i _ V

%	G°°

PRCrt^

Regulatory Impact Analysis for the Proposed
Repeal of Greenhouse Gas Emissions Standards
for Fossil Fuel-Fired Electric Generating Units


-------

-------
EPA-452/R-25-002
June 2025

Regulatory Impact Analysis for the Proposed Repeal of Greenhouse Gas Emissions Standards for

Fossil Fuel-Fired Electric Generating Units

U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Health and Environmental Impacts Division
Research Triangle Park, NC


-------
CONTACT INFORMATION

This document has been prepared by staff from the Office of Air and Radiation, U.S.
Environmental Protection Agency. Questions related to this document should be addressed to the
Air Economics Group in the Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Office of Air and Radiation, Research Triangle Park, North Carolina 27711
(email: OAQPSeconomics@epa.gov).

ACKNOWLEDGEMENTS

In addition to U.S. EPA staff from the Office of Air and Radiation, personnel from the Office of
Policy of the U.S. Environmental Protection Agency contributed data and analysis to this
document.


-------
TABLE OF CONTENTS

Table of Contents	i

List of Tables	hi

List of Figures	iv

1	Executive Summary	1-1

1.1	Introduction	1-1

1.2	Compliance Cost Savings	1-2

1.3	Emissions Changes of the Regulated Pollutant	1-3

1.4	Economic Impacts	1-3

1.5	Net Benefits from the Proposed Action	1-5

1.6	References	1-6

2	Introduction and Background	2-1

2.1	Introduction	2-1

2.2	Purpose of RIA	2-1

2.3	Overview of Regulatory Impact Analysis	2-2

2.3.1	Analysis from 2024 CPS RIA	2-2

2.3.2	Analysis in this RIA	2-3

2.4	References	2-5

3	Compliance Costs, Emissions, and Energy Impacts	3-1

3.1	Introduction	3-1

3.2	Economic Impacts	3-1

3.2.1	Representing the Proposed Action with Modeling from 2024 CPS RIA	3-2

3.2.2	Emissions Changes Assessment	3-3

3.2.3	Monitoring, Reporting, and Recordkeeping Costs	3-6

3.2.4	Compliance, Real Resource, and Social Cost Estimates	3-7

3.2.5	Impacts on Fuel Use, and Prices	3-12

3.2.6	Total Compliance Costs	3-13

3.2.7	References	3-16

4	Benefits Analysis	4-1

4.1	Introduction	4-1

4.2	PM2.5 and Ozone-Related Human Health Benefits	4-1

4.3	References	4-9

5	Social Costs and Economic Impacts	5-1

5.1	Energy Market Impacts	5-1

5.2	Economy-wide Social Costs and Economic Impacts	5-1

5.2.1	Linking IPM Partial Equilibrium Model to SAGE CGE Model	5-3

5.2.2	Results	5-4

5.2.3	Limitations	5-13

5.3	Small Entity Analysis	5-13

5.4	LaborImpacts	5-13

5.4.1	Overview of Methodology	5-13

5.4.2	Overview of Power Sector Employment	5-14

5.4.3	Projected Sectoral Employment Changes	5-14

5.4.4	Conclusions	5-16

5.5	References	5-18

l


-------
6 Comparison of Benefits and Costs	6-1

6.1	Introduction	6-1

6.2	Methods	6-1

6.3	Results	6-2

6.4	Uncertainties and Limitations	6-5

li


-------
LIST OF TABLES

Table

1-1

Table

1-2

Table

1-3

Table

3-1

Table

3-2

Table

3-3

Table

3-4

Table

3-5

Table

3-6

Table

3-7

Table

3-8

Table

4-1

Table

4-2

Table

4-3

Table

4-4

Table

5-1

Table

5-2

Table

5-3

Table

5-4

Table

5-5

Table

6-1

Table

6-2

Present Value and Equivalent Annualized Value Estimates of Compliance Cost Savings (billion 2024

dollars, discounted to 2025)	 1-3

Carbon Dioxide Emissions Changes3	1-3

Summary of Certain Energy Market Impacts	1-4

EGU Annual CO2 Emissions and Emission Changes (million metric tons)	3-3

EGU Annual Emissions and Emissions Changes for NOx, SO2, Hg, and PM2.5, and Ozone Season

NOx	3-5

Summary of State and Industry Annual Respondent Cost of Reporting and Recordkeeping

Requirements (million 2024 dollars)	3-7

National Power Sector Costs (billion 2024$)	3-8

Real Resource Costs Avoided under the Proposed Action (billion 2024 dollars)	3-10

National Impacts on Fuel Prices, Fuel Consumption, and Electricity Prices	3-13

Total Costs of the Proposed Action (billion 2024 dollars, undiscounted)	3-14

Present Value and Equivalent Annualized Value Estimates of Costs (billion 2024 dollars, discounted

to 2025)a	3-15

Estimated PM2 5 and Ozone-Related Avoided Premature Mortality a	4-5

Estimated Economic Value of Avoided PM2 5 and Ozone-Related Attributable Premature Mortality

and Illnesses for the Proposed Action (95 percent confidence interval; billion 2024 dollars)a	4-5

PM2 5 and 03-related Health Benefits (billion 2024 dollars, undiscounted)1:1	4-7

Present Value and Equivalent Annualized Value Estimates of PM2 5 and 03-related Health Benefits

(billion 2024 dollars, discounted to 2025)a-b	4-8

Summary of Certain Energy Market Impacts	5-1

SAGE Dimensional Details	5-3

Compliance Costs, Transfers, and Social Costs (billion 2024 dollars)	5-6

Changes in Labor Utilization: Construction-Related (number of job-years of employment in a single

year	5-15

Changes in Labor Utilization: Recurring Non-Construction (number of job-years of employment in a

single year)	5-16

Net Benefits of the Proposed Action (billion 2024 dollars, undiscounted)a	6-3

Net Benefits, Present Value, and Equivalent Annualized Value Estimates of the Proposed Action
(billion 2024 dollars, discounted to 2025)a	6-4

ill


-------
LIST OF FIGURES

Figure 5-1 Percent Change in Real GDP and Components	5-7

Figure 5-2 Percent Change in Sectoral Output and Real Output Prices	5-10

Figure 5-3 Percent Change in Labor Demand in Select Sectors	5-11

Figure 5-4 Distribution of General Equilibrium Social Costs	5-12

iv


-------
1 EXECUTIVE SUMMARY

1.1 Introduction

In this action, the U.S. Environmental Protection Agency (EPA) is proposing to repeal all
greenhouse gas (GHG) standards for fossil fuel-fired power plants. The EPA is proposing that
the Clean Air Act requires it to make a finding that GHG emissions from fossil fuel-fired power
plants contribute significantly to dangerous air pollution, as a predicate to regulating GHG
emissions from those plants. The EPA is further proposing to make a finding that GHG
emissions from fossil fuel-fired power plants do not contribute significantly to dangerous air
pollution.1 The EPA is also proposing, as an alternative, to repeal a narrower set of requirements
that include the emission guidelines for existing fossil fuel-fired steam generating units, the
carbon capture and seque strati on/storage (CCS)-based standards for coal-fired steam generating
units undertaking a large modification, and the CCS-based standards for new base load stationary
combustion turbines.2

In accordance with Executive Orders (E.O). 12866 and 13563, the guidelines of OMB
Circular A-4, and EPA's Guidelines for Preparing Economic Analyses (U.S. EPA, 2024a), this
Regulatory Impact Analysis (RIA) analyzes the regulatory compliance costs and benefits
associated with this proposed action, with the proposed action containing both a proposal and
alternative proposal as described above. This RIA builds upon modeling prepared for the 2024
final CPS rule regulatory impact analysis (2024 CPS RIA), as that modeling reflects the
projected impact of the CPS relative to a baseline without the CPS.3 In this RIA, we draw upon
the "final rules" illustrative scenario and baseline modeling from the 2024 CPS RIA.

The "baseline" is a business-as-usual scenario that ordinarily represents the behavior of
the regulated sector under market and regulatory conditions in the absence of a regulatory action.
The baseline for the 2024 CPS RIA included numerous rules that had been finalized at the time
of that analysis. From the perspective of this proposed action, the 2024 CPS RIA is now in the
baseline, and this proposed action is the policy case. Additionally, there are significant market

1	For more information on the proposal, see Section IV. A of the preamble for this proposed action.

2	For more information on the alternative proposal, see Section V.A of the preamble for this proposed action.

3	The 2024 CPS RIA (U.S. EPA, 2024b) is found here: https://www.regulations.gov/document/EPA-HO-OAR-

2023-0072-8913

1-1


-------
and regulatory changes that have occurred since the 2024 CPS RIA was developed. As such,
EPA commits to conduct additional analysis in the future.

In absence of updated baseline modeling for comparison to projections under this
proposal, the compliance cost estimates presented in the 2024 CPS RIA are the EPA's best
available estimate of the reduction of compliance costs under this proposed action. Similarly, the
projected emission changes of the CPS in the final rules illustrative scenario in the 2024 CPS
RIA are the EPA's best available estimate of the emissions changes that will be reversed under
this proposed action, along with associated benefits under this proposed action. We have also
determined that these projected impacts are the Agency's best available estimate of the total
projected impacts of the alternative proposal. Modeling the differences between the proposal and
the alternative proposal requirements would likely not result in a meaningful difference in
modeling projections, as described in Section 3.2.1. Therefore, throughout this RIA we generally
present one set of results for this proposed action unless otherwise noted.

In this RIA, we evaluate the potential impacts of the proposed action using the present
value (PV) and equivalent annual value (EAV) of costs, benefits, and net benefits, calculated for
the years 2026 to 2047, discounted to 2025 using 3 and 7 percent discount rates.4 In addition, in
Sections 3 and 4, the Agency presents the assessment for specific snapshot years, consistent with
historical practice. These snapshot years are 2028, 2030, 2035, 2040 and 2045. We present
information about potential impacts of the proposed action on electricity markets, employment,
and markets outside the electricity sector. The RIA also presents a discussion of uncertainties
and limitations of the analysis. While the results are described and presented in more detail
throughout the RIA, we present the high-level results of the analysis here.

1.2 Compliance Cost Savings

The compliance cost estimates presented in this RIA are based on Integrated Planning
Model (IPM) projections supplemented with cost estimates for monitoring, reporting, and

4 Values in the 2024 CPS RIA were converted from 2019 dollars to 2024 dollars by multiplying by 1.204, which
was derived from the annual GDP Implicit Price Deflator values in the U.S. Bureau of Economic Analysis'
NIPA Table 1.1.9. Adjusting to 2024 dollars accounts for the majority of the change in values between the 2024
CPS RIA and this RIA. Some difference in the change in values is also due to discounting over a timeframe one
year shorter (i.e., to 2025 instead of 2024).

1-2


-------
recordkeeping (MR&R).5 Table 1-1 presents the projected compliance cost savings of this
proposed action with adjustments to account for updated estimates of MR&R costs.6

Table 1-1 Present Value and Equivalent Annualized Value Estimates of Compliance
Cost Savings (billion 2024 dollars, discounted to 2025)	



3% Discount Rate



7% Discount Rate

PV

EAV

PV

EAV

19

1.2

9.6

0.87

Note: Values have been rounded to two significant figures.

1.3 Emissions Changes of the Regulated Pollutant

Table 1-2 presents the CO2 emission changes associated with the proposed action, which
are projected to increase relative to the baseline.

Table 1-2 Carbon Dioxide Emissions Changes a

CO2 (million metric tons)

2028

38

2030

50

2035

123

2040

54

2045

42

a This analysis is limited to the geographically contiguous lower 48 states.

1.4 Economic Impacts

As a result of the change in compliance costs incurred by the regulated sector, the
proposed action has economic and energy market implications. The energy market impact
estimates presented here reflect the opposite of the projected impacts of the CPS in the 2024 CPS
RIA. Table 1-3 presents a variety of energy market impact estimates for 2028, 2030, 2035, 2040,
and 2045 for the proposed action. However, there are several key areas of uncertainty inherent in
these projections as outlined in Section 3.7 of the 2024 CPS RIA.

5	Information on IPM can be found at the following link: https://www.epa.gov/power-sector-modeling

6	Compliance costs refer to the difference between policy and baseline IPM projected capital, operations and

maintenance, fuel, transmission, CO2 storage and transportation costs, and net tax payments. Other costs are not
accounted for. Please see Sections 3.2 and 5.2 for further discussion of the differences between compliance costs
and social costs.

1-3


-------
Table 1-3 Summary of Certain Energy Market Impacts



2028

2030

2035

2040

2045

Retail electricity prices

1%

-0%

-1%

-0%

-1%

Average price of coal delivered to the power sector

1%

1%

-0%

-0%

32%

Coal production for power sector use

6%

4%

21%

-15%

84%

Price of natural gas delivered to power sector

2%

-0%

-3%

-0%

-0%

Price of average Henry Hub (spot)

2%

1%

-3%

-0%

-0%

Natural gas use for electricity generation

1%

2%

-4%

-0%

-2%

A more detailed version of this table is found in Section 3.2.5 of the 2024 CPS RIA, along
with additional discussion of energy market impacts.

More broadly, changes in production in a directly regulated sector may affect other
markets when output from that sector is used in the production of other goods. In particular,
electricity, the directly regulated sector in this proposed action, is an input to many goods and
services throughout the economy. The proposed action may also affect production in upstream
industries that supply inputs to the electricity sector and household consumption patterns as a
result of electricity, natural gas, and other final good price changes. Changes in firm and
household behavior could also interact with pre-existing market distortions, such as taxes. A
computable general equilibrium (CGE) model can be used to evaluate the broad economy-wide
impacts of a regulatory action and its social cost by accounting for these interactions and
feedback.

The EPA used the peer-reviewed CGE model SAGE to evaluate the economy-wide social
costs and economic impacts of the proposed action (U.S. EPA Science Advisory Board, 2020;
Marten et al., 2024). The annualized social cost estimated in SAGE for the proposed action is
approximately -$1.58 billion (2024 dollars) between 2026 and 2047, and -$1.81 billion (2024
dollars) over the period from 2026 to 2081. Note that SAGE does not currently account for the
effects of changing environmental quality as a result of this proposed action. Section 5.2 of the
RIA provides additional explanation of how the social cost was estimated using SAGE, as well
as the potential household distributional impacts.

Environmental regulation may affect groups of workers differently, as changes in
abatement and other compliance activities cause labor and other resources to shift. An
employment impact analysis describes the characteristics of groups of workers potentially

1-4


-------
affected by a regulation, as well as labor market conditions in affected occupations, industries,
and geographic areas. Employment impacts of the proposed action are discussed in Section 5 of
this RIA.

