Benefit and Cost Analysis for Revisions to the Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source Category


United States
Environmental Protection
Agency
Office of Water
Washington, DC 20460
EPA-821 -R-20-003
August 28, 2020
SERA	Benefit and Cost Analysis for
Revisions to the Effluent
Limitations Guidelines and
Standards for the Steam
Electric Power Generating
Point Source Category

-------
o-EPA
United States
Environmental Protection
Agency
Benefit and Cost Analysis for Revisions to the
Effluent Limitations Guidelines and Standards
for the Steam Electric Power Generating Point
Source Category
EPA-821 -R-20-003
August 28, 2020
U.S. Environmental Protection Agency
Office of Water (4303T)
Engineering and Analysis Division
1200 Pennsylvania Avenue, NW
Washington, DC 20460

-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
Acknowledgements and Disclaimer
This report was prepared by the U.S. Environmental Protection Agency. Neither the United States
Government nor any of its employees, contractors, subcontractors, or their employees make any warranty,
expressed or implied, or assume any legal liability or responsibility for any third party's use of or the results
of such use of any information, apparatus, product, or process discussed in this report, or represents that its
use by such party would not infringe on privately owned rights.

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Table of Contents
Table of Contents
Table of Contents	i
List of Figures	v
List of Tables	vii
Abbreviations	xii
Executive Summary	1
1	Introduction	1-1
1.1	Steam Electric Power Plants	1-2
1.2	Baseline and Regulatory Options Analyzed	 1-2
1.3	Analytic Framework	 1-5
1.3.1	Constant Prices	 1-6
1.3.2	Discount Rate and Year	1-6
1.3.3	Period of Analysis	1-6
1.3.4	Timing of Technology Installation and Loading Reductions	1-7
1.3.5	Annualization of future costs and benefits	 1-7
1.3.6	Direction of Environmental Changes and Benefits	1-7
1.3.7	Population and Income Growth	 1-7
1.4	Organization of the Benefit and Cost Analysis Report	1-8
2	Benefits Overview	2-1
2.1	Human Health Impacts Associated with Changes in Surface Water Quality	2-4
2.1.1	Drinking Water	2-4
2.1.2	Fish Consumption	2-6
2.1.3	Complementary Measure of Human Health Impacts	2-8
2.2	Ecological and Recreational Impacts Associated with Changes in Surface Water Quality	2-8
2.2.1	Changes in Surface Water Quality	2-9
2.2.2	Impacts on Threatened and Endangered Species	2-10
2.2.3	Changes in Sediment Contamination	2-11
2.3	Economic Productivity	2-11
2.3.1	Marketability of Coal Ash for Beneficial Use	2-11
2.3.2	Water Supply and Use	2-12
2.3.3	Reservoir Capacity	2-13
2.3.4	Sedimentation Changes in Navigational Waterways	2-14
2.3.5	Commercial Fisheries	2-14
2.3.6	Tourism	2-14
2.3.7	Property Values	2-15

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs	Table of Contents
2.4
Changes in Air Pollution	
2-15
2.5
Changes in Water Withdrawals	
2-17
2.6
Summary of Benefits Categories	
2-17
3
Water Quality Effects of Regulatory Options	
3-1
3.1
Waters Affected by Steam Electric Power Plant Discharges	
3-1
3.2
Changes in Pollutant Loadings	
3-?

3.2.1 Implementation Timing	
3-?

3.2.2 Results	
3-5
3.3
Water Quality Downstream from Steam Electric Power Plants	
3-7
3.4
Overall Water Quality Changes	
3-8

3.4.1 WQI Data Sources	
3-8

3.4.2 WQI Calculation	
3-10

3.4.3 Baseline WQI	
3-11

3.4.4 Estimated Changes in Water Quality (AWQI) from the Regulatory Options	
3-12
3.5
Limitations and Uncertainty	
3-12
4
Human Health Benefits from Changes in Pollutant Exposure via the Drinking Water Pathway 4-1
4.1
Estimates of Changes in Halogen Concentrations in Source Water	
4-1

4.1.1 Bromide Bromide and Iodine Concentrations in Surface Water	
4-1

4.1.2 Changes in Bromide and Iodine Levels in Source Water	
4-2
4.2
Additional Measures of Human Health Effects from Exposure to Steam Electric Pollutants via


Drinking Water Pathway	
4-10
4.3
Limitations and Uncertainties	
4-1?
5
Human Health Effects from Changes in Pollutant Exposure via the Fish Ingestion Pathway .
...5-1
5.1
Population in Scope of the Analysis	
5-2
5.2
Pollutant Exposure from Fish Consumption	
5-4

5.2.1 Fish Tissue Pollutant Concentrations	
5-4

5.2.2 Average Daily Dose	
5-5
5.3
Health Effects in Children from Changes in Lead Exposure	
5-6

5.3.1 Methods	
5-7

5.3.2 Results	
5-9
5.4
Heath Effects in Children from Changes in Mercury Exposure	
5-10

5.4.1 Methods	
5-10

5.4.2 Results	
5-11
5.5
Estimated Changes in Cancer Cases from Arsenic Exposure	
5-12
5.6
Total Monetary Values of Estimated Changes in Human Health Effects	
5-12
ii

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Table of Contents
5.7	Additional Measures of Potential Changes in Human Health Effects	5-13
5.8	Limitations and Uncertainties	5-14
6	Nonmarket Benefits from Water Quality Changes	6-1
6.1	Estimated Total WTP for Water Quality Changes	6-1
6.2	Limitations and Uncertainties	6-4
7	Impacts and Benefits to Threatened and Endangered Species	7-1
7.1	Introduction	7-1
7.2	Baseline Status of Freshwater Fish Species	7-2
7.3	T&E Species Potentially Affected by the Regulatory Options	7-2
7.3.1	Identifying T&E Species Potentially Affected by the Regulatory Options	7-2
7.3.2	Estimating Effects of the Rule on T&E Species	7-4
7.4	Limitations and Uncertainties	7-5
8	Air Quality-Related Benefits	8-1
8.1	Changes in Air Emissions	8-3
8.2	Climate Change Benefits	8-6
8.2.1	Data and Methodology	8-6
8.2.2	Results	8-9
8.3	Human Health Benefits	8-9
8.3.1	Data and Methodology	8-9
8.3.2	Results	8-13
8.4	Annualized Air Quality-Related Benefits of Regulatory Options	8-19
8.5	Limitations and Uncertainties	8-23
9	Changes in Water Withdrawals	9-1
9.1	Methods	9-1
9.2	Results	9-1
9.3	Limitations and Uncertainties	9-2
10	Estimated Changes in Dredging Costs	10-1
10.1	Methods	10-1
10.1.1 Estimated Changes in Navigational Dredging Costs	10-1
10.1.2Estimated Changes in Reservoir Dredging Costs	10-2
10.2	Limitation and Uncertainty	10-3
11	Summary of Estimated Total Monetized Benefits	11-1
12	Summary of Total Social Costs	12-1
12.1 Overview of Costs Analysis Framework	12-1
iii

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Table of Contents
12.2 Key Findings for Regulatory Options	 12-2
13	Benefits and Social Costs	13-1
13.1	Comparison of Benefits and Costs by Option	13-1
13.2	Analysis of Incremental Benefits and Costs	13-1
14	Environmental Justice	14-1
14.1	Socioeconomic Characteristics of Populations Residing in Proximity to Steam Electric Power Plants 14-
1
14.2	Distribution of Human Health Impacts and Benefits	 14-4
14.2.1	Socioeconomic Characteristics of Populations Affected by Changes in Pollutant Levels in
Drinking Water Sources	 14-4
14.2.2	Socioeconomic Characteristics of Populations Affected by Changes in Exposure to Pollutants via
the Fish Ingestion Pathway	 14-8
14.2.3	Socioeconomic Characteristics of Populations Affected by Changes in Exposure to Air
Pollutants	 14-15
14.3	EJ Analysis Findings	14-16
14.4	Limitations and Uncertainties	 14-17
15	Cited References	15-1
A	Changes to Benefits Methodology since 2019 Proposed Rule Analysis	A-l
B	WQI Calculation and Regional Subindices	B-l
C Derivation of Ambient Water and Fish Tissue Concentrations in Receiving and Downstream
Reaches	C-l
D	Georeferencing Surface Water Intakes to the Medium-resolution Reach Network	D-l
E	Estimation of Exposed Population for Fish Ingestion Pathway	E-l
F	Sensitivity Analysis for IQ Point-based Human Health Effects	F-l
G	Methodology for Estimating WTP for Water Quality Changes	G-l
H Identification of Threatened and Endangered Species Potentially Affected by the Final Rule
Regulatory Options	H-l
I Uncertainty Associated with Estimating the Social Cost of Carbon	1-1
J Methodology for Modeling Air Quality Changes for the Final Rule	J-l
iv

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Figures
Figure 2-1: Summary of Benefits Resulting from the Regulatory Options	2-3
Appendix Figures
Figure D-l: PWS Intakes Review Subset
	D-2
Figure E-l: Illustration of Intersection of CBGs and Reaches	E-l
Figure 1-1: Frequency Distribution of Interim Domestic SC-CO2 Estimates for 2030 (in 2016$ per Metric Ton
C02)	1-4
Figure J-l: Air Quality Modeling Domain	J-2
Figure J-2: Map of Pennsylvania Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)	J-3
Figure J-3: Map of Pennsylvania Non-Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)
	J-4
Figure J-4: Map of Texas Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)	J-4
Figure J-5: Map of Texas Non-Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)	J-5
Figure J-6: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January-March) Nitrate
U'g/m )	J-6
Figure J-7: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September) Nitrate
(ug/m:)	J-6
Figure J-8: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January-March) Sulfate
(Lig/m:)	J-7
Figure J-9: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September) Sulfate
(Lig/m:)	J-7
Figure J-10: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January-March) Primary
PM2 5 (Lig/m )	J-8
Figure J-l 1: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September) Primary
PM2 5 (lig/m3)	J-8
Figure J-12: Map of Change in May-Sep MDA8 Ozone (ppb): 2021 Option A - Baseline	J-28
Figure J-13: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2021 Option A - Baseline	J-28
Figure J-14: Map of Change in Annual Mean PM2 5 (|a,g/m3): 2021 Option A - Baseline	J-29
Figure J-15: Map of Change in May-Sep MDA8 Ozone (ppb): 2023 Option A - Baseline	J-29
Figure J-16: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2023 Option A - Baseline	J-29
Figure J-17: Map of Change in Annual Mean PM2 5 (|a,g/m3): 2023 Option A - Baseline	J-30
Figure J-l 8: Map of Change in May-Sep MDA8 Ozone (ppb): 2025 Option A - Baseline	J-30
V

-------
BCA for Revisions to the Steam Electric Power Generating ELGs	List of Figures
Figure
J-19:
Map
of Change
in
Apr-Oct MDA1 Ozone (ppb): 2025 Option A - Baseline	
	J-30
Figure
J-20:
Map
of Change
in
Annual Mean PM2 5 (|a,g/m3): 2025 Option A - Baseline	
	J-31
Figure
J-21:
Map
of Change
in
May-Sep MDA8 Ozone (ppb): 2030 Option A - Baseline	
	J-31
Figure
J-22:
Map
of Change
in
Apr-Oct MDA1 Ozone (ppb): 2030 Option A - Baseline	
	J-31
Figure
J-23:
Map
of Change
in
Annual Mean PM2 5 (|a,g/m3): 2030 Option A - Baseline	
	J-32
Figure
J-24:
Map
of Change
in
May-Sep MDA8 Ozone (ppb): 2035 Option A - Baseline	
	J-32
Figure
J-25:
Map
of Change
in
Apr-Oct MDA1 Ozone (ppb): 2035 Option A - Baseline	
	J-32
Figure
J-26:
Map
of Change
in
Annual Mean PM2 5 (|a,g/m3): 2035 Option A - Baseline	
	J-3 3
Figure
J-27:
Map
of Change
in
May-Sep MDA8 Ozone (ppb): 2040 Option A - Baseline	
	J-3 3
Figure
J-28:
Map
of Change
in
Apr-Oct MDA1 Ozone (ppb): 2040 Option A - Baseline	
	J-3 3
Figure
J-29:
Map
of Change
in
Annual Mean PM2 5 (|a,g/m3): 2040 Option A - Baseline	
	J-3 4
Figure
J-30:
Map
of Change
in
May-Sep MDA8 Ozone (ppb): 2045 Option A - Baseline	
	J-3 4
Figure
J-31:
Map
of Change
in
Apr-Oct MDA1 Ozone (ppb): 2045 Option A - Baseline	
	J-3 4
Figure
J-32:
Map
of Change
in
Annual Mean PM2 5 (|a,g/m3): 2045 Option A - Baseline	
	J-3 5

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Tables
List of Tables
Table 1-1: Regulatory Options	 1-4
Table 2-1: Estimated Annual Pollutant Loadings and Changes in Loadings for Baseline and Regulatory
Options Under Technology Implementation	2-1
Table 2-2: Drinking Water Maximum Contaminant Levels and Goals for Selected Pollutants in Steam Electric
FGD Wastewater or Bottom Ash Transport Water Discharges	2-4
Table 2-3: Estimated Welfare Effects of Changes in Pollutant Discharges from Steam Electric Power Plants 2-
17
Table 3-1: Strahler Stream Order Designation for Reaches Receiving Steam Electric Power Plant Discharges
	3-2
Table 3-2: Implementation Schedule by Wastestream and Regulatory Option	3-4
Table 3-3: Annual Average Changes in Total Pollutant Loading in Period 1 (2021-2028) and Period 2 (2029-
2047) for Selected Pollutants in Steam Electric Power Plant Discharges, Compared to Baseline (lb/year)
	3-6
Table 3-4: Estimated Exceedances of National Recommended Water Quality Criteria under the Baseline and
Regulatory Options	3-9
Table 3-5: Water Quality Data used in Calculating WQI for the Baseline and Regulatory Options	3-10
Table 3-6: Estimated Percentage of Potentially Affected Reach Miles by WQI Classification: Baseline
Scenario	3-11
Table 3-7: Ranges of Estimated Water Quality Changes for Regulatory Options, Compared to Baseline ... 3-12
Table 3-8: Limitations and Uncertainties in Estimating Water Quality Effects of Regulatory Options	3-13
Table 4-1: Estimated Reaches, Surface Water Intakes, Public Water Systems, and Populations Potentially
Affected by Steam Electric Power Plant Discharges	4-4
Table 4-2: Estimated Distribution of Changes in Source Water and PWS-Level Bromide Concentrations by
Period and Regulatory Option, Compared to Baseline	4-6
Table 4-3: Estimated Distribution of Changes in Source Water and PWS-Level Iodine Concentrations by
Period and Regulatory Option, Compared to Baseline	4-8
Table 4-4: Estimated Distribution of Changes in Source Water and PWS-Level Arsenic, Lead, and Thallium
Concentrations by Period and Regulatory Option, Compared to Baseline	4-11
Table 4-5: Limitations and Uncertainties in the Analysis of Human Health Benefits from Changes in
Discharges of Halogens and Other Pollutants Via the Drinking Water Pathway	4-12
Table 5-1: Summary of Population Potentially Exposed to Contaminated Fish Living within 50 Miles of
Affected Reaches (as of 2017)	5-4
Table 5-2: Summary of Group-specific Consumption Rates for Fish Tissue Consumption Risk Analysis .... 5-5
Table 5-3: Value of an IQ Point (2018$) based on Expected Reductions in Lifetime Earnings	5-9
Table 5-4: Estimated Monetary Value of Changes in IQ Points for Children Exposed to Lead under the
Regulatory Options, Compared to Baseline	5-9
vii

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Tables
Table 5-5: Estimated Monetary Values of Changes in IQ Points for Infants from Mercury Exposure under the
Regulatory Options, Compared to Baseline	5-12
Table 5-6: Total Monetary Values of Changes in Human Health Outcomes Associated with Fish Consumption
under the Regulatory Options, Compared to Baseline (Millions of2018$)	5-13
Table 5-7: Estimated Number of Reaches Exceeding Human Health Criteria for Steam Electric Pollutants 5-14
Table 5-8: Limitations and Uncertainties in the Analysis of Human Health Effects via the Fish Ingestion
Pathway	5-15
Table 6-1: Estimated Household Willingness-to-Pay for Water Quality Changes under the Regulatory
Options, Compared to Baseline	6-3
Table 6-2: Estimated Total Annualized Willingness-to-Pay for Water Quality Changes under the Regulatory
Options, Compared to Baseline (Millions of 2018$)	6-4
Table 6-3: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits	6-5
Table 7-1: Number of T&E Species with Habitat Range Intersecting Reaches Immediately Receiving or
Downstream of Steam Electric Power Plant Discharges, by Group	7-3
Table 7-2: Higher Vulnerability T&E Species with Habitat Intersecting Waters with Estimated Changes in
NRWQC Exceedance Status under the Regulatory Options, Compared to Baseline	7-4
Table 7-3: Higher Vulnerability T&E Species Whose Habitat May be Affected by the Regulatory Options
Compared to Baseline in Period 1	7-5
Table 7-4: Higher Vulnerability T&E Species Whose Habitat May be Affected by the Regulatory Options
Compared to Baseline in Period 2	7-5
Table 7-5: Limitations and Uncertainties in the Analysis of T&E Species Impacts and Benefits	7-6
Table 8-1: IPM Run Years	8-3
Table 8-2: Estimated Changes in Air Pollutant Emissions Due to Increase in Power Requirements and
Trucking at Steam Electric Power Plants 2021-2047, Compared to Baseline	8-4
Table 8-3: Estimated Changes in Annual CO2, NOx, SO2, and Primary PM2 5 Emissions Due to Changes in
Electricity Generation Profile, Compared to Baseline	8-5
Table 8-4: Estimated Changes in Annual Primary PM10, Hg and HC1 Emissions Due to Changes in Electricity
Generation Profile, Compared to Baseline	8-5
Table 8-5: Estimated Net Changes in Air Pollutant Emissions Due to Changes in Power Requirements,
Trucking, and Electricity Generation Profile, Compared to Baseline	8-5
Table 8-6: Interim Domestic Social Cost of Carbon Values (2018$/Metric Tonne CO2)	8-7
Table 8-7: Estimated Domestic Climate Benefits from Changes in CO2 Emissions for Selected Years under
the Final Rule, Compared to Baseline (Millions of 2018$)	8-9
Table 8-8: Estimated Total Annualized Domestic Climate Benefits from Changes in CO2 Emissions under the
Final Rule, Compared to Baseline (Millions of 2018$)	8-9
Table 8-9: Human Health Effects of Ambient PM2 5 and Ozone	8-12
Table 8-10: Estimated Avoided PM2 5 and Ozone-Related Premature Deaths and Illnesses by Year for the
Final Rule, Compared to Baseline (95% Confidence Interval)	8-15

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Tables
Table 8-11: Estimated Avoided PM2 5 and Ozone-Related Premature Deaths and Illnesses for the Final Rule,
Compared to Baseline, Using Alternative Approaches to Quantifying Avoided PM2 5-Attributable Deaths
(95% Confidence Interval)	8-16
Table 8-12: Estimated Economic Value of Avoided PM2.5 and Ozone-Attributable Deaths and Illnesses for
the Final Rule, Compared to Baseline, Using Alternative Approaches to Represent PM2 5 Mortality Risk
Effects (95% Confidence Interval; Million of 2018$)	8-17
Table 8-13: Estimated Percent of Avoided PM2 5-related Premature Deaths Above and Below PM2 5
Concentration Cut Points for the Final Rule, Compared to Baseline	8-19
Table 8-14: Total Annualized Air Quality-Related Benefits of Regulatory Options, Compared to the Baseline,
2021-2047 (Millions of 2018$)	8-21
Table 8-15: Total Annualized Air Quality-Related Benefits of Regulatory Options, Compared to the Baseline,
2021-2047, Showing Only PM2 5 Related Premature Mortality Risk Benefits above the Lowest Measured
Level of Each Long-Term PM25 Mortality Study (Millions of 2018$)	8-21
Table 8-16: Total Annualized Air Quality-Related Benefits of Regulatory Options Compared to the Baseline,
2021-2047, showing only PM2 5 Related Premature Mortality Risk Benefits above PM2 5 National
Ambient Air Quality Standard (Millions of 2018$)	8-22
Table 8-17: Limitations and Uncertainties in Analysis of Air Quality-Related Benefits	8-23
Table 9-1: Industry-level Total Changes in Water Withdrawals under the Regulatory Options, Compared to
Baseline (Both Surface Water and Aquifers)	9-1
Table 9-2: Estimated Annualized Benefits from Changes in Groundwater Withdrawals under the Regulatory
Options, Compared to Baseline (Millions of 2018$)	9-2
Table 9-3: Limitations and Uncertainties in Analysis of Changes in Groundwater Withdrawals	9-2
Table 10-1: Estimated Annualized Navigational Dredging Costs at Affected Reaches Based on Historical
Averages (Millions of 2018$)	10-2
Table 10-2: Estimated Annualized Changes in Navigational Dredging Costs under the Regulatory Options,
Compared to Baseline	10-2
Table 10-3: Estimated Annualized Reservoir Dredging Volume and Costs based on Historical Averages.. 10-3
Table 10-4: Estimated Total Annualized Changes in Reservoir Dredging Volume and Costs under the
Regulatory Options, Compared to Baseline	10-3
Table 10-5: Limitations and Uncertainties in Analysis of Changes in Dredging Costs	10-4
Table 11-1: Summary of Estimated Total Annualized Benefits of the Regulatory Options, Compared to
Baseline, at 3 Percent (Millions of 2018$)	11-2
Table 11-2: Summary of Estimated Total Annualized Benefits of the Regulatory Options, Compared to
Baseline, at 7 Percent (Millions of 2018$)	11-3
Table 12-1: Summary of Estimated Annualized Costs (Millions of 2018$)	12-3
Table 12-2: Time Profile of Costs to Society (Millions of 2018$)	12-3
Table 13-1: Total Estimated Annualized Benefits and Costs by Regulatory Option and Discount Rate,
Compared to Baseline (Millions of 2018$)	13-1

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Tables
Table 13-2: Analysis of Estimated Incremental Net Benefit of the Regulatory Options, Compared to Baseline
and to Other Regulatory Options (Millions of 2018$)	 13-2
Table 14-1: Socioeconomic Characteristics of Communities Living in Proximity to Steam Electric Power
Plants and Associated Immediate Receiving Reach, Compared to National Average	14-3
Table 14-2: Socioeconomic Characteristics of Communities Living in Proximity to Steam Electric Power
Plants and Associated Immediate Receiving Reach, Compared to National and State Averages	14-4
Table 14-3: Socioeconomic Characteristics of Counties in Service Areas of Potentially Affected PWS,
Compared to State Average	14-5
Table 14-4: Socioeconomic Characteristics of Affected Tribal Areas, Compared to State Average	14-7
Table 14-5: Socioeconomic Characteristics of Communities Living in Proximity to Reaches with Changes to
Selected Pollutant Concentrations under the Regulatory Options, Compared to Baseline (Period 1)... 14-9
Table 14-6: Socioeconomic Characteristics of Communities Living in Proximity to Reaches with Changes to
Selected Pollutant Concentrations under the Regulatory Options, Compared to Baseline (Period 2). 14-10
Table 14-7: Characteristics of Children Potentially Exposed to Steam Electric Power Plant Pollutants via
Consumption of Self-caught Fish	 14-11
Table 14-8: Estimated Distribution of IQ Point Changes from Lead and Mercury Exposure Via Self-caught
Fish Consumption Under the Regulatory Options, Compared to Baseline (2021 to 2047)	 14-13
Table 14-9: Estimated Distribution of Changes in IQ Point Changes from Lead and Mercury Exposure Via
Self-caught Fish Consumption under the Regulatory Options, Compared to Baseline, by Fishing Mode
(2021 to 2047)	 14-14
Table 14-10: Socioeconomic Characteristics of Populations Projected to see Net Increases and Decreases in
Adverse Health Outcomes from Changes in Exposure to PM2 5 and Ground-level Ozone Under the Final
Rule, Compared to Baseline, in 2021-2047	 14-16
Table 14-11: Limitations and Uncertainties in EJ Analysis	14-17
Appendix Tables
Figure D-l: PWS Intakes Review Subset	A-l
Table B-l: Freshwater Water Quality Subindices	B-2
Table B-2: Freshwater Water Quality Subindex for Toxics	B-3
Table B-3: Suspended Solids Subindex Curve Parameters, by Ecoregion	B-4
Table B-4: TN Subindex Curve Parameters, by Ecoregion	B-5
Table B-5: TP Subindex Curve Parameters, by Ecoregion	B-7
Table C-l: Background Fish Tissue Concentrations, based on 10th percentile	C-3
Table C-2: Imputed and Validated Fish Tissue Concentrations by Regulatory Option	C-3
Table D-l: Summary of Intakes Potentially Affected by Steam Electric Power Plant	D-l
Table F-l: Value of an IQ Point (2018$) based on Expected Reductions in Lifetime Earnings	F-l
Table F-2: Estimated Monetary Value of Changes in IQ Losses for Children Exposed to Lead	F-l
X

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
List of Tables
Table F-3: Estimated Monetary Values from Changes in IQ Losses for Infants from Mercury Exposure	F-2
Table G-l: Independent Variable Assignments for Surface Water Quality Meta-Analysis	G-6
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric Power
Plant Outfalls	H-2
Table J-l: Ozone scaling factors for coal EGU tags in the Baseline scenario	J-12
Table J-2: Ozone scaling factors fornon-coal EGU tags in the Baseline scenario	J-13
Table J-3: Ozone scaling factors for coal EGU tags in the Option A scenario	J-14
Table J-4: Ozone scaling factors for non-coal EGU tags in the Option A scenario	J-15
Table J-5: Nitrate scaling factors for coal EGU tags in the Baseline scenario	J-16
Table J-6: Nitrate scaling factors for non-coal EGU tags in the Baseline scenario	J-17
Table J-7: Nitrate scaling factors for coal EGU tags in the Option A scenario	J-18
Table J-8: Nitrate scaling factors fornon-coal EGU tags in the Option A scenario	J-19
Table J-9: Sulfate scaling factors for coal EGU tags in the Baseline scenario	J-20
Table J-10: Sulfate scaling factors for non-coal EGU tags in the Baseline scenario	J-21
Table J-l 1: Sulfate scaling factors for coal EGU tags in the Option A scenario	J-22
Table J-12: Sulfate scaling factors for non-coal EGU tags in the Option A scenario	J-23
Table J-13: Primary PM2.5 scaling factors for coal EGU tags in the Baseline scenario	J-24
Table J-14: Primary PM2.5 scaling factors for non-coal EGU tags in the Baseline scenario	J-25
Table J-15: Primary PM2 5 scaling factors for coal EGU tags in the Option A scenario	J-26
Table J-16: Primary PM2 5 scaling factors for non-coal EGU tags in the Option A scenario	J-27

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Abbreviations
Abbreviations
ACE	Affordable Clean Energy
ACS	American Community Survey
ADD	Average daily dose
As	Arsenic
ATSDR	Agency for Toxic Substances and Disease Registry
BA	Bottom ash
BAT	Best available technology economically achievable
BCA	Benefit-cost analysis
BEA	Bureau of Economic Analysis
BenMAP-CE Environmental Benefits Mapping and Analysis Program—Community Edition
BLS	Bureau of Labor Statistics
BMP	Best management practices
BOD	Biochemical oxygen demand
BW	Body weight
CAMx	Comprehensive Air Quality Model with Extensions
CBG	Census Block Group
CCI	Construction Cost Index
CCME	Canadian Council of Ministers of the Environment
CCR	Coal combustion residuals
CDC	Center for Disease Control
CFR	Code of Federal Regulations
CO2	Carbon dioxide
COD	Chemical oxygen demand
COI	Cost-of-illness
COPD	Chronic obstructive pulmonary disease
CPI	Consumer Price Index
CWA	Clean Water Act
D-FATE	Downstream Fate and Transport Equations
DBP	Disinfection byproduct
DBPR	Disinfectants and Disinfection Byproduct Rule
DCN	Document Control Number
DICE	Dynamic Integrated Climate and Economy
DO	Dissolved oxygen
E2RF1	Enhanced River File 1
EA	Environmental Assessment
EC	Elemental carbon
ECI	Employment Cost Index
ECOS	Environmental Conservation Online System
EGU	Electricity generating unit
EJ	Environmental justice
ELGs	Effluent limitations guidelines and standards
EO	Executive Order
EPA	United States Environmental Protection Agency
EROM	Enhanced Runoff Method
xii

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Abbreviations
ESA
Endangered Species Act
FC
Fecal coliform
FCA
Fish consumption advisories
FGD
Flue gas desulfurization
FUND
Climate Framework for Uncertainty, Negotiation, and Distribution
FR
Federal Register
GDP
Gross Domestic Product
GHG
Greenhouse gas
GIS
Geographic Information System
HAP
Hazardous air pollutant
HC1
Hydrogen chloride
Hg
Mercury
HRTR
High Residence Time Reduction
HUC
Hydrologic unit code
IAM
Integrated assessment model
IBI
Index of biotic integrity
IEUBK
Integrated Exposure, Uptake, and Biokinetics
IPCC
Intergovernmental Panel on Climate Change
IPM
Integrated Planning Model
ISA
Integrated science assessment
IRIS
Integrated Risk Information System
IQ
Intelligence quotient
LADD
Lifetime average daily dose
LML
Lowest measured level
LRTR
Low Residence Time Reduction
MATS
Mercury and Air Toxics Standards
MCL
Maximum contaminant level
MCLG
Maximum contaminant level goal
MDA1
Maximum daily 1-hour average
MDA8
Maximum daily 8-hour average
MGD
Million gallons per day
MRM
Meta-regression model
MWTP
Marginal willingness-to-pay
NAAQS
National Ambient Air Quality Standards
NEI
National Emissions Inventory
NERC
North American Electric Reliability Corporation
NHD
National Hydrography Dataset
NLCD
National Land Cover Dataset
NLFA
National Listing Fish Advisory
NO A A
National Oceanic and Atmospheric Administration
NOx
Nitrogen oxides
NPDES
National Pollutant Discharge Elimination System
NRWQC
National Recommended Water Quality Criteria
NWIS
National Water Information System
03
Ozone
03V
Ozone formed in VOC-limited chemical regimes
xiii

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Abbreviations
03N	Ozone formed in NOx-limited chemical regimes
OA	Organic aerosol
O&M	Operation and maintenance
OMB	Office of Management and Budget
OSAT/APCA Ozone Source Apportionment Technique/Anthropogenic Precursor Culpability Assessment
PACE	Policy Analysis of the Greenhouse Gas Effect
Pb	Lead
PbB	Blood lead concentration
PM2 5	Particulate matter (fine inhalable particles with diameters 2.5 |a,m and smaller)
PM10	Particulate matter (inhalable particles with diameters 10 |a,m and smaller)
ppm	parts per million
PSAT	Particulate Source Apportionment Technique
PSES	Pretreatment Standards for Existing Sources
PV	Present value
PWS	Public water system
QA	Quality assurance
QC	Quality control
RIA	Regulatory Impact Analysis
SAB-HES	Science Advisory Board Health Effect Subcommittee
SBREFA	Small Business Regulatory Enforcement Fairness Act
SC-CO2	Social cost of carbon
SDWIS	Safe Drinking Water Information System
Se	Selenium
SO2	Sulfur dioxide
SPARROW SPAtially Referenced Regressions On Watershed attributes
SSC	Suspended solids concentration
SWFSC	Southwest Fisheries Science Center
T&E	Threatened and endangered
TDD	Technical Development Document
TDS	Total dissolved solids
TEC	Threshold effect concentration
TN	Total nitrogen
TP	Total phosphorus
TRI	Toxics Release Inventory
TSD	Technical support document
TSS	Total suspended solids
TTHM	Total trihalomethanes
TWTP	Total willingness-to-pay
U.S. FWS	United States Fish and Wildlife Service
USGS	United States Geological Survey
VIP	Voluntary Incentive Program
VOC	Volatile organic compounds
VSL	Value of a statistical life
WBD	Watershed Boundary Dataset
WQ	Water quality
WQI	Water quality index
xiv

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
WQI-BL	Baseline water quality index
WQI-PC	Post-technology implementation water quality index
WQL	Water quality ladder
WTP	Willingness-to-pay
Abbreviations
xv

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Executive Summary
The U.S. Environmental Protection Agency (EPA) is finalizing revisions to the technology-based effluent
limitations guidelines and standards (ELGs) for the steam electric power generating point source category,
40 Code of Federal Regulations (CFR) part 423, which EPA proposed in November 2019 (84 FR 64620). The
final rule revises certain best available technology economically achievable (BAT) effluent limitations and
pretreatment standards for existing sources (PSES) for two wastestreams: flue gas desulfurization (FGD)
wastewater and bottom ash (BA) transport water.
Regulatory Options
EPA presents four main regulatory options, summarized in Table ES-1. The four main regulatory options
analyzed at proposal (1, 2, 3, and 4), the details of which were discussed in the proposed rule (84 FR 64620),
correspond generally to regulatory options D, A, B, and C here, but do contain differences as detailed in the
preamble. The availability and achievability of technologies with better pollutant removals, as well as the
general lack of public comments in support for proposed regulatory Option 1, led EPA to focus updates to the
Agency's analysis on the remaining three regulatory options. EPA did not update the analyses for regulatory
Option D, but rather retained the results of the proposed rule analyses for this option (see the 2019 Benefits
and Costs Analysis [BCA; U.S. EPA, 2019a]).
The baseline for the benefit and social cost analyses reflects ELG requirements in absence of this final EPA
action.1 As detailed in this report, EPA calculated the difference between the baseline and regulatory options
A, B, and C to determine the net incremental effect (as positive or negative change) of the regulatory options.
EPA is finalizing Option A. For a description of Option A and other regulatory options EPA analyzed, see
Table ES-1.
In general, the estimated incremental effects of the final rule, Option A, are small compared to baseline (see
U.S. EPA, 2015a).
This includes the 2015 rule as well as the September 2017 postponement rule which delayed the earliest technology
implementation date for the ELGs applicable to FGD wastewater and bottom ash transport water.
ES-1

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Table ES-1: Regulatory Options
Wastestream
Subcategory
Technology Basis for BAT/PSES Regulatory Options3
2015 Rule
(Baseline)
Option D
Option A
(Final Rule)
Option B
Option C
FGD
Wastewater
NA (default unless in
subcategory)15
Chemical Precipitation
+ HRTR Biological
Treatment
Chemical Precipitation
Chemical Precipitation
+ LRTR Biological
Treatment
Chemical Precipitation
+ LRTR Biological
Treatment
Membrane Filtration
High FGD Flow Facilities
NS
NS
Chemical Precipitation
NS
NS
Low Utilization Boilers
NS
NS
Chemical Precipitation
NS
NS
Boilers permanently
ceasing the combustion of
coal by 2028
NS
NS
Surface Impoundment
NS
NS
FGD Wastewater Voluntary Incentives
Program (Direct Dischargers Only)
Evaporation
Membrane Filtration
Membrane Filtration
Membrane Filtration
NA
Bottom Ash
Transport
Water
NA (default unless in
subcategory)15
Dry Handling / Closed
Loop
High Recycle Rate
Systems
High Recycle Rate
Systems
High Recycle Rate
Systems
High Recycle Rate
Systems
Low Utilization Boilers
NS
NS
Surface Impoundment
+ BMP Plan
NS
NS
Boilers permanently
ceasing the combustion of
coal by 2028
NS
NS
Surface Impoundment
NS
NS
Abbreviations: BMP = Best Management Practice; HRTR = High Residence Time Reduction; LRTR = Low Residence Time Reduction; NS = Not subcategorized (default technology basis
applies); NA = Not applicable
a.	See Supplemental TDD for a description of these technologies (U.S. EPA, 2020g).
b.	The table does not present existing subcategories included in the 2015 rule as EPA did not reopen the existing subcategorization of oil-fired units or units with a nameplate capacity
of 50 MW or less.
Source: U.S. EPA Analysis, 2020
ES-2

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Benefits of Regulatory Options
EPA estimated the potential social welfare effects of the regulatory options and, where possible, quantified
and monetized the benefits (see Chapters 3 through 11 for details of the methodology and results). Table ES-2
and Table ES-3 summarize the benefits that EPA quantified and monetized using 3 percent and 7 percent
discounts, respectively. In the tables, positive values indicate improvements in social welfare, relative to the
baseline, whereas negative values reflect forgone benefits of the regulatory options, i.e., social welfare losses.
In general, the estimated effects of implementing the regulatory options are comparable to those estimated at
proposal (see U.S. EPA, 2019a), and are small compared to those estimated in 2015 (see U.S. EPA, 2015a).
EPA quantified but did not monetize other welfare effects of the regulatory options and discusses other effects
only qualitatively. Chapter 2 presents additional information on these welfare effects.
3

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Table ES-2: Summary of Total Annualized Benefits for Regulatory Options, Compared to Baseline, at 3 Percent (Millions of 2018$)
Benefit Category
Option Da,b
Option Ab
(Final Rule)
Option Bb
Option Cb
Human Health

-$0.3
-$0.3
-$0.1
Changes in IQ losses in children from exposure to leadd

<$0.0
<$0.0
<$0.1
Changes in IQ losses in children from exposure to
mercury
-$0.3
-$0.3
-$0.3
-$0.1
Ecological Conditions and Recreational Uses Changes

-$15.3 to -$7.4
-$10.4 to -$5.5
-$9.9 to -$4.8
Use and nonuse values for water quality changes6

-$15.3 to -$7.4
-$10.4 to -$5.5
-$9.9 to -$4.8
Market and Productivity Effectsd

<$0.0
<$0.0
$0.0
Changes in dredging costsd

<$0.0
<$0.0
<$0.0
Reduced water withdrawal

<$0.0
<$0.0
<$0.0
Air Quality-Related Effects

$14 to $51
$11 to $41
-$8.5 to -$2.4
Climate change effects from changes in C02 emissions'

-$14
t—1
t—1
1
$2.3
Human health effects from changes in NOx, S02, and
PM2.5 emissions8
Not estimated
$28 to $65
$23 to $52
-$11 to -$4.7
Total8,11

-$1.7 to $43.3
$0.3 to $35.7
-$12.4 to -$13.4
a. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
b. Negative values represent forgone benefits and positive values represent realized benefits.
c. Total includes $0.4 million of benefits due to changes in bladder cancer risk from disinfection byproducts in drinking water as estimated for the 2019 proposed rule (U.S. EPA,
2019a).
d. "<$0.0" indicates that monetary values are greater than -$0.1 million but less than $0.0 million.
e. The range reflects the lower and upper bound willingness-to-pay estimates. See Chapter 6 for details.
f. Values for air-quality related effects are rounded to two significant figures. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air quality-
related benefits for Options B and C from the estimate for Option A that is based on IPM outputs. See Chapter 8 for details.
g. Values for air-quality related effects are rounded to two significant figures. The range reflects the lower and upper bound estimates of human health effects from changes in PM2.5
and ozone levels. See Chapter 8 for details.
h. Values for individual benefit categories may not sum to the total due to independent rounding. Range is based on the low and high willingness to pay estimates and air quality-
related effects.
i. Value reflects midpoint willingness-to-pay estimate. See 2019 BCA for details (U.S. EPA, 2019a).
Source: U.S. EPA Analysis, 2020
4

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Table ES-3: Summary of Total Annualized Benefits for Regulatory Options, Compared to Baseline, at 7 Percent (Millions of 2018$)
Benefit Category
Option D°'b
Option Ab
(Final Rule)
Option Bb
Option Cb
Human Health

-$0.1
-$0.1
-$0.1
Changes in IQ losses in children from exposure to leadd

<$0.0
<$0.0
<$0.0
Changes in IQ losses in children from exposure to
mercury
-$0.1
-$0.1
-$0.1
-$0.1
Ecological Conditions and Recreational Uses Changes

-$16.4 to -$8.0
-$12.0 to -$5.8
-$13.9 to -$6.7
Use and nonuse values for water quality changes6

-$16.4 to -$8.0
-$12.0 to -$5.8
-$13.9 to -$6.7
Market and Productivity Effectsd

<$0.0
<$0.0
<$0.0
Changes in dredging costsd

<$0.0
<$0.0
<$0.0
Reduced water withdrawal

<$0.0
<$0.0
<$0.0
Air Quality-Related Effects

$23 to $54
$19 to $44
$2.7 to $6.4
Climate change effects from changes in C02 emissions'

-$2.3
-$1.9
-$0.27
Human health effects from changes in NOx, S02, and
PM2.5 emissions8
Not estimated
$25 to $56
$21 to $46
$3.0 to $6.6
Total&h

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Executive Summary
Social Costs of Regulatory Options
Table ES-4 presents the incremental social costs attributable to the regulatory options, calculated as the
difference between each option and the baseline. The regulatory options generally result in cost savings across
regulatory options and discount rates, except for Option C which results in additional costs at the
three percent discount rate. Chapter 12 describes the social cost analysis. The compliance costs of the
regulatory options are detailed in the Regulatory Impact Analysis (RIA) (U.S. EPA, 2020d).
Comparison of Benefits and Social Costs of Regulatory Options
In accordance with the requirements of Executive Order 12866: Regulator}! Planning and Review and
Executive Order 13563: Improving Regulation and Regulatory Review, EPA compared the benefits and costs
of each regulatory option. Table ES-5 presents the incremental monetized benefits and incremental social
costs attributable to the regulatory options, calculated as the difference between each option and the baseline.
Table ES-4: Total Annualized Benefits and Social Costs by Regulatory
Option and Discount Rate (Millions of 2018$)
Regulatory Option
Total Monetized Benefits3
Total Social Costs
3% Discount Rate

Option A (Final Rule)
-$1.7 to $43.3
-$127.1
Option B
$0.3 to $35.7
-$103.2
Option C
-$12.4 to -$13.4
$21.4
7% Discount Rate

Option A (Final Rule)
$6.5 to $45.9
-$153.4
Option B
$6.9 to $38.1
-$126.4
Option C
-$11.3 to -$0.4
-$18.2
a. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air
quality-related benefits for Options B and C from the estimate for Option A that is based on IPM
outputs. The range of benefits reflects the lower and upper bound estimates of human health
effects from changes in PM2.5 and ozone levels. See Chapter 8 for details.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final
rule. All results shown for Option D are based on the 2019 analysis, as detailed in the 2019 BCA
(U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and
other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020.
6

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
1 Introduction
EPA is finalizing a regulation that revises the technology-based ELGs for the steam electric power generating
point source category, 40 CFR part 423, which EPA proposed in November 2019 (84 FR 64620). The final
rule revises certain effluent limitations based on BAT and pretreatment standards for existing sources for two
wastestreams: FGD wastewater and bottom ash (BA) transport water.
This document presents an analysis of the benefits and social costs of the regulatory options, including the
final rule option (Option A), and complements other analyses EPA conducted in support of the final rule,
described in separate documents:
•	Supplemental Environmental Assessment for Revisions to the Effluent Guidelines and Standards for
the Steam Electric Power Generating Point Source Category (Supplemental EA; U.S. EPA, 2020f).
The Supplemental EA summarizes the potential environmental and human health impacts that are
estimated to result from implementation of the final rule.
•	Supplemental Technical Development Document for Revisions to the Effluent Guidelines and
Standards for the Steam Electric Power Generating Point Source Category (Supplemental TDD; U.S.
EPA, 2020g). The Supplemental TDD summarizes the technical and engineering analyses supporting
the final rule. The Supplemental TDD presents EPA's updated analyses supporting the revisions to
limitations and standards applicable to discharges of FGD wastewater and bottom ash transport water.
These updates include additional data collection that has occurred since the signature of the 2015 rule,
updates to the industry (e.g., retirements, updates to FGD treatment and bottom ash handling), cost
methodologies, pollutant removal estimates, corresponding non-water quality environmental impacts
associated with updated FGD and bottom ash methodologies, and explanations for the calculation of
the effluent limitations and standards.
•	Regulatory Impact Analysis for Revisions to the Effluent Limitations Guidelines and Standards for the
Steam Electric Power Generating Point Source Category (RIA; U.S. EPA, 2020d). The RIA describes
EPA's analysis of the costs and economic impacts of the regulatory options. This analysis provides
the basis for social cost estimates presented in Chapter 12 of this document. The RIA also provides
information pertinent to meeting several legislative and administrative requirements, including the
Regulatory Flexibility Act of 1980 (as amended by the Small Business Regulatory Enforcement
Fairness Act [SBREFA] of 1996), the Unfunded Mandates Reform Act of 1995, Executive Order
13211 on Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution, or
Use, and others.
The rest of this chapter discusses aspects of the regulatory options that are salient to EPA's analysis of the
benefits and social costs of the final rule and summarizes key analytic inputs used throughout this document.
The analyses of the regulatory options are based on data generated or obtained in accordance with EPA's
Quality Policy and Information Quality Guidelines. EPA's quality assurance (QA) and quality control (QC)
activities for this rulemaking include the development, approval and implementation of Quality Assurance
Project Plans for the use of environmental data generated or collected from all sampling and analyses, existing
databases and literature searches, and for the development of any models which used environmental data.
Unless otherwise stated within this document, the data used and associated data analyses were evaluated as
1-1

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
described in these quality assurance documents to ensure they are of known and documented quality, meet
EPA's requirements for objectivity, integrity and utility, and are appropriate for the intended use.
1.1	Steam Electric Power Plants
The ELGs for the Steam Electric Power Generating Point Source Category apply to a subset of the electric
power industry, namely those plants "with discharges resulting from the operation of a generating unit by an
establishment whose generation of electricity is the predominant source of revenue or principal reason for
operation, and whose generation of electricity results primarily from a process utilizing fossil-type fuel (coal,
oil, or gas), fuel derived from fossil fuel (e.g., petroleum coke, synthesis gas), or nuclear fuel in conjunction
with a thermal cycle employing the steam water system as the thermodynamic medium" (40 Code of Federal
Regulations [CFR] 423.10).
As described in the RIA, of the 914 steam electric power plants in the universe identified by EPA, only those
coal fired power plants that discharge bottom ash transport water or FGD wastewater may incur compliance
costs under the final rule. See Supplemental TDD and RIA for details (U.S. EPA, 2020d; 2020g). In total, EPA
estimated that 112 steam electric power plants generate the wastestreams subject to the final rule.
1.2	Baseline and Regulatory Options Analyzed
EPA presents four regulatory options (see Table 1-1). These options differ in the stringency of controls and
applicability of these controls to generating units or plants based on generation capacity utilization, retirement
or repowering status, and scrubber purge flow (see Supplemental TDD for a detailed discussion of the options
and the associated treatment technology bases). Additionally, under Options A and B, steam electric power
plants may elect to participate in the Voluntary Incentive Program (VIP) which requires them to meet more
stringent limits for FGD wastewater in exchange for additional time to comply with those limits.
The baseline for this analysis reflects applicable requirements (in absence of the final rule). The baseline
includes the 2015 rule (80 FR 67838) as well as the September, 2017 postponement rule (82 FR 43494) which
postpones the earliest compliance date for the new more stringent BAT effluent limitations and PSES for
FGD wastewater and bottom ash transport water in the 2015 rule. As discussed further in Section 2.2.2 of the
RIA, the baseline for this analysis also includes the effects of the 2020 CCR Part A rule.2
The Agency estimated and presents in this report the water quality and other environmental effects of FDG
wastewater and bottom ash transport water discharges under both the 2015 rule baseline and regulatory
options A through D presented in Table 1-1.3 The Agency calculated the difference between the baseline and
2	In the 2015 CCR rule RIA (U.S. EPA, 2014), EPA explicitly accounts for the baseline closure of all surface impoundments
(including composite lined surface impoundments) at the end of their useful life (40 years). At the end of a surface
impoundment's useful life, facilities are projected to face a decision between multiple replacement disposal alternatives. EPA
modeled these alternatives and selected the least-cost alternative for each facility (see section 3.2.4.2 of the 2015 CCR RIA).
Based on EPA's cost estimates, the Agency found that the least-cost alternative universally involved some form of converting
away from disposal surface impoundments and incurring the costs of making a "wet-dry conversion."
In light of the changes from the USWAG and Waterkeeper mandates, the 2020 CCR Part A RIA revises cost estimates to reflect
the new timing and number of surface impoundment closures and wet to dry conversions (U.S. EPA, 2020e). All unlined surface
impoundments are now required by these court decisions to close. EPA estimated the increase in annualized costs as
$40.5 million in the adjusted baseline costs in Section 2.5 of the CCR Part A RIA.
3	As noted above, option D is presented in this report, but the option D analysis has not been updated since proposal.
1-2

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs	1: Introduction
the regulatory options to determine the net effect of the regulatory options. The changes attributable to the
regulatory options are the difference between each option and the baseline.
1-3

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
Table 1-1: Regulatory Options
Wastestream
Subcategory
Technology Basis for BAT/PSES Regulatory Options3
2015 Rule
(Baseline)
Option D
Option A
(Final Rule)
Option B
Option C
FGD
Wastewater
NA (default unless in
subcategory)15
Chemical Precipitation
+ HRTR Biological
Treatment
Chemical Precipitation
Chemical Precipitation
+ LRTR Biological
Treatment
Chemical Precipitation
+ LRTR Biological
Treatment
Membrane Filtration
High FGD Flow Facilities
NS
NS
Chemical Precipitation
NS
NS
Low Utilization Boilers
NS
NS
Chemical Precipitation
NS
NS
Boilers permanently
ceasing the combustion of
coal by 2028
NS
NS
Surface Impoundment
NS
NS
FGD Wastewater Voluntary Incentives
Program (Direct Dischargers Only)
Evaporation
Membrane Filtration
Membrane Filtration
Membrane Filtration
NA
Bottom Ash
Transport
Water
NA (default unless in
subcategory)15
Dry Handling / Closed
Loop
High Recycle Rate
Systems
High Recycle Rate
Systems
High Recycle Rate
Systems
High Recycle Rate
Systems
Low Utilization Boilers
NS
NS
Surface Impoundment
+ BMP Plan
NS
NS
Boilers permanently
ceasing the combustion of
coal by 2028
NS
NS
Surface Impoundment
NS
NS
Abbreviations: BMP = Best Management Practice; HRTR = High Residence Time Reduction; LRTR = Low Residence Time Reduction; NS = Not subcategorized (default technology basis
applies); NA = Not applicable
a.	See Supplemental TDD for a description of these technologies (U.S. EPA, 2020g).
b.	The table does not present existing subcategories included in the 2015 rule as EPA did not reopen the existing subcategorization of oil-fired units or units with a nameplate capacity
of 50 MW or less.
Source: U.S. EPA Analysis, 2020
1-4

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
1.3 Analytic Framework
The analytic framework of this benefit-cost analysis (BCA) includes basic components used consistently
throughout the analysis of benefits and social costs4 of the regulatory options:
1.	All values are presented in 2018 dollars;
2.	Future benefits and costs are discounted using rates of 3 percent and 7 percent back to 2020;
3.	Benefits and costs are analyzed over a 27-year period (2021 to 2047);
4.	Technology installation and the resulting pollutant loading changes occur at the end of the estimated
wastewater treatment technology implementation year;
5.	Benefits and costs are annualized;
6.	Positive values represent an increase in benefits (improvements in environmental conditions or
incremental social costs) compared to baseline, whereas negative values represent forgone benefits
(or social cost savings) compared to the baseline; and
7.	Future values account for annual U.S. population and income growth, unless noted otherwise.
These components are discussed in the sections below.
EPA's analysis of the regulatory options generally follows the methodology the Agency used previously to
analyze the 2015 rule (see Benefit and Cost Analysis for the Final Effluent Limitations Guidelines and
Standards for the Steam Electric Power Generating Point Source Category (U.S. EPA, 2015a) and 2019
proposed rule (see U.S. EPA, 2019a). In analyzing the regulatory options, however, EPA made several
changes relative to the analysis of the 2019 proposal:
•	EPA used revised inputs that reflect the costs and loads estimated for regulatory options A through C
(see Supplemental TDD and RIA for details) and estimated loading reductions for two distinct periods
during the overall period of analysis to account for transitional conditions when different plants are in
the process of installing technologies to meet the requirements under the final rule.
•	EPA updated the baseline industry information to incorporate changes in the universe and operational
characteristics of steam electric power plants such as electricity generating unit retirements and fuel
conversions since the analysis of the 2019 proposal. EPA also incorporated updated information on
the technologies and other controls that plants employ. See the Supplemental TDD for details on the
changes (U.S. EPA, 2020g).
•	Finally, EPA made certain changes to the methodologies to be consistent with approaches used by the
Agency for other rules and/or incorporate recent advances in environmental assessment, health risk,
and resource valuation research.
These changes are described in the relevant sections of this document, and summarized in Appendix A.
Unless otherwise noted, costs represented in this document are social costs.
1-5

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
1.3.1	Constant Prices
This BCA applies a year 2018 constant price level to all future monetary values of benefits and costs. Some
monetary values of benefits and costs are based on actual past market price data for goods or services, while
others are based on other measures of values, such as household willingness-to-pay (WTP) surveys used to
monetize ecological changes resulting from surface water quality changes. This BCA updates market and
non-market prices using the Consumer Price Index (CPI), Gross Domestic Product (GDP) implicit price
deflator, or Construction Cost Index (CCI).5
1.3.2	Discount Rate and Year
This BCA estimates the annualized value of future benefits using two discount rates: 3 percent and 7 percent.
The 3 percent discount rate reflects society's valuation of differences in the timing of consumption; the
7 percent discount rate reflects the opportunity cost of capital to society. In Circular A-4, the Office of
Management and Budget (OMB) recommends that 3 percent be used when a regulation affects private
consumption, and 7 percent in evaluating a regulation that would mainly displace or alter the use of capital in
the private sector (OMB, 2003; updated 2009). The same discount rates are used for both benefits and costs.
All future cost and benefit values are discounted back to 2020.6
1.3.3	Period of Analysis
Benefits are projected to begin accruing when each plant implements the control technologies needed to
comply with any applicable BAT effluent limitations or pretreatment standards. As discussed in the RIA (in
Chapter 3), for the purpose of the economic impact and benefit analysis, EPA generally estimates that plants
will implement bottom ash transport water control technologies to meet the applicable rule limitations and
standards as their permits are renewed over the period of 2021 through 2025. However, some regulatory
options provide a longer period to meet FGD effluent limits. Under Options A and B, plants may implement
FGD wastewater controls as late as 20287 and under Option C, plants have until 2028 to meet FGD
wastewater controls based on the membrane technology.8 This schedule reflects differing levels of controls
that may be needed to meet limits under different options as compared to the baseline and recognizes that
control technology implementation is likely to be staggered over time across the universe of steam electric
power plants.
The period of analysis extends to 2047 to capture the estimated life of the compliance technology at any
steam electric power plant (20 or more years), starting from the year of technology implementation, which
can be as late as 2028.
5	To update the value of a Statistical Life (VSL), EPA used the GDP deflator and the elasticity of VSL with respect to income of
0.4, as recommended in EPA's Guidelines for preparing Economic Analysis (U.S. Environmental Protection Agency, 2010a).
EPA used the GDP deflator to update the value of an IQ point, CPI to update the WTP for surface water quality improvements,
cost of illness (COI) estimates, and the price of water purchase, and the CCI to update the cost of dredging navigational
waterways and reservoirs.
6	In its analysis of the 2015 rule, EPA presented benefits in 2013 dollars and discounted these benefits and costs to 2015 (see U.S.
EPA, 2015a), whereas the analysis of the 2019 proposed rule and this analysis used 2018 dollars and discounted benefits and
costs to 2020.
7	The VIP program under Options A and B allows facilities to implement FGD controls as late as 2028. Plants that are not
participating in the VIP program may implement FGD controls as late as 2025.
8	Different dates may apply to subcategories of facilities as described in Section 3.2.1.
1-6

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
The different compliance years between options, wastestreams, and plants means that environmental changes
may occur in a staggered fashion over the analysis period as plants implement control technologies to meet
applicable limits under each option. To analyze environmental changes from the baseline and resulting
benefits, EPA used the annual average of loadings or other environmental changes (e.g., air emissions, water
withdrawals) projected during two distinct periods (2021-2028 and 2029-2047) within the overall analysis
period (2021-2047). Section 3.2 provides further details on the breakout of the analysis periods.
1.3.4 Timing of Technology Installation and Loading Reductions
For the purpose of estimating benefits and social costs, EPA estimates that plants meet revised applicable
limitations and standards by the end of their estimated technology implementation year and that any resulting
changes in loadings will be in effect at the start of the following year.
1.3.5 Annualization of future costs and benefits
Consistent with the timing of technology installation and loading reductions described above, EPA uses the
following equation to annualize the future stream of costs and benefits:
Equation 1-1.
r(PV)
AV = —
(1 + r) [1 — (1 + r) n]
Where A V is the annualized value, PVis the present value, r is the discount rate (3 percent or 7 percent), and n
is the number of years (27 years).
1.3.6 Direction of Environmental Changes and Benefits
The technology bases or subcategorizations shown in Table 1-1 for some regulatory options yield effluent
limitations and standards that may be less stringent than the baseline. This is true, for example, for discharges
of pollutants in bottom ash transport water or for subcategories under which FGD effluent limitations and
standards are based on chemical precipitation only. Additionally, the delayed compliance deadline for FGD
limitations and standards under some options, such as the 2028 deadline for meeting FGD wastewater
limitations and standards based on membrane filtration technology under Option C, prolongs the period when
plants would continue to operate their existing systems and discharge at current levels. The combination of
these factors means that some options can be expected to provide negative benefits (i.e.. disbenefits or
forgone benefits) when compared to the baseline. This document uses the generic term "benefits" whether the
changes are truly beneficial or are detrimental to society (reduce social welfare). The sign, positive or
negative, communicates the direction of the effects. Under this convention, positive benefit values indicate
improvements in social welfare under the option as compared to the baseline. This effect is typically in the
opposite direction as the change in environmental effects. For example, lower effluent pollutant
concentrations (negative changes) reduce the incidence of the health effects being quantified (negative
changes) and avoid excess mortality resulting from the exposure (positive changes).
1.3.7 Population and Income Growth
To account for future population growth or decline, EPA used the U.S. Census Bureau population forecasts
for the United States (U.S. Census Bureau, 2017). EPA used the growth projections for each year to adjust
affected population estimates for future years (i.e.. from 2021 to 2047).
Because WTP is expected to increase as income increases, EPA accounted for income growth for estimating
the value of avoided premature mortality based on the value of a statistical life (VSL) and WTP for water
1-7

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
quality improvements. To develop adjustment factors for VSL, EPA first used income growth factors in the
Environmental Benefits Mapping and Analysis Program—Community Edition (BenMAP-CE) database
between 1990 and 2025 to estimate a linear regression model, which the Agency then used to extrapolate the
income growth factors for years 2026-2047. EPA applied the projected income data along with the income
elasticity for the respective models (VSL and meta-regression) to adjust the VSL and meta-analysis estimates
of WTP for water quality changes in future years.9'10
1.4 Organization of the Benefit and Cost Analysis Report
This BCA report presents EPA's analysis of the benefits of the regulatory options, assessment of the total
social costs, and comparison of the social costs and monetized benefits.
The remainder of this report is organized as follows:
• Chapter 2 provides an overview of the main benefits expected to result from the implementation of
the main regulatory options analyzed for the final rule.
• Chapter 3 describes EPA's estimates of the environmental changes resulting from the regulatory
options, including water quality modeling that underlays estimates of several categories of benefits.
• Chapters 4 and 5 details the methods and results of EPA's analysis of human health benefits from
changes in pollutant exposure via the drinking water and fish ingestion pathways, respectively.
• Chapter 6 discusses EPA's analysis of the nonmarket benefits of changes in surface water quality
resulting from the regulatory options.
• Chapter 7 discusses EPA's analysis of changes in benefits to threatened and endangered (T&E)
species.
• Chapter 8 describes EPA's analysis of benefits associated with changes in emissions of air pollutants
associated with energy use, transportation, and the profile of electricity generation for the regulatory
options.
• Chapter 9 discusses benefits arising from changes in water withdrawals.
• Chapter 10 describes benefits from changes in maintenance dredging of navigational channels and
reservoirs.
• Chapter 11 summarizes monetized benefits across benefit categories.
• Chapter 12 summarizes the social costs of the regulatory options.
9 These extrapolated income elasticity factors were originally developed for EPA's COBRA tool. The latest public version is 4.0
released in June 2020 ('https://www.epa.gov/statelocalenergv').
10 There is a relatively strong consensus in economic literature that income elasticities of approximately "1" are appropriate for
adjusting WTP for water quality improvements in future years (Johnston et al., 2019; Tyllianakis & Skuras, 2016). Therefore,
EPA used an income elasticity of "1" in this analysis.
1-8

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
1: Introduction
•	Chapter 13 addresses the requirements of Executive Orders that EPA is required to satisfy for the
final rule, notably Executive Order (EO) 12866, which requires EPA to compare the benefits and
social costs of its actions.
•	Chapter 14 details EPA's analysis of the distribution of benefits across socioeconomic groups to
fulfill requirements under EO 12898 on Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations.
•	Chapter 15 provides references cited in the text.
Several appendices provide additional details on selected aspects of analyses described in the main text of the
report.
1-9

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2 Benefits Overview
This chapter provides an overview of the welfare effects to society resulting from changes in pollutant
loadings due to implementation of the main regulatory options analyzed for the final rule. EPA expects the
regulatory options to change discharge loads of various categories of pollutants when fully implemented. The
categories of pollutants include conventional (such as suspended solids, biochemical oxygen demand (BOD),
and oil and grease), priority (such as mercury [Hg], arsenic [As], and selenium [Se]), and non-conventional
pollutants (such as total nitrogen [TN], total phosphorus [TP], chemical oxygen demand [COD] and total
dissolved solids [TDS]).
Table 2-1 presents estimated annual pollutant loads under full implementation of the effluent limitations and
standards for the baseline and the regulatory options. The Supplemental TDD provides further detail on the
loading changes (U.S. EPA, 2020g). As described in Section 3.2, loadings during interim years before all
plants meet the requirements under the final rule differ from these values.
Table 2-1: Estimated Annual Pollutant Loadings and Changes in Loadings for Baseline and
Regulatory Options Under Technology Implementation

Estimated Total Industry Pollutant
Estimated Changes3 in Pollutant
Regulatory Option
Loadings
Loadings from Baseline

(pounds per year)
(pounds per year)
Baseline
1,530,000,000
NA

Option A (Final Rule)
1,530,000,000
-972,000
Option B
1,510,000,000
-14,700,000
Option C
15,600,000
-1,510,000,000
NA: Not applicable to the baseline
Note: Pollutant loadings and removals are rounded to three significant figures, so figures do not sum due to independent rounding.
For example, estimated changes in pollutant loadings from baseline for Option A are calculated as 1,528,154,581 Ib/yr -
1,529,126,625 Ib/yr = -972,044 Ib/yr which when rounded to three significant figures becomes 1,530,000,000 - 1,530,000,000 in
this table but still results in -972,000 Ib/yr. See Supplemental TDD (U.S. EPA, 2020g) and Document Control Number (DCN) SE08644
for details.
a. Negative values represent loading reductions and positive values represent loading increases, compared to the baseline.
b. Regulatory Option D reflects the population, methodology, and pollutant loadings for Option 1 in the 2019 proposed rule (see
Section 6.4 of the 2019 TDD (U.S. EPA, 2019k). The values do not reflect changes in the baseline, plant universe, and other
analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
As discussed in Section 1.3.4, some of the options may increase pollutant loadings for some plants,
wastestreams, pollutants, or years, when compared to the baseline. Technology options resulting in an overall
increase in pollutant loadings would result in forgone benefits to society while options resulting in loading
reductions would result in realized benefits. Furthermore, whether a regulatory option increases or reduces
loadings depends on the particular plant, pollutant, and timing of the comparison to baseline conditions.
Section 3.2 discusses the temporal profile of pollutant loads in further detail.
Changes estimated for Option A and Option B include effects of the VIP. Because participation in the VIP is
voluntary, the number of plants that may participate in the program is uncertain. For the purposes of the costs
and benefits analyses, EPA estimated VIP participants by comparing the discounted total annualized cost of
chemical precipitation + LRTR biological treatment and membrane filtration for each plant, with the
2-1

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
expectation that a plant owner would select the less costly of the two. The Agency estimated that eight steam
electric power plants may choose to participate in the VIP under Option A and 14 plants may choose to
participate in the VIP under Option B. The plants for which EPA estimates VIP may be the least-cost option
vary in FGD wastewater flows, nameplate capacity, capacity utilization, and location. For these plants, EPA
retained the membrane filtration costs for estimating economic impacts in the RIA and social costs in Chapter
12, and the membrane filtration loadings for the benefits analysis.
Effects of the regulatory options in comparison to the 2015 rule also include other effects of the
implementation of control technologies or other changes in plant operations, such as changes in emissions of
air pollutants (e.g., carbon dioxide [CO2], fine particulate matter [PM2.5], nitrogen oxides [NOx], and sulfur
dioxide [SO2]) which result in benefits to society in the form of changes in morbidity and mortality and CO2
impacts on environmental quality and economic activities. Other effects include changes in water use, which
provide benefits in the form of changes in the availability of surface water and groundwater.
This chapter also provides a brief discussion of the effects of pollutants found in bottom ash transport water
and FGD wastewater addressed by the regulatory options on human health and ecosystem services, and a
framework for understanding the benefits expected to be achieved by these options. For a more detailed
description of steam electric wastewater pollutants, their fate, transport, and impacts on human health and
environment, see the Supplemental EA (U.S. EPA, 2020f).
Figure 2-1 summarizes the potential effects of the regulatory options, the expected environmental changes,
and categories of social welfare effects as well as EPA's approach to analyzing those welfare effects. EPA
was not able to bring the same depth of analysis to all categories of social welfare effects because of imperfect
understanding of the link between discharge changes or other environmental effects of the regulatory options
and welfare effect categories, and how society values some of these effects. EPA was able to quantify and
monetize some welfare effects, quantify but not monetize other welfare effects, and assess still other welfare
effects only qualitatively. The remainder of this chapter provides a qualitative discussion of the social welfare
effects applicable to this rule, including human health effects, ecological effects, economic productivity, and
changes in air pollution, solid waste generation, and water withdrawals. Some estimates of the monetary value
of social welfare changes presented in this document rely on models with a variety of limitations and
uncertainties, as discussed in more detail in Chapters 3 through 10 for the relevant benefit categories.
2-2

-------
BCAfor Revisions to Steam Electric Power Generating ELGs
2: Benefits Overview
Figure 2-1: Summary of Benefits Resulting from the Regulatory Options.
Changes in water
Surface water
withdrawals
Avoided cost of water purchase
Qualitative discussion
Fish tissue
contamination
Change in air
emissions of CO2,
PM2.5, NOx, and S0;
Surface water
quality
Effect of
Regulatory
Options
Environmental
Change
Changes in groundwater availability
Avoided cost of dredging
Qualitative discussion
Qualitative discussion
Changes in impingement and entrainment mortality
Changes in toxic,
bioaccumulative,
and other harmful
pollutants to surface
waters
CQ
VSL
COI
Social cost of carbon
Ability to market ash
for beneficial use
Groundwater
withdrawals
WTP for water quality improvements
Count of aquatic life criteria
exceedances (non-monetized)
Qualitative discussion
Valuation
Changes in disposal costs
Changes in life-cycle impacts and costs of virgin raw materials
Change in:
•	Power
requirements
•	Trucking
•	Electricity
generation
Conversion to dry
systems
Value of an IQ point
Count of human health criteria
exceedances (non-monetized)
Qualitative discussion
Benefit Category
Change in premature mortality, non-fatal heart attacks, hospital
admissions, emergency department visits, upper and lower
respiratory symptoms, acute bronchitis, aggravated asthma, lost
work days and acute respiratory symptoms
Changes in C02 impacts
Economic Productivity
Changes in dredging costs for maintaining navigational
waterways and reservoir capacity
Changes in drinking water treatment costs
Changes in property values
Changes in tourism
Changes in commercial fishery yields
Human Health Effects (Fish Consumption Exposure)
Changes in IQ losses in children from mercury and lead exposure
Changes in cardiovascular disease from lead exposure
Changes in cancer cases from arsenic exposure
Changes in cases of other cancer and non-cancer health effects
Human Health Effects (Drinking Water Exposure)
Changes in cancer incidence (bladder, colon, and rectal) and
reproductive and developmental effects from exposure to
halogenated DBPs in treated water
Ecological Conditions
Changes in recreational and non-use values
Changes in threatened and endangered (T&E) species protection
DBP= Disinfection byproducts; WTP = Willingness to Pay; VSL = Value of Statistical Life; COI =Cost of illness
Source: U.S. EPA Analysis, 2020.
2-3

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2.1 Human Health Impacts Associated with Changes in Surface Water Quality
Pollutants present in steam electric power plant wastewater discharges can cause a variety of adverse human
health effects. Chapter 3 describes the approach EPA used to estimate changes in pollutant levels in waters.
More details on the fate, transport, and exposure risks of steam electric pollutants are provided in the EA
(U.S. EPA, 2015b, 2020f).
Human health effects are typically analyzed by estimating the change in the expected number of adverse
human health events in the exposed population resulting from changes in effluent discharges. While some
health effects (e.g., cancer) are relatively well understood and can be quantified in a benefits analysis, others
are less well characterized and cannot be assessed with the same rigor, or at all.
The regulatory options affect human health risk by changing exposure to pollutants in water via two principal
exposure pathways discussed below: (1) treated water sourced from surface waters affected by steam electric
power plant discharges and (2) fish and shellfish taken from waterways affected by steam electric power plant
discharges. The regulatory options also affect human health risk by changing air emissions of pollutants via
shifts in the profile of electricity generation, changes in auxiliary electricity use, and transportation; these
effects are discussed separately in Section 2.4.
2.1.1 Drinking Water
Pollutants discharged by steam electric power plants to surface waters may affect the quality of water used for
public drinking supplies. People may then be exposed to harmful constituents in treated water through
ingestion, as well as inhalation and dermal absorption (e.g., showering, bathing). The pollutants may not be
removed adequately during treatment at a drinking water treatment plant, or constituents found in steam
electric power plant discharges may interact with drinking water treatment processes and contribute to the
formation of disinfection byproducts (DBPs).
Public drinking water supplies are subject to legally enforceable maximum contaminant levels (MCLs)
established by EPA (U.S. EPA, 2018b). As the term implies, an MCL for drinking water specifies the highest
level of a contaminant that is allowed in drinking water. The MCL is based on the MCL Goal (MCLG), which
is the level of a contaminant in drinking water below which there is no known or expected risk to human
health. EPA sets the MCL as close to the MCLG as possible, with consideration for the best available
treatment technologies and costs. Table 2-2 shows the MCL and MCLG for selected constituents or
constituent derivatives of steam electric power plant effluent.
Table 2-2: Drinking Water Maximum Contaminant Levels and Goals for Selected Pollutants in Steam
Electric FGD Wastewater or Bottom Ash Transport Water Discharges
Pollutant
MCL
MCLG

(mg/L)
(mg/L)
Antimony
0.006
0.006
Arsenic
0.01
0
Barium
2.0
2.0
Beryllium
0.004
0.004
Bromate
0.010
0
Cadmium
0.005
0.005
Chromium (total)
0.1
0.1
Copper3
1.3
1.3
Cyanide (free cyanide)
0.2
0.2
Lead3
0.015
0
2-4

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 2: Benefits Overview
Table 2-2: Drinking Water Maximum Contaminant Levels and Goals for Selected Pollutants in Steam
Electric FGD Wastewater or Bottom Ash Transport Water Discharges
Pollutant
MCL
MCLG

(mg/L)
(mg/L)
Mercury
0.002
0.002
Nitrate-Nitrite as N
10 (Nitrate); 1 (Nitrite)
10 (Nitrate); 1 (Nitrite)
Selenium
0.05
0.05
Thallium
0.002
0.0005
Total trihalomethanes15
0.080
Not applicable
bromodichloromethane
Not applicable
0
bromoform
Not applicable
0
dibromochloromethane
Not applicable
0.06
chloroform
Not applicable
0.07
a. MCL value is based on action level.
b. Bromide, a constituent found in steam electric power plant effluent, is a precursor is a precursor for Total Trihalomethanes and
three of its subcomponents. Additional trihalomethanes may also be formed in the presence of iodine, a constituent also found in
steam electric power plant wastewater discharges.
Source: 40 CFR 141.53 as summarized in U.S. EPA (2018b): National Primary Drinking Water Regulation, EPA 816-F-09-004
Pursuant to MCLs, public drinking water supplies are tested and treated for pollutants that pose human health
risks. For the purpose of analyzing the human health benefits of the regulatory options, EPA estimates that
treated water meets applicable MCLs in the baseline. Table 2-2 shows that for arsenic, bromate, lead, and
certain trihalomethanes, the MCLG is zero. For these pollutants and for those that have an MCL above the
MCLG (thallium), there may be incremental benefits from reducing concentrations below the MCL.
EPA used a mass balance approach to estimate the changes in halogen (bromide and iodine) levels in surface
waters downstream from steam electric power plant outfalls. Halogens can be precursors for halogenated
disinfection byproduct formation in treated drinking water, including trihalomethanes addressed by the total
trihalomethanes (TTHM) MCL. The occurrence of TTHM and other halogenated disinfection byproducts in
downstream drinking water depends on a number of environmental factors and site-specific processes at
drinking water treatment plants. There is evidence of linkages between adverse human health effects,
including bladder cancer, and exposure to halogenated disinfection byproducts in drinking water. For
additional information on these topics, see the Supplemental EA (U.S. EPA, 2020f). For the 2019 proposed
rule, EPA quantitatively estimated the effect of changes in surface water bromide levels on drinking water
TTHM levels and bladder cancer incidence in exposed populations. EPA also monetized associated changes
in human mortality and morbidity. EPA received public comments that further evaluation of certain DBPs
should be completed and that the analysis at proposal should be subject to peer review. The Agency
acknowledges that further study in this area should be conducted, including peer review of the model used at
proposal. EPA did not update this analysis for the final rule beyond updating the downstream surface water
concentrations of bromide and iodine but will continue to evaluate the scientific data on the health impacts of
disinfection byproducts.11.
To the extent the proposed rule analysis accurately quantified human health effects, the final rule's
quantitative benefits analysis may underestimate human health-related benefits.
11 Where information is available on actual or expected concentrations for particular DBPs, the human health impacts can be
monetized for those specific DBPs as was done in the Stage 2 Disinfection Byproduct Rule.
2-5

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
To assess potential for changes in health risk from exposure to arsenic, lead, and thallium in drinking water,
EPA estimated changes in pollutant levels in source waters downstream from steam electric power plants
under each policy option. This analysis is discussed in Section 4.2. EPA did not quantify or monetize benefits
from reduced exposure to arsenic, lead, and thallium via drinking water due to the very small concentration
changes in source waters downstream from steam electric plants. EPA notes that lead found in supplied water
is generally associated with the water distribution infrastructure rather than source water quality.
2.1.2 Fish Consumption
Recreational and subsistence fishers (and their household members) who consume fish caught in the reaches
downstream of steam electric power plants may be affected by changes in pollutant concentrations in fish
tissue. EPA analyzed the following direct measures of change in risk to human health from exposure to
contaminated fish tissue:
1. Neurological effects to children ages 0 to 7 from exposure to lead;
2. Neurological effects to infants from in-utero exposure to mercury;
3. Incidence of skin cancer from exposure to arsenic12; and
4. Reduced risk of other cancer and non-cancer toxic effects.
The Agency evaluated changes in potential intellectual impairment, or intelligence quotient (IQ), resulting
from changes in childhood and in-utero exposures to lead and mercury. The EPA also translated changes in
the incidence of skin cancer into changes in the number of skin cancer cases.
For constituents with human health ambient water quality criteria, the change in the risk of other cancer and
non-cancer toxic effects from fish consumption is addressed indirectly in EPA's assessment of changes in
exceedances of these criteria (see Section 5.1).
EPA used a cost-of-illness (COI) approach to estimate the value of changes in the incidence of skin cancer,
which are generally non-fatal (see Section 5.5). The COI approach allows valuation of a particular type of
non-fatal illness by placing monetary values on measures, such as lost productivity and the cost of health care
and medications that can be monetized. Some health effects of changes in exposure to steam electric
pollutants, such as neurological effects to children and infants exposed to lead and mercury, are measured
based on avoided IQ losses. Changes in IQ cannot be valued based on WTP approaches because the available
economic research provides little empirical data on society's WTP to avoid IQ losses. Instead, EPA calculated
monetary values for changes in neurological and cognitive damages based on the impact of an additional IQ
point on an individual's future earnings and the cost of compensatory education for children with learning
disabilities. These estimates represent only one component of society's WTP to avoid adverse neurological
effects and therefore produce a partial measure of the monetary value from changes in exposure to lead and
mercury. Employed alone, these monetary values would underestimate society's WTP to avoid adverse
neurological effects. See Sections 5.3 and 5.4 for applications of this method to valuing health effects in
children and infants from changes in exposure to lead and mercury. Although EPA performed a screening
analysis for the 2019 proposal, which indicated very small changes in cardiovascular disease mortality for the
12 EPA is currently revising its cancer assessment of arsenic to reflect new data on internal cancers including bladder and lung
cancers associated with arsenic exposure via ingestion (U.S. EPA, 2010b). Because cancer slope factors for internal organs have
not been finalized, the Agency did not consider these effects in the analysis of the final rule.
2-6

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
proposed rule options compared to those estimated in the analysis for the 2015 rule, EPA is not estimating
avoided cardiovascular mortality that may result from the final rule. EPA acknowledges the scientific
understanding of the relationship between lead exposure and cardiovascular mortality is evolving and
scientific questions remain. (See also U.S. EPA, 2019c).
EPA received comments that it did not evaluate potential health impacts via the fish consumption pathway
arising from changes in discharges of other steam electric pollutants, such as aluminum, boron, cadmium,
hexavalent chromium, manganese, selenium, thallium, and zinc U.S. EPA, 2020f. Analyses of these health
effects require data and information on the relationships between ingestion rate and potential adverse health
effects and on the economic value of potential adverse health effects. Thus, due to data limitations and
uncertainty in these quantitative relationships, for the final rule EPA did not quantify, nor was it able to
monetize, changes in health effects associated with exposure to these pollutants. Despite numerous studies
conducted by EPA and other researchers, dose-response functions are available for only a subset of health
endpoints associated with steam electric wastewater pollutants. In addition, the available research does not
always allow complete economic evaluation, even for quantifiable health effects. For example, sufficient data
are not available to evaluate and monetize the following potential health effects from fish consumption: low
birth weight and neonatal mortality from in-utero exposure to lead and other impacts to children from
exposure to lead, such as decreased postnatal growth in children ages one to 16, delayed puberty,
immunological effects, and decreased hearing and motor function (Cleveland etal., 2008; NTP, 2012; U.S.
EPA, 2013c; 2019c); effects to adults from exposure to lead such as cardiovascular diseases, decreased kidney
function, reproductive effects, immunological effects, cancer and nervous system disorders (Aoki et al., 2016;
Chowdhury et al., 2018; Lanphear et al., 2018; NTP, 2012; U.S. EPA, 2013c; 2019c); neurological effects to
children from exposure to mercury after birth (Grandjean et al. ,2014); effects to adults from exposure to
mercury, including vision defects, hand-eye coordination, hearing loss, tremors, cerebellar changes, and
others (Mergler et al, 2007; Center for Disease Control and Prevention (CDC), 2009); and other cancer and
non-cancer effects from exposure to other steam electric pollutants (e.g., kidney, liver, and lung damage from
exposure to cadmium, reproductive and developmental effects from exposure to arsenic, boron, and thallium,
liver and blood effects from exposure to hexavalent chromium, and neurological effects from exposure to
manganese) (California EPA, 2011; Oulhote et al., 2014; Roels et al., 2012; U.S. Department of Health and
Human Services, 2012; U.S. EPA, 2020f).
EPA received comments that its analyses supporting the proposal didn't fully consider cumulative or
synergistic effects. Data and resource limitations preclude a full analysis cumulative or synergistic effects of
pollutants that share the same toxicity mechanism, affect the same body organ or system, or result in the same
health endpoint. For example, exposure to several pollutants discharged by steam electric plants (i.e., lead,
mercury, manganese, and aluminum) is associated with adverse neurological effects, in particular in fetuses
and small children (Agency for Toxic Substances and Disease Registry (ATSDR), 2009; Grandjean et al.,
2014; NTP, 2012; Oulhote etal., 2014; U.S. EPA, 2013c). A weight of evidence approach is typically used in
qualitatively evaluating the cumulative effect of a chemical mixture. Cumulative effects often depend on
exposure doses as well as potential threshold effects (ATSDR, 2004; 2009).
Due to these limitations, the total monetary value of changes in human health effects included in this analysis
represent only a subset of the potential health benefits (or forgone benefits) that are expected to result from
the regulatory options.
2-7

