United States Office of Water EPA-821-R-15-005
Environmental Protection Washington, DC 20460 September 2015
Agency
&EPA Benefit and Cost Analysis
for the Effluent Limitations
Guidelines and Standards
for the Steam Electric Power
Generating Point Source
Category
-------
v-xEPA
United States
Environmental Protection
Agency
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
September 2015
U.S. Environmental Protection Agency
Office of Water (4303T)
Engineering and Analysis Division
1200 Pennsylvania Avenue, NW
Washington, DC 20460
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Acknowledgments 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 warrant,
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.
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
Table of Contents
1 INTRODUCTION 1-1
1.1 STEAM ELECTRIC POWER PLANTS 1-2
1.2 REGULATORY OPTIONS ANALYZED FOR THE FINAL ELGs 1-2
1.3 ANALYSIS SCENARIOS 1-3
1.4 LOADING AND WITHDRAWAL REDUCTIONS 1-4
1.4.1 Loading Reductions 1-4
1.4.2 Water Withdrawal Reductions 1-4
1.4.3 Loading Reductions Used in Estimating Benefits 1-4
1.5 ANALYTIC FRAMEWORK 1-6
1.5.1 Constant Prices 1-6
1.5.2 Discount Rate and Year 1-7
1.5.3 Period of Analysis 1-7
1.5.4 Population and Income Growth 1-7
1.6 ORGANIZATION OF THE BENEFIT AND COST ANALYSIS REPORT 1-8
2 BENEFITS OVERVIEW 2-1
2.1 HUMAN HEALTH BENEFITS ASSOCIATED WITH IMPROVEMENTS IN SURFACE WATER
QUALITY 2-3
2.1.1 Fish Consumption 2-3
2.1.2 Drinking Water Consumption 2-4
2.1.3 Complementary Measure of Human Health Benefits 2-5
2.2 ECOLOGICAL BENEFITS ASSOCIATED WITH IMPROVEMENTS IN SURFACE WATER QUALITY 2-5
2.2.1 Improved Surface Water Quality 2-5
2.2.2 Benefits to Threatened and Endangered Species 2-7
2.2.3 Reduced Sediment Contamination 2-8
2.3 BENEFITS ASSOCIATED WITH IMPROVEMENTS IN GROUNDWATER QUALITY 2-8
2.4 ECONOMIC PRODUCTIVITY BENEFITS 2-9
2.4.1 Reduced Impoundment Releases 2-9
2.4.2 Enhanced Marketability of Coal Ash for Beneficial Use 2-9
2.4.3 Water Supply and Use 2-10
2.4.4 Reduced Sedimentation in Navigational Waterways 2-11
2.4.5 Commercial Fisheries 2-11
2.4.6 Tourism 2-11
2.4.7 Property Values 2-12
2.5 REDUCED AIR POLLUTION 2-12
2.6 REDUCED WATER WITHDRAWALS 2-13
2.7 SUMMARY OF BENEFITS CATEGORIES 2-13
3 HUMAN HEALTH BENEFITS 3-1
3.1 AFFECTED POPULATION 3-2
3.2 POLLUTANT EXPOSURE FROM FISH CONSUMPTION 3-4
September 29, 2015 T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
3.2.1 Fish Tissue Pollutant Concentrations 3-4
3.2.2 Average Daily Dose 3-4
3.3 BENEFITS TO CHILDREN FROM REDUCED LEAD EXPOSURE 3-6
3.3.1 Methods 3-6
3.3.2 Results 3-9
3.4 BENEFITS TO ADULTS FROM REDUCED LEAD EXPOSURE 3-10
3.4.1 Methods 3-10
3.4.2 Results 3-14
3.5 BENEFITS TO CHILDREN FROM REDUCED MERCURY EXPOSURE 3-14
3.5.1 Methods 3-15
3.5.2 Results 3-16
3.6 REDUCED CANCER CASES FROM ARSENIC EXPOSURE 3-16
3.6.1 Methods 3-16
3.6.2 Results 3-19
3.7 TOTAL MONETIZED HUMAN HEALTH BENEFITS 3-19
3.8 ADDITIONAL MEASURES OF HUMAN HEALTH BENEFITS 3-20
3.9 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 3-20
3.10 LIMITATIONS AND UNCERTAINTIES 3-21
4 NONMARKET BENEFITS FROM WATER QUALITY IMPROVEMENTS 4-1
4.1 WATER QUALITY 4-1
4.1.1 WQI Calculation 4-2
4.1.2 Sources of Data on Ambient Water Quality 4-5
4.1.3 Baseline WQI 4-8
4.1.4 Estimated Changes in Water Quality (AWQI) from the ELG Rule 4-9
4.2 METHODOLOGY FOR ESTIMATING WTP FOR WATER QUALITY IMPROVEMENTS 4-11
4.3 TOTAL WTP FOR WATER QUALITY IMPROVEMENTS 4-17
4.4 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 4-20
4.5 LIMITATIONS AND UNCERTAINTIES 4-22
5 IMPACTS AND BENEFITS TO THREATENED AND ENDANGERED SPECIES 5-1
5.1 INTRODUCTION 5-1
5.2 BASELINE STATUS OF FRESHWATER FISH SPECIES 5-1
5.3 T&E SPECIES AFFECTED BY THE FINAL ELGs 5-2
5.3.1 Identifying T&E Species Potentially Affected by the Final ELGs 5-2
5.3.2 Estimating Benefits of T&E Species Improvements from the Final ELGs 5-4
5.4 ESTIMATING WTP FOR T&E SPECIES POPULATION INCREASES 5-5
5.4.1 Economic Valuation Methods 5-5
5.4.2 Estimating WTP for Improved Protection of T&E Species 5-5
5.5 RESULTS 5-7
5.6 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 5-9
5.7 LIMITATIONS AND UNCERTAINTIES 5-9
6 BENEFITS FROM AVOIDED IMPOUNDMENT FAILURES 6-1
6.1 METHODS AND DATA 6-1
September 29, 2015 iT
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
6.1.1 Baseline Failure Probability and Release Quantity 6-2
6.1.2 Effects of the ELGs 6-4
6.1.3 Costs of a Release 6-6
6.2 RESULTS 6-10
6.3 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 6-13
6.4 LIMITATIONS AND UNCERTAINTIES 6-13
7 AIR-RELATED BENEFITS 7-1
7.1 DATA AND METHODOLOGY 7-2
7.1.1 Changes in Air Emissions 7-2
7.1.2 NOxandSO2 7-4
7.1.3 CO2 7-7
7.1.4 Estimating Total Air-Related Benefits 7-9
7.2 RESULTS 7-10
7.3 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 7-11
7.4 LIMITATIONS AND UNCERTAINTIES 7-11
8 BENEFITS FROM REDUCED WATER WITHDRAWALS 8-1
8.1 GROUNDWATER WITHDRAWALS 8-1
8.1.1 Methods 8-1
8.1.2 Results 8-2
8.1.3 Implications of Revised Steam Electric Plant Loading Estimates 8-2
8.1.4 Limitations and Uncertainties 8-2
9 BENEFITS FROM AVOIDED DREDGING COSTS 9-1
10 BENEFITS FROM ENHANCED MARKETABILITY OF COAL COMBUSTION RESIDUALS 10-1
10.1 METHODS 10-1
10.1.1 Beneficial Use Applications 10-1
10.1.2 Marketable CCR by State 10-2
10.1.3 Value to Society from Re-Using CCR 10-5
10.2 RESULTS 10-8
10.3 IMPLICATIONS OF REVISED STEAM ELECTRIC PLANT LOADING ESTIMATES 10-10
10.4 LIMITATIONS AND UNCERTAINTIES 10-10
11 SUMMARY OF TOTAL MONETIZED BENEFITS 11-1
11.1 TOTAL ANNUALIZED BENEFITS 11-1
11.2 TIME PROFILE OF BENEFITS 11-6
12 SUMMARY OF TOTAL COSTS 12-1
12.1 OVERVIEW OF COSTS ANALYSIS FRAMEWORK 12-1
12.2 KEY FINDINGS FOR REGULATORY OPTIONS 12-2
13 BENEFITS AND COSTS 13-1
13.1 COMPARISON OF BENEFITS AND COSTS BY OPTION 13-1
13.2 ANALYSIS OF INCREMENTAL BENEFITS AND COSTS 13-2
14 ENVIRONMENTAL JUSTICE 14-1
September 29, 2015 Mi
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
14.1 SOCIO-ECONOMIC CHARACTERISTICS OF POPULATIONS RESIDING IN PROXIMITY TO
STEAM ELECTRIC POWER PLANTS 14-1
14.2 DISTRIBUTION OF HUMAN HEALTH IMPACTS AND BENEFITS 14-3
14.3 DISTRIBUTION OF BENEFITS ACROSS BENEFITTING POPULATIONS 14-4
14.4 SUBSISTENCE FISHERS 14-6
14.5 EJ ANALYSIS FINDINGS 14-7
14.6 LIMITATIONS AND UNCERTAINTIES 14-7
15 REFERENCES 15-1
APPENDIX A. CHANGES TO BENEFITS ANALYSIS SINCE PROPOSAL A-l
APPENDIX B. ANALYSIS FOR SCENARIO WITHOUT CPP RULE B-l
B.I HUMAN HEALTH BENEFITS B-l
B.I.I Benefits to Children from Reduced Lead Exposure B-l
B.1.2 Benefits to Adults from Reduced Lead Exposure B-2
B.I.3 Benefits to Children from Reduced Mercury Exposure B-3
B.1.4 Reduced Cancer Cases from Arsenic Exposure B-3
B.I.5 Total Monetized Human Health Benefits B-4
B.I.6 Additional Measures of Human Health Benefits B-4
B.2 NON-MARKET BENEFITS FOR WATER QUALITY IMPROVEMENTS B-5
B.3 IMPACTS AND BENEFITS TO THREATENED AND ENDANGERED SPECIES B-5
B.4 BENEFITS FROM AVOIDED IMPOUNDMENT FAILURES B-7
B.5 AIR-RELATED BENEFITS B-7
B.6 BENEFITS FROM REDUCED WATER WITHDRAWALS B-ll
B.7 BENEFITS FROM ENHANCED MARKETABILITY OF COAL COMBUSTION RESIDUALS B-ll
B.8 TOTAL MONETIZED BENEFITS B-14
B.9 TOTAL COSTS B-19
B.lOMoNETizED BENEFITS AND COSTS COMPARISON B-21
B.10.1 Comparison of Total Monetized Benefits and Costs B-21
B.10.2 Analysis of Incremental Monetized Benefits and Costs B-21
APPENDIX C. ESTIMATION OF EXPOSED POPULATION C-l
APPENDIX D. DERIVATION OF AMBIENT WATER AND FISH TISSUE CONCENTRATIONS IN
RECEIVING AND DOWNSTREAM REACHES D-l
D.I METALS D-l
D.I.I Estimating Water Concentrations in each Reach D-l
D.I.2 Estimating Fish Tissue Concentrations in each Reach D-2
D.2 NUTRIENTS AND SUSPENDED SEDIMENT D-3
APPENDIX E. DETAILS ON MODELING OF CARDIOVASCULAR DISEASE INCIDENCE AND
MORTALITY E-l
E.I BENEFITS TO ADULTS FROM REDUCED LEAD EXPOSURE E-l
E.I.I Hazard Reduction under Final ELGs E-l
E.I.2 Estimating Premature Deaths Avoided Over Multiple Years E-2
APPENDIX F. HUMAN HEALTH BENEFITS SENSITIVITY ANALYSIS F-l
September 29, 2015 iv
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
F.I ALTERNATIVE FISHING DISTANCE ASSUMPTION F-l
F.2 LOG-LINEAR CONCENTRATION RESPONSE FUNCTION FOR IQ IMPACTS TO CHILDREN
FROM LEAD EXPOSURE F-2
F.3 ALTERNATIVE CANCER SLOPE FACTOR AND CASE VALUATION FOR ARSENIC ANALYSIS F-3
APPENDIX G. WQI REGIONAL SUBINDICES 1
APPENDIX H. DEVELOPMENT OF META-REGRESSION MODELS OF WILLINGNESS TO PAY FOR
WATER QUALITY IMPROVEMENTS H-l
H.I LITERATURE REVIEW TO IDENTIFY ADDITIONAL STUDIES H-l
H.2 VARIABLE DEVELOPMENT AND CODING H-5
H.2.1 Study Methodology and Year H-9
H.2.2 Region and Surveyed Populations H-9
H.2.3 Reconciliation of Water Quality Baseline and Change H-10
H.2.4 Delineation of the Affected Resource and Substitutes H-12
H.3 MODEL SPECIFICATION H-15
H.3.1 Marginal Willingness to Pay (Model 1) H-15
H.3.2 Marginal Willingness to Pay (Model 2) H-16
H.4 REGRESSION RESULTS H-17
H.5 LIMITATIONS AND UNCERTAINTY H-20
APPENDIX I. IMPACTS OF STEAM ELECTRIC POLLUTANTS ON AQUATIC SPECIES 1-1
APPENDIX J. SUPPORTING DATA FOR IMPOUNDMENT RELEASE ANALYSIS J-l
APPENDIX K. METHODOLOGY FOR BENEFITS FROM AVOIDED DREDGING COSTS K-l
K.I REVIEW AND ANALYSIS OF HISTORICAL DREDGING DATA K-l
K. 1.1 Dredging Location, Recurrence Interval, and Dredging Volumes K-2
K. 1.2 Determining Affected Navigational Dredging Jobs and Unit Costs K-3
K.I.3 Determining Reservoir Dredging Locations K-4
K.2 SEDIMENT DEPOSITION IN NAVIGABLE WATERWAYS AND RESERVOIRS K-4
K. 3 ESTIMATING DREDGING COSTS UNDER BASELINE AND REGULATORY OPTIONS K-4
K. 3.1 Estimating Annualized Dredging Costs for Navigational Waterbodies K-4
K.3.2 Estimating Annualized Dredging Costs for Reservoirs K-6
K.4 RESULTS K-7
K.4.1 Navigational Dredging Benefits K-7
K.4.2 Reservoir Dredging Benefits K-7
K.5 LIMITATIONS AND UNCERTAINTIES K-8
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
List of Figures
Figure 2-1: Summary of Benefits Resulting from the ELGs 2-2
Figure 3-1: Example changes in PbB through time under baseline, option B, and option D for cohorts in
the 20s and 60s (age as of 2014). Differences between cohorts are driven by differences in
body mass. Example data represent modeled PbBs of male subsistence fishers residing in
CBG 560419753004. The majority of cohorts nationwide experience a smaller reduction in
PbB due to the rule 3-12
Figure 4-1: Comparison between the NHD and E2RF1 Network in a Single Watershed 4-7
Figure 6-1: Baseline Release Probability and Capacity Factor Assumptions 6-6
Figure 10-1: Cost Accounting Framework for the Beneficial Reuse Analysis 10-5
Figure E-l: Illustration of a Hypothetical Policy Effect on a Survival Curve E-3
Figure E-2: A Recursive Illustration of a Policy Effect on Survival Rates E-4
List of Tables
Table 1-1: Steam Electric ELG Regulatory Options 1-3
Table 1-2: Pollutant Removal for Final ELGs Regulatory Options 1-4
Table 1-3: Pollutant Removal for Final ELGs Regulatory Options Used in Estimating Benefits 1-5
Table 1-4: Impacts of Loading Reduction Revisions on Benefit Estimates 1-5
Table 2-1: Estimated Benefits of Reduced Pollutant Discharges from Steam Electric Power Plants 2-13
Table 3-1: Summary of Potentially Affected Population Living within 50 Miles of Affected Reaches
(baseline, as of 2010) 3-4
Table 3-2: Summary of group-specific assumptions for human health benefit analysis3 3-5
Table 3-3: Value of an IQ Point3 (2013$) 3-8
Table 3-4: Estimated Benefits from Avoided IQ Losses for Children Exposed to Lead 3-9
Table 3-5: Estimated Avoided Cost of Compensatory Education for Children with Blood Lead
Concentrations above 20 |o,g/dL and IQ Less than 70 3-10
Table 3-6: Parameters Used to Apply the Leggett+ Model 3-11
Table 3-7: Sample of Inputs for Age- and Sex-Specific Hazard Functions 3-13
Table 3-8: Summary of Estimated Health Benefits due to Decreased Risk of CVD Mortality during
2019-2042 based on the Economic Value of Avoided Premature Mortality (VSL) 3-14
Table 3-9: Estimated Benefits from Avoided IQ Losses for Infants from Mercury Exposure 3-16
Table 3-10: Total Costs of Illness for Skin Cancer3 3-18
Table 3-11: Weighted Average Skin Cancer Cost of Illness3 3-18
Table 3-12: Annual Benefits from Reduced Cancer Cases due to Arsenic Exposure 3-19
Table 3-13: Total Monetized Human Health Benefits for ELG Options (millions of 2013$) 3-19
Table 3-14: Estimated Number of Reaches Exceeding Human Health Criteria for Steam Electric
Pollutants 3-20
Table 3-15: Estimated Aggregate Changes in Pollutant Loadings for Lead, Mercury and Arsenic
(Pounds per Year) 3-21
Table 3-16: Limitations and Uncertainties in the Analysis of Human Health Benefits 3 -22
Table 4-1: Freshwater Water Quality Subindices 4-3
Table 4-2: Freshwater Water Quality Subindex for Heavy Metals 4-4
Table 4-3: Water Quality Classifications 4-5
September 29, 2015 vi
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
Table 4-4: Water Quality Modeling Data used in Calculating the Baseline and Policy Metal, Nutrient
and Sediment Concentrations 4-6
Table 4-5: Water Quality Data used in Calculating the Baseline and Policy WQI 4-8
Table 4-6: Percentage of Potentially Affected Inland Reach Miles by WQI Classification: Baseline
Scenario 4-9
Table 4-7: Water Quality Improvements from Final ELGs in All Benefiting Reaches 4-10
Table 4-8: Independent Variable Assignments for Surface Water Quality Meta-Analysis 4-15
Table 4-9: Household Willingness-to-Pay for Water Quality Improvements 4-18
Table 4-10: Total Willingness-to-Pay for Water Quality Improvements 4-19
Table 4-11: Estimated Aggregate Changes in Pollutant Loadings for Metals, Nutrients, and Suspended
Solids 4-21
Table 4-12: Number of Reaches with Baseline AWQC Exceedances Based on Initial Loadings 4-21
Table 4-13: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits 4-22
Table 5-1: T&E Species with Habitat Occurring within Waterbodies Affected by Steam Electric Power
Plants 5-3
Table 5-2: T&E Species Whose Recovery May Benefit from the Final ELGs 5-4
Table 5-3: Independent Variable Assignments for the T&E Meta-Regression 5-6
Table 5-4: Estimated Annualized Benefits to T&E Species from WQ Improvements (Millions 2013$)a'b 5-8
Table 5 -5: Limitations and Uncertainties in the Analysis of T&E Species Benefits 5-9
Table 6-1: Probabilities of CCR impoundment releases, based on analysis of 49 historical release events
1995-2008 6-3
Table 6-2: Release Probability and Capacity Factor Assumptions for the ELGs 6-4
Table 6-3: Steam Electric Impoundments by Size in ELG Baseline 6-5
Table 6-4: Documented Cleanup Costs from Impoundment Releases 6-7
Table 6-5: NRD Settlements Summary Statistics 6-9
Table 6-6: Studies Summarizing Transaction Costs as a Share of Superfund Spending (for potentially
responsible parties) 6-9
Table 6-7: Unit Costs for Impoundment Releases (2013$) 6-10
Table 6-8: Total Expected Number of Releases in 2019 through 2042, by Failure Type 6-11
Table 6-9: Expected Total Coal Combustion Residuals Volume Released Annually, by Failure Type
(Million Gallons) 6-12
Table 6-10: Estimated Annualized Benefits of Avoided Impoundment Failures by Release Type
(Millions; 2013$)a 6-13
Table 6-11: Limitations and Uncertainties in Analysis of Avoided Risk of Impoundment Failure
Benefits 6-14
Table 7-1: Estimated Changes in Electricity Consumption and Air Pollutant Emissions due to Increase
in Auxiliary Service at Steam Electric Power Plants, Relative to Baseline 7-2
Table 7-2: Estimated Changes in Annual Air Pollutant Emissions due to Increased Trucking at Steam
Electric Power Plants, Relative to Baseline 7-3
Table 7-3: Estimated Changes in Annual Air Pollutant Emissions due to Changes in Electricity
Generation Profile, Relative to Baseline 7-3
Table 7-4: Estimated Net Changes in Air Pollutant Emissions due to Increase in Auxiliary Service at
Steam Electric Power Plants, Increased Trucking at Steam Electric Power Plants, and
Changes in Electricity Generation Profile, Relative to Baseline 7-4
Table 7-5: National Benefits per Ton Estimates for NOx and SO2 Emissions (2013$/ton) from the
Benefits per Ton Analysis Reported by Fann et al. (2012)a'b'° 7-7
Table 7-6: Social Cost of Carbon Values (2013$/metric ton CO2) 7-9
September 29, 2015 viT
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Table of Contents
Table 7-7: Estimated Benefits from Reduced Air Emissions for Selected Years (millions; 2013$)a 7-10
Table 7-8: Estimated Annualized Benefits from Reduced Air Emissions (Millions; 2013$) 7-11
Table 7-9: Limitations and Uncertainties in Analysis of Air-related Benefits 7-12
Table 8-1: Estimated Annualized Benefits from Reduced Groundwater Withdrawals (Millions; 2013$) 8-2
Table 8-2: Limitations and Uncertainties in Analysis of Reduced Groundwater Withdrawals 8-2
Table 10-1: Applicable Beneficial Use Applications of CCRs, by CCR Category 10-1
Table 10-2: State-level Market Approximation (Short Tons) 10-4
Table 10-3: Economic Value of CCR Handling Costs per Unit (2013$) 10-7
Table 10-4: Avoided Resource and Environmental Impacts per Ton of Virgin Material Produced 10-7
Table 10-5: Economic Value of Avoided Resource and Environmental Impacts per Unit of Impact
(2013$) 10-8
Table 10-6: Estimated Beneficial Use Applications of CCRs, by CCR Category 10-8
Table 10-7: Annual Avoided Resource and Environmental Impacts Given CCR Reuse in Concrete and
Fill Applications 10-9
Table 10-8: Annualized Economic Value of Estimated Changes in Beneficial Use (Million 2013$)a>b 10-10
Table 11-1: Summary of Total Annualized Benefits at 3 Percent (Millions; 2013$) 11-2
Table 11-2: Summary of Total Annualized Benefits at 7 Percent (Millions; 2013$) 11-4
Table 11-3: Time Profile of Benefits at 3 Percent (Millions; 2013$) (Including Air-Related Benefits for
Options B and D) 11-7
Table 11-4: Time Profile of Benefits at 7 Percent (Millions; 2013$) (Including Air-Related Benefits for
Options B and D) 11-8
Table 12-1: Summary of Annualized Costs (Millions; $2013) 12-2
Table 12-2: Time Profile of Costs to Society (Millions; $2013) 12-3
Table 13-1: Total Annualized Benefits and Costs by Regulatory Option and Discount Rate (Millions;
2013$) 13-1
Table 13-2: Incremental Net Benefit Analysis (Millions; 2013$) 13-3
Table 14-1: Socio-economic Characteristics of Communities Living in Proximity to Receiving Reaches .. 14-2
Table 14-2: Socio-economic Characteristics of Affected Communities, Compared to State Average 14-3
Table 14-3: Characteristics of Population Potentially Exposed to Steam Electric Pollutants via
Consumption of Self-caught Fish 14-4
Table 14-4: Distribution of Baseline IQ Point Decrements by Pollutant (2021 to 2042) 14-5
Table 14-5: Distribution of Avoided IQ Point Decrements Relative to the Baseline, by Pollutant (2021
to 2042) 14-5
Table 14-6: Distribution of Baseline IQ Point Decrements by Pollutant and Fishing Mode (2021 to
2042) 14-6
Table 14-7: Distribution of Avoided IQ Point Decrements Relative to the Baseline by Fishing Mode,
and Pollutant (2021 to 2042) 14-6
September 29, 2015 viii
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
List of Abbreviations
List of Abbreviations
ADD Average daily dose
ASMFC Atlantic States Marine Fisheries Commission
As Arsenic
AFSC Alaskan Fisheries Science Center
AWQC Ambient water quality criteria
BAT Best available technology economically achievable
BCA Benefit cost analysis
BEA Bureau of Economic Analysis
BenMAP Environmental Benefits Mapping and Analysis Program
BLS Bureau of Labor Services
BOD Biochemical oxygen demand
BPJ Best professional judgment
BPT Best practicable control technology currently available
C&D Construction and development
Cal OEFiF£A California Office of Environmental Health Hazard Assessment
CBG Census Block Group
CBI Confidential Business Information
CCI Construction Cost Index
CCME Canadian Council of Ministers of the Environment
CCR Coal combustion residuals
CMAQ Community Multiscale Air Quality
CO2 Carbon dioxide
COD Chemical oxygen demand
COI Cost of illness
COPD Chronic obstructive pulmonary disease
CPI Consumer Price Index
CSF Cancer slope factor
CVD Cardiovascular disease
DCN Document Control Number
DHHS Department of Health and Human Services
DO Dissolved oxygen
DOJ Department of Justice
EA Environmental Assessment
ECI Employment Cost Index
EGU Electric Generating Unit
EJ Environmental justice
ELGs Effluent limitations guidelines and standards
EPA United States Environmental Protection Agency
ESA Endangered Species Act
FC Fecal coliform
FCA Fish consumption advisories
FGD Flue gas desulfurization
FGMC Flue gas mercury control
FR Federal Register
GDP Gross domestic product
GIS Geographic Information System
September 29, 2015 \x
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
List of Abbreviations
Hg Mercury
HR Hazard ratio
HUC Hydrologic unit code
IEUBK Integrated Exposure, Uptake, and Biokinetics
IPM Integrated Planning Model
IRIS Integrated Risk Information System
IQ Intelligence quotient
LADD Lifetime average daily dose
MCL Maximum contaminant level
MCLG Maximum contaminant level goal
MDNR Minnesota Department of Natural Resources
MEPS Medical Expenditure Panel Survey
NAAQS National Ambient Air Quality Standards
NEFSC Northeast Fisheries Science Center
NERC North American Electric Reliability Corporation
NESHAP National Emissions Standards for Hazardous Air Pollutants
NHD National hydrography dataset
NMFS National Marine Fisheries Service
NOAA National Oceanic and Atmospheric Administration
NOX Nitrogen oxides
NPDES National Pollutant Discharge Elimination System
NRC National Research Council
NRD Natural resource damages
NSPS New Source Performance Standards
NWIS National Water Information System
OAQPS Office of Air Quality Planning and Standards
O&M Operation and maintenance
OMB Office of Management and Budget
OPPT Office of Pollution Prevention and Toxics
ORCR Office of Resource Conservation and Recovery
OSC On-scene coordinator
PbB Blood lead concentration
PIFSC Pacific Islands Fisheries Science Center
ppm parts per million
PSES Pretreatment Standards for Existing Sources
PSNS Pretreatment Standards for New Sources
QA Quality assurance
QC Quality control
RIA Regulatory Impact Analysis
RSEI Risk-Screening Environmental Indicators
SCC Social cost of carbon
Se Selenium
SEFSC Southeast Fisheries Science Center
SO2 Sulfur dioxide
SPARROW SPAtially Referenced Regressions On Watershed attributes
STORET STOrage and RETrieval Data Warehouse
SWFSC Southwest Fisheries Science Center
T&E Threatened and endangered
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
List of Abbreviations
TDD Technical Development Document
TDS Total dissolved solids
THMs Trihalomethanes
TN Total nitrogen
TP Total phosphorus
TRI Toxic Release Inventory
TSD Technical Support Document
TSS Total suspended solids
TVA Tennessee Valley Authority
U.S. FWS United States Fish and Wildlife Service
USGS United States Geological Survey
VSL Value of a statistical life
WQI Water quality index
WQL Water quality ladder
WTP Willingness to pay
September 29, 2015
xi
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 1: Introduction
1 Introduction
The U.S. Environmental Protection Agency (EPA) is promulgating a regulation that would strengthen the
existing controls on discharges from steam electric power plants by revising technology-based effluent
limitations guidelines and standards (ELGs) for the steam electric power generating point source category, 40
CFR part 423.
The analyses supporting the final ELGs for the Steam Electric Power Generating Point Source Category 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 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.
This document presents an analysis of the social benefits and social costs of the final rule and complements
other analyses EPA conducted in support of the ELGs, described in separate documents:
> Environmental Assessment for the Effluent Limitations Guidelines and Standards for the Steam
Electric Power Generating Point Source Category (EA) (U.S. EPA, 2015a; DCN SE04527). The EA
summarizes the environmental and human health improvements that are expected to result from
implementation of the ELGs.
> Technical Development Document for the Effluent Limitations Guidelines and Standards for the
Steam Electric Power Generating Point Source Category (TDD) (U.S. EPA, 2015b; DCN SE05904).
The TDD provides background on the final ELGs; applicability and summary of the ELGs; industry
description; wastewater characterization and identification of pollutants of concern; and treatment
technologies and pollution prevention techniques. It also documents EPA's engineering analyses to
support the final rule including facility specific compliance cost estimates, pollutant loadings, and
non-water quality impact assessment.
> Regulatory Impact Analysis for the Effluent Limitations Guidelines and Standards for the Steam
Electric Power Generating Point Source Category (RIA) (U.S. EPA, 2015c; DCN SE05976). The
RIA describes EPA's analysis of the costs and economic impacts of the final rule. This analysis
provides the basis for social cost estimates presented in 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 final ELGs that are salient to EPA's analysis of the social
benefits and social costs of the rule and summarizes key analytic assumptions used throughout this document.
September 29, 2015 1-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
1: Introduction
team Electric Power P ants
1.1
The final rule establishes new limitations and standards for plants subject to the previously established ELGs
for the Steam Electric Power Generating Point Source Category The ELGs apply to a subset of the electric
power industry, namely those plants with discharges resulting from the operation of a generating unit
"primarily engaged in the generation of electricity for distribution and/or sale, which results primarily from a
process utilizing fossil-type fuels (coal, petroleum coke, oil, gas) or nuclear fuel in conjunction with a thermal
cycle employing the steam water system as the thermodynamic medium."1
Based on data EPA obtained from the 2010 Questionnaire for the Steam Electric Power Generating Effluent
Guidelines (industry survey; U.S. EPA, 2010c) and other sources (see TDD), EPA estimates that there were
1,080 steam electric power plants in 2009.2 EPA projects that a subset of these plants will implement changes
to meet the final limitations (refer to the TDD and RIA for details).
1.2 Regulatory Options Analyzed for the Final ELGs
EPA presents six regulatory options for the final rule (see Table 1-1). These options differ in the wastestreams
controlled by the regulation, the size of the units controlled, and the stringency of controls (see TDD for a
detailed discussion of the options and the associated treatment technology bases). Thus, EPA evaluated
revising or establishing Best Available Technology Economically Achievable (BAT), New Source
Performance Standards (NSPS), Pretreatment Standards for Existing Sources (PSES), and Pretreatment
Standards for New Sources (PSNS) that apply to discharges of up to seven wastestreams: FGD wastewater,
fly ash transport water, bottom ash transport water, combustion residual leachate from landfills and surface
impoundments, wastewater from FGMC systems, wastewater from gasification systems, and nonchemical
metal cleaning wastes.
EPA is establishing limitations and standards for existing sources (BAT/PSES) based on the technologies in
Option D. For new sources, EPA selected the technologies in Option F as the basis for the NSPS and PSNS.
The preamble that accompanies the final rule explains the rationale for EPA's decision.
The final rule contains three minor modifications to the wording of the previously established applicability
provision in the steam electric power generating ELGs to reflect EPA's longstanding interpretation and
implementation of the rule. These revisions do not alter the universe of generating units regulated by the ELGs,
nor do they impose compliance costs on the industry. Instead, they remove potential ambiguity in the
regulations by revising the text to more clearly reflect EPA's longstanding interpretation. See Section VIII of
the preamble for more details.
2 The industry survey EPA conducted in 2010 requested data for several years of operation, up to the most recent
complete calendar year at the time the survey was conducted: 2009.
September 29, 2015 T^2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
1: Introduction
Table 1-1: Steam Electric ELG Regulatory Options
Technology Basis for BAT/NSPS/PSES/PSNS
Regulatory Options
Wastestreams
FGD
Wastewater
Fly Ash Transport
Water
Bottom Ash
Transport Water
FGMC
Wastewater
Gasification
Wastewater
Combustion
Residual Leachate
Nonchemical
Metal Cleaning
Wastes
A
Chemical
Precipitation
Dry Handling
Impoundment
(Equal to BPT)
Dry Handling
Evaporation
Impoundment
(Equal to BPT)
[Reserved]
B
Chemical
Precipitation +
Biological
Treatment
Dry Handling
Impoundment
(Equal to BPT)
Dry Handling
Evaporation
Impoundment
(Equal to BPT)
[Reserved]
C
Chemical
Precipitation +
Biological
Treatment
Dry Handling
Dry handling /
Closed loop (for
units >400
MW);
Impoundment
(Equal to
BPT)(for units
<400 MW)
Dry Handling
Evaporation
Impoundment
(Equal to BPT)
[Reserved]
D
Chemical
Precipitation +
Biological
Treatment
Dry Handling
Dry Handling /
Closed loop
Dry Handling
Evaporation
Impoundment
(Equal to BPT)
[Reserved]
E
Chemical
Precipitation +
Biological
Treatment
Dry Handling
Dry Handling /
Closed loop
Dry Handling
Evaporation
Chemical
Precipitation
[Reserved]
F
Evaporation
Dry handling
Dry handling /
Closed loop
Dry handling
Evaporation
Chemical
Precipitation
[Reserved]
Source: U.S. EPA, 2015
In the remainder of this document, EPA presents the analytical results only for Options A through E for
existing sources. During development of the final rule, EPA decided not to base the final rule on Option F for
existing sources due primarily to the high cost of that option, particularly in light of the costs associated with
other rulemakings expected to impact the steam electric industry (see Section VIII.C. 1 of the preamble). As a
result, EPA chose not to conduct particular analyses for Option F to the same extent that it did for some of the
other options considered.
While EPA calculated the cost impacts of New Sources Performance Standards (NSPS) and Pretreatment
Standards for New Sources (PSNS) for the final ELGs, no new coal steam plants have been announced nor
are projected (see RIA Section 3.2; U.S. EPA 2015c). The lack of any new coal steam plant in the near future
and the site-specific nature of environmental effects and benefits make the assessment of load reductions and
benefits associated with new source requirements hypothetical and speculative. Accordingly, EPA focused the
analysis of the benefits of the final ELGs on the BAT/PSES requirements for existing sources.
1.3 Analysis Scenarios
EPA made every effort to appropriately account for other rules in its many analyses for this rule. Since
proposal, EPA has promulgated several other rules affecting the steam electric industry: the Cooling Water
Intake Structures (CWIS) rule for existing facilities (79 FR 48300), the CCR rule (80 FR 21302), the CPP rule
(FR publication forthcoming). At the time it conducted these analyses, the CPP rule had not yet been
September 29, 2015
1-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
1: Introduction
finalized, and thus EPA used the proposed CPP rule for its analyses as a proxy for the final CPP rule
requirements. In some cases, EPA performed two sets of parallel analyses to demonstrate how the other rules
affected the final ELGs. For example, EPA conducted an assessment of the final ELGs both with and without
accounting for the CPP rule. EPA approached analyses associated with each rulemaking carefully. EPA also
recognizes that the steam electric industry complying with three regulations cumulatively in a very short
period time may choose different compliance path than assumed in the analyses. The cumulative effect
introduces uncertainty on the compliance path, and thereby on the benefits and costs associated with these
rules.
The results presented in the main body of this document are based on this scenario with the CPP rule. The
results of EPA's analyses without accounting for the CPP rule are presented in Appendix B.
1.4 Loading and Withdrawal Reductions
1.4.1 Loading Reductions
EPA expects that final rule will reduce discharge loads of various categories of pollutants including
conventional (such as total suspended solids (TSS), biochemical oxygen demand (BOD), and oil and grease),
priority (such as mercury (Hg), arsenic (As), and selenium (Se)), and non-conventional pollutants (such as
phosphorus (TP), chemical oxygen demand (COD) and total dissolved solids (TDS)). Table 1-2 summarizes
the estimated pollutant reductions under each of the five regulatory options for existing sources.
Table 1-2: Pollutant Removal for Final ELGs Regulatory Options
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Pollutant Load Reduction
(pounds per year)
123,814,202
132,342,281
306,198,515
371,152,958
382,032,630
Source: U.S. EPA Analysis, 2015
1.4.2 Water Withdrawal Reductions
The regulatory options are also expected to eliminate or reduce water withdrawals associated with wet ash
transport and wet FGD scrubbers. EPA estimates that the final BAT/PSES option (Option D) will reduce
surface water withdrawals at steam electric power plants by 57 billion gallons per year (155 million gallons
per day), and reduce withdrawals of 8 million gallons of groundwater per year (21,971 gallons per day).
1.4.3 Loading Reductions Used in Estimating Benefits
EPA revised the estimated steam electric power generating plant discharge loads to incorporate data
submitted via public comments and following additional review conducted after completing the benefit
analyses described in this report.3 The revisions affect baseline loadings as well as loading reductions
3 EPA reevaluated the bottom ash dataset for the final rule, including the addition of new data submitted via
public comments. EPA subjected all data to its revised data editing criteria and as such, removed or replaced
some of the data included in the 1982 TDD. See TDD for details (U.S. EPA, 2015b).
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
1: Introduction
estimated for each of the regulatory options. The load reductions presented in Table 1-2 above reflect these
revisions.
Table 1-3 shows loading reductions calculated from the original loading data used in estimating the benefits
of the regulatory options; these are 0.01 percent to 0.17 percent greater than the reductions shown Table 1-2.
The changes affect most pollutants of concern, including arsenic, cadmium, chromium, lead, mercury, nickel,
total nitrogen, total phosphorus, selenium, total suspended solids, and zinc. See TDD for details (U.S. EPA,
2015b).
Table 1-3: Pollutant Removal for Final ELGs Regulatory Options Used
in Estimating Benefits
Regulatory Option
Pollutant Load Reduction
(pounds per year)
Option A
Option B
Option C
Option D
Option E
124,030,659
132,558,737
306,236,745
371,220,336
382,110,008
Source: U.S. EPA Analysis, 2015
Revisions to the plant loadings may affect estimated benefits in several categories. Implication of the changes
are summarized in Table 1-4, and discussed at greater length in each of the relevant chapters.
Table 1-4: Impacts of Loading Reduction Revisions on Benefit Estimates
Benefit Category
Human health benefits (from fish
consumption)
Nonmarket benefits from water quality
improvements
Benefits to threatened and endangered
species
Benefits from avoided impoundment
failures
Air-related benefits
Benefits from reduced water withdrawals
Benefits from avoided dredging costs
Benefits from enhanced marketability of
coal combustion residuals
Impact of Revised Loading Reduction on Benefit
Estimate
Revisions may reduce benefits, relative to estimates
presented in Chapter 3.
Revisions may reduce benefits, relative to estimates
presented in Chapter 4.
Revisions may reduce benefits, relative to estimates
presented in Chapter 5.
No impact. BCA estimates are unchanged.
No impact. BCA estimates are unchanged.
No impact. BCA estimates are unchanged.
Revisions may reduce benefits, relative to estimates
presented in Chapter 9.
No impact. BCA estimates are unchanged.
Report
Chapter
3
4
5
6
7
8
9
10
Source: U.S. EPA Analysis, 2015
September 29, 2015
1-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 1: Introduction
1.5 Analytic Framework
The analytic framework of this benefit-cost analysis (BCA) includes four basic components used consistently
throughout the analysis of social benefits and social costs4 of the final ELGs:
1. All values are presented in 2013 dollars;
2. Future benefits and costs are discounted using rates of 3 percent and 7 percent back to 2015;
3. Benefits and costs are analyzed over a 24-year period (2019 to 2042); and
4. Future values account for annual U.S. population and income growth.
These components are discussed in the sections below.
EPA's analysis of the final ELGs generally follows the methodology the Agency used previously to analyze
the proposed ELGs (see Benefit and Cost Analysis for the Proposed Effluent Limitations Guidelines and
Standards for the Steam Electric Power Generating Point Source Category (U.S. EPA, 2013a; DCN
SE03172). In analyzing the final ELGs, however, EPA made several important changes relative to the
analysis of the proposed rule:
> EPA used revised inputs that reflect the costs and loads estimated for the final regulatory options (see
TDD and RIA for details).
> EPA updated the universe and characteristics of steam electric power plants to reflect generating units
that have been announced to retire or convert (e.g., to natural gas) during the period of analysis as
well as the anticipated effects of other regulations affecting the power sector and which may change
operations and wastestreams of steam electric power plants.
> EPA updated the baseline scenario to reflect the projected effects of the CCR rule, including
regarding the residual environmental risks from CCR impoundments.
> EPA revised assumptions to use more recent data (e.g., analysis year, compliance period, dollar year
adjustments).
> Finally, EPA made certain changes to the methodologies to address comments EPA received on the
proposed rule (e.g., Environmental Justice analysis), to be consistent with approaches used by the
Agency for other rules, and to incorporate recent advances in the health risk and resource valuation
research.
These changes are described in the relevant sections of this document, and summarized in. Appendix A.
The benefits and social cost analyses presented in this document are generally based on loading reductions
and other inputs generated for individual steam electric power plants (for all plants that were surveyed). These
inputs are used to determine surface waters and other resources affected by steam electric power plant
discharges, estimate changes in pollutant levels, identify populations exposed to steam electric pollutants, etc.
1.5.1 Constant Prices
This BCA applies a year 2013 constant price level to all future annual monetary values of costs and benefits.
Some monetary values of benefits and costs are based on actual past market price data (i.e., prior to 2013),
and in those instances, EPA updated the prices to 2013 by multiplying them by appropriate indexes, or
specific sub-components of these general indexes (index-updated prices). However, not all dollar-monetized
Unless otherwise noted, costs represented in this document are social costs.
September 29, 2015 1-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 1: Introduction
benefits and costs in this BCA are based on actual market prices of goods or services. Several categories of
benefits presented in this report are estimated based on household willingness-to-pay (WTP) surveys, such as
WTP for surface water quality improvements for monetizing ecological benefits of the final ELGs. This BCA
updates these non-market prices as needed using appropriate indexes (e.g., Consumer Price Index (CPI)).
1.5.2 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 will mainly displace or alter the use of capital in
the private sector (U.S. 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 2015.
1.5.3 Period of A nalysis
Benefits are expected to begin accruing when each plant implements the control technologies needed to
comply with any applicable new effluent limits or standards. As discussed in the RIA (in Chapter 3:
Compliance Costs), for the purpose of the economic impact and benefit analysis, EPA assumes that plants
will implement control technologies to meet the final rule limitations and standards as their permits are
renewed over the period of 2019 through 2023. This schedule recognizes that control technology
implementation is likely to be staggered over time across the universe of steam electric power plants.
As discussed in the relevant sections of this document, for several benefit categories where environmental
changes are not provided on a plant-specific basis (e.g., reduced air emissions or changes in surface water
quality attributed to multiple plants), EPA was not able to use plant-specific assumed compliance years in the
benefits analysis but instead used the mid-point of the compliance period (2021) as the assumed starting year
when benefits begin accruing. As presented in the RIA (Table 3-1; U.S. EPA 2015c), over half of the plants
potentially incurring costs for the final ELGs (under Option E) have their technology implementation year in
2021 or earlier.
The period of analysis extends to 2042 to capture the life of the longest-lived compliance technology at any
steam electric power plant (20 years), and the last year of technology implementation (2023).
1.5.4 Population and Income Growth
To account for future population growth or decline, EPA used the population forecasts in Woods & Poole
(2012), which developed county-level forecasts for each year from 2000 through 2040, by age and gender for
non-Hispanic White, African-American, Asian-American, and Native-American and for all Hispanics.5 EPA
aggregated the population forecasts across all ages, genders, races, and ethnicities for the entire U.S. and used
the aggregated growth projections to adjust affected population estimates for future years (i.e., from 2019 to
2040). EPA used a linear extrapolation approach to forecast population values between 2040 and 2042.
Also, since WTP is expected to increase as income increases, EPA took into account income growth for
estimating the value of avoided premature mortality based on the value of a statistical life (VSL) and WTP for
Woods and Poole (2012); the detailed documentation can be found at
http://www.woodsandpoole.com/pdfs/CED12.pdf.
September 29, 2015 1-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 1: Introduction
water quality (WQ) improvements. To develop adjustment factors for VSL, EPA first used income growth
factors in the Environmental Benefits Mapping and Analysis Program (BenMAP) database between 1990 and
2024 to estimate a linear regression model. Using coefficient estimates from the linear regression, EPA
extrapolated the income growth factors for years 2025-2042. EPA applied the projected income data along
with the income elasticity for the respective models (VSL and meta-regression) to adjust the VSL and WQ
meta-analysis estimates of WTP in future years.6
1.6 Organization of the Benefit and Cost Analysis Report
This BCA report presents EPA's analysis of the benefits of the ELGs, assessment of the total costs, and
comparison of the costs and monetized benefits.
The remainder of this report is organized as follows:
> Chapter 2: Benefits Overview provides an overview of the main benefits expected to result from the
implementation of the ELGs.
> Chapter 3: Human Health Benefits details the methods and results of EPA's analysis of the human
health benefits.
> Chapter 4: Nonmarket Benefits from Water Quality Improvements discusses EPA's analysis of the
surface water quality improvements resulting from the ELGs.
> Chapter 5: Impacts and Benefits to Threatened and Endangered Species discusses expected benefits
to threatened and endangered (T&E) species.
> Chapter 6: Benefits from Avoided Impoundment Failures assesses the benefits of reducing the
impacts of any future CCR releases from impoundments used by some steam electric power plants to
manage their CCR waste.
> Chapter 7: Air-Related Benefits describes EPA's analysis of benefits associated with changes in
emissions of air pollutants due to increased electricity consumption, transportation, and changes in the
profile of electricity generation.
> Chapter 8: Benefits from Reduced Water Withdrawals discusses benefits arising from reduced surface
water intake and groundwater use.
> Chapter 9: Benefits from Avoided Dredging Costs describes benefits from reduced maintenance
dredging of navigational channels and reservoirs.
> Chapter 10: Benefits from Enhanced Marketability of Coal Combustion Residuals discusses benefits
arising from the enhanced ability by plants to market dry coal combustion ashes.
> Chapter 11: Summary of Total Monetized Benefits summarizes results across benefit categories.
> Chapter 12: Summary of Total Costs summarizes costs of the ELGs.
> Chapter 13: Benefits and addresses the requirements of Executive Orders that EPA is required to
satisfy for this proposal, notably Executive Order 12866, which requires EPA to compare the benefits
and costs of its actions.
6 These extrapolated income growth factors were originally developed for EPA's COBRA tool
(http://epa.gov/statelocalclimate/resources/cobra.html). The latest public version is 2.613 released in September
2014.
September 29, 2015 T^
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 1: Introduction
> Chapter 14: EnvironmentalJustice details EPA's analysis of the distribution of benefits across
socioeconomic groups to fulfill requirements under Executive Order (E.O.) 12898.
> 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. In particular, Appendix B presents the results of EPA's analysis of the benefits for an alternate scenario
using a baseline that excludes the incremental conversions, retirements, and other changes projected to occur
in response to the CPP rule.
September 29, 2015 1-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
2: Benefits Overview
2 Benefits Overview
This chapter provides an overview of the potential benefits to society resulting from implementation of the
ELGs. EPA expects that benefits will accrue to society in several broad categories, including enhanced
surface water quality, reduced health risks, and increased productivity in economic activities that are
adversely affected by steam electric discharges. These effects follow directly from changes in effluent limits
and standards, which will reduce pollutant loadings to receiving waters. Benefits of the ELGs also include
other effects of the implementation of control technologies or other changes in plant operations, such as
reduction in emissions of air pollutants (e.g., carbon dioxide (CO2), nitrogen oxides (NOX), and sulfur dioxide
(SO2)) which provide benefits in the form of reduced mortality and CO2 impacts on environmental quality and
economic activities; reduction in water use, which provide benefits in the form of increased availability of
surface water and groundwater; and reduction in the use of surface impoundment to manage CCR wastes,
with benefits in the form of avoided cleanup and other costs associated with impoundment releases.
This chapter also provides a brief discussion of the steam electric pollutants, their human health and
ecological effects, and a framework for understanding the benefits expected to be achieved by the steam
electric ELGs. For a more detailed description of steam electric pollutants, their fate, transport, and impacts
on human health and environment, see the Environmental Assessment document (U.S. EPA, 2015a).
Figure 2-1 summarizes the potential effects of the ELGs, the expected environmental changes, and categories
of benefits, as well as EPA's approach to analyzing those benefits. EPA was not able to bring the same depth
of analysis to all categories of benefits because of imperfect understanding of the link between discharge
reductions or other environmental effects of the ELGs and benefit categories, and how society values some of
the benefits. EPA was able to quantify and monetize some benefits, quantify but not monetize other benefits,
and assess still other benefits only qualitatively. The remainder of this chapter provides a qualitative
discussion of the benefit categories applicable to this rule, including human health benefits, ecological
benefits, improved groundwater quality, economic productivity, and reduced air pollution and water
withdrawals. Some benefits estimates presented in this document rely on complex models that embed a
variety of assumptions, limitations and uncertainties discussed in more details in chapters 3 through 10 for the
relevant benefit categories.
September 29, 2015
2-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
2: Benefits Overview
Effect of Final 1 Environmental MM Hi
Jg ^^ Benefit ^y Valuation
Reduced toxic,
bioaccumulative,
and other harmful
pollutants to surface
waters
/
\
Reduced fish tissue
contamination
Improved surface
water quality
/
\
Improved Human Health
• Avoided cardiovascular disease from lead and arsenic exposure
• Avoided cancer cases from arsenic exposure
• Avoided IQ losses in children from mercury and lead exposure
• Reduced cases of other cancer and non-cancer health effects
Improved Ecological Conditions
• Threatened and endangered (T&E) species protection
Reduced sediment deposition in channels and reservoirs
Improved Economic Productivity
• Improved tourism
• Increased commercial fishery yields
• Reduced need for water treatment
• Enhanced property values
• COI
• VSL
• Value of an IQ point
• Count of human health criteria
exceedances (non-monetized)
• WTP for use and non-use values of
surface water quality improvements
• WTP for T&E population increases
Avoided maintenance dredging costs
Qualitative discussion
Reduced reliance on
impoundments to
manage CCR
Change in:
• Auxiliary power
use
• Transportation
• Electricity
generation
Conversion to dry
systems
Reduced water use
^
\
Reduced CCR waste
managed in
impoundments
Reduced air
emissions of COj,
NOx, and SOx
Enhanced ability to
market ash for
beneficial use
Reduced
groundwater
withdrawals
Reduced surface
water withdrawals
Fewer and less consequential impoundment failures
• Reduced 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
• Reduced CO2 impacts
• Avoided disposal costs
• Avoided life-cycle impacts and costs of virgin raw materials
Increased groundwater availability
Reduced vulnerability to drought
Reduced impingement and entrainment mortality
Avoided cleanup costs, natural resource
damage, and transaction costs
• Benefits-per-ton
• Social cost of carbon
• Avoided disposal costs
• Avoided raw material costs and life-
cycle production impacts (various)
Avoided cost of water purchase
Qualitative discussion
COI = Cost of illness; VSL * Value of Statistical Life; WTP = Willingness to Pay; CCR = Coal Combustion Residuals
Figure 2-1: Summary of Benefits Resulting from the ELGs.
September 29, 2015
2-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
2.1 Human Health Benefits Associated with Improvements in Surface Water Quality
Pollutants present in steam electric plant discharges can cause a wide variety of adverse human health effects
arising, for example, from consuming contaminated fish tissue. Toxic bioaccumulative pollutants are of
particular concern because they do not volatilize, do not biodegrade, can be toxic to plants, invertebrates and
fish, adsorb to sediments, and bio-concentrate in fish tissues (U.S. EPA, 2003). More details on the fate,
transport, and exposure risks of steam electric pollutants are provided in the EA (U.S. EPA, 2015a).
Reducing pollutant discharges to the nation's waterways provides human health benefits by several
mechanisms. The most important and readily analyzed benefits stem from reduced risk of illness associated
with the consumption of water, fish, shellfish, and other aquatic organisms that are taken from waterways
affected by steam electric discharges. Human health benefits are typically analyzed by estimating the change
in the expected number of adverse human health events in the exposed population resulting from a reduction
in effluent discharges. While some health effects (e.g., cancer or mortality from cardiovascular disease
(CVD)) 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 ELGs provide human health benefits by reducing exposure to pollutants in water via two principal
exposure pathways discussed below: (1) consumption offish and shellfish taken from waterways affected by
steam electric discharges, and (2) consumption of water from surface waters affected by steam electric power
plant discharges. The ELGs also provide human health benefits by reducing air emissions of pollutants via
changes in the profile of electricity generation; these benefits are discussed separately in Section 2.5.
2.1.1 Fish Consumption
Recreational anglers and subsistence fishers (and their household members) who consume fish caught in the
reaches receiving steam electric power plant discharges are expected to benefit from reduced pollutant
concentrations in fish tissue. EPA analyzed the following five direct measures of change in risk to human
health from exposure to contaminated fish tissue:
1. Neurological effects to children ages 0 to 7 and incidence of cardiovascular disease in adults from
exposure to lead;
2. Neurological effects to infants from in-utero exposure to mercury;
3. Incidence of skin cancer and cardiovascular disease from exposure to arsenic; and
4. Reduced risk of other cancer and non-cancer toxic effects.
EPA was able to monetize only the first three of these four measures. The Agency evaluated lead and mercury
impacts to children in terms of potential intellectual impairment as measured by estimated changes in
intelligence quotient (IQ). Incidence of cardiovascular diseases was translated into an expected level of
avoided early mortality and, on that basis, monetized. Incidence of cancer was translated into an expected
number of avoided cases and monetized based on avoided costs. Chapter 3 of this report provides details on
these analyses.
The fifth effect (reduced risk of other cancer and non-cancer toxic effects from fish consumption) is addressed
indirectly in EPA's assessment of changes in exceedances of ambient water quality criteria (see Section 3.8).
The value of health benefits is the monetary value that society is willing to pay to avoid the adverse health
effects. WTP to avoid morbidity or mortality is generally considered to be a comprehensive measure of the
costs of health care, losses in income, and pain and suffering of affected individuals and their caregivers. For
September 29, 2015 2^
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
example, the value of a statistical life (VSL) (see Section 3.4) is based on estimates of society's WTP to avoid
the risk of premature mortality. Alternatively, the cost-of-illness (COI) approach, which is used to estimate
the value of avoided skin cancer cases (see Section 3.6), is a less comprehensive measure of cost: it allows
valuation of a particular type of non-fatal illness by placing monetary values on metrics, such as lost
productivity and the cost of health care and medications that can be monetized.
Some health benefits of reduced 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 since available economic research provides little empirical data on
society's WTP to avoid IQ losses. Instead, EPA calculated monetary values for avoided 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
benefits from reduced exposure to lead and mercury. Employed alone, these monetized benefits will
underestimate society's WTP, and perhaps significantly so. See Sections 3.3 and 3.5 for applications of this
method to valuing benefits to children and infants from reduced exposure to lead and mercury.
EPA expects that there could also be material health benefits via the fish consumption pathway arising from
reduced discharges of other steam electric pollutants, such as cadmium, selenium, and zinc. Analyses of these
health benefits are not possible due to lack of data on a quantitative relationship between ingestion rate and
potential adverse health effects.
Despite numerous studies conducted by EPA and other researchers, dose-response functions are available
only for a handful of health endpoints associated with steam electric pollutants. In addition, the available
research does not always allow complete economic evaluation, even for quantifiable health effects. For
example, EPA's analysis of health benefits omits the following health effects: low birth weight and neonatal
mortality from in-utero exposure to lead (U.S. EPA, 2009d); additional effects to adults from exposure to lead
(e.g., nervous system disorders, anemia and blood disorders) (U.S. EPA, 2009d; 2013a); effects to adults from
exposure to mercury, including vision defects, hand-eye coordination, hearing loss, tremors, cerebellar
changes, and others (Mergler, et al., 2007; CDC, 2009); and other cancer and non-cancer effects from
exposure to other steam electric pollutants. Therefore, the total monetized human health benefits included in
this analysis represent only a subset of the potential health benefits that are expected to result from the ELGs.
2.1.2 Drinking Water Consumption
Steam electric pollutants discharged to surface waters may affect the quality of water used for public drinking
supplies. However, public drinking water supplies are subject to legally enforceable maximum contaminant
levels (MCLs) established by EPA (U.S. EPA, 2012a). 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.
Pursuant to MCLs, public drinking water supplies are already treated for pollutants that pose human health
risks. Detection of the pollutants is subject to imperfect monitoring and treatment may not remove all
contaminants from the drinking water supplies, as evidenced by reported MCL violations for inorganic and
other contaminants at community water systems (U.S. EPA, 2013d). There may therefore be some
incremental health-related benefits associated with reduced concentrations arising from the final ELGs.
However, EPA's screening level analysis suggests that these benefits would not be substantial. As a first step
in assessing the potential benefits, EPA determined that 27 directly receiving reaches have metal
September 29, 2015 Z4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
concentrations above the MCL in the baseline. None of those reaches, however, has an associated drinking
water intake. Based on the analysis of the larger set of reaches (including downstream reaches) there are 817
reaches with MCL exceedances in the baseline; 250 of these reaches improve under the final ELGs. However,
again, none of these improving reaches has an associated drinking water intake. Accordingly, EPA restricted
the analysis of monetized health benefits from improved surface water quality to benefits arising from the
consumption of contaminated fish tissue.
2.1.3 Complementary Measure of Human Health Benefits
EPA quantified but did not monetize the expected reduction of pollutant concentrations in excess of human
health-based ambient water quality criteria (AWQC) limits. 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
AWQC for steam electric pollutants in the baseline to the number exceeding AWQC under the regulatory
options (Section 3.8).
Because AWQC are set at levels to protect human health through ingestion of water and aquatic organisms,
reducing the frequency at which human health-based AWQC are exceeded should translate into reduced risk
to human health. This measure should be viewed as an indirect indicator of reduced risk to human health
because it does not reflect the size of the exposed population and does not quantify changes in human health
risk per se.
