 (CSAPR) and EPA ICR Number
1633.16/OMB Control Number 2060-0258 (ARP). The burden and costs of the information
collection requirements covered under the CSAPR ICR are estimated as incremental to the
information collection requirements covered under the ARP ICR. Most of the information used
to estimate burden and costs in this ICR was developed for the existing CSAPR and ARP ICRs.
This proposed rule would change the universe of sources subject to certain information
collection requirements under CSAPR but would not change the substance of any CSAPR
information collection requirements. The burden and costs associated with the proposed changes
in the reporting universe are estimated as reductions from the burden and costs under the existing
CSAPR ICR. (This proposed rule would not change any source's information collection
requirements with respect to the ARP.) The EPA intends to incorporate the burden and costs
associated with the proposed changes in the reporting universe under this rule into the next
renewal of the CSAPR ICR.
Respondents/affected entities: Entities potentially affected by this proposed action are
EGUs in the states of Florida, Kansas, and South Carolina that meet the applicability criteria for
the CSAPR NOx Ozone Season Trading Program in 40 CFR 97.404.
Respondent's obligation to respond: Mandatory (sections 110(a) and 301 (a) of the Clean
Air Act).
Estimated number of respondents: 116 sources in Florida, Kansas, and South Carolina
with one or more EGUs.
Frequency of response: Quarterly, occasionally.
7-2
-------
Total estimated burden: reduction of 14,064 hours (per year). Burden is defined at 5 CFR
1320.3(b).
Total estimated cost: reduction of $1,472,047 (per year), includes reduction of $450,951
operation and maintenance costs.
The burden and cost estimates above reflect the reduction in burden and cost for Florida
sources with EGUs that would no longer be required to report NOx mass emissions and heat
input data for the ozone season to the EPA under the proposed rule and that are not subject to
similar information collection requirements under the Acid Rain Program. Because these EGUs
would no longer need to collect NOx emissions or heat input data under 40 CFR part 75, the
estimates above also reflect the reduction in burden and cost to collect and quality assure these
data and to maintain the associated monitoring equipment.
The EPA estimates that the proposed rule would cause no change in information
collection burden or cost for EGUs in Kansas that would be required to report NOx mass
emissions and heat input data for the ozone season to the EPA or for EGUs in South Carolina
that would no longer be required to report NOx emissions and heat input data for the ozone
season to the EPA. The EGUs in both Kansas and South Carolina already are and would remain
subject to requirements to report NOx mass emissions and heat input data for the entire year to
the EPA under the CSAPR NOx Annual Trading Program, and the requirements related to ozone
season reporting are a subset of the requirements related to annual reporting. Similarly, the EPA
estimates that the proposed rule would cause no change in information collection burden or cost
for EGUs in Florida that are subject to the Acid Rain Program because of the close similarity
between the information collection requirements under CSAPR and under the Acid Rain
Program.
More information on the ICR analysis is included in the docket for the proposed rule.
An agency may not conduct or sponsor, and a person is not required to respond to, a
collection of information unless it displays a currently valid OMB control number. The OMB
control numbers for the EPA's regulations in 40 CFR are listed in 40 CFR part 9.
7-3
-------
7.3 Regulatory Flexibility Act
The EPA certifies that this action will not have a significant economic impact on a
substantial number of small entities under the 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 7 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 proposed rulemaking, it must prepare and make
available an initial regulatory flexibility analysis, unless it certifies that the proposed rule, if
promulgated, will not have a significant economic impact on a substantial number of small
entities (5 U.S.C. 605[b]). Small entities include small businesses, small organizations, and
small governmental jurisdictions.
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.1
This analysis draws on the "parsed" unit-level estimates using IPM results for 2018,2 as well as
ownership, employment, and financial information for the potentially affected small entities
drawn from other resources described in more detail below.
The EPA identified the size of ultimate parent entities by using the Small Business
Administration (SBA) size threshold guidelines.3 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 7-1)
1 Detailed documentation for IPM v.5.15 is available at: http://www.epa.gov/airmarkets/powersectormodeling.html.
2 For this analysis, the 2018 parsed file is used to represent 2017 for the purposes of RIA analysis.
3 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.
7-4
-------
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 SB A
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.4
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.
7.3.1 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 proposed 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 proposal, in comparison to
the baseline.
4 Certain affected EGUs are owned by ultimate parent entities whose primary business is not electric power
generation.