1.5 Net Benefits from the Proposed Action

The net benefits associated with the regulated pollutants are the cost savings of this
proposed action presented above in Table 1-1. Consistent with E.O. 14154 "Unleashing
American Energy" (90 FR 8353, January 20, 2025) and the memorandum titled "Guidance
Implementing Section 6 of Executive Order 14154, Entitled 'Unleashing American Energy'", the
EPA did not monetize benefits associated with CO2 emissions changes in Table 1-2.7

The rest of the RIA presents a full discussion of the projected costs, benefits, and net
benefits of this proposed action, as well as a discussion of uncertainty and additional impacts that
the EPA could not monetize.

This RIA follows the EPA's historical practice of using a technology-rich partial
equilibrium model of the electricity and related fuel sectors to estimate the incremental costs of
producing electricity under the requirements of proposed and final major EPA power sector
rules. In Section 5.2 of this RIA, the EPA has also included an economy-wide analysis that
considers additional facets of the economic response to the proposed action. This analysis
includes estimates of the full resource requirements the power sector will no longer be required
to spend as a result of this proposed action, some of which were paid for through subsidies.

7 The memorandum titled "Guidance Implementing Section6 of Executive Order 14154, Entitled 'Unleashing
American Energy'" is found here: https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-27-Guidance-
Implementing-Section-6-of-Executive-Order-14154-Entitled-Unleashing-American-Energy.pdf

1-5


-------
1.6 References

Marten, A., Schreiber, A., and Wolverton, A. (2024). SAGE Model Documentation (2.1.1).
Washington, DC. https://www.epa.gov/environmental-economics/cge-modeling-
regulatory-analysis.

OMB. (2003). Circular A-4: Regulatory Analysis. Washington DC.

https://www.whitehouse.gov/wpcontent/uploads/legacy_drupal_files/omb/circulars/A4/a-
4.pdf.

U.S. EPA. (2024a). Guidelines for Preparing Economic Analyses (3rd edition). EPA-240-R-24-
001. Washington, DC.

U.S. EPA. (2024b). Regulatory Impact Analysis for the New Source Performance Standards for
Greenhouse Gas Emissions from New, Modified, and Reconstructed Fossil Fuel-Fired
Electric Generating Units; Emission Guidelines for Greenhouse Gas Emissions from
Existing Fossil Fuel-Fired Electric Generating Units; and Repeal of the Affordable
Clean Energy Rule. EPA-452/R-24-009. Office of Air Quality Planning and Standards,
Health and Environmental Impacts Division, Research Triangle Park, NC.
https://www.regulations.gov/document/EPA-HQ-OAR-2023-0072-8913.

U.S. EPA Science Advisory Board. (2020). Technical Review ofEPA's Computable General
Equilibrium Model, SAGE. EPA-SAB-20-010. Washington, DC.

1-6


-------
2 INTRODUCTION AND BACKGROUND

2.1	Introduction

In this action, the U.S. Environmental Protection Agency (EPA) is proposing to repeal all
greenhouse gas (GHG) standards for fossil fuel-fired power plants. The EPA is proposing that
the Clean Air Act requires it to make a finding that GHG emissions from fossil fuel-fired power
plants contribute significantly to dangerous air pollution, as a predicate to regulating GHG
emissions from those plants. The EPA is further proposing to make a finding that GHG
emissions from fossil fuel-fired power plants do not contribute significantly to dangerous air
pollution.8 The EPA is also proposing, as an alternative, to repeal a narrower set of requirements
that include the emission guidelines for existing fossil fuel-fired steam generating units, the
carbon capture and seque strati on/storage (CCS)-based standards for coal-fired steam generating
units undertaking a large modification, and the CCS-based standards for new base load stationary
combustion turbines.9

2.2	Purpose of RIA

In accordance with Executive Orders (E.O). 12866 and 13563, the guidelines of OMB
Circular A-4, and EPA's Guidelines for Preparing Economic Analyses (U. S. EPA, 2024), the
EPA prepared this RIA for this "significant regulatory action." This action is an economically
significant regulatory action because it may have an annual effect on the economy of $100
million or more or adversely affect in a material way the economy, a sector of the economy,
productivity, competition, jobs, the environment, public health or safety, or state, local, or tribal
governments or communities. This RIA addresses the regulatory compliance costs, emission
impacts, and benefits of this proposed action. Additionally, this RIA includes information about
potential impacts of the proposed action on electricity markets, employment, and markets outside
the electricity sector.

8	For more information on the proposal, see Section IV. A of the preamble for this proposed action.

9	For more information on the alternative proposal, see Section V.A of the preamble for this proposed action.

2-1


-------
2.3 Overview of Regulatory Impact Analysis

The starting point for the analysis in this RIA is the modeling conducted for the 2024
CPS RIA. The baseline from the 2024 CPS RIA reflected projected conditions without the CPS,
and the final rules illustrative scenario from that RIA reflected projected conditions with the
CPS. The comparison of these results showed the projected impact of the CPS. These impacts
included evaluation of the benefits, costs, and certain impacts of compliance with an illustrative
scenario titled Final Rules (final rules illustrative scenario), which represented compliance with
the CPS. This RIA draws upon that analysis to provide estimates of the impacts of this proposed
action.

2.3.1 Analysis from 2024 CPS RIA

2.3.1.1 Baseline

The "baseline" is a business-as-usual scenario that, in the context of this analysis,
represents expected behavior in the power industry sector under market and regulatory
conditions in the absence of a regulatory action. The baseline for the 2024 CPS RIA included
numerous rules that had been finalized at the time of that analysis. The version of the IPM model
used for the 2024 CPS RIA also included state and federal legislation affecting the power sector,
including the Inflation Reduction Act of 2022 (IRA). The modeling documentation, available in
the docket, includes a summary of all legislation reflected in that version of the model as well as
a description of how that legislation is implemented in the model.10

Please see Section 3 of the 2024 CPS RIA for details of the baseline modeling. However,
from the perspective of this proposed action, the 2024 CPS RIA is now in the baseline, and there
are additional significant market and regulatory changes that have occurred since the 2024 CPS

10 IPM is a cost minimizing, perfect foresight, technology-rich partial equilibrium model. Because the model
incorporates sector specific information, EPA finds the model useful in examining the likely effects of
regulations on the power sector. Recognizing that complexity of such models may reduce public understanding
of IPM, including how the model takes into account behavioral changes likely to occur in the power sector as
well as the validity of the model projection (e.g., how well internal projection of natural gas price in IPM tracks
the observed natural gas price), EPA works to increase transparency of the IPM model, facilitating public
understanding of the strengths and limitations of the model in order to improve public discourse around its
results. While the model is proprietary, it periodically undergoes peer reviews, which are available at:
https://www.epa.gov/power-sector-modeling/ipm-peer-reviews. and EPA provides detailed documentation of
model logic and input assumptions, available at: https://www.epa.gov/power-sector-modeling/2023-reference-
case

2-2


-------
RIA was developed. We have not updated the baseline for this proposed action to reflect these
regulatory and other subsequent changes since the CPS was promulgated in 2024. Rather, we
rely on the 2024 CPS RIA policy case analysis as the baseline for this action. Similarly, there
may be other regulatory changes before the promulgation of this proposed action, and these too
are not accounted for in the baseline for this action. These facts introduce important uncertainties
in the analysis within this RIA. Because of these uncertainties, EPA commits to conducting
additional analysis that incorporates these changes in the baseline. EPA will also consider
whether to include a new air quality assessment to take into account potential changes in the
energy market.

2.3.1.2 Years of Analysis

In the 2024 CPS RIA, the EPA evaluated the potential costs, benefits, and net benefits of
the final rules illustrative scenario for the years 2024 to 2047 from the perspective of 2024, using
the discount rates of two percent, three percent, and seven percent. In addition, the Agency
presented an assessment of costs, benefits, and net benefits for specific snapshot years, consistent
with historical practice. These snapshot years were 2028, 2030, 2035, 2040 and 2045. The
Agency believed that these specific years were each representative of several surrounding years,
which enabled the analysis of costs and benefits over the timeframe of 2024 to 2047. The year
2028 was the first year of detailed power sector modeling for the 2024 CPS RIA. However,
because the Agency estimated that some monitoring, reporting, and recordkeeping (MR&R)
costs would be incurred beginning in 2024, the EPA analyzed compliance costs in years before
2028. Therefore, while the MR&R costs analysis was presented beginning in the year 2024, the
detailed assessment of costs, emissions impacts, and benefits began in the year 2028. The
analysis timeframe concluded in 2047, as this was the last year that could be represented with the
analysis conducted for the specific year of 2045.

2.3.2 Analysis in this RIA

2.3.2.1 Representing the Proposed Action with Modeling from the 2024 CPS RIA

In absence of updated baseline modeling for comparison to projections under this
proposal, the compliance cost estimates presented in the 2024 CPS RIA are the EPA's best
available estimate of the reduction of compliance costs under this proposal. Similarly, the

2-3


-------
projected emission changes of the CPS in the final rules illustrative scenario in the 2024 CPS
RIA are the EPA's best available estimate of the emissions changes that will be reversed under
this proposed action, along with associated benefits under this proposed action. We have also
determined that these projected impacts are the Agency's best available estimate of the total
projected impacts of the alternative proposal, as described in Section 3.2.1. Therefore,
throughout this RIA we generally present one set of results for this proposed action unless
otherwise noted.

2.3.2.2 Years of Analysis in this RIA

In this RIA, we evaluate the potential impacts of the proposed action using the present
value (PV) of costs, benefits, and net benefits, calculated for the years 2026 to 2047, discounted
to 2025. We estimate that 2026 is the first year for which MR&R costs would not be incurred
under this proposed action, which is why the timeframe of this RIA begins in 2026. All other
impact analyses begin in 2028, as in the 2024 CPS RIA. In addition, in Sections 3 and 4, the
Agency presents the assessment for specific snapshot years of 2028, 2030, 2035, 2040 and 2045.

2-4


-------
2.4 References

OMB. (2003). Circular A-4: Regulatory Analysis. Washington DC.

https://www.whitehouse.gov/wpcontent/uploads/legacy_drupal_files/omb/circulars/A4/a-
4.pdf.

U.S. EPA. (2024). Guidelines for Preparing Economic Analyses (3rd edition). EPA-240-R-24-
001. Washington, DC.

2-5


-------
3 COMPLIANCE COSTS, EMISSIONS, AND ENERGY IMPACTS

3.1	Introduction

This section presents the projected regulatory cost impacts of this proposed action. Given
that while Phase 1 of the NSPS requirements is in effect, the requirements on existing sources
and Phase 2 of the NSPS are not yet in effect and in the absence of updated projections of the
power sector under this proposed action compared to an updated baseline, this analysis assumes
the costs of the CPS as previously estimated upon original promulgation will be cost reductions
under this proposed action.

The cost estimates provided for this proposed action are presented in 2024 dollars,
whereas those presented in the 2024 CPS RIA were presented in 2019 dollars.11 Additionally, the
cost estimates provided here reflect the period 2026 to 2047, whereas those provided in the 2024
CPS RIA reflected the period 2024 to 2047. The reduction in costs for this proposed action
reflects the previously calculated change in the projected electric power system costs between
the base case and the final rules illustrative scenario as presented in the 2024 CPS RIA.12 These
costs include the change in capital costs, variable costs, fixed costs, transmission costs, fuel
costs, and MR&R costs13 that are expected to not be incurred as a result of this proposed action.

3.2	Economic Impacts

As outlined in the 2024 CPS RIA, the projected changes in costs and emissions were
derived using IPM, which is a system-wide, least-cost optimization model that projects EGU
behavior across the geographically contiguous U.S.14 IPM projects one possible combination of
compliance outcomes under a given policy scenario. The change in production costs and

11	Values are adjusted to 2024 dollars using the annual GDP Implicit Price Deflator values in the U.S. Bureau of

Economic Analysis' (BEA) NIPA Table 1.1.9, last revised February 27, 2025, which is available at
https://apps.bea.gov/iTable/?reqid= 19&step=3&isuri=l& 192 l=survey& 1903=13.

12	See Section 3 of the CPS RIA for a detailed description of the modeled policy.

13	See Section 3 of the CPS RIA for a detailed discussion of the compliance cost impacts.

14IPM uses model years to represent the full planning horizon being modeled. By mapping multiple calendar years
to a run year, the model size is kept manageable. IPM considers the costs in all years in the planning horizon
while reporting results only for model run years. For this analysis, IPM maps the calendar years 2028 and 2029
to run year 2028, calendar years 2030-31 to run year 2030, calendar years 2032-37 to run year 2035, calendar
years 2038-41 to run year 2040, calendar years 2042-47 to run year 2045 and calendar years 2048-52 to run year
2050. For model details, please see Chapter 2 of the IPM documentation, available at:

https://www.epa.gov/airmarkets/power-sector-modeling Please see Section 3 of the 2024 CPS RIA for detailed
description of IPM scenarios modeled.

3-1


-------
associated emissions consistent with the operation of the compliant generating portfolio capture
the costs of installation and operation of controls needed to comply with a regulatory action, the
costs of building and operating any new resources added to the grid, as well as costs associated
with shifts in the production and bulk transmission of electricity and other compliance costs.
These impacts are summarized in the following sections.

3.2.1 Representing the Proposed Action with Modeling from 2024 CPS RIA

The 2024 CPS RIA reflected a baseline that included the 2015 NSPS requirements. In
other words, the representative generating portfolio under both the baseline and policy scenarios
were compliant with the 2015 NSPS.15 The 2015 NSPS set standards for new, modified, and
reconstructed fossil fuel-fired steam generating units based on a highly efficient, supercritical
pulverized coal EGU that implements post-combustion partial carbon capture and
sequestration/storage technology. For new and reconstructed baseload combustion turbines the
2015 NSPS set standards based on efficient natural gas combined cycle (NGCC) technology and
for non-baseload and multifuel-fired units the required use of lower emitting fuels.

Given a confluence of market trends, most notably declining natural gas prices and
increasing renewable penetration, the 2015 NSPS requirements have been largely non-binding.
The relative economics of coal-fired generation have remained challenging as evidenced by
continued retirement, while investors have an incentive to build the most efficient NGCC units
possible in order to minimize fuel costs - i.e., building EGUs that are compliant with the 2015
NSPS. Simple cycle turbines are designed to operate more infrequently, resulting in operation
that tends to remain below the baseload threshold as defined in the 2015 NSPS.

While demand growth is projected to be higher under the baseline presented in the 2024
CPS RIA than over the prior ten years, a combination of greater renewable penetration (driven
by both incentives under the IRA as well as improving cost and performance characteristics over
the period) and lower natural gas prices continue to weaken the relative economics of coal-fired
EGUs, while capacity factors for the natural gas fleet are projected to decline over the period. In

15 "Standards of Performance for Greenhouse Gas Emissions From New, Modified, and Reconstructed Stationary
Sources: Electric Utility Generating Units", (80 FR 64510; October 23, 2015) (2015 NSPS) codified in 40 CFR
part 60, subpart TTTT.