2.2 Ecological Benefits Associated with Improvements 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 used, and waste management techniques used; wastewater
often contains metals such as aluminum, arsenic, boron, cadmium, chromium, copper, iron, lead, manganese,
mercury, nickel, selenium, thallium, vanadium, and zinc (U.S. EPA, 2015a). Discharges of these pollutants to
surface water has 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, 2015a). 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.
EPA expects the ecological benefits from the ELGs to include enhanced habitat 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 reduction in pollutant loadings is expected to reestablish
productive ecosystems in damaged waterways and to protect resident species, including threatened and
endangered species. EPA expects the regulation to enhance the general health offish and invertebrate
populations, increase their propagation to waters currently impaired, and expand fisheries for both
commercial and recreational purposes. Improvements in water quality will also favor recreational activities
such as swimming, boating, fishing, and water skiing. Finally, the Agency expects the regulation to augment
nonuse values (e.g., option, existence, and bequest values) of the affected water resources.
2.2.1 Improved Surface Water Quality
The steam electric ELGs are expected to provide ecological benefits through improvements in the habitats or
ecosystems (aquatic and terrestrial) that are affected by steam electric power plant discharges. Society values
such ecological improvements by a number of mechanisms, including increased frequency and value of use of
September 29, 2015 2-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
the improved habitat for recreational activities. In addition, individuals also value the protection of habitats
and species that are adversely affected by effluent discharges, even when those individuals do not use or
anticipate future use of the affected waterways for recreational or other purposes, resulting in nonuse values.
Recreational activities that may be enhanced by reducing steam electric discharges to surface waters include:
> Recreational Fishing. Degraded water can reduce fish populations by inhibiting reproduction, growth,
and survival of an aquatic species (Friedman et al. 1996; Niimi and Kissoon 1994; U.S. EPA, 2009d;
U.S. EPA, 201 la) resulting in fewer and smaller fish and thereby reducing the value of a fishing trip.
Reducing pollutant loads in steam electric power plant discharges is expected to improve aquatic
habitat and thus increase the number, size, diversity, and health of recreational fish species and, as a
result, the value of recreational fishing. Studies have shown that the value of water resources for
recreational fishing increases with declining level of toxic contamination in fish tissue (Phaneuf et al.,
1998; and Jakus et al., 1997). In addition, improved aesthetic qualities of the waterbody (e.g., from
reduced nutrient loadings) and knowledge that the water is cleaner and does not contain any or
contains fewer pollutants that harm humans and aquatic life, increases individuals' enjoyment of their
recreational experience.
> Outings. Participants in other recreational activities such as hiking, jogging, picnicking, and wildlife
viewing also benefit from improved abundance and diversity of aquatic and terrestrial species. For
example, wildlife viewers benefit from improved abundance of piscivorous birds (e.g., osprey, eagle)
and waterfowl whose populations are likely to increase due to a reduction of mercury and other heavy
metals in the food web and an increase in the forage fish populations (Schoch et al., 2011; U.S. EPA,
201 la). In addition, improved aesthetic quality of surface waters (e.g., clarity and odors) enhances the
recreational experience of wildlife viewers and other recreational users. (Schoch et al., 2011; U.S.
EPA, 2011 a).
> Boating. Boaters benefit from enhanced opportunities for companion activities, such as fishing and
wildlife viewing (e.g., piscivorous birds), and from improved aesthetic quality.
> Swimming. Swimmers benefit from improved aesthetic quality of surface waters including water
clarity and odor thereby enhancing swimmer's aesthetic enjoyment of a waterbody.
> Hunting. Waterfowl hunters benefit from improved aesthetic enjoyment of a water resource, an
increase in the number and quality of game available, and the removal of waterfowl consumption
advisories. Reducing nutrient loadings from steam electric power plants is likely to benefit diving
ducks populations by reducing eutrophication and turbidity in the affected waters and improving their
food sources. Diving ducks rely upon undisturbed and abundant plant and invertebrate sources to
prepare for migration. Excessive nutrient loadings can lead to eutrophic and turbid waters, with few
plants and invertebrates food sources (Minnesota Department of Natural Resources (MDNR), 2010).
Waterfowl populations are adversely affected by consuming contaminated fish or invertebrates; zebra
mussels are an attractive food source for ducks and have been found to have high concentrations of
methyl mercury (MDNR, 2010). High mercury levels have led to duck consumption advisories (Utah
Department of Natural Resources, 2005). Reduction in metal loading to surface waters and of their
presence in the food web may benefit waterfowl reproduction and lead to removal of duck
consumption advisories.
EPA quantified potential ecological impacts from the final ELG options by estimating in-waterway
concentrations of nutrients and other harmful pollutants discharged by steam electric power plants and
translating water quality measurements into a single numerical indicator (water quality index (WQI)). EPA
used the estimated change in WQI as a quantitative measure of ecological benefit for this regulatory analysis.
September 29, 2015 24
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
Section 4.1 of this report provides detail on the parameters used in formulating the WQI and the WQI
methodology and calculations.
A variety of primary methods exist for estimating recreational use values, including both revealed and stated
preference methods (Freeman, 2003). Where appropriate data are available or can be collected, revealed
preference methods can represent a preferred set of methods for estimating use values. These 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 (U.S. EPA, 2010a; U.S. OMB, 2003). Stated preference methods rely on carefully
designed surveys, which either (1) ask people about their WTP for particular ecological improvements, such
as increased protection of aquatic species or habitats with particular attributes, or (2) ask people to choose
between competing hypothetical "packages" of ecological 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). Thus, EPA developed a benefit transfer approach based on a meta-analysis
of surface water valuation studies to evaluate the use and non-use benefits of improved surface water quality
resulting from the final rule. This analysis is presented in Chapter 4. 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" (Smith et al. 2002, p. 134). It involves adapting
research conducted for another purpose to estimate values within a particular policy context (Bergstrom and
De Civita, 1999). In the benefit transfer used for analyzing non-market benefits associated with water quality
improvements, EPA used a regression-based meta-analysis of 140 estimates of total WTP (including both use
and nonuse values) for water quality improvements, provided by 51 original studies conducted between 1981
and 2011.7 The estimated econometric model allows calculation of total WTP for improvements in a variety
of environmental services affected by water quality and valued by humans, including enhanced recreational
fishing, other water-based recreation, and 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.
2.2.2 Benefits to Threatened and Endangered Species
For threatened and endangered (T&E) species vulnerable to future extinction, even minor changes to
reproductive rates and small levels of mortality may represent a substantial portion of annual population
growth. Consequently, steam electric power plant discharges may either lengthen recovery time, or hasten the
demise of these species. By reducing the discharge of steam electric pollutants to aquatic habitats, the ELGs
are expected to enhance 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
7 Although the potential limitations and challenges of benefit transfer are well established (Desvousges et al.,
1998), benefit transfers are a nearly universal component of benefit cost analyses conducted by and for
government agencies. As noted by Smith et al. (2002; p. 134), "nearly all benefit cost analyses rely on benefit
transfers, whether they acknowledge it or not."
September 29, 2015 2~
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
use activities general constitute take, which is illegal unless permitted, the majority of the economic value for
T&E species comes from nonuse values.
Species-specific estimates of nonuse values held for the protection of T&E species can be derived only by
primary research using stated preference techniques. As a second-best alternative,8 EPA used a benefit
transfer approach that relies on information from existing studies (U.S. EPA, 2010a). This benefit transfer
approach is based on a meta-analysis of 31 stated preference studies valuing threatened, rare, or endangered
fish, bird or mammal species (Richardson and Loomis, 2009). EPA used the estimated WTP equation
provided in this meta-analysis to estimate the monetary value of the potential increases in T&E populations
resulting from the ELGs. This analysis and results are presented in Chapter 5. WTP values for improvements
in water quality discussed in the preceding section may inherently include benefits to T&E species. Although
there may be some overlap between WTP estimates for T&E species and the WTP estimates for
improvements in water quality, this overlap is likely to be minimal, however, since none of the studies in
EPA's meta-analysis of WTP for water quality improvements specifically mentioned or otherwise prompted
respondents to include benefits to T&E species populations (see Chapter 4).
2.2.3 Reduced Sediment Contamination
Effluent discharges from steam electric power plants can also contaminate waterbody sediments. For
example, adsorption of arsenic, selenium, and other pollutants found in steam electric power plant discharges
can result in accumulation of contaminated sediment on stream and lake beds (Ruhl, et al., 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 steam
electric 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).
By reducing discharges of pollutants to receiving reaches, the ELGs are expected to reduce the future
contamination of waterbody sediments, thereby mitigating impacts to benthic organisms and reducing the
probability that the pollutants would 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 this benefit.
2.3 Benefits Associated with Improvements in Groundwater Quality
Impoundments used by steam electric power plants to manage their wastewater can leach pollutants into
groundwater aquifers, degrading water quality and potentially creating health hazards to households drawing
drinking water from affected aquifers. The operational changes prompted by the ELGs are expected to result
in plants ceasing or significantly reducing their use of impoundments to manage coal combustion residuals
(CCR). EPA estimated benefits from reducing the risk of groundwater contamination as part of its analysis of
the proposed ELG options (U.S. EPA, 2013a). In December 2014, EPA promulgated the CCR rule which
specifically addresses risks to groundwater quality from leaking impoundments. The CCR rule establishes
technical requirements for CCR surface impoundments, including composite liners, groundwater monitoring,
corrective action, and closure/post closure care, among others. For example, the rule requires any existing
unlined CCR surface impoundment that is contaminating groundwater above a regulated constituent's
groundwater protection standard to stop receiving CCR and either retrofit or close, and that corrective action
be taken to address contamination from leaking clay- or composite-lined impoundments. The final ELGs may
still provide groundwater protection benefits, however, by reducing the potential impacts of future
The cost, administrative burden, and time required to develop primary research estimates to value effects of the
regulation on T&E species were beyond the schedule and resources available for this rulemaking.
September 29, 2015 2-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
impoundment leaks, thereby avoiding future corrective action costs. EPA does not have sufficient data to
model the avoidance of future leaks of clay- or composite-lined impoundments that may no longer receive
CCR as a result of the final ELGs. EPA's analysis of the CCR rule showed significantly lower lifetime risk of
groundwater contamination from lined impoundments than from unlined impoundments (4 and 300 times
smaller risk for clay- and composite-lined impoundments respectively), which suggests that corrective actions
associated with lined impoundments should be infrequent. Accordingly, EPA estimates that the residual risk
to aquifers after implementation of the CCR rule is small and did not estimate incremental benefits for the
ELGs.
2.4 Economic Productivity Benefits
The economic productivity benefits expected to result from the ELGs include reduced impacts of
impoundment releases of CCR and the reduction in the costs associated with the resulting cleanup,
environmental damages, and transaction costs. Conversion to dry handling systems to comply with the ELG
are expected to provide economic benefits by enhancing the ability of steam electric power plants to market
the ash for beneficial use (e.g., in concrete or fill), thereby reducing the disposal costs otherwise incurred by
steam electric power plants and displacing resource intensive virgin materials. Other economic productivity
benefits may stem from reduced contamination of public drinking water supplies and irrigation water;
increased tourism; increased commercial fish harvests; and increased property values.
2.4.1 Reduced Impoundment Releases
Steam electric power plants manage CCR such as fly ash and bottom ash through either wet or dry handling.
For plants that use wet handling, the waste is typically sluiced to one or more surface impoundments (e.g.,
settling ponds), where the solids settle out of the water. Many plants also use surface impoundments to
manage their flue gas desulfurization (FGD) wastewater. In addition to solids associated with the ash and
FGD wastes, these impoundments typically contain water with high concentrations of steam electric
pollutants, including dissolved metals.
The operational changes prompted by the ELGs, such as conversion to dry handling, are expected to cause
some plant owners to reduce their reliance on impoundments to handle CCR. These changes could affect the
volume of CCR released in the event of a failure and/or the future probability of impoundment releases.
Benefits arising from the reduced risk of impoundment releases include avoided cleanup costs, environmental
damage, and transaction costs.
EPA quantified and monetized these benefits based on expected future impoundment release rates, the
volumes of CCR that would be released in an incident, and the costs of cleanup, natural resource damages,
and transaction costs. Chapter 6 describes this analysis.
2.4.2 Enhanced Marketability of Coal Ash for Beneficial Use
EPA anticipates that the final ELGs will prompt certain plants to convert from wet handling of fly ash, bottom
ash, and/or FGD waste to dry handling. This change would in turn allow plants to more readily market the
CCR to beneficial uses. EPA quantified and monetized changes in the marketability of two CCR
wastestreams, and two end uses: (1) fly ash as a substitute for Portland cement in concrete production and (2)
fly- and bottom ashes as substitutes for sand and gravel in fill applications. The changes are based on the
tonnage of fly and bottom ash handled dry instead of wet, with benefits derived from plants avoiding certain
costs associated with disposing of the ashes as waste and society or users of the ash avoiding the cost and life-
cycle effects associated with the displaced virgin material. Chapter 10 describes this analysis.
September 29, 2015 2-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
2.4.3 Water Supply and Use
The ELGs are expected to reduce loading of steam electric pollutants to surface waters and thus enhance uses
of these waters for drinking water supply and agriculture:
> Drinking water treatment costs. The ELGs have the potential to reduce costs of drinking water
treatment (e.g., filtration and chemical treatment) by reducing metal concentrations and
eutrophication 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 screening-level assessment to evaluate the potential
for cost saving to public drinking water systems and concluded that such savings, while they exist,
may not be significant. The assessment involved identifying the pollutants for which treatment costs
may vary depending on source water quality, and using data from EPA's ELG analysis and the
location of drinking water intakes to determine whether modeled water quality improvements have
the potential to reduce drinking water treatment costs. During the first step in the assessment, EPA
determined that water utilities may see reduced costs for the removal of arsenic, cadmium, copper,
lead, mercury, selenium, thallium, nitrogen, and total suspended solids. During the second step in the
assessment, EPA determined that few drinking water systems are currently drawing water at levels
that exceed one or more MCLs and would improve under the policy options. And similarly, few
reaches with elevated total nitrogen (TN) levels will see those levels decline significantly under the
regulatory options to result in substantive cost savings. Accordingly, EPA did not conduct detailed
analysis of cost savings to publicly operated treatment systems.
> Reduction in bromide concentrations. Public drinking water sources do not always effectively remove
bromides (a steam electric pollutant) from raw surface waters. While bromide itself is not thought to
be toxic at levels present in the environment, lab studies and case reports show that bromide found in
source water can react during routine drinking water treatment to generate harmful disinfection
byproducts (DBFs) (Richardson, etal., 2007; U.S. EPA, 2012a). For example, McTigue etal. (2014)
estimate that 96 drinking water treatment plants using surface water are downstream of 57 coal-fired
power plants using wet scrubbers. If existing water treatment is not sufficient, an alternate water
source needs to be substituted or developed, or alternate disinfection technologies need to be adopted
(Watson, et al., 2012). Long-term solutions might require the development of new raw water supplies,
which would involve costs for the acquisition of land (if available), regulatory review and permitting,
development of infrastructure (dams, pumps, pipes), and watershed protection. Thus, increased
bromide levels in raw source water could translate into additional drinking water treatment costs at
some plants, and potentially pose human health risk. In this Final Rule, EPA is not proposing
technology-based effluent limits for bromide for steam electric industry. However, NPDES permit
developers could specify effluent limits for bromide on a plant-by-plant basis based on site-specific
considerations that include the potential to affect public drinking water sources. Benefits of any plant-
specific limits in terms of avoided treatment costs or human health risk would need to be determined
based on more detailed information about the receiving waters and drinking water system.
> Irrigation and other agricultural uses: Reducing steam electric pollutants discharges can improve
agricultural productivity by improving water quality used for irrigation and livestock watering (Clark
et al., 1985). Although elevated nutrient concentrations in irrigation water would not adversely affect
its usefulness for plants, concerns exist for potential residual effects due to steam electric pollutants
entering the food chain. Further, eutrophication promotes cyanobacteria blooms that can kill livestock
September 29, 2015 2-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
and wildlife that drink the contaminated surface water. EPA did not quantify or monetize benefits
from enhanced quality agricultural water sources arising from the ELGs due to data limitations.
> 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 silt layers overtime, at a recorded average rate of
1.2 billion kilograms per reservoir every year (USGS, 2007b). Sedimentation reduces reservoir
capacity (Graf et al., 2010) and the useful life of reservoirs unless measures such as dredging are
taken to reclaim capacity (Clark et al., 1985). EPA expects that by reducing total suspended solids
(TSS) concentrations, the ELGs will provide modest cost savings by reducing dredging activity to
reclaim capacity at existing reservoirs.
2.4.4 Reduced Sedimentation in Navigational Waterways
Navigable waterways, including rivers, lakes, bays, shipping channels and harbors, are an integral part of the
United States' transportation network. Navigable 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). For many
navigable waters, periodic dredging is necessary to remove sediment and keep them passable. Dredging of
navigable waterways can be costly.
EPA expects that the ELGs will reduce sediment loadings to surface waters and reduce dredging of
navigational waterways and reservoirs. EPA quantified and monetized these benefits based on the avoided
cost for expected future dredging volumes. Chapter 9 describes this analysis.
2.4.5 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 offish, 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 closures 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 harvest, which in
turn lead to an increase in producer and consumer surplus.
EPA did not quantify or monetize benefits to commercial fisheries from the ELGs. EPA's EA (see U.S. EPA,
2015a) shows that a small number of steam electric power plants discharge to estuaries or marine waters.
Although benefits to local fish populations and commercial harvest may be positive, the overall benefits to
commercial fisheries arising from the ELGs are likely to be negligible. Most species offish 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
consumers welfare (consumer surplus) is unlikely to change as a result of small changes in fish landings, such
as those EPA expects under the final rule.
2.4.6 Tourism
The ELGs may also benefit local economies by contributing to the tourism industries (e.g., sales of fishing
equipment) in the areas surrounding affected waters due to improved 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 the vicinity of steam electric power plants may lead to a reduction in tourism
September 29, 2015 2^TT
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 2: Benefits Overview
in other locations. Due to these factors EPA believes that benefit from an increase in tourism would be limited
to communities in the vicinity of steam electric power plants; although tourism revenue is potentially
important to these communities, the overall societal benefits are likely to be small. Therefore, EPA did not
quantify or monetize this benefit category.
2.4.7 Property Values
The ELGs are expected to improve the aesthetic quality of land and water resources by reducing pollutant
discharges and thus enhancing water clarity, odor, and color in the receiving and downstream reaches. Several
studies (Boyle et al., 1999; Poor et al., 2001; Leggett and Bockstael, 2000; Bin and Czakowski, 2013; Walsh
et al., 2011; Turtle and Heintzelman, 2014) suggest that waterfront property is more desirable when located
near unpolluted water. Therefore, the value of properties located in proximity to waters contaminated with
steam electric pollutants may increase due to reduced steam electric discharges. Although this benefit would
accrue to the current property owners only, it represents an overall increase in societal wealth.
Due to data limitations, EPA was not able to quantify or monetize the potential increase in property values
associated with the ELGs. The magnitude of the potential increase depends on many factors, including the
number of housing units located in the vicinity of the affected waterbodies, community (e.g., residential
density) and housing stock (e.g., single family or multiple family) and the effects of steam electric pollutants
on aesthetic quality of surface water. Given that the main benefit of the final ELG is reduction in metal
concentrations that do not affect aesthetic quality of surface water, changes in property values are expected to
be small. In addition, there may be overlap between changes in property values and the estimated total WTP
for surface water quality improvements summarized in Section 2.2.1.
2.5 Reduced Air Pollution
The ELGs are expected to affect air pollution through three main mechanisms: 1) additional auxiliary
electricity use by steam electric power plants to operate wastewater treatment, ash handling, and other
systems needed to comply with the new effluent limits and standards; 2) additional transportation-related
emissions due to the increased trucking of CCR 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 for the ELGs. The different 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 at the market
level is the most significant. Small reductions in coal-based electricity generation as a result of the final ELGs
are compensated by increases in generation using other fuels or energy sources - biomass, landfill gas, natural
gas, nuclear power, oil, and wind power. The 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 a net reduction in CO2 and
SO2, and a slight increase in NOX emissions.
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). To estimate
benefits of reducing NOX, and SO2 emissions, EPA used estimates of national monetized benefits per ton of
emissions avoided. CO2is an important greenhouse gas that is linked to climate change effects including
global warming, sea level rise, increased frequency of extreme weather events, ocean acidification, etc. EPA
September 29, 2015 2-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
2: Benefits Overview
used estimates of the social cost of carbon (SCC) obtained from the Interagency Working Group on Social
Cost of Carbon (see Interagency Working Group on Social Cost of Carbon (IWGSCC), 2013b) to derive
benefits per ton for CO2 The SCC reflects a broad range of climate change impacts, including changes in
agricultural productivity, human health risks, property damage from increased flood frequencies, the loss of
ecosystem services, and others. Chapter 7 details this analysis.
2.6 Reduced Water Withdrawals
Steam electric power plants use water for wet ash transport and for operating wet FGD scrubbers. By
eliminating or reducing water used in sluicing operations or prompting the recycling of water in FGD
wastewater treatment systems, the ELGs are expected to reduce demand on aquifers by plants that rely on
groundwater sources.
Additionally, reduced surface water intake would reduce impingement and entrainment mortality. Due to data
limitations, EPA did not quantify and monetize these benefits as part of this analysis. For more details on the
impacts of surface water withdrawals, see paper titled "Water Withdrawals in Water Stressed Areas: Impacts
of the Steam Electric Effluent Limitations Guidelines" (DCN SE05943) in the record for the final steam
electric ELG rule.
Reduced water use from groundwater sources by steam electric power plants would result in greater
availability of groundwater supplies for alternative uses. EPA used the state specific prices for bulk drinking
water supplies to value the increased quantity of groundwater. This analysis is presented in Chapter 8.
Benefits Categories
Table 2-1 summarizes the benefits of the ELGs and the level of analysis applied to each category. As
indicated in the table, only a subset of anticipated benefits can be quantified and monetized (in which case the
table identifies the section of the report that discusses the analysis). The monetized benefits include reductions
in some human health risks, use and non-use values from improved surface water quality, benefits to T&E
species, reduced impacts from impoundment releases, increase in the amount of ash marketed for beneficial
uses, reduced costs for dredging navigational waterways, reduced air pollution, and reduced water
withdrawals. Other benefit categories, including expected reduction of pollutant concentrations in excess of
human health-based AWQC limits, can be quantified but not monetized. Finally, EPA was not able to
quantify or monetize other benefits, including drinking water treatment costs and benefits to commercial
fisheries; EPA evaluated these benefits qualitatively as discussed above in Sections 2.1 through 2.6.
Table 2-1: Estimated Benefits of Reduced Pollutant Discharges from Steam Electric Power Plants
Category
Effect of ELGs
Benefits Analysis
Quantified
Monetized
Methods (Report
Chapter or Section
where Analysis is
Detailed)
Human Health Benefits from Surface Water Quality Improvements
Reduced IQ losses to
children ages 0 to 7
Reduced need for
specialized education
Reduced incidence of
cardiovascular disease
Reduced childhood exposure to lead
from fish consumption
Reduced childhood exposure to lead
from fish consumption
Reduced exposure to lead from fish
consumption
•/
•/
•/
v'
v'
v'
IQ point valuation
(Section 3.3)
Avoided cost
(Section 3.3)
VSL (Section 3.4)
September 29, 2015
2-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
2: Benefits Overview
Table 2-1: Estimated Benefits of Reduced Pollutant Discharges from Steam Electric Power Plants
Category
Reduced IQ losses to
infants
Reduced incidence of
cancer
Reduced other adverse
health effects (cancer
and non-cancer)
Reduced adverse health
effects
Effect of ELGs
Reduced in-utero mercury exposure
from maternal fish consumption
Reduced exposure to arsenic from
fish consumption
Reduced exposure to other
pollutants (arsenic, lead, etc.) via
fish consumption
Reduced exposure to pollutants from
recreational water uses
Benefits Analysis
Quantified
•/
•/
v'
Monetized
^
^
Methods (Report
Chapter or Section
where Analysis is
Detailed)
IQ point valuation
(Section 3.5)
COI (Section 3. 6)
Human health
criteria exceedances
(Section 3.8)
Qualitative
discussion
Ecological Conditions and Recreational use Benefits from Surface Water Quality Improvements
Improved aquatic and
wildlife habitat3
Water-based recreation3
Aesthetics3
Non-use values3
Aquatic and wildlife
Improved protection of
T&E species
Reduced sediment
contamination
Improved ambient water quality in
receiving reaches
Enhanced swimming, fishing,
boating, and near-water activities
from improved water quality
Increased aesthetics from improved
water clarity, color, odor, including
nearby site amenities (residing,
working, traveling)
Enhanced existence, option, and
bequest values from improved
ecosystem health
Reduced risks to aquatic life from
exposure to steam electric pollutants
Improved T&E habitat and thus
potential increase in T&E population
Reduced deposition of toxic
pollutants to sediment
•/
v'
v'
v'
Benefit transfer
(Chapter 4)
Benefit transfer
(Chapter 5)
Qualitative
discussion
Groundwater Quality Benefits
Groundwater quality
Reduced groundwater contamination
Qualitative
discussion
Market and Productivity Benefits
Impoundment releases
Reduced dredging costs
Reduced risk of impoundment
releases due to changes in the use of
impoundment
Reduced costs for maintaining
navigational waterways and
reservoir capacity
v'
•/
v'
•/
Avoided cost of
cleanup, natural
resource damages,
and transaction
costs (Chapter 6)
Avoided dredging
costs (Chapter 9)
September 29, 2015
2-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
2: Benefits Overview
Table 2-1: Estimated Benefits of Reduced Pollutant Discharges from Steam Electric Power Plants
Category
Beneficial use of ash
Reduced water treatment
costs for drinking water
and irrigation water
Commercial fisheries
Benefits to tourism
industries
Property values
Effect of ELGs
Reduced disposal costs and avoided
life-cycle impacts from displaced
virgin material
Improved quality of source water
used for drinking and irrigation
Improved fisheries yield and harvest
quality due to aquatic habitat
improvement
Increased participation in water-
based recreation
Increased property values from
water quality improvements
Benefits Analysis
Quantified
•/
Monetized
•/
Methods (Report
Chapter or Section
where Analysis is
Detailed)
Avoided disposal
cost and avoided
resource use and
environmental
impacts (Chapter
10)
Qualitative
discussion
Qualitative
discussion
Qualitative
discussion
Qualitative
discussion
Air-Related Benefits
Reduced air emissions of
NOX, SO2
Reduced air emissions of
C02
Reduced mortality and morbidity
from exposure to NOX, SO2 and
paniculate matter (PM2 5)
Avoided climate change /global
warming impacts
•/
•/
•/
•/
Benefit per ton of
air pollutant
removed (Chapter
7)
Social cost of
carbon (SCC)
(Chapter 7)
Reduced Water Withdrawal Benefits
Reduced groundwater
withdrawals
Reduced surface water
withdrawals
Increased availability of
groundwater resources
Reduced vulnerability to drought
and reduced impingement and
entrainment mortality
v'
v'
Avoided cost
(Chapter 8)
Qualitative
discussion
a. These values are implicit in the total WTP for water quality improvements.
September 29, 2015
2-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
3 Human Health Benefits
EPA expects that the ELGs will yield a range of human health benefits by reducing effluent discharges to
surface waters and, as a result, ambient pollutant concentrations in the receiving reaches. EPA's EA (U.S.
EPA, 2015a) provides details on the health effects of steam electric pollutants. Recreational anglers and
subsistence fishers (and their household members) who consume fish caught in the reaches receiving steam
electric discharges are expected to benefit from reduced pollutant concentrations in fish tissue. This chapter
presents EPA's analysis of human health benefits from reduced exposure to steam electric pollutants via the
fish consumption pathway.9 The analyzed health benefits include:
> Reduced exposure to lead:
- Avoided 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
- Reduced incidence of CVD in adults
> Reduced exposure to mercury:
- Reduced neurological and cognitive damages in infants from exposure to mercury in-utero
> Reduced exposure to arsenic:
- Reduced incidence of cancer cases.
The total quantified human health benefits included in this analysis represent only a subset of the potential
health benefits expected to result from the final ELGs. While additional adverse health effects are also
associated with steam electric pollutants (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 relationships10 between ingestion rates and these effects precluded EPA from quantifying the
associated benefits.
EPA's analysis of the monetary value of human health benefits is based on data and methodologies described
in the EA (U.S. EPA, 2015a). The relevant data include COMIDs11 for receiving waters, baseline and post-
compliance annual plant-level loadings of each discharged pollutant, ambient pollutant concentrations in
receiving reaches and downstream reaches, pollutant concentrations in fish tissue, fish consumption rates
among different racial and ethnic cohorts for affected recreational anglers and subsistence fishers, and the
average daily dose (ADD) or lifetime average daily dose (LADD) of pollutants for each age cohort for
recreational anglers and subsistence fishers.
Section 3.1 describes how EPA identified the population potentially exposed to pollutants from steam electric
discharges via fish consumption. Section 3.2 describes the methods for determining fish tissue pollutant
concentrations and exposure via fish consumption in the affected population. Sections 3.3 to 3.7 describe
The analysis of human health benefits focuses on the fish consumption pathway only, since EPA assumed that
drinking water is treated to reduce pollutant concentrations below MCLs. See Chapter 2 for further discussion.
10 A dose response relationship is an increase in incidences of an adverse health outcome per unit increase in
exposure to a toxin.
11 A COMID is a unique numeric identifier for a given waterbody, assigned by a joint effort of the United States
Geological Survey, EPA, and Horizon Systems, Inc.
September 29, 2015 3T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
EPA's analysis of benefits of various human health endpoints affected by the final ELGs. The human health
benefits estimates rely on models of concentration-response and exposure pathways that embed assumptions
and involve limitations and uncertainties. Section 3.9 describes the limitations and uncertainties in the health
benefit analyses.
3.1 Affected Population
The affected population (i.e., individuals potentially exposed to steam electric pollutants via consumption of
contaminated fish tissue) includes recreational anglers and subsistence fishers who fish reaches affected by
steam electric 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, presence of substitute fishing locations, data on the locations and status offish consumption
advisories (FCAs) for affected reaches, and information on anglers' awareness and adherence to FCAs. Since
fish consumption rates vary across different racial and ethnic groups and across fishing mode (recreational
versus subsistence fishing), EPA estimated benefits separately for a number of age-, ethnicity-, and mode-
specific cohorts.
First, for each Census Block Group (CBG) within 50 miles of an affected reach, EPA pulled 2010 Census
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 or
higher), and then subdivided each group according to 7 racial/ethnic categories:12 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 Hispanic13.
Within each racial/ethnic group, EPA further subdivided the population according to recreational and
subsistence groups, assuming that 5 percent of the population practices subsistence fishing.14 Finally, 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.15 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
exposure cohorts [recreational vs. subsistence fishing] x 2 poverty status designations).
EPA distinguished the exposed population by racial/ethnic group and poverty status to support analysis of
potential Environmental Justice (EJ) considerations in 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 data.
Equation 3-1 shows how EPA estimated the affected population, ExPop(i)(s)(c), for CBG / in state s for
cohort c.
12 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.
13 The Mexican Hispanic and Hispanic block group populations were calculated by applying the Census tract
percent Mexican Hispanic and Hispanic to the underlying block-group populations, since these data were not
available at the block-group level.
14 Data is not available on the share of the fishing population that practices subsistence fishing. EPA assumed that
5% of people who fish practice subsistence fishing, based on the assumed 95th percentile fish consumption rate
for this population (see U.S. EPA, 201 Ib).
15 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.
September 29, 2015 3-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
Equation 3-1. £*Pop(0(s)(c) = Pop(0(c) x %Ftsh(s) x A(f) x Caff(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 U.S. Census Bureau's 2010 Census. The Census data provides
population numbers for each CBG broken out by age and racial/ethnic group separately. 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 the percentages to the population in each age group.
%Fish(s) = Fraction of people who live in households with anglers. To determine what percentage of the
total population participates in fishing, EPA used region-specific U.S. Fish and Wildlife Service (U.S.
FWS, 2011) estimates of the population 16 and older who fish.16 EPA assumed that the share of
households that includes anglers is equal to the fraction of people over 16 who are anglers.
A(i) = Adjustment for fish consumption advisories. EPA further adjusted the affected angler population
to reflect the presence of FCAs, where applicable for agiven reach. Based on EPA's review of studies
documenting anglers' awareness of FCA and their behavioral responses to FCAs, 57.0 percent to
61.2 percent of anglers are aware of FCAs, and 71.6 percent to 76.1 percent of those who are aware
ignore FCAs. Conservatively assuming that 61.2 percent of anglers are aware of applicable FCAs and
that 71.6 percent of aware anglers ignore them, the number of anglers exposed to steam electric
pollutants would be 17.4 percent lower for reaches with FCAs.17 Therefore, for receiving reaches with
FCAs, EPA reduced the affected populations by 17.4 percent.
CaR(c) = Adjustment for catch-and-release practices. According to U.S. FWS (2006) data,
approximately 23.3 percent of anglers release all the fish they catch ("catch-and-release" anglers).
Anglers practicing "catch-and-release" would not be exposed to steam electric pollutants via
consumption of contaminated fish. For all recreational anglers, EPA reduced the affected population by
23.3 percent. EPA assumed that subsistence fishers do not practice "catch-and-release" fishing.
Table 3-1 summarizes the population living within 50 miles of reaches affected by steam electric power plant
discharges (see Section 3.2.1 for a discussion of this distance buffer) and EPA's estimate of the population
potentially exposed to the pollutants via consumption of recreationally-caught fish (based on 2010 population
data and not adjusted for population growth during the analysis period). Of the total population, 14 percent
live within 50 miles of an affected reach and participate in recreational and/or subsistence fishing, and
10 percent are potentially exposed to fish contaminated by steam electric pollutants.
The share of the population who fishes ranges from 9 percent in the Pacific region to 23 percent in the West
North Central region. Other regions include the Middle Atlantic (11 percent), New England (12 percent), South
Atlantic (13 percent), Mountain (15 percent), West South Central (16 percent), East North Central (16 percent),
and East South Central (17 percent).
17 This is calculated as 61.2 percent aware of advisories times 28.4 percent (100%-71.6%) who choose not to fish
or otherwise don't eat fish caught in waterbodies affected by advisories.
September 29, 2015 3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
Table 3-1: Summary of Potentially Affected Population Living within 50 Miles of Affected
Reaches (baseline, as of 2010)
Total population
Total angler population3
Population potentially exposed to contaminated fishb> °
306,707,864
42,710,492
29,655,404
Source: U.S. EPA Analysis, 2015
a. Total population living within 50 miles of an affected reach times the state-specific share of the population who
fishes based on U.S. FWS (2011; between 9% and 23%).
b. Total angler population adjusted to reflect lower fishing/consumption rates for reaches with fish consumption
advisories and catch-and-release practices.
c. Analysis accounts for projected population growth so that the average affected population over the period of 2019
through 2042 is 21.3 percent higher than population in 2010 presented in the table, or 36 million people.
3.2 Pollutant Exposure from Fish Consumptior
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 ADD and LADD consumed via the fish consumption pathway.
3.2.1 Fish Tissue Pollutant Concentrations
The set of reaches that may represent a source of contaminated fish for recreational anglers and subsistence
fishers in each CBG depends on the typical travel distance anglers travel to fish. EPA assumed that anglers
typically travel up to 50 miles to fish, using this distance to estimate the relevant fishing sites for the
population of anglers in each CBG. See Appendix F for a sensitivity analysis using a travel distance of
100 miles. Based on data from the National Survey on Recreation and the Environment, about 80 percent of
all water-based recreation occurs with 100 miles of the users' homes (Viscusi et al., 2008).
Anglers may have several fishable sites to choose from within 50 miles of travel. To account for the effect of
substitute sites, EPA assumed that anglers are uniformly distributed among all the available fishing sites
within 50 miles from the CBG (travel zone) and alternate their travels across all the sites. For each CBG, EPA
identified all fishable COMIDs within 50 miles (where distance was determined based on the Pythagorean
distance between the centroid of the CBG and the midpoint of the COMID) and the COMID length in miles
EPA then calculated, for each CBG, the reach-length weighted average fish tissue concentration of arsenic
(As), mercury (Hg), and lead (Pb) based on all fishable sites within the 50 mile buffer. Appendix D describes
the approach used to derive ambient water and fish tissue concentrations of steam electric pollutants in the
baseline and under each of the regulatory options.
For each CBG, EPA then calculating the reach length (Length^ weighted fish filet concentration (C Fish_FUet
(CBG)) based on all fishable COMIDS within the 50 mile radius according to Equation 3-2:
Fnnatirm ** 9
Equation 3-2.
3.2.2 Average Daily Dose
Exposure to steam electric pollutants via fish consumption depends on the cohort-specific fish consumption
rates. Table 3-2 summarizes the fish consumption rates, expressed in daily grams per kilogram of body
September 29, 2015 3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
weight, according to the race/ethnicity and fishing mode. For more details on these fish consumption rates,
seeU.S.EPA(2015a).
Table 3-2: Summary of group-specific assumptions for human health benefit analysis3
Race/ Ethnicity
White (non-Hispanic)
African American (non-Hispanic)
Asian/Pacific Islander (non-Hispanic)
Tribal/Native Alaskan (non-Hispanic)
Other non-Hispanic
Mexican Hispanic
Other Hispanic
EA Cohortb
Non-Hispanic White
Non-Hispanic Black
Other, including Multiple Races
Other, including Multiple Races
Other, including Multiple Races
Mexican Hispanic
Other Hispanic
Consumption Rate (g/ kg BW/day)
Recreational
0.67
0.77
0.96
0.96
0.96
0.93
0.82
Subsistence
1.9
2.1
3.6
3.6
3.6
2.8
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. U.S. EPA (2015a).
Equation 3-3 and Equation 3-4 show the cohort- and CBG-specific calculation of the ADD and LADD based
on fish tissue concentrations, consumption rates, and other assumptions about exposure duration and
averaging periods, as shown below.
Equation 3-3. ADD(c~)(i) =
CFlsh_Fllet (0
1000
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)
Cfishjiiet (0 = average fish filet pollutant concentration consumed by humans for CBG / (mg per kg)
CRfisb(c) = Consumption rate offish for cohort c (grams per kg BW per day); see Table 3-2.
Ffish = fraction offish from contaminated source (percent; assumed value of 100%)
Equation 3-4. LADD =
ADD XED XEF
47x365
Where:
LADD(i)(c) = Lifetime average daily dose (mg per kg BW per day) for cohort c in CBG /'
ADD(i)(c) = 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; assumed value of 350)
AT= averaging time (years; assumed value of 70)
EPA used the doses of steam electric pollutants from recreational caught fish thus obtained in its analysis of
benefits associated with the various human health endpoints described below.
September 29, 2015
3-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
3.3 Benefits to Children from Reduced Lead Exposure
Lead is a highly toxic pollutant that can damage neurological, cardiovascular, and other major organ systems.
In particular, elevated lead exposure may induce a number of adverse neurological effects in children,
including hyperactivity, behavioral and attentional difficulties, delayed mental development, and motor and
perceptual skill deficits (see U.S. EPA, 2015a). Elevated blood lead (PbB) concentrations in children may
also result in metabolic effects such as impaired heme synthesis, anemia, and slowed growth (U.S. EPA,
1990; National Academy of Sciences, 1993; Kim et al. 1995). Severe lead poisoning may result in renal
effects, seizures, impaired coordination, recurrent vomiting, coma, and acute lead encephalopathy, a
potentially fatal condition (Piomelli et al., 1984; National Academy of Sciences, 1993; CDC, 2005). Studies
have also found a relationship between lead exposure in expectant mothers and lower birth weight in
newborns (Zhu et al., 2010; Bornschein et al., 1989; and Dietrich et al. 1987). Because of data limitations,
EPA estimated only the benefits from reducing neurological and cognitive damages to pre-school (ages 0 to
7) children using the dose-response relationship for IQ decrements (Schwartz 1994).
EPA estimated benefits from reduced exposure to lead to preschool children using PbB as a biomarker of lead
exposure. EPA first modeled PbB under the baseline and post-compliance scenarios, and then used a
concentration-response relationship between PbB and IQ loss to estimate avoided IQ losses in the affected
population of children and reduced incidences of extremely low IQ scores (less than 70, or two standard
deviations below the mean). EPA calculated the monetary value of benefits to children 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 |o,g/dL).
EPA used the methodology described in Section 3.1 to estimate the population of pre-school children who
live in recreational angler and subsistence fisher households and are potentially exposed to lead via
consumption of contaminated fish tissue. Since this benefit category applies to children up to the seventh
birthday only, EPA restricted the analysis to the relevant age cohorts of angler household members.
3.3.1 Methods
This analysis considers children who are born after implementation of the ELGs and live in recreational
angler and subsistence fisher households. It relies on EPA's Integrated Exposure, Uptake, and Biokinetics
(IEUBK) Model for Lead in Children (U.S. EPA, 2009c), 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. The BCA report for
the proposed ELG provides a more detailed description of the IEUBK model and describes EPA's application
of the model to estimating benefits to pre-school children from reduced exposure to lead contaminated fish
(U.S.EPA,2013a).
For each CBG, EPA used the cohort-specific ADD based on Equation 3-3. Lead bioavailability and uptake
after consumption varies 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 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-compliance scenarios. Note the IEUBK model processes daily intake to two
decimal places (ug/day). For this analysis, this means that some of the change between the baseline and
regulatory options is missed by using the model (i.e., it does not capture very small changes), since the
September 29, 2015 3-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
benefits are driven by very small changes across large populations.18 As such, the benefits shown in this
section are likely to underestimate the actual lead-related benefits to children arising from the ELG.
3.3.1.1 Estimating of Avoided IQ Point Losses
In a pooled data analysis, Lanphear et al. (2005; as cited in Agency for Toxic Substances and Disease
Registry (ATSDR), 2007) found that the greatest IQ losses per 1 |og/dL occur at the lowest ranges of PbB.
When the authors grouped IQ losses data for children with PbB below and above 7.5 |og/dL, they found that
the IQ losses were 2.94 points per |og/dL for children with PbB concentrations below 7.5 |og/dL and
0.16 points per |o,g/d for children with PbB concentrations above 7.5 |og/dL.
Given the baseline PbB levels estimated using the IEUBK model (mean PbB of approximately 2.7 (ig/dL),
EPA used the dose-response factor of 2.94 points per |og/dL from Lanphear et al. to estimate changes in IQ
losses between the baseline and post-compliance scenarios.19 Comparing the baseline and post-compliance
results provides the avoided IQ loss per child. Multiplying the result by the number of affected pre-school
children yields the total increase 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 U.S. Census Bureau (2010) 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 final rule, 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 avoided IQ loss is thus an annual value (i.e., it would apply to the cohort of children born each year
after implementation).20 Equation 3-5 shows this calculation for the annual increase in total IQ points.
Equation 3-5. A/0(Q(c) = (AGM(Q(c) * CRF
Where:
AIQ(i)(c) = the difference in total IQ points between the baseline and regulatory option scenarios for
cohort c in CBG /
AGM(i)(c) = the change in the average PbB in affected population of children ((ig/dL) in cohort c in
CBG/
CRF= concentration response function (2.94 IQ points lost per ug/dL increase in PbB)
ExCh(i) = the number of affected children aged 0 to 7 for CBG /'
The available economic literature provides little empirical data on society's overall WTP to avoid a decrease
in children's IQ. To determine 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.
18 For example, the average intake across all affected children is reduced from 0.333 ug/kg BW/day under the
baseline to 0.332 under Option D.
19 See Appendix F for a sensitivity analysis using a log-linear concentration-response function.
20 Dividing by seven undercounts overall benefits. Children from ages 1 to 7 are not accounted for in the base year
of the analysis, although they are presumably affected by lead exposure.
September 29, 2015 3^7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
Salkever (1995) and Schwartz (1994) estimate that a one point IQ reduction reduces expected lifetime
earnings by 2.38 percent and 1.76 percent, respectively. Data from the U.S. Census Bureau for 2009 indicate
that lifetime earnings are approximately $147,462 when discounting future earnings at 7 percent, and
$656,737 when discounting future earnings at 3 percent.21 The resulting estimated values of an IQ point are
summarized in Table 3-3.
Table 3-3: Value of an IQ Point3 (2013$)
Discount Rate
3 percent
7 percent
Assumed Reduction in Expected Lifetime Earnings (percent per IQ point)
1.76 percent/IQ point (Schwartz, 1994 )
$11,559
$2,559
2.38 percent/IQ point (Salkever, 1995)
$15,630
$3,510
a. Values are not adjusted for the cost of education.
Decreased IQ also results in less education and, therefore, reduced education costs. Data from the National
Center for Education Statistics (2014) indicate that the average expenditure per student in 2010/2011 was
$12,600 (in 2013$). Schwartz (1994) and Salkever (1995) estimate that a one IQ point reduction results in
0.131 and 0.101 fewer education years, respectively; this represents lifetime cost savings between $377 and
$970, discounting the avoided costs over 18 years. EPA subtracted these education costs from the value of
lifetime earnings per IQ point in Table 3-3. Subtracting education costs is done for accounting purposes only
and does not suggest that this is a desirable, positive outcome.
The value of an IQ point reduction adjusted for the avoided cost of education ranges between $2,107
(following Schwarz (1994) and discounting at 7 percent) to $14,883 (following Salkever (1995) and
discounting at 3 percent). This effect represents only one component of society's WTP to avoid IQ decreases,
and thus underestimates the total value of benefits to children from reduced exposure to lead.
3.3.1.2 Reduced Expenditures on Compensatory Education
Children whose PbB exceeds 20 ng/dL are more likely to have IQs less than 70, which means that they would
require compensatory education tailored to their specific needs. EPA's IEUBK model can generate
probabilities that a child would have a PbB in excess of a specific threshold.
EPA estimated the number of children that would have PbB above 20 ng/dL for each CBG under the baseline
and each analyzed regulatory option. EPA assumed that 20 percent of children with PbB above 20 ug/dL
would have IQs less than 70 and require compensatory education.22 Equation 3-6 shows the calculation of the
number of children requiring compensatory education for each cohort and CBG. Summing across all cohorts
and CBGs provides the total number of children who would require special education.
21 EPA updated lifetime earnings to 2013 dollars using the Consumer Price Index (2013=236.384; 2009 =
214.537).
22 This assumption follows the methodology used by EPA in Economic and Environmental Benefits Analysis
Document for the Final Effluent Limitations Guidelines and Standards for the Metal Products and Machinery
Point Source Category (U.S. EPA, 2003).
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Equation 3-6. Comp£d(i)(c) = ExCh(i} * Pr20(i)(c) * S
Where:
CompEd(i) = the number of children with PbB over 20 |o,g/dL and IQ less than 70 (who would need
compensatory education) for cohort c in CBG /
ExCh(i) = the number of affected children for cohort c in CBG /'
Pr20(i) = the probability that a child's PbB is above 20 |og/dL for cohort c in CBG /
S = Share of children with PbB over 20 |o,g/dL that would have IQ scores less than 70 (20%)
The U.S. Department of Education (Chambers, et al., 2003) estimated that average annual expenditures for a
student with mental retardation are approximately $8,484 higher than for an average student. Updating to
2013 dollars23 yields annual costs of $16,075 per child. EPA assumed that children with IQ less than 70
would incur these additional costs each year for 12 years. Discounting future costs using a 3 percent discount
rate yields a total compensatory education cost of approximately $164,806 per child with an IQ score less
than 70 ($136,613 per child if using a 7 percent discount rate).
3.3.2 Results
Table 3-4 shows the benefits associated with avoided IQ losses from lead exposure via fish consumption, and
Table 3-5 shows the benefits associated with a reduced need for specialized education. The final BAT/PSES
option will generate annualized benefits of $0.8 million to $1.1 million using a 3 percent discount rate, and
$0.1 million to $0.2 million using a 7 percent discount rate.
Table 3-4: Estimated Benefits from Avoided IQ Losses for Children Exposed to Lead
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Average
Annual
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
3,326,127
3,326,127
3,326,127
Total
Avoided
IQ Losses,
2021 to
2042
853
853
1,285
1,985
1,985
Annualized Value of Avoided IQ Point Losses" (Millions 2013$)
3% Discount Rate
Low Bound
$0.34
$0.34
$0.51
$0.79
$0.79
High Bound
$0.48
$0.48
$0.72
$1.11
$1.11
7% Discount Rate
Low Bound
$0.05
$0.05
$0.08
$0.13
$0.13
High Bound
$0.08
$0.08
$0.12
$0.19
$0.19
Source: U.S. EPA Analysis, 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76% of lifetime earnings (following Schwartz, 1994);
high bound assumes that the loss of one IQ point results in the loss of 2.38% of lifetime earnings (following Salkever, 1995).
Updated to 2013 dollars using the Consumer Price Index for Education (2013 = 224.521; 1999 = 107.0).
September 29, 2015
3-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Table 3-5: Estimated Avoided Cost of Compensatory Education for Children with Blood Lead
Concentrations above 20 ug/dL and IQ Less than 70
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
3,326,127
3,326,127
3,326,127
Decrease in Number
of Cases of IQ< 70,
in 2021 to 2042
1
1
1
2
2
Avoided Annual Cost (Millions; 2013$)a
3% Discount Rate
$0.00
$0.00
$0.01
$0.01
$0.01
7% Discount Rate
$0.00
$0.00
$0.00
$0.01
$0.01
Source: U.S. EPA Analysis, 2015
a. "-" indicates that a value was not estimated and "$0.00" indicates that avoided annual cost is less than !
101 million.
3.4 Benefits to Adults from Reduced Lead Exposure
The public health literature suggests that a wide spectrum of adverse health outcomes can occur from lead
exposure in people of all ages. For example, recent evidence has suggested that exposure to lead in adults can
result in CVD impacts; specifically, increases in hypertension, coronary heart disease, CVD, and CVD-related
mortality (National Toxicology Program, 2012; U.S. EPA, 2013b). EPA has developed and externally peer-
reviewed concentration response functions for CVD-related mortality resulting from adult exposure to lead
(Abt Associates, 2015 and U.S. EPA, 2015f). EPA is using these concentration response functions to estimate
human health benefits from reduced adult exposure to lead following the implementation of the final ELGs.
3.4.1 Methods
This section summarizes EPA's approach for estimating the benefits to human health due to reductions in
adult exposure to lead following implementation of the final ELGs. EPA relied on two models to estimate
baseline and post-compliance exposure for each population cohort:
> The first model estimates PbB in adults as a function of air-based background lead exposure and
consumption of contaminated fish tissues.
> The second model is a population life table model that estimates the gains in life years due to
decreased risk of CVD mortality from the PbB reductions.
EPA then aggregated the resulting cohort-specific gains in life expectancy to represent the total magnitude of
the expected human health benefits for each ELG option. EPA estimated monetized benefits by applying a
constant value per statistical life (VSL) to the estimated number of premature deaths avoided in each analysis
year.
3.4.1.1 Modeling Blood Lead Concentrations in Adults
The Leggett model (Leggett, 1993), a physiologically-based pharmacokinetic model, has been used to
estimate bone and blood lead levels in adults by U.S. EPA and others (California Office of Environmental
Health Hazard Assessment (Cal OEHHA), 2013a, 2013b).24 The model predicts PbB by explicitly modeling
lead absorption, distribution, metabolism, and excretion for 21 body tissue compartments on a daily time step.
Therefore, the model is able to address time-dependent conditions (e.g., changing exposures through time).
EPA Office of Pollution Prevention and Toxics (OPPT) is using the Leggett+ model for the upcoming
renovation and repair rule.
September 29, 2015
3-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Since its original publication, the model has been updated and improved to better fit observed data. Most
recently, the Cal OEHHA has published an updated version of the Leggett model, Leggett+, which they have
used for occupational health standards in California. EPA implemented the updated model in the R
programming language using publicly available documentation and MATLAB code (Cal OEHHA, 2013a,
2013b). The Agency modified the model inputs to more easily alter lead exposure due to contaminated fish
consumption, and to reduce the complexity (and therefore computational time) of background exposure from
air. Internal model parameters with respect to the transit of lead through the 21 body tissues are unchanged.
Table 3-6 presents and describes parameters used to apply the Leggett+ model.
Table 3-6: Parameters Used to Apply the Leggett+ Model
Parameter
Age at Start of
Exposure
Age in 2014
Duration of
Exposure to
Contaminated
Fish
Initial Blood
Lead
Concentration
Background Lead
Intake
Body Mass
Lead Intake from
Contaminated
Fish Tissue
Initial Tissue
Compartment
Parameters
Units
Years
Years
Years
Hg/dL
Hg/day
kg
mg/kg body
mass / day
Various
Value
18
25 - 65 in 10-year increments
Within age cohort, number of
years between 18 and year of
ELG implementation.
1.5
1.8
Age
20s
30s
40s
50s
60s
Men
82.5
85.7
88.0
88.5
88.2
Women
71.4
76.3
76.3
77.8
75.6
Average daily doses calculated
as described in Section 3.2.2
Various
Notes
The Leggett+ model is parameterized to model
PbB only for adults
Data used for all individuals within decadal
cohort
Pre-2014 exposures limited to 27 years (based
upon examination of model results). Pre-2014
exposure calculated as:
Duration = Age in 2014 - 18.
CDC 2009; Schober 2006
Calculated to maintain PbB at 1.5 ng/dL with
no fish exposure for a body mass of 74 kg (Cal
OEHHA 20 13a)
Values from Table 8-24 of EPA (201 Ib).
Values are held constant for all ages within
each decade. Affects quantity of contaminated
fish tissue consumed (see below).
Baseline (i.e., no ELG) values used until 2021;
thereafter, option-specific values.
Values used were obtained from CAL OEHHA
(2013a).
EPA ran the models for several population cohorts defined by sex, age, exposure (recreational and subsistence
fishers), and geography (215,460 CBGs) to characterize variability in the relevant population characteristics.
The model was run on a daily time-step for each of the combinations of input parameters (6 scenarios
[baseline + 5 ELG options)] x 2 sexes [male/female] x 215,460 CBGs x 2 exposure cohorts [recreational vs.
subsistence fishing] x 5 age cohorts [20s - 60s, by decade]). Simulation duration included two separate
The Leggett+ model was designed to model airborne workplace exposure to air; in adapting the model to this
analysis, EPA excised all parameters directly related to workplace exposure (including workplace airborne lead
concentrations, breathing rate, transfer fraction of inhaled lead to blood, etc.) to improve model runtime.
Background airborne ingestion remained consistent with the baseline Leggett+ model inputs.
September 29, 2015
3-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
components: exposures pre- and post-2015. Exposure occurring before 2015 (beginning at the age of 18 for
each decadal age group) was run once for each combination of input parameters. These parameter values were
then used as initial parameters for all model runs (baseline and the five ELG options) simulating exposure
after 2015. Exposure prior to 2015 was limited to 27 years in duration, because inspection of model results
suggested that additional lead accumulation in body tissues was minimal beyond 25 years. Consequently,
simulations under no-ELG conditions ranged in duration from 34 years (for the 20s age cohort: 7 years of pre-
2015 exposure plus 2015 - 2042, inclusive) to 54 years (for 50s and 60s age cohorts, 27 years of pre-2015
exposure plus 2015- 2042, inclusive), while the duration for all ELG-option simulations was 27 years (2015-
2042, inclusive).
For each model run, EPA averaged forecasted PbB outputs within forecast year. Figure 3-1 illustrates changes
in PbB over time for two selected male subsistence fishing cohorts within a single CBG under the baseline
and two of the regulatory options.
OJ -
CN
LO
o
•
•
Baseline - 20s
Baseline - 60s
Option B - 20s
Option B - 60s
Option D - 20s
Option D - 60s
2015
2020
2025 2030
Year
2035
2040
Figure 3-1: Example changes in PbB through time under baseline, option B, and option D for cohorts
in the 20s and 60s (age as of 2014). Differences between cohorts are driven by differences in body
mass. Example data represent modeled PbBs of male subsistence fishers residing in CBG
560419753004. The majority of cohorts nationwide experience a smaller reduction in PbB due to the
rule.
Relative to baseline, PbB was reduced for approximately 1.1 percent of the simulated population under
options A and B, 2.0 percent of simulations under option C, and 2.3 percent of simulated individuals under
options D and E. The extent of the reductions followed the spatial distribution and magnitude of loading
reductions across the options. When reductions did occur, they were typically small (e.g., < 0.03 (ig/dL in
2042 for options A - C, and < 0.08 (ig/dL in 2042 for options D - E).
Relative to baseline, PbB was reduced by more than 1 percent of baseline - a decrease EPA considers to be a
meaningful reduction level based upon CVD mortality rates - for between 0.03 percent to 1.18 percent of the
simulated population under options A to E, respectively, with the final BAT/PSES (Option D) showing 1.18
percent of simulated individuals with a reduction of greater than 1 percent of baseline PbB.
September 29, 2015
3-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
3.4.1.2 Estimating Hazard Reduction under the Final ELGs
For each sex, single-year age, and exposure cohort (hereafter cohort), the analysis characterizes two basic
survival analyses under the baseline and option scenarios: a death hazard function and a survival function
(details of the analysis are provided in Appendix E)26 To estimate changes in mortality following ELG
implementation, EPA characterized:
1. the baseline CVD death hazard function;
2. the baseline non-CVD death hazard function; and
3. an option-specific CVD death hazard function.
The main source of data for hazard estimation in key simulation elements (1) and (2) above is a life table, a
collection of statistics that shows age-specific probabilities of survival and fecundity.27 EPA obtained life
table data from two sources: the CDC's National Center for Health Statistics (Arias, 2014) and the CDC's
Underlying Cause of Death Database (CDC 2012).28 After obtaining all necessary data, EPA calculated age
and sex- specific baseline hazards for CVD mortality; a sample of life table data used in calculations is
presented in Table 3-7.
Table 3-7: Sample of Inputs for Age- and Sex-Specific Hazard Functions
Age
20s
30s
40s
50s
60s
20s
30s
40s
50s
60s
Sex
M
M
M
M
M
F
F
F
F
F
Proportion of
Population
0.0875
0.0906
0.0977
0.0925
0.0630
0.0852
0.0909
0.0995
0.0973
0.0693
Mean Body Mass
(kg)
82.5
85.7
88.0
88.5
88.2
71.4
76.3
76.3
77.8
75.6
CVD Rate
(per 1,000)
0.072149
0.242882
0.833447
2.209863
4.633724
0.037076
0.123911
0.392972
0.940202
2.303575
Using CVD rates collected, EPA then calculated option-specific CVD death hazard functions based on the
relationship between PbB and the CVD hazard ratio. To do this, EPA used a concentration-response function
from apeer reviewed study which found an adjusted relative hazards of CVD mortality of 1.53 (1.21-1.94)
per 3.4-fold increase in PbB (Menke et al. 2006).