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2.1.3 Complementary Measure of Human Health Impacts
EPA quantified but did not monetize changes in pollutant concentrations in excess of human health-based
national recommended water quality criteria (NRWQC). This analysis provides a measure of the change in
cancer and non-cancer health risk by comparing the number of receiving reaches exceeding health-based
NRWQC for steam electric pollutants in the baseline to the number exceeding NRWQC under the regulatory
options (Section 5.7).
Because the NRWQC in this analysis are set at levels to protect human health through ingestion of water and
aquatic organisms, changes in the frequency at which human health-based NRWQC are exceeded could
translate into changes in risk to human health. This analysis should be viewed as an indirect indicator of
changes in risk to human health because it does not reflect the magnitude of human health risk changes or the
population over which those changes would occur.
2.2 Ecological and Recreational Impacts Associated with Changes in Surface Water Quality
The composition of steam electric power plant wastewater depends on a variety of factors, such as fuel
composition, air pollution control technologies, and wastewater management techniques. Wastewater often
contains toxic pollutants such as aluminum, arsenic, boron, cadmium, chromium, copper, iron, lead,
manganese, mercury, nickel, selenium, thallium, vanadium, molybdenum, and zinc (U.S. EPA, 2020f).
Discharges of these pollutants to surface water can have a wide variety of environmental effects, including
fish kills, reduction in the survival and growth of aquatic organisms, behavioral and physiological effects in
wildlife, and degradation of aquatic habitat in the vicinity of steam electric power plant discharges (U.S. EPA,
2020f). The adverse effects associated with releases of steam electric pollutants depend on many factors such
as the chemical-specific properties of the effluent, the mechanism, medium, and timing of releases, and site-
specific environmental conditions.
The modeled changes in environmental impacts are quite small. Still, EPA expects the ecological impacts
from the regulatory options could include habitat changes for fresh- and saltwater plants, invertebrates, fish,
and amphibians, as well as terrestrial wildlife and birds that prey on aquatic organisms exposed to steam
electric pollutants. The change in pollutant loadings has the potential to result in changes in ecosystem
productivity in waterways and the health of resident species, including threatened and endangered (T&E)
species. Loadings projected under the rule have the potential to impact the general health of fish and
invertebrate populations, their propagation to waters, and fisheries for both commercial and recreational
purposes. Changes in water quality also have the potential to impact recreational activities such as swimming,
boating, fishing, and water skiing. Finally, the final rule has the potential to impact nonuse values (e.g.,
option, existence, and bequest values) of the waters that receive steam electric power plant discharges.
EPA's analysis is intended to isolate possible effects of the regulatory options and the final rule on aquatic
ecosystems and organisms, including T&E species, however, it does not take into account the fact that the
National Pollutant Discharge Elimination System (NPDES) permit for each steam electric power plant, like
all NPDES permits, is required to have limits more stringent than the technology-based limits established by
an ELG, wherever necessary to protect water quality standards. Because this analysis does not project where a
permit will have more stringent limits than those required by the ELG, it may overestimate any negative
impacts to aquatic ecosystems and T&E species, including impacts that will not be realized at all because the
permits will be written to include limits as stringent as necessary to meet water quality standards as required
by the Clean Water Act (CWA).
2-8

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2.2.1 Changes in Surface Water Quality
The regulatory options may affect the value of ecosystem services provided by surface waters through
changes in the habitats or ecosystems (aquatic and terrestrial) that receive steam electric power plant
discharges. Society values changes in ecosystem services by a number of mechanisms, including increased
frequency of use and improved quality of the habitat for recreational activities (e.g., fishing, swimming, and
boating). Individuals also value the protection of habitats and species that may reside in waters that receive
FGD wastewater and bottom ash transport water discharges, even when those individuals do not use or
anticipate future use of such waters for recreational or other purposes, resulting in nonuse values.
EPA quantified potential environmental impacts from the regulatory options by estimating in-waterway
concentrations of bottom ash transport water and FGD wastewater pollutants and translating water quality
estimates into a single numerical indicator, a water quality index (WQI). EPA used the estimated change in
WQI as a quantitative estimate of ecological changes for this regulatory analysis. Section 3.4 of this report
provides details on the parameters used in formulating the WQI and the WQI methodology and calculations.
In addition to estimating changes using the WQI, EPA compared estimated pollutant concentrations to
freshwater NRWQC for aquatic life (see Section 3.4.1.1). The Supplemental EA (U.S. EPA, 2020f) details
comparisons of the estimated concentrations in immediate receiving and downstream reaches to the
freshwater acute and chronic NRWQC for aquatic life for individual pollutants.
A variety of primary methods exist for estimating recreational use values, including both revealed and stated
preference methods (Freeman III, 2003). Where appropriate data are available or can be collected, revealed
preference methods can represent a preferred set of methods for estimating use values. Revealed preference
methods use observed behavior to infer users' values for environmental goods and services. Examples of
revealed preference methods include travel cost, hedonic pricing, and random utility (or site choice) models.
In contrast to direct use values, nonuse values are considered more difficult to estimate. Stated preference
methods, or benefit transfer based on stated preference studies, are the generally accepted techniques for
estimating these values (OMB, 2003; U.S. EPA, 2010a). Stated preference methods rely on carefully designed
surveys, which either (1) ask people about their WTP for particular environmental improvements, such as
increased protection of aquatic species or habitats with particular attributes, or (2) ask people to choose
between competing hypothetical "packages" of environmental improvements and household cost (Bateman et
al., 2006). In either case, values are estimated by statistical analysis of survey responses.
Although the use of primary research to estimate values is generally preferred because it affords the
opportunity for the valuation questions to closely match the policy scenario, the realities of the regulatory
process often dictate that benefit transfer is the only option for assessing certain types of non-market values
(Rosenberger and Johnston, 2007). Benefit transfer is described as the "practice of taking and adapting value
estimates from past research ... and using them ... to assess the value of a similar, but separate, change in a
different resource" (V. K. Smith et al., 2002, p. 134). It involves adapting research conducted for another
purpose to estimate values within a particular policy context (Bergstrom & De Civita, 1999). EPA followed
the same methodology used in analyzing the 2015 rule and 2019 proposal (U.S. EPA, 2015a) and relied on a
benefit transfer approach based on an updated meta-analysis of surface water valuation studies to estimate the
use and non-use benefits of improved surface water quality resulting from the final rule. The updates
2-9

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
consisted of incorporating WTP estimates from more recent peer review studies into EPA's existing
econometric model.13 This analysis is presented in Chapter 6.
2.2.2 Impacts on Threatened and Endangered Species
For T&E species, even minor changes to reproductive rates and small mortality levels may represent a
substantial portion of annual population growth. By changing the discharge of steam electric pollutants to
aquatic habitats, the regulatory options have the potential to impact the survivability of some T&E species
living in these habitats. These T&E species may have both use and nonuse values. However, given the
protected nature of T&E species and the fact that use activities, such as fishing or hunting, generally
constitute "take" which is illegal unless permitted, the majority of the economic value for T&E species comes
from nonuse values.14
EPA quantified but did not monetize the potential effects of the regulatory options on T&E species. EPA
constructed databases to determine which species have habitat ranges that intersect waters downstream from
steam electric power plants. EPA then queried these databases to identify "affected areas" of those habitats
where 1) receiving waters do not meet aquatic life-based NRWQC under the baseline conditions; and
2) receiving waters do meet aquatic life-based NRWQC under regulatory options, or vice versa. Because
NRWQC are set at levels to protect aquatic organisms, reducing the frequency at which aquatic life-based
NRWQC are exceeded should translate into reduced effects to T&E species and potential improvement in
species populations. Conversely, increasing the frequency of exceedances could potentially impact T&E
species. Therefore, to estimate the benefits of the regulatory options, EPA identified the waterbodies that
overlap with T&E species habitat ranges that see changes in achievement of wildlife NRWQC as a
consequence of the regulatory options and used these data as a proxy for benefits to T&E species.15 This
analysis and results are presented in Chapter 7.
EPA was unable to monetize the final rule's effects on T&E species due to challenges in quantifying the
response of T&E populations to changes in water quality conditions. Although a relatively large number of
economic studies have estimated WTP for T&E protection, these studies focused on estimating WTP to avoid
species loss/extinction, increase in the probability of survival, or an increase in species population levels
(Richardson & Loomis, 2009). These studies suggest that people attach economic value to protection of T&E
species ranging from $10.4 per household per year (in 2018$) for avoiding loss of the striped shiner (a fish
species) to $172 (in 2018$) for doubling salmon population levels.16 In addition, T&E species may serve as a
focus for eco-tourism and provide substantive economic benefit to local communities. For example, Solomon
et al. (2004) estimate that manatee viewing provides a net benefit (tourism revenue minus the cost of manatee
protection) of $12 - $13 million (in 2018$) per year for Citrus County, Florida.17 EPA's analysis does not
account for the potential for the NPDES permit issuance process to establish more stringent site-specific
13	See ICF (2020) for additional detail on updating the meta-analysis.
14	The Federal Endangered Species Act (ESA) defines "take" to mean "to harass, harm, pursue, hunt, shoot, wound, kill, trap,
capture, or collect, or to attempt to engage in any such conduct." 16 U.S. Code § 1532
15	EPA is not required by Section 7(a)(2) of the Endangered Species Act to consult with the Fish and Wildlife Service and National
Marine Fisheries Service prior to promulgating this technology-based rule (see Executive Summary) because the Agency lacks
discretion to account for effects on species when issuing a technology-based rule under sections 301(b), 304(b), 306 and 307(b)
oftheCWA.
16	Values adjusted from $8.32 and $138 per household per year (in 2006$), respectively, using the CPI.
17	Range adjusted from $8.2 - $9 million (in 2001$), using the CPI.
2-10

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
controls to meet applicable water quality standards (/'. e., water quality-based effluent limits issued under
Section 301(b)(1)(C)), relative to baseline. The analysis may therefore overestimate any potential negative
impacts to T&E species and associated forgone benefits.
2.2.3 Changes in Sediment Contamination
Effluent discharges from steam electric power plants can also contaminate waterbody sediments. For
example, sediment adsorption of arsenic, selenium, and other pollutants found in FGD wastewater and bottom
ash transport water discharges can result in accumulation of contaminated sediment on stream and lake beds
(Ruhl etal., 2012), posing a particular threat to benthic (i.e.. bottom-dwelling) organisms. These pollutants
can later be re-released into the water column and enter organisms at different trophic levels. Concentrations
of selenium and other pollutants in fish tissue of organisms of lower trophic levels can bio-magnify through
higher trophic levels, posing a threat to the food chain at large (Ruhl et al., 2012).
In waters receiving direct discharges from steam electric power plants, EPA examined potential exposures of
ecological receptors (/'. e., sediment biota) to pollutants in contaminated sediment. Benthic organisms can be
affected by pollutant discharges such as mercury, nickel, selenium, and cadmium (U.S. EPA, 2015b; 2020f).
The pollutants in steam electric power plant discharges may accumulate in living benthic organisms that
obtain their food from sediments and pose a threat to both the organism and humans consuming the organism.
As discussed in the Supplemental EA, EPA modeled sediment pollutant concentrations in immediate receiving
waters and compared those concentrations to threshold effect concentrations (TECs) for sediment biota (U.S.
EPA, 2020f). In 2015, EPA also evaluated potential risks to fish and waterfowl that feed on aquatic organisms
with elevated selenium levels and found that steam electric power plant selenium discharges elevated the risk
of adverse reproduction impacts among fish and mallards in immediate receiving waters (U.S. EPA, 2015b).
By changing discharges of pollutants to receiving reaches, the final rule may affect the contamination of
waterbody sediments, thereby impacting benthic organisms and changing the probability that pollutants could
later be released into the water column and affect surface water quality and the waterbody food chain. Due to
data limitations, EPA did not quantify or monetize the associated benefits.
2.3 Economic Productivity
The economic productivity changes estimated to result from the regulatory options may include changes in
beneficial use of coal ash and the resulting reduction in disposal costs. Other potential economic productivity
effects may stem from changes in the quality of public drinking water supplies and irrigation water; changes
in sediment deposition in reservoirs and navigational waterways; and changes in tourism, commercial fish
harvests, and property values. Due to the small magnitude of the estimated changes in water quality (see
Chapter 3 for details), only changes in sediment deposition in reservoirs and navigational waterways are
quantified and monetized. Other benefit categories (e.g., effects on drinking water treatment costs) are
discussed qualitatively in the following sections.
2.3.1 Marketability of Coal Ash for Beneficial Use
The regulatory options may prompt certain plants to convert from wet handling of bottom ash to dry handling.
This change could in turn allow plants to more readily market the coal combustion residuals (CCR) to
beneficial uses. In particular, bottom ash can be used as a substitute for sand and gravel in fill applications.
There are economic productivity benefits from plants avoiding certain costs associated with disposing of the
ash as waste and from society or users of the ash avoiding the cost and life-cycle effects associated with the
displaced virgin material. In the analysis of the 2015 rule, EPA quantified the benefits from increased dry
handling of fly ash and bottom ash (see Chapter 10 in U.S. EPA, 2015c). That analysis showed that the
2-11

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
economic value was greatest for fly ash used in concrete production, and smallest for fly ash or bottom ash
used as fill material.
Among the regulatory options considered for the final rule, Options A, B, and C could affect fly ash to the
extent facilities decide to encapsulate membrane filtration brine with fly ash that is currently beneficially
used. Since EPA could not estimate with certainty which facilities might use fly ash for encapsulation versus
an alternative brine management method (e.g., deep well injection), this potential change in fly ash beneficial
use was not quantified and represents an uncertainty in the analysis. With respect to bottom ash, EPA
estimates that only Option A would affect the quantity of bottom ash handled wet when compared to the
baseline, and for that option the estimated increase in bottom ash handled wet is small (246,871 tons per year
at five plants). See the Supplemental TDD for details (U.S. EPA, 2020g). Given the uncertainties associated
with changes in fly ash, the small changes in the quantity of bottom ash handled wet, and the uncertainty
associated with projecting plant-specific changes in marketed bottom ash, EPA did not quantify this benefit
category in the analysis of the final rule.
2.3.2 Water Supply and Use
The regulatory options are projected to change loadings of steam electric pollutants to surface waters by small
amounts relative to baseline, and thus may have small effects on the uses of these waters for drinking water
supply and agriculture.
2.3.2.1 Drinking Water Treatment Costs
The regulatory options have the potential to affect costs of drinking water treatment (e.g., filtration and
chemical treatment) by changing eutrophication levels and pollutant concentrations in source waters.
Eutrophication is one of the main causes of taste and odor impairment in drinking water, which has a major
negative impact on public perceptions of drinking water safety. Additional treatment to address foul tastes and
odors can significantly increase the cost of public water supply.
The Agency conducted a screening-level assessment to evaluate the potential for changes in costs incurred by
public drinking water systems and concluded that such changes, while they may exist, are likely to be
negligible. The assessment involved identifying the pollutants for which treatment costs may vary depending
on source water quality, estimating changes in downstream concentrations of these pollutants at the location
of drinking water intakes, and determining whether modeled water quality changes have the potential to affect
drinking water treatment costs. Based on this analysis, EPA determined that there are no drinking water
systems drawing water at levels that exceed an MCL for metals and other toxics listed in Table 2-2 such as
selenium and cyanide under either the baseline or the regulatory options (see Section 4.2 for details). EPA
estimated no changes in MCL exceedances under the regulatory options. At many drinking water treatment
facilities, treatment system operations do not generally respond to small incremental changes in source water
quality for one pollutant or a small subset of pollutants. Furthermore, associated operations costs are not
expected to change significantly due to small incremental changes in water quality. Accordingly, EPA did not
conduct an analysis of cost changes in publicly operated treatment systems.
Potential effects of the estimated changes in the levels of halogens downstream from steam electric power
plant outfalls on drinking water treatment costs are uncertain for several reasons including that there can be
other environmental sources of halogens and existing treatment technologies in the majority of PWS are not
designed to remove halogens from raw surface waters. Halogens found in source water can react during
routine drinking water treatment to generate harmful DBPs at levels that vary with site-specific conditions
(Good & VanBriesen, 2017, 2019; Regli et al., 2015; U.S. EPA, 2016b). EPA estimated the costs of
2-12

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
controlling DBP levels to the MCL in treated water as part of the Stage 2 Disinfectants and Disinfection
Byproduct Rule (DBPR). These costs include treatment technology changes as well as non-treatment costs
such as routine monitoring and operational evaluations. Public water systems (PWS) may adjust their
operations to control DBP levels, such as changing disinfectant dosage, moving the chlorination point, or
enhancing coagulation and softening. These changes carry "negligible costs" (U.S. EPA, 2005a, page 7-19).
Where those low-cost changes are insufficient to meet the MCL, PWS may need to incur irreversible capital
costs to upgrade their treatment process to use alternative disinfection technologies such as ozone, ultraviolet
light, or chloride dioxide; switch to chloramines for residual disinfection; or add a pre-treatment stage to
remove DBP precursors (e.g., microfiltration, ultrafiltration, aeration, or increased chlorine levels and contact
time). Some drinking water treatment facilities have already had to upgrade their treatment systems as a direct
result of halogen discharges from steam electric power plants (United States of America v. Duke Energy,
2015; Rivin, 2015). However, not all treatment technologies remove sufficient organic matterto control DBP
formation to required levels (Watson etal., 2012 ). Thus, increased halogens levels in raw source water could
translate into permanently higher drinking water treatment costs at some plants, in addition to posing
increased human health risk. Conversely, reducing halogen levels in source waters can reduce the health risk,
even where treatment changes have already occurred.18 In some cases, operation and maintenance (O&M)
costs may also be reduced. EPA did not have data on drinking water treatment technologies at potentially
affected PWS or estimates of how costs for those technologies vary with changes in halogens concentrations
in source water. Since cost data were insufficient, the Agency assessed only the changes in levels of halogens
downstream from steam electric power plant outfalls and the number of people served by PWS with changes
in halogen levels in their source waters (see Section 2.1.1 for a discussion of this benefit category and Chapter
4 for a discussion of the analysis).
2.3.2.2 Irrigation and Other Agricultural Uses
Irrigation accounts for 42 percent of the total U.S. freshwater withdrawals and approximately 80 percent of
the Nation's consumptive water use. Irrigated agriculture provides important contributions to the U.S.
economy accounting for approximately 40 percent of the total farm sales (Hellerstein etal., 2019). Pollutants
in steam electric power plant discharges can affect the quality of water used for irrigation and livestock
watering. Although elevated nutrient concentrations in irrigation water would not adversely affect its
usefulness for plants, other steam electric pollutants, such as arsenic, mercury, lead, cadmium, and selenium
have the potential to affect soil fertility and enter the food chain (National Research Council, 1993). Nutrients
can increase eutrophication, however, promoting cyanobacteria blooms that can kill livestock and wildlife that
drink the contaminated surface water. TDS can impair the utility of water for both irrigation and livestock use.
EPA did not quantify or monetize effects of quality changes in agricultural water sources arising from the
regulatory options due to data limitations and small estimated changes in water quality.
2.3.3 Reservoir Capacity
Reservoirs serve many functions, including storage of drinking and irrigation water supplies, flood control,
hydropower supply, and recreation. Streams can carry sediment into reservoirs, where it can settle and cause
buildup of sediment layers over time, reducing reservoir capacity (Graf et al., 2010) and the useful life of
reservoirs unless measures such as dredging are taken to reclaim capacity (Hargrove et al., 2010; Miranda,
2017). EPA expects that changes in suspended solids discharges under the regulatory options could affect
reservoir maintenance costs by changing the frequency or volume of dredging activity. Changes in sediment
18 Regli et al. (2015) estimated benefits of reducing bromide across various types of water treatment systems.
2-13

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
loads could result in a modest increase in dredging costs in reservoirs under all regulatory options. See
Chapter 10 for details.
2.3.4 Sedimentation Changes in Navigational Waterways
Navigable waterways, including rivers, lakes, bays, shipping channels and harbors, are an integral part of the
United States' transportation network. E. Clark et al., 1985E. Clark et al., 1985E. Clark et al., 1985E. Clark et
al., 1985E. Clark et al., 1985E. Clark et al., 1985E. Clark et al., 1985Navigable channels are prone to reduced
functionality due to sediment build-up, which can reduce the navigable depth and width of the waterway
(Clark et al., 1985; Marc Ribaudo & Johansson, 2006). For many navigable waters, periodic dredging is
necessary to remove sediment and keep them passable. Dredging of navigable waterways can be costly.
EPA estimated that Option C would reduce sediment loadings to surface waters and reduce dredging of
navigational waterways. EPA quantified and monetized these benefits based on the avoided cost for projected
changes in future dredging volumes. Conversely, EPA estimated that small increases in sediment loads under
Options A and B would result in a small increase in dredging costs in navigational waterways. Chapter 10
describes this analysis.
2.3.5 Commercial Fisheries
Pollutants in steam electric power plant discharges can reduce fish populations by inhibiting reproduction and
survival of aquatic species. These changes may negatively affect commercial fishing industries as well as
consumers of fish, shellfish, and fish and seafood products. Estuaries are particularly important breeding and
nursery areas for commercial fish and shellfish species. In some cases, excessive pollutant loadings can lead
to the closure of shellfish beds, thereby reducing shellfish harvests. Improved water quality due to reduced
discharges of steam electric pollutants would enhance aquatic life habitat and, as a result, contribute to
reproduction and survival of commercially harvested species and larger fish and shellfish harvests, which in
turn could lead to an increase in producer and consumer surplus. Conversely, an increase in pollutant loadings
could lead to negative impacts on fish and shellfish harvest.
EPA did not quantify or monetize impacts to commercial fisheries under the regulatory options. EPA
estimated that six steam electric power plants discharge bottom ash transport water or FGD wastewater
directly to the Great Lakes or to estuaries. Although estimated increases or decreases in annual average
pollutant loads under the regulatory options may affect local fish populations and commercial harvest, the
overall effects to commercial fisheries arising from the regulatory options are likely to be negligible. Most
species of fish have numerous close substitutes. The literature suggests that when there are plentiful substitute
fish products, numerous fishers, and a strong ex-vessel market, individual fishers are generally price takers.
Therefore, the measure of consumer welfare (consumer surplus) is unlikely to change as a result of small
changes in fish landings, such as those EPA expects under the regulatory options.
2.3.6 Tourism
Discharges of pollutants may also affect the tourism and recreation industries (e.g., boat rentals, sales at local
restaurants and hotels) and, as a result, local economies in the areas surrounding affected waters due to
changes in recreational opportunities. The effects of water quality on tourism are likely to be highly localized.
Moreover, since substitute tourism locations may be available, increased tourism in one location (e.g., the
vicinity of steam electric power plants) may lead to a reduction in tourism in other locations or vice versa.
Due to the estimated small magnitude of water quality changes expected from the regulatory options (see
Section 3.4 for details) and availability of substitute sites, the overall effects on tourism and, as a result, social
welfare is likely to be negligible. Therefore, EPA did not quantify or monetize this benefit category.
2-14

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2.3.7 Property Values
Discharges of pollutants may affect the aesthetic quality of water resources by changing pollutant discharges
and thus altering water clarity, odor, and color in the receiving and downstream reaches. Technologies
implemented by steam electric power plants to comply with the regulatory options remove nutrients and
sediments to varying degrees and have varying effects on water eutrophication, algae production, and water
turbidity, and other surface water characteristics. Several studies (e.g., K.J. Boyle etal., 1999; Leggett &
Bockstael, 2000; Gibbs etal., 2002; Bin & Czajkowski, 2013; Walsh etal, 2011; Tuttle &Heintzelman,
2014; Netusil etal, 2014; Liu etal, 2017) suggest that both waterfront and non-waterfront properties are
more desirable when located near unpolluted water. Therefore, the value of properties located in proximity to
waters contaminated with steam electric pollutants may increase or decrease due to changes in discharges of
bottom ash transport water or FGD wastewater.
Due to data limitations, EPA was not able to quantify or monetize the potential change in property values
associated with the regulatory options. The magnitude of the potential change depends on many factors,
including the number of housing units located in the vicinity of the affected waterbodies, community
characteristics (e.g., residential density) and housing stock (e.g., single family or multiple family) and the
effects of steam electric pollutants on the aesthetic quality of surface water. Given the small changes in the
aesthetic quality of surface waters that may result from the small changes in pollutant concentrations under
the regulatory options, EPA expects impacts of the final rule on property values to be small. In addition, there
may be an overlap between shifts in property values and the estimated total WTP for surface water quality
changes discussed in Section 2.2.1.
2.4 Changes in Air Pollution
The final rule is expected to affect air pollution through three main mechanisms: 1) changes in energy use by
steam electric power plants to operate wastewater treatment, ash handling, and other systems needed to
comply with the final rule; 2) changes in transportation-related emissions due to changes in trucking of CCR
and other waste to on-site or off-site landfills; and 3) the change in the profile of electricity generation due to
relatively higher cost to generate electricity at plants incurring compliance costs (or conversely, lower
generation costs for plants incurring cost savings). The altered profile of generation can result in lower or
higher air pollutant emissions due to differences in emission factors for coal or natural gas combustion, or
nuclear or hydroelectric power generation.
Of the three mechanisms above, the change in the emissions profile of electricity generation is the only one
that increases emissions under the final rule. As described in Chapter 5 of the 1(1 A. EPA used the Integrated
Planning Model (IPM®), a comprehensive electricity market optimization model that can evaluate impacts
within the context of regional and national electricity markets, to analyze impacts of the final rule (i.e.,
Option A).
Electricity market analyses using IPM indicate that, under the final rule, coal fired electric power generation
may increase by 0.6 percent in 2030 and by 0.4 percent in 2035 and 2040, when compared to the baseline (see
RIA; U.S. EPA, 2020d). These small changes in generation generally result in air emsission increases that are
also relatively small. Changes in coal-based electricity generation as a result of the final rule are compensated
by changes in generation using other fuels or energy sources, such as natural gas, nuclear power, solar, and
wind power. The net changes in air emissions reflect the differences in emissions factors for these other fuels,
as compared to coal-fueled generation. Overall for the three mechanisms (auxiliary services, transportation,
and market-level generation), EPA estimates changes in CO2, SO2, and NOx emissions as compared to the
2-15

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
baseline. EPA also estimates changes in direct emissions of PM2 5, PM10, Hg, and hydrogen chloride (HC1)
from electricity generating units.
CO2 is the most prevalent of the greenhouse gases, which are air pollutants that EPA has determined endanger
public health and welfare through their contribution to climate change. EPA used estimates of the domestic
social cost of carbon (SC-CO2) to monetize the benefits of changes in CO2 emissions as a result of the final
rule The SC-C02 is a metric that estimates the monetary value of projected impacts associated with marginal
changes in CO2 emissions in a given year. It includes a wide range of anticipated climate impacts, such as net
changes in agricultural productivity and human health, property damage from increased flood risk, and
changes in energy system costs, such as reduced costs for heating and increased costs for air conditioning.
Chapter 8 details this analysis.
NOx, and SO2 are known precursors to PM2 5, a criteria air pollutant that has been associated with a variety of
adverse health effects, including premature mortality and hospitalization for cardiovascular and respiratory
diseases (e.g., asthma, chronic obstructive pulmonary disease [COPD], and shortness of breath). EPA
quantified changes in direct PM2 5 emissions and in emissions of PM2 5 and ozone19 precursors NOx and SO2
and assessed impacts of those emission changes on air quality changes across the country using the
Comprehensive Air Quality Model with Extensions (CAMx) (Ramboll Environ International Corporation,
2016). This is the same modeling approach EPA used in analyses of the final Affordable Clean Energy (ACE)
rule (U.S. EPA, 2019i) and of the "National Emission Standards for Hazardous Air Pollutants: Coal- and Oil-
Fired Electric Utility Steam Generating Units - Subcategory of Certain Existing Electric Utility Steam
Generating Units Firing Eastern Bituminous Coal Refuse for Emissions of Acid Gas Hazardous Air
Pollutants" (85 FR 20838; U.S. EPA, 2020c). EPA then used spatial fields of baseline and post-compliance
air pollutant concentrations as input to Benefits Mapping and Analysis Program—Community Edition
(BenMAP-CE) to estimate incremental human health effects (including the potential for premature mortality
and morbidity) from changes in ambient air pollutant concentrations (U.S. EPA, 2018a). Chapter 8 details this
analysis.
The final rule may also affect air quality through changes in electricity generation units emissions of larger
particulate matter (PM10) and hazardous air pollutants (HAP) including mercury and hydrogen chloride. The
health effects of mercury are detailed in the Supplemental EA (U.S. EPA, 2020f). Hydrogen chloride is a
corrosive gas that can cause irritation of the mucous membranes of the nose, throat, and respiratory tract. For
more information about the impacts of mercury and hydrogen chloride emissions, see the Final Mercury and
Air Toxics Standards (MATS) for Power Plants,20 including 2020 revisions to the 2012 Coal- and Oil-Fired
Electric Utility Steam Generating Units National Emission Standards for Hazardous Air Pollutants (85 FR
31286).
In addition to health effects from air emissions, air pollution can create a haze that affects visibility. Reduced
visibility could impact views in national parks by softening the textures, fading colors, and obscuring distant
features and therefore reduce the value of recreational activities (e.g., K. J. Boyle etal., 2016; Pudoudyal et
al., 2013). A number of studies (e.g., Bayer etal., 2006; Beron etal, 2001; Chay & Greenstone, 1998) also
found that reduced air quality and visibility can negatively affect residential property values.
19 Emissions of nitrogen oxides (NOx) lead to formation of both ozone and PM2.5 while SO2 emissions lead to formation of PM2.5
only.
20 See https://www.epa.gov/mats/regulatorv-actions-fmal-mercurv-and-air-toxics-standards-mats-Dower-plants.
2-16

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
2.5 Changes in Water Withdrawals
The final rule may change water withdrawals associated with wet bottom ash transport and wet FGD
scrubbers. In comparison to the baseline, these changes are estimated to be small. The regulatory options are
expected to increase water withdrawals from surface waterbodies under Option A and Option B, and from
aquifers under all regulatory options. Overall, the estimated increase in water withdrawal ranges from
1.4 billion gallons per year (3.94 million gallons per day) under Option A to 1.6 billion gallons per year
(4.49 million gallons per day) under Option B (see Supplemental TDD for details, U.S. EPA, 2020g). EPA
estimates that power plants would decrease water withdrawals by 3.6 billion gallons per year (9.93 million
gallons per day) under Option C.
Increased water use from groundwater sources by steam electric power plants under the regulatory options
could reduce availability of groundwater supplies for alternative uses. One power plant affected by the final
rule relies on groundwater sources. EPA's analysis of potential forgone benefits associated with an increase in
groundwater withdrawal are presented in Chapter 9.
A change in surface water intake would affect impingement and entrainment mortality. An overall increase in
surface water withdrawal under Options A and B would increase impingement and entrainment mortality.
Although the overall increase in water withdrawal is modest, the significance of local ecological impacts is
uncertain and will depend on the overall health of the affected species population as well as species
vulnerability to impingement and entrainment (e.g., if water intakes affect a nursery habitat). A reduction in
water withdrawal under Option C may benefit fish species affected by impingement and entrainment
mortality. Due to data limitations and uncertainty, EPA did not quantify and monetize these benefits as part of
this analysis.
2.6 Summary of Benefits Categories
Table 2-3 summarizes the potential social welfare effects of the regulatory options analyzed for the final rule
and the level of analysis applied to each category. As indicated in the table, only a subset of potential effects
can be quantified and monetized. The monetized welfare effects include changes in some human health risks,
use and non-use values from changes in surface water quality, changes in costs for dredging reservoirs and
navigational waterways, changes in air pollution, and changes in water withdrawals. Other welfare effect
categories, including expected changes of pollutant concentrations in excess of human health-based NRWQC
limits and changes in halogen levels in PWS source waters downstream from steam electric power plants,
were quantified but not monetized. Finally, EPA was not able to quantify or monetize other welfare effects,
including impacts to commercial fisheries or changes in the marketability of coal ash for beneficial use. EPA
evaluated these effects qualitatively as discussed above in Sections 2.1 through 2.5.
Table 2-3: Estimated Welfare Effects of Changes in Pollutant Discharges from Steam Electric Power
Plants
Category
Effect of Regulatory Options
Benefits Analysis
Quantified
Monetized
Methods (Report
Chapter where
Analysis is Detailed)
Human Health Benefits from Surface Water Quality Improvements
Changes in halogen
levels in drinking water
treatment plant source
waters
Changes in halogen levels in PWS source
water
V

Halogen
concentrations in PWS
source water (Chapter
4)
2-17

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
Table 2-3: Estimated Welfare Effects of Changes in Pollutant Discharges from Steam Electric Power
Plants
Benefits Analysis
Category
Effect of Regulatory Options
Quantified
Monetized
Methods (Report
Chapter where
Analysis is Detailed)
Changes in human health
Changes in exposure to halogenated


Qualitative discussion
effects (e.g., bladder
DBPs in drinking water


(Chapter 2)
cancer) associated with




halogenated DBP




exposure via drinking




water




IQ losses to children ages
Changes in childhood exposure to lead
V
V
IQ point valuation
Oto 7
from fish consumption
(Chapter 5)
Need for specialized
Changes in childhood exposure to lead
V
V
Avoided cost (Chapter
education
from fish consumption
5)
Incidence of
Changes in exposure to lead from fish


Qualitative discussion
cardiovascular disease
consumption


(Chapter 2)
IQ losses in infants
Changes in in-utero mercury exposure
V
V
IQ point valuation

from maternal fish consumption
(Chapter 5)
Incidence of cancer
Changes in exposure to arsenic from fish


COI (Chapter 5);

consumption
V
V
Qualitative discussion
(Chapter 2)
Other adverse health
Changes in exposure to toxic pollutants


Human health criteria
effects (cancer and non-
(lead, cadmium, thallium, etc.) via fish
V

exceedances(Chapter
cancer)
consumption or drinking water

5); Qualitative
discussion (Chapter 2)
Reduced adverse health
Changes in exposure to pollutants from


Qualitative discussion
effects
recreational water uses


(Chapter 2)
Ecological Condition and Recreational Use Effects from Surface Water Quality Changes
Aquatic and wildlife
Changes in ambient water quality in



habitat3
receiving reaches



Water-based recreation3
Changes in swimming, fishing, boating,
and near-water activities from water
quality changes



Aesthetics3
Changes in aesthetics from shifts in


Benefit transfer

water clarity, color, odor, including
V
V
(Chapter 6);

nearby site amenities for residing,
Qualitative discussion

working, and traveling


(Chapter 2)
Non-use values3
Changes in existence, option, and
bequest values from improved
ecosystem health



Aquatic organisms and
Changes in risks to aquatic life from



other wildlife3
exposure to steam electric pollutants



Protection of T&E
Changes in T&E species habitat and


Habitat range
species
potential effects on T&E species


intersecting with

populations
V

reaches with NRWQC



exceedances(Chapter
7); Qualitative
discussion (Chapter 2)
Sediment contamination
Changes in deposition of toxic pollutants
to sediment


Qualitative discussion
(Chapter 2)
2-18

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
2: Benefits Overview
Table 2-3: Estimated Welfare Effects of Changes in Pollutant Discharges from Steam Electric Power
Plants
Category
Effect of Regulatory Options
Benefits Analysis
Quantified
Monetized
Methods (Report
Chapter where
Analysis is Detailed)
Market and Productivity Effects
Dredging costs
Changes in costs for maintaining
navigational waterways and reservoir
capacity
V
V
Cost of dredging
(Chapter 10);
Qualitative discussion
(Chapter 2)
Beneficial use of ash
Changes in disposal costs and avoided
lifecycle impacts from displaced virgin
material


Qualitative discussion
(Chapter 2)
Water treatment costs
for drinking water and
irrigation and other
agricultural uses
Changes in quality of source water used
for drinking and irrigation and other
agricultural uses


Qualitative discussion
(Chapter 2)
Commercial fisheries
Changes in fisheries yield and harvest
quality due to aquatic habitat changes


Qualitative discussion
(Chapter 2)
Tourism industries
Changes in participation in water-based
recreation


Qualitative discussion
(Chapter 2)
Property values
Changes in property values from
changes in water quality


Qualitative discussion
(Chapter 2)
Air Quality-Related Effects
Air emissions of PM2.5,
NOx and S02
Changes in mortality and morbidity from
exposure to particulate matter (PM25)
emitted directly or linked to changes in
NOx and S02 emissions (precursors to
PM2.5 and ozone)
V
V
VSL and COI (Chapter
8); Qualitative
discussion (Chapter 2)
Air emissions of NOx and
S02
Changes in ecosystem effects; visibility
impairment; and human health effects
from direct exposure to N02, S02, and
hazardous air pollutants.


Qualitative discussion
(Chapters 2 and 8)
Air emissions of C02
Changes in climate change effects
V
V
Social cost of carbon
(SC-C02) (Chapter 8;
Appendix 1)
Changes in Water Withdrawal
Groundwater
withdrawals
Changes in availability of groundwater
resources
V
V
Cost per gallon of
water withdrawn
(Chapter 9);
Qualitative discussion
(Chapter 2)
Surface water
withdrawals
Changes in vulnerability to impingement
and entrainment mortality


Qualitative discussion
(Chapter 2)
a. These values are implicit in the total WTP for water quality improvements.
Source: U.S. EPA Analysis, 2020
2-19

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
3 Water Quality Effects of Regulatory Options
Changes in the quality of surface waters, aquatic habitats and ecological functions due to the final rule
depend on a number of factors, including the operational characteristics of steam electric power plants,
treatment technologies implemented to control pollutant levels, the timing of treatment technology
implementation, and the hydrography of reaches receiving steam electric pollutant discharges, among
others. This chapter describes the surface water quality changes projected under the regulatory options.
EPA modeled water quality based on loadings estimated for the baseline and for each of the regulatory
options (Options A, B, and C).21 The differences in predicted concentrations between the baseline and
option scenarios represent the changes attributable to the regulatory options. These changes inform the
analysis of several of the benefits described in Chapter 2 and detailed in later chapters of this report.
The analyses use pollutant loading estimates detailed in the Supplemental TDD (U.S. EPA, 2020g) and
expand upon the analysis of immediate receiving waters described in the Supplemental EA (U.S. EPA,
2020f) by estimating changes in both receiving and downstream reaches. The Supplemental EA provides
additional information on the effects of steam electric power plant discharges on surface waters and how
they may change under the regulatory options.
3.1 Waters Affected by Steam Electric Power Plant Discharges
EPA estimates the regulatory options potentially affect 112 steam electric power plants.22 EPA used the
United States Geological Survey (USGS) medium-resolution National Hydrography Dataset (NHD)
(USGS, 2018) to represent and identify waters affected by steam electric power plant discharges, and
used additional attributes provided in version 2 of the NHDPlus dataset (U.S. EPA, 2019f) to characterize
these waters.
Of the total 112 plants represented in the analysis, EPA estimated that 102 have non-zero pollutant
discharges under the baseline or the regulatory options. In the aggregate, these 102 plants with modeled
bottom ash transport water or FGD wastewater discharge to 108 waterbodies (as categorized in
NHDPlus), including lakes, rivers, and estuaries.23 NHDPlus also provides the Strahler Stream Order24
for each reach, where the order increases as one moves from headwaters (order 1) to downstream
segments (orders 2-9). Table 3-1 summarizes the Strahler Stream Order for the 108 reaches receiving
loadings from steam electric plants under the baseline or the regulatory options. Stream order is one of the
factors considered in evaluating potential uses of reaches (e.g., whether the reach is likely to be fishable),
when estimating benefits of water quality changes.
21 For more details about Option D, see the 2019 BCA (U.S. EPA, 2019a).
22 EPA analyzed a total of 112 plants that generate the wastestreams within the scope of the final rule. Not all 112 plants have
costs and/or loads under the baseline or regulatory options. For example, of the 112 plants analyzed, 108 plants are
estimated to incur technology implementation costs under the baseline and 75 plants are estimated to incur technology
implementation costs under Option A (see the Supplemental TDD for details [U.S. EPA, 2020g]). The modeling scope is all
112 plants, but as discussed in this section, some plants have zero loads whereas others discharge to waters that lack a valid
flow path (e.g., Great Lakes and estuaries).
23 One plant discharges waste streams to two different receiving waters and one reach receives discharges from two separate
plants.
24 Strahler Stream Order is a numerical measure of stream branching complexity. First order streams are the origin or
headwaters of a flowline. The confluence of two first order streams forms a second order stream, the confluence of two
second order streams forms a third order stream, and so on.
3-1

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-1: Strahler Stream Order Designation for Reaches Receiving
Steam Electric Power Plant Discharges
Stream Order
Number of Reaches
1
15
2
10
3
6
4
8
5
9
6
17
7
14
8
20
9
3
Not classified3
6
Total
108
a. Receiving reaches without a valid stream order include four reaches in the Great Lakes, one
reach in Hillsborough Bay, and one reach in Washington state.
Receiving reaches that lack NHD classification for both waterbody area type and stream order generally
correspond to reaches that do not have valid flow paths25 for analysis of the fate and transport of steam
electric power plant discharges (see Section 3.3). While six steam electric power plants discharge bottom
ash transport water and/or FGD wastewater to tidal reaches or the Great Lakes,26 EPA did not assess
pollutant loadings and water quality changes associated with these waterbodies because of the lack of a
defined flow path in NHDPlus, the complexity of flow patterns, and the relatively small changes in
concentrations expected.27 EPA did not quantify the water quality changes and resulting benefits (or
forgone benefits) to these systems. Thus, the total number of plants for which EPA estimated downstream
water quality changes is 96 (102 plants with nonzero pollutant discharges minus six plants discharging to
the Great Lakes or tidal waterbodies).
3.2 Changes in Pollutant Loadings
EPA estimated post-technology implementation pollutant loadings for each plant under the baseline and
the regulatory options. The Supplemental TDD details the methodology (U.S. EPA, 2020g). The sections
below discuss the approach EPA used to develop a profile of loading changes over time under the
baseline and each regulatory option and summarize the results.
3.2.1 Implementation Timing
Benefits analyses account for the temporal profile of environmental changes as the public values changes
occurring in the future less than those that are more immediate (OMB, 2003). As described in the final
rule, the regulatory options incorporate varying technology implementation deadlines for meeting the
revised limits depending on the wastestream and technology basis, including providing more time to
25 In NHDPlus, the flow path represents the distance traveled as one moves downstream from the reach to the terminus of the
stream network. An invalid flow path suggests that a reach is disconnected from the stream network.
20 Four reaches, one of which receives non-zero discharges from two steam electric power plants, are located in the Great
Lakes (three reaches along or near Lake Michigan and one reach along Lake Erie). One additional reach is located in
Hillsborough Bay and is influenced by tidal processes.
27 EPA looked at the changes in pollutant loadings and impacts to these systems in selected case studies as part of the analysis
of the 2015 rule (see 2015 EA for details; U.S. EPA, 2015b).
3-2

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
plants that participate in the VIP under Options A and B to meet more stringent FGD wastewater effluent
limits.
Table 3-2 summarizes the estimated technology implementation schedules for the baseline and the
regulatory options. This implementation schedule means that plants may be installing wastewater
treatment technologies in different years across the industry and potentially even within a given plant
(e.g., complying with bottom ash transport water requirements in 2021 and FGD wastewater requirements
in 2028). This in turn can translate into variations in pollutant loads to waters overtime.
To estimate the benefits of the regulatory options, EPA first developed a time profile of loadings for each
scenario (i.e.. baseline and each regulatory option), electricity generating unit (EGU), wastestream, and
pollutant that reflects the current loadings, the estimated loadings under the applicable technology basis,
the estimated technology implementation year for the plant, and the timing of any retirements or
repowerings. Specifically, EPA used current loadings starting in 2021 through the applicable technology
implementation year, technology-based loadings for all years following the implementation year, and zero
loadings following a unit's retirement or repowering.
EPA then used this year-explicit time profile to calculate the annual average loadings discharged by each
plant for two distinct periods within the overall period of analysis of 2021 through 2047:
•	Period 1, which extends from 2021 through 2028, when the universe of plants would transition
from current treatment practices to practices that achieve the revised limits, and
•	Period 2, which extends from 2029 through 2047 and is the post-transition period during which
the full universe of plants is projected to employ treatment practices that achieve the revised
limits.
The analysis accounts for each plant's technology implementation year(s) and for announced unit
retirements or repowerings. Using average annual values for two distinct periods instead of a single
average over the entire period of analysis improves the representation of rule implementation and enables
EPA to better capture the transitional effects of the regulatory options, including the temporary increases
in loadings relative to the 2015 rule baseline due to an extended status quo from delayed implementation
of new requirements. While using an annual average does not show the differences between the baseline
and regulatory options for individual years within Period 1, EPA considers that the average provides a
reasonable measure of the transitional effects of the regulatory options given the categories of benefits
that EPA is analyzing, which generally result from changes in multi-year processes.
As discussed in the RIA (U.S. EPA, 2020d), there is uncertainty in the exact timing of when individual
steam electric power plants would be implementing technologies to meet the final rule or the other
regulatory options. This benefits analysis uses the same plant- and wastestream-specific technology
installation years used in the cost and economic impact analyses. To the extent that technologies are
implemented earlier or later, the annualized loading values presented in this section may under- or
overstate the annual loads during the analysis period.
3-3

-------
BCAfor Revisions to Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-2: Implementation Schedule by Wastestream and Regulatory Option

Bottom Ash Transport Water
FGD Wastewater
Year(s)
Baseline
Option D
Option A
Option B
Option C
Baseline
Option D
Option A
Option B
Option C
2020
Current
Current
Current
Current
Current
Current
Current
Current
Current
Current
2021
Transition
Transition
Transition
Transition
Transition
Transition
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Current
2022
Transition
Transition
Transition
Transition
Transition
Transition
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Current
2023
Transition
Transition
Transition
Transition
Transition
Transition
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Transition (non-
VIP plants)
Current
2024
Full
Full
Transition3
Transition3
Transition3
Full
Transition (non-
Transition (non-
Transition (non-
Transition3

Implementation
Implementation



Implementation
VIP plants)3
VIP plants)3
VIP plants)3

2025
Full
Full
Transition3
Transition3
Transition3
Full
Transition (non-
Transition (non-
Transition (non-
Transition3

Implementation
Implementation



Implementation
VIP plants)3
VIP plants)3
VIP plants)3

2026
Full
Full
Full
Full
Full
Full
Interim Loads
Interim Loads
Interim Loads
Transition

Implementation
Implementation
Implementation
Implementation
Implementation
Implementation




2027
Full
Full
Full
Full
Full
Full
Interim Loads
Interim Loads
Interim Loads
Transition

Implementation
Implementation
Implementation
Implementation
Implementation
Implementation




2028
Full
Full
Full
Full
Full
Full
Transition (VIP
Transition (VIP
Transition (VIP
Transition

Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
plants)
plants)
plants)

2029-
Full
Full
Full
Full
Full
Full
Full
Full
Full
Full
2047
Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
Implementation
a. Indirect dischargers must meet the revised PSES limits by the end of 2023.
Current = Current loadings
Transition = Some plants meet the revised limits, based on their permitting schedule (see Section 3.1.3 in the RIA (U.S. EPA, 2020d) for details on the modeled plant-specific technology
implementation schedule). Aggregate loadings are lower than under Current conditions but greater than under the Full Implementation conditions.
Interim Loads = Non-VIP plants have reached the steady-state post-technology implementation loadings, but loadings for VIP plants are still at the Current level.
Full Implementation = All plants meet revised limits. Loadings are at their lowest steady-state post-technology implementation level.
Source: U.S. EPA Analysis, 2020
3-4

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
3.2.2 Results
Differences in the stringency of effluent limits and pretreatment standards and the timing of their applicability
to steam electric power plants (and the resulting treatment technology implementation) mean that changes in
pollutant loads between the regulatory options and the baseline vary over the period of analysis. Within the
period of analysis, the years 2021-2028 represent a period of transition as plants implement treatment
technologies to meet the revised limits under the baseline and regulatory options, whereas years 2029 through
2047 have steady state loadings that reflect implementation of technologies across all plants.28
Table 3-3 summarizes the average annual changes during Period 1 and Period 2 in FGD wastewater, bottom
ash transport water, and total loads for selected pollutants that inform EPA's analysis of the benefits discussed
in Chapters 4 through 7 and in Chapter 10. Negative values in the table indicate reductions in pollutant
loadings under an option as compared to the baseline, whereas positive values indicate increases in pollutant
loadings.
As shown in the table, total annual average pollutant loads increase under Options A and B across all
pollutants during Period 1, whereas Option C is estimated to result in net reductions of total bromide, iodine,
phosphorus and thallium loads during that same period, but net increases in other pollutants.
Under Options A, loadings of the pollutants in FGD wastewater generally decline in Period 2 as a result of
plants participating in the VIP program, but bottom ash loadings increase. Under Option A, there are net
estimated reductions in bromide, cadmium, iodine, and nickel loadings.
While not apparent from the total values, the direction of the changes for a particular pollutant is not
necessarily uniform across all plants under a given option. For example, plants that participate in the VIP
program under Options A and B may see reduced pollutant loadings in their FGD wastewater when compared
to the baseline, whereas pollutant loads may increase for non-VIP plants implementing chemical precipitation
with LRTR biological treatment control technologies as compared to baseline. Additionally, while Option C
reduces total bromide loads through treatment of FGD wastewater, plants with only bottom ash transport
water discharges may discharge greater loads under Option C compared to baseline. These differences will
have varying impacts on benefit estimates depending on the location of the plants and their proximity to
sensitive populations or environmental receptors.
28 This steady state reflects unit retirements and repowerings. EPA accounted for unit retirements and repowerings by zeroing out
the loadings starting in the year following the change in status.
3-5

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-3: Annual Average Changes in Total Pollutant Loading in Period 1 (2021-2028) and Period 2
(2029-2047) for Selected Pollutants in Steam Electric Power Plant Discharges, Compared to Baseline
(lb/year)
Pollutant
Option Aa
(Final Rule)
Option Ba
Option C
FGD
Bottom
Ash
Totalb
FGD
Bottom
Ash
Totalb
FGD
Bottom
Ash
Totalb
Period 1 (2021-2028)
Antimony
18
465
483
18
211
229
27
211
238
Arsenic
4
250
254
4
113
117
-69
113
44
Barium
346
2,851
3,197
346
1,293
1,639
-412
1,293
880
Beryllium
1
C
1
1
C
1
-14
C
-14
Boron
38,100
142,300
180,400
38,100
64,600
102,600
-3,306,600
64,600
-3,242,000
Bromide
36
137,000
137,000
36
62,000
62,000
-5,823,000
62,000
-5,761,000
Cadmium
230
19
249
230
9
239
1,085
9
1,094
Chromium
24
136
161
24
62
86
25
62
87
Copper
38
106
144
38
48
86
134
48
182
Cyanide
85
C
85
85
C
85
-13,840
C
-13,840
Iodine
3
C
3
3
C
3
-167,920
C
-167,920
Lead
3
279
282
3
127
129
-37
127
89
Manganese
171,400
4,100
175,500
171,400
1,859
173,300
664,200
1,859
666,100
Mercury
18
3
20
17
1
18
83
1
84
Nickel
1,887
468
2,355
1,854
212
2,067
9,182
212
9,394
TN
1,546,000
71,000
1,616,000
423,000
32,000
455,000
2,674,000
32,000
2,707,000
TP
2
5,945
5,947
2
2,693
2,696
-4,950
2,693
-2,260
Selenium
15,440
328
15,770
4,576
149
4,725
29,290
149
29,440
Thallium
8
30
39
8
14
22
-107
14
-94
TSS
38,530
358,190
396,720
38,530
162,380
200,910
100,220
162,380
262,600
Zinc
2,912
906
3,818
2,912
411
3,323
14,240
411
14,650
Period 2 (2029-2047)
Antimony
-9
332
323
-10
198
188
-261
198
-64
Arsenic
-8
179
170
-9
106
97
-354
106
-248
Barium
-250
2,039
1,789
-271
1,212
941
-8,580
1,212
-7,370
Beryllium
-2
C
-2
-2
C
-2
-81
C
-81
Boron
-293,800
101,800
-192,100
-327,800
60,500
-267,200
-13,727,800
60,500
-13,667,200
Bromide
-2,951,000
98,000
-2,853,000
-2,954,000
58,000
-2,896,000
-23,828,000
58,000
-23,770,000
Cadmium
-53
14
-39
-54
8
-45
-302
8
-294
Chromium
-13
97
84
-14
58
44
-395
58
-337
Copper
-13
76
63
-13
45
32
-236
45
-191
Cyanide
85
C
85
85
C
85
-13,840
C
-13,840
Iodine
-116,540
C
-116,540
-116,760
C
-116,760
-684,730
C
-684,730
Lead
-5
200
195
-5
119
113
-206
119
-87
Manganese
-51,400
2,933
-48,400
-53,300
1,744
-51,500
-794,700
1,744
-793,000
Mercury
-2
2
0
-3
1
-2
-6
1
-5
Nickel
-352
335
-17
-391
199
-192
-765
199
-566
TN
1,246,000
51,000
1,297,000
-52,000
30,000
-22,000
-497,000
30,000
-467,000
TP
-406
4,250
3,844
-454
2,526
2,071
-19,480
2,526
-16,960
Selenium
12,030
235
12,270
-518
140
-378
-858
140
-718
Thallium
-14
22
8
-16
13
-3
-596
13
-584
TSS
-5,880
256,270
250,380
-5,880
152,200
146,320
-397,370
152,200
-245,170
Zinc
-627
648
20
-630
385
-245
-1,810
385
-1,420
3-6

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-3: Annual Average Changes in Total Pollutant Loading in Period 1 (2021-2028) and Period 2
(2029-2047) for Selected Pollutants in Steam Electric Power Plant Discharges, Compared to Baseline
(lb/year)
Pollutant
Option Aa
(Final Rule)
Option Ba
Option C
FGD
Bottom
Ash
Total"
FGD
Bottom
Ash
Total"
FGD
Bottom
Ash
Total"
a. Negative values represent a reduction in pollutant loadings as compared to the baseline.
b. FGD and bottom ash loadings may not add up to the total due to independent rounding.
c. EPA did not estimate changes in beryllium, cyanide, and iodine loadings associated with bottom ash transport water.
TN = Nitrogen, total (as N); TP = Phosphorus, total (as P); TSS = Total suspended solids
Source: U.S. EPA Analysis, 2020.
3.3 Water Quality Downstream from Steam Electric Power Plants
EPA used the estimated annual average changes in total pollutant loadings for Periods 1 and 2 to estimate
concentrations downstream from each plant. The methodology uses two main models to estimate downstream
concentrations from each plant for each period:
• A dilution model to estimate pollutant concentrations downstream from the plants. The approach,
which for the purpose of this analysis is referred to as the D-FATE model (Downstream Fate And
Transport Equations), involves calculating concentrations in each downstream medium-resolution
NHD reach using annual average Enhanced Runoff Method (EROM) flows from NHDPlus v2 and
mass conservation principles.
• USGS's SPAtially Referenced Regressions On Watershed attributes (SPARROW) to estimate flow-
weighted nutrient (TN and TP) and suspended sediment concentrations. The SPARROW models
provide baseline and regulatory option concentrations of TN, TP, and suspended solids concentration
(SSC). For this analysis, EPA used the most recent calibrated regional models published by the USGS
(Ator, 2019; Hoos & Roland Ii, 2019; Robertson & Saad, 2019; Wise, 2019; Wise etal., 2019). These
models define the stream network using the same medium-resolution NHD reaches used in D-FATE.
The models represent only discharges to reaches represented in the NHD, which include the vast majority of
plants within the scope of the rule (106 plants out of 112 plants within the scope of the rule). As discussed in
Section 3.1, EPA omitted six steam electric power plants that discharge to the Great Lakes or to estuaries
from this analysis.
In the D-FATE model, EPA used stream routing and flow attribute information from the medium-resolution
NHDPlus v2 to track masses of pollutants from steam electric power plant discharges and other pollutant
sources as they travel through the hydrographic network. For each point source discharger, the D-FATE
model estimates pollutant concentrations for the receiving reach and all downstream reaches based on NHD
mean annual flows. In-stream flows are kept constant (i.e.. discharges have no effect on flows). EPA notes
that steam electric power plant discharges frequently constitute a return of flow withdrawn for plant use from
3-7

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
the same surface water. In addition, FGD and BA wastewater discharges generally comprise a very small
fraction of annual mean flows in the NHDPlus v2 dataset.29
Following the approach used in the analysis of the 2015 rule and 2019 proposal (U.S. EPA, 2015a, 2019a) to
estimate pollutant concentrations, EPA included loadings from major dischargers (in addition to the steam
electric power plants) that reported to the 2016 Toxics Release Inventory (TRI).30 TRI data were available for
a subset of toxics: arsenic, barium, chromium, copper, lead, manganese, mercury, nickel, selenium, thallium,
and zinc. EPA summed reach-specific concentrations from TRI dischargers and concentration estimates
resulting from steam electric power plant loadings to represent water quality impacts from multiple sources.
The pollutant concentrations calculated in the D-FATE model are used to derive fish tissue concentrations
used to analyze human health effects from consuming self-caught fish (see Chapter 5), analyze nonmarket
benefits of water quality improvements (see Chapter 6), and assess potential impacts to T&E species whose
habitat ranges intersect with waters affected by steam electric plant discharges (see Chapter 7).
3.4 Overall Water Quality Changes
Following the approach used in the analysis of the 2015 rule and 2019 proposal (U.S. EPA, 2015a, 2019a),
EPA used a WQI to link water quality changes from reduced toxics, nutrient and sediment discharges to
effects on human uses and support for aquatic and terrestrial species habitat. The WQI translates water quality
measurements, gathered for multiple parameters (e.g., dissolved oxygen [DO], nutrients) that are indicative of
various aspects of water quality, into a single numerical indicator. The WQI ranges from 0 to 100 with low
values indicating poor quality and high values indicating good water quality.
As detailed in U.S. EPA (2015a), the WQI includes seven parameters: DO, BOD, fecal coliform (FC), TN,
TP, suspended solids, and one aggregate sub index for toxics. The pollutants considered in the aggregate
subindex for toxics are those that are discharged by modeled steam electric power plants or 2016 TRI
dischargers and that have chronic aquatic life-based NRWQC. Following the approach used for the 2019
proposal analysis, pollutants that meet these qualifications include arsenic, hexavalent chromium, copper,
lead, manganese, mercury, nickel, selenium, and zinc. See the Supplemental EA for details on NRWQC (U.S.
EPA, 2020f). The subindex curve for toxics assigns the lowest WQI value of 0 to waters where exceedances
are observed for the nine toxics analyzed, and a maximum WQI value of 100 to waters where there are no
exceedances. Intermediate values are distributed between 100 and 0 in proportion to the number of
exceedances.
3.4.1 WQI Data Sources
To calculate the WQI, EPA used modeled NRWQC exceedances for toxics (using concentrations from D-
FATE) and modeled concentrations for TN, TP, and SSC from the respective SPARROW regional models.
Following the approach used for the 2019 proposal, the USGS National Water Information System (NWIS)
provided concentration data from 2007-2017 for three parameters that are held constant between the baseline
29 Steam electric power plant FGD discharge rates are typically approximately 1 million gallons per day (MGD), whereas the
annual mean stream flows in receiving waters average approximately 15,000 MGD.
30 According to EPA TRI National Analysis, TRI releases to water reported in 2018 were approximately 1 percent higher, in the
aggregate, than releases reported in 2016 (195.0 million pounds versus. 192.3 million pounds). See
https://www.epa.gov/trinationalanalvsis/water-releases for details.
3-8

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
and regulatory options: 1) fecal coliform, 2) dissolved oxygen, and 3) biochemical oxygen demand (see
Section 3.4.1.2).31
3.4.1.1 Exceedances of Water Quality Standards and Criteria
For each regulatory option, EPA identified reaches that do not meet NRWQC for aquatic life in Periods 1 and
2.32 Table 3-4 summarizes the number of reaches with estimated exceedances of NRWQC in the baseline and
under the regulatory options.
Table 3-4: Estimated Exceedances of National Recommended
Water Quality Criteria under the Baseline and Regulatory Options
Regulatory Option
Number of Reaches with at Least One
NRWQC Exceedance
Chronic
Acute
Period 1 (2021-2028)
Baseline
19
4
Option A (Final Rule)
22
4
Option B
22
4
Option C
23
4
Period 2 (2029-2047)
Baseline
3
3
Option A (Final Rule)
0
0
Option B
0
0
Option C
0
0
Source: U.S. EPA Analysis, 2020
Refer to the Supplemental EA for additional discussion of comparisons of receiving and downstream water
pollutant concentrations to acute and chronic aquatic NRWQC (U.S. EPA, 2020f).
3.4.1.2 Sources for Ambient Water Quality Data
Following the approach used for the 2019 proposal analysis, EPA used average monitoring values for fecal
coliform, dissolved oxygen, and biochemical oxygen demand for 2007-2017 where available. Where more
recent data were not available, EPA used the same averages as for the 2015 rule analysis. EPA used a
successive average approach to assign average values for the three WQI parameters not explicitly modeled
(i.e.. DO, BOD, fecal coliform). The approach, which adapts a common sequential averaging imputation
technique, involves assigning the average of ambient concentrations for a given parameter within a hydrologic
unit to reaches within the same hydrologic unit with missing data, and progressively expanding the
31 USGS's NWIS provides information on the occurrence, quantity, quality, distribution, and movement of surface and underground
waters based on data collected at approximately 1.5 million sites in all 50 States, the District of Columbia, and U.S. territories.
More information on NWIS can be found at http://waterdata.usgs.gov/nwis/.
32 Aquatic life criteria are the highest concentration of pollutants in water that are not expected to pose a significant risk to the
majority of species in a given environment. For most pollutants, aquatic NRWQC are more stringent than human health NRWQC
and thus provide a more conservative estimate of potential water quality impairment. Chronic criteria are derived using longer
term (7-day to greater than 28-day) toxicity tests if available, or an acute-to-chronic ratio procedure where the acute criteria is
derived using short term (48-hour to 96-hour) toxicity tests (U.S. EPA, 2017a). More information on aquatic NRWQC can be
found at https://www.epa.gov/wac/national-recornmended-water-amlitv-criteria-aauatic-life-criteria-table and in the
Supplemental EA (U.S. EPA, 2020f).
3-9

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
geographical scope of the hydrologic unit (Hydrologic unit code (HUC)8, HUC6, HUC4, and HUC2) to fill in
all missing data.33 This approach is based on the assumption that reaches located in the same watershed
generally share similar characteristics. Using this estimation approach, EPA compiled ambient water quality
data and/or estimates for all analyzed NHD reaches. As discussed below, the values of the three WQI
parameters not explicitly modeled are kept constant for the baseline and regulatory policy scenarios. This
approach has not been peer reviewed, but it has been used by EPA for several prior rules and reviewed by the
public during the associated comment periods.
The water quality analysis included a total of 16,169 medium-resolution NHD reaches that are potentially
affected by steam electric power plants under the baseline. Of these 16,169 NHD reaches, EPA estimated
concentrations for 15,159 reaches affected by non-zero loadings from steam electric power plants. Table 3-5
summarizes the data sources used to estimate baseline and regulatory option values by water quality
parameter.
Table 3-5: Water Quality Data used in Calculating WQI for the Baseline and Regulatory
Options

Parameter
Baseline
Regulatory Option
TN
Concentrations calculated using
Concentrations calculated using

SPARROW (baseline run)
SPARROW (regulatory option run)
TP
Concentrations calculated using
Concentrations calculated using

SPARROW (baseline run)
SPARROW (regulatory option run)
Suspended
Concentrations calculated using
Concentrations calculated using
sediment
SPARROW (baseline run)
SPARROW (regulatory option run)
DO
Observed values averaged at the WBD
No change. Regulatory option value set

watershed level
equal to baseline value
BOD
Observed values averaged at the WBD
No change. Regulatory option value set

watershed level
equal to baseline value
Fecal Coliform
Observed values averaged at the WBD
No change. Regulatory option value set

watershed level
equal to baseline value
Toxics
Baseline exceedances calculated using
Regulatory option exceedances

D-FATE model
calculated using D-FATE model
WBD = Watershed Boundary Dataset. The WBD is a companion dataset to the NHD
Source: U.S. EPA Analysis, 2020.
3.4.2 WQI Calculation
EPA used the approach described in the BCA for the 2015 rule and 2019 proposal (U.S. EPA, 2015a, 2019a)
to estimate WQI values for each reach under the baseline and each option, and used revised subindex curves
for TN, TP, and SSC34 that reflect data from the most current SPARROW regional models (Ator, 2019; Hoos
33 Hydrologic Unit Codes (HUCs) are cataloguing numbers that uniquely identify hydrologic features such as surface drainage
basins. The HUCs consist of 8 to 14 digits, with each set of 2 digits giving more specific information about the hydrologic
feature. The first pair of values designate the region (of which there are 21), the next pair the subregion (total of 222), the third
pair the basin or cataloguing unit (total of 352), and the fourth pair the subbasin, or accounting unit (total of 2,262) (U.S.
Geological Survey, 2007). Digits after the first eight offer more detailed information, but are not always available for all waters.
In this discussion, a HUC level refers to a set of waters that have that number of HUC digits in common. For example, the HUC6
level includes all reaches for which the first six digits of their HUC are the same.
34 The 2015 WQI includes a subindex for TSS. For this analysis, EPA developed a curve for SSC based on more recent SPARROW
regional models which estimates SSC rather than TSS concentrations (Ator, 2019; Hoos & Roland Ii, 2019; Robertson & Saad,
2019; Wise, 2019; Wise etal., 2019). This bypasses translation of SSC to TSS values and any associated uncertainty.
3-10

-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
& Roland Ii, 2019; Robertson & Saad, 2019; Wise, 2019; Wise etal., 2019). Implementing the WQI
methodology involves three key steps: 1) obtaining water quality levels for each of seven parameters included
in the WQI; 2) transforming parameter levels to subindex values expressed on a common scale; and 3)
aggregating the individual parameter subindices to obtain an overall WQI value that reflects waterbody
conditions across the seven parameters. These steps are repeated to calculate the WQI value for the baseline,
and for each analyzed regulatory option. See details of the calculations in Appendix B, including the subindex
curves used to transform levels of individual parameters. The scope of this analysis is the same as that for the
analysis of nonmarket benefits of water quality improvements discussed in Chapter 6, which focuses on
reaches within 300 km of a steam electric plant outfall.35
3.4.3 Baseline WQI
The WQI value can be related to suitability for potential uses. Vaughan (1986) developed a water quality
ladder (WQL) that can be used to indicate whether water quality is suitable for various human uses (i.e..
boating, rough fishing, game fishing, swimming, and drinking without treatment). Vaughan identified
"minimally acceptable parameter concentration levels" for each of the five potential uses. Vaughan used a
scale of zero to 10 instead of the WQI scale of zero to 100 to classify water quality based on its suitability for
potential uses. Therefore, the WQI value corresponding to a given water quality use classification equals the
WQL value multiplied by 10.
Based on the estimated WQI value under the baseline scenario (WQI-BL), EPA categorized each of the
10,454 NHD reaches using five WQI ranges (WQI < 25, 25
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-6: Estimated Percentage of Potentially Affected Reach Miles by WQI Classification:
Baseline Scenario
Water Quality Classification
Baseline WQ
Number of
Reaches
Percent of
Affected
Reaches
Number of
Reach Miles
Percent of
Affected
Reach Miles
Suitable for Game Fishing
50
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
the NHDPlus v2 documentation (U.S. EPA, 2019f). Regarding the uncertainties associated with estimated
loads, see the Supplemental. TDD (U.S. EPA, 2020g).
Table 3-8: Limitations and Uncertainties in Estimating Water Quality Effects of Regulatory Options
Uncertainty/Limitation
Effect on Water
Quality Effects
Estimation
Notes
Limited data are available to validate
water quality concentrations
estimated in D-FATE
Uncertain
The modeled concentrations reflect only a subset of
pollutant sources [e.g., steam electric power plant
discharges and TRI releases) whereas monitoring data
also reflect other sources such as bottom sediments,
air deposition, and other point and non-point sources
of pollution. EPA comparisons of D-FATE estimates to
monitoring data available for selected locations and
parameters [e.g., bromide concentrations downstream
of steam electric power plant discharges) confirmed
that D-FATE provides reasonable values. Also refer to
the 2015 EA for discussion of model validation for
selected case studies (U.S. EPA, 2015b)
Steam electric power plant
discharges have no effects on reach
annual average flows
Overestimate
The degree of overestimation in the estimation of
pollutant concentrations, if any, would be small given
that steam electric power plant discharge flows tend to
be very small as compared to flows in modeled
receiving and downstream reaches.
Ambient water toxics concentrations
are based only on loadings from
steam electric power plants and
other TRI discharges.
Uncertain
Concentration estimates do not account for
background concentrations of these pollutants from
other sources, such as legacy pollution in sediments,
non-point sources, point sources that are not required
to report to TRI, air deposition, etc. Not including other
contributors to background toxics concentrations in
the analysis is likely to result in understatement of
baseline concentrations of these pollutants and
therefore of NRWQC exceedances. The effect on WQI
calculations is uncertain.
Annual loadings are estimated based
on EPA's estimated plant-specific
technology implementation years
Uncertain
To the extent that technologies are implemented
earlier or later, the Period 1 annualized loading values
presented in this section may under- or overstate the
annual loads during the analysis period. The effect of
this uncertainty is limited to Period 1 since loads reach
a steady-state level by the technology implementation
deadlines applicable to the regulatory options (e.g., by
the end of 2028)
Changes in WQI reflect only
reductions in toxics, nutrient, and
suspended solids concentrations.
Underestimate
The estimated changes in WQI reflect only water
quality changes resulting directly from changes in
toxics, nutrient and sediment concentrations. They do
not include changes in other water quality parameters
(e.g., BOD, dissolved oxygen) that are part of the WQI
and for which EPA used constant values. Because the
omitted water quality parameters are also likely to
respond to changes in pollutant loads (e.g., dissolved
oxygen levels respond to changes in nutrient levels),
the analysis underestimates the water quality changes.
3-13
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
3: Water Quality Effects
Table 3-8: Limitations and Uncertainties in Estimating Water Quality Effects of Regulatory Options
Uncertainty/Limitation
Effect on Water
Quality Effects
Estimation
Notes
EPA used regional averages of
monitoring data from 2007-2017 for
fecal coliform, dissolved oxygen, and
biochemical oxygen demand, when
location-specific data were not
available. In cases where more
recent data were not available, EPA
used the same averages as used in
the 2015 rule analysis (U.S. EPA,
2015a).
Uncertain
The monitoring values were averaged over
progressively larger hydrologic units to fill in any
missing data. As a result, WQI values may not reflect
certain constituent fluctuations resulting from the
various regulatory options and/or may be limited in
their temporal and spatial relevance. Note that the
analysis keeps these parameters constant under both
the baseline and regulatory options. Modeled changes
due to the regulatory options are not affected by this
uncertainty.
Use of nonlinear subindex curves
Uncertain
The methodology used to translate suspended solids
and nutrient concentrations into subindex scores (see
Section 3.4.2 and Appendix B) employs nonlinear
transformation curves. Water quality changes that fall
outside of the sensitive part of the transformation
curve (i.e., above/below the upper/lower bounds,
respectively) yield no change in the analysis and no
benefits in the analysis described in Chapter 6.
3-14
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
4 Human Health Benefits from Changes in Pollutant Exposure via the
Drinking Water Pathway
As described in Section 2.1, human health benefits deriving from changes in pollutant loadings to receiving
waters include those associated with changes in exposure to pollutants via treated drinking water use and fish
consumption. This chapter addresses the first exposure pathway: drinking water. Chapter 5 addresses the fish
consumption pathway.
The small changes in pollutant loadings from the regulatory options relative to the 2015 analysis (U.S. EPA,
2015b) could affect human health by changing halogen and other pollutant discharges to surface waters and,
as a result, pollutant concentrations in the reaches that serve as sources of drinking water. The Supplemental
EA presents background information regarding the potential impacts of halogen discharges on drinking water
quality and human health (U.S. EPA, 2020f). Section 4.1 presents EPA's analysis of the modeled changes in
halogen concentrations in public drinking water systems' source waters. Section 4.2 summarizes potential
impacts on source waters from changes in other pollutant discharges. Section 4.3 discusses uncertainty and
limitations associated with the analysis presented in this chapter.
In general, EPA estimated small impacts on source waters under the final rule relative to the baseline,
compared to those estimated in the 2015 rule (see U.S. EPA, 2015a).
4.1 Estimates of Changes in Halogen Concentrations in Source Water
For the final rule, EPA estimated the change in halogen levels in the source water for PWS that have intakes
downstream from steam electric power plants.37 Halogens such as bromide and iodine are precursors for
halogenated disinfection byproduct formation in treated drinking water, including certain trihalomethanes
addressed by the TTHM MCL. Higher halogen levels in PWS source waters have been associated with higher
levels of halogenated DBPs in treated drinking water. The formation of DBPs varies with site-specific factors.
In vitro toxicology studies with bacteria and mammalian cells have documented evidence of genotoxic
(including mutagenic), cytotoxic, tumorigenic, and developmental toxicity properties of iodinated DBPs, but
the available data are insufficient at this time to determine the extent of iodinated DBP's contribution to
adverse human health effects from exposure to treated drinking water. Populations exposed to changes in
halogenated disinfection byproduct levels in their drinking water under the regulatory options could
experience changes in the incidence of adverse health effects. For additional information on these issues, see
the Supplemental EA (U.S. EPA, 2020f).
In this section, the Agency presents the number of PWS with modeled changes in bromide and iodine
concentration in their source water, the magnitude and direction of these changes, and the PWS service
population estimated to experience a change in DBP exposure levels due to changes in source water bromide
and iodine levels.
4.1.1 Bromide Bromide and Iodine Concentrations in Surface Water
As described in the Supplemental TDD (U.S. EPA, 2020g), EPA estimated steam electric power plant-level
bromide and iodine loadings associated with bottom ash transport water and FGD wastewater for the baseline
and the regulatory options. Total plant loadings are calculated as the sum of bottom ash transport water and
37 These analyses correspond to steps 1 and 2 of the methodology EPA used for the 2019 proposal (see Chapter 4 in U.S. EPA,
2019a)
4-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
FGD wastewater loadings under each scenario. Data on iodine is more limited and loading estimates are
available for FGD wastewater. See Section 6 of the Supplemental TDD for a discussion of the uncertainties
associated with iodine data generally and the resulting uncertainties propagated within this analysis. This
chapter presents EPA's best estimate of changes in bromide and iodine loadings under each of the regulatory
options.
EPA used the D-FATE model described in Section 3.3 to estimate in-stream bromide concentrations
downstream from 102 steam electric power plants that EPA estimated have non-zero bromide loads (i.e..
discharge FGD wastewater and/or bottom ash transport water) under the baseline or regulatory options. EPA
used the same approach to estimate in-stream iodine concentrations downstream from the subset of 61 plants
that EPA estimated have non-zero iodine loads (i.e.. plants discharging FGD wastewater). EPA first estimated
the annual average bromide and iodine loads in Period 1 and Period 2 (see Section 3.2.1). EPA then estimated
concentrations in the receiving reach and each downstream reach, using conservation of mass principles, until
the load reaches the network terminus (e.g., Great Lake, estuary).38 EPA summed individual contributions
from all plants to estimate total in-stream concentrations under the baseline and the regulatory options.
Finally, EPA estimated the change in bromide and iodine concentrations in each reach as the difference
between each regulatory option and the baseline. This change is not dependent on bromide or iodine
contributions from other sources (e.g., receiving waterbody background levels).
The bromide and iodine loading estimates represent two independent scenarios that each assume the subset of
plants using coal additives (30 plants) rely exclusively on either bromide-based- or iodine-based additives,
respectively. The two scenarios would therefore not occur concurrently. For example, no plant using bromide
additives at the level used in developing the bromide loadings estimate would simultaneously use iodine
additives at the level used in developing the iodine loadings estimate (and vice versa). The two scenarios are
therefore best interpreted as bounding cases that represent maximum potential discharges of either
constituent. At this time, more coal-fired facilities use bromide-based rather than iodine-based additives
(Tinuum Group LLC, 2020). However, information on the additive type used by individual facilities in this
analysis is limited.
4.1.2 Changes in Bromide and Iodine Levels in Source Water
4.1.2.1 Affected Public Water Systems
For the final rule, EPA updated the universe of PWS potentially affected by steam electric plant discharges to
reflect adjustments to the universe of plants projected to be subject to the rule and their associated
downstream reaches. EPA also collected more recent information about the operating characteristics of the
water systems (e.g., population served, facility status, wholesale water purchases).
38 As discussed in Section 3.1, EPA did not estimate concentration changes in the Great Lakes or estuaries.
4-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
EPA's Safe Drinking Water Information System (SDWIS) database39 provides the latitude and longitude of
surface water facilities40, including source water intakes for public drinking water treatment systems. To
identify potentially affected PWS, the Agency georeferenced each permanent surface water facility associated
with non-transient community water systems to the NHD medium-resolution stream network used in D-
FATE.41 Appendix E describes the methodology EPA used to identify the NHD water feature for each facility.
The SDWIS database also includes information on PWS primary sources (e.g., whether a PWS relies
primarily on groundwater or surface water for their source water), operational status, and population served,
among other attributes. For this analysis, EPA used the subset of facilities that identify surface water as their
primary water source (specifically surface water intakes and reservoirs) and were categorized as "active" and
"permanent" in SDWIS. This subset of facilities corresponds to PWS that are more likely to be affected by
upstream bromide and/or iodine releases on an ongoing basis, as compared to other systems that may use
surface water sources only sporadically. This approach identifies populations most likely to experience
changes in long-term halogenated DBP exposures and associated health effects due to the regulatory options.
PWS can be either directly or indirectly affected by steam electric power plant discharges. Directly affected
PWS are systems with surface water intakes drawing directly from reaches downstream from steam electric
power plants discharging bromide or iodine.42 Other PWS are indirectly affected because they purchase their
source water from another PWS via a "consecutive connection" instead of withdrawing directly from a
surface water or groundwater source. For these systems, SDWIS provides information on the PWS that
supplies the purchased water. EPA used SDWIS data to identify PWS that may be indirectly affected by
steam electric power plant discharges because they purchase water from a directly affected PWS. The total
potentially exposed population consists of the people served by both directly and indirectly affected systems.
Table 4-1 summarizes the intakes, PWS, and populations potentially affected by steam electric power plant
discharges. Sixteen PWS may be directly and indirectly affected. In this analysis, the average distance from
the steam electric power plant discharge point to the drinking water treatment plant intake is approximately
286 miles and more than a quarter of the intakes are located within 50 miles of a steam electric power plant
outfall. A subset of these PWS are downstream of iodine discharges, specifically 208 reaches have intakes
used by 764 PWS serving a total of 25.8 million people.
39 EPA used intake locations as of January 2018 and PWS data as of April 2020, which reflects the first quarter report for 2020.
Intake location data are protected from disclosure due to security concerns. SDWIS public data records are available from the
Federal Reporting Services system at https://ofmpub.epa.gov/apex/sfdw/.
40 Surface water facilities include any part of a PWS that aids in obtaining, treating, and distributing drinking water. Facilities in the
SDWIS database may include groundwater wells, consecutive connections between buyer and seller PWS, pump stations,
reservoirs, and intakes, among others.
41 This analysis does not include intakes that draw from the Great Lakes or other water bodies not analyzed in the D-FATE model.
42 To identify potentially affected PWS, EPA looked at all downstream reaches starting from the immediate reach receiving the
steam electric power plant discharge to the reach identified as the terminus of the stream network.
4-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-1: Estimated Reaches, Surface Water Intakes, Public Water Systems, and Populations
Potentially Affected by Steam Electric Power Plant Discharges
PWS Impact Category
Number of Reaches
with Drinking Water
Intakes
Number of Intakes
Downstream of
Steam Electric Power
Plants
Number of PWS
Total Population
Served (Million
People)
Direct3
255
370
272
20.3
Indirect
Not applicable
Not applicable
677
11.3
Total
255
370
949
31.6
a. Includes 16 systems with intakes downstream of steam electric power plant discharges and that purchase water from other
systems with intakes downstream of steam electric power plant discharges.
Source: U.S. EPA analysis, 2020
4.1.2.2 System-Level Changes in Bromide and Iodine Concentrations in Source Water
EPA estimated the change in bromide and iodine concentrations in the source water for each PWS that could
result from the regulatory options. In this discussion, the term "system" refers to PWS and their associated
drinking water treatment operations, whereas the term "facility" refers to the intake that is drawing untreated
water from a source reach for treatment at the PWS level.
To estimate changes in bromide and iodine concentrations at the PWS level, EPA obtained the number of
active permanent surface water sources used by each PWS based on SDWIS data. SDWIS does not provide
information on respective source flow contributions from surface water and groundwater facilities for a given
PWS. For drinking water treatment systems that have both surface water and groundwater facilities, EPA
assessed changes from surface water sources only. This approach is reasonable given that the analysis is
limited to the PWS for which SDWIS identifies surface water as primary source.
For intakes located on reaches modeled in D-FATE, EPA calculated the reach-level change in bromide and
iodine concentration as the difference between the regulatory option and the baseline conditions. Some PWS
rely on a single intake facility for their source water supply. If the source water reach associated with this
single intake is affected by steam electric power plant bromide or iodine discharges, the system-level changes
in bromide or iodine concentration at the PWS would equal the estimated change in bromide or iodine
concentration of the source water reach. Other PWS rely on multiple intake facilities that may be located
along different source water reaches. System-level changes in bromide or iodine concentrations at these PWS
are an average of the estimated changes in bromide or iodine concentrations associated with each source
water reach. For any additional intakes not located on the modeled reaches and for intakes relying on
groundwater sources, EPA estimated zero change in bromide or iodine concentration. Because SDWIS does
not provide information on source flows contributed by intake facilities used by a given PWS, EPA calculated
the system-level change in bromide or iodine concentration assuming each active permanent source facility
contributes equally to the total volume of water treated by the PWS. For example, the PWS-level change in
bromide concentration for a PWS with three intakes, of which one intake is directly affected by steam electric
power plant discharges, is estimated as one third of the modeled reach concentration change ([ABr + 0 + 0]/3).
EPA addressed water purchases similarly, but with the change in bromide or iodine concentration associated
with the consecutive connection set equal to the PWS-level change estimated for the seller PWS instead of a
reach-level change. For facilities affected only indirectly by steam electric power plant discharges, EPA
assumed zero change in bromide and iodine concentrations for any other unaffected source facility associated
4-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
with the buyer. EPA also assumed that each permanent source facility contributes an equal share of the total
volume of water distributed by the buyer. For the 16 intakes classified as both directly and indirectly affected
by steam electric power plant bromide and iodine discharges, EPA assessed the total change in bromide or
iodine concentration as a blended average of the change in concentration from both directly-drawn and
purchased water.
Table 4-2 summarizes the distribution of changes in bromide concentrations under the regulatory options for
the two analysis periods. The direction of the changes depends on the Period, option, source water reach, and
PWS but is generally consistent with the changes in bromide loadings associated with FGD and bottom ash
transport wastewaters under each regulatory option (see Table 3-3). During Period 1, Options A and B show
either increases or no changes in bromide concentrations for all source waters and PWS and Option C shows
both increases and decreases in bromide concentrations across locations. During Period 2, all regulatory
options show both estimated increases and decreases in bromide concentrations with both the magnitude and
scope (the number of reaches, PWS, and population served) of the decreases larger than during Period 1.
Table 4-3 provides a similar summary of the distribution of changes in iodine concentrations under the
regulatory options. As was the case for bromide, the direction of the changes is generally consistent with the
changes in iodine loadings (see Table 3-3). However, because these changes arise only from FGD wastewater,
they are more uniform during Period 2 across the options. Thus, during Period 1, Options A and B show an
estimated increase in iodine concentrations and Option C shows both increases and decreases in iodine
concentrations. During Period 2, the three options show decreases in iodine concentrations.
4-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-2: Estimated Distribution of Changes in Source Water and PWS-Level Bromide Concentrations by Period and Regulatory Option,
Compared to Baseline
ABr Range
(Hg/L)
Number of Source Water Reaches
Number of PWSa
Population Served by
PWS
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Option iy