7-5
-------
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 318 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.5 Majority owners of power plants with affected
EGUs were categorized as one of the seven ownership types.6 These ownership types are:
1. Investor-Owned Utility (IOU): Investor-owned assets (e.g., a marketer, independent
power producer, financial entity) and electric companies owned by stockholders, etc.
2. Cooperative (Co-Op): Non-profit, customer-owned electric companies that generate
and/or distribute electric power.
3. Municipal: A municipal utility, responsible for power supply and distribution in a small
region, such as a city.
4. Sub-division: Political subdivision utility is a county, municipality, school district,
hospital district, or any other political subdivision that is not classified as a municipality
under state law.
5. Private: Similar to an investor-owned utility, however, ownership shares are not openly
traded on the stock markets.
6. State: Utility owned by the state.
7. Federal: Utility owned by the federal government.
5 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.
6 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.
7-6
-------
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).7 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 7-1. 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.
7 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
7-7
-------
Table 7-1. SBA Size Standards by NAICS Code
NAICS
Code
221112
221118
221122
221210
238210
324110
325180
325320
331313
333613
424720
486210
522110
522220
523120
523910
523930
524126
525120
525990
541611
551112
NAICS Description
Fossil Fuel Electric Power Generation
Other Electric Power Generation
Electric Power Distribution
Natural Gas Distribution
Electrical Contractors and Other Wiring Installation Contractors
Petroleum Refineries
Other Basic Inorganic Chemical Manufacturing
Pesticide and Other Agricultural Chemical Manufacturing
Alumina Refining and Primary Aluminum Production
Mechanical Power Transmission Equipment Manufacturing
Petroleum and Petroleum Products Merchant Wholesalers
(except Bulk Stations and Terminals)
Pipeline Transportation of Natural Gas
Commercial Banking
Sales Financing
Securities Brokerage
Miscellaneous Intermediation
Investment Advice
Direct Property and Casualty Insurance Carriers
Health and Welfare Funds
Other Financial Vehicles
Administrative Management and General Management
Consulting Services
Offices of Other Holding Companies
SBA Size Standard
750 employees
250 employees
1,000 employees
500 employees
$15 million in revenue
1,500 employees
1,000 employees
500 employees
1,000 employees
500 employees
100 employees
$27.5 million in revenue
$550 million in assets
$38.5 million in revenue
$38.5 million in revenue
$38.5 million in revenue
$38.5 million in revenue
1,500 employees
$32.5 million in revenue
$32.5 million in revenue
$15 million in revenue
$20.5 million in revenue
Note: Based on size standards effective at the time the EPA conducted this analysis (SBA size standards, effective
July 14, 2014)
Source: SBA, 2014
The EPA compared the relevant entity size criterion for each ultimate parent entity to the
SBA threshold value noted in Table 7-1. 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.8 In parallel, the EPA also considered estimated revenues from affected EGUs
based on analysis of parsed-file estimates for the proposal. The EPA assumed that the
ultimate parent entity revenue was the larger of the two revenue estimates. In limited
! Estimates of sales were used in lieu of revenue estimates when revenue data was unavailable.
7-8
-------
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 16 potentially affected EGUs owned by 7 small entities included in the EPA's
Base Case.
7.3.2 Overview of Analysis and Results
This section presents the methodology and results for estimating the impact of the
proposed CSAPR update to small entities in 2017 based on the following endpoints:
annual economic impacts of the proposed CSAPR update on small entities, and
ratio of small entity impacts to revenues from electricity generation.
7.3.2.1 Methodology for Estimating Impacts of the Proposed CSAPR update on Small
Entities
An entity can comply with the proposed CSAPR update through some combination of the
following: optimizing existing SCR and SNCRs, 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:
7-9
-------
^Compliance A ^Operating+Retrofit ' A (--Fuel ' A ^-Allowances ' A ^-Transaction * Z\ K.
where C represents a component of cost as labeled, and// R represents the value of foregone
electricity generation, calculated as the difference in revenues between the base case and the
proposed CSAPR update in 2017.
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 proposed
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 proposed 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 proposed 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 proposed CSAPR update, the EPA identified units that install control
technology under the proposed CSAPR update, 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 units projected to optimize and/or turn
on existing idled SCR or SNCR.
(2) Sale or purchase of allowances: To estimate the value of allowances holdings,
allocated allowances were subtracted from projected emissions, and the difference
7-10
-------
was then multiplied by $1,300 per ton. $1,300 per ton is the marginal cost of NOx
reductions used to set the proposed budgets in the proposed CSAPR update.