3-2


-------
other words, market forces are leading to compliance with the 2015 NSPS, and this proposed
action is not expected to result in any changed decisions or compliance costs changes.

Phase 1 of the NSPS in the CPS requires an efficient generation standard of 800 lbs.
CCh/MWh-gross for baseload combustion turbines operating above a 40 percent capacity factor,
1,170 lbs. CO2 /MWh-gross for intermediate load combustion turbines operating between a 20 to
40 percent capacity factor, and 160 lbs. CCh/MMbtu for low load combustion turbines operating
at below a 20 percent capacity factor. Within the modeling of the 2024 CPS RIA, performance
assumptions taken from the Annual Energy Outlook (AEO) 2023 resulted in new combustion
turbine build options that can comply with the standards. Since the generating portfolio under
both the baseline and final rules illustrative scenario in the 2024 CPS RIA is compliant with the
phase 1 standards, the final rules illustrative scenario modeling is representative of the alternative
proposal in this proposed action. This is true even though phase 1 of the CPS remains in the
alternative proposal of this action but is not explicitly modeled as part of the baseline scenario of
the 2024 CPS RIA.

3.2.2 Emissions Changes Assessment

Table 3-1 shows the total EGU annual emissions of CO2 in the 2024 CPS RIA and the
emissions changes of this proposed action. Positive values for the emission changes column
reflect emissions increases.

Table 3-1 EGU Annual CO2 Emissions and Emission Changes (million metric tons)16

Annual CO2

Total Emissions

(million metric tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

1,121

1,159

38

2030

1,048

1,098

50

2035

601

724

123

2040

406

459

54

2045

265

307

42

16 This analysis is limited to the geographically contiguous lower 48 states.

3-3


-------
Table 3-2 shows the total EGU annual emissions of NOx, PM2.5, Hg and SO2 in the 2024
CPS RIA, and the emissions changes of this proposed action. Positive values for the emission
changes column reflect emissions increases.

3-4


-------
Table 3-2 EGU Annual Emissions and Emissions Changes for NOx, SO2, Hg, and
PM2.5, and Ozone Season NOx	

Annual NOx

Total Emissions

(Thousand Tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

441

461

20

2030

374

393

20

2035

210

259

49

2040

166

173

6

2045

83

107

24

Ozone Season NOxa

Total Emissions

(Thousand Tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

183

189

6

2030

168

175

7

2035

100

119

19

2040

82

88

6

2045

45

59

14

Annual SO2

Total Emissions

(Thousand Tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

420

454

34

2030

313

334

20

2035

150

240

90

2040

139

143

4

2045

13

55

41

Annual
Mercury

Total Emissions

(Tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

3.0

3.1

0.1

2030

2.8

2.9

0.1

2035

2.4

2.5

0.1

2040

2.3

2.0

-0.2

2045

1.2

1.4

0.2

Direct PM2.5

Total Emissions

(Thousand Tons)

Baseline with CPS

Proposed Action

Emissions Change

2028

69

71

2

2030

65

66

2

2035

49

51

1

2040

39

37

-2

2045

22

24

2

a Ozone season is the May through September period in this analysis.

3-5


-------
3.2.3 Monitoring, Reporting, and Recordkeeping Costs

In this RIA, we estimate that 2026 is the first year for which MR&R costs would be
reduced under this proposed action. Under the proposal, the estimated MR&R costs are the sum
of MR&R costs for the CPS and the 2015 NSPS that would be reduced with the proposed repeal
of these rules. The MR&R cost reductions associated with repealing the CPS are estimated to be
equivalent to the MR&R costs as found in Section 3.3 of the 2024 CPS RIA. The MR&R cost
reductions associated with repealing the 2015 NSPS are estimated to be equivalent to those in the
Information Collection Request for the 2015 NSPS, with an analysis assumption that these costs
continue through 2047. Table 3-3 presents a summary of the MR&R costs. For details on these
estimates made for purposes of this economic analysis, see the associated spreadsheet in the
docket titled P VEA V and Net Benefits Analysis - 2025 Proposed Action.xlsx.

3-6


-------
Table 3-3 Summary of State and Industry Annual Respondent Cost of Reporting and

Recordkeeping Requirements (million 2024 dollars)

Year

2015 NSPS
Industry State

Total

2024 CPS

Total

NSPS for New,
Modified, and
Reconstructed
Fossil Fuel-Fired
Electric Generating
Units

EGs for Existing
Fossil Fuel Fired
Electric Generating
Units

Total

Industry State

Industry State

2026

0.95

0.95

0.04

13

13

14

2027

1.0

1.0

0.08

-

0.08

1.1

2028

1.1

1.1

0.13

-

0.13

1.2

2029

1.2

1.2

0.17

-

0.17

1.3

2030

1.2

1.2

0.21

-

0.21

1.5

2031

1.3

1.3

0.25

-

0.25

1.6

2032

1.4

1.4

0.30

-

0.30

1.7

2033

1.5

1.5

0.34

-

0.34

1.8

2034

1.5

1.5

0.38

-

0.38

1.9

2035

1.6

1.6

0.42

-

0.42

2.0

2036

1.7

1.7

0.46

-

0.46

2.2

2037

1.8

1.8

0.51

-

0.51

2.3

2038

1.8

1.8

0.55

-

0.55

2.4

2039

1.9

1.9

0.59

-

0.59

2.5

2040

2.0

2.0

0.63

-

0.63

2.6

2041

2.1

2.1

0.68

-

0.68

2.7

2042

2.1

2.1

0.72

-

0.72

2.9

2043

2.2

2.2

0.76

-

0.76

3.0

2044

2.3

2.3

0.80

-

0.80

3.1

2045

2.4

2.4

0.84

-

0.84

3.2

2046

2.4

2.4

0.89

-

0.89

3.3

2047

2.5

2.5

0.93

-

0.93

3.4

Notes: Values have been rounded to two significant figures, and some are presented to no more than two decimal
places. Values may not appear to add correctly due to rounding. These estimates are for purposes of the analysis in
this RIA. For additional information, please see the spreadsheet in the docket titled PVEA V and Net Benefits
Analysis - 2025 Proposed Action, xlsx.

The difference in the estimated MR&R costs between the proposal and alternative
proposal is the 2015 NSPS costs, which are avoided costs under the proposal but not the
alternative proposal. This difference is less than $3 million in any year, as shown in Table 3-3.
These values are smaller than the rounding throughout the total table at the end of this section
(Table 3-7) and the net benefit tables in Section 6.

3.2.4 Compliance, Real Resource, and Social Cost Estimates

The CPS was projected to incur significant compliance costs that are no longer necessary
as a result of this proposed action and therefore reflect a reduction in costs in PV and EAV
terms. Table 3-4 presents these changes in costs for the representative IPM run years 2028, 2030,

3-7


-------
2035, 2040, and 2045 in real 2024 dollars. A negative cost reflects a projected reduction in
compliance cost expenditures as a result of this proposed action, while a positive cost denotes an
increase. Additionally, annualized costs are presented for the periods 2026 to 2042 and 2026 to
2047 in 2024 dollars. Costs for these periods are annualized using IPM's real discount rate of
3.76 percent and exclude costs for years outside of each of the respective multiple-year time
periods.17 For a detailed description of these cost trends, please see Section 3 of the 2024 CPS
RIA.

Table 3-4 National Power Sector Costs (billion 2024$)	

2026 to 2042 (Annualized)	-0.58

2026 to 2047 (Annualized)	-1.14

2028 (Annual)	1.56

2030 (Annual)	0.27

2035 (Annual)	-1.54

2040 (Annual)	-0.71

2045 (Annual)	-4.02

IPM provides the EPA's best estimate of the change in costs due to the proposed action to
the electricity sector and related energy sectors (i.e., natural gas, coal mining). The projected
change in the IPM system cost shown in Table 3-4 is the change in costs paid by the power
sector as a result of the EPA actions analyzed in the 2024 CPS RIA. The projected change in
IPM system cost includes the cost of additional resources that are used by the sector because of
the regulation, such as the cost of additional generation equipment, pollution controls, labor,

17 The objective function of IPM minimizes the present value of system costs, and a discount rate is used in IPM to
convert all future costs to a present value. The private discount rate adopted for modeling investment behavior
should reflect the rate at which investors are willing to invest in the sector. For a general discussion of the risk
and temporal preferences, tax treatments, and costs of borrowing that inform discount rates, Section 6.4 of the
EPA's Guidelines for Preparing Economic Analyses (U.S. EPA, 2024a). The real discount rate used in EPA's
Power Sector Modeling Platform 2023 Using the Integrated Planning Model, 3.76 percent, equals the real
weighted average after-tax cost of capital for various ownership types and technologies. The discount rate used
in EPA's modeling is invariant over time. For more information, see Chapter 10 of the Documentation for EPA's
Power Sector Modeling Platform 2023 Using the Integrated Planning Model 2023 Reference Case, available in
the docket (U.S. EPA, 2024b). The private discounting used in IPM to simulate industry behavior differs from
the social discounting used to estimate the social net benefits of the regulatory action. The social discount rates
used in the net benefits analysis in this RIA reflect the intertemporal preferences of society as a whole, with 3
percent representing the consumption rate of interest and 7 percent representing the social opportunity cost of
capital (OMB Circular A-4 (2003), and Section 6.2 of the EPA Guidelines (2024a)). As discussed in Section
5.2.1, the discount rate in SAGE varies over time and household depending on intertemporal consumption
preferences. The social costs estimates from SAGE apply a discount rate of 4.5 percent, which is consistent with
the internal discount rate and average capital rate in the model.

3-8


-------
materials, fuel, transport, and storage costs. These "real resources" constitute the additional
physical and labor inputs the sector purchases because of the regulation.

The projected change in system cost also accounts for changes in tax and subsidy
payments, financing charges for new capital, and insurance. For example, when IPM projects
that a new generator will be built, the system cost accounts for the cost to purchase and install
and operate the generator as well as the cost to finance the generator and expected taxes and
insurance that will be paid on the investment. The system cost also accounts for any production
and investment subsidies or tax credits that may offset expenditures on real resources.

Most tax, tax credits, and subsidy payments are transfers (OMB, 2003; U.S. EPA,
2024a).18 Transfers are shifts in money or resources from one part of the economy (e.g., a group
of individuals, firms, or institutions) to another in a way that does not affect the total resources
that are available to society. In other words, the loss to one part of the economy is exactly offset
by the gain to another. Transfers should be excluded from estimates of the benefits and costs of a
regulatory action (Ibid). There are two important sources of tax-related transfers accounted for in
the IPM system cost analysis for the 2024 CPS rule: taxes and tax credits on the production and
investment of generating resources and air pollution control equipment, such as federal and state
income taxes, and the tax credit for sequestration of carbon dioxide.19

18	See "The Difference between Costs (or Benefits) and Transfer Payments" in OMB (2003) and sections 6.4.2 and

8.2.2.2 of U.S. EPA (2024a).

19	See Chapters 6 and 10 of the IPM documentation (U.S. EPA 2024b). The total value of the transfer associated

with taxes accounts for the revenue forgone due to any tax credits (i.e., the value equals the total change in net
tax payments). Certain credits may be limited by the total tax obligation of reporting entities or the ability to
transfer them to other entities. The IPM analysis in this RIA assumes that the full value of any credits is not
limited by total tax obligation of reporting entities or their ability to transfer their value.

3-9


-------
Table 3-5 Real Resource Costs Avoided under the Proposed Action (billion 2024
dollars)	



[1]

[2]

[3]

[4]





Change in Transfers





Change in IPM
System Costs

CO2 Storage
Tax Credits

Production and
Investment Taxes
and Credits

Change in Real
Resource Costs

2028

1.63

-0.00

-0.12

1.75

2030

0.31

-0.18

-0.25

0.38

2035

-1.61

-7.52

-1.77

-7.38

2040

-0.70

-6.88

-1.54

-6.05

2045

-3.96

-0.00

-1.63

-2.34

3% Discount Rate

PV (2026 to 2047)

-19.06

-52.33

-18.19

-53.32

EAV (2026 to 2047)

-1.20

-3.28

-1.14

-3.35

7% Discount Rate

PV (2026 to 2047)

-9.57

-34.46

-11.11

-33.01

EAV (2026 to 2047)

-0.87

-3.12

-1.00

-2.98

Table 3-5 reports the changes in IPM system costs, transfers, and incremental real
resource costs for each IPM model year.20 The values in Table 3-5 are not incurred under the
proposed action. To estimate the incremental resource costs of the CPS, we must identify the
portion of the changes in IPM system costs that are due to net changes in production and
investment taxes and credits for generation and tax credits for CO2 storage (i.e., the 45Q tax
credit).21

To start, Column 1 of Table 3-5 shows the change in IPM system costs (drawn from
Table 3-4 in this RIA), which represents the expenditures that the power sector will not have to
make as a result of the proposed action. Column 2 reports the projected change in the CO2
storage tax credits paid to the electricity sector under the CPS rule and will not be paid under the
proposed action. The total value of the tax credit for carbon storage is reported separately given
its scale relative to the changes in the amount of the other transfers. Column 3 reports the other

20	The IPM system cost in this table excludes changes in payments for energy transmission costs and capacity

transfer costs (although the cost of operating reserve capacity is accounted for); this accounts for the difference
in compliance cost values between those in Table 3-4 and those in Table 3-5. The change in real resource costs
similarly in this table accounts for payments to capital amortized over the financing period of an investment at
IPM's internal discount rate.

21	It is not currently possible to separately report the changes in various investment and production taxes and credits

as a result of the proposed repeal of the CPS. However, on net, the avoided taxes exceed the value of the avoided
tax credits in those years where the proposed regulation decreases capital investment. (The converse is true in
those years where the proposed regulation increases capital investment).

3-10


-------
net tax-related transfers paid by the sector, which also will be avoided under the proposed action.
The projected incremental change in real resource costs presented in Column 4 is calculated as
the power system cost change plus the net change in CO2 tax credits minus the net change in
production taxes and tax credits. That is, Column 4 equals the sum of Columns 1 and 2 less
Column 3. This is the estimate of the incremental real resource costs that will not be incurred as
a result of the proposed action.

All capital costs in Column 4 are amortized at the private cost of borrowing over the
assumed financing period (i.e., book life) of the capital investment. The financing period varies
by technology and owner-type (Chapter 10, U.S. EPA 2024b).22

To estimate the change in social costs of this proposed action, the EPA used information
from IPM as an input into the Agency's economy-wide computable general equilibrium model,
SAGE. As described in Section 5.2.2.1, the inputs to the SAGE model matched both the real
resource requirements for the expected compliance pathway and the impact of the IRA tax
credits on the compliance expenditures for the electricity sector. To accomplish this, the real
resource requirements that would have been paid for by the IRA tax credits are included in the
avoided incremental costs of the CPS (i.e., the incremental costs exclude the production and
investment taxes and credits) as similarly presented in Table 3-5.