Not estimated
Not estimated
Not estimated

Not estimated
Not estimated
Not estimated
>75
0
0
0
0
0
0
Not estimated
Not estimated
Not estimated
Period 1
Option A (Final Rule)
Oto 10
245
0
10
894
0
55
30,510,519
0
1,102,458
10 to 30
0
0
0
0
0
0
0
0
0
30 to 50
0
0
0
0
0
0
0
0
0
50 to 75
0
0
0
0
0
0
0
0
0
>75
0
0
0
0
0
0
0
0
0
Option B
Oto 10
245
0
10
894
0
55
30,510,519
0
1,102,458
10 to 30
0
0
0
0
0
0
0
0
0
30 to 50
0
0
0
0
0
0
0
0
0
50 to 75
0
0
0
0
0
0
0
0
0
>75
0
0
0
0
0
0
0
0
0
Option C
Oto 10
67
157
3
249
600
13
6,796,937
21,787,841
202,550
10 to 30
0
19
0
0
61
0
0
2,140,443
0
30 to 50
0
8
0
0
15
0
0
286,635
0
50 to 75
0
1
0
0
11
0
0
398,571
0
>75
0
0
0
0
0
0
0
0
0
Period 2
Option A (Final Rule)
Oto 10
174
60
13
568
279
58
23,258,818
6,314,975
1,088,640
10 to 30
0
7
0
0
41
0
0
803,727
0
30 to 50
0
0
0
0
0
0
0
0
0
50 to 75
0
0
0
0
0
0
0
0
0
>75
0
1
0
0
3
0
0
146,817
0
4-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-2: Estimated Distribution of Changes in Source Water and PWS-Level Bromide Concentrations by Period and Regulatory Option,
Compared to Baseline
ABr Range
(Hg/L)
Number of Source Water Reaches
Number of PWSa
Population Served by
PWS
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Positive13 ABr
Negative13 ABr
No ABr
(ABr = 0)
Option B
Oto 10
167
67
13
549
298
58
18,203,137
11,370,656
1,088,640
10 to 30
0
7
0
0
41
0
0
803,727
0
30 to 50
0
0
0
0
0
0
0
0
0
50 to 75
0
0
0
0
0
0
0
0
0
>75
0
1
0
0
3
0
0
146,817
0
Option C
Oto 10
46
124
5
188
449
13
6,105,600
14,857,638
194,799
10 to 30
0
36
0
0
169
0
0
5,042,608
0
30 to 50
0
18
0
0
55
0
0
3,239,910
0
50 to 75
0
15
0
0
43
0
0
1,217,991
0
>75
0
11
0
0
32
0
0
954,431
0
a. Includes systems potentially directly and/or indirectly affected by steam electric power plant discharges.
b. Positive values indicate higher estimated bromide concentrations under the regulatory option as compared to the baseline, whereas negatives values indicate lower bromide
concentrations under the regulatory option.
c. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. In the 2019
analysis, EPA calculated annual average changes over the entire period of analysis (2021-2047) instead of the two periods used for the final rule analysis.
Source: U.S. EPA Analysis, 2020.
4-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-3: Estimated Distribution of Changes in Source Water and PWS-Level Iodine Concentrations by Period and Regulatory Option,
Compared to Baseline
Al Range
(Hg/L)
Number of Source Water Reaches
Number of PWSa
Population Served by PWS
Positive13 Al
Negative13 Al
No Al (Al = 0)
Positive13 Al
Negative13 Al
No Al (Al = 0)
Positive13 Al
Negative13 Al
No Al (Al = 0)
Period 1
Option A (Final Rule)
Oto 0.3
17
0
191
49
0
715
3,307,761
0
22,458,258
0.3 to 0.6
0
0
0
0
0
0
0
0
0
0.6 to 0.9
0
0
0
0
0
0
0
0
0
0.9 to 1.2
0
0
0
0
0
0
0
0
0
>1.2
0
0
0
0
0
0
0
0
0
Option B
Oto 0.3
17
0
191
49
0
715
3,307,761
0
22,458,258
0.3 to 0.6
0
0
0
0
0
0
0
0
0
0.6 to 0.9
0
0
0
0
0
0
0
0
0
0.9 to 1.2
0
0
0
0
0
0
0
0
0
>1.2
0
0
0
0
0
0
0
0
0
Option C
Oto 0.3
9
157
14
30
598
47
442,035
21,335,359
710,494
0.3 to 0.6
0
21
0
0
52
0
0
1,814,351
0
0.6 to 0.9
0
2
0
0
7
0
0
176,180
0
0.9 to 1.2
0
3
0
0
19
0
0
995,383
0
>1.2
0
2
0
0
11
0
0
292,217
0
Period 2
Option A (Final Rule)
Oto 0.3
0
60
140
0
273
441
0
6,309,989
18,500,500
0.3 to 0.6
0
7
0
0
46
0
0
365,713
0
0.6 to 0.9
0
0
0
0
1
0
0
443,000
0
0.9 to 1.2
0
0
0
0
0
0
0
0
0
>1.2
0
1
0
0
3
0
0
146,817
0
Option B
Oto 0.3
0
67
133
0
292
422
0
11,365,670
13,444,819
0.3 to 0.6
0
7
0
0
46
0
0
365,713
0
0.6 to 0.9
0
0
0
0
1
0
0
443,000
0
0.9 to 1.2
0
0
0
0
0
0
0
0
0
>1.2
0
1
0
0
3
0
0
146,817
0
4-8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-3: Estimated Distribution of Changes in Source Water and PWS-Level Iodine Concentrations by Period and Regulatory Option,
Compared to Baseline
Al Range
Number of Source Water Reaches
Number of PWSa
Population Served by PWS
(Hg/L)
Positive13 Al
Negative13 Al
No Al (Al = 0)
Positive13 Al
Negative13 Al
No Al (Al = 0)
Positive13 Al
Negative13 Al
No Al (Al = 0)
Option C
Oto 0.3
0
110
4
0
402
16
0
14,086,251
453,441
0.3 to 0.6
0
33
0
0
139
0
0
5,025,163
0
0.6 to 0.9
0
23
0
0
75
0
0
2,262,198
0
0.9 to 1.2
0
17
0
0
54
0
0
601,109
0
>1.2
0
21
0
0
78
0
0
3,337,857
0
a. Includes systems potentially directly and/or indirectly affected by steam electric power plant discharges.
b. Positive values indicate higher estimated iodine concentrations under the regulatory option as compared to the baseline, whereas negatives values indicate lower iodine
concentrations under the regulatory option.
Option D is omitted from this table because EPA did not conduct this analysis at proposal. See the 2019 BCA (U.S. EPA, 2019a).
Source: U.S. EPA Analysis, 2020.
4-9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
4.2 Additional Measures of Human Health Effects from Exposure to Steam Electric Pollutants
via Drinking Water Pathway
The regulatory options may result in small changes to source water quality for additional parameters that can
adversely affect human health (see Section 2.1.1). Many pollutants in steam electric power plant discharges
have MCLs that set allowable levels in treated water. For some pollutants that have an MCL above the
MCLG, there may be incremental benefits from reducing concentrations below the MCL.
Estimated concentrations of arsenic and lead in drinking water source reaches downstream of steam electric
facilities do not exceed typical detection limits for these contaminants. The results show thallium
concentrations in source waters that exceed levels detectable by standard methods (0.005 j^ig/L) in one source
water reach but are below 0.005 (ig/L in all other modeled source waters. Relative to baseline concentrations,
the changes in arsenic, lead, and thallium concentrations are very small.
Table 4-4 summarizes the direction of changes in arsenic, lead, and thallium concentrations under the
regulatory options43 for the two analysis periods. The direction of the changes depends on the Period,
regulatory option, source water reach, and PWS but is generally consistent with the changes in halogen
loadings associated with FGD wastewater and bottom ash transport water under each analyzed regulatory
option (see Table 3-3). During Period 1, Options A and B show either increases or no changes in arsenic,
lead, and thallium concentrations for all source waters and PWS and Option C shows both increases and
decreases in arsenic, lead, and thallium concentrations across locations. During Period 2, the three options
show estimated increases and decreases in arsenic, lead, and thallium concentrations with both the magnitude
and scope (the number of reaches, PWS, and population served) of the decreases larger than during Period 1.
To assess potential additional drinking water-related health benefits, EPA estimated the changes in the
number of receiving reaches with drinking water intakes that have modeled pollutant concentrations in excess
of MCLs or MCLGs. EPA did this analysis for all of the pollutants listed in Table 2-2, except bromate and
TTHM. This analysis showed no changes in the number of MCL or MCLG exceedances under the regulatory
options, when compared to the baseline. In addition, EPA found no reaches with drinking water intakes that
had modeled lead, arsenic, or thallium concentrations in excess of MCLs or MCLGs under either the baseline
or the regulatory options, even where concentrations increased as summarized in Table 4-4.44 The Agency
concluded, based on these screening analyses, that any additional benefits from changes in exposure to the
pollutants examined in this analysis via the drinking water pathway would be very small.
43 Option D is omitted from this table because EPA did not conduct this analysis at proposal. See the 2019 BCA (U.S. EPA, 2019a).
44 EPA also found that there are no reaches with drinking water intakes that have pollutant concentrations in excess of human health
ambient water quality criteria for either the consumption of water and organism or the consumption of organism only.
4-10
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
Table 4-4: Estimated Distribution of Changes in Source Water and PWS-Level Arsenic, Lead, and Thallium Concentrations by Period and
Regulatory Option, Compared to Baseline

Number of Source Water Reaches

Number of PWSa
Population Served by PWS
Regulatory Option
Positive13
Change
Negative13
Change
No
Change
Positive13
Change
Negative13
Change
No Change
Positive13
Change
Negative13
Change
No Change
Period 1
Arsenic
Option A (Final Rule)
245
0
10
894
0
55
30,510,519
0
1,102,458
Option B
245
0
10
894
0
55
30,510,519
0
1,102,458
Option C
201
51
3
790
146
13
27,760,901
3,649,526
202,550
Lead
Option A (Final Rule)
245
0
10
894
0
55
30,510,519
0
1,102,458
Option B
245
0
10
894
0
55
30,510,519
0
1,102,458
Option C
215
37
3
818
118
13
28,604,368
2,806,059
202,550
Thallium
Option A (Final Rule)
245
0
10
894
0
55
30,510,519
0
1,102,458
Option B
245
0
10
894
0
55
30,510,519
0
1,102,458
Option C
113
139
3
495
441
13
8,936,558
22,473,869
202,550
Period 2
Arsenic
Option A (Final Rule)
215
27
13
690
201
58
28,650,054
1,874,283
1,088,640
Option B
215
27
13
690
201
58
28,650,054
1,874,283
1,088,640
Option C
76
174
5
270
666
13
12,591,406
18,826,772
194,799
Lead
Option A (Final Rule)
219
23
13
735
156
58
29,514,216
1,010,121
1,088,640
Option B
219
23
13
735
156
58
29,514,216
1,010,121
1,088,640
Option C
130
120
5
489
447
13
22,539,805
8,878,373
194,799
Thallium
Option A (Final Rule)
200
42
13
640
251
58
25,892,193
4,632,144
1,088,640
Option B
193
49
13
621
270
58
20,836,512
9,687,825
1,088,640
Option C
53
197
5
204
732
13
6,248,250
25,169,928
194,799
a. Includes systems potentially directly and/or indirectly affected by steam electric power plant discharges.
b. Positive values indicate higher estimated concentrations under the regulatory option as compared to the baseline, whereas negatives values indicate lower concentrations under
the regulatory option.
Source: U.S. EPA Analysis, 2020.
4-11
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
4: Human Health Benefits via Drinking Water
4.3 Limitations and Uncertainties
Table 4-5 summarizes principal limitations and sources of uncertainties associated with the estimated changes
in pollutant levels in source waters downstream from steam electric power plant discharges. Additional
limitations and uncertainties are associated with the estimation of pollutant loadings (see U.S. EPA, 2020f).
Note that the effect on benefits estimates indicated in the second column of the table refers to the magnitude
of the benefits rather than the direction (i.e.. a source of uncertainty that tends to underestimate benefits
indicates expectation for either larger forgone benefits or larger realized benefits).
Table 4-5: Limitations and Uncertainties in the Analysis of Human Health Benefits from Changes in
Discharges of Halogens and Other Pollutants Via the Drinking Water Pathway
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
For PWS with multiple
Uncertain
Data on the flow rates of individual source facilities are not
sources of water, the analysis

available and EPA therefore estimated that all permanent
uses equal contributions from

active sources contribute equally to a PWS's total supply.
each source.

Effects of the regulatory option may be greater or smaller
than estimated, depending on actual supply shares.
Changes in bromide and
Underestimate
The analysis includes only permanent active surface water
iodine concentrations are

facilities associated with non-transient PWS classified as
analyzed for active

"community water systems" that use surface water as primary
permanent surface water

source. To the extent that PWS using surface waters as
intakes and reservoirs only.

secondary source or other non-permanent surface water
facilities are affected, this approach understates the effects of
the regulatory options.
Discharge monitoring data for
Uncertain
Limited bromide monitoring data are available to assess
bromide from steam electric

bromide source water concentration estimates.
power plants are limited and

demonstrate significant

variability based on site-

specific factors.

Discharge monitoring data for
Uncertain
No iodine monitoring data are available to assess source
iodine from steam electric

water iodine concentration estimates.
power plants are unavailable.

Source water monitoring data
Uncertain
While some bromide monitoring data are available to assess
are unavailable to confirm

source water bromide concentration estimates, no iodine
estimated iodine

monitoring data are available to assess iodine concentration
concentrations associated

estimates.
with steam electric power

plant discharges in PWS

source waters.

The analysis does not
Underestimate
The analysis of other pollutants does not account for natural
consider pollutant sources

background and anthropogenic sources that do not report to
beyond those associated with

TRI. This results in a potential underestimate of the number of
steam electric power plants

waters exceeding the MCL or MCLG.
orTRI dischargers.

4-12
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
5 Human Health Effects from Changes in Pollutant Exposure via the
Fish Ingestion Pathway
EPA expects the regulatory options to affect human health risk by changing effluent discharges to surface
waters and, as a result, ambient pollutant concentrations in the receiving reaches. The Supplemental EA (U.S.
EPA, 2020f) provides details on the health effects of steam electric pollutants. Recreational and subsistence
fishers (and their household members) who consume fish caught in the reaches receiving steam electric power
plant discharges could benefit from reduced pollutant concentrations in fish tissue. This chapter presents
EPA's analysis of human health effects resulting from changes in exposure to pollutants in bottom ash
transport water and FGD wastewater via the fish consumption pathway. The analyzed health effects include:
• Changes in exposure to lead: This includes changes in neurological and cognitive damages in children
(ages 0-7) based on the impact of an additional IQ point on an individual's future earnings and the
cost of compensatory education for children with learning delays.
• Changes in exposure to mercury: Changes in neurological and cognitive damages in infants from
exposure to mercury in-utero based on the impact of an additional IQ point on an individual's future
earnings.
• Changes in exposure to arsenic: Changes in incidence of cancer cases and the COI associated with
treating skin cancer.
The total quantified human health effects included in this analysis represent only a subset of the potential
health effects estimated to result from the regulatory options. While additional adverse health effects are
associated with pollutants in bottom ash transport water and FGD wastewater (such as kidney damage from
cadmium or selenium exposure, gastrointestinal problems from zinc, thallium, or boron exposure, and others),
the lack of data on dose-response relationships45 between ingestion rates and these effects precluded EPA
from quantifying the associated health effects.
EPA's analysis of the monetary value of human health effects utilizes data and methodologies described in
Chapter 3 and in the Supplemental EA (U.S. EPA, 2020f). The relevant data include the set of immediate and
downstream reaches that receive steam electric power plant discharges (i.e.. affected reaches), as defined by
the NHD COMID46, the estimated ambient pollutant concentrations in receiving reaches, and estimated fish
consumption rates among different age and ethnic cohorts for affected recreational and subsistence fishers.
Section 5.1 describes how EPA identified the population potentially exposed to pollutants from steam electric
power plant discharges via fish consumption. Section 5.2 describes the methods for estimating fish tissue
pollutant concentrations and potential exposure via fish consumption in the affected population. Sections 5.3
to 5.5 describe EPA's analysis of various human health endpoints potentially affected by the regulatory
options, which are then summarized in Section 5.6. Section 5.7 provides additional measures of human health
benefits. Section 5.8 describes limitations and uncertainties.
45 A dose response relationship is an increase in incidences of an adverse health outcome per unit increase in exposure to a toxin.
46 A COMID is a unique numeric identifier for a given waterbody (reach), assigned by a joint effort of the United States Geological
Survey and EPA.
5-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
In general, the estimated human health effects of the final rule, Option A, are small compared to baseline (see
U.S. EPA, 2015a).
5.1 Population in Scope of the Analysis
The population in scope of the analysis (/'. e., individuals potentially exposed to steam electric pollutants via
consumption of contaminated fish tissue) includes recreational and subsistence fishers who fish reaches
affected by steam electric power plant discharges (including receiving and downstream reaches), as well as
their household members. EPA estimated the number of people who are likely to fish affected reaches based
on typical travel distances to a fishing site and presence of substitute fishing locations. EPA notes that the
universe of sites potentially visited by recreational and subsistence fishers includes reaches subject to fish
consumption advisories (FCA).47 EPA expects that recreational fishers responses to FCA presence are
reflected in their catch and release practices, as discussed below.
Since fish consumption rates vary across different age, racial and ethnic groups, and fishing mode
(recreational versus subsistence fishing), EPA estimated potential health effects separately for a number of
age-, ethnicity-, and mode-specific cohorts. For each Census Block Group (CBG) within 50 miles of an
affected reach, EPA assembled 2017 American Community Survey data on the number of people in 7 age
categories (0 to 1, 2, 3 to 5, 6 to 10, 11 to 15, 16 to 21, and 21 years or higher), and then subdivided each
group according to 7 racial/ethnic categories:48 1) White non-Hispanic; 2) African-American non-Hispanic; 3)
Tribal/Native Alaskan non-Hispanic; 4) Asian/Pacific Islander non-Hispanic; 5) Other non-Hispanic
(including multiple races); 6) Mexican Hispanic; and 7) Other Hispanic49. Within each racial/ethnic group,
EPA further subdivided the population according to recreational and subsistence fisher groups. The Agency
assumed that the 95th percentile of the general population fish consumption rate is representative of the
subsistence fisher consumption rate. Accordingly, the Agency assumed that 5 percent of the total fishers
population practices subsistence fishing.50 EPA also subdivided the affected population by income into
poverty and non-poverty groups, based on the share of people below the federal poverty line.51 After
subdividing population groups by age, race, fishing mode, and the poverty indicator, each CBG has
196 unique population cohorts (7 age groups x 7 ethnic/racial groups x 2 fishing modes [recreational versus
subsistence fishing] x 2 poverty status designations).
47 Based on EPA's review of studies documenting fishers' awareness of FCA and their behavioral responses to FCA, 57.0 percent
to 61.2 percent of fishers are aware of FCA, and 71.6 percent to 76.1 percent of those who are aware ignore FCA (Burger, 2004,
Jakus etal., 1997; Jakus etal., 2002; R. L. Williams etal., 2000). Therefore, only 17.4 percent of fishers may adjust their
behavior in response to FCA (U.S. EPA, 2015a). The analysis reflects EPA's expectations that fishers responses to FCA are
reflected in their catch and release practices.
48 The racial/ethnic categories are based on available fish consumption data as well as the breakout of ethnic/racial populations in
Census data, which distinguishes racial groups within Flispanic and non-Hispanic categories.
49 The Mexican Flispanic and Flispanic block group populations were calculated by applying the Census tract percent Mexican
Flispanic and Flispanic to the underlying block-group populations, since these data were not available at the block-group level.
50 Data are not available on the share of the fishing population that practices subsistence fishing. EPA assumed that 5 percent of
people who fish practice subsistence fishing, based on the assumed 95th percentile fish consumption rate for this population in
EPA's Exposure Factors Flandbook (see U.S. Environmental Protection Agency, 2011).
51 Poverty status is based on data from the Census Bureau's American Community Survey which determines poverty status by
comparing annual income to a set of dollar values called poverty thresholds that vary by family size, number of children, and the
age of the householder.
5-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
EPA distinguished the exposed population by racial/ethnic group and poverty status to support analysis of
potential environmental justice (EJ) considerations from baseline exposure to pollutants in steam electric
power plant discharges, and to allow evaluation of the effects of the regulatory options on mitigating any EJ
concerns. See Chapter 14 for details of the EJ analysis. As noted below, distinguishing the exposed
population in this manner also allows the Agency to account for differences in exposure among demographic
groups, where supported by available data.
Equation 5-1 shows how EPA estimated the population potentially exposed to steam electric pollutants,
ExPop(i)(s)(c), for CBG i in state 5 for cohort c.
Equation 5-1. ExPop(i)(s)(c) = Pop(i)(c)x %Fish(s) x CaR(c)
Where:
Pop(i)(c) = Total CBG population in cohort c. Age and racial/ethnicity-specific populations in each
CBG are based on data from the 2017 American Community Survey, which provides
population numbers for each CBG broken out by age and racial/ethnic group. To
estimate the population in each age- and ethnicity/race-specific group, EPA calculated
the share of the population in each racial/ethnic group and applied those percentages to
the population in each age group.
%Fish(s) = Fraction of people who live in households with fishers. To estimate what percentage of the
total population participates in fishing, EPA used region-specific U.S. Fish and Wildlife
Service (U.S. FWS, 2018) estimates of the population 16 and older who fish.52 EPA
assumed that the share of households that includes fishers is equal to the fraction of
people over 16 who participate in recreational fishing.
CaR(c) = Adjustment for catch-and-re lease practices. According to U.S. FWS (U.S. FWS, 2006) data,
approximately 23.3 percent of recreational fishers release all the fish they catch ("catch-
and-release" fishers). Fishers practicing "catch-and-release" would not be exposed to
steam electric pollutants via consumption of contaminated fish. For all recreational
fishers, EPA reduced the affected population by 23.3 percent. EPA assumed that
subsistence fishers do not practice "catch-and-release" fishing.
Table 5-1 summarizes the population living within 50 miles of reaches affected by steam electric power plant
discharges (see Section 5.2.1 for a discussion of this distance buffer) and EPA's estimate of the population
potentially exposed to the pollutants via consumption of subsistence- and recreationally-caught fish (based on
2017 population data and not adjusted for population growth during the analysis period). Of the total
population, 16.0 percent live within 50 miles of an affected reach and participate in recreational and/or
subsistence fishing, and 12.4 percent are potentially exposed to fish contaminated by steam electric pollutants
in bottom ash transport water and/or FGD wastewater discharges.
52 The share of the population who fishes ranges from 8 percent in the Pacific region to 20 percent in the East South Central region.
Other regions include the Middle Atlantic (10 percent), New England (11 percent), South Atlantic (15 percent), Mountain
(15 percent), West South Central (17 percent), East North Central (17 percent), and West North Central (18 percent).
5-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 5: Human Health Benefits via Fish Ingestion
Table 5-1: Summary of Population Potentially Exposed to Contaminated Fish Living within 50 Miles
of Affected Reaches (as of 2017)
Total population
133,802,146
Total fishers population3
21,338,805
Population potentially exposed to contaminated fishb c
16,615,461
a. Total population living within 50 miles of an affected reach multiplied by the state-specific share of the population who fishes
based on U.S. FWS (2018; between 8 percent and 20 percent, depending on the state).
b. Total fishers population adjusted to remove fishers practicing catch-and-release and who therefore do not consume self-caught
fish.
c. Analysis accounts for projected population growth so that the average population in scope of the analysis over the period of
2021 through 2047 is 11.1 percent higher than the population in 2017 presented in the table, or 18.5 million people. The analysis
estimates that the fraction of the U.S. population engaged in recreational and subsistence fishing remains constant from 2021
through 2047.
Source: U.S. EPA Analysis, 2020
5.2 Pollutant Exposure from Fish Consumption
EPA calculated an average fish tissue concentration for each pollutant for each CBG based on a length-
weighted average concentration for all reaches within 50 miles. For each combination of pollutant, cohort and
CBG, EPA calculated the average daily dose (ADD) and lifetime average daily dose (LADD) consumed via
the fish consumption pathway.
5.2.1 Fish Tissue Pollutant Concentrations
The set of reaches that may represent a source of contaminated fish for recreational and subsistence fishers in
each CBG depends on the typical distance fishers travel to fish. EPA assumed that fishers typically travel up
to 50 miles to fish53, and used this distance to estimate the relevant fishing sites for the population of fishers
in each CBG.
Fishers may have several fishable sites to choose from within 50 miles of travel. To account for the effect of
substitute sites, EPA assumed that fishing efforts are uniformly distributed among all the available fishing
sites within 50 miles from the CBG (travel zone). For each CBG, EPA identified all fishable reaches within
50 miles (where distance was determined based on the Euclidean distance between the centroid of the CBG
and the midpoint of the reach) and the reach length in miles.
EPA then calculated, for each CBG within the 50-mile buffer of a fishable reach, the fish tissue concentration
of As, Hg, and lead (Pb). Appendix D describes the approach used to calculate fish tissue concentrations of
steam electric pollutants in the baseline and under each of the regulatory options.
For each CBG, EPA then calculated the reach length (Lengthy) weighted fish fillet concentration (C /.¦„/, hvia
(CBG)) based on all fishable reaches within the 50 mile radius according to Equation 5-2:
53 Studies of fishers behavior and practices have made similar observations (e.g., Sohngen eta/., 2015 and Sea Grant - Illinois-
Indiana, 2018).
5-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
_ Zi=icFishpm^l-engthi
Equation 5-2. CpishptUete(CBG) - rULength.
5.2.2 Average Daily Dose
Exposure to steam electric pollutants via fish consumption depends on the cohort-specific fish consumption
rates. Table 5-2 summarizes the average fish consumption rates, expressed in daily grams per kilogram of
body weight (BW), according to the race/ethnicity and fishing mode. The rates reflect recommended values
for consumer-only intake of finfish in the general population from all sources, based on EPA's Exposure
Factors Handbook (U.S. EPA, 2011). For more details on these fish consumption rates, see the Supplemental
EA (U.S. EPA, 2020f) and the uncertainty discussion in Section 5.8.
Table 5-2: Summary of Group-specific Consumption Rates for Fish Tissue Consumption Risk
Analysis
Race/ Ethnicity3
EA Cohort Nameb
Consumption Rate (g/kg BW/day)
Recreational
Subsistence
White (non-Hispanic)
Non-Hispanic White
0.67
1.9
African American (non-Hispanic)
Non-Hispanic Black
0.77
2.1
Asian/Pacific Islander (non-Hispanic)
Other, including Multiple Races
0.96
3.6
Tribal/Native Alaskan (non-Hispanic)
Other, including Multiple Races
0.96
3.6
Other non-Hispanic
Other, including Multiple Races
0.96
3.6
Mexican Hispanic
Mexican Hispanic
0.93
2.8
Other Hispanic
Other Hispanic
0.82
2.7
a. Each group is also subdivided into seven age groups (0-1, 2, 3-5, 6-10,11-15,16-20, Adult [21 or higher] and two income groups
[above and below the poverty threshold]).
b. See Supplemental EA for details (U.S. EPA, 2020f).
Source: U.S. EPA Analysis, 2020
Equation 5-3 and Equation 5-4 show the cohort- and CBG-specific ADD and LADD calculations based on
fish tissue concentrations, consumption rates, and exposure duration and averaging periods from U.S. EPA
(2020f).
, . ... ^FisllFillet (^^RFiskt^XpFish
Equation 5-3. ADD(c)(i) = ———
Where:
ADD(c)(i) = average daily dose of pollutant from fish consumption for cohort c in CBG /
(milligrams[mg] per kilogram [kg] body weight [BW] per day)
Cfishjmetij) = average fish fillet pollutant concentration consumed by humans for CBG /' (mg per kg)
CRfhh(c) = consumption rate of fish for cohort c (grams per kg BW per day); see Table 5-2.
FfiSh = fraction of fish from reaches within the analyzed distance from the CBG (percent; estimated value
of 100%)54
54 Given the uncertainty inherent in this estimate, EPA conducted a sensitivity analysis using an alternative estimate. These results
are summarized in DCN SE09336: Alternative Value for Fraction of Fish Consumed from a Contaminated Source.
5-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
r- X- f A . -r.™/ WN ADD(c)(i)xED(c)xEF
Equation 5-4. LADD(c)(i) = ——
^ v yw 4TX365
Where:
LADD (c)(j) = lifetime average daily dose (mg per kg BW per day) for cohort c in CBG /'
ADD (c)(j) = average daily dose (mg per kg BW per day) for cohort c in CBG /'
ED(c) = exposure duration (years) for cohort c
EF= exposure frequency (days; set to 350)
AT= averaging time (years; set to 70)
EPA used the doses of steam electric pollutants as calculated above from fish caught through recreational and
subsistence fishing in its analysis of benefits associated with the various human health endpoints described
below.
5.3 Health Effects in Children from Changes in Lead Exposure
EPA's estimated changes in lead exposure relative to the baseline as a result of the regulatory options are
small compared to those estimated in the 2015 analysis (see U.S. EPA, 2015a).
Lead is a highly toxic pollutant that can cause a variety of adverse health effects in children of all ages. In
particular, elevated lead exposure may induce a number of adverse neurological effects in children, including
decline in cognitive function, conduct disorders, attentional difficulties, internalizing behavior55, and motor
skill deficits (see National Toxicology Program [NTP], 2012; U.S. EPA, 2013c, 2019c, and 2020f). Elevated
blood lead (PbB) concentrations in children may also slow postnatal growth in children ages one to 16, delay
puberty in 8- to 17-year-olds, and decrease hearing and motor function (NTP, 2012; U.S. EPA, 2019c). Lead
exposure is also associated with adverse health outcomes related to the immune system, including atopic and
inflammatory responses (e.g., allergy and asthma) and reduced resistance to bacterial infections. Studies have
also found a relationship between lead exposure in expectant mothers and lower birth weight in newborns
(NTP, 2012; U.S. EPA, 2019c; Zhu etal., 2010). Because of data limitations, EPA estimated only the effects
of changes in neurological and cognitive damages to pre-school (ages 0 to 7) children using the dose-response
relationship for IQ decrements (Crump et al., 2013).
EPA estimated health effects from changes in exposure to lead to preschool children using PbB as a
biomarker of lead exposure. EPA modeled PbB under the baseline and regulatory option scenarios, and then
used a concentration-response relationship between PbB and IQ loss to estimate changes in IQ losses in the
affected population of children and changes in incidences of extremely low IQ scores (less than 70, or two
standard deviations below the mean). EPA calculated the monetary value of changes in children's health
effects based on the impact of an additional IQ point on an individual's future earnings and the cost of
compensatory education for children with learning disabilities (including children with IQ less than 70 and
PbB levels above 20 ng/dL).
EPA used the methodology described in Section 5.1 to estimate the population of children from birth to age
seven who live in recreational fisher and subsistence fisher households and are potentially exposed to lead via
consumption of contaminated fish tissue. EPA notes that fish tissue is not the only route of exposure to lead
55 Behavioral difficulties in children may include both externalizing behavior (e.g., inattention, impulsivity, conduct disorders), and
internalizing behaviors (e.g., withdrawn behaviors, symptoms of depression, fearfulness, and anxiety).
5-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
among children. Other routes of exposure may include drinking water, dust, and other food. EPA used
reference exposure values for these other routes of lead exposures and held these values constant for the
baseline and regulatory options scenarios. Since this health effect applies to children up to the seventh
birthday only, EPA restricted the analysis to the relevant age cohorts of fisher household members.
5.3.1 Methods
This analysis considers children who are born after implementation of the regulatory options and live in
recreational fisher and subsistence fisher households. It relies on EPA's Integrated Exposure, Uptake, and
Biokinetics (IEUBK) Model for Lead in Children (U.S. EPA, 2009b), which uses lead concentrations in a
variety of media - including soil, dust, air, water, and diet - to estimate total exposure to lead for children in
seven one-year age cohorts from birth through the seventh birthday. Based on this total exposure, the model
generates a predicted geometric mean PbB for a population of children exposed to similar lead levels. See the
2013 BCA report (U.S. EPA, 2013a) for details.
For each CBG, EPA used the cohort-specific ADD based on Equation 5-3. EPA then multiplied the cohort-
specific ADD by the average body weight for each age group56 to calculate the "alternative source" input for
the IEUBK model. Lead bioavailability and uptake after consumption vary for different chemical forms.
Many factors complicate the estimation of bioavailability, including nutritional status and timing of meals
relative to lead intake. For this analysis, EPA used the default media-specific bioavailability factor for the
"alternative source" provided in the IEUBK model, which is 50 percent for oral ingestion.
EPA used the IEUBK model to generate the geometric mean PbB for each cohort in each CBG under the
baseline and post-technology implementation scenarios. The IEUBK model processes daily intake to two
decimal places ((.ig/day). For this analysis, this means that some of the change between the baseline and
regulatory options is not accounted for by using the model (i.e., IEUBK does not capture very small changes),
since the estimated changes in health effects are driven by very small changes across large populations. This
aspect of the model contributes to potential underestimation of the lead-related health effects in children
arising from the regulatory options.
5.3.1.1 Estimating Changes in IQ Point Losses
EPA used the Crump et al. (2013) dose-response function to estimate changes in IQ losses between the
baseline and regulatory options. Comparing the baseline and regulatory option results provides the changes in
IQ loss per child. Crump et al. (2013) concluded that there was statistical evidence that the exposure-response
is non-linear over the full range of PbB. Equation 5-5 shows an exposure-response function that represents
this non-linearity:
Equation 5-5. A IQ = p1 x ln(PbB + 1)
Where:
Pi = -3.315 (log-linear regression coefficient on the lifetime blood lead level57)
56 The average body weight values are 11.4 kg for ages 0 to 2, 13.8 kg for ages 2 to less than 3, 18.6 kg for ages 3 to less than 6, and
31.8 kg for ages 6 to 7.
57 The lifetime blood lead level in children ages 0 to 7 is defined as a mean from six months of age to present (Crump etal., 2013).
5-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Multiplying the result by the number of affected pre-school children yields the total change in the number of
IQ points for the affected population of children for the baseline and each regulatory option.
The IEUBK model estimates the mean of the PbB distribution in children, assuming a continuous exposure
pattern for children from birth through the seventh birthday. The 2017 American Community Survey
indicates that children ages 0 to 7 are approximately evenly distributed by age. To get an annual estimate of
the number of children that would benefit from implementation of the regulatory options, EPA divided the
estimated number of affected pre-school children by 7. This division adjusts the equation to apply only to
children age 0 to 1. The estimated changes in IQ loss is thus an annual value (i.e.. it would apply to the cohort
of children born each year after implementation).58 Equation 5-6 shows this calculation for the annual
increase in total IQ points.
\IO(i)(c) = the difference in total IQ points between the baseline and regulatory option scenarios for
cohort c in CBG i
Ln(AGM(i)(c)) = the log-linear change in the average PbB in affected population of children (j^ig/dL) for
cohort c in CBG i
CRF= -3.315, the log-linear regression coefficient from Crump el al. (2013)
ExCh(i)(c) = the number of affected children aged 0 to 7 for cohort c in CBG i
The available economic literature provides little empirical data on society's overall WTP to avoid a decrease
in children's IQ. To estimate the value of avoided IQ losses, EPA used estimates of the changes in a child's
future expected lifetime earnings per one IQ point reduction and the cost of compensatory education for
children with learning disabilities.
EPA estimated the value of an IQ point using the methodology presented in Salkever (1995) but with more
recent data from the 1997 National Longitudinal Survey of Youth (U.S. EPA, 2019b). Updated results based
on Salkever (1995) indicate that a one-point IQ reduction reduces expected lifetime earnings by 2.63 percent.
Table 5-3 summarizes the estimated values of an IQ point based on the updated Salkever (1995) analysis
using 3 percent and 7 percent discount rates. These values are discounted to the third year of life to represent
the midpoint of the exposed children population. EPA also used an alternative value of an IQ point from Lin
el al. (2018) in a sensitivity analysis (see Appendix G).
58 Dividing by seven undercounts overall benefits. Children from ages 1 to 7 (i.e., born prior to the base year of the analysis) are not
accounted for in the analysis, although they are also affected by changes in lead exposure.
Equation 5-6.
Where:
A/Q(i)(c) = (ln(AGM(i)(c))
x CRF x

5-8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Table 5-3: Value of an IQ Point (2018$) based on Expected
Reductions in Lifetime Earnings
Discount Rate
Value of an IQ Pointa b (2018$)
3 percent
$20,832
7 percent
$4,358
a. Values are adjusted for the cost of education.
b. EPA adjusted the value of an IQ point to 2018 dollars using the GDP
deflator.
Source: U.S. EPA, 2019b re-analysis of data from Salkever (1995)
5.3.1.2 Reduced Expenditures on Compensatory Education
Children whose PbB exceeds 20 |a,g/dL are more likely to have IQs less than 70, which means that they would
require compensatory education tailored to their specific needs. Costs of compensatory education and special
education are not reflected in the IQ point dollar value. Reducing exposure to lead at an early age is expected
to reduce the incidence of children requiring compensatory and/or special education, which would in turn
lower associated costs. Though these costs are not a substantial component of the overall benefits, they do
represent a potential benefit of changes in lead exposure. EPA quantitatively assessed this benefit category
using the methodology from the 2015 BCA (U.S. EPA, 2015a). The estimated cost savings from the estimated
changes in the need of compensatory education are negligible and are not included in the total monetized
benefits.
5.3.2 Results
Table 5-4 shows the monetary values associated with changes in IQ losses from lead exposure via fish
consumption. EPA estimated that regulatory options A and B lead to slight increases in lead exposure and, as
a result, forgone benefits, whereas Option C results in slight decreases in exposure to lead. The total net
change in IQ point losses over the entire population of children with changes in lead exposure ranges from -
19 (Option A) points to 12 points (Options C). Annualized monetary values of changes in IQ losses range
from approximately -$16,000 (Option A) to $7,000 (Option C) using a 3 percent discount, and approximately
-$4,000 (Option A) to $1,000 (Option C) using a 7 percent discount.
Table 5-4: Estimated Monetary Value of Changes in IQ Points for Children Exposed to Lead under
the Regulatory Options, Compared to Baseline
Regulatory Option
Average Annual
Number of Children 0
to 7 in Scope of the
Analysisc
Total Change in IQ
Point Losses, 2021 to
2047 in All Children 0 to
7 in Scope of the
Analysis'*
Annualized Value of Changes in IQ Pointsa b
(Thousands 2018$)
3% Discount Rate
7% Discount Rate

Option A (Final Rule)
1,615,629
-19
-$15.8
-$3.9
Option B
1,615,629
-11
-$10.5
-$2.9
Option C
1,615,629
12
$6.5
$0.7
a. Based on estimate that the loss of one IQ point results in the loss of 2.63 percent of lifetime earnings, following updated
Salkever (1995) values from U.S. EPA (2019b).
b. Negative values represent forgone benefits.
c. The number of children in scope of the analysis is based on reaches analyzed across the regulatory options. Some of the
children included in this count see no changes in exposure under some options.
5-9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 5: Human Health Benefits via Fish Ingestion
Table 5-4: Estimated Monetary Value of Changes in IQ Points for Children Exposed to Lead under
the Regulatory Options, Compared to Baseline
Regulatory Option
Average Annual
Number of Children 0
to 7 in Scope of the
Analysisc
Total Change in IQ
Point Losses, 2021 to
2047 in All Children 0 to
7 in Scope of the
Analysis'*
Annualized Value of Changes in IQ Pointsa b
(Thousands 2018$)
3% Discount Rate
7% Discount Rate
d. EPA notes that the IQ point losses are very small. EPA further notes that the IEUBK model does not analyze blood lead level
changes beyond two decimal points. EPA presents these estimates primarily for comparison to the 2015 final rule estimates.
e. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
5.4 Heath Effects in Children from Changes in Mercury Exposure
EPA estimated small changes in mercury exposure as a result of the regulatory options, compared to baseline
(U.S. EPA, 2015a).
Mercury can have a variety of adverse health effects on adults and children (U.S. EPA, 2020f). The regulatory
options may change the discharge of mercury to surface waters by steam electric power plants and therefore
affect a range of human health effects. Due to data limitations, however, EPA estimated only the monetary
value of the changes in IQ losses among children exposed to mercury in-ntero as a result of maternal
consumption of contaminated fish.
EPA identified the population of children exposed in-ntero starting from the CBG-specific population in
scope of the analysis described in Section 5.1. Because this analysis focuses only on infants born after
implementation of the regulatory options, EPA further limited the analyzed population by estimating the
number of women between the ages of 15 and 44 potentially exposed to contaminated fish caught in the
affected waterbodies, and multiplying the result by ethnicity-specific average fertility rates.59 This yields the
cohort-specific annual number of births for each CBG.
The U.S. Department of Health and Human Services provides fertility rates by race for 2015 in the National
Vital Statistics Report (Martin etal., 2019). The fertility rate measures the number of births occurring per
1,000 women between the ages of 15 and 44 in a particular year. Fertility rates were highest for Hispanic
women at 71.7, followed by African Americans at 64.1, Caucasians at 59.3, Asian or Pacific Islanders at 58.5,
and Tribal/Other at 43.9.
5.4.1 Methods
EPA used the ethnicity- and mode-specific consumption rates shown in Table 5-2 and calculated the CBG-
and cohort-specific mercury ADD based on Equation 5-3. In this analysis, EPA used a linear dose-response
relationship between maternal mercury hair content and subsequent childhood IQ loss from Axelrad et al.
(2007). Axelrad et al. (2007) developed a dose-response function based on data from three epidemiological
studies in the Faroe Islands, New Zealand, and Seychelle Islands. According to their results, there is a 0.18-
point IQ loss for each 1 part-per-million (ppm) increase in maternal hair mercury.
59 EPA acknowledges that fertility rates vary by age. However, the use of a single average fertility rate for all ages is not expected
to bias results because the average fertility rate reflects the underlying distribution of fertility rates by age.
5-10
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
To estimate maternal hair mercury concentrations based on the daily intake (see Section 5.2.2), EPA used the
median conversion factor derived by Swartout and Rice (2000), who estimated that a 0.08 jj.g/kg body weight
increase in daily mercury dose is associated with a 1 ppm increase in hair concentration. Equation 5-7 shows
EPA's calculation of the total annual IQ changes for a given receiving reach.
IQL(i) = IQ changes associated with in-utero exposure to mercury from maternal consumption of fish
contaminated with mercury for cohort c in CBG i
InExPop(i)(cj = population of infants in scope of the analysis for cohort c in CBG /' (the number of
births)
MADD(i)(c) = maternal ADD for cohort c in CBG i (|_ig/kg BW/day)
Conv = conversion factor for hair mercury concentration based on maternal mercury exposure
(0.08 (ig/kg BW/day per 1 ppm increase in hair mercury)
1)111'' = dose response function for IQ decrement based on marginal increase in maternal hair mercury
(0.18 point IQ decrement per 1 ppm increase in hair mercury)
Summing estimated IQ changes across all analyzed CBGs yields the total changes in the number of IQ points
due to in-utero mercury exposure from maternal fish consumption under each analyzed regulatory option. The
benefits of the regulatory options are calculated as the change in IQ points between the baseline and modeled
post-technology implementation conditions under each of the regulatory options.
The available economic literature provides little empirical data on society's overall WTP to avoid a decrease
in children's IQ. To estimate the value of avoided IQ losses, EPA used estimates of the changes in a child's
future expected lifetime earnings per one IQ point reduction. The values of an IQ point presented in Section
5.3.1 are discounted to the third year of life to represent the midpoint of the exposed children population of
interest for that analysis. EPA further discounted the present value of lifetime income differentials three
additional years to reflect the value of an IQ point at birth and better align the benefits of reducing exposure to
mercury with in-utero exposure (U.S. EPA, 2019d). The IQ values discounted to birth range from $3,704 to
$19,064. EPA also used an alternative value of an IQ point from Lin et al. (2018) in a sensitivity analysis (see
Appendix G).
5.4.2 Results
Table 5-5 shows the estimated changes in IQ point losses for infants exposed to mercury in-utero and the
corresponding monetary values, using 3 percent and 7 percent discount rates. Regulatory options A and B
result in a small net increase in IQ losses and, as a result, in forgone benefits to society. Option C results in a
small net decrease in IQ point losses , with decreases in Period 2 larger than initial increases in Period 1. The
annualized value of changes in IQ losses for Option C is negative despite the overall decrease in IQ point
losses due to discounting. Using a 3 percent discount rate, the monetary values of increased IQ losses range
from -$0.32 million (Option A) to -$0.11 million (Option C). Using a 7 percent discount rate, estimates range
from -$0.11 million (Option A) to -$0.07 million (Option C).
Equation 5-7.
Where:
IQL(i)(c) = InExPop(i)(c) * MADD(i)(c) * * DRF
5-11
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Table 5-5: Estimated Monetary Values of Changes in IQ Points for Infants from Mercury Exposure
under the Regulatory Options, Compared to Baseline
Regulatory Option
Number of Infants in
Scope of the Analysis per
Yearc
Total Change in IQ Point
Losses, 2021 to 2047 in
All Infants in Scope of the
Analysis
Annualized Value of Changes in IQ
Points3" (Millions 2018$)
3% Discount Rate
7% Discount Rate

Option A (Final Rule)
225,537
-201
-$0.32
-$0.11
Option B
225,537
-144
-$0.28
-$0.10
Option C
225,537
a;
t—1
r*>
-$0.11
-$0.07
a. Based on the estimate that the loss of one IQ point results in the loss of 2.63 percent of lifetime earnings discounted to birth,
following updated Salkever (1995) values from U.S. EPA (2019d).
b. Negative values represent forgone benefits.
c. The number of infants in scope of the analysis is based on reaches analyzed across the regulatory options. Some of the children
included in this count see no changes in exposure under some options.
d. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
e. Although Option C results in a small net decrease in IQ point losses (or positive benefits) due to larger decreases in Period 2 than
initial increases in Period 1, the annualized value for Option C is negative due to discounting.
Source: U.S. EPA Analysis, 2020
5.5 Estimated Changes in Cancer Cases from Arsenic Exposure
Among steam electric pollutants that can contaminate fish tissue and are analyzed in the Supplemental EA,
arsenic is the only confirmed carcinogen with a published dose response function (see U.S. EPA, 2010b).6"
EPA used the methodology presented in Section 3.6 of the 2015 BCA (U.S. EPA, 2015a) to estimate the
number of annual skin cancer cases associated with consumption of fish contaminated with arsenic from
steam electric power plant discharges under the baseline and the change corresponding to each regulatory
option and the associated monetary values. EPA's analysis shows no changes in skin cancer cases from
exposure to arsenic via fish consumption are expected under the regulatory options. Accordingly, the
estimated benefits are zero under all regulatory options.
5.6 Total Monetary Values of Estimated Changes in Human Health Effects
Table 5-6 presents the estimated monetary value of changes in adverse human health outcomes under the
regulatory options. Using a 3 percent discount rate, the estimated monetary values range from -$0.34 million
(Option A) to -$0.10 million (Option C). Using a 7 percent discount rate, the estimated monetary values range
from -$0.11 million (Option A) to -$0.07 million (Option C). Negative values reflect forgone benefits.
Changes in mercury exposure for children account for the majority of total monetary values from increases in
adverse health outcomes.
o0 Although other pollutants, such as cadmium, are also likely to be carcinogenic (see U.S. Department of Health and Human
Services, 2012), EPA did not identify dose-response functions to quantify the effects of changes in these other pollutants.
5-12
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Table 5-6: Total Monetary Values of Changes in Human Health Outcomes Associated with Fish
Consumption under the Regulatory Options, Compared to Baseline (Millions of 2018$)
Discount
Rate
Regulatory Option
Changes in Lead
Exposure for
Child renabc
Changes in
Mercury
Exposure for
Children313
Changes in
Cancer Cases
from Arsenic
Totalab

3%
Option A (Final Rule)
-$0.02
-$0.32
$0.00
-$0.34
Option B
-$0.01
-$0.28
$0.00
-$0.29

Option C
$0.01
-$0.11
$0.00
-$0.10

7%
Option A (Final Rule)
<$0.00
-$0.11
$0.00
-$0.11
Option B
<$0.00
-$0.10
$0.00
-$0.10

Option C
>$0.00
-$0.07
$0.00
-$0.07
a. Negative values represent forgone benefits and positive values represent realized benefits.
b. Based on the estimate that the loss of one IQ point results in the loss of 2.63 percent of lifetime earnings, following updated
Salkever (1995) values from U.S. EPA (2019b).
c. "<$0.00" indicates monetary values greater than -$0.01 million but less than $0.00 million. ">$0.00" indicates monetary values
greater than $0.00 million but less than $0.01 million.
d. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for
Option D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect
changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
5.7 Additional Measures of Potential Changes in Human Health Effects
As noted in the introduction to this chapter, untreated pollutants in steam electric power plant discharges have
been linked to additional adverse human health effects. EPA compared immediate receiving water
concentrations to human health-based NRWQC in U.S. EPA (2020f). To provide an additional measure of the
potential health effects of the regulatory options, EPA also estimated the changes in the number of receiving
and downstream reaches with pollutant concentrations in excess of human health-based NRWQC. This
analysis compares pollutant concentrations estimated for the baseline and each analyzed regulatory option in
receiving reaches and downstream reaches to criteria established by EPA for protection of human health. EPA
compared estimated in-water concentrations of antimony, arsenic, barium, cadmium, chromium, cyanide,
copper, lead, manganese, mercury, nitrate-nitrite as N, nickel, selenium, thallium, and zinc to EPA's NRWQC
protective of human health used by states and tribes (U.S. EPA, 2018c) and to MCLs.61 Estimated pollutant
concentrations in excess of these values indicate potential risks to human health. This analysis and its findings
are not additive to the preceding analyses in this chapter, but instead represent another way of characterizing
potential health effects resulting from changes in exposure to steam electric pollutants.
Table 5-7 shows the results of this analysis.62 During Period 1, EPA estimates that with baseline steam
electric pollutant discharges, concentrations of steam electric pollutants exceed human health criteria for at
least one pollutant in 161 reaches based on the "consumption of water and organism" criteria, and 38 reaches
based on the "consumption of organism only" criteria nationwide. EPA estimates that the total number of
01 For pollutants that do not have NRWQC protective of human health, EPA used MCLs. These pollutants include cadmium,
chromium, lead, and mercury.
02 Only reaches designated as fishable (i.e., Strahler Stream Order larger than 1) were included in the NRWQC exceedances
analysis.
5-13
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
reaches with exceedances during Period 1 will increase under all regulatory options. Under Option C, some
reaches are also estimated to experience a reduction in the number of exceedances relative to baseline. During
Period 2, concentrations of steam electric pollutants exceed human health criteria for at least one pollutant in
68 reaches based on the "consumption of water and organism" criteria, and 23 reaches based on the
"consumption of organism only" criteria nationwide under the baseline scenario. The estimated number of
reaches with exceedances of "consumption water and organism" criteria during Period 2 increases under
Options A and B and decreases under Option C. The total number of reaches with exceedances of
"consumption of organism only" criteria decreases under the three options.63
Table 5-7: Estimated Number of Reaches Exceeding Human Health Criteria for Steam Electric
Pollutants
Regulatory Option
Number of Reaches with
Ambient Concentrations
Exceeding Human Health
Criteria for at Least One
Pollutant3
Number of Reaches with
Higher Number of
Exceedances, Relative to
Baselineb
Number of Reaches with
Lower Number of
Exceedances, Relative to
Baselinec
Consumption
of Water +
Organism
Consumption
of Organism
Only
Consumption
of Water+
Organism
Consumption
of Organism
Only
Consumption
of Water +
Organism
Consumption
of Organism
Only
Period 1
Baseline
161
38
Not applicable
Not applicable
Not applicable
Not applicable
Option A (Final Rule)
223
71
71
40
0
0
Option B
223
70
71
39
0
0
Option C
230
65
79
36
10
4
Period 2
Baseline
68
23
Not applicable
Not applicable
Not applicable
Not applicable
Option A (Final
Rule)
88
17
26
0
6
6
Option B
88
17
26
0
6
6
Option C
34
2
26
0
65
23
a. Pollutants for which there was at least one exceedance in the baseline or regulatory options include antimony, arsenic,
cadmium, cyanide, lead, manganese, and thallium in Period 1 and arsenic, cyanide, manganese, and thallium in Period 2.
b. Pollutants for which there was at least one reach with higher number of exceedances relative to baseline include antimony,
arsenic, cadmium, cyanide, manganese, and thallium in Period 1 and arsenic in Period 2.
c. Pollutants for which there was at least one reach with lower number of exceedances relative to baseline include arsenic,
manganese, and thallium in Period 1 and arsenic, cyanide, manganese, and thallium in Period 2.
Source: U.S. EPA Analysis, 2020
5.8 Limitations and Uncertainties
The analysis presented in this chapter does not include all possible human health effects associated with post-
technology implementation changes in pollutant discharges due to lack of data on a dose-response
relationship between ingestion rates and potential adverse health effects. Therefore, the total quantified human
03 EPA's analysis does not take into account the fact that the NPDES permit for each steam electric power plant, like all NPDES
permits, is required to have limits more stringent than the technology-based limits established by an ELG, wherever necessary to
protect water quality standards. Because this analysis does not project where a permit will have more stringent limits than those
required by the ELG, it may overestimate any negative impacts to aquatic ecosystems and T&E species, including impacts that
will not be realized at all because the permits will be written to include limits as stringent as necessary to meet water quality
standards as required by the CWA.
5-14
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
health effects included in this analysis represent only a subset of the potential health effects estimated to result
from the regulatory options. Section 2.1 provides a qualitative discussion of health effects omitted from the
quantitative analysis.
The methodologies and data used in the analysis of adverse health outcomes due to consumption of fish
contaminated with steam electric pollutants involve limitations and uncertainties. Table 5-8 summarizes the
limitations and uncertainties and indicates the direction of the potential bias. Note that the effect on benefits
estimates indicated in the second column of the table refers to the magnitude of the benefits rather than the
direction (i.e.. a source of uncertainty that tends to underestimate benefits indicates expectation for either
larger forgone benefits or larger realized benefits). Additional limitations and uncertainties associated with the
EA analyses and data are discussed in the Supplemental EA (see U.S. EPA, 2020f).
Table 5-8: Limitations and Uncertainties in the Analysis of Human Health Effects via the Fish
Ingestion Pathway
Uncertainty/Limitation
Effect on Benefits Estimate
Notes
Fishers are estimated to
evenly distribute their
activity over all available
fishing sites within the 50-
mile travel distance.
Uncertain
EPA estimated that all fishers travel up to 50 miles
and distribute their visits over all fishable sites within
the area. In fact, recreational and subsistence fishers
may have preferred sites [e.g., a site located closer
to their home) that they visit more frequently. The
characteristics of these sites, notably ambient water
concentrations and fishing advisories, affects
exposure to pollutants, but EPA does not have data
to support a more detailed analysis of fishing visits.
The impact of this approach on monetary estimates
is uncertain since fewer/more fishers may be
exposed to higher/lower fish tissue concentrations
than estimated by EPA.
The exposed population is
estimated based on
households in proximity to
affected reaches and the
fraction of the general
population who fish.
Uncertain
EPA estimated the share of households that includes
fishers to be equal to the fraction of people over 16
who are fishers. This may double-count households
with more than one fisher over 16. However, the
exposed population may also include non-household
members who also consume the catch.
Fish intake rates used in
estimating exposure are
based on recommended
values for the general
consumer population.
Uncertain
The fish consumption rates used in the analysis are
based on the general consumer population which
may understate or overstate the amount offish
consumed by fishers who may consume fish at
higher or lower rates than the general population
(e.g., Burger, 2013; U.S. EPA, 2011, 2013b)
100 percent of fish
consumed by recreational
fishers is self- caught.
Overestimate
The fish consumption rates used in the analysis
account for all fish sources, i.e., store-bought or self-
caught fish. Assuming that recreational fishers
consume only self-caught fish may overestimate
exposure to steam electric pollutants from fish
consumption. The degree of the overestimate is
unknown as the fraction offish consumed that is
self-caught varies significantly across different
locations and population subgroups (e.g., U.S. EPA,
2013b).
5-15
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Table 5-8: Limitations and Uncertainties in the Analysis of Human Health Effects via the Fish
Ingestion Pathway
Uncertainty/Limitation
Effect on Benefits Estimate
Notes
The number of subsistence
fishers was set to equal
5 percent of the total
number of fishers fishing the
affected reaches.
Uncertain
The magnitude of subsistence fishing in the United
States or individual states is not known. Using
5 percent may understate or overstate the number
of potentially affected subsistence fishers (and their
households) overall, and ignores potential variability
in subsistence fishing rates across racial/ethnic
groups and different geographic locations.
There is a 0.18 point IQ loss
for each 1 ppm increase in
maternal hair mercury (i.e.,
the relationship is assumed
to be linear).
Uncertain
The exact form of the relationship between maternal
body mercury burden and IQ losses is uncertain.
Using a linear relationship may understate or
overstate the IQ losses resulting from a given change
in mercury exposure.
For the mercury- and lead-
related health impact
analyses, EPA assessed IQ
losses to be an appropriate
endpoint for quantifying
adverse cognitive and
neurological effects resulting
from childhood or in-utero
exposures to lead and
mercury (respectively).
Underestimate
IQ may not be the most sensitive endpoint.
Additionally, there are deficits in cognitive abilities
that are not reflected in IQ scores, including
increased incidence of attention-related and
problem behaviors (NTP, 2012 and U.S. EPA, 2005b).
To the extent that these impacts create
disadvantages for children exposed to mercury and
lead in the absence of (or independent from)
measurable IQ losses, this analysis may
underestimate the social welfare effects of the
regulatory options of changes in lead and mercury
exposure.
The IEUBK model processes
daily intake from "alternative
sources" to 2 decimal places
(Hg/day).
Underestimate
Since the fish-associated pollutant intakes are small,
some variation is missed by using this model (i.e., it
does not capture very small changes).
EPA did not monetize the
health effects associated
with changes in adult
exposure to lead or mercury.
Underestimate
The scientific literature suggests that exposure to
lead and mercury may have significant adverse
health effects for adults (e.g., Aoki et ai, 2016;
Chowdhury et al., 2018; Lanphear et ai, 2018). If
measurable effects are occurring at current exposure
levels, excluding the effects of increased adult
exposure results in an underestimate of benefits.
EPA did not quantify other
health effects in children
from exposure to lead or
mercury.
Underestimate
As discussed in Section 2.1, exposure to lead could
result in additional adverse health effects in children
(e.g., low birth weight and neonatal mortality from
in-utero exposure to lead, or neurological effects in
children exposed to lead after age seven) (NTP,
2012; U.S. EPA, 2013c; U.S. EPA, 2019c). Additional
neurological effects could also occur in children from
exposure to mercury after birth (Mergler et ai,
2007; CDC, 2009). If measurable effects are
occurring at current exposure levels, excluding
additional health effects of increased children
exposure results in an underestimate of benefits.
5-16
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
5: Human Health Benefits via Fish Ingestion
Table 5-8: Limitations and Uncertainties in the Analysis of Human Health Effects via the Fish
Ingestion Pathway
Uncertainty/Limitation
Effect on Benefits Estimate
Notes
EPA did not assess combined
health risk of multiple
pollutants.
Uncertain
The combined health risk of multiple pollutants
could be greater than from a single pollutant (Evans
et ai, 2020). However, quantifying cumulative risk is
challenging because a mixture of pollutants could
affect a wide range of target organs and endpoints
(ATSDR, 2004, 2009). For example, different
carcinogens found in steam electric power plant
discharges may affect different organs [e.g., arsenic
is linked to skin cancer while cadmium is linked to
kidney cancer). Other synergistic effects may
increase or lessen the risk.
5-17
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
6 Nonmarket Benefits from Water Quality Changes
As discussed in the Supplemental EA (U.S. EPA, 2020f), heavy metals, nutrients, and other pollutants
discharged by steam electric power plants can have a wide range of effects on water resources downstream
from the plants. These environmental changes affect environmental goods and services valued by humans,
including recreation; commercial fishing; public and private property ownership; navigation; water supply
and use; and existence services such as aquatic life, wildlife, and habitat designated uses. Some environmental
goods and services (e.g., commercially caught fish) are traded in markets, and thus their value can be directly
observed. Other environmental goods and services (e.g., recreation and support of aquatic life) cannot be
bought or sold directly and thus do not have observable market values. This second type of environmental
goods and services are classified as "nonmarket." The estimated changes in the nonmarket values of the water
resources affected by the regulatory options (hereafter nonmarket benefits or disbenefits) are additive to
market values (e.g., avoided costs of producing various market goods and services).
The analysis of the nonmarket value of water quality changes resulting from the regulatory options follows
the same approach EPA used in the analysis of the 2015 rule and 2019 proposal (U.S. EPA, 2015a, 2019a).
This approach, which is briefly summarized below, involves:
• characterizing the change in water quality under the regulatory options relative to the baseline using a
WQI and linking these changes to ecosystem services or potential uses that are valued by society (see
Section 3.4.2),
• monetizing changes in the nonmarket value of affected water resources under the regulatory options
using a meta-analysis of surface water valuation studies that provide data on the public's WTP for
water quality changes (see Section 6.1).
The analysis accounts for changes in water quality resulting from changes in nutrient, sediment, and toxics
concentrations in reaches potentially affected by bottom ash transport water and FGD wastewater discharges.
The assessment uses the CBG as the geographic unit of analysis, assigning a radial distance of 100 miles from
the CBG centroid. EPA estimates that households residing in a given CBG value water quality changes in all
modeled reaches within this range, with all unaffected reaches being viable substitutes for affected reaches
within the area around the CBG. Appendix E describes EPA's approach.
In general, the analysis shows that the estimated effects of the final rule on the nonmarket value of water
quality result in small forgone benefits when compared to those estimated under baseline (see U.S. EPA,
2015a).
6.1 Estimated Total WTP for Water Quality Changes
EPA estimated economic values of water quality changes at the CBG level using results of a meta-analysis of
168 estimates of total WTP (including both use and nonuse values) for water quality improvements, provided
by 65 original studies conducted between 1981 and 2017.64 The estimated econometric model allows
calculation of total WTP for changes in a variety of environmental services affected by water quality and
valued by humans, including changes in recreational fishing opportunities, other water-based recreation, and
64 Although the potential limitations and challenges of benefit transfer are well established (Desvousges eta/., 1987), benefit
transfers are a nearly universal component of benefit cost analyses conducted by and for government agencies. As noted by V. K.
Smith et al. (2002, p. 134), "nearly all benefit cost analyses rely on benefit transfers, whether they acknowledge it or not."
6-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
existence services such as aquatic life, wildlife, and habitat designated uses. The model also allows EPA to
adjust WTP values based on the core geospatial factors predicted by theory to influence WTP, including:
scale (the size of affected resources or areas), market extent (the size of the market area over which WTP is
estimated), and the availability of substitutes. The meta-analysis regression is based on two models: Model 1
provides EPA's central estimate of non-market benefits and Model 2 develops a range of estimates that
account for uncertainty in the WTP estimates. Appendix G provides details on how EPA used the meta-
analysis to predict household WTP for each CBG and year as well as the estimated regression equation,
intercept and variable coefficients for the two models used in this analysis. The appendix also provides the
corresponding independent variable names and assigned values.
Based on the meta-analysis results, EPA multiplied the coefficient estimates for each variable (see Model 1
and Model 2 in Table G-l) by the variable levels calculated for each CBG or fixed at the levels indicated in
the "Assigned Value" column in Table G-l. The sum of these products represents the predicted natural log of
marginal household WTP (In MWTP) for a representative household in each CBG. Equation 6-1 provides the
equation used to calculate household benefits for each CBG.
Equation 6-1. HWTPYB = MWTPYB x AWQIB
where:
HWTPy,b = Annual household WTP in 2018$ in year 7 for households located in
the CBG (B),
MWTPy.b = Marginal WTP for water quality for a given year (7) and the CBG (B)
estimated by the meta-analysis function and evaluated at the midpoint
of the range over which water quality is changed,
AWQIb = Estimated annual average water quality change for the CBG (B).
To estimate WTP for water quality changes under the regulatory options, EPA first estimated water quality
changes for each year within Period 1 and Period 2 (see Section 3.2.1 for details) and then applied the meta-
regression model (MRM) to estimate per household WTP for water quality improvements in a given year.
Monetary values of water quality changes are estimated for all years from 2021 through 2047. As summarized
in Table 6-1, average annual household WTP estimates for the regulatory options range from -$0.40 under
Option A (low estimate) to -$0.14 under Option C (high estimate), for the regulatory options EPA analyzed.
6-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
Table 6-1: Estimated Household Willingness-to-Pay for Water Quality Changes under the
Regulatory Options, Compared to Baseline
Regulatory Option
Nu mber of Affected
Households (Millions)c
Average Annual WTP Per Household (2018$)a b
Low
Central
High

Option A (Final Rule)
82.4
-$0.20
-$0.31
-$0.40
Option B
78.5
-$0.16
-$0.25
-$0.32
Option C
84.6
-$0.14
-$0.22
-$0.28
a. Negative values represent forgone benefits
b. Model 2 provides low and high estimates for each option, while Model 1 provides central estimates. EPA used AWQI equal to
5 units to develop low estimates and to 50 units to develop high estimates based on Model 2 (See Appendix G for details). The
central estimate does not fall at the midpoint of the range, but instead represents the value from Model 1 which falls between
the low and high bound estimates provided by Model 2.
c. The number of affected households varies across options because of differences in the number of reaches that have non-zero
changes in water quality.
d. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for
Option D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes
in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
To estimate total WTP (TWTP) for water quality changes for each CBG, EPA multiplied the per-household
WTP values for the estimated water quality change by the number of households within each block group in a
given year and calculated the present value (PV) of the stream of WTP over the 27 years in EPA's period of
analysis. EPA then calculated annualized total WTP values for each CBG using 3 percent and 7 percent
discount rates as shown in Equation 6-2.
Equation 6-2.
where:
TWTPe
HWTP
HHy3
T
i
n
Y,B
2047
twtpb = ( ^
HWTPy b X hhy b
\T = 2021
(l + 0
,'^7-2020
X
i x (1 + i)T
(1 + i)n+1 -
Annualized total household WTP in 2018$ for households located in
the CBG (5),
Annual household WTP in 2018$ for households located in the CBG
(B) in year (Y),
the number of households residing in the CBG (B) in year (Y),
Year when benefits are realized
Discount rate (3 or 7 percent)
Duration of the analysis (27 years)65
05 See Section 1.3.3 for details on the period of analysis.
6-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
EPA generated annual household counts for each CBG through the period of analysis based on projected
population growth following the method described in Section 1.3.7. Table 6-2 presents the results for the
3 percent and 7 percent discount rates.
Table 6-2: Estimated Total Annualized Willingness-to-Pay for Water Quality Changes under the
Regulatory Options, Compared to Baseline (Millions of 2018$)
Regulatory Option
Number of Affected
Households
(Millions)'
3% Discount Ratea,b
7% Discount Ratea b
Low
Central
High
Low
Central
High

-$8.6
Option A (Final Rule)
82.4
-$15.3
-$11.8
-$7.4
-$16.4
-$12.5
o
00
1
Option B
78.5
-$10.4
-$7.8
o
LO
i
-$12.0
-$9.0
00
LO
1
Option C
84.6
-$9.9
-$7.4
CO
1
-$13.9
-$10.3
-$6.7
a. Negative values represent forgone benefits.
b. Model 2 provides low and high estimates for each option, while Model 1 provides central estimates. EPA used AWQI equal to 5
units to develop low estimates and to 50 units to develop high estimates based on Model 2 (see Appendix G for details). The
central estimate does not fall at the midpoint of the range, but instead represents the value from Model 1 which falls between the
low and high bound estimates provided by Model 2.
c. The number of affected households varies across options because of differences in the number of reaches that have non-zero
changes in water quality.
d. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
The total annualized values of water quality changes resulting from changes in toxics, nutrient and sediment
discharges in these reaches range from -$16.4 million under Option A (7 percent discount rate)
to -$4.8 million under Option C (3 percent discount rate). The negative values indicate that all regulatory
options result in net forgone benefits.
6.2 Limitations and Uncertainties
Table 6-3 summarizes the limitations and uncertainties in the analysis of benefits associated with changes in
surface water quality and indicates the direction of any potential bias. Note that the effect on benefits
estimates indicated in the second column of the table refers to the magnitude of the benefits rather than the
6-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
direction (i.e.. a source of uncertainty that tends to underestimate benefits indicates expectation for larger
forgone benefits or for larger realized benefits).
Table 6-3: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
Use of 100-mile buffer
for calculating water
quality benefits for each
CBG
Underestimate
The distance between the surveyed households and the affected
waterbodies is not well measured by any of the explanatory variables
in the meta-regression model. EPA would expect values for water
quality changes to diminish with distance (all else equal) between the
home and affected waterbody. The choice of 100 miles is based on
typical driving distance to recreational sites (i.e., 2 hours or 100 miles).
Therefore, EPA used 100 miles to approximate the distance decay
effect on WTP values. The analysis effectively assumes that people
living farther than 100 miles place no value on water quality
improvements for these waterbodies despite literature that shows
that while WTP tends to decline with distance from the waterbody,
people place value on the quality of waters outside their region.
Selection of the WQI
parameter value for
estimating low and high
WTP values
Uncertain
EPA set AWQI to 5 and 50 units to estimate low and high benefit
values using Model 2. These values were based on the lowest and
highest water quality changes included in the meta-data. To the
extent that AWQI = 50 is significantly larger than the change in water
quality expected from the regulatory options, it is likely to significantly
understate the estimated WTP value in absolute terms. AWQI = 5 is
more consistent with the magnitude of water quality changes
resulting from the regulatory options.
Potential hypothetical
bias in underlying stated
preference results
Uncertain
Following standard benefit transfer approaches, this analysis proceeds
under the assumption that each source study provides a valid,
unbiased estimate of the welfare measure under consideration (cf.
Moeltner et al., 2007; Rosenberger and Phipps, 2007). To minimize
potential hypothetical bias underlying stated preference studies
included in meta-data, EPA set independent variable values to reflect
best benefit transfer practices.
Use of different water
quality measures in the
underlying meta-data
Uncertain
The estimation of WTP may be sensitive to differences in the
environmental water quality measures across studies in the meta
data. Studies that did not use the WQI were mapped to the WQI so a
comparison could be made across studies. In developing the 2015
meta-regression models, EPA tested a binary variable (WQI) that
captures the effect of a study using (WQI=1) or not using (WQI=0) the
WQI. The variable coefficient was not statistically different from zero,
indicating no evidence of systematic bias in the mapping of studies
that did not use the WQI. However, the 2020 update of the meta-
regression, which added 14 new studies to the 2015 meta-data,
accounts for potential effects of the use of a different water quality
metric (i.e., index of biotic integrity (IBI)) on the interpretation of the
baseline and water quality and improvements (see Appendix G for
details).
6-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
6: Nonmarket Benefits
Table 6-3: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
Transfer error
Uncertain
Transfer error may occur when benefit estimates from a study site are
adopted to forecast the benefits of a policy site. Rosenberger and
Stanley (2006) define transfer error as the difference between the
transferred and actual, generally unknown, value. While meta-analysis
is fairly accurate when estimating benefit function, transfer error may
be a problem in cases where the sample size is small. Meta-analyses
have been shown to outperform other function-based transfer
methods in many cases, but this result is not universal (Shrestha et ai,
2007). This notwithstanding, meta-analyses results are "very
promising" for the performance of meta-analytic benefit transfers
relative to alternative transfer methods (Rosenberger and Phipps,
2007).
Omission of Great Lakes
and estuaries from
analysis of benefits from
water quality changes
Underestimate
Six out of 112 (5 percent) steam electric power plants discharge to the
Great Lakes or estuaries. Due to limitations of the water quality
models used in the analysis of the regulatory options, these
waterbodies were excluded from the analysis. This omission likely
underestimates benefits of water quality changes from the regulatory
options.
The water quality model
accounts for only a
subset of sources of
toxic pollutants
contributing to baseline
concentrations
Uncertain
The overall impact of this limitation on the estimated WTP for water
quality changes is uncertain but is expected to be small since the
estimated WTP is a function of a mid-point between the baseline and
post-technology implementation water quality. Therefore, the
difference in WTP between the baseline and post-technology
implementation would be more sensitive to the estimated water
quality changes.
6-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
7 Impacts and Benefits to Threatened and Endangered Species
7.1 Introduction
T&E species are species vulnerable to future extinction or at risk of extinction in the near future, respectively.
These designations reflect low or rapidly declining population levels, loss of essential habitat, or life history
stages that are particularly vulnerable to environmental alteration or other stressors. In many cases, T&E
species are given special protection due to inherent vulnerabilities to habitat modification, disturbance, or
other impacts of human activities. This chapter examines the projected change in environmental impacts of
steam electric power plant discharges on T&E species and the estimated benefits associated with the projected
changes resulting from the regulatory options.
As described in the 2015 EA and in the 2020 Supplemental EA (U.S. EPA, 2015b, 2020f), the untreated
chemical constituents of steam electric power plant wastestreams can pose serious threats to ecological health
due to the bioaccumulative nature of many pollutants, high concentrations, and high loadings. Pollutants such
as selenium, arsenic and mercury have been associated with fish kills, disruption of growth and reproductive
cycles and behavioral and physiological alterations in aquatic organisms. Additionally, high nutrient loads can
lead to the eutrophication of waterbodies. Eutrophication can lead to increases in the occurrence and intensity
of water column phytoplankton, including harmful algal blooms (e.g., nuisance and/or toxic species), which
have been found to cause fatal poisoning in other animals, fish, and birds. Eutrophication may also result in
the loss of critical submerged rooted aquatic plants (or macrophytes), and reduced DO levels, leading to
anoxic or hypoxic waters.
For species vulnerable to future extinction, even minor changes to growth and reproductive rates and small
levels of mortality may represent a substantial portion of annual population growth. To quantify the estimated
effects of the regulatory options compared to baseline, EPA conducted a screening analysis using changes in
projected attainment of freshwater NRWQC as an indicator. Specifically, EPA identified the reaches that are
projected to see changes in achievement of freshwater aquatic life NRWQC as a consequence of the
regulatory options, assuming no more stringent controls are established to meet applicable water quality
standards (i.e.. water-quality-based effluent limits issued under Section 301(b)(1)(C)), relative to the
baseline. Using these projections, EPA then estimated the number of T&E species whose recovery could be
affected based on the species' habitat range. Because NRWQC are recommended at levels to protect aquatic
organisms, reducing the frequency at which aquatic life-based NRWQC are exceeded could translate into
reduced risk to T&E species and potential improvements in species populations.66 Conversely, increasing the
frequency of exceedances may increase risk to T&E species and decrease their survival or recovery.
In this chapter, EPA examines the current conservation status of species belonging to freshwater taxa and
identifies the extent to which the regulatory options, independent of consideration of water quality-based
controls, may benefit or adversely impact T&E species. Specifically, EPA estimated the changes in potential
impacts of steam electric power plant discharges on surface waters intersecting habitat ranges of T&E species,
to provide a quantitative, but unmonetized proxy for the benefits associated with the regulatory options.
66 Criteria are developed based on the 1985 Guidelines methods (U.S. EPA, 1985) and generally reflect high quality toxicity data
from at least 8 different taxa groups that broadly represent aquatic organisms. To the extent that more stringent levels are
required to protect organisms in a particular location, that is addressed during the water quality standard development process for
that location.
7-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
The analysis generally follows the approach EPA used in 2015 and 2019 (U.S. EPA, 2015a, 2019a).
However, in response to comments EPA received on the 2019 proposal, EPA updated inputs for the analysis
from the critical habitat range data used in 2019 to the most current total habitat range data from the U.S.
FWS in order to provide a more comprehensive assessment of the potential overlap between species ranges
and affected reaches. EPA has also provided additional details on the identified overlaps (see Appendix H)
and revised its description of the analysis below to clarify the methodology, assumptions, and inputs.
In general, the analysis shows the estimated effects of the final rule, Option A, on T&E species to be small
compared to baseline (see U.S. EPA, 2015a).
7.2 Baseline Status of Freshwater Fish Species
Reviews of aquatic species' conservation status over the past three decades have documented the effect of
cumulative stressors on freshwater aquatic ecosystems, resulting in a significant decline in the biodiversity
and condition of indigenous communities (Deacon etal., 1979; J. E. Williams etal., 1989; J. D. Williams el
al., 1993; Taylor etal., 1996; Taylor etal., 2007; Jelks et al., 2008). Overall, aquatic species are
disproportionately imperiled relative to terrestrial species. For example, while 39 percent of freshwater and
diadromous fish species are imperiled (Jelks et al., 2008), a similar status review found that only 7 percent of
North American bird and mammal species are imperiled (Wilcove & Master, 2005). Recent studies of threats
and extinction trends in freshwater taxa also concluded that biodiversity is much more at risk in freshwater
compared to marine ecosystems (Winemiller, 2018).
Approximately 39 percent of described fish species in North America are imperiled, with 700 fish taxa
classified as vulnerable (230), threatened (190), or endangered (280) in addition to 61 taxa presumed extinct
or functionally extirpated from nature (Jelks et al., 2008). These data show that the number of T&E species
have increased by 98 percent and 179 percent when compared to similar reviews conducted by the American
Fisheries Society in 1989 (J. E. Williams et al., 1989) and 1979 (Deacon et al., 1979), respectively. Despite
recent conservation efforts, including the listing of several species under the Endangered Species Act (ESA),
only 6 percent of the fish taxa assessed in 2008 had improved in status since the 1989 inventory (Jelks et al.,
2008).
Several families of fish have high proportions of T&E species. Approximately 46 percent and 44 percent of
species within families Cyprinidae (carps and true minnows) and Percidae (darters and perches) are imperiled,
respectively. Some families with few, wide-ranging species have even higher rates of imperilment, including
the Acipenseridae (sturgeons; 88 percent) and Polyodontidae (paddlefish; 100 percent). Families with species
important to sport and commercial fisheries have imperilment levels ranging from a low of 22 percent for
Centrarchidae (sunfishes) to a high of 61 percent for Salmonidae (salmon) (Jelks et al., 2008).
7.3 T&E Species Potentially Affected by the Regulatory Options
To assess the potential effects of the regulatory options on T&E species, EPA used the U.S. FWS
Environmental Conservation Online System (ECOS) to construct a database to analyze which species have
habitats that overlap with waters projected to improve or degrade due to changes in pollutant discharge from
steam electric power plants. The database includes all animal species currently listed or under consideration
for listing under the ESA (U.S. FWS, 2020d).
7.3.1 Identifying T&E Species Potentially Affected by the Regulatory Options
To estimate the effects of the regulatory options on T&E species, EPA first compiled data on habitat ranges
for all species currently listed or under consideration for listing under the ESA. EPA obtained the
7-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
geographical distribution of T&E species in geographic information system (GIS) format from ECOS (U.S.
FWS, 2020b).
EPA constructed a screening database using the spatial data on species habitat ranges and all NHD reaches
downstream from steam electric power plants. This database included all T&E species whose habitat ranges
intersect reaches immediately receiving or downstream of steam electric power plant discharges. EPA used a
200-meter buffer on either side of each reach when estimating the intersection to account for waterbody
widths and any minor errors in habitat maps. This initial analysis identified a total of 197 T&E species.
EPA then classified these species on the basis of their vulnerability to changes in water quality for the purpose
of assessing potential impacts of the regulatory options. EPA obtained species life history data from a wide
variety of sources to assess T&E species" vulnerability to water pollution. For the purpose of this analysis,
species were classified as follows:
• Higher vulnerability - species living in aquatic habitats for several life history stages and/or species
that obtain a majority of their food from aquatic sources.
• Moderate vulnerability - species living in aquatic habitats for one life history stage and/or species that
obtain some of their food from aquatic sources.
• Lower vulnerability - species whose habitats overlap bodies of water, but whose life history traits and
food sources are terrestrial.
Table 7-1 summarize the results of this assessment. Appendix H lists all T&E species whose habitat ranges
intersect reaches immediately receiving or downstream of steam electric power plant discharges.
Table 7-1: Number of T&E Species with Habitat Range Intersecting Reaches Immediately Receiving
or Downstream of Steam Electric Power Plant Discharges, by Group
Species Group
Species Vulnerability