3.4.1.3 Estimating Premature Deaths Avoided Over Multiple Years
The VSL is the marginal rate of substitution between wealth and mortality risk in a defined time period,
usually taken to be one year. Therefore, the product of VSL and the estimated aggregate reduction in risk of
Collett (2003), pp. 10-12.
An extensive discussion of life tables can be found in Shryock et al. (1980) Chapter 15.
Database query parameters included: Dataset: Underlying Cause of Death, 1999-2010; Autopsy: All; Gender:
Female, Male; ICD-10 Codes: 100-199 (Diseases of the Circulatory System); Place of Death: All; Race: All;
Single-year ages 20 - 100+, inclusive; Years: 2006-2010; States: All; Urbanization: All; Calculate Rates Per:
100,000; Rate Options: Default intercensal populations for years 2001-2009.
September 29, 2015
3-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
premature death represents the affected population's aggregate WTP to reduce its probability of death in one
year. EPA estimated the benefits of multi-year mortality risk as the product of:
1. The reduction in initial age-specific mortality rate (i.e., the proportion of people alive at exact age x,
who will die before attaining exact age x + 1 in year t); and
2. The number of individuals surviving to the beginning of year t. This value is calculated as the initial
cohort population size in 2015 multiplied by the probability that these individuals survive to age x,
and are alive at the beginning of year t to enjoy the benefits of the year's mortality risk reduction.
It is important to note that WTP for a specific reduction in risk may depend on (i) the time of payment, (ii) the
conditions under which the individual can save and borrow against future income, and (iii) whether the
individual knows about the change in survival probability ahead of time (Hammitt 2007). The earlier an
individual learns of the altered probability of survival, the earlier she "can adjust by reallocating her planned
future consumption and risk-reducing expenditures." (Hammitt 2007).
3.4.2 Results
Table 3-8 summarizes the magnitude and economic value of human health benefits from reduced incidence of
cardiovascular disease connected to adult lead exposure due to ELG implementation during the period 2019-
2042. At a 3 percent discount rate, the final BAT/PSES (Option D) has annualized benefits of $12.8 million;
the annualized benefits are $10.7 million at a 7 percent discount rate.
Table 3-8: Summary of Estimated Health Benefits due to Decreased Risk of CVD Mortality
during 2019-2042 based on the Economic Value of Avoided Premature Mortality (VSL)
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Aggregate Reduction
in Risk of Premature
Death
10.7
10.7
23.6
36.2
36.2
Annualized Benefits (millions 2013$)
3 Percent
$3.77
$3.77
$8.35
$12.80
$12.80
7 Percent
$3.14
$3.14
$6.97
$10.68
$10.68
Source: EPA Analysis, 2015
3.5 Benefits to Children from Reduced Mercury Exf
Mercury can have a variety of adverse health effects on adults and children (see U.S. EPA, 2015a). The final
ELGs are expected to reduce the discharge of mercury to surface waters by steam electric power plants and
therefore provide a range of human health benefits. Due to data limitations, however, EPA estimated only the
benefits from reduced IQ losses among children exposed to mercury in-utero as a result of maternal
consumption of contaminated fish.
EPA identified the population of children exposed in-utero starting from the CBG-specific affected
population described in Section 3.1. Because this analysis focuses only on infants born after implementation
of the ELGs, EPA further limited the affected population by estimating the number of women between the
ages of 15 and 42 potentially exposed to contaminated fish caught in the affected waterbodies, and
September 29, 2015
3-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
multiplying the result by ethnicity-specific average fertility rates.29 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 2012 in the National
Vital Statistics Report (Martin, et al., 2013). The fertility rate measures the number of births occurring per
1,000 women between the ages of 15-44 in a particular year. Fertility rates were highest for Hispanic women
at 74.4, followed by African Americans at 65.0, Asian or Pacific Islanders at 62.2, Caucasians at 58.6, and
Tribal/Other at 47.0.
3.5.1 Methods
EPA used the same ethnicity- and mode-specific consumption rates shown in Table 3-2 and calculated the
CBG- and cohort-specific mercury ADD based on Equation 3-3. 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.
To estimate maternal hair mercury concentrations based on the daily intake (see Section 3.2.2), EPA used the
median conversion factor derived by Swartout and Rice (2000), who estimated that a 0.08 (ig/kg body weight
increase in daily mercury dose is associated with a 1 ppm increase in hair concentration. Equation 3-7 shows
EPA's calculation of the total annual IQ decrement for a given receiving reach.
Equation 3-7. /
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
3.5.2 Results
Table 3-9 shows the estimated benefits of avoided IQ point losses for infants exposed to mercury in-utero.
The final BAT/PSES option will generate annualized benefits of $2.9 million to $4.0 million at a 3 percent
discount rate, and $0.5 million to $0.7 million at a 7 percent discount rate.
Table 3-9: Estimated Benefits from Avoided IQ Losses for Infants from Mercury Exposure
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of
Affected
Infants per
Year
418,953
418,953
418,953
418,953
418,953
Total Avoided
IQ Losses,
2021 to 2042
3,239
3,311
6,001
7,219
7,898
Annualized Value of Avoided IQ Point Losses" (Millions 2013$)
3% Discount Rate
Low Bound
$1.29
$1.32
$2.38
$2.87
$3.14
High Bound
$1.81
$1.85
$3.35
$4.03
$4.41
7% Discount Rate
Low Bound
$0.21
$0.21
$0.39
$0.47
$0.51
High Bound
$0.31
$0.32
$0.58
$0.69
$0.76
Source: U.S. EPA Analysis, 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76 percent of lifetime earnings (following Schwartz,
1994); high bound assumes that the loss of one IQ point results in the loss of 2.38 percent of lifetime earnings (following Salkever,
1995).
3.6 Reduced Cancer Cases from Arsenic Exposur
Among steam electric pollutants analyzed in the EA, arsenic is the only confirmed carcinogen with a
published dose response function (see U.S. EPA, 2010b).30 EPA estimated the number of annual cancer cases
associated with consumption offish contaminated with arsenic from steam electric discharges under the
baseline and each regulatory option. The reduction in the number of cancer cases from the baseline to post-
compliance represents human health benefits attributable to the final ELGs.
3.6.1 Methods
EPA used the cohort-specific arsenic LADD (see Section 3.2.2) to calculate the total number of cancer cases
for each cohort for each CBG under the baseline and each of the regulatory options, based on Equation 3-8.
Equation 3-8. CC(Q(c) = ExPop(i}(c) * CSF * LADD(i)(c)
Where:
CC(i)(c) = the number of cancer cases for cohort c in CBG /
ExPop(i)(c) = the number of people affected for cohort c in CBG /'
CSF= Cancer Slope Factor for skin cancer from arsenic [1.5 (mg/kg BW/day)"1].
LADD(i)(c) = Lifetime Average Daily Dose of arsenic for cohort c in CBG /' (mg/kg BW/day).
For this analysis, EPA used the current Integrated Risk Information System (IRIS) CSF, which is based on
incidences of skin cancer. EPA is currently revising its cancer assessment of arsenic to reflect new data on
30 Although other pollutants, such as cadmium, are also likely to be carcinogenic (see U.S. Department of Health
and Human Services (U.S. DHHS), 2008), EPA did not identify dose-response functions to quantify the effects
of changes in these other pollutants.
September 29, 2015
3-16
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 3: Human Health Benefits
internal cancers. It is possible that the revised combined (lung and bladder cancer) CSF would be higher (e.g.,
the draft value is 25.7 per mg/kg BW/day), suggesting that the use of the current IRIS value may bias benefits
downward.
Summing the number of cancer cases across all cohorts and all CBGs yields the total number of annual cancer
cases under the baseline and each of the regulatory options. To estimate the number of avoided cancer cases,
EPA subtracted the estimated number of cancer cases for each analyzed regulatory option from the estimated
number of cancer cases under the baseline.
In the analysis of the proposed Steam Electric ELGs, EPA had used VSL to place a monetary value on
avoiding a cancer case. This monetization approach inherently assumes that all cases would be fatal. Given
that the majority of skin cancer cases (which are the basis for the health benefits arising from reduced arsenic
exposure) are not fatal, this approach is likely to bias benefit estimates upward. For the analysis of the final
rule, therefore, EPA revised the monetization approach to value skin cancer cases based on a COI approach.
Based on a literature review, EPA developed COI estimates for the skin cancer health endpoint associated
with oral arsenic exposure (Abt Associates, 2014). The Agency found the following distribution of non-
melanoma skin cancers associated with arsenic:
> Basal cell carcinoma: 15%;
> Invasive squamous cell carcinoma: 19%;
> Non-invasive squamous cell carcinoma: 58%; and
> Combined: 8%.
These types of skin cancers have very low fatality rates. Diagnosis involves medical histories, physical
exams, and skin biopsies, while treatments consist of minor surgeries and periodic follow-up visits.
The COI estimates for skin cancers associated with arsenic include both direct medical expenditures and
indirect opportunity costs. The direct medical costs were based on Medical Expenditure Panel Survey
(MEPS) data on office-based provider visits and outpatient visits between 1996 and 2010, and represent the
mean expenditures per patient.31 Diagnosis and surgery are one-time costs, while follow-up visits are periodic
for all years after the surgery, with frequency depending on the type of cancer. The other component of the
cost of an illness is the opportunity cost - i.e., the value of time lost during the illness. For non-melanoma
skin cancer, EPA assumes that all patients in this analysis incur opportunity costs for diagnosis, surgery, and
follow-up doctor visits. Table 3-10 shows the total costs for different types of skin cancer.
Expenditures are for entire procedures; for example, expenditures for a surgery include the surgery itself as well
as the associated hospital stay.
September 29, 2015 3^7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Table 3-10: Total Costs of Illness for Skin Cancer
Type of Cost
Cost in First Year
Cost in Subsequent Years
Basal Cell Skin Cancer
Medical care
Opportunity cost
Total cost for BCC
$1,112
$282
$1,394
$307
$39
$347
Non-invasive Squamous Cell Skin Cancer
Medical care
Opportunity cost
Total cost for non-invasive SCC
$1,419
$322
$1, 741
$266
$34
$300
Invasive Squamous Cell Skin Cancer
Medical care
Opportunity cost
Total cost for invasive SCC
$3,020
$460
$3,480
$388
$50
$439
a. Updated to 2013$ using the consumer price index for medical care
Table 3-11 shows the weighted average skin cancer cost of illness estimate, based on the proportion of cases
for each type of skin cancer. Overall, a skin cancer case from arsenic exposure results in costs of
approximately $2,056 in the first year and $338 in the subsequent 14 years. Using a 3 percent discount rate,
this equates to a $5,877 value of a skin cancer case, and using a 7 percent discount rate, the value is $5,015.
Table 3-11: Weighted Average Skin Cancer Cost of Illness3
Type
Non-invasive squamous cell
Invasive squamous cell
Basal cell
Combination0
Total/Weighted Average
Proportion of Cases
58%
19%
15%
8%
100%
COI (first year)
$1,741
$3,480
$1,394
$2,205
$2,056
COI (subsequent years)b
$300
$439
$347
$362
$338
a. updated to 2013$ using the consumer price index for medical care.
b. Assumes 14 subsequent years.
c. For "combination," EPA calculated COI based on the average of basal cell, invasive squamous cell, and non-invasive squamous
cell cases.
A reduction in pollutant loadings does not immediately result in cessation of adverse health effects. There is a
lag between the time when exposures are reduced and the time when a reduction in risk occurs. Additionally,
there may be a latency period between the initial exposure and the onset of the illness. The latency period
between low-dose arsenic exposure and skin cancer is unknown (Karagas et al, 2001; Shannon and Strayer,
1989), though some researchers postulate it could range from several years to decades (Karagas et al., 1998).
U.S. EPA (2010b) notes that the cessation lag for skin cancer from arsenic is unknown, but that the cessation
lag for internal cancers from arsenic may be longer than for skin cancer, ranging from 15 to 50 years. For this
analysis, EPA assumed that cancer cases resulting from arsenic would not occur for ten years after exposure
and discounted the value of avoided cancer cases by an additional ten years.
September 29, 2015
3-18
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
3.6.2 Results
Table 3-12 shows the estimated changes in incidence of cancer cases from exposure to arsenic in fish tissue
under the ELGs and the annualized benefits calculated using 3 percent and 7 percent discount rates. Under
both discount rates, annualized benefits are under $0.01 million.
Table 3-12: Annual Benefits from Reduced Cancer Cases due to Arsenic Exposure
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Annual
Affected
Population
35,972,005
35,972,005
35,972,005
35,972,005
35,972,005
Reduced Cancer Cases,
2019 to 2042
0.03
0.03
0.06
0.09
0.1
Benefits (Millions 2013$)a
3% Discount
$0.00
$0.00
$0.00
$0.00
$0.00
7% Discount
$0.00
$0.00
$0.00
$0.00
$0.00
Source: U.S. EPA Analysis, 2015
a. "-" indicates that a value was not estimated and "$0.00" indicates that annual benefits are less than $0.01 million.
btal Monetized Human Heal
Table 3-13 presents total monetized human health benefits for the final BAT/PSES (Option D) and alternate
regulatory options. Using a 3 percent discount rate, benefits under Option D range from $16.5 million to
$18.0 million ($11.3 million to $11.6 million using a 7 percent discount rate). Reduced lead exposure for
adults and reduced mercury exposure for children represent the majority of total monetized human health
benefits.
Table 3-13: Total Monetized Human Health Benefits for ELG Options (millions of 2013$)
Discount
Rate
3%
7%
Option
A
B
C
D
E
A
B
C
D
E
Reduced Lead
Exposure for
Children3
Low
$0.34
$0.34
$0.52
$0.80
$0.80
$0.05
$0.05
$0.08
$0.14
$0.14
High
$0.48
$0.48
$0.73
$1.12
$1.12
$0.08
$0.08
$0.12
$0.20
$0.20
Reduced
Lead
Exposure
for Adults
$3.77
$3.77
$8.35
$12.80
$12.80
$3.14
$3.14
$6.97
$10.68
$10.68
Reduced Mercury
Exposure for
Children3
Low
$1.29
$1.32
$2.38
$2.87
$3.14
$0.21
$0.21
$0.39
$0.47
$0.51
High
$1.81
$1.85
$3.35
$4.03
$4.41
$0.31
$0.32
$0.58
$0.69
$0.76
Reduced
Cancer
Cases from
Arsenicb
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
Total3
Low
$5.40
$5.43
$11.25
$16.47
$16.74
$3.40
$3.40
$7.44
$11.29
$11.33
High
$6.06
$6.10
$12.43
$17.95
$18.33
$3.53
$3.54
$7.67
$11.57
$11.64
Source: U.S. EPA Analysis, 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76 percent of lifetime earnings (following Schwartz,
1994); high bound assumes that the loss of one IQ point results in the loss of 2.38 percent of lifetime earnings (following Salkever,
1995).
b. "$0.00" indicates that annual benefits are less than $0.01 million.
September 29, 2015
3-19
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
3.8
Additional Measures of Human Health Benefits
The benefits described above are only some of the human health effects expected to improve as a result of the
final ELGs for which EPA was able to identify dose-response relationships. As noted in the introduction to
this Chapter, pollutants in steam electric power plant discharges have been linked to additional adverse human
health effects. To provide an additional measure of the potential health benefits of the final ELGs, EPA also
estimated the expected reduction in the number of receiving reaches with pollutant concentrations in excess of
human health-based AWQC. This analysis and its findings are not additive to the preceding analyses in this
chapter, but represent another way of characterizing potential health benefits resulting from reduced exposure
to steam electric pollutants. This analysis compares in-stream pollutant concentrations estimated for the
baseline and each analyzed ELG option in receiving reaches and downstream reaches (see the EA; U.S. EPA,
2015a) to criteria established by EPA for protection of human health. EPA compared in-water concentrations
of arsenic, copper, nickel, selenium, thallium, and zinc to EPA's national recommended water quality criteria
protective of human health used by states and tribes (U.S. EPA, 2012b). Pollutant concentrations in excess of
these values indicate potential risks to human health. For another four steam electric pollutants (cadmium,
chromium, lead, and mercury) for which there are no recommended criteria, EPA instead compared
concentrations to MCLs (U.S. EPA, 2012a).
Table 3-14 shows the results of this analysis. EPA estimates that in-stream concentrations of steam electric
pollutants exceed human health criteria or MCLs for at least one pollutant in 3,673 reaches nationwide as a
result of baseline steam electric pollutant discharges. EPA expects that the final BAT/PSES (Option D) will
reduce the occurrence of concentrations in excess of human health-based criteria for 1,973 of the reaches, and
eliminate all exceedances for 1,870 of those reaches. While Option D reduces concentrations in the remaining
103 reaches relative to the baseline levels, and thereby improves water quality in these reaches, the reductions
are not sufficient to bring concentrations below the human health criteria or MCLs.
Table 3-14: Estimated Number of Reaches Exceeding Human Health Criteria for Steam Electric
Pollutants
Regulatory
Option
Baseline
Option A
Option B
Option C
Option D
Option E
Number of Reaches with
Steam Electric Pollutant"
Concentrations Exceeding
Human Health Criteria for at
Least One Pollutant
3,673
3,128
3,128
2,270
1,803
1,625
Number of Reaches with Improved Water Quality,
Relative to Baseline
Number of Reaches with
Reduced Number of
Exceedancesb
~
631
633
1,540
1,973
2,088
Number of Reaches with
All Exceedances
Eliminated
~
545
545
1,403
1,870
2,048
Source: U.S. EPA Analysis, 2015
a. Pollutants include arsenic, copper, nickel, selenium, thallium, zinc, cadmium, chromium, lead, and mercury.
b. The number of reaches with exceedances reduced includes those reaches where all exceedances are eliminated.
c>.y Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. The revisions affect baseline discharges of the three
September 29, 2015
3-20
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
pollutants explicitly modeled to quantify and monetize human health benefits in this Chapter: lead, mercury,
and arsenic. Table 3-15 summarizes changes in loading estimates between the original loads used in
estimating benefits in this Chapter and the revised loads. The table shows total industry loads of lead, mercury
and arsenic in the baseline and under the final ELGs (Option D), as well as pollutant removals achieved by
the final ELGs.
Table 3-15: Estimated Aggregate Changes in Pollutant Loadings for Lead, Mercury and Arsenic
(Pounds per Year).
Pollutant
Lead
Mercury
Arsenic
Baseline
Original
Values
14,588
1,176
22,219
Revised
Values
7,674
992
20,138
%
Change
-47%
-16%
-9%
Option D
Original
Values
350
31
1,483
Revised
Values
331
31
1,480
%
Change
-5%
-1%
0%
Removal under Option D
Original
Values
14,238
1,145
20,737
Revised
Values
7,343
961
18,658
%
Change
-48%
-16%
-10%
Source: U.S. EPA Analysis, 2015
As shown in Table 3-15, the revisions mainly reduced the baseline loadings of lead, mercury and arsenic, and
had a smaller relative impact on loadings post-compliance. Because baseline pollutant loads are lower than
previously estimated, EPA expects the adverse health effects estimated in the baseline to also be lower than
previously calculated, which correspondingly reduces the incidence of adverse health effects avoided by
implementing the final ELGs and therefore expected improvements from the final ELGs.
For several reasons — notably the fact that revisions do not affect all plants equally, exposure also depends
on other point sources, and many model functions are not linear —it would be inappropriate to simply scale
the monetized benefits based on the aggregate changes in loadings. While EPA cannot readily recalculate the
benefits, the direction and magnitude of the change in pollutant removals in Table 3-15 indicate that it is
likely that revisions to the loadings affect benefit estimates and that new benefit estimates would be lower
than presented in this Chapter. The magnitude of this reduction is, however, uncertain.
3.10 Limitations and Uncertainties
The analysis presented in this chapter does not include all possible human health benefits associated with
post-compliance reductions 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 health
benefits included in this analysis represent only a subset of the potential health benefits expected to result
from the final rule.
Additionally, the methodologies and data used in the analysis of health benefits associated with reduced
incidences of adverse health outcomes due to consumption offish contaminated with steam electric pollutants
involve limitations and uncertainties, that add to the limitations and uncertainties inherited from the EA
analysis and data (see U.S. EPA, 2015a). Table 3-16 summarizes the limitations and uncertainties and
indicates the direction of the potential bias.
September 29, 2015
3-21
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Table 3-16: Limitations and Uncertainties in the Analysis of Human Health Benefits
Uncertainty/Assumption
, i • • i j i j • i
1 tie analysis is based loadings that
were subsequently revised by EPA
The analysis does not consider the
suitability of alternate fishing
sites.
Anglers are assumed to be
distributed evenly (over the reach
miles) over all available fishing
sites within the 50-mile travel
distance.
The number of subsistence fishers
was assumed to equal 5 percent of
the total number of anglers fishing
the affected reaches.
EPA used a CSF for arsenic of 1.5
cases per mg/kg BW/day based on
skin cancer only.
There is a linear 0.18 point IQ loss
for each 1 ppm increase in
maternal hair mercury.
Effect on Benefits
Estimate
Overestimate
Uncertain
Uncertain
Uncertain
Uncertain
Uncertain
Notes
Revised loadings of lead, mercury, and arsenic are
lower than the loadings used to estimate benefits. The
changes indicate that the incidence of adverse health
effects are likely lower in the baseline, and therefore
that improvements due to the ELGs may be
correspondingly lower.
Estimating the number of anglers fishing on receiving
and downstream reaches based on the ratio of reach
length to the total number of reach miles within the
same 50-mile buffer area recognizes the effects of the
quantity of competing fishing opportunities on the
likelihood of fishing a given reach, but does not
account for the differential quality of fishing sites. If
the quality of substitute sites is distinctly worse or
better (e.g., some sites have better access or designated
fishing areas), the estimated benefits may be overstated
or understated.
EPA assumed that all anglers travel up to 50 miles and
distribute their visits over all fishable sites within the
area. In fact, recreational anglers 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 the assumption on
benefit estimates is uncertain since fewer/more anglers
may be exposed to higher/lower fish tissue
concentrations than assumed by EPA in the analysis.
The magnitude of subsistence fishing in the United
States or individual states is not known. Assuming
5 percent may understate or overstate the number of
potentially affected subsistence fishers (and their
households) overall, and ignores potential variability in
subsistence rates across racial/ethnic groups.
This is the current IRIS value and was based on
incidences of skin cancer. EPA is currently revising its
cancer assessment of arsenic to reflect new data on
internal cancers.
This dose-response function may over- or
underestimate IQ impacts arising from mercury
exposure if a linear function is not the best
representation of the relationship between maternal
body burden and IQ losses.
September 29, 2015
3-22
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
3: Human Health Benefits
Table 3-16: Limitations and Uncertainties in the Analysis of Human Health Benefits
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
For the mercury- and lead-related
benefits analyses, EPA assumed
that IQ losses are 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 acquisition
and retention of information presented verbally and
many motor skills (U.S. EPA, 2005). To the extent that
these impacts create disadvantages for children
exposed to mercury at current exposure levels or result
in the absence of (or independent from) measurable IQ
losses, this analysis may underestimate the benefits of
the ELGs of reduced lead and mercury exposure.
The IEUBK model processes daily
intake from "alternative sources"
to 2 decimal places (ug/day).
Underestimate
Since the intakes are very small, some variation is
missed by using the model (/'. e., it does not capture
very small changes).
EPA did not quantify the benefits
associated with reduced adult
exposure to mercury.
Underestimate
The scientific literature suggests that exposure to
mercury may have significant adverse health effects for
adults; if measurable effects are occurring at current
exposure levels, excluding the benefits of reduced adult
exposure results in an underestimate of benefits.
EPA assumed constant body mass
for all males and females in the
adult Pb and As analyses
Uncertain
Male and female body mass estimates used in analyses
are national estimates obtained from the CDC. The
extent to which CBG-specific body masses varies will
have an unknown effect on benefits
Uniform application of data from
national life tables in adult Pb and
As analyses
Uncertain
By applying national averages, EPA assumes that
mortality rates of all modeled cohorts (specific to
location, fishing cohort, etc.) do not differ from the
national mortality experience. EPA also assumes that
age-specific mortality rates are constant through time.
CVD-related benefits of reduced
Pb exposure are not tracked for
individuals younger than age 20 in
2014
Underestimate
Benefits accruing to younger individuals are not
tracked because the Leggett+ model is unable to model
PbBs in youth and adolescents.
EPA assumed no cessation lag for
PbB effects
Uncertain
EPA assumed no time lag between changes in PbB and
risk reductions. The accuracy of this assumption is
unknown.
Concentration-response functions
used in adult Pb and As analyses
Uncertain
The overall applicability of the concentration-response
functions are subject to uncertainty. Generalized
concentration-response functions are used to model
age-specific hazard ratios.
September 29, 2015
3-23
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
4 Nonmarket Benefits from Water Quality Improvements
As discussed in the EA document (U.S. EPA, 2015a), heavy metals, nutrients, and other pollutants discharged
by steam electric power plants have a wide range of effects on water resources located in the vicinity and
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. These second types of
environmental goods and services are classified as "nonmarket". The expected changes in the nonmarket
values of the water resources affected by the final ELGs (hereafter nonmarket benefits) are additive to the
market benefits (e.g., avoided costs of producing various market goods and services) and benefits from
improved groundwater quality estimated in other chapters (Freeman, 2003).
EPA's approach to estimating the nonmarket benefits from water quality improvements resulting from the
final ELGs involves 1) characterizing baseline and post-compliance water quality using a water quality index
and 2) monetizing changes in the nonmarket value of affected water resources attributable to the final ELGs
using a meta-analysis of surface water valuation studies that provide data on the public's WTP for water
quality improvements. The analysis accounts for improvements in water quality resulting from changes in
nutrient, sediment, and metals concentrations in reaches affected by discharges from steam electric power
plants.
4.1 Water Quality
To link water quality changes from reduced metal, nutrient and sediment discharges to effects on human uses
and support for aquatic and terrestrial species habitat, EPA used a water quality index (WQI) which translates
water quality measurements, gathered for multiple parameters that are indicative of various aspects of water
quality, into a single numerical indicator.
The WQI provides the link between specific pollutant levels, as reflected in individual index parameters (e.g.,
dissolved oxygen (DO) concentrations), and the presence of aquatic species and suitability for particular uses.
The WQI value, which is measured on a scale from 0 to 100, reflects varying water quality, with 0 for poor
quality and 100 for excellent.
The WQI used in this analysis is the same as was used for the proposed rule analysis.32 The WQI includes the
six parameters of the WQI previously used for the Final Construction and Development Rule - DO,
biochemical oxygen demand (BOD), fecal coliform (FC), total nitrogen (TN), total phosphorus (TP), and total
suspended solids (TSS) - and one additional subindex for metals, for a total of seven parameters.33'34 As
The WQI modifies the WQI used by EPA in the Environmental Impact and Benefits Assessment for Final
Effluent Guidelines and Standards for the Construction and Development Category (also referred to as the C&D
rule; U.S. EPA, 2009a), which builds on McClelland (1974) and on the methodology developed by Dunnette
(1979) and subsequently updated by Cude (2001) to better account for spatial and morphologic variability in the
natural characteristics of streams. A more detailed discussion of the history of the WQI framework is found in
Chapter 10 of the C&D report (U.S. EPA, 2009b).
EPA modified the WQI for freshwater waterbodies from the C&D analysis to include metals. This was done by
incorporating elements of the WQI developed by the Canadian Council of Ministers of the Environment
September 29, 2015 4T
33
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
discussed in Chapter 4 of the EA document (U.S. EPA, 2015a), lotic systems such as rivers and streams,
account for the vast majority (82 percent) of water bodies receiving direct discharges from steam electric
power plants, with 183 of the 222 immediately receiving waters. Lentic freshwater systems such as lakes
(with the exception of the Great Lakes), ponds, and reservoirs, represent 12 percent, the Great Lakes another
5 percent, and estuaries the remaining 1 percent. EPA focused the national level model on rivers/streams and
lakes/ponds/reservoirs as the most common affected waterbodies. Because of the specific hydrodynamics and
scale of the analysis required to appropriately model and quantify receiving water concentrations in the Great
Lakes and estuary systems, EPA looked at the changes in pollutant loadings and impacts to these systems in
selected case studies (see EA document for details; U.S. EPA, 2015a). EPA did not quantify the benefits to
these systems, leading to an underestimate of the benefits discussed in this chapter.
4.1.1 WQI Calculation
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 (/'. e., in the absence of the final ELGs), and for each analyzed regulatory option.
The first step in the implementation of the WQI involves obtaining water quality levels for each parameter,
and for each waterbody, under both baseline conditions and post-compliance conditions (see Section 0). 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 TSS, 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 the variability in geologic and other
region-specific conditions, and to reflect the national context of the analysis. TSS, TN, and TP subindex
curves were developed for each Level III ecoregion (U.S. EPA, 2009a) using baseline TSS, TN, and TP
concentrations calculated in SPARROW at the E2RF1 reach level.35'36'37 For each of the 85 Level III
ecoregions intersected by the E2RF1 reach network, EPA derived the transformation curves by assigning a
(CCME) wherein the index values are calculated based on the scope, frequency, and amplitude of exceedances
of specified numeric thresholds (CCME, 2001).
34 EPA analyzed changes to water quality resulting from the implementation of the final ELGs on receiving
freshwater reaches. While steam electric plants also discharge to estuarine and coastal reaches, EPA did not
estimate benefits from reducing pollutant loadings to these waterbodies due to the relatively small changes in
concentrations expected.
35 The SPARROW (SPAtially Referenced Regressions On Watershed attributes) model was developed by the
United States Geological Survey (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
36 EPA's E2RF1 (Enhanced River File Version 2.0) is a digital stream networks used in SPARROW models. This
dataset extends over the continental United States and includes approximately 62,000 stream reaches.
37 Following the approach EPA used for the C&D analysis, the selected data exclude outlier TSS concentrations,
defined as values that exceed the 95th percentile based on the universe of all E2RF1 reaches modeled in
SPARROW (U.S. EPA, 2009a). In the C&D analysis, the USGS and EPA had determined that these outlier
values corresponded to headwater reaches and were an artifact of the model rather than expected concentrations.
September 29, 2015 4^
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
score of 100 to the 25th percentile of the reach-level TSS concentrations 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 metals-specific subindex curve based on the number of Ambient Water
Quality Criteria (AWQC) exceedances for metals in each waterbody. National freshwater chronic AWQC
values are available for arsenic, cadmium, chromium, lead, mercury, nickel, selenium, and zinc. 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 metal
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 (see Section 0
and Appendix D for details), and therefore any exceedance of an AWQC may indicate that ambient
concentrations exceed AWQCs 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 AWQC then the water is
impaired for that constituent and the level of exceedance is of secondary concern. Using this approach, the
subindex curve for metals assigns the lowest subindex score of 0 to waters where exceedances are observed
for all eight metals analyzed, and a maximum score of 100 to waters where there are no exceedances.
Intermediate values are distributed evenly between 0 and 100.
Table 4-1 presents parameter-specific functions used for transforming water quality data into water quality
subindices for freshwater waterbodies for the six traditional pollutants. Table 4-2 presents the subindex values
for metals. The equation parameters for each of the 85 ecoregion-specific TSS, TN, and TP subindex curves
are provided in Appendix G.
Table 4-1 : Freshwater Water Quality Subindices
Parameter
Concentrations
Concentration
Unit
Subindex
Dissolved Oxygen (DO)
DO saturation 10.5
mg/L
mg/L
mg/L
10
-80.29+3 1.88 xDO-1.401 xDO2
100
100% < DO saturation < 275%
DO
N/A
mg/L
100 x exp((DOsat - 100) x -1.197xlO"2)
275% < DO saturation
DO
N/A
mg/L
10
Fecal Coliform (FC)
FC
FC
FC
FC > 1,600
50
< -9.9178X10'4)
98
September 29, 2015
4-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-1: Freshwater Water Quality Subindices
Parameter
Concentrations
Concentration
Unit
Subindex
Total Nitrogen (TN)a
TN
TN
TN
TN > TN10
TN10o < TN < TN10
TN < TN10()
mg/L
mg/L
mg/L
10
a x exp(TNxb); where a and b are ecoregion-
specific values in Appendix G
100
Total Phosphorus (TP)b
TP
TP
TP
TP>TP10
TP100TSS10
TSS1008
BOD<8
mg/L
mg/L
10
100
< exp(BOD
x -0.1993)
a. TN10 and TNI 00 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. TSS10 and TSS100 are ecoregion-specific TSS 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 using methodology in Cude (2001).
Table 4-2: Freshwater Water Quality Subindex for Heavy Metals
Number of Metals with AWQC
Exceedances
Subindex
0
1
2
o
3
4
5
6
7
8
100.0
87.5
75.0
62.5
50.0
37.5
25.0
12.5
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
September 29, 2015
4-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
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 4-1 presents EPA's calculation of the overall WQI score.
Equation 4-1.
WQIr =
W;
n =
the multiplicative water quality index (from 0 to 100) for reach r
the water quality subindex measure for parameter /
the weight of the /'-th parameter (0.143)
the number of parameters (i.e., seven)
Once an overall WQI value is calculated, it 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. Table 4-3 presents water use classifications and the
corresponding WQL and WQI values.
Table 4-3: Water Quality Classifications
Water Quality Classification
. . . drinking without treatment
... swimming
. . . game fishing
... rough fishing
. . . boating
WQL Value
9.5
7.0
5.0
4.5
2.5
WQI Value
95
70
50
45
25
Source: Vaughan (1986)
4.1.2 Sources of Data on Ambient Water Quality
EPA used the following data sources to obtain ambient concentrations for the seven parameters included in
the WQI:
> Outputs from USGS's SPARROW models provided baseline and post-compliance concentrations of
total nitrogen, total phosphorus, and total suspended solids. These calibrated national models are the
same models previously used by EPA in the C&D rule analysis (U.S. EPA, 2009c). See Appendix D
for further details.
> EPA estimated baseline and post-compliance metal concentrations using the water quality model
component of EPA's Risk-Screening Environmental Indicators (RSEI) model (U.S. EPA, 2012c).
EPA used estimates of metal loadings discharged from steam electric plants to directly receiving
reaches under the baseline and the five analyzed options (see EA for discussion of directly receiving
reaches; U.S. EPA, 2015a). EPA input the loadings from steam electric plants in the RSEI model to
estimate the long-term average concentrations in directly receiving and downstream reaches. These
September 29, 2015
4-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
loadings are used to complement discharges already included in RSEI for other facilities that report to
TRI. See Appendix D for details. The number of exceedances per waterbody (each reach) was
calculated by comparing baseline and post-compliance concentrations with EPA's freshwater chronic
aquatic life criteria values for each metal.38 If the concentration was greater than the aquatic life
criteria value for a given metal, EPA categorized the waterbody as having an AWQC exceedance for
that metal. EPA then summed the total number of metals with AWQC exceedances (up to eight) for
each waterbody under the baseline and under each analyzed regulatory option.
> The USGS National Water Information System (NWIS) provided concentration data for three
parameters: 1) fecal coliform, 2) dissolved oxygen, and 3) biochemical oxygen demand. 39 EPA's
Storage and Retrieval (STORET) data warehouse provided additional data on fecal coliform counts
and biochemical oxygen demand where NWIS data was unavailable (U.S. EPA, 2008a).40
Note that the concentration data input into the WQI typically represent long-term average concentrations.
Table 4-4 summarizes the water quality modeling data used for estimating baseline and post-compliance
metal, nutrient and sediment concentrations for reaches directly receiving steam electric plant discharge and
for downstream reaches.
Table 4-4: Water Quality Modeling Data used in Calculating the Baseline and Policy Metal, Nutrient
and Sediment Concentrations
Reach
Input Data
Water Quality Model
Model Output
Reaches directly
receiving steam
electric plant
discharge and
downstream
reaches (18,622
NHD reaches total)
Baseline and policy metal loadings
to inland reaches directly receiving
steam electric plant discharges
(U.S. EPA, 2015a).
Metal loadings from other TRI
dischargers in 2012.
Concentrations modeled in
RSEI
In-steam metal
concentrations at the NHD
level
Baseline and policy nutrient and
sediment loadings to inland
reaches directly receiving steam
electric plant discharges (U.S.
EPA, 2015a)
Baseline values for other nutrient
and sediment sources in inland
reaches (e.g., urban, agricultural
and forested land, animal
agriculture (nutrients), streambed
(sediment), atmospheric deposition
(nitrogen)).
Concentrations modeled in
SPARROW
In-stream nutrient and
sediment concentrations at
the E2RF1 level
RSEI utilizes the USGS's National Hydrology Dataset (NHD) which 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. The NHD
reaches in this analysis range from 0.003 miles to 9.11 miles in length.
USGS's NWIS dataset 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/
EPA's STORET (STOrage and RETrieval) Data Warehouse is a repository for water quality, biological, and
physical water monitoring data. More information can be found at http://www.epa.gov/storet/
September 29, 2015
4-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
EPA used two different reach classification frameworks to assess in-stream water quality under the baseline
and each of the regulatory options: the National Hydrography Dataset (NHD) network and the USGS's
Enhanced River File 1 (E2RF1). Metal concentrations were estimated for reaches indexed to the NHD
network. In contrast, the SPARROW, NWIS, and STORET data are available for reaches indexed to the
E2RF1 network and to USGS's Hydrological Unit Code (HUC) watersheds. The WQI and benefits are
ultimately calculated at the resolution of NHD reaches, but with adjustments made to data available only at
the E2RF1 level to reflect differences in spatial scale. Thus, to reconcile the two levels of resolution, EPA
mapped all modeled reaches from the E2RF1 to the NHD network using GIS and assigned the closest E2RF1
ID to each NHD reach. Figure 4-1 illustrates the differences in scale between the E2RF1 network and the
NHD network.
Legend
— NHD Medium Resolution FlowlinesVsi (1 100.000 scale)
^^~ Enhanced River Fi!e 1 Reach Network
^ USGS HUC 8-Dlgit Watershed
5 10
20
30
40
• Km
Figure 4-1: Comparison between the NHD and E2RF1 Network in a Single Watershed.
The water quality analysis included a total of 18,622 NHD reaches, totaling 27,421 miles, that are potentially
affected by steam electric plants under baseline conditions. Baseline concentrations for all WQI parameters
were available for over 95 percent of the potentially affected NHD reaches. EPA used a successive average
approach to address the data gaps in WQI parameters not described above (i.e., DO, BOD, fecal coliform) in
the remaining inland reaches. The approach involves assigning the average of ambient concentrations for a
WQI parameter within a hydrologic unit to reaches within the same hydrologic unit with missing data, and
progressively expanding the geographical scope of the hydrologic unit (HUC8, HUC6, HUC4, and HUC2) to
September 29, 2015
4-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
fill in all missing data.41 This approach assumes that reaches located in the same watershed generally share
similar characteristics. Using this estimation approach, EPA compiled baseline water quality data for all
analyzed NHD reaches. Table 4-5 summarizes the data sources used to estimate baseline and post-
compliance values by water quality parameters.
Table 4-5: Water Quality Data used in Calculating the Baseline and Policy WQI
Parameter
TN
TP
TSS
DO
BOD
Fecal Coliform
Metals
Baseline value
From SPARROW output (baseline run)
matched to NHD level
From SPARROW output (baseline run)
matched to NHD level
From SPARROW output (baseline run)
matched to NHD level
Baseline value at the E2RF1 level matched to
NHD level3
Baseline value at the E2RF1 level matched to
NHD level3
Baseline value at the E2RF1 level matched to
NHD level3
Baseline exceedances at NHD level based on
RSEI model outputs
Post-compliance value
From SPARROW output (regulatory option run)
matched to NHD level
From SPARROW output (regulatory option run)
matched to NHD level
From SPARROW output (regulatory option run)
matched to NHD level
No change. Regulatory option value equal
baseline value
No change. Regulatory option value equal
baseline value
No change. Regulatory option value equal
baseline value
Regulatory option exceedances at NHD level
based on RSEI model outputs.
a. Values based on STORET and NWIS data, complemented with data available for progressively larger geographical units
(HUCS, HUC6, HUC4, and HUC2), as needed to fill in all missing data.
4.1.3 Baseline WQI
The water quality analysis included a total of 18,622 NHD reaches that are potentially affected by steam
electric power generating plants under baseline conditions. Based on the estimated WQI value under the
baseline scenario (WQI-BL), EPA categorized each of these 18,622 NHD reaches using five WQI ranges
(WQI < 25, 25
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-6: Percentage of Potentially Affected Inland Reach Miles by WQI Classification: Baseline
Scenario
Water Quality
Classification
Unusable
Suitable for
Boating
Suitable for Rough
Fishing
Suitable for Game
Fishing
Suitable for
Swimming
Baseline WQ
WQK25
25 0 under Option D, comprising a
total of 19,573 reach miles, or 71 percent of all reach miles. Among other options analyzed, the largest
number of reaches affected by the ELGs occurs under Option E. Under Option E, there are 13,537 inland
reaches with AWQI > 0, totaling 20,300 miles. Note that the changes are based on annual average
concentrations and represent changes expected after compliance with the final ELGs. As discussed in Section
1.5.3, the changes are assumed to start in 2021 and remain constant thereafter over the period of analysis
through 2042.
September 29, 2015
4-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-7: Water Quality Improvements from Final ELGs in All Benefiting Reaches
Change in WQI
Number of Inland
Reaches
Percentage of
Potentially
Affected Inland
Reaches
(18,622 Reaches)
Reach Miles
Percentage of
Potentially Affected
Inland Reach Miles
(27,421 Miles)
Option A
AWQI = 0
0
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-7: Water Quality Improvements from Final ELGs in All Benefiting Reaches
Change in WQI
Number of Inland
Reaches
Percentage of
Potentially
Affected Inland
Reaches
(18,622 Reaches)
Reach Miles
Percentage of
Potentially Affected
Inland Reach Miles
(27,421 Miles)
Option E
AWQI = 0
0
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
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.
Using this meta-dataset, EPA then 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. In other words, the 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 improvements and other specifics of the ELGs where possible, and best practices where not
possible. EPA developed two versions of the meta-regression model. 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
uncertainty in the resulting WTP values. The two models differ only in how they account for the magnitude of
the water quality improvements 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 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 final ELGs. 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 et al. (2007), and Rosenberger and Phipps (2007). The benefit
September 29, 2015 4-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
transfer approach uses census block groups (CBGs) as the geographic unit of analysis.44 The transfer approach
involved projecting benefits in each CBG and year, based on the following general benefit function:
Equation 4-2. ln(MWTPYB] = Intercept + ^(coefficient^ x (independent variable value J
Where
ln(MWTPY:B) = The predicted natural log of marginal household WTP for a given year (7)
and CBG (5).
coefficient = A vector of variable coefficients from the meta-regression.
independent = A vector of independent variable values. Variables include baseline water
variable values quality level (WQI-BLYB) and expected water quality under the regulatory
option (WQI-PCY,B) for a given year and CBG.
Here, ln(MWTPY:B) is the dependent variable in the meta-analysis—the log of approximated marginal WTP
per household, in a given CBG B for water quality in a given year Y45 The baseline water quality level (WQI-
BLY,B) and expected water quality under the regulatory option (WQI-PCY:B) were based on water quality at
waterbodies within a 100-mile buffer of the centroid of each CBG. A buffer of 100 miles is consistent with
Viscusi et al. (2008) and with the assumption that the majority of recreational trips will occur within a 2-hour
drive from home. Because marginal WTP is assumed to depend, according to Equation 4-2, on both baseline
water quality level (WQI-BLYiB) and expected water quality under the regulatory option (WQI-PCYiB), EPA
estimated the marginal WTP for water quality improvements resulting from the final ELGs at the mid-point of
the range over which water quality was changed, WQIYiB = (1/2) ( WQI-BLYiB + WQI-PCYiB)).
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 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.46 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 C 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 improvements (i.e., WQI-
PCYtB), the scale of resource improvements relative to the size of the buffer and relative to available
substitutes, the characteristics of surveyed populations (e.g., users, nonusers), and other methodological
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.
45 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.
46 Population double-counting issues can arise when using "distance to waterbody" to assess simultaneous
improvements to many waterbodies.
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
variables. For example, EPA assumed that household income (an independent variable) changes over time,
resulting in household WTP values that vary by year.
Table 4-8 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 intercept, variable coefficients (coefficient) for the
two models, and the corresponding independent variable 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 final ELGs' policy context. EPA assigned a value to each model variable
corresponding with theory, characteristics of the water resources, and sites affected by the final ELGs and the
policy context. 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.
In this instance, EPA assigned six study and methodology variables, (thesis, volunt, nonparam, non_reviewed,
lump_sum, and WTP_mediari) 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 3.0796, 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 (ce) was set to 1 to reflect recent
trends in the use of choice experiments within the environmental valuation literature. Model 2 includes an
additional variable, water quality change (ln_quality_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
improvements resulting from the final ELGs, EPA estimated marginal WTP using two alternative settings of
the ln_quality 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.
For median household income, EPA used CBG-level median household income data from the U.S. Census
2010 (American Community Survey 5-year data) and used a stepwise autoregressive forecasting method to
estimate future annual state level median household income. EPA set the variable nonusers_only to zero
because water quality improvements 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). EPA set the variable river to 1 and multjype to 0 because the analysis
focuses only on rivers and streams. Other waterbody types (e.g., lakes and estuaries) are excluded from the
analysis.
The geospatial variables corresponding to the sampled market and scale of the affected resources (ln_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 18,622 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 18,622 reaches that fall 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 improvements resulting
from the final ELGs, and thus does not vary across affected CBGs.
September 29, 2015 4-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Because data on specific recreational uses of the water resources affected by the final ELGs 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.47 Water quality variables (Q and lnquality_ch) vary
across CBGs and regulatory options based on the magnitude of the reach-length weighted average water
quality improvement at affected resources within the 100-mile buffer of each CBG.
Table 4-8: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable
Type
Study
Methodology
and Year
Variable
intercept
Ce
thesis
Inyear
volunt
outliers trim
nonparam
non reviewed
Coefficient
Model
1
-1.040
0.377
0.866
-0.412
-1.390
-0.367
-0.408
-0.709
Model
2
-6.14
0.423
0.774
-0.5
-1.184
-0.291
-0.39
-0.871
Assigned
Value
1
0
3.0796
0
1
0
0
Explanation
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.
Binary variable indicating that the study is a thesis
or dissertation. Set to zero because studies
published in peer-reviewed journals are preferred.
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 (21.7) 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.
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).
Binary variable indicating that outlier bids were
excluded when estimating WTP. Set to one
because WTP estimates that exclude outlier bids
are preferable.
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.
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.
If a particular recreational use was not specified in the survey instrument, EPA assumed that survey respondents
were thinking of all relevant uses.
September 29, 2015
4-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-8: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable
Type
Region and
Surveyed
Population
Sampled
Market and
Affected
Resource
Variable
lump sum
wtp median
northeast
central
south
nonusers only
Inincome
mult type"
River
swim use
Gamefish
Coefficient
Model
1
0.843
-0.161
1.180
0.561
1.400
-0.586
0.333
-0.827
-0.079
-0.234
0.233
Model
2
0.773
-0.151
0.593
0.726
1.563
-0.54
0.96
-0.63
-0.174
-0.27
-0.01
Assigned
Value
0
0
Varies
Varies
Varies
0
Varies
0
1
0
0
Explanation
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.
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 will
mathematically yield total benefits if the
distribution of WTP is not perfectly symmetrical.
Binary variable indicating that the affected
population is located in a Northeast U.S. state,
defined as ME, NH, VT, MA, RI, CT, and NY. Set
based on the state in which the CBG is located.
Binary variable indicating that the affected
population is located in a Central U.S. state,
defined as OH, MI, IN, IL, WI, MN, IA, MO, ND,
SD, ME, KS, MT, WY, UT, and CO. Set based on
the state in which the CBG is located.
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 MM. Set based on the state
in which the CBG is located.
Dummy 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
improvements for all households, including users
and nonusers.
Natural log of median household income values
assigned separately for each CBG. Varies by year
based on the estimated income growth in future
years.
Binary variable indicating that multiple waterbody
types are affected (e.g., river and lakes). Set to zero
because calculations are based exclusively on
rivers.
Binary variable indicating that rivers are affected.
Set to one because calculations are based
exclusively on stream miles. EPA did not estimate
water quality changes for other waterbody types
(e.g., lakes and estuaries).
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
September 29, 2015
4-16
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-8: Independent Variable Assignments for Surface Water Quality Meta-Analysis
Variable
Type
Water
Quality
Variable
boat use
In ar agr
In ar ratio
sub _proportion
Q
Inquality ch
Coefficient
Model
1
-0.725
-0.271
-0.034
1.100
-0.015
NA
Model
2
-0.32
-0.413
-0.057
0.607
-0.004
-0.746
Assigned
Value
0
Varies
1.238
Varies
Varies
ln(5) or
ln(50)
Explanation
electric plant discharges are not available, set to
zero, which corresponds to all recreational uses.
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
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
In 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). Set to the
mean value from the CBG's with 100-mile buffers
containing waters affected by the final ELGs.
The size of the affected resources relative to
available substitutes. Calculated as the ratio of
affected reaches miles to the total number of reach
miles within the buffer that are the same order(s)
as the affected reaches within the buffer. Its value
can range from 0 to 1 .
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
changes due to the final ELGs, WQIYB =
(l/2)(WQI-BLYf + WQI-PCYtB). Calculated as the
length-weighted average WQI score for all
potentially affected COMIDs within the 100-mile
buffer of each CBG.
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.
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.
4.3 Total WTP for Water Quality Improvements
EPA estimated economic values of water quality improvements at the CBG level. For each block group, EPA
multiplied the coefficient estimates for each variable, taken from meta-analysis results (Table 4-8,
Coefficients: Model 1 and Model 2), by the variable levels calculated for each CBG or fixed at the levels
September 29, 2015
4-17
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
indicated above (Table 4-8, column 4). The sum of these products represents the predicted natural log of
marginal household WTP (In MWTP) for a representative household in each CBG, as indicated by Table 4-8.
Equation 4-3 provides the discount formula used to calculate household benefits for each CBG.
Equation 4-3.
HWTPYiB = MWTPYiB X
where:
HWTPY,B
MWTPY:B
AWQIE
Annual household WTP in 2013$ in year Y for households located in
the CBG (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,
Estimated annual average water quality change for the CBG (B).
As summarized in Table 4-9, average annual household WTP estimates for the final ELGs (Option D) range
from $0.32 on the low end (Model 2) to $1.77 on the high end (Model 2), with a central estimate of $0.35
(Model 1). The average is calculated based on the 84.5 million households affected by Option D. We note that
the central estimate does not fall at the midpoint of the range, but instead represents the value from Model 1
which falls between the low and high bound estimates provided by Model 2.
Table 4-9: Household Willingness-to-Pay for Water Quality Improvements
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Improving Reach
Miles
8,177
13,431
16,417
19,573
20,300
Number of Affected
Households (Millions)
30.3
60.3
67.9
84.5
94.3
Average Annual WTP Per Household (2013$)
Low
$0.17
$0.28
$0.33
$0.32
$0.31
Central
$0.20
$0.37
$0.46
$0.45
$0.45
High
$0.92
$1.58
$1.85
$1.77
$1.75
Source: U.S. EPA Analysis, 2015
To estimate total WTP (TWTP) for water quality improvements for each CBG, EPA multiplied the per-
household WTP values for the estimated water quality improvement by the number of households within each
block group in a given year. EPA then calculated annualized total WTP values for each CBG with both a
3 percent and 7 percent discount rate as shown below in Equation 4-4. As discussed in Chapter 1, benefits
from water quality changes are estimated for all years between 2021 and 2042. For reaches directly receiving
steam electric plant discharges, benefits are expected to begin accruing according to the technology
implementation period of 2019 through 2023 described in Chapter 3 of the RIA (U.S. EPA, 2015b).
Downstream reaches, however, can be affected not only by discharges from a given plant discharging directly
to the reach, but also by any change in discharges from plants located upstream, which may have different
compliance years. Therefore, for this analysis, EPA used a simplified assumption that all benefits begin
accruing in 2021, which is the midpoint of the compliance period.
September 29, 2015
4-18
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Equation 4-4.
2042
TWTPB = ( V
\ t—l
HWTPYiB X HHYiB
'
X
NT=2021
i x (1 + j)
(1 + i)
n+1 -
where:
TWTPB
HWTPy
HHy,B
T
i
n
Total household WTP in 2013$ for households located in the CBG
(B),
Annual household WTP in 2013$ 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 (22 years)48
EPA generated annual household counts for each CBG through the period of analysis based on projected
population growth following the method described in Section 1.5.4. Table 4-10 presents the results for the
3 percent and 7 percent discount rates.
Table 4-10: Total Willingness-to-Pay for Water Quality Improvements
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of Affected
Households
(Millions)
30.3
60.3
67.9
84.5
94.3
3% Discount Rate
Low
$4.2
$15.0
$19.6
$23.2
$25.1
Central
$4.9
$18.9
$26.0
$31.3
$34.0
High
$23.4
$83.7
$109.4
$129.5
$140.0
7% Discount Rate
Low
$3.3
$12.0
$15.7
$18.5
$20.0
Central
$3.9
$15.2
$20.9
$25.1
$27.3
High
$18.6
$66.7
$87.3
$103.4
$111.7
Source: U.S. EPA Analysis, 2015
EPA estimated that 19,573 reach miles would improve under the final ELGs. The total annualized benefits of
water quality improvements resulting from reduced metal, nutrient and sediment discharges in these reaches
range from $23.2 million to $ 129.5 million with a central estimate of $31.3 million using a 3 percent discount
rate and $18.5 million to $103.4 million with a central estimate of $25.1 million using a 7 percent discount
rate.
Readers may wonder why the central estimate is closer to the low end of the range than the high end. EPA
tested several different functional forms for Model 2, and found that the model has the highest explanatory
power (R-squared) when water quality change is included in logged form. This implies that water quality
change has a nonlinear effect on MWTP. In particular, small initial increases in the scale of the water quality
change scenario have a larger effect on MWTP than subsequent increases. This is the reason why the central
estimate of MWTP (based on a water quality change scenario of approximately +20 units) is closer to the low
MWTP estimate (based on a water quality change scenario of+50) than to the high MWTP estimate (based
on a water quality change scenario of+5). In addition, when Model 2 is used in a benefits transfer application
with a water quality change of+20, the mean of the meta-data, the results are very close to the results of
1 See Section 1.5.3 for detail on the period of analysis.
September 29, 2015
4-19
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
Model 1. This sensitivity analysis is included because a water quality change of +5 is closer to the size of
water quality changes projected to result from the ELGs than the +20 analog to the central estimate, while the
+50 represents the upper end of water quality changes in existing surveys (and the lower end of the sensitivity
benefits range), for completeness.
In addition, EPA views its revisions to the meta-analysis employed here as a work in progress. As part of the
revisions, EPA has identified some issues that require further analysis, and intends to continue this work after
the Steam Electric ELGs are promulgated. EPA also intends to have the SAB review the results of that
progress when it has reached a point of completion where SAB review is appropriate. In particular, EPA has
identified the following issues to address: conducting additional robustness tests and cross validation,
investigating model over-fitting, investigating whether a distance decay effect can be gleaned from the meta-
data (to substitute for the 100-mile radius assumption used in the benefits transfer application here), and to
consider employing Bayesian estimation techniques. However, EPA views the results presented here as a
sufficient improvement over the results presented at proposal to warrant their inclusion in this report.
4.4 Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. The revisions affect baseline discharges of the several
pollutants explicitly modeled to quantify and monetize the water quality improvement benefits in this
Chapter: the eight metals, TN, TP and TSS. Changes in baseline metal loadings range from a 47 percent
reduction (for lead) to a 1 percent increase (for selenium) while changes in TN, TP, and TSS loadings are all
less than 1 percent. The overall effect of changes in TN, TP, and TSS loadings is likely to be trivial given the
magnitude of loading reductions (i.e., less than 1 percent). Impacts of the revised metal loadings on the WQI
score is difficult to assess since the score is not based on a continuous scale, but depends instead on whether
the modeled instream concentrations metals exceed the relevant AWQCs. To the extent that baseline
concentrations calculated based on the original loads exceed the relevant AWQCs only slightly, small
reductions in these concentrations due to the revised loads could increase the baseline WQI score, all else
being equal, and therefore reduce the magnitude of improvements attributable to the ELGs.
For several reasons — notably the fact that revisions do not affect all plants equally, the metals subindex
score is based on counts of exceedances, and the model functions are not linear — it would be inappropriate
to simply scale the monetized benefits based on the aggregate changes in loadings. However, comparison of
the initial and revised baseline loads, on the one hand, and the ratios of modeled concentrations based on the
initial loads to AWQCs, on the other hand, provides insight on the potential significance of the revisions on
benefit estimates.
Table 4-11 summarizes changes in loading estimates between the original loads used in estimating benefits
for Option D in this Chapter and the revised loads. The table shows total industry loads of the eight metals
included in the WQI subindex, for the baseline and under the final ELGs (Option D).
September 29, 2015 4-20
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
Table 4-11: Estimated Aggregate Changes in Pollutant Loadings for Metals, Nutrients, and
Suspended Solids.
Pollutant
Arsenic
Cadmium
Chromium (VI)
Lead
Mercury
Nickel
Selenium
Zinc
Total nitrogen
Total phosphorus
Total suspended
solids
Baseline Loadings
Initial
22,219
10,925
119
14,588
1,176
94,201
112,999
145,045
13,134,797
153,972
14,182,480
Revised
20,138
8,289
119
7,674
992
61,933
114,533
124,333
13,134,797
154,519
14,286,431
% Change
-9%
-24%
0%
-47%
-16%
-34%
1%
-14%
0%
0%
1%
Option D
Initial
1,483
636
0
350
31
1,781
3,122
6,951
69,969
31,630
1,661,665
Revised
1,480
630
0
331
31
1,697
3,129
6,893
70,285
31,667
1,665,345
% Change
0%
-1%
0%
-5%
-1%
-5%
0%
-1%
0%
0%
0%
Source: U.S. EPA Analysis, 2015
While EPA cannot readily recalculate the benefits, the direction and relative magnitude of the change in
pollutant removals in Table 4-11 suggest there could be reaches where modeled concentrations would no
longer exceed the applicable AWQS after revising the loadings. The changes are most significant for the
baseline and are most likely to affect benefit estimates in instances where the concentrations were only
slightly above the AWQC. However, review of baseline exceedances shows that, of the 1,023 exceedances
present in the baseline (in 567 reaches), 774 exceedances result from instream concentrations being more than
1.5 times the relevant AWQC and 690 exceedances result from instream concentrations that are more than
twice the AWQC. Instream concentrations would need to decline by at least 33 percent and 50 percent,
respectively, to eliminate such baseline exceedances. Based on this data review, EPA concludes that the
values presented in this Chapter based on the initial loadings provide reasonable insight into the water quality
improvement benefits expected from the final ELGs.