While this is a reasonable approximation, it is possible that the actual allowance
price could be lower. Units were assumed to purchase or sell allowances to
exactly cover their projected emissions under the proposed CSAPR update.
(3) Fuel costs: The change in fuel expenditures under the proposed CSAPR update
was estimated by taking the difference in projected fuel expenditures between the
IPM estimates for the proposed CSAPR update and the base case.
(4) Value of electricity generated: To estimate the value of electricity generated, the
projected level of electricity generation is multiplied by the regional-adjusted
retail electricity price ($/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 (TPP). 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.
7.3.2.2 Results
The potential impacts of the proposed CSAPR update on small entities are summarized in
Table 7-2. All costs are presented in $2011. The EPA estimated the annual net compliance cost
to small entities to be approximately - $38.3 million in 2017 or savings of $38.3 million. The fact
that the net compliance costs for all entities are actually net savings does not mean that each
small entity would benefit from the proposal to update CSAPR. The net savings are driven by
entities that are able to increase their revenues by increasing generation.
7-11
-------
Table 7-2. Projected Impact of the Proposed CSAPR Update on Small Entities in 2017
ECU
Ownership
Type
Cooperative
Investor-Owned
Utility
Municipal
Private
Total
Number of
Potentially
Affected
Entities
2
1
2
2
7
Total Net
Compliance
Cost ($2011
millions)
$0.2
-$38
-$0.1
-$0.4
-$38.3
Number of Small
Entities with
Compliance Costs
>1% of Generation
Revenues
0
0
*
0
*
Number of Small
Entities with
Compliance Costs
>3% of Generation
Revenues
0
0
*
0
*
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 proposed CSAPR update.
* One entity may experience compliance costs greater than 1 or 3 percent of generation revenues in 2017.
Since this entity is not projected to operate in the base case, we are unable to compare the estimated
compliance costs to base case generation revenues.
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 7 small entities considered in this analysis, 1 entity may experience compliance
costs greater than 1 or 3 percent of generation revenues in 2017. Since this entity is not projected
to operate in the base case, we are unable to compare the estimated compliance costs to base case
generation revenues. However, we note that this entity is located in a cost of service market,
where typically we expect entities should be able to recover all of their costs of complying with
the proposed CSAPR update. Entities that experience negative net costs under the proposed
CSAPR update are excluded from these totals. The EPA has concluded that there is no
7-12
-------
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 7-2.
The distribution across entities of economic impacts as a share of base case revenue is
summarized in Table 7-3. 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.
Note that one municipal entity is not projected to operate in the base case, and we are therefore
unable to compare the estimated compliance costs to generation revenues. We estimate a positive
compliance cost for this entity.
Table 7-3. Summary of Distribution of Economic Impacts of the Proposed CSAPR Update
on Small Entities in 2017
ECU Ownership
Type
Cooperative
Investor-owned utility
Municipal*
Private
Capacity- Weighted
Average Economic
Impacts as a % of
Generation Revenues
0.3%
-60.1%
-0.5%
-2.3%
Min
0.1%
-60.1%
-0.5%
-4.2%
Max
0.5%
-60.1%
N/A*
-1.5%
*Note: One municipal entity is not projected to operate in the base case, and we are therefore unable to compare
the estimated compliance costs to generation revenues.
Source: IPM analysis
7-13
-------
The separate components of annual costs to small entities under the proposed CSAPR
update are summarized in Table 7-4. The most significant components of incremental cost to
these entities under the proposed CSAPR update are due to lower electricity revenues and
increased fuel costs. Fuel costs increase over all ownership groups except the ones under the
ownership type "Private" because an entity with the second largest generation under "Private" is
projected to cut its generation by 25 percent under the proposed CSAPR update, which translates
to lower fuel costs for the whole group. Additionally, increases in electricity generation revenue,
shown as cost savings or negative costs are experienced by cooperative, investor-owned utility,
municipal and subdivision entities. This is due largely to the projected increase in electricity
prices under the proposed CSAPR update. Among the private category, however, reduced
generation by the one entity with a large share of generation leads to higher net costs for the
entire category. Our data suggests this entity owns a group of combined cycle units and which
are presumably marginal units in their respective load segments under the base case.