Lower
Moderate
Higher
Species Count
Amphibians
3
2
3
8
Arachnids
6
0
0
6
Birds
18
6
1
25
Clams
0
0
62
62
Crustaceans
0
2
3
5
Fishes
0
0
35
35
Insects
9
0
1
10
Mammals
14
1
1
16
Reptiles
15
1
3
19
Snails
2
0
9
11
Total
67
12
116
197
Source: U.S. EPA Analysis, 2020.
To estimate the potential impacts of the regulatory options, EPA focused the analysis on species with higher
vulnerability potentials based upon life history traits. EPA's further review of this subset of species resulted in
the removal from further analysis of those species endemic to isolated headwaters and natural springs, as
these waters are unlikely to receive steam electric power plant discharges in the scope of the final rule (see
Appendix H for details). Review of life history data for the remaining species shows pollution or water quality
7-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
issues as one of the factors influencing species decline. This suggests that water quality issues may be
important to species recovery even if not listed explicitly in species recovery plans.
7.3.2 Estimating Effects of the Rule on T&E Species
EPA used the results of the water quality model described in Chapter 3 to flag those reaches where estimated
pollutant concentrations exceed the freshwater NRWQC under the baseline or the regulatory options (see
Section 3.4.1.1). EPA estimated exceedances for two distinct periods (2021-2028 and 2029-2047) within the
overall analysis period (2021-2047). As described in Section 3.2.1, Period 1 corresponds to the years when
the steam electric power plants would be transitioning to treatment technologies to comply with the revised
limits, whereas Period 2 reflects post-technology implementation conditions when all plants meet applicable
revised limits.
EPA then linked the water quality model outputs with the species database described in the section above to
identify potentially "affected T&E species habitats" where the water quality analysis shows changes under the
regulatory options, meaning either: 1) the reaches intersecting the habitat range of a T&E species meet the
NRWQC under baseline conditions but do not meet the NRWQC under one or more of the regulatory options
(i.e., potential forgone benefits); or 2) the reaches intersecting the habitat range of a T&E species do not meet
the NRWQC under baseline conditions but do meet the NRWQC under one or more of the regulatory options
(i.e., potential positive benefits). EPA compared dissolved concentration estimates for eight pollutants to the
freshwater acute and chronic NRWQC values67 to assess the exceedance status of the reaches under the
baseline and each regulatory option. The first condition occurs in a subset of reaches during Period 1, whereas
the second condition is met for a subset of reaches during Period 2.
EPA identified a total of five species, listed in Table 7-2, whose habitat ranges intersect reaches that show
changes in NRWQC exceedance status under the regulatory options during Period 1 and/or Period 2.
Table 7-2: Higher Vulnerability T&E Species with Habitat Intersecting Waters with Estimated
Changes in NRWQC Exceedance Status under the Regulatory Options, Compared to Baseline
Species Group
Species Count
Species
Common Name
Clams
1
Pleurobema clava
Clubshell
Fishes
3
Etheostoma trisella
Trispot darter
Gila cypha
Humpback chub
Ptychocheilus lucius
Colorado pikeminnow (squawfish)
Mammals
1
Trichechus manatus
West Indian Manatee
Total
5

Source: U.S. EPA Analysis, 2020.
Table 7-3 and Table 7-4 summarizes changes in exceedance status for Period 1 and Period 2, respectively.
EPA's analysis shows that in Period 1 (2021-2028) a total of seven reaches within the habitat range of four
T&E species have projected water quality degradation compared to baseline, as indicated by the NRWQC
exceedance status. EPA estimated that no reaches would newly exceed the aquatic life NRWQC in Period 2
(2029-2047) under any of the regulatory options, when compared to the baseline. Further, the Period 1
exceedances in Table 7-3 are not present in Period 2.
07 The eight pollutants are arsenic, cadmium, copper, lead, mercury, nickel, selenium, and zinc. For more information about the
aquatic life NRWQC, see Table C-7 in the Supplemental EA (U.S. EPA, 2020f).
7-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
EPA's analysis also indicates that three reaches that intersect habitat ranges of one T&E species (trispot
darter) exceed NRWQC under the baseline conditions. The baseline exceedances are present in Period 1
(2021-2028) under all options, whereas water quality improvements in Period 2 (2029-2047) in three reaches
under all regulatory options eliminate the estimated baseline exceedances and result in this species potentially
benefiting from all three regulatory options (see Table 7-4).
Table 7-3: Higher Vulnerability T&E Species Whose Habitat May be Affected by the Regulatory
Options Compared to Baseline in Period 1
Species Common Name
State
Number of Reaches with NRWQC Exceedances3 for at
Least One Pollutant in Period 1 (2021-2028)
Baseline
Option A
(Final Rule)
Option B
Option C
Clubshell
KY
0
1
1
1
Colorado pikeminnow (=squawfish)
WY
0
5
5
5
Humpback chub
WY
0
5
5
5
West Indian Manatee
SC
0
0
0
1
Number of unique reaches with NRWQC exceedances
0
6
6
7
a. Exceedance counts are based on comparison of dissolved pollutant concentrations to NRWQC. Option D exceedances based on
total pollutant concentrations are summarized in U.S. EPA (2019a).
Source: U.S. EPA Analysis, 2020.
Table 7-4: Higher Vulnerability T&E Species Whose Habitat May be Affected by the Regulatory
Options Compared to Baseline in Period 2
Species Common Name
State
Number of Reaches with NRWQC Exceedances3 for at
Least One Pollutant in Period 2 (2029-2047)3
Baseline
Option A
(Final Rule)
Option B
Option C
Trispot darter
GA
3
0
0
0
Number of unique reaches with NRWQC exceedances
3
0
0
0
a. Exceedance counts are based on comparison of dissolved pollutant concentrations to NRWQC. Option D exceedances based on
total pollutant concentrations are summarized in U.S. EPA (2019a).
Source: U.S. EPA Analysis, 2020.
7.4 Limitations and Uncertainties
One limitation of EPA's analysis of the regulatory options" impacts on T&E species and their habitat is the
lack of data necessary to quantitively estimate population changes of T&E species and to monetize these
effects. The data required to estimate the response of T&E species populations to improved habitats are rarely
available. In addition, understanding the contribution of T&E species to ecosystem functions can be
challenging because: (1) it is often difficult to detect the location of T&E species, (2) experimental studies
including rare or threatened species are limited; and (3) ecologists studying relationships between biodiversity
and ecosystem functions typically focus on overall species diversity or estimate species contribution to
ecosystem functions based on abundance (Dee etal., 2019). Finally, much of the wildlife economic literature
focuses on recreational benefits that are not relevant for many protected species (i.e.. use values) and the
existing T&E valuation studies tend to focus on species that many people consider to be "charismatic" (e.g.,
spotted owl, salmon) (Richardson & Loomis, 2009). Although a relatively large number of economic studies
have estimated WTP for T&E protection, these studies focused on estimating WTP to avoid species
loss/extinction, reintroduction, increase in the probability of survival, or a substantial increase in species
7-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
7: Threatened & Endangered Species Benefits
population (Richardson and Loomis, 2006). In addition, Richardson and Loomis (2006) developed a meta-
analysis of 31 stated preference studies valuing a variety of threatened, rare, or endangered species that allow
estimation of WTP for avoiding species loss or changes in species population levels. However, use of this
meta-regression model is not feasible for this analysis due to the challenges associated with estimating T&E
population changes from the final rule. Table 7-5 summarizes limitations and uncertainties known to affect
EPA's assessment of the impacts of the final rule on T&E species. Note that the effect on benefits estimates
indicated in the second column of the table refers to the magnitude of the benefits rather than the direction
(i.e.. a source of uncertainty that tends to underestimate benefits indicates expectation for larger forgone
benefits or for larger realized benefits).
Table 7-5: Limitations and Uncertainties in the Analysis of T&E Species Impacts and Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
The analysis does not
account for water quality
based effluent limits
Overestimate
This screening analysis is intended to isolate possible effects of
the regulatory options on T&E species, however, it does not
take into account the fact that the NPDES permits for each
steam electric power plant, like all NPDES permits, are required
to have limits more stringent than the technology-based limits
established by an ELG wherever necessary to protect water
quality standards. Because this analysis does not project where
a permit will have more stringent limits than those required by
the ELG, it may overestimate any negative impacts to T&E
species, including impacts that will not be realized because the
permits will be written to include limits as stringent as
necessary to meet water quality standards as required by the
CWA.
Intersection of T&E species
habitat with reaches affected
by steam electric plant
discharges is used as proxy
for exposure to steam
electric pollutants
Overestimate
EPA used the habitat range as the basis for assessing the
potential for impacts to the species from water quality
changes. This approach is reasonable given the lack of reach-
specific population data to support a national-level analysis,
but the Agency acknowledges that the habitat range of a
species does not necessarily indicate that the species is found
in individual reaches within the habitat range.
The change in T&E species
populations due to the effect
of the regulatory options is
uncertain
Uncertain
Data necessary to quantitatively estimate population changes
are unavailable. Therefore, EPA used the methodology
described in Section 7.3.1 as a screening-level analysis to
estimate whether the regulatory options could contribute to a
change in the recovery of T&E species populations.
Only those T&E species listed
as threatened or endangered
under the ESA are included
in the analysis
Underestimate
The databases used to conduct this analysis include only
species protected under the ESA. Additional species may be
considered threatened or endangered by scientific
organizations but are not protected by the ESA (e.g., the
American Fisheries Society [J. D. Williams et ai, 1993; Taylor et
al., 2007; Jelks et al., 2008]). The magnitude of the
underestimate is unknown. Although the proportion of
imperiled freshwater fish and mussel species is high (e.g., Jelks
et al., 2008; Taylor et al., 2007) the geographic distribution of
these species may or may not overlap with reaches affected by
steam electric discharges.
7-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 7: Threatened & Endangered Species Benefits
Table 7-5: Limitations and Uncertainties in the Analysis of T&E Species Impacts and Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
The potential for impact to
T&E species is also present
for changes in pollutant
concentrations that don't
result in changes in NRWQC
exceedances
Underestimate
EPA's analysis quantifies changes in whether a NRWQC is
exceeded in a given reach that intersects T&E species habitat
ranges. However, changes in pollutant concentrations (either
positive or negative) have the potential to result in impacts to
T&E species even where they do not result in changes in
NRWQC exceedance status. There are also potential impacts to
T&E species from changes in pollutants for which freshwater
NRWQC are not available (e.g., salinity).
EPA's water quality model
does not capture all sources
of pollutants with a potential
to impact aquatic T&E
species
Uncertain
EPA's water quality model focuses on toxic pollutant discharges
from steam electric power plants and certain other point
sources, but does not account for other pollution sources (e.g.,
historical contamination) or background levels. Adding these
other sources or background levels could result in additional
NRWQC exceedances under the baseline and/or regulatory
options, but it is uncertain how the regulatory options would
change the exceedance status of the intersected reaches.
7-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
8 Air Quality-Related Benefits
The regulatory options evaluated may affect air quality through three main mechanisms: 1) changes in energy
used by steam electric power plants to operate wastewater treatment, ash handling, and other systems needed
to meet the limitations and standards under the regulatory options; 2) transportation-related emissions due to
the changes in trucking of CCR and other waste to on-site or off-site landfills; and 3) changes in the
electricity generation profile from changes in wastewater treatment costs (and savings compared to the
baseline) and the resulting changes in EGU relative operating costs. With respect to the third mechanism, the
Integrated Planning Model (IPM) projects a 0.6 percent increase in electricity generation from coal in 2030
(+4,699 GWh) under the final rule compared to baseline. Because electricity demand is constant, this increase
is offset by a 0.2 percent decline in generation from natural gas (-5,695 GWh) and renewable sources
(-1,726 GWh), and a 0.4 percent increase in nuclear power generation (+2,292 GWh). See details in Chapter 5
of the RIA (U.S. EPA, 2020d). The changes in air emissions reflect the differences in EGU emissions factors
for these other fuels or sources of energy, as compared to coal.
EPA estimated the climate-related benefits of changes in CO2 emissions, as well as the human health benefits
resulting from net changes in emissions of NOx, SO2, and directly emitted fine particulate matter (PM2 5), also
referred to as primary PM2 5 emissions.
CO2 is the most prevalent of the greenhouse gases, which are air pollutants that EPA has determined endanger
public health and welfare through their contribution to climate change. EPA used estimates of the domestic
social cost of carbon (SC-CO2) to monetize the benefits of changes in CO2 emissions as a result of the final
rule The SC-CO2 is a metric that estimates the monetary value of projected impacts associated with marginal
changes in CO2 emissions in a given year. It includes a wide range of anticipated climate impacts, such as net
changes in agricultural productivity and human health, property damage from increased flood risk, and
changes in energy system costs, such as reduced costs for heating and increased costs for air conditioning.
PM2 5 air pollution has been associated with a variety of adverse health effects detailed in the Integrated
Science Assessments for Particulate Matter (PM ISA), including premature mortality and a variety of
morbidity effects associated with acute and chronic exposures (U.S. EPA, 2009a, 2019e). In addition to
primary PM2 5 emitted directly by electricity generating units and other sources, NOx and SOx (which include
SO2 emissions quantified in this analysis) are known precursors to PM2 5 air pollution. In addition, in the
presence of sunlight, NOx and volatile organic compounds (VOCs) can undergo a chemical reaction in the
atmosphere to form ozone (O3). Depending on localized concentrations of VOCs, changes in NOx emissions
also change human exposure to ozone. EPA's Integrated Science Assessments for Ozone and Related
Photochemical Oxidants (Ozone ISA) identify a variety of potential health effects associated with acute and
chronic ozone exposures, including premature mortality and a variety of morbidity effects (U.S. EPA, 2013d,
2020b). For the purpose of this analysis, EPA performed gridded photochemical air quality modeling and
quantified the health benefits attributable to changes in PM2 5 and ground-level ozone.68 This BCA follows
EPA's recent practice which has been to estimate the impact on total non-accidental premature mortality
associated with the change in ozone exposure. However, the 2020 Ozone ISA concludes that the currently
available evidence for cardiovascular effects and total mortality is suggestive of, but not sufficient to infer, a
causal relationship with short-term (as well as long-term) ozone exposures (U.S. EPA, 2020b, sections
Changes in emissions of SO2 and NOx would also change ambient exposure to SO2 and NO2, respectively, but EPA did not
quantify health effects from these exposures.
8-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
IS.4.3.4 and IS.4.3.5). As such, EPA is in the process of recalibrating its benefits estimates to model only
premature mortality from respiratory causes (/'. e., non-respiratory causes of premature mortality associated
with ozone exposure would no longer be estimated). Until a replacement method that only estimates the
benefits associated with respiratory causes of premature mortality has been developed, EPA will be removing
the estimate of the impact of reduced ozone exposure on premature mortality from its benefits estimates from
subsequent rulemakings.
For the analysis of the 2015 rule, the EPA relied on estimates of national monetized benefits per ton of
emissions avoided, which represented the total monetized human health benefits from changes in the adverse
outcomes mentioned above (U.S. EPA, 2015a). For the proposed rule, the Agency quantified, but did not
monetize, changes in emissions of PM2 5 precursors NOx and SO2. For this final rule, EPA leveraged available
photochemical modeling outputs that were created as part of the ACE rule RIA (U.S. EPA, 2019h). The full-
scale modeling used in this analysis included annual model simulations for a 2011 base year and a 2023 future
year to provide hourly concentrations of ozone and primary and secondarily formed PM2 5 component species
(e.g., sulfate, nitrate, ammonium, elemental carbon (EC), organic aerosol (OA), and crustal material) for both
years nationwide. EPA tracked the impact of specific emissions sources on ozone and PM2 5 in the 2023
modeled case using a tool called "source apportionment." This air quality modeling approach provides
spatially explicit estimates of concentration changes, which is required for characterizing uncertainty in
mortality risk from changes in PM2 5 concentrations at different levels of baseline PM2 5 exposure. This is an
important improvement in analytic method compared to the 2019 proposal and the 2015 rule because the air
quality model is based on finer than national resolution of data where impacts are broken out based on state-
level emission information and coal versus non-coal emissions for a subset of the fleet. This modeling also
takes into account elemental carbon, organic carbon and crustal emissions, rather than elemental carbon and
organic carbon used in the benefits per ton approach.
In addition, this air quality modeling also used a 2011 emission inventory projected to 2023 that reflects the
current fleet of coal-fired power plants. The benefits-per-ton approach used a 2005 emission inventory
projected to 2016. Changes in the location and emissions of facilities occurring between 2005 and 2011
would affect the size and distribution of PM changes, which will in turn affect the size of the estimated
benefits. These differences in data and modeling can result in substantial differences in estimates of PM
benefits or disbenefits.
As such, EPA will continue its current efforts to evaluate the usefulness of Reduced Form Tools (RFT),
including a benefits per ton approach, in regulatory impact analysis and how they compare to Full Form
Models (FFM). The areas of further evaluation between FFMs and RFTs would include for example,
comparing the effect of differences in emissions including speciation of emissions (e.g., crustal emissions
versus elemental and organic carbon emissions), the impact of differences in model and data resolution,
among other relevant areas, on the results from FFMs and RFTs.
The regulatory options evaluated may also affect air quality through changes in emissions of larger particulate
matter (PM10) and hazardous air pollutants (HAP) including Hg and HC1. The effects of mercury are detailed
in the Supplemental EA (U.S. EPA, 2020f). HC1 is a corrosive gas that can cause irritation of the mucous
membranes of the nose, throat, and respiratory tract.
The following sections summarize the estimated changes in air emissions, describe the modeling and
quantification methods, and present estimated benefits for two categories of benefits: climate change
(Section 8.2) and human health (Section 8.3). More details about the methodology used to value benefits of
8-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
CO2 changes and to model air quality changes can be found in Appendix I and Appendix J, respectively.
Section 8.4 presents total annualized air benefits.
Section 8.5 summarizes major limitations and sources of uncertainty in the analysis of air quality-related
benefits. Data, resource, and methodological limitations prevent EPA from estimating all domestic climate
benefits and health and environmental benefits, including those from health effects from direct exposure to
SO2, NO2, PM10, and HAP, and ecosystem effects and visibility impairment. Chapter 2 discusses these
unquantified effects.
In general, the analysis shows the estimated effects of the final rule on air quality to be smaller than those
estimated for the baseline (see U.S. EPA, 2015a).
8.1 Changes in Air Emissions
As discussed in the 1(1 A. EPA used IPM to estimate the electricity market-level effects of the final rule
(Option A; see Chapter 5 in RIA [U.S. EPA, 2020d]). IPM outputs include estimated C02,N0x, SO2, Hg, and
HC1 emissions to air from EGUs. EPA also used IPM outputs to estimate EGU emissions of primary PM2.5
and PM10 based on emission factors described in U.S. EPA (2020a). Specifically, EPA estimated primary
PM2.5 and PM10 emissions by multiplying the generation predicted for each IPM plant type (ultrasupercritical
coal without carbon capture and storage, combined cycle, combustion turbine, etc.) by a type-specific
empirical emission factor derived from the 2016 National Emissions Inventory (NEI) and other data sources.
The emission factors reflect the fuel type (including coal rank), FGD controls, and state emission limits for
each plant type, where applicable.
Comparing emissions projected under Option A to those projected for the baseline provides an assessment of
the changes in air emissions resulting from changes in the profile of electricity generation under the final
rule.69 EPA used seven run years, shown in Table 8-1, to represent the 2021-2047 period of analysis.
Table 8-1: IPM Run Years
IPM Run Year
Years Represented
2021
2021
2023
2022-2023
2025
2024-2027
2030
2028-2032
2035
2033-2037
2040
2038-2042
2045
2043-2047
Source: U.S. EPA, 2018b
As part of its analysis of non-water quality environmental impacts, EPA developed separate estimates of
changes in energy requirements for operating wastewater treatment systems and ash handling systems, and
changes in transportation needed to landfill solid waste and CCR (see Supplemental TDD for details; U.S.
EPA, 2020g). EPA estimated NOx, SO2, and CO2 emissions associated with changes in energy requirements
to power wastewater treatment systems by multiplying plant-specific changes in electricity consumption by
plant- or North American Electric Reliability Corporation (NERC)-specific emission factors obtained from
09 While EPA only ran IPM for the final rule (Option A), the Agency extrapolated the benefits estimated using these IPM outputs to
Options B and C to provide insight on the potential air quality-related effects of the other regulatory options. See Section 8.4 for
details.
8-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
IPM for each run year. EPA estimated air emissions associated with changes in transportation by multiplying
the number of miles traveled by average emission factors.
Table 8-2 summarizes the estimated changes in emissions associated with changes in power requirements to
operate treatment systems and with the transportation of CCR and solid waste under the regulatory options.
EPA estimates that changes in power requirements and transportation would result in a decrease in emissions
under Options A and B, and an increase in emissions under Option C. These values reflect full technology
implementation under the regulatory options, which is projected to occur by the end of 2028.70
Table 8-2: Estimated Changes in Air Pollutant Emissions Due to Increase in Power Requirements and
Trucking at Steam Electric Power Plants 2021-2047, Compared to Baseline
Regulatory Option
C02 (Million
Tons/Year)3
NOx (Thousand
Tons/Year)3
S02 (Thousand
Tons/Year)3
Primary PM2.5
(Thousand
Tons/Year)3
Option Db
-0.073
-0.049
-0.082
Not estimated
Power Requirements
Option A (Final Rule)
-0.023
-0.013
-0.016
Not estimated
Option B
-0.020
-0.012
-0.015
Not estimated
Option C
0.180
0.079
0.063
Not estimated
Transportation
Option A (Final Rule)
-0.0098
-0.0090
-0.000082
Not estimated
Option B
-0.0098
-0.0090
-0.000082
Not estimated
Option C
-0.0080
-0.0073
-0.000067
Not estimated
a. Values rounded to two significant figures. Negative values indicate a reduction in emissions and positive values indicate an
increase in emissions.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. Values are the net total change
associated with power requirements and transportation (see Table 7-3 in the 2019 Supplemental TDD [U.S. EPA, 2019k])
Source: U.S. EPA Analysis, 2020
Table 8-3 and Table 8-4 summarize the estimated changes in pollutant emissions from electricity generation
under the final rule (i.e., Option A).71 As shown in the two tables, projected changes in the profile of
electricity generation under Option A, compared to the baseline, generally lead to increased pollutant
emissions starting with the 2023 run year. Within this general trend, there are a few run years within the
period of analysis when emissions ofNOx(in 2035 and 2040), SO2(in2023, 2035 and 2040), and primary
PM2 5 (in 2040) decrease relative to the baseline. These changes in air emissions reflect the differences in
emissions factors for coal as compared to other fuels. As presented in the R1A (U.S. EPA, 2020d; see
Section 5.2), IPM projects increases in electricity generation from coal as a result of the final rule
(approximately 0.6 percent in 2030), while decreases are projected for generation from other fuels or energy
sources, specifically natural gas and renewables.
70 For the purpose of this analysis, EPA developed a time profile of air emissions changes based on plants' estimated technology
implementation years during the period of 2021 through 2028.
71 EPA did not run IPM for Options B and C.
8-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-3: Estimated Changes in Annual CO2, NOx, SO2, and Primary PM2.5 Emissions Due to
Changes in Electricity Generation Profile, Compared to Baseline
Regulatory
Option
Year
C02 (Million
Tons/Year)3
NOx (Thousand
Tons/Year)3
S02 (Thousand
Tons/Year)3
Primary PM2.5
(Thousand
Tons/Year)3

2021
-0.079
-0.25
-1.4
-0.028

2023
2.9
3.0
-2.6
0.45
Option A
(Final Rule)
2025
2.2
1.6
-0.70
0.91
2030
2.7
0.69
1.7
0.48
2035
0.88
-0.57
1.8
0.81

2040
1.0
-1.6
-2.9
-0.22

2045
2.8
0.15
0.92
0.44
a. Values rounded to two significant figures. Negative values indicate a reduction in emissions and positive values indicate an
increase in emissions.
Source: U.S. EPA Analysis, 2020; See Chapter 5 in RIAfor details on IPM (U.S. EPA, 2020d).
Table 8-4: Estimated Changes in Annual Primary PM10, Hg and HCI Emissions Due to Changes in
Electricity Generation Profile, Compared to Baseline
Regulatory
Option
Year
Primary PM10 (Thousand
Tons/Year)3
Hg (Tons/Year)3
HCI (Tons/Year)3

2021
0.0035
0.0043
-0.41

2023
0.58
0.015
17
Option A
(Final Rule)
2025
1.1
0.015
15
2030
0.43
0.010
24
2035
0.80
0.0030
11

2040
-0.47
0.0027
14

2045
0.28
0.0018
13
a. Values rounded to two significant figures. Negative values indicate a reduction in emissions and positive values indicate an
increase in emissions.
Source: U.S. EPA Analysis, 2020; See Chapter 5 in RIAfor details on IPM (U.S. EPA, 2020d).
The rest of this chapter quantifies benefits associated with changes in emissions of CO2, SO2, NOx, and
primary PM2 5 Table 8-5 presents the net changes in emissions of these four pollutants for the final rule
across the three mechanisms compared to baseline. The largest effect on projected air emissions is due to the
change in the emissions profile of electricity generation at the market level.
Table 8-5: Estimated Net Changes in Air Pollutant Emissions Due to Changes in Power
Requirements, Trucking, and Electricity Generation Profile, Compared to Baseline
Regulatory
Option
Year
C02 (Million
Tons/Year)3
NOx (Thousand
Tons/Year)3
S02 (Thousand
Tons/Year)3
Primary PM2.5
(Thousand
Tons/Year)3

2021
-0.088
-0.25
-1.4
-0.028

2023
2.9
2.9
-2.6
0.45
Option A
(Final Rule)
2025
2.2
1.6
-0.73
0.91
2030
2.6
0.67
1.6
0.48
2035
0.85
-0.59
1.7
0.81

2040
0.97
-1.6
-3.0
-0.22

2045
2.8
0.12
0.90
0.44
a. Values rounded to two significant figures. Negative values indicate a reduction in emissions and positive values indicate an
increase in emissions.
8-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-5: Estimated Net Changes in Air Pollutant Emissions Due to Changes in Power
Requirements, Trucking, and Electricity Generation Profile, Compared to Baseline
Regulatory
Option
Year
C02 (Million
Tons/Year)3
NOx (Thousand
Tons/Year)3
S02 (Thousand
Tons/Year)3
Primary PM2.5
(Thousand
Tons/Year)3
Source: U.S. EPA Analysis, 2020
8.2 Climate Change Benefits
8.2.1 Data and Methodology
EPA estimated the monetary value of CO2 emission changes using a measure of the domestic social cost of
carbon (SC-CO2). The SC-CO2 is a metric that estimates the monetary value of projected impacts associated
with marginal changes in CO2 emissions in a given year. It includes a wide range of anticipated climate
impacts, such as net changes in agricultural productivity and human health, property damage from increased
flood risk, and changes in energy system costs, such as reduced costs for heating and increased costs for air
conditioning. It is used to assess the change in damages as a result of regulatory actions (e.g., benefits of
rulemakings that lead to an incremental reduction in cumulative global CO2 emissions). The SC-CO2
estimates used in this analysis focus on the projected impacts of climate change that are anticipated to directly
occur within U.S. borders.
The SC-CO2 estimates used in this analysis are interim values developed under EO 13783 for use in
regulatory analyses until an improved estimate of the impacts of climate change to the U.S. can be developed
based on the best available science and economics. EO 13783 directed agencies to ensure that estimates of the
social cost of greenhouse gases used in regulatory analyses "are based on the best available science and
economics" and are consistent with the guidance contained in OMB Circular A-4, "including with respect to
the consideration of domestic versus international impacts and the consideration of appropriate discount rates"
(EO 13783, Section 5(c)). In addition, EO 13783 withdrew the technical support documents (TSDs) used in
the benefits analysis of the 2015 rule for describing the global social cost of greenhouse gas estimates
developed under the prior Administration as no longer representative of government policy.
Regarding the two analytical considerations highlighted in EO 13783 - how best to consider domestic versus
international impacts and appropriate discount rates - current guidance in OMB Circular A-4 is as follows.
Circular A-4 states that analysis of economically significant proposed and final regulations "should focus on
benefits and costs that accrue to citizens and residents of the United States." (OMB, 2003) EPA follows this
guidance by adopting a domestic perspective in the central analysis. Regarding discount rates, Circular A-4
states that regulatory analyses "should provide estimates of net benefits using both 3 percent and 7 percent."
(OMB, 2003) The 7 percent rate is intended to represent the average before-tax rate of return to private capital
in the U.S. economy. The 3 percent rate is intended to reflect the rate at which society discounts future
consumption, which is particularly relevant if a regulation is expected to affect private consumption directly.
EPA follows this guidance by presenting estimates based on both 3 and 7 percent discount rates in the main
analysis. See Appendix I for a discussion of the modeling steps involved in estimating the domestic SC-CO2
estimates based on these discount rates. These SC-CO2 estimates developed under EO 13783 and presented
below will be used in regulatory analysis until more comprehensive domestic estimates are developed, which
would take into consideration recent recommendations from the National Academies of Sciences and
Medicine (2017) to further update the current methodology to ensure that the SC-CO2 estimates reflect the
best available science.
8-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-6T presents the average domestic SC-CO2 estimate across all of the integrated assessment model runs
used to estimate the SC-CO2 for each discount rate for the years 2015 to 2050.72 As with the global SC-CO2
estimates, the domestic SC-CO2 increases overtime because future emissions are expected to produce larger
incremental damages as economies grow and physical and economic systems become more stressed in
response to climate change.
EPA estimated the dollar value of the C02-related effects for each analysis year between 2021 and 2047 by
applying the SC-CO2 estimates, shown in Table 8-6T, to the estimated changes in CO2 emissions in the
corresponding year under the regulatory options. EPA then calculated the present value and annualized
benefits from the perspective of 2020 by discounting each year-specific value to the year 2020 using the same
3 percent and 7 percent discount rates.
Table 8-6: Interim Domestic Social Cost of Carbon Values (2018$/Metric Tonne CO2)
Year
3% Discount Rate, Average
7% Discount Rate, Average
2020
$7
$1
2025
$7
$1
2030
$8
$1
2035
$9
$2
2040
O
t—I
-c/>
$2
2045
O
t—I
-c/>
$2
2050
T—I
T—I
-c/>
$2
Note: These SC-C02 values are stated in $/metric tonne C02 and rounded to the nearest dollar (1 metric tonne
equals 1.102 short tons). The estimates vary depending on the year of C02 emissions and are defined in real
terms, i.e., adjusted for inflation using the GDP implicit price deflator. Values were updated from 2016 dollars to
2018 dollars using the GDP deflator (1.030). EPA interpolated annual values for intermediate years.
Source: U.S. EPA Analysis, 2020 based on U.S. EPA (2019i)
The limitations and uncertainties associated with the SC-CO2 analysis, which were discussed in the 2015 and
2019 BCAs (U.S. EPA, 2015a, 2019a), likewise apply to the domestic SC-CO2 estimates presented in this
chapter. Some uncertainties are captured within the analysis, as discussed in Appendix I, while other areas of
uncertainty have not yet been quantified in a way that can be modeled. For example, limitations include the
incomplete way in which the integrated assessment models capture catastrophic and non-catastrophic impacts,
their incomplete treatment of adaptation and technological change, the incomplete way in which inter-
regional and intersectoral linkages are modeled, uncertainty in the extrapolation of damages to high
temperatures, and inadequate representation of the relationship between the discount rate and uncertainty in
economic growth over long time horizons. The science incorporated into these models understandably lags
behind the most recent research, and the limited amount of research linking climate impacts to economic
damages makes this comprehensive global modeling exercise even more difficult. These individual
limitations and uncertainties do not all work in the same direction in terms of their influence on the SC-CO2
estimates. In accordance with guidance in OMB Circular A-4 on the treatment of uncertainty, Appendix I
provides a detailed discussion of the ways in which the modeling underlying the development of the SC-CO2
72 The SC-CO2 estimates rely on an ensemble of three integrated assessment models (IAMs): Dynamic Integrated Climate and
Economy (DICE) 2010; Climate Framework for Uncertainty, Negotiation, and Distribution (FUND) 3.8; and Policy Analysis of
the Greenhouse Gas Effect (PAGE) 2009.
8-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
estimates used in this BCA addressed quantified sources of uncertainty and presents a sensitivity analysis to
show consideration of the uncertainty surrounding discount rates over long time horizons.
Recognizing the limitations and uncertainties associated with estimating the SC-CO2, the research community
has continued to explore opportunities to improve SC-CO2 estimates. Notably, the National Academies of
Sciences, Engineering, and Medicine conducted a multidiscipline, multi-year assessment to examine potential
approaches, along with their relative merits and challenges, for a comprehensive update to the current
methodology. The task was to ensure that the SC-CO2 estimates that are used in Federal analyses reflect the
best available science, focusing on issues related to the choice of models and damage functions, climate
science modeling assumptions, socioeconomic and emissions scenarios, presentation of uncertainty, and
discounting. In January 2017, the Academies released their final report, "Assessing Approaches to Updating
the Social Cost of Carbon," and recommended specific criteria for future updates to the SC-CO2 estimates, a
modeling framework to satisfy the specified criteria, and both near-term updates and longer-term research
needs pertaining to various components of the estimation process (National Academies of Sciences &
Medicine, 2017).
The Academies' 2017 report also discussed the challenges in developing domestic SC-CO2 estimates, noting
that current integrated assessment models do not model all relevant regional interactions - i.e.. how climate
change impacts in other regions of the world could affect the United States, through pathways such as global
migration, economic destabilization, and political destabilization. The Academies concluded that it "is
important to consider what constitutes a domestic impact in the case of a global pollutant that could have
international implications that impact the United States. More thoroughly estimating a domestic SC-CO2
would therefore need to consider the potential implications of climate impacts on, and actions by, other
countries, which also have impacts on the United States." (National Academies of Sciences & Medicine,
2017, pg. 12-13). In addition to requiring reporting of impacts at a domestic level, Circular A-4 states that
when an agency "evaluate [s] a regulation that is likely to have effects beyond the borders of the United States,
these effects should be reported separately" (OMB, 2003; page 15). This guidance is relevant to the valuation
of damages from CO2 and other greenhouse gases (GHGs), given that GHGs contribute to damages around
the world independent of the country in which they are emitted. Therefore, in accordance with this guidance
in OMB Circular A-4, Appendix I presents the global climate benefits from this final rule using global SC-
CO2 estimates based on both 3 and 7 percent discount rates. EPA did not quantitatively project the full impact
of the final rule on international trade and the location of production, so it is not possible to present analogous
estimates of international costs resulting from the regulatory options. However, to the extent that the
electricity market analysis endogenously models international electricity and natural gas trade (see 1(1 A.
Chapter 5; U.S. EPA, 2020d), and to the extent that affected firms have some foreign ownership, some of the
costs accruing to entities outside U.S. borders is captured in the technology implementation costs presented in
the RIA (U.S. EPA, 2020d).
8-8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
8.2.2 Results
Table 8-7 shows the estimated monetary value of the estimated changes in CO2 emissions in each of several
selected years for Option A, the final rule. Negative values indicate forgone benefits as compared to the
baseline.
Table 8-7: Estimated Domestic Climate Benefits from Changes in CO2 Emissions for Selected Years
under the Final Rule, Compared to Baseline (Millions of 2018$)
Regulatory Option
Year
3% Discount Rate3
7% Discount Rate3

2021
$0.55
$0.08

2025
-$15
-$2.4
Option A
(Final Rule)
2030
-$19
-$3.3
2035
-$6.7
-$1.3
2040
-$8.4
-$1.7

2045
-$26
-$5.3

2047
-$26
-$5.6
a. Values rounded to two significant figures. Negative values indicate forgone benefits (i.e., the number of avoided cases under
the final rule is smaller than in the baseline).
Source: U.S. EPA Analysis, 2020
Table 8-8 shows the total annualized monetary values associated with changes in CO2 emissions for the final
rule by category of emissions. EPA annualized monetary value estimates to enable consistent reporting across
benefit categories (e.g., benefits from improvement in water quality). The annualized values are -$14 million
and -$2.3 million, using discount rates of 3 and 7 percent, respectively. The values are negative, indicating
that the final rule results in forgone benefits when compared to the baseline. The vast majority of the forgone
benefits arise from changes in the profile of electricity generation.
Table 8-8: Estimated Total Annualized Domestic Climate Benefits from Changes in CO2 Emissions
under the Final Rule, Compared to Baseline (Millions of 2018$)
Regulatory Option
Category of Air Emissions
3% Discount Rate3
7% Discount Rate3

Electricity Generation
-$14
-$2.3
Option A
Trucking
$0.07
$0.01
(Final Rule)
Energy use
$0.18
$0.03

Total
-$14
-$2.3
a. Values rounded to two significant figures. Negative values indicate forgone benefits (i.e., the number of avoided cases under the
final rule is smaller than in the baseline).
Source: U.S. EPA Analysis, 2020
8.3 Human Health Benefits
8.3.1 Data and Methodology
As summarized in Table 8-5, the final rule is estimated to influence the level of pollutants emitted in the
atmosphere that adversely affect human health, including directly emitted PM2.5, as well as SO2 and NOx,
which are both precursors to ambient PM2 5 NOx emissions are also a precursor to ambient ground-level
ozone. The change in emissions in turn alters the ambient concentrations, which in turn leads to changes in
8-9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
population exposure. In this document we estimate changes in the human health impacts associated with
PM2 5 and ozone.73
This section summarizes EPA's approach to estimating the incidence and economic value of the potential
PM2.5 and ozone-related benefits estimated for this final rule. The approach entails two major steps: (1)
Developing spatial fields of air quality across the U.S. using nationwide photochemical modeling and related
analyses; and (2) Using these spatial fields in BenMAP-CE to quantify the benefits under Option A as
compared to the baseline.
Under this approach, EPA used IPM projections of EGU air emissions under the baseline and final rule. EPA
then adjusted outputs from this analysis to account for the effects of incremental emissions from
transportation and energy use, and to estimate the benefits of Options B and C. See Section 8.4 for a
description of the methodology for these estimates.
8.3.1.1 Air Quality Modeling Methodology
To create annual PM2 5 and ozone spatial fields representing the baseline and Option A, EPA leveraged
available photochemical modeling outputs that were created as part of the Regulatory Impact Analysis for the
Repeal of the Clean Power Plan, and the Emission Guidelines for Greenhouse Gas Emissions from Existing
Electric Utility Generating Units (U.S. EPA, 2019i), also referred to the Affordable Clean Energy (ACE) rule.
These PM2 5 and ozone spatial fields were used as input to BenMAP-CE which, in turn, was used to quantify
the benefits from this final rule. The analysis supporting this rule used outputs from several full-scale
photochemical model simulations.
EPA prepared spatial fields of air quality for the baseline and the final rule for each of the following health-
impact metrics: annual mean PM2.5, May through September seasonal average 8-hour daily maximum
(MDA8) ozone, and April through October seasonal average 1-hour daily maximum (MDA1) ozone. The
EGU emissions for the baseline and the final rule, consisting of total NOx, SO2, and primary PM2 5 emissions
summarized by year, state, and generation type, were obtained from the outputs of the corresponding IPM
runs, as described in Section 8.1 of this document and Chapter 5 of the RIA (U.S. EPA, 2020d). As such, the
spatial fields do not account for changes in emissions associated with power requirements to operate treatment
systems or transportation. See Section 8.5 regarding limitations and uncertainty associated with this analysis.
The photochemical model simulations as well as the basic methodology for determining air quality changes
are the same as those used in the ACE RIA. Appendix J provides an overview of the air quality modeling and
the methodologies EPA used to develop spatial fields of annual PM2 5 and seasonal ozone concentrations. The
appendix also provides selected figures showing the geographical and temporal distribution of air quality
changes. Additional information on the air quality modeling platform (inputs and set-up), model performance
evaluation for PM2 5 and ozone, emissions processing for this analysis, and additional details and numerical
examples of the methodologies for developing PM25 and ozone spatial fields are available in U.S. EPA
(2019i; Chapter 8) .
EPA used air quality modeling to estimate health benefits associated with changes in particulate matter and
ozone concentrations that may occur because of the final rule relative to the baseline, with the air quality
73 Ambient concentrations of both SO2 and NOx also pose health risks independent of PM2.5 and ozone, though EPA does not
quantify these impacts in this analysis (U.S. EPA, 2016a, 2017c)
8-10
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
modeling baseline including emissions from all sources. Consequently, in addition to rules and economic
conditions included in IPM, the baseline for this analysis included emissions from, and rules for, non-EGU
point sources, on-road vehicles, non-road mobile equipment and marine vessels.74 While the air quality model
includes a range of pollution sources, contributions from non-EGU point sources, on-road vehicles, non-road
mobile equipment and marine vessels are held constant in this analysis, and the only changes are those
associated with the projected impacts of the final rule on the profile of electricity generation and EGU
emissions, as compared to the baseline. The modeled air quality changes do not include other potential effects
of the final rule, such as changes in power requirements to run treatment systems or changes in CCR
transportation, which were estimated separately as described in section 8.1 and were found to be negligible as
described in section 8.4.
8.3.1.2 PM2.5 and Ozone Related Health Impacts
EPA estimated the benefits of the final rule using the open-source environmental Benefits Mapping and
Analysis Program—Community Edition (BenMAP-CE) (Sacks et al., 2018). The procedure for calculating
and valuing air pollution-related impacts is described in detail in Fann et al. (2018), Sacks el al. (2018), and
U.S. EPA (2012).
The BenMAP-CE tool uses health impact functions to quantify excess cases of air pollution-attributable
premature deaths and illnesses. When used to quantify PM2 5- or ozone-related effects, the functions combine
an effect estimates (i.e.. the (3 coefficients) from epidemiological studies, which portray the relationship
between a change in air quality and a health effect, such as mortality, with estimated PM2 5 or ozone
concentrations (supplied using the model simulations described above), population data, and baseline death
rates for each county in each year. The Agency estimates the incidence of air pollution effects for those health
endpoints which the relevant Integrated Science Assessment (ISA) classified as either causal or likely-to-be-
causal. Table 8-9 reports the effects EPA quantified (and monetized) and those the Agency did not quantify.75
74 The air quality modeling techniques used for this analysis reflect non-EGU emissions as of 2023, so implementation or effects of
any changes in non-EGU emissions expected to occur after 2023 are not accounted for in this analysis. However, the effect of
non-EGU emissions on changes in pollution concentrations due to the final rule is likely to be small.
75 EPA is evaluating the adequacy of the PM and O3 National Ambient Air Quality Standards (NAAQS). Once EPA promulgates
final PM and O3 NAAQS, the Agency will revisit its approach for estimating benefits for each pollutant.
8-11
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 8: Air Quality-Related Benefits
Table 8-9: Human Health Effects of Ambient PM2.5 and Ozone
Category
Effect
Effect
Quantified
Effect
Monetized
More
Information
Premature
mortality from
exposure to PM2.5
Adult premature mortality based on cohort study
estimates and expert elicitation estimates (age >25 or age
>30)
V
V
PM ISA
Infant mortality (age <1)
V
V
PM ISA
Morbidity from
exposure to PM2.5
Non-fatal heart attacks (age > 18)
V
V
PM ISA
Hospital admissions—respiratory (all ages)
V
V
PM ISA
Hospital admissions—cardiovascular (age >20)
V
V
PM ISA
Emergency room visits for asthma (all ages)
V
V
PM ISA
Acute bronchitis (age 8-12)
V
V
PM ISA
Lower respiratory symptoms (age 7-14)
V
V
PM ISA
Upper respiratory symptoms (asthmatics age 9-11)
V
V
PM ISA
Exacerbated asthma (asthmatics age 6-18)
V
V
PM ISA
Lost work days (age 18-65)
V
V
PM ISA
Minor restricted-activity days (age 18-65)
V
V
PM ISA
Chronic Bronchitis (age >26)
—
—
PM ISA3
Emergency room visits for cardiovascular effects (all ages)
—
—
PM ISA3
Strokes and cerebrovascular disease (age 50-79)
—
—
PM ISA3
Other cardiovascular effects (e.g., other ages)
—
—
PM ISAb
Other respiratory effects (e.g., pulmonary function, non-
asthma ER visits, non-bronchitis chronic diseases, other
ages and populations)

PM ISAb
Reproductive and developmental effects (e.g., low birth
weight, pre-term births)
—
—
PM ISAbc
Cancer, mutagenicity, and genotoxicity effects
—
—
PM ISAbc
Mortality from
exposure to ozone
Premature mortality based on short-term study estimates
(all ages)
V
V
Ozone ISA
Premature mortality based on long-term study estimates
(age 30-99)
V
V
Ozone ISA3
Morbidity from
exposure to ozone
Hospital admissions—respiratory causes (age > 65)
V
V
Ozone ISA
Emergency department visits for asthma (all ages)
V
V
Ozone ISA
Exacerbated asthma (asthmatics age 6-18)
V
V
Ozone ISA
Minor restricted-activity days (age 18-65)
V
V
Ozone ISA
School absence days (age 5-17)
V
V
Ozone ISA
Decreased outdoor worker productivity (age 18-65)
—
—
Ozone ISA3
Other respiratory effects (e.g., premature aging of lungs)
—
—
Ozone ISAb
Cardiovascular and nervous system effects
—
—
Ozone ISAb
Reproductive and developmental effects
—
—
Ozone ISAb c
a. EPA assesses these benefits qualitatively due to data and resource limitations for this analysis. In other analyses EPA quantified
these effects as a sensitivity analysis.
b. EPA assesses these benefits qualitatively because of insufficient confidence in available data or methods.
c. EPA assesses these benefits qualitatively because current evidence is only suggestive of causality or there are other significant
concerns over the strength of the association.
Source: U.S. EPA Analysis, 2020
After having quantified PM2.5- and ozone-attributable cases of premature death and illness, EPA estimated the
economic value of these cases using WTP and COI measures. For this analysis, EPA used version 1.5.0.4 of
BenMAP-CE (March 2019 release). The Appendix to the BenMAP-CE user manual and the RIA for the
8-12
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Particulate Matter National Ambient Air Quality Standards each detail the sources of the above input
parameters (U.S. EPA, 2012, 2018a).
EPA estimated the number of PM2 5-attributable premature deaths using effect estimates from two
epidemiology studies examining two large population cohorts: the American Cancer Society (Krewski etal.,
2009) and the Harvard Six Cities (Lepeule et al., 2012) cohorts. Consistent with the ACE RIA (U.S. EPA,
2019i), EPA reports the estimated number of PM2 5-attributable deaths according to alternative PM2 5
concentration cutpoints. This approach allows readers to determine the portion of the population exposed to
annual mean PM2 5 levels at or above different concentrations. However, EPA does not view these
concentration cutpoints as thresholds below which there are no quantifiable human health impacts attributable
to PM2 5. EPA reports the ozone-attributable deaths as a range reflecting the intensity of the relationship
between ozone levels and health effects as reflected in concentration-response parameters from R. L. Smith et
al. (2009) on the low end to Jerrett et al., 2009 on the high end.
Projected impacts of the final rule show both decreased and increased levels of PM2 5 and ozone, depending
on the year and location, compared to the baseline (see maps in Appendix J for details). Some portion of the
air quality and health benefits from the final rule occur in areas not attaining the PM2 5 or Ozone National
Ambient Air Quality Standards (NAAQS), the requirements of which should be accounted for in the baseline.
The analysis does not account for possible interactions between NAAQS compliance and the final rule, which
introduces uncertainty into the benefits (and forgone benefits) estimates. If the final rule increases or
decreases primary PM2 5, SO2 and NOx emissions and consequentially PM2 5 and/or ozone concentrations,
these changes may affect compliance with existing NAAQS standards and subsequently affect the actual
benefits (and forgone benefits) of the final rule. For example, in the case of areas that do not meet the
NAAQS that see decreased concentrations of PM2 5 or ozone, states may be able to avoid applying certain
other measures to assure NAAQS attainment. As a result, there would be avoided costs and the PM2 5 and
ozone health and ecological benefits of the final rule would likely be lessened. In areas not attaining the
NAAQS where PM2 5 or ozone concentrations may increase due to the final rule, states may need to identify
additional approaches to reduce emissions from local sources relative to the baseline, thus mitigating any
increased PM2 5 and ozone concentrations. In this case, the health benefits would be higher and there would be
additional social costs associated with these additional approaches.
8.3.2 Results
EPA reports below the estimated number of reduced or increased PM2 5 and ozone-related premature deaths
and illnesses in each year for Option A, the final rule, relative to the baseline along with the 95% confidence
interval (see Table 8-10). The number of reduced or increased estimated deaths and illnesses under the final
rule are calculated from the sum of individual reduced mortality and illness risk across the population in a
given year. Table 8-11 provides the estimated number of avoided or increased PM2 5- related premature deaths
calculated using different approaches to help the reader determine the fraction of PM2 5 attributable deaths
occurring at lower ambient concentrations. Table 8-12 summarizes the dollar value of these impacts for the
final rule across all PM2 5 and ozone-related premature deaths and illnesses, using alternative approaches to
representing and quantifying PM2 5 mortality risk effects. Because total benefits are a function of both
increases and decreases in PM2 5 and ozone exposures depending on the year of analysis, the percentage of
total benefits attributable to reducing PM2 5 exposure may, for example, outweigh the percentage of foregone
benefits attributable to increasing ozone exposure during one year, while the percentage of total forgone
benefits of increasing PM2 5 exposure may outweigh the percentage of benefits attributable to decreasing
ozone exposure in another year.
8-13
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
The alternative approaches to quantifying and presenting mortality risk effects include both different means
for quantifying expected impacts using concentration-response functions over the entire domain of exposure
(i.e.. the no-threshold model) along with different means of presenting impacts by limiting consideration to
only those impacts at exposures above the lowest measured level (LML) or above the NAAQS (Table 8-13).
The estimated number of deaths above and below the LML varies considerably according to the epidemiology
study used to estimate risk. Thus, for four out of seven years analyzed, EPA estimated a larger fraction of
PM2 5-related deaths above the LML of the Krewski et al. (2009) study than the Lepeule et al., 2012 study as
shown in Table 8-13. Likewise, EPA estimated a greater percentage of PM2 5-related deaths below the LML
of the Lepeule et al. (2012) study than the Krewski et al. (2009) study for four out of seven years analyzed.
Table 8-13 also shows a very small percentage of PM2 5-related premature deaths occurring above the
NAAQS in any future year using either of these two studies.
8-14
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-10: Estimated Avoided PM2.5 and Ozone-Related Premature Deaths and Illnesses by Year for the Final Rule, Compared to Baseline
(95% Confidence Interval)
Category
Basis
2021a
2023a
2025a
2030a
2035a
2040a
2045a
Avoided premature death among
adults"

PM2.5
Krewski et al. (2009)
15
(10 to 20)
8
(6 to 11)
3
(2 to 4)
-4
(-5 to -3)
-11
(-15 to -8)
27
(18 to 36)
-8
(-11 to -5)
Lepeule et al. (2012)
34
(17 to 50)
19
(10 to 29)
7
(3 to 11)
-9
(-14 to-5)
-25
(-38 to-13)
61
(30 to 91)
-18
(-27 to -9)
Ozone (03)
Smith et al. (2009)
1
(0 to 1)
-1
(-2 to 0)
-1
(-1 to 0)
0
(0 to 0)
0
(0 to 0)
2
(1 to 3)
1
(0 to 1)
Jerrett et al. (2009)
3
(1 to 5)
-3
(-6 to -1)
-3
(-6 to -1)
0
(0 to 1)
0
(0 to 1)
7
(2 to 11)
2
(1 to 4)
PM2.5-related non-fatal heart attacks
among adults

Peters et al. (2001)
15
(4 to 27)
10
(2 to 17)
5
(1 to 8)
-6
(-10 to -1)
-15
(-26 to -4)
26
(6 to 45)
-12
(-21 to -3)
Pooled estimate
2
(1 to 4)
1
(0 to 3)
0
(0 to 1)
-1
(-2 to 0)
-2
(-4 to -1)
3
(1 to 7)
-1
(-4 to 0)
All other morbidity effects

Hospital admissions—cardiovascular
(PM2.5)
4
(2 to 7)
2
(1 to 4)
1
(0 to 2)
-1
(-2 to -1)
-4
(-6 to -3)
6
(3 to 12)
-3
(-5 to -2)
Hospital admissions—respiratory
(PM2.5 &03)
5
(-2 to 10)
2
(-1 to 3)
0
(0 to 1)
-2
(-2 to -1)
-5
(-6 to -2)
7
(-3 to 13)
-5
(-5 to -4)
ER visits for asthma (PM2.5 & 03)
11
(-2 to 28)
-3
(-27 to 10)
-6
(-22 to 2)
-1
(-5 to 7)
-2
(-10 to 11)
28
(-4 to 72)
4
(-7 to 23)
Exacerbated asthma (PM2.5&O3)
2100
(-1500 to 5100)
-1700
(-5100 to 2700)
-1800
(-4900 to 2100)
630
(-960 to 1900)
350
(-1300 to 1700)
5100
(-3600 to 12000)
1400
(-1900 to 4000)
Minor restricted-activity days (PM2.5&
O3)
13000
(9500 to 17000)
3100
(1900 to 4300)
-1300
(-3200 to 560)
-1600
(-1900 to -1300)
-5300
(-5700 to -4900)
26000
(18000 to 34000)
-1200
(-2400 to -18)
Acute bronchitis (PM2.5)
18
(-4 to 41)
16
(-4 to 36)
7
(-2 to 15)
-6
(-13 to 1)
-14
(-30 to 3)
36
(-9 to 81)
-9
(-21 to 2)
Upper resp. symptoms (PM2.5)
330
(60 to 600)
290
(53 to 530)
120
(21 to 220)
-100
(-190 to-19)
-250
(-450 to -45)
660
(120 to 1200)
-170
(-310 to -31)
Lower resp. symptoms (PM2.5)
230
(89 to 380)
210
(78 to 330)
84
(32 to 140)
-73
(-120 to -28)
-170
(-280 to -65)
460
(180 to 750)
-120
(-190 to -45)
Lost work days (PM2.5)
1700
(1400 to 1900)
1300
(1100 to 1500)
470
(400 to 500)
-540
(-620 to -450)
-1100
(-1300 to -970)
3100
(2600 to 3500)
-820
(-940 to -690)
8-15
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-10: Estimated Avoided PM2.5 and Ozone-Related Premature Deaths and Illnesses by Year for the Final Rule, Compared to Baseline
(95% Confidence Interval)
Category
Basis
2021a
2023a
2025a
2030a
2035a
2040a
2045a
Avoided premature death among
adults"

School absence days (03)
1000
(370 to 2300)
-1300
(-2900 to -460)
-1300
(-2800 to -450)
490
(170 to 1100)
430
(150 to 960)
2600
(930 to 5800)
1100
(370 to 2300)
a. Values rounded to two significant figures. Negative values indicate forgone benefits (i.e., the number of avoided cases under the final rule is smaller than in the baseline). Lower
bound of confidence interval represents the 95% confidence estimate that is lower in value than the point estimate, while upper bound represents the estimate that is higher in value
than the point estimate.
b. EPA also quantified changes in premature infant mortality from exposure to PM2.5 but the estimated change was less than 1 for all years analyzed.
Source: U.S. EPA Analysis, 2020
Table 8-11: Estimated Avoided PM2.5 and Ozone-Related Premature Deaths and Illnesses for the Final Rule, Compared to Baseline, Using
Alternative Approaches to Quantifying Avoided PM2.5-Attributable Deaths (95% Confidence Interval)
Category
Basis
2021
2023
2025
2030
2035
2040
2045
Avoided premature death among adults

Log-linear no-threshold
model
Krewski et al. (2009)
15
(10 to 20)
8
(6 to 11)
3
(2 to 4)
-4
(-5 to -3)
-11
(-15 to -8)
27
(18 to 36)
-8
(-11 to-5)
Lepeule et al. (2012)
34
(17 to 50)
19
(10 to 29)
7
(3 to 11)
-9
(-14 to -5)
-25
(-38 to -13)
61
(30 to 91)
-18
(-27 to-9)
Quantifying effect of PM2.5
above the LML in each
study
Krewski et al. (2009)
13
(9 to 17)
9
(6 to 13)
4
(3 to 6)
0
(0 to 0)
-9
(-12 to -6)
26
(17 to 34)
-6
(-8 to -4)
Lepeule et al. (2012)
18
(13 to 40)
5
(3 to 8)
-1
(-2 to -1)
-11
(-16 to -5)
-21
(-31 to -10)
38
(19 to 56)
-16
(-24 to -8)
a. Values rounded to two significant figures. Negative values indicate forgone benefits (i.e., the number of avoided cases under the final rule is smaller than in the baseline). Lower
bound of confidence interval represents the 95% confidence estimate that is lower in value than the point estimate, while upper bound represents the estimate that is higher in value
than the point estimate.
Source: U.S. EPA Analysis, 2020
8-16
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-12: Estimated Economic Value of Avoided PM2.5 and Ozone-Attributable Deaths and Illnesses for the Final Rule, Compared to
Baseline, Using Alternative Approaches to Represent PM2.5 Mortality Risk Effects (95% Confidence Interval; Million of 2018$)
Year
No-threshold Modelb
Limited to Above LMLC
Effects Above NAAQSd
3% Discount Rate
2021
$160 $370
($15 to $430) ° ($33 to $1000)
$140 $300
($14 to $380) ° ($27 to $850)
$14 $38
($2 to $36) ($3.9 to $110)
2023
$77 $160
(-$20 to $230) (-$80 to $540)
$88 $24
(-$19 to $260) (-$92 to $160)
-$6 -$28
(-$27 to $8.7) (-$97 to $8.3)
2025
$21 $33
(-$29 to $86) (-$110 to $200)
$34 -$49
(-$28 to $120) (-$150 to $4.4)
-$9.2 -$35
(-$32 to $4.9) (-$110 to $5.6)
2030
-$41 -$91
(-$120 to $0.33) (-$270 to $2.5)
-$0.4 -$110
(-$6.2 to $5.5) (-$310 to $4.5)
$0.43 $3.2
(-$3.9 to $5.6) ° (-$3.7 to $14)
2035
-$120 -$260
(-$330 to -$4.4) ° (-$760 to -$7.9)
-$96 -$220
(-$270 to -$2) ° (-$630 to -$3)
-$1.5 $1.6
(-$9.9 to $6.7) (-$9.7 to $16)
2040
$320 $740
($31 to $870) ° ($66 to $2100)
$300 $490
($29 to $830) ° ($44 to $1400)
$34 $93
($4.3 to $91) ($9 to $270)
2045
-$81 -$170
(-$250 to $19) (-$570 to $64)
-$60 -$150
(-$190 to $22) ° (-$500 to $69)
$6.8 $25
(-$7.1 to $28) (-$5.6 to $84)
7% Discount Rate
2021
$140 $330
($14 to $390) ° ($30 to $950)
$130 $270
($13 to $350) ° ($25 to $770)
$14 $38
($1.9 to $36) ($3.9 to $110)
2023
$69 $140
(-$20 to $210) (-$82 to $490)
$78 $19
(-$19 to $240) (-$93 to $140)
-$6.1 -$28
(-$27 to $8.6) (-$97 to $8)
2025
$18 $26
(-$30 to $78) (-$110 to $180)
$29 -$48
(-$29 to $110) (-$150 to $4)
-$9.3 -$35
(-$32 to $4.6) (-$110 to $5.1)
2030
-$37 -$82
(-$100 to $0.72) (-$240 to $3.3)
-$0.35 -$95
(-$5.9 to $5.4) ° (-$280 to $5.1)
$0.4 $3.2
(-$3.9 to $5.4) (-$3.7 to $14)
2035
-$110 -$240
(-$290 to-$3.3) ° (-$680 to-$5.6)
-$86 -$190
(-$240 to-$1.1) ° (-$570 to-$1.1)
-$1.4 $1.6
(-$9.8 to $6.7) (-$9.6 to $16)
2040
$290 $670
($28 to $790) ° ($60 to $1900)
$280 $450
($27 to $750) ° ($41 to $1300)
$33 $92
($4.3 to $90) ($8.9 to $270)
2045
-$73 -$150
(-$220 to $20) (-$510 to $66)
-$53 -$130
(-$170 to $22) ° (-$450 to $70)
$6.8 $25
(-$7 to $28) (-$5.6 to $84)
a. Values rounded to two significant figures. Negative values indicate forgone benefits. Lower bound of confidence interval represents the 95% confidence estimate that is lower in
value than the point estimate, while upper bound represents the estimate that is higher in value than the point estimate.
b. PM2.5 effects quantified using a no-threshold model. Low end of range reflects dollar value of effects quantified using concentration-response parameter from Krewski et al. (2009)
and Smith et al. (2008) studies; upper end quantified using parameters from Lepeule et al. (2012) and Jerrett et al. (2009). Full range of ozone effects is included.
8-17
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-12: Estimated Economic Value of Avoided PM2.5 and Ozone-Attributable Deaths and Illnesses for the Final Rule, Compared to
Baseline, Using Alternative Approaches to Represent PM2.5 Mortality Risk Effects (95% Confidence Interval; Million of 2018$)
Year
No-threshold Modelb
Limited to Above LMLC
Effects Above NAAQSd
c. PM2.5 effects quantified at or above the Lowest Measured Level of each long-term epidemiological study. Low end of range reflects dollar value of effects quantified down to LML
of Krewski et al. (2009) study (5.8 iag/m3); high end of range reflects dollar value of effects quantified down to LML of Lepeule et al. (2012) study (8 iag/m3). Full range of ozone
effects is still included.
d. PM effects only quantified at or above the annual mean of 12 |ig/m3 to provide insight regarding the fraction of benefits occurring above the NAAQS. Range reflects effects
quantified using concentration-response parameters from Smith et al. (2008) study at the low end and Jerrett et al. (2009) at the high end. Full range of ozone effects is still included.
Source: U.S. EPA Analysis, 2020
8-18
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-13: Estimated Percent of Avoided PM2.5-related Premature Deaths Above and Below PM2.5
Concentration Cut Points for the Final Rule, Compared to Baseline
Year
Epidemiological Study
Total
Mortality
Avoided PM2.5-related Premature Deaths Reported by Air
Quality Cutpointa
Above NAAQS
Below NAAQS and
Above LMLC
Below LMLC
2021
Krewski et al. (2009)
15
<1 (<1%)
13 (86%)
2 (12%)
Lepeule et al. (2012)
34
<1 (<1%)
26 (78%)
7 (21%)
2023
Krewski et al. (2009)
8
<1 (<1%)
9 b(118%)
-1 (-13%)
Lepeule et al. (2012)
19
<1 (<1%)
5 (28%)
14 (72%)
2025
Krewski et al. (2009)
3
<1 (3%)
4 b(143%)
-1 (-43%)
Lepeule et al. (2012)
7
<1 (2%)
-1 (-20%)
8 b(117%)
2030
Krewski et al. (2009)
-4
<1 (-1%)
>-1 (2%)
-4 (100%)
Lepeule et al. (2012)
-9
<1 (-1%)
-11 b(119%)
1 (-16%)
2035
Krewski et al. (2009)
-11
<1 (<1%)
-9 (82%)
-2 (19%)
Lepeule et al. (2012)
-25
<1 (<1%)
-20 (83%)
-4 (18%)
2040
Krewski et al. (2009)
27
<1 (<1%)
25 (94%)
1 (5%)
Lepeule et al. (2012)
61
<1 (<1%)
37 (61%)
23 (38%)
2045
Krewski et al. (2009)
-8
<1 (<1%)
-6 (75%)
-2 (24%)
Lepeule et al. (2012)
-18
<1 (<1%)
-16 (88%)
-2 (12%)
a. Values rounded to the nearest integer.
b. Avoided premature deaths below a threshold may be negative, while avoided premature deaths above a different threshold may
be positive. This can result in the percent of avoided PM2.5-related premature deaths above a certain threshold exceeding 100%.
c. The LML of the Krewski et al. (2009) study is 5.8 ng/m3 and 8 pg/m3 for Lepeule et al. (2012) study.
Source: U.S. EPA Analysis, 2020
8.4 Annualized Air Quality-Related Benefits of Regulatory Options
EPA calculated the present value of estimated air quality-related benefits and annualized these values using 3
percent and 7 percent discount rates to provide a measure that is comparable to the way other benefit
categories and social costs are reported.
Sections 8.2 and 8.3 provide benefit estimates for Option A, the final rule, based on the changes in the
electricity generation profile projected in IPM. As discussed in Section 8.3.1.1, the analysis of human health
benefits does not account for other changes in pollutant emissions associated with power requirements to
operate wastewater treatment systems or transport CCR or other solid waste. EPA examined the effects of
adjusting the estimated benefits in proportion to the average ratio between total air emissions of NOx and SO2
(Table 8-5) and EGU emissions associated with changes in the electricity generation profile (Table 8-3) for
the final rule and found that such an adjustment would have a negligible effect on human health benefit
estimates given interannual variability and discounting effects. Therefore, EPA is presenting unadjusted
values for the final rule below.
Because EPA did not run IPM for Options B and C, EPA did not analyze domestic climate and human health
benefits for Options B and C using the same modeling approach used for Option A. To provide insight into
the potential air quality-related benefits across regulatory options, EPA estimated benefits for Options B and
C by scaling Option A benefits in proportion to the social costs of the respective options (see Section 12.2).
This scaling factor is appropriate since changes in the profile of electricity generation account for the majority
of changes in air emissions (see Table 8-3 and Table 8-5) and this generation profile is affected most directly
by the incremental technology implementation costs. Specifically, EPA calculated the ratio of the benefits to
8-19
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs 8: Air Quality-Related Benefits
total social costs for Option A, then multiplied total social costs for Options B and C by this ratio. Table 8-14
summarizes the annualized air quality-related benefits of the regulatory options. Table 8-15 and Table 8-16
present results using alternative cut-points for PM2 5 related mortality risk benefits.
8-20
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-14: Total Annualized Air Quality-Related Benefits of Regulatory Options, Compared to the Baseline, 2021-2047 (Millions of 2018$)

3% Discount Rate3
7% Discount Rate3
Regulatory Option
Climate
Human Health
Total Benefits
Climate
Human Health
Total Benefits
Change
Krewski /
Lepeule/
Lower
Upper
Change
Krewski /
Lepeule/
Lower
Upper

Smith
Jerrett
Boundb
Boundc

Smith
Jerrett
Boundb
Boundc
Option A (Final Rule)
-$14
$28
$65
$14
$51
-$2.3
$25
$56
$23
$54
Option Bd
-$11
$23
$52
$11
$41
-$1.9
$21
$46
$19
$44
Option Cd
$2.3
-$4.7
t—1
t—1
i
-$2.4
LO
00
i
-$0.27
$3.0
$6.6
$2.7
$6.4
a. Values rounded to two significant figures. Negative values indicate forgone benefits.
b. Lower bound based on human health benefit point estimates using Krewski et al. (2009) for PM2.5 and Smith et al (2009) for ozone.
c. Upper bound based on human health benefit point estimates using Lepeule et al. (2012) for PM2.5 and Jerrett et al. (2009) for ozone.
d. EPA estimated air quality-related benefits for Options B and C by multiplying the total social costs for each option (see Section 12.2) by the ratio of [air quality-related benefits / total
social costs] for Option A.
Source: U.S. EPA Analysis, 2020
Table 8-15: Total Annualized Air Quality-Related Benefits of Regulatory Options, Compared to the Baseline, 2021-2047, Showing Only PM2.5
Related Premature Mortality Risk Benefits above the Lowest Measured Level of Each Long-Term PM2.5 Mortality Study (Millions of 2018$)

3% Discount Rate3
7% Discount Rate3
Regulatory Option
Climate
Human Health
Total Benefits
Climate
Human Health
Total Benefits
Change
Krewski /
Smith
Lepeule/
Jerrett
Lower
Boundb
Upper
Boundc
Change
Krewski /
Smith
Lepeule/
Jerrett
Lower
Boundb
Upper
Boundc
Option A (Final Rule)
-$14
$4.1
t—1
t—1
-c/>
-$9.7
-$3.3
-$2.3
$1.9
$3.4
-$0.40
$1.1
Option Bd
-$11
$3.3
LO
00
-$7.9
-$2.7
-$1.9
$1.6
$2.8
-$0.33
$0.91
Option Cd
$2.3
-$0.69
-$1.8
$1.6
$0.56
-$0.27
$0.23
$0.40
-$0,047
$0.13
a. Values rounded to two significant figures. Negative values indicate forgone benefits.
b. Lower bound based on human health benefit point estimates using Krewski et al. (2009) for PM2.5 and Smith et al (2009) for ozone.
c. Upper bound based on human health benefit point estimates using Lepeule et al. (2012) for PM2.5 and Jerrett et al. (2009) for ozone.
d. EPA estimated air quality-related benefits for Options B and C by multiplying the total social costs for each option (see Section 12.2) by the ratio of [air quality-related benefits / total
social costs] for Option A.
Source: U.S. EPA Analysis, 2020
8-22
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
8.5 Limitations and Uncertainties
Table 8-17 summarizes the limitations and uncertainties associated with the analysis of the air quality-related
benefits. The effect on benefits estimates indicated in the second column of the table refers to the magnitude
of the benefits rather than the direction (i.e.. a source of uncertainty that tends to underestimate benefits
indicates expectation for either larger forgone benefits or larger realized benefits). The analysis also
incorporates uncertainties associated with IPM modeling, which are discussed in Chapter 5 in the RIA (U.S.
EPA, 2020d). See Appendix I and Appendix J for additional discussions of the uncertainty associated with the
climate change benefit estimates and air quality modeling methodology, respectively.
Table 8-17: Limitations and Uncertainties in Analysis of Air Quality-Related Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
EPA extrapolated Option A
benefits to Options B and C.
Uncertain
EPA ran IPM only for Option A and used the results to
extrapolate benefits of Options B and C, based on the ratios
of annualized benefits and annualized social costs. Air
emissions and air quality changes are unlikely to follow
differences in social costs in a linear fashion, however, given
how marginal changes in operating costs for individual units
may affect dispatch of EGUs within the broader regional and
national electricity markets. Projected benefits for Options B
and C are therefore uncertain, with the uncertainty
expected to be greatest for Option C.
Domestic SC-C02 estimates
do not capture the full range
of impacts from climate
change.
Underestimate
Current integrated assessment models (lAMs) used in
developing the SC-C02 do not model all relevant regional
interactions - i.e., how climate change impacts in other
regions of the world could affect the United States, through
pathways such as global migration, economic
destabilization, and political destabilization.
The modeled air quality
surfaces used in the analysis
of human health benefits
only reflect changes in
emissions associated with
changes in the electricity
generation profile.
Uncertain
EPA developed the spatial fields based on IPM projected
emissions changes for Option A. These projections do not
include additional changes in NOx and S02 emissions
associated with power requirements to operate wastewater
treatment systems or transport CCR and other solid waste.
While these emissions changes could affect human health
benefit estimates, such effects are expected to be minimal
given that these emissions generally represent less than
1 percent of total NOx and S02 emissions changes.
The health impact function
for fine particles is log-linear
without a threshold.
Uncertain
The estimates include health benefits from reducing fine
particles in areas with different concentrations of PM25,
including both areas that do not meet the fine particle
standard and those areas that are in attainment and reflect
the full distribution of PM25 air quality simulated above.
All fine particles, regardless
of their chemical
composition, are equally
potent in causing premature
mortality.
Uncertain
The PM ISA concluded that "many constituents of PM2 5 can
be linked with multiple health effects, and the evidence is
not yet sufficient to allow differentiation of those
constituents or sources that are more closely related to
specific outcomes" (U.S. EPA, 2009a). The 2019 PM ISA
reaffirmed this conclusion (U.S. EPA, 2019e).
8-23
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
8: Air Quality-Related Benefits
Table 8-17: Limitations and Uncertainties in Analysis of Air Quality-Related Benefits
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
There is a "cessation" lag
between the change in PM
exposures and the total
realization of changes in
mortality effects.
Uncertain
The approach distributes the incidences of premature
mortality related to PM2.5 exposures over the 20 years
following exposure based on the advice of the EPA Science
Advisory Board Health Effect Subcommittee (SAB-HES) (U.S.
EPA, 2004). This distribution affects the valuation of
mortality benefits at different discount rates. The actual
distribution of effects over time is uncertain.
Climate changes may affect
ambient concentrations of
pollutants.
Uncertain
Estimated health benefits do not account for the influence
of future changes in the climate on ambient concentrations
of pollutants (U.S. Global Change Research Program, 2016).
For example, recent research suggests that future changes
to climate may create conditions more conducive to forming
ozone; the influence of changes in the climate on PM2.5
concentrations are less clear (Fann et a!., 2015). The
estimated health benefits also do not consider the potential
for climate-induced changes in temperature to modify the
relationship between ozone and the risk of premature death
(Jhun et al., 2014; Ren, Williams, Mengersen, et al., 2008;
Ren, Williams, Morawska, et al., 2008).
EPA did not analyze all
benefits of changes in NOx,
S02, and other pollutants
emitted by EGUs.
Underestimate
The analysis focused on adverse health effects related to
PM2.5 and ozone levels. There are additional direct benefits
from changes in levels of NOx, S02 and other air pollutants
emitted by EGUs (e.g., Hg, HCI). As described in U.S. EPA
(2019f), these include health benefits from changes in
ambient N02 and S02 exposure, health benefits from
changes in mercury deposition, ecosystem benefits
associated with changes in emissions of NOx, S02, PM, and
mercury, and visibility impairment.
8-24
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
9: Water Withdrawals
9 Changes in Water Withdrawals
Steam electric power plants use water for ash transport and for operating wet FGD scrubbers. The regulatory
options are estimated to change water withdrawal from surface waterbodies and aquifers by affecting sluicing
operations or incentives to recycle water within the plants.
Table 9-1 shows estimated changes in water withdrawals for each regulatory option.
Table 9-1: Industry-level Total Changes in Water Withdrawals under the
Regulatory Options, Compared to Baseline (Both Surface Water and Aquifers)
Regulatory Option
Change in Water Withdrawals
(Million Gallons per Day)a

Option Ac (Final Rule)
3.94
Option Bc
4.49
Option Cc
-9.93
a. Negative values represent a decrease in water withdrawals compared to the baseline, whereas
positive values represent an increase in water use.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule.
All results shown for Option D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA,
2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical
inputs for the analysis of Options A, B, and C.
c. Groundwater withdrawals are included in the total and are estimated to increase by 12,300 gallons
per day under regulatory options A, B, and C.
Source: U.S. EPA Analysis, 2020
The sections below discuss the benefits resulting specifically from estimated changes in groundwater
withdrawals. Benefits associated with surface water withdrawals are discussed qualitatively in Chapter 2.
The analysis shows very small effects from the final rule on water withdrawals compared to baseline (see U.S.
EPA, 2015a).
9.1 Methods
The analysis follows the same general methodology EPA used in the analysis of the 2015 rule and the 2019
proposal (U.S. EPA, 2015a; 2019a). Changes in water withdrawal from groundwater sources by steam electric
power plants may affect availability of groundwater for local municipalities that rely on aquifers for drinking
water supplies. These municipalities may incur incremental costs for supplementing drinking water supplies
through alternative means, such as bulk water purchases as water withdrawals by steam electric power plants
change. EPA estimated the monetary value of changes in groundwater withdrawals based on costs of
purchasing drinking water during periods of shortages in groundwater supply.
9.2 Results
EPA estimated that regulatory options A, B, and C would result in one plant increasing the volume of
groundwater withdrawn. See details in the Supplemental TDD (U.S. EPA, 2020g).
The estimated increase in groundwater withdrawals is 12,300 gallons per day (4.5 million gallons per year)
under Options A, B, and C. EPA estimated that demand for additional water supply exists in the affected
9-1
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
9: Water Withdrawals
areas due to potential drought (Tetra Tech, 2010). To estimate the value of reduced groundwater supply, EPA
used state-specific prices of bulk drinking water supplies, given that municipalities may need to purchase
supplementary supplies in response to any change in groundwater availability arising from additional
withdrawals by steam electric plants. This analysis provides screening-level indication of the potential
forgone benefits.
To estimate the monetary value of the changes in groundwater withdrawals, EPA multiplied the increase in
groundwater withdrawal (in gallons per year) by the estimated retail price of drinking water ($947.73 per
acre-foot for the affected location; Lincoln Public Works and Utilities, 2018) times a conversion factor of
325,851 to convert acre foot to gallons.76
Table 9-2 shows estimated annual forgone benefits from increased groundwater withdrawals under the
regulatory options. The annual forgone benefits are $0.01 million using both 3 percent and 7 percent discount
rates under regulatory options A, B, and C.
Table 9-2: Estimated Annualized Benefits from Changes in Groundwate
under the Regulatory Options, Compared to Baseline (Millions of 20182
;r Withdrawals
5)
Regulatory Option
Change in Groundwater
Withdrawals (Million
Gallons per Year)3
3% Discount Rate
7% Discount Rate