Like the real resource cost analysis summarized in Table 3-5, the economy-wide analysis
is a complement to the more detailed evaluation of sector costs produced by IPM. Both the real
resource cost and SAGE analyses improve the characterization of regulatory costs by accounting
for changes outside of the directly regulated sector. For example, the partial equilibrium IPM
model accounts for impacts in the directly regulated electricity sector as well as the closely-
related fuels sectors, but IPM does not account for impacts in the rest of the economy.

22 Components of private financing costs may reflect omitted real resource costs. If this is the case, excluding these
costs may lead to an underestimate of avoided real resource costs. For example, as discussed in Section 6.4.2 of
U.S. EPA (2024a), an interest payment reflects a transfer between a borrower and a lender that would net out
with social discounting. However, in some contexts, interest rates and insurance may account for risk and
uncertainty, that, in expectation, reflect additional resource costs (e.g., actuarily fair insurance premia). Section
6.1.6.2 of U.S. EPA (2024a) cautions, "While it is technically possible to adjust the discount rate to account for
uncertainty, doing so may hide important assumptions and information about the relative effects of discounting
and uncertainty from decision-makers. Uncertainty about future values should be treated separately when
discounting."

3-11


-------
As shown below in Section 5.2, Table 5-3, the annualized reduction in social cost
estimated in SAGE for the proposed action is approximately -$1.58 billion (2024 dollars)
between 2026 and 2047 using a 4.5 percent discount rate that is consistent with the internal
discount rate in SAGE. Under the assumption that compliance costs from IPM in 2056 continue
until 2081, the equivalent annualized value for reduction in social costs in the SAGE model is -
$1.81 billion (2024 dollars) over the period from 2026 to 2081 using the same discount rate. The
change in social cost estimate reflects the combined effect of the rules' requirements and
interactions with IRA23 tax credits for specific technologies that are expected to see decreased
use in response to the proposed action. We are currently not able to identify their relative roles.
Note that SAGE does not currently estimate the change in social benefits and their effects on the
economy from changing environmental quality. See Section 5.2 for more discussion on the
economy-wide analysis with SAGE and estimates of private and social costs.

3.2.5 Impacts on Fuel Use, and Prices

The CPS may have had important energy market implications that will no longer occur as
a result of this proposed action. Table 3-6 presents a variety of important national average energy
market impacts that were projected for the final rules illustrative scenario in the 2024 CPS RIA
(updated to 2024 dollars, where appropriate). The proposed action would reverse these potential
impacts, so the signs on these projected impacts are the opposite from what they were in the
2024 CPS RIA.

The projected energy market and electricity retail rate impacts of the CPS are discussed
more extensively in Section 3 of the 2024 CPS RIA, which also presents projections of power
sector generation and capacity changes by technology and fuel type. The change in wholesale
energy prices and the changes in power generation were projected using IPM. The change in

23 The Inflation Reduction Act of 2022 (IRA) contains tax credit provisions that affect power sector operations,
which are incorporated into the IPM modeling. Details are included the IPM documentation. The Clean
Electricity Investment and Production Tax Credits (provisions 48E and 45Y of the IRA) are described in more
detail in Section 4 of the IPM documentation (U.S. EPA 2024b). The credit for Carbon Capture and
Sequestration (provision 45Q) is described in Section 3. The impacts of the Zero-Emission Nuclear Power
Production Credit (provision 45U) are reflected through modifying nuclear retirement limits, as described in
Section 4 of U.S. EPA (2024b). The Credit for the Production of Clean Hydrogen (provision 45 V) is reflected
through the inclusion of an exogenously delivered price of hydrogen fuel, see Section 9 of U.S. EPA (2024b).
The Advanced Manufacturing Production Tax Credit (45X) was reflected through adjustments to the short-term
capital cost added for renewable technologies, see Section 4 of U.S. EPA (2024b). Documentation available at:
https://www.epa.gov/power-sector-modeling

3-12


-------
retail electricity prices reported Table 3-6 is a national average. The change in electricity retail
prices were projected using outputs of IPM. The average regional electricity price is projected to
decrease up to 6 percent or increase as much as 1 percent in 2035 as a result of the proposed
action.

Table 3-6 National Impacts on Fuel Prices, Fuel Consumption, and Electricity Prices





2028

2030

2035

2040

2045

Retail electricity

Baseline with CPS

116

119

117

115

112

prices
(2024
mills/kWh)

Proposed Action
Percentage Change (%)

117
0.7%

120
0.5%

115
-1.4%

115
-0.2%

111
-0.7%

Average price of
coal delivered to
the power sector
(2024 $/MMBtu)

Baseline with CPS
Proposed Action
Percentage Change (%)

1.8

1.9
1.4%

1.9
1.9
1.1%

1.9
1.9
0.5%

1.9
1.9
-0.5%

1.1
1.7
31.6%

Coal production
for power sector
use

(million tons)

Baseline with CPS
Proposed Action

236
250

209
218

112
141

104
90

4

26

Percentage Change (%)

6%

4%

21%

-15%

84%

Price of natural
gas delivered to

power sector
(2024$/MMBtu)

Baseline with CPS
Proposed Action
Percentage Change (%)

3.4
3.4
1.5%

3.5
3.5
0.5%

3.5
3.5
-3.0%

3.4
3.4
-0.0%

3.5
3.5
-0.1%

Price of average
Henry Hub (spot)
(2024$/MMBtu)

Baseline with CPS
Proposed Action
Percentage Change (%)

3.3
3.3
1.6%

3.4

3.5
0.6%

3.6
3.5
-2.9%

3.4
3.4
-0.1%

3.5
3.5
-0.0%

Natural gas use
for electricity
generation
(TCF)

Baseline with CPS
Proposed Action
Percentage Change (%)

11

12
1.0%

12
12
1.7%

10

9

-4.4%

6
6

-0.0%

4
4

-1.8%

3.2.6 Total Compliance Costs

Table 3-7 presents the undiscounted power sector generating costs, MR&R costs, and
total costs of the proposed action for the 2026 to 2047 timeframe. Table 3-8 presents the present
values (PVs) and equivalent annualized values (EAVs), calculated for the 2026 to 2047
timeframe.

3-13


-------
Table 3-7

Total Costs of the Proposed Action (billion 2024 dollars, undiscounted)



Power Sector Generating
Costsa

Monitoring, Reporting, and
Recordkeeping Costs b

Total Costs b

2026

-

-0.01

-0.01

2027

-

0.00

0.00

2028

1.6

0.00

1.6

2029

1.6

0.00

1.6

2030

0.3

0.00

0.27

2031

0.3

0.00

0.27

2032

-1.5

0.00

-1.5

2033

-1.5

0.00

-1.5

2034

-1.5

0.00

-1.5

2035

-1.5

0.00

-1.5

2036

-1.5

0.00

-1.5

2037

-1.5

0.00

-1.5

2038

-0.71

0.00

-0.71

2039

-0.71

0.00

-0.71

2040

-0.71

0.00

-0.71

2041

-0.71

0.00

-0.71

2042

-4.0

0.00

-4.0

2043

-4.0

0.00

-4.0

2044

-4.0

0.00

-4.0

2045

-4.0

0.00

-4.0

2046

-4.0

0.00

-4.0

2047

-4.0

0.00

-4.0

Notes: Values are undiscounted. Values have been rounded to two significant figures, and some are presented to no
more than two decimal places. Values may not appear to add correctly due to rounding.
a The discount rate in IPM is 3.76 percent, as described in Section 3.

b While the MR&R costs differ slightly between the proposal and the alternative proposal, the difference is smaller
than the rounding reflected here.

3-14


-------
Table 3-8 Present Value and Equivalent Annualized Value Estimates of Costs (billion
2024 dollars, discounted to 2025) a	

Power Sector Generating Costsa

Monitoring, Reporting, and
Recordkeeping Costs b



Total Costs b

PV EAV

PV EAV

PV

EAV

3% Discount Rate

-19 -1.2

-0.05 -0.00

-19

-1.2

7% Discount Rate

-9.6 -0.87

-0.03 -0.00

-9.6

-0.87

Notes: Values have been rounded to two significant figures, and some are presented to no more than two decimal
places. Values may not appear to add correctly due to rounding.
a The discount rate in IPM is 3.76 percent, as described in Section 3.

b While the MR&R costs differ slightly between the proposal and the alternative proposal, the difference is smaller
than the rounding reflected here.

3-15


-------
3.2.7 References

OMB. (2003). Circular A-4: Regulatory Analysis. Washington DC.

https://www.whitehouse.gov/wpcontent/uploads/legacy_drupal_files/ornb/circulars/A4/a-
4.pdf

U.S. EPA. (2024a). Guidelines for Preparing Economic Analyses (3rd edition). EPA-240-R-24-
001. Washington, DC. https://www.epa.gov/environmental-economics/guidelines-
preparing-economic-analyses-3rd-edition

U.S. EPA. (2024b). Documentation for EPA 's Power Sector Modeling Platform 2023 Using the
Integrated Planning Model 2023 Reference Case. Washington, DC.
https://www.epa.gov/system/files/documents/2025-02/epa-2023-reference-case.pdf

3-16


-------
4 BENEFITS ANALYSIS

4.1	Introduction

In this section, we present the monetized health impact estimates associated with the
emissions changes for the proposed action. The benefit estimates provided for this proposal are
presented in 2024 dollars, whereas the benefits estimates presented in the 2024 CPS RIA were
presented in 2019 dollars. Similar to Section 3, this section relies on the emissions changes
produced for the 2024 CPS RIA analysis to assess the health impacts of the proposed action.

The 2024 CPS RIA provides a detailed discussion of the methods used to estimate the
human health impacts of projected changes in the concentrations of PM2.5 and ozone resulting
from projected emissions changes under the rule. See Section 4 of the 2024 CPS RIA for details
on quantifying PM2.5 and ozone-related health benefits. Also, see Appendix B of the 2024 CPS
RIA for additional details on the air quality modeling and analysis used to create PM2.5 and
ozone air quality surfaces as well as a presentation of the uncertainties and limitations associated
with the methodologies. See also Section 6.4 for a brief discussion of these uncertainties and
limitations.

Consistent with E.O. 14154 "Unleashing American Energy" (90 FR 8353, January 20,
2025) and the memorandum titled "Guidance Implementing Section 6 of Executive Order 14154,
Entitled 'Unleashing American Energy'", the EPA did not monetize benefits associated with
CO2 emissions changes.24 For a brief discussion of uncertainties and limitations associated with
monetizing C02-related domestic climate benefits, see Section 6.4.

4.2	PM2.5 and Ozone-Related Human Health Benefits

The 2024 CPS RIA estimated the human health benefits of the reduced exposure to PM2.5
and ground-level ozone for the CPS by quantifying the number and economic value of avoided
PM2.5- and ground-level ozone-related premature deaths, illnesses, and related adverse effects.
Under this proposed action, the PM2.5 and ozone-related health benefits quantified in the 2024

24 The memorandum titled "Guidance Implementing Section 6 of Executive Order 14154, Entitled 'Unleashing
American Energy'" is found here: https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-27-Guidance-
Implementing-Section-6-of-Executive-Order-14154-Entitled-Unleashing-American-Energy.pdf

4-1


-------
CPS RIA are no longer expected. This RIA includes the estimates from the 2024 CPS RIA to
quantify the number and economic value of the human health impacts from this proposed action.

The health impacts are derived from estimates originally provided in the 2024 CPS RIA
that were calculated using a benefits transfer approach that adapts studies relating changes in
PM2.5 and ground-level ozone concentrations to incidences of premature death, illness, and
related adverse effects that are then monetized using a valuation function. This benefits transfer
approach makes it possible to assign a dollar value to the changes in PM2.5 and ground-level
ozone concentrations that cannot otherwise be directly valued. The 2024 CPS RIA used the
Benefits Mapping and Analysis Program - Community Edition (BenMAP-CE) software program
to conduct that analysis. For a full description of the methods used, see Section 4 of the 2024
CPS RIA and the BenMAP-CE technical support document titled Estimating PM2.5- and Ozone-
Attributable Health Benefits: 2024 Update (Health Benefits TSD) (U.S. EPA, 2024).

The EPA's methods for estimating health benefits due to changes in PM2.5 and ground-
level ozone concentrations were reviewed by an EPA Science Advisory Board (SAB) in 2023
(U.S. EPA Science Advisory Board, 2024). This SAB panel concluded that EPA's methods are
"scientifically robust and appropriate for regulatory analyses". The panel made several
recommendations for improvements, including valuing changes in nonfatal health risks with
willingness-to-pay measures or broader measures of the cost of illness, using scenario-based
demographic projections, and updating inputs into the calculation of the value of a statistical life.

To determine which PM2.5 and ozone-related human health impacts to quantify using the
BenMAP-CE model, the Agency consults the Integrated Science Assessment for Particulate
Matter (PM ISA) (U.S. EPA, 2019), the Supplement to the ISA for Particulate Matter (U.S. EPA,
2022), and the Integrated Science Assessment for Ozone and Related Photochemical Oxidants
(Ozone ISA) (U.S. EPA, 2020). Section 4.3.2 in the 2024 CPS RIA describes the process of
selecting air pollution health endpoints to quantify. The Health Benefits TSD fully describes the
Agency's approach for quantifying the number and value of estimated air pollution-related
impacts as well as the demographic, health and economic data used, and our techniques for
quantifying uncertainty.

Table 4-4 in the 2024 CPS RIA reports the PM2.5 and ozone-related human health effects
that were quantified and those that were not quantified in that RIA. Because the present analysis

4-2


-------
assumes that all estimated benefits reported for the final rule illustrative scenario in the 2024
CPS RIA will no longer occur for this proposed action, the PM2.5 and ozone-related human
health effects quantified in the 2024 CPS RIA are the same health effects relevant to this
proposed action. Table 4-4 in the 2024 CPS RIA does not contain an exhaustive list of benefit
categories that were not quantified. And, among the effects quantified, it was not always possible
to quantify completely either the full range of human health impacts or economic values. Section
4.4 and Table 4-24 of the 2024 CPS RIA report other omitted health and environmental benefits
expected from the emissions and effluent changes that would have resulted from that rule, such
as health effects associated with NO2 and SO2, and welfare effects such as acidification and
nutrient enrichment.

For more information on health impact functions, see Sections 4.3.3 and 4.4 of the 2024
CPS RIA and the Health Benefits TSD. For more information on the data inputs to the BenMAP-
CE model that enable estimation of avoided adverse health effect incidence, see Section 4.3 of
the 2024 CPS RIA.

Once incidence of avoided adverse health effects is estimated, BenMAP-CE calculates
the expected economic value of the avoided deaths, illnesses, and related adverse effects using
valuation functions. For a discussion of valuation functions, see Section 4.3.4 of the 2024 CPS
RIA. For some health effects, such as hospital admissions, valuation functions use the cost of
treating or mitigating the effect to economically value the health impact. For example, for the
valuation of hospital admissions, the avoided medical costs are used as an estimate of the value
of avoiding the health effects causing the admission. These cost-of-illness (COI) estimates
generally (although not in every case) understate the true value of reductions in risk of a health
effect. The COI estimates tend to reflect the direct expenditures related to treatment but not the
value of avoided pain and suffering from the health effect.