Table 7-4. Incremental Annual Costs under the Proposed CSAPR Update Summarized by
Ownership Group and Cost Category in 2017 (2011$ millions)
ECU
Ownership
Type
Cooperative
Investor-
Owned Utility
Municipal
Private
Operating
Cost
$0.1
$1.9
$0.1
-$0.5
Net Purchase
of Allowances
$0.2
$0.0
$0.0
-$0.3
Fuel Cost
-$0.5
$26.8
-$0.2
-$3.9
Lost
Electricity
Revenue
$0.4
-$66.7
-$0.1
$4.4
Administrative
Cost
$0.00
$0.00
$0.00
$0.00
Source: IPM analysis.
7.3.3 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 7 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
7-14
-------
1 percent of revenues have some potential for significant impact resulting from implementation
of the proposed CSAPR update. However, as noted above, it is the EPA's position that because
this entity does not operate in a competitive market environment, they should generally be able
to pass the costs of complying with the proposed CSAPR update on to rate-payers.
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 23 states for which the EPA is proposing FIPs.
7.4 Unfunded Mandates Reform Act
Title II of the UMRA of 1995 (Public Law 104-4)(UMRA) establishes requirements for
federal agencies to assess the effects of their regulatory actions on state, local, and Tribal
governments and the private sector. Under Section 202 of the UMRA, 2 U.S.C. 1532, 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
7-15
-------
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 proposal, the EPA is not directly establishing any
regulatory requirements that may significantly or uniquely affect small governments, including
Tribal governments. Thus, under the proposed 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 proposed CSAPR update on
government entities, however. This analysis does not examine potential indirect economic
impacts associated with the proposed 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.
7.4.1 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 proposed 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 proposal, 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 proposed CSAPR update
region. The EPA then used IPM-parsed output to associate these plants with individual
generating units. The EPA identified nine municipality-owned utilities that are potentially
affected by the proposed CSAPR update.
7-16
-------
7.4.2 Overview of Analysis and Results
After identifying potentially affected government entities, the EPA estimated the impact
of the proposed 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 proposed CSAPR update is composed
of compliance and administrative costs. This section outlines the compliance and administrative
costs for the nine potentially affected government-owned units in the proposed CSAPR update
region.
7.4.2.1 Methodology for Estimating Impacts of the proposed CSAPR update on
Government Entities
An entity can comply with the proposed CSAPR update through any combination of the
following: optimizing existing SCR and SNCRs, 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:
(_ Compliance Zl C- Operating+Retrofit ' /i (^Fuel ' A ^-Allowances ' A (- Transaction * Z\ K.
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
proposed 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
7-17
-------
the projected increase in electricity prices under the proposed CSAPR update, 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 proposed
CSAPR update to estimate compliance cost at the unit level. These costs were then summed
for each small 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 electricity generation revenues under the proposed 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 proposed CSAPR update, the EPA identified units that install control
technology under the proposed CSAPR update 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 optimize and/or turn on existing
idled SCR or 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 $1,300 per ton. $1,300 per ton is the marginal annualized
cost of NOx reductions used to set the proposed budgets. While this is a
reasonable approximation, it is possible that the actual allowance price could be
lower. Units were assumed to purchase or sell allowances to exactly cover their
projected emissions under the proposed CSAPR update.
7-18
-------
(3) Fuel costs: The change in fuel expenditures under the proposed CSAPR update
was estimated by taking the difference in projected fuel expenditures between the
proposed CSAPR update and the base case.
(4) Value of electricity generated: To estimate the value of electricity generated, the
projected level of electricity generation is multiplied by the regional-adjusted
retail electricity price ($/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 (TPP). 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.
7.4.2.2 Results
A summary of economic impacts on government owned entities is presented in
Table 7-5. 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 negative in 2014.9 Note that we expect the proposed CSAPR update to
potentially have an impact on only one category of government-owned entities
(municipality-owned entities).
All costs are reported in 2011 dollars.
7-19
-------
Table 7-5. Summary of Potential Impacts on Government Entities under the Proposed
CSAPR Update in 2017
ECU Ownership
Type
Municipal
Total
Potentially
Affected Entities
9
9
Projected
Annualized
Costs ($2011
millions)
-$4.6
-$4.6
Number of
Government
Entities with
Compliance
Costs >1% of
Generation
Revenues
4
4
Number of
Government
Entities with
Compliance
Costs >3% of
Generation
Revenues
2
2
Note: The total number of entities with costs greater than 1 percent or 3 percent of revenues includes only entities
experiencing positive costs, and includes one entity for which we are unable to estimate base case generation
revenues due to projected closure in the base case. A negative cost value implies that the group of entities
experiences a net savings under the proposed CSAPR update.