Option A (Final Rule)
4.5
-$0.01
-$0.01
Option B
4.5
-$0.01
-$0.01
Option C
4.5
-$0.01
-$0.01
a. Reflects changes after implementation of technologies to meet the regulatory option.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results
shown for Option D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the
values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options
A, B, and C.
Source: U.S. EPA Analysis, 2020
9.3 Limitations and Uncertainties
Table 9-3 summarizes the limitations and uncertainties in the analysis of benefits associated with changes in
groundwater withdrawals. Note that the effect on benefits estimates indicated in the second column of the
table refers to the magnitude of the benefits rather than the direction (i.e.. a source of uncertainty that tends to
underestimate benefits indicates expectation for larger forgone benefits or for larger realized benefits).
Table 9-3: Limitations and Uncertainties in Analysis of Changes in Groundwater Withdrawals
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
EPA estimated that municipalities
would need to replace lost
groundwater supplies with bulk
drinking water purchases.
Uncertain
Municipalities may not need to replace groundwater
withdrawn by steam electric power plants (in which
case the benefits of the final rule may be overstated),
or they may choose to replace the groundwater
through other means.
70 Water prices are uncertain. Average prices for irrigation water within the same geographic area range from approximately $50 to
$300 per acre-foot, with some water trades reaching $1,800 per acre-foot. Using these alternative prices would not have a large
impact on EPA's overall benefit estimates, however, given the small change in withdrawals under the regulatory options.
9-2
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
9: Water Withdrawals
Table 9-3: Limitations and Uncertainties in Analysis of Changes in Groundwater Withdrawals
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
EPA estimated a direct relationship
between groundwater withdrawals
in water-stressed states and
groundwater shortages, i.e., that
reducing demand for limited
groundwater supplies would result
in avoided costs for purchased
water.
Overestimate
EPA estimated that demand for additional water
supply exists in the affected areas due to potential
drought. However, the extent of this demand is
uncertain.
EPA estimated cost of bulk water
purchases based on state-wide
averages.
Uncertain
Costs of water may vary within a state and using the
average value may result in under- or overstating of
the cost for any given location. This uncertainty is
more significant in cases where there are few affected
locations, as is the case for this analysis which shows
only one plant with changes in groundwater
withdrawals.
Data on the characteristics of
affected aquifers are not available.
Uncertain
If the affected aquifers are used for private wells only,
the estimated benefits of improved groundwater
recharge could be under- or overstated, depending on
households' WTP for protecting groundwater quantity.
9-3
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
10: Dredging
10 Estimated Changes in Dredging Costs
As summarized in Table 3-3, the regulatory options could result in small changes in suspended solid
discharges by steam electric power plants, which could have an impact on the rate of sediment deposition in
affected reaches, including navigable waterways and reservoirs that require dredging for maintenance.
Navigable waterways, including rivers, lakes, bays, shipping channels and harbors, are an integral part of the
United States' transportation network. They are prone to reduced functionality due to sediment build-up,
which can reduce the navigable depth and width of the waterway (Clark etal., 1985; M. Ribaudo, 2011). In
many cases, costly periodic dredging is necessary to keep them passable. The regulatory options could
increase or reduce costs for government and private entities responsible for maintenance of navigable
waterways by changing the need for dredging.
Reservoirs serve many functions, including water storage for drinking, irrigation, and hydropower uses, flood
control, and recreation. Streams and rivers carry sediment into reservoirs, where it can settle and build up at a
recorded average rate of 1.2 billion kilograms per reservoir every year (USGS, 2009). Sedimentation reduces
reservoir capacity (Graf etal., 2010) and the useful life of reservoirs unless measures such as dredging are
taken to reclaim capacity (Clark etal., 1985; Hargrove etal., 2010; Miranda, 2017).
EPA estimated that the final rule, Option A, will have a small effect on historical average dredging costs
when compared to those estimated in 2015 for the 2015 rule (see U.S. EPA, 2015a).
10.1 Methods
In this analysis, EPA followed the same general methodology for estimating changes in costs associated with
changes in sediment depositions in navigational waterways and reservoirs that EPA used in the 2015 rule and
2019 proposal (U.S. EPA, 2015a [see Appendix K]; 2019a).77 The methodology utilizes information on
historic dredging locations, frequency of dredging, the amount of sediment removed, and dredging costs in
conjunction with the estimated changes in net sediment deposition (sedimentation minus erosion) in dredged
waterways and reservoirs under the regulatory options. Benefits are equal to avoided costs, calculated as the
difference from historical averages in total annualized dredging costs due to changes between the baseline and
the regulatory options. Negative values represent cost increases (i.e., forgone benefits to society).
10.1.1 Estimated Changes in Navigational Dredging Costs
EPA identified 250 unique dredging jobs and 393 dredging occurrences78 within the affected reaches. This
corresponds to approximately 13 percent of the dredging occurrences with coordinates reported in the
Dredging Information System (U.S. Army Corps of Engineers, 2013). The recurrence interval for dredging
jobs ranged from one to 15 years across affected reaches and averaged 13.3 years. Dredging costs vary
considerably across geographic locations and dredging jobs from approximately $0.11 per cubic yard at
77 For the final rule analysis, EPA made one change to the methodology used to estimate net sediment deposition at any given
location in the reach network by using the TOTAL YIELD output variable from the SPARROW models instead of
INCTOTALYIELD.
78 Dredging jobs refer to unique sites/locations defined by the U.S. Army Corps of Engineers where dredging was conducted,
whereas dredging occurrences are unique instances when dredging was conducted and may include successive dredging at the
same location.
10-1
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
10: Dredging
Sardine Point in Louisiana to $80.87 per cubic yard at Service Point in Illinois. The median unit cost of
dredging for the entire conterminous United States is $6.44 per cubic yard.
Table 10-1 presents low, mean, and high estimates of dredged sediment volume and dredging costs during the
period of 2021 through 2047 in navigational waterways that may be affected by steam electric plant
discharges, based on historical averages. EPA generated low, medium, and high estimates for navigational
dredging by varying the projected future dredging occurrence, including dredging frequency and job start as
well as cost of dredging for locations that did not report location specific costs (see U.S. EPA, 2015a,
Appendix K for details). Estimated total navigational dredging costs based on historical averages range from
$66.1 million to $155.0 million per year, using a 3 percent discount rate, and from $57.2 million to
$162.1 million using a 7 percent discount rate.
Table 10-1: Estimated Annualized Navigational Dredging Costs at Affected Reaches Based on
Historical Averages (Millions of 2018$)
Total Sediment Dredged
(Millions Cubic Yards)
Costs at 3% Discount Rate
(Millions of 2018$ per Year)
Costs at 7% Discount Rate
(Millions of 2018$ per Year)
Low
Mean
High
Low
Mean
High
Low
Mean
High
897.6
907.9
1,365.0
$66.1
$69.2
$155.0
$57.2
$59.6
$162.1
Source: U.S. EPA analysis, 2020.
The difference between the estimated dredging costs using historical averages and costs resulting from the
incremental sediment deposition under a regulatory option as compared to baseline represents the avoided
costs (or forgone benefits) of the regulatory option. Table 10-2 presents estimated changes in navigational
dredging costs for the regulatory options.
Table 10-2: Estimated Annualized Changes in Navigational Dredging Costs under the Regulatory
Options, Compared to Baseline

Total Reduction in Sediment
3% Discount Rate
7% Discount Rate
Regulatory Option
Dredged (Thousands Cubic
Yards)
(Thousands of 2018$ per
Year)3
(Thousands of 2018$ per
Year)3

Low
Mean
High
Low
Mean
High
Low
Mean
High

Option A (Final Rule)
-0.9
-0.9
-1.3
-$0.1
-$0.1
-$0.2
-$0.1
-$0.1
-$0.2
Option B
-0.9
-0.9
-1.2
-$0.1
-$0.1
-$0.1
-$0.1
-$0.1
-$0.1
Option C
14.8
14.8
15.1
$5.5
$5.5
$5.8
$4.8
$4.8
$5.3
a. Positive values represent cost savings; negative values represent cost increases.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C (including changes in the sediment
models and basis for estimating sediment depositions).
Source: U.S. EPA analysis, 2020.
10.1.2 Estimated Changes in Reservoir Dredging Costs
EPA identified 2,747 reservoirs within the affected reaches with changes in sediment loads under at least one
of the regulatory options, corresponding to approximately one percent of the reservoirs represented in the
SPARROW models (Ator, 2019; Hoos & Roland Ii, 2019; Robertson & Saad, 2019; Wise, 2019; Wise etal.,
2019). EPA used regional estimates of median dredging costs to calculate changes in reservoir dredging costs
under the regulatory options. The median cost per cubic yard ranges from $2.72 in EPA Region 2 to $31.38 in
EPA Region 5, with a national median value of $6.44. Table 10-3 presents low, mean, and high estimates of
10-2
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
10: Dredging
the projected volume of sediment to be dredged during the period of 2021 through 2047 from these reservoirs
and estimated annualized dredging costs, based on historical averages. Estimated reservoir dredging costs
based on historical averages range between $144.2 million and $195.7 million using a 3 percent discount rate
and $120.7 million and $179.9 million using a 7 percent discount rate.
Table 10-3: Estimated Annualized Reservoir Dredging Volume and Costs based on Historical
Averages
Total Sediment Dredged
(Millions Cubic Yards)
Costs at 3% Discount Rate (Millions
of 2018$ per Year)
Costs at 7% Discount Rate (Millions
of 2018$ per Year)
Low
Mean
High
Low
Mean
High
Low
Mean
High
708.3
850.0
920.8
$144.2
$174.6
$195.7
$120.7
$151.5
$179.9
Source: U.S. EPA analysis, 2020.
The difference between the estimated dredging costs using historical averages and costs resulting from the
incremental sediment deposition under a regulatory option as compared to baseline represents the avoided
costs for that regulatory option. Table 10-4 presents estimated cost changes for reservoir dredging under the
regulatory options, including low, mean, and high estimates.
Table 10-4: Estimated Total Annualized Changes in Reservoir Dredging Volume and Costs under the
Regulatory Options, Compared to Baseline

Total Reduction in Sediment
Costs at 3% Discount Rate3
Costs at 7% Discount Rate3

Dredged (Thousands Cubic
(Thousands of 2018$ per
(Thousands of 2018$ per

Yards)

Year)

Regulatory Option
Low
Mean
High
Low
Mean
High
Low
Mean
High

Option A (Final Rule)
-2.0
-2.7
-3.0
-$0.4
-$0.5
-$0.6
-$0.3
-$0.5
-$0.6
Option B
-1.3
-1.6
-1.8
-$0.3
-$0.4
-$0.4
-$0.2
-$0.3
-$0.4
Option C
-0.3
-1.1
-1.4
-$0.1
-$0.3
-$0.4
-$0.1
-$0.3
-$0.4
a. Positive values represent cost savings; negative values represent cost increases.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C (including changes in the sediment
models and basis for estimating sediment depositions).
Source: U.S. EPA analysis, 2020.
10.2 Limitation and Uncertainty
Table 10-5 summarizes key uncertainties and limitations in the analysis of sediment dredging benefits. A
more detailed description is provided in Appendix K of the 2015 BCA (U.S. EPA, 2015a). Note that the effect
on benefits estimates indicated in the second column of the table refers to the magnitude of the benefits rather
than the direction (i.e.. a source of uncertainty that tends to underestimate benefits indicates expectation for
larger forgone benefits or for larger realized benefits). Uncertainties and limitations associated with
SPARROW model estimates of sediment deposition are discussed in the respective regional model reports
(Ator, 2019; Hoos & Roland Ii, 2019; Robertson & Saad, 2019; Wise, 2019; Wise etal., 2019).
10-3
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
10: Dredging
Table 10-5: Limitations and Uncertainties in Analysis of Changes in Dredging Costs
Uncertainty/Limitation
Effect on Benefits
Estimate
Notes
The analysis scales dredging volumes
and costs in proportion to the
percent change in sediment
deposition in navigational
waterways and reservoirs.
Uncertain
EPA estimated a linear relationship between changes
in sediment deposition and dredging volumes and
costs which may not capture non-linear dynamics in
the relationships between sediment deposition and
dredging volumes and between dredging volumes and
costs.
The analysis of navigational
waterways includes only jobs
reported for 1998 through 2012
(U.S. Army Corps of Engineers,
2013).
Underestimate
Because some dredging jobs included in the U.S. Army
Corps of Engineers Database lack latitude and
longitude and the database does not use standardized
job names, EPA was only able to map approximately
71 percent of all recorded dredging occurrences. This
may lead to potential underestimation of historical
costs and changes in dredging costs under the
regulatory options.
The analysis of reservoir dredging is
limited to reservoirs identified on
the NHD reach network.
Underestimate
The omission of other reservoirs could understate the
magnitude of estimated historical costs and changes in
reservoir dredging benefits if there are additional
reservoirs located downstream from steam electric
power plants.
10-4
-------
BCAfor Proposed Revisions to the Steam Electric Power Generating ELGs
11: Total Monetized Benefits
ii Summary of Estimated Total Monetized Benefits
Table 11-1 and Table 11-2, on the next two pages, summarize the total annualized monetized benefits using
3 percent and 7 percent discount rates, respectively.
The monetized benefits do not account for all effects of the regulatory options, including changes in certain
cancer and non-cancer health risk (e.g., effects of halogenated disinfection byproducts in drinking water,
effects of cadmium on kidney functions and bone density), impacts of pollutant load changes on T&E species
habitat, etc. See Chapter 2 for a discussion of categories of benefits EPA did not monetize. Chapter 4 through
Chapter 10 provide more detail on the estimation methodologies for each benefit category.
11-1
-------
BCAfor Proposed Revisions to Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-1: Summary of Estimated Total Annualized Benefits of the Regulatory Options, Compared to Baseline, at 3 Percent (Millions of
2018$)
Benefit Category
Option D°'b
Option A
(Final Rule)
Option B
Option C
Human Health

-$0.3
-$0.3
-$0.1
Changes in IQ losses in children from exposure to leadd

<$0.0
<$0.0
<$0.1
Changes in IQ losses in children from exposure to mercury

-$0.3
-$0.3
-$0.1
Ecological Conditions and Recreational Uses Changes

-$15.3 to -$7.4
-$10.4 to -$5.5
-$9.9 to -$4.8
Use and nonuse values for water quality changes6

-$15.3 to -$7.4
-$10.4 to -$5.5
-$9.9 to -$4.8
Market and Productivity Effectsd

<$0.0
<$0.0
$0.0
Changes in dredging costsd

<$0.0
<$0.0
<$0.0
Reduced water withdrawal

<$0.0
<$0.0
<$0.0
Air Quality-Related Effects

$14 to $51
$11 to $41
-$8.5 to -$2.4
Climate change effects from changes in C02 emissions'

-$14
t—1
t—1
1
$2.3
Human health effects from changes in NOx, S02, and
PM2.5 emissions8
Not estimated
$28 to $65
$23 to $52
-$11 to -$4.7
Total8,11

-$1.7 to $43.3
$0.3 to $35.7
-$12.4 to -$13.4
a. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
b. Negative values represent forgone benefits and positive values represent realized benefits.
c. Total includes $0.4 million of benefits due to changes in bladder cancer risk from disinfection byproducts in drinking water as estimated for the 2019 proposed rule (U.S. EPA, 2019a).
d. "<$0.0" indicates that monetary values are greater than -$0.1 million but less than $0.0 million.
e. The range reflects the lower and upper bound willingness-to-pay estimates. See Chapter 6 for details.
f. Values for air-quality related effects are rounded to two significant figures. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air quality-
related benefits for Options B and C from the estimate for Option A that is based on IPM outputs. See Chapter 8 for details.
g. Values for air-quality related effects are rounded to two significant figures. The range reflects the lower and upper bound estimates of human health effects from changes in PM2.5
and ozone levels. See Chapter 8 for details.
h. Values for individual benefit categories may not sum to the total due to independent rounding. Range is based on the low and high willingness to pay estimates and air quality-related
effects.
i. Value reflects midpoint willingness-to-pay estimate. See 2019 BCA for details (U.S. EPA, 2019a).
Source: U.S. EPA Analysis, 2020
11-2
-------
BCAfor Proposed Revisions to Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-2: Summary of Estimated Total Annualized Benefits of the Regulatory Options, Compared to Baseline, at 7 Percent (Millions of
2018$)
Benefit Category
Option D°'b
Option A
(Final Rule)
Option B'
Option C
Human Health

-$0.1
-$0.1
-$0.1
Changes in IQ losses in children from exposure to leadd

<$0.0
<$0.0
<$0.0
Changes in IQ losses in children from exposure to mercury

-$0.1
-$0.1
-$0.1
Ecological Conditions and Recreational Uses Changes

-$16.4 to -$8.0
-$12.0 to -$5.8
-$13.9 to -$6.7
Use and nonuse values for water quality changes6

-$16.4 to -$8.0
-$12.0 to -$5.8
-$13.9 to -$6.7
Market and Productivity Effectsd

<$0.0
<$0.0
<$0.0
Changes in dredging costsd

<$0.0
<$0.0
<$0.0
Reduced water withdrawal

<$0.0
<$0.0
<$0.0
Air Quality-Related Effects

$23 to $54
$19 to $44
$2.7 to $6.4
Climate change effects from changes in C02 emissions'

-$2.3
-$1.9
-$0.27
Human health effects from changes in NOx, S02, and PM2.5
emissions8
Not estimated
$25 to $56
$21 to $46
$3.0 to $6.6
Total&h

$6.5 to $45.9
$6.9 to $38.1
-$11.3 to -$0.4
a. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
b. Negative values represent forgone benefits and positive values represent realized benefits.
c. Total includes $0.2 million of benefits due to changes in bladder cancer risk from disinfection byproducts in drinking water as estimated for the 2019 proposed rule (U.S. EPA, 2019a).
d. "<$0.0" indicates that monetary values are greater than -$0.1 million but less than $0.0 million.
e. The range reflects the lower and upper bound willingness-to-pay estimates. See Chapter 6 for details.
f. Values for air-quality related effects are rounded to two significant figures. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air quality-
related benefits for Options B and C from the estimate for Option A that is based on IPM outputs. See Chapter 8 for details.
g. Values for air-quality related effects are rounded to two significant figures. The range reflects the lower and upper bound estimates of human health effects from changes in PM2.5
and ozone levels. See Chapter 8 for details.
h. Values for individual benefit categories may not sum to the total due to independent rounding. Range is based on the low and high willingness to pay estimates and air quality-
related effects.
i. Value reflects midpoint willingness-to-pay estimate. See 2019 BCA for details (U.S. EPA, 2019a).
Source: U.S. EPA Analysis, 2020
11-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
12: Total Social Costs
12 Summary of Total Social Costs
This chapter discusses EPA's estimates of the costs to society under the regulatory options. Social costs
include costs incurred by both private entities and the government (e.g., in implementing the regulation). As
described further in Chapter 10 of the RIA (U.S. EPA, 2020d), EPA did not evaluate incremental cost to state
governments to evaluate and incorporate best professional judgment into NPDES permits. Consequently, the
only category of costs used to calculate social costs are estimated technology implementation costs for steam
electric power plants. As discussed below, these costs may be positive or negative, with the latter occurring
when a regulatory option provides savings as compared to the baseline.
12.1 Overview of Costs Analysis Framework
The RIA (Chapter 3) presents EPA's development of costs for the estimated 914 steam electric power plants
within the scope of the final rule (U.S. EPA, 2020d). These costs (pre-tax) are used as the basis of the social
cost analysis. A subset of these plants incur non-zero costs under the baseline or the regulatory options.
As described in Chapter 1, EPA estimated that steam electric power plants, in the aggregate, will implement
control technologies between 2021 and 2028, with the technology implementation schedule varying across
wastestreams and regulatory options. For the analysis of social costs, EPA estimated a plant- and year-explicit
schedule of technology implementation cost outlays over the period of 2021 through 2047.79 This schedule
accounts for retirements and repowerings by zeroing-out O&M costs to operate treatment systems in years
following unit retirement or repowering. After creating a cost-incurrence schedule for each cost component,
EPA summed the costs expected to be incurred in each year for each plant, then aggregated these costs to
estimate the total costs for each year in the analysis period. Following the approach used for the 2015 rule
analysis and 2019 proposal (U.S. EPA, 2015c, 2019g), after technology implementation costs were assigned
to the year of occurrence, the Agency adjusted these costs for change between 2018 (the year when costs were
estimated) and the year(s) of their incurrence as follows:
• All technology costs, except planning, were adjusted to their incurrence year(s) using the
Construction Cost Index (CCI) from McGraw Hill Construction and the Gross Domestic Product
(GDP) deflator index published by the U.S. Bureau of Economic Analysis (BEA);
• Planning costs were adjusted to their incurrence year(s) using the Employment Cost Index (ECI)
Bureau of Labor Statistics (BLS) and GDP deflator.
The CCI and ECI adjustment factors were developed only through the year 2027; after these years, EPA
assumed that the real change in prices is zero - that is, costs are expected to change in line with general
inflation. EPA judges this to be a reasonable approach, given that capital expenditures will occur by 2028 and
the uncertainty of long-term future price projections.
After developing the year-explicit schedule of total costs and adjusting them for predicted real change to the
year of their incurrence, EPA calculated the present value of these cost outlays as of the rule promulgation
year by discounting the cost in each year back to 2020, using both 3 percent and 7 percent discount rates.
These discount rate values reflect guidance from the OMB regulatory analysis guidance document, Circular
A-4 (OMB, 2003). EPA calculated the constant annual equivalent value (annualized value), again using the
79 The period of analysis extends through 2047 to capture a substantive portion of the life of the wastewater treatment technology at
any steam electric power plant (20 or more years), and the last year of technology implementation (2028).
12-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
12: Total Social Costs
two values of the discount rate, 3 percent and 7 percent, over a 27-year social cost analysis period. EPA
assumed no re-installation of wastewater treatment technology during the period covered by the social cost
analysis.
To assess the economic costs of the regulatory options to society, EPA relied first on the estimated costs to
steam electric power plants for the labor, equipment, material, and other economic resources needed to
comply with the regulatory options (see U.S. EPA, 2020d for details). In this analysis, the market prices for
labor, equipment, material, and other compliance resources represent the opportunity costs to society for use
of those resources in regulatory compliance. EPA assumed in its social cost analysis that the regulatory
options do not affect the aggregate quantity of electricity that will be sold to consumers and, thus, that the
rule's social cost will include no changes in consumer and producer surplus from changes in electricity sales
by the electricity industry in aggregate. Given the small impact of the regulatory options on electricity
production cost for the total industry, this approach is reasonable for the social cost analysis (for more details
on the impacts of the regulatory options on electricity production cost, see RIA Chapter 5). The social cost
analysis considers costs on an as-incurred, year-by-year basis - that is, this analysis associates each cost
component to the year(s) in which they are assumed to occur relative to the assumed rule promulgation and
technology implementation years.80
Finally, as discussed in Chapter 10 of the RIA (U.S. EPA, 2020d; see Section 10.7: Paperwork Reduction Act
of 1995), the regulatory options will not result in additional administrative costs for plants to implement, and
state and federal NPDES permitting authorities to administer, the final rule. As a result, the social cost
analysis focuses on the resource cost of compliance as the only direct cost incurred by society as a result of
the regulatory options.
12.2 Key Findings for Regulatory Options
Table 12-1 presents annualized costs for the baseline and the analyzed regulatory options. The table also
provides the incremental costs attributable to the regulatory options, calculated as the difference between each
option and the baseline. As shown in the table, the regulatory options generally result in cost savings across
the options and discount rates, with the exception of Option C which results in incremental costs at 3 percent
discount rate. Thus, incremental costs range from -$127.1 million to $21.4 million at a 3 percent discount rate,
and from -$153.4 million to -$18.2 million at a 7 percent discount rate.
The specific assumptions of when each cost component is incurred can be found in Chapter 3 of the RIA.
12-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
12: Total Social Costs
Table 12-1: Summary of Estimated Annualized Costs (Millions of 2018$)

Annualized Costs
Incremental Costs
Regulatory Option
3% Discount Rate
7% Discount Rate
3% Discount Rate
7% Discount Rate
Baseline
$309.6
$347.8

Option A (Final Rule)
$182.5
$194.4
-$127.1
-$153.4
Option B
$206.4
$221.4
-$103.2
-$126.4
Option C
$331.1
$329.6
$21.4
-$18.2
a. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. Incremental costs for Option D are
relative to the 2019 analysis baseline versus the baseline to which Options A, B, and C are compared. For these reasons, the values
should not be used for direct comparisons to the final rule.
Source: U.S. EPA Analysis, 2020.
Table 12-2 provides additional detail on the social cost calculations. The table compiles, for the baseline and
each regulatory option, the assumed time profiles of technology implementation costs incurred. The table also
reports the estimated annualized values of costs at 3 percent and 7 percent discount rates (see bottom of the
table). The maximum technology implementation outlays differ across the options but are incurred over the
years 2021 through 2028, i.e., during the estimated window (defined as Period 1 in Section 3.2.1) when steam
electric power plants are expected to implement wastewater treatment technologies.
Table 12-2: Time Profile of Costs to Society (Millions of 2018$)

Technology Implementation Costs
Incremental Costs

Option A

(Final

Year
Baseline
Option D°
Rule)
Option B
Option C
Option D°
Rule)
Option B
Option C
2020
$0.0

$0.0
$0.0
$0.0

$0.0
$0.0
$0.0
2021
$998.2

$518.4
$595.6
$223.1

-$479.8
-$402.5
-$775.1
2022
$540.0

$193.1
$288.4
$229.0

-$346.9
-$251.5
-$310.9
2023
$1,551.8

$246.6
$273.7
$118.4

-$1,305.2
-$1,278.1
-$1,433.4
2024
$201.8

$305.2
$321.2
$393.0

$103.3
$119.4
$191.1
2025
$201.6

$448.8
$494.5
$609.4

$247.2
$292.8
$407.7
2026
$196.5

$117.2
$137.3
$1,036.2

-$79.4
-$59.2
$839.6
2027
$198.8

$121.5
$139.1
$384.9

-$77.3
-$59.7
$186.1
2028
$186.7

$238.7
$272.6
$707.4

$52.1
$85.9
$520.7
2029
$189.7

$130.9
$143.9
$271.9

-$58.9
-$45.8
$82.2
2030
$187.6

$129.8
$142.9
$271.5

-$57.8
-$44.6
$83.9
2031
$188.2

$134.3
$147.4
$273.6

-$53.9
-$40.8
$85.4
2032
$188.4

$132.5
$146.4
$275.0

-$55.9
-$42.0
$86.5
2033
$191.0

$130.9
$144.3
$273.4

-$60.0
-$46.6
$82.4
2034
$186.9

$132.9
$146.0
$274.8

-$54.0
-$40.9
$87.9
2035
$188.9

$133.7
$146.7
$275.3

-$55.2
-$42.2
$86.3
2036
$183.8

$130.0
$143.1
$273.3

-$53.8
-$40.7
$89.5
2037
$185.5

$131.4
$144.7
$273.6

-$54.1
-$40.8
$88.1
2038
$180.5

$130.8
$143.8
$273.5

-$49.7
-$36.7
$93.0
2039
$187.5

$131.4
$144.6
$272.6

-$56.0
-$42.9
$85.2
2040
$186.9

$130.2
$143.3
$272.1

-$56.7
-$43.5
$85.2
2041
$190.8

$133.9
$146.9
$274.0

-$56.9
-$43.9
$83.2
2042
$188.6

$133.6
$147.4
$275.6

-$55.1
-$41.2
$87.0
12-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
12: Total Social Costs
Table 12-2: Time Profile of Costs to Society (Millions of 2018$)

Technology Implementation Costs
Incremental Costs

Option A

(Final

Year
Baseline
Option D°
Rule)
Option B
Option C
Option D°
Rule)
Option B
Option C
2043
$189.8

$131.9
$145.3
$274.0

-$57.9
-$44.5
$84.2
2044
$186.5

$132.1
$145.3
$274.4

-$54.4
-$41.3
$87.8
2045
$187.1

$133.1
$146.2
$275.1

-$54.0
-$40.9
$88.0
2046
$186.3

$131.5
$144.7
$274.0

-$54.8
-$41.6
$87.7
2047
$186.9

$131.5
$144.7
$273.8

-$55.4
-$42.2
$86.9
Annualized
$309.6
$234.3
$182.5
$206.4
$331.1
-$130.6
-$127.1
-$103.2
$21.4
Costs, 3%

Annualized
$347.8
$263.0
$194.4
$221.4
$329.6
-$154.0
-$153.4
-$126.4
-$18.2
Costs, 7%

a. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option
D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the
baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. Incremental costs for Option D are
relative to the 2019 analysis baseline versus the baseline to which Options A, B, and C are compared. For these reasons, the values
should not be used for direct comparisons to the final rule.
Source: U.S. EPA Analysis, 2020.
12-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
13: Benefits and Social Costs
13 Benefits and Social Costs
This chapter compares total monetized benefits and costs for the regulatory options. Benefits and costs are
compared on two bases: (1) incrementally for each of the options analyzed as compared to the baseline and
(2) incrementally across options. The comparison of benefits and costs also satisfies the requirements of
Executive Order 12866: Regulatory Planning and Review and Executive Order 13563: Improving Regulation
and Regulatory Review (see Chapter 9 in the RIA; U.S. EPA, 2020d).
13.1 Comparison of Benefits and Costs by Option
Chapters 11 and 12 present estimates of the benefits and costs, respectively, for the regulatory options as
compared to the baseline. Table 13-1 presents EPA's estimates of benefits and costs of the regulatory options,
at 3 percent and 7 percent discount rates, and annualized over 27 years.
Table 13-1: Total Estimated Annualized Benefits and Costs by Regulatory Option and
Discount Rate, Compared to Baseline (Millions of 2018$)
Regulatory Option
Total Monetized Benefits3
Total Costs
Low
High
3% Discount Rate

Option A (Final Rule)
-$1.7
$43.3
-$127.1
Option B
$0.3
$35.7
-$103.2
Option C
-$12.4
-$13.4
$21.4
7% Discount Rate

Option A (Final Rule)
$6.5
$45.9
-$153.4
Option B
$6.9
$38.1
-$126.4
Option C
-$11.3
-$0.4
-$18.2
a. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air quality-related benefits
for Options B and C from the estimate for Option A that is based on IPM outputs. The Low and High values reflect the
lower and upper bound estimates of air quality-related human health benefits. See Chapter 8 for details.
b. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown
for Option D are based on the 2019 analysis, as detailed in the 2019 BCA (U.S. EPA, 2019a). As such, the values do not
reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. Total
benefits for this option include benefits from changes in cancer incidence associated with disinfection byproducts in
drinking water and do not include human health benefits associated with changes in air quality.
Source: U.S. EPA Analysis, 2020.
13.2 Analysis of Incremental Benefits and Costs
In addition to comparing estimated benefits and costs for each regulatory option relative to the baseline, as
presented in the preceding section, EPA also estimated the benefits and costs of the options on an incremental
basis. The comparison in the preceding section addresses the simple quantitative relationship between
estimated benefits and costs for each option and determines whether costs or benefits are greater for a given
option and by how much. In contrast, incremental analysis looks at the differential relationship of benefits and
costs across options and poses a different question: as increasingly more costly options are considered, by
what amount do benefits, costs, and net benefits (/'. e., benefits minus costs) change from option to option?
Incremental net benefit analysis provides insight into the net gain to society from imposing increasingly more
costly requirements.
13-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
13: Benefits and Social Costs
EPA conducted the incremental net benefit analysis by calculating, for regulatory options A through C, the
change in net benefits, from option to option, in moving from the least stringent option to successively more
stringent options, where stringency is determined based on total pollutant loads. As described in Chapter 1,
the regulatory options differ in the technology basis for different wastestreams. Thus, the difference in
benefits and costs across the options derives from the characteristics of the wastestreams controlled by an
option, the relative effectiveness of the control technology in reducing pollutant loads, the timing of control
technology implementation, and the distribution and characteristics of steam electric power plants and of the
receiving reaches.
As reported in Table 13-2, Options A and B have positive benefits and cost savings, with net annual
monetized benefits ranging from $125.7 million to $170.1 million under Option A and from $104.0 million to
$139.1 million under Option B (3 percent discount rate). Option C has forgone benefits and positive costs,
with net annual monetized benefits of -$33.8 million to -$34.9 million using a 3 percent discount rate. Among
the regulatory options, the final rule (Option A) results in the highest net annual monetized benefits.
Using a 3 percent discount rate, the incremental net annual monetized benefits of moving from Option A to
Option B ranges from -$31.5 million to -$22.0 million. The negative values indicate that net annual
monetized benefits are higher for Option A than for Option B. Moving from Option B to Option C, the
change is negative, at -$173.7 million to -$137.3 million, indicating that the net annual monetized benefits are
higher from Option B than for Option C.
Table 13-2: Analysis of Estimated Incremental Net Benefit of the Regulatory Options,
Compared to Baseline and to Other Regulatory Options (Millions of 2018$)
Regulatory Option
Net Annual Monetized Benefitsa b
Incremental Net Annual Monetized
Benefits'"
Low
High
Low
High
3% Discount Rate
Option A (Final Rule)
$125.5
$170.4
NA
NA
Option B
$103.5
$138.9
-$22.0
-$31.5
Option C
-$33.8
-$34.8
-$137.3
-$173.7
7% Discount Rate
Option A (Final Rule)
$159.9
$199.3
NA
NA
Option B
$133.3
$164.5
-$26.6
-$34.8
Option C
$6.9
$17.8
-$126.4
-$146.7
NA: Not applicable for Option A
a. Net benefits are calculated by subtracting total annualized costs from total annual monetized benefits, where both
costs and benefits are measured relative to the baseline.
b. EPA estimated the air quality-related benefits for Option A. EPA extrapolated estimates of air quality-related benefits
for Options B and C from the estimate for Option A that is based on IPM outputs. See Chapter 8 for details.
c. Incremental net benefits are equal to the difference between net benefits of an option and net benefits of the
previous, less stringent option.
Source: U.S. EPA Analysis, 2020.
13-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
14 Environmental Justice
Executive Order (EO) 12898 (59 FR 7629, February 11, 1994) requires that, to the greatest extent practicable
and permitted by law, each Federal agency must make the achievement of EJ part of its mission. EO 12898
provides that each Federal agency must conduct its programs, policies, and activities that substantially affect
human health or the environment in a manner that ensures such programs, policies, and activities do not have
the effect of (1) excluding persons (including populations) from participation in, or (2) denying persons
(including populations) the benefits of, or (3) subjecting persons (including populations) to discrimination
under such programs, policies, and activities because of their race, color, or national origin.
To meet the objectives of EO 12898, EPA examined whether the change in benefits from the regulatory
options may be differentially distributed among population subgroups in the affected areas. EPA considered
the following factors in this analysis: population characteristics, proximity to affected waters, exposure
pathways, cumulative risk exposure, and susceptibility to environmental risk. For example, subsistence fishers
rely on self-caught fish for a larger share of their food intake than do recreational fishermen, and as such may
incur a larger share of effects arising from the regulatory options.
As described in the following sections, EPA conducted two types of analyses to evaluate the EJ implications
of the regulatory options:
1. Summarizing the demographic characteristics of the households living in proximity to steam
electric power plants, plant air emissions and surface water discharges, and to the downstream
reaches affected by plant discharges; and
2. Analyzing the distribution of human health impacts among minority and/or low-income
populations from changes in exposure to pollutants in drinking water, self-caught fish, and the air.
The first analysis provides insight on the distribution of regulatory option effects (e.g., effects on water
quality and air pollutant emissions) on communities in proximity to steam electric power plants. The second
analysis seeks to provide more specific insight on the distribution of estimated changes in adverse health
effects and benefits and to assess whether minority and/or low-income populations incur disproportionately
high environmental impacts and/or will be disproportionately excluded from realizing benefits under the
regulatory options.
14.1 Socioeconomic Characteristics of Populations Residing in Proximity to Steam Electric
Power Plants
For the first analysis, EPA assessed the demographic characteristics of the populations within specified
distances of steam electric power plants. The analysis is analogous to the profile EPA developed to support
the 2015 rule (U.S. EPA, 2015a) and updated for the 2019 proposal (U.S. EPA, 2019a).
14-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
EPA collected population-specific U.S. Census Bureau's American Community Survey (ACS) data on:
• the percent of the population below the poverty threshold,81 referred to as low-income population for
the purpose of this analysis, and
• the population categorized in various racial/ethnic groups, from which EPA calculated the percent of
the total population that belongs to a minority racial/ethnic group.82
EPA compiled these data for CBGs located within specified distances (e.g., one mile, three miles, 15 miles,
30 miles, and 50 miles) of steam electric power plants and within 50 miles of reaches downstream from steam
electric power plant outfalls. EPA compared demographic metrics of these buffer areas to state and national
averages to identify communities where EJ concerns may exist. EJ concerns may exist in areas where the
percent of the population that is low-income and/or minority is higher than the respective state or national
averages.
Specifically, this first analysis considers the spatial distribution of low-income and minority groups to
determine whether these groups are more or less represented in the populations living in proximity to steam
electric power plants that discharge bottom ash transport water or FGD wastewater. The distance buffers from
the steam electric power plants and their associated immediate receiving reaches83 are denoted below as the
"analysis region." Populations within the regions included in the analysis may be affected by steam electric
power plant discharges and other environmental impacts in the baseline and by environmental changes
resulting from the regulatory options, whether those changes are beneficial or detrimental. If the population
within a given region has a larger proportion of low-income or minority than the state average, it may indicate
that the regulatory options may affect communities that have been historically exposed to a disproportionate
share of environmental impacts and the final rule may thus contribute to redressing or exacerbating existing
EJ concerns, depending on the direction of the changes under the regulatory options.
EPA used the U.S. Census Bureau's ACS data for 2013 to 2017 to identify minority and income status at the
CBG, analysis region, and state levels. Table 14-1 summarizes the socioeconomic characteristics of the
analysis regions defined using radial distances of one, three, 10, 15, 30, and 50 miles from the steam electric
power plants.
81 Poverty status is based on data from the Census Bureau's American Community Survey which determines poverty status by
comparing annual income to a set of dollar values called poverty thresholds that vary by family size, number of children, and the
age of the householder.
82 The racial/ethnic categories are based on available fish consumption data as well as the breakout of ethnic/racial populations in
Census data, which distinguishes racial groups within Hispanic and non-Hispanic categories. Minority groups include: African
American (non-Hispanic); Asian (non-Hispanic); Native Hawaiian/Pacific Islander (non-Hispanic); American Indian/Alaska
Native (non-Hispanic); Other non-Hispanic; Hispanic/Latino.
83 In this analysis, EPA used the coordinates of each steam electric plant as the basis to define analysis regions using various
distance buffers.
14-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-1: Socioeconomic Characteristics of Communities Living in Proximity to Steam Electric
Power Plants and Associated Immediate Receiving Reach, Compared to National Average
Distance from Steam Electric
Power Plant
Total Population
(Millions)"
Percent Minority
Percent Low-
Income
Demographic
Index"
1 mile
0.5
18.4%
13.2%
15.8%
3 miles
1.7
22.8%
13.1%
18.0%
15 miles
25.0
32.9%
13.6%
23.3%
30 miles
87.7
34.7%
13.4%
24.1%
50 miles
199.7
34.4%
13.7%
24.1%
United States
325.7
39.2%
14.9%
27.1%
a. Total population is based on the ensemble of CBGs within the specified distance of one or more of the 102 steam electric
power plants with non-zero pollutant loads under the baseline or the regulatory options.
b. The demographic index is an average of the two demographic indicators explicitly named in EO 12898: low-income and
minority.
Source: U.S. EPA analysis, 2020
As shown in Table 14-1 approximately 500,000 people live within one mile of at least one steam electric
power plant currently discharging bottom ash transport water or FGD wastewater to surface waters,
approximately 1.7 million live within three miles, and approximately 88 million people live within 30 miles.
The socioeconomic statistics show that a smaller fraction of communities that live within all analyzed regions
is minority or low-income, when compared to the national average. Of the analyzed regions, communities
within 30 or 50 miles of steam electric power plants have the highest demographic index (24.1 percent),
which is still lower than the national average (27.1 percent). As one moves farther away from the steam
electric power plants, the fraction of the community that is low-income and the percent minority generally
increases.
The simple comparison to the national average may not account for important differences, however, between
states, particularly given the non-uniform geographical distribution of steam electric power plants across the
country. EPA therefore also compared the demographic profile of communities around each plant to that of
the states intersected by each analysis region. Table 14-2 summarizes the results of this comparison, as well as
the results of comparing the demographic profile of each community to the national average. Although the
results in Table 14-1 show that low-income and minority percentages within the various radial distances from
steam electric plants are below the national average when considered as a group, the comparison of individual
analysis regions around each plant to the national and state averages shows that varying shares of
communities within each distance buffer have greater low-income or minority percentages than the national
and state averages. For example, although communities within all distance buffers from steam electric plants
were, in the aggregate, below the national low-income and minority percentages, there are communities
around individual plants with higher proportions of low-income households and/or minority population than
the national or respective state averages. Details of this analysis are included in the docket for this final rule
(DCN SE09377: Environmental Justice Analysis: Code, Inputs, and Outputs). These results highlight the
potential for localized differences based on socioeconomic factors, and do not show uniformly higher
proportions of low-income and minority population across analysis regions relative to state and national
averages.
14-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-2: Socioeconomic Characteristics of Communities Living in Proximity to Steam Electric
Power Plants and Associated Immediate Receiving Reach, Compared to National and State
Averages
Distance
Number
Number of Communities3 Living in Proximity to Steam Electric Plants that...
from
Steam
Electric
Power
Plant
of Steam
Electric
Power
Plants
Have a
Higher
Proportion
of Low
Income
Population
Have a
Higher
Proportion
of Minority
Population
Have a
Higher
Proportion
of Low
Income and
Minority
Population
Have a
Higher
Proportion
of Low
Income
Population
Have a
Higher
Proportion
of Minority
Population
Have a
Higher
Proportion
of Low
Income and
Minority
Population

... than the National Average
... than the State Average13
1 mile
102
40
13
8
39
14
10
3 miles
102
39
10
6
35
14
10
15 miles
102
44
17
10
45
26
15
30 miles
102
49
17
10
39
37
13
50 miles
102
47
21
10
40
39
11
a. In this analysis, a "community" consists of the population associated with the CBGs within the specified distance of each of the
102 steam electric power plants with non-zero pollutant loads under the baseline or the regulatory options.
b. The state average is based on the states intersected by the analysis region around each plant. In cases where an analysis region
intersects multiple states, EPA weighted state statistics based on each state's share of the total population within the analysis
region.
Source: U.S. EPA analysis, 2020
14.2 Distribution of Human Health Impacts and Benefits
The analysis described in Section 14.1 characterizes populations living in proximity to power plants but does
not account for differences across plants in the magnitude of pollutant releases and population exposure.
The second type of analysis looks at the distribution of environmental effects and benefits to further inform
understanding of the potential EJ concerns and the extent to which the regulatory options may mitigate or
exacerbate these concerns. This analysis allows the Agency to report the distribution of benefits or forgone
benefits across population subgroups, including subgroups who may have been historically exposed to a
disproportionate share of environmental impacts.
A significant share of the benefits of the regulatory options comes from the small estimated changes in the
discharges of harmful pollutants to surface waters and associated changes in drinking water quality84 and fish
tissue contamination, and from small estimated changes in air emissions. The sections below discuss the
distribution of health effects for these pathways.
14.2.1 Socioeconomic Characteristics of Populations Affected by Changes in Pollutant Levels in Drinking
Water Sources
EPA estimated the changes in halogen concentrations in PWS source waters affected by steam electric power
plants" discharges, and characterized the populations served by the PWS directly or indirectly affected by
these changes. Chapter 4 discusses the analysis and the approach used to identify the affected population
84 Although EPA did not monetize the benefits of changes in treated drinking water quality, the Agency did look at the distribution
of changes in source water quality used by PWS and uses these results here to assess impacts to different communities for this
pathway.
14-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
based on the county or tribal service areas. That analysis indicates that at the national level, source water
halogen levels decrease in the long term under the final rule (Option A), as well as under Options B and C.85
Table 14-3 summarizes the estimated population potentially affected by changes in drinking water quality
resulting from changes in halogen levels in source waters. The analysis is conducted at the county level and
compares the demographic profile of the affected counties (based on the service areas of affected PWS) to
that of the respective states where each county is located. More than 31 million people, across 286 counties
and 27 states, are estimated to be potentially affected by the small estimated changes in source water quality
under the regulatory options. Most of the 27 states have PWS that serve at least one county with higher
proportions of low-income and/or minority populations than the state average, indicating a potential for
localized effects. Overall, however, counties in service areas potentially affected by changes in source water
quality are not uniformly more low-income and/or minority than their state averages. Details of this analysis
are included in the docket for this final rule (DCN SE09377: Environmental Justice Analysis: Code, Inputs,
and Outputs).
Table 14-3: Socioeconomic Characteristics of Counties in Service Areas of Potentially Affected PWS,
Compared to State Average
State
Number of
Counties in
Service Areas of
Potentially
Affected PWS
Population Served
by Affected PWSa
Number of Counties in Service Areas of Potentially
Affected PWS that...
Have a Higher
Proportion of
Low Income
Population
Have a Higher
Proportion of
Minority
Population
Have a Higher
Proportion of Low
Income and
Minority
Population
... than the State Average
Alabama
24
1,668,000
11
8
6
Arizona
1
9,000
1
1
1
Delaware
1
309,000
0
1
0
District of Columbia
1
649,000
0
0
0
Georgia
12
701,000
7
3
2
Illinois
13
715,000
6
1
1
Indiana
4
201,000
1
0
0
Iowa
5
269,000
4
1
1
Kansas
7
825,000
3
1
1
Kentucky
27
1,469,000
6
6
1
Louisiana
8
992,000
2
5
1
Maryland
9
4,155,000
3
3
1
Massachusetts
2
376,000
0
1
0
Michigan
8
3,440,000
2
4
2
Missouri
16
2,615,000
6
3
2
Nebraska
5
573,000
2
1
1
North Carolina
11
1,555,000
6
4
3
North Dakota
5
33,000
1
0
0
Ohio
9
1,183,000
5
1
1
Oklahoma
3
62,000
1
0
0
Pennsylvania
16
3,860,000
4
3
1
South Carolina
12
1,000,000
5
3
3
85 Halogen levels increase in the short term under Options A and B before decreasing in the long term and decrease under Option C
in both the short and long term. See Section 4.1 for details.
14-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-3: Socioeconomic Characteristics of Counties in Service Areas of Potentially Affected PWS,
Compared to State Average
State
Number of
Counties in
Service Areas of
Potentially
Affected PWS
Population Served
by Affected PWSa
Number of Counties in Service Areas of Potentially
Affected PWS that...
Have a Higher
Proportion of
Low Income
Population
Have a Higher
Proportion of
Minority
Population
Have a Higher
Proportion of Low
Income and
Minority
Population
... than the State Average
South Dakota
43
187,000
18
13
12
Tennessee
20
2,116,000
11
3
1
Utah
2
1,000
1
0
0
Virginia
10
2,345,000
5
9
4
West Virginia
12
305,000
3
5
2
Total
286
31,610,000
114
80
47
a The affected population is based on the total population served reported by SDWIS for affected PWS within each state. However,
not all reported individuals may reside within the designated county and state in cases where a PWS service area extends over
multiple counties (or states).
Source: U.S. EPA analysis, 2020
Table 14-4 summarizes the estimated tribal area population potentially affected by small changes in drinking
water quality as a result of steam electric power plant discharges. The analysis compares the demographic
profile of the affected tribal areas to that of the state where they are located. As shown in the table, affected
tribal areas consistently have a higher minority population than the state average; half of the tribal areas have
minority population percentages greater than 90 percent.
14-6
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-4: Socioeconomic Characteristics of Affected Tribal Areas, Compared to State Average
Affected Tribal Areas
States with
Affected
Tribal
Areas
Total Population
Percent Minority
Percent Low-Income
Demographic Index
Affected
Population3
Total for Tribal
Area
State(s)
Tribal Area
State
Tribal Area
State
Tribal Area
State
Crow Creek
Reservation
SD
1,873
2,151
855,444
94.9%
17.3%
41.4%
13.9%
68.2%
15.6%
Lake Traverse
Reservation
ND; SD
230
10,967
1,600,919
48.1%
15.9%
19.4%
12.6%
33.8%
14.3%
Lower Brule
Reservation
SD
2,116bc
1,594
855,444
94.0%
17.3%
42.7%
13.9%
68.4%
15.6%
Navajo Nation
AZ; NM; UT
1,190
175,005
11,888,715
98.3%
41.6%
40.5%
16.1%
69.4%
28.8%
Pine Ridge
Reservation
NE; SD
8,713
19,779
2,749,365
89.5%
19.3%
50.4%
12.6%
70.0%
16.0%
Rosebud Indian
Reservation
SD
5,619
11,354
855,444
92.0%
17.3%
54.3%
13.9%
73.2%
15.6%
Standing Rock
Reservation
ND; SD
6,839
8,616
1,600,919
79.2%
15.9%
42.3%
12.6%
60.8%
14.3%
Yankton Reservation
SD
1,064
6,676
855,444
50.4%
17.3%
27.3%
13.9%
38.9%
15.6%
a. The affected population is based on the population served by the PWS, as reported in SDWIS. In some cases, the PWS serves both the tribal area and surrounding counties.
b. PWS ID 84690026 serves several reservations and counties. Therefore, SDWIS reported population served was equally distributed over the three reservations served: Lower Brule
Reservation, Pine Ridge Reservation, and Rosebud Indian Reservation.
c. PWS ID 84690441 serves the Lower Brule Reservation and surrounding South Dakota counties. As a result, the SDWIS reported population served exceeds the Census reported total
population of the reservation. The affected percentage of tribal area was adjusted to 80 percent to reflect that the majority of the reservation is likely served by the affected PWS.
Source: U.S. EPA analysis, 2020
14-7
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
14.2.2 Socioeconomic Characteristics of Populations Affected by Changes in Exposure to Pollutants via
the Fish Ingestion Pathway
This section first evaluates the socioeconomic characteristics of communities within 50 miles of immediate
reaches that receive discharges from steam electric power plants, as well as of downstream reaches.86 The
section then presents the distribution of EPA's quantified human health effects resulting from the small
estimated changes in exposure to selected pollutants via consumption of self-caught fish.87 Chapter 5 provides
more details on the approach used to identify the affected recreational and subsistence fisher population,
estimate exposure based on race and ethnicity-specific data, quantify health effects, and monetize benefits.
As shown in Table 14-5 and Table 14-6 (for Period 1 and Period 2, respectively), the community living in
proximity to reaches with increases in pollutant levels under Options A, B, and C has a smaller proportion of
low-income and minority population than the national average. For many pollutants, the community living in
proximity to reaches with decreases or no change in pollutant levels under Options A, B, and C has a greater
proportion of low-income and minority population than the national average in many cases.
86 The analysis focuses on selected pollutants that have at least one exceedance of human health criteria across the options and
periods (antimony, arsenic, cadmium, cyanide, lead, manganese, and thallium) plus mercury to provide additional insight on the
distribution of benefits discussed in Chapter 5. Table 5-7 provides additional information on reaches with exceedances of human
health criteria.
87 The first analysis defines "communities in proximity to reaches" as the aggregate populations residing in CBGs within 50 miles
of all reaches within 300 km of steam electric power plant outfalls. This analysis provides total population and does not make
adjustments for the fraction of this population that consumes self-caught fish.
14-8
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-5: Socioeconomic Characteristics of Communities Living in Proximity to Reaches with Changes to Selected Pollutant
Concentrations under the Regulatory Options, Compared to Baseline (Period 1)
Pollutant
Changes in
Concentrations
Number of Reaches
Percent Minority
Percent Low-Income
Demographic Index
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Antimony
Decreases
0
0
811
0.0%
0.0%
37.3%
0.0%
0.0%
15.0%
0.0%
0.0%
26.2%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
8,839
33.1%
33.1%
33.2%
14.0%
14.0%
14.0%
23.6%
23.6%
23.6%
Arsenic
Decreases
0
0
1,329
0.0%
0.0%
37.5%
0.0%
0.0%
14.7%
0.0%
0.0%
26.1%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
8,321
33.1%
33.1%
32.9%
14.0%
14.0%
14.0%
23.6%
23.6%
23.5%
Cadmium
Decreases
0
0
2,854
0.0%
0.0%
38.1%
0.0%
0.0%
13.9%
0.0%
0.0%
26.0%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
6,796
33.1%
33.1%
31.0%
14.0%
14.0%
14.3%
23.6%
23.6%
22.7%
Cyanide
Decreases
0
0
4,729
0.0%
0.0%
35.2%
0.0%
0.0%
14.3%
0.0%
0.0%
24.8%
No changes
5,711
5,711
1,114
35.0%
35.0%
34.7%
14.0%
14.0%
15.2%
24.5%
24.5%
25.0%
Increases
524
524
392
29.8%
29.8%
27.4%
16.6%
16.6%
11.6%
23.2%
23.2%
19.5%
Lead
Decreases
0
0
960
0.0%
0.0%
40.2%
0.0%
0.0%
15.0%
0.0%
0.0%
27.6%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
8,690
33.1%
33.1%
32.5%
14.0%
14.0%
13.9%
23.6%
23.6%
23.2%
Manganese
Decreases
0
0
3,052
0.0%
0.0%
37.3%
0.0%
0.0%
13.9%
0.0%
0.0%
25.6%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
6,598
33.1%
33.1%
31.3%
14.0%
14.0%
14.3%
23.6%
23.6%
22.8%
Mercury
Decreases
0
0
331
0.0%
0.0%
37.6%
0.0%
0.0%
15.9%
0.0%
0.0%
26.8%
No changes
1,287
1,259
794
32.2%
33.9%
26.0%
16.0%
15.6%
14.7%
24.1%
24.8%
20.4%
Increases
9,167
9,204
9,329
33.7%
33.5%
33.5%
14.0%
13.9%
14.0%
23.9%
23.7%
23.8%
Thallium
Decreases
0
0
3,846
0.0%
0.0%
36.2%
0.0%
0.0%
14.2%
0.0%
0.0%
25.2%
No changes
1,403
1,346
804
35.8%
35.9%
25.6%
15.1%
15.1%
14.7%
25.5%
25.5%
20.2%
Increases
9,051
9,108
5,804
33.1%
33.1%
31.6%
14.0%
14.0%
14.1%
23.6%
23.6%
22.9%
United States
39.2%
14.9%
27.1%
Source: U.S. EPA analysis, 2020
14-9
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-6: Socioeconomic Characteristics of Communities Living in Proximity to Reaches with Changes to Selected Pollutant
Concentrations under the Regulatory Options, Compared to Baseline (Period 2)
Pollutant
Changes in
Concentrations
Number of Reaches
Percent Minority
Percent Low-Income
Demographic Index
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Option
A (Final
Rule)
Option
B
Option
C
Antimony
Decreases
292
292
2,966
56.8%
56.8%
36.6%
17.5%
17.5%
15.6%
37.2%
37.2%
26.1%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
8,100
8,100
6,024
31.3%
31.3%
31.2%
14.1%
14.1%
13.6%
22.7%
22.7%
22.4%
Arsenic
Decreases
588
588
4,993
52.0%
52.0%
34.3%
16.9%
16.9%
15.0%
34.5%
34.5%
24.7%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
7,804
7,804
3,997
31.4%
31.4%
31.3%
14.1%
14.1%
13.5%
22.8%
22.8%
22.4%
Cadmium
Decreases
1,150
1,351
5,463
37.2%
42.1%
37.2%
15.9%
13.3%
14.5%
26.6%
27.7%
25.9%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
7,242
7,041
3,527
31.8%
30.5%
24.5%
14.0%
14.4%
13.8%
22.9%
22.5%
19.2%
Cyanide
Decreases
1,895
2,096
5,841
32.8%
36.4%
35.6%
15.8%
14.1%
14.5%
24.3%
25.3%
25.1%
No changes
4,340
4,139
394
35.0%
33.8%
28.5%
13.8%
14.3%
12.4%
24.4%
24.1%
20.5%
Increases
0
0
0
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
Lead
Decreases
513
513
3,612
55.0%
55.0%
35.9%
17.2%
17.2%
15.6%
36.1%
36.1%
25.8%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
7,879
7,879
5,378
31.3%
31.3%
31.2%
14.1%
14.1%
13.5%
22.7%
22.7%
22.4%
Manganese
Decreases
1,711
1,912
5,841
35.3%
39.5%
36.7%
15.9%
13.9%
14.5%
25.6%
26.7%
25.6%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
6,681
6,480
3,149
31.8%
30.4%
24.7%
13.9%
14.3%
13.8%
22.9%
22.4%
19.3%
Mercury
Decreases
588
588
4,740
52.0%
52.0%
35.4%
16.9%
16.9%
15.4%
34.5%
34.5%
25.4%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
7,804
7,804
4,250
31.4%
31.4%
30.7%
14.1%
14.1%
13.2%
22.8%
22.8%
22.0%
Thallium
Decreases
1,150
1,351
5,579
37.2%
42.1%
37.0%
15.9%
13.3%
14.5%
26.6%
27.7%
25.8%
No changes
2,062
2,062
1,464
37.0%
37.0%
36.4%
14.0%
14.0%
13.6%
25.5%
25.5%
25.0%
Increases
7,242
7,041
3,411
31.8%
30.5%
24.4%
14.0%
14.4%
13.8%
22.9%
22.5%
19.1%
United States
39.2%
14.9%
27.1%
Source: U.S. EPA analysis, 2020
14-10
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
As described in Chapter 5, EPA quantified the distribution of human health effects associated with lead and
mercury exposure among two types of fishers (recreational and subsistence) and their families.88 Because
quantified health effects are limited to changes in IQ losses from exposure to lead and mercury, the following
discussion focuses on the population of infants and children ages 0 to 7 potentially exposed to steam electric
pollutants via consumption of contaminated fish. The quantified effects relative to the baseline are small,
overall.
Table 14-7 summarizes the estimated number of children ages 0 to 7 exposed to lead and infants exposed to
mercury in-utero via the consumption of self-caught fish in the total population of fishers and in population
subgroups that may be indicative of EJ concerns.89 As shown in the table, of the approximately 1.6 million
children ages 0 to 7 potentially exposed to steam electric power plant wastewater pollutants through fish
tissue consumption, an estimated 15.9 percent are low-income, 64.0 percent are minority, and 11.1 percent are
both low-income and minority. Overall, 68.8 percent of potentially exposed children are categorized in at
least one or more EJ subgroup based on household income or race/ethnicity, while 31.2 percent are neither
minority nor low-income.9" EPA estimates that approximately 151,000 infants are potentially exposed to
mercury in-utero. The potentially exposed infant population and its characteristics are based on the number of
women of child-bearing age (15 to 44 years old) multiplied by the average, ethnic group-based fertility rates,
and their socioeconomic characteristics.
Table 14-7: Characteristics of Children Potentially Exposed to Steam Electric Power Plant
Pollutants via Consumption of Self-caught Fish
Subgroup
Minority
Non-Minority
Total
Lead (Children Ages 0-7)
Low-income
179,693
11.1%
77,559
4.8%
257,252
15.9%
Minority
853,552
52.8%
504,825
31.2%
1,358,378
84.1%
Total
1,033,245
64.0%
582,384
36.0%
1,615,629
100.0%
Mercury (Infants)3
Low-income
25,710
11.4%
10,294
4.6%
36,003
16.0%
Minority
125,393
55.6%
64,140
28.4%
189,534
84.0%
Total
151,103
67.0%
74,434
33.0%
225,537
100.0%
a. Potentially exposed infant population is based on the number of women of child-bearing age (15 to 44 years old) multiplied
by the average, ethnic group-based fertility rates. Therefore, it reflects socio-demographic characteristics of women of child-
bearing age.
Source: U.S. EPA Analysis, 2020
88 As discussed in Chapter 5, the regulatory options did not result in material changes in arsenic-related health effects.
89 Because data on socioeconomic characteristics of freshwater fishers are not available at the CBG level, EPA used the same
socioeconomic characteristics for fishers as those of the general population residing within a 50-mile radius from the affected
reaches.
90 In the discussion, EPA uses minority/low-income percentages based on the population potentially exposed to lead because the
population potentially exposed to lead (children aged 0-7) encompasses the population potentially exposed to mercury (infants).
14-11
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
The distribution of adverse health effects is a function of the characteristics of the affected population (Table
14-7), including age and sex,91 ethnicity-specific exposure factors,92 and water quality. Table 14-8 shows the
distribution of changes in adverse health effects under each of the regulatory options for minority and low-
income subgroups, as well as non-minority and not low-income subgroups. The first two subgroups are the
primary interest of this analysis as potentially indicative of EJ concerns.
The quantified effects relative to the baseline are small, both overall and for the subgroups. The distribution of
the small changes in IQ points across the subgroups shows that the changes (positive or negative) estimated
over the total population of exposed children predominantly affect children in the minority subgroup.
In the analysis of health benefits for the fish ingestion pathway (see Chapter 5), EPA assumed that 5 percent
of the exposed population are subsistence fishers, and that the remaining 95 percent are recreational fishers.
Subsistence fishers consume more self-caught fish than recreational fishers and can therefore be expected to
experience higher health risks associated with exposure to steam electric pollutants in fish tissue. Table 14-9
shows the distribution of changes in adverse health effects among children in households of subsistence
fishers and recreational fishers.
Here also, the quantified effects relative to the baseline are small, both overall and across fisher subgroups.
The distribution of changes in IQ points across children of the two fisher subgroups shows that effects on
children in subsistence fishers' households from exposure to lead-contaminated fish tissue are
disproportionate to their share of the affected population (see Table 14-9 for details). While children from
subsistence fishers' households account for 5 percent of the total exposed populations, they incur 6.4 percent
of the effects (positive or negative) under the regulatory options, compared to baseline.
91 Some adverse health effects are analyzed only for individuals in certain age groups. For example, IQ point decrements from
exposure to lead are calculated for children 0 to 7 years old and the baseline exposure therefore depends on the number of
children within this age group in the affected population in each socioeconomic subgroup. IQ point decrements from exposure to
mercury are calculated for infants born within the analysis period and baseline exposure depends on the number of women of
childbearing age (and fertility rates) in the affected population.
92 Ethnicity-specific factors that determine exposure to pollutants in fish tissue include the assumed fish consumption rates and
average fertility rate. For example, Asian/Pacific Islander fishers have fish consumption rates that are 1.4 times and 1.9 times
those of While (non-Hispanic) fishers for recreational and subsistence fishing modes, respectively.
14-12
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-8: Estimated Distribution of IQ Point Changes from Lead and Mercury Exposure Via Self-caught Fish Consumption Under the
Regulatory Options, Compared to Baseline (2021 to 2047)
Pollutant
and
Population
Regulatory
Option
Percent Minority
Percent Low-Income
Percent Non-Minority, Not Low-
Income
Total
Positive IQ
Change (% of
Exposed
Population)
Negative IQ
Change (% of
Exposed
Population)
Positive IQ
Change (% of
Exposed
Population)
Negative IQ
Change (% of
Exposed
Population)
Positive IQ
Change (% of
Exposed
Population)
Negative IQ
Change (% of
Exposed
Population)
Positive IQ
Change (% of
Exposed
Population)
Negative IQ
Change (% of
Exposed
Population)
% of Children Ages 0 to 7
Exposed to Lead
64.0%
15.9%
31.2%
100.0%
Children
Exposed to
Leadab

- -
- -
- -
- -
- -
- -

Option A
(Final Rule)
0 (0.0%)
-18 (64.0%)
0 (0.0%)
-5 (15.9%)
0 (0.0%)
>-1 (31.2%)
0 (0.0%)
-19 (100.0%)
Option B
0 (0.0%)
-11 (64.0%)
0 (0.0%)
-3 (15.9%)
0 (0.0%)
>-1 (31.2%)
0 (0.0%)
-11 (100.0%)
Option C
13 (8.8%)
-1 (55.2%)
3 (2.2%)
0 (13.7%)
0 (0.0%)
>-1 (31.2%)
13 (8.8%)
-2 (91.2%)
% of Infants Exposed to
Mercury In-utero
67.0%
16.0%
28.4%
100.0%
Infants
Exposed to
Mercuryab

- -
- -
- -
- -
- -
- -

Option A
(Final Rule)
122 (35.9%)
-201 (31.1%)
34 (6.3%)
-46 (9.7%)
0 (0.0%)
-105 (28.4%)
122 (35.9%)
-323 (64.1%)
Option B
151 (35.9%)
-190 (31.1%)
38 (6.3%)
-42 (9.7%)
0 (0.0%)
-90 (28.4%)
151 (35.9%)
-296 (64.1%)
Option C
199 (36.1%)
-78 (30.9%)
46 (6.6%)
-15 (9.4%)
0 (0.0%)
-43 (28.4%)
199 (36.1%)
-128 (63.9%)
-Not estimated
a. Negative values represent forgone benefits and positive values represent realized benefits.
b. EPA estimates that options A and B will result in an overall increase in exposure to lead and mercury and thus an increase in IQ losses (i.e., negative changes). Option C results in an
overall decrease in IQ losses (i.e., positive changes).
c. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C. Only the total
value is provided due to changes in the presentation of the EJ subgroups.
Source: U.S. EPA Analysis, 2020
14-13
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-9: Estimated Distribution of Changes in IQ Point Changes from Lead and Mercury Exposure Via Self-caught Fish Consumption
under the Regulatory Options, Compared to Baseline, by Fishing Mode (2021 to 2047)

Subsistence Fishers

Recreational Fishers

Total

Pollutant and
Exposed
Population

(5% of Population)

(95% of Population)

Regulatory
Option
Positive IQ
Change(% of
Exposed
Population)
Negative IQ Change
(% of Exposed
Population)
Positive IQ
Change(% of
Exposed
Population)
Negative IQ
Change (% of
Exposed
Population)
Positive IQ Change
(% of Exposed
Population)
Negative IQ
Change(% of
Exposed
Population)

Children
Exposed to
Option A
(Final Rule)
0
(0.0%)
>-1
(6.4%)
0
(0.0%)
-18
(93.6%)
0
(0.0%)
-19
(100.0%)
Lead3
Option B
0
(0.0%)
>-1
(6.4%)
0
(0.0%)
-11
(93.6%)
0
(0.0%)
-11
(100.0%)

Option C
0
(0.0%)
>-1
(6.4%)
13
(8.8%)
-1
(84.8%)
13
(8.8%)
-2
(91.2%)

Infants
Exposed to
Option A
(Final Rule)
20
(2.3%)
-55
(4.1%)
102
(33.6%)
-268
(60.0%)
122
(35.9%)
-323
(64.1%)
Mercuryab
Option B
26
(2.3%)
-51
(4.1%)
126
(33.6%)
-245
(60.0%)
151
(35.9%)
-296
(64.1%)