Estimates of avoided premature mortalities due to this proposed action quantified using
the described approach are provided in Table 4-1. Negative numbers indicate avoided premature
mortalities that are estimated to no longer occur under this proposed action. The estimated
number of reduced premature deaths in each year relative to the baseline is presented along with
the 95 percent confidence interval. For the CPS, for a full list of the total count of avoided

4-3


-------
premature deaths, illnesses, and related adverse effects for each of the representative years 2028,
2030, 2035, 2040, and 2045, please refer to tables 4-5 to 4-14 in the 2024 CPS RIA.

Estimates of the economic value of the avoided premature deaths, illnesses, and related
adverse effects are provided in Table 4-2. Ninety-five percent confidence intervals are also
reported. Negative numbers indicate estimated PM2.5 and ozone-related health benefits that will
no longer occur under this proposed action. When estimating the value of improved air quality
over a multi-year time horizon, the analysis applies population growth and income growth
projections for each future year through 2047 and estimates of baseline mortality incidence rates
at five-year increments. Two estimates of economic value are reported for each year and
discount rate combination. These estimates were quantified using two different epidemiological
estimates for the mortality impact of ozone and two different epidemiological estimates for the
mortality impact of PM, as well as their sum. The smaller estimate reports the economic value of
all avoided premature mortalities, illnesses, and related effects using a pooled short-term ozone
exposure risk estimate based on Katsouyanni et al. (2009) and Zanobetti et al. (2008) and a small
estimate of long-term PM2.5 exposure mortality risk based on Wu et al. (2020). The larger
estimate reports the same economic value but instead uses a long-term ozone exposure risk
estimate based on Turner et al. (2016) and a high estimate of long-term PM2.5 exposure mortality
risk based on Pope et al. (2019). These estimates should not be thought of as representing lower
and upper bounds. To see the economic value broken down by pollutant for each of the
representative years, please see tables 4-15 to 4-20 in the 2024 CPS RIA. To see the annual
stream of economic values discounted using 3 and 7 percent discount rates, see tables 4-22 and
4-23 in the 2024 CPS RIA. Please note that the 2024 CPS RIA provides estimates in 2019
dollars.

4-4


-------
Table 4-1 Estimated PJVh.s and Ozone-Related Avoided Premature Mortality a



Ozone-related Avoided Premature Mortality b

PIVh.s-related Avoided Premature
Mortality c

2028

-2.7 (-1.1 to -4.3) and -60 (-42 to -78)

-210 (-180 to -230) and -450 (-320 to -570)

2030

-2.7 (-1.1 to -4.3) and -60 (-42 to -78)

-140 (-120 to -150) and -290 (-200 to -360)

2035

-5.6 (-2.2 to -8.8) and -120 (-85 to -160)

-560 (-490 to -620) and -1,100 (-820 to -1,400)

2040

0.32 (0.50 to 0.13) and 7.0 (4.9 to 9.1)

11 (9.2 to 12) and 22 (16 to 28)

2045

-5.8 (-2.3 to -9.2) and -130 (-89 to -170)

-270 (-240 to -300) and -530 (-380 to -670)

a Values rounded to two significant figures. The two benefits estimates are separated by the word "and" to signify
that they are two separate estimates. The estimates do not represent lower- and upper-bound estimates and should
not be summed.

b The first ozone mortality estimate uses the pooled Katsouyanni et al. (2009) and Zanobetti et al. (2008) short-term
ozone exposure risk estimate and the second ozone mortality estimate uses the Turner et al. (2016) long-term ozone
exposure risk estimate.

0 The first PM2 5 mortality estimate uses the Wu et al. (2020) long-term PM2 5 exposure mortality risk estimate and
the second PM2 5 mortality estimate uses the Pope et al. (2019) long-term PM2 5 exposure mortality risk estimate.

Table 4-2 Estimated Economic Value of Avoided PM2.5 and Ozone-Related
Attributable Premature Mortality and Illnesses for the Proposed Action (95 percent
confidence interval; billion 2024 dollars) a	



Discount Rate b

PM2.5 and Ozone-Related Health Benefits c'd

2028

3%
7%

-$3 (-$0.46 to -$7.6) and -$6.8 (-$0.8 to -$18)
-$2.7 (-$0.35 to -$6.8) and -$6 (-$0.65 to -$16)

2030

3%
7%

-$2.1 (-$0.33 to -$5.1) and -$4.7 (-$0.57 to -$12)
-$1.8 (-$0.25 to -$4.5) and -$4.2 (-$0.46 to -$11)

2035

3%
7%

-$8 (-$1.1 to -$20) and -$17 ($-1.9 to -$46)
-$7.1 (-$0.88 to -$18) and -$15 (-$1.6 to -$41)

2040

3%
7%

$0.17 ($0.42 to $0,024) and $0.41 ($1.1 to $0,046)
$0.15 ($0.37 to $0,019) and $0.36 ($0.97 to $0,039)

2045

3%
7%

-$4.3 (-$0.62 to -$11) and -$9.5 (-$1.1 to -$25)
-$3.7 (-$0.48 to -$9.5) and -$8.5 (-$0.9 to -$23)

a Values rounded to two significant figures. The two benefits estimates are separated by the word "and" to signify
that they are two separate estimates. The estimates do not represent lower- and upper-bound estimates and should
not be summed.

b Estimates represent sums of all future benefit streams discounted back to the scenario year (2028, 2030, 2035,
2040, or 2045) to account for lags in the onset of health effects. These estimates have not been discounted to 2025.
0 The first estimate is the sum of ozone mortality estimated using the pooled short-term ozone exposure risk estimate
and the Wu et al. (2020) long-term PM2 5 exposure mortality risk estimate.

dThe second estimate is the sum of the Turner et al. (2016) long-term ozone exposure risk estimate and the Pope et
al. (2019) long-term PM2 5 exposure mortality risk estimate.

Data, time, and resource limitations prevented the EPA from quantifying the estimated
health impacts or monetizing estimated benefits for the 2024 CPS RIA associated with

4-5


-------
incremental changes in direct exposure to NO2 and SO2, independent of the role NO2 and SO2
play as precursors to PM2.5 and ozone, as well as ecosystem effects, and visibility impairment
that might result from emissions changes associated with compliance with the final requirements.
While all health benefits and welfare benefits were not quantified, it does not imply that there
were not additional benefits associated with reductions in human exposures to NO2 or SO2 and
ecosystem exposure to air pollutants would have potentially resulted from emissions changes
under the CPS.

Table 4-3 presents the undiscounted stream of annual monetized PM2.5 and ozone-related
health benefits and total benefits. Table 4-4 presents the present values (PVs) and equivalent
annualized values (EAVs), calculated for the 2026 to 2047 timeframe.

4-6


-------
Table 4-3 PM2.5 and Cb-related Health Benefits (billion 2024 dollars, undiscounted) a'b

Discount Rate



3%

7%

2026

-

-

2027

-

-

2028

-6.8

-6.0

2029

-6.9

-6.2

2030

-4.7

-4.2

2031

-4.8

-4.2

2032

-16

-14

2033

-16

-15

2034

-17

-15

2035

-17

-15

2036

-17

-16

2037

-18

-16

2038

0.39

0.35

2039

0.40

0.36

2040

0.41

0.36

2041

0.42

0.37

2042

-9.2

-8.2

2043

-9.3

-8.3

2044

-9.4

-8.4

2045

-9.5

-8.5

2046

-9.6

-8.6

2047

-9.7

-8.6

Non-Monetized Disbenefits

From increases in HAP emissions and GHG emissions
To water quality and availability
To ecosystems from increases in emissions of CO2, NOx, SO2, PM, and HAP
From increases in exposure to ambient NO2 and SO2
Decreased visibility (increased haze) from PM2 5 emissions increases
Notes: Benefits analysis begins in 2028.

a Values have been rounded to two significant figures. Values may not appear to add correctly due to rounding.
b The PM2.5 and Ch-related health benefits estimates use the larger of the two benefits estimates presented in Table
4-1. Monetized health benefits include those related to public health associated with changes in PM2 5 and ozone
concentrations. The health benefits are associated with several point estimates.

4-7


-------
Table 4-4 Present Value and Equivalent Annualized Value Estimates of PM2.5 and O3-
related Health Benefits (billion 2024 dollars, discounted to 2025) a'b	



3% Discount Rate



7% Discount Rate

PV

EAV

PV

EAV

-130

-8.0

-76

-6.9

a Values have been rounded to two significant figures. Values may not appear to add correctly due to rounding.
b The PM2.5 and Ch-related health benefits estimates use the larger of the two benefits estimates presented in Table
4-1. Monetized health benefits include those related to public health associated with changes in PM2 5 and ozone
concentrations. The health benefits are associated with several point estimates.

4-8


-------
4.3 References

Katsouyanni, K., Samet, J. M., Anderson, H. R., Atkinson, R., Le Tertre, A., Medina, S., . . .
Committee, H. E. I. H. R. (2009). Air pollution and health: a European and North
American approach (APHENA). Res Rep Health Eff Inst (142), 5-90. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/20Q73322

Pope, C. A., Lefler, J. S., Ezzati, M., Higbee, J. D., Marshall, J. D., Kim, S.-Y., . . . Robinson, A.
L. (2019). Mortality risk andfine particulate air pollution in a large, representative
cohort of US adults. Environmental Health Perspectives, 127(7), 077007.

Turner, M. C., Jerrett, M., Pope, A., Ill, Krewski, D., Gapstur, S. M., Diver, W. R., . . . Burnett,
R. T. (2016). Long-term ozone exposure and mortality in a large prospective study.
American Journal of Respiratory and Critical Care Medicine, 193(10), 1134-1142.
doi:10.1164/rccm.201508-16330C

U.S. BEA. (2025). Table 1.1.9. Implicit Price Deflators for Gross Domestic Product.
Washington, DC.

https://apps.bea.gov/iTable/?reqid=19&step=3&isuri=l&1921=survey& 1903=13

U.S. EPA. (2019). Integrated Science Assessment (ISA) for Particulate Matter (Final Report).
(EPA/600/R-19/188). Research Triangle Park, NC: U.S. Environmental Protection
Agency, Office of Research and Development, Center for Public Health and
Environmental Assessment. Retrieved from https://www.epa.gov/naaqs/particulate-
matter-pm-standardsintegrated-science-assessments-current-review

U.S. EPA. (2020). Integrated Science Assessment (ISA) for Ozone and Related Photochemical
Oxidants (FinalReport). (EPA/600/R-20/012). Washington DC: U.S. Environmental
Protection Agency. https://cfpub.epa.gOv/ncea/isa/recordisplav.cfm?deid=348522

U.S. EPA. (2022). Supplement to the 2019 Integrated Science Assessment for Particulate Matter
(FinalReport). (EPA/600/R-22/028). Research Triangle Park, NC: U.S. Environmental
Protection Agency, Office of Research and Development, Center for Public Health and
Environmental Assessment.

https://cfpub.epa.gOv/ncea/i sa/recordisplay.cfm?deid=354490

U.S. EPA. (2024). EstimatingPM2.5- and Ozone-Attributable Health Benefits: 2024 Update.
Research Triangle Park, NC: U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Health and Environmental Impact Division. Retrieved
from https://www.epa.gov/system/files/documents/2024-06/estimating-pm2.5-and-ozone-
attributable-health-benefits-tsd-2024.pdf

U.S. EPA Science Advisory Board. (2024). Review ofBenMAP and Benefits Methods. (EPA-
SAB-24-003). U.S. Environmental Protection Agency.

https://sab.epa.gov/ords/sab/f?p=l 14:18:11364624237840: ::18:P18_ID:2617#report

4-9


-------
Wu, X., Braun, D., Schwartz, J., Kioumourtzoglou, M. A., & Dominici, F. (2020). Evaluating
the impact of long-term exposure to fine particulate matter on mortality among the
elderly. Sci Adv, 6(29), eaba5692. doi:10.1126/sciadv.aba5692

Zanobetti, A., & Schwartz, J. (2008). Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. Am J Respir Crit Care Med,
177(2), 184-189. doi:10.1164/rccm.200706-8230C

4-10


-------
5 SOCIAL COSTS AND ECONOMIC IMPACTS

This section discusses potential energy market impacts, economy-wide social costs and
economic impacts, small entity impacts, and labor impacts associated with the proposed action.
The social cost and economy-wide impacts are estimated using EPA's SAGE model. Note that
SAGE does not currently estimate changes in emissions nor account for environmental impacts
resulting from the proposed action. For additional discussion of impacts on fuel use and
electricity prices, see Section 3.

5.1 Energy Market Impacts

The energy sector impacts presented in Section 3 of this RIA include potential changes in
the prices for electricity (the change in retail electricity prices reported here is a national average
across residential, commercial, and industrial consumers), natural gas, and coal resulting from
the proposed action. Table 5-1 summarizes the impact of these potential changes on other
markets. We refer to these changes as secondary market impacts.

The projected energy market and electricity retail rate impacts of the proposed action are
discussed more extensively in Section 3 of the 2024 CPS RIA, which also presents projections of
power sector generation and capacity changes by technology and fuel type.

Table 5-1 Summary of Certain Energy Market Impacts



2028

2030

2035

2040

2045

Retail electricity prices

1%

-0%

-1%

-0%

-1%

Average price of coal delivered to the power sector

1%

1%

-0%

-0%

32%

Coal production for power sector use

6%

4%

21%

-15%

84%

Price of natural gas delivered to power sector

2%

-0%

-3%

-0%

-0%

Price of average Henry Hub (spot)

2%

1%

-3%

-0%

-0%

Natural gas use for electricity generation

1%

2%

-4%

-0%

-2%

5.2 Economy-wide Social Costs and Economic Impacts

This section analyzes the potential economy-wide impacts of the proposed action using a
computable general equilibrium (CGE) model. CGE models are designed to capture substitution
possibilities between production, consumption, and trade; interactions between economic
sectors; and interactions between a policy shock and pre-existing market distortions, such as

5-1


-------
taxes that have altered consumption, investment, and labor decisions. As such, CGE models can
provide insights into the effects of regulation that occur outside of the directly regulated sector
because they are able to represent the entire economy in equilibrium in the baseline and under a
regulatory or policy scenario. A CGE model can also be used to estimate the social cost of a
regulation (or the social benefit of removing a regulation).25

For this analysis, we use version 2.1.1 of the EPA's SAGE (SAGE is an Applied General
Equilibrium) model of the U.S. economy (Marten et al., 2024).26 The SAGE model is a forward-
looking, intertemporal CGE model that assumes that for some discrete period of time an
economy can be characterized by a set of conditions in which supply equals demand in all
markets (referred to as equilibrium). When the imposition of a regulation alters conditions in one
or more markets, the SAGE model estimates a new set of relative prices and quantities for all
markets that return the economy to a new equilibrium. The social cost of the regulation is
estimated as the change in economic welfare in the post-regulation simulated equilibrium from
the pre-regulation "baseline" equilibrium. Table 5-2 summarizes model dimensions (time
periods, sectors, regions, representative households, and capital stocks). More details on the
SAGE model including complete documentation, source code, and build-stream are available on
the EPA's website.27

25	As discussed in EPA's Guidelines for Preparing Economic Analyses, social costs are the total economic burden of

a regulatory action (U.S. EPA, 2024). 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 because of reallocating some resources towards pollution mitigation.