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
measure is greater than 1 percent.10 The EPA projects that four government entities will have
compliance costs greater than 1 percent of revenues from electricity generation in 2017. One
municipal entity is not projected to operate in the base case, and we are therefore unable to
compare the estimated compliance costs to generation revenues. We include this entity in the
>3% category. 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 either a decrease in electricity
revenue or increased fuel cost, coupled with a relatively low base case revenue estimate. The
EPA notes that increased fuel costs are often passed through to rate-payers as common practice
in many areas of the U.S. due to fuel adder arrangements instituted by state public utility
commissions. Entities that are projected to experience negative compliance costs under the
proposed CSAPR update are not included in those totals. This approach is more indicative of a
10 Neither the costs nor the revenues of units that retire under the proposed CSAPR update are included in this
portion of the analysis. Because these units are better off retiring under the proposed CSAPR update than continuing
operation, the true cost of the rule on these units is not represented by our modeling. The true cost of the proposed
CSAPR update for these units is the differential between their costs in the base case and the costs of meeting their
customers' demand under the rule.
7-20
-------
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 proposed 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-owned entities, the capacity-weighted average economic impact as a
share of base case revenue is 4.6 percent. This average is heavily influenced by an outlier, for
which the maximum economic impact as a share of base case revenues is 470 percent. This
entity holds two small combustion turbines that are not projected to make any operational
changes (e.g., turn on controls), but rather is projected to operate more under the proposed
CSAPR update in 2017 than it was projected to operate in the base case. The relatively large
impact estimate on a percent basis results from the low level of generation and thus revenue in
the base case. This is the only entity that experiences economic impacts that are significantly
higher than the capacity-weighted average for this group. The next highest economic impact in
this group is 2.1 percent.
Additionally, four of the nine entities are projected to experience a negative economic
impact as a share of base case revenue, which implies that this group of four entities experiences
a net savings under the proposed CSAPR update.
The various components of annual incremental cost under the proposed CSAPR update to
government entities are summarized in Table 7-6. In 2017, municipals are a net purchaser of
allowances, and experience both an increase in fuel expenditures and an increase in electricity
revenue under the proposed CSAPR update. Incremental fuel costs are positive because these
entities are projected to increase generation and face higher fuel prices. Overall, increases in total
electricity revenue by government entities under the proposed CSAPR update exceed the
increases in fuel and operating costs.
7-21
-------
Table 7-6. Incremental Annual Costs under the Proposed CSAPR Update Summarized by
Ownership Group and Cost Category (2011$ millions) in 2017
ECU
Ownership
Type
Municipal
Retrofit +
Operating
Cost
$2.7
Net Purchase
of Allowances
$0.7
Fuel Cost
$1.4
Lost
Electricity
Revenue
-$9.4
Administrative
Cost
$0.0
Source: IPM analysis
7.4.3 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 -$4.6 million in 2017 or a net
savings of $4.6 million. This does not mean that each government entity will experience net
savings as the overall net savings is driven by some entities garnering savings. Of the nine
government entities considered in this analysis, four 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
proposed 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.
7.5 Executive Order 13132: Federalism
This action does not have federalism implications. It will not have substantial direct
effects on the states, on the relationship between the national government and the states, or on
the distribution of power and responsibilities among the various levels of government.
7-22
-------
7.6 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments
This action has tribal implications as specified in Executive Order 13175.
However, it will neither impose substantial direct compliance costs on federally
recognized tribal governments, nor preempt tribal law.
This action proposes to implement EGU NOx ozone-season emissions
reductions in 23 eastern states. However, at this time, none of the existing or
planned EGUs affected by this rule are owned by tribes or located in Indian
country. This action may have tribal implications if a new affected EGU is built in
Indian country. Additionally, tribes have a vested interest in how this proposed
rule would affect air quality.
In developing CSAPR, which was promulgated on July 6, 2011 to address
interstate transport of ozone pollution under the 1997 ozone NAAQS,11 the EPA
consulted with tribal officials under the EPA Policy on Consultation and
Coordination with Indian Tribes early in the process of developing that regulation
to permit them to have meaningful and timely input into its development. A
summary of that consultation is provided in 76 FR 48346 (August 8, 2011).
The EPA received comments from several tribal commenters regarding
the availability of CSAPR allowance allocations to new units in Indian country.