Option C
34
(2.3%)
-22
(4.1%)
165
(33.8%)
-106
(59.8%)
199
(36.1%)
-128
(63.9%)
a. Negative values represent forgone benefits and positive values represent realized benefits.
c. Option D corresponds to the proposed Option 1. EPA did not reanalyze this option for the final rule. All results shown for Option D are based on the 2019 analysis, as detailed in the
2019 BCA (U.S. EPA, 2019a). As such, the values do not reflect changes in the baseline, plant universe, and other analytical inputs for the analysis of Options A, B, and C.
Source: U.S. EPA Analysis, 2020
14-14
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
14.2.3 Socioeconomic Characteristics of Populations Affected by Changes in Exposure to Air Pollutants
EPA quantified the human health effects resulting from the small estimated changes in EGU emissions of
NOx, SO2, and PM2 5 under the final rule (Option A), compared to baseline. This analysis, which is detailed in
Chapter 8, included estimated changes in the incidence of adverse health effects resulting from exposure to
PM2 5 and ozone using BenMAP-CE. To provide insight into the potential EJ implications of these changes,
EPA reviewed the distribution of changes in adverse health effects projected by BenMAP-CE across counties
and, using ACS data, summarized the socioeconomic characteristics of the populations living in the counties
projected to see changes in the number of cases of adverse health outcomes during the period of analysis
(2021-2047). This analysis entailed: (1) exporting county-level projected health outcomes for each analysis
year; (2) summing cases in each county over the 27-year analysis period; (3) categorizing counties according
to whether they see a net increase or net decrease in adverse health outcomes over the period of analysis under
the final rule as compared to the baseline, and (4) summarizing the socioeconomic characteristics of people
living in the counties.
This analysis reflects the geographical distribution of air quality changes (both in terms of direction and
magnitude of the changes) and differences in the socioeconomic characteristics of the counties that see these
projected changes in air quality. EPA's approach to characterizing risks among these subgroups did not
account for differences in susceptibility or vulnerability among subgroups according to income or
race/ethnicity by, for example, using concentration-response relationships that account for population race or
ethnicity. The approach distributes the estimated changes in the number of adverse health outcomes within a
county uniformly across all people residing within the county.
Table 14-10 summarizes the estimated distribution of selected health outcomes quantified in BenMAP-CE.
The presentation is similar to that used above for characterizing populations living near reaches with
improving or degrading water quality, except that in this analysis, the analyzed population corresponds to the
population of the counties with net total increase or net total decrease in the number of projected adverse
health outcomes due to EGU emissions changes and resulting changes in air quality under the final rule as
compared to the baseline. As shown in Table 14-10, a larger share of the U.S. population see positive benefits
(a net reduction in adverse health outcomes) than forgone benefits (a net increase in adverse health outcomes)
under the final rule.
Minority populations are estimated to accrue a disproportionate share of the benefits from the final rule;
whereas 39.2 percent of the U.S. population is minority, between 42.1 percent and 43.8 percent of the net
avoided adverse health outcomes are estimated to accrue to minority populations. Conversely, minority
populations see an estimated 32.3 percent to 35.2 percent of the net increases in adverse outcomes, which is
less than their 39.2 percent share of the general U.S. population.
The distribution of net changes in health outcomes relative to income is more uniform. The shares of net
changes in adverse health outcomes (increase or decrease) accruing to the low-income subgroup approach the
14.9 percent of the general U.S. population that is low-income. However, a slightly larger share of the
benefits (i.e.. net decrease in adverse health outcomes), ranging between 15.1 and 15.4 percent, accrues to the
low-income subgroup than the subgroup's representation in the general U.S. population (14.9 percent).
14-15
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-10: Socioeconomic Characteristics of Populations Projected to see Net Increases and
Decreases in Adverse Health Outcomes from Changes in Exposure to PM2.5 and Ground-level
Ozone Under the Final Rule, Compared to Baseline, in 2021-2047
Quantified Health Outcome
Net Direction
of Change3
% of Total
Population
% of Change
Accruing to
Minority
Population
% of Change
Accruing to Low
Income Population
Premature mortality15
Net Increase
46.9%
33.4%
13.8%
Net Decrease
53.1%
42.8%
15.3%
Non-fatal heart attacks (age > 18)
Net Increase
45.9%
33.8%
13.8%
Net Decrease
54.1%
42.2%
15.3%
Hospital admissions—respiratory
(all ages)
Net Increase
46.0%
32.3%
13.7%
Net Decrease
54.0%
43.5%
15.4%
Hospital admissions-
cardiovascular (age >20)
Net Increase
45.9%
33.9%
13.8%
Net Decrease
54.1%
42.1%
15.3%
Emergency room visits for asthma
(all ages)
Net Increase
47.9%
33.2%
14.0%
Net Decrease
52.1%
43.1%
15.2%
Acute bronchitis (age 8-12)
Net Increase
46.0%
33.8%
13.8%
Net Decrease
54.0%
42.2%
15.3%
Lower respiratory symptoms (age
7-14)
Net Increase
46.0%
33.8%
13.8%
Net Decrease
54.0%
42.2%
15.3%
Upper respiratory symptoms
(asthmatics age 9-11)
Net Increase
46.0%
33.8%
13.8%
Net Decrease
54.0%
42.2%
15.3%
Exacerbated asthma (asthmatics
age 6-18)
Net Increase
52.4%
33.5%
14.0%
Net Decrease
47.6%
43.8%
15.3%
Lost work days (age 18-65)
Net Increase
46.0%
33.8%
13.8%
Net Decrease
54.0%
42.2%
15.3%
Minor restricted-activity days (age
18-65)
Net Increase
47.3%
33.0%
13.9%
Net Decrease
52.7%
43.2%
15.3%
School absence days (age 5 17)
Net Increase
54.3%
35.2%
14.2%
Net Decrease
45.7%
42.1%
15.1%
Subgroup as % of U.S. Population
39.2%
14.9%
a. Reflects the net direction of total changes in cases over the period of 2021 through 2047. Some individual years may have
negative changes and other years may have positive changes.
b. Reported percentages for premature mortality reflect the upper bound of mortality incidence estimates from Jerret et al.
(2009) and Lepeule et al. (2012). The difference between the percentages for the upper bound and lower bound (not reported)
of mortality incidence estimates is very small (<0.3%).
Source: U.S. EPA analysis, 2020
14.3 EJ Analysis Findings
Overall, the various analyses show that environmental changes under the regulatory options analyzed,
including the final rule, may affect minority and/or low income populations to different degrees across
environmental media, exposure pathways, and over time, but the effects (positive or negative) of the changes
will be small.
Communities living near steam electric power plants {i.e., up to 50 miles) tend to have a lower proportion of
low-income households and minority population than the national average, when considered in the aggregate,
but there may be localized EJ considerations for some communities near individual plants that have higher
proportions of low-income or minority populations than the national and/or state average (see Table 14-2).
14-16
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
EPA's analysis considered the distribution of effects on populations near both immediate and downstream
reaches, in downstream PWS service areas, and in adjacent airsheds to assess whether low-income and/or
minority populations may be disproportionately affected by changes under the final rule (Option A) and other
regulatory options. The analysis shows that the EJ population subgroups are not excluded from the benefits
associated with the regulatory options, including the final rule. For example, projected air quality changes
under the final rule (Option A) may disproportionately benefit minority and low-income populations based on
the socioeconomic characteristics of populations of counties with changes in PM2.5 and ozone levels during
the period of analysis. Additionally, estimated foregone benefits related to water quality may
disproportionately affect minority and subsistence fisher populations. However, the magnitude of the changes
(positive and negative) and associated benefits (including foregone benefits) is small, relative to the baseline,
both overall across the exposed population, and across socioeconomic and fisher subgroups.
14.4 Limitations and Uncertainties
This EJ analysis incorporates the limitations and uncertainties associated with the human health effects
analyses (see Chapter 4, Chapter 5, and Chapter 8) regarding pollutant exposure, and incidence of adverse
health outcomes. In addition, the EJ analysis embeds uncertainty derived from the application of uniform
inputs across the estimated population exposed to pollutant discharges when factors may instead vary across
socioeconomic characteristics. In summary, use of average values across the entire population of the United
States (or within a state or a county associated with a PWS service area) instead of inputs that reflect specific
socioeconomic factors may over- or understate inequities present in the baseline and the differential impacts
or benefits to low-income or minority populations from changes due to the regulatory options.
Note that the effect on benefits estimates indicated in the second column of the table refers to the magnitude
of the benefits rather than the direction (i.e.. a source of uncertainty that tends to underestimate benefits
indicates expectation for larger forgone benefits or for larger realized benefits).
Table 14-11: Limitations and Uncertainties in EJ Analysis
Uncertainty/Limitation
Effect on EJ Analysis
Notes
EPA estimated that all fishers travel
up to 50 miles.
Uncertain
Certain EJ subpopulations may tend to fish closer to
home (e.g., low-income and subsistence fishers). To the
extent that these people fish predominantly from waters
receiving discharges from steam electric power plants,
they may be exposed to relatively higher concentrations
of pollutants. Conversely, people who live farther from
steam electric power plants may predominantly fish
from waters not affected by pollutants in steam electric
power plant discharges and be exposed to relatively
lower concentrations of pollutants.
EPA estimated that subsistence fishers
are 5 percent of all fishers, and
applied this estimate uniformly across
all socioeconomic groups.
Underestimate
A relatively higher share of EJ groups may be subsistence
fishers. This could increase inequities in the baseline and
affect the extent to which the regulatory options may
increase or decrease these inequities.
EPA applied uniform fishing
participation rates, FCA responses,
and catch and release practices across
the entire population.
Uncertain
Differences in behavior across socioeconomic groups
may result in a different distribution of baseline and
regulatory option impacts.
14-17
-------
BCAfor Revisions to Steam Electric Power Generating ELGs
14: Environmental Justice
Table 14-11: Limitations and Uncertainties in EJ Analysis
Uncertainty/Limitation
Effect on EJ Analysis
Notes
EPA used the counties served by PWS,
as reported in the SDWIS database, as
representative of the population
potentially affected by changes in
halogenated DBPs due to steam
electric power plant discharges.
Uncertain
Counties and tribal areas can be served by multiple PWS
and some PWS serve people across multiple counties,
such that the affected population may have different
socioeconomic characteristics.
EPA used the SDWIS database to
identify counties served by affected
PWS. For any PWS IDs without any
associated county information, EPA
used the PWS Name and the PWS
latitude and longitude to identify
associated tribal areas.
Uncertain
There may be some PWSs that serve counties and tribal
areas. However, if only the county was listed in SDWIS,
the EJ analysis does not account for the associated tribal
area.
The IEUBK model does not capture
very small changes.
Underestimate
The human health effects from changes in lead exposure
analysis is based on IEUBK model geometric mean PbB
values for each cohort in each CBG under the baseline
and the regulatory options. The IEUBK model processes
daily intake to two decimal places (ng/day), so some of
the change between the baseline and regulatory options
is not accounted for by using the model (i.e., IEUBK does
not capture very small changes) since the estimated
changes in health effects are driven by very small
changes across large populations. This aspect of the
model contributes to potential underestimation of the
lead-related health effects in children in the different
subgroups.
EPA's approach to characterizing risks
among subgroups did not account for
differences in susceptibility or
vulnerability among subgroups
according to income or race/ethnicity
by, for example, using air pollutant
concentration-response relationships
that account for population race or
ethnicity.
Uncertain
This analysis reflects solely the geographical distribution
of air quality changes and differences in socioeconomic
characteristics of the counties that see projected
changes in air quality. People in different subgroup may
be more or less susceptible to changes in pollutant
exposure.
The spatial resolution of information
used in the analysis on changes in air
quality limits the degree to which
changes in population subgroup
exposure can be characterized.
Uncertain
EPA used county-level input data to assess the
distribution of changes in air quality and their impacts
on different populations. This is a fairly coarse resolution
for detecting differences in exposure or risk among
population subgroups.
14-18
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
15 Cited References
United States of America v. Duke Energy Business Services, LLC, Duke Energy Carolinas, LLC, Duke
Energy Progress, Inc., No. No. 5:15-CR-62-H, No. 5:15-CR-67-H, No. 5:15-CR-68-H, Joint Factual
Statement (United States District Court for the Eastern District of North Carolina Western Division
2015).
Agency for Toxic Substances and Disease Registry. (2004). Interaction Profiles for Toxic Substance -
Arsenic, Cadmium, Chromium, Lead. Final Interaction Profile - May 2004. Retrieved from
https: //www .atsdr. cdc. gov/interactionprofiles/ip04 .html
Agency for Toxic Substances and Disease Registry. (2009). Interaction Profiles for Toxic Substances - Lead,
Manganese, Zinc, and Copper. Final Interaction Profile - May 2004. Retrieved from
https://www.atsdr.cdc.gov/interactionprofiles/ip06.html
Anthoff, D., & Tol, R. S. J. (2010). On international equity weights and national decision making on climate
change. Journal of Environmental Economics and Management, (50(1), 14-20.
Anthoff, D., & Tol, R. S. J. (2013). The uncertainty about the social cost of carbon: A decomposition analysis
using fund. Climatic Change, 777(3), 515-530.
Aoki, Y., Brody, D. J., Flegal, K. M., Fakhouri, T. H. I., Parker, J. D., & Axelrad, D. A. (2016). Blood lead
and other metal biomarkers as risk factors for cardiovascular disease mortality. Medicine, 95(1),
e2223. doi: 10.1097/MD.0000000000002223
Arrow, K., Cropper, M., Gollier, C., Groom, B., Heal, G., Newell, R., . . . Portney, P. (2013). Determining
benefits and costs for future generations. Science, 347(6144), 349-350.
Ator, S. W. (2019). Spatially referenced models of streamflow and nitrogen, phosphorus, and suspended-
sediment loads in streams of the northeastern United States (2019-5118). Retrieved from Reston,
VA: http://pubs.er.usgs.gov/publication/sir20195118
Axelrad, D. A., Bellinger, D. C., Ryan, L. M., & Woodruff, T. J. (2007). Dose-response relationship of
prenatal mercury exposure and IQ: an integrative analysis of epidemiologic data. Environmental
health perspectives, 775(4), 609-615.
Baker, K. R., & Kelly, J. T. (2014). Single source impacts estimated with photochemical model source
sensitivity and apportionment approaches. Atmospheric Environment, 96, 266-274.
doi:https://doi.org/10.1016/i.atmosenv.2014.07.042
Bayer, P., Keohane, N., & Timmins, C. (2006). Migration and hedonic valuation: The case of air quality.
National Bureau of Economic Research Working Paper Series, Working Paper No. 12106.
Bergstrom, J. C., & De Civita, P. (1999). Status of benefits transfer in the United States and Canada: a review.
Canadian Journal of Agricultural Economics/Revue canadienne d'agroeconomie, 47(1), 79-87.
Bergstrom, J. C., & Taylor, L. O. (2006). Using meta-analysis for benefits transfer: Theory and practice.
Ecological economics, 60(2), 351-360.
Beron, K., Murdoch, J., & Thayer, M. (2001). The Benefits of Visibility Improvement: New Evidence from
the Los Angeles Metropolitan Area. Journal of Real Estate Finance and Economics, 22(2/3), 319-
337.
Bin, O., & Czajkowski, J. (2013). The impact of technical and non-technical measures of water quality on
coastal waterfront property values in South Florida. Marine Resource Economics, 28( 1), 43-63.
Boyle, K. J., Paterson, R., Carson, R., Leggett, C., Kanninen, B., Molenar, J., & Neumann, J. (2016). Valuing
shifts in the distribution of visibility in national parks and wilderness areas in the United States.
Journal of environmental management, 173, 10-22. doi: 10.1016/j.jenvman.2016.01.042
15-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Boyle, K. J., Poor, P. J., & Taylor, L. O. (1999). Estimating the demand for protecting freshwater lakes from
eutrophication. American journal of agricultural economics, 81(5), 1118-1122.
Brown, T. C., & Gregory, R. (1999). Why the WTA-WTP disparity matters. Ecological economics, 28(3),
323-335. doi:https://doi.org/10.1016/S0921-8009(98)00050-0
Burger, J. (2004). Fish consumption advisories: knowledge, compliance and why people fish in an urban
estuary. Journal of Risk Research, 7(5), 463-479.
Burger, J. (2013). Role of self-caught fish in total fish consumption rates for recreational fishermen: Average
consumption for some species exceeds allowable intake. J Risk Red, 7(5(8), 1057-1075.
doi:10.1080/13669877.2013.788546.
California Environmental Protection Agency. (2011). Public Health Goal for Hexavalent Chromium (CR VI)
in Drinking Water.
Canadian Council of Ministers of the Environment. (2001). Canadian water quality guidelines for the
protection of aquatic life: CCME Water Quality Index 1.0, Technical Report. Retrieved from
Canadian Council of Ministers of the Environment, Winnipeg:
Carruthers, T., & Wazniak, C. (2003). Development of a water quality index for the Maryland Coastal Bays.
Maryland's Coastal Bays: Ecosystem Health Assessment, 4-59.
Centers for Disease Control and Prevention, & Department of Health and Human Services. (2009). Fourth
National Report on Human Exposure to Environmental Chemicals. Retrieved from
https://www.cdc.gov/exposurereport/pdf/fourthreport.pdf
Chay, K. Y., & Greenstone, M. (1998). Does air quality matter? Evidence from the housing market. National
Bureau of Economic Research Working Paper Series, Working Paper No. 6826.
Chowdhury, R., Ramond, A., O'Keeffe, L. M., Shahzad, S., Kunutsor, S. K., Muka, T., . . . Di Angelantonia,
E. (2018). Environmental toxic metal contaminants and risk of cardiovascular disease: systematic
review and meta-analysis. BMJ (Clinical research ed), 362, k3310.
doi: https ://doi. org/10.113 6/bmi ,k3 310
Clark, E. H., Haverkamp, J. A., & Chapman, W. (1985). Eroding soils. The off-farm impacts: Conservation
Foundation.
Cleveland, L. M., Minter, M. L., Cobb, K. A., Scott, A. A., & German, V. F. (2008). Lead Hazards for
Pregnant Women and Children: Part 1: Immigrants and the poor shoulder most of the burden of lead
exposure in this country. Part 1 of a two-part article details how exposure happens, whom it affects,
and the harm it can do. The American Journal of Nursing, 108(10), 40-49.
Cohan, D., & Napelenok, S. (2011). Air Quality Response Modeling for Decision Support. Atmosphere, 2,
407-425. doi: 10.3390/atmos2030407
Cohan, D. S., Hakami, A., Hu, Y., & Russell, A. G. (2005). Nonlinear Response of Ozone to Emissions:
Source Apportionment and Sensitivity Analysis. Environmental Science & Technology, 39(11), 6739-
6748. doi:10.1021/es048664m
Crump, K. S., Van Landingham, C., Bowers, T. S., Cahoy, D., & Chandalia, J. K. (2013). A statistical
reevaluation of the data used in the Lanphear et al. pooled-analysis that related low levels of blood
lead to intellectual deficits in children. Critical reviews in toxicology, 43(9), 785-799.
Cude, C. G. (2001). Oregon water quality index a tool for evaluating water quality management effectiveness
1. JAWRA Journal of the American Water Resources Association, 37(1), 125-137.
Deacon, J. E., Kobetich, G., Williams, J. D., & Contreras, S. (1979). Fishes of North America endangered,
threatened, or of special concern: 1979. Fisheries, 4(2), 29-44.
Dee, L. E., Cowles, J., Isbell, F., Pau, S., Gaines, S. D., & Reich, P. B. (2019). When do ecosystem services
depend on rare species. Trends in Ecology and Evolution, doi: 10.1016/j.tree.2019.03.010
15-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Desvousges, W. H., Smith, V. K., & Fisher, A. (1987). Option price estimates for water quality
improvements: a contingent valuation study for the Monongahela River. Journal of Environmental
Economics and Management, 14, 248-267. doi:httos://doi.org/10.1016/0095-0696(87)90019-2
Ding, D., Zhu, Y., Jang, C., Lin, C.-J., Wang, S., Fu, J., . . . Qiu, X. (2016). Evaluation of health benefit using
BenMAP-CE with an integrated scheme of model and monitor data during Guangzhou Asian Games.
Journal of Environmental Sciences, 42, 9-18. doi:https://doi.org/10.1016/i.ies.2015.06.003
Dunker, A. M., Yarwood, G., Ortmann, J. P., & Wilson, G. M. (2002). The Decoupled Direct Method for
Sensitivity Analysis in a Three-Dimensional Air Quality Model Implementation, Accuracy, and
Efficiency. Environmental Science & Technology, 36(13), 2965-2976. doi: 10.1021/esOl 12691
Dunnette, D. A. (1979). A geographically variable water quality index used in Oregon. Water Pollution
Control Federation, 57(1), 53-61.
Evans, S., Campbell, C., & Naidenko, O. V. (2020). Analysis of Cumulative Cancer Risk Associated with
Disinfection Byproducts in United States Drinking Water. Int J Environ Res Public Health, 17(6).
doi: 10.3 3 90/ij erph 17062149
Fann, N., Coffman, E., Timin, B., & Kelly, J. (2018). The Estimated Change in the Level and Distribution of
PM2.5-Attributable Health Impacts in the United States: 2005 to 2014. Environmental Research, 167.
doi: 10.1016/j.envres.2018.08.018
Fann, N., Nolte, C. G., Dolwick, P., Spero, T. L., Brown, A. C., Phillips, S., & Anenberg, S. (2015). The
geographic distribution and economic value of climate change-related ozone health impacts in the
United States in 2030. J Air Waste Manag Assoc, 65(5), 570-580.
doi: 10.1080/10962247.2014.996270
Fraas, A., Lutter, R., Dudley, S., Gayer, T., Graham, J., Shogren, J. F., & Viscusi, W. K. (2016). Social cost
of carbon: Domestic duty. Science, 351(6213), 569-569.
Freeman III, A. M. (2003). The measurement of environmental and resource values: theory and methods.
Washington, DC: Resources for the Future.
Froese, R., & Pauly., D. (2019). Fishbase (version 04/2019). In (2009, July ed.).
Gayer, T., & Viscusi, W. K. (2016). Determining the proper scope of climate change policy benefits in US
regulatory analyses: Domestic versus global approaches. Review of Environmental Economics and
Policy, 10(2), 245-263.
Gibbs, J. P., Halstead, J. M., Boyle, K. J., & Huang, J.-C. (2002). An hedonic analysis of the effects of lake
water clarity on New Hampshire lakefront properties. Agricultural and Resource Economics Review,
31(\), 39-46.
Good, K. D., & VanBriesen, J. M. (2017). Power plant bromide discharges and downstream drinking water
systems in Pennsylvania. Environmental Science & Technology, 57(20), 11829-11838.
Good, K. D., & VanBriesen, J. M. (2019). Coal-Fired Power Plant Wet Flue Gas Desulfurization Bromide
Discharges to U.S. Watersheds and Their Contributions to Drinking Water Sources. Environmental
Science & Technology, 53(1), 213-223. doi:10.1021/acs.est.8b03036
Graf, W. L., Wohl, E., Sinha, T., & Sabo, J. L. (2010). Sedimentation and sustainability of western American
reservoirs. Water Resources Research, 46(12).
Grandjean, P., Weihe, P., Debes, F., Choi, A. L., & Budtz-Jorgensen. E. (2014). Neurotoxicity from prenatal
and postnatal exposure to methylmercury. Neurotoxicology and teratology, 43, 39-44.
doi:10.1016/j.ntt.2014.03.004.
Greco, S. L., Belova, A., Haskell, J., & Backer, L. (2019). Estimated burden of disease from arsenic in
drinking water supplied by domestic wells in the United States. Journal of water and health, 17(5),
801-812.
15-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Hargrove, W. L., Johnson, D., Snethen, D., & Middendorf, J. (2010). From dust bowl to mud bowl:
sedimentation, conservation measures, and the future of reservoirs. Journal of Soil and Water
Conservation, 65(1), 14A-17A. doi: 10.2489/jswc.65.1.14A
Hellerstein, D., Vilorio, D., & Ribaudo, M. e. (2019). Agricultural Resources and Environmental Indicators.
(EIB-208).
Hoos, A. B., & Roland Ii, V. L. (2019). Spatially referenced models of streamflow and nitrogen, phosphorus,
and suspended-sediment loads in the southeastern United States (2019-5135). Retrieved from Reston,
VA: http://pubs.er.usgs.gov/publication/sir20195135
Hope, C. (2013). Critical issues forthe calculation of the social cost of CO 2: why the estimates from
PAGE09 are higherthan those from PAGE2002. Climatic Change, 117(3), 531-543.
Horowitz, J. K., & McConnell, K. E. (2002). A Review ofWTA/WTP Studies. Journal of Environmental
Economics and Management, 44(3), 426-447. doi:https://doi.org/10.1006/ieem.2001.1215
ICF (2020, August 19, 2020). [2020 Meta-regression Update Results].
Institute of Medicine. (2013). Environmental Decisions in the Face of Uncertainty. In Washington DC: The
National Academies Press (Ed.). doi:https://doi.org/10.17226/12568
Interagency Working Group on Social Cost of Carbon United States Government. (2015). Technical Support
Document: Technical Update of the Social Cost of Carbon for Regulatory Impact ANalysis Under
Executive Order 12866.
Jakus, P. M., Downing, M., Bevelhimer, M. S., & Fly, J. M. (1997). Do sportfish consumption advisories
affect reservoir anglers' site choice? Agricultural and Resource Economics Review, 26(2), 196-204.
Jakus, P. M., McGuinness, M., & Krupnick, A. J. (2002). The benefits and costs of fish consumption
advisories for mercury. Retrieved from
Jelks, H. L., Walsh, S. J., Burkhead, N. M., Contreras-Balderas, S., Diaz-Pardo, E., Hendrickson, D. A., . . .
Nelson, J. S. (2008). Conservation status of imperiled North American freshwater and diadromous
fishes. Fisheries, 33(8), 372-407.
Jerrett, M., Burnett, R. T., Pope, C. A., Ito, K., Thurston, G., Krewski, D., . . . Thun, M. (2009). Long-Term
Ozone Exposure and Mortality. New England Journal of Medicine, 360(11), 1085-1095.
doi: 10.1056/NEJMoa0803894
Jhun, I., Fann, N., Zanobetti, A., & Hubbell, B. (2014). Effect modification of ozone-related mortality risks
by temperature in 97 US cities. Environment international, 73, 128-134.
doi:https://doi.org/10.1016/i.envint.2014.07.009
Johnston, R. J., Besedin, E. Y., & Holland, B. (2019). Modeling Distance Decay within Valuation Meta-
Analysis. Environmental and resource economics, 72(3), 657-690.
doi:httPs://doi.org/10.1007/s 10640-018-0218-z
Johnston, R. J., Besedin, E. Y., Iovanna, R., Miller, C. J., Wardwell, R. F., & Ranson, M. H. (2005).
Systematic variation in willingness to pay for aquatic resource improvements and implications for
benefit transfer: A meta-analysis. Canadian Journal of Agricultural Economics/Revue canadienne
d'agroeconomie, 53(2-3), 221-248.
Koo, B., Dunker, A. M., & Yarwood, G. (2007). Implementing the Decoupled Direct Method for Sensitivity
Analysis in a Particulate Matter Air Quality Model. Environmental Science & Technology, 41(%),
2847-2854. doi: 10.1021/es0619962
Kopp, R. J., Krupnick, A. J., & Toman, M. (1997). Cost-benefit analysis and regulatory reform: An
assessment of the science and the art. Retrieved from
15-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Krewski, D., Jerrett, M., Burnett, R. T., Ma, R., Hughes, E., Shi, Y., . . . Calle, E. E. (2009). Extended follow-
up and spatial analysis of the American Cancer Society study linking particulate air pollution and
mortality. Health Effects Institute Boston, MA.
Lanphear, B. P., Rauch, S., Auinger, P., Allen, R. W., & Hornung, R. W. (2018). Low-level lead exposure
and mortality in US adults: a population-based cohort study. The Lancet Public Health, 3(A), q\11-
el84. doi:https://doi.org/10.1016/S2468-2667(18)30025-2
Leggett, C. G., & Bockstael, N. E. (2000). Evidence of the effects of water quality on residential land prices.
Journal of Environmental Economics and Management, 39(2), 121-144.
Lepeule, J., Laden, F., Dockery, D., & Schwartz, J. (2012). Chronic exposure to fine particles and mortality:
an extended follow-up of the Harvard Six Cities study from 1974 to 2009. Environmental health
perspectives, 120(7), 965-970.
Lin, D., Lutter, R., & Ruhm, C. J. (2018). Cognitive performance and labour market outcomes. Labour
Economics, 51, 121-135.
Lincoln Public Works and Utilities. (2018). 2018/19 Water and Wastewater Rates. In L. W. a. W. System
(Ed.). 555 South 10th Street, Suite 203, Lincoln, NE 68508.
Liu, T., Opaluch, J. J., & Uchida, E. (2017). The impact of water quality in Narragansett Bay on housing
prices. Water Resources Research, 53(8), 6454-6471.
Martin, J. A., Hamilton, B. E., Osterman, M. J. K., & Driscoll, A. K. (2019). Births: Final Data for 2018.
National Vital Statistics Reports, Volume <5S(Number 13).
Mergler, D., Anderson, H. A., Chan, L. H. M., Mahaffey, K. R., Murray, M., Sakamoto, M., & Stern, A. H.
(2007). Methylmercury exposure and health effects in humans: a worldwide concern. AMBIO: A
Journal of the Human Environment, 36(1), 3-11.
Miranda, L. E. (2017). Section 3: Sedimentation. In Reservoir fish habitat management (pp. 35-60).
Moeltner, K., Boyle, K. J., & Paterson, R. W. (2007). Meta-analysis and benefit transfer for resource
valuation-addressing classical challenges with Bayesian modeling. Journal of Environmental
Economics and Management, 53(2), 250-269.
Napelenok, S. L., Cohan, D. S., Hu, Y., & Russell, A. G. (2006). Decoupled direct 3D sensitivity analysis for
particulate matter (DDM-3D/PM). Atmospheric Environment, 40(32), 6112-6121.
doi:https://doi.org/10.1016/i.atmosenv.2006.05.039
National Academies of Sciences, E., & Medicine. (2017). Valuing Climate Damages: Updating Estimation of
the Social Cost of Carbon Dioxide. Washington, DC: The National Academies Press.
National Research Council. (1993). Soil and Water Quality: An Agenda for Agriculture. Washington, DC:
The National Academies Press.
National Toxicology Program. (2012). NTP Monograph on Health Effects of Low-Level Lead.
NatureServe. (2020). "Welcome to the New NatureServe Explorer.". Retrieved from
https://explorer.natureserve.org/
Nelson, J. P. K., P.E. (2009). The Use (and Abuse) of Meta-analysis in Environmental and Natural Resource
Economics: An Assessment. Environmental and resource economics, 42(3), 345-377.
Netusil, N. R., Kincaid, M., & Chang, H. (2014). Valuing water quality in urban watersheds: A comparative
analysis of Johnson Creek, Oregon, and Burnt Bridge Creek, Washington. Water Resources Research,
50(5), 4254-4268.
Nordhaus, W. D. (2014). Estimates of the social cost of carbon: concepts and results from the DICE-2013R
model and alternative approaches. Journal of the Association of Environmental and Resource
Economists, 7(1/2), 273-312.
15-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Nordhaus, W. D. (2017). Revisiting the social cost of carbon. Proceedings of the National Academy of
Sciences, 114(1), 1518-1523.
Office of Management and Budget. (2003). Circular A-4: Regulatory Analysis. Retrieved from
https: //www. whitehouse. gov/sites/whitehouse. gov/file s/omb/circulars/A4/a-4 .pdf
Omernik, J. M., & Griffith, G. E. (2014). Ecoregions of the Conterminous United States: Evolution of a
Hierarchical Spatial Framework. Environmental Management, 54(6), 1249-1266.
doi: 10.1007/s00267-014-0364-l
Oulhote, Y., Mergler, D., Barbeau, B., Bellinger, D. C., Bouffard, T., Brodeur, M.-E., . . . Bouchard, M. F.
(2014). Neurobehavioral function in school-age children exposed to manganese in drinking water.
Environmental health perspectives, 122(12), 1343-1350. doi:http://dx.doi.org/10.1289/ehp. 1307918
Pudoudyal, N. C., Paudel, B., & Green, G. T. (2013). Estimating the impact of impaired visibility on the
demand for visits to national parks. Tourism Economics, 19(2), 433-452. doi: 10.5367/te.2013.0204
Ramboll Environ International Corporation. (2016). User's Guide: Comprehensive Air Quality Model with
Extensions, version 6.40. Retrieved from 773 San Marin Drive, Suite 2115, Novato, California,
94998:
Regli, S., Chen, J., Messner, M., Elovitz, M. S., Letkiewicz, F. J., Pegram, R. A., . . . Wright, J. M. (2015).
Estimating potential increased bladder cancer risk due to increased bromide concentrations in sources
of disinfected drinking waters. Environmental Science & Technology, 49(22), 13094-13102.
Ren, C., Williams, G. M., Mengersen, K., Morawska, L., & Tong, S. (2008). Does temperature modify short-
term effects of ozone on total mortality in 60 large eastern US communities? — An assessment using
the NMMAPS data. Environment international, 34(4), 451-458.
doi:https://doi.org/10.1016/i.envint.2007.10.001
Ren, C., Williams, G. M., Morawska, L., Mengersen, K., & Tong, S. (2008). Ozone modifies associations
between temperature and cardiovascular mortality: analysis of the NMMAPS data. Occup Environ
Med, 65(4), 255-260. doi:10.1136/oem.2007.033878
Revesz, R. L., Schwartz, J. A., Howard, P. H., Arrow, K., Livermore, M. A., Oppenheimer, M., & Sterner, T.
(2017). The social cost of carbon: A global imperative. Review of Environmental Economics and
Policy, 77(1), 172-173.
Ribaudo, M. (2011). Chapter 2.3 Water quality: Impacts of agriculture. In Agricultural Resources and
Environmental Indicators (pp. 201-209).
Ribaudo, M., & Johansson, R. (2006). Water Quality: Impacts on Agriculture. In Keith Wiebe & Noel
Gollehon (Eds.), Agricultural Resources and Environmental Indicators, 2006 Edition (EIB-16 ed.,
pp. 33-41).
Richardson, L., & Loomis, J. (2009). The total economic value of threatened, endangered and rare species: an
updated meta-analysis. Ecological economics, 68(5), 1535-1548.
Rivin, G. (2015). Duke Energy Required to Pay Towns, Cities for Degraded Water. North Carolina Health
News, 2019.
Robertson, D. M., & Saad, D. A. (2019). Spatially referenced models of streamflow and nitrogen,
phosphorus, and suspended-sediment loads in streams of the midwestern United States (2019-5114).
Retrieved from Reston, VA: http://pubs.er.usgs.gov/publication/sir20195114
Roe, G. H., & Baker, M. B. (2007). Why is climate sensitivity so unpredictable? Science, 375(5850), 629-
632.
Roels, H. A., Bowler, R. M., Kim, Y., Henn, B. C., Mergler, D., Hoet, P., . . . Harris, M. G. (2012).
Manganese exposure and cognitive deficits: a growing concern for manganese neurotoxicity.
Neurotoxicology, 33(4), 872-880. doi:doi: 10.1016/j.neuro.2012.03.009
15-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Rosenberger, R. S., & Johnston, R. J. (2007). Selection Effects inMeta-Valuation Function Transfers. Paper
presented at the Meta Analysis in Economics, Conference., Sandbjerg Manor, Sonderborg. Denmark.
Rosenberger, R. S., & Phipps, T. (2007). Correspondence and convergence in benefit transfer accuracy: meta-
analytic review of the literature. In Environmental value transfer: Issues and methods (pp. 23-43):
Springer.
Rosenberger, R. S., & Stanley, T. D. (2006). Measurement, generalization, and publication: Sources of error
in benefit transfers and their management. Ecological economics, 60(2), 372-378.
Ruhl, L., Vengosh, A., Dwyer, G. S., Hsu-Kim, H., Schwartz, G., Romanski, A., & Smith, S. D. (2012). The
impact of coal combustion residue effluent on water resources: a North Carolina example.
Environmental Science & Technology, 46(21), 12226-12233.
Sacks, J. D., Lloyd, J. M., Zhu, Y., Anderton, J., Jang, C. J., Hubbell, B., & Fann, N. (2018). The
Environmental Benefits Mapping and Analysis Program - Community Edition (BenMAP-CE): A tool
to estimate the health and economic benefits of reducing air pollution. Environ Model Softw, 104,
118-129.
Salkever, D. S. (1995). Updated estimates of earnings benefits from reduced exposure of children to
environmental lead. Environmental Research, 70(1), 1-6.
Sea Grant - Illinois-Indiana. (2018). Lake Michigan anglers boost local Illinois and Indiana economies.
Shrestha, R., Rosenberger, R., & Loomis, J. (2007). Benefit transfer using meta-analysis in recreation
economic valuation. In Environmental value transfer: Issues and methods (pp. 161-177): Springer.
Smith, R. L., Xu, B., & Switzer, P. (2009). Reassessing the relationship between ozone and short-term
mortality in U.S. urban communities. Inhalation Toxicology, 27(sup2), 37-61.
doi: 10.1080/08958370903161612
Smith, V. K., Van Houtven, G., & Pattanayak, S. K. (2002). Benefit transfer via preference
calibration:"Prudential algebra" for policy. Land Economics, 75(1), 132-152.
Sohngen, B., Zhang, W., Bruskotter, J., & Sheldon, B. (2015). Results from a 2014 survey of Lake Erie
anglers. Columbus, OH: The Ohio State University, Department of Agricultural, Environmental and
Development Economics and School of Environment & Natural Resources.
Solomon, B. D., Corey-Luse, C. M., & Halvorsen, K. E. (2004). The Florida manatee and eco-tourism:
toward a safe minimum standard. Ecological economics, 50(1), 101-115.
doi:https://doi.org/10.1016/i.ecolecon.2004.03.025
Southwest Fisheries Science Center (SWFSC) of NOAA National Marine Fisheries Service. (2019).
"Southwest Fisheries Science Center Publication Database. ".
Stanley, T. D. (2005). Beyond publication Bias. Journal of Economic Surveys, 19(3), 309-345.
doi: 10.1111/j .0950-0804.2005.00250.x
Swartout, J., & Rice, G. (2000). Uncertainty analysis of the estimated ingestion rates used to derive the
methylmercury reference dose. Drug and chemical toxicology, 23( 1), 293-306.
Taylor, C. A., Schuster, G. A., Cooper, J. E., DiStefano, R. J., Eversole, A. G., Hamr, P., . . . Thoma, R. F.
(2007). A reassessment of the conservation status of crayfishes of the United States and Canada after
10+ years of increased awareness. Fisheries, 32(8), 372-389.
Taylor, C. A., Warren Jr, M. L., Fitzpatrick Jr, J. F., Hobbs Iii, H. H., Jezerinac, R. F., Pflieger, W. L., &
Robison, H. W. (1996). Conservation status of crayfishes of the United States and Canada. Fisheries,
27(4), 25-38.
Tetra Tech. (2010). Evaluating Sustainability of Projected Water Demands Under Future Climate Change
Scenarios. Prepared for Natural Resources Defense Council. Prepared by Tetra Tech.
15-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Tinuum Group LLC. (2020). Comment submitted by Ron Eller, CEO and Jim Zerefos, General Counsel,
Tinuum Group, LLC. In Docket ID No. EPA-HQ-OW-2009-0819; EPA's Proposed Rule titled
"Effluent Limitation Guidelines and Standards for the Steam Electric Power Generating Point Source
Category" 84 FR 64620 (November 22, 2019) (pp. 4).
Tuttle, C., & Heintzelman, M. D. (2014). A Loon on Every Lake: A Hedonic Analysis of Lake Quality in the
Adirondacks. Available at SSRN 2467745.
Tyllianakis, E., & Skuras, D. (2016). The income elasticity of Willingness-To-Pay (WTP) revisited: A meta-
analysis of studies for restoring Good Ecological Status (GES) of water bodies under the Water
Framework Directive (WFD). Journal of environmental management, 182, 531-541.
doi: 10.1016/j.jenvman.2016.08.012
U.S. Army Corps of Engineers. (2013). Dredging Information System.
U.S. Census Bureau. (2017). National Population Projections Tables: 2017 to 2060. Retrieved from
httos://www. census. gov/data/tables/2017/demo/popproi/2017-summarv-tables.html
U.S. Department of Health and Human Services. (2012). Toxicological Profile for Cadmium.
U.S. Department of the Interior. (2019). Endangered and Threatened Wildlife and Plants; Endangered
Species Status for Barrens Topminnow. (FWS-R4-ES-2017-0094). Retrieved from
https://www.govinfo.gov/content/pkg/FR-2019-10-21/pdf/2Q19-22857.pdf
U.S. Department of the Interior, U.S. Fish and Wildlife Service, U.S. Department of Commerce, & U.S.
Census Bureau. (2018). 2016 National Survey of Fishing, Hunting, and Wildlife-Associated
Recreation. (FHW/16-NAT(RV)).
U.S. Environmental Protection Agency. (1985). Guidelines for Deriving Numerical National Water Quality
Criteria for the Protection Of Aquatic Organisms and Their Uses. (PB85-227049). Retrieved from
https://www.epa.gov/sites/production/files/2016-02/documents/guidelines-water-qualitv-criteria.pdf
U.S. Environmental Protection Agency. (2004). Advisory on Plans for Health Effects Analysis in the
Analytical Plan for EPA's Second Prospective Analysis - Benefits and Costs of the Clean Air Act,
1990-2020. (EPA-SAB-COUNCIL-ADV-04-002).
U.S. Environmental Protection Agency. (2005a). Economic Analysis for the Final Stage 2 Disinfectants and
Disinfection Byproducts Rule. (EPA 815-R-05-010).
U.S. Environmental Protection Agency. (2005b). Regulatory Impact Analysis of the Final Clean Air Mercury
Rule. (EPA-452/R-05-003). Research Triangle Park,N.C. 27711
U.S. Environmental Protection Agency. (2009a). Integrated Science Assessment for Particular Matter.
(EPA/600/R-08/139F). Research Triangle Park, NC
U.S. Environmental Protection Agency. (2009b). Software and User's Manuals. Retrieved from
http://www.epa. gov/superfund/lead/products .htm
U.S. Environmental Protection Agency. (2010a). Guidelines for Preparing Economic Analyses. (EPA 240-R-
10-001).
U.S. Environmental Protection Agency. (2010b). Integrated Risk Information System (IRIS) Toxicological
Review of Inorganic Arsenic (Cancer) (External Draft Review). Retrieved from
http://cfpub.epa.gov/ncea/iris drafts/recordisplav.cfm?deid=219111
U.S. Environmental Protection Agency. (2011). Exposure Factors Handbook, 2011 Edition (Final). (EPA-
600-R-09-025F). U.S. Environmental Protection Agency, Washington, DC
U.S. Environmental Protection Agency. (2012). Regulatory Impact Assessment for the Particulate Matter
National Ambient Air Quality Standards. (EPA-452/R-12-005). Retrieved from
https://www3 .epa.gov/ttnecas 1/regdata/RIAs/finalria.pdf
15-8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
U.S. Environmental Protection Agency. (2013a). Benefit and Cost Analysis for the Proposed Effluent
Limitation Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. (EPA-821-R-13-004). U.S. Environmental Protection Agency, Washington, DC 204640
U.S. Environmental Protection Agency. (2013b). Fish Consumption in Connecticut, Florida, Minnesota, and
North Dakota. (EPA/600/R-13/098F). U.S. Environmental Protection Agency, Washington, DC
20460
U.S. Environmental Protection Agency. (2013c). Integrated Science Assessment for Lead. (EPA/600/R-
10/075F). Research Triangle Park, NC
U.S. Environmental Protection Agency. (2013d). Integrated Science Assessment for Ozone and Related
Photochemical Oxidants. (EPA/600/R-10/076F). Research Triangle Park, NC
U.S. Environmental Protection Agency. (2014). Regulatory Impact Analysis (RIA) for EPA's 2015 Coal
Combustion Residuals (CCR) Final Rule Retrieved from https://downloads.regulations.gov/EPA-HQ-
RCRA-2009-0640-12034/content.pdf
U.S. Environmental Protection Agency. (2015a). Benefit and Cost Analysis for the Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source Category. (EPA-
821-R-15-005).
U.S. Environmental Protection Agency. (2015b). Environmental Assessment for the Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source Category. (EPA
821-R-15-006).
U.S. Environmental Protection Agency. (2015c). Regulatory Impact Analysis for the Effluent Limitations
Guidelines and Standards for the Steam Electric Power Generating Point Source Category. (EPA-
821-R-15-004).
U.S. Environmental Protection Agency. (2016a). Integrated Science Assessment for Oxides of Nitrogen: Final
Report. (EPA/600/R-15/068). Retrieved from
http://ofmpub.epa.gov/eims/eimscomm.getfile7p download id=526855
U.S. Environmental Protection Agency. (2016b). Six-Year Review 3 Technical Support Document for
Disinfectants/Disinfection Byproducts Rules. (EPA-810-R-16-012).
U.S. Environmental Protection Agency. (2017a). Chapter 3: Water Quality Criteria. Water Quality Standards
Handbook. (EPA 823-B-17-001). Retrieved from https://www.epa.gov/sites/production/files/2Q14-
10/documents/handbook-chapter3 .pdf
U.S. Environmental Protection Agency. (2017b). Documentation for the EPA's Preliminary 2028 Regional
Haze Modeling. Retrieved from
https://www3.epa.gov/ttn/scram/reports/2028 Regional Haze Modeling-TSD.pdf
U.S. Environmental Protection Agency. (2017c). Integrated Science Assessment for Sulfur Oxides: Health
Criteria. (EPA/600/R-17/451). Retrieved from
http://ofmpub.epa.gov/eims/eimscomm.getfile7p download id=533653
U.S. Environmental Protection Agency. (2018a). Environmental Benefits Mapping and Analysis Program
(BenMAP) - Community Edition, User's Manual.
U.S. Environmental Protection Agency. (2018b). National Primary Drinking Water Regulations. Retrieved
from https://www.epa.gov/ground-water-and-drinking-water/national-primarv-drinking-water-
regulations
U.S. Environmental Protection Agency. (2018c). National Recommended Water Quality Criteria - Human
Health Criteria Table. Retrieved from https://www.epa.gov/wac/national-recommended-water-
qualitv-criteria-human-health-criteria-table
15-9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
U.S. Environmental Protection Agency. (2019a). Benefit and Cost Analysis for Proposed Revisions to the
Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category.
U.S. Environmental Protection Agency. (2019b). Economic Analysis of the Final Rule to Revise the TSCA
Lead-Dust Hazard Standards.
U.S. Environmental Protection Agency. (2019c). Economic Analysis of the Proposed Lead and Copper Rule
Revisions.
U.S. Environmental Protection Agency. (2019d). Health Risk Reduction and Cost Analysis of the Proposed
Per chlorate National Primary Drinking Water Regulation. (EPA 816-R-19-004).
U.S. Environmental Protection Agency. (2019e). Integrated Science Assessment for Particulate Matter.
(EPA/600/R-19/188). Retrieved from https://cfpub.epa.gov/ncea/isa/recordisplav.cfm?deid=347534
U.S. Environmental Protection Agency. (2019f). NHDPlus Version 2: User Guide (DataModel Version 2.1).
Retrieved from https://s3 .amazonaws.com/edap-
nhdplus/NHDPlusV21/Documentation/NHDPlusV2 User Guide.pdf
U.S. Environmental Protection Agency. (2019g). Regulatory Impact Analysis for Proposed Revisions to the
Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. (EPA-821-R-19-012).
U.S. Environmental Protection Agency. (2019h). Regulatory Impact Analysis for Repeal of the Clean Power
Plan; Emission Guidelines for Greenhouse Gas Emissions from Existing Electric Utility Generating
Units; Revisions to Emission Guidelines Implementing Regulations. (EPA-452/R-19-003). Research
Triangle Park, NC
U.S. Environmental Protection Agency. (2019i). Regulatory Impact Analysis for the Repeal of the Clean
Power Plan, and the Emission Guidelines for Greenhouse Gas Emissions from Existing Electric
Utility Generating Units. (EPA-452/R-19-003).
U.S. Environmental Protection Agency. (2019j). Supplemental Environmental Assessment for Proposed
Revisions to Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating
Point Source Category (EA).
U.S. Environmental Protection Agency. (2019k). Supplemental Technical Development Document for
Proposed Revisions to Effluent Limitations Guidelines and Standards for the Steam Electric Power
Generating Point Source Category.
U.S. Environmental Protection Agency. (2020a). Flat file generation methodology (Version: January 2020
Reference Case using EPA Platform v6). Retrieved from
https://www.epa.gov/sites/production/files/2020-
02/documents/flat file methodology ianuarv 2020.pdf
U.S. Environmental Protection Agency. (2020b). Integrated Science Assessment for Ozone and Related
Photochemical Oxidants. (EPA/600/R-20/012).
U.S. Environmental Protection Agency. (2020c). National Emission Standards for Hazardous Air Pollutants:
Coal- and Oil-FIred Electric Utility Steam Generating Units — Subcategory of Certain Existing
Electric Utility Steam Generating Units Firing Eastern Bituminous Coal Refuse for Emissions of Acid
Gas Hazardous Air Pollutants. (EPA-HQ-OAR-2018-0749; FRL-10007-26-OAR).
U.S. Environmental Protection Agency. (2020d). Regulatory Impact Analysis for Revisions to the Effluent
Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. (EPA-821-R-20-004).
U.S. Environmental Protection Agency. (2020e). Regulatory Impact Analysis, Hazardous and Solid Waste
Management System: Disposal of Coal Combustion Residuals from Electric Utilities; A Holistic
Approach to Closure Part A: Deadline to Initiate Closure.
15-10
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
U.S. Environmental Protection Agency. (2020f). Supplemental Environmental Assessment for Revisions to the
Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category.
U.S. Environmental Protection Agency. (2020g). Supplemental Technical Development Document for
Revisions to the Effluent Limitations Guidelines and Standards for the Steam Electric Power
Generating Point Source Category. (EPA-821-R-20-001).
U.S. Fish and Wildlife Service. (2006). National Survey of Fishing, Hunting, and Wildlife-Associated
Recreation. (FHW/06-NAT).
U.S. Fish and Wildlife Service. (2019a). Grotto Sculpin (Cottus specus). Retrieved from
https://www.fws.gov/midwest/endangered/fishes/grottosculpin/grottosculpinfactsheet.html
U.S. Fish and Wildlife Service. (2019b). Ozark Hellbender (Cryptobranchus alleganiensis bishopi) Fact
Sheet. Retrieved from
https://www.fws.gov/midwest/endangered/amphibians/ozhe/ozheFactSheet.html
U.S. Fish and Wildlife Service. (2019c). Questions and Answers about the Topeka Shiner in Minnesota.
Retrieved from https://www.fws.gov/midwest/endangered/fishes/TopekaShiner/tosh mn.html
U.S. Fish and Wildlife Service. (2019d). Rayed Bean (Villosa fabalis) Fact Sheet. Retrieved from
https://www.fws.gov/midwest/Endangered/clams/ravedbean/RavedBeanFactSheet.html
U.S. Fish and Wildlife Service. (2019e). Sheepnose (a freshwater mussel) Plethobasus cyphys - Fact Sheet.
Retrieved from
https://www.fws.gov/midwest/Endangered/clams/sheepnose/SheepnoseFactSheetMarch2012.html
U.S. Fish and Wildlife Service. (2019f). Snuffbox (freshwater mussel) Epioblasma triquetra Fact Sheet.
Retrieved from https://www.fws.gov/midwest/Endangered/clams/snuffbox/SnuffboxFactSheet.html
U.S. Fish and Wildlife Service. (2019g). Spectaclecase (a freshwater mussel) Cumberlandia mondota.
Retrieved from
https://www.fws.gov/midwest/endangered/clams/spectaclecase/SpectaclecaseFactSheet.html
U.S. Fish and Wildlife Service. (2020a). Birdwing Pearlymussel, Lemiox rimosus (Rafmesque, 1831), 5-Year
Review: Summary and Evaluation. Retrieved from
https://ecos.fws.gov/docs/five year review/doc6386.pdf
U.S. Fish and Wildlife Service. (2020b). ECOS Complete Current Species Range.
U.S. Fish and Wildlife Service. (2020c). ECOS Habitat Conservation Plans - All Regions. Retrieved from
https://ecos.fws. gov/ecpO/conservationPlan/region/summarv?region=9&tvpe=HCP
U.S. Fish and Wildlife Service. (2020d). Environmental Conservation Online System (ECOS) Species
Reports. Retrieved from https://ecos.fws.gov/ecp/
U.S. Fish and Wildlife Service. (2020e). Freshwater Fish of America: Atlantic Sturgeon. Retrieved from
https://www.fws.gov/fisheries/freshwater-fish-of-america/atlantic sturgeon.html
U.S. Fish and Wildlife Service. (2020f). Freshwater Fish of America: Humpback Chup. Retrieved from
https://www.fws.gov/fisheries/freshwater-fish-of-america/humpback chub.html
U.S. Fish and Wildlife Service. (2020g). Mountain Prairie Region - Endangered Species. Retrieved from
https: //www.fws. gov/mountain-prairie/
U.S. Fish and Wildlife Service. (2020h). Neosho Mucket (Lampsilis rafmesqueana) 5-Year Review: Summary
and Evaluation. Retrieved from https://ecos.fws.gov/docs/five year review/doc6400.pdf
U.S. Fish and Wildlife Service. (2020i). "North Florida Ecological Services Office: Loggerhead Sea Turtle
(Caretta caretta)". Retrieved from
https://www.fws.gov/northflorida/SeaTurtles/Turtle%20Factsheets/loggerhead-sea-turtle.htm
15-11
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
U.S. Fish and Wildlife Service. (2020j). Northeast Region Endangered Species Program - Ecological
Services. Retrieved from https://www.fws.gov/northeast/ecologicalservices/endangeredspecies.html
U.S. Fish and Wildlife Service. (2020k). Topeka Shiner (Notropis topeka). Retrieved from
https://www.fws.gov/midwest/endangered/fishes/TopekaShiner/index.html
U.S. Geological Survey. (2007). National Hydrography Dataset (NHD). Retrieved from
http://nhd.usgs. gov/data.html
U.S. Geological Survey. (2009). RESISII: An Updated Version of the Original Reservoir Sedimentation
Survey Information System (RESIS). U.S. Geological Survey, Reston, Virginia
U.S. Geological Survey. (2018). National Hydrography Dataset (NHD). Retrieved from
https://www.usgs.gov/core-science-svstems/ngp/national-hvdrographv/national-hvdrographv-
dataset?qt-science support page related con=Q#qt-science support page related con
U.S. Global Change Research Program. (2016). The Impacts of Climate Change on Human Health in the
United States: A Scientific Assessment. doi:10.7930/J0R49NQX
Upper Colorado River Endangered Fish Recovery Program. (2020). Bonytail (Gila elegans). Retrieved from
https://www.coloradoriverrecoverv.org/general-information/the-fish/bonvtail.html
Viscusi, W. K., Huber, J., & Bell, J. (2008). The economic value of water quality. Environmental and
resource economics, 41(2), 169-187.
Walsh, P. J., Milon, J. W., & Scrogin, D. O. (2011). The spatial extent of water quality benefits in urban
housing markets. Land Economics, 87(4), 628-644.
Wathen, J. B., Lazorchak, J. M., Olsen, A. R., & Batt, A. (2015). A national statistical survey assessment of
mercury concentrations in fillets of fish collected in the US EPA national rivers and streams
assessment of the continental USA. Chemosphere, 122, 52-61.
Watson, K., Farre, M. J., & Knight, N. (2012). Strategies forthe removal of halides from drinking water
sources, and their applicability in disinfection by-product minimisation: a critical review. Journal of
environmental management, 110, 276-298.
Whittington, D., & MacRae Jr, D. (1986). The issue of standing in cost-benefit analysis. Journal of Policy
Analysis and Management, 5(4), 665-682.
Wilcove, D. S., & Master, L. L. (2005). How many endangered species are there in the United States?
Frontiers in Ecology and the Environment, 3(8), 414-420.
Williams, J. D., Warren Jr, M. L., Cummings, K. S., Harris, J. L., & Neves, R. J. (1993). Conservation status
of freshwater mussels of the United States and Canada. Fisheries, 18(9), 6-22.
Williams, J. E., Johnson, J. E., Hendrickson, D. A., Contreras-Balderas, S., Williams, J. D., Navarro-
Mendoza, M., . . . Deacon, J. E. (1989). Fishes of North America endangered, threatened, or of
special concern: 1989. Fisheries, 14(6), 2-20.
Williams, R. L., O'Leary, J. T., Sheaffer, A. L., & Mason, D. (2000). An examination of fish consumption by
Indiana recreational anglers: an on-site survey. West Lafayette, IN: Purdue University.
Winemiller, K. O. (2018). Trends in Biodiversity: Freshwater. The Encyclopedia of the Anthropocene, 3, 151-
161.
Wise, D. R. (2019). Spatially referenced models of streamflow and nitrogen, phosphorus, and suspended-
sediment loads in streams of the Pacific region of the United States (2019-5112). Retrieved from
Reston, VA: http://pubs.er.usgs.gov/publication/sir20195112
Wise, D. R., Anning, D. W., & Miller, O. L. (2019). Spatially referenced models of streamflow and nitrogen,
phosphorus, and suspended-sediment transport in streams of the southwestern United States (2019-
5106). Retrieved from Reston, VA: http://pubs.er.usgs.gov/publication/sir20195106
15-12
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
15: Cited References
Younjun, K., Kling, C. L., & Zhao, J. (2015). Understanding Behavioral Explanations of the WTP-WTA
Divergence Through a Neoclassical Lens: Implications for Environmental Policy. Annual Review of
Resource Economics, 7(1), 169-187.
Zavala, M., Lei, W., Molina, M. J., & Molina, L. T. (2009). Modeled and observed ozone sensitivity to
mobile-source emissions in Mexico City. Atmos. Chem. Phys., 9(1), 39-55. doi:10.5194/acp-9-39-
2009
Zhu, M., Fitzgerald, E. F., Gelberg, K. H., Lin, S., & Druschel, C. M. (2010). Maternal low-level lead
exposure and fetal growth. Environmental health perspectives, 775(10), 1471-1475.
15-13
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix A: Changes to Benefits Analysis
A Changes to Benefits Methodology since 2019 Proposed Rule Analysis
The table below summarizes the principal methodological changes EPA made to analyses of the benefits of
the final rule regulatory options, as compared to the analyses of the 2019 proposed rule (U.S. EPA, 2019a).
Table A-1: Changes to Benefits Analysis Since 2019 Proposed Rule
Benefits Category
Analysis Component
[2019 proposed rule analysis value]
Changes to Analysis for regulatory options
[2020 rule analysis value]
General inputs and pollutant loads
Regulatory options
analyzed
EPA analyzed the four proposed options
(Options, 1, 2, 3, and 4)
EPA conducted new analyses only for Options
A, B, and C, and did not re-analyze Option D,
which corresponds to proposed Option 1.
Universe of plants,
EGUs, and receiving
reaches
Analysis includes loadings for only coal-
fired units operating as of December 31,
2028.
Analysis includes loadings for all coal-fired units
operating as of 2020. The analysis also reflects
other updates to the steam electric industry
profile through the end of 2019, including the
timing of projected retirements and refueling
projects and existing treatment technologies.
See Supplemental TDD for details (U.S. EPA,
2020g).
General pollutant
loadings and
concentrations
Affected reaches based on immediate
receiving reaches and flow paths in
medium-resolution NHD.
Updated immediate receiving reaches for
selected plants.
SPARROW modeling of nutrient and
sediment concentrations in receiving and
downstream reaches based on national
SPARROW models and Enhanced River
File 1 (E2RF1) stream network.
SPARROW modeling of nutrient and sediment
concentrations in receiving and downstream
reaches based on the most recent five regional
SPARROW models that use the medium-
resolution NHD stream network.
Uses the annual average loadings for
analysis period [2021-2047], with pre-
technology implementation loads set
equal to current loads.
Uses the annual average loadings for two
distinct periods during the analysis: 2021-2028
and 2029-2047, with pre-technology
implementation loads set equal to current loads
and post-retirement or repowering loads set to
zero.
Water quality index
Expresses overall water quality changes
using a seven-parameter index that
includes subindex curve parameters for
nutrients and sediment based on the
national SPARROW models.
Expresses overall water quality changes using a
seven-parameter index that includes subindex
curve parameters for nutrients and sediment
based on the regional SPARROW models.
Human health benefits from changes in exposure to halogenated disinfection byproducts in drinking water
Public water systems
affected by bromide
discharges
Modeled changes in bromide
concentrations in source water of public
water systems and total trihalomethane
concentrations in drinking water.
Modeled changes in bromide concentrations in
source water of public water systems.
Public water systems
affected by iodine
discharges
Not analyzed. Referred to qualitative
discussion in Supplemental EA (U.S. EPA,
2019j)
Modeled changes in iodine concentrations in
source water of public water systems.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix A: Changes to Benefits Analysis
Table A-1: Changes to Benefits Analysis Since 2019 Proposed Rule
Benefits Category
Analysis Component
[2019 proposed rule analysis value]
Changes to Analysis for regulatory options
[2020 rule analysis value]
Lifetime changes in
incidence of bladder
cancer
Applied lifetime risk model to estimate
changes in bladder cancer incidence in
population served by public water
systems.
Qualitative discussion. EPA received public
comments that further evaluation of certain
DBPs should be completed and that the analysis
at proposal should be subjected to peer review.
EPA acknowledges that further study in this
area should be conducted, including peer
review of the model used at proposal. EPA will
continue to evaluate the scientific data on the
health impacts of DBPs.
Monetization of
changes in incidence of
bladder cancer
Mortality valued using VSL (U.S. EPA,
2010a). Morbidity valued based on COI
(Greco et a!., 2019).
Because EPA did not calculate changes in
incidence of bladder cancer, the Agency was
unable to monetize this effect.
Non-market benefits from water quality improvements
WTP for water quality
improvements
Meta-regression model
EPA added 14 new studies to the 2015 meta-
data and re-estimated the meta-regression
model (see Appendix G for details). Similar to
the 2015 meta-regression, the model includes
spatial characteristics of the affected water
resources: size of the market, waterbody
characteristics (length and flow), availability of
substitute sites, and land use type in the
adjacent counties.
Variables characterizing the availability of
substitute sites, size of the market, and land-
use were revised based on changes in the
universe of receiving reaches and CBGs
included in the analysis.
Effects on T&E species
Categorical analysis based on designated
critical habitat overlap/proximity to
reaches with estimated changes in
NRWQC exceedances.
EPA updated the list of species included in the
analysis based on the 2020 ECOS online
database (U.S. FWS, 2020d). EPA also relied on
the habitat range of T&E species in determining
whether reaches downstream from steam
electric power plant outfalls intersect species
habitat (U.S. FWS, 2020b), rather than "critical
habitat" as the term is defined in the ESA. EPA
included all species categorized as having
higher vulnerability to water pollution in its
analysis (see Chapter 7 and Appendix H for
details). The only exception is species endemic
to springs and headwaters.
Air quality-related effects
Emissions changes
Emissions from changes in electricity
generation profile from 2018 and 2019
IPM runs.
Energy use-associated emissions based on
emission factors estimated using the 2018
and 2019 IPM runs.
Emissions from changes in electricity
generation profile from 2020 IPM runs.
Energy use-associated emissions were updated
to reflect emission factors estimated using the
2020 IPM runs.
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix A: Changes to Benefits Analysis
Table A-1: Changes to Benefits Analysis Since 2019 Proposed Rule
Benefits Category
Analysis Component
[2019 proposed rule analysis value]
Changes to Analysis for regulatory options
[2020 rule analysis value]
Air quality changes
Qualitative discussion.
Used the ACE modeling methodology to
estimate changes in air pollutant
concentrations.
Monetization
Qualitative discussion.
Used BenMAP-CE model to estimate associated
human health benefits.
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
WQI Calculation and Regional Subindices
B.1 WQI Calculation
The first step in the implementation of the WQI involves obtaining water quality levels for each parameter,
and for each waterbody, under both the baseline conditions and each regulatory option. Some parameter levels
are field measurements while others are modeled values.
The second step involves transforming the parameter measurements into subindex values that express water
quality conditions on a common scale of 0 to 100. EPA used the subindex transformation curves developed by
Dunnette (1979) and Cude (2001) for the Oregon WQI for BOD, DO, and FC. For suspended sediment, TN,
and TP concentrations, EPA adapted the approach developed by Cude (2001) to account for the wide range of
natural or background nutrient and sediment concentrations that result from variability in geologic and other
region-specific conditions, and to reflect the national context of the analysis. Suspended sediment, TN, and
TP subindex curves were developed for each Level III ecoregion (Omernik & Griffith, 2014) using pre-
compliance (before the implementation of the 2015 rule) SSC and TN and TP concentrations modeled in
SPARROW at the medium-resolution NHD reach level.93 For each of the 84 Level III ecoregions intersected
by the NHD reach network, EPA derived the transformation curves by assigning a score of 100 to the 25th
percentile of the reach-level SSC level in the ecoregion (i.e., using the 25th percentile as a proxy for
"reference" concentrations), and a score of 70 to the median concentration. An exponential equation was then
fitted to the two concentration points following the approach used in Cude (2001).
For this analysis, EPA also used a toxics-specific subindex curve based on the number of NRWQC
exceedances for toxics in each waterbody. National freshwater chronic NRWQC values are available for
arsenic, cadmium, chromium, copper, lead, mercury, nickel, selenium, and zinc. See the Supplemental EA for
details on the NRWQC (U.S. EPA, 2020f). To develop this subindex curve, EPA used an approach developed
by the Canadian Council of Ministers of the Environment (CCME, 2001). The CCME water quality index is
based on three attributes of water quality that relate to water quality objectives: scope (number of monitored
parameters that exceed water quality standard or toxicological benchmark); frequency (number of individual
measurements that do not meet objectives, relative to the total number of measurements for the time period of
interest) and amplitude (i.e.. amount by which measured values exceed the standards or benchmarks).
Following the CCME approach, EPA's toxics subindex considers the number of parameters with exceedances
of the relevant water quality criterion. With regards to frequency, EPA modeled long-term annual average
concentrations in ambient water, and therefore any exceedance of an NRWQC may indicate that ambient
concentrations exceed NRWQC most of the time (assumed to be 100 percent of the time). EPA did not
consider amplitude, because if the annual average concentration exceeds the chronic NRWQC then the water
is impaired for that constituent and the level of exceedance is of secondary concern. Using this approach, the
subindex curve for toxics assigns the lowest subindex score of 0 to waters where exceedances are observed
for all nine of the toxics analyzed, and a maximum score of 100 to waters where there are no exceedances.
Intermediate values are distributed evenly between 0 and 100.
93 The SPARROW model was developed by the USGS for the regional interpretation of water-quality monitoring data. The model
relates in-stream water-quality measurements to spatially referenced characteristics of watersheds, including contaminant sources
and factors influencing terrestrial and aquatic transport. SPARROW empirically estimates the origin and fate of contaminants in
river networks and quantifies uncertainties in model predictions. More information on SPARROW can be found at
http://water.usgs.gOv/nawqa/sparrow/FAQs/faq.html#l
B-1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-l presents parameter-specific functions used for transforming water quality data into water quality
subindices for freshwater waterbodies for the six pollutants with individual subindices. Table B-2 presents the
subindex values for toxics. The equation parameters for each of the 84 ecoregion-specific SSC, TN, and TP
subindex curves are provided in the next section. The curves include threshold values below or above which
the subindex score does not change in response to changes in parameter levels. For example, improving DO
levels from 10.5 mg/L to 12 mg/L or from 2 mg/L to 3.3 mg/L would result in no change in the DO subbindex
score.
Table B-1: Freshwater Water Quality Subindices
Parameter
Concentrations
Concentration
Subindex

Unit

Dissolved Oxygen (DO)
DO saturation <100%
DO
DO < 3.3
mg/L
10
DO
3.3 < DO < 10.5
mg/L
-80.29+31.88xDO-1.401xDO2
DO
DO > 10.5
mg/L
100
100% < DO saturation < 275%
DO
NA
mg/L
100 x exp((DOsat - 100) x -1.197xl0"2)
275% < DO saturation
DO
NA
mg/L
10
Fecal Coliform (FC

FC
FC > 1,600
cfu/100 mL
10
FC
50 < FC < 1,600
cfu/100 mL
98 x exp((FC - 50) x -9.9178xl0"4)
FC
FC < 50
cfu/100 mL
98
Total Nitrogen (TN
a
TN
TN >TNio
mg/L
10
TN
TNioo < TN < TNio
mg/L
a x exp(TNxb); where a and b are ecoregion-
specific values
TN
TN < TNioo
mg/L
100
Total Phosphorus (TP)b
TP
TP > TPio
mg/L
10
TP
TPioo < TP < TPio
mg/L
a x exp(TPxb); where a and b are ecoregion-
specific values
TP
TP < TPioo
mg/L
100

Suspended Solids

SSC
SSC > SSCio
mg/L
10
SSC
SSCioo ^ SSC < SSCio
mg/L
a x exp(SSCxb); where a and b are ecoregion-
specific values
SSC
SSC < SSCioo
mg/L
100
Biochemical Oxygen Demand, 5-day (BOD)
BOD
BOD >8
mg/L
10
BOD
BOD <8
mg/L
100 xexp(BODx-0.1993)
a. TN10 and TN100 are ecoregion-specific TN concentration values that correspond to subindex scores of 10 and 100,
respectively. Use of 10 and 100 for the lower and upper bounds of the WQI subindex score follow the approach in Cude (2001)
b. TP10 and TP100 are ecoregion-specific TP concentration values that correspond to subindex scores of 10 and 100, respectively.
Use of 10 and 100 for the lower and upper bounds of the WQI subindex score follow the approach in Cude (2001)
c. SSC10 and SSC100 are ecoregion-specific SSC concentration values that correspond to subindex scores of 10 and 100,
respectively. Use of 10 and 100 for the lower and upper bounds of the WQI subindex score follow the approach in Cude (2001)
Source: EPA analysis, 2020, based on methodology in Cude (2001).
B-2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-2: Freshwater Water Quality Subindex for Toxics
Number of Toxics with NRWQC
Subindex
Exceedances

0
100.0
1
88.9
2
77.8
3
66.7
4
55.6
5
44.4
6
33.3
7
22.2
8
11.1
9
0.0
The final step in implementing the WQI involves combining the individual parameter subindices into a single
WQI value that reflects the overall water quality across the parameters. EPA calculated the overall WQI for a
given reach using a geometric mean function and assigned all WQ parameters an equal weight of 0.143 (l/7th
of the overall score). Unweighted scores for individual metrics of a WQI have previously been used in Cude
(2001), CCME, 2001, and Carruthers and Wazniak (2003).
Equation B-l presents EPA's calculation of the overall WQI score.
Equation B-1.
WQir = Uf=iQiWi
WQIr = the multiplicative water quality index (from 0 to 100) for reach r
Qi = the water quality subindex measure for parameter /'
Wi = the weight of the z-th parameter (0.143)
n = the number of parameters (z. e., seven)
B.2 Regional Subindices
The following tables provide the ecoregion-specific parameters used in estimating the suspended solids, TN,
or TP water quality subindex, as follows:
- If [WQ Parameter] < WQ Parameter 100 Subindex =100
- If WQ Parameter 100 < [WQ Parameter] < WQ Parameter 10 Subindex = a exp(b [WQ Parameter])
- If [WQ Parameter] > WQ Parameter 10 Subindex = 10
Where [WQ Parameter] is the measured concentration of either suspended solids, TN, or TP and WQ
Parameter io, WQ Parameter ioo, a, and b are specified in Table B-3 for suspended solids, Table B-4 for TN,
and Table B-5 for TP.
B-3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-3: Suspended Solids Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
SSCioo
SSCio
ECOL3_01
Coast Range
140.44
-0.0069
49.5
385.0
ECOL3_02
Strait of Georgia/Puget Lowland
131.95
-0.0044
62.5
581.9
ECOL3_03
Willamette Valley
131.91
-0.0046
59.8
556.9
ECOL3_04
Cascades
108.63
-0.0080
10.4
299.7
ECOL3_05
Sierra Nevada
109.47
-0.0108
8.3
220.7
ECOL3_06
California Coastal Sage, Chaparral, and Oak
Woodlands
117.59
-0.0042
38.6
587.6
ECOL3_07
Central California Valley
105.23
-0.0012
42.0
1,940.7
ECOL3_08
Southern and Baja California Pine-Oak Mountains
122.49
-0.0062
32.8
404.8
ECOL3_09
Eastern Cascades Slopes and Foothills
110.36
-0.0053
18.6
453.5
ECOL3_10
Columbia Plateau
105.57
-0.0006
88.8
3,858.9
ECOL3_ll
Blue Mountains
118.33
-0.0026
64.2
943.1
ECOL3_12
Snake River Plain
105.49
-0.0012
45.1
1,988.9
ECOL3_13
Central Basin and Range
101.85
-0.0008
22.9
2,901.7
ECOL3_14
Mojave Basin and Range
100.33
-0.0012
2.9
1,999.7
ECOL3_15
Columbia Mountains/Northern Rockies
154.23
-0.0085
50.9
321.4
ECOL3_16
Idaho Batholith
149.46
-0.0111
36.0
242.6
ECOL3_17
Middle Rockies
102.71
-0.0057
4.7
411.9
ECOL3_18
Wyoming Basin
102.05
-0.0005
41.8
4,792.9
ECOL3_19
Wasatch and Uinta Mountains
103.18
-0.0025
12.5
929.9
ECOL3_20
Colorado Plateaus
101.57
-0.0001
111.8
16,595.3
ECOL3_21
Southern Rockies
102.90
-0.0033
8.7
712.1
ECOL3_22
Arizona/New Mexico Plateau
100.30
-0.0001
31.6
24,144.6
ECOL3_23
Arizona/New Mexico Mountains
100.62
-0.0009
6.8
2,562.6
ECOL3_24
Chihuahuan Desert
101.79
-0.0014
12.8
1,671.6
ECOL3_25
High Plains
102.70
-0.0004
66.5
5,806.3
ECOL3_26
Southwestern Tablelands
103.35
-0.0004
74.0
5,239.0
ECOL3_27
Central Great Plains
103.49
-0.0004
94.9
6,462.6
ECOL3_28
Flint Hills
111.64
-0.0012
90.3
1,979.5
ECOL3_29
Cross Timbers
106.31
-0.0017
36.9
1,425.3
ECOL3_30
Edwards Plateau
106.83
-0.0070
9.4
336.3
ECOL3_31
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
100.74
-0.0008
8.7
2,731.7
ECOL3_32
Texas Blackland Prairies
110.38
-0.0011
91.6
2,226.9
ECOL3_33
East Central Texas Plains
106.96
-0.0008
84.8
2,987.0
ECOL3_34
Western Gulf Coastal Plain
103.78
-0.0012
31.1
1,964.6
ECOL3_35
South Central Plains
117.84
-0.0050
32.7
491.8
ECOL3_36
Ouachita Mountains
175.85
-0.0157
36.0
182.8
ECOL3_37
Arkansas Valley
124.25
-0.0060
35.9
416.7
ECOL3_38
Boston Mountains
240.61
-0.0252
34.8
126.1
ECOL3_39
Ozark Highlands
137.77
-0.0034
95.1
778.1
ECOL3_40
Central Irregular Plains
116.98
-0.0008
193.2
3,030.6
ECOL3_41
Canadian Rockies
102.38
-0.0064
3.7
364.9
ECOL3_42
Northwestern Glaciated Plains
101.25
-0.0002
49.9
9,287.6
ECOL3_43
Northwestern Great Plains
102.30
-0.0004
50.8
5,192.4
ECOL3_44
Nebraska Sand Hills
108.78
-0.0073
11.5
327.0
ECOL3_45
Piedmont
123.28
-0.0043
48.5
582.1
B-4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-3: Suspended Solids Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
SSCioo
SSCio
ECOL3_46
Aspen Parkland/Northern Glaciated Plains
106.80
-0.0005
121.8
4,382.1
ECOL3_47
Western Corn Belt Plains
113.45
-0.0008
150.6
2,899.9
ECOL3_48
Lake Manitoba and Lake Agassiz Plain
106.32
-0.0009
66.3
2,558.1
ECOL3_49
Northern Minnesota Wetlands
104.69
-0.0047
9.7
498.9
ECOL3_50
Northern Lakes and Forests
101.64
-0.0302
0.5
76.8
ECOL3_51
North Central Hardwood Forests
101.18
-0.0063
1.9
367.1
ECOL3_52
Driftless Area
113.90
-0.0025
51.8
968.9
ECOL3_53
Southeastern Wisconsin Till Plains
107.87
-0.0015
50.0
1,569.9
ECOL3_54
Central Corn Belt Plains
126.49
-0.0018
132.9
1,434.9
ECOL3_55
Eastern Corn Belt Plains
137.96
-0.0013
238.5
1,945.4
ECOL3_56
Southern Michigan/Northern Indiana Drift Plains
104.69
-0.0049
9.4
482.9
ECOL3_57
Huron/Erie Lake Plains
110.27
-0.0022
45.0
1,105.5
ECOL3_58
Northern Appalachian and Atlantic Maritime
Highlands
105.30
-0.0220
2.3
106.9
ECOL3_59
Northeastern Coastal Zone
109.98
-0.0213
4.5
112.6
ECOL3_60
Northern Allegheny Plateau
112.39
-0.0059
19.7
408.7
ECOL3_61
Erie Drift Plain
115.53
-0.0021
69.3
1,174.2
ECOL3_62
North Central Appalachians
122.90
-0.0192
10.7
130.6
ECOL3_63
Middle Atlantic Coastal Plain
105.17
-0.0077
6.6
306.4
ECOL3_64
Northern Piedmont
124.31
-0.0048
45.0
521.0
ECOL3_65
Southeastern Plains
118.94
-0.0065
26.8
382.9
ECOL3_66
Blue Ridge
108.09
-0.0080
9.7
297.3
ECOL3_67
Ridge and Valley
115.89
-0.0049
30.1
500.8
ECOL3_68
Southwestern Appalachians
124.64
-0.0070
31.5
360.3
ECOL3_69
Central Appalachians
121.03
-0.0113
16.9
220.7
ECOL3_70
Western Allegheny Plateau
120.20
-0.0030
61.8
835.8
ECOL3_71
Interior Plateau
137.46
-0.0038
84.8
698.8
ECOL3_72
Interior River Valleys and Hills
116.26
-0.0011
135.9
2,212.1
ECOL3_73
Mississippi Alluvial Plain
105.34
-0.0008
63.4
2,866.1
ECOL3_74
Mississippi Valley Loess Plains
115.94
-0.0026
56.1
930.1
ECOL3_75
Southern Coastal Plain
100.33
-0.0113
0.3
204.7
ECOL3_77
North Cascades
140.30
-0.0083
40.9
318.7
ECOL3_78
Klamath Mountains
142.69
-0.0124
28.6
213.7
ECOL3_79
Madrean Archipelago
100.41
-0.0021
1.9
1,078.2
ECOL3_80
Northern Basin and Range
102.69
-0.0010
26.5
2,319.2
ECOL3_81
Sonoran Desert
100.09
-0.0021
0.4
1,072.2
ECOL3_82
Acadian Plains and Hills
110.65
-0.0302
3.4
79.7
ECOL3_83
Eastern Great Lakes Lowlands
103.55
-0.0031
11.4
764.8
ECOL3_84
Atlantic Coastal Pine Barrens
105.25
-0.0173
3.0
135.8
ECOL3_85
California Coastal Sage, Chaparral, and Oak
Woodlands
104.56
-0.0005
95.8
5,039.6

Table B-4: TN Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TNioo
TN 10
ECOL3_01
Coast Range
117.12
-1.576
0.10
1.56
ECOL3_02
Strait of Georgia/Puget Lowland
115.02
-0.618
0.23
3.95
B-5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-4: TN Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TNioo
TN 10
ECOL3_03
Willamette Valley
124.45
-0.626
0.35
4.03
ECOL3_04
Cascades
140.20
-4.890
0.07
0.54
ECOL3_05
Sierra Nevada
147.87
-5.172
0.08
0.52
ECOL3_06
California Coastal Sage, Chaparral, and Oak
Woodlands
115.62
-0.753
0.19
3.25
ECOL3_07
Central California Valley
106.36
-0.182
0.34
13.02
ECOL3_08
Southern and Baja California Pine-Oak Mountains
132.91
-1.449
0.20
1.79
ECOL3_09
Eastern Cascades Slopes and Foothills
124.23
-2.589
0.08
0.97
ECOL3_10
Columbia Plateau
107.54
-0.213
0.34
11.13
ECOL3_ll
Blue Mountains
128.88
-1.825
0.14
1.40
ECOL3_12
Snake River Plain
112.05
-0.421
0.27
5.74
ECOL3_13
Central Basin and Range
142.81
-1.582
0.23
1.68
ECOL3_14
Mojave Basin and Range
168.00
-1.527
0.34
1.85
ECOL3_15
Columbia Mountains/Northern Rockies
162.78
-6.219
0.08
0.45
ECOL3_16
Idaho Batholith
175.32
-6.599
0.09
0.43
ECOL3_17
Middle Rockies
125.63
-1.555
0.15
1.63
ECOL3_18
Wyoming Basin
133.37
-0.991
0.29
2.61
ECOL3_19
Wasatch and Uinta Mountains
182.10
-3.323
0.18
0.87
ECOL3_20
Colorado Plateaus
139.56
-1.074
0.31
2.45
ECOL3_21
Southern Rockies
125.73
-1.312
0.17
1.93
ECOL3_22
Arizona/New Mexico Plateau
164.67
-1.394
0.36
2.01
ECOL3_23
Arizona/New Mexico Mountains
196.35
-2.556
0.26
1.16
ECOL3_24
Chihuahuan Desert
178.59
-1.966
0.29
1.47
ECOL3_25
High Plains
128.76
-0.238
1.06
10.73
ECOL3_26
Southwestern Tablelands
117.79
-0.402
0.41
6.14
ECOL3_27
Central Great Plains
122.53
-0.161
1.26
15.57
ECOL3_28
Flint Hills
172.99
-0.487
1.13
5.85
ECOL3_29
Cross Timbers
127.67
-0.539
0.45
4.73
ECOL3_30
Edwards Plateau
275.43
-2.830
0.36
1.17
ECOL3_31
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
134.52
-1.349
0.22
1.93
ECOL3_32
Texas Blackland Prairies
140.22
-0.528
0.64
5.00
ECOL3_33
East Central Texas Plains
147.35
-0.877
0.44
3.07
ECOL3_34
Western Gulf Coastal Plain
108.99
-0.486
0.18
4.91
ECOL3_35
South Central Plains
166.55
-1.506
0.34
1.87
ECOL3_36
Ouachita Mountains
549.75
-3.223
0.53
1.24
ECOL3_37
Arkansas Valley
177.73
-0.855
0.67
3.37
ECOL3_38
Boston Mountains
280.85
-1.715
0.60
1.94
ECOL3_39
Ozark Highlands
163.12
-0.707
0.69
3.95
ECOL3_40
Central Irregular Plains
180.12
-0.386
1.53
7.50
ECOL3_41
Canadian Rockies
168.86
-4.873
0.11
0.58
ECOL3_42
Northwestern Glaciated Plains
112.01
-0.198
0.57
12.19
ECOL3_43
Northwestern Great Plains
128.64
-0.450
0.56
5.67
ECOL3_44
Nebraska Sand Hills
130.07
-0.440
0.60
5.83
ECOL3_45
Piedmont
184.09
-1.008
0.61
2.89
ECOL3_46
Aspen Parkland/Northern Glaciated Plains
131.56
-0.109
2.52
23.65
ECOL3_47
Western Corn Belt Plains
135.26
-0.101
3.00
25.87
B-6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-4: TN Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TNioo
TN 10
ECOL3_48
Lake Manitoba and Lake Agassiz Plain
121.75
-0.137
1.44
18.24
ECOL3_49
Northern Minnesota Wetlands
223.00
-1.380
0.58
2.25
ECOL3_50
Northern Lakes and Forests
146.53
-1.166
0.33
2.30
ECOL3_51
North Central Hardwood Forests
119.82
-0.244
0.74
10.17
ECOL3_52
Driftless Area
143.37
-0.237
1.52
11.25
ECOL3_53
Southeastern Wisconsin Till Plains
130.76
-0.155
1.73
16.60
ECOL3_54
Central Corn Belt Plains
141.14
-0.110
3.14
24.13
ECOL3_55
Eastern Corn Belt Plains
122.49
-0.109
1.86
23.00
ECOL3_56
Southern Michigan/Northern Indiana Drift Plains
129.61
-0.236
1.10
10.86
ECOL3_57
Huron/Erie Lake Plains
118.83
-0.103
1.68
24.11
ECOL3_58
Northern Appalachian and Atlantic Maritime
Highlands
180.97
-2.805
0.21
1.03
ECOL3_59
Northeastern Coastal Zone
139.63
-1.023
0.33
2.58
ECOL3_60
Northern Allegheny Plateau
135.73
-0.742
0.41
3.52
ECOL3_61
Erie Drift Plain
174.63
-0.463
1.20
6.18
ECOL3_62
North Central Appalachians
173.28
-1.578
0.35
1.81
ECOL3_63
Middle Atlantic Coastal Plain
117.16
-0.371
0.43
6.63
ECOL3_64
Northern Piedmont
127.21
-0.327
0.74
7.78
ECOL3_65
Southeastern Plains
192.15
-1.201
0.54
2.46
ECOL3_66
Blue Ridge
276.75
-1.954
0.52
1.70
ECOL3_67
Ridge and Valley
141.88
-0.720
0.49
3.69
ECOL3_68
Southwestern Appalachians
256.93
-1.490
0.63
2.18
ECOL3_69
Central Appalachians
675.15
-3.064
0.62
1.37
ECOL3_70
Western Allegheny Plateau
340.07
-1.467
0.83
2.40
ECOL3_71
Interior Plateau
152.97
-0.594
0.72
4.59
ECOL3_72
Interior River Valleys and Hills
123.32
-0.196
1.07
12.84
ECOL3_73
Mississippi Alluvial Plain
119.35
-0.337
0.53
7.37
ECOL3_74
Mississippi Valley Loess Plains
161.09
-1.056
0.45
2.63
ECOL3_75
Southern Coastal Plain
150.19
-0.711
0.57
3.81
ECOL3_77
North Cascades
161.05
-5.800
0.08
0.48
ECOL3_78
Klamath Mountains
144.12
-5.333
0.07
0.50
ECOL3_79
Madrean Archipelago
184.29
-2.163
0.28
1.35
ECOL3_80
Northern Basin and Range
118.17
-1.049
0.16
2.36
ECOL3_81
Sonoran Desert
134.26
-1.398
0.21
1.86
ECOL3_82
Acadian Plains and Hills
153.19
-3.186
0.13
0.86
ECOL3_83
Eastern Great Lakes Lowlands
124.57
-0.396
0.55
6.37
ECOL3_84
Atlantic Coastal Pine Barrens
113.96
-0.612
0.21
3.97
ECOL3_85
California Coastal Sage, Chaparral, and Oak
Woodlands
108.05
-0.149
0.52
16.00