26	To ensure that SAGE is consistent with economic theory and reflects the latest science, EPA initiated a separate

SAB panel to conduct a technical review of SAGE, completed in August 2020 (U.S. EPA Science Advisory
Board, 2020). Peer review of SAGE was in accordance with requirements laid out for a Highly Influential
Science Assessment (HISA) consistent with OMB guidelines. The report included recommendations for refining
and improving the model, including several changes that the SAB advised EPA to incorporate before using the
model in regulatory analysis (denoted as Tier 1 recommendations by the SAB). These Tier 1 recommendations
are incorporated into the model version used in this analysis (v2.1.1), as are several of the SAB's other medium-
and long-run recommendations.

27	https://www.epa.gov/environmental-economics/cge-modeling-regulatory-analysis

5-2


-------
Table 5-2 SAGE Dimensional Details

Time

Sectors

Census

Households

Capital

Periods

Regions

(income)

Vintage

2016-2081

Agriculture, forestry, fishing, and hunting

Northeast

<3 0k

Extant

(5-year

Crude oil

South

30-50k

New

time steps)

Coal mining

Midwest

50-70k





Metal ore and nonmetallic mineral mining

West

70-150k





Electric power



>150k





Natural gas









Water, sewage, and other utilities









Construction









Food and beverage manufacturing









Wood product manufacturing









Petroleum refineries









Chemical manufacturing









Plastics and rubber products manufacturing









Cement manufacturing









Primary metal manufacturing









Fabricated metal product manufacturing









Electronics and technology manufacturing









Transportation equipment manufacturing









Other manufacturing









Transportation









Truck transportation









Services









Healthcare services







5.2.1 Linking IPM Partial Equilibrium Model to SA GE CGE Model

The approach for linking the model outputs from IPM to SAGE model inputs is described
in Section 5.2.3 of the 2024 CPS RIA and in Schreiber et al., (2023). IPM incremental costs are
translated into the SAGE framework by: (1) mapping IPM model years to SAGE model years;28
(2) mapping IPM regions to SAGE regions; (3) splitting delivered fuel costs to separate

28 IPM year 2028 is mapped to SAGE model year 2026. Subsequent IPM years (2030-2055) are mapped to the
SAGE model year that is one year later (2031-2056). Because SAGE has a longer time horizon than IPM (to
2081), IPM incremental costs in 2055 are expected to continue into the future and are mapped to SAGE model
years 2061-2081. Due to this mapping methodology, the different timeframes of the models, and the internal
discount rate in SAGE, the discounted value of the resource costs used as inputs to the SAGE model in this
section may differ from the corresponding results in other sections of this RIA. For a final action, the EPA is
considering an alternative mapping between IPM and SAGE model years in which the IPM inputs to SAGE are
further adjusted to improve consistency in how the information from IPM is used as an input to SAGE.

5-3


-------
transportation costs; (4) mapping variable operations and maintenance costs to specific inputs in
SAGE according to the reference cost structure in the model; (5) attributing fixed operations and
maintenance costs to labor; (6) attributing incremental costs on existing and new generation to
production with extant and new capital, respectively;29 and (7) removing taxes and other transfers
from capital payments using the difference between the capital charge rate and the capital
recovery factor to recover the real resource costs.

Because SAGE does not include an explicit representation of the Inflation Reduction Act
of 2022 (IRA) in the baseline, the model linkage methodology must be adjusted to account for
IRA investment and production tax credits (i.e., ITC/PTC and 45Q). We calibrate the model to
match both the real resource requirements for the expected compliance pathway and the impact
of the IRA tax credits on the compliance expenditure for the electricity sector. The SAGE model
is closed by assuming the government budget is balanced through lump sum transfers with
households. Aggregate changes in government budgets can occur in model simulations due to
changes in the use of the IRA tax credits and changes in revenues from other taxes (e.g., output,
capital, and labor) as the economy adjusts in response to the proposed rules.

5.2.2 Results

This section summarizes the estimated economy-wide impacts of the proposed action.
SAGE model results include aggregate social costs, macroeconomic impacts, sectoral impacts,
and distributional impacts. Note that SAGE does not currently estimate changes in emissions nor
account for the effects of changes in environmental quality on the economy (i.e., the social
benefits of reducing environmental externalities).

5.2.2.1 Economy-wide Social Costs

Table 5-3 presents the economy-wide, general equilibrium social costs of the proposed
action, calculated as equivalent variation. In this context, equivalent variation is an estimate of
the amount of money that society would be willing to pay to avoid the compliance requirements

29 Production with extant and new capital is not equivalent to differentiating existing and new generation in the IPM
modeling framework. For example, the lifespan of existing generators in IPM can be extended through
investments in ways that are not directly comparable to production with extant capital in the SAGE model. In
this analysis, we attribute all incremental costs associated with existing generation to production with extant
capital until 2051. Incremental costs on existing generation in model years after 2051 are levied on production
with new capital.

5-4


-------
of the proposed action, setting aside health, climate, and other benefits (quantified or described
qualitatively elsewhere in the RIA). For comparison, Table 5-3 also presents the compliance
costs estimated by IPM to be paid by the electricity sector for real resources - and which exclude
all transfer payments - mapped to the SAGE model years. For both the compliance costs and the
general equilibrium social costs, Table 5-3 presents the present value and annualized costs using
a discount rate of 4.5 percent, which is consistent with the internal discount rates in the SAGE
model. Compliance costs and transfer changes are presented as they are input into the SAGE
model. Present value and equivalent annualized value estimates of the IPM inputs to SAGE
reported in Table 5-3 are not comparable to those reported in Sections 3 or 6 and are provided
here for transparency and as a point of comparison for the social cost estimates.

The annualized social cost estimated in SAGE for the proposed action is approximately -
$1.58 billion (2024 dollars) between 2026 and 2047 using the 4.5 percent discount rate that is
consistent with the internal discount rates in the model.30 Under the assumption that the
compliance costs from IPM in 2056 continue until 2081, the equivalent annualized value for
social costs in the SAGE model is -$1.81 billion (2024 dollars) over the period from 2026 to
2081, again using a 4.5 percent discount rate. This social cost estimate reflects the combined
effects of the proposed action and interactions with IRA tax credits for specific technologies that
are expected to see decreased use in response to the proposed action. We are currently not able to
identify their relative roles. See Section 5.2.4.1 of the 2024 CPS RIA for a discussion of the
differences between social cost and compliance cost estimates.

30 The SAGE model estimates the present value of costs (i.e., equivalent variation) for each representative household
in the model and sums those estimates to calculate the present value of social costs. The present value of costs
for a representative household is based on its calibrated intertemporal utility function and the equilibrium
solution. Implicit in those estimates are endogenous discount rates that vary by household and over time. The
intertemporal preferences of households are calibrated such that their average discount rate over the first 20 years
of the model is consistent with a discount rate of 4.5 percent, which based on the effective marginal capital tax
rate in the model, is consistent with a 7.0 percent social rate of return to capital. See Section 3.4 of the SAGE
model documentation at https://www.epa.gov/environmental-economics/cge-modeling-regulatory-analysis.

5-5


-------
Table 5-3 Compliance Costs, Transfers, and Social Costs (billion 2024 dollars)

SAGE Model Year

Compliance Costs - Input to
SAGE (Excluding Transfers)

Change in Transfers
- Input to SAGE

General Equilibrium
Social Costs

2026

1.75

0.12

-1.37

2031

0.38

-0.07

-1.52

2036

-7.38

5.76

-1.65

2041

-6.05

5.34

-1.77

2046

-2.34

-1.63

-1.91

2051

-1.13

-0.59

-2.07

2056

-0.89

-1.17

-2.24

Present Value

-31.67
-2.00

28.71
1.81

-25.00
-1.58

(2026 to 2047, 4.5%)
Equivalent Annualized
Value (2026 to 2047, 4.5%)

Present Value

-38.40

21.82

-40.58

(2026 to 2081, 4.5%)

Equivalent Annualized
Value (2026 to 2081, 4.5%)

-1.72

0.98

-1.81

Notes: Social costs are calculated as equivalent variation. Present value and annualized cost estimates are calculated
by interpolating between SAGE model years and use a discount rate of 4.5 percent, which is consistent with the
internal discount rate in SAGE. Compliance costs and the change in the transfer amounts are calculated from the
IPM outputs. Transfers include changes in tax payments on capital and production and investment tax credits (e.g.,
the 45Q tax credit). Negative transfer values reflect decreases in net additional payments out of the sector or
increases in payments into the sector (e.g., subsidies) due to the proposed action. Incremental monitoring and
reporting costs are not accounted for in this analysis. Compliance costs and transfer changes are reported as they are
input into the SAGE model. Section 5.2.1 discusses assumptions on mapping IPM model years to SAGE model
years. Present value and equivalent annualized value estimates are based on this mapping and are therefore not
directly comparable to estimates in Sections 3 and 6.

Equivalent variation is a theoretically appropriate measure of social cost (U.S. EPA
Science Advisory Board, 2017). However, equivalent variation by definition does not provide
detail on how the social costs of a regulation are borne over time.31 Changes in real full
consumption, which is the value of changes in the consumption of goods, services, and leisure,
better communicates the temporal evolution of impacts from a regulation. As a result, changes in
full consumption may also more closely align with the temporal pattern of compliance costs in
Table 5-3. Furthermore, the present value of this cost metric closely approximates equivalent
variation. Changes in real full consumption has been used to approximate social cost in CGE
models, for example with alternative expectations that constrain the intertemporal utility
maximization problem (McKibbin and Wilcoxen, 1999). For these reasons, the EPA is
considering presenting changes in full consumption from SAGE as an alternative point of

31 See Table 5-3 note for how we allocated equivalent variation over time.

5-6


-------
comparison to the (negative) social benefits of a regulation in future regulatory analyses to better
reflect the temporal pattern of impacts and provide a better point of comparison to compliance
cost estimates calculated over a limited time horizon.

5.2.2.2 Macroeconomic Impacts

The estimated percent change in real gross domestic product (GDP), or the real value of
the goods and services produced by the U.S. economy, and its components are presented in
Figure 5-1. GDP is defined as the sum of the value (price times quantity) of all market goods and
services produced in the economy and is equal to Consumption (C) + Investment (I) +
Government (G) + (Exports (X) - Imports (M)). The proposed action is estimated to decrease
GDP in 2026 and 2031 by 0.015 percent and 0.020 percent due to decreases in near term
investment, but subsequently increase GDP with a peak increase of 0.017 percent in 2036.32

0.10%-

2020	2030	2040	2050

Year

Figure 5-1 Percent Change in Real GDP and Components

32 GDP is a measure of economic output and not a measure of social welfare. Thus, the expected social cost of a
regulation will generally not be the same as the expected change in GDP (U.S. EPA, 2015). U.S. EPA Science
Advisory Board (2017) notes: "GE models are strongly grounded in economic theory, which allows social costs
to be evaluated using equivalent variation or other economically-rigorous approaches. Simpler measures, such as
changes in gross domestic product or in household consumption, do not measure welfare accurately and are
inappropriate for evaluating social costs."

5-7


-------
Figure 5-1 also reports changes in the components of GDP from the expenditure side.
The proposed action is expected to decrease investments in the electricity sector, leading to a
reduction in aggregate investment in 2026 and 2031 (0.09 percent and 0.13 percent,
respectively). A reduction in investment reallocates resources toward consumption and as a
result, the proposed action increases consumption throughout the model time horizon. Aggregate
investment is expected to then increase and then fall again in later model years. The proposed
action is also expected to impact the net trade balance, including a modest increase in net exports
in the initial years through changes in domestic relative prices due to avoided compliance.

5.2.2.3 Sectoral and Labor Impacts

Figure 5-2 presents the percent change in output and real output prices for each sector in
model years 2026, 2031, 2036, and 2041.33 Changes in output reflect the estimated shifts in
generation sources in addition to an economy-wide demand response to increases in electricity
prices. Similarly, the estimated changes in sector output prices reflect compliance cost reductions
associated with the proposed action as well as demand side increases in electricity use from both
firms and households. As the price of electricity falls, the economy is expected to increase
demand for electricity through a variety of pathways. Measured in terms of percent change from
the baseline, output and price changes in the electricity, coal mining, and natural gas sectors are
expected to be relatively larger than in other sectors of the economy. Modest output increases are
estimated in some relatively more energy-intensive sectors (e.g., chemical manufacturing) and
those that support coal use in the electricity sector (e.g., transportation), whereas output
decreases in sectors associated with capital formation in 2026 and 2031 due to reductions in
investments attributable to the proposed action.

Impacts to sectoral labor demand follows similar trends across sectors to changes in
sectoral output. Labor demand impacts are relatively larger in the electricity, coal mining, and
natural gas sectors, whereas labor demand responses in other sectors of the economy tend to be
driven by increases in demand for labor in sectors that are more energy-intensive and reductions

33 CGE models report prices in relative terms. We denominate output prices in terms of a consumer price index
(CPI) internal to the SAGE model, which reflects the overall change in end-use prices for the bundle of goods
demanded by households. Characterizing prices relative to this CPI allows a comparison of changes in the
magnitude of output prices to overall trends in the economy (i.e., a percentage change that is positive reflects a
price that increases more than the average price changes across the economy).

5-8


-------
in sectors associated with capital formation. Figure 5-3 presents the percent change in select
sectors for 2026, 2031, 2036, and 2041. Across model years, the percent change in total
economy-wide labor demand ranges from -0.005 percent to 0.004 percent. As with many other
CGE models, SAGE assumes an economy with full employment, meaning that the labor market
in the model adjusts to the new equilibrium such that there is no involuntary unemployment (i.e.,
all workers that want to work at the new prevailing wage can find a job). Any net changes in
employment levels are associated with voluntary changes in labor.34 In contrast, Section 5.4 of
this RIA characterizes employment impacts of the proposed action in the power and fuels
sectors, providing detailed estimates by capacity and fuel type, without accounting for changes in
prices, wages, and interactions with other sectors in the economy.

34 While SAGE does not capture any near-term transition dynamics in the labor market, recent economics research
suggests that they likely are a small component of overall welfare costs of environmental regulation (Rogerson
2015; Hafstead and Williams 2018).