The EPA responded to these comments by instituting Indian country new unit set-
asides in the final CSAPR. In order to protect tribal sovereignty, these set-asides
are managed and distributed by the federal government regardless of whether
CSAPR in the adjoining or surrounding state is implemented through a FIP or
SIP. While there are no existing affected EGUs in Indian country covered by this
proposal, the Indian country set-asides will ensure that any future new units built
in Indian country will be able to obtain the necessary allowances. This proposal
11 CSAPR also addressed interstate transport of fine paniculate matter (PM2.5) under the 1997 and 2006 PM2.5
NAAQS.
7-23
-------
maintains the Indian country new unit set-aside and adjusts the amounts of
allowances in each set-aside according to the same methodology of the original
CSAPR rule.
The EPA has informed tribes of our development of this proposal through
a National Tribal Air Association - EPA air policy conference call (January 29,
2015). The EPA plans to further consult with tribal officials under the EPA Policy
on Consultation and Coordination with Indian Tribes early in the process of
developing this regulation to permit them to have meaningful and timely input
into its development. The EPA will facilitate this consultation before finalizing
this proposed rule.
As required by section 7(a), the EPA's Tribal Consultation Official has
certified that the requirements of the executive order have been met in a
meaningful and timely manner. A copy of the certification is included in the
docket for the proposed rule.
7.7 Executive Order 13045: Protection of Children from Environmental Health &
Safety Risks
The EPA interprets EO 13045 (62 FR 19885, April 23, 1997) as applying to those
regulatory actions that concern health or safety risks, such that the analysis required under
section 5-501 of the Order has the potential to influence the regulation. This action is not subject
to EO 13045 because it does not involve decisions on environmental health or safety risks that
may disproportionately affect children. The EPA believes that the ozone-related benefits, PlVb.s-
related co-benefits, and CO2-related co-benefits would further improve children's health.
7.8 Executive Order 13211: Actions that Significantly Affect Energy Supply,
Distribution, or Use
This action, which is a significant regulatory action under EO 12866, is likely to have a
significant effect on the supply, distribution, or use of energy. The EPA notes that one aspect of
this proposal that may affect energy supply, disposition or use is the EPA's proposing and taking
comment on a range of options with respect to use of 2015 vintage and 2016 vintage CSAPR
NOx ozone-season allowances for compliance with 2017 and later ozone-season requirements.
7-24
-------
The EPA has prepared a Statement of Energy Effects for the proposed regulatory control
alternative as follows. We estimate a much less than 1 percent change in retail electricity prices
on average across the contiguous U.S. in 2017, and a much less than 1 percent change in coal-
fired electricity generation in 2017 as a result of this rule. The EPA projects that utility power
sector delivered natural gas prices will change by less than 1 percent in 2017. For more
information on the estimated energy effects, please see chapter 5 of this RIA.
7.9 National Technology Transfer and Advancement Act
The proposed rulemaking does not involve technical standards.
7.10 Executive Order 12898: Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations
The EPA believes the human health or environmental risk addressed by
this proposal will not have potential disproportionately high and adverse human
health or environmental effects on minority, low-income or indigenous
populations.
The EPA notes this action proposes to update CSAPR to reduce interstate
ozone transport with respect to the 2008 ozone NAAQS. The rule uses EPA's
authority in CAA section 110(a)(2)(d) to reduce (nitrogen oxides) NOx pollution
that significantly contributes to downwind ozone nonattainment or maintenance
areas. As a result, the rule will reduce exposures to ozone in the most-
contaminated areas (i.e., areas that are not meeting the 2008 ozone National
Ambient Air Quality Standards (NAAQS)). In addition, the rule separately
identifies both nonattainment areas and maintenance areas. This requirement
reduces the likelihood that areas close to the level of the standard will exceed the
current health-based standards in the future. The EPA proposes to implement
these emission reductions using the CSAPR EGU NOx ozone-season emissions
trading program with assurance provisions.
The EPA recognizes that many environmental justice communities have
voiced concerns in the past about emission trading and the potential for any
emission increases in any location. The EPA believes that CSAPR mitigated these
concerns and that this proposal, which applies the CSAPR framework to reduce
7-25
-------
interstate ozone pollution and implement these reductions, will also minimize
community concerns. The EPA seeks comment from communities on this
proposal.
Ozone pollution from power plants have both local and regional
components: part of the pollution in a given locationeven in locations near
emission sourcesis due to emissions from nearby sources and part is due to
emissions that travel hundreds of miles and mix with emissions from other
sources.