Table B-5: TP Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TPioo
TP io
ECOL3_01
Coast Range
120.62
-11.18
0.017
0.223
ECOL3_02
Strait of Georgia/Puget Lowland
116.41
-7.23
0.021
0.340
ECOL3_03
Willamette Valley
122.02
-4.53
0.044
0.552
ECOL3_04
Cascades
127.84
-19.74
0.012
0.129
B-7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-5: TP Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TPioo
TP 10
ECOL3_05
Sierra Nevada
120.03
-31.12
0.006
0.080
ECOL3_06
California Coastal Sage, Chaparral, and Oak
Woodlands
111.64
-5.08
0.022
0.475
ECOL3_07
Central California Valley
109.69
-2.16
0.043
1.110
ECOL3_08
Southern and Baja California Pine-Oak Mountains
109.66
-5.64
0.016
0.424
ECOL3_09
Eastern Cascades Slopes and Foothills
114.91
-8.82
0.016
0.277
ECOL3_10
Columbia Plateau
106.54
-0.98
0.064
2.409
ECOL3_ll
Blue Mountains
112.26
-4.21
0.027
0.575
ECOL3_12
Snake River Plain
104.86
-1.19
0.040
1.975
ECOL3_13
Central Basin and Range
106.44
-8.32
0.007
0.284
ECOL3_14
Mojave Basin and Range
102.55
-6.82
0.004
0.341
ECOL3_15
Columbia Mountains/Northern Rockies
119.55
-26.30
0.007
0.094
ECOL3_16
Idaho Batholith
124.76
-11.69
0.019
0.216
ECOL3_17
Middle Rockies
107.73
-5.56
0.013
0.427
ECOL3_18
Wyoming Basin
106.78
-1.31
0.050
1.810
ECOL3_19
Wasatch and Uinta Mountains
109.62
-15.21
0.006
0.157
ECOL3_20
Colorado Plateaus
107.19
-4.62
0.015
0.514
ECOL3_21
Southern Rockies
110.45
-6.82
0.015
0.352
ECOL3_22
Arizona/New Mexico Plateau
103.18
-4.06
0.008
0.575
ECOL3_23
Arizona/New Mexico Mountains
104.60
-13.34
0.003
0.176
ECOL3_24
Chihuahuan Desert
109.07
-12.20
0.007
0.196
ECOL3_25
High Plains
113.62
-0.57
0.225
4.282
ECOL3_26
Southwestern Tablelands
107.60
-1.24
0.059
1.913
ECOL3_27
Central Great Plains
112.74
-0.48
0.250
5.055
ECOL3_28
Flint Hills
129.43
-1.39
0.185
1.837
ECOL3_29
Cross Timbers
108.32
-3.40
0.023
0.700
ECOL3_30
Edwards Plateau
110.37
-26.58
0.004
0.090
ECOL3_31
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
102.67
-7.15
0.004
0.326
ECOL3_32
Texas Blackland Prairies
112.92
-1.99
0.061
1.221
ECOL3_33
East Central Texas Plains
106.42
-2.53
0.025
0.934
ECOL3_34
Western Gulf Coastal Plain
100.87
-1.57
0.006
1.469
ECOL3_35
South Central Plains
120.39
-7.58
0.024
0.328
ECOL3_36
Ouachita Mountains
133.54
-15.66
0.018
0.165
ECOL3_37
Arkansas Valley
112.48
-2.72
0.043
0.891
ECOL3_38
Boston Mountains
131.47
-9.61
0.028
0.268
ECOL3_39
Ozark Highlands
114.84
-3.37
0.041
0.724
ECOL3_40
Central Irregular Plains
164.67
-2.20
0.227
1.274
ECOL3_41
Canadian Rockies
134.76
-33.85
0.009
0.077
ECOL3_42
Northwestern Glaciated Plains
110.26
-0.62
0.158
3.877
ECOL3_43
Northwestern Great Plains
117.40
-1.13
0.142
2.186
ECOL3_44
Nebraska Sand Hills
105.59
-1.69
0.032
1.392
ECOL3_45
Piedmont
132.98
-5.22
0.055
0.496
ECOL3_46
Aspen Parkland/Northern Glaciated Plains
128.82
-0.76
0.332
3.353
ECOL3_47
Western Corn Belt Plains
172.45
-1.54
0.355
1.854
ECOL3_48
Lake Manitoba and Lake Agassiz Plain
112.93
-0.92
0.131
2.622
ECOL3_49
Northern Minnesota Wetlands
120.81
-12.32
0.015
0.202
B-8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix B: WQI Calculation
Table B-5: TP Subindex Curve Parameters, by Ecoregion
ID
Ecoregion Name
a
b
TPioo
TP 10
ECOL3_50
Northern Lakes and Forests
118.45
-14.48
0.012
0.171
ECOL3_51
North Central Hardwood Forests
111.56
-2.39
0.046
1.008
ECOL3_52
Driftless Area
139.72
-2.09
0.160
1.263
ECOL3_53
Southeastern Wisconsin Till Plains
132.83
-1.83
0.155
1.411
ECOL3_54
Central Corn Belt Plains
178.81
-2.30
0.253
1.255
ECOL3_55
Eastern Corn Belt Plains
186.94
-2.86
0.219
1.025
ECOL3_56
Southern Michigan/Northern Indiana Drift Plains
130.88
-3.90
0.069
0.659
ECOL3_57
Huron/Erie Lake Plains
142.40
-3.19
0.111
0.832
ECOL3_58
Northern Appalachian and Atlantic Maritime
Highlands
132.90
-30.01
0.009
0.086
ECOL3_59
Northeastern Coastal Zone
125.36
-13.84
0.016
0.183
ECOL3_60
Northern Allegheny Plateau
126.26
-9.88
0.024
0.257
ECOL3_61
Erie Drift Plain
134.57
-3.24
0.092
0.803
ECOL3_62
North Central Appalachians
148.98
-21.89
0.018
0.123
ECOL3_63
Middle Atlantic Coastal Plain
112.32
-4.26
0.027
0.568
ECOL3_64
Northern Piedmont
141.23
-5.01
0.069
0.528
ECOL3_65
Southeastern Plains
130.40
-7.65
0.035
0.336
ECOL3_66
Blue Ridge
117.13
-8.26
0.019
0.298
ECOL3_67
Ridge and Valley
113.75
-5.34
0.024
0.455
ECOL3_68
Southwestern Appalachians
127.64
-7.37
0.033
0.345
ECOL3_69
Central Appalachians
141.58
-19.20
0.018
0.138
ECOL3_70
Western Allegheny Plateau
154.57
-6.77
0.064
0.404
ECOL3_71
Interior Plateau
119.63
-2.12
0.085
1.172
ECOL3_72
Interior River Valleys and Hills
134.24
-1.63
0.181
1.595
ECOL3_73
Mississippi Alluvial Plain
102.40
-1.04
0.023
2.229
ECOL3_74
Mississippi Valley Loess Plains
115.53
-2.27
0.064
1.078
ECOL3_75
Southern Coastal Plain
113.24
-6.14
0.020
0.395
ECOL3_77
North Cascades
118.69
-17.30
0.010
0.143
ECOL3_78
Klamath Mountains
117.21
-28.37
0.006
0.087
ECOL3_79
Madrean Archipelago
104.02
-18.29
0.002
0.128
ECOL3_80
Northern Basin and Range
103.35
-2.23
0.015
1.048
ECOL3_81
Sonoran Desert
101.23
-8.38
0.001
0.276
ECOL3_82
Acadian Plains and Hills
113.37
-25.58
0.005
0.095
ECOL3_83
Eastern Great Lakes Lowlands
114.01
-3.62
0.036
0.673
ECOL3_84
Atlantic Coastal Pine Barrens
109.88
-11.65
0.008
0.206
ECOL3_85
California Coastal Sage, Chaparral, and Oak
Woodlands
104.34
-1.37
0.031
1.717
B-9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix C: Water Quality and Fish Tissue
Derivation of Ambient Water and Fish Tissue Concentrations in Receiving
and Downstrec aches
This appendix describes the methodology EPA used to estimate water and fish tissue concentrations under the
baseline and each of the regulatory options. The concentrations are used as inputs to estimate the water
quality changes and human health benefits of the regulatory options. Specifically, EPA used ambient water
toxics concentrations to derive fish tissue concentrations used to analyze human health effects from
consuming self-caught fish (see Chapter 5) and to analyze non-use benefits of water quality changes (see
Chapter 6). Nutrient and suspended solids concentrations are used to support analysis of non-use benefits
from water quality changes (see Chapter 6).
The overall modeling methodology builds on data and methods described in the Supplemental EA and
Supplemental TDD for the regulatory options (U.S. EPA, 2020f, 2020g). The following sections discuss
calculations of the toxics concentrations in ambient water and fish tissue and nutrient and sediment
concentrations in ambient water.
C.1 Toxics
Estimating Water Concentrations in each Reach
EPA first estimated the baseline and regulatory option toxics concentrations in reaches receiving steam
electric power plant discharges and downstream reaches.
The D-FATE model (see Chapter 3) was used to estimate water concentrations. The model tracks the fate and
transport of discharged pollutants through a reach network defined based on the medium resolution NHD.94
The hydrography network represented in the D-FATE model consists of 11,369 reaches within 300 km of a
steam electric power plant, 10,454 of which are estimated to be potentially fishable.95
The analysis involved the following key steps for the baseline and each of the regulatory options:
• Summing plant-level loadings to the receiving reach. EPA summed the estimated plant-level
annual average loads for each unique reach receiving plant discharges from steam electric power
plants in the baseline and under the regulatory options. For a description of the approach EPA used to
identify the receiving waterbodies, see U.S. EPA, 2020f.
• Performing dilution and transport calculations. The D-FATE model calculates the concentration
of the pollutant in a given reach based on the total mass transported to the reach from upstream
sources and the EROM flows for each reach from NHDPlus v2. In the model, a plant is assumed to
94 The USGS's National Hydrology Dataset (NHD) defines a reach as a continuous piece of surface water with similar hydrologic
characteristics. In the NHD each reach is assigned a reach code; a reach may be composed of a single feature, like a lake or
isolated stream, but reaches may also be composed of a number of contiguous features. Each reach code occurs only once
throughout the nation and once assigned a reach code is permanently associated with its reach. If the reach is deleted, its reach
code is retired.
95 Reaches represented in the D-FATE model are those estimated to be potentially fishable based on type and physical
characteristics. Because the D-FATE model calculates the movement of a chemical release downstream using flow data, reaches
must have at least one downstream or upstream connecting reach and have a non-negative flow and velocity. The D-FATE model
does not calculate concentrations for certain types of reaches, such as coastlines, treatment reservoirs, and bays; the downstream
path of any chemical is assumed to stop if one of these types of reach is encountered. Additionally, some types of reaches are
excluded from the set of fishable reaches, such as those designated as having Strahler Stream Order 1 in the NHDPlus, because
they do not have the flow rates and species diversity to support trophic level 3 and 4 species.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix C: Water Quality and Fish Tissue
release its annual load at a constant rate throughout the year. Each source-pollutant release is tracked
throughout the NHD reach network until the terminal reach.96
• Specifying concentrations in the water quality model. The D-FATE model includes background
data on estimated annual average pollutant concentrations to surface waters from facilities that
reported to the TRI in 2016. EPA added background concentrations where available to concentration
estimates from steam electric power plant dischargers.
EPA used the approach above to estimate annual average concentrations of ten toxics: arsenic, cadmium,
hexavalent chromium, copper, lead, mercury, nickel, selenium, thallium, and zinc.
Estimating Fish Tissue Concentrations in each Reach
To support analysis of the human health benefits associated with water quality improvements (see Chapter 4),
EPA estimated concentrations of arsenic, lead, and mercury in fish tissue based on the D-FATE model
outputs discussed above.
The methodology follows the same general approach described in the Supplemental EA for estimating fish
tissue concentrations for receiving reaches (U.S. EPA, 2020f), but applies the calculations to the larger set of
reaches modeled using D-FATE, which include not only the receiving reaches analyzed in the EA, but also
downstream reaches. Further, the calculations use D-FATE-estimated concentrations as inputs, which account
not only for the steam electric power plant discharges, but also other major dischargers that report to TRI. The
methodology follows the same general approach described in the Supplemental EA for estimating fish tissue
concentrations for receiving reaches (U.S. EPA, 2020f), but applies the calculations to the larger set of
reaches modeled using D-FATE, which include not only the immediate receiving reaches analyzed in the EA,
but also downstream reaches. Further, the calculations use D-FATE-estimated concentrations as inputs, which
account not only for the steam electric power plant discharges, but also other major dischargers that report to
TRI.
The analysis involved the following key steps for the baseline and each of the regulatory options:
1. Obtaining the relationship between water concentrations and fish tissue concentrations.
EPA used the results of the Immediate Receiving Water (IRW) model (see Supplemental EA,
U.S. EPA, 2020f) to parameterize the linear relationship between water concentrations in
receiving reaches and composite fish tissue concentrations (representative of trophic levels 3 and
4 fish consumed) in these same reaches for each of the three toxics.
2. Calculating fish tissue data for affected reaches. For reaches for which the D-FATE model
provides non-zero water concentrations (i. e., reaches affected by steam electric power plants or
other TRI dischargers), EPA used the relationship obtained in Step 1 to calculate a preliminary
fish tissue concentration for each pollutant.
3. Imputing the fish tissue concentrations for all other modeled reaches. For reaches for which
the D-FATE model calculates water concentrations, EPA added background fish tissue
concentrations based on the 10th percentile of the distribution of reported concentrations in fish
96 For some analyses, EPA limits the scope of reaches to 300 km (186 miles) downstream from steam electric power plant outfalls.
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix C: Water Quality and Fish Tissue
tissue samples in the National Listing Fish Advisory (NLFA) data97 (see Table C-l). EPA found
that the distribution of these samples was consistent with values reported in Wathen et cil. (2015)
and used the 10th percentile as representative of background levels in "clean" reaches not affected
by point source discharges.
4. Validating and adjusting the fish tissue concentrations based on empirical data, if needed.
EPA then applied the same method used to validate and adjust estimated fish tissue data in the
IRW model to ensure that the fish tissue concentrations calculated based on the D-FATE model
outputs are reasonable when compared to measured data. The approach involves applying order-
of-magnitude adjustments in cases where the preliminary concentrations are greater than
empirical measurements for a given reach or geographic area by an order of magnitude or more.
The Supplemental EA describes the methodology in greater detail (U.S. EPA, 2020f).
The analysis provides background toxic-specific composite fish fillet concentrations for each reach modeled
in the D-FATE model. Total fish tissue concentrations (D-FATE modeled concentrations plus background
concentrations) are summarized in Table C-2.
Table C-1: Background Fish Tissue Concentrations,
based on 10th percentile
Parameter
Pollutant Concentration (mg/kg)
As
0.039
Hg
0.058
Pb
0.039
Source: U.S. EPA Analysis, 2020
Table C-2: Imputed and Validated Fish Tissue Concentrations by Regulatory Option
Regulatory
Option
Fish Fillet Concentration (mg/kg)
Arsenic
Lead
Mercury
Min
Max
Mean
Min
Max
Mean
Min
Max
Mean
Period 1
Baseline
0.0390
0.2174
0.0391
0.0390
2.3318
0.0400
0.0580
7.2341
0.0669
Option A
0.0390
0.3363
0.0391
0.0390
3.8604
0.0406
0.0580
12.0181
0.0755
Option B
0.0390
0.3363
0.0391
0.0390
3.8604
0.0406
0.0580
12.0181
0.0743
Option C
0.0390
0.3363
0.0391
0.0390
3.8604
0.0406
0.0580
12.0181
0.0760
Period 2
Baseline
0.0390
0.0469
0.0390
0.0390
0.0918
0.0390
0.0580
1.6952
0.0590
Option A
0.0390
0.0471
0.0390
0.0390
0.0942
0.0390
0.0580
1.7436
0.0593
Option B
0.0390
0.0471
0.0390
0.0390
0.0942
0.0390
0.0580
1.7436
0.0592
Option C
0.0390
0.0399
0.0390
0.0390
0.0503
0.0390
0.0580
0.4824
0.0583
Source: U.S. EPA Analysis, 2020.
C.2 Nutrients and Suspended Sediment
EPA used the USGS's regional SPARROW models to estimate nutrient and sediment concentrations in
receiving and downstream reaches. The regional models used for this analysis are the five regional models
developed for the Pacific, Southwest, Midwest, Southeast, and Northeast regions for flow, total nitrogen
97 See https://fishadvisoryonline.epa.gov/general.aspx.
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix C: Water Quality and Fish Tissue
(TN), total phosphorus (TP), and suspended sediment (Ator, 2019; Hoos & Roland Ii, 2019; Robertson &
Saad, 2019; Wise, 2019; Wise et al., 2019). EPA adjusted the models to include a variable for steam electric
discharges using the following steps:
• Specifying a source load parameter for steam electric discharges. The regional SPARROW
models do not include an explicit explanatory variable for point sources related to industrial
dischargers (non publicly owned treatment works). EPA recalibrated the regional models by adding a
variable for steam electric loadings, initially setting all loadings for this parameter equal to zero,
assigning this new variable a calibration coefficient value of 1, and specifying zero land-to-water
delivery effects associated with this new variable.
• Appending steam electric TN, TP, and TSS loadings to regional input data. Once the regional
SPARROW models were recalibrated to include the steam electric loadings variable, EPA added the
steam electric TN, TP, and TSS98 loadings to the model input data and ran each regional model for
each pollutant to obtain catchment-level TN, TP, and SSC predictions.
For Periods 1 and 2, the SPARROW models output predicted annual average baseline and regulatory option
concentrations in each reach. EPA compared the baseline predictions to the predictions obtained for each of
the regulatory options to estimate changes in concentrations.
TSS loadings are converted to SSC values at this step by using location-specific relationships built into the SPARROW regional
models.
4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix D: Georeferencing Intakes
D Georeferencing Surface Water Intakes to the Medium-resolution Reach
Network
For the 2019 proposal analysis, EPA used the following steps to assign PWS surface water intakes to waters
represented in the medium-resolution NHD Plus version 2 dataset and identify those intakes potentially
affected by steam electric power plant discharges.
1. Identify the closest (simple cartesian distance) medium-resolution NHD feature (including Flowline,
Area, and Waterbody) to each PWS intake.
2. If the closest feature to a given intake was an NHD Flowline, reference the intake to this Flowline.
3. If the closest feature to a given intake was an NHD Area or Waterbody, consider the Flowlines
contained within or intersected by the Area/Waterbody.
a. If any of the Flowlines associated with the Area/Waterbody were on the flowpath
downstream from a steam electric power plant, select the Flowline within this set and closest
to the intake.
b. If none of the Flowlines were on the flowpath downstream from a steam electric power
plant, select the Flowline closest to the intake.
c. If there were no Flowlines associated with the Area/Waterbody, select the closest Flowline.
EPA then compared the set of Flowline COMIDs identified in steps 2 and 3 to NHD COMIDs in the
downstream flowpath of steam electric power plant discharges. COMIDs that georeferenced directly to the
downstream flowpath received a "Category 1" designation. Intakes that were georeferenced to COMIDs
within 10 km of the downstream flowpath received a "Category 2" designation. EPA included all intakes
within 10 km of the discharge flow path to account for cases where georeferencing did not select the correct
COMID based on uncertainty in the flow direction or stream network connectivity. For example, if a PWS
intake was located on a wide reach like the Mississippi River, the above methods may assign that intake to a
tributary COMID.
As discussed in Chapter 3, EPA did not model more complex waterbodies (e.g., Great Lakes) explicitly.
Therefore, the Agency reviewed all intakes within 50 miles of the plants discharging to the Great Lakes or
other non-modeled waterbodies to classify intakes that withdraw directly from the non-modeled waterbodies
as "Category 3". These intakes are excluded from the subsequent analysis.
Table D-l summarizes the intake categorization following the above steps.
Table D-1: Summary of Intakes Potentially Affected by Steam Electric Power Plant Discharges
Categorization
Number of Intakes
Category 1 (on flow path)
297
Category 2 (within 10 km of flow path) but not Category 1
313
Category 3 (on Great Lakes or other non-modeled waterbodies)
67
Total all categories
677
Source: U.S. EPA Analysis, 2019.
Figure D-l summarizes how EPA subset Category 1 and 2 PWS intakes for a more targeted categorization
review.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs Appendix D: Georeferencing Intakes
Figure D-1: PWS Intakes Review Subset
247 Category 1
facilities
Additional review
271 Category 2
faciities
610 unique PWS facilities
518 facilities georeferenced to a COMID
with non-zero changes in bromide
concentrations
554 active "community water systems
with a "permanent" supply
Source: U.S. EPA Analysis, 2020.
EPA evaluated the "Category 2" PWS intakes further using spatial reference to any steam electric
downstream flow paths and SDWIS facility information, namely facility name.
EPA excluded intakes from the benefits analysis if they were:
• on an upstream or visually unconnected body of water from the steam electric downstream flow path,
• did not sit on a visible body of water when looking at the topographical maps and/or orthophotos,
• had a PWS facility name indicating that it was not a surface water intake (i.e.. included the word
"welF').99
EPA recategorized intakes as Category 1 if they were:
• on the same NHD waterbody as the steam electric downstream flow path (prominent examples
include intakes on Lake Norman, Upper or Lower Potomac River, and Missouri River) or
• the PWS facility name in SDWIS corresponded with the named reach of the steam electric
downstream flow path.
99 This criterion resulted in the omission of only one facility in Tennessee.
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix D: Georeferencing Intakes
Of the 271 Category 2 facilities that EPA reviewed, 102 facilities were recategorized into
Category 1. Therefore, EPA included a total number of 349 PWS intakes100 in the analysis of the 2019
proposal.
For the final rule analysis, EPA updated the set of surface water intakes potentially affected by steam electric
power plant discharges by adding intakes associated with additional reaches identified after the 2019
proposal. This analysis identified one additional intake located on the flowpath downstream from receiving
reaches, which EPA included in the analysis described in Section 4.
100 Only intakes with facility types categorized by SDWIS as "Intake" or "Reservoir" were retained in the human health benefits
analysis. One of the 349 PWS intakes (PWS ID IA9778045) was categorized as "Infiltration Gallery" and was thus not included,
bringing the total number of PWS intakes included in the analysis to 348 (see Table 4-1).
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix E: Estimation of Exposed Population
E Estimation of Exposed Population for Fish Ingestion Pathway
The assessment uses the CBG as the geographic unit of analysis, assigning a radial distance {i.e., 50 miles)
from the CBG centroid. EPA assumes that all modeled reaches within this range are viable fishing sites, with
all unaffected reaches serving as viable substitutes for affected reaches within the area around the Census
Block Group.
By focusing on distance from the CBG, rather than distances from affected reaches, each household is only
included in the assessment once, eliminating the potential for double-counting of households that are near
multiple affected waterbodies.
Figure E-l presents a hypothetical example focusing on two CBGs (square at the center of each circular area),
each near five reaches with water quality changes under the regulatory options (thick red lines).
Figure E-1: Illustration of Intersection of CBGs and Reaches.
0
0
0
D
li1
I?
W
V
p
I?
0

0
5
¥
li'
0
Source: U.S. EPA, 2015a.
Note that a similar approach is used to identify populations for the analysis of non-market benefits in
Chapter 6. In that case, the circles represent the outer edge of the 100-mile buffer around each CBG.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix F: IQ Sensitivity Analysis
F Sensitivity Analysis for IQ Point-based Human Health Effects
EPA monetized the value of an IQ point based on the methodology from Salkever (1995) but with more
recent data from the 1997 National Longitudinal Survey of Youth (U.S. EPA, 2019b). As a sensitivity
analysis of the benefits of changes in lead and mercury exposure, EPA used alternative, more conservative
estimates provided in Lin et al. (2018) which indicate that a one-point IQ reduction reduces expected lifetime
earnings by 1.39 percent, as compared to 2.63 percent based on Salkever (1995). As noted in Sections 5.3 and
5.4, values of an IQ point used in the analysis of health effects in children from lead exposure are discounted
to the third year of life to represent the midpoint of the exposed children population, and values of an IQ point
used in the analysis of health effects associated with in-utero exposure to mercury are discounted to birth.
Table F-l summarizes the estimated values of an IQ point based on Lin et al. (2018), using 3 percent and
7 percent discount rates.
Table F-1: Value of an IQ Point (2018$) based on
Expected Reductions in Lifetime Earnings
Discount Rate
Value of an IQ Point3 (2018$)

Value of an IQ point Discounted to Age 3
3 percent
$11,279
7 percent
$2,371

Value of an IQ point Discounted to Birth
3 percent
$10,322
7 percent
$1,936
a. Values are adjusted for the cost of education.
Source: U.S. EPA, 2019b and 2019c analysis of data from Lin et al. (2018)
F.1 Health Effects in Children from Changes in Lead Exposure
Table F-2 shows the monetary values associated with changes in IQ losses from lead exposure via fish
consumption. EPA estimated that regulatory options A and B lead to small increases in lead exposure and, as
a result, forgone benefits, whereas Option C results in small reductions. The total net change in IQ point
losses over the entire population of children with changes in lead exposure ranges from -19 points to 12
points. Annualized monetary values of changes in IQ losses from differences in lead exposure, based on the
Lin et al. (2018) IQ point value, range from approximately -$9,000 to $4,000 (3 percent discount rate) and
from approximately -$2,000 to $400 (7 percent discount rate).
Table F-2: Estimated Monetary Value of Changes in IQ Losses for Children Exposed to Lead under
the Regulatory Options, Compared to Baseline
Regulatory Option
Average Annual
Number of Children 0
to 7 in Scope of the
Analysis'3
Total Change in IQ Point
Losses, 2021 Through 2047
in All Children 0 to 7 in
Scope of the Analysis
Annualized Value of Changes in IQ Point
Losses3
(Thousands of 2018$)

3 Percent Discount
Rate
7 Percent Discount
Rate
Option A
1,615,629
-19
LO
00
1
-$2.0
Option B
1,615,629
-11
-$5.7
-$1.5
Option C
1,615,629
12
$3.5
$0.4
1
-------
BCA for Revisions to the Steam Electric Power Generating ELGs Appendix F: IQ Sensitivity Analysis
Table F-2: Estimated Monetary Value of Changes in IQ Losses for Children Exposed to Lead under
the Regulatory Options, Compared to Baseline
Regulatory Option
Average Annual
Number of Children 0
to 7 in Scope of the
Analysis'3
Total Change in IQ Point
Losses, 2021 Through 2047
in All Children 0 to 7 in
Scope of the Analysis
Annualized Value of Changes in IQ Point
Losses3
(Thousands of 2018$)
3 Percent Discount
Rate
7 Percent Discount
Rate
a. Based on estimates that the loss of one IQ point results in the loss of 1.39 percent of lifetime earnings (following Lin et al. (2018)
values from U.S. EPA, 2019b).
b. The number of affected children is based on reaches analyzed across the regulatory options. Some of the children included in
this count see no changes in exposure under some options.
Source: U.S. EPA Analysis, 2020
F.2 Heath Effects in Children from Changes in Mercury Exposure
Table F-3 shows the estimated changes in IQ point losses for infants exposed to mercury in-utero and the
corresponding monetary values, using 3 percent and 7 percent discount rates. Regulatory options A and B
result in a small net increase in IQ losses and, as a result, in forgone benefits to society. Option C results in a
small net decrease in IQ point losses (positive benefits), with decreases in Period 2 larger than initial increases
in Period 1. However, the annualized monetary value for Option C is negative despite the overall decrease in
IQ point losses due to discounting. Annualized monetary values of changes in IQ losses from changes in
mercury exposure, based on the Lin et al. (2018) IQ point value, range from -$0.17 million (Option A) to -
$0.06 million (Option C) using a 3 percent discount rate.
Table F-3: Estimated Monetary Values from Changes in IQ Losses for Infants from Mercury Exposure
under the Regulatory Options, Compared to Baseline
Regulatory Option
Number of Infants in
Scope of the Analysis
per Yearc
Total Changes in IQ Point
Losses, 2021 to 2047 in All
Infants in Scope of the
Analysis
Annualized Value of Changes in IQ Point
Lossesab (Millions 2018$)
3 Percent Discount
Rate
7 Percent Discount
Rate
Option A
225,537
-201
-$0.17
-$0.06
Option B
225,537
-144
-$0.15
-$0.05
Option C
225,537
7 Id
-$0.06
-$0.04
a. Based on estimates that the loss of one IQ point results in the loss of 1.39 percent of lifetime earnings (following Lin et al. (2018)
values from U.S. EPA, 2019b and 2019c).
b. Negative values represent forgone benefits.
c. The number of affected children is based on reaches analyzed across the regulatory options. Some of the children included in this
count see no changes in exposure under some options.
d. Although Option C results in a small net decrease in IQ point losses (or positive benefits) due to larger decreases in Period 2 than
initial increases in Period 1, the annualized value for Option C is slightly negative due to discounting.
Source: U.S. EPA Analysis, 2020
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
Methodology for Estimating t Water Quality Changes
To estimate the nonmarket benefits of the water quality changes resulting from the regulatory options, EPA
used updated results from a meta-analysis of stated preference studies described in detail in Appendix H in the
2015 BCA (U.S. EPA, 2015a). The final rule is estimated to have mixed water quality effects at the reach-
level, either positive or negative, depending on the reach and regulatory option. Because the appropriate
welfare measure depends on property rights, WTP is the appropriate measure for water improvements,
whereas willingness-to-accept (WTA) compensation is the appropriate measure for water quality degradation.
EPA used WTP to value both positive and negative water quality changes due to the limited studies
measuring WTA. In theory, WTP and WTA should be close to each other for moderate environmental
changes. In practice, however, there is a significant divergence between WTA and WTP (Younjun el al,
2015). In particular, WTA for environmental goods tends to be significantly higher compared to WTP (Brown
& Gregory, 1999; Horowitz & McConnell, 2002). Brown and Gregory (1999) lists two dozen studies that
compared WTA to WTP for environmental goods (e.g., visibility and tree density in parks) and non-
environmental goods. The eleven studies of environmental goods included in the paper report WTA:WTP
ratios ranging from 2:1 to 5:1, with some ratios substantially higher. Horowitz and McConnell (2002) report a
mean ratio of WTA to WTP for non-market goods of 10.41 (standard error 2.53) based on 17 studies.
Therefore, using WTP to estimate the monetary value of all water quality changes (positive or negative) under
the final rule likely underestimates forgone benefits associated with the final rule. The magnitude of this
underestimate is uncertain.
To update results of the 2015 meta-analysis, EPA first conducted a literature review and identified 14 new
studies to augment the existing meta-data. EPA then re-estimated the 2015 meta-regression model and made
additional improvements to the model by introducing explanatory variables to account for potential
publication bias and differences in water quality metrics used in some of the added studies. A memorandum
titled 2020 Meta-regression Update Results (ICF, 2020, DCN SE09335) summarizes EPA's literature review
to identify additional studies, meta-data development and coding, model specification, and regression results
based on the 2020 meta-data. The 2020 meta-regression results are also briefly summarized below.
Like the 2015 meta-regression model, the updated meta-model satisfies the adding-up condition, a
theoretically desirable property.101 This condition ensures that if the model were used to estimate WTP for the
cumulative water quality change resulting from a number of CWA regulations, the benefits estimates would
be equal to the sum of benefits from using the model to estimate WTP for water quality changes separately for
each rule.
The meta-analysis is based on a meta-dataset of 65 stated preference studies, published between 1985 and
2017. Each of these studies used a stated preference approach to elicit survey respondents' willingness to pay
for water quality changes. The variables in the 2015 meta-data fall into four general categories:
101 For a WTP function WTP (WQIo, WQh, Yo) to satisfy the adding-up property, it must meet the simple condition that
WTP(WQIo, WQIi , Yo) + WTP( WQIi, WQh, Yo - WTP(WQIo, WQIi, Yo)) = WTP( WQIo, WQh , Yo) for all possible values
of baseline water quality (WQIo), potential future water quality levels (WQIi and WQh), and baseline income (Yo).
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
1. Study methodology and year variables characterize such features as the year in which a study was
conducted, payment vehicle and elicitation formats, WTP estimation method, and publication type.
These variables are included to explain differences in WTP across studies but are not expected to vary
across benefit transfer for different policy applications.
2. Region and surveyed populations variables characterize such features as the geographical region
within the United States in which the study was conducted, the average income of respondent
households and the representation of users and nonusers within the survey sample.
3. Sampled market and affected resource variables characterize features such as the geospatial scale (or
size) of affected waterbodies, the size of the market area over which populations were sampled, as
well as land cover and the quantity of substitute waterbodies.
4. Water quality (baseline and change) variables characterize baseline conditions and the extent of the
water quality change. To standardize the results across these studies, EPA expressed water quality
(baseline and change) in each study using the 100-point WQI, if they did not already employ the WQI
orWQL.
In addition to the variables included in the 2015 meta-regression, EPA included two new variables to estimate
the updated models. Six of the 14 studies added to the meta-data used a 100-point IBI to specify the baseline
and policy scenario conditions of the waterbody in question. These measures differ from the water quality
metrics used in the studies included the 2015 meta-data. To account for potential effects of the use of IBI on
interpretation of the baseline water quality and expected improvements, EPA included an interaction of a
binary variable indicating studies that use the IBI (IBI=1) as the water quality metric with the Q variable
(IBI Q).
Following best practices in economic meta-analysis literature, EPA also included the inverse of the square
root of sample size (n) as an independent variable on the right-hand side of the model to test for potential
publication bias in the meta-regression model. This variable serves as an instrumental variable (IV) or proxy
for the standard error (SE) of the welfare estimate (Stanley, 2005; Nelson, 2009).
The two additional variables allowed EPA to more accurately use the new studies while retaining the same
meta regression format from the 2015 rule. The variables were termed as follows:
• Interaction of a binary variable {i.e., a variable taking a value of 0 or 1) indicating studies that use the
index of biotic integrity (IBI) as the water quality metric with the Q variable (IBI Q). IBI Q = 1
when IBI was used by the study or = 0 otherwise.
• The inverse root of sample size as a proxy for the standard error of the estimate (inv rootsz).
Using the updated meta-dataset, EPA developed a meta-regression model that predicts how marginal WTP for
water quality improvements depends on a variety of methodological, population, resource, and water quality
change characteristics. The estimated meta-regression model predicts the marginal WTP values that would be
generated by a stated preference survey with a particular set of characteristics chosen to represent the water
quality changes and other specifics of the regulatory options where possible, and best practices in economic
literature (e.g., excluding outlier responses from estimating WTP). As with the 2015 meta-analysis, EPA
developed two versions of the meta-regression model (U.S. EPA, 2015a). Model 1 is used to provide EPA's
central estimate of non-market benefits and Model 2 is used to develop a range of estimates to account for
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
uncertainty in the resulting WTP values. The two models differ only in how they account for the magnitude of
the water quality changes presented to respondents in the original stated preference studies:
• Model 1 assumes that individuals' marginal WTP depends on the level of water quality, but not on
the magnitude of the water quality change specified in the survey. This restriction means that the
meta-model satisfies the adding-up condition, a theoretically desirable property.
• Model 2 allows marginal WTP to depend not only on the level of water quality but also on the
magnitude of the water quality change specified in the survey. The model allows for the possibility
that marginal WTP for improving from, for example, 49 to 50 on the water quality index depends on
whether respondents were asked to value a total water quality change of 10, 20, or 50 points on a
WQI scale. This model provides a better statistical fit to the meta-data, but it satisfies the adding-up
conditions only if the same magnitude of the water quality change is considered (e.g., 10 points). To
uniquely define the demand curve and satisfy the adding-up condition using this model, EPA treats
the water quality change variable as a methodological variable and therefore must make an
assumption about the size of the water quality change that would be appropriate to use in a stated
preference survey designed to value water quality changes resulting from the regulatory options.
When the water quality change is fixed at the mean of the meta-data, the predicted WTP is very close
to the central estimate from Model 1.
EPA used the two meta-regression models in a benefit transfer approach that follows standard methods
described by (Johnston et al. (2005), Shrestha el al. (2007), and Rosenberger and Phipps (2007). In particular,
literature on benefit transfer recommends selecting values for methodological variables included in the
regression equation with the goal of providing conservative WTP estimates, subject to consistency with
methodological guidance in the literature. The literature also recommends setting variables representing
policy outcomes and policy context (i.e.. resource and population characteristics) at the levels that might be
expected from a regulation. The benefit transfer approach uses CBGs as the geographic unit of analysis.102
The transfer approach involved projecting benefits in each CBG and year, based on the following general
benefit function:
Equation H-1.
\x\(MWTPYib) = Intercept + ^' (coefficientx (independent variable valuet)
Where
ln(MWTP y,b)
coefficient
independent
variable values
= The predicted natural log of marginal household WTP for a given year (7)
and CBG (E).
= A vector of variable coefficients from the meta-regression.
= A vector of independent variable values. Variables include baseline water
quality level ( WQI-BLy,b) and expected water quality under the regulatory
option (WQI-PCy.b) for a given year and CBG.
102 A Census Block group is a group of Census Blocks (the smallest geographic unit for the Census) in a contiguous area that never
crosses a State or county boundary. A block group typically contains a population between 600 and 3,000 individuals. There are
217,740 block groups in the 2010 Census. See http://www.census.gov/geo/maps-data/data/tallies/tractblock.html.
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
Here, ln(MWTPr,B) is the dependent variable in the meta-analysis—the natural log of approximated marginal
WTP per household, in a given CBG B for water quality in a given year Y.103 The baseline water quality level
( WQI-BLy,b) and expected water quality under the regulatory option ( WQI-PCy.b) were based on water quality
in waterbodies within a 100-mile buffer of the centroid of each CBG. A buffer of 100 miles is consistent with
Viscusi el al. (2008) and with the assumption that the majority of recreational trips would occur within a 2-
hour drive from home. Because marginal WTP is assumed to depend, according to Equation H-l, on both
baseline water quality level ( WQI-BLy,b) and expected water quality under the regulatory option ( WQI-PCy,b),
EPA estimated the marginal WTP for water quality changes resulting from the regulatory options at the mid-
point of the range over which water quality was changed, WQIy.b = (1/2)(WQI-BLy,b + WQI-PCy,b) .
In this analysis, EPA estimated WTP for the households in each CBG for waters within a 100-mile radius of
that CBG's centroid. EPA chose the 100 mile-radius because households are likely to be most familiar with
waterbodies and their qualities within the 100-mile distance. However, this assumption may be an
underestimate of the distance beyond which households have familiarity with and WTP for waterbodies
affected by steam electric power plant discharges and their quality. By focusing on a buffer around the CBG
as a unit of analysis, rather than buffers around affected waterbodies, each household is included in the
assessment exactly once, eliminating the potential for double-counting of households.104 Total national WTP
is calculated as the sum of estimated CBG-level WTP across all block groups that have at least one affected
waterbody within 100 miles. Using this approach, EPA is unable to analyze the WTP for CBGs with no
affected waters within 100 miles. Appendix E describes the methodology used to identify the relevant
populations.
In each CBG and year, predicted WTP per household is tailored by choosing appropriate input values for the
meta-analysis parameters describing the resource(s) valued, the extent of resource changes (i.e.. WQI-PCy.b),
the scale of resource changes relative to the size of the buffer and relative to available substitutes, the
characteristics of surveyed populations (e.g., users, nonusers), and other methodological variables. For
example, EPA projected that household income (an independent variable) changes over time, resulting in
household WTP values that vary by year.
Table G-l provides details on how EPA used the meta-analysis to predict household WTP for each CBG and
year. The table presents the estimated regression equation intercepts and variable coefficients (coefficient,) for
the two models, and the corresponding independent variables names and assigned values. The meta-regression
allows the Agency to forecast WTP based on assigned values for model variables that are chosen to represent
a resource change in the context of the regulatory options. EPA assigned a value to each model variable
corresponding with theory, characteristics of the water resources, and sites potentially affected by the
regulatory options. This follows general guidance from Bergstrom and Taylor (2006) that meta-analysis
benefit transfer should incorporate theoretical expectations and structures, at least in a weak form.
103 To satisfy the adding-up condition, as noted above, EPA normalized WTP values reported in the studies included in the meta-
data so that the dependent variable is MWTP per WQI point. This 'average' marginal WTP value is an approximation of the
MWTP value elicited in each survey scenario.
104 Population double-counting issues can arise when using "distance to waterbody" to assess simultaneous improvements to many
waterbodies.
4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
In this instance, EPA assigned six study and methodology variables, (thesis, volunt, nonparam, nonreviewed,
lump sum, and WTP median) a value of zero. One methodological variable, outliers trim, was included with
an assigned value of 1. Because the interpretation of the study year variable (Lnyear) is uncertain, EPA gave
the variable a value of 3.2189, which is the 75th percentile of the year values in the meta-data. This value
assignment reflects an equal probability that the variable represents a real time trend (in which case its value
should be set to the most recent year of the analysis) and spurious effects (in which case its values should be
set to the mean value from the meta-data). The choice experiment variable fee) was set to 1 to reflect recent
trends in the use of choice experiments within the environmental valuation literature. Finally, the inverse root
of sample size (inv rootsz) variable is set to the mean value for studies in the metadata. Model 2 includes an
additional variable, water quality change (Inquality_ch), which as discussed above allows the function to
reflect differences in marginal WTP based on differences in the magnitude of changes presented to survey
respondents when eliciting values. To ensure that the benefit transfer function satisfies the adding-up
condition, this variable was treated as a demand curve shifter, similar to the methodological control variables,
and held fixed for the benefit calculations. To estimate low and high values of WTP for water quality changes
resulting from the regulatory options, EPA estimated marginal WTP using two alternative settings of the
Inquality variable: AWQI = 5 units and AWQI = 50 units, which represent the low and high end of the range
of values observed in the meta-data.
All but one of the region and surveyed population variables vary based on the characteristics of each CBG.
EPA set the variable nonusersonly to zero for all CBGs because water quality changes are expected to
enhance both use and non-use values of the affected resources and thus benefit both users and nonusers (a
nonuser value of 1 implies WTP values that are representative of nonusers only, whereas the default value of
0 indicates that both users and nonusers are included in the surveyed population). For median household
income, EPA used CBG-level median household income data from the 2017 American Community Survey
(5-year data) and used a stepwise autoregressive forecasting method to estimate future annual state level
median household income.
The geospatial variables corresponding to the sampled market and scale of the affected resources (In ar agr,
ln_ar ratio , sub proportion) vary based on attributes of the CBG and attributes of the nearby affected
resources. For all options, the affected resource is based on the 10,454 NHD reaches potentially affected by
steam electric power generating plant discharges under baseline conditions. The affected resource for each
CBG is the portion of the 10,454 reaches that falls within the 100-mile buffer of the CBG. Spatial scale is
held fixed across regulatory options. The variable corresponding to the sampled market (ln_ar ratio) is set to
the mean value across all CBGs included in the analysis of benefits from water quality changes resulting from
the regulatory options, and thus does not vary across affected CBGs. To reflect characteristics of the resources
included in the analysis (i.e., rivers and streams), EPA set the variable river to 1 and mult type to 0. Other
waterbody types (e.g., Great Lakes, estuaries, enclosed lakes and ponds) are excluded from the analysis.
Because data on specific recreational uses of the water resources affected by the regulatory options are not
available, the recreational use variables (swim use, gamefish. boat use) are set to zero, which corresponds to
"unspecified" or "all" recreational uses in the meta-data.105 Water quality variables (Q and Inquality_ch) vary
across CBGs and regulatory options based on the magnitude of the reach-length weighted average water
quality changes in resources within scope of the analysis within the 100-mile buffer of each CBG. Interaction
105 If a particular recreational use was not specified in the survey instrument, EPA assesssed that survey respondents were thinking
of all relevant uses.
5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
of a binary variable indicating studies that use the IBI as the water quality metric with the Q variable (IBIO)
is set to zero because EPA's analysis of the final rule's benefits relies on the WQI as the water quality metric,
not the IBI.
Table G-1: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable Type
Variable
Coefficient
Assigned
Value
Explanation
Model 1
Model 2
Study
Methodology
and Year
intercept
-1.578
-1.646

Ce
0.488
0.329
1
Binary variable indicating that the study is a choice
experiment. Set to one to reflect that choice
experiments represent current state-of-art
methods in stated preference literature.
thesis
0.634
0.713
0
Binary variable indicating that the study is a thesis
or dissertation. Set to zero because studies
published in peer-reviewed journals are preferred.
Inyear
-0.157
-0.209
3.2189
Natural log of the year in which the study was
conducted (i.e., data were collected), converted to
an index by subtracting 1980. Set to the natural log
of the 75th percentile of the year index value for
studies in the metadata (25.0) to reflect
uncertainty in the variable interpretation. If the
variable represents a real time trend, the
appropriate value should reflect the most recent
year of the analysis. If it represents spurious
effects, the values should reflect the mid-point
from meta-data. Both interpretations are equally
probable.
volunt
-0.991
-0.842
0
Binary variable indicating that WTP was estimated
using a payment vehicle described as voluntary as
opposed to, for example, property taxes. Set to
zero because hypothetical voluntary payment
mechanisms are not incentive compatible (Mitchell
and Carson 1989).
outliers_trim
-0.385
-0.338
1
Binary variable indicating that outlier bids were
excluded when estimating WTP. Set to one
because WTP estimates that exclude outlier bids
are preferable.
nonparam
-0.577
-0.531
0
Binary variable indicating that regression analysis
was not used to model WTP. Set to zero because
use of the regression analysis to estimate WTP
values is preferred.
non_reviewed
-0.506
-0.550
0
Binary variable indicating that the study was not
published in a peer-reviewed journal. Set to zero
because studies published in peer-reviewed
journals are preferred.
lump_sum
0.542
0.486
0
Binary variable indicating that the study provided
WTP as a one-time, lump sum or provided annual
WTP values for a payment period of five years or
less. Set to zero to reflect that the majority of
studies from the meta-data estimated an annual
WTP, and to produce an annual WTP prediction.
6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
Table G-1: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable Type
Variable
Coefficient
Assigned
Value
Explanation
Model 1
Model 2
wtp_median
0.156
0.111
0
Binary variable indicating that the WTP measure
from the study is the median. Set to zero because
only average or mean WTP values in combination
with the number of affected households would
mathematically yield total benefits if the
distribution of WTP is not perfectly symmetrical.
inv_rootsz
-0.0282
-0.517
0.052
Inverse root of sample size [1 / square root(sample
size)], used as a proxy for the standard error of the
estimate. Set to the mean value for studies in the
metadata.
Region and
Surveyed
Population
northeast
0.644
0.443
Varies
Binary variable indicating that the affected
population is located in a Northeast U.S. state,
defined as ME, NH, VT, MA, Rl, CT, and NY. Set
based on the state in which the CBG is located.
central
0.672
0.665
Varies
Binary variable indicating that the affected
population is located in a Central U.S. state,
defined as OH, Ml, IN, IL, Wl, MN, IA, MO, ND, SD,
NE, KS, MT, WY, UT, and CO. Set based on the
state in which the CBG is located.
south
1.489
1.538
Varies
Binary variable indicating that the affected
population is located in a Southern U.S. state,
defined as NC, SC, GA, FL, KY, TN, MS, AL, AR, LA,
OK, TX, and NM. Set based on the state in which
the CBG is located.
nonusers_only
-0.355
-0.355
0
Binary variable indicating that the sampled
population included nonusers only; the alternative
case includes all households. Set to zero to
estimate the total value for aquatic habitat
changes for all households, including users and
nonusers.
Inincome
0.312
0.398
Varies
Natural log of median household income values
assigned separately for each CBG. Varies by year
based on the estimated income growth in future
years.
Sampled
Market and
Affected
Resource
mult_typea
-0.648
-0.617
0
Binary variable indicating that multiple waterbody
types are affected (e.g., river and lakes). Set to
zero because calculations are based exclusively on
rivers.
River
0.0196
-0.0213
1
Binary variable indicating that rivers are affected.
Set to one because calculations are based
exclusively on reach miles. EPA did not estimate
water quality changes for other waterbody types
(e.g., Great Lakes, estuaries, and enclosed lakes
and ponds).
swim use
0.0110
0.0405
0
Binary variables that identify studies in which
swimming, gamefish, and boating uses are
specifically identified. Since data on specific
recreational uses of the reaches affected by steam
electric power plant discharges are not available,
set to zero, which corresponds to all recreational
uses.
Gamefish
0.557
0.475
0
boat_use
-0.889
-0.786
0
7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix G: WTP Estimation Methodology
Table G-1: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable Type
Variable
Coefficient
Assigned
Explanation
Model 1
Model 2
Value

Natural log of the proportion of the affected

resource area which is agricultural based on

National Land Cover Database, reflecting the

nature of development in the area surrounding the

resource. Used Census county boundary layers to

ln_ar_agr
-0.621
-0.630
Varies
identify counties that intersect affected resources

within the 100-mile buffer of each CBG. For
intersecting counties, calculated the fraction of
total land area that is agricultural using the
National Land Cover Dataset (NLCD). The ln_ar_agr
variable was coded in the metadata to reflect the
area surrounding the affected resources.

The natural log of the ratio of the sampled area

(sa_area) relative to the affected resource area

(defined as the total area of counties that intersect

the affected resource[s]) (ar_total_area). In the

context of the steam electric scenario, sa_area is

ln_ar_ratio
-0.0891
-0.0939
1.491
set based on the total area within the 100-mile

buffer from the CBGs in scope of the analysis
(31,415 mi2) and the area of counties that intersect
affected reaches (COMIDs) within the 100-mile
radius. ln_ar_ratio is set to the mean value from
the all CBG's containing waters within the scope of
the analysis.

The size of the resources within the scope of the

analysis relative to available substitutes. Calculated

as the ratio of affected reaches miles to the total

sub_proportion
1.261
0.975
Varies
number of reach miles within the buffer that are
the same or greater than the order(s) of the
affected reaches within the buffer. Its value can
range from 0 to 1.
Water Quality

Because marginal WTP is assumed to depend on
both baseline water quality and expected water
quality under the regulatory option, this variable is
set to the mid-point of the range of water quality

Q
-0.0215
-0.0167
Varies
changes due to the regulatory options, WQI y,b =
(1/2)(WQI-BLy,b + WQI-PCy,b)- Calculated as the
length-weighted average WQI score for all
potentially affected reaches within the 100-mile
buffer of each CBG.

lnquality_ch
NA
-0.314
ln(5) or
ln(50)
Ln_quality_ch was set to the natural log of
AWQI=5 or AWQI=50 for high and low estimates of

the marginal WTP, respectively.

Interaction of a Binary variable indicating studies

that use the IBI as the water quality metric with

IBI_Q
-0.0502
-0.0463
0
the Q variable. Set to zero because the meta-
regression uses the WQI as the water quality
metric, not the IBI.
8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs Appendix G: WTP Estimation Methodology
Table G-1: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable Type
Variable
Coefficient
Assigned
Value
Explanation
Model 1
Model 2
a. The meta-data includes six waterbody categories (1) river and stream, (2) lake, (3) all freshwater, (4) estuary, (5) river and lake,
(6) salt pond/marshes, Variable multi-type takes on a value of 1 if the study focused on waterbody categories (3) and (6). EPA notes
that the overall effect of this variable should be considered in conjunction with the regional dummies (e.g., a study of the Lake
Okeechobee basin in Florida) and that only eight percent of all observations in the meta-data fall in the multiple waterbody
categories.
Source: U.S. EPA Analysis, 2020
The estimates for total WTP are shown in Table 6-2. EPA presents the results as a range because a water
quality change of +5 is closer to the size of water quality changes projected to result from the regulatory
options than the +20 analog to the central estimate, while the +50 represents the upper end of water quality
changes in existing surveys. In estimating forgone benefits (i.e., negative WTP estimates), +5 represents the
lower end of the sensitivity range, while the +50 represents the higher end of the sensitivity range.
9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
H Identification of Threatened and Endangered Species Potentially Affected by
the Final Rule Regulatory Options
As discussed in Chapter 7, EPA identified a total of 197 T&E species whose habitat range intersects reaches
affected by steam electric power plant discharges. These species include amphibians, arachnids, birds, clams,
crustaceans, fishes, insects, mammals, reptiles, and snails. Table H-l summarizes the number of species
within each group that have habitat ranges intersecting reaches with NRWQC exceedances for at least one
pollutant under the baseline or regulatory options in Period 1 (2021-2028) or Period 2 (2029-2047). As shown
in the table, several species of birds, clams, fishes, mammals, and snails have habitat ranges overlapping
reaches with baseline exceedances in Period 1. Additional species have exceedances under regulatory options
(Option C), but not the baseline (e.g., one species of amphibians, 2 species of birds, and 1 species of
mammals).
Water quality improvements in Period 2 generally eliminate exceedances under the baseline and regulatory
options, with the exception of one species of fish1"6 whose habitat range intersects reaches with remaining
baseline exceedances in Period 2.
Table H-1: Number of T&E Species with Habitat Range Intersecting Reaches Downstream from
Steam Electric Power Plant Outfalls, by Species Group
Species Group
Species
Count
Number of Species with Habitat Range Intersecting Reaches with NRWQC
Exceedances for at Least One Pollutant

Period 1
Period 2

a>
<
CO
u
a>
<
CO
u

£
£
£
£
£
£
£
£

a>
o
o
o
a>
o
o
o

CO
a.
a.
a.
CO
a.
a.
a.

CO
o
o
o
CO
o
o
o
Amphibians
8
0
0
0
1
0
0
0
0
Arachnids
6
0
0
0
0
0
0
0
0
Birds
25
3
3
3
5
0
0
0
0
Clams
62
16
17
17
17
0
0
0
0
Crustaceans
5
0
0
0
0
0
0
0
0
Fishes
35
7
7
7
7
l
0
0
0
Insects
10
0
0
0
0
0
0
0
0
Mammals
16
3
3
3
4
0
0
0
0
Reptiles
19
0
0
0
0
0
0
0
0
Snails
11
1
1
1
1
0
0
0
0
Total
197
30
31
31
35
l
0
0
0
Source: U.S. EPA Analysis, 20202
Table H-2 provides further details on the 197 T&E species whose habitat range intersects reaches affected by
steam electric power plant discharges. The table denotes, for each species, the number of reaches with at least
one reported exceedance of a NRWQC in the baseline or regulatory options in Period 1 and Period 2. The
table also includes the results of EPA's assessment of species vulnerability to water pollution. As noted in
Chapter 7, EPA classified species as follows:
100 As shown in Table H-2, Etheostoma trisella (Trispot darter) has baseline exceedances in Period 2.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
• Higher vulnerability - species living in aquatic habitats for several life history stages and/or species
that obtain a majority of their food from aquatic sources.
• Moderate vulnerability - species living in aquatic habitats for one life history stage and/or species that
obtain some of their food from aquatic sources.
• Lower vulnerability - species whose habitats overlap bodies of water, but whose life history traits and
food sources are terrestrial.
EPA obtained species life history data from a wide variety of sources to assess T&E species vulnerability to
water pollution. These sources included U.S. DOI, 2019; Froese and Pauly, 2019; NatureServe, 2020; NOAA
Fisheries, 2020; Southwest Fisheries Science Center (SWFSC), 2019; U.S. FWS, 2019a, 2019b, 2019c,
2019d, 2019e, 2019f, 2019g, 2020a, 2020b, 2020c, 2020e, 2020f, 2020g, 2020h, 2020i, 2020j, 2020k; Upper
Colorado River Endangered Fish Recovery Program, 2020.
Section 7.3.2 discusses impacts on five higher vulnerability species whose habitat ranges intersect reaches
with estimated changes in NRWQC exceedance status under the regulatory options.
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Group
Species
Count
Species Name
Vulnerability
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant
Period 1
Period 2
Baseline
Option A
Option B
Option C
Baseline
Option A
Option B
Option C
Amphibians
8
Ambystoma bishopi
Moderate
0
0
0
0
0
0
0
0
Ambystoma cingulatum
Moderate
0
0
0
1
0
0
0
0
Cryptobranchus alleganiensis
bishopi
Higher
0
0
0
0
0
0
0
0
Necturus alabamensis
Higher
0
0
0
0
0
0
0
0
Phaeognathus hubrichti
Lower
0
0
0
0
0
0
0
0
Plethodon nettingi
Lower
0
0
0
0
0
0
0
0
Rana pretiosa
Higher
0
0
0
0
0
0
0
0
Ranasevosa
Lower
0
0
0
0
0
0
0
0
Arachnids
6
Cicurina baronia
Lower
0
0
0
0
0
0
0
0
Cicurina madia
Lower
0
0
0
0
0
0
0
0
Cicurina venii
Lower
0
0
0
0
0
0
0
0
Cicurina vespera
Lower
0
0
0
0
0
0
0
0
Neoleptoneta microps
Lower
0
0
0
0
0
0
0
0
Texella cokendolpheri
Lower
0
0
0
0
0
0
0
0
Birds
25
Ammodramus savannarum
floridanus
Lower
0
0
0
0
0
0
0
0
Aphelocoma coerulescens
Lower
0
0
0
0
0
0
0
0
Brachyramphus marmoratus
Moderate
0
0
0
0
0
0
0
0
Calidris canutus rufa
Lower
0
0
0
0
0
0
0
0
Campephilus principalis
Lower
0
0
0
0
0
0
0
0
Charadrius melodus
Moderate
5
5
5
5
0
0
0
0
Coccyzus americanus
Lower
7
9
9
9
0
0
0
0
Dendroica chrysoparia
Lower
0
0
0
0
0
0
0
0
Empidonax traillii extimus
Lower
0
0
0
0
0
0
0
0
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Count
Species Name
Vulnerability
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant

Period 1
Period 2
Baseline
Option A
Option B
Option C
Baseline
Option A
Option B
Option C
Eremophila alpestris strigata
Lower
0
0
0
0
0
0
0
0
Falcofemoralis sept en trion alis
Lower
0
0
0
0
0
0
0
0
Grus americana
Moderate
0
0
0
0
0
0
0
0
Grus canadensis pulla
Moderate
0
0
0
0
0
0
0
0
Gymnogyps californianus
Lower
0
0
0
0
0
0
0
0
Mycteria americana
Moderate
0
0
0
1
0
0
0
0
Numenius borealis
Lower
0
0
0
0
0
0
0
0
Phoebastria (=Diomedea)
albatrus
Lower
0
0
0
0
0
0
0
0
Picoides borealis
Lower
0
0
0
1
0
0
0
0
Polyborus plancus audubonii
Lower
0
0
0
0
0
0
0
0
Rostrhamus sociabilis
plumbeus
Lower
0
0
0
0
0
0
0
0
Sterna antillarum
Higher
5
5
5
5
0
0
0
0
Sterna dougallii dougallii
Lower
0
0
0
0
0
0
0
0
Strix occidentalis lucida
Lower
0
0
0
0
0
0
0
0
Tympanuchus cupido
attwateri
Lower
0
0
0
0
0
0
0
0
Vermivora bachmanii
Moderate
0
0
0
0
0
0
0
0
Amblema neislerii
Higher
0
0
0
0
0
0
0
0
Cumberlandia monodonta
Higher
1
1
1
1
0
0
0
0
Cyprogenia stegaria
Higher
1
1
1
1
0
0
0
0
Dromus dromas
Higher
1
1
1
1
0
0
0
0
Elliptio chipolaensis
Higher
0
0
0
0
0
0
0
0
Elliptio lanceolata
Higher
0
0
0
0
0
0
0
0
Elliptio spinosa
Higher
0
0
0
0
0
0
0
0
Elliptoideus sloatianus
Higher
0
0
0
0
0
0
0
0
Epioblasma brevidens
Higher
0
0
0
0
0
0
0
0
Epioblasma capsaeformis
Higher
0
0
0
0
0
0
0
0
Epioblasma florentina
florentina
Higher
0
0
0
0
0
0
0
0
Epioblasma florentina walkeri
(=E. walkeri)
Higher
0
0
0
0
0
0
0
0
Epioblasma metastriata
Higher
0
0
0
0
0
0
0
0
Epioblasma obliquata
obliquata
Higher
0
0
0
0
0
0
0
0
Epioblasma othcaloogensis
Higher
0
0
0
0
0
0
0
0
Epioblasma torulosa
gubernaculum
Higher3
1
1
1
1
0
0
0
0
Epioblasma torulosa rangiana
Higher
0
0
0
0
0
0
0
0
Epioblasma torulosa torulosa
Higher
0
0
0
0
0
0
0
0
Epioblasma triquetra
Higher
0
0
0
0
0
0
0
0
Epioblasma turgidula
Higher
0
0
0
0
0
0
0
0
Fusconaia cor
Higher
1
1
1
1
0
0
0
0
62
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Count
Species Name
Vulnerability

a>
c
~aj
<
£
o
CO
£
o
u
£
o
a>
_£
~aj
<
£
o
CO
£
o
u
£
o

ru
BO
a.
O
a.
O
a.
o
ru
BO
a.
o
a.
o
a.
o
Fusconaia cuneolus
Higher
1
1
1
1
0
0
0
0
Hemistena lata
Higher
1
1
1
1
0
0
0
0
Lampsilis abrupta
Higher
1
1
1
1
0
0
0
0
Lampsilis altilis
Higher
0
0
0
0
0
0
0
0
Lampsilis higginsii
Higher
0
0
0
0
0
0
0
0
Lampsilis perovalis
Higher
0
0
0
0
0
0
0
0
Lampsilis rafinesqueana
Higher
0
0
0
0
0
0
0
0
Lampsilis subangulata
Higher
0
0
0
0
0
0
0
0
Lampsilis virescens
Higher
1
1
1
1
0
0
0
0
Lasmigona decorata
Higher
0
0
0
0
0
0
0
0
Lemiox rimosus
Higher
1
1
1
1
0
0
0
0
Leptodea leptodon
Higher
0
0
0
0
0
0
0
0
Margaritifera hembeli
Higher
0
0
0
0
0
0
0
0
Margaritifera marrianae
Higher
0
0
0
0
0
0
0
0
Medionidus acutissimus
Higher
0
0
0
0
0
0
0
0
Medionidus parvulus
Higher
0
0
0
0
0
0
0
0
Medionidus penicillatus
Higher
0
0
0
0
0
0
0
0
Obovaria retusa
Higher
1
1
1
1
0
0
0
0
Plethobasus cicatricosus
Higher
1
1
1
1
0
0
0
0
Plethobasus cooperianus
Higher
1
1
1
1
0
0
0
0
Plethobasus cyphyus
Higher
1
1
1
1
0
0
0
0
Pleurobema clava
Higher
0
1
1
1
0
0
0
0
Pleurobema collina
Higher
0
0
0
0
0
0
0
0
Pleurobema decisum
Higher
0
0
0
0
0
0
0
0
Pleurobema furvum
Higher
0
0
0
0
0
0
0
0
Pleurobema georgianum
Higher
0
0
0
0
0
0
0
0
Pleurobema hanleyianum
Higher
0
0
0
0
0
0
0
0
Pleurobema perovatum
Higher
0
0
0
0
0
0
0
0
Pleurobema plenum
Higher
1
1
1
1
0
0
0
0
Pleurobema pyriforme
Higher
0
0
0
0
0
0
0
0
Pleurobema taitianum
Higher
0
0
0
0
0
0
0
0
Pleuronaia dolabelloides
Higher
0
0
0
0
0
0
0
0
Potamilus capax
Higher
0
0
0
0
0
0
0
0
Potamilus inflatus
Higher
0
0
0
0
0
0
0
0
Ptychobranchus greenii
Higher
0
0
0
0
0
0
0
0
Quadrula cylindrica cylindrica
Higher
0
0
0
0
0
0
0
0
Quadrula cylindrica strigillata
Higher"
1
1
1
1
0
0
0
0
Quadrula fragosa
Higher
0
0
0
0
0
0
0
0
Quadrula intermedia
Higher
0
0
0
0
0
0
0
0
Villosa fabatis
Higher"
0
0
0
0
0
0
0
0
Villosa perpurpurea
Higher
0
0
0
0
0
0
0
0
Antrolana lira
Higher
0
0
0
0
0
0
0
0
Cambarus aculabrum
Higher
0
0
0
0
0
0
0
0
Gammarus acherondytes
Moderate
0
0
0
0
0
0
0
0
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant
Period 1
Period 2
4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Group
Species
Count
Species Name
Vulnerability
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant
Period 1
Period 2
Baseline
Option A
Option B
Option C
Baseline
Option A
Option B
Option C
Orconectes shoupic
Higher
0
0
0
0
0
0
0
0
Palaemonias alabamae
Moderate
0
0
0
0
0
0
0
0
Fishes
35
Acipenser oxyrinchus
(=oxyrhynchus) desotoi
Higher
0
0
0
0
0
0
0
0
Amblyopsis rosae
Higher
0
0
0
0
0
0
0
0
Chrosomus saylori
Higher"
0
0
0
0
0
0
0
0
Cottus specus
Higher"
0
0
0
0
0
0
0
0
Cyprinella caerulea
Higher
0
0
0
0
0
0
0
0
Elassoma alabama
Higher"
0
0
0
0
0
0
0
0
Erimonax monachus
Higher
0
0
0
0
0
0
0
0
Erimystax cahni
Higher
1
1
1
1
0
0
0
0
Etheostoma boschungi
Higher
0
0
0
0
0
0
0
0
Etheostoma chienense
Higher
0
0
0
0
0
0
0
0
Etheostoma etowahae
Higher
0
0
0
0
0
0
0
0
Etheostoma nianguae
Higher
0
0
0
0
0
0
0
0
Etheostoma osburni
Higher"
0
0
0
0
0
0
0
0
Etheostoma phytophilum
Higher
0
0
0
0
0
0
0
0
Etheostoma rubrum
Higher
0
0
0
0
0
0
0
0
Etheostoma scotti
Higher
0
0
0
0
0
0
0
0
Etheostoma sellare
Higher
0
0
0
0
0
0
0
0
Etheostoma trisella
Higher
4
4
4
4
3
0
0
0
Fundulus julisia
Higher"
1
1
1
1
0
0
0
0
Gila cypha
Higher
7
9
9
9
0
0
0
0
Gila elegans
Higher
0
0
0
0
0
0
0
0
Notropis cahabae
Higher
0
0
0
0
0
0
0
0
Notropis girardi
Higher
0
0
0
0
0
0
0
0
Notropis topeka (=tristis)
Higher
5
5
5
5
0
0
0
0
Noturus flavipinnis
Higher
1
1
1
1
0
0
0
0
Oncorhynchus clarkii stomias
Higher
0
0
0
0
0
0
0
0
Percina aurora
Higher
0
0
0
0
0
0
0
0
Percina rex
Higher
0
0
0
0
0
0
0
0
Percina tana si
Higher
0
0
0
0
0
0
0
0
Ptychocheilus lucius
Higher
7
9
9
9
0
0
0
0
Salvelinus confluentus
Higher
0
0
0
0
0
0
0
0
Scaphirhynchus albus
Higher
0
0
0
0
0
0
0
0
Scaphirhynchus suttkusi
Higher
0
0
0
0
0
0
0
0
Speoplatyrhinus poulsoni
Higher"
0
0
0
0
0
0
0
0
Xyrauchen texanus
Higher
0
0
0
0
0
0
0
0
Insects
10
Batrisodes venyivi
Lower
0
0
0
0
0
0
0
0
Bombus affinis
Lower
0
0
0
0
0
0
0
0
Cicindelidia floridana
Lower
0
0
0
0
0
0
0
0
Hesperia dacotae
Lower
0
0
0
0
0
0
0
0
Lycaeides melissa samuelis
Lower
0
0
0
0
0
0
0
0
5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Group
Species
Count
Species Name
Vulnerability
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant
Period 1
Period 2
Baseline
Option A
Option B
Option C
Baseline
Option A
Option B
Option C
Neonympha mitchellii
mitchellii
Lower
0
0
0
0
0
0
0
0
Nicrophorus americanus
Lower
0
0
0
0
0
0
0
0
Rhadine exilis
Lower
0
0
0
0
0
0
0
0
Rhadine infernalis
Lower
0
0
0
0
0
0
0
0
Somatochlora hineana
Higher
0
0
0
0
0
0
0
0
Mammals
16
Canis lupus
Lower
0
0
0
0
0
0
0
0
Corynorhinus (=Plecotus)
townsendii ingens
Lower
0
0
0
0
0
0
0
0
Corynorhinus (=Plecotus)
townsendii virginianus
Lower
0
0
0
0
0
0
0
0
Herpailurus (=Felis)
yagouaroundi cacomitli
Lower
0
0
0
0
0
0
0
0
Leopardus (=Felis) pardalis
Lower
0
0
0
0
0
0
0
0
Lynx canadensis
Lower
0
0
0
0
0
0
0
0
Mustela nigripes
Lower
0
0
0
0
0
0
0
0
Myotis grisescens
Moderate
1
2
2
2
0
0
0
0
Myotis septentrionalis
Lower
8
9
9
10
0
0
0
0
Myotis sodalis
Lower
3
4
4
4
0
0
0
0
Peromyscus polionotus
phasma
Lower
0
0
0
0
0
0
0
0
Puma (=Felis) concolor coryi
Lower
0
0
0
0
0
0
0
0
Thomomys mazama
pugetensis
Lower
0
0
0
0
0
0
0
0
Thomomys mazama tumuli
Lower
0
0
0
0
0
0
0
0
Thomomys mazama yelmensis
Lower
0
0
0
0
0
0
0
0
Trichechus manatus
Higher
0
0
0
1
0
0
0
0
Reptiles
19
Caretta caretta
Lower
0
0
0
0
0
0
0
0
Chelonia mydas
Lower
0
0
0
0
0
0
0
0
Clemmys muhlenbergii
Moderate
0
0
0
0
0
0
0
0
Crocodylus acutus
Lower
0
0
0
0
0
0
0
0
Dermochelys coriacea
Lower
0
0
0
0
0
0
0
0
Drymarchon corais couperi
Lower
0
0
0
0
0
0
0
0
Eretmochelys imbricata
Lower
0
0
0
0
0
0
0
0
Eumeces egregius lividus
Lower
0
0
0
0
0
0
0
0
Gopherus polyphemus
Lower
0
0
0
0
0
0
0
0
Graptemys flavimaculata
Higher
0
0
0
0
0
0
0
0
Lepidochelys kempii
Lower
0
0
0
0
0
0
0
0
Neoseps reynoldsi
Lower
0
0
0
0
0
0
0
0
Pituophis melanoleucus
lodingi
Lower
0
0
0
0
0
0
0
0
Pituophis ruthveni
Lower
0
0
0
0
0
0
0
0
Pseudemys alabamensis
Higher
0
0
0
0
0
0
0
0
Sistrurus catenatus
Lower
0
0
0
0
0
0
0
0
6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix H: T&E Species
Table H-2: T&E Species with Habitat Range Intersecting Reaches Downstream from Steam Electric
Power Plant Outfalls
Species
Group
Species
Count
Species Name
Vulnerability
Number of Intersected Reaches Exceeding
NRWQC for at Least One Pollutant
Period 1
Period 2
Baseline
Option A
Option B
Option C
Baseline
Option A
Option B
Option C
Sternotherus depressus
Higher
0
0
0
0
0
0
0
0
Thamnophis eques megalops
Lower
0
0
0
0
0
0
0
0
Thamnophis rufipunctatus
Lower
0
0
0
0
0
0
0
0
Snails
11
Athearnia anthonyi
Higher
1
1
1
1
0
0
0
0
Campeloma decampi
Higher
0
0
0
0
0
0
0
0
Discus macclintocki
Lower
0
0
0
0
0
0
0
0
Elimia crenatella
Higher
0
0
0
0
0
0
0
0
Leptoxis foremani
Higher
0
0
0
0
0
0
0
0
Leptoxis taeniata
Higher
0
0
0
0
0
0
0
0
Lioplax cyclostomaformis
Higher
0
0
0
0
0
0
0
0
Pleurocera foremani
Higher
0
0
0
0
0
0
0
0
Pyrgulopsis ogmorhaphe
Higher
0
0
0
0
0
0
0
0
Triodopsis platysayoides
Lower
0
0
0
0
0
0
0
0
Tulotoma magnifica
Higher
0
0
0
0
0
0
0
0
a This species is presumed extinct.
b While this species is categorized as highly vulnerable to water quality changes, it is endemic to waters (headwater streams and
springs) that are not likely to receive discharges from steam electric plants or be affected by upstream discharges. EPA did not
include this species in the set of T&E species with benefits or forgone benefits as a result of the final rule.
c U.S. Fish and Wildlife Service proposed delisting this species on 11/26/2019. See notice of proposed rulemaking "Endangered and
Threatened Wildlife and Plants: Removal of the Nashville Crayfish from the Federal List of Endangered and Threatened Wildlife."
(84 FR 65098)
Source: U.S. EPA Analysis, 2020
7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon
I Uncertainty Associated with Estimating the Social Cost of Carbon
The methodology used to develop interim domestic SC-CO2 estimates and uncertainty associated with the
interim SC-CO2 values are the same as described in the RIA for the Affordable Clean Energy (ACE) final rule
(see U.S. EPA, 2019i). This appendix applies the methodology to the analysis of the climate benefits of
changes in CO2 emissions under the regulatory options described in Chapter 8.
1.1 Overview of Methodology Used to Develop Interim Domestic SC-CO2 Estimates
The domestic SC-CO2 estimates rely on the same ensemble of three integrated assessment models (IAMs) that
were used to develop the global SC-CO2 estimates (DICE 2010, FUND 3.8, and PAGE 2009)107 used in the
benefits analysis of the 2015 rule (see U.S. EPA, 2015a). The three IAMs translate emissions into changes in
atmospheric greenhouse concentrations, atmospheric concentrations into changes in temperature, and changes
in temperature into economic damages. The emissions projections used in the models are based on specified
socioeconomic (GDP and population) pathways. These emissions are translated into atmospheric
concentrations, and concentrations are translated into warming based on each model's simplified
representation of the climate and a key parameter, equilibrium climate sensitivity. The effect of the changes is
estimated in terms of consumption-equivalent economic damages. As in the estimation of SC-CO2 estimates
used in the 2015 benefits analysis (U.S. EPA, 2015a), three key inputs were harmonized across the three
models: a probability distribution for equilibrium climate sensitivity; five scenarios for economic, population,
and emissions growth; and discount rates.108 All other model features were left unchanged. Future damages
are discounted using constant discount rates of both 3 and 7 percent, as recommended by OMB Circular A-4.
The domestic share of the global SC-CO2 - i. e., an approximation of the climate change impacts that occur
within U.S. borders - are calculated directly in both FUND and PAGE. However, DICE 2010 generates only
global SC-CO2 estimates. Therefore, EPA approximated U.S. damages as 10 percent of the global values from
the DICE model runs, based on the results from a regionalized version of the model (RICE 2010) reported in
Table 2 ofNordhaus (2017).
The steps involved in estimating the social cost of CO2 are as follows. The three integrated assessment models
(FUND, DICE, and PAGE) are run using the harmonized equilibrium climate sensitivity distribution, five
socioeconomic and emissions scenarios, and constant discount rates described above. Because the climate
sensitivity parameter is modeled probabilistically, and because PAGE and FUND incorporate uncertainty in
other model parameters, the final output from each model run is a distribution over the SC-CO2 in year t
based on a Monte Carlo simulation of 10,000 runs. For each of the IAMs, the basic computational steps for
calculating the social cost estimate in a particular year t are:
1. calculate the temperature effects and (consumption-equivalent) damages in each year resulting
from the baseline path of emissions;
2. adjust the model to reflect an additional unit of emissions in year t;
107 The full model names are as follows: Dynamic Integrated Climate and Economy (DICE); Climate Framework for Uncertainty,
Negotiation, and Distribution (FUND); and Policy Analysis of the Greenhouse Gas Effect (PAGE).
108 See the summary of the methodology in the 2015 Clean Power Plan docket, document ID number EPA-HQ-OAR-2013-0602-
37033, "Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866,
Interagency Working Group on Social Cost of Carbon United States Government, 2015. See also National Academies of
Sciences & Medicine, 2017 for a detailed discussion of each of these modeling assumptions.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon
3. recalculate the temperature effects and damages expected in all years beyond t resulting from this
adjusted path of emissions, as in step 1; and
4. subtract the damages computed in step 1 from those in step 3 in each model period and discount
the resulting path of marginal damages back to the year of emissions. In PAGE and FUND step 4
focuses on the damages attributed to the US region in the models. As noted above, DICE does not
explicitly include a separate US region in the model and therefore, EPA approximates U.S.
damages in step 4 as 10 percent of the global values based on the results of Nordhaus (2017).
This exercise produces 30 separate distributions of the SC-CO2 for a given year, the product of 3 models, 2
discount rates, and 5 socioeconomic scenarios. Following the approach used by the IWG, the estimates are
equally weighted across models and socioeconomic scenarios in order to reduce the dimensionality of the
results down to two separate distributions, one for each discount rate.
1.2 Treatment of Uncertainty in Interim Domestic SC-CO2 Estimates
There are various sources of uncertainty in the SC-CO2 estimates used in this BCA. Some uncertainties
pertain to aspects of the natural world, such as quantifying the physical effects of greenhouse gas emissions
on Earth systems. Other sources of uncertainty are associated with current and future human behavior and
well-being, such as population and economic growth, GHG emissions, the translation of Earth system
changes to economic damages, and the role of adaptation. It is important to note that even in the presence of
uncertainty, scientific and economic analysis can provide valuable information to the public and decision
makers, though the uncertainty should be acknowledged and when possible taken into account in the analysis
(Institute of Medicine, 2013). OMB Circular A-4 also requires a thorough discussion of key sources of
uncertainty in the calculation of benefits and costs, including more rigorous quantitative approaches for higher
consequence rules. This section summarizes the sources of uncertainty considered in a quantitative manner in
the domestic SC-CO2 estimates.
The domestic SC-CO2 estimates consider various sources of uncertainty through a combination of a multi-
model ensemble, probabilistic analysis, and scenario analysis. EPA provides a summary of this analysis here;
more detailed discussion of each model and the harmonized input assumptions can be found in the 2017
National Academies report. For example, the three IAMs used collectively span a wide range of Earth system
and economic outcomes to help reflect the uncertainty in the literature and in the underlying dynamics being
modeled. The use of an ensemble of three different models at least partially addresses the fact that no single
model includes all of the quantified economic damages. It also helps to reflect structural uncertainty across
the models, which is uncertainty in the underlying relationships between GHG emissions, Earth systems, and
economic damages that are included in the models. Bearing in mind the different limitations of each model
and lacking an objective basis upon which to differentially weight the models, the three integrated assessment
models are given equal weight in the analysis.
Monte Carlo techniques were used to run the IAMs a large number of times. In each simulation the uncertain
parameters are represented by random draws from their defined probability distributions. In all three models
the equilibrium climate sensitivity is treated probabilistically based on the probability distribution from Roe
and Baker (2007) calibrated to the Intergovernmental Panel on Climate Change (IPCC) AR4 consensus
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon
statement about this key parameter.109 The equilibrium climate sensitivity is a key parameter in this analysis
because it helps define the strength of the climate response to increasing GHG concentrations in the
atmosphere. In addition, the FUND and PAGE models define many of their parameters with probability
distributions instead of point estimates. For these two models, the model developers' default probability
distributions are maintained for all parameters other than those superseded by the harmonized inputs (/'. e.,
equilibrium climate sensitivity, socioeconomic and emissions scenarios, and discount rates). More
information on the uncertain parameters in PAGE and FUND is available upon request.
For the socioeconomic and emissions scenarios, uncertainty is included in the analysis by considering a range
of scenarios selected from the Stanford Energy Modeling Forum exercise, EMF-22. Given the dearth of
information on the likelihood of a full range of future socioeconomic pathways at the time the original
modeling was conducted, and without a basis for assigning differential weights to scenarios, the range of
uncertainty was reflected by simply weighting each of the five scenarios equally for the consolidated
estimates. To better understand how the results vary across scenarios, results of each model run are available
in the docket for the ACE final rule (Docket ID EPA-HQ-OAR-2017-0355).
The outcome of accounting for various sources of uncertainty using the approaches described above is a
frequency distribution of the SC-CO2 estimates for emissions occurring in a given year for each discount rate.
Unlike the approach taken for consolidating results across models and socioeconomic and emissions
scenarios, the SC-CO2 estimates are not pooled across different discount rates because the range of discount
rates reflects both uncertainty and, at least in part, different policy or value judgements; uncertainty regarding
this key assumption is discussed in more detail below. The frequency distributions reflect the uncertainty
around the input parameters for which probability distributions were defined, as well as from the multi-model
ensemble and socioeconomic and emissions scenarios where probabilities were implied by the equal
weighting assumption. It is important to note that the set of SC-CO2 estimates obtained from this analysis
does not yield a probability distribution that fully characterizes uncertainty about the SC-CO2 due to impact
categories omitted from the models and sources of uncertainty that have not been fully characterized due to
data limitations.
Figure 1-1 presents the frequency distribution of the domestic SC-CO2 estimates for emissions in 2030 for
each discount rate. Each distribution represents 150,000 estimates based on 10,000 simulations for each
combination of the three models and five socioeconomic and emissions scenarios. In general, the distributions
are skewed to the right and have long right tails, which tend to be longer for lower discount rates. To highlight
the difference between the impact of the discount rate on the SC-CO2 and other quantified sources of
uncertainty, the bars below the frequency distributions provide a symmetric representation of quantified
variability in the SC-CO2 estimates conditioned on each discount rate. The full set of SC-CO2 results through
2050 is available in the docket for the ACE final rule (Docket ID EPA-HQ-OAR-2017-0355).
109 Specifically, the Roe and Baker (2007) distribution for the climate sensitivity parameter was bounded between 0 and 10 with a
median of 3 °C and a cumulative probability between 2 and 4.5 °C of two-thirds.
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon

£=
O
03
=3
E
in
o
03
LD
O
O
CO
o
CM
d
o
o
7% Average = $1
3% Average = $8
~eL
DQqc
I
Discount Rate
~ 7%
~ 3%
5<". 95th Percentile
of Simulations
i r~i i iii
x
iii r
16 20

I I I I
32
T
0 4 8 12 16 20 24 28 32 36
Interim U.S. Domestic Social Cost of Carbon in 2030 [2016$ / metric ton C02]
40
Figure 1-1: Frequency Distribution of Interim Domestic SC-CO2 Estimates for 2030 (in 2016$ per Metric
Ton CO2)
As illustrated by the frequency distributions in Figure 1-1, the assumed discount rate plays a critical role in the
ultimate estimate of the social cost of carbon. This is because CO2 emissions today continue to impact society
far out into the future, so with a higher discount rate, costs that accrue to future generations are weighted less,
resulting in a lower estimate. Circular A-4 recommends that costs and benefits be discounted using the rates
of 3 percent and 7 percent to reflect the opportunity cost of consumption and capital, respectively. Circular A-
4 also recommends quantitative sensitivity analysis of key assumptions11", and offers guidance on what
sensitivity analysis can be conducted in cases where a rule will have important intergenerational benefits or
costs. To account for ethical considerations of future generations and potential uncertainty in the discount rate
over long time horizons, Circular A-4 suggests "further sensitivity analysis using a lower but positive
discount rate in addition to calculating net benefit using discount rates of 3 and 7 percent" (page 36) and notes
that research from the 1990s suggests intergenerational rates "from 1 to 3 percent per annum" (OMB, 2003).
EPA considers the uncertainty in this key assumption by calculating the domestic SC-CO2 based on a
2.5 percent discount rate, in addition to the 3 and 7 percent used in the main analysis. Using a 2.5 percent
discount rate, the average domestic SC-CO2 estimate across all the model runs for emissions occurring over
2020-2045 ranges from $10 to $14 per metric ton of CO2 (in 2018 dollars). In this case the forgone domestic
climate benefits under Option A in 2025 are $21 million; by 2035, the estimated forgone benefits decrease to
$9.4 million; and by 2045, the estimated forgone benefits are $35 million.
110 "If benefit or cost estimates depend heavily on certain assumptions, you should make those assumptions explicit and carry out
sensitivity analyses using plausible alternative assumptions." (OMB, 2003, page 42).
4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon
In addition to the approach to accounting for the quantifiable uncertainty described above, the scientific and
economics literature has further explored known sources of uncertainty related to estimates of the SC-CO2.
For example, researchers have published papers that explore the sensitivity of IAMs and the resulting SC-CO2
estimates to different assumptions embedded in the models (e.g., Hope, 2013, Anthoff & Tol, 2013, and
Nordhaus, 2014). However, there remain additional sources of uncertainty that have not been fully
characterized and explored due to remaining data limitations. Additional research is needed in order to expand
the quantification of various sources of uncertainty in estimates of the SC-CO2 (e.g., developing explicit
probability distributions for more inputs pertaining to climate impacts and their valuation). On the issue of
intergenerational discounting, some experts have argued that a declining discount rate would be appropriate to
analyze impacts that occur far into the future (Arrow etal., 2013). However, additional research and analysis
is still needed to develop a methodology for implementing a declining discount rate and to understand the
implications of applying these theoretical lessons in practice. The 2017 National Academies report also
provides recommendations pertaining to discounting, emphasizing the need to more explicitly model the
uncertainty surrounding discount rates over long time horizons, its connection to uncertainty in economic
growth, and, in turn, to climate damages using a Ramsey-like formula (National Academies of Sciences &
Medicine, 2017). These and other research needs are discussed in detail in the 2017 National Academies'
recommendations for a comprehensive update to the current methodology, including a more robust
incorporation of uncertainty.
1.3 Forgone Global Climate Benefits
In addition to requiring reporting of impacts at a domestic level, OMB Circular A-4 states that when an
agency "evaluate[s] a regulation that is likely to have effects beyond the borders of the United States, these
effects should be reported separately" (OMB, 2003; page 15).111 This guidance is relevant to the valuation of
damages from CO2 and other GHGs, given that GHGs contribute to damages around the world independent of
the country in which they are emitted. Therefore, this section presents the forgone global climate benefits in
2030 from this final rule using the global SC-CO2 estimates corresponding to the model runs that generated
the domestic SC-CO2 estimates used in the main analysis. The average global SC-CO2 estimate across all the
model runs for emissions occurring over 2025-2045 range from $6 to $13 per metric ton of CO2 emissions (in
2018 dollars) using a 7 percent discount rate, and $55 to $76 per metric ton of CO2 emissions (in 2018
dollars) using a 3 percent discount rate. The domestic SC-CO2 estimates presented above are approximately
18 percent and 14 percent of these global SC-CO2 estimates for the 7 percent and 3 percent discount rates,
respectively.
Applying these estimates to the forgone CO2 emission reductions results in estimated forgone global climate
benefits in 2025 of $13 million using a 7 percent discount rate and $110 million using a 3 percent discount
111 While Circular A-4 does not elaborate on this guidance, the basic argument for adopting a domestic only perspective for the
central benefit-cost analysis of domestic policies is based on the fact that the authority to regulate only extends to a nation's own
residents who have consented to adhere to the same set of rules and values for collective decision-making, as well as the
assumption that most domestic policies will have negligible effects on the welfare of other countries' residents (U.S. EPA, 2010a;
Kopp et al., 1997; Whittington & MacRae Jr, 1986). In the context of policies that are expected to result in substantial effects
outside of U.S. borders, an active literature has emerged discussing how to appropriately treat these impacts for purposes of
domestic policymaking (e.g., Gayer & Viscusi, 2016, 2017; Anthoff & Tol, 2010; Fraas etal., 2016; Revesz etal., 2017). This
discourse has been primarily focused on the regulation of GHGs, for which domestic policies may result in impacts outside of
U.S. borders due to the global nature of the pollutants.
5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix I: Uncertainty in Social Cost of Carbon
rate. By 2045, the estimated forgone global climate benefits are $33 million using a 7 percent discount rate
and $190 million using a 3 percent discount rate.
Under the sensitivity analysis considered above using a 2.5 percent discount rate, the average global SC-CO2
estimate across all the model runs for emissions occurring over 2025-2045 ranges from $80 to $105 per
metric ton of CO2 (2018 dollars); in this case the forgone global climate benefits in 2025 are $160 million; by
2045, the forgone global benefits in this sensitivity case increase to $260 million.
6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
J Methodology for liocleli ¦ Quality Changes for the Final Rule
As described in Chapter 8, EPA applied photochemical modeling to create air quality surfaces that were then
used in air pollution co-benefits calculations of the final rule (i.e.. Option A). The photochemical modeling-
based surfaces captured air pollution impacts resulting from changes in electricity generation profile due to
the incremental costs to generate electricity at plants incurring water treatment costs and did not simulate the
impact of emissions changes resulting from changes in energy use by steam electric power plants or resulting
from changes in trucking of CCR and other waste. This appendix describes methods used to create air quality
surfaces for the baseline scenario and a scenario representing water treatment technology implementation-
driven EGU profile changes for Option A for 7 years: 2021, 2023, 2025, 2030, 2035, 2040 and 2045. EPA
created air quality surfaces for the following pollutants and metrics: Annual average PM2 5; May-September
average of 8-hr daily maximum (MDA8) ozone; and April-October average of 1-hr daily maximum (MDA1)
ozone.
The photochemical model simulations as well as the basic methodology for determining air quality changes
are the same as those used in the Regulatory Impact Analysis for the Repeal of the Clean Power Plan, and the
Emission Guidelines for Greenhouse Gas Emissions from Existing Electric Utility Generating Units (U.S.
EPA, 2019i), also referred to the Affordable Clean Energy (ACE) rule. EPA calculated baseline and Option A
scenario EGU emissions estimates of NOx and SO2 for all 7 years using the Integrated Planning Model (IPM)
(Chapter 5 of the RIA; U.S. EPA, 2020d). EPA also used IPM outputs to estimate EGU emissions of PM25
and PM10 based on emission factors described in U.S. EPA (2020a). This appendix provides an overview of
the data and methods used to translate these emissions scenarios into air quality surfaces. Additional
information on the air quality modeling platform (inputs and set-up), model performance evaluation for ozone
and PM2 5, emissions processing for the photochemical modeling, and additional details and numerical
examples of the methodologies for developing ozone and PM2 5 spatial fields are available in the ACE rule
RIA (U.S. EPA, 2019i; see Chapter 8).
J.1 Air Quality Modeling Simulations
To create PM2 5 and ozone spatial fields representing the baseline and Option A, EPA leveraged available
photochemical modeling outputs that were created as part of the ACE rule RIA (U.S. EPA, 2019i). The full-
scale modeling used in this analysis included annual model simulations for a 2011 base year and a 2023 future
year to provide hourly concentrations of ozone and primary and secondarily formed PM2 5 component species
(e.g., sulfate, nitrate, ammonium, elemental carbon (EC), organic aerosol (OA), and crustal material112) for
both years nationwide. The photochemical modeling results for 2011 and 2023, in conjunction with modeling
to characterize the air quality impacts from groups of emissions sources (i.e., source apportionment modeling)
and emissions data for the baseline and Option A, were used to construct the air quality spatial fields that
reflect the influence of ELG-induced changes on ozone and PM2 5 concentrations over the period of 2021
through 2047 (represented by IPM run years 2021 through 2045).
EPA performed the air quality model simulations (i.e., model runs) using the Comprehensive Air Quality
Model with Extensions (CAMx) (Ramboll Environ International Corporation, 2016). The CAMx nationwide
112 Crustal material refers to elements that are commonly found in the earth's crust such as Aluminum, Calcium, Iron, Magnesium,
Manganese, Potassium, Silicon, Titanium and the associated oxygen atoms.
1
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
modeling domain (i.e.. the geographic area included in the modeling) covers all lower 48 states plus adjacent
portions of Canada and Mexico using a horizontal grid resolution of 12 x 12 km shown in Figure J-l.
Figure J-1: Air Quality Modeling Domain
—ft —'——