Table 4-12: Number of Reaches with Baseline AWQC Exceedances Based on Initial Loadings.
Ratio of [baseline concentration]/[AWQC]
ruiiuiaiu
Arsenic
Cadmium
Chromium (VI)
Lead
Mercury
Nickel
Selenium
Zinc
>1.0
167
151
10
124
17
216
91
247
>1.5
128
100
7
97
11
147
68
216
>2.0
124
75
5
62
7
138
47
167
Source: U.S. EPA Analysis, 2015
September 29, 2015
4-21
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
4: Non-Market Benefits
4.5 Limitations and Uncertainties
Table 4-13 summarizes the limitations and uncertainties in the analysis of benefits associated with improved
surface water quality and indicates the direction of any potential bias.
Table 4-13: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits
Issue
Effect on Benefits
Estimate
Notes
Revisions to pollutant loads used to estimate water quality improvements
The analysis is based
loadings that were
subsequently revised
by EPA
Overestimate
Revised loadings of parameters included in the WQI (notably the
eight metals), are lower than the loadings used to estimate
benefits. The changes indicate that number of exceedances of
AWQCs may be lower in the baseline, and therefore the resulting
improvements due to the ELGs may be lower. As discussed in
Section 4.4, however, the magnitude of the overstatement may be
small.
Limitations inherent to the meta-analysis model and benefit transfer
Use of 100-mile buffer
for calculating water
quality benefits for
each CBG
Selection of the WQI
parameter value for
estimating low and
high WTP values
Whether potential
hypothetical bias is
present in underlying
stated preference
results
Use of different water
quality measures in
the underlying meta-
data
Underestimate
Uncertain
Uncertain
Uncertain
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 improvements 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.
EPA set AWQI to 5 and 50 units to estimate high and low benefit
values based on 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 water
quality expected from the final ELGs it is likely to significantly
understate the estimated WTP value. AWQI = 5 is more
consistent with the magnitude of water quality changes resulting
from the final ELGs.
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.
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 preliminary
model runs, EPA tested a dummy 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 systematic bias in the mapping of studies that
did not use the WQI.
September 29, 2015
4-22
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 4: Non-Market Benefits
Table 4-13: Limitations and Uncertainties in the Analysis of Nonmarket Water Quality Benefits
Issue
Transfer error
Effect on Benefits
Estimate
Uncertain
Notes
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 al. 2007). This notwithstanding, results
reviewed by Rosenberger and Phipps (2007) are "very
promising" for the performance of meta-analytic benefit transfers
relative to alternative transfer methods.
Use of the WQI to link water quality changes to effects on human uses and support for aquatic and terrestrial
species
Omission of lakes and
estuaries from analysis
of benefits from water
quality improvements
Changes in WQI
reflect only reductions
in metal, nutrient, and
total suspended
sediment
concentrations
In-stream metal
concentrations are
based only on loadings
from steam electric
power generating
plants and other TRI
dischargers
Use of nonlinear
subindex curves
Underestimate
Uncertain
Uncertain
Underestimate
12 percent of steam electric power generating plants discharge to
the Great Lakes or estuaries. Due to limitations of the water
quality models used in the analysis of the final ELGs, these
waterbodies were excluded from the analysis. This omission is
likely to underestimate benefits of water quality improvements
from the final ELGs.
The estimated changes in WQI reflect only water quality
improvements resulting directly from reductions in metal,
nutrient and sediment concentrations. They do not include
improvements in other water quality parameters (e.g., BOD,
dissolved oxygen) that are part of the WQI. If the omitted water
quality parameters also improve, then the analysis underestimates
the expected water quality changes.
In-stream concentrations for heavy metals were estimated based
on loadings from steam electric plant and other TRI dischargers
only and, as a result, do not account for background
concentrations of these pollutants from other sources, such as
contaminated sediments, non-point sources, point sources that are
not required to report to TRI, air deposition, etc. Not including
other contributors to background metal concentrations in the
analysis is likely to result in understatement of baseline
concentrations of these pollutants and therefore of AWQC
exceedances. The overall impact of this limitation on the
estimated WTP for water quality improvement is uncertain but is
expected to be small since the WTP function used in this analysis
is most sensitive to the change in water quality.
The methodology used to translate in-stream sediment and
nutrient concentrations into subindex scores employs nonlinear
transformation curves. Water quality changes that fall outside of
the sensitive part of the transformation curve (i.e., above^elow
the upper/lower bounds, respectively) yield no benefit in the
analysis.
September 29, 2015
4-23
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
5: T&E Species Benefits
5 Impacts and Benefits to Threatened and Endangered Species
5.1
Introduction
Threatened and endangered (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. In many
cases, T&E species are given special protection due to inherent vulnerabilities to habitat modification,
disturbance, or other human impacts. This chapter examines the environmental impacts of steam electric
power plant discharges on T&E species and the benefits associated with improvements resulting from the
final ELGs and other regulatory options.
As described in the EA (U.S. EPA, 2015a), the chemical constituents of steam electric waste streams 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 psychological alterations in aquatic
organisms (U.S. EPA, 2009d; Appendix H). 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 (Williams et al., 2001). 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. Consequently, steam
electric power plant discharges may either lengthen recovery time, or hasten the demise of these species. For
this reason, the final ELGs may have a significant impact on T&E species populations.
From an economic perspective, T&E species affected by steam electric power plant discharges may have both
use and nonuse values. However, given the protected nature of T&E species and the fact that use activities
generally constitute take, which is illegal unless permitted, the majority of T&E species do not have direct
uses, the majority of the economic value for T&E species comes from nonuse values. Species-specific
estimates of nonuse values held for the protection of T&E species can be most accurately derived by primary
research using stated preference techniques. However, the cost, administrative burden, and time required to
develop primary research estimates to value effects of the final rule on T&E species are beyond the schedule
and resources available to EPA for this rulemaking. As an alternative, EPA used a benefit transfer approach
that relies on information from existing studies (U.S. EPA, 2010c).
In this chapter, EPA explores the current status of major freshwater taxa, identifies the extent to which the
final ELGs can be expected to benefit species protected by the Endangered Species Act (ESA), and applies
economic valuation studies to these T&E species to estimate WTP for these benefits.
5.2 Baseline Status of Freshwater Fish Specie
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 et al., 1979; Williams et al., 1989; Williams et al., 1993;
Taylor et al., 1996; Taylor et al., 2007; Jelks et al., 2008). Overall, aquatic species are disproportionately
September 29, 2015 5T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 5: T&E Species Benefits
imperiled relative to terrestrial species. For example, while 39 percent of freshwater and diadromous fish
species (Jelks et al., 2008) are classified as T&E, a similar status review found that only 7 percent of North
American bird and mammal species are currently imperiled (Wilcove and Master, 2005).
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 (Williams, Johnson et al. 1989) and 1979 (Deacon et al., 1979), respectively.
Despite recent conservation efforts, including the listing of several species under the 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 offish have strikingly 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 ranged from a low of
22 percent for Centrarchidae (sunfishes) to ahigh of 61 percent for Salmonidae (salmon) (Jelks et al., 2008).
5.3 T&E Species Affected by the Final ELGs
To assess the potential effects of the final ELGs on T&E species, EPA constructed databases to determine
which species are found in waters expected to improve due to a reduction in pollutant discharge from steam
electric power plants. Notably, these databases exclude all species considered threatened or endangered by
scientific organizations [e.g., the American Fisheries Society (Williams et al., 1993; Taylor et al., 2007; Jelks
et al., 2008)] but not protected by the ESA. These databases allowed EPA to estimate the potential for adverse
impacts of steam electric power plant discharges on T&E species, as well as benefits associated with the final
ELGs.
5.3.1 Identifying T&E Species Potentially Affected by the Final ELGs
To estimate the effects of the final ELGs on T&E species, all affected species must first be identified. EPA
identified all species currently listed or in consideration for listing under the ESA using the U.S. FWS
Environmental Conservation Online System (U.S. FWS, 2014a). Whenever possible, EPA obtained the
geographical distribution of T&E species in geographic information system (GIS) format as polygon (shape)
files, line files (for inhabitants of small creeks and rivers) and as a subset of geodatabase files. Data sources
include U.S. FWS (2014b), the National Oceanic and Atmospheric Administration's (NOAA's) Office of
Response and Restoration (NOAA, 2010), NatureServe (NatureServe, 2014), and NOAA National Marine
Fisheries Service (NMFS, 2014a; NMFS, 2014b; NMFS, 2014c). For several freshwater species, geographic
ranges were available only as 6-digitHUCs (NatureServe, 2014; U.S. FWS, 2014b). For these species, EPA
created GIS data layers using a GIS HUC database obtained from the USGS (Steeves and Nebert, 1994).
To determine the probability that individual T&E species could benefit from the final ELGs, EPA compiled
data on locations of steam electric power plants and receiving waterbodies. The Agency used plant and outfall
coordinates it had obtained through its 2010 Questionnaire for the Steam Electric Power Generating Effluent
Guidelines (the industry survey) and georeferenced these coordinates to waterbodies (see EA for details; U.S.
EPA, 2015a). The result of this analysis consists of the NHD Plus (COMIDs) identifiers of waterbodies that
receive discharges from steam electric power plants and indicators of water quality under the baseline and
each analyzed regulatory option. EPA queried these data to identify "affected areas" as those habitats where
1) receiving waters do not meet water quality metrics recognized to cause harm in organisms under baseline
September 29, 2015 5-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
5: T&E Species Benefits
conditions; and 2) receiving waters exceed water quality metrics under the most stringent regulatory option
EPA analyzed (Option E). EPA used these data in ArcGIS to determine the T&E species with habitat extents
overlapping the affected areas.
EPA constructed a screening database using the spatial data. This database included all T&E species whose
habitat overlaps those waterbodies receiving effluent discharges from steam electric power plants. A buffer of
500 m was chosen when constructing this database to account for any minor errors in outfall location and
habitat maps.
After identifying T&E species potentially affected by the final ELGs, EPA classified the species on the basis
of their vulnerability to changes in water quality. Species were classified as follows:
> High 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.
> Low vulnerability - species whose habitats overlap bodies of water, but whose life history traits and
food sources are terrestrial.
Life history data used to classify species were obtained from a wide variety of sources (Froese and Pauly,
2009; NatureServe, 2014; Alaska Fisheries Science Center (AFSC), 2010; Atlantic States Marine Fisheries
Commission (ASMFC), 2010; Northeast Fisheries Science Center (NEFSC), 2010; Pacific Islands Fisheries
Science Center (PIFSC), 2010a; PIFSC, 201 Ob; Southeast Fisheries Science Center (SEFSC), 2010;
Southwest Fisheries Science Center (SWFSC), 2010; U.S. FWS, 2010).
The results of the spatial analysis and vulnerability classification process (as described above) are presented in
Table 5-1. Appendix I lists all T&E species potentially affected by the final ELGs.
Table 5-1: T&E Species with Habitat Occurring within Waterbodies Affected by Steam Electric Power
Plants
Species Group
Amphibians
Arachnids
Birds
Clams
Crustaceans
Fishes
Insects
Mammals
Reptiles
Snails
Total
Species Vulnerability
Low
0
4
9
0
0
0
10
17
3
8
51
Moderate
2
0
2
0
1
0
2
4
1
0
12
High
2
0
1
33
1
17
1
1
5
14
75
Species Count
4
4
12
33
2
17
13
22
9
22
138
Source: U.S. EPA Analysis, 2015
For the purposes of estimating benefits, EPA excluded all species with low and moderate vulnerability
potentials based upon life history traits. For all species with high potential vulnerability, EPA conducted
further analyses to identify those species likely to be affected by the final ELGs, rather than all species whose
September 29, 2015
5-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 5: T&E Species Benefits
life histories make them vulnerable. High vulnerability species meeting the following criteria were removed
from further consideration:
> Species presumed to be extinct, including those not collected for a minimum of 30 years (e.g.,
Noturus trautmani).
> Endemic species living in waterbodies (e.g., isolated headwaters, natural springs) unlikely to be
affected by steam electric power plant discharges (e.g., Gambusia georgei).
> Species protected by the ESA whose recovery plans i) do not include pollution or water quality issues
as factors preventing recovery, and ii) identify habitat destruction (due to damming, stream
channelization, water impoundments, wetland drainage, etc.) as a primary factor preventing recovery
(e.g., Erimystax cahni).
> Listings due to non-native species introductions and/or hybridization with native or non-native
congeners (e.g., Oncorhynchus clarki somias)
> Listings where water quality issues are identified as the primary issue preventing recovery, but where
a specific industry or entity not within the scope of the final ELGs is identified as the culprit, (e.g.,
Erimystax cahni due to siltation from coal mining activity).
> Species about which very little is known, including geographic distribution.
After eliminating the T&E species meeting these criteria, EPA identified a total of 15 species whose recovery
may be enhanced by the final ELGs.
5.3.2 Estimating Benefits of T&E Species Improvements from the Final ELGs
The final ELGs have the potential to positively affect the recovery trajectory for 15 T&E species. For each of
these species, EPA estimated the magnitude of potential benefits by identifying inhabited waterbodies likely
to meet AWQC for aquatic life as a consequence of the final ELGs and comparing these areas to the overall
area of habitat occupied by T&E species.
First, for each T&E species affected by steam electric power plant discharges, EPA examined water quality in
each of the waterbodies inhabited by each T&E species under baseline conditions, and under conditions
projected to exist following implementation of the final ELGs. For each analyzed regulatory option, EPA
identified waterbodies that 1) do not meet AWQC for wildlife under baseline conditions, but 2) have no
wildlife AWQC exceedances following implementation of the final ELGs. For 11 species, there were no
waterbodies that met these conditions, leaving three T&E fish species and one salamander species in nine
states that may experience increases in population growth rates as a result of the final ELGs (Table 5-2).
Table 5-2: T&E Species Whose Recovery May Benefit from the Final ELGs
Species
Acipenser brevirostrum
Cryptobranchus alleganiensis
Etheostoma chermocki
Etheostoma etowahae
Common Name
Shortnose Sturgeon
Hellbender salamander
Vermilion darter
Etowah darter
State(s)
MD, SC
IL, KY, MO, OH, PA
AL
GA
Source: U.S. EPA Analysis, 2015
EPA did not identify data sufficient to explicitly model population growth rates as a function of water quality
for any of these species. Therefore, to estimate proportionate population increases as a result of the final
ELGs, EPA identified the fraction of inhabited waterbodies that meet wildlife AWQC as a consequence of the
September 29, 2015 5-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 5: T&E Species Benefits
final ELGs. This fraction was used to estimate relative population changes in estimating the WTP for T&E
species recovery.
5.4 Estimating WTP for T&E Species Population Increases
5.4.1 Economic Valuation Methods
For several reasons, it is difficult to estimate the benefits of improving T&E species habitats resulting from
the final ELGs. First, data required to estimate the response of T&E populations to improved habitats are
rarely available. Second, the contribution of T&E species to ecosystem stability, ecosystem function, and life
history remains relatively unknown. Third, much of the wildlife economic literature focuses on commercial
and recreational benefits that are not relevant for many protected species (/'. e., use values) .There is a paucity
of economic data focused on the benefits of preserving habitat for T&E species because nonuse values
comprise the principal source of benefit estimates for most T&E species.
Analysis of nonuse benefits for T&E species affected by changes in pollutant discharges from steam electric
power plants stemming from the final ELGs involves the following two steps: 1) quantifying the impacts of
pollutant discharges from steam electric power plants on T&E species and estimating the change in these
impacts as a consequence of reducing steam electric discharges; and 2) estimating an economic value of
improving T&E habitats and populations as a consequence of the final ELGs.
Benefit transfer involves extrapolating existing estimates of nonmarket values to the policy sites that
potentially differ from the original analytical situation in terms of geographic locations or affected species.
Ideally, the resource in question (i.e., T&E species), policy variables (change in species status, recovery
interval, population size, etc.), and the geographic location and benefitting population (i.e., defined human
population) are identical. Such a match rarely occurs. Despite differences in these variables, however, a
benefit transfer approach can provide useful insights into the social benefits gained by reducing impacts on
T&E species.
5.4.2 Estimating WTP for Improved Protection of T&E Species
To estimate the potential economic values of increased T&E species populations affected by the final ELGs,
EPA used a benefit transfer approach based on a meta-analysis of 31 stated preference studies eliciting WTP
for changes in T&E populations (Richardson and Loomis, 2009). This meta-analysis is based on studies
conducted in the United States that valued threatened, rare, or endangered fish, bird, reptile, or mammal
species. Because the underlying meta-data does not contain amphibian valuation studies, EPA was unable to
monetize any benefits for potential population increases of Hellbender salamander as a result of the final
ELGs. Equation 5-1 contains the estimated WTP equation from the Richardson and Loomis (2009) paper that
EPA used to monetize potential population increases resulting from the final ELGs.
Equation 5-1. (n WTP (2006$) = -153.231 + 0.870 In CHANGESIZE + 1.256VISITOR + 1.020 FISH +
0.772MARINE + 0.826 BIRD - 0.603 In RESPONSERATE + 2.767 CONJOINT +
1.024 CHARISMATIC - 0.903 MAIL + 0.078 STUDYYEAR.
September 29, 2015 5-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
5: T&E Species Benefits
Table 5-3 lists the assigned variable values and definitions used in estimating per household WTP for
improved protection of T&E species resulting from the final ELGs.
Table 5-3: Independent Variable Assignments for the T&E Meta-Regression
Variable
Intercept
In ChangeSize
Visitor
Fish
Marine
Bird
Charismatic
Conjoint
In ResponseRate
Mail
StudyYear
Description
Intercept
Natural log of percentage change in the
population of the species of interest
Dummy variable indicating if survey
respondents are visitors rather than full-
time residents
Dummy variable indicating population
increases for fish species
Dummy variable indicating population
increases for marine mammals
Dummy variable indicating population
increases for bird
Dummy variable indicating a
charismatic species
Dummy variable indicating conjoint
method surveys
Natural log of the survey response rate
Indicates mail surveys
Year of study
Value
-153.231
Varies
0
1
0
0
Varies
0
3.912
0.851
1992
Explanation
-
Log of percentage change in fish
population
Primary beneficiaries are expected to be
full-time state residents
Only freshwater T&E fish species are
expected to be affected
Sturgeon species are considered
charismatic; minnow species are not
Default value from Richardson and
Loomis (2009) as only one underlying
meta-study used conjoint analysis; the rest
were CV studies
Mean value from Richardson and Loomis
(2009) following the Johnston et al. (2006)
approach where values for methodological
attributes are set at mean values from the
metadata
Source: U.S. EPA Analysis, 2015
EPA does not currently have either species-specific estimates of the population effects of the final ELGs or
population models to estimate future population changes for the affected T&E species due to improved
aquatic habitat conditions. In the absence of such estimates, EPA used best professional judgment to assign a
range of potential improvements in the T&E populations based on the expected reductions in AWQC
exceedances under the post-compliance scenario. To estimate total population increases as a result of each
analyzed regulatory option, EPA assumed minimal increases in population size of 0.5, 1, or 1.5 percent. EPA
then weighted these population growth estimates within states by the proportion of reaches used by T&E
species expected to meet wildlife-based AWQC under each option. The natural log of these weighted
population growth estimates under each scenario was used to assign a value to the ChangeSize parameter
estimate. EPA used the approach described in Johnston et al. (2006) and assigned mean study values from
Richardson and Loomis (2009) for the methodological variables (In ResponseRate, Mail, and StudyYear).
Although it could not find published literature to support dose-response relationships for any of the species
assessed or any other numerical estimates of benefit that might occur to T&E species because of the rule, EPA
believes that its low (0.5 percent), medium (1.0 percent) and high (1.5 percent) estimates of population
growth for T&E species (occurring within affected reaches exceeding AWQC at baseline) because of the final
ELG are reasonable. This is because of the high number of species filtered from further assessment (as
described in Sections 5.3.1 and 5.3.2, only four species met all criteria for inclusion), because population
increases were estimated to occur only in reaches meeting AWQC because of the rule (the majority of habitat
September 29, 2015
5-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
5: T&E Species Benefits
used by T&E species assessed meets AWQC at baseline), and because few individuals must be saved to attain
these growth estimates. For example, consider a T&E species with a state-level population of 10,000
individuals (reasonable for threatened species, likely an over-estimate for endangered species and endemic
species), residing in 10 reaches - only one of which does not meet AWQC criteria at baseline (but does under
rule options), and a population growth rate of 0 in the baseline. This species would achieve EPA's low,
medium and high population increases if the ELG results in one fewer premature mortality every 5 years (low
growth) and approximately one fewer premature mortality every 1.5 years (high growth) between 2019 and
2042. This low level of effect needed to meet growth assumptions, when combined with known effects of
wildlife living in areas with poor water quality, make this level of population increase reasonable.
Because population growth was assessed at the state level, EPA was unable to attribute benefits to a specific
steam electric power plant and therefore to account for the timing of benefits based on the assumed control
technology implementation year. EPA assumed that benefits begin accruing in 2021 for all states. This year is
the midpoint of the period of 2019 through 2023 when plants are assumed to implement control technologies
to comply with the revised effluent limits and standards.
For each state, EPA estimated household WTP for improved protection of T&E species resulting from the
final ELGs using Equation 5-1 and the independent variable assignments presented in Table 5-3. EPA
estimated total annual benefits for the years between 2021 and 2042 by multiplying household WTP by the
number of households in each state for a given year. EPA then calculated the value of benefits for each year
and the annualized total WTP values for each state using 3 percent and 7 percent discount rates.
5.5
Results
Table 5-4 presents the annualized total benefits calculated using discount rates of 3 percent and 7 percent. The
monetized benefits to T&E species of Option D are concentrated in two states: Alabama (AL) and Georgia
(GA). The annualized benefits of Option D are $0.02 million for the medium population increase using a
3 percent discount rate.
September 29, 2015
5-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
5: Threatened and Endangered Species Benefits
Table 5-4: Estimated Annualized Benefits to T&E Species from WQ Improvements (Millions 2013$)
a,b
Discount
Rate
3%
7%
State
AL
GA
MD
SC
Total
AL
GA
MD
SC
Total
Option A
Low
<$0.01
<$0.01
$0.00
$0.00
$0.01
<$0.01
<$0.01
$0.00
$0.00
$0.01
Medium
<$0.01
$0.01
$0.00
$0.00
$0.01
<$0.01
$0.01
$0.00
$0.00
$0.01
High
<$0.01
$0.01
$0.00
$0.00
$0.02
<$0.01
$0.01
$0.00
$0.00
$0.01
Option B
Low
<$0.01
<$0.01
$0.00
$0.00
$0.01
<$0.01
<$0.01
$0.00
$0.00
$0.01
Medium
<$0.01
$0.01
$0.00
$0.00
$0.01
<$0.01
$0.01
$0.00
$0.00
$0.01
High
<$0.01
$0.01
$0.00
$0.00
$0.02
<$0.01
$0.01
$0.00
$0.00
$0.01
Option C
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
Option D
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
Option E
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
Source: U.S. EPA Analysis, 2015
September 29, 2015
5-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 5: Threatened and Endangered Species Benefits
Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
••••diseharges after completing the benefit analyses. The revisions affect baseline discharges of several pollutants
with wildlife AWQC exceedances. For several reasons — notably the fact that revisions do not affect all
plants equally, and that species vulnerability to steam electric pollutants is determined based on threshold
effects — it would be inappropriate to simply scale the monetized benefits based on the aggregate changes in
loadings. The impacts of the loading revisions on estimated benefits to E&T specifies may be similar to that
discussed in Section 4.4 for water quality improvement benefits.
5.7 Limitations and Uncertainties
Table 5-5 summarizes the caveats, omissions, biases, and uncertainties known to affect EPA's estimates of
the benefits to T&E species and indicates the direction of the potential bias.
Table 5-5: Limitations and Uncertainties in the Analysis of T&E Species Benefits
Issue
The analysis is based
loadings that were
subsequently revised by EPA
Change in T&E populations
due to the effect of steam
electric ELGs is uncertain
Only those T&E species
listed as threatened or
endangered on the
Endangered Species Act are
included in the analysis
Benefit estimates do not
include monetized values for
potential population increases
in Hellbender salamander
(Cryptobranchus
alleganiensis a)
Effect on
Benefits
Estimate
Overestimate
Uncertain
Underestimate
Underestimate
Notes
Revised loadings are lower than the loadings used to estimate
benefits. The changes indicate that number of exceedances of
AWQCs may be lower in the baseline, and therefore the
improvements resulting from the ELGs may be lower. As
discussed in Section 4.4, however, the magnitude of the
overstatement may be small.
Data necessary to quantitatively estimate population changes are
unavailable. Therefore, EPA used best professional judgment to
assess reasonable changes in T&E populations. Actual effects of
the final rule may be larger or smaller than projected changes in
the population of T&E species assumed in this analysis.
The databases used to estimate benefits to T&E species exclude
all species considered threatened or endangered by scientific
organizations but not protected by the ESA. The magnitude of
the underestimate is likely to be significant, since the proportion
of imperiled fish and mussel species is high (e.g., Jelks et al
2008, Taylor et al 2007).
It is likely that population increases in Cryptobranchus
alleganiensis have value to the public. In addition to bequest,
altruistic, and existence values, salamanders may have aesthetic
or cultural values. Salamanders also provide beneficial
ecological services through indirect biotic control of species
diversity and ecosystem processes, connection of energy and
matter between aquatic and terrestrial landscapes, contributing
to soil dynamics, and providing available stores of energy and
nutrients for tertiary consumers (David and Welsh 2004).
September 29, 2015
5-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 5: Threatened and Endangered Species Benefits
Table 5-5: Limitations and Uncertainties in the Analysis of T&E Species Benefits
Issue
Benefit estimates are likely to
include only a subset of
species that may be affected
Benefit transfer introduces
uncertainties
Ecological roles filled by
T&E species
Overlap between WTP
estimates for T&E species
and the WTP estimates for
improvements in water
quality
WTP estimates do not take
into account possible
substitution for effects for
similar species.
Effect on
Benefits
Estimate
Underestimate
Uncertain
Underestimate
Overestimate
Overestimate
Notes
EPA did not consider species for which water quality was not
listed as an important factor to species recovery. Because water
quality issues may be important to species recovery even if not
listed explicitly in species recovery plans this analysis may omit
species that are likely to benefit from the final ELG.
Value may over- or understate true WTP values (See Section 4
for more details).
WTP values are unlikely to include changes to food-webs and
ecosystem stability as a consequence of the restoration (or loss)
of T&E species.
There may be some overlap between WTP estimates for T&E
species and the WTP estimates for improvements in water
quality because WTP values for improvements in water quality
may inherently include benefits to T&E species. However, none
of the studies in EPA' s meta-analysis of WTP for water quality
improvements specifically mentioned or otherwise prompted
respondents to include benefits to T&E species populations (see
Chapter 4); therefore, any overlap is likely to be minimal.
WTP estimates may be affected by the availability of related
species (Hoehn and Loomis, 1993), however Kahneman and
Knetsch (1992) argue that substitution effects may not apply to
the values associated with endangered species, because their
uniqueness is the essence of their existence value.
September 29, 2015
5-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
6 Benefits from Avoided Impoundment Failures
EPA has promulgated several rules affecting the steam electric industry recently: the Cooling Water Intake
Structures (CWIS) rule for existing facilities (79 FR 48300), the CCR rule (80 FR 21302), and the CPP rule
(FR publication forthcoming). EPA approached analyses associated with each rulemaking carefully. EPA also
recognizes that the steam electric industry complying with three regulations cumulatively in a very short
period of time may choose a different compliance path than assumed in the analyses. The cumulative effect
introduces uncertainty on the compliance path, and thereby on the benefits and costs associated with these
rules.
EPA expects that the operational changes prompted by the final ELGs will cause some plant owners to reduce
their reliance on impoundments to manage coal combustion residuals. The CPP rule is likely to have similar
effect on plant owners' reliance on impoundments. These changes will affect the future probability and/or
magnitude of impoundment failures and the resulting accidental, and sometimes catastrophic, releases of coal
combustion residuals. This rule takes the CPP rule into account in the baseline whereas the CCR rule did not,
though it included the proposed CPP rule as a sensitivity analysis. Because the timing was such that the CCR
rule did not include compliance with CPP in its main analysis, the benefits and costs for the ELGs of
complying with the full set of rules may be higher than reported here, although EPA has done its best to
incorporate the effect of all three rules in this analysis.
Benefits from the reduced risk of impoundment failures include avoided cleanup costs, environmental
damage, and transaction costs. EPA's analysis of the monetary value of avoided impoundment failures is
based on the identification of impoundments that would be affected by each of the regulatory options. EPA
estimated benefits from avoided impoundment failures based on the probability of a release for each
impoundment in a given year, the capacity of the impoundment, and the cost (including cleanup costs, natural
resource damages (NRD) and transaction costs) per gallon of coal combustion residuals slurry spilled.
Benefits are calculated as the difference between expected failure costs for a regulatory option and expected
failure costs under baseline conditions, over the period of 2019 through 2042.
This section describes the methodology and data used to determine the baseline and post-compliance
probability of impoundment failures, assign costs to the releases, and estimate the total present and annualized
values of benefits resulting from the final ELGs.
As described below, the ELG analysis follows an approach similar to that used in the final CCR rule analysis
(U.S. EPA, 2014), but uses ELG-specific data and assumptions that reflect differences in the impoundment
universe and failure probability. For example, the final ELG analysis considers a universe of 1,070
impoundments at 1,080 steam electric plants to which the ELGs apply (see Section 6.1.2), which is larger
than the 735 impoundments analyzed for the CCR rule. Fifty-three of the impoundments in the Steam Electric
universe are projected to close as a result of the CCR rule, leaving 1,017 impoundments in the ELG analysis
baseline. The ELG analysis also uses assumptions that reflect implementation of the CCR rule in the ELG
baseline. Thus, the risk of an impoundment failure applied to big impoundments subject to wall breaches
(0.044 percent; see Section 6.1.1) reflects the residual risk remaining after implementation of the CCR rule, as
compared to the higher historical rate of 0.09 percent used as baseline in the CCR rule analysis.
September 29, 2015
6-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
6.1.1 Baseline Failure Probability and Release Quantity
EPA estimated the future probability of coal combustion residuals releases from impoundments based on
historical trends, accounting for the primary types of releases applicable to different types of impoundments.
The approach builds on the methodology used by EPA Office of Resource Conservation and Recovery
(ORCR) for the analysis of final regulations governing the disposal of CCRin impoundments (i.e., "CCR
rule"; U.S. EPA, 2014).
To determine the frequency of releases, EPA used data from a survey of impoundments conducted in support
of the CCR rule (U.S. EPA, 2012d). The surveyed plants had atotal of 656 CCR impoundments, and the
survey obtained information from owners and operators about impoundment releases between 1999 and 2008.
Two of the survey respondents also provided data for two additional releases that occurred in 1995 and 1998.
In total, the survey provides data for 49 relevant historical releases49 over a total of 6,565 impoundment-year
observations.50
In response to the benefit analysis for the proposed ELGs, EPA received several comments noting that
seepage events, which account for 13 of the 49 release events, are not likely to result in significant cleanup or
other types of costs. EPA excluded these 13 seepage events from the historical release data used to determine
the frequency of relevant releases. The remaining 36 releases are summarized in Appendix J.
EPA sorted the 36 releases into two categories based on the circumstances and cause of the release: wall
breaches (4 releases), which are structural failures of external perimeter embankments, and other releases (32
releases). These other releases include overtopping (8 failures); miscellaneous causes (13 failures) such as
sink hole, stack failure, pump failure, hydraulic dredging pipe failure, liner perforation, internal dike breach
(not perimeter), seal failure, discharge structure disturbed during maintenance, and embankment slough; or
unknown causes (11 releases).
EPA assumes, and the historical release data shows, that the potential for wall breaches exists only for large
impoundments that meet certain structural integrity "factor of safety" design criteria specified in the CCR
rule. These impoundments (labeled big for the purpose of the analysis) are those with:
> Height (impounding elevation) of five feet or more above the upstream toe, and storage volume of 20
acre-feet or more, or
> Height (impounding elevation) of 20 feet or more above the upstream toe.
Impoundments that do not meet either criterion (labeled small) are assumed not susceptible to failure by wall
breach, but susceptible only to other types of releases. Based on these criteria, the ORCR survey (U.S. EPA,
2012d) includes 444 big impoundments and 212 small impoundments. EPA calculated the frequency of
releases based on the 36 historic releases for each type of failure (wall breach vs. other) and type of
impoundment (big vs. small).
49 This differs from the risk of failure estimated for the proposed ELG. For the proposed rule analysis, EPA had
counted 42 CCR pond damage cases rather than the current count of 49 cases. EPA corrected the survey
database for nine cases that had been entered as single incidents but were in fact separate incidents, leading to
an interim total of 52 cases. Two of these 52 cases did not involve coal combustion residuals ponds (they
involved a metal cleaning waste basin and a coal pile sump) and one case did not involve the release of
impounded coal combustion residuals material (it involved an oil and grease exceedance due to servicing of a
pump located near a coal combustion residuals pond).
50 The survey responses provided 10 years of data for 656 impoundments, an additional 4 years of data for one
impoundment, and an additional year of data for another impoundment, for a total of 6,565 observations (656 x
10+1x4 + 1x1).
September 29, 2015 6-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
The probability of a release is assumed to be uniform over time and across all impoundments within a
category (big or small), irrespective of other impoundment characteristics such as age, amount of coal
combustion residuals managed, etc. In practice, the probability of a release may depend on impoundment
characteristics and could therefore change as a result of the final rule. However, EPA did not have sufficient
data to model the probability as a function of impoundment characteristics.
The impoundment survey conducted by ORCR (see U.S. EPA, 2012d) provides data on the release volume
and impoundment capacity for 17 of the 36 documented releases. For each type of release and impoundment
category, EPA calculated a "capacity factor" that represents the ratio of gallons of coal combustion residuals
released compared to the design capacity of the impoundment involved in the release.
Table 6-1 summarizes the historical probability of impoundment release and the capacity factor by release and
impoundment type.
Table 6-1: Probabilities of CCR impoundment releases, based on analysis of 49 historical release
events 1995-2008
Impoundment Type - Big
(at least 5 feet high AND at least 20 acre-feet OR at least 20 feet high independent of volume)
Impoundments with observations for 10 years
Additional observations (impoundment-years)
Total number of observations
444
4
4,444
Release Type
Wall Breach
Other
Number of releases
Probability of a release - number of releases per impoundment per year
Capacity factor - volume released as percent of capacity
4
0.09%
27.42%
26
0.59%
2.91%
Impoundment Type - Small
(Does not fit "big" criteria)
Impoundments with observations for 10 years
Additional observations (impoundment-years)
Total number of observations
212
1
2,121
Release Type
Wall Breaches
Other
Number of releases
Probability of a release - number of releases per impoundment per year
Capacity factor - volume released as percent of capacity
Not applicable3
Not applicable3
Not applicable3
6
0.28%
0.41%
a. Small impoundments only incurred "other" releases (no wall breaches).
Source: U.S. EPA Analysis, 2015
EPA adjusted the release probabilities shown in Table 6-1 to account for implementation of the final CCR
rule.51 The rule establishes minimum national criteria for the storage of CCR in surface impoundments at
coal-fired electric utility plants, regulating CCR as a non-hazardous waste. It requires standardized pollution
control measures as well as monitoring and corrective actions in the event of a leak. Additionally, it subjects
all impoundments to location restrictions, design and operating conditions, groundwater monitoring, closure
requirements, and post-closure care. For more details on the CCR rule, see U.S. EPA (2014).
As detailed in U.S. EPA (2014), some plant owners may close their existing impoundments in response to the
CCR rule, effectively eliminating the probability of a future release from these impoundments. For
The CCR rule is not expected to affect the capacity factors in the event of a release.
September 29, 2015
6-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
impoundments that remain operational, the CCR rule is expected to reduce the release rate through the
implementation of structural integrity programs, including regular inspections and other safeguards. For these
impoundments, EPA estimated that the CCR rule would reduce the probability of releases from big
impoundments to 0.044 percent annually for wall breaches and to 0.16 percent for other releases (compared
with 0.09 percent and 0.59 percent, respectively, in Table 6-1). For small impoundments, EPA expects that
the CCR rule will reduce the probability of a release from 0.28 percent (Table 6-1) to 0.07 percent. See U.S.
EPA (2014) for more detail on these release probability assumptions.
Table 6-2 summarizes the release probability and capacity factor assumptions used for big and small
impoundments in this analysis, taking the effects of the CCR rule into account.
Table 6-2: Release Probability and Capacity Factor Assumptions for the ELGs
Type
Big
Small
Annual Probability of Release
Wall Breach
0.044%
NA
Other Release
0.16%
0.07%
Capacity Factor
Wall Breach
27.42%
NA
Other Release
2.91%
0.41%
Source: U.S. EPA Analysis, 2015
6.1.2 Effects of the ELGs
The 1,080 steam electric power plants subject to the Steam Electric ELGs have a total of 1,070
impoundments. Following implementation of the CCR rule, EPA expects 1,017 of these impoundments to
continue operation as of 2023. As discussed in Section 1.3, EPA included the effects of the CPP rule in the
baseline for the ELG analysis. EPA projects that the CPP rule will result in some generating unit ceasing
operation (retiring) or converting, which will also affect impoundments that receive wastestream from the
units. Due to the uncertainty in plant owners' decision to continue to operate impoundments at plants where
some but not all generating units are converted or retire, EPA estimated the impacts of the CPP in two ways:
> Low bound estimate: EPA adjusted impoundment capacity only for those plants where all units that
send wastewater streams to an impoundment retire as a result of the CPP rule.
> High bound estimate: EPA adjusted impoundment capacity considering the share of the generating
capacity of units that retire as a result of the CPP rule.
This CPP-adjusted universe of impoundments (883 to 925 impoundments) represents the baseline for the ELG
analysis, i.e., EPA evaluates the incremental impacts of the ELG on those impoundments EPA determined
would handle coal combustion residuals wastestreams by the time the plant would comply with the ELGs
after accounting for both the CCR and CPP rules.
EPA categorized these impoundments as big or small following the criteria outlined in Section 6.1.1. EPA
used the Steam Electric Industry survey (U.S. EPA, 2010c) as the primary source of impoundment size data,
supplemented by information collected in the CCR survey (see U.S. EPA 2012d) for impoundments that are
common to both data sets.52 Table 6-3 summarizes the results.
52 To categorize each of the impoundments as big or small, EPA imputed impoundment height based on the
volume, difference in elevation, and the reported maximum berm height from U.S. EPA (2010b). Together with
this height data, EPA used the reported capacity to categorize impoundments that are clearly either big or small.
EPA used ORCR impoundment data (from U.S. EPA 2010c) to categorize an additional 69 impoundments
where there was clear overlap between the two impoundment data sets. EPA categorized an additional 187
impoundments as big based on at least one height indicator being over 20 feet or at least one height indicator
September 29, 2015 &4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-3: Steam Electric Impoundments by Size in ELG Baseline
Type
Big
Small
Total
Number of Impoundments
Count
Low Bound
632
251
883
High Bound
668
257
925
Percent of
Total
72%
28%
100%
Impoundment Capacity
Total (million gallons)
Low Bound
687,758
34,336
722,094
High Bound
713,317
34,348
747,665
Percent of
Total
95%
5%
100%
Source: U.S. EPA Analysis, 2015
For each of these impoundments, EPA estimated the expected number of magnitude of releases for each year
between 2019 and 2042, and estimated the expected costs associated with the releases. EPA did this
calculation for the baseline and under each of the five regulatory options.
Specifically, EPA used the probability of release established based on the historic data and revised to account
for implementation of the CCR rule (Table 6-2) to estimate the expected number of releases from each of the
steam electric power plant impoundments in the baseline. EPA used the applicable capacity factor to estimate
the volume involved in a release from each impoundment. For example, EPA assumes that an "other" release
from a big impoundment involves a volume of coal combustion residuals equal to 2.91 percent of the
impoundment's capacity. Figure 6-1 illustrates assumptions used for the baseline scenario.
The Steam Electric ELG is anticipated to change how plant owners or operator handle their coal combustion
residuals, potentially reducing the quantity of FGD solids or ash managed using impoundments. This lower
coal combustion residuals volume is in turn expected to reduce the amount of coal combustion residuals
accumulating in the impoundments in any given year. For this analysis, EPA assumed that the amount of coal
combustion residuals released in the event of a wall breach or other release is reduced in proportion to the
reduction in the amount of coal combustion residuals handled by the impoundment. Thus, for each scenario,
EPA assumed that an impoundment has the same expected number of releases post-compliance, but
calculated the volume released by multiplying the capacity factor by an adjusted volume that reflects the
reduction in the amount of coal combustion residuals handled wet (see U.S. EPA, 2015g; DCN SE05831(CBI
version / DCN SE05832 (non-CBI version) for the calculations). EPA assumed that the reduction in volume
occurs in the same year as the technology implementation year assumed in other parts of the analysis (see
Section 1.5.3).
being over 5 feet together with capacity over 20 acre-feet. For impoundments for which no height data was
provided, EPA categorized impoundments based on capacity only.
September 29, 2015
6-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Figure 6-1: Baseline Release Probability and Capacity Factor Assumptions
Impoundment
Type
(289)
Big
(776)
I
?!
/
/
\
\
\.
ii
,
\
\
\
Close
(4)
f'n.ntiniia +f\ f\ne*rt\+t*
(285)
Close
(49)
Continue to Operate
(732)
-^
/>
/
/
X
s
:i
/'
/
x
\
ii
NA
(4)
Affected by CPP1
(28 - 34)
^
(251 -257)
NA
(49)
Affected by CPP1
(64 - 100)
(632 - 668)
^^^
t^^
^^.
^^^
•z^
^^^^
Annual Release
Probability
D=0
Wall Breach
p=0
Other Release
p=0.065%
All Releases
n-Q
Wall Breach
p=0.044%
Other Release
p=0.16%
-^
^
'
Capacity
Factor
0.41%
27.42%
2.91%
1 Changes in the risk of impoundment failure in impoundments affected by the CPP rule are not attributable to the ELGs and do not
generate benefits attributable to the ELGs in this analysis.
6.1.3 Costs of a Release
The following sections discuss three categories of costs associated with impoundment releases: cleanup,
NRD, and transaction costs. All dollar values are presented in year 2013 dollars.53
6.7.3.7 Cleanup Costs
EPA estimated per-gallon cleanup costs based on five historical impoundment failures. The average unit cost
associated with these historical incidents is $1.35 per gallon released (see Table 6-4). Details on the five
incidents are provided below and reflect additional research EPA conducted since the proposed ELG analysis.
As needed, costs were updated to 2013 dollars using the Construction Cost Index from the Engineering News
Report (unless otherwise indicated).
September 29, 2015
6-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-4: Documented Cleanup Costs from Impoundment Releases
Incident
Oak Creek
TVA Widows Creek
Martins Creek
Massey
TVA Kingston
Average
Volume Spilled
(million gallons)
5
6
100
230
1,100
288
Cleanup Cost
(millions; 2013$)
$13
$10
$62
$89
$1,379
$311
Unit Cleanup
Cost (2013$)
$2.83
$1.64
$0.62
$0.39
$1.25
$1.35
The Oak Creek release occurred in October 2011 when a bluff collapsed near an ongoing construction project
at the Wisconsin Energy Oak Creek power plant, releasing 22,720 cubic yards (or 4.6 million gallons) of coal
ash into Lake Michigan. The collapse also carried "debris from the construction worksite, including vehicles,
heavy machinery, a filter press, a frac tank and four miscellaneous conex boxes filled with unknown amounts
of miscellaneous equipment, down the bluff and into Lake Michigan." (Wisconsin Department of Justice
(DOJ), 2013). Wisconsin Energy's immediate response to the spill included the placement of 1,500 feet of
containment berms and booms and a geotechnical analysis of the stability of the bluff (Wisconsin DOJ, 2013;
Jones and Behm, 2011). Additional response actions included excavation and removal of the material that had
been spilled, with cleanup being complete by the end of November 2011 (Wisconsin DOJ, 2013). According
to the U.S. EPA On-Scene Coordinator (U.S. EPA OSC, 2012), Wisconsin Energy reported spending
$12.1 million on cleanup and restoration. After updating to 2013$ (for a total of $13 million), unit costs are
$2.83 per gallon spilled.
The Tennessee Valley Authority (TVA) Widows Creek spill involved the release of 6.1 million gallons of
gypsum, water, and fly ash. Updated to 2013$, cleanup costs for dredging of the creek and other activities
were approximately $10 million (TVA, 2009), or $1.64 per gallon spilled.
The Martins Creek release involved the discharge of 100 million gallons of slurry over the course of 3 days,
resulting from the failure of a wooden stop log. In its 2006 annual report (PPL Corporation, 2006), PPL
Corporation estimated that the costs of the remediation effort were $48 million. After updating the revised
costs to 2013$ (for atotal cost of $62 million), the unit cost of this spill is $0.62 per gallon spilled.
The Massey Coal slurry spill involved the collapse of a 2,000 acre-foot (651.7 million gallon) surface
impoundment on top of an idled underground mine. The Massey Energy 2002 annual financial report (10-K;
Massey Energy, 2002) states that the release involved 230 million gallons and that Massey incurred atotal of
$58.3 million of cleanup costs in connection with the spill. Updated to 2013$ (for a total cost of $89 million),
the unit cost for this spill is $0.39 per gallon spilled.
The TVA Kingston release involved approximately 1.1 billion gallons of slurry being released from an ash
pond onto 300 acres, primarily the Watts Bar Reservoir and shoreline property. The spill also damaged three
homes, interrupted utility services, and blocked a local road (TVA, 2009). Cleanup activities included ash
dredging and processing, ash disposition, infrastructure repair, dredge cell repair, dike reinforcement,
construction of temporary ash storage basins, and others. Updated to 2013$, the cleanup costs were
approximately $1.38 billion (TVA, 2010; 2011), or $1.25 per gallon spilled.
In early February 2014, a coal ash release occurred at the Dan River power plant operated by Duke Energy in
North Carolina. The release of 24 to 27 million gallons of ash and ash pond water was caused by a break in a
48-inch stormwater pipe beneath the ash basin (Duke Energy, 2014). Cleanup involved the removal of 2,500
September 29, 2015
6-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
tons of ash and contaminated sediment that settled against a dam, and an additional 500 tons that settled in
other parts of the river and municipal water treatment settling tanks (Associated Press, 2014).
Duke Energy has stated that company investors and insurers will pay for the cleanup rather than ratepayers
(Duke Energy, 2014). Duke Energy's second quarterly financial report in 2014 reports that the company spent
approximately $20 million in repairs and remediation related to the spill through June 2014, and completed
cleanup in July. Additionally, according to a plea agreement signed by Duke Energy subsidiaries on May 14,
2015, Duke Energy agreed to spend $34 million for environmental damages, including a $24 million payment
to the National Fish and Wildlife Foundation to benefit riparian areas of North Carolina and Virginia, plus
$10 million in wetland mitigation bank credits. Total cleanup costs, transaction costs, and NRD were not
available at the time that EPA conducted this analysis, however, and this incident was therefore not included
in the cleanup costs shown in Table 6-4, nor used in this analysis.
6.7.3.2 Natural Resource Damages
Israel (2006) provides a detailed state-by-state summary of NRD programs, including some prominent cases
(arising from oil spills, chemical spills, and other incidents) in each state. Israel (2013) provides an updated
version of the accounting with additional cases identified since the 2006 study. Appendix J lists the 137 NRD
cases from Israel (2006; 2013) that provide quantitative estimates of NRD restoration and compensation
costs.54
Releases resulting from impoundment failures may affect resources similar to those that were affected by
some of the NRD cases identified in Israel (2006; 2013) and therefore would be expected to have similar
NRD costs. Of the 137 NRD settlements identified in Israel (2006; 2013), EPA identified 65 cases that are
relevant to this analysis based on the resources affected and the general circumstances of the releases. In
particular, EPA excluded as potentially less relevant settlements for NRD, which, based on the description
provided in Israel, involved damage to groundwater only or to ocean/coastal resources, or resulted from
legacy pollution associated with Superfund sites.
Table 6-5 shows summary statistics for the 65 NRD settlements EPA retained as relevant to this analysis,
compared to the summary statistics for the cleanup costs in Table 6-4. As shown in the table, the mean NRD
settlement is approximately $13.7 million, or 4 percent of the mean total cleanup cost.
54 NRD does not include cleanup costs (or legal and transaction costs, if reported) but includes only the resource
restoration and compensation values. For example, in one case, Israel (2006) reported that "In total, the State's
claim was $764 million, $342 million of which was restoration cost damages, $410 million of which was
compensable value damages, and $ 12 million of which was assessment and legal costs." For this case, EPA
used the sum of $342 million and $410 million as NRD (i.e., excluded assessment and legal costs). EPA used
values for individual cases discussed in the study, focusing exclusively on NRD and excluding or subtracting
assessment costs when those costs were reported separately. EPA also excluded NRD values that represent the
aggregate of several cases when it was not possible to discern NRD for individual cases.
September 29, 2015 6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-5: NRD Settlements Summary Statistics
Minimum
Median
Mean
Maximum
Natural Resource Damages
(2013$)a
$44,000
$1,965,000
$13,723,000
$439,545,000
Cleanup Costs (2013$)b
$10,000,000
$62,000,000
$311,000,000
$1,379,000,000
NRD as a Percent of
Cleanup Costs
0%
3%
4%
32%
a. Based on Israel (2006; 2013); updated using changes in the GDP. If a year is not provided for a case, EPA updated the value
based on the year of the study that identified the case (i.e., 2006 or 2013).
b. Based on cleanup costs in Table 6-4.
To estimate expected NRD costs for future impoundment releases, EPA assumed that the NRD varies
depending on the magnitude of the release in proportion to cleanup costs. To derive a per-gallon estimate of
the expected NRD value for future releases, EPA divided the mean NRD settlement of $13.7 million by the
mean cleanup cost ($311 million). This calculation yields an NRD estimate of 4 percent of cleanup costs, or
$0.05 per gallon spilled (4 percent of $1.35 /gallon).
Note that this estimate provides an approximate value for NRD resulting from impoundment failures as a
function of the volume of coal combustion residuals released, and is appropriate for analyses that look at
NRD over a range of locations and circumstances. EPA expects that actual damages from any given release
would be highly location- and release-specific.
6.1.3.3 Transaction Costs
For this analysis, transaction costs include the costs associated with negotiating NRD, determining
responsibility among potentially responsible parties, and litigating details regarding settlements and
remediation.55 EPA estimated transaction costs based on data showing transaction costs as a share of total
cleanup costs at Superfund sites and the share of spending that represents total transaction costs. Table 6-6
shows the data sources.
Table 6-6: Studies Summarizing Transaction Costs as a Share of Superfund
Spending (for potentially responsible parties)
Acton (1995)
Acton and Dixon (1992)
Dixon, etal. (1993)
Stemhardt et al. (1994)
Average
Multiplier (transaction costs as a share of cleanup cost)3
27%
17%
32%
33%
27%
37%
a. Multiplier is calculated as Average/(l - Average)
These data indicate that, on average, transaction costs represent 27 percent of total costs. As the purpose is to
estimate unit costs per gallon spilled, transactions costs represent 37 percent of the subset of costs that are
These activities involve services, whether performed by the complying entity or other parties, that EPA expects
would be required in the absence of this final rule in the event of an impoundment failure. Accordingly, it is
appropriate to account for the avoided resource cost of these services as social benefits in the benefit-cost
analysis. Note that the transaction costs do not include fines, cleanup costs, damages, or other costs that
constitute transfers or are already accounted for in the other categories analyzed separately.
September 29, 2015
6-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
cleanup costs, calculated as 0.277(1-0.27). Therefore, the estimated transaction costs per gallon of coal
combustion residuals slurry spilled are $0.50 (37 percent of $1.35/gallon).
6.1.3.4 Total Release Costs
Table 6-7 summarizes unit costs for cleanup costs, NRD, and transaction costs. Total impoundment release
costs are $1.90 per gallon spilled.56
Table 6-7: Unit Costs for Impoundment Releases (2013$)
Cost Component
Cleanup costs
NRD
Transaction costs
Total costs
Unit Cost ($/gallon spilled)
$1.35
$0.05
$0.50
$1.90
6.2 Results
The final ELGs will provide benefits by reducing the impact of releases from impoundments that are expected
to see reduced utilization as a result of the final rule, but would continue to operate in the absence of the final
rule. For each of the impoundments included in the analysis, EPA calculated the difference between the
annualized costs from future expected failures under the baseline and each analyzed regulatory option. The
calculation involves the following steps:
> Multiplying the release rate for a given impoundment and year by the capacity factor, impoundment
volume (adjusted if needed to reflect the effects of the ELG), and total unit costs for the release
(including cleanup, NRD, and transaction costs);
> For each regulatory scenario, subtracting the costs of expected releases from the cost of expected
releases under the baseline; and
> Discounting for future years, aggregating across the analysis time horizon (2019 to 2042), and
annualizing over a 24-year period using rates of 3 percent and 7 percent.
Table 6-8 shows the total number of impoundment failures estimated over the period of 2019-2042 in the
baseline and under each of the five regulatory options. Expected failures are reported as a range to reflect the
range of the anticipated effects of the CPP rule discussed in Section 6.1.2. These values reflect the number
and types of impoundments (big and small) and the associated failure probability for each type of failure (wall
breach and other). Table 6-9 shows the estimated volume of coal combustion residuals released annually in
these expected failures, after full compliance by all steam electric power generating plants. Estimates of the
coal combustion residuals volume released reflect the size of the impoundments involved in the expected
failures and the capacity factor for each type of failure.
The number of failures estimated in the ELG baseline (approximately 8 wall breaches and 33 other failures
over 24 years) is consistent with the post-compliance failures estimated in the final CCR rule analysis (15
wall breaches and 91 other releases over 100 years; see U.S. EPA 2014), given differences in the
impoundment universe noted in the introduction to this Chapter. Similarly, the number of avoided failures
estimated to result from implementation of the final ELGs (Option D), which is approximately 2 wall
56 In the economic analysis of the proposed rule, EPA capped release costs for any single incident at $1.3 billion
based on the total estimated cleanup costs of the TVA Kingston spill. However, EPA removed this cap for the
final rule analysis, since future incidents could feasibly exceed the damages associated with Kingston.
September 29, 2015 (M
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
breaches and 7 other releases over 24 years, seems consistent with the number of avoided failures estimated
for the final CCR rule (32 wall breaches and 429 other releases over 100 years; see U.S. EPA 2014), given the
differences in the universe of impoundments and in assumptions regarding the failure rate for the ELG
analysis.
Table 6-8: Total Expected Number of Releases in 2019 through 2042, by Failure Type
Regulatory
Option
Expected Number of Failures
Wall Breaches
Other Releases
Reduction in Expected Number of Failures
Wall Breaches
Other Releases
Low Bound Estimate of CPP Effects
Baseline
Option A
Option B
Option C
Option D
Option E
7.7
7.3
7.3
6.7
6.1
6.1
32.6
30.8
30.8
28.1
25.7
25.7
0.0
0.4
0.4
1.0
1.6
1.6
0.0
1.7
1.7
4.5
6.9
6.9
High Bound Estimate of CPP Effects
Baseline
Option A
Option B
Option C
Option D
Option E
7.7
7.2
7.2
6.5
5.8
5.8
32.6
30.4
30.4
27.4
24.4
24.4
0.0
0.5
0.5
1.2
1.9
1.9
0.0
2.1
2.1
5.2
8.1
8.1
Source: U.S. EPA Analysis, 2015
September 29, 2015
6-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-9: Expected Total Coal Combustion Residuals Volume Released Annually, by Failure Type
(Million Gallons)
Regulatory
Option
Estimated CCR Volume Released
Wall
Breaches
Other
Releases
All Releases
Estimated Reduction in CCR Volume
Released"
Wall
Breaches
Other
Releases
All Releases
Low Bound Estimate of CPP Effects
Baseline
Option A
Option B
Option C
Option D
Option E
125
115
115
88
82
82
48
45
45
34
32
32
173
160
160
122
114
114
0
9
9
37
42
42
0
4
4
14
16
16
0
13
13
51
59
59
High Bound Estimate of CPP Effects
Baseline
Option A
Option B
Option C
Option D
Option E
125
114
114
86
79
79
48
44
44
33
31
31
173
158
158
119
110
110
0
10
10
39
45
45
0
4-4
4-4
14-15
16-18
16-18
0
14
14
54
63
63
Source: U.S. EPA Analysis, 2015
a. The values reflect reductions in the volume of coal combustion residuals released in expected failures for years 2023-2042, after
compliance by all steam electric power generating plants. Reductions in years 2019-2022 are less than reported in this table due to
the assumed distribution of permit renewals and control technologies implementation over the period of 2019 through 2023.
Table 6-10 shows the total benefits of avoided impoundment failures, calculated as the sum of the avoided
release costs across all impoundments expected to be affected by the final rule under each analyzed regulatory
option. The range of benefits reflects the range of the anticipated effects of the CPP rule discussed in Section
6.1.2. For each option, avoided wall breaches account for 72 percent of benefits, while other releases account
for the remaining 28 percent.
September 29, 2015
6-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-10: Estimated Annualized Benefits of Avoided Impoundment Failures by Release Type
(Millions; 2013$)a
Discount
Rate
3%
7%
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Option A
Option B
Option C
Option D
Option E
Wall Breaches
Low Bound
$14.5
$14.5
$59.3
$68.8
$68.8
$11.6
$11.6
$47.9
$55.9
$55.9
High Bound
$16.5
$16.5
$62.7
$74.1
$74.1
$13.3
$13.3
$50.7
$60.3
$60.3
Other Releases
Low Bound
$5.6
$5.6
$22.9
$26.7
$26.7
$4.5
$4.5
$18.5
$21.7
$21.7
High Bound
$6.4
$6.4
$24.2
$28.8
$28.8
$5.1
$5.1
$19.6
$23.4
$23.4
All Releases
Low Bound
$20.1
$20.1
$82.2
$95.6
$95.6
$16.1
$16.1
$66.4
$77.7
$77.7
High Bound
$22.9
$22.9
$86.9
$102.9
$102.9
$18.4
$18.4
$70.3
$83.7
$83.7
Source: U.S. EPA Analysis, 2015
a. Baseline value of total failure costs minus option value of total failure costs.