It is important to note that the section of the Clean Air Act providing
authority for this rule, section 110(a)(2)(D), unlike some other provisions, does
not dictate levels of control for particular facilities. CSAPR allows sources to
trade allowances with other sources in the same or different states while firmly
constraining any emissions shifting that may occur by requiring a strict emission
ceiling in each state (the assurance level). In addition, assurance provisions in the
rule outline the allowance surrender penalties for failing to meet the assurance
level; there are additional allowance penalties as well as financial penalties for
failing to hold an adequate number of allowances to cover emissions.
This approach reduces EGU emissions in each state that significantly
contribute to downwind nonattainment or maintenance areas, while allowing
power companies to adjust generation as needed and ensure that the country's
electricity needs will continue to be met. The EPA maintains that the existence of
these assurance provisions, including the penalties imposed when triggered, will
ensure that state emissions will stay below the level of the budget plus variability
limit.
In addition, all sources must hold enough allowances to cover their
emissions. Therefore, if a source emits more than its allocation in a given year,
either another source must have used less than its allocation and be willing to sell
some of its excess allowances, or the source itself had emitted less than its
allocation in one or more previous years (i.e., banked allowances for future use).
In summary, the CSAPR minimizes community concerns about localized
hot spots and reduces ambient concentrations of pollution where they are most
7-26
-------
needed by sensitive and vulnerable populations by: considering the science of
ozone transport to set strict state emissions budgets to reduce significant
contributions to ozone nonattainment and maintenance (i.e., the most polluted)
areas; implementing air quality-assured trading; requiring any emissions above
the level of the allocations to be offset by emission decreases; and imposing strict
penalties for sources that contribute to a state's exceedance of its budget plus
variability limit. In addition, it is important to note that nothing in this final rule
allows sources to violate their title V permit or any other federal, state, or local
emissions or air quality requirements.
Finally, it is also important to note that CAA section 110(a)(2)(d), which
addresses transport of criteria pollutants between states, is only one of many
provisions of the CAA that provide EPA, states, and local governments with
authorities to reduce exposure to ozone in communities. These legal authorities
work together to reduce exposure to these pollutants in communities, including
for minority, low-income, and tribal populations, and provide substantial health
benefits to both the general public and sensitive sub-populations.
The EPA has informed communities of our development of this proposal through an
Environmental Justice community call (January 28, 2015) and a National Tribal Air Association
- EPA air policy conference call (January 29, 2015). The EPA plans to further consult with
communities early in the process of developing this regulation to permit them to have
meaningful and timely input into its development. The EPA will facilitate this engagement
before finalizing the proposed rule.
7-27
-------
-------
-------
CHAPTER 8: 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.
The EPA is proposing EGU NOx ozone season emissions budgets for 23 states.1 The emissions
reductions evaluated in the proposal reflect EGU NOx reduction strategies that are achievable for
the 2017 ozone season. The EPA proposes to quantify EGU NOx ozone-season emissions
budgets reflecting EGU NOx reduction strategies that are widely available at a uniform
annualized cost of $1,300 per ton (2011$). For the RIA, in order to implement the OMB Circular
A-4 requirement to assess one less stringent and one more stringent alternative to the proposal,
the EPA is also analyzing EGU NOx ozone season emissions budgets reflecting NOx reduction
strategies that are widely available at a uniform cost of $500 per ton (2011$) and strategies that
are widely available at a uniform cost of $3,400 per ton (2011$). This chapter summarizes these
results.
8.1 Results
As shown in Chapter 5, the estimated annualized costs to implement the proposal, as
described in this document, are approximately $93 million (2011 dollars). As shown in Chapter
6, the total estimated combined benefits from implementation of the proposal are approximately
$660 to $1,300 million in 2017 (2011 dollars, based on a discount rate of 3 percent and 7 percent
for health benefits (ozone benefits and PIVb.s co-benefits), a range of 2.5% to 5% for climate co-
benefits, and rounded to two significant figures). EPA can thus calculate the net benefits of the
proposal by subtracting the estimated annualized costs from the estimated benefits in 2017. The
net benefits of the proposal are approximately $600 to $1,100 million (based on air quality
benefits discounted at 3 percent, the central estimate of CO2 co-benefits, and annualized cost
estimates) or $580 to $1,100 million (based on air quality benefits discounted at 7 percent, the
central estimate of CO2 co-benefits, and annualized cost estimates). Therefore, the EPA expects
that implementation of this rule, based solely on economic efficiency criteria, will provide
1 Alabama, Arkansas, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Mississippi,
Missouri, New Jersey, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, Tennessee, Texas, Virginia,
West Virginia, and Wisconsin.