- -
*
m \
V-
t 1 t"- ~ 7 jC
t d
i
\ ^ t*
'< { \
1JUSJ Oom**. '
m f»».2u 7 y \ r
l
v - 'X
EPA tracked the impact of specific emissions sources on ozone and PM; 5 in the 2023 modeled case using a
tool called "source apportionment/' In general, source apportionment modeling quantifies the air quality
concentrations formed from individual, user-defined groups of emissions sources or "tags". These source tags
are tracked through the transport, dispersion, chemical transformation, and deposition processes within the
model to obtain hourly gridded113 contributions from the emissions in each individual tag to hourly modeled
concentrations of ozone and PMj j.114 Thus, the source apportionment method provides an estimate of the
effect of changes in emissions from each group of emissions sources {i.e., each tag) to changes in ozone and
PM; i concentrations. For this analysis EPA applied outputs from source apportionment modeling for ozone
and PM2 5 using the 2023 modeled case to obtain the contributions from EGU emissions as well as other
sources to ozone and to PM25 component species concentrations.11' EPA modeled ozone contributions using
the Ozone Source Apportionment Technique/Anthropogenic Precursor Culpability Assessment
(OSAT/APCA) tool and modeled PM25 component species contributions using the Particulate Source
Apportionment Technique (PSAT) tool116. The source apportionment modeling, which was already available
from analysis performed to support the ACE rule RIA (U.S. EPA, 2019i) was used to quantify the
contributions from EGU emissions on a state-by-state or, in some cases, on a multi-state basis. For ozone,
EPA modeled the contributions from the 2023 EGU sector emissions of NOx and VOC to hourly ozone
concentrations for the penod April through October to provide data for developing spatial fields for the two
seasonal ozone benefits metrics identified above (i.e.. for the May-September seasonal average MDA8 ozone
and the April-October seasonal average MDA1 ozone). For PM25, EPA modeled the contributions from the
113 Hourly contribution information is provided for each grid cell to provide spatial patterns of the contributions from each tag.
114 The sum of the contributions in a model grid cell from each tag for a pollutant equals the total concentration of that pollutant in
the grid cell.
113 I11 the source apportionment modeling for PM2.5 EPA tracked the source contributions from primary, but not secondary organic
aerosols (SOA). The method for treating SOA concentrations is described in U.S. EPA (2019i), Chapter 8.
116 OSAT/APCA and PSAT tools are described in Ramboll Environ International Corporation (2016).
2
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
2023 EGU sector emissions of SO2, NOx, and directly emitted PM2.5 for the entire year to inform the
development of spatial fields of annual mean PM2.5. For each state, or multi-state group, the Agency
separately tagged EGU emissions depending on whether the emissions were from coal-fired units or non-coal
units.117 In addition to tagging coal-fired and non-coal EGU emissions EPA also tracked the ozone and PM2.5
contributions from all other sources.
Examples of the magnitude and spatial extent of ozone tagged contributions are provided in Figure J-2
through Figure J-5 for coal and non-coal EGUs in Pennsylvania and Texas. These figures show how both the
magnitude and the spatial patterns of contributions can differ between coal and non-coal EGU units within a
state and downwind. In addition, the figures demonstrate that the spatial extent of contributions can vary
substantially from state to state depending on the location of sources, the magnitude of their emissions, and
meteorology. Moreover, day to day variations in meteorology can have a substantial impact on day to day
patterns in contributions, which are captured in the analysis. While EPA used the daily contributions in the
calculations, seasonal average contributions are presented here to provide a general illustration of the
differential spatial patterns of contribution.
Figure J-2: Map of Pennsylvania Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)
PA Coal EGU Ozone Contributions
May-Sep Mean of MDA8
1 h
1 396ppb
Min = O.OOE+O at (1,1), Max = 2.039 at (327,151)
117 For the purposes of this analysis non-coal units include natural gas, oil, biomass, municipal waste combustion and waste coal
EGUs.
3
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-3: Map of Pennsylvania Non-Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone
(ppb)
PA Non-coal EGU Ozone Contributions
May-Sep Mean of MDA8
jl
1 396 ppb
Min = 0.00E+0 at (1,1), Max = 1.808 at (344,152)
Min = O.OOE+0 at (1,1), Max = 4.818 at (221,G6)
Figure J-4: Map of Texas Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)
TX Coal EGU Ozone Contributions
May-Sep Mean of MDA8
4
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-5: Map of Texas Non-Coal EGU Tag Contribution to Seasonal Average MDA8 Ozone (ppb)
TX Non-coal EGU Ozone Contributions
May-Sep Mean of MDA8
jl
1 396ppb
Min = 0.00E+0 at (1,1), Max = 2.310 at (227,44)
Examples of the magnitude and spatial extent of tagged contributions for PM2.5 component species are
provided in Figure J-6 through Figure J-l 1. Examples are provided for coal-fired EGUs in Indiana. These
figures show how both the magnitude and the spatial patterns of contributions can differ by season and by
PM2.5 component species. The species which are formed through chemical reactions in the atmosphere
(sulfate and nitrate) have a more regional signal than directly emitted primary PM2.5 (OA, EC, and crustal
material) whose impact is more local in nature. In addition, the chemistry and transport can vary by season
with nitrate contributions being higher in the winter than in the summer and sulfate contributions being higher
in the summer than in the winter.
5
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-6: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January-March) Nitrate
(pg/m3)
IN Coal EGU Nitrate Contributions
Quarterly Avg - Winter
Min = 0.00E+0 at (1,1), Max = 0.151 at (274,124)
Min = 0.0QE+0 at (1,1), Max = 0.022 at (274,125)
396 ug/m3
Figure J-7: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September)
Nitrate (pg/m3)
IN Coal EGU Nitrate Contributions
Quarterly Avg ¦ Summer
6
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
396 Uf*/m3
Min = O.OOE-H) at (1,1), Max = 0.123 at (269,120)
-March)
Figure J-8: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January
Sulfate (|jg/m3)
IN Coal EGU Sulfate Contributions
Quarterly Avg - Winter
Figure J-9: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September)
Sulfate (pg/m3)
IN Coal EGU Sulfate Contributions
396 ug/m3
Quarterly Avg - Summer
7
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs Appendix J: Air Quality Modeling Methodology
Min = O.OOE+O at (1,1), Max = 0.108 at (272,120)
396 ug/m3
Figure J-10: Map of Indiana Coal EGU Tag Contributions to Wintertime Average (January-March)
Primary PM2.5 (|jg/m3)
IN Coal EGU Primary PM Contributions
Quarterly Avg - Winter
Figure J-11: Map of Indiana Coal EGU Tag Contributions to Summertime Average (July-September)
Primary PM2.5 (pg/m3)
IN Coal EGU Primary PM Contributions
Quarterly Avg - Sum m er
Min = O.OOE+O at (1,1), Max = 0.099 at (272,120)
8
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
J.2 Applying Modeling Outputs to Create Spatial Fields
To create the air quality surfaces, EPA used the 2023 source apportion modeling outputs in combination with
the endogenous IPM estimates of EGU SO2, NOx, and exogenously calculated PM2.5118 for each scenario in
each of the 7 years. EGU emissions were first aggregated according to the sources associated with each tag
(i.e., coal and non-coal EGU emissions on a state-by-state basis). EPA scaled contributions from each "tag"
based on the ratio of emissions in the year/scenario being evaluated to the emissions in the modeled 2023
scenario119. Scaling ratios for PM2 5 components that are emitted directly from the source (OA, EC, crustal)
were based on the relative changes in annual primary PM2 5 emissions between the 2023 emissions case and
the baseline and the Option A scenarios. Scaling ratios for components that are formed through chemical
reactions in the atmosphere were created as follows: scaling ratios for sulfate were based on relative changes
in annual SO2 emissions; scaling ratios for nitrate were based on relative changes in annual NOx emissions;
and scaling ratios for ozone formed in NOx-limited regimes120 ("03N") were based on relative changes in
ozone season (May-September) NOx emissions. The scaling ratios that were applied to each species and
scenario are provided in Table J-l through Table J-16. EPA held tags representing sources other than EGUs
constant at 2023 levels between the baseline and Option A for all years. For each year and scenario, EPA
summed the scaled contributions from all sources to create a gridded surface of total modeled O3 or total
modeled PM2 5. Finally, the Agency created "fused" fields based on the combination of this modeled data with
observed concentrations at air quality monitoring locations which were the bases for the BenMAP-CE runs.
Steps in this process are described below.
EPA used the following data to create the spatial fields of ozone and PM2 5 concentrations for the baseline and
Option A scenario in each year:
1. Emissions totals used in the 2023 source apportionment modeling: 2023 annual EGU SO2, NOx,
and directly emitted PM2 5 emissions121 and 2023 ozone season122 EGU NOx emissions for each
EGU tag as described in Chapter 8 of U.S. EPA (2019i);
2. 2021, 2023, 2025, 2030, 2035, 2040 and 2045 annual EGU emissions of SO2, NOx, and directly
emitted PM2 5 and EGU ozone season NOx emissions for the baseline and Option A scenario that
correspond to each of the EGU tags defined for the 2023 source apportionment modeling;
118 As described in Chapter 8, EPA estimated PM2.5 and PM10 emissions by multiplying the generation predicted for each IPM plant
type (ultrasupercritical coal without carbon capture and storage, combined cycle, combustion turbine, etc.) by a type-specific
empirical emission factor derived from the 2016 National Emissions Inventory (NEI) and other data sources. The emission
factors reflect the fuel type (including coal rank), FGD controls, and state emission limits for each plant type, where applicable.
See U.S. EPA (2020a) for details.
119 Note that while there were no EGU emissions from Washington D.C. in the 2023 source apportionment simulations, there were
small emissions predicted in the baseline and Option A scenarios (<10 ton per year of NOx and 0 tons per year of SO2,). Since
the emissions were small and there was no associated source apportionment tag to scale to, we did not include any impact of
Washington D.C. EGU emissions in the air quality surfaces. We also note that changes in Washington D.C. EGU emissions
between the Option A and baseline scenarios for all years were less than 1 tpy for all pollutants
120 The CAMx model internally determines whether the ozone formation regime is NOx-limited or VOC-limited depending on
predicted ratios of indicator chemical species
121 See footnote 118.
122 "Ozone season NOx emissions" refers to total NOx (tons) emitted during the period of May-September.
9
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
3. Daily 2011 and 2023 modeling-based concentrations of 24-hour average PM25 component species
and MDA1 and MDA8 ozone;
4. 2023 daily contributions to 24-hour average PM2 5 component species and MDA1 and MDA8
ozone from each of the various source tags; and
5. Base period (2011) "fused surfaces" of measured and modeled air quality123 representing
quarterly average PM2 5 component species concentrations and ozone concentrations for the two
seasonal average ozone metrics. These "fused surfaces" use the ambient data to adjust modeled
fields to match observed data at locations of monitoring sites. Details on the methods for creating
fused surfaces are provided in Chapter 8 of U.S. EPA (2019i).
Next, we identify the general process for developing the spatial fields for PM2 5 using the 2025 baseline as an
example to illustrate the procedure. The steps in this process are as follows:
1. Use the EGU annual SO2, NOx, and directly emitted PM2 5 emissions124 for the 2025 baseline and
the corresponding modeled 2023 SO2, NOx, and directly emitted PM2 5 emissions to calculate the
ratio of 2025 baseline emissions to modeled 2023 emissions for each of these pollutants for each
EGU tag (i.e.. a scaling ratio for each pollutant and each tag).
2. Multiply the tag-specific 2025 to 2023 EGU emissions-based scaling ratios from step (1) by the
corresponding 365 gridded daily 24-hour average PM2 5 component species contributions from
the 2023 contribution modeling. Apply the emissions ratios for SO2 to sulfate contributions; apply
the ratios for annual NOx to nitrate contributions; and apply the ratios for directly emitted PM2 5
to the EGU contributions to primary OA, EC and crustal material. This step results in 365
adjusted gridded daily PM2 5 component species contributions for each EGUs tag that reflects the
emissions in the 2025 baseline.
3. For each individual PM2 5 component species, sum the adjusted gridded contributions for each
EGU tag from step (2) to produce a gridded daily EGU tag total.
4. Combine the daily total EGU contributions for each PM2 5 component species from step (3) with
the species contributions from source tags representing all other sources of PM25. As part of this
step also add the total secondary organic aerosol concentrations from the 2023 modeling to the
net EGU contributions of primary OA. Note that the secondary organic aerosol concentration
does not change between scenarios. This step results in 24-hour average PM2 5 component species
concentrations for the 2025 baseline in each model grid cell, nationwide for each day in the year.
5. For each PM2 5 component species, average the daily concentrations from step (4) for each quarter
of the year.
123 In this analysis, a "fused surface" represents a spatial field of concentrations of a particular pollutant that was derived by applying
the Enhanced Voronoi Neighbor Averaging with adjustment using modeled and measured air quality data (i. e., eVNA) technique
(Ding etal., 2016).
124 The 2021, 2023, 2025, 2030, 2035, 2040 and 2045 EGU SO2, NOx and directly emitted PM2.5 emissions for the baseline and
Option A scenarios were obtained from IPM outputs described in Chapter 8 of this BCA and in Chapter 5 of the RIA (U.S. EPA,
2020d).
10
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
6. Divide the quarterly average PM2 5 component species concentrations from step (5)125 by the
corresponding quarterly average species concentrations from the 2011 CAMx model run. This
step provides a Relative Response Factor (i.e.. RRF) between 2011 and the 2025 baseline for each
species in each model grid cell.
7. Multiply the species-specific quarterly RRFs from step (6) by the corresponding species-specific
quarterly average concentrations from the base period (2011) fused surfaces to produce quarterly
average species concentrations for the 2025 baseline.
8. Sum the 2025 baseline quarterly average species concentrations from step (7) over the species to
produce total PM2 5 concentrations for each quarter. Finally, average the total PM2 5
concentrations for the four quarters of the year to produce the spatial field of annual average
PM2 5 concentrations for the 2025 baseline that are input to BenMAP-CE.
EPA repeated the steps above for the baseline in each of the 6 other analysis years as well as for the Option A
scenario in each year.
For generating the spatial fields for each of the two ozone concentration metrics (MDA1 and MDA8), EPA
followed steps similar to those above for PM2 5. Again, we use the 2025 baseline to illustrate the steps for
producing ozone spatial fields for each of the cases we analyzed.
1. Use the EGU May through September (i.e., Ozone Season - OS) NOx for the 2025 baseline126 and
the corresponding modeled 2023 OS NOx emissions to calculate the ratio of 2025 baseline
emissions to modeled 2023 emissions for each EGU tag (i.e., an ozone-season scaling factor for
each tag).
2. The source apportionment modeling provided separate ozone contributions for ozone formed in
VOC-limited chemical regimes (O3V) and ozone formed in NOx-limited chemical regimes
(O3N).127 Multiply the tag-specific 2025 to modeled 2023 EGU NOx emissions-based scaling
ratios from step (1) by the corresponding O3N gridded daily contributions to MDA1 and MDA8
concentrations from the 2023 contribution modeling. This step results in adjusted gridded daily
MDA1 and MDA8 contributions due to NOx changes for each EGUs tag that reflect the
emissions in the 2025 baseline.
3. For MDA1 and MDA8, sum the adjusted contributions for each EGU tag from step (2) to produce
a daily adjusted EGU tag total. Since IPM does not output VOC from EGUs, there are no
predicted changes in VOC emissions in these scenarios so the O3V contributions remain
unchanged. The contributions from the unaltered 2023 O3V tags are added to the summed
adjusted O3N EGU tags.
125 Ammonium concentrations are calculated assuming that the degree of neutralization of sulfate ions remains at 2011 levels (see
Chapter 8 of U.S. EPA [2019i] for details).
126 The 2021, 2023,2025, 2030, 2035, 2040, 2045 and 2045 EGU NOx emissions for the baseline and Option A scenario were
obtained from IPM outputs described in Chapter 5 of the RIA (U.S. EPA, 2020d).
127 Information on the treatment of ozone contributions under NOx-limited and VOC-limited chemical regimes in the CAMx APCA
source apportionment technique can be found in the CAMx v6.40 User's Guide (Ramboll Environ International Corporation,
2016).
11
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
4. Combine the daily total EGU contributions for MDA1 and MDA8 from step (3) with the
contributions to MDA1 and MDA8 from all other sources. This step results in MDA1 and MDA8
concentrations for the baseline 2025 scenario in each model grid cell, nationwide for each day in
the ozone season.
5. For MDA1, average the daily concentrations from step (4) across all the days in the period April
1 through October 31. For MDA8, average the daily concentrations across all days in the period
May 1 through September 30.
6. Divide the seasonal mean concentrations from step (5) by the corresponding seasonal mean
concentrations from the 2011 CAMx model run. This step provides a Relative Response Factor
(i.e., RRF) between 2011 and the 2025 baseline for MDA1 and MDA8 in each model grid cell.
7. Finally, multiply the RRFs for the seasonal mean metrics from step (6) by the corresponding
seasonal mean concentrations from the base period (2011) MDA1 and MDA8 fused surfaces to
produce seasonal mean concentrations for MDA1 and MDA8 for the 2025 baseline that are input
to BenMAP-CE.
As with PM2.5, EPA repeated the steps outlined for ozone for the baseline in each of the 6 other analysis years
as well as for the Option A scenario in each year.
Selected maps showing changes in air quality concentrations between the Option A and the baseline are
provided later in this appendix.
Scaling Ratio Applied to Source Apportionment Tags
Scaling ratios for PM2.5 components that are emitted directly from the source (OA, EC, crustal) were based on
relative changes in annual primary PM2.5 emissions between the 2023 emissions case and the baseline or
Option A. EPA created scaling ratios for components that are formed through chemical reactions in the
atmosphere as follows: scaling ratios for sulfate were based on relative changes in annual SO2 emissions;
scaling ratios for nitrate were based on relative changes annual NOx emissions; and scaling ratios for ozone
formed in NOx-limited regimes128 (""03N") were based on relative changes in ozone season (May-September)
NOx emissions. The scaling ratios that were applied to each tag and year are provided in separate tables by
species, scenario, and EGU fuel-type in Table J-l through Table J-16.
Table J-1: Ozone scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.452
0.45
0.447
0.598
0.623
0.623
0.623
AR
0.896
1.204
1.226
1.206
0.458
0.46
0.461
AZ
1.011
1.073
0.773
0.84
0.865
0.878
0.905
CA
0.064
0.064
0.064
0
0
0
0
CO
1.048
0.771
0.759
0.667
0.667
0.667
0.661
CTRI
NA
NA
NA
NA
NA
NA
NA
DENJ
0
0
0
0
0
0
0
FL
0.555
0.62
0.567
0.618
0.844
0.756
0.756
GA
0.88
0.841
0.807
1.124
1.133
1.236
1.236
IA
1.114
1.165
1.123
1.107
1.096
1.116
1.033
128 The CAMx model internally determines whether the ozone formation regime is NOx-limited or VOC-limited depending on
predicted ratios of indicator chemical species.
12
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-1: Ozone scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.686
0.738
0.7
0.728
0.672
0.728
0.67
IN
0.925
0.864
0.851
0.863
0.825
0.821
0.735
KS
1.092
1.182
1.162
1.162
1.186
1.252
1.186
KY
0.383
0.494
0.346
0.477
0.307
0.484
0.358
LA
0.202
0.494
0.481
0.521
0.555
0.589
0.589
MD
0
0.166
0.129
0.103
0
0.168
0.16
MEMANHVT
0
0
0
0
0
0
0
Ml
0.999
1.033
1.04
1.139
1.059
1.059
1.007
MN
1.204
1.279
1.262
0.86
0.527
0.558
0.494
MO
1.104
1.135
1.093
1.093
1.11
1.126
1.119
MS
0.225
0.294
0.286
0.286
0.29
0.286
0.291
MT
1.066
1.066
1.046
1.046
1.046
1.046
1.046
NC
0.957
1.091
0.936
0.781
0.434
0.488
0.45
NDSD
0.687
0.782
0.754
0.753
0.743
0.775
0.732
NE
1.325
1.137
1.122
1.122
1.122
1.122
1.122
NM
0.847
0.846
0.81
0.828
0.828
0.831
0.831
NV
4.275
6.262
0.907
0
0
0
0
NY
0
0
0
0
0
0
0
OH
1.108
1.19
1.175
1.175
0.964
0.963
0.894
OK
1.586
2.047
2.171
2.043
2.104
2.198
2.108
PA
0.231
0.25
0.247
0.185
0.14
0.14
0.14
SC
1.091
1.169
1.135
1.157
1.098
1.156
1.156
TB
0.546
0.478
0.468
0.472
0.472
0.472
0.469
TN
0.309
0.35
0.344
0.354
0.356
0.29
0.288
TX
0.978
1.026
1.009
1.074
1.132
1.057
1.059
UT
0.968
0.787
0.781
0.781
0.781
0.781
0.772
VA
0
0.062
0
0
0
0
0
Wl
0.512
0.605
0.57
0.559
0.533
0.62
0.488
WV
1.214
1.098
1.06
1.078
0.926
0.911
0.846
WY
1.05
1.08
1.043
0.825
0.806
0.818
0.82
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-2: Ozone scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
1.075
0.867
0.99
0.85
0.89
0.758
0.798
AR
0.676
0.677
0.636
0.615
0.738
0.768
0.757
AZ
0.653
0.641
0.672
0.644
0.696
0.775
0.608
CA
0.699
0.652
0.604
0.191
0.169
0.191
0.071
CO
0.399
0.517
0.56
0.79
0.805
0.805
0.553
CTRI
1.127
1.115
1.091
1.091
1.069
1.094
1.098
DENJ
0.867
1.06
1.136
1.172
1.14
1.051
0.905
FL
0.828
0.822
0.847
0.824
0.843
0.826
0.796
GA
0.711
0.68
0.74
0.717
0.752
0.789
0.822
IA
0.872
0.937
1.058
1.197
1.398
1.343
1.102
IDORWA
0.44
0.44
0.457
0.42
0.44
0.432
0.462
IL
0.672
0.782
0.813
0.93
1.011
0.874
0.796
13
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-2: Ozone scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
IN
0.837
0.916
0.972
0.996
1.188
1.23
1.329
KS
1.1
1.508
1.919
1.034
1.697
1.108
0.83
KY
1.034
1.096
1.221
1.351
1.305
1.079
1.127
LA
0.42
0.406
0.419
0.389
0.272
0.267
0.263
MD
1.066
1.069
1.075
1.114
1.077
1.088
0.988
MEMANHVT
0.6
0.596
0.597
0.58
0.571
0.565
0.566
Ml
1.165
1.106
1.089
1.117
1.168
1.145
1.033
MN
0.6
0.615
0.644
0.601
0.918
0.917
0.671
MO
0.417
0.473
0.558
0.496
0.712
0.611
0.514
MS
0.367
0.32
0.38
0.384
0.332
0.31
0.32
MT
0.027
0.027
0.029
0.037
0.054
0.063
0.063
NC
0.898
0.809
0.961
1.069
0.937
0.694
0.682
NDSD
0.567
0.991
0.73
0.747
0.824
0.78
0.756
NE
0.828
0.921
0.864
0.831
0.914
0.772
0.653
NM
0.629
0.596
0.621
0.45
0.268
0.25
0.077
NV
0.748
0.728
0.721
0.72
0.738
0.955
1.795
NY
0.863
0.865
0.847
0.794
0.705
0.63
0.642
OH
1.26
1.453
1.477
1.693
1.683
1.585
1.64
OK
0.673
0.649
0.694
0.635
0.772
0.961
0.634
PA
0.986
0.976
0.949
0.908
0.919
0.866
0.711
SC
0.708
0.689
0.812
0.844
0.853
0.855
0.83
TB
0.233
0.23
0.247
0.237
0.257
0.285
0.223
TN
0.741
0.845
0.864
0.974
0.791
0.9
0.975
TX
0.924
0.909
0.908
0.82
0.804
0.819
0.423
UT
0.493
0.491
0.493
0.372
0.374
0.424
0.326
VA
0.749
0.831
0.849
1
0.922
0.832
0.759
Wl
0.659
0.733
0.758
0.766
0.835
0.837
0.781
WV
0.194
0.164
0.164
0.25
0.853
0.98
1.209
WY
0.015
0.03
0.015
0.075
0.239
0.239
0.239
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-3: Ozone scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.452
0.45
0.447
0.598
0.623
0.623
0.623
AR
0.902
1.16
1.178
1.198
0.458
0.46
0.461
AZ
1.011
1.073
0.773
0.84
0.865
0.878
0.905
CA
0.064
0.064
0.064
0
0
0
0
CO
1.05
0.771
0.759
0.667
0.667
0.667
0.661
CTRI
NA
NA
NA
NA
NA
NA
NA
DENJ
0
0
0
0
0
0
0
FL
0.568
0.62
0.564
0.641
0.838
0.756
0.756
GA
0.88
0.841
0.817
0.951
1.03
1.063
1.063
IA
1.121
1.174
1.13
1.121
1.101
1.121
1.037
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.624
0.674
0.634
0.66
0.633
0.662
0.633
IN
0.937
0.879
0.864
0.879
0.826
0.821
0.732
KS
1.092
1.19
1.163
1.181
1.186
1.252
1.186
KY
0.395
0.507
0.373
0.471
0.307
0.453
0.358
LA
0.26
0.494
0.481
0.521
0.555
0.589
0.586
14
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-3: Ozone scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
MD
0
0.158
0.129
0.099
0
0.168
0.148
MEMANHVT
0
0
0
0
0
0
0
Ml
0.99
1.028
1.044
1.139
1.059
1.059
1.007
MN
1.208
1.279
1.266
0.86
0.528
0.56
0.494
MO
1.107
1.164
1.095
1.095
1.109
1.126
1.119
MS
0.237
0.294
0.286
0.286
0.286
0.281
0.291
MT
1.066
1.066
1.046
1.046
1.046
1.046
1.046
NC
0.93
1.13
0.931
0.813
0.434
0.494
0.446
NDSD
0.688
0.782
0.753
0.753
0.743
0.775
0.733
NE
1.325
1.137
1.122
1.122
1.122
1.122
1.122
NM
0.847
0.846
0.81
0.828
0.828
0.831
0.831
NV
4.281
5.911
0.907
0
0
0
0
NY
0
0
0
0
0
0
0
OH
1.108
1.19
1.175
1.175
0.966
0.963
0.894
OK
1.584
2.032
2.188
2.04
2.104
2.198
2.108
PA
0.277
0.316
0.294
0.185
0.14
0.14
0.134
SC
1.091
1.145
1.14
1.14
1.103
1.156
1.156
TB
0.546
0.478
0.468
0.472
0.472
0.472
0.469
TN
0.236
0.24
0.235
0.52
0.52
0.585
0.562
TX
0.977
1.026
1.017
1.075
1.132
1.057
1.059
UT
0.968
0.787
0.781
0.781
0.781
0.781
0.772
VA
0
0.076
0
0
0
0
0
Wl
0.51
0.609
0.567
0.557
0.528
0.622
0.487
WV
1.199
1.098
1.078
1.078
0.928
0.915
0.846
WY
1.042
1.081
1.05
0.825
0.806
0.818
0.821
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-4: Ozone scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
1.077
0.864
1.008
0.936
0.967
0.753
0.801
AR
0.676
0.681
0.638
0.615
0.735
0.77
0.75
AZ
0.659
0.64
0.673
0.645
0.696
0.775
0.61
CA
0.699
0.651
0.6
0.191
0.168
0.188
0.071
CO
0.357
0.513
0.584
0.79
0.805
0.81
0.566
CTRI
1.126
1.114
1.09
1.095
1.07
1.096
1.095
DENJ
0.865
1.056
1.124
1.186
1.133
1.05
0.897
FL
0.825
0.822
0.847
0.826
0.844
0.827
0.792
GA
0.725
0.677
0.802
0.787
0.833
0.821
0.841
IA
0.835
0.941
1.084
1.162
1.391
1.343
1.104
IDORWA
0.44
0.44
0.46
0.42
0.44
0.432
0.465
IL
0.676
0.797
0.818
0.923
0.997
0.872
0.802
IN
0.836
0.958
1.007
0.994
1.171
1.249
1.342
KS
1.1
1.536
1.963
1.031
1.694
1.132
0.831
KY
1.051
1.095
1.21
1.329
1.271
1.057
1.109
LA
0.42
0.407
0.423
0.389
0.274
0.271
0.265
MD
1.065
1.069
1.074
1.115
1.076
1.088
0.992
MEMANHVT
0.598
0.594
0.595
0.579
0.57
0.563
0.565
15
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-4: Ozone scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
Ml
1.174
1.092
1.097
1.128
1.166
1.145
1.046
MN
0.602
0.617
0.65
0.601
0.891
0.916
0.638
MO
0.421
0.488
0.549
0.5
0.683
0.595
0.517
MS
0.366
0.325
0.365
0.37
0.333
0.307
0.313
MT
0.027
0.029
0.029
0.037
0.054
0.063
0.063
NC
0.899
0.811
0.976
1.082
0.937
0.698
0.677
NDSD
0.568
0.949
0.721
0.73
0.824
0.785
0.78
NE
0.837
0.92
0.876
0.833
0.926
0.773
0.627
NM
0.629
0.596
0.621
0.435
0.268
0.25
0.077
NV
0.748
0.729
0.72
0.72
0.738
0.954
1.873
NY
0.864
0.863
0.844
0.791
0.704
0.628
0.643
OH
1.261
1.456
1.478
1.701
1.684
1.588
1.605
OK
0.694
0.654
0.697
0.634
0.775
0.952
0.644
PA
0.986
0.976
0.949
0.91
0.92
0.868
0.712
SC
0.708
0.69
0.814
0.844
0.851
0.848
0.824
TB
0.235
0.228
0.247
0.237
0.257
0.285
0.225
TN
0.742
1.009
1.135
0.707
0.733
0.799
0.906
TX
0.921
0.906
0.905
0.82
0.807
0.817
0.428
UT
0.492
0.491
0.492
0.373
0.374
0.426
0.328
VA
0.749
0.833
0.848
1.012
0.921
0.83
0.758
Wl
0.66
0.739
0.756
0.764
0.838
0.836
0.774
WV
0.194
0.164
0.164
0.258
0.836
0.965
1.21
WY
0.015
0.03
0.015
0.075
0.239
0.239
0.239
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-5: Nitrate scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.33
0.327
0.346
0.416
0.355
0.382
0.384
AR
0.782
1.045
1.19
1.062
0.333
0.336
0.34
AZ
1.031
1.039
0.765
0.637
0.609
0.615
0.627
CA
0.075
0.056
0.075
0
0
0
0
CO
1.022
0.752
0.741
0.651
0.622
0.64
0.571
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.367
0.421
0.42
0.391
0.483
0.446
0.439
GA
0.563
0.47
0.481
0.59
0.503
0.537
0.544
IA
1.236
1.273
1.256
1.231
1.126
1.128
1.046
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.662
0.71
0.69
0.71
0.665
0.686
0.609
IN
0.804
0.753
0.741
0.744
0.701
0.679
0.559
KS
1.034
1.119
1.197
1.029
0.894
0.969
0.895
KY
0.274
0.339
0.27
0.326
0.198
0.27
0.22
LA
0.137
0.365
0.371
0.392
0.401
0.396
0.396
MD
0.017
0.117
0.099
0.085
0
0.277
0.218
MEMANHVT
0
0
0
0
0
0
0
Ml
0.947
1.059
1.133
1.178
0.961
0.956
0.9
MN
1.166
1.258
1.274
0.836
0.457
0.486
0.425
MO
1.034
1.153
1.129
1.056
1.003
0.964
0.826
MS
0.181
0.217
0.229
0.232
0.231
0.229
0.232
16
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-5: Nitrate scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
MT
0.951
0.951
0.934
0.934
0.934
0.934
0.934
NC
0.864
0.919
0.856
0.69
0.362
0.387
0.369
NDSD
0.668
0.736
0.716
0.686
0.668
0.685
0.641
NE
1.313
1.128
1.116
1.076
1.042
1.044
0.958
NM
0.764
0.764
0.751
0.646
0.517
0.497
0.487
NV
4.959
7.095
1.354
0.379
0.379
0.379
0.379
NY
0
0
0
0
0
0
0
OH
1.149
1.212
1.194
1.126
0.789
0.696
0.664
OK
1.248
1.702
1.887
1.358
1.348
1.518
1.373
PA
0.2
0.227
0.213
0.139
0.104
0.104
0.104
SC
1.123
1.177
1.168
1.096
0.833
0.864
0.87
TB
0.534
0.466
0.461
0.431
0.374
0.358
0.348
TN
0.265
0.322
0.341
0.308
0.293
0.237
0.232
TX
0.925
1.002
1.056
0.942
0.859
0.837
0.768
UT
0.929
0.755
0.749
0.733
0.661
0.632
0.61
VA
0
0.053
0.025
0
0
0
0
Wl
0.577
0.698
0.691
0.59
0.537
0.592
0.429
WV
1.065
0.951
0.922
0.883
0.598
0.575
0.548
WY
1.054
1.069
1.049
0.813
0.803
0.81
0.813
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-6: Nitrate scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.7224
0.6346
0.6822
0.7176
0.9528
0.8221
0.7229
AR
0.7654
0.7854
0.767
0.8019
0.9865
0.9689
0.9278
AZ
0.7901
0.7752
0.8465
0.8106
0.8798
0.9989
0.77
CA
0.8245
0.7756
0.6872
0.2187
0.2076
0.2176
0.1133
CO
0.2743
0.3995
0.4415
0.603
0.701
0.7478
0.5169
CTRI
1.1378
1.124
1.12
1.0821
1.0644
1.0694
1.0842
DENJ
1.0132
1.237
1.2665
1.3849
1.3657
1.2833
1.11
FL
0.8984
0.8967
0.9006
0.8825
0.9184
0.8994
0.874
GA
0.7836
0.7636
0.7948
0.8473
0.9811
0.938
0.9916
IA
0.7402
0.7961
0.8476
0.9896
1.1419
1.091
0.8687
IDORWA
0.5435
0.5398
0.5472
0.525
0.5335
0.5394
0.5588
IL
0.6181
0.6802
0.6277
0.8198
0.8734
0.7982
0.7158
IN
0.7144
0.771
0.7711
0.8556
1.0969
1.1166
1.2091
KS
0.8034
1.0743
1.3381
0.7854
1.231
0.8555
0.6366
KY
0.9515
1.0506
1.1487
1.4607
1.5971
1.4229
1.4767
LA
0.4008
0.397
0.3835
0.3561
0.3372
0.3078
0.3018
MD
1.1702
1.1933
1.1979
1.2382
1.2598
1.3143
1.2545
MEMANHVT
0.5943
0.591
0.5913
0.5717
0.5658
0.5649
0.5674
Ml
1.1477
1.0598
1.0409
1.1511
1.159
1.1482
1.0307
MN
0.5391
0.5517
0.5458
0.5466
0.7045
0.7024
0.5426
MO
0.3655
0.408
0.4727
0.4538
0.7392
0.713
0.4788
MS
0.3945
0.3668
0.4028
0.4206
0.4611
0.4151
0.4187
MT
0.01
0.01
0.0108
0.0141
0.0204
0.0243
0.0243
NC
0.7542
0.7324
0.792
0.8889
0.8718
0.6818
0.662
17
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-6: Nitrate scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
NDSD
0.3605
0.6096
0.4619
0.4461
0.594
0.57
0.4937
NE
0.7964
0.8836
0.8376
0.7995
0.8713
0.7625
0.6677
NM
0.5007
0.4876
0.4876
0.2455
0.1917
0.1864
0.0901
NV
0.9766
0.9659
0.9386
1.0079
1.0579
1.0764
1.5046
NY
0.9176
0.9231
0.9129
0.8708
0.7774
0.6904
0.694
OH
1.2905
1.4225
1.4123
1.7681
1.7036
1.7142
1.8185
OK
0.5194
0.497
0.5215
0.5286
0.7589
0.8601
0.6078
PA
0.9857
1.1102
1.1479
1.1538
1.2547
1.1082
0.9329
SC
0.6111
0.6119
0.6653
0.7243
0.8134
0.7929
0.7828
TB
0.0961
0.0951
0.103
0.099
0.1079
0.1218
0.0941
TN
0.8409
0.932
0.9587
1.1883
1.1385
1.1608
1.2251
TX
0.8789
0.8367
0.8084
0.7709
0.8254
0.8342
0.4252
UT
0.6888
0.7056
0.7072
0.6535
0.6868
0.7293
0.6833
VA
0.8064
0.8795
0.8669
1.0732
1.0763
0.9804
0.8838
Wl
0.7513
0.7947
0.8048
0.8195
0.8917
0.8716
0.8293
WV
0.1123
0.1126
0.1229
0.2284
0.8747
1.0451
1.2785
WY
0.0097
0.0194
0.0097
0.0485
0.1747
0.1942
0.1942
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-7: Nitrate scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.328
0.327
0.347
0.416
0.355
0.382
0.384
AR
0.775
1.01
1.156
1.06
0.334
0.336
0.339
AZ
1.031
1.039
0.765
0.637
0.609
0.615
0.627
CA
0.075
0.056
0.075
0
0
0
0
CO
1.023
0.752
0.741
0.651
0.622
0.64
0.574
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.373
0.421
0.415
0.402
0.48
0.446
0.437
GA
0.563
0.47
0.488
0.534
0.467
0.481
0.488
IA
1.24
1.313
1.26
1.239
1.129
1.13
1.05
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.622
0.667
0.647
0.654
0.624
0.628
0.567
IN
0.811
0.761
0.749
0.752
0.701
0.678
0.561
KS
1.034
1.134
1.205
1.038
0.897
0.976
0.899
KY
0.282
0.349
0.283
0.319
0.195
0.251
0.21
LA
0.168
0.376
0.372
0.392
0.401
0.396
0.394
MD
0.016
0.114
0.101
0.084
0
0.242
0.213
MEMANHVT
0.138
0.138
0.136
0.136
0.136
0.136
0.136
Ml
0.945
1.078
1.137
1.179
0.961
0.956
0.901
MN
1.168
1.265
1.276
0.836
0.458
0.489
0.425
MO
1.038
1.163
1.123
1.056
1.005
0.964
0.827
MS
0.187
0.218
0.229
0.232
0.229
0.227
0.232
MT
0.951
0.951
0.934
0.934
0.934
0.934
0.934
NC
0.836
0.955
0.863
0.725
0.363
0.39
0.367
NDSD
0.669
0.737
0.715
0.687
0.669
0.685
0.641
NE
1.313
1.13
1.116
1.076
1.042
1.044
0.958
NM
0.764
0.764
0.751
0.646
0.517
0.493
0.487
18
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-7: Nitrate scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
NV
4.968
6.918
1.354
0.379
0.379
0.379
0.379
NY
0
0
0
0
0
0
0
OH
1.145
1.212
1.194
1.133
0.79
0.695
0.664
OK
1.247
1.694
1.894
1.429
1.348
1.518
1.373
PA
0.243
0.288
0.267
0.139
0.104
0.104
0.101
SC
1.123
1.167
1.187
1.088
0.835
0.868
0.87
TB
0.534
0.466
0.461
0.431
0.374
0.357
0.347
TN
0.207
0.258
0.258
0.441
0.392
0.462
0.435
TX
0.924
1.002
1.06
0.942
0.859
0.837
0.77
UT
0.929
0.755
0.749
0.733
0.661
0.632
0.61
VA
0
0.06
0.025
0
0
0
0
Wl
0.578
0.698
0.688
0.581
0.531
0.59
0.422
WV
1.059
0.949
0.929
0.884
0.598
0.576
0.548
WY
1.051
1.069
1.052
0.812
0.803
0.81
0.814
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-8: Nitrate scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.7229
0.6326
0.688
0.7548
0.9913
0.7505
0.7201
AR
0.7707
0.7899
0.7697
0.793
0.9863
0.9613
0.9235
AZ
0.7931
0.7746
0.8488
0.8123
0.8808
0.9998
0.7765
CA
0.8249
0.7752
0.6859
0.2187
0.2074
0.2165
0.1132
CO
0.26
0.3976
0.4516
0.6074
0.7014
0.7585
0.5277
CTRI
1.1373
1.1214
1.1158
1.0835
1.0646
1.0692
1.0823
DENJ
1.01
1.235
1.2612
1.3888
1.3567
1.2741
1.101
FL
0.8976
0.8967
0.8991
0.8858
0.92
0.899
0.8724
GA
0.7925
0.7622
0.8338
0.8801
1.0247
0.9582
1.0017
IA
0.7243
0.7994
0.8803
0.9743
1.1414
1.0963
0.8642
IDORWA
0.5429
0.5399
0.5485
0.5249
0.5335
0.5394
0.5603
IL
0.6219
0.683
0.6265
0.814
0.8624
0.7979
0.7192
IN
0.7122
0.7856
0.7868
0.8562
1.0957
1.1342
1.2296
KS
0.8032
1.0913
1.3659
0.7879
1.2289
0.8709
0.6373
KY
0.9623
1.0544
1.1487
1.4547
1.584
1.4376
1.4714
LA
0.4009
0.3984
0.3877
0.3542
0.3355
0.311
0.3042
MD
1.1695
1.1917
1.1853
1.2386
1.2601
1.3162
1.2558
MEMANHVT
0.593
0.5889
0.5902
0.5714
0.565
0.5636
0.5663
Ml
1.1478
1.0492
1.0452
1.1571
1.1582
1.1477
1.0456
MN
0.54
0.5525
0.5486
0.5467
0.6929
0.7019
0.5282
MO
0.3667
0.4157
0.4649
0.4564
0.7222
0.6816
0.505
MS
0.3949
0.3696
0.3956
0.412
0.4526
0.3927
0.4214
MT
0.01
0.0108
0.0108
0.0141
0.0203
0.0234
0.0243
NC
0.7549
0.7351
0.7929
0.8933
0.8788
0.6797
0.66
NDSD
0.3607
0.5851
0.4569
0.4363
0.594
0.5713
0.5077
NE
0.8042
0.8829
0.8473
0.8006
0.8812
0.7633
0.646
NM
0.5007
0.4876
0.4877
0.2394
0.1917
0.1863
0.09
NV
0.9763
0.9644
0.9559
1.0081
1.0578
1.0762
1.5421
NY
0.9169
0.9199
0.9087
0.8676
0.7749
0.689
0.6939
19
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-8: Nitrate scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
OH
1.2905
1.4176
1.4124
1.7761
1.7048
1.7186
1.8026
OK
0.5331
0.5009
0.5235
0.5261
0.7615
0.8593
0.6141
PA
0.9841
1.1048
1.1474
1.1403
1.2561
1.1171
0.9433
SC
0.6111
0.6131
0.663
0.7204
0.8114
0.79
0.78
TB
0.097
0.0941
0.103
0.099
0.1069
0.1218
0.0951
TN
0.8413
1.0311
1.1143
0.9858
1.0961
1.0665
1.173
TX
0.8761
0.8343
0.8055
0.7701
0.8271
0.8335
0.4296
UT
0.6849
0.705
0.7071
0.6543
0.6865
0.7302
0.6848
VA
0.8058
0.8795
0.8641
1.0783
1.079
0.9742
0.883
Wl
0.7518
0.7978
0.8045
0.8197
0.8931
0.8722
0.8247
WV
0.1123
0.1126
0.1227
0.2334
0.8569
1.0288
1.2799
WY
0.0097
0.0194
0.0097
0.0485
0.1747
0.1942
0.1942
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-9: Sulfate scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.568
0.604
0.61
0.695
0.601
0.616
0.622
AR
0.985
1.424
1.65
1.392
0.137
0.139
0.143
AZ
1.698
1.71
1.469
1.385
1.336
1.342
1.353
CA
0.631
0.471
0.631
0
0
0
0
CO
0.878
0.781
0.77
0.718
0.693
0.708
0.627
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.425
0.548
0.566
0.416
0.638
0.542
0.53
GA
0.788
0.634
0.7
0.644
0.549
0.588
0.596
IA
0.641
0.48
0.469
0.454
0.431
0.434
0.39
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.515
0.538
0.532
0.538
0.514
0.519
0.467
IN
0.88
0.864
0.853
0.836
0.778
0.75
0.605
KS
0.885
0.962
0.997
0.883
0.739
0.822
0.735
KY
0.172
0.194
0.173
0.181
0.14
0.173
0.162
LA
0.199
0.423
0.426
0.443
0.443
0.419
0.419
MD
0.013
0.087
0.074
0.064
0
0.119
0.104
MEMANHVT
0
0
0
0
0
0
0
Ml
0.65
0.569
0.847
0.892
0.665
0.663
0.619
MN
0.955
0.887
1.172
1.158
0.528
0.543
0.514
MO
1.172
1.254
1.229
1.283
1.324
1.213
1.208
MS
0.509
0.61
0.645
0.652
0.651
0.645
0.652
MT
0.525
0.525
0.515
0.599
0.599
0.599
0.599
NC
0.535
0.565
0.549
0.435
0.296
0.347
0.33
NDSD
0.58
0.597
0.581
0.574
0.571
0.579
0.553
NE
1.034
0.978
0.97
0.94
0.922
0.924
0.866
NM
1.213
1.214
1.205
1.096
1.016
0.929
0.923
NV
12.171
17.44
2.322
0.649
0.649
0.649
0.649
NY
0
0
0
0
0
0
0
OH
0.641
0.698
0.695
0.741
0.494
0.422
0.379
OK
0.831
1.177
1.225
0.842
0.791
0.872
0.798
PA
0.08
0.096
0.086
0.061
0.048
0.048
0.048
20
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-9: Sulfate scaling factors for coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
sc
1.478
1.627
1.616
1.462
1.051
1.099
1.106
TB
1.056
1.056
1.05
0.962
0.906
0.836
0.817
TN
0.285
0.347
0.366
0.257
0.224
0.242
0.243
TX
0.793
0.914
0.931
0.8
0.747
0.728
0.622
UT
1.347
1.347
1.357
1.318
1.408
1.388
1.205
VA
0
0.044
0.021
0
0
0
0
Wl
0.515
0.659
0.652
0.512
0.463
0.511
0.396
WV
1.387
0.864
0.838
0.814
0.518
0.509
0.471
WY
0.734
0.784
0.776
0.52
0.511
0.517
0.599
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-10: Sulfate scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.00
0.00
0.00
0.00
0.00
0.00
0.00
AR
0.00
0.00
0.00
0.00
0.00
0.00
0.00
AZ
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CA
0.17
0.17
0.21
0.00
0.00
0.01
0.01
CO
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CTRI
2.00
2.00
2.00
2.00
2.00
2.00
2.00
DENJ
2.50
2.50
2.50
2.50
2.50
2.50
2.50
FL
0.68
0.68
0.68
0.67
0.67
0.67
0.67
GA
0.04
0.09
0.10
0.10
0.28
0.32
0.31
IA
0.00
0.00
0.00
0.00
0.00
0.00
0.00
IDORWA
0.07
0.07
0.07
0.07
0.07
0.07
0.07
IL
0.00
0.00
0.00
0.00
0.00
0.00
0.00
IN
0.20
0.20
0.20
0.20
0.20
0.20
0.20
KS
0.00
0.00
0.00
0.00
0.00
0.00
0.00
KY
0.03
0.03
0.03
0.03
0.02
0.02
0.02
LA
0.06
0.06
0.06
0.06
0.06
0.06
0.06
MD
0.45
0.45
0.45
0.45
0.45
0.45
0.45
MEMANHVT
0.31
0.31
0.31
0.31
0.31
0.31
0.31
Ml
0.26
0.07
0.07
0.06
0.06
0.06
0.06
MN
0.29
0.29
0.29
0.29
0.29
0.29
0.29
MO
0.00
0.00
0.00
0.00
0.00
0.00
0.00
MS
0.00
0.00
0.00
0.00
0.00
0.00
0.00
MT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NC
0.01
0.01
0.01
0.01
0.00
0.00
0.00
NDSD
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NE
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NM
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NV
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NY
1.90
1.90
1.90
0.64
0.64
0.64
0.64
OH
0.08
0.08
0.08
0.08
0.08
0.08
0.08
OK
0.00
0.00
0.00
0.00
0.00
0.00
0.00
PA
0.04
0.04
0.04
0.04
0.04
0.04
0.04
SC
0.01
0.01
0.01
0.01
0.00
0.00
0.00
TB
0.00
0.00
0.00
0.00
0.00
0.00
0.00
21
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-10: Sulfate scaling factors for non-coal EGU tags in the Baseline scenario

2021
2023
2025
2030
2035
2040
2045
TN
0.00
0.00
0.00
0.00
0.00
0.00
0.00
TX
0.04
0.04
0.04
0.04
0.04
0.04
0.04
UT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
VA
0.20
0.21
0.20
0.20
0.20
0.20
0.20
Wl
0.00
0.00
0.00
0.00
0.00
0.00
0.00
WV
0.00
0.00
0.00
0.00
0.00
0.00
0.00
WY
0.00
0.00
0.00
0.00
0.00
0.00
0.00
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-11: Sulfate scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.563
0.604
0.61
0.695
0.601
0.616
0.621
AR
0.965
1.341
1.568
1.388
0.137
0.139
0.142
AZ
1.698
1.71
1.469
1.385
1.336
1.342
1.353
CA
0.631
0.471
0.631
0
0
0
0
CO
0.878
0.781
0.77
0.718
0.693
0.708
0.621
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.432
0.548
0.556
0.459
0.625
0.542
0.528
GA
0.787
0.632
0.709
0.637
0.566
0.583
0.589
IA
0.642
0.499
0.471
0.458
0.432
0.435
0.392
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.504
0.527
0.521
0.523
0.504
0.487
0.456
IN
0.88
0.866
0.855
0.837
0.778
0.749
0.605
KS
0.885
0.967
1.006
0.887
0.74
0.829
0.74
KY
0.175
0.199
0.183
0.178
0.14
0.162
0.153
LA
0.227
0.433
0.427
0.443
0.443
0.419
0.418
MD
0.012
0.085
0.075
0.063
0
0.115
0.1
MEMANHVT
0.404
0.404
0.396
0.396
0.396
0.396
0.396
Ml
0.649
0.581
0.851
0.891
0.665
0.664
0.619
MN
0.894
0.889
1.157
1.158
0.528
0.544
0.514
MO
1.168
1.255
1.227
1.234
1.322
1.214
1.209
MS
0.526
0.615
0.645
0.652
0.645
0.639
0.652
MT
0.525
0.525
0.515
0.599
0.599
0.599
0.599
NC
0.523
0.589
0.554
0.459
0.296
0.35
0.328
NDSD
0.58
0.597
0.581
0.575
0.571
0.579
0.553
NE
1.034
0.978
0.97
0.94
0.922
0.924
0.866
NM
1.213
1.214
1.205
1.096
1.016
0.926
0.923
NV
12.191
16.982
2.322
0.649
0.649
0.649
0.649
NY
0
0
0
0
0
0
0
OH
0.639
0.698
0.694
0.744
0.495
0.422
0.382
OK
0.831
1.177
1.226
0.913
0.791
0.872
0.796
PA
0.094
0.115
0.108
0.061
0.048
0.048
0.047
SC
1.481
1.608
1.638
1.446
1.054
1.104
1.106
TB
1.056
1.056
1.05
0.962
0.906
0.834
0.817
TN
0.286
0.349
0.357
0.391
0.327
0.386
0.372
TX
0.791
0.889
0.923
0.814
0.751
0.711
0.625
UT
1.347
1.347
1.357
1.318
1.407
1.387
1.206
22
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-11: Sulfate scaling factors for coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
VA
0
0.05
0.021
0
0
0
0
Wl
0.516
0.659
0.651
0.496
0.46
0.509
0.391
WV
1.363
0.857
0.844
0.813
0.519
0.51
0.471
WY
0.733
0.784
0.779
0.521
0.511
0.517
0.584
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-12: Sulfate scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.00
0.00
0.00
0.00
0.00
0.00
0.00
AR
0.00
0.00
0.00
0.00
0.00
0.00
0.00
AZ
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CA
0.17
0.17
0.21
0.00
0.00
0.01
0.01
CO
0.00
0.00
0.00
0.00
0.00
0.00
0.00
CTRI
2.00
2.00
2.00
2.00
2.00
2.00
2.00
DENJ
2.50
2.50
2.50
2.50
2.50
2.50
2.50
FL
0.68
0.68
0.68
0.67
0.67
0.67
0.67
GA
0.04
0.09
0.09
0.10
0.28
0.32
0.31
IA
0.00
0.00
0.00
0.00
0.00
0.00
0.00
IDORWA
0.07
0.07
0.07
0.07
0.07
0.07
0.07
IL
0.00
0.00
0.00
0.00
0.00
0.00
0.00
IN
0.20
0.20
0.20
0.20
0.20
0.20
0.20
KS
0.00
0.00
0.00
0.00
0.00
0.00
0.00
KY
0.03
0.03
0.03
0.03
0.02
0.02
0.02
LA
0.06
0.06
0.06
0.06
0.06
0.06
0.06
MD
0.45
0.45
0.45
0.45
0.45
0.45
0.45
MEMANHVT
0.31
0.31
0.31
0.31
0.31
0.31
0.31
Ml
0.26
0.07
0.07
0.06
0.06
0.06
0.06
MN
0.29
0.29
0.29
0.29
0.29
0.29
0.29
MO
0.00
0.00
0.00
0.00
0.00
0.00
0.00
MS
0.00
0.00
0.00
0.00
0.00
0.00
0.00
MT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NC
0.01
0.01
0.01
0.01
0.00
0.00
0.00
NDSD
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NE
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NM
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NV
0.00
0.00
0.00
0.00
0.00
0.00
0.00
NY
1.90
1.90
1.90
0.64
0.64
0.64
0.64
OH
0.08
0.08
0.08
0.08
0.08
0.08
0.08
OK
0.00
0.00
0.00
0.00
0.00
0.00
0.00
PA
0.04
0.04
0.04
0.04
0.04
0.04
0.04
SC
0.01
0.01
0.01
0.01
0.00
0.00
0.00
TB
0.00
0.00
0.00
0.00
0.00
0.00
0.00
TN
0.00
0.00
0.00
0.00
0.00
0.00
0.00
TX
0.04
0.04
0.04
0.04
0.04
0.04
0.04
UT
0.00
0.00
0.00
0.00
0.00
0.00
0.00
VA
0.20
0.21
0.20
0.20
0.20
0.20
0.20
Wl
0.00
0.00
0.00
0.00
0.00
0.00
0.00
23
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-12: Sulfate scaling factors for non-coal EGU tags in the Option A scenario

2021
2023
2025
2030
2035
2040
2045
wv
0.00
0.00
0.00
0.00
0.00
0.00
0.00
WY
0.00
0.00
0.00
0.00
0.00
0.00
0.00
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-13: Primary PM2.5 scaling factors for coal EGU tags in the Baseline
scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.148
0.142
0.15
0.178
0.158
0.168
0.169
AR
0.715
1.184
1.32
0.686
0.286
0.289
0.301
AZ
1.11
1.139
1.04
0.766
0.734
0.735
0.735
CA
0.632
0.472
0.632
0
0
0
0
CO
1.939
1.694
1.694
1.612
1.537
1.575
1.41
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.513
0.675
0.654
0.504
0.732
0.628
0.606
GA
0.361
0.302
0.311
0.35
0.297
0.321
0.325
IA
0.702
0.62
0.62
0.603
0.56
0.562
0.499
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.525
0.555
0.554
0.597
0.558
0.59
0.547
IN
1.905
1.797
1.804
1.827
1.769
1.716
1.535
KS
0.971
1.054
1.104
1.017
0.849
0.924
0.831
KY
1.156
1.314
1.221
1.369
1.127
1.413
1.345
LA
0.11
0.278
0.292
0.314
0.321
0.302
0.302
MD
0.02
0.134
0.115
0.099
0
0.278
0.206
MEMANHVT
0
0
0
0
0
0
0
Ml
7.479
8.428
9.333
9.574
7.742
7.722
7.343
MN
1.279
1.478
1.523
1.141
0.85
0.896
0.7
MO
1.022
1.105
1.098
1.042
1.012
1.017
0.914
MS
0.212
0.253
0.276
0.279
0.279
0.276
0.279
MT
1.059
1.059
1.059
1.059
1.059
1.059
1.059
NC
0.38
0.433
0.336
0.25
0.169
0.174
0.17
NDSD
0.919
0.963
0.969
0.937
0.93
0.95
0.899
NE
0.587
0.526
0.532
0.488
0.462
0.465
0.431
NM
0.453
0.453
0.453
0.331
0.252
0.229
0.217
NV
0.783
1.108
0.651
0.181
0.181
0.181
0.181
NY
0
0
0
0
0
0
0
OH
0.424
0.443
0.442
0.421
0.291
0.268
0.253
OK
1.096
1.519
1.669
1.142
1.119
1.224
1.126
PA
0.296
0.342
0.318
0.202
0.155
0.155
0.155
SC
0.551
0.588
0.575
0.521
0.416
0.442
0.446
TB
0.574
0.574
0.577
0.577
0.577
0.577
0.577
TN
0.207
0.252
0.271
0.237
0.221
0.247
0.242
TX
1.095
1.231
1.355
1.187
1.12
1.154
1.095
UT
0.376
0.376
0.376
0.361
0.301
0.289
0.278
VA
0
0.161
0.075
0
0
0
0
Wl
0.439
0.485
0.488
0.46
0.455
0.461
0.342
WV
0.707
0.611
0.584
0.562
0.333
0.327
0.312
WY
0.469
0.518
0.518
0.448
0.438
0.446
0.449
24
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-13: Primary PM2.5 scaling factors for coal EGU tags in the Baseline
scenario

2021
2023
2025
2030
2035
2040
2045
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-14: Primary PM2.5 scaling factors for non-coal EGU tags in the Baseline
scenario

2021
2023
2025
2030
2035
2040
2045
AL
1.226
1.162
1.193
1.191
1.325
1.345
1.45
AR
0.675
0.719
0.728
0.75
0.832
1.135
1.481
AZ
0.832
0.8
0.92
0.874
0.956
1.059
0.803
CA
1.719
1.595
1.492
0.661
0.64
0.666
0.414
CO
0.918
1.266
1.334
1.681
1.888
1.942
1.578
CTRI
1.794
1.726
1.715
1.484
1.404
1.42
1.451
DENJ
1.238
1.567
1.581
1.78
1.741
1.595
1.362
FL
1.236
1.238
1.247
1.236
1.27
1.318
1.354
GA
1.776
1.744
1.766
1.749
1.839
1.994
2.129
IA
2.096
2.258
2.223
2.548
3.028
2.827
2.034
IDORWA
1.495
1.495
1.512
1.471
1.495
1.551
1.635
IL
0.661
0.723
0.595
0.91
1.008
0.896
0.782
IN
1.274
1.325
1.295
1.359
1.783
2.179
2.419
KS
1.698
2.22
2.57
1.704
2.936
2.472
2.169
KY
1.143
1.208
1.84
2.299
3.035
3.37
4.146
LA
0.67
0.67
0.67
0.673
0.745
0.801
0.819
MD
2.525
2.55
2.55
2.59
2.7
2.8
2.843
MEMANHVT
0.928
0.903
0.892
0.786
0.715
0.702
0.729
Ml
1.443
1.504
1.483
1.581
1.673
1.875
1.814
MN
0.742
0.773
0.757
0.771
1.02
1.019
0.722
MO
0.505
0.567
0.602
0.575
0.932
0.994
0.802
MS
1.054
1.033
1.043
1.05
1.218
1.201
1.245
MT
2.706
2.706
2.707
2.717
2.728
2.732
2.732
NC
1.963
1.935
2.006
2.164
2.033
2.07
2.164
NDSD
0.797
1.022
0.904
0.855
1.208
1.143
0.964
NE
0.75
0.802
0.763
0.697
0.79
0.721
0.696
NM
0.852
0.856
0.857
0.725
0.818
0.825
0.502
NV
2.3
2.268
2.166
2.19
2.325
2.405
2.846
NY
0.845
0.863
0.859
0.809
0.651
0.463
0.47
OH
1.224
1.31
1.265
1.485
1.659
1.75
1.937
OK
0.982
0.992
1.033
1.064
1.364
1.519
1.152
PA
1.529
1.639
1.662
1.698
1.807
1.725
1.852
SC
1.013
1.009
1.037
1.118
1.435
1.55
1.604
TB
0.033
0.116
0.31
0.307
0.32
1.125
0.38
TN
1.814
1.869
1.89
1.95
2.136
2.515
2.992
TX
1.065
1.027
1.018
1.004
1.079
1.137
0.634
UT
1.777
1.827
1.834
1.826
1.896
2.023
1.952
VA
1.636
1.788
1.711
2.128
2.215
2.074
1.91
Wl
0.472
0.488
0.489
0.497
0.517
0.517
0.504
WV
0.032
0.032
0.034
0.065
5.385
7.522
9.484
WY
0.021
0.021
0.021
0.063
0.208
0.229
0.229
25
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-14: Primary PM2.5 scaling factors for non-coal EGU tags in the Baseline
scenario

2021
2023
2025
2030
2035
2040
2045
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
Table J-15: Primary PM2.5 scaling factors for coal EGU tags in the Option A
scenario

2021
2023
2025
2030
2035
2040
2045
AL
0.147
0.142
0.15
0.178
0.158
0.168
0.169
AR
0.674
1.088
1.201
0.684
0.286
0.289
0.299
AZ
1.11
1.139
1.04
0.766
0.734
0.735
0.735
CA
0.632
0.472
0.632
0
0
0
0
CO
1.942
1.694
1.694
1.612
1.537
1.575
1.413
CTRI
0
0
0
0
0
0
0
DENJ
0
0
0
0
0
0
0
FL
0.531
0.675
0.647
0.545
0.72
0.628
0.602
GA
0.361
0.302
0.316
0.345
0.306
0.316
0.32
IA
0.705
0.639
0.623
0.607
0.562
0.563
0.501
IDORWA
NA
NA
NA
NA
NA
NA
NA
IL
0.416
0.435
0.435
0.446
0.442
0.438
0.43
IN
1.91
1.807
1.813
1.836
1.767
1.72
1.539
KS
0.971
1.059
1.11
1.019
0.85
0.937
0.834
KY
1.178
1.344
1.297
1.297
1.123
1.311
1.264
LA
0.137
0.29
0.293
0.314
0.321
0.302
0.302
MD
0.019
0.13
0.116
0.097
0
0.256
0.2
MEMANHVT
0.696
0.696
0.696
0.696
0.696
0.696
0.696
Ml
7.47
8.629
9.374
9.584
7.744
7.723
7.354
MN
1.279
1.491
1.536
1.141
0.85
0.901
0.7
MO
1.023
1.113
1.097
1.042
1.015
1.019
0.915
MS
0.219
0.256
0.276
0.279
0.276
0.274
0.279
MT
1.059
1.059
1.059
1.059
1.059
1.059
1.059
NC
0.355
0.443
0.341
0.266
0.169
0.175
0.17
NDSD
0.92
0.964
0.968
0.937
0.93
0.95
0.898
NE
0.587
0.529
0.532
0.488
0.462
0.465
0.431
NM
0.453
0.453
0.453
0.331
0.252
0.223
0.217
NV
0.784
1.091
0.651
0.181
0.181
0.181
0.181
NY
0
0
0
0
0
0
0
OH
0.421
0.443
0.442
0.423
0.292
0.267
0.253
OK
1.095
1.51
1.671
1.214
1.119
1.224
1.126
PA
0.364
0.437
0.41
0.202
0.155
0.155
0.153
SC
0.551
0.57
0.595
0.509
0.417
0.445
0.447
TB
0.574
0.574
0.577
0.577
0.577
0.577
0.577
TN
0.209
0.261
0.266
0.505
0.451
0.533
0.499
TX
1.095
1.234
1.357
1.187
1.122
1.154
1.097
UT
0.376
0.376
0.376
0.361
0.301
0.288
0.278
VA
0
0.18
0.075
0
0
0
0
Wl
0.439
0.485
0.488
0.459
0.445
0.46
0.342
WV
0.7
0.601
0.587
0.562
0.333
0.328
0.312
WY
0.468
0.518
0.522
0.448
0.438
0.446
0.449
26
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-15: Primary PM2.5 scaling factors for coal EGU tags in the Option A
scenario

2021
2023
2025
2030
2035
2040
2045
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
**NAs are shown where the modeled 2023 emissions were = 0 for any source apportionment tag
Table J-16: Primary PM2.5 scaling factors for non-coal EGU tags in the Option A
scenario

2021
2023
2025
2030
2035
2040
2045
AL
1.226
1.159
1.196
1.206
1.316
1.327
1.438
AR
0.68
0.722
0.733
0.75
0.833
1.116
1.482
AZ
0.84
0.804
0.924
0.88
0.956
1.06
0.817
CA
1.721
1.595
1.487
0.661
0.64
0.648
0.41
CO
0.872
1.265
1.362
1.692
1.89
1.956
1.602
CTRI
1.794
1.722
1.703
1.492
1.412
1.419
1.454
DENJ
1.234
1.56
1.575
1.772
1.725
1.591
1.355
FL
1.235
1.238
1.243
1.236
1.273
1.318
1.354
GA
1.781
1.744
1.772
1.775
1.867
2.024
2.149
IA
2.096
2.275
2.28
2.552
3.038
2.795
2.003
IDORWA
1.493
1.495
1.515
1.47
1.495
1.551
1.64
IL
0.667
0.73
0.606
0.903
0.994
0.89
0.774
IN
1.271
1.329
1.289
1.363
1.801
2.209
2.464
KS
1.698
2.248
2.621
1.709
2.903
2.551
2.149
KY
1.149
1.208
1.864
2.109
3.078
3.346
4.096
LA
0.67
0.67
0.672
0.673
0.744
0.8
0.821
MD
2.524
2.549
2.536
2.592
2.703
2.809
2.841
MEMANHVT
0.923
0.895
0.889
0.781
0.712
0.697
0.722
Ml
1.442
1.499
1.487
1.585
1.673
1.875
1.816
MN
0.744
0.776
0.764
0.768
1.009
1.017
0.719
MO
0.509
0.57
0.6
0.575
0.91
0.985
0.823
MS
1.053
1.034
1.042
1.026
1.197
1.174
1.24
MT
2.706
2.707
2.707
2.717
2.728
2.73
2.732
NC
1.964
1.935
2.01
2.151
2.033
2.056
2.16
NDSD
0.799
1.003
0.92
0.853
1.208
1.143
0.982
NE
0.758
0.802
0.786
0.701
0.794
0.723
0.677
NM
0.852
0.856
0.858
0.72
0.818
0.825
0.501
NV
2.297
2.263
2.179
2.19
2.325
2.406
2.86
NY
0.842
0.858
0.853
0.801
0.645
0.463
0.469
OH
1.225
1.297
1.265
1.482
1.659
1.758
1.939
OK
0.997
0.999
1.034
1.053
1.365
1.494
1.178
PA
1.528
1.637
1.661
1.701
1.815
1.743
1.859
SC
1.012
1.012
1.028
1.103
1.431
1.541
1.599
TB
0.033
0.116
0.31
0.307
0.32
1.125
0.38
TN
1.815
1.914
1.932
1.841
1.991
2.237
2.804
TX
1.064
1.026
1.016
1.004
1.08
1.136
0.641
UT
1.756
1.824
1.834
1.827
1.896
2.026
1.945
VA
1.636
1.776
1.705
2.129
2.216
2.066
1.908
Wl
0.472
0.488
0.488
0.497
0.517
0.519
0.501
WV
0.032
0.032
0.035
0.074
5.245
7.397
9.496
WY
0.021
0.021
0.021
0.063
0.208
0.229
0.229
27
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Table J-16: Primary PM2.5 scaling factors for non-coal EGU tags in the Option A
scenario

2021
2023
2025
2030
2035
2040
2045
*CTRI = CT and Rl; DENJ = DE and NJ; IDORWA = ID, OR, and WA; MEMAVTNH = ME, MA, VT, and NH; NDSD
= ND and SD; TB = tribal lands
J.3 Air Quality Surface Results
Figure J-12 through Figure J-32 present the model-predicted changes in May-Sep MDA8 ozone, Apr-Oct
MDA1 ozone and annual mean PM2.5 concentrations between the baseline and Option A for 2021, 2023,
2025, 2030, 2035, 2040 and 2045 calculated as Option A minus the baseline. The spatial patterns shown in
the figures are a result of (1) of the spatial distribution of EGU sources that are predicted to have changes in
emissions and (2) of the physical or chemical processing that the model simulates in the atmosphere. The
spatial fields used to create these maps serve as an input to the benefits analysis.
Figure J-12: Map of Change in May-Sep MDA8 Ozone (ppb): 2021 Option A - Baseline
May 01, 2021 00:00:00.000 UTC
Min (271, 106) = -0.193, Max (228, 62) = 0.100
Figure J-13: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2021 Option A - Baseline
28
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-14: Map of Change in Annual Mean PM2.5 (jxg/m3): 2021 Option A - Baseline
1 57 113 169 225 281 337 393
January 01, 2021 00:00:00.000 UTC
Min (315, 143) = -0.010, Max (228, 62) = 0,009
- Baseline
Figure J-15: Map of Change in May-Sep MDA8 Ozone (ppb): 2023 Option A
0.210
0.150
0.090
2 0.030
Q-
CL
-0.030
-0.090
-0.150
-0.210
67 133 199 265 331
May 01, 2023 00:00:00.000 UTC
Min (271, 106) = -0.277, Max (321, 106) = 0.198
Figure J-16: Map of Change
241
217
193
169
in Apr-Oct MDA1 Ozone (ppb): 2023 Option
A - Baseline
0.210 •
0.150
0.090
> 0.030
Q.
Q.
-0.030
-0.210
73
49
25
1 T T T T T T
1 67 133 199 265 331
April 01, 2023 00:00:00.000 UTC
Min (271, 106) = -0.304, Max (321, 106) = 0.220
29
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Baseline
0.014
0.010
0.006
0.002
-0.002
-0.006
-0.010
-0.014
Figure J-17: Map of Change in Annual Mean PM2.5 (ng/m3): 2023 Option A -
1 57 113 169 225 281 337 393
January 01, 2023 00:00:00.000 UTC
Min (213, 41) = -0.017, Max (327, 151) = 0.012
221
199
177
Figure J-18: Map
of Change in May-Sep MDA8 Ozone (ppb): 2025
243
Option A - Baseline
i
i ¦
i ¦
i
May 01, 2025 00:00:00.000 UTC
Min (271, 106) = -0.272, Max (327, 151) = 0.096
Figure J-19: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2025 Option A - Baseline
0.210 "
0.150
0.090
0.030
-0.030
-0.090
-0.150
-0.210
April 01, 2025 00:00:00.000 UTC
Min (271, 106) = -0.299, Max (327, 151) = 0.102
30
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-20: Map of Change in Annual Mean PM2.5 (ng/m3): 2025 Option A - Baseline
January 01, 2025 00:00:00.000 UTC
Min (243, 97) = -0.012, Max (327, 151) = 0.015
-0.006
-0.010
-0.014
1
> -
IB
I
Figure J-21: Map of Change
243 "
221 "
199 -
177 "
in May-Sep MDA8 Ozone (ppb): 2030 Option A - Baseline
155 -
133 "
111 "
89 "
> 0.030
CL
CL
-0.030
1
May 01, 2030 00:00:00.000 UTC
Mill (294, S7) = -0.508, Max (271, 106) = 0.413
A - Baseline
0.210 •
0.150
0.090
-0.090
-0.150
-0.210
Figure J-22: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2030 Option
73
49
25
1 T T T T T
1 67 133 199 265 331
April 01, 2030 00:00:00.000 UTC
Min (294, 87) = -0.599, Max (271, 106) = 0.456
31
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-23: Map of Change in Annual Mean PM2.5 (jxg/m3): 2030 Option A - Baseline
1 57 113 169 225 281 337 393
January 01, 2030 00:00:00.000 UTC
Mln (246, 124) = -0.019, Max (271, 106) = 0.034
1
¦§ 0.002 J
A - Baseline
0.210 "
0.150
0.090
> 0.030
-0.030
-0.090
7
_
Figure J-24: Map of Change in May-Sep MDA8 Ozone (ppb): 2035 Option
199 "
177 "
155 "
133 "
111 "
89 -
67 "
45 "
23 "
1 r
1 67 133 199 265 331
May 01, 2035 00:00:00.000 UTC
Min (294, 87) = -0.292, Max (271, 106) = 0.429
Figure J-25: Map of Change
241 •
217 -
193 ¦
169 ¦
145 -
121 -
97 -
73 ¦
49 -
25 -
April 01, 2035 00:00:00.000 UTC
Min (294, 87) = -0.347, Max (271, 106) = 0.468
in Apr-Oct MDA1 Ozone (ppb): 2035 Option A - Baseline
0.210 •
0.150
0.090
> 0.030
-0.030
-0.090
-0.150
-0.210
I
32
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
21 1
1 '
1
January 01, 2035 00:00:00.000 UTC
Min (320, 143) = -0.014, Max (271, 106) = 0.032
Figure J-26: Map of Change in Annual Mean PM2.5 (^g/m3): 2035 Option A - Baseline
241 1
Figure J-27: Map of Change in May-Sep MDA8 Ozone (ppb): 2040
Option A - Baseline
0.210 "
0.150
0.090
2 0.030
Cl
Q.
-0.030
-0.090
-0.150
-0.210
67 133 199 265 331
April 30, 2040 00:00:00.000 UTC
Min (294, 87) = -0.508, Max (271, 106) = 0.763
-0.150
2040
Option A - Baseline
0.210 •
0.090
-0.030
73
49
25
1 i t t t t
1 67 133 199 265 331
March 31, 2040 00:00:00.000 UTC
Min (294, 87) = -0.605, Max (271, 106) = 0.834
Figure J-28: Map of Change in Apr-Oct MDA1 Ozone (ppb):
33
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
I
Figure J-29: Map of Change in Annual Mean PM2.5 (ng/m3): 2040 Option A - Baseline
241
221 -j]
20 i I
181 I
161 |
141 1
121 \
101 "
ni \
61
A1 |
21
1 i-
January 01, 2040 00:00:00.000 UTC
Min (259, 137) = -0.019, Max (271, 106) = 0.037
Figure J-30: Map of Change in May-Sep MDA8 Ozone (ppb): 2045 Option A - Baseline
May 01, 2045 00:00:00.000 UTC
Min (294, 87) = -0.508, Max (271, 106) = 0.721
0.210
0.150
0.090
^ 0.030
CL
CL
-0.030
-0.090
-0.150
-0.210
Option A -
0.210
Baseline
0.150
0,090
0.030
-0.030
-0,090
-0.150
-0.210
Figure J-31: Map of Change in Apr-Oct MDA1 Ozone (ppb): 2045
193
169
145
121
97
73
49
25
1 T T T T T
1 67 133 199 265 331
April 01, 2045 00:00:00.000 UTC
Min (294, 87) = -0.605, Max (271, 106) = 0.789
34
-------
BCAfor Revisions to the Steam Electric Power Generating ELGs
Appendix J: Air Quality Modeling Methodology
Figure J-32: Map of Change in Annual Mean PM2.5 (|ag/m3): 2045 Option A - Baseline
1 57 113 169 225 281 337 393
January 01, 2045 00:00:00.000 utc
Min (282, 122) = -0.010, Max (271, 106) = 0.037
J.4 Uncertainties and Limitations of Air Quality Methodology
One limitation of the scaling methodology for creating PM2 5 surfaces associated with the baseline or
Option A scenarios described above is that it treats air quality changes from the tagged sources as linear and
additive. It therefore does not account for nonlinear atmospheric chemistry and does not account for
interactions between emissions of different pollutants and between emissions from different tagged sources.
This is consistent with how air quality estimations have been treated in past regulatory analyses (U.S. EPA,
2012, 2019i, 2020c). Air quality is calculated in the same manner for the baseline and the Option A scenario,
so any uncertainty associated with these assumptions is carried through both sets of scenarios in the same
manner and is thus not expected to impact the air quality differences between scenarios. In addition,
emissions changes between scenarios are relatively small compared to modeled 2023 totals. Previous studies
have shown that air pollutant concentrations generally respond linearly to small emissions changes of up to
30 percent (D. Cohan &Napelenok, 2011; D. S. Cohan etal., 2005; Dunker et al, 2002; Koo etal., 2007;
Napelenok et al., 2006; Zavala etal., 2009) and that linear scaling from source apportionment can do a
reasonable job of representing impacts of 100 percent of emissions from individual sources (Baker & Kelly,
2014). Therefore, it is reasonable to expect that the differences between the baseline and Option A scenarios
can be adequately represented using this methodology.
A second limitation is that the source apportionment PM2.5 contributions represent the spatial and temporal
distribution of the emissions from each source tag as they occur in the 2023 modeled case. Thus, the
contribution modeling results do not allow EPA to represent any changes to "within tag" spatial distributions.
As a result, the method does not account for any changes of spatial patterns that would result from changes in
the relative magnitude of sources within a source tag in the scenarios investigated here.
Finally, the 2023 modeled concentrations themselves have some uncertainty. While all models have some
level of inherent uncertainty in their formulation and inputs, the base-year 2011 model outputs have been
evaluated elsewhere against ambient measurements (U.S. EPA, 2017b, 2019i) and have been shown to
adequately reproduce spatially and temporally varying ozone and PM2.5 concentrations.
35
-------