6.3 Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. Estimated benefits from avoided impoundment failures do
not depend on pollutant loadings and therefore the monetized benefit estimates provided in this Chapter are
unaffected by revisions to steam electric plant loading estimates.
6.4 Limitations and Uncertainties
Table 6-11 summarizes the limitations and uncertainties in the analysis of benefits associated with reduced
impoundment failures arising from the final rule. The methodologies used in this analysis involve several
simplifications and sources of limitations and uncertainties, as described below. These uncertainties add to the
limitations and uncertainties inherited from the EA analysis and data (see U.S. EPA, 2015a). Whether these
limitations and uncertainties, taken together, are likely to result in an understatement or overstatement of the
estimated benefits is not known.
September 29, 2015
6-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 6: Benefits from Avoided Impoundment Failures
Table 6-11: Limitations and Uncertainties in Analysis of Avoided Risk of Impoundment Failure
Benefits
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
The analysis assumes that, in the
absence of the final rule, all
impoundments would continue to
operate in the baseline during the entire
period of analysis.
Overestimate
Plant owners may close existing impoundments or
make other changes to their operations that would
reduce the baseline probability of failure. Not
accounting for these baseline conditions may
overstate the benefits of the final rule.
The analysis accounts for projected
closures of steam electric plant
impoundments due to the CCR rule as of
2023 (53 impoundments). The analysis
does not account for additional projected
closures in 2024-2042 due to the CCR
rule.
Overestimate
To the extent that additional impoundments are
projected to close due to the CCR rule after
implementation of the ELGs and EPA estimated
benefits for these closures as part of the CCR rule
analysis, EPA may be overstating the incremental
benefits of the ELGs. The magnitude of
overstatement is unknown but is expected to be small
given the relatively few closures projected overall in
the CCR rule analysis (98 impoundments close by
2114 out of the total of 735 impoundments analyzed;
see U.S. EPA 2014), the timing of these projected
future closures, and the effects of discounting on
annualized benefit estimates. Furthermore, EPA's
adjustment to the ELG benefit analysis to reflect
impoundment closures due to the CCR rule is
consistent with the adjustment to ELG compliance
costs.
EPA estimated expected future
impoundment releases from big and
small impoundments based on uniform
release rates for wall breaches and other
releases. In practice, the probability of
failure may depend on impoundment
characteristics and management
practices.
Uncertain
Using a uniform failure rate may understate benefits
of the final rule. Conversely, the historical failure
rate may overstate projected failures under baseline
conditions by not reflecting the effects of any recent
changes in impoundment management practices
(e.g., revised inspection and monitoring programs).
The analysis uses a uniform cost of
$1.90 per gallon spilled, including
cleanup costs, natural resource damages,
and transaction costs.
Uncertain
There is significant uncertainty involved in
estimating the costs of unknown future release
incidents, and these estimates are based on a small
and highly variable sample of historic releases. The
costs of future releases may be substantially higher
or lower than the cost estimates applied in this
analysis, depending on site-specific factors,
including the ecosystems, infrastructure, and other
resources damaged by the release.
September 29, 2015
6-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 7: Air-Related Benefits
Air-Related Benefits
The final rule is expected to affect air pollution through three main mechanisms: 1) additional auxiliary
electricity use by steam electric power plants to operate wastewater treatment, ash handling, and other
systems needed to comply with the new effluent limits and standards; 2) additional transportation-related
emissions due to the increased trucking of CCR 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 for the ELGs. The different profile of generation can result in lower or higher air pollutant
emissions due to differences in emission factors. Thus, small reductions in coal-based electricity generation as
a result of the ELGs are compensated by increases in generation using other fuels or energy sources -
biomass, landfill gas, natural gas, nuclear power, oil, and wind power. For example, as detailed in Section
10.6 of the RIA (U.S. EPA, 2015c), IPM projects a 0.3 percent decline in electricity generation from coal
(3,276 GWh), as a result of the final ELGs (Option D); this decline is offset by a 0.1 percent increase in
natural gas generation (1,964 GWh) and additional increases in electricity generation from waste coal, wind,
and biomass. The changes in air emissions reflect the differences in emissions factors for these other fuels or
sources of energy, as compared to coal.
In this analysis, EPA estimated the human health and other benefits resulting from net changes in emissions
of three pollutants: NOX, SO2, and CO2.
NOX and SOX (which include SO2 emissions quantified in this analysis) are known precursors to fine particles
(PM2 5) air pollution, a criteria air pollutant that has been associated with a variety of adverse health effects -
most notably, premature mortality.57 In addition, in the presence of sunlight, NOX and VOCs can undergo a
chemical reaction in the atmosphere to form ozone. Depending on localized concentrations of volatile organic
compounds (VOCs), reducing NOX emissions would also reduce human exposure to ozone and the incidence
of ozone-related health effects. Reducing emissions of SO2 and NOx would also reduce ambient exposure to
SO2 and NO2, respectively. For the purpose of this analysis, EPA quantified only those benefits from
associated reductions PM2 5.58
CO2 is an important greenhouse gas that is linked to climate change effects, including: an increase in
temperature; sea level rise; changes in weather patterns toward an intensified water cycle with stronger floods
and droughts; and stress on ecosystems, especially in the Arctic, mountain and tropical areas, resulting in the
shift of species habitat range. The expected economic losses from climate change include reduced agricultural
yields, human health risks, property damages from increased flood frequencies, the loss of ecosystem
services, etc. Increased CO2 levels also affect biological systems independent of climate change. For example,
oceans become markedly more acidic, endangering coral reefs and potentially harming fisheries and other
marine life (Intergovernmental Panel on Climate Change (IPCC), 2014).
Sulfur oxides (SOx) include sulfur monoxide (SO), sulfur dioxide (SO2), sulfur trioxide (SO3) and other sulfur
oxides. In this analysis, EPA analyzed changes in emissions of SO2 only.
58 The Integrated Science Assessment for Paniculate Matter (PM ISA) (U.S. EPA, 2009b) identified the human
health effects associated with ambient PM2.5 exposure, which include premature morality and a variety of
morbidity effects associated with acute and chronic exposures. Similarly, the Integrated Science Assessment for
Ozone and Related Photochemical Oxidants (Ozone ISA) (U.S. EPA, 2013c) identified the human health effects
associated with ambient ozone exposure, which include premature morality and a variety of morbidity effects
associated with acute and chronic exposures.
September 29, 2015 ?T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
7.1.1 Changes in A ir Emissions
As discussed in the RIA (Chapter 5: Electricity Market Analyses), EPA used the Integrated Planning Model
(IPM) to estimate the electricity market-level effects of two of the five regulatory options (Options B and D;
see Chapter 5 in RIA (U.S. EPA, 2015c)). IPM outputs include NOX, SO2, and CO2 emissions to air from
electricity generating units (EGU). Comparing these emissions to those projected for the base case provides
an assessment of the changes in air emissions resulting from changes in the profile of electricity generation
underthe final rule. EPA used four run years, 2020, 2025, 2030, and 2040, to represent the periods of 2019-
2022, 2023-2027, 2028-2033, and 2034-2042, respectively (for a more detailed discussion of the IPM analysis
years, refer to Chapter 5 in RIA).
EPA developed separate estimates of air emissions associated with increases in electricity generation to power
wastewater treatment systems by multiplying plant-specific additional electricity consumption estimated as
part of the engineering analysis by plant- or North American Electric Reliability Corporation (NERC)-specific
emission factors obtained from IPM for each analysis year. EPA estimated air emissions associated with
increased trucking by multiplying the number of miles by average emission factors. Details of these two
analyses are provided in the TDD.
Table 7-1 through Table 7-3 summarize the estimated changes in emissions for the three mechanisms, the
three pollutants, and the two regulatory options covered in this particular analysis. As shown in the tables,
EPA estimates that changes in auxiliary service (Table 7-1) and transportation (Table 7-2) would result in an
increase in emissions (positive values), while changes in the profile of electricity generation (Table 7-3)
would reduce CO2, SO2 and NOx emissions (negative values). Table 7-4 presents the net emissions changes
across the three mechanisms.
The largest effect on projected air emissions is due to the change in the emissions profile of electricity
generation at the market level. As presented in the RIA (Section 10.6: Executive Order 13211: Actions
Concerning Regulations That Significantly Affect Energy Supply, Distribution, or Use), IPM projects small
reductions in electricity generation coming from coal as a result of the ELGs (less than 0.1 percent for
Option B; 0.3 percent for Option D), which is compensated with increases in generation using other fuels or
energy sources - biomass, landfill gas, natural gas, nuclear power, and wind power. The changes in air
emissions reflect the differences in emissions factors for these other fuels, as compared to coal.
Table 7-1: Estimated Changes in Electricity Consumption and Air Pollutant Emissions due to
Increase in Auxiliary Service at Steam Electric Power Plants, Relative to Baseline
Regulatory
Option
Option B
Year
2015-2018
2019
2020
2021
2022
2023-2042
Electricity
Consumption
(MWh)
0.0
30,859.8
41,529.5
72,651.5
86,488.4
102,168.7
CO2 (Metric
Tonnes/Year)
0.0
27,409.7
36,972.7
63,257.8
75,759.4
89,632.5
NOx (Tons/Year)
0.0
13.9
20.5
35.4
45.2
59.6
SO2 (Tons/Year)
0.0
35.9
45.4
69.4
75.6
86.7
September 29, 2015
7-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
Table 7-1: Estimated Changes in Electricity Consumption and Air Pollutant Emissions due to
Increase in Auxiliary Service at Steam Electric Power Plants, Relative to Baseline
Regulatory
Option
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2042
Electricity
Consumption
(MWh)
0.0
60,680.1
94,172.8
152,706.0
192,404.2
237,367.0
CO2 (Metric
Tonnes/Year)
0.0
55,222.0
83,855.3
134,385.6
169,583.1
208,881.9
NOx (Tons/Year)
0.0
35.3
64.7
95.7
128.3
164.9
SO2 (Tons/Year)
0.0
59.2
84.4
132.0
162.3
195.8
Source: U.S. EPA Analysis, 2015; see TDD for details.
Table 7-2: Estimated Changes in Annual Air Pollutant Emissions due to Increased Trucking at
Steam Electric Power Plants, Relative to Baseline
Option
Option B
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2042
2015-2018
2019
2020
2021
2022
2023-2042
CO2 (Metric Tonnes/Year)
0.0
550.9
691.5
865.3
925.2
982.5
0.0
2,111.5
2,480.3
3,244.6
3,641.2
3,767.8
NOx (Tons/Year)
0.0
0.2
0.3
0.4
0.4
0.4
0.0
0.9
1.1
1.4
1.6
1.6
SO2 (Tons/Year)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Source: U.S. EPA Analysis, 2015; see TDD for details.
Table 7-3: Estimated Changes in Annual Air Pollutant Emissions due to Changes in Electricity
Generation Profile, Relative to Baseline
Regulatory
Option
Option B
Option D
Year
2015-2018
2019-2022
2023-2027
2028-2033
2034-2042
2015-2018
2019-2022
2023-2027
2028-2033
2034-2042
CO2 (Metric Tonnes/Year)
0.0
-2,057,293.5
-1,437,164.1
-246,295.3
-1,186,462.9
0.0
-4,869,524.3
-2,555,361.0
-2,089,591.4
-3,193,009.0
NOx (Tons/Year)
0.0
-3,534.5
-3,323.9
-1,361.2
-1,500.2
0.0
-14,614.0
-11,615.0
-8,826.0
-10,638.8
SO2 (Tons/Year)
0.0
-4,620.6
-527.5
1,315.7
-1,608.4
0.0
-5,662.8
2,238.4
-984.2
-4,243.9
Source: U.S. EPA Analysis, 2015; see TDD for details.
September 29, 2015
7-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
Table 7-4: Estimated Net Changes in Air Pollutant Emissions due to Increase in Auxiliary Service at
Steam Electric Power Plants, Increased Trucking at Steam Electric Power Plants, and Changes in
Electricity Generation Profile, Relative to Baseline
Regulatory
Option
Option B
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2027
2028-2033
2034-2042
2015-2018
2019
2020
2021
2022
2023-2027
2028-2033
2034-2042
CO2 (Metric Tonnes/Year)
0.0
-2,029,332.9
-2,019,629.3
-1,993,170.4
-1,980,609.0
-1,346,549.1
-155,680.3
-1,095,847.9
0.0
-4,812,190.8
-4,783,188.7
-4,731,894.1
-4,696,300.0
-2,342,711.3
-1,876,941.7
-2,980,359.3
NOx (Tons/Year)
0.0
-3,520.3
-3,513.7
-3,498.7
-3,488.9
-3,263.9
-1,301.2
-1,440.2
0.0
-14,577.7
-14,548.2
-14,516.9
-14,484.1
-11,448.4
-8,659.4
-10,472.3
SO2 (Tons/Year)
0.0
-4,584.6
-4,575.1
-4,551.2
-4,545.0
-440.8
1,402.5
-1,521.7
0.0
-5,603.6
-5,578.4
-5,530.8
-5,500.5
2,434.3
-788.3
-4,048.0
Source: U.S. EPA Analysis, 2015
7.1.2 NOxandSO2
Detailed human health benefits analyses for air regulations typically involve the use of a sophisticated air
quality model, such as the Community Multiscale Air Quality (CMAQ) Model, and BenMAP, EPA's
principal air pollution benefits analysis modeling tool. The air quality model estimates the changes in
concentrations of criteria air pollutants in each cell of a grid resulting from changes in emissions to air (e.g.,
of NOX and SO2) under various policy scenarios. These criteria air pollutant changes are then input to
BenMAP, which estimates the resulting changes in incidence in the exposed population of the adverse health
effects associated with the pollutants and the corresonding monetized benefits (see (Abt Associates, 2012) for
additional description of BenMAP). This detailed approach for human health benefits analysis of air
regulations tends to be time- and resource-intensive.
Recognizing that a less resource- and time-intensive approach is sometimes desirable, EPA relied on the best
available methods of benefits transfer, which is the science and art of adapting primary research from similar
contexts to estimate benefits for the environmental quality change under analysis. EPA's Air Office
developed estimates of national monetized benefits per ton of emissions avoided for use in estimating benefits
without the need to conduct detailed air quality and human health benefits modeling. The benefit per ton
values repsent the total monetized human health co-benefits, premature mortaility and premature morbidity,
from the reduction in one ton of PM2 5 (or PM2 5 precursor such as NOX or SO2). Because the benefits per ton
of emissions depend on both the type of emissions (e.g., NOX vs. SO2) and the geographic distribution
(relative to population centers) of the emitting sources, EPA developed benefits per ton estimates for specific
combinations of emission source categories and PM2 5 precursors. EPA used this approach, for example, in its
September 29, 2015
7-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 7: Air-Related Benefits
assessment of the benefits of PM and SO2 reductions for the Industrial Boiler and Process Heaters National
Emissions Standards for Hazardous Air Pollutants (NESHAP) rule (U.S. EPA 2004a) and its analysis of the
Mobile Source Area Toxics Rule (U.S. EPA 2004b) (See also Fann et al., 2012; Fann et al., 2009; Levy et al,
2009; and Muller and Mendelsohn, 2009).
EPA's calculation of the benefits per ton values involved three principal steps, as described by Fann et al.
(2009) and the Technical Support Document for the calculation of benefit per-ton estimates (U.S. EPA
2008b):
1. Using an air quality model to estimate the changes in ambient PM2 5 concentrations resulting from
specified precursor emissions reductions under various scenarios and then calculating the total tons
(of the precursor emissions) reduced under each scenario;
2. Using BenMAP to estimate the changes in incidence of the associated health effects and the
monetized benefits of those incidence reductions under each scenario; and
3. Estimating national benefits per ton by dividing the national monetized benefits by the total tons of
emissions reduced under each scenario.
In the current analysis, benefits per ton estimates are needed for four combinations of emission type and
source category involving NOX or SOX:
> NOX from EGUs (to be applied to changes in market-level NOX emissions projected by IPM, and
changes in emissions from auxiliary service);
> SOX from EGUs (to be applied to changes in market-level SOX emissions projected by IPM, and
changes in emissions from auxiliary service);
> NOX from mobile sources (to be applied to changes in NOX emissions associated with transporting
CCR waste to landfills); and
> SOX from mobile sources (to be applied to changes in SOX emissions associated with transporting
CCR waste to landfills).
As described by Fann et al. (2009), "ambient PM2 5 is a complex mixture of primary and secondarily formed
particles, resulting from interactions in the atmosphere and physical transport of emissions of particulate
matter precursors, including available SO2, NOX, and NH3, meteorology (particularly temperature), and
baseline levels and composition of PM25" (Fann etal. 2009, p. 170). NOxand SOX differ in their propensity for
becoming PM25. The benefits per ton estimates are based on the assumption that all fine particulates have the
same potency for causing premature mortality (U.S. EPA 201 la).59
Fann et al. (2012) reported benefits per ton estimates for a variety of emission type/source category
combinations, including all of those listed above, that are relevant to the current analysis, for the years 2005
and 2016. Although they are not reported in Fann et al. (2012), EPA also obtained benefits per ton estimates
for each of these categories for the years 2020, 2025, and 2030 directly from one of the study co-authors.60
For these additional years, the benefits per ton were calculated assuming the same change in ambient air
quality as the author's forecast for 2016, but accounting for the 2020, 2025 and 2030 projected baseline
mortality and population to estimate the change in mortality risk
Benefits per ton estimates are available for other pollutants, such as direct PM2 5 emissions, but they were not
included in this analysis because emissions factors were not available. The chemistry of PM formation is
complex and nonlinear.
60 Provided in personal communication with Charles Fulcher, EPA Office of Air Quality Planning and Standards
(OAQPS), on October 19, 2012.
September 29, 2015 T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 7: Air-Related Benefits
To be consistent with the rest of the analysis of the costs and benefits of the final ELGs, benefits per ton
estimates are needed for each year from 2019 through 2042. Because the benefits per ton estimates for the
years 2016, 2020, 2025, and 2030 are almost linear as a function of year, EPA interpolated benefits per ton
values for the intermediate years (e.g., between 2020 and 2025) and projected values for the years from 2031
through 2042 by linear regression, using (year, benefits per ton) data points for the years 2016, 2020, 2025,
and 2030. Note, however, that the approximate linearity of the (year, benefits per ton) data points may be an
artifact of the inability to project meteorological changes and thus changes in air quality for all years after
2016, noted above. Thus, additional uncertainty was generated by using benefits per ton estimates for the
future years that did not account for meteorological and air quality changes.
Assuming that the geographic distribution of controlled emitting sources in a source category (e.g., EGUs)
and of emissions reductions in the current analysis are similar to the geographic distribution of emitting
sources and emissions reduction in the analysis in Fann et al. (2012), EPA can derive a rough estimate of
benefits from changes in air emissions by applying these benefits per ton estimates to the changes (in tons) of
emissions resulting from compliance with the final rule. For example, the benefits from reduced emissions of
NOX from EGUs under Option D can be estimated by multiplying emissions avoided under the regulatory
option by the appropriate benefits per ton value.
As noted above, NOX and SOX are known precursors to PM2 5 Several adverse health effects have been
associated with PM2 5 including 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. All of these health effects were included in the estimation of benefits that
went into the calculation of benefits per ton in Fann et al. (2012).
A very large percentage, 98 percent, of the total monetized benefits of reducing PM2 5 concentrations are
attributable to avoided premature mortality. The appropriate method for valuing the reductions in premature
mortality and development of an accepted value for projected reduction in the risk of premature mortality is
still a discussion in the economics and public policy communities. Fann et al. (2012) used data from Krewski
et al. (2009), a study of mortality and long-term exposure to PM2 5, to estimate the change in incidence of
premature mortality associated with a given change in PM2 5 concentrations. This study is one of several
credible peer-reviewed long-term exposure studies that EPA has used in benefits analyses of PM25.
When using long-term exposure studies, EPA has traditionally assumed that premature mortality avoided as a
result of a reduction in PM2 5 concentrations in a given year do not all occur in that year. Instead, EPA
assumes that the avoided PM2 5-related premature mortalities are distributed over a 20-year period, with most
occurring in the earlier years. EPA values avoided premature mortality using VSL and then discounts that
value back to the year of the analysis. Thus the numerator (the benefits) of the benefits per ton estimate for a
given year is the value of morbidity avoided in that year plus the present discounted value of the stream of
avoided premature mortalities over a twenty year period, discounted back to that year. For example, a benefits
per ton estimate from Fann et al. (2012) for a ton removed in 2016 is the value of avoided morbidity in 2016
plus the present discounted value of the stream of avoided premature mortalities associated with that ton from
2016 to 2036, discounted back to 2016.
EPA obtained two sets of benefits per ton estimates for this analysis for the years 2005, 2016, 2020, 2025, and
2030: one set using a 3 percent discount rate and the other using a 7 percent discount rate.61 All benefits per
ton estimates for years 2019 through 2042 were further discounted back to the year 2015 (using a 3 percent or
7 percent discount rate, as appropriate). Because avoided premature mortalities are assumed to occur over a
twenty-year period, and real income is likely to increase overtime, these benefits per ton estimates reflect
61 Provided in personal communication with Charles Fulcher, EPA/OAQPS, on October 19, 2012
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
EPA's estimated increases in the value for mortal risk reductions with respect to increases in real income. The
income growth adjustment factors used are those in BenMAP (Abt Associates Inc., 2012). Table 7-5
summarizes the benefits per ton estimates EPA used for the different emission type and source category
combinations involving NOX and SO2 in the analysis of the final rule.
Table 7-5: National Benefits per Ton Estimates for NOx and SO2 Emissions (2013$/ton) from the
Benefits per Ton Analysis Reported by Fann et al. (2012)a'b'°
Discount
Rate
3%
7%
Year
2005
2016
2020
2025
2030
2005
2016
2020
2025
2030
ECU
NOx
$3,791
$5,475
$5,686
$6,107
$6,528
$3,475
$4,844
$5,159
$5,475
$5,897
S02
$27,377
$36,853
$38,959
$42,118
$45,277
$25,271
$32,641
$34,747
$37,906
$41,065
Mobile Source (Onroad)
NOx
$4,633
$7,687
$8,108
$8,845
$9,582
$4,107
$6,949
$7,476
$8,002
$8,634
S02
$21,059
$20,006
$22,112
$24,218
$27,377
$18,953
$17,900
$20,006
$22,112
$24,218
Source: U.S. EPA Analysis, 2015 based on Fann etal. (2012)
a. Provided for this analysis by Charles Fulcher, EPA/OAQPS on October 19, 2012.
b. Mortality benefits based on Krewski et al. (2009).
c. Estimation of benefits per ton for 2016, 2020, 2025, and 2030 were based on year 2016 emissions modeling.
7.1.3 CO2
EPA estimated the global social benefits of CO2 emission reductions using the social cost of carbon (SCC)
estimates developed by the Interagency Working Group on the Social Cost of Carbon (IWGSCC, 2010,
2013a, 2013b, 2015a). This document refers to these estimates, which were developed by the U.S.
government, as "SCC estimates." The SCC is a metric that estimates the monetary value of impacts associated
with marginal changes in CO2 emissions in a given year. It includes a wide range of anticipated climate
impacts, such as net changes in agricultural productivity and human health, property damage from increased
flood risk, and changes in energy system costs, such as reduced costs for heating and increased costs for air
conditioning. It is used to quantify the benefits of reducing CO2 emissions, or the disbenefit from increasing
emissions, in regulatory impact analyses.
The SCC estimates were developed over many years, using the best science available, and with input from the
public. Specifically, an interagency working group (IWG) that included EPA and other executive branch
agencies and offices used three integrated assessment models (lAMs) to develop the SCC estimates and
recommended four global values for use in regulatory analyses. The SCC estimates were first released in
February 2010 and updated in 2013 using new versions of each IAM. The 2013 update did not revisit the
2010 modeling decisions with regards to the discount rate, reference case socioeconomic and emission
scenarios, and equilibrium climate sensitivity distribution. Rather, improvements in the way damages are
modeled are confined to those that have been incorporated into the latest versions of the models by the
developers themselves and published in the peer-reviewed literature. The 2010 SCC Technical Support
Document (TSD) (IWGSCC, 2010) provides a complete discussion of the methods used to develop these
estimates and the current SCC TSD presents and discusses the 2013 update (including recent minor technical
September 29, 2015
7-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 7: Air-Related Benefits
corrections to the estimates) (IWGSCC, 2013b). In July 2015, IWGSCC published technical correction to the
estimates (IWGSCC, 2015a); EPA uses these most current values for the benefit estimates presented in this
chapter.
The 2010 SCC TSD noted a number of limitations to the SCC analysis, including the incomplete way in
which the lAMs capture catastrophic and non-catastrophic impacts, their incomplete treatment of adaptation
and technological change, uncertainty in the extrapolation of damages to high temperatures, and assumptions
regarding risk aversion. Currently lAMs do not assign value to all of the important physical, ecological, and
economic impacts of climate change recognized in the climate change literature due to a lack of precise
information on the nature of damages and because the science incorporated into these models understandably
lags behind the most recent research. Nonetheless, these estimates and the discussion of their limitations
represent the best available information about the social benefits of CO2 reductions to inform benefit-cost
analysis. The new versions of the models offer some improvements in these areas, although further work is
warranted.
Accordingly, EPA and other agencies continue to engage in research on modeling and valuation of climate
impacts with the goal to improve these estimates. The EPA and other agencies also continue to consider
feedback on the SCC estimates from stakeholders through a range of channels, including public comments on
Agency rulemakings that use the SCC in supporting analyses and through regular interactions with
stakeholders and research analysts implementing the SCC methodology used by the IWG. In addition, OMB's
Office of Information and Regulatory Affairs sought public comment on the approach used to develop the
SCC estimates through a separate comment period that ended on February 26, 2014. See response to comment
document (IWGCC, 2015b).
After careful evaluation of the full range of comments, the IWG continues to recommend the use of the SCC
estimates in regulatory impact analysis. With the release of the response to comments, the IWG announced
plans to obtain expert independent advice from the National Academy of Sciences to ensure that the SCC
estimates continue to reflect the best available scientific and economic information on climate change. The
NRC review will be informed by the public comments received and focus on the technical merits and
challenges of potential approaches to improving the SCC estimates in future updates.
Concurrent with OMB's publication of the response to comments on SCC and announcement of the NRC
process, OMB posted a revised TSD that includes two minor technical corrections to the current estimates.
One technical correction addressed an inadvertent omission of climate change damages in the last year of
analysis (2300) in one model and the second addressed a minor indexing error in another model. On average
the revised SCC estimates are one dollar less than the mean SCC estimates reported in the November 2013
TSD. The change in the estimates associated with the 95th percentile estimates when using a 3 percent
discount rate is slightly larger, as those estimates are heavily influenced by the results from the model that
was affected by the indexing error.
The four SCC estimates are: $13, $46, $68, and $130 per metric ton of CO2 emissions in the year 2020 (2013
dollars).62 The first three values are based on the average SCC from the three lAMs, at discount rates of 5, 3,
and 2.5 percent, respectively. Estimates of the SCC for several discount rates are included because the
literature shows that the SCC is sensitive to assumptions about the discount rate, and because no consensus
exists on the appropriate rate to use in an intergenerational context (where costs and benefits are incurred by
different generations). The fourth value is the 95th percentile of the SCC across all three models at a 3 percent
62 The SCC TSDs provide SCC in 2007 dollars, which are adjusted to 2013 dollars using the GDP Implicit Price
Deflator. While the SCC values reported in Table 7-6 have been rounded to two significant digits, unrounded
numbers were used to calculate the CO2 benefits.
September 29, 2015 T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
discount rate. It is included to represent higher-than-expected impacts from temperature change further out in
the tails of the SCC distribution. The SCC 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 greater climate change.
These estimates are then discounted back to the year 2015 using the same discount rate used to estimate the
SCC. For internal consistency, the annual benefits are discounted back to net present value terms using the
same discount rate as each SCC estimate (i.e. 5 percent, 3 percent, and 2.5 percent) rather than the discount
rates of 3 percent and 7 percent used to derive the net present value of other streams of costs and benefits of
the final rule.63
EPA estimates the dollar value of the CO2-related benefits for each analysis year between 2019 and 2042 by
applying the global SCC estimates, shown in Table 7-6, to the estimated reductions in CO2 emissions under
the final rule.
Table 7-6: Social Cost of Carbon Values (2013$/metric ton CO2)
Year
2019
2020
2021
2022
2025
2030
2035
2040
5% Discount Rate,
Average
$13
$13
$13
$14
$15
$18
$20
$23
3% Discount Rate,
Average
$45
$46
$46
$47
$50
$55
$60
$66
2.5% Discount
Rate, Average
$67
$68
$69
$70
$74
$80
$85
$92
3% Discount Rate,
95th Percentile
$130
$130
$140
$140
$150
$170
$180
$200
Source: IWGSCC, 2013b (values updated to 2013 dollars using GDP deflator (1.095)).
7.1.4 Estimating Total Air-Related Benefits
EPA calculated the monetized air-related benefits of the final rule, under options B and D, in any given year
(discounted back to the year 2015) by (1) multiplying the tons of emissions avoided for a given emissions
type/source category combination in that year by the benefits per ton for that emissions type/source category
combination for that year, and then (2) summing the benefits across all emissions type/source category
combinations. The total benefit for year y, then, is calculated using Equation 7-1.
Equation 7-1.
Where:
^ (Tons avoided) y j
j = 1,2, and 3 denote NOX, SO2, and CO2, respectively, from market-level EGUs;
j = 4, 5, and 6 denote NOX, SO2, and CO2, respectively, associated with auxiliary service;
j = 7, 8, and 9 denote NOX, SO2, and CO2, respectively, associated with transportation; and
BPTy j is the present discounted value, discounted to the year 2015, of the benefits per ton for the/th
emissions type/source category combination.
63 See more discussion on the appropriate discounting of climate benefits using SCC in the 2010 SCC TSD
(IWGSCC, 2010).
September 29, 2015
7-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
The total present discounted value of benefits, discounted to the year 2015, PDV2ois, is calculated using
Equation 7-2.
2042 9
Equation 7-2. PDV2015= J^(Tons avoided)yj xBPT™
^2015
>J '
y=2015j=l
7.2 Results
Table 7-7 shows the estimated benefits from reductions in emissions of NOX, SO2, and CO2 in each of several
selected years for the two regulatory options EPA analyzed.
Table 7-7: Estimated Benefits from Reduced Air Emissions for Selected Years (millions; 2013$)a
Option
Option B
Option D
Year
2019b
2020
2025
2030
2019b
2020
2025
2030
3% Discount Rate
$287.7
$291.1
$106.3
-$46.5
$514.5
$520.0
$85.4
$195.0
7% Discount Rate
$266.3
$270.0
$102.4
-$41.4
$482.2
$488.8
$88.4
$186.2
Source: U.S. EPA Analysis, 2015
a. EPA used SCC values based on a 3 percent (average) discount rate to calculate total benefit values presented for both the 3
percent and 7 percent discount rate.
b. The benefits per ton values used for year 2019 benefit calculation is assumed to be the same as the 2020 benefits per ton values.
Table 7-8 shows the annualized benefits from reductions in emissions of NOX, SO2, and CO2 for the two
regulatory options EPA analyzed. EPA annualized benefit estimates to enable consistent reporting across
benefit categories (e.g., benefits from improvement in water quality). The total air-related benefits include
benefits from CO2 emissions reductions calculated using average SCC values (average at 2.5 percent, average
at 5 percent, and the 95th percentile at 3 percent).
The annualized benefits of Options B and D are $110.2 million and $284.5 million, respectively, using a
discount rate of 3 percent ($103.6 million and $248.6 million, respectively, using a discount rate of 7
percent).64
The 3 percent average SCC estimate was used to value reductions in CO2 everywhere total benefits of the final
rule are reported.
September 29, 2015
7-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
Table 7-8: Estimated Annualized Benefits from Reduced Air Emissions (Millions; 2013$)
ELG Option
Option B
Option D
Pollutant
NOX
S02
CO2
TOTAL
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
NOX
S02
C02
TOTAL
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Discount Rate
$12.7
$45.0
$52.5
$15.3
$77.4
$157.7
$110.2
$73.1
$135.2
$215.5
$63.1
$81.6
$139.8
$40.4
$206.8
$421.1
$284.5
$185.1
$351.4
$565.8
7% Discount Rate
$10.6
$40.5
$52.5
$15.3
$77.4
$157.7
$103.6
$66.5
$128.6
$208.9
$49.4
$59.4
$139.8
$40.4
$206.8
$421.1
$248.6
$149.2
$315.6
$529.9
Source: U.S. EPA Analysis, 2015
7.3 Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. Estimated air-related benefits do not depend on pollutant
loadings to receiving waters and therefore the monetized benefit estimates provided in this Chapter are
unaffected by revisions to steam electric plant loading estimates.
imitations and Uncertainties
This analysis is subject to the standard sources of uncertainty found in any air pollution benefits analysis -
uncertainties surrounding the estimated emissions changes, the estimated changes in air pollutant
concentrations resulting from changes in emissions, the estimated concentration-response relationships
between the air pollutant and various health effects in the exposed population, and the estimated value of each
health effect avoided. There is additional uncertainty in the SCC estimates, which reflect the projection of
future harm from climate change, and the benefits per ton estimates. More details about the limitations and
uncertainties associated with the air-benefit analysis are discussed in Table 7-9.
September 29, 2015
7-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
Table 7-9: Limitations and Uncertainties in Analysis of Air-related Benefits
Issue
Uncertainty in projecting the
;&.j4.».>*?p....l*Qi*rn "Pmtn r'livniitp
-itltulU iiclllll llUlll L-llllld.LC
change.
There is uncertainty
associated with the effects of
compliance costs on the
forecast change in emissions
from the electricity sector.
Effect on Benefits
Estimate
Uncertain
Uncertain
Notes
When attempting to assess the incremental economic impacts
of carbon dioxide emissions, the analyst faces a number of
serious challenges. A report from the National Research
Council (NRC, 2009) points out that any assessment will
suffer from uncertainty, speculation, and lack of information
about (1) future emissions of greenhouse gases, (2) the effects
of past and future emissions on the climate system, (3) the
impact of changes in climate on the physical and biological
environment, and (4) the translation of these environmental
impacts into economic damages. As a result, any effort to
quantify and monetize the harms associated with climate
change will raise serious questions of science, economics, and
ethics and should be viewed as provisional.
The Interagency Working Group on Social Cost of Carbon
(IWGSCC, 2010, 2013a, 2013b) noted a number of
limitations to the SCC analysis, including the incomplete way
in which the integrated assessment models capture
catastrophic and noncatastrophic impacts, the modeling of
inter-regional and inter-sectoral linkages, their incomplete
treatment of adaptation and technological change, uncertainty
in the extrapolation of damages to high temperatures, and
assumptions regarding risk aversion. The limited amount of
research linking climate impacts to economic damages makes
the interagency modeling exercise even more difficult. This
said, the SCC estimates were developed using a defensible set
of input assumptions that are grounded in the existing
literature.
Compliance costs (capital, fixed or variable) will influence
marginal generation decisions of plants affected by the final
rule. In order to model the electricity market effects of the
final rule, EPA made certain modeling assumptions that may
influence the pattern of generation across the electricity
sector, and therefore emissions. For example, EPA converted
engineering capital costs to annual fixed operation and
maintenance (O&M) costs in order to model the cost of
complying with the final rule in IPM.
See PJA Chapter 5, and Section 10.6: Executive Order 13211:
Actions Concerning Regulations That Significantly Affect
Energy Supply, Distribution, or Use for additional discussion
of how modeling assumptions may influence the forecast air
pollution changes from the IPM modeling.
Differences between modeled and actual quantities of
electricity generated and emission factors of dispatched
generating units would affect the changes in air pollutants
emissions and therefore the benefits resulting from these
changes. EPA does not have information to quantify the
magnitude of this uncertainty.
September 29, 2015
7-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
7: Air-Related Benefits
Table 7-9: Limitations and Uncertainties in Analysis of Air-related Benefits
Issue
There is uncertainty
associated with the effects of
the Clean Power Plan rule
EPA used a reduced form
approach (benefits per ton) to
value air-related benefits of
emissions changes.
EPA used year-specific
benefits per ton estimates to
derive values for each year
within the analysis period.
Effect on Benefits
Estimate
Uncertain
Uncertain
Uncertain
Notes
The final Clean Power Plan provides states considerable
flexibility in developing state implementation plans to meet
the rate or mass targets. This flexibility provides states great
leeway to meet key priorities. However, it induces a
considerable degree of uncertainty in what the future electric
power market will look like and the overall economic impacts
of the final ELGs. For example, states may choose to comply
with the Clean Power Plan in ways that will lead to fewer or
more coal -fired steam ELG plant retirements than the IPM
runs would indicate. Such differences may have an important
impact on dispatch profiles and emission changes.
As Fann et al. (20 12) note, "... implicit in the benefit per ton
assessment is that the key attributes of the modeling — e.g.
population distribution, source parameters, etc. — are not so
different from the policy scenario as to affect the estimated
benefits appreciably. Reduced form approaches assume a
linear relationship between changes in emissions and benefits,
an assumption that may not be valid for large changes in
emissions" (Fann et al., 2012, p. 142).
Use of year-specific benefits per ton estimates from which to
generate annual estimates introduces another layer of
uncertainty into the analysis. In particular, because actual air
quality modeling was carried out only for 2005 and 2016, the
approximate linearity seen in the benefits per ton estimates for
2016, 2020, 2025, and 2030 may be an artifact of assuming
that air quality remains constant at 2016 levels. The benefits
per ton estimates for intermediate years also do not take into
account the likely non-linearity involved. If each year-specific
benefits per ton is uncertain, then an annual estimate
incorporating benefits per ton-based estimates may be more
uncertain. As a result, the annual estimates can be considered
only rough estimates.
September 29, 2015
7-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 8: Benefits from Reduced Water Withdrawals
8 Benefits from Reduced Water Withdrawals
Steam electric power plants use vast quantities of water for ash transport and for operating wet FGD
scrubbers.
By eliminating or reducing water used in sluicing operations or prompting the recycling of water in FGD
wastewater treatment systems, the ELGs are expected to reduce water withdrawal from both surface
waterbodies and aquifers. The reduction in water use depends on the regulatory option.65 EPA estimates that
power plants would reduce water withdrawals from between 45 billion gallons per year (0.12 million gallons
per day) under Option and 209 to 222 billion gallons per year (0.57 to 0.61 million gallons per day) under
Option E (see Chapter 11 of TDD for details). The final BAT/PSES option (Option D) will reduce water
withdrawals at steam electric power plants by 143 to 155 billion gallons per year (0.39 to 0.42 million gallons
per day).
The section below discusses the benefits resulting specifically from reductions in groundwater withdrawals
(Section 8.1). Benefits associated with surface water withdrawals are discussed qualitatively in Chapter 2 and
in a separate report provided in the final rule record (see DCN SE05943).
roundwater Withdrawal
Reduced water intake from groundwater sources by steam electric power plants are expected to result in
increased availability of groundwater for local municipalities that rely on groundwater aquifers for drinking
water supplies. These municipalities are expected to avoid the cost of supplementing drinking water supplies
through alternative means, such as bulk drinking water purchases. The following sections describe EPA's
estimate of reduced groundwater withdrawal benefits.
EPA estimated the benefits of reduced groundwater withdrawals based on avoided costs of purchasing
drinking water during periods of shortages in groundwater supply.
8.1.1 Methods
EPA's analysis of the final ELG options (U.S. EPA, 2015b) indicate that one plant located in Nebraska will
reduce the volume of groundwater withdrawn as a result of the ELGs. EPA estimated that the plant will avoid
withdrawing a total of 21,971 gallons per day (8 million gallons per year) by converting to dry handling for its
bottom ash under Options D and E. Because the state is potentially or currently water-stressed (Tetra Tech,
2011), the ELGs are likely to generate benefits from improved groundwater recharge. To estimate the value of
improved groundwater supply, EPA relied on state-specific prices of bulk drinking water supplies, since
municipalities may need to purchase supplementary supplies in response to groundwater shortages arising
from excessive withdrawals. EPA recognizes that the assumption that a reduction in groundwater withdrawals
in the water-stressed states may result in reduced groundwater shortages is somewhat speculative, but used
this assumption to provide screening-level estimates of the potential benefits.
To estimate the monetary value of reduced groundwater withdrawal, EPA relied on current state-specific
water prices ($1,192.06 per acre/foot for Nebraska). For each affected plant and regulatory option, EPA
65
Depending on the BAT/PSES technology basis for fly and bottom ash wastewater, the regulatory option may
eliminate or reduce water use associated with current wet sluicing operating systems at steam electric power
plants. Specifically, reductions in intake flow are expected to occur at plants which convert to dry handling or
recycle FGD wastewater.
September 29, 2015
8-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 8: Benefits from Reduced Water Withdrawals
multiplied the reduction in groundwater withdrawal (in gallons per year) by the estimated price of drinking
water per gallon. EPA used a conversion factor of 325,851 to convert acre foot to gallons.
8.1.2 Results
Table-8-1 shows estimated annual benefits from reduced groundwater withdrawals. The annual benefits from
the BAT/PSES option (Option D) for existing sources are $0.02 million using a 3 percent discount rate
($0.02 million using a 7 percent discount rate).
Table 8-1: Estimated Annualized Benefits from Reduced Groundwater Withdrawals (Millions; 2013$)
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Reduction in
Groundwater Intakes
(million gallons per year;
full implementation)
0.0
0.0
0.0
8.0
8.0
3% Discount Rate
$0.00
$0.00
$0.00
$0.02
$0.02
7% Discount Rate
$0.00
$0.00
$0.00
$0.02
$0.02
Source: U.S. EPA Analysis, 2015
8.1.3 Implications of Revised Steam Electric Plant Loading Estimates
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. Estimated water withdrawal benefits do not depend on
pollutant loadings and therefore the monetized benefit estimates provided in this Chapter are unaffected by
revisions to steam electric plant loading estimates.
8.1.4 Limitations and Uncertainties
Table 8-2 summarizes the limitations and uncertainties in the analysis of benefits associated with reduced
groundwater withdrawals.
Table 8-2: Limitations and Uncertainties in Analysis of Reduced Groundwater Withdrawals
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
EPA assumed that municipalities
would need to replace lost
groundwater supplies with bulk
drinking water purchases.
Uncertain
See below.
Municipalities may not need to replace groundwater
withdrawn by steam electric power plants (in which case
the benefits of the ELG may be overstated), or they may
choose to replace the groundwater through other means,
such as desalinization (in the case of Florida, in which
case the benefits of the ELG may be understated).
EPA assumed 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 assumed that demand for additional water supply
exists in the affected areas (Florida and Nebraska) due to
potential draughts. However, the extent of this demand is
uncertain.
September 29, 2015
8-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 8: Benefits from Reduced Water Withdrawals
Table 8-2: Limitations and Uncertainties in Analysis of Reduced Groundwater Withdrawals
Uncertainty/Assumption
Affected aquifer characteristics
Effect on Benefits
Estimate
Uncertain
Notes
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.
September 29, 2015
8-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
9: Avoided Dredging Costs
9 Benefits from Avoided Dredging Costs
EPA expects that the final rule will reduce discharge loads of various categories of pollutants including total
suspended solids (TSS), thereby reducing the rate of sediment deposition to affected waterbodies, 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, Haverkamp, and Chapman, 1985). In
many cases, costly periodic dredging is necessary to keep them passable. The final steam electric ELG could
provide cost savings to government and private entities responsible for maintenance of navigable waterways
by reducing the dredging volume.
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 silt layers over time, at a recorded average rate of 1.2 billion kilograms per reservoir every year
(USGS, 2007b). Sedimentation reduces reservoir capacity (Graf, Wohl, Sinha, and Sabo, 2010) and the useful
life of reservoirs unless measures such as dredging are taken to reclaim capacity (Clark, et al., 1985). EPA
expects that the final ELG will provide cost savings by reducing dredging activity to reclaim capacity at
existing reservoirs.
EPA estimates that the final ELG would result in modest cost savings from reducing the amount of sediment
dredged from navigational waterways and reservoirs affected by pollutant discharges from steam electric
power plants. Under Option D, benefits from reduced navigational waterway dredging are less than one
thousand dollars, with both 3 and 7 percent discount rates; benefits from reduced reservoir dredging are less
than two thousand dollars, with both 3 and 7 percent discount rates. EPA has revised sediment loadings
following the completion of the benefit analysis However, because sediment loadings changed by
approximately 1 percent the overall effect on the estimated benefits from reduced dredging of navigational
waterways and reservoirs is likely to be trivial.
Appendix K provides a more detailed description of the methodology and results of this analysis. This
appendix presents EPA's analysis of the avoided dredging costs for navigable waterways and reservoirs under
the five regulatory options. First, it describes EPA's analysis of historic dredging locations and frequency.
Next, it presents EPA's approach for estimating sediment deposition and removal in dredged waterways and
reservoirs under the ELG regulatory options and associated costs. It then presents estimated benefits.
September 29, 2015
9-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
10 Benefits from Enhanced Marketability of Coal Combustion Residuals
EPA expects that installation of waste stream treatment technologies to comply with final ELGs will affect
the type of coal combustion residue (CCR) by prompting plants to convert from wet handling of fly ash,
bottom ash, and/or flue gas desulfurization (FGD) waste to dry handling. Relative to wet ash, dry ash's
chemical and physical properties make it more suitable for re-use in a variety of applications like structural
fill, concrete, and wallboard (U.S. EPA, 2015d; American Coal Ash Association, 2012). This change would in
turn allow plants to more readily market CCR to beneficial uses.
There are two main economic benefits to society of re-using CCR. First, plants that are able to beneficially re-
use CCRs are able to offset the CCR disposal costs (e.g., trucking and landfills) that EPA counted in its cost
and economic impact analyses of this final rule. These avoided costs are net of added costs that some plants
may incur in preparing dry CCRs for reuse. Second, by replacing raw or virgin inputs with re-used CCR
materials during the production of structural fill and concrete, society avoids the need for and cost of
extracting and preparing the raw or virgin inputs. These benefits include both direct costs (e.g., operating
machinery, costs of transport), and indirect costs (e.g., downstream environmental benefits).
This Chapter describes the methodology EPA used to estimate the avoided cost benefits resulting from the
enhanced marketability of CCR. The approach builds on the methodology EPA used in analyzing changes in
beneficial use of CCRs under the final RCRA Final Rule for Disposal of Coal Combustion Residuals
Generated by the Electric Utility Industry (U.S. EPA, 2014).
i
0.1 Methods
10.1.1 Beneficial Use Applications
The methodology focuses on two CCR wastestreams, and two end uses that may be affected by the final
ELGs:66 (1) fly ash as a substitute for Portland cement in concrete production and (2) fly- and bottom ashes as
substitutes for sand and gravel in fill applications (Table 10-1)61
In this analysis, EPA assumes that the ELGs do not materially affect the total quantity of CCR generated by
steam electric power plants but instead change only the relative shares of the CCR handled dry instead of wet.
Following the central assumptions of EPA's cost and economic impact analyses for this rule (U.S. EPA,
2015c), EPA assumes that plants generate a constant amount of CCR each year in the analytic time period.
Table 10-1: Applicable Beneficial Use Applications of CCRs, by CCR Category
CCR Category
Fly ash
Bottom ash
Beneficial Use Application
Concrete Production
•/
Fill
s
•/
EPA does not expect the final ELGs to affect the quantity or handling of FGD gypsum, and this CCR use is
therefore excluded from the benefit analysis.
67 Note that this assumed pattern simplifies actual uses. The American Coal Ash Association (ACAA) reports
other important uses (i.e., each more than 1 million short tons in 2012) for fly ash, bottom ash, and FGD
gypsum, including waste stabilization (fly ash), blended cement/feed for clinker (fly ash, bottom ash, and FGD
gypsum), and mining applications (fly ash, FGD gypsum) (http://www.acaa-
usa.org/Portals/9Mles/PDFs/revisedTINAL2012CCPSurveyReport.pdT).
September 29, 2015 loT
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 10: Ash Marketing
10.1.2 Marketable OCR by State
The demand for concrete and fill constrains the extent to which steam electric power plants will be able to
shift CCRs from disposal to beneficial use. Transportation costs are another main constraint and make
marketing of CCRacross long distances less likely. USGS Minerals Commodity Summaries (e.g., Bolen,
2014; van Oss, 2013) show that both sand and cement production vary across regions of the United States;
thus, Steam Electric facilities in different parts of the country likely face different markets for CCR.
CCR end use sites include a mix of industrial facilities (e.g., to produce cement) and dispersed sites (e.g., as
structural fill). Without precise information about the location of these beneficial use sites within regions,
EPA used sand and cement production within the state surrounding a plant to approximate the regional market
demand each plant may face in marketing dry CCR. This approach assumes that plants only market their CCR
to sites within the plant's state, and that state-level demand constrains the total amount of CCR that steam
electric power plants in that state, as a group, will be able to market to various end uses. To determine the
maximum quantity of CCRs that steam electric power plants located in a given state will be able to market
("marketable CCRs"), EPA compared state-level estimates of supply and demand for CCR.68 Procedures to
estimate demand and supply are described below. Due to heterogeneity in state sizes, this assumption implies
the geographic extent of, and volume of, CCR markets vary across states.
10.1.2.1 Demand for CCR
EPA assumed that the annual quantities of end products (e.g., cement, and construction sand and gravel)
produced and used in the United States represent the maximum annual demand for CCR in those applications,
accounting for the share of the end products that may be replaced with CCR.
> Total End Product Production. EPA estimated concrete and fill production based on U.S. Geological
Survey "Minerals Commodity Summaries," and developed 3-year average production statistics for
cement and construction sand and gravel, respectively (Bolen, 2011, 2012, 2013, 2014; van Oss,
2009, 2010, 2013). These are the same sources as used for the final CCR rule analysis, but with more
current data. In estimating production, EPA used state-level sand and gravel production statistics
directly (e.g., Bolen, 2010). Annual cement production statistics are presented by multi-state regions,
and separately report production for Portland and masonry cements (e.g., Van Oss, 2010). Because
Portland cement constitutes the majority of total cement production by weight (97% in 2013), EPA
used total cement production as a proxy for Portland cement production. EPA downscaled regional
cement production statistics to individual state(s) by state shares of regional population (US Census,
2012).
> End Product Input Replaceable with CCR. For each end product, EPA then estimated the portion of
total production that could be replaced with CCR. For cement, EPA assumed that:
- Fly ash could replace 25% of cement used in concrete. This percentage represents the low
end of several replacement rates reported in industry literature while remaining consistent
with the final CCR rule and other federal government documents. For example, ERMCO
(2013) reports that fly ash additives constitute 21.7% of cement in the United States, and We
Energies (2013, p. 65) concluded that fly ash could be substituted for up to 40% of Portland
concrete by weight in making structural grade concrete. Government replacement rate
68 The CCR Final Rule approach (Appendix S) used linear programming to estimate incremental changes in
beneficial reuse due to the rule, accounting for county-level capacity constraints (maximum demand).Although
our state-level market approximations more coarsely delineate markets, they also serve to constrain maximum
reuse to estimated market capacity.
September 29, 2015 lo
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 10: Ash Marketing
estimates tend to be somewhat lower than industry estimates; for example, the CCR rule
assumed that "that fly ash makes up no more than 30% of... cementitious material"
(Appendix S in US EPA, 2014), based on resources from US EPA (2008)69, and the Federal
Highway Administration Materials Group recommends a substitution rate of no more than 15
to 25 percent in pavement applications (FHWA, 2015).
- 73 percent of cement is used in ready-mix concrete (ERMCO, 2013).Of the cement that fly
ash can replace, only a portion is used in concrete; mortar and grout are other uses for
cement.
For fill, EPA assumed that:
- Fill is 36% of total sand and gravel production by state. In 2012, 36% of national
construction sand and gravel production was used in fill and fill-like applications (Bolen,
2013), which include concrete aggregates (including concrete sand); asphaltic concrete
aggregates and other bituminous mixtures; road base and coverings; fill; and snow and ice
control.
- Beneficial reuse ofCCRs infill carries no stigma, and could potentially replace 100 percent
of sand/gravel use as fill. Consistent with the analysis of the CCR final rule, EPA assumed no
stigma around the reuse of CCRs in fill applications. EPA notes, however, that there may
remain some potential for the public to negatively perceive CCR re-used in fill or concrete
because they are unfamiliar with CCR and/or with re-use practices, but consume media
information that incorrectly characterizes these products and practices as "toxic" (U.S. EPA,
2010d) . Because EPA did not identify information estimating the national extent or impact of
stigma around CCR reuse, EPA assumed no stigma in residential applications. This approach
is consistent with the non-hazardous determination applied in the Final CCR Rule, which is
designed to avoid stigma around uses of "hazardous" products. To the extent that stigma or
perceptional effects does exist and leads to reduced marketability of CCRs, the analysis may
over-state the re-use of CCRs in some applications (e.g., residential contexts).
70.7.2.2 CCR Supply
> Baseline Production. EPA estimated state-level baseline CCR supply by ash type, including (1) the
sum of plant-level dry fly and bottom ash production marketed, as reported by Steam Electric plants
located within the state in response to the 2010 Questionnaire for the Steam Electric Power
Generating Effluent Guidelines (U.S. EPA, 2010a); and (2) The state-level sum of additional CCRs
available for beneficial reuse due to the final CCR rule. The CCR final rule analysis projected that in
2025, the 3-year rolling average of the annual increase in national CCR beneficial uses will total 0.26
million tons of fly ash in concrete, and 5.65 million tons of CCRs in structural fill.70 EPA allocated
these national total supply of CCR to states with baseline cement and sand production. Using
population as a proxy for fill and concrete use (and thereby, a proxy for production), EPA allocated
the additional CCR to states based on each state's population.
> Change in Dry CCR Supply due to Final Steam Electric ELGs. EPA used data on annual plant-level
changes in the quantity of CCRs handled dry, by category - fly ash and bottom ash - and which could
U.S. EPA (2008). Study on Increasing the Usage of Recovered Mineral Components in Federally Funded
Projects Involving Procurement of Cement or Concrete to Address the Safe, Accountable, Flexible, Efficient
Transportation Equity Act: A Legacy for Users, Report to Congress, June 3, 2008, EPA530-R-08-007.
70 CCR RIA, Exhibit 5-L.
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
be more readily marketed for beneficial use after making changes to meet the final ELGs under each
of the regulatory options (see TDD for details).71
Consistent with the screening-level cost and economic impact analyses for the final rule which assume that
steam electric power plants generate a constant amount of electricity throughout the analysis period (see RIA;
U.S. EPA 2015c), EPA assumed that steam electric power plants will produce the same quantity of dry CCRs
every year in the analysis period, beginning in the technology implementation year. EPA determined the
amount of plant-level changes in fly- and bottom ash CCRs marketable to the two beneficial use categories
assuming that fly ash is first marketed to concrete, as it is the use with higher market value, and then the
remaining fly and bottom ash CCRs marketed to fill, as possible given remaining demand.
70.7.2.3 Marketable OCR
Thirty six states produced cement in 2010, 2011, or 2012 (van Oss, 2009; 2010; 2013). Given baseline CCR
supply, EPA estimated that additional CCR production due to conversion to dry ash handling systems to meet
the final ELGs would be marketable in 14 (39 percent) of these states. Note that none of the steam electric
power plants expected to increase dry CCR production due to the ELGs are located in states without cement
production in the baseline.
All 50 states produce sand and gravel, and EPA estimated that the CCR supply (from existing marketing of
CCR by steam electric power plants and projected CCR reuse) exceeds demand for fill in only one state (West
Virginia). In all other states, demand data suggest that steam electric power plants can market all the
additional CCR handled dry.
Table 10-2 reports "marketability" estimates for all states combined. Marginal changes in dry CCR represent
approximately 5 percent of unmet demand for CCR as a substitute for Portland cement in concrete and
2 percent of unmet demand for CCR as a substitute for sand and gravel in fill. EPA expects that some plants
will market CCR to both cement and fill, and that a lack of demand for fly ash in cement will induce many to
market fly ash to fill. Nearly all of the expected conversions to dry-handled fly and bottom ash could
potentially be marketed for beneficial reuse within the plant's state.
Table 10-2: State-level Market Approximation (Short Tons).
Application
Concrete
Fill
Baseline Production
Production
13,400,062
320,701,683
Unmet Demand (%
of production)
3,277,184 (24%)
312,705,505(98%)
Marketable Changes due to ELGs
Fly Ash
(% Total ACCR)
162,491 (12.2%)
1,166,260(87.8%)
Bottom Ash (% Total
ACCR)
NA
4,314,946(91.7%)
Source: U.S. EPA Analysis 2015
EPA excluded from subsequent valuation of tallied "incremental" marketable CCR any changes that occur at
steam electric power plants which already market the wet ash in the baseline, and only calculated benefits
associated with plants that do not currently market their CCR wet. While these plants may convert to dry
handling and may therefore find it easier to market their CCR in the future, EPA assumed for simplicity that
the amount of ash they could market for beneficial use will not change as a result of this conversion. As a
Fly ash tonnages are a dry-basis (not moisture conditioned), and were only calculated for steam electric units
flagged for transport and disposal costs, excluding those units flagged for costs associated with additional
conveyance capacity. Bottom ash tonnages were estimated for all bottom ash handling conversions excluding
units marked for bottom ash management costs. Tonnage estimates include the CCR population; conversions at
plants flagged as a fly ash or bottom ash dry conversion under the CCR rule were zeroed.
September 29, 2015
10-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
result of this assumption, the analysis values 100 percent of fly ash conversions marketed to concrete, and
93 percent of fly ash and bottom ash conversions marketed to fill.
10.1.3 Value to Society from Re-Using OCR
Replacing virgin materials like fill and cement with CCR wastes avoids some costs while introducing other
costs. EPA used the social cost accounting framework summarized in Error! Reference source not found, to
track benefits and costs of CCR re-use, notably changes in production and disposal pre- and post-ELGs. The
remainder of this section provides more detail on cost and benefit categories.
Figure 10-1: Cost Accounting Framework for the Beneficial Reuse Analysis.