5-9


-------
Healthcare services
Services

Truck transportation
T ransportation
Other manufacturing

Transportation equipment manufacturing
Electronics and technology manufacturing
Fabricated metal product manufacturing
Primary metal manufacturing
Cement manufacturing

Plastics and rubber products manufacturing

o

+-> Chemical manufacturing
u	6

$ Petroleum refineries

Wood product manufacturing
Food and beverage manufacturing
Construction

Water, sewage, and other utilities
Natural gas

Metal ore and nonmetalic mineral mining
Coal mining
Electric power
Crude oil

Agriculture, forestry, fishing and hunting

-1

Year # 2026 # 2031 2036 # 2041

Real Output Price

.5% 1.0% -1.5% -1.
% Change

Figure 5-2 Percent Change in Sectoral Output and Real Output Prices

5-10


-------
Electronics and tech
nology manufacturing

Transportation equip
ment manufacturing

Primary metal manufa
cturing

Construction

Cement manufacturing

Other manufacturing

Fabricated metal pro
duct manufacturing

Plastics and rubber
products manufacturing

Agriculture, forestr
y, fishing and hunting

^ Wood product manufac
turing

Truck transportation

Metal ore and nonmet
alic mineral mining

Chemical manufacturi

ng

Petroleum refineries

Food and beverage ma
nufacturing

Healthcare services

Water, sewage, and o
ther utilities

T ransportation

n

]
i
]

]

-0.2% -0.1%

0.1% -0.2% -0.1% 0.0% 0.1% -0.2% -0.1% 0.0% 0.1% -0.2% -0.1%

% Change

Figure 5-3 Percent Change in Labor Demand in Select Sectors

5.2.2.4 Distributional Impacts

The social costs of regulation are ultimately borne by households through changes in
final goods prices or changes in labor, capital, and resource income. For model years, overall
consumer prices are expected to fall by 0.01 percent to 0.03 percent. The SAGE model also
characterizes representative households by income quintiles in each of the four Census regions.
This allows the social costs to be separately estimated across the income distribution and for

5-11


-------
different regions of the country, as presented in Figure 5-4.35 In general, avoided household
social costs are expected to increase with income. Estimates in Figure 5-4 reflect a combined
effect of the proposed action requirements and interactions with IRA subsidies that are expected
to see decreased use in response to the proposed action.36

s

I

<30k	30-50k	50-70k	70-150k

Income Quintile

>150k

Regions

Midwest
^ Northeast

South
A West

Figure 5-4 Distribution of General Equilibrium Social Costs

SAGE's representation in the incidence of regulatory changes on domestic households is
affected by its ability to distinguish between the sources and uses of income for domestic and
foreign consumers and asset owners. For example, SAGE distinguishes between domestic and
foreign asset owners of government debt (e.g., bonds) and thus the change in the value of debt

35	Distributional cost estimates are calculated for the period 2025 to 2047 and divided by the total number of

households of a given income quintile and region using 2016 estimates from the Census' Current Population
Survey.

36	A regulation may affect the value of government expenditures through relative prices of goods and services

purchased by the government. In addition, it may affect tax revenues through impacts on the value of the base for
ad valorem taxes (e.g., labor and capital taxes). In these cases, a CGE model must implement a closure rule to
ensure that the government has the funds necessary to support its expenditures. A common assumption in CGE
models is to balance the government's budget through lump sum transfers between households and the
government as a non-distortionary approach to closing the model. This is the approach used in the SAGE model.
Given uncertainties in the accounting for the IRA subsidies in this analysis, we are unable to determine the
relative role of this effect in the distributional estimates at this time.

5-12


-------
due to a regulation. It further represents the effects of trade on income earned in the U.S. and the
costs of goods and services to domestic households. However, SAGE does not distinguish
between the ownership of physical capital in the U.S. between domestic and foreign investment
by sector. SAGE also does not account for how the value domestic owned assets outside the U.S.
may be affected by a regulatory change. Further refining how changes in U.S. regulation affect
the welfare of domestic households in SAGE is an area of ongoing work.

5.2.3 Limitations

The SAGE model and methodology for aligning IPM outputs for use as inputs in SAGE
reflect the best available science for conducting economy-wide modeling of the proposed action.
However, both the use of SAGE in a regulatory analysis and the framework for linking IPM with
the SAGE model are subject to uncertainty and limitations. See Section 5.2.5 of the 2024 CPS
RIA for an overview of these potential limitations.

5.3	Small Entity Analysis

As outlined in Section 5 of the 2024 CPS RIA, the EPA assessed the impact of the 2024
CPS on small entities by 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. Of the 14 entities that own NGCC units considered in this analysis, three were projected
to experience compliance costs greater than or equal to 1 percent of generation revenues in 2035
and none were projected to experience compliance costs greater than or equal to 3 percent of
generation revenues in 2035. This proposed action will no longer require these compliance costs
and will therefore not have a significant economic impact on a substantial number of small
entities.

5.4	Labor Impacts

5.4.1 Overview of Methodology

See the 2024 CPS RIA and the U.S. EPA Methodology for Power Sector-Specific
Employment Analysis in the docket for a detailed overview of the methodology followed for this
analysis, including all underlying assumptions and the types of employment represented. This
analysis is a complement to the economy-wide analysis presented in Section 5.2 of the 2024 CPS

5-13


-------
RIA using the SAGE model, which projects medium- to long-run shifts in the expected use of
labor across aggregate sectors as a result of those rules.

5.4.2	Overview of Power Sector Employment

Employment in electric power generation, as well as coal and natural gas extraction, is
the focus of this analysis using available data. Other segments not discussed here include the
extraction or production of other fuels (e.g., hydrogen), energy efficiency, transmission, and
distribution. To contextualize the analysis, this section presents national data from the 2020
United States Energy and Employment Report (USEER), which reports employment data from
2019. See Section 5.4.2 of the 2024 CPS RIA for an explanation on the importance of using this
data instead of data from more recent years.

In 2019, the electric power generation sector employed nearly 900,000 people. Relative
to 2018, this sector grew by over 2 percent. Job losses related to nuclear and coal generation
were offset by increases in employment related to other generating technologies, including
natural gas, solar, and wind. The largest component of total 2019 employment in this sector was
construction (33 percent). Other components of the electric power generation workforce include
utility workers (20 percent), professional and business service employees (20 percent),
manufacturing (13 percent), wholesale trade (8 percent), and other (5 percent). In 2019, jobs
related to solar and wind generation represented 31 percent and 14 percent of total jobs,
respectively, and jobs related to coal generation represented 10 percent of total employment.

In addition to generation-related employment, we also look at employment related to coal
and natural gas in the electric power sector. In 2019, the coal industry employed about 75,000
workers. Mining and extraction jobs represented most coal-related employment in 2019 (74
percent). Likewise, about 60 percent of the 276,000 jobs in the 2019 natural gas fuel sector were
related to mining and extraction.

5.4.3	Projected Sectoral Employment Changes

Electric generating units subject to the proposed action will not implement various GHG
mitigation measures originally projected under the modeling of the 2024 CPS RIA. See Section
3.6.3 of the 2024 CPS RIA for more details on these power sector modeling projections.

5-14


-------
Based on these power sector modeling projections, we estimate a reduction in 7,900
construction-related job-years related to the reduced installation of new pollution controls under
the proposed action in 2030 and a reduction of 10,200 construction-related job-years for the
reduction of new pollution controls in 2035. We estimate a reduction of 45,300 job-years in 2028
related to the reduction of constructed new capacity in that year, and a reduction of
approximately 181,300 construction-related job-years in 2035 related to the reduction of
constructed battery storage systems. In 2030 and 2040, however, we estimate increases of 21,000
construction-related job-years and 107,500 construction-related job-years, respectively.

The changes in the year-over-year results primarily from relatively small temporal
changes in the projected deployment of renewable energy and battery storage capacity in the
modeling. Of note is that the employment factors related to battery storage are relatively high
and relatively uncertain, as it is a relatively new technology for which there is limited data to
base assumptions. Without including battery storage in the total estimate, we estimate the
proposed action to result in decreases in 2028, 2030, 2035, and 2045 of 46,100, 7,300, 18,300,
and 42,600 job-years, respectively, related to the decreased construction of new capacity in those
years and an increase of 11,300 job-years in 2040.

Construction-related job-year changes are one-time impacts, occurring each year of the
multi-year periods during which construction of new capacity is completed. Construction-related
figures in Table 5-4 represent a point estimate of incremental changes in construction jobs for
each year (e.g., for a three-year construction projection, this table presents one-third of the total
jobs for that project).

Table 5-4 Changes in Labor Utilization: Construction-Related (number of job-years of
employment in a single year	



2028

2030

2035

2040

2045

New Pollution Controls

<100

-7,900

-10,200

<100

<100

New Capacity

-45,300

21,000

-181,300

107,500

-41,800

Note: These values describe changes under the proposed action relative to a projected baseline. A large share of the
construction-related job years is attributable to construction of battery storage, a relatively new technology on which
there is limited data to base labor assumptions.

We also estimate changes in the number of job-years related to recurring non-
construction employment. Recurring employment changes are job-years associated with annual
recurring operating, maintenance, and fuel extraction jobs. Newly built generating capacity

5-15


-------
creates a recurring stream of positive job-years, while retiring generating capacity and avoiding
capacity builds create a stream of negative job-years. The proposed action is projected to
generally forgo or delay the replacement of relatively labor-intensive coal capacity with less
labor-intensive capacity, which results in an overall increase of non-construction jobs between
2028 and 2045. The total net estimated increase of recurring employment in any given analysis
year is a small percentage of total 2019 power sector employment reported in the 2020 USEER
(approximately 900,000 generation-related jobs, 75,000 coal-related jobs, and 276,000 natural
gas-related jobs). Table 5-5 provides detailed estimates of recurring non-construction
employment changes.

Table 5-5 Changes in Labor Utilization: Recurring Non-Construction (number of job-
years of employment in a single year)	



2028

2030

2035

2040

2045

Pollution Controls

<100

200

300

<100

-100

Existing Capacity

2,000

3,900

8,700

5,100

7,600

New Capacity

-3,000

-3,400

-4,100

-3,000

-6,300

Fuels (Coal, Natural Gas, Uranium)

1,200

400

800

-1,100

1,100

Coal

900

300

1,800

-1,100

1,200

Natural Gas

300

<100

-1,100

<100

-100

Uranium

<100

<100

<100

<100

<100

Note: These values describe changes under the proposed action relative to a projected baseline. "<100" denotes an
increase or decrease of less than 100 job-years; Numbers may not sum due to rounding

5.4.4 Conclusions

Generally, there are significant challenges when trying to evaluate the employment
effects from the repeal of an environmental regulation due to a wide variety of other economic
changes to labor markets and the state of the macroeconomy. The analysis of employment
impacts in this section evaluates first-order employment effects at a detailed level for
construction and recurring non-construction employment utilization for pollution control
equipment and operation of different capacity and fuel types. For EGUs, the proposed action
may result in decreases and shifts over time of construction-related jobs related to the installation
of new pollution controls and construction of new capacity. The proposed action also is projected
to result, generally, in an increase of relatively labor-intensive coal capacity compared to less
labor-intensive generating capacity, which results in an overall increase of non-construction jobs.
It is important to note that this analysis does not estimate the employment losses likely to result

5-16


-------
from the expected decrease in development and construction of new transmission and
distribution capacity throughout the U.S. due to the proposed action.

5-17


-------
5.5 References

Hafstead, M. A. C., & Williams, R. C. (2018). Unemployment and environmental regulation in
general equilibrium. Journal of Public Economics, 160, 50-65.
doi:https://doi.org/10.1016/j.jpubeco.2018.01.013.

Marten, A., Schreiber, A., and Wolverton, A. (2024). SAGE Model Documentation (2.1.1).
Washington, DC. https://www.epa.gov/environmental-economics/cge-modeling-
regulatory-analysis.

McKibbin, W. J., & Wilcoxen, P. J. (1999). The theoretical and empirical structure of the G-
Cubed model. Economic Modelling, 16(1), 123-148.

OMB. (2004). Issuance of OMB's 'Final Information Quality Bulletin for Peer Review.'
Washington, DC. https://cfpub.epa.gov/si/m05-03.pdf.

Rogerson, R. (2015). A Macroeconomic Perspective on Evaluating Environmental Regulations.
Review of Environmental Economics and Policy, 9(2), 219-238. doi:10.1093/reep/rev005

Schreiber, A., Evans, D., Marten, A., Wolverton, A., Davis, W. (2023). Evaluating Economy-
wide Effects of Power Sector Regulations Using the SAGE Model. Working Paper.
Retrieved from https://www.epa.gov/environmental-economics/evaluating-economy-
wide-effects-power-sector-regulations-using-sage-model.

U.S. EPA. (2015). Economy-Wide Modeling: Social Cost and Welfare White Paper.
https://www.epa.gov/system/files/documents/2023-
02/CGE%20social%20cost%20white%20paper%20final.pdf.

U.S. EPA. (2024). Guidelines for Preparing Economic Analyses (3rd edition). Report number
EPA-240-R-24-001. Washington, DC.

U.S. EPA Science Advisory Board. (2017). SAB Advice on the Use of Economy-Wide Models in
Evaluating the Social Costs, Benefits, and Economic Impacts of Air Regulations. (EPA-
SAB-17-012). Washington, DC.

U.S. EPA Science Advisory Board. (2020). Technical Review of EPA's Computable General
Equilibrium Model, SAGE. (EPA-SAB-20-010). Washington, DC.

5-18


-------
6 COMPARISON OF BENEFITS AND COSTS

6.1	Introduction

This section provides the estimates of the costs, benefits, and net benefits of the proposed
action, as well as discusses unquantified impacts. The reduced expenditures on compliance costs
reported in this section are not social costs; instead, we use compliance costs as a proxy for
social costs. The projected real resource costs and economy-wide social costs are separately
estimated and discussed in Section 3.2.4 (real resource costs) and Section 5.2 (economy-wide
social costs), respectively, but those estimates are not applied in this section. Therefore, in this
section, we do not account for changes in costs and benefits due to changes in economic welfare
in the broader economy arising from shifts in production and consumption that may be induced
by the proposed action. Furthermore, costs and benefits due to interactions with pre-existing
market distortions outside the electricity sector are omitted, as are changes in social costs that
may be associated with the net change in transfers attributable to the proposed action. Additional
limitations of the analysis and sources of uncertainty are described throughout the RIA and
summarized later in this section.

6.2	Methods

The EPA calculated the PV and EAV of costs, benefits, and net benefits for the years
2026 through 2047, using the discount rates of 3 percent and 7 percent from the perspective of
2025. The calculations of PV and EAV use an annual stream of values from 2026 to 2047
timeframe.