8-1
-------
society with a significant net gain in social welfare, notwithstanding the expansive set of health
and environmental effects we were unable to quantify. Further quantification of acidification-,
and eutrophication-related impacts would increase the estimated net benefits of the rule. Table
8-1 presents a summary of the benefits, costs, and net benefits of the proposal and also the more
and less stringent alternatives.
Table 8-1. Total Costs, Total Monetized Benefits, and Net Benefits of the Proposal and
More or Less Stringent Alternatives for 2017 for U.S. (millions of 2011$)a'b'c
Proposal More Stringent Less Stringent
Climate Co-Benefits $23 $23 $27
Air Quality Health Benefits $670 to $1200 $690 to $1,300 $190 to $340
Total Benefits $700 to $1200 $720 to $1300 $210 to $360
Annualized Compliance Costs $93 $96 $4.7
Net Benefits $600 to $1100 $620 to $1200 $210 to $360
Non-Monetized Benefits'1 Non-quantified climate benefits
Reductions in exposure to ambient NC>2 and SC>2
Ecosystem benefits assoc. with reductions in emissions of NOx, SC>2, and PM
Visibility impairment
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 6 for additional detail and explanation. The costs presented in this table reflect compliance costs annualized
at 4.77 percent discount rate possible and monitoring, recordkeeping, and reporting costs. See Chapter 5 for
additional detail and explanation.
0 All costs and benefits are rounded to two significant figures; columns may not appear to add correctly.
d 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 proposed rule. These
alternatives are illustrative of the cost and benefit impacts of varying program stringency. They
are designed to show the effects of more stringent and less stringent NOx reduction requirements
in a regulatory structure that is otherwise the same as the proposed NOx emissions budgets. Air
8-2
-------
quality modeling was not conducted for the proposal or the more and less stringent alternatives.
EPA applied a benefit-per-ton approach appropriate for deriving benefits for the evaluated
regulatory control alternatives, as described in Chapter 6. The proposal and alternatives'
compliance costs are estimated using the IPM model, as described in Chapter 5. Table 8-2
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 proposed rule and the
more and less stringent alternatives.
Table 8-2. Projected 2017* Reductions in Emissions of NOx, SOi, and CCh with the
proposed NOx Emissions Budgets and More or Less Stringent Alternatives
(Tons)
NOx (annual)
NOx (ozone season)
SO2 (annual)
CO2 (annual short tons)
Proposal
90,000
85,000
1,200
660,000
More Stringent
Alternative
93,000
87,000
1,200
710,000
Less Stringent
Alternative
24,000
24,000
1,100
770,000
* 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.
8.2 Net Present Value of a Stream of Costs and Benefits
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 three 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, and
(iii) adjusting the cost of illness for non-fatal heart attacks to adjust for lags in follow up costs
are all appropriate. We explain our discounting of benefits in Chapter 6 of the RIA, specifically
8-3
-------
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 5. 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 rulemaking requiring emissions
reductions. 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 are limited by
underlying assumptions and data. Calculating a present value (PV) of the stream of future
benefits also poses special challenges, which we describe below. In addition, the method requires
definition of the length of that future time period, 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 5.
Further, because we do not know when a facility will stop using a control measure or
change to another measure based on economic or other reasons, the EPA assumes the control
equipment and measures applied in the proposed option, and in the more and less stringent
options, remain in service for their full useful life. As a result, the annualized cost of controls in
a single future year is the same throughout the lifetimes of control measures analyzed, allowing
the EPA to compare the annualized control costs with the benefits in a single year for consistent
comparison.
The EPA's RIAs for air quality rules generally report the estimated net benefits of
improved air quality for a single year. The estimated NPV can better characterize the stream of
benefits and costs over a multi-year period. However, calculating the PV of improved air quality
is generally quite data-intensive and costly. Further, the results are sensitive to assumptions
regarding the time period over which the stream of benefits is discounted.
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
8-4
-------
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 approach2, 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 to model air quality levels for each
year and challenging 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.
2 The pulse approach assumes changes in air pollution in a single year and affects mortality estimates over a 20-year
period.
8-5
-------
8-1
-------
------- |