Without Reuse
Raw Materials (sand,
Extraction/ mining
Processing
Transport to market
With Reuse
cement)
No extraction/ Mining
No processing
No transport to market
Net Change
4* Raw material
extraction
4* Raw materials
processing
^ Raw materials
transportation
Valuation Framework
Life Cycle Analysis3
CCR from Steam Electric Power Generation
Prepare CCR for
disposal
Transport SE CCR to
disposal site
Disposal
Beneficiate
CCR for re-use
Transport SE CCR to
reuse site
No Disposal
1s Beneficiation
No net change in SE
CCR transport costs"
^ Disposal costs to
meet SE ELG
4 Estimate beneficiation
costs
1 Qualitative uncertainty
analysis
Estimate avoided
^| disposal costs
a: As was done in the analysis of the final CCR rule, EPA selected the LCA approach to estimate costs that society avoids by
replacing virgin materials with CCR, including environmental impacts. The market value of avoided raw materials is an alternative
cost proxy which captures some, but not all of these production costs. LCA captures some, but not all costs from a market price
framework, and thereby is also an approximate avoided social cost.
b: On net, EPA assumed no change in transportation cost between the base case and policy scenario. In the base case, some plants
dispose of CCR on-site, while others dispose of it off-site. Re-use applications are assumed to occur off-site; however, EPA also
assumed that, on net to society, total costs of transportation to disposal and re-use sites are approximately equal given our in-state
market framework.
Avoided Costs. Re-using CCRs avoids several costs, including the environmental costs of extracting virgin
materials and costs to steam electric power plants of disposing of their CCR.
> Compliance cost offset. By redirecting the CCR to beneficial re-use rather than disposing of it, steam
electric power plants can avoid disposal O&M costs attributed to meeting the ELGs in the economic
impact analysis (see RIA; U.S. EPA 2015c). EPA's analysis of the Final CCR Rule assumed plants
marketing CCR could avoid certain per-ton transportation and disposal costs based on the volume of
CCR marketed, but that plants could not offset certain other costs of the rule (e.g., general liability
insurance required under the CCR rule, but not under the Steam Electric ELGs). For this Final Rule,
EPA also assumed plants could not offset capital costs associated with disposal systems (e.g.,
landfills), but could offset O&M costs include tipping fees at off-site landfills. In this analysis, EPA
assumes Steam Electric power plants market 100% of dry CCR produced following the rule. Since
EPA assumes that plants that market dry CCR are able to re-use 100% of ash by volume, EPA
assumed plants could offset 100% of these O&M costs. To the extent that plants may not market
100% of ash in all years, this assumption produces an upper bound estimate of potential cost offsets
due to beneficial reuse.
September 29, 2015
10-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 10: Ash Marketing
> Production cost offset. This offset is the avoided cost of virgin materials costs72 resulting from using
CCR in place of the virgin materials (e.g., Portland cement, sand and gravel). Depending on who uses
the CCR for beneficial purposes, these cost offsets (transfers) may accrue to the steam electric power
plant, the secondary CCR user who either paid for or received the CCR for free, a third-party reseller,
etc.
> Environmental benefits. By avoiding the production of virgin materials needed for Portland cement
and fill, society also gains a reduction in total environmental damages associated with extracting,
transporting, and processing these raw materials (e.g., air emissions and resource consumption). Life
Cycle Analysis (LCA) is a framework to assess the total environmental impact of a product based on
an inventory of energy and material inputs and outputs associated with each step in producing the
product. Typically, LCA addresses raw material acquisition, materials manufacture, production,
use/reuse/maintenance, and waste management (SAIC, 2006), and estimates environmental impacts
per unit of product produced (e.g., changes in energy consumption, water consumption, and air
pollutant emissions). To estimate the total environmental benefits of avoided Portland cement and fill
production in this analysis, EPA applied the existing LCA impacts for cement and sand and gravel
used in EPA's CCR Final Rule.
Table 10-4 lists environmental impacts per ton of avoided virgin material, and Table 10-5 lists the
estimated unit values of the avoided economic impacts of these environmental impacts. EPA applied
benefit transfer of unit economic values based on the Final CCR Rule's LCA approach, the Social
Cost of Carbon (IWGSCC, 2013b), and human health benefits from reducing emissions of PM25
precursors (U.S. EPA based on Fann et al., 2012).
As shown in Error! Reference source not found., production cost offsets overlap with benefits from avoided
life-cycle impacts. Specifically, the life-cycle impact framework captures many of the energy, water and other
production costs that fill and concrete producers can avoid by re-using CCR products and avoiding the
production of virgin materials. Since the life-cycle framework also captures avoided environmental impacts of
production, the remainder of this analysis reports only the monetized benefits from the life-cycle framework
to avoid potentially double-counting benefits.
Additional Costs. Steam electric power plants that market CCRs for beneficial re-use do so when the re-use is
economically preferable to disposal. Some plants may find it necessary to undertake additional preparation of
dry CCRs before beneficial use (e.g., to bring CCR to ASTM standards), or may need to consider the cost of
transporting CCR to a beneficial use destination. There is uncertainty in the degree to which steam electric
power plants incur one or more of these additional costs. To provide a conservative estimate of avoided costs
(and benefits), EPA included additional costs for beneficiation.
> Beneficiation of fly ash prior to use in concrete: EPA assumed that 13 percent of fly ash requires
beneficiation prior to use in concrete.73 Electric Power Research Institute (2005) reports a range of
beneficiation costs per ton of fly ash; the CCR Final Rule analysis estimated that a reasonable central
tendency value for the cost of beneficiation is approximately $ 10 per ton (in year 2004$). EPA
updated this cost to 2013 dollars and applied this cost to 13 percent of marketable fly ash CCRs at
each steam electric power plant.
For reference, the 3-year average (2008 to 2012) market price of sand and gravel used as fill is $4.36/ton (range: $2.38 -
$9.30 by state) (van Oss/USGS, 2010, 2011, 2013). The 3-year average (2008 to 2012) market price of Portland cement
is $44.50/ton ($76.24 - $148.02 by state) (Bolen, 2011, 2012, 2013, 2014).
73 In the Final CCR rule analysis document, ORCR cites personal communication with David Goss, ACCA (June
25, 2008) as the source of this assumption.
September 29, 2015 1(M5"
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
EPA calculated monetary values of estimated changes in beneficial use of marketable CCRs based on the
avoided costs by steam electric power plants, as well as the avoided resource impacts from displacing virgin
materials. Table 10-3 lists rates and ratios for avoided costs and additionally-incurred costs associated with re-
using CCR for beneficial uses. Table 10-4 and Table 10-5 list the quantity and value of avoided resource and
environmental impacts.
Table 10-3: Economic Value of CCR Handling Costs per Unit (2013$).
Cost Type
Avoided Costs of Disposal
Additionally Incurred Costs of Beneficiation
Beneficial Use
Sand and Gravel Used as Fill;
Concrete
Concrete
Cost/Benefit
Varies by Steam Electric plant3
$11.95pertonb
a. EPA estimated the annual cost of CCR disposal by Steam Electric plant and year (U.S. EPA, 2015).
b. Electric Power Research Institute (2005).
Table 10-4: Avoided Resource and Environmental Impacts per Ton of Virgin Material Produced.
Impact Category
Energy (MMBtu)
Water (gal)
Air Emissions
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
Portland Cementa'c
3.52-3.62*
201-246,947*
0.90-0.93*
0.0019
2.43 x 10"4
Fillb'c
0.04
1,208
0.0032
2.68 x 10"5
8.50 x 10"6
a. Assumptions are based on Final CCR rule analysis. For energy use, water use, greenhouse gases, carbon monoxide, and PM10
associated with portland cement, the range of values presented in this Table reflect data from Ecoinvent v2.2 as included in SimaPro
7.2 and the National Renewable Energy Laboratory's U.S. Life Cycle Inventory Database, the latter of which is available at
http://www.nrel.gov/lci/. The Portland cement values for NOx, PM2.5, SOx, and mercury were all derived from emissions data
presented in U.S. EPA, Regulatory Impact Analysis: Amendments to the National Emission Standards for Hazardous Air Pollutants
and New Source Performance Standards (NSPS) for the Portland Cement Manufacturing Industry Final Report, August 2010. The
value for lead was derived from Ecoinvent v2.2 as included in SimaPro 7.2.
b. Assumptions are based on Final CCR rule analysis. Values for fill reflect the average of values for sand and clay from Ecoinvent
v2.2 as included in SimaPro 7.2.
* Denotes cases in which EPA used the lower limit of the range for this scoping analysis.
September 29, 2015
10-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
Table 10-5: Economic Value of Avoided Resource and Environmental Impacts per Unit of Impact
(2013$).
Impact Category
Energy ($/MMBtu)a
Water ($/gal)b
Greenhouse gases ($/ton)°
NOx ($/ton)d
SOx ($/ton)d
Portland Cement
$4.94
Fill
$19.15
$0.00003
$36 to $62
$5,694 to $8,836
$43,856 to $68,109
$5,694 to $8,836
$43,856 to $68, 109
a. Assumptions are based on Final CCR rule analysis. Derived from U.S. Department of Energy, Energy Consumption by
Manufacturers, May 2013. Table 7.2. This is the same data source as that used for the final CCR rule, updated to 2013$ using GDP
deflator.
b. Assumptions are based on Final CCR rule analysis, updated to 2013$ using GDP deflator. Value provided by H. Scott Matthews,
Professor of Civil and Environmental Engineering and Engineering and Public Policy at Carnegie Mellon University and Research
Director of Carnegie Mellon University's Green Design Institute, October 18, 2011.
c. Values shown are the range of yearly Social Cost of Carbon values from 2015 to 2042, using the Average Value at 3% discount rate
(adjusted to 2013$ and converted from tonnes to tons). Values are derived from Interagency Working Group on Social Cost of
Carbon, United States Government, Technical Support Document: Technical Update of the Social Cost of Carbon for Regulatory
Impact Analysis Under Executive Order 12866,2013 (revised June 2015). EPA also calculated benefits using the 5% average ($ 10 to
$22), 2.5% average ($50 to $85), and 3% 95th percentile ($85 to $188) SCC estimates.
d. Values shown are the range of yearly benefit per ton estimates (2015 to 2043) linearly interpolated and extrapolated from BenMap
benefit per ton estimates for 2016,2020, 2025, and 2030. Benefit per ton unit impacts were provided in personal communication with
Charles Fulcher, EPA/OAQPS, on October 19, 2012 and updated to 2013$ using GDP deflator. EPA applied values for the "Cement
Kilns" sector to Portland cement, and values for the "area sources" aggregate sector to fill. Values shown are the central tendency by
sector, at 3% discount rate.
suits
Table 10-6 summarizes the estimated change in dry and marketable CCRs beneficially used in concrete and
fill applications due to final ELGs and other regulatory options, for steam electric power plants not marketing
wet ash in the baseline. The analysis suggests that most marketable CCRs would be beneficially used in fill
applications.
Table 10-6: Estimated Beneficial Use Applications of CCRs, by CCR Category
ELG Option and CCR Category
Beneficial Use Application (1,000 short tons)
Concrete
Structural Fill
Option A or Option B
Fly ash
Bottom ash
Total
162
NA
162
1,166
0
1,166
Option C
Fly ash
Bottom ash
Total
162
NA
162
1,166
2,813
3,979
Option D or Option E
Fly ash
Bottom ash
Total
162
NA
162
1,166
3,763
4,929
September 29, 2015
10-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
Table 10-7 reports the estimated life cycle benefits of avoiding virgin materials in concrete and fill
production. Table 10-8 reports the estimated annualized economic value of increased marketable CCRs by
value category. Relative changes in total CCR beneficial use and total life cycle impacts under the final ELGs
are consistent with the projected effects of the CCR rule on beneficial reuse (shown in last column of the
table; U.S. EPA, 2014).
Table 10-7: Annual Avoided Resource and Environmental Impacts Given CCR Reuse in Concrete
and Fill Applications.
Impact Category
Fly Ash
Bottom Ash
Option Total
CCR RIA, 2030 3-
yr rolling avg
Option A or Option B
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
618,618
1,442
149,974
340
49
0
0
0
0
0
618,618
1,442
149,974
340
49
910,000
21,000
75,000
310
160
Option C
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
618,618
1,442
149,974
340
49
112,536
3,399
9,003
75
24
731,154
4,840
158,977
415
73
910,000
21,000
75,000
310
160
Option D or Option E
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
618,618
1,442
149,974
340
49
150,515
4,546
12,041
101
32
769,133
5,987
162,015
441
81
910,000
21,000
75,000
310
160
Note: Values in this table represent annual changes in a full-compliance year (e.g., starting in 2023).
September 29, 2015
10-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
10: Ash Marketing
Table 10-8: Annualized Economic Value of Estimated Changes in Beneficial Use (Million 2013$)
a,b
Regulatory Option
Option A or
Option B
Option D or
Option E
Impact Category
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials
Net Social Value
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials
Net Social Value
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials0
Net Social Value
3%
$0.80
-$0.21
$10.74
$11.33
$13.03
-$0.21
$12.18
$25.00
$18.49
-$0.21
$12.52
$30.80
7%
$0.68
-$0.16
$13.79
$14.31
$10.87
-$0.16
$15.43
$26.14
$15.46
-$0.16
$15.81
$31.11
Notes: a. Annualized over 24 years (2015 - 2042). Values escalated using CCI and GDP through 2022; thereafter, assume no real
change in prices above inflation. Avoided disposal costs to steam electric power plants include are annual O&M costs. B.
b. EPA used SCC values based on a 3 percent (average) discount rate to calculate total benefit values presented for both the 3
percent and 7 percent discount rate.
c. EPA also estimated life cycle benefits for Options D orE, at other SCC estimates (IWGSCC, 2013 (revised June 2015)). Results
of LCA benefits at 5% average, 2.5% average, and 3% 9th percentile SCC estimates range from $5.12 million (5% average) to
$19.90 million (3% 95thpercentile) when discounted at 3%, and from $3.81 million (5% average) to $14.87 million (3% 95th
percentile).
As shown in Table 10-8, for Options requiring both fly and bottom ash technologies, much of the annualized
economic value from changes in beneficial use comes from avoided disposal costs incurred by steam electric
plants, even after accounting for any increased beneficiation costs.
The estimates above translate into net annualized benefits of approximately $30.8 million (3 percent discount
rate) for Options D or E. Estimated benefits for Option C are lower due to the reduction in the amount of
marketable bottom ash relative to Options D and E (i.e., BAT/PSES standards for bottom ash apply only to
units with greater than 400 MW capacity), and benefits for Options A and B are lower yet, as they derive
from the incremental marketability of fly ash only.
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. Estimated benefits from the enhanced marketability of CCR
do not depend on pollutant loadings to receiving waters and therefore the monetized benefit estimates
provided in this Chapter are unaffected by revisions to steam electric plant loading estimates.
10.4 Limitations and Uncertainties
Key uncertainties and limitations include:
> Benefits from marketed CCR are sensitive to assumptions about CCR generation. The analysis
assumes that the amount of CCR generated by steam electric power plants is constant throughout the
analysis, i.e., coal-fired plants generate a constant quantity of electricity. This is consistent with the
screening-level cost and economic impact analysis framework (see RIA report). However, the
September 29, 2015
10-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 10: Ash Marketing
Department of Energy and EPA both project that generation from steam electric plants will decline
over time, due to a variety of market factors, including environmental regulations. This decline would
reduce the quantity of CCR generated, available to be marketed for beneficial use, and the associated
benefits estimated in this memo. The Agency accounted for disposal costs in its economic impact
analysis assuming that a constant quantity of CCR would need to be disposed of. Therefore, the
assumption of constant CCR generation does not result in an overstatement of avoided disposal
benefits since they offset costs already calculated elsewhere. In contrast, non-disposal related benefits
could be lower when accounting for the reduction in CCR generation.
> As discussed above, EPA assumed that conversions to dry handling at plants already marketing wet
ash in the baseline are not incremental economic benefits because there is no change in the total
amount of CCR that is beneficially used. Relaxing this assumption could increase estimated cost
offsets to the extent that these plants avoid certain compliance or operational costs that have been
counted against the ELGs.
> Steam electric plants may not market all of their CCR within a given year; for example, they may
instead temporarily store it. This affects both the time profile and value of beneficial reuse benefits.
To the extent that steam electric plants may or may not be able to completely avoid the annual O&M
costs of disposal, EPA's analysis is an upper bound.
> Certain other dimensions of this analysis may lead to under-estimated benefits of CCR re-use. For
example, EPA assumed that facilities only offset O&M costs of disposal. To the extent that some
plants are actually able to avoid or offset capital costs of compliance or other longer-term costs given
reduced need for landfilling (e.g., costs of securing and developing additional disposal sites), EPA's
analysis is a lower-bound on the benefits of dry handling CCR.
September 29, 2015 10-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
11 Summary of Total Monetized Benefits
11.1 Total Annualized Benefits
Table 11-1 and Table 11-2, on the next two pages, summarize the total annual monetized benefits using 3
percent and 7 percent discount rates, respectively. Table 11-3 and Table 11-4 compile, for each of the five
analyzed regulatory options, the time profiles of total (non-discounted) monetized benefits. The tables also
report the calculated present and annualized values of benefits at 3 percent and 7 percent discount rates,
respectively.
The estimated total monetized benefits of the five regulatory options EPA analyzed range from $41.1 million
to $565.6 million per year using a 3 percent discount rate, depending on the option ($37.2 million to
$478.4 million per year using a 7 percent discount rate), and whether the analysis includes air-related benefits.
Option D has total benefits (including air-related benefits) of approximately $450.6 million to $565.6 million
using a 3 percent discount rate and $387.3 million to $478.4 million using a 7 percent discount rate.
The monetized benefits of the final rule do not account for all benefits because they omit various sources of
benefits to society from reduced steam electric pollutant discharges, such as reduction in certain non-cancer
health risk (e.g., effects of cadmium on kidney functions and bone density) and reduced cost of drinking water
treatment. See Chapter 2 for a discussion of categories of benefits EPA did not monetize.
Finally, EPA was able to estimate air-related benefits for Options B and D only (see Chapter 7). Benefits for
options A, C and E are therefore understated to a greater degree than the other options; in particular, EPA
expects that the benefits for Option E would be higher than those for Option D if air-related benefits were
included. See Section 13.1 for an extrapolation of potential air-related benefits for these options.
As described in Section 1.4.3, EPA revised its estimates of pollutant loadings in steam electric power plant
discharges after completing the benefit analyses. These revisions are likely to reduce benefit estimates in the
following four categories: human health, nonmarket water quality benefits, benefits to threatened and
endangered species and avoided dredging costs. For several reasons — notably the fact that revisions do not
affect all plants equally and the model functions are not linear — it would be inappropriate to simply scale the
monetized benefits based on the aggregate changes in loadings. Therefore, EPA qualitatively assessed effects
of pollutant loading reductions on the relevant benefit categories (see Chapters 3, 4, 5, and 9 for detail). The
Agency concluded that although the revised loadings are likely to reduce the estimated benefit in four benefit
categories, the magnitude of these reductions is likely to be small compared to the total benefits of the rule.
Chapters 3 through 10 provide more detail on limitation and uncertainty inherent in the analysis of each
benefit category.
September 29, 2015
11-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-1: Summary of Total Annualized Benefits at 3 Percent (Millions; 2013$)
Benefit Category
Human Health Benefits8
Reduced IQ losses in children
from exposure to lead3'8
Reduced CVD in adults from
exposure to lead8
Reduced IQ losses in children
from exposure to mercury3'8
Avoided cancer cases from
exposure to arsenicb'8
Improved Ecological
Conditions and Recreational
Uses8
Use and nonuse values for
water quality improvements8
Nonuse values of T&E
species'3'8
Market and Productivity
Benefits
Avoided impoundment failures
Reduced dredging costs'3
Ash marketing benefits
Air-related benefits
Reduced human health effects
Reduced CO2 emissions0
Reduced water withdrawals'"
Option A
Low
$5.4
$0.3
Midf
$5.7
$0.4
High
$6.1
$0.5
$3.8
$1.3
$1.6
$1.8
<$0.1
$4.2
$4.2
<$0.1
$31.5
$20.1
$4.9
$4.9
<$0.1
$32.8
$21.5
$23.4
$23.4
<$0.1
$34.2
$22.9
<$0.1
$11.3
NE
NE
NE
<$0.1
Option B
Low
$5.4
$0.3
Midf
$5.8
$0.4
High
$6.1
$0.5
$3.8
$1.3
$1.6
$1.9
<$0.1
$15.0
$15.0
<$0.1
$31.5
$20.1
$18.9
$18.9
<$0.1
$32.8
$21.5
$83.7
$83.7
<$0.1
$34.2
$22.9
<$0.1
$11.3
$110.2
$57.8
$52.5
<$0.1
Option C
Low
$11.3
$0.5
Midf
$11.8
$0.6
High
$12.4
$0.7
$8.4
$2.4
$2.9
$3.4
<$0.1
$19.6
$19.6
<$0.1
$107.2
$82.2
$26.0
$26.0
<$0.1
$109.5
$84.5
$109.4
$109.4
<$0.1
$111.9
$86.9
<$0.1
$25.0
NE
NE
NE
<$0.1
Option D
Low
$16.5
$0.8
Midf
$17.2
$1.0
High
$17.9
$1.1
$12.8
$2.9
$o c
3.5
$4.0
<$0.1
$23.3
$23.2
<$0.1
$126.4
$95.6
$31.3
$31.3
<$0.1
$130.0
$99.2
$129.5
$129.5
<$0.1
$133.7
$102.9
<$0.1
$30.8
$284.5
$144.7
$139.8
<$0.1
Option E
Low
$16.7
$0.8
Midf
$17.5
$1.0
High
$18.3
$1.1
$12.8
$3.1
$3.8
$4.4
<$0.1
$25.1
$25.1
<$0.1
$126.4
$95.6
$34.0
$34.0
<$0.1
$130.0
$99.2
$140.0
$140.0
<$0.1
$133.7
$102.9
<$0.1
$30.8
NE
NE
NE
<$0.1
September 29, 2015
11-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-1: Summary of Total Annualized Benefits at 3 Percent (Millions; 2013$)
Benefit Category
Total (excluding air-related
Benefits)d
Total (including air-related
Benefits)"1'6
Option A
Low
$41.1
NE
Midf
$43.4
NE
High
$63.7
NE
Option B
Low
$51.9
$162.2
Midf
$57.5
$167.8
High
$124.0
$234.2
Option C
Low
$138.1
NE
Midf
$147.4
NE
High
$233.7
NE
Option D
Low
$166.1
$450.6
Midf
$178.5
$463.0
High
$281.2
$565.6
Option E
Low
$168.3
NE
Midf
$181.6
NE
High
$292.0
NE
Source: U.S. EPA Analysis, 2015
"NE" indicates that EPA did not estimate the benefits. Air-related benefits of Option A are expected to be less than those for Option B; air-related benefits for Option C are expected to be
between those of Options B and D; and air-related benefits of Option E are expected to be greater than those for Option D.
a. Value includes reduced IQ losses and avoided cost of compensatory education in children from exposure to lead. For details see Chapter 3.
b. "< $0.1" indicates that the monetized annual benefits are positive but less than $0.1 million.
c. For the valuation of benefits from reductions in CO2 emissions EPA relied on the 3 percent average social cost of carbon estimate.
d. Values for individual benefit categories may not sum to the total due to independent rounding.
e. The total monetized benefits for options A, C, and E do not include air-related benefits. This category of benefits was analyzed for Options B and D only (see Chapter 7).
f EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of willingness-to-pay. One model provides the low and high bounds
while a different model provides a central estimate (included in this table in the mid-range column). For this reason, the mid-range estimate differs from the midpoint of the range for this
benefit category. For details, see Chapter 4.
g. Estimates for this benefit category do not reflect revised pollutant loadings, which could result in lower monetized benefits. See Section 1.4.3 for details.
September 29, 2015
11-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-2: Summary of Total Annualized Benefits at 7 Percent (Millions; 2013$)
Benefit Category
Human Health Benefits8
Reduced IQ losses in children
from exposure to lead3'8
Reduced CVD in adults from
exposure to lead8
Reduced IQ losses in children
from exposure to mercury3'8
Avoided cancer cases from
exposure to arsenicb'8
Improved Ecological
Conditions and Recreational
Uses8
Use and nonuse values for
water quality improvements8
Nonuse values of T&E
species'3'8
Market and Productivity
Benefits
Avoided impoundment failures
Reduced dredging costs'3
Ash marketing benefits
Air-Related Benefits
Reduced human health effects
Reduced CO2 emissions0
Reduced water withdrawals'"
Option A
Low
$3.4
$0.1
Midf
$3.5
$0.1
High
$6.7
$0.1
$3.1
$0.2
$0.3
$o c
3.5
<$0.1
$3.3
$3.3
<$0.1
$30.4
$16.1
$3.9
$3.9
<$0.1
$31.6
$17.3
$18.6
$18.6
<$0.1
$32.7
$18.4
<$0.1
$14.3
NE
NE
NE
<$0.1
Option B
Low
$3.4
$0.1
Midf
$3.5
$0.1
High
$14.2
$0.1
$3.1
$0.2
$0.3
$11.0
<$0.1
$12.0
$12.0
<$0.1
$30.4
$16.1
$15.2
$15.2
<$0.1
$31.6
$17.3
$66.7
$66.7
<$0.1
$32.7
$18.4
<$0.1
$14.3
$103.6
$51.1
$52.5
<$0.1
Option C
Low
$7.4
$0.1
Midf
$7.6
$0.1
High
$7.7
$0.1
$7.0
$0.4
$0.5
$0.6
<$0.1
$15.7
$15.7
<$0.1
$92.5
$66.4
$20.9
$20.9
<$0.1
$94.4
$68.3
$87.4
$87.3
<$0.1
$96.4
$70.3
<$0.1
$26.1
NE
NE
NE
<$0.1
Option D
Low
$11.3
$0.1
Midf
$11.4
$0.2
High
$11.6
$0.2
$10.7
$0.5
$0.6
$0.7
<$0.1
$18.6
$18.5
<$0.1
$108.8
$77.7
$25.1
$25.1
<$0.1
$111.8
$80.7
$103.4
$103.4
<$0.1
$114.8
$83.7
<$0.1
$31.1
$248.6
$108.8
$139.8
<$0.1
Option E
Low
$11.3
$0.1
Midf
$11.5
$0.2
High
$11.6
$0.2
$10.7
$0.5
$0.6
$0.8
<$0.1
$20.1
$20.0
<$0.1
$108.8
$77.7
$27.3
$27.3
<$0.1
$111.8
$80.7
$111.7
$111.7
<$0.1
$83.7
$83.7
<$0.1
$31.1
NE
NE
NE
<$0.1
September 29, 2015
11-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-2: Summary of Total Annualized Benefits at 7 Percent (Millions; 2013$)
Benefit Category
Total (excluding air-related
Benefits)d
Total (including air-related
Benefits)"1'6
Option A
Low
$37.2
NE
Midf
$38.9
NE
High
$58.0
NE
Option B
Low
$45.8
$149.4
Midf
$50.2
$153.8
High
$113.7
$217.3
Option C
Low
$115.6
NE
Midf
$122.9
NE
High
$191.4
NE
Option D
Low
$138.7
$387.3
Midf
$148.4
$397.0
High
$229.8
$478.4
Option E
Low
$140.2
NE
Midf
$150.6
NE
High
$207.1
NE
Source: U.S. EPA Analysis, 2015
"NE" indicates that EPA did not estimate the benefits. Air-related benefits of Option A are expected to be less than those for Option B; air-related benefits for Option C are expected to be
between those of Options B and D; and air-related benefits of Option E are expected to be greater than those for Option D.
a. Value includes reduced IQ losses and avoided cost of compensatory education in children from exposure to lead. For details see Chapter 3.
b. "< $0.1" indicates that the monetized annual benefits are positive but less than $0.1 million.
c. For the valuation of benefits from reductions in CO2 emissions EPA relied on the 3 percent average social cost of carbon estimate.
d. Values for individual benefit categories may not sum to the total due to independent rounding.
e. The total monetized benefits for options A, C, and E do not include air-related benefits. This category of benefits was analyzed for Options B and D only (see Chapter 7).
f EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of willingness-to-pay. One model provides the low and high bounds
while a different model provides a central estimate (included in this table in the mid-range column). For this reason, the mid-range estimate differs from the midpoint of the range for this
benefit category. For details, see Chapter 4.
g. Estimates for this benefit category do not reflect revised pollutant loadings, which could result in lower monetized benefits. See Section 1.4.3 for details.
September 29, 2015
11-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 11: Total Monetized Benefits
11.2 Time Profile of Benefits
Table 11-3 and Table 11-4 compile the time profiles of total (non-discounted) monetized benefits including
air-related benefits for Options B and D, and the calculated annualized values of benefits at 3 percent and 7
percent discount rates, respectively. The time profile is based on mid-range (or central) estimates for benefits
for which benefits are presented as a range in Table 11-1 and Table 11-2.
As shown in the tables, benefits under Option D increase from 2019 until 2022, then decline in 2023 before
gradually increasing. This 2023 decline is due to an IPM-projected increase in SO2 relative to baseline
(negative benefits). For Option B, there is a similar decline in 2023 as well as a larger decline in benefits in
2028. As with Option D, these declines are attributable to IPM-projected increases in SO2 (negative benefits),
but also occur due to smaller estimated CO2 and NOX reductions when compared to earlier years. For both
Options B and D (excluding the increases in SO2 in the given periods), IPM indicates a decline in net air
emissions reductions between 2019 and 2033 followed by an increase in net reductions in 2034. For more
details on the IPM projections, see Chapter 7.
September 29, 2015 11-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-3: Time Profile of Benefits at 3 Percent (Millions; 2013$) (Including Air-Related Benefits for
Options B and D)
Year
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Benefits,
3%b
Option Aa
$0.0
$0.0
$0.0
$0.0
$3.8
$5.5
$18.7
$31.0
$57.2
$58.0
$58.4
$58.6
$58.7
$58.7
$58.6
$58.5
$58.6
$58.6
$58.6
$58.6
$58.7
$58.7
$58.8
$58.9
$59.0
$59.2
$59.3
$59.3
$43.4
Option B
$0.0
$0.0
$0.0
$0.0
$291.4
$296.6
$327.6
$344.4
$176.1
$179.1
$181.5
$184.0
$186.3
$31.1
$30.4
$29.9
$29.6
$29.2
$28.8
$224.3
$226.7
$229.2
$231.7
$234.2
$236.8
$239.3
$241.9
$243.2
$167.8
Option Ca
$0.0
$0.0
$0.0
$0.0
$16.4
$39.9
$83.3
$155.8
$189.9
$191.6
$192.3
$192.6
$192.7
$192.7
$192.5
$192.3
$192.3
$192.2
$192.1
$192.1
$192.1
$192.1
$192.2
$192.4
$192.5
$192.7
$192.9
$193.1
$147.4
Option D
$0.0
$0.0
$0.0
$0.0
$544.8
$578.7
$633.5
$718.1
$310.5
$314.9
$317.6
$319.9
$321.8
$422.6
$423.2
$426.2
$428.8
$431.6
$434.4
$670.1
$676.4
$682.7
$689.1
$695.6
$702.1
$708.6
$715.3
$718.6
$463.0
Option Ea
$0.0
$0.0
$0.0
$0.0
$30.3
$58.7
$115.1
$193.2
$232.6
$235.1
$235.9
$236.2
$236.2
$236.0
$235.5
$235.1
$234.8
$234.4
$234.1
$233.8
$233.6
$233.5
$233.4
$233.4
$233.4
$233.5
$233.7
$233.8
$181.6
Source: U.S. EPA Analysis, 2015
a. Estimates for Options A, C and E do not include air-related benefits. This category of benefits was only estimated for Options B and
D (see Chapter 7).
b. Total annualized year-specific benefits may not sum to the total annualized benefits due to rounding.
September 29, 2015
11-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
11: Total Monetized Benefits
Table 11-4: Time Profile of Benefits at 7 Percent (Millions; 2013$) (Including Air-Related Benefits for
Options B and D)
Year
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Benefits,
7%b
Option Aa
$0.0
$0.0
$0.0
$0.0
$3.8
$5.5
$16.9
$29.2
$55.4
$56.2
$56.5
$56.7
$56.8
$56.8
$56.6
$56.5
$56.6
$56.6
$56.6
$56.6
$56.6
$56.7
$56.8
$56.8
$56.9
$57.0
$57.2
$57.2
$38.9
Option B
$0.0
$0.0
$0.0
$0.0
$270.0
$275.5
$304.5
$321.3
$170.3
$173.3
$175.7
$178.2
$180.4
$34.2
$33.5
$33.0
$32.7
$32.3
$31.9
$214.9
$217.3
$219.8
$222.3
$224.8
$227.3
$229.8
$232.4
$233.7
$153.8
Option Ca
$0.0
$0.0
$0.0
$0.0
$16.4
$39.9
$80.1
$152.5
$186.6
$188.3
$188.9
$189.2
$189.3
$189.3
$189.1
$188.9
$188.8
$188.6
$188.5
$188.5
$188.5
$188.5
$188.5
$188.6
$188.8
$188.9
$189.1
$189.2
$122.9
Option D
$0.0
$0.0
$0.0
$0.0
$512.5
$547.5
$597.4
$682.1
$309.7
$314.0
$316.4
$318.8
$320.7
$409.6
$410.2
$413.0
$415.6
$418.3
$421.1
$641.9
$648.1
$654.4
$660.7
$667.1
$673.5
$680.0
$686.6
$689.8
$397.0
Option Ea
$0.0
$0.0
$0.0
$0.0
$30.3
$58.7
$110.8
$188.8
$228.2
$230.6
$231.4
$231.6
$231.6
$231.3
$230.8
$230.4
$230.0
$229.6
$229.2
$228.9
$228.7
$228.5
$228.4
$228.4
$228.3
$228.4
$228.5
$228.5
$150.6
Source: U.S. EPA Analysis, 2015
a. Estimates for Options A, C and E do not include air-related benefits. This category of benefits was only estimated for Options B and
D (see Chapter 7).
b. Total annualized year-specific benefits may not sum to the total annualized benefits due to rounding. Additionally, note that SCC
values are annualized at 3 percent, whereas other benefits are annualized at 7 percent.
September 29, 2015
11-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
12: Total Social Costs
12 Summary of Total Costs
This chapter develops EPA's estimates of the costs to society resulting from the final ELGs. As analyzed in
this chapter, the costs of regulatory actions are the opportunity costs to society of employing resources to
prevent the environmental damage otherwise occurring from discharges of wastewater containing metals,
nutrients, and other pollutants.
12.1 Overview of Costs Analysis Framework
RIA Chapter 3: Compliance Costs presents EPA's development of costs to the 1,080 steam electric power
plants subject to the final ELGs (U.S. EPA, 2015c). These costs are used as the basis of the social cost
analysis. However, the compliance costs used to estimate total costs differ in their consideration of taxes from
those reported in RIA Chapter 3, which were calculated for the purpose of estimating the private costs and the
economic impacts of the ELGs. In the analysis of costs to society, compliance costs are considered without
accounting for any tax effects. The costs to society are the full value of the resources used, whether they are
paid for by the regulated plants, by taxpayers in the form of lost tax revenues, or by some combination.74
As described in Chapter 1, EPA assumed that steam electric power plants, in the aggregate, would implement
control technologies during a 5-year period from 2019 to 2023. For this analysis, EPA developed a year-
explicit schedule of compliance outlays over the period of 2019 through 2042.75 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.
After compliance costs were assigned to the year of occurrence, the Agency adjusted these costs for real
change between their stated year 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.
Note that the CCI and ECI adjustment factors were developed only through the year 2022; 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 assumption, given 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 promulgation year by
discounting the cost in each year back to 2015, 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
For the impact analyses, compliance costs are measured as they affect the financial performance of the
regulated plants and firms. The economic impact analyses therefore consider the tax deductibility of compliance
expenditures, as appropriate depending on the tax status of the complying entity.
75 The end of the analysis period, 2042, was determined based on the life of the longest-lived compliance
technology implemented at any steam electric power plant (20 years), and the last year of technology
implementation (2023).
September 29, 2015 12~T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 12: Total Social Costs
(U.S. OMB, 2003). EPA calculated the constant annual equivalent value (annualized value), again using the
two values of the discount rate, 3 percent and 7 percent, over a 24-year social cost analysis period. EPA
assumed no re-installation of compliance technology during the period covered by the social cost analysis.
To assess the economic costs of the ELGs 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
ELGs. In this analysis, EPA assumed that 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. Finally, EPA assumed in its social cost analysis that the final rule does not affect the aggregate
quantity of electricity that would be sold to consumers and, thus, that the rule's social cost would include no
loss in consumer and producer surplus from lost electricity sales by the electricity industry in aggregate.
Given the small impact of the final rule on electricity production cost for the total industry, EPA believes that
this assumption is reasonable for the social cost analysis (for more details on the impacts of the final rule on
electricity production cost, see RIA Chapter 5: Electricity Market Analyses) ^ThQ 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 promulgation and technology
implementation years.76
Finally, as discussed in RIA Chapter 10 (Section 10.7: Paperwork Reduction Act of 1995), the final ELGs are
not expected to result in additional administrative costs for plants to implement, and state and federal NPDES
permitting authorities to administer, the final ELGs. 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 ELGs.
12.2 Key Findings for Regulatory Options
Table 12-1 presents annualized costs for each of the five regulatory options. At a 3 percent discount rate,
estimated annualized costs range between $120.5 million under Option A and $536.0 million under Option E.
The final BAT/PSES options for existing sources (Option D) have annualized costs of $479.5 million at a
3 percent discount rate, and $471.2 million at a 7 percent discount rate.77
Table 12-1: Summary of Annualized Costs (Millions; $2013)
Regulatory Option
Option A
Option B
Option C
Option D
Option E
3% Discount Rate
$120.5
$198.7
$383.5
$479.5
$536.0
7% Discount Rate
$116.9
$194.7
$379.9
$471.2
$525.8
Source: U.S. EPA Analysis, 2015.
Table 12-2 provides additional detail on the social cost calculations. The table compiles, for each of the five
regulatory options EPA analyzed for the final ELG, the time profiles of compliance costs incurred. The table
The specific assumptions of when each cost component is incurred can be found in Chapter 3: Compliance
Costs of the RIA.
77 Similarities in the values obtained when discounting costs using 3 percent and 7 percent are due to the time
profile of the costs, specifically the timing and relative magnitude of upfront capital costs versus ongoing O&M
expenditures.
September 29, 2015 12~T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
12: Total Social Costs
also reports the calculated annualized values of costs at 3 percent and 7 percent discount rates.78 The
maximum compliance outlays are incurred over the years 2019 through 2023, i.e., during the estimated
window when steam electric power plants are expected to implement compliance technologies.
Table 12-2: Time Profile of Costs to Society (Millions; $2013)
Year
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Costs, 3%
Annualized Costs, 7%
Option A
$0.0
$0.0
$0.0
$0.0
$388.8
$132.3
$390.2
$201.7
$286.6
$75.7
$83.0
$78.3
$80.9
$79.0
$84.9
$83.2
$85.4
$84.6
$85.1
$77.8
$83.1
$79.0
$82.6
$81.4
$84.4
$83.9
$83.8
$83.3
$120.5
$116.9
Option B
$0.0
$0.0
$0.0
$0.0
$641.8
$276.7
$620.2
$374.2
$512.9
$120.2
$127.5
$122.7
$125.4
$123.4
$129.3
$127.7
$129.8
$129.0
$129.5
$122.2
$127.5
$123.5
$127.1
$125.8
$128.8
$128.3
$128.2
$127.7
$198.7
$194.7
Option C
$0.0
$0.0
$0.0
$0.0
$1,143.4
$631.1
$1,243.4
$882.5
$1,058.8
$215.3
$222.9
$219.0
$219.9
$218.6
$230.9
$227.1
$232.1
$229.5
$230.7
$218.0
$223.8
$221.5
$224.0
$224.0
$229.4
$226.5
$226.9
$224.9
$383.5
$379.9
Option D
$0.0
$0.0
$0.0
$0.0
$1,336.6
$856.7
$1,364.9
$1,123.1
$1,332.7
$282.3
$291.2
$289.5
$289.1
$286.6
$304.2
$298.6
$303.7
$303.4
$302.9
$286.8
$293.4
$290.7
$295.7
$295.2
$300.3
$298.0
$297.2
$293.9
$479.5
$471.2
Option E
$0.0
$0.0
$0.0
$0.0
$1,427.1
$950.3
$1,498.0
$1,331.8
$1,497.4
$317.9
$327.5
$326.0
$325.9
$324.1
$341.2
$334.2
$340.0
$340.0
$339.7
$324.2
$330.5
$326.3
$332.0
$331.5
$336.7
$334.5
$333.2
$329.5
$536.0
$525.8
Source: U.S. EPA Analysis, 2015.
Whereas EPA calculated the time profile of benefits using discount rate-specific social cost of carbon estimates
and therefore obtained time profiles that are specific to the discount rates (see Table 11-3 and Table 11-4), the
time profile of costs does not depend on the discount rate.
September 29, 2015
12-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
13: Benefits and Social Costs
13 Benefits and Costs
This chapter compares total monetized benefits and costs for the five regulatory options analyzed for the final
ELGs. Benefits and costs are compared on two bases: (1) for each of the options analyzed 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: Other Administrative Requirements of the RIA; U.S. EPA, 2015c).
13.1 Comparison of Benefits and Costs by Option
Chapter 11 and Chapter 12 present estimates of the benefits and costs, respectively, for the regulatory options
evaluated in developing the final ELGs.
Table 13-1 presents EPA's estimates of benefits and costs of the regulatory options for existing steam electric
power plants, at 3 percent and 7 percent discount rates, and annualized over 24 years. These values are all in
2013 dollars and are based on the discounting of costs and benefits to 2015, the rule promulgation year.
As discussed in Chapter 11, EPA did not analyze air-related benefits for Options A, C, and E. The total
monetized benefits for those options are therefore understated. To compare the costs and benefits of these
options, EPA calculated the average ratio of total benefits with air-related benefits to total benefits without air-
related benefits for Options B and D,79 then applied the average ratio to Options A, C, and E to extrapolate
total monetized benefits including air-related benefits for these options. These extrapolated, approximate
benefits are shown in Table 13-1.
Table 13-1: Total Annualized Benefits and Costs by Regulatory Option and Discount Rate (Millions;
2013$)
Regulatory
Option
Total Monetized Benefits
Low
Midb
High
Total Monetized Benefits, Including
Extrapolated Values"
Low
Midb
High
Total
Costs
3% Discount Rate
Option A
Option B
Option C
Option D
Option E
$41.1
$162.2
$138.1
$450.6
$168.3
$43.4
$167.8
$147.4
$463.0
$181.6
$63.7
$234.2
$233.7
$565.6
$292.0
$122.1
$162.2
$410.4
$450.6
$500.2
$122.2
$167.8
$414.5
$463.0
$510.8
$125.7
$234.2
$461.3
$565.6
$576.3
$120.5
$198.7
$383.5
$479.5
$536.0
This ratio averaged 2.97 for low benefits, 2.81 for mid-range, and 1.97 for high.
September 29, 2015
13-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
13: Benefits and Social Costs
table 13-1: Total Annualized Benefits and Costs by Regulatory Option and Discount Rate (Millions;
013$)
Regulatory
Option
Total Monetized Benefits
Low
Midb
High
Total Monetized Benefits, Including
Extrapolated Values"
Low
Midb
High
Total
Costs
7% Discount Rate
Option A
Option B
Option C
Option D
Option E
$37.2
$149.4
$115.6
$387.3
$140.2
$38.9
$153.8
$122.9
$397.0
$150.6
$58.0
$217.3
$191.4
$478.4
$207.1
$110.5
$149.4
$343.7
$387.3
$416.8
$109.5
$153.8
$345.6
$397.0
$423.7
$114.4
$217.3
$377.8
$478.4
$408.7
$116.9
$194.7
$379.9
$471.2
$525.8
Source: U.S. EPA Analysis, 2015.
a. EPA did not analyze air-related benefits for Options A, C, and E. This category of benefits was only estimated for Options B and
D (see Chapter 7). EPA adjusted the total benefits estimated for Options A, C and E by multiplying the totals without air-related
benefits by the average ratio of [total with air-related benefits]/[total without air-related benefits] for Options B and D.
b. EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of WTP. One
model provides the low and high bounds while a different model provides a central estimate (included in this table under the mid-
range column). For this reason, the mid benefit estimate differs from the midpoint of the benefits range. For details, see Chapter 4.
13.2 Analysis of Incremental Benefits and Costs
In addition to comparing benefits and costs for each regulatory option, as presented in the preceding section,
EPA also analyzed 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 by itself: for a given option, which is greater - costs or benefits - 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 (i.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.
EPA conducted the incremental net benefit analysis by calculating, for the five regulatory options, the change
in net benefits, from option to option, in moving from the least stringent option to successively more stringent
options. As described in Chapter 1, the regulatory options differ in the technology basis used to determine
effluent limits and standards 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, and the distribution and characteristics of steam electric
power plants that would implement the technologies and of the receiving waterbodies.
As noted previously, however, the total monetized benefits for Options A, C, and E do not include air-related
benefits; this benefit category is included in results for Options B and D only. Therefore, to allow for
consistent calculation of incremental benefits as one moves from one option to the next, EPA used the
extrapolated total benefits estimated for Options A, C, and D to include the air-related benefits for these
options (see Table 13-1).
As reported in Table 13-2, EPA estimates that the annual monetized costs exceed the mid-range annual
monetized benefits for the final ELG by $16.5 million using a 3 percent discount rate and $74.2 million using
a 7 percent discount rate.
September 29, 2015
13-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
13: Benefits and Social Costs
Using a 3 percent discount rate, the incremental net annual monetized benefits moving from Option A to
Option B is -$33 million (the negative value indicates that the increase in costs is larger than the increase in
benefits). Moving from Option B to Option C, the change is $62 million, with the positive value indicating
that the increase in benefits is larger than the increase in costs. Moving from Option C to Option D, and from
Option D to Option E, the change is negative, at -$48 million and -$9 million, respectively.
Table 13-2: Incremental Net Benefit Analysis (Millions; 2013$)
Regulatory
Option"
Total Annual Monetized
Benefits, Including
Adjusted or Inferred
Low
Values
Midd
High
Total
Social
Costs
Net Annual Monetized
Low
Benefits"
Midd
High
Incremental Net Annual
Monetized Benefits'"
Low
Midd
High
3% Discount Rate
Option A
Option B
Option C
Option D
Option E
$122.1
$162.2
$410.4
$450.6
$500.2
$122.2
$167.8
$414.5
$463.0
$510.8
$125.7
$234.2
$461.3
$565.6
$576.3
$120.5
$198.7
$383.5
$479.5
$536.0
$1.6
-$36.5
$26.9
-$28.9
-$35.8
$1.6
-$30.9
$31.0
-$16.5
-$25.2
$5.2
$35.6
$77.8
$86.1
$40.3
-
-$38.1
$63.4
-$55.8
-$6.8
-
-$32.6
$61.9
-$47.5
-$8.7
-
$30.4
$42.2
$8.3
-$45.8
7% Discount Rate
Option A
Option B
Option C
Option D
Option E
$110.5
$149.4
$343.7
$387.3
$416.8
$109.5
$153.8
$345.6
$397.0
$423.7
$114.4
$217.3
$377.8
$478.4
$408.7
$116.9
$194.7
$379.9
$471.2
$525.8
-$6.4
-$45.3
-$36.2
-$83.9
-$109.0
-$7.4
-$40.9
-$34.3
-$74.2
-$102.1
-$2.5
$22.5
-$2.1
$7.2
-$117.1
-
-$38.9
$9.1
-$47.7
-$25.1
-
-$33.5
$6.6
-$39.9
-$27.9
-
$25.0
-$24.6
$9.2
-$124.3
Source: U.S. EPA Analysis, 2015.
a. EPA did not analyze air-related benefits for Options A, C, and E. This category of benefits was only estimated for Options B and D
(see Chapter 7). EPA adjusted the total benefits estimated for Options A, C, and E by multiplying the totals without air-related
benefits by the average ratio of [total with air-related benefits]/[total without air-related benefits] for Options B and D.
b. Net benefits are calculated by subtracting total annualized costs from total annual monetized benefits.
c. Incremental net benefits are equal to the difference between net benefits of an option and net benefits of the previous, less stringent
option.
d. EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of WTP. One
model provides the low and high bounds while a different model provides a central estimate (included in this table under the mid-
range column). For this reason, the mid benefit estimate differs from the midpoint of the benefits range. For details, see Chapter 4.
September 29, 2015
13-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 14: Environmental Justice
14 Environmental Justice
Executive Order (E.G.) 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 environmental justice
(EJ) part of its mission. E.O. 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 E.O. 12898, EPA examined whether the benefits from the final ELGs 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 benefits arising from the final ELGs.
As described in the following sections, EPA conducted two types of analyses to evaluate the EJ implications
of the final ELGs: (1) summarizing the demographic characteristics of the households living in proximity to
reaches that receive steam electric power plant discharges; and (2) analyzing the human health impacts from
consuming self-caught fish on minority and/or low-income populations located within 50 miles of reaches
affected by steam electric power plant discharges.80 This second analysis seeks to provide more specific
insight on the distribution of adverse health effects and benefits and to assess whether minority and/or low-
income populations incur disproportionally high environmental impacts and/or are disproportionally excluded
from realizing the benefits of this final rule.
The following two sections describe (1) a comparison of the socio-economic characteristics of the populations
that live in proximity to steam electric power plants to state and national averages, and (2) the evaluation of
human health effects and benefits that accrue to populations in different socio-economic cohorts.
14.1 Socio-economic 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 reaches that receive steam electric power plant discharges. The receiving reaches are those to
which plants discharge directly in the baseline; for this first analysis, EPA did not include additional reaches
located downstream from the receiving reaches (see BCA for a discussion of receiving and downstream
reaches, U.S. EPA 2015a). The analysis is similar to the profile EPA had developed to support the proposed
ELG, but looks at populations living within different, closer distances of steam electric power plants. The
change was made in part to addresses comments EPA received on the need to look at communities in closer
proximity to the plants, instead of the 100-mile buffer the Agency had used at proposal (see comment
response document; U.S. EPA 2015d).81
As detailed in the Chapter 3, EPA used a distance of 50 miles to determine the affected population.
81 Commenters recommended that EPA consider the characteristics of populations within 4 miles, 5 miles,
15 miles and 30 miles of the plants (see U.S. EPA, 2015d).
September 29, 2015 14T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
EPA collected population-specific Census data on:
> the percent of the population below the poverty threshold,82 and
> the population categorized in various racial/ethnic groups, from which EPA calculated the percent of
the population that belongs to a minority racial/ethnic group.83
EPA compiled these data for CBGs located within specified distances (e.g., 1 mile, 3 miles, 15 miles, 30
miles, and 50 miles) of the reaches receiving steam electric power plant discharges. EPA compared
demographic metrics 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 below the poverty threshold is higher than the
state or national average or the percent of the population that is minority is above the state or national
average.
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 in proximity to reaches receiving steam electric
power plant discharges. The specified distance buffers from the reaches are denoted below as the "benefit
region." Populations within the benefit regions may be affected by steam electric power plant discharges and
other environmental impacts in the baseline, and would be expected to benefit from environmental
improvements resulting from the final ELGs. If the population within a benefit region has a larger proportion
of minority or low-income families than the state average, it may indicate that the final ELGs may benefit
communities that have been historically exposed to a disproportionate share of environmental impacts and
thus contribute to redressing existing EJ concerns.
EPA used the U.S. Census Bureau's American Community Survey (ACS) data for 2006 to 2010 to identify
poverty status (Table C17002) at the state and CBG levels. EPA also used 2010 U.S. Census data (Summary
File 1; Table 8 - P3) to identify the percent of the population that is minority at the CBG and state levels.
EPA overlaid the data with GIS data of buffer zones of specified distances from receiving reaches to
characterize the affected communities living in proximity to the reaches. Table 14-1 summarizes the socio-
economic characteristics of benefit regions defined using radial distances of 1,3, 10, 15,30 and 50 miles from
the receiving reaches.
Table 14-1: Socio-economic Characteristics of Communities Living in Proximity to Receiving
Reaches
Distance from receiving
reach
1 mile
3 miles
15 miles
30 miles
50 miles
United States
Total population
(millions)
0.2
1.1
14.0
37.4
57.3
306.3
Percent minority
20.7%
23.4%
29.8%
31.5%
29.9%
36.0%
Percent below poverty
level
16.4%
15.3%
13.6%
13.1%
13.1%
13.9%
Source: U.S. EPA analysis, 2015
82 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.
83 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/Pacific Islander (non-Hispanic);
Tribal/Native Alaskan (non-Hispanic); Other non-Hispanic; Mexican Hispanic, and Other Hispanic.
September 29, 2015
14-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
As shown in Table 14-1, approximately 200,000 people live within 1 mile of steam electric power plants
currently discharging to surface waters, over 1.1 million live within 3 miles, and nearly 37.4 million people
live within 30 miles. The statistics also show that a greater fraction of the communities living in close
proximity to steam electric power plants is poor, when compared to the national average. Approximately
16.4 percent of households in communities within 1 mile of steam electric power plants have income below
the poverty level as compared to a national average of 13.9 percent. A smaller fraction of the population
within 1 mile of the plants belongs to minority racial or ethnic groups (20.7 percent), than the national
average (36.0 percent). As one moves further away from the steam electric power plants, the fraction of the
community that is below the poverty threshold goes down while the percent minority increases, so that the
overall composition of the communities approaches that of the U.S. population overall. Thus, looking at
communities within 30 miles of steam electric power plants, 13.1 percent of the population is below the
poverty level (vs. 13.9 percent nationally) while 31.5 percent belong to a minority group (vs. 36.0 percent
nationally).
The simple comparison to the national average masks important differences, however, between states,
particularly given the non-uniform geographical distribution of plants across the country. EPA therefore also
compared the demographic profile of affected communities to that of the state where they are located. Table
14-2 summarizes the results of this comparison. Of the 37 states with communities within 1 mile of steam
electric power plants, 11 states have communities with a higher percentage of households below the poverty
threshold than the overall state, 17 have a higher percent of the population that is minority, and 10 have a
higher proportion of poor and minority households. These results show the potential for localized differences
indicative of potential EJ concern, but the overall comparison reveals no systematic difference in demographic
characteristics of the populations living in proximity to the steam electric power plants that would indicate
that any communities with EJ concern would be precluded from the benefits of the final ELGs.
Table 14-2: Socio-economic Characteristics of Affected Communities, Compared to State Average
Distance from
receiving
reach
1 mile
3 miles
15 miles
30 miles
50 miles
Number of States
with Affected
Communities3
37
37
38
42
45
Number of States where Affected Communities...
are Poorer
have a Higher
Proportion of
Minority Population
are Poorer and have a
Higher Proportion of
Minority Population
... than the State Average
11
10
21
22
19
17
17
16
16
9
10
7
19
21
12
Source: U.S. EPA analysis, 2015
a. "Affected communities" are Census Block Groups within the specified distance of one or more receiving reaches.
14.2 Distribution of Human Health Impacts and Benefits
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 final rule may mitigate them.
A significant share of the benefits of the final ELGs comes from reducing discharges of harmful pollutants to
surface waters and associated reductions in fish tissue contamination. EPA quantified the human health
September 29, 2015
14-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
benefits resulting from reducing exposure to pollutants in fish tissue in individuals who consume fish caught
in reaches immediately receiving or downstream from steam electric plant discharges. The analysis relied on
CBG- level data to estimate the number and characteristics of individuals exposed to steam electric pollutants
through the consumption of self-caught fish, and race and ethnicity-specific assumptions to estimate
exposure. This analysis allows the agency to report the distribution of benefits across population subgroups,
including subgroups who may have been historically exposed to a disproportionate share of environmental
impacts.
This section presents results for the two types of anglers analyzed: recreational anglers and subsistence
fishers. Chapter 3 provides more details on the approach used to identify the affected population, derive
exposure, quantify health effects and monetize benefits.
EPA limited its analysis of the distribution of health effects and benefits to two pollutants (lead and mercury)
due to the small benefits resulting from reducing arsenic discharges. Further, for recreational anglers, EPA
focused on benefits accruing to infants and children due to the complexity of carrying the detailed socio-
economic data through the models used to quantify the changes in premature mortality from cardiovascular
disease (CVD) in adults. The outputs from the CVD benefit analyses do not provide sufficient information to
assess changes across socio-economic subgroups. However, EPA did account for changes in the incidence of
CVD attributable to the final ELGs when comparing recreational and subsistence fishers (see Section 14.4).
Table 14-3 summarizes the estimated number of individuals exposed to steam electric pollutants through
consumption of self-caught fish in the general population and in population subgroups that may be indicative
of EJ concerns. As shown in the table, of the approximately 36.0 million people exposed to steam electric
pollutants, 13.9 percent are poor, 34.4 percent are minority, and 6.5 percent are both poor and minority.
Overall, 41.9 percent of potentially exposed individuals are categorized in at least one or more EJ subgroup
based on their poverty level or race/ethnicity, while 58.1 percent are neither minority nor poor.
Table 14-3: Characteristics of Population Potentially Exposed to Steam Electric Pollutants via
Consumption of Self-caught Fish
Subgroup
Poor
Non-Poor
Total
Minority
2,345,972
10,042,390
12,388,362
6.5%
27.9%
34.4%
Non-Minority
2,668,083
20,915,560
23,583,643
7.4%
58.1%
65.6%
Total
5,014,055
30,957,950
35,972,005
13.9%
86.1%
100%
Source: U.S. EPA Analysis, 2015
The distribution of adverse health effects is a function of the characteristics of the affected population (Table
14-3), including age and sex,84 ethnicity-specific exposure factors,85 and reach water quality. Table 14-4
84
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 socio-
economic 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.
Ethnicity-specific factors that determine exposure to pollutants in fish tissue include the assumed fish
consumption rates and average fertility rate. For example, as described in Chapter 2 of the BCA, Asian/Pacific
September 29, 2015
14-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
shows the distribution of selected adverse health effects in the baseline. Table 14-5 shows the distribution of
adverse health effects avoided under each of the five options EPA analyzed in developing the final ELGs.
Note that benefits follow the same distribution as avoided adverse health effects since each case is valued
equally, irrespective of the socio-economic subgroup.
The two tables show results for three selected subgroups:
> Poor and minority (6.5 percent of the exposed population),
> Poor or minority (/'. e., but not both; 35.3 percent of the exposed population), and
> All others (i.e., non-poor white; 58.1 percent of the exposed population).
The first two subgroups are the primary interest of this analysis as potentially indicative of EJ concerns.
Table 14-4: Distribution of Baseline IQ Point Decrements by Pollutant (2021 to 2042)
Pollutant
Lead
Mercury
Poor & Minority
(6.5% of
Population)
6,266,344 8.0%
91,838 9.3%
Poor or Minority
(35.3% of Population)
29,663,242 38.0%
439,403 44.6%
All Others
(58.1% of Population)
42,214,217 54.0%
454,081 46.1%
Total
78,143,802 100%
985,322 100%
Source: U.S. EPA Analysis, 2015
E fable 14-5: Distribution of Avoided IQ Point Decrements Relative to the Baseline, by Pollutant
2021 to 2042)
Pollutant
and
Exposed
Population
Children
Exposed to
Lead
Infants
Exposed to
Mercury
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Option A
Option B
Option C
Option D
Option E
Poor & Minority
(6.5% of
Population)
78 9.2%
78 9.2%
128 9.9%
155 7.8%
155 7.8%
287 8.9%
293 8.9%
471 7.9%
540 7.5%
591 7.5%
Poor or Minority
(35.3% of
Population)
313 36.6%
313 36.6%
518 40.3%
693 34.9%
693 34.9%
1,300 40.1%
1,331 40.2%
2,216 36.9%
2,603 36.1%
2,839 35.9%
All Others
(58.1% of
Population)
462 54.2%
462 54.2%
640 49.8%
1,137 57.3%
1,137 57.3%
1,652 51.0%
1,687 50.9%
3,314 55.2%
4,076 56.5%
4,469 56.6%
Total
853 100%
853 100%
1,285 100%
1,985 100%
1,985 100%
3,239 100%
3,311 100%
6,001 100%
7,219 100%
7,898 100%
Source: U.S. EPA Analysis, 2015
The distribution of baseline health effects and health improvements resulting from the ELGs can be compared
to the relative share of the population exposed to steam electric pollutants (from Table 14-3) to assess the
Islander anglers have daily consumption rates that are 1.4 times and 1.9 times those of While (non-Hispanic)
anglers for recreational and subsistence fishing modes, respectively.
September 29, 2015
14-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
degree to which the regulatory options contribute to mitigating any EJ concerns that may be present in the
baseline.
The poor and minority subgroup represents 6.5 percent of the potentially affected population, but accounts for
8.0 percent and 9.3 percent of the estimated IQ point decrements in the exposed population in the baseline.
Under Option D, the poor and minority subgroup also sees a disproportional share of the improvements, with
7.8 percent and 7.5 percent of the avoided IQ point decrements from exposure to lead and mercury,
respectively. Findings are similar for the poor or minority subgroup, i.e., this subgroup also incurs a
disproportionate share of baseline adverse health effects, and of the improvements arising from the final
ELGs.
1
4.4 Subsistence Fishers
In the analysis of health benefits (see Chapter 3), EPA assumed that 5 percent of the exposed population are
subsistence fishers, and that the remaining 95 percent are recreational anglers. This is based on the assumed
95th percentile fish consumption rate for subsistence fishers. Subsistence fishers consume more self-caught
fish than recreational anglers and can therefore be expected to experience higher health risks associated with
steam electric pollutants in fish tissue.
The results of the human health analysis suggest that subsistence fishers are disproportionally exposed to
pollutants in steam electric power plant discharges via fish consumption and disproportionally incur adverse
health effects from this exposure. As shows in Table 14-6 and Table 14-7, subsistence fishers incur 7 percent
to 17 percent of the baseline IQ decrements, even though they represent only 5 percent of the overall
population, and account for 17 percent to 30 percent of the avoided health effects and benefits of the final
rule.
Table 14-6: Distribution of Baseline IQ Point Decrements by Pollutant and Fishing Mode (2021 to
2042)
Pollutant and Exposed
Population
Children Exposed to Lead
Infants Exposed to Mercury
Subsistence Fishers
(5 percent of population)
5,302,873 6.8%
166,415 16.9%
Recreational Fishers
(95 percent of
population)
72,840,929 93.2%
818,907 83.1%
Total
78,143,802 100%
985,322 100%
Source: U.S. EPA Analysis, 2015
Table 14-7: Distribution of Avoided IQ Point Decrements Relative to the Baseline by Fishing Mode,
and Pollutant (2021 to 2042)
Pollutant
and
Exposed
Population
Children
Exposed to
Lead
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Subsistence Fishers
(5 percent of population)
186
186
390
605
605
21.8%
21.8%
30.4%
30.5%
30.5%
Recreational Fishers
(95 percent of
population)
667 78.2%
667 78.2%
895 69.6%
1,381 69.5%
1,381 69.5%
Total
853
853
1,285
1,985
1,985
100%
100%
100%
100%
100%
September 29, 2015
14-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
14: EnvironmentalJustice
Table 14-7: Distribution of Avoided IQ Point Decrements Relative to the Baseline by Fishing Mode,
and Pollutant (2021 to 2042)
Pollutant
and
Exposed
Population
Infants
Exposed to
Mercury
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Subsistence Fishers
(5 percent of population)
536 16.5%
548 16.5%
992 16.5%
1,194 16.5%
1,306 16.5%
Recreational Fishers
(95 percent of
population)
2,703 83.5%
2,763 83.5%
5,009 83.5%
6,025 83.5%
6,592 83.5%
Total
3,239
3,311
6,001
7,219
7,898
100%
100%
100%
100%
100%
Source: U.S. EPA Analysis, 2015
'
4.5 EJ Analysis Findings
Based on the EJ analyses discussed above, EPA determined that the final ELGs will not deny communities
from the benefits of environmental improvements expected to result from compliance with the more stringent
effluent limits. In fact, the distribution of avoided adverse health outcomes and benefits suggests that poor and
minority communities may receive a greater share of the benefits from the final ELGs than their
representation in the affected populations. The final ELGs may thus help redress environmental inequities that
may be present in the baseline.
By reducing exposures to pollutants discharged by steam electric power plants, all of the evaluated options
would benefit communities with potential EJ concerns. Of the five options, Options C, D, and E are expected
to provide significantly greater human health benefits than Options A and B, as indicated by higher reductions
in adverse health effects from lead and mercury exposure. Improvements under Option C accrue to poor and
minority population to a larger relative extent than do improvements under Options D and E, but Option D
(on which the final limitations and standards are based) provides greater benefits overall than Option C,
including to poor and minority populations exposed to steam electric pollutants through consumption of self-
caught fish. Thus, the final rule will further environmental justice objectives.
14.6 Limitations and Uncertainties
This EJ analysis inherits the limitations and uncertainties of the human health benefit analysis (see Chapter 3)
regarding pollutant exposure, health effects, and valuation. In addition, however, the analysis also embeds
uncertainty derived from the application of uniform assumptions across the population exposed to pollutant
discharges when factors may instead vary across socioeconomic characteristics, including:
> EPA assumed that all fishers travel up to 50 miles; in fact, some anglers stay closer to home and
certain EJ or sensitive subpopulations may tend to stay closer to home (e.g., poor people and
subsistence fishers). These people may be exposed to relatively higher concentrations of pollutants.
> EPA assumed that subsistence fishers are 5 percent of all anglers, with this assumption applied
uniformly across all socioeconomic groups. In fact, a relatively higher share of EJ groups may be
subsistence fishers. This would tend to increase the inequities already in the baseline and further
increase the mitigating effect of the ELGs in addressing these inequities.
September 29, 2015
14-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 14: Environmental Justice
> EPA applied uniform fishing participation rates, FCAs, and catch and release practices across the
entire population. However, differences in behavior across socioeconomic groups may result in
different distribution of baseline impacts and benefits.
In summary, use of average values across the entire US population (or within a state) instead of assumptions
that reflect specific socioeconomic conditions may understate inequities present in the baseline and benefits to
poor or minority populations, to the extent that different socioeconomic groups may be more likely to be
exposed to pollutants from steam electric power plant discharges.
September 29, 2015 14-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
15 References
Abt Associates Inc., 2012. BenMAP: Environmental Benefits Mapping and Analysis Program, User's
Manual, Prepared for U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards.
Abt Associates Inc., 2014. Disease and Economic Burden of Exposure to Arsenic in Drinking Water.
Prepared forthe National Center for Environmental Health, CDC. Under Contract No. 200-2013-
57290.
Abt Associates Inc. 2015. Developing a Concentration-Response Function for Pb Exposure and
Cardiovascular Disease-Related Mortality. Final Report. Prepared for the U.S. EPA National Center
for Environmental Economics. 27 June, 2015.
Acton, J.P. 1995. CBO Testimony on the Comprehensive Environmental Response, Compensation, and
Liability Act before the Subcommittee on Superfund, Waste Control, and Risk Assessment
Committee on Environment and Public Works, United States Senate. Congressional Budget Office.
Acton J.P., L.S. Dixon. 1992. Superfund and Transaction Costs: The Experiences of Insurers and Very Large
Industrial Firms. RAND Institute for Civil Justice.
Agency for Toxic Substances and Disease Registry (ATSDR), United States Department of Health and
Human Services. 2007. Toxicological Profile for Lead.
Aiken, R.A. 1985. Public Benefits of Environmental Protection in Colorado. Masters thesis, Colorado State
University.
Alaska Fisheries Science Center (AFSC) of NOAA National Marine Fisheries Service (2010). Alaska
Fisheries Science Center Publications Database.
Alan Plumer Associates. 2005. Dredging versus New Reservoirs, Report to the Texas Water Development
Board, TWDB Contract #2004-483-534, 91 p.
American Coal Ash Association. 2012. Beneficial Use of Coal Combustion Products: an American Recycling
Success Story. Retrieved from http://www.acaa-usa.Org/Portals/9/Files/PDFs/ACAA-Brochure-
Web.pdf
Anderson, G.D. and S.F. Edwards. 1986. "Protecting Rhode Island's Coastal Salt Ponds: An Economic
Assessment of Downzoning." Coastal Zone Management, 14( Yi): 67-91.
Arias, E. 2014. United States Life Tables, 2009. National Vital Statistics Reports, 62(7). Hyattsville, MD:
National Center for Health Statistics.
Associated Press. 2014. North Carolina: Coal Ash Cleanup Ends.
http://www.nytimes.com/2014/07/18/us/north-carolina-coal-ash-cleanup-
ends.html?mabReward=RI%3A5& r=0 (accessed 8-19-2014)
Atlantic States Marine Fisheries Commission (ASMFC). (2010). "Managed Species." from
http ://www.asmfc .org/managedspecies .htm.
Axelrad, D.A., D.C. Bellinger, L.M. Ryan, and T.J. Woodruff. 2007. Dose-Response Relationship of Prenatal
Mercury Exposure and IQ: An Integrative Analysis of Epidemiological Data. Environmental Health
Perspectives 115 (4): 609-615.
September 29, 2015 1ST
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Azevedo, C., J.A. Herriges, and C.L. Kling. 2001. "Valuing Preservation and Improvements of Water Quality
in Clear Lake." Staff Report 01-SR 94, Center for Agricultural and Rural Development (CARD),
Iowa State University.
Banzhaf, H.S., D. Burtraw, S. Chung, D.A. Evans, A. Krupnik, and J. Siikamaki. 2011. "Valuation of
Ecosystem Services in the Southern Appalachian Mountains." Paper Presented at the Annual Meeting
of the Association of Environmental and Resource Economists (AERE).
Banzhaf, Spencer H, Dallas Burtraw, David Evans, and Alan Krupnick. 2006. "Valuation of Natural Resource
Improvements in the Adirondacks." Land Economics, 82(3): 445-464.
Bateman, I.J., B.H. Day, S. Georgiou and I. Lake. 2006. "The Aggregation of Environmental Benefit Values:
Welfare Measures, Distance Decay and Total WTP." Ecological Economics 60(2): 450-460.
Bergstrom, J.C. and L.O. Taylor. 2006. "Using Meta-Analysis for Benefits Transfer: Theory and Practice."
Ecological Economics 60(2):351-360.
Bergstrom, J.C. and P. De Civita. 1999. "Status of Benefits Transfer in the United States and Canada: A
Review." Canadian Journal of Agricultural Economics 47(1): 79-87.
Beyersmann, J., Latouche, A., Buchholz, A., & Schumacher, M. 2009. Simulating competing risks data in
survival analysis. Statistics in Medicine, 28(6), 956-971.
Bin, O., and J. Czajkowski. 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.
Bockstael, N.E., McConnell, K.E., and Strand, I.E. 1988. "Benefits from improvements in Chesapeake Bay
Water Quality." Department of Agricultural & Resource Economics, University of Maryland, 1988.
Bockstael, N.E., K.E. McConnell and I.E. Strand. 1989. "Measuring the Benefits of Improvements in Water
Quality: The Chesapeake Bay." Marine Resource Economics 6: 1-18.
Bolen, Wallace P. 2011. USGS Minerals Yearbook 2009. v.I, Metals and Minerals: Sand and Gravel,
Construction. Retrieved from
http://minerals.usgs. gov/minerals/pubs/commodity/sand_&_gravel_construction/
Bolen, Wallace P. 2012. USGS Minerals Yearbook 2010. v.I, Metals and Minerals: Sand and Gravel,
Construction. Retrieved from
http://minerals.usgs. gov/minerals/pubs/commodity/sand_&_gravel_construction/
Bolen, Wallace P. 2013. USGS Minerals Yearbook 2011. v.I, Metals and Minerals: Sand and Gravel,
Construction. Retrieved from
http://minerals.usgs.gov/minerals/pubs/commodity/sand & gravel construction/
Bolen, Wallace P. 2014. USGS Minerals Yearbook 2012. v.I, Metals and Minerals: Sand and Gravel,
Construction. Retrieved from
http://minerals.usgs.gov/minerals/pubs/commodity/sand & gravel construction/
Borisova, Tatiana, Alan Collins, Gerard D'Souza, Matthew Benson, Mary Leigh Wolfe, and Brian Benham,
2008. "A Benefit-Cost Analysis of Total Maximum Daily Load Implementation." Journal of the
American Water Resources Association, 44(4): 1009-1023.
Bornschein, R.L., J. Grote, T. Mitchell, P.A. Succop, K.N. Dietrich, K.M Krafft, and P.B. Hammond. 1989.
"Effects of Prenatal Lead Exposure on Infant Size at Birth". In M. A. Smith, L. D. Grant, and A. I.
Sors (Eds.), Lead Exposure and Child Development (pp. 307-319): Springer Netherlands.
September 29, 2015 15-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Boyle, Kevin J., Christopher C. Parmeter, Bent Boehlert, and Robert Paterson. 2013. "Due Diligence in Meta-
Analyses to Support Benefit Transfers." Environmental and Resource Economics, 55(3): 357-386.
Boyle, K.J., P.J. Poor, and L.O. Taylor. 1999. Estimating the Demand for Protecting Freshwater Lakes from
Eutrophication. American Journal of Agricultural Economics 81: 1118-1122.
Brander, L., and R. Brouwer. 2011. "A multi-level meta-analysis regression model of contingent values for
water ecosystem services." Paper presented at the European Association of Environmental and
Resource Economists 18th Annual Conference, 29 June - 2 July 2011, Rome.
California Environmental Protection Agency, Office of Environmental Health Hazard Assessment (Cal
OEHHA). 2013a. Estimating Workplace Air and Worker Blood Lead Concentration using an
Updated Physiologically-based Pharmacokinetic (PBPK) Model. Available at:
http://oehha.ca.gov/air/risk assess/PBPK2013.pdf.
California Environmental Protection Agency, Office of Environmental Health Hazard Assessment (Cal
OEHHA). 2013b. Legget Model Code. Available at: http://oehha.ca.gov/air/legget.html.
Cameron, T.A. and D.D. Huppert. 1989. "OLS versus ML Estimation of Non-market Resource Values with
Payment Card Interval Data." Journal of Environmental Economics and Management 17, 230-246.
Canadian Council of Ministers of the Environment (CCME). 2001. Canadian water quality guidelines for the
protection of aquatic life: CCME Water Quality Index 1.0, Technical Report. In: Canadian
environmental quality guidelines, 1999, Canadian Council of Ministers of the Environment,
Winnipeg.
Carruthers, T. and Wazniak, C. 2003. Development of a Water Quality Index for the Maryland Coastal Bays.
Maryland's Coastal Bays: Ecosystem Health Assessment, Chapter 4.4. 2003.
Carson, R.T., and R.C. Michell. 1993. "The Value of Clean Water: The Public's Willingness to Pay for
Beatable, Fishable, and Swimmable Water Quality." Water Resources Research, 29(70), 2445-2454.
Carson, R.T., W.M. Hanemann, R.J. Kopp, J.A. Krosnick, RC. Mitchell, S. Presser, P.A. Ruud, and V.K.
Smith. 1994. Prospective Interim Lost Use Value due to DDT and PCB Contamination in the
Southern California Bight. Volume 2. Report to the National Oceanic and Atmospheric
Administration, Produced by Natural Resources Damage Assessment Inc., LA Jolla, CA.
Centers for Disease Control and Prevention (CDC), Department of Health and Human Services. 2005.
Preventing Lead Poisoning in Young Children. A Statement by the Centers for Disease Control and
Prevention. U.S. Department of Health and Human Services, Public Health Service.
Center for Disease Control and Prevention (CDC), Department of Health and Human Services. 2009. Fourth
National Report on Human Exposure to Environmental Chemicals.
Centers for Disease Control and Prevention (CDC), National Center for Health Statistics. 2012. Underlying
Cause of Death 1999-2010. Data are from the Multiple Cause of Death Files, 1999-2010, as compiled
from data provided by 57 Vital Statistics Jurisdictions through the Vital Statistics Cooperative
Program. CDC Wonder Online Database. Available at http://wonder.cdc.gov/ucd-icd 10.html.
Accessed March 21, 2014.
Chambers, J.F., J. Shkolnik, and M. Perez. 2003. Total Expenditures for Students with Disabilities, 1999-
2000: Spending Variation by Disability. Special Education Expenditure Project, United States
Department of Education.
September 29, 2015 15-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Clonts, H.A., and J.W. Malone. 1990. "Preservation Attitudes and Consumer Surplus in Free Flowing
Rivers." In J. Vining (ed.), Social Science and Natural Resource Recreation Management, Westview
Press: Boulder, CO, pp. 301-317.
Collett, D. (2003). Modelling Survival Data in Medical Research (2nd ed.). Boca Raton: Chapman &
Hall/CRC.
Clark, E., J.A. Haverkamp, and W.Chapman. 1985. "Eroding Soils: The Off-Farm Impacts." Washington,
DC: The Conservation Foundation.
Collins, Alan R. and Rosenberger, R.S. 2007. "Protest Adjustments in the Valuation of Watershed Restoration
Using Payment Card Data." Agricultural and Resource Economics Review, 36(2): 321-335.
Collins, A.R., R.S. Rosenberger and J.J. Fletcher. 2009. "Valuing the restoration of acidic streams in the
Appalachian Region: A stated choice method." In H.W. Thurstone, M.T. Heberling and A.
Schrecongost (eds.), Environmental Economics for Watershed Restoration. Boca Raton, FL:
CRC/Taylor Francis, pp.29-52.
Corrigan, J.R., C.L. Kling, and J. Zhao. 2008. "Willingness to Pay and the Cost of Commitment: An
Empirical Specification and Test." Environmental and Resource Economics, 40: 285-298.
Croke, Kevin, R.G. Fabian, and G. Brenniman. 1986-87. "Estimating the Value of Improved Water Quality in
an Urban River System." Journal of Environmental Systems 16(1): 13-24.
Crowder, B. M. (1987). Economic costs of reservoir sedimentation: A regional approach to estimating
cropland erosion damage. Journal of soil and water conservation, 42(3), 194-197.
Cude, C. G. 2001. Oregon Water Quality Index: "A Tool for Evaluating Water Quality Management
Effectiveness." Journal of the American Water Resources Association 37(1): 125-137.
Davic RD, Welsh HH Jr. 2004. On the ecological roles of salamanders. Annual Review of Ecology,
Evolution, and Systematics 35: 405-434.
Deacon, J. E., J. D. Kobetich, et al. 1979. "Fishes of North America endangered, threatened, or of special
concern: 1979." Fisheries 4(2): 1-69.
Desvousges, W.H., F.R. Johnson, and H.S. Banzhaf. 1998. Environmental Policy Analysis with Limited
Information: Principles and Applications of the Transfer Method. Northampton, MA: Edward Elgar
Publishers.
Desvousges, W.H., V.K. Smith and A. Fisher. 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.
De Zoysa, A.D.N. 1995. A Benefit Evaluation of Programs to Enhance Groundwater Quality, Surface Water
Quality and Wetland Habitat in Northwest Ohio. Dissertation, Ohio State University.
Diamond, P.A. 1996. "Testing the Internal Consistency of Contingent Valuation Surveys." Journal of
Environmental Economics and Management, 30: 337-347.
Dietrich, K.N., K.M. Krafft, R. Shukla, R.L. Bornschein, and P.A. Succop. 1987. "The Neurobehavioral
Effects of Early Lead Exposure." In S. R. Schroeder (Ed.), Toxic Substances and Mental Retardation:
Monographs of the American Association on Mental Deficiency, 8: 71-95.
Dixon L.S., D.S. Drezner, J.K. Hammitt. 1993. Private-Sector Cleanup Expenditures and Transaction Costs at
18 Superfund Sites. RAND Corporation.
September 29, 2015
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Downstream Strategies LLC. 2008. An Economic Benefit Analysis for Abandoned Mine Drainage
Remediation in the West Branch Susquehanna River Watershed, Pennsylvania. Prepared for Trout
Unlimited.
Duke Energy. 2014. Form 10-Q for the Quarterly Period Ended June 20, 2014.
Dunnette, D. A. 1979. A Geographically Variable Water Quality Index Used in Oregon. As cited in Cude
2001.
Electric Power Research Institute. 2005. Beneficiation of Fly Ash Containing Mercury and Carbon. Retrieved
fromhttp://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000000001004267
ERMCO. 2014. Ready-mixed concrete industry statistics. Year 2013. [Table 2A. Materials-Cement.] 20pp.
Retrieved from http://www.ermco.eu/documents/statistics/ermco-statistics-v-2013-final-version.pdf
Fann N, Baker K.R. and Fulcher C. 2012. Characterizing the PM2.5-related health benefits of emission
reductions for 17 industrial, area and mobile emission sectors across the U.S. Environment
International 49: 141-151.
Fann N, Fulcher C, and Hubbell B. 2009. The influence of location, source, and emission type in estimates of
the human health benefits of reducing a ton of air pollution. Air Qual Atmos Health 2:169-76.
Farber, S., and Griner, B. 2000. "Using Conjoint Analysis to Value Ecosystem Change," Environmental
Science and Technology, 34(8): 1407-1412.
Federal Highway Administration Materials Group. 2015. Fly Ash. [Webpage] Retrieved from
http://www.fhwa.dot.gov/infrastructure/materialsgrp/flyash.htm
Freeman III, A.M. 2003. The Measurement of Environmental and Resource Values: Theory and Methods.
Washington, DC: Resources for the Future.
Friedmann, A.S., M.C. Watzin, T. Brinck-Johnsen, J.C. Leiter. 1996. Low levels of dietary methylmercury
inhibit growth and gonadal development in juvenile walleye (Stizostedion vitreum). Aquatic
Toxicology 35: 265-278.
Froese, R. and D. Pauly. 2009. "Fishbase (version 07/2009)." http://www.fishbase.org.
Ge, J., C. Kling, and J. Herriges. 2013. "How much is Clean Water Worth? Valuing Water Quality
Improvement Using a Meta Analysis." Iowa State University, WP #13016, March 2013.
Ghermandi, A. and P.A.L.D. Nunes. 2013. "A Global Map of Coastal Recreation Values: Results from a
Spatially Explicit Meta-analysis." Ecological Economics, 86(1): 1-15.
Ghermandi, A., J.C.J.M. van den Bergh, L.M. Brander, H.L.F. de Groot, and P.A.L.D. Nunes. 2010. "Values
of Natural and Human-made Wetlands: A Meta-analysis." Water Resources Research, 46: W12516.
Graf, W. L., E. Wohl, T. Sinha, and J.L. Sabo. 2010. "Sediment and sustainability of western American
reservoirs." Water Resources Research 46(W12535).
Hammitt, J. K. 2007. Valuing changes in mortality risk: Lives saved versus life years saved. Review of
Environmental Economics and Policy, 1(2), 228-240.
Hanley, N., F. Schlapfer, and J. Spurgeon. 2003. Aggregating the benefits of environmental improvements:
distance-decay function for use and non-use values. Journal of Environmental Management, 68: 297-
304.
September 29, 2015 15-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Hargrove, W. W. (2007). Brashear Island Volumentric Interpolation Retrieved April 4, 2008, from
http://research.esd.ornl.gov/CRERP/SEDIMENT/BRASHEAR.HTM
Hayes, K.M., T.J. Tyrell and G. Anderson. 1992. "Estimating the Benefits of Water Quality Improvements in
the Upper Narragansett Bay." Marine Resource Economics 7: 75-85.
Herriges, J.A. and J.F. Shogren. 1996. "Starting point bias in dichotomous choice valuation with followup
questioning." Journal of Environmental Economics and Management, 30(1):112-131.
Kite, D. 2002. Willingness to Pay for Water Quality Improvements: The Case of Precision Application
Technology. Department of Agricultural Economics and Rural Sociology, Auburn University,
Auburn, AL; 8 August 2002.
Hoehn, J.P., Loomis, J.B., 1993. Substitution effects in the valuation of multiple environmental programs.
Journal of Environmental Economics and Management 25 (1), 56-75.
Huang, J.C., T.C. Haab and J.C. Whitehead. 1997. "Willingness to Pay for Quality mprovements: Should
Revealed and Stated Preference Data Be Combined?" Journal of Environmental Economics and
Management, 34(3): 240-255.
Interagency Working Group on Social Cost of Carbon (IWGSCC), United States Government. 2010.
Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis under Executive
Order 12866. February 2010.
Interagency Working Group on Social Cost of Carbon (IWGSCC), United States Government. 2013a.
Technical Update of the Social Cost for Regulatory Impact Analysis under Executive Order 12866.
Revised November 2013.
Interagency Working Group on Social Cost of Carbon (IWGSCC), United States Government. 2013b.
Technical Update of the Social Cost for Regulatory Impact Analysis under Executive Order 12866.
Revised June 2015.
Interagency Working Group on Social Cost of Carbon (IWGSCC), United States Government. 2015a.
Technical Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact
Analysis Under Executive Order 12866. Revised July 2015.
Interagency Working Group on Social Cost of Carbon (IWGSCC), United States Government. 2015b.
Response to Comments: Social Cost of Carbon for Regulatory Impact Analysis Under Executive
Order 12866.July 2015.
Intergovernmental Panel on Climate Change (IPCC). 2014. Climate Change 2014: Mitigation of Climate
Change. Working Group III contribution of the Fifth Assessment report of the Intergovernmental
Panel on Climate Change.
Irvin, S., T. Haab, and F.J. Hitzhusen. 2007. "Estimating willingness to pay for additional protection of Ohio
surface waters: contingent valuation of water quality". In F.J. Hitzhusen (ed.), Economic Valuation of
River Systems, Cheltenham: Edward Elgar, pp. 35-51.
Israel, B.D. 2006. State-by-State Guide to NRD Programs in All 50 States. Arnold & Porter LLP.
Israel, B.D. 2013. State-by-State Guide to NRD Programs in All 50 States. Arnold & Porter LLP.
Jakus, P., M. Downing, M.S. Bevelhimer, J.M. Fly. 1997. Do Sportfish Consumption Advisories Affect
Reservoir Anglers' Site Choice? Agricultural and Resource Economics Review. Vol. 26, 2, Oct.
1997.
September 29, 2015 15-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Jelks, H. L., S. J. Walsh, N.M. Burkhead, S. Contreras-Balderas, E. Diaz-Pardo, D.A. Hendrickson, J. Lyons,
N.E. Mandrak, F. McCormick, J.S. Nelson, S.P. Platania, B.A. Porter, C.B. Renaud, J.J. Schmitter-
Soto, E.B. Taylor, and M.L. Warren, Jr. 2008. "Conservation Status of Imperiled North American
Freshwater and Diadromous Fishes." Fisheries 33(8): 372-407.
Johnston, R.J., E.Y. Besedin, R. lovanna, C.J. Miller, R.F. Wardwell and M.H. Ranson. 2005. "Systematic
Variation in Willingness to Pay for Aquatic Resource Improvements and Implications for Benefit
Transfer: A Meta-Analysis." Canadian Journal of Agricultural Economics 53(2-3): 221-248.
Johnston, R.J. and J.M. Duke. 2009. "Willingness to Pay for Land Preservation Across States and
Jurisdictional Scale: Implications for Benefit Transfer." Land Economics, 85(2): 217-237.
Johnston, R.J. and M. Ramachandran. 2014. "Modeling Spatial Patchiness and Hot Spots in Stated Preference
Willingness to Pay." Environmental and Resource Economics, 59(3): 363-387.
Johnston, R.J., M.H. Ranson, E.Y. Besedin and E.G. Helm. 2006. "What Determines Willingness To Pay Per
Fish? A Meta-Analysis of Recreational Fishing Values." Marine Resource Economics 21: 1-32.
Johnston, S.K. Swallow, and T.F. Weaver. 1999. "Estimating Willingness to Pay and Resource Tradeoffs with
Different Payment Mechanisms: An Evaluation of a Funding Guarantee for Watershed Management."
Journal of Environmental Economics and Management, 38: 97-120.
Jones, M. and D. Behm. 2011. Bluff Collapse at Power Plant Sends Dirt, Coal Ash into Lake. Milwaukee
Journal Sentinel. October 31, 2011. http://www.jsonline.com/news/milwaukee/authorities-investigate-
bluff-collapse-at-we-energies-plant-13292953 8 .html
J0rgensen, S.L., S.B. Olsen, J. Ladenburg, L. Martinsen, S.R. Svenningsen and B. Hasler. 2013. "Spatially
Induced Disparities in Users' and Non-users' WTP for Water Quality Improvements - Testing the
Effect of Multiple Substitutes and Distance Decay." Ecological Economics, 92(1): 58-66.
Kahneman, D. and Knetsch, J. 1992. "Valuing Public Goods: The Purchase of Moral Satisfaction" Journal of
Environmental Economics and Management 22: 57-70.
Kaoru, Y. 1993. "Differentiating Use and Nonuse Values for Coastal Pond Water Quality Improvements."
Environmental and Resource Economics, 3: 487-494.
Karagas, M. R., Stukel, T. A., Morris, J. S., Tosteson, T. D., Weiss, J. E., Spencer, S. K., & Greenberg, E. R.
2001. Skin cancer risk in relation to toenail arsenic concentrations in a US population-based case-
control study. American journal of epidemiology, 153(6), 559-565.
Karagas, M. R, Tosteson, T. D., Blum, J., Morris, J. S., Baron, J. A., & Klaue, B. 1998. Design of an
epidemiologic study of drinking water arsenic exposure and skin and bladder cancer risk in a US
population. Environmental Health Perspectives, 106(Suppl 4), 1047.
Kim, R., H. Hu, A. Rotnitzky, D. Bellinger, and H. Needleman. 1995. "A Longitudinal Study of Chronic Lead
Exposure and Physical Growth in Boston Children." Environmental Health Perspectives 103(10):
952-56.
Krewski D, Jerrett M, Burnett RT, Ma R, Hughes E, Shi Y, et al. 2009. Extended follow-up and spatial
analysis of the American Cancer Society study linking particulate air pollution and mortality.
Research report. Health Effects Institute; 2009. p. 5-114. (discussion 115-36).
Lanphear, B.P., R. Hornung, J. Khoury, et al. 2005. Low-level Environmental Lead Exposure and children's
Intellectual Function: An International Pooled Analysis. Environmental Health Perspectives 113(7):
894-899. As cited in ATSDR (2007).
September 29, 2015 1ST
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Lant, C.L., and R. S. Roberts. 1990. "Greenbelts in the Cornbelt: Riparian Wetlands, Intrinsic Values, and
Market Failure." Environment and Planning 22: 1375-1388.
Lant, C.L., and G.A. Tobin. 1989. "The Economic Value of Riparian Corridors in Cornbelt Floodplains: A
Research Framework." Professional Geographer (41): 337-349.
Leggett R.W. 1993. An age-specific kinetic model of lead metabolism in humans. Environ Health Perspect
101(7): 598-616.
Leggett, C.G. and N.E. Bockstael. 2000. Evidence of the Effects of Water Quality on Residential Land Prices.
Journal of Environmental Economics and Management 39: 121-144.
Levy JI, Baxter LK, Schwartz J. 2009. Uncertainty and variability in health-related damages from coal-fired
power plants in the United States. Risk Anal 29(7): 1000-14.
Lichtkoppler, F.R. and T.W. Elaine. 1999. "Environmental Awareness and Attitudes of Ashtabula County
Voters Concerning the Ashtabula River Area of Concern: 1996-1997." Journal of Great
LakesResources 25: 500-514.
Lindsey, G. 1994. "Market Models, Protest Bids, and Outliers in Contingent Valuation." Journal of
WaterResources Planning and Management 12: 121-129.
Lipton, D. 2004. "The Value of Improved Water Quality to Chesapeake Bay Boaters." Marine Resource
Economics 19: 265-270
Londono Cadavid, C. and A. W. Ando .2013. "Valuing preferences over stormwater management outcomes
including improved hydrologic function." Water Resources Research 49: 4114-4125.
Loomis, J.B. 1996. "How Large is the Extent of the Market for Public Goods: Evidence from a Nation Wide
Contingent Valuation Survey." Applied Economics, 28(7): 779-782.
Loomis, J.B. 2000. "Vertically Summing Public Good Demand Curves: An Empirical Comparison of
Economic versus Political Jurisdictions." Land Economics, 76: 312-321.
Loomis, J.B. and R.S. Rosenberger. 2006. "Reducing barriers in future benefit transfers: needed
improvements in primary study design and reporting." Ecological Economics, 60: 343-350.
Lyke, A.J. 1993. Discrete Choice Models to Value Changes in Environmental Quality: A Great Lakes Case
Study. Dissertation submitted to the Graduate School of the University of Wisconsin, Madison.
Martin, J.A., B.E. Hamilton, M.J.K. Osterman, S.C. Curtin, and T.J. Matthews. 2013. National Vital Statistics
Reports: Volume 62, Number 9. Births: Final Data for 2012.
Massey Energy Company. 2002. Form 10-K: Annual Report.
Matthews, L.G., F.R. Homans, and K.W. Easter. 1999. "Reducing Phosphorous Pollution in the Minnesota
River: How Much is it Worth?" Staff Paper. Department of Applied Economics, University of
Minnesota.
McClelland, N. I. 1974. Water Quality Index Application in the Kansas River Basin. U.S. Environmental
Protection Agency Region 7, Kansas City, Missouri.
McTigue, N., Graf, K., & Brown, R. (2014). Occurrence and consequences of increased bromide in drinking
water sources. Journal of the American Waterworks Association, 106(11), E492-E508.
September 29, 2015 15-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Menke, A., Muntner, P., Batuman, V., Silbergeld, E. K., & Guallar, E. 2006. Blood lead below 0.48
micromol/L (10 microg/dL) and mortality among US adults. Circulation, 114(13), 1388-1394. doi:
10.1161/circulationaha.l06.628321
Mergler, D., H.A. Anderson, L.H.M. Chan, K.R. Mahaffey, M. Murray, M. Sakamoto, and A.H. Stern. 2007.
Methylmercury Exposure and Health Effects in Humans: A Worldwide Concern. AMBIO: A Journal
of the Human Environment 36: 3-11.
Mitchell, R.C. and R.T. Carson. 1989. Using Surveys to Value Public Goods: The Contingent Valuation
Method. Resources for the Future, Washington, DC.
Minnesota Department of Natural Resources (MDNR). 2010. Managing Minnesota's Shallow Lakes for
Waterfowl and Wildlife. Shallow Lakes Program, Minnesota Division of Fish and Wildlife.
Moeltner, K., K.J. Boyle, and R.W. Paterson. 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.
Morris, G., & Fan, J. (1997). Reservoir Sedimentation Handbook. U.S.A.: McGraw-Hill Professional.
Muller N.Z. and R. Mendelsohn. 2009. Efficient pollution regulation: getting the prices right. American
Economic Review 99: 1714-1739.
National Academy of Sciences. 1993. Measuring lead exposure in infants, children, and other sensitive
populations. National Academy Press: Washington, D.C.
National Center for Education Statistics. 2014. Fast Facts: Expenditures. Available at
http://nces.ed.gov/fastfacts/display.asp?id=66
National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration
(NOAA). 2014a. Fish Stock Sustainability Index: 2013 Quarter 4 Update Through December 31,
2013. Silver Spring, Maryland, U.S. Department of Commerce, NOAA, National Marine Fisheries
Service. Available
at: http://www.nmfs.noaa.gov/sfa/fisheries_eco/status_of_fisheries/status_updates.html
National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration
(NOAA). 2014b. NOAA Fisheries Geographical Information Systems, Fisheries Data: Critical
Habitat. Available at: http://www.nmfs.noaa.gov/gis/data/listed.htm
National Marine Fisheries Service (NMFS) of the National Oceanic and Atmospheric Administration
(NOAA). 2014c. "NOAA National Marine Fisheries Service Southwest Regional Office, GIS Data."
from http://swr.nmfs.noaa.gov/salmon/layers/finalgis.htm.
National Oceanic and Atmospheric Administration (NOAA). 2010. Environmental Sensitivity Index (ESI)
Maps, Office of Response and Restoration.
National Research Council (NRC). 2009. Informing Decisions in a Changing Climate: Panel on Strategies and
Methods for Climate-Related Decision Support. Committee on the Human Dimensions of Global
Change Division of Behavioral and Social Sciences and Education, 198 pages.
National Toxicology Program. 2012. NTP monograph on health effects of low-level lead. US Department of
Health and Human Services, National institutes of Health, National Institute of Environmental Health
Sciences, Office of Health Assessment and Translation. Research Triangle Park, NC. NIH Publication
No. 12-5996, 148 pp.
September 29, 2015 15-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
NatureServe. 2014. "NatureServe Explorer: An Online Encyclopedia of Life." from
http://www.natureserve.org/explorer/.
Navrud, S. and R. Ready. 2007. "Review of Methods for Value Transfer." In: Navrud, S. and R. Ready (eds.),
Environmental Value Transfer: Issues and Methods. Dordrecht, The Netherlands: Springer, pp.1-10.
Nelson, J.P. and P.E. Kennedy. 2009. "The Use (and Abuse) of Meta-Analysis in Environmental and
Resource Economics: An Assessment." Environmental and Resource Economics, 42(3): 345-377.
Niimi, A. J. and G. P. Kissoon. 1994. Evaluation of the critical body burden concept based on inorganic and
organic mercury toxicity to Rainbow Trout (Oncorhynchus mykiss). Archives of Environmental
Contamination and Toxicology 26: 169 - 178.
Northeast Fisheries Science Center (NEFSC) of NOAA National Marine Fisheries Service. 2010. "Northeast
Fisheries Science Center Publications." from http://www.nefsc.noaa.gov/publications/.
Olsen, D., Richards, Jack, and Scott, R. D. 1991. "Existence and Sport Values for Doubling the Size of
Columbia River Basin Salmon and Steelhead Runs." Rivers. 2(1): 44-56.
Opaluch, J.J., T. Origanums, M.J. Mazzotta, J. Diamantides, and R. Johnston. 1998. Resource and
Recreational Economic Values for the Peconic Estuary. Report prepared for Peconic Estuary
Program, Suffolk County Department of Health Services, Riverhead, NY, by Economic Analysis,
Inc., Peace Dale, Rhode Island.
Pacific Islands Fisheries Science Center (PIFSC) of the NOAA National Marine Fisheries Service. 2010a.
"Fishery Biology and Stock Assessment Division." from http://www.pifsc.noaa.gov/fbsad/index.php.
Pacific Islands Fisheries Science Center (PIFSC) of the NOAA National Marine Fisheries Service. 2010b.
Pacific Islands Fisheries Science Center Staff Publications Database.
Phaneuf, D. J., C. L. Kling, and J. A. Herriges. 1998. Valuing Water quality Improvements Using Revealed
Preference Methods When Corner Solutions are Present. American Journal of Agricultural
Economics 80: 1025-1031.
Piomelli S., J.F. Rosen, J.J. Chisolm Jr, and J.W. Graef. 1984. "Management of childhood lead poisoning."
Journal of Pediatrics 105: 523-32.
Poor PJ, Boyle KJ, Taylor LO, Bouchard R. 2001. Water clarity in hedonic property value models. Land
Economics 77:482A93.
PPL Corporation. 2006. Annual Report.
Richardson, S. D., Plewa, M. J., Wagner, E. D., Schoeny, R., & Demarini, D. M. 2007. Occurrence,
genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking
water: a review and roadmap for research. Mutation Research, 636(1-3), 178-242.
Richardson, L. and J. Loomis. 2009. "The total economic value of threatened, endangered and rare species:
An updated meta-analysis." Ecological Economics 68(5): 1535-1548.
Roberts, L.A., and J.A. Leitch. 1997. "Economic Valuation of Some Wetland Outputs of Mud Lake."
Agricultural Economics Report No. 381, Department of Agricultural Economics, North Dakota
Agricultural Experiment Station, North Dakota State University.
Rosenberger, R.S. and R.J. Johnston. 2007. Selection Effects in Meta-Valuation Function Transfers. Meta-
Analysis in Economics Conference, September 27-30, 2007, Sandbjerg Manor, S0nderborg,
Denmark.
September 29, 2015 15-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Rosenberger, R.S. and T.D. Stanley. 2006. "Measurement, Generalization, and Publication: Sources of Error
in Benefit Transfers and Their Management." Ecological Economics 60(2): 372-378.
Rosenberger, R.S. and T.T. Phipps. 2007. "Correspondence and Convergence in Benefit Transfer Accuracy:
Meta-Analytic Review of the Literature." In: Navrud, S. and R. Ready (eds.), Environmental Value
Transfer: Issues and Methods. Dordrecht, The Netherlands: Springer, pp.23-43.
Rowe, R.D., W.D. Schulze, B. Hurd and D. Orr. 1985. Economic Assessment of Damage Related to the Eagle
Mine Facility. Energy and Resource Consultants, Inc., Boulder, CO.
Ruhl, L., A. Vengosh, G.S. Dwyer, H. Hsu-Kim, G. Schwartz, A. Romanski, S.D. Smith. 2012. The Impact of
Coal Combustion Residue Effluent on Water Resources: A North Carolina Example. Environmental
Science and Technology 2012 Oct 15.
SAIC. 2006. Life Cycle Assessment: Principles and Practice. Report submitted to U.S. Environmental
Protection Agency (Contract No. 68-C02-067, WA 3-15). Retrieved from
http://www.epa.gov/nrmrl/std/lca/lca.html
Salkever, D.S. 1995. Updated Estimates of Earnings Benefits from Reduced Exposure of Children to
Environmental Lead. Environmental Research 70: 1-6.
Sanders, L. B., R.G. Walsh, and J.B. Loomis. 1990. "Toward Empirical Estimation of the Total Value of
Protecting Rivers." Water Resources Research. 26(7): 1345-1357.
Schaafsma, M., R. Brouwer and J. Rose. 2012. "Directional Heterogeneity in WTP models for Environmental
Valuation." Ecological Economics, 79(1): 21-31.
Schober, S. E., et al. 2006. 'Blood lead levels and death from all causes, cardiovascular disease, and cancer:
Results from the NHANES III Mortality Study', Environmental Health Perspectives, 114 (10), 1538-
41.
Schoch, N., M. Glennon, D. Evers, M. Duron, A. Jackson, C. Driscoll, X. Yu, H. Simonin. 2011. Long-term
monitoring and assessment of mercury based on integrated sampling efforts using the common loon,
prey fish, water, and sediment. New York State Energy Research and Development Authority Report
No. 12-06, 116pp.
Schulze, W.D., RD. Rowe, W.S. Breffle, RR Boyce, and G.H. McClelland. 1995. Contingent Valuation of
Natural Resource Damages Due to Injuries to the Upper Clark Fork River Basin. State of Montana,
Natural Resource Damage Litigation Program. Prepared by: RCG/Hagler Bailly, Boulder, CO.
Schwartz, J. 1994. Low-Level Lead Exposure and Children's IQ: A Meta-analysis and Search for a
Threshold. Environmental Research 65: 42-55.
Shannon, R. L., & Strayer, D. S. 1989. Arsenic-induced skin toxicity. Human & Experimental
Toxicology, 8(2), 99-104.
Shrestha, R. K., J.R.R. Alavalapati. 2004. "Valuing Environmental Benefits of Silvopasture Practice: A Case
Study of the Lake Okeechobee Watershed in Florida." Ecological Economics 49: 349-359.
Shrestha, R., R. Rosenberger and J. Loomis. 2007. "Benefit Transfer Using Meta-Analysis in Recreation
Economic Valuation, in Benefit Transfer Accuracy: Meta-Analytic Review of the Literature." In:
Navrud, S. and R. Ready (eds.). Environmental Value Transfer: Issues and Methods. Dordrecht, The
Netherlands: Springer, pp. 161-177.
September 29, 2015 15-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Shryock, Henry S., Jacob S. Siegel, and Elizabeth A. Larmon. 1980. The methods and materials of
demography. Vol. 2. US Dept. of Commerce, Bureau of the Census: for sale by the Supt. of Docs. US
Govt. Print. Off, 1980.
Smith, V.K. and S.K. Pattanayak. 2002. "Is Meta-Analysis a Noah's Ark for Non-Market Valuation?"
Environmental and Resource Economics, 22(1-2): 271-296.
Smith, V.K., G. Van Houtven, and S.K. Pattanayak. 2002. "Benefit Transfer via Preference Calibration:
'Prudential Algebra' for Policy." Land Economics 78(1): 132-152.
Sohngen, B., & Rausch, J. (1998). Soil Erosion in the Maumee River Basin: A Case Study Using Market
Methods to Value Environmental Externalities. Draft. Department of Agricultural Economics. Ohio
State University.
Southeast Fisheries Science Center (SEFSC) of NOAA National Marine Fisheries Service. 2010. SouthEast
Data, Assessment and Review (SEDAR).
Southwest Fisheries Science Center (SWFSC) of NOAA National Marine Fisheries Service. 2010.
"Southwest Fisheries Science Center Fisheries Biology Publications Database." from
http://swfsc.noaa.gov//publications/fedbin/qrypublications.asp?ParentMenuId=54.
Stanley, T.D., H. Doucouliagos, M. Giles, J. H. Heckemeyer, R. J. Johnston, P. Laroche, J. P. Nelson, M.
Paldam, J. Foot, and G. Pugh. 2013. "Meta-Analysis of Economic Research Reporting Guidelines."
Journal of Economic Surveys, 27(2): 390-394.
Steeves, P. and D. Nebert. 1994. l:250,000-scale Hydrologic Units of the United States. U.S. Geological
Survey, Reston, VA.
Steinhardt, B., J.F. Donaghy, F.A. Featherston. 1994. Report to Congressional Requesters: Superfund Legal
Expenses for Cleanup-Related Activities of Major U.S. Corporations. United States General
Accounted Office (U.S. GAO).
Stumborg, B.E., K.A. Baerenklau and R.C. Bishop. 2001. "Nonpoint Source Pollution and Present Values: A
Contingent Valuation of Lake Mendota." Review of Agricultural Economics. 23(1): 120-132.
Sutherland, R. J., and R.G. Walsh. 1985. "Effect of Distance on the Preservation Value of Water Quality."
Land Economics 61(3): 282-29.
Swartout, J. and G. Rice. 2000. Uncertainty Analysis of the Estimated Ingestion Rates Used to Derive the
Methylmercury Reference Dose. Drug and Chemical Toxicology 23 (1): 293-306.
Takatsuka, Yuki. 2004. Comparison of the Contingent Valuation Method and the Stated Choice Model for
Measuring Benefits of Ecosystem Management: A Case Study of the Clinch River Valley, Tennessee.
PhD dissertation, University of Tennessee.
Taylor, C. A., G. A. Schuster, et al. 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., M. L. Warren, et al. 1996. "Conservation Status of Crayfishes of the United States and
Canada." Fisheries 21(4): 25-38.Tennessee Valley Authority (TVA). 2009. Form 10-K (Annual
Report).
Tennessee Valley Authority (TVA). 2010. Kingston Ash Recovery Project Non-Time-Critical Removal
Action Embayment/Dredge Cell Action Memorandum. Document No. EPA-AO-024.
September 29, 2015 15-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Tennessee Valley Authority (TVA). 2011. Form 10-Q: Quarterly Report Pursuant to Sections 13 or 15(d).
Filing Date: 2011-08-11, Period of Report: 2011-06-30.
Tetra Tech. 2010. Evaluating Sustainability of Projected Water Demands on Future Climate Change
Scenarios. Prepared for Natural Resources Defense Council. Prepared by Tetra Tech.
Turtle, C.M., and M.D. Heintzelman. 2014. "A Loon on Every Lake: A Hedonic Analysis of Lake Water
Quality in the Adirondacks." Resource and Energy Economics 39: 1-15.
U.S. Army Corps of Engineers (USAGE). (2002). Final Dredged Material Management Plan and
Environmental Impact Statement.
U.S. Army Corps of Engineers (USAGE). (2013). Dredging Information System.
United States Census Bureau. 2010. QT-P2 Single Years of Age and Sex: 2010, Summary File.
http://factfinder2.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=DEC 10 SF1QT
P2&prodType=table
United States Census Bureau. 2012. American Community Survey Demographic and Housing Estimates.
DP05: 2008-2012 American Community Survey 5-Year Estimates. Retrieved from
http ://factfinder.census .gov/
United States Department of Health and Human Services (U.S. DHHS). 2008. Draft Toxicological Profile for
Cadmium.
United States Environmental Protection Agency (U.S. EPA). 1990. Review of the National Ambient Air
Quality Standards for Lead: Assessment of Scientific and Technical Information. OAQPS Staff
Paper. EPA-450/2-89-022.
United States Environmental Protection Agency (U.S. EPA). 2003. Economic and Environmental Benefits
Analysis Document for the Final Effluent Limitations Guidelines and Standards for the Metal
Products and Machinery Point Source Category. EPA-821-B-03-002.
United States. Environmental Protection Agency (U.S. EPA). 2004a. Final Regulatory Impact Analysis for
the Industrial Boilers and Process Heaters NESHAP: Final Report. EPA-452/R-04-002.
United States. Environmental Protection Agency (U.S. EPA). 2004b. Final Regulatory Analysis: Control of
Emissions fromNonroad Diesel Engines. EPA420-R-04-007.
United States Environmental Protection Agency (U.S. EPA). 2005. Regulatory Impact Analysis of the Clean
Air Mercury Rule. EPA-452/R-05-003.
United States Environmental Protection Agency (U.S. EPA). 2008a. Welcome to STORET, EPA's largest
computerized environmental data system. Available at http://www.epa.gov/storet/.
United States Environmental Protection Agency (U.S. EPA). 2008b. Technical Support Document:
Calculating Benefit Per-Ton Estimates. Final Ozone NAAQS regulatory impact analysis. Office of
Air Quality Planning and Standards, Office of Air and Radiation.
United States Environmental Protection Agency (U.S. EPA). 2009a. Environmental Impact and Benefits
Assessment for Final Effluent Guidelines and Standards for the Construction and Development
Category. November, 2009. http://www.epa.gov/guide/construction/files/environment.pdf.
United States Environmental Protection Agency (U.S. EPA). 2009b. Integrated Science Assessment for
Particulate Matter (Final Report). U.S. Environmental Protection Agency, Washington, DC,
EPA/600/R-08/139F, 2009.
September 29, 2015 15-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
United States Environmental Protection Agency (U.S. EPA). 2009c. Software and User's Manuals.
http://www.epa.gov/superfund/lead/products.htm
United States Environmental Protection Agency (U.S. EPA). 2009d. Steam Electric Power Generating Point
Source Category: Final Detailed Study. EPA 821-R-09-008.
United States Environmental Protection Agency (U.S. EPA). 2010a. Guidelines for Preparing Economic
Analyses. EPA 240-R-10-001.
United States Environmental Protection Agency (U.S. EPA). 2010b. IRIS Toxicological Review of Inorganic
Arsenic (Cancer) (External Draft
Review).http://cfpub.epa.gov/ncea/iris drafts/recordisplay.cfm?deid=219111
United States Environmental Protection Agency (U.S. EPA). 2010c. Questionnaire for the Steam Electric
Power Generating Effluent Guidelines, OMB Control 2040-0281, May 20, 2010.
United States Environmental Protection Agency (U.S. EPA). 2010d. Hazardous and Solid Waste Management
System; Identification and Listing of Special Wastes; Disposal of Coal Combustion Residuals From
Electric Utilities. [Proposed Rule]. Docket Number EPA-HQ-RCRA-2009-0640FRL-9149-4.
https://www.federalregister.gov/articles/2010/06/21/2010-12286/hazardous-and-solid-waste-
management-system-identification-and-listing-of-special-wastes-disposal-of#h-79
United States Environmental Protection Agency (U.S. EPA). 201 la. "Ecological Toxicity Information"
Available at: http://www.epa.gov/reg5sfun/ecology/toxprofiles.htm
United States Environmental Protection Agency (U.S. EPA). 201 Ib. Exposure Factors Handbook, 2011
Edition (Final). U.S. Environmental Protection Agency, Washington, DC. EPA-600-R-09-025F.
United States Environmental Protection Agency (U.S. EPA). 2012a. National Primary Drinking Water
Regulations, http://water.epa.gov/drink/contaminants/index.cfm
United States Environmental Protection Agency (U.S. EPA). 2012b. National Recommended Water Quality
Criteria, http://water.epa.gov/scitech/swguidance/standards/criteria/current/index.cfmtfhhtable
United States Environmental Protection Agency (U.S. EPA). 2012c. Risk-Screening Environmental Indicators
(RSEI). http://www.epa.gov/oppt/rsei/
United States Environmental Protection Agency (U.S. EPA). 2012d. Information Request Responses from
Electric Utilities. Database of Survey Response (April 12, 2012).
http://www.epa.gov/osw/nonhaz/industrial/special/fossil/surveys/index.htm
United States Environmental Protection Agency (U.S. EPA). 2013a. Benefit and Cost Analysis for the
Proposed Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating
Point Source Category. EPA-821-R-13-004.
United States Environmental Protection Agency (U.S. EPA). 2013b. Integrated Science Assessment for Lead.
(EPA/600/R-10/075F). Research Triangle Park, NC.
United States Environmental Protection Agency (U.S. EPA). 2013c. Integrated Science Assessment of Ozone
and Related Photochemical Oxidants (Final Report). U.S. Environmental Protection Agency,
Washington, DC, EPA/600/R-10/076F, 2013.
United States Environmental Protection Agency (U.S. EPA). 2013d. Fiscal Year 2011 Drinking Water and
Ground Water Statistics. Washington, DC, EPA 816-R-13-003, March 2013
September 29, 2015 15-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
United States Environmental Protection Agency (U.S. EPA). 2014. Regulatory Impact Analysis for EPA's
2015 RCRA Final Rule Regulating Coal Combustion Residue (CCR) Landfills and Surface
Impoundments at Coal-Fired Electric Utility Power Plants.
United States Environmental Protection Agency (U.S. EPA). 2015a. Environmental Assessment for the
Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category (40 CFR423). EPA 821-R-15-006
United States Environmental Protection Agency (U.S. EPA). 2015b. Technical Development Document for
the Effluent Limitations Guidelines and Standards for the Steam Electric Power Generating Point
Source Category. EPA 821-R-15-007
United States Environmental Protection Agency (U.S. EPA). 2015c. Regulatory Impact Analysis for Effluent
Limitations Guidelines and Standards for the Steam Electric Power Generating Point Source
Category. EPA 821-R-15-004
United States Environmental Protection Agency (U.S. EPA). 2015d. "Resources: Industrial Materials, Coal
Combustion Products." [Website]. Accessed on June 1, 2015 from
http://www.epa.gov/epawaste/conserve/imr/indust.htmtfcoal
United States Environmental Protection Agency (U.S. EPA). 2015e. Peer Review Package for Meta-analysis
of the Willingness-to-Pay for Water Quality Improvements.
United States Environmental Protection Agency (U.S. EPA). 2015f. External Peer Review of EPA's
Approach for Estimating Exposures and Incremental Health Effects from Lead due to Renovation,
Repair, and Painting Activities in Public and Commercial Buildings. Prepared by Versar, Inc.,
Springfield, VA.
United States Environmental Protection Agency (U.S. EPA). 2015g. Incremental Costs and Pollutant
Removals for Effluent Limitation Guidelines and Standards for the Steam Electric Generating Point
Source Category Report.
United States Environmental Protection Agency On-Scene Coordinator (U.S. EPA OSC) 2012. Oak Creek
Bluff Pollution Situation Report #4 (Final).
United States Fish and Wildlife Service (U.S. FWS). 2006. National Survey of Fishing, Hunting, and
Wildlife-Associated Recreation. FHW/06-NAT.
United States Fish and Wildlife Service (U.S. FWS). 2010. "North Florida Ecological Services Office:
Loggerhead Sea Turtle (Caretta caretta)." from
http://www.fws.gov/northflorida/seaturtles/turtle%20factsheets/loggerhead-sea-turtle.htm.
United States Fish and Wildlife Service (U.S. FWS). 2011. National Survey of Fishing, Hunting, and
Wildlife-Associated Recreation.
United States Fish and Wildlife Service (U.S. FWS). 2014a. "Environmental Conservation Online System."
Available at: http://ecos.fws.gov/
United States Fish and Wildlife Service (U.S. FWS). 2014b. "Critical Habitat Portal." Available
at: http://criticalhabitat.fws.gov/
United States Geological Survey (USGS). 2007. National Hydrography Dataset (NHD). Available at
http ://nhd.usgs .gov/data.html
United States Geological Survey (USGS). 2009. RESIS II: An Updated Version of the Original Reservoir
Sedimentation Survey Information System (RESIS).
September 29, 2015 15-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
United States Office of Management and Budget (U.S. OMB). 2003. Circular A-4: Regulatory Analysis.
Utah Department of Natural Resources (DNR), 2005. Utah Department of Natural Resources. "Utah
Waterfowl Advisories" http://waterfowladvisories.utah.gov/advisories.htm
VandenBerg, T. P., Poe, G. L. and Powell, J. R. 2001. "Assessing the accuracy of benefits transfers: evidence
from amulti-site contingent valuation study of groundwater quality." In J. C. Bergstrom, K. J.
Van Houtven, G., J. Powers and S.K. Pattanayak. 2007. "Valuing Water Quality Improvements in the United
States Using Meta-analysis: Is the Glass Half-full or Half-empty for National Policy Analysis?"
Resource and Energy Economics, 29(3): 206-228.
van Oss, Henrik G. 2009. USGS Minerals Yearbook 2009. v.I, Metals and Minerals: Cement. Retrieved from
http://minerals.usgs.gov/minerals/pubs/commodity/cement/
van Oss, Henrik G. 2010. USGS Minerals Yearbook 2010. v.I, Metals and Minerals: Cement. Retrieved from
http://minerals.usgs. gov/minerals/pubs/commodity/cement/
van Oss, Henrik G. 2013. USGS Minerals Yearbook 2011. v.I, Metals and Minerals: Cement. Retrieved from
http://minerals.usgs. gov/minerals/pubs/commodity/cement/
Vaughan, W.J. 1986. "The RFF Water Quality Ladder." In: Mitchell, R.C. and R.T. Carson. The Use of
Contingent Valuation Data for Benefit/Cost Analysis in Water Pollution Control, Final Report.
Washington: Resources for the Future. Appendix B.
Viscusi, K.W, J. Huber and J. Bell. 2008. "The Economic Value of Water Quality." Environmental and
Resource Economics 41: 169-187.
Walsh, P.J., J.W. Milon, and D.O. Scrogin. 2011. "The Spatial Extent of Water Quality Benefits in Urban
Housing Markets." Land Economics 87(4): 628-644.
Wathen, J.B., J. M. Lazorchak, A.R. Olsen, and A. Bart. 2015. "A national statistical survey assessment of
mercury concentrations in fillets offish collected in the U.S. EPA national rivers and streams
assessment of the continental USA." Chemosphere 122(March 2015): 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, 770(0), 276-298.
Wattage, P. M. 1993. Measuring the benefits of water resource protection from agricultural contamination:
Results from a contingent valuation study. PhD dissertation, Forestry, Iowa State University.
We Energies. 2013. Coal Combustion Products Utilization Handbook. 3rd Edition.
Welle, P.G. 1986. Potential Economic Impacts of Acid Deposition: A Contingent Valuation Study of
Minnesota. Dissertation, University of Wisconsin-Madison.
Welle, P.G., and J.B. Hodgson. 2011. Property Owner's Willingness to Pay for Water Quality Improvements:
Contingent Valuation Estimates in Two Central Minnesota Watersheds. Journal of Applied Business
and Economics 12(1): 81-94.
Wey, K.A. 1990. Social Welfare Analysis of Congestion and Water Quality of Great Salt Pond, Block Island,
Rhode Island. Dissertation, University of Rhode Island.
Whitehead, J.C. 2006. "Improving Willingness to Pay Estimates for Quality Improvements through Joint
Estimation with Quality Perceptions." Southern Economic Journal, 73(1): 100-111.
September 29, 2015 15-16
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs 15: References
Whitehead, J.C., G.C. Blomquist, T.J. Hoban and W.B. Clifford. 1995. "Assessing the Validity and
Reliability of Contingent Values: A Comparison of On-Site Users, Off-Site Users, and Nonusers."
Journal of Environmental Economics and Management, 29: 238-251.
Whitehead, J.C., and P.A. Groothuis. 1992. "Economic Benefits of Improved Water Quality: a case study of
North Carolina's Tar-Pamlico River." Rivers, 3: 170-178.
Whitehead, J.C., T. Hoban and W. Clifford. 2002. "Landowners' Willingness to Pay for Water Quality
Improvements: Jointly Estimating Contingent Valuation and Behavior with Limited Information."
White paper developed in part by U.S. EPA, NCDENR and the College of Agriculture and Life
Sciences at NSCU.
Whittington, D., G. Cassidy, D. Amaral, E. McClelland, H. Wang and C. Poulos. 1994. "The Economic Value
of Improving the Environmental Quality of Galveston Bay." Department of Environmental Sciences
and Engineering, University of North Carolina at Chapel Hill. GBNEP-38, 6/94.
Wilcove, D. S. and L. L. Master. 2005. "How many endangered species are there in the United States?"
Frontiers in Ecology and the Environment 3(8): 414-420.
Williams, C.D., Burns, J.W., Chapman, A.D., Flewelling, L., Pawlowicz, M. and W. Carmichael. 2001.
Assessment of cyanotoxins in Florida's lakes, reservoirs and rivers. White paper produced for Florida
Harmful Algal Bloom Task Force.
Williams, J. D., M. L. Warren, et al. 1993. "Conservation Status of Freshwater Mussels of the United States
and Canada." Fisheries 18(9): 6-22.
Williams, J. E., J. E. Johnson, et al. 1989. "Fishes of North America Endangered, Threatened, or of Special
Concern: 1989." Fisheries 14(6): 2-20.
Wisconsin Department of Justice. 2013. Memo to Roseann Lee from Steven E. Tinker, Assistant Attorney
General. Re: State of Wisconsin v. Wisconsin Electric Power Company.
Woods & Poole Economics Inc. 2012. Complete Demographic Database. Washington, DC.
Zhu, M., E.F. Fitzgerald, K.H. Gelberg, S. Lin, and C.M. Druschel. 2010. "Maternal low-level lead exposure
and fetal growth." Environmental Health Perspectives 118(10): 1471-1475.
September 29, 2015 15-17
-------
*** E.O. 12866 Review-Revised Version-Do Not Cite, Quote, or Release During Review ***
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix A: Changes to Benefits Analysis
Appendix A. Changes to Benefits Analysis since Proposal
The table below summarizes the principal changes EPA made to its benefits analysis for the final rule, as
compared to the analysis of the proposed rule (in addition to changes to inputs such as costs and pollutant
loads). EPA made changes to address comments it received on the proposed rule analysis and to incorporate
updated, more recent data.
Table A-1: Changes to Benefits Analysis Since Proposal
Report Section or
Benefit Category
General assumptions
General pollutant
loadings and
concentration
assumptions
Analysis Component
[Proposed rule analysis value]
Dollar year [all costs and benefits expressed
in 20 10 dollars]
Promulgation year [all costs and revenue
streams discounted back to 2014]
Period of social-costs and benefits analyses
[2017-2040]
Technology implementation years [2017-
2021]
EA modeling of metal concentrations in
immediately receiving reaches (national
model)
RSEI modeling of metal concentrations in
immediate and downstream reaches model
across nation
SPARROW modeling of nutrient
concentrations in receiving and downstream
reaches [estimate changes in ambient
pollutant (nutrient) concentrations in
receiving and downstream reaches based on
regression model]
Changes to Analysis for Final Rule
[Final rule analysis value ]
Updated dollar year [2013]
Updated promulgation year [2015]
Updated period of social-costs and benefits
analyses [2019-2042]
Updated technology implementation years
[2019-2023]
Updated immediately receiving reaches (see
EA for details)
Adjusted loadings to reflect partitioning of
pollutants in immediate reach, based on the
EA national model.
Used updated baseline that reflects TRI
releases for 2012.
No changes for nutrients.
Estimated changes in suspended sediments
and channel deposition using SPARROW
Human health benefits associated with reductions in fish tissue contamination
Assignment of
populations to affected
waterbodies to determine
potential exposure
Pollutant exposure via
consumption of self-
caught fish
Avoided IQ losses in
children from lead
exposure
Potentially exposed population [people
residing within 100 miles of an affected
waterbody]
Fishing practices [addressed by consumption
estimate]
Scope [immediately receiving reaches only]
Consumption rates [uniform consumption
rates for recreational/subsistence fishers]
Exposure estimates and monetization
Revised the analysis to focus on Census
Block Groups. Subdivided Census Block
Group population according to
socioeconomic indicators. Used travel
distance of 50 miles
For recreational anglers, adjusted exposed
population to account for catch and release
practices
Included downstream reaches
Consumption rates [ethnicity and fishing
mode-specific consumption rates]
Updated/verified the dose-response
relationship between PbB and IQ based on
most recent data for low level exposure (to
address comment)
September 29, 2015
A-1
-------
*** E.O. 12866 Review-Revised Version-Do Not Cite, Quote, or Release During Review ***
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix A: Changes to Benefits Analysis
Table A-1: Changes to Benefits Analysis Since Proposal
Report Section or
Benefit Category
Avoided IQ losses in
infants from mercury
exposure
Avoided cancer cases
from arsenic exposure
Avoided cardiovascular
disease in adults from
lead exposure
Avoided exposure in
drinking water
Analysis Component
[Proposed rule analysis value]
Monetization
Exposure estimates [exposed population is
infants born to women (15-50 years of age)
in households that fish recreationally]
Monetization
Monetization [valuation based on value of
statistical life (VSL)]
N/A
N/A
Changes to Analysis for Final Rule
[Final rule analysis value ]
Used updated estimates of the cost of
education
Modified definition of cohort of women of
child-bearing age (15-42 years old)
Used updated estimates of the cost of
education
Changed valuation to use cost of illness
(COI) as basis for monetizing avoided cancer
cases since skin cancer is generally not fatal
Estimated changes in number of CVD
(Leggett model) based on ingestion of
recreationally -caught fish. Monetization
based on VSL
Expanded discussion of the potential benefits
of reducing pollutant concentrations below
MCLs
Other benefits
Non-market benefits from
surface water quality
improvements
Data sources
Water quality index
Willingness-to-pay function
Benefitting population
Added changes in suspended solid
concentration (from SPARROW)
New data for monetization to reflect
parameters in new meta-analysis function
No change
Revised meta-analysis to include spatial
characteristics of the affected water
resources: size of the market, waterbody
characteristics (length and flow), availability
of substitute sites, land use type in the
abutting counties
Test alternative definition of the unit of the
analysis as the Census Block Group; all
households in a given Census Block Group
value all water quality changes in a 100 mile
radius
Benefits from
groundwater quality
improvements
Willingness-to-pay for water quality
Avoided human health hazards
Did not estimate benefits given promulgation
of the final CCR rule
Benefits to threatened and
endangered species
Categorical analysis based on habitat
overlap/proximity
Monetization [willingness to pay]
No change
Benefits from avoided
impoundment failures
Projected impoundment use at steam electric
power plants
Updated to account for projected effects of
the CCR rule
Average failure rate and expected impacts of
failures
Expressed risk of failure and impacts for two
impoundment categories (small, big)
depending on size and failure type
Residual failure rate
Revised to reflect effects of the CCR rule
Cleanup costs
Updated to include more recent data and/or
more detailed review
September 29, 2015
A-2
-------
*** E.O. 12866 Review-Revised Version-Do Not Cite, Quote, or Release During Review ***
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix A: Changes to Benefits Analysis
Table A-1: Changes to Benefits Analysis Since Proposal
Report Section or
Benefit Category
Reduced emissions of
NOX and SO2
Reduced emissions of
C02
Benefits from reduced
water withdrawals
Enhanced CCR
marketability for
beneficial uses
Avoided maintenance
dredging costs
Reduced water treatment
costs
Analysis Component
[Proposed rule analysis value]
Transaction costs
Natural resources damages
Changes in air pollutant levels from IPM and
engineering analysis
Benefit-per-ton estimates for avoided human
health impacts
Changes in air pollutant levels from IPM and
engineering analysis
Monetization [based on social cost of carbon
(SCC) (Interagency Working Group on
Social Cost of Carbon, 2010)]
Surface water withdrawals
Groundwater withdrawals
N/A
N/A
Qualitative discussion
Changes to Analysis for Final Rule
[Final rule analysis value ]
Updated to include more recent data and/or
more detailed review
Updated to include more recent data and/or
more detailed review
No change
No change
Updated SCC to reflect most recent value in
Interagency Working Group on Social Cost
of Carbon (2013)
Expanded qualitative discussion
No change
Estimated changes in beneficial uses of CCR
due to conversions from wet to dry handling.
Estimated benefits based on approach used
for the final CCR rule analysis.
Estimated reductions in volume of sediment
dredged from waterways and reservoirs.
Estimated benefits based on the avoided
costs for maintenance dredging.
Expanded qualitative discussion to provide
more information on potential impacts of
bromide discharges on disinfection
byproducts (and benefits from reducing
discharges)
Environmental Justice
Environmental justice
Profile of affected populations [compare
demographic characteristics of population
affected by steam electric plant discharges to
the state population, using 100-mile buffer
distance]
N/A
Use Census Block Group as the unit of
analysis; use various distance buffers that
reflect distance travelled by different
socioeconomic groups (1, 3, 15, 30, and 50
miles), based on comments
Evaluated EJ considerations across the
regulatory options by explicitly analyzing
distribution of human health benefits by
socioeconomic subgroup
September 29, 2015
A-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Appendix B. Analysis for Scenario without CPP Rule
EPA included the anticipated effects of the Clean Power Plan (CPP) rule in its baseline for the final ELG rule
analysis,, as described in Chapter 1. This baseline was developed based on information about conversions,
retirements, and other changes EPA projected in response to the CPP rule, as proposed by EPA in June 2014.
In particular, EPA updated its steam electric power plants profile to account for additional changes due to
CPP implementation that affect plant costs for meeting the ELGs and pollutant loads (see TDD for details).
The results presented in the main body of this document are based on this scenario with CPP.
This appendix presents the results of an alternative benefits analysis using a baseline that does not include the
incremental conversions, retirements, and other changes projected to occur in response to the CPP rule. All
other assumptions match those described in Chapters 3 through 10.
Table B-1. Pollutant Removal for Final ELGs Regulatory Options for Scenario without CPP
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Pollutant Load Reduction (pounds per year)
Conventional
Pollutants3
5,810,000
5,810,000
13,600,000
18,500,000
19,100,000
Priority Pollutants
285,000
386,000
528,000
618,000
665,000
Nonconventional
Pollutants'1
151,000,000
161,000,000
377,000,000
514,000,000
526,000,000
Toxic- Weighted Pound
Equivalent
1,110,000
1,250,000
1,630,000
1,870,000
1,910,000
a. The loadings reduction for conventional pollutants includes BOD and TSS.
b. The loadings reduction for nonconventional pollutants excludes TDS and COD to avoid double-counting removals for certain
pollutants that would also be measured by these bulk parameters (e.g., sodium, magnesium).
Additionally, under the scenario without CPP, EPA estimates that the final BAT/PSES option (Option D) will
reduce surface water withdrawals at steam electric power plants by 209 to 222 billion gallons per year (0.57 to
0.61 million gallons per day) and will avoid withdrawals of 8 million gallons of groundwater per year
(21,971 gallons per day).
B.1
Human Health Benefit
B.1.1 Benefits to Children from Reduced Lead Exposure
Table B-2. Estimated Benefits from Avoided IQ Losses for Children Exposed to Lead for Scenario
without CPP
Regulatory
Option
Option A
Option B
Option C
Average
Annual
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
3,326,127
Total
Avoided IQ
Losses, 2021
to 2042
893
893
3,940
Annualized Value of Avoided IQ Point Losses" (Millions
2013$)
3% Discount Rate
Low Bound
$0.35
$0.35
$1.57
High Bound
$0.50
$0.50
$2.20
7% Discount Rate
Low Bound
$0.06
$0.06
$0.25
High Bound
$0.09
$0.09
$0.38
September 29, 2015
B-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-2. Estimated Benefits from Avoided IQ Losses for Children Exposed to Lead for Scenario
without CPP
Regulatory
Option
Option D
Option E
Average
Annual
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
Total
Avoided IQ
Losses, 2021
to 2042
4,693
4,693
Annualized Value of Avoided IQ Point Losses" (Millions
2013$)
3% Discount Rate
Low Bound
$1.86
$1.86
High Bound
$2.62
$2.62
7% Discount Rate
Low Bound
$0.30
$0.30
High Bound
$0.45
$0.45
Source: U.S. EPA Analysis, June 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76% of lifetime earnings (following Schwartz, 1994);
high bound assumes that the loss of one IQ point results in the loss of 2.38% of lifetime earnings (following Salkever, 1995).
Table B-3. Estimated Avoided Cost of Compensatory Education for Children with Blood Lead
Concentrations above 20 |ag/dL and IQ Less than 70a for Scenario without CPP
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
3,326,127
3,326,127
3,326,127
Decrease in Number
of Cases of IQ< 70,
in 2021 to 2042
1
1
2
3
3
Avoided Annual Cost (Millions; 2013$)
3% Discount Rate
$0.00
$0.00
$0.01
$0.02
$0.02
7% Discount Rate
$0.00
$0.00
$0.01
$0.01
$0.01
Source: U.S. EPA Analysis, June 2015
a. "-" indicates that a value was not estimated and "$0.00" indicates that avoided annual cost is less than $0.01.
B.1.2 Benefits to Adults from Reduced Lead Exposure
Table B-4. Summary of Health Benefits due to Decreased Risk of CVD Mortality during 2019-2042
based on the Economic Value of Avoided Premature Mortality (VSL) for Scenario without CPP
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Avoided premature
deaths
11.4
11.4
48.0
62.1
62.1
Annualized Benefits (millions 2013$)
3 Percent
$4.03
$4.03
$16.98
$21.93
$21.93
7 Percent
$3.36
$3.36
$14.17
$18.30
$18.30
Source: EPA Analysis, June 2015
September 29, 2015
B-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.1.3 Benefits to Children from Reduced Mercury Exposure
Table B-5. Estimated Benefits from Avoided IQ Losses for Infants from Mercury Exposure for
Scenario without CPP
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of
Affected
Infants per
Year
418,953
418,953
418,953
418,953
418,953
Total Avoided
IQ Losses,
2021 to 2042
4,519
4,610
11,851
13,351
14,189
Annualized Value of Avoided IQ Point Losses" (Millions 2013$)
3% Discount Rate
Low Bound
$1.80
$1.83
$4.71
$5.31
$5.64
High Bound
$2.52
$2.57
$6.62
$7.46
$7.92
7% Discount Rate
Low Bound
$0.29
$0.30
$0.76
$0.86
$0.92
High Bound
$0.43
$0.44
$1.14
$1.28
$1.36
Source: U.S. EPA Analysis, 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76 percent of lifetime earnings (following Schwartz,
1994); high bound assumes that the loss of one IQ point results in the loss of 2.38 percent of lifetime earnings (following Salkever,
1995).
B.1.4 Reduced Cancer Cases from Arsenic Exposure
Table B-6. Annual Benefits from Reduced Cancer Cases due to Arsenic Exposure3 for Scenario
without CPP
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Annual
Affected
Population
35,972,005
35,972,005
35,972,005
35,972,005
35,972,005
Reduced Cancer Cases,
2019 to 2042
0.04
0.04
0.13
0.16
0.17
Benefits (Millions 2013$)
3% Discount
$0.00
$0.00
$0.00
$0.00
$0.00
7% Discount
$0.00
$0.00
$0.00
$0.00
$0.00
Source: U.S. EPA Analysis, 2015
a. "-" indicates that a value was not estimated and "$0.00" indicates that annual benefits are less than $0.01 million.
September 29, 2015
B-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.1.5 Total Monetized Human Health Benefits
Table B-7. Total Monetized Human Health Benefits for ELG Options (millions of 2013$)a'b for Scenario
without CPP
Discount
Rate
3%
7%
Option
A
B
C
D
E
A
B
C
D
E
Reduced Lead
Exposure for
Children
Low
$0.35
$0.35
$1.58
$1.88
$1.88
$0.06
$0.06
$0.26
$0.31
$0.31
High
$0.50
$0.50
$2.21
$2.64
$2.64
$0.09
$0.09
$0.39
$0.46
$0.46
Reduced
Lead
Exposure
for Adults
$4.03
$4.03
$16.98
$21.93
$21.93
$3.36
$3.36
$14.17
$18.30
$18.30
Reduced Mercury
Exposure for
Children
Low
$1.80
$1.83
$4.71
$5.31
$5.64
$0.29
$0.30
$0.76
$0.86
$0.92
High
$2.52
$2.57
$6.62
$7.46
$7.92
$0.43
$0.44
$1.14
$1.28
$1.36
Reduced
Cancer
Cases from
Arsenic
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
$0.00
Total
Low
$6.18
$6.21
$23.27
$29.12
$29.45
$3.71
$3.72
$15.19
$19.47
$19.53
High
$7.05
$7.10
$25.81
$32.03
$32.49
$3.88
$3.89
$15.70
$20.04
$20.12
Source: U.S. EPA Analysis, June 2015
a. "$0.00" indicates that annual benefits are less than $0.01 million.
b. Low bound assumes that the loss of one IQ point results in the loss of 1.76 percent of lifetime earnings (following Schwartz,
1994); high bound assumes that the loss of one IQ point results in the loss of 2.38 percent of lifetime earnings (following Salkever,
1995).
B.1.6 Additional Measures of Human Health Benefits
Table B-8. Reaches Exceeding Human Health Criteria for Steam Electric Pollutants for Scenario
without CPP
Regulatory
Option
Baseline
Option A
Option B
Option C
Option D
Option E
Number of Reaches with
Steam Electric Pollutant3
Concentrations Exceeding
Human Health Criteria for at
Least One Pollutant
3,959
,3,334
3,334
2,400
1,904
1,753
Number of Reaches with Improved Water Quality,
Relative to Baseline
Number of Reaches with
Fewer Exceedancesb
-
712
714
1,716
2,215
2,303
Number of Reaches with All
Exceedances Eliminated
-
625
625
1,559
2,055
2,206
Source: U.S. EPA Analysis, 2015
a. Pollutants include arsenic, copper, nickel, selenium, thallium, zinc, cadmium, chromium, lead, and mercury.
b. The number of reaches with exceedances reduced includes those reaches where all exceedances are eliminated.
September 29, 2015
B-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.2 Non-Market Benefits for Water Quality Improvements
Table B-9: Total Willingness-to-Pay for Water Quality Improvements
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Number of Affected
Households
(Millions)
39.9
70.4
76.3
94.8
101.4
3% Discount Rate
Low
$5.1
$15.1
$20.7
$25.8
$27.6
Mean
$6.0
$19.2
$27.8
$35.3
$38.0
High
$28.4
$84.4
$115.5
$143.6
$154.0
7% Discount Rate
Low
$4.0
$12.1
$16.6
$20.6
$22.1
Mean
$4.8
$15.4
$22.3
$28.3
$30.5
High
$22.6
$67.3
$92.2
$114.6
$122.9
Source: U.S. EPA Analysis, 2015
B.3 Impacts and Benefits to Threatened and Endangered Species
Table B-10. T&E Species with Habitat Occurring within Waterbodies Affected by Steam Electric Power
Plants for Scenario without CPP
Species Group
Amphibians
Arachnids
Birds
Clams
Crustaceans
Fishes
Insects
Mammals
Reptiles
Snails
Total
Species Vulnerability
Low
0
4
10
0
0
0
10
17
3
8
52
Moderate
4
0
2
0
1
0
2
6
1
0
16
High
2
0
1
37
2
21
1
1
5
14
84
Species Count
6
4
13
37
3
21
13
24
9
22
152
Source: U.S. EPA Analysis, 2015
September 29, 2015
B-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-11. Estimated Annualized Benefits to T&E Species from WQ Improvements (Millions 2013$)a' for Scenario without CPP
Discount
Rate
3%
7%
State
AL
GA
MD
SC
Total
AL
GA
MD
SC
Total
Option A
Low
<$0.01
<$0.01
$0.00
$0.00
$0.01
<$0.01
<$0.01
$0.00
$0.00
$0.01
Medium
<$0.01
$0.01
$0.00
$0.00
$0.01
<$0.01
$0.01
$0.00
$0.00
$0.01
High
<$0.01
$0.01
$0.00
$0.00
$0.02
<$0.01
$0.01
$0.00
$0.00
$0.01
Option B
Low
<$0.01
<$0.01
$0.00
$0.00
$0.01
<$0.01
<$0.01
$0.00
$0.00
$0.01
Medium
<$0.01
$0.01
$0.00
$0.00
$0.01
<$0.01
$0.01
$0.00
$0.00
$0.01
High
<$0.01
$0.01
$0.00
$0.00
$0.02
<$0.01
$0.01
$0.00
$0.00
$0.01
Option C
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
Option D
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
Option E
Low
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
<$0.01
<$0.01
<$0.01
<$0.01
$0.01
Medium
<$0.01
$0.01
$0.01
<$0.01
$0.02
<$0.01
$0.01
$0.01
<$0.01
$0.02
High
<$0.01
$0.01
$0.01
$0.01
$0.03
<$0.01
$0.01
$0.01
<$0.01
$0.03
September 29, 2015
B-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.4 Benefits from Avoided Impoundment Failures
Table B-12: Steam Electric Impoundments by Size in ELG Baseline for Scenario without CPP
Type
Big
Small
Total
Number of Impoundments
Count
732
285
1,017
Percent of Total
72%
28%
100%
Impoundment Capacity
Total (million gallons)
1,031,833
44,202
1,076,036
Percent of Total
96%
4%
100%
Table B-13. Estimated Annualized Benefits of Avoided Impoundment Failures by Release Type for
Scenario without CPP (Millions; 2013$)a
Discount
Rate
3%
7%
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Option A
Option B
Option C
Option D
Option E
Wall Breaches
$25.3
$25.3
$101.7
$136.3
$136.3
$20.3
$20.3
$82.6
$110.6
$110.6
Other Releases
$9.8
$9.8
$39.4
$52.9
$52.9
$7.9
$7.9
$31.9
$42.9
$42.9
All Releases
$35.2
$35.2
$141.1
$189.1
$189.1
$28.1
$28.1
$114.5
$153.5
$153.5
Source: U.S. EPA Analysis, 2015
a. Baseline value of total failure costs minus option value of total failure costs.
B.
5 Air-Related Benefits
Table B-14. Estimated Changes in Electricity Consumption and Air Pollutant Emissions due to
Increase in Auxiliary Service at Steam Electric Power Plants, Relative to Baseline for Scenario
without CPP
Regulatory
Option
Option B
Year
2015-2018
2019
2020
2021
2022
2023-2042
Electricity
Consumption
(MWh)
0.0
51,163.2
69,022.4
105,250.1
119,561.0
139,648.1
CO2 (Metric
Tonnes/Year)
0.0
35,363.6
47,657.8
73,942.8
86,906.2
102,655.1
NOx (Tons/Year)
0.0
20.4
28.2
43.1
54.3
70.9
SO2 (Tons/Year)
0.0
37.1
51.5
75.4
81.7
93.0
September 29, 2015
B-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-14. Estimated Changes in Electricity Consumption and Air Pollutant Emissions due to
Increase in Auxiliary Service at Steam Electric Power Plants, Relative to Baseline for Scenario
without CPP
Regulatory
Option
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2042
Electricity
Consumption
(MWh)
0.0
95,434.4
146,933.1
223,365.9
270,636.7
339,202.2
CO2 (Metric
Tonnes/Year)
0.0
68,584.9
101,292.8
153,405.3
191,064.6
238,871.5
NOx (Tons/Year)
0.0
48.3
80.3
112.5
146.4
193.5
SO2 (Tons/Year)
0.0
66.1
93.3
144.4
176.0
212.4
Source: U.S. EPA Analysis, 2015; see TDD for details.
table B-15. Estimated Changes in Annual Air Pollutant Emissions due to Increased Trucking at
team Electric Power Plants, Relative to Baseline for Scenario without CPP
Regulatory
Option
Option B
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2042
2015-2018
2019
2020
2021
2022
2023-2042
CO2 (Metric Tonnes/Year)
0.0
775.3
918.4
1,092.2
1,143.4
1,218.8
0.0
2,401.2
2,772.4
3,551.3
3,972.7
4,277.3
NOx (Tons/Year)
0.0
0.3
0.4
0.5
0.5
0.5
0.0
1.0
1.2
1.5
1.7
1.9
SO2 (Tons/Year)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Source: U.S. EPA Analysis, 2015; see TDD for details.
Table B-16. Estimated Changes in Annual Air Pollutant Emissions due to Changes in Electricity
Generation Profile, Relative to Baseline for Scenario without CPP
Regulatory
Option
Year
CO2 (Metric Tonnes/Year)
NOx (Tons/Year)
SO2 (Tons/Year)
With CPP (Also Applied to Without CPP Scenario)
Option B
2015-2018
2019-2022
2023-2027
2028-2033
2034-2042
0.0
-2,057,293.5
-1,437,164.1
-246,295.3
-1,186,462.9
0.0
-3,534.5
-3,323.9
-1,361.2
-1,500.2
0.0
-4,620.6
-527.5
1,315.7
-1,608.4
September 29, 2015
B-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-16. Estimated Changes in Annual Air Pollutant Emissions due to Changes in Electricity
Generation Profile, Relative to Baseline for Scenario without CPP
Regulatory
Option
Option D
Year
2015-2018
2019-2022
2023-2027
2028-2033
2034-2042
CO2 (Metric Tonnes/Year)
0.0
-4,869,524.3
-2,555,361.0
-2,089,591.4
-3,193,009.0
NOx (Tons/Year)
0.0
-14,614.0
-11,615.0
-8,826.0
-10,638.8
SO2 (Tons/Year)
0.0
-5,662.8
2,238.4
-984.2
-4,243.9
Source: U.S. EPA Analysis, 2015; see TDD for details.
Table B-17. Estimated Net Changes in Air Pollutant Emissions due to Increase in Auxiliary Service
at Steam Electric Power Plants, Increased Trucking at Steam Electric Power Plants, and Changes in
Electricity Generation Profile, Relative to Baseline for Scenario without CPP
Regulatory
Option
Option B
Option D
Year
2015-2018
2019
2020
2021
2022
2023-2027
2028-2033
2034-2042
2015-2018
2019
2020
2021
2022
2023-2027
2028-2033
2034-2042
CO2 (Metric Tonnes/Year)
0.0
-2,021,154.6
-2,008,717.3
-1,982,258.5
-1,969,243.9
-1,333,290.2
-142,421.36
-1,082,588.96
0.0
-4,798,538.2
-4,765,459.0
-4,712,567.7
-4,674,486.9
-2,312,212.2
-1,846,442.6
-2,949,860.2
NOx (Tons/Year)
0.0
-3,513.7
-3,505.9
-3,490.9
-3,479.6
-3,252.5
-1,289.73
-1,428.78
0.0
-14,564.6
-14,532.5
-14,499.9
-14,465.9
-11,419.6
-8,630.6
-10,443.5
SO2 (Tons/Year)
0.0
-4,583.5
-4,569.1
-4,545.2
-4,538.8
-434.5
1,408.69
-1,515.47
0.0
-5,596.7
-5,569.4
-5,518.4
-5,486.8
2,450.9
-771.7
-4,031.4
Source: U.S. EPA Analysis, 2015
September 29, 2015
B-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-18. Estimated Benefits from Reduced Air Emissions for Selected Years for Scenario without
CPP (millions; 2013$)a'b
Option
Option B
Option D
Year
2019C
2020
2025
2030
2019C
2020
2025
2030
3% Discount Rate
$287.2
$290.3
$105.3
-$47.5
$513.5
$518.7
$83.0
$192.4
7% Discount Rate
$265.8
$269.2
$101.4
-$42.4
$481.3
$487.6
$86.1
$183.7
Source: U.S. EPA Analysis, 2015
a. EPA used SCC values based on a 3 percent (average) discount rate to calculate total benefit values presented for both the 3
percent and 7 percent discount rate.
b. EPA used the changes in annual air pollutant emissions due to changes in electricity generation profile for scenario with CPP
(from IPM analysis) when calculating benefits for the scenario without CPP
c. The benefits per ton values used for year 2019 benefit calculation is assumed to be the same as the 2020 benefits per ton values.
Table B-19. Estimated Annualized Benefits from Reduced Air Emissions for Scenario without CPP
(Millions; 2013$)a
ELG Option
Option B
Option D
Pollutant
NOX
S02
C02
TOTAL
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
NOX
SO2
C02
TOTAL
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Avg
5% Avg
2.5% Avg
3% 95th Percentile
3% Discount Rate
$12.7
$44.8
$51.9
$15.1
$76.5
$155.9
$109.3
$72.6
$134.0
$213.4
$62.9
$81.0
$138.4
$40.0
$204.7
$416.9
$282.4
$183.9
$348.6
$560.8
7% Discount Rate
$10.6
$40.4
$51.9
$15.1
$76.5
$155.9
$102.8
$66.1
$127.4
$206.8
$49.3
$58.9
$138.4
$40.0
$204.7
$416.9
$246.7
$148.3
$313.0
$525.2
Source: U.S. EPA Analysis, 2015
a. EPA used the changes in annual air pollutant emissions due to changes in electricity generation profile for scenario with CPP
(from IPM analysis) when calculating benefits for the scenario without CPP
September 29, 2015
B-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B
.6 Benefits from Reduced Water Withdrawals
Table B-20. Estimated Annualized Benefits from Reduced Groundwater Withdrawals (Millions;
2013$)
Regulatory Option
Option A
Option B
Option C
Option D
Option E
Reduction in
Groundwater Intakes
(million gallons per year;
full implementation)
0.0
0.0
0.0
8.0
8.0
3% Discount Rate
$0.00
$0.00
$0.00
$0.02
$0.02
7% Discount Rate
$0.00
$0.00
$0.00
$0.02
$0.02
Source: U.S. EPA Analysis, 2015
B.7 Benefits from Enhanced Marketability of Coal Combustion Residuals
Table B-212. State-level Market Approximation (Short Tons).
Application
Concrete
Fill
Baseline Production
Production
13,400,062
320,701,683
Unmet Demand (%
of production)
3,277,184
(24%)
312,705,505(98%)
Changes due to ELGs
Fly Ash
(% ACCR)
166,535
(9.6%)
1,575,450
(90.4%)
Bottom Ash
(% ACCR)
-
5,480,176(92.7%)
Source: U.S. EPA Analysis 2015
Table B-23. Estimated Beneficial Use Applications of CCRs, by CCR Category
Beneficial Use Application (1,000 short tons)
ELG Option and CCR Category
Concrete
Structural Fill
Option A or Option B
Fly ash
Bottom ash
Total
167
NA
767
1,575
0
7,575
Option C
Fly ash
Bottom ash
Total
167
NA
767
1,575
3,431
5,006
Option D or Option E
Fly ash
Bottom ash
Total
167
NA
767
1,575
4,821
6,397
September 29, 2015
B-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-224. Annual Avoided Resource and Environmental Impacts Given CCR Reuse in Concrete
and Fill Applications for Scenario Without CPP
Impact Category
Fly Ash
Bottom Ash
Option Total
CCR RIA, 2030
3-yr rolling avg
Option A or Option B
Concrete (mill, tons)
Structural fill (mill, tons)
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
0.167
1.58
657,548
1,937
1,442,070
359
54
NA
0
0
0
0
0
0
0.167
1.58
657,548
1,937
1,442,070
359
54
0.21
6.93
910,000
21,000
75,000
310
59
Option C
Concrete (mill, tons)
Structural fill (mill, tons)
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
0.167
1.58
657,548
1,937
1,442,070
359
54
NA
3.43
137,232
4,144
10,979
92
29
0.167
5.06
794,780
6,081
1,453,048
451
83
0.21
6.93
910,000
21,000
75,000
310
160
Option D or Option E
Concrete (mill, tons)
Structural fill (mill, tons)
Energy (MMBtu)
Water (million gal)
Greenhouse gases (tons)
NOx (tons)
SOx (tons)
0.167
1.58
657,548
1,937
1,442,070
359
54
NA
4.82
192,848
5,824
15,428
129
41
0.167
6.39
850,396
7,761
1,457,498
488
95
0.21
6.93
910,000
21,000
75,000
310
160
Note: Values in this table represent annual changes in a full-compliance year (e.g., starting in 2023).
September 29, 2015
B-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-25. Annualized Economic Value of Estimated Changes in Beneficial Use (Million 2013$).
Regulatory Option
Option A or
Option B
Option C
Option D or
Option E
Impact Category
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials
Net Social Value
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials
Net Social Value
Avoided Disposal Costs to Steam Electric Plants
Beneficiation Costs
Avoided Life Cycle Costs of Virgin Materials
Net Social Value
3%
$6.44
-$0.21
$11.34
$17.57
$21.97
-$0.21
$13.25
$35.01
$34.43
-$0.21
$13.94
$48.16
7%
$5.46
-$0.16
$14.52
$19.82
$18.39
-$0.16
$16.69
$34.92
$28.59
-$0.16
$17.48
$45.90
Notes: Annualized over 24 years (2015 - 2042). Values escalated using CCI and GDP through 2022; thereafter, assume no real
change in prices above inflation. Avoided disposal costs to steam electric power plants include annual O&M costs.
September 29, 2015
B-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.8 T<
btal Monetized Benefits
Table B-26. Summary of Total Annualized Benefits at 3 Percent (Millions; 2013$)
Benefit Category
Human Health Benefits
Reduced IQ losses in children
from exposure to lead3
Reduced CVD in adults from
exposure to lead
Reduced IQ losses in children
from exposure to mercury3
Avoided cancer cases from
exposure to arsenicb
Improved Ecological
Conditions and Recreational
Uses
Use and nonuse values for
water quality improvements
Nonuse values of T&E species'3
Market and Productivity
Benefits
Avoided impoundment failures
Reduced dredging costs'3
Ash marketing benefits
Air-related benefits
Reduced human health effects
Reduced CO2 emissions0
Reduced water withdrawals'"
Total (excluding air-related
Benefits)"1
Total (including air-related
Benefits)"'6
Option A
Low
$6.2
$0.4
Midf
$6.6
$0.4
High
$7.1
$0.5
$4.0
$1.8
$2.2
$2.5
$0.0
$5.1
$5.1
$0.0
$6.1
$6.0
$0.0
$28.4
$28.4
$0.0
$52.7
$35.2
$0.0
$17.6
NE
NE
NE
$0.0
$64.0
NE
$65.4
NE
$88.2
NE
Option B
Low
$6.2
$0.4
Midf
$6.7
$0.4
High
$7.1
$0.5
$4.0
$1.8
$2.2
$2.6
$0.0
$15.2
$15.1
$0.0
$19.2
$19.2
$0.0
$84.4
$84.4
$0.0
$52.7
$35.2
$0.0
$17.6
$109.4
$57.5
$51.9
$0.0
$74.1
$183.5
$78.6
$188.0
$144.3
$253.6
Option C
Low
$23.3
$1.6
Midf
$24.5
$1.9
High
$25.8
$2.2
$17.0
$4.7
$5.7
$6.6
$0.0
$20.7
$20.7
$0.0
$27.8
$27.8
$0.0
$115.5
$115.5
$0.0
$176.1
$141.1
$0.0
$35.0
NE
NE
NE
$0.0
$220.1
NE
$228.4
NE
$317.4
NE
Option D
Low
$29.1
$1.9
Midf
$30.6
$2.3
High
$32.0
$2.6
$21.9
$5.3
$6.4
$7.5
$0.0
$25.8
$25.8
$0.0
$35.3
$35.3
$0.0
$143.7
$143.6
$0.0
$237.3
$189.1
$0.0
$48.2
$282.4
$143.9
$138.4
$0.0
$292.2
$574.6
$303.2
$585.5
$413.0
$695.4
Option E
Low
$29.4
$1.9
Midf
$31.0
$2.3
High
$32.5
$2.6
$21.9
$5.6
$6.8
$7.9
$0.0
$27.6
$27.6
$0.0
$38.0
$38.0
$0.0
$154.0
$154.0
$0.0
$237.3
$189.1
$0.0
$48.2
NE
NE
NE
$0.0
$294.4
NE
$306.3
NE
$423.8
NE
Source: U.S. EPA Analysis, 2015
September 29, 2015
B-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-26. Summary of Total Annualized Benefits at 3 Percent (Millions; 2013$)
Benefit Category
Option A
Low
Midf
High
Option B
Low
Midf
High
Option C
Low
Midf
High
Option D
Low
Midf
High
Option E
Low
Midf
High
"NE" indicates that EPA did not estimate the benefits. Air-related benefits of Option A are expected to be less than those for Option B; air-related benefits for Option C are expected to be
between those of Options B and D; and air-related benefits of Option E are expected to be greater than those for Option D.
a. Value includes reduced IQ losses and avoided cost of compensatory education in children from exposure to lead. For details see Chapter 3.
b. "< $0.1" indicates that the monetized annual benefits are positive but less than $0.1 million.
c. For the valuation of benefits from reductions in CO2 emissions EPA relied on the 3 percent average social cost of carbon estimate.
d. Values for individual benefit categories may not sum to the total due to independent rounding.
e. The total monetized benefits for options A, C, and E do not include air-related benefits. This category of benefits was analyzed for Options B and D only (see Chapter 7).
f EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of willingness-to-pay. One model provides the low and high bounds
while a different model provides a central estimate (included in this table in the mid-range column). For this reason, the mid-range estimate differs from the midpoint of the range for this
benefit category. For details, see Chapter 4.
September 29, 2015
B-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-27. Summary of Total Annualized Benefits at 7 Percent (Millions; 2013$)
Benefit Category
Human Health Benefits
Reduced IQ losses in children
from exposure to lead3
Reduced CVD in adults from
exposure to lead
Reduced IQ losses in children
from exposure to mercury3
Avoided cancer cases from
exposure to arsenicb
Improved Ecological
Conditions and Recreational
Uses
Use and nonuse values for
water quality improvements
Nonuse values of T&E species'3
Market and Productivity
Benefits
Avoided impoundment failures
Reduced dredging costs'3
Ash marketing benefits
Air-Related Benefits
Reduced human health effects
Reduced CO2 emissions0
Reduced water withdrawals'"
Total (excluding air-related
Benefits)d
Total (including air-related
Benefits)"1'6
Option A
Low
$3.7
$0.1
Midf
$3.8
$0.1
High
$7.2
$0.1
$3.4
$0.3
$0.4
$3.8
$0.0
$4.1
$4.0
$0.0
$4.8
$4.8
$0.0
$22.6
$22.6
$0.0
$47.9
$28.1
$0.0
$19.8
NE
NE
NE
$0.0
$55.7
NE
$56.6
NE
$77.7
NE
Option B
Low
$3.7
$0.1
Midf
$3.8
$0.1
High
$22.2
$0.1
$3.4
$0.3
$0.4
$18.7
$0.0
$12.1
$12.1
$0.0
$15.4
$15.4
$0.0
$67.3
$67.3
$0.0
$47.9
$28.1
$0.0
$19.8
$102.8
$50.92
$51.85
$0.0
$63.7
$166.5
$67.1
$169.9
$137.4
$240.2
Option C
Low
$15.2
$0.3
Midf
$15.4
$0.3
High
$15.7
$0.4
$14.2
$0.8
$1.0
$1.1
$0.0
$16.6
$16.6
$0.0
$22.3
$22.3
$0.0
$92.2
$92.2
$0.0
$149.4
$114.5
$0.0
$34.9
NE
NE
NE
$0.0
$181.2
NE
$187.2
NE
$257.4
NE
Option D
Low
$19.5
$0.3
Midf
$19.8
$0.4
High
$20.0
$0.5
$18.3
$0.9
$1.1
$1.3
$0.0
$20.6
$20.6
$0.0
$28.3
$28.3
$0.0
$114.6
$114.6
$0.0
$199.4
$153.5
$0.0
$45.9
$246.7
$108.28
$138.42
$0.0
$239.5
$486.2
$247.5
$494.2
$334.1
$580.8
Option E
Low
$19.5
$0.3
Midf
$19.8
$0.4
High
$20.1
$0.5
$18.3
$0.9
$1.1
$1.4
$0.0
$22.1
$22.1
$0.0
$30.5
$30.5
$0.0
$122.9
$122.9
$0.0
$199.4
$153.5
$0.0
$45.9
NE
NE
NE
$0.0
$241.0
NE
$249.8
NE
$296.5
NE
Source: U.S. EPA Analysis, 2015
"NE" indicates that EPA did not estimate the benefits. Air-related benefits of Option A are expected to be less than those for Option B; air-related benefits for Option C are expected to be
between those of Options B and D; and air-related benefits of Option E are expected to be greater than those for Option D.
a. Value includes reduced IQ losses and avoided cost of compensatory education in children from exposure to lead. For details see Chapter 3.
b. "< $0.1" indicates that the monetized annual benefits are positive but less than $0.1 million.
September 29, 2015
B-16
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-27. Summary of Total Annualized Benefits at 7 Percent (Millions; 2013$)
Benefit Category
Option A
Low
Midf
High
Option B
Low
Midf
High
Option C
Low
Midf
High
Option D
Low
Midf
High
Option E
Low
Midf
High
c. For the valuation of benefits from reductions in CO2 emissions EPA relied on the 3 percent average social cost of carbon estimate.
d. Values for individual benefit categories may not sum to the total due to independent rounding.
e. The total monetized benefits for options A, C, and E do not include air-related benefits. This category of benefits was analyzed for Options B and D only (see Chapter 7).
f EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of willingness-to-pay. One model provides the low and high bounds
while a different model provides a central estimate (included in this table in the mid-range column). For this reason, the mid-range estimate differs from the midpoint of the range for this
benefit category. For details, see Chapter 4.
September 29, 2015
B-17
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-28: Time Profile of Benefits at 3 Percent (Millions; 2013$) (Including Air-Related Benefits for
Options B and D) for Scenario Without CPP
Year
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Benefits,
3%
Option Aa
$0.0
$0.0
$0.0
$0.0
$10.6
$13.2
$30.6
$48.9
$85.6
$86.5
$86.9
$87.1
$87.2
$87.3
$87.2
$87.1
$87.2
$87.2
$87.2
$87.3
$87.3
$87.4
$87.5
$87.7
$87.8
$87.9
$88.1
$88.2
$65.4
Option B
$0.0
$0.0
$0.0
$0.0
$297.9
$303.5
$337.8
$360.5
$202.5
$205.5
$208.0
$210.4
$212.7
$57.5
$56.9
$56.3
$56.0
$55.6
$55.2
$250.6
$253.1
$255.6
$258.1
$260.6
$263.1
$265.7
$268.3
$269.6
$188.0
Option Ca
$0.0
$0.0
$0.0
$0.0
$27.7
$83.6
$138.0
$244.3
$295.2
$298.2
$299.0
$299.1
$298.8
$298.3
$297.4
$296.6
$295.9
$295.1
$294.4
$293.8
$293.3
$292.9
$292.5
$292.2
$291.9
$291.8
$291.7
$291.5
$228.4
Option D
$0.0
$0.0
$0.0
$0.0
$555.8
$630.0
$714.3
$833.9
$471.8
$477.4
$480.2
$482.3
$483.7
$583.9
$583.9
$586.2
$588.0
$590.1
$592.3
$827.4
$833.1
$838.9
$844.8
$850.8
$856.8
$863.0
$869.2
$872.2
$585.5
Option Ea
$0.0
$0.0
$0.0
$0.0
$42.3
$111.3
$197.5
$310.8
$396.2
$400.0
$401.0
$401.1
$400.7
$399.9
$398.8
$397.7
$396.7
$395.7
$394.8
$393.9
$393.2
$392.6
$392.1
$391.6
$391.3
$391.0
$390.8
$390.6
$306.3
Source: U.S. EPA Analysis, 2015
a. Estimates for Options A, C and E do not include air-related benefits. This category of benefits was only estimated for Options B and
D (see Chapter 7).
Table B-29: Time Profile of Benefits at 7 Percent (Millions; 2010$) (Including Air-Related Benefits for
Options B and D) for Scenario Without CPP
Year
2015
2016
2017
2018
2019
Option Aa
$0.0
$0.0
$0.0
$0.0
$10.6
Option B
$0.0
$0.0
$0.0
$0.0
$276.5
Option Ca
$0.0
$0.0
$0.0
$0.0
$27.7
Option D
$0.0
$0.0
$0.0
$0.0
$523.6
Option Ea
$0.0
$0.0
$0.0
$0.0
$42.3
September 29, 2015
B-18
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-29: Time Profile of Benefits at 7 Percent (Millions; 2010$) (Including Air-Related Benefits for
Options B and D) for Scenario Without CPP
Year
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Benefits,
7%
Option Aa
$13.2
$28.2
$46.5
$83.2
$84.0
$84.4
$84.6
$84.7
$84.7
$84.6
$84.5
$84.6
$84.6
$84.6
$84.6
$84.6
$84.7
$84.8
$84.9
$85.0
$85.1
$85.3
$85.3
$56.6
Option B
$282.4
$314.1
$336.8
$196.2
$199.2
$201.6
$204.1
$206.3
$60.1
$59.4
$58.8
$58.5
$58.1
$57.7
$240.6
$243.1
$245.5
$248.0
$250.5
$253.0
$255.5
$258.1
$259.4
$169.9
Option Ca
$83.6
$131.1
$237.2
$288.1
$291.0
$291.8
$291.8
$291.5
$290.8
$289.9
$289.0
$288.2
$287.4
$286.7
$286.0
$285.4
$284.9
$284.5
$284.1
$283.8
$283.6
$283.4
$283.2
$187.2
Option D
$598.9
$674.4
$794.0
$467.1
$472.6
$475.1
$477.2
$478.6
$566.8
$566.7
$568.8
$570.6
$572.6
$574.7
$794.8
$800.4
$806.1
$811.9
$817.8
$823.8
$829.8
$836.0
$838.8
$494.2
Option Ea
$111.3
$189.2
$302.4
$387.7
$391.4
$392.3
$392.4
$391.9
$391.0
$389.9
$388.7
$387.6
$386.5
$385.5
$384.6
$383.8
$383.1
$382.5
$382.0
$381.5
$381.2
$380.9
$380.6
$249.8
Source: U.S. EPA Analysis, 2015
a. Estimates for Options A, C and E do not include air-related benefits. This category of benefits was only estimated for Options B and
D (see Chapter 7).
B.9 Total Costs
Table B-30: Summary of Annualized Costs for Scenario without CPP (Millions; $2013)
Regulatory Option
Option A
Option B
Option C
Option D
Option E
3% Discount Rate
$141.1
$238.9
$465.3
$640.5
$711.8
7% Discount Rate
$137.1
$234.9
$461.4
$626.1
$695.8
Source: U.S. EPA Analysis, 2015.
September 29, 2015
B-19
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-31: Time Profile of Costs to Society (Millions; $2013), Scenario Without CPP
Year
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
Annualized Costs, 3%
Annualized Costs, 7%
Option A
$0.0
$0.0
$0.0
$0.0
$474.6
$169.2
$431.3
$223.1
$355.8
$85.8
$95.9
$91.3
$93.8
$90.2
$98.4
$96.3
$99.3
$97.9
$98.7
$88.3
$95.7
$91.8
$96.2
$93.8
$97.8
$97.0
$97.0
$96.3
$141.1
$137.1
Option B
$0.0
$0.0
$0.0
$0.0
$820.7
$377.9
$696.7
$415.8
$618.1
$141.8
$151.9
$147.3
$149.8
$146.2
$154.4
$152.3
$155.2
$153.9
$154.7
$144.2
$151.7
$147.8
$152.2
$149.8
$153.8
$153.0
$153.0
$152.3
$238.9
$234.9
Option C
$0.0
$0.0
$0.0
$0.0
$1,420.3
$855.0
$1,347.4
$1,043.8
$1,367.7
$258.5
$268.5
$265.2
$265.6
$263.8
$279.0
$275.0
$279.7
$277.7
$279.8
$261.7
$269.6
$267.7
$271.2
$271.1
$277.2
$273.9
$273.7
$271.7
$465.3
$461.4
Option D
$0.0
$0.0
$0.0
$0.0
$1,758.7
$1,218.1
$1,657.5
$1,367.5
$1,843.9
$391.0
$401.8
$401.5
$400.0
$398.0
$419.2
$414.1
$419.0
$416.8
$420.4
$395.6
$404.1
$404.2
$407.8
$408.9
$415.4
$411.3
$409.1
$405.2
$640.5
$626.1
Option E
$0.0
$0.0
$0.0
$0.0
$1,900.2
$1,388.1
$1,809.6
$1,597.5
$2,035.9
$434.6
$446.9
$446.8
$445.1
$443.8
$464.6
$457.7
$464.1
$462.1
$465.5
$441.3
$449.5
$447.8
$452.9
$453.9
$460.0
$456.0
$453.4
$448.8
$711.8
$695.8
Source: U.S. EPA Analysis, 2015.
September 29, 2015
B-20
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
B.10.1 Comparison of Total Monetized Benefits and Costs
Table B-32. Total Annualized Benefits and Costs by Regulatory Option and Discount Rate (Millions;
2013$)
Regulatory
Option
Total Monetized Benefits
Low
Mid
High
Total Monetized Benefits, Including
Extrapolated Values"
Low
Mid
High
Total Costs
3% Discount Rate
Option A
Option B
Option C
Option D
Option E
$64.0
$183.5
$220.1
$574.6
$294.4
$65.4
$188.0
$228.4
$585.5
$306.3
$88.2
$253.6
$317.4
$695.4
$423.8
$145.4
$183.5
$499.9
$574.6
$668.6
$144.7
$188.0
$505.3
$585.5
$677.6
$152.8
$253.6
$549.8
$695.4
$734.0
$141.1
$238.9
$465.3
$640.5
$711.9
7% Discount Rate
Option A
Option B
Option C
Option D
Option E
$55.7
$166.5
$181.2
$486.2
$241.0
$56.6
$169.9
$187.2
$494.2
$249.8
$77.7
$240.2
$257.4
$580.8
$296.5
$126.5
$166.5
$411.5
$486.2
$547.4
$125.1
$169.9
$414.1
$494.2
$552.6
$134.6
$240.2
$445.8
$580.8
$513.6
$137.1
$234.9
$461.4
$626.1
$695.8
Source: U.S. EPA Analysis, 2015.
a. EPA did not analyze air-related benefits for Options A, C, and E. This category of benefits was only estimated for Options B and D
(see Chapter 7). EPA adjusted the total benefits estimated for Options A, C and E by multiplying the totals without air-related benefits
by the average ratio of [total with air-related benefits]/[total without air-related benefits] for Options B and D.
B.10.2 Analysis of Incremental Monetized Benefits and Costs
Table B-33. Incremental Net Benefit Analysis for Scenario without CPP (Millions; 2013$)
Regulatory
Option"
Total Annual Monetized
Benefits, Including
Adjusted or Inferred
Low
Values
Midd
High
Total
Social
Costs
Net Annual Monetized
Low
Benefits3
Midd
High
Incremental Net Annual
Monetized Benefits'"
Low
Midd
High
3% Discount Rate
Option A
Option B
Option C
Option D
Option E
$145.4
$183.5
$499.9
$574.6
$668.6
$144.7
$188.0
$505.3
$585.5
$677.6
$152.8
$253.6
$549.8
$695.4
$734.0
$141.1
$238.9
$465.3
$640.5
$711.9
$4.3
-$55.4
$34.6
-$65.9
-$43.3
$3.6
-$50.9
$40.0
-$55.0
-$34.3
$11.7
$14.7
$84.5
$54.9
$22.1
-
-$59.7
$90.0
-$100.5
$22.7
-
-$54.5
$90.9
-$95.0
$20.7
-
$3.0
$69.8
-$29.7
-$32.7
7% Discount Rate
Option A
Option B
Option C
$126.5
$166.5
$411.5
$125.1
$169.9
$414.1
$134.6
$240.2
$445.8
$137.1
$234.9
$461.4
-$10.6
-$68.4
-$49.9
-$12.0
-$65.0
-$47.3
-$2.5
$5.3
-$15.6
-
-$57.8
$18.5
-
-$53.0
$17.6
-
$7.7
-$20.9
September 29, 2015
B-21
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix B: Analysis without CPP Rule
Table B-33. Incremental Net Benefit Analysis for Scenario without CPP (Millions; 2013$)
Regulatory
Option3
Option D
Option E
Total Annual Monetized
Benefits, Including
Adjusted or Inferred
Values
Low
$486.2
$547.4
Midd
$494.2
$552.6
High
$580.8
$513.6
Total
Social
Costs
$626.1
$695.8
Net Annual Monetized
Benefits"
Low
-$139.9
-$148.4
Midd
-$131.9
-$143.2
High
-$45.3
-$182.2
Incremental Net Annual
Monetized Benefits'"
Low
-$90.0
-$8.5
Midd
-$84.6
-$11.3
High
-$29.7
-$136.9
Source: U.S. EPA Analysis, 2015.
a. EPA did not analyze air-related benefits for Options A, C, and E. This category of benefits was only estimated for Options B and D
(see Chapter 7). EPA adjusted the total benefits estimated for Options A, C, and E by multiplying the totals without air-related
benefits by the average ratio of [total with air-related benefits]/[total without air-related benefits] for Options B and D.
b. Net benefits are calculated by subtracting total annualized costs from total annual monetized benefits.
c. Incremental net benefits are equal to the difference between net benefits of an option and net benefits of the previous, less stringent
option.
d. EPA estimated use and nonuse values for water quality improvements using two different meta-regression models of WTP. One
model provides the low and high bounds while a different model provides the mid estimate. For this reason, the mid benefit estimate
differs from the midpoint of the benefits range. For details, see Chapter 4.
September 29, 2015
B-22
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix C: Estimation of Exposed Population
Appendix C. Estimation of Exposed Population
The assessment uses the Census Block Group as the geographic unit of analysis, assigning a radial distance
(e.g., 50 miles or 100 miles) from the Census Block Group centroid. EPA assumes that all modeled reaches
within this range are viable fishing sites, with all unaffected reaches viable substitutes for affected reaches
within the area around the Census Block Group.
By focusing on distance from the Census Block Group, 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 C-l presents a hypothetical example focusing on two Census Block Groups (square at the center of
each circular area), each near five waterbodies with water quality changes under the ELGs (thick red lines).
The same approach is used to identify populations for the analysis of non-market benefits in Chapter 4. In
that case, the circles represent the outer edge of the 100-mile buffer around each block group. Highlighted in
red are the affected NHD reaches under regulatory options for which baseline WQI and AWQI would be
estimated
Figure C-1. Illustration of Intersection of Census Block Groups and COMIDs.
September 29, 2015
C-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix D: Water and Fish Tissue
Appendix D. Derivation of Ambient Water and Fish Tissue
Concentrations in Receiving and Downstream Reaches
This appendix describes the methodology EPA used to estimate in-stream and fish tissue concentrations under
the baseline and each of the five ELG regulatory options. The concentrations are used as inputs to estimate the
water quality improvements and human health benefits of the final rule. Specifically, EPA in-stream metal
concentrations to analyze non-use benefits of water quality improvements (see Chapter 4), and to derive fish
tissue concentrations used to analyze human health effects from consuming self-caught fish (see Chapter 3).
Nutrient and suspended sediment concentrations are used to support analysis of non-use benefits from water
quality improvements (see Chapter 4).
The overall modeling methodology is similar to that used at proposal (see Chapter 4 in U.S. EPA (2013a))
and builds on data and methods described in the Technical Development Document and Environmental
Assessment documents for the final ELGs (U.S. EPA, 2015a; 2015b). The following sections discuss
calculations of the metal concentrations in streams and fish tissue and nutrient and sediment concentrations in
streams.
o.
D.1.1 Estimating Water Concentrations in each Reach
EPA first estimated the baseline and post-compliance metal concentrations in reaches receiving steam electric
power plant discharges and downstream reaches.
The water quality model component of EPA's Risk-Screening Environmental Indicators (RSEI) model (U.S.
EPA, 2012c) 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 National Hydrographic Dataset