The EPA used IPM to estimate cost and emission changes for the projection years 2028,
2030, 2035, 2040 and 2045. In the IPM modeling for this RIA, the 2028 projection year is
representative of 2028 and 2029, the 2030 projection year is representative of 2030 and 2031, the
2035 projection year is representative of 2032 to 2037, the 2040 projection year is representative
of 2038 to 2041, and the 2045 projection year is representative of 2042 through 2047. Estimates
of costs and emission changes in other years are determined from the mapping of projection
years to the calendar years that they represent. Consequently, the cost and emission estimates
from IPM in each projection year are applied to the years which it represents.

6-1


-------
PM2.5 and ozone-related health benefits are based on projection year emission estimates
and also account for year-specific variables that influence the size and distribution of the
benefits. These variables include population growth, income growth, and baseline mortality
rates.

The 2024 CPS RIA followed the EPA's historical practice of using a technology-rich
partial equilibrium model of the electricity and related fuel sectors to estimate the incremental
costs of producing electricity under the requirements of major EPA power sector rules. In
Section 5.2 of this RIA, we also included an economy-wide analysis that considers additional
facets of the economic response to the proposed action, including the full resource requirements
of the expected compliance pathways, some of which were paid for through subsidies in the
partial equilibrium analysis in the 2024 CPS RIA final rules illustrative scenario.

6.3 Results

Table 6-1 and Table 6-2 compare benefits with costs as well as present net benefits
estimated from this proposed action. For comparison, the social cost estimates in the economy-
wide analysis discussed in Section 5.2 also exceed the estimated benefits of the proposed action,
such that the net benefits are negative using the economy-wide cost estimate.

6-2


-------
Table 6-1 Net Benefits of the Proposed Action (billion 2024 dollars, undiscounted)

PM2.5 and 03-related Health „	,, ^	^

Benefits b	Compliance Costs c	Net Benefits



3%

7%



3%

7%

2026

-

-

-0.01

0.01

0.01

2027

-

-

0.00

0.00

0.00

2028

-6.8

-6.0

1.6

-8.3

-7.6

2029

-6.9

-6.2

1.6

-8.5

-7.7

2030

-4.7

-4.2

0.27

-5.0

-4.4

2031

-4.8

-4.2

0.27

-5.1

-4.5

2032

-16

-14

-1.5

-15

-13

2033

-16

-15

-1.5

-15

-13

2034

-17

-15

-1.5

-15

-13

2035

-17

-15

-1.5

-16

-14

2036

-17

-16

-1.5

-16

-14

2037

-18

-16

-1.5

-16

-14

2038

0.39

0.35

-0.71

1.10

1.10

2039

0.40

0.36

-0.71

1.10

1.10

2040

0.41

0.36

-0.71

1.10

1.10

2041

0.42

0.37

-0.71

1.10

1.10

2042

-9.2

-8.2

-4.0

-5.2

-4.2

2043

-9.3

-8.3

-4.0

-5.3

-4.3

2044

-9.4

-8.4

-4.0

-5.4

-4.3

2045

-9.5

-8.5

-4.0

-5.5

-4.4

2046

-9.6

-8.6

-4.0

-5.6

-4.5

2047

-9.7

-8.6

-4.0

-5.7

-4.6

Non-Monetized Disbenefits

From increases in HAP emissions and GHG emissions

To water quality and availability

To ecosystems from increases in emissions of CO2, NOx, SO2, PM, and HAP

From increases in exposure to ambient NO2 and SO2

Decreased visibility (increased haze) from PM2 5 emissions increases

a Annual values from 2026 to 2047 are not discounted. PV and EAV values discounted to 2025. Values have been
rounded to two significant figures and are presented no smaller than two decimal places. Values may not appear to
add correctly due to rounding.

b The PM2.5 and 03-related health benefits estimates use the larger of the two benefits estimates presented in
Table 4-1. Monetized health benefits include those related to public health associated with changes in PM2 5 and
ozone concentrations. The health benefits are associated with several point estimates.

0 The discount rate in IPM is 3.76 percent, as described in Section 3.

6-3


-------
Table 6-2 Net Benefits, Present Value, and Equivalent Annualized Value Estimates of
	the Proposed Action (billion 2024 dollars, discounted to 2025) a	



PM2.5 and 03-related Health
Benefits b

Compliance Costs c



Net Benefits



3%

7%

3%

7%

3%

7%

2026

-

-

-0.01

-0.01

0.01

0.01

2027

-

-

0.00

0.00

0.00

0.00

2028

-6.2

-4.9

1.4

1.3

-7.6

-6.2

2029

-6.2

-4.7

1.4

1.2

-7.5

-5.9

2030

-4.1

-3.0

0.23

0.19

-4.3

-3.2

2031

-4.0

-2.8

0.23

0.18

-4.2

-3.0

2032

-13

-8.9

-1.3

-0.96

-12

-7.9

2033

-13

-8.5

-1.2

-0.90

-12

-7.6

2034

-13

-8.1

-1.2

-0.84

-12

-7.3

2035

-13

-7.8

-1.1

-0.78

-12

-7.0

2036

-13

-7.4

-1.1

-0.73

-12

-6.7

2037

-12

-7.0

-1.1

-0.68

-11

-6.4

2038

0.27

0.1

-0.5

-0.29

0.75

0.44

2039

0.26

0.1

-0.5

-0.28

0.73

0.41

2040

0.26

0.1

-0.5

-0.26

0.72

0.39

2041

0.26

0.1

-0.4

-0.24

0.70

0.37

2042

-5.60

-2.6

-2.4

-1.3

-3.2

-1.3

2043

-5.50

-2.5

-2.4

-1.2

-3.1

-1.3

2044

-5.40

-2.3

-2.3

-1.1

-3.1

-1.2

2045

-5.30

-2.2

-2.2

-1.0

-3.1

-1.1

2046

-5.20

-2.1

-2.2

-0.97

-3.0

-1.1

2047

-5.10

-1.9

-2.1

-0.91

-3.0

-1.0



PM2.5 and 03-related Health
Benefits

Compliance Costs c



Net Benefits



3%

7%

3%

7%

3%

7%

PV

-130

-76

-19

-9.6

-110

-67

EAV

-8.0

-6.9

-1.2

-0.87

-6.8

-6.0

Non-Monetized Disbenefits

From increases in HAP emissions and GHG emissions
To water quality and availability
To ecosystems from increases in emissions of CO2, NOx, SO2, PM, and HAP
From increases in exposure to ambient NO2 and SO2
Decreased visibility (increased haze) from PM2 5 emissions increases

a Annual values from 2026 to 2047 are not discounted. PV and EAV values discounted to 2025. Values have been
rounded to two significant figures and are presented no smaller than two decimal places. Values may not appear to
add correctly due to rounding.

b The PM2.5 and Ch-related health benefits estimates use the larger of the two benefits estimates presented in Table
4-1. Monetized health benefits include those related to public health associated with changes in PM2 5 and ozone
concentrations. The health benefits are associated with several point estimates.

0 The discount rate in IPM is 3.76 percent, as described in Section 3.

6-4


-------
The compliance cost values in Table 6-1 and Table 6-2 above differ from the compliance
cost values in Table 1-1. In Table 1-1, we present positive compliance cost values as cost
savings.

6.4 Uncertainties and Limitations

Throughout the RIA, we considered several sources of uncertainty, both quantitatively
and qualitatively, regarding the emissions changes, benefits, and costs estimated for the proposed
repeal. We summarize these discussions as well as other important uncertainties here.

Compliance costs of the baseline: The IPM-proj ected cost estimates of private
compliance costs provided in this analysis and based on the modeling performed for the 2024
CPS RIA is intended to estimate the change in production and transmission costs to the power
sector in response, in this RIA, to the proposed action. As discussed in more detail in Section 3.7
of the 2024 CPS RIA, there are several key areas of uncertainty related to the electric power
sector that are worth noting, including assumptions about electricity demand, natural gas supply
and demand, longer-term planning by utilities, and assumptions about the cost and performance
of controls. Additional uncertainties in the cost analysis are introduced by the fact that the "true"
baseline in this RIA is different than the baseline modeling that informed the 2024 CPS RIA
which provides the estimates of compliance cost here.

Uncertainty in modeled CCS costs and CO2 reduction efficiency of the baseline: As

explained in Section V of the preamble, the EPA is proposing to determine that CCS with 90
percent capture is not the BSER for long-term existing coal-fired steam generating units because
it has not been adequately demonstrated and the costs are unreasonable. Furthermore, because it
is unlikely that infrastructure necessary for CCS can be deployed by the January 1, 2032,
compliance date, the EPA is proposing to determine that the degree of emission limitation in the
CPS for long-term coal-fired steam generating units is not achievable. Consequently, the EPA is
proposing to repeal the requirements in the emission guidelines pertaining to long-term existing
coal-fired steam generating units.

In the RIA for this action, which is based upon the 2024 CPS RIA, the EPA compliance
modeling indicated substantial uptake of CCS under the CPS relative to the baseline, particularly

6-5


-------
in analysis year 2035.37 If CCS is more costly or less effective at CO2 removal than modeled in
the 2024 CPS RIA, it is possible that there would have been less CCS uptake under the CPS. As
result, costs and benefits of the CPS final rule may have been different had alternative CCS cost
and removal efficiency assumptions been used, and we note this conclusion as an important
uncertainty in this proposal RIA.

Additionally, the 2024 CPS RIA baseline contained assumptions based on the latest
available data as of Summer 2023 around the total demand for electricity as well as announced
fleet retirements and additions. In light of changing market conditions, current projections point
towards a significantly higher electricity demand environment, while some owners and operators
have announced plans to delay retirement decisions. All else equal, both these trends may result
in cost savings and disbenefits that are higher than those estimated in this analysis, should the
EPA revisit its earlier assumptions.

As noted earlier, the EPA relies on the modeling conducted in support of the 2024 CPS in
this RIA. That modeling included CCS cost and performance assumptions for newNGCC builds
that included a 45Q subsidy stream that was available for twelve years, with model plants
choosing compliance pathways and dispatching based on relative economics. In that modeling,
we observe that under the final rules, 18 GW of economic NGCC additions occur by 2035, and
of these units 870 MW of NGCCs install CCS in 2035.

Model plants that are projected to install CCS are projected to continue to operate at
higher capacity factors post expiration of credits since they are either located in markets subject
to a CCS price or can take advantage of EOR opportunities. As noted in the preamble, the
majority of NGCC units would likely not be able to benefit from these market conditions, and
would therefore revert to lower levels of utilization post 45Q availability. This is consistent with
the lower uptake of CCS projected within this source category.

Monetizing CCh-related domestic climate benefits of the baseline: There are
significant uncertainties related to the monetization of greenhouse gases that include, but are not
limited to: the magnitude of the change in climate due to a change in GHG emissions; the

37 To account for parasitic load as a result of installation of CCS, IPM includes a heat rate penalty and capacity
penalty on model plants that incorporate CCS as well as CCS retrofit options. These assumptions are detailed in
chapter 6 of the IPM documentation, available at: https://www.epa.gov/system/files/documents/2024-04/chapter-
6-co2-capture-storage-and-transport.pdf

6-6


-------
relationship between changes in the climate and the economy and therefore, the resulting
economic impacts; future economic and population growth which are important for estimating
vulnerability, willingness to pay to avoid impacts, and the ability to adapt to future changes;
future technological advancements that would reduce vulnerability and impacts; the share of
impacts from GHG emissions that affect citizens and residents of the United States; and the
appropriate discount rates to use when discounting in an intergenerational context. Consistent
with the memorandum titled "Guidance Implementing Section 6 of Executive Order 14154,
entitled 'Unleashing American Energy'", the EPA did not monetize impacts from changes in
GHG emissions from this proposed action. Monetizing these impacts could potentially result in
flawed decision-making due to overreliance on highly uncertain values.

Monetized PM2.5 and ozone-related benefits of the baseline: The analysis of
monetized PM2.5 and ozone-related impacts includes many data sources as inputs that are each
subject to uncertainty. Input parameters include projected emissions inventories, projected
compliance methods, projected emissions changes, air quality data from models (with their
associated parameters and inputs), population data, population estimates, health effect estimates
from epidemiology studies, economic data, and assumptions regarding the future state of the
world (i.e., regulations, technology, and human behavior). When compounded, even small
uncertainties can greatly influence the size of the total quantified benefits. The uncertainty in
health effect estimates from epidemiology studies may be larger for ambient pollution levels
with fewer observations, such as very low levels. In order to quantitatively characterize the
uncertainty in the relationship between exposure and mortality, EPA reports multiple estimates
of (here forgone) avoided mortality based on different estimates of the exposure-mortality
relationship. To quantitively characterize additional sources of uncertainty, EPA reports the 2.5th
and 97.5th percentiles of a Monte Carlo simulation (see Section 4.3 of the 2024 CPS RIA and
Section 6 of the Health Benefits TSD).

Non-quantified benefits of reductions in carbon capture and sequestration (CCS)
system installation and operation. The installation and operation of CCS systems can result in
emissions of pollutants. A CO2 capture plant can impact emissions of HAP and VOC. A
reduction in installation and operation of CCS under this rule could reduce HAP and VOC
emissions from CCS systems. The EPA is unable to quantify the health benefits of this change.
The installation and operation of CCS systems can also impact emissions of criteria pollutants

6-7


-------
including SO2, PM, and NOx. In the CPS, the EPA assumed that these releases would be
controlled by installation of SCR and/or FGD. This proposed action would therefore not be
expected to alter the emissions of SO2, PM, or NOx due to the reduction in installation and
operation of CCS.

Interaction of the proposed action with NAAQS attainment: Had the CPS been
implemented, the projected emissions changes under the action would likely have affected
ambient of PM2.5 and ozone concentrations in parts of the U.S. Affected areas may have included
locations both meeting and exceeding the NAAQS for PM2.5 and ozone. States with
nonattainment areas designated as moderate or higher are required to achieve concentration
reductions in those areas sufficient to attain the NAAQS. The 2024 CPS RIA did not account for
how interaction with NAAQS compliance would affect the benefits and costs projected under the
rule. The emissions reductions projected under the CPS for most years of analysis may have
contributed to concentration reductions that aided states in achieving attainment. As these
emissions reductions will not occur under the proposed action, states may need to pursue
emissions reductions from other sources to obtain the standards, incurring costs for those
sources. Similarly, in the analysis years where emissions increased until the CPS, states may
have needed to identify additional approaches to reduce emissions from local sources relative to
the baseline to comply with the NAAQS. If this is the case, from a nationwide perspective, the
projections of avoided compliance costs and forgone emissions impacts and associated health
impacts under this proposed rule may be under- or over-estimated depending on the specifics of
how this proposed action interacts with NAAQS compliance.

6-8


-------
United States	Office of Air Quality Planning and Standards	Publication No. EPA-452/R-25-002

Environmental Protection	Health and Environmental Impacts Division	June 2025

Agency	Research Triangle Park, NC


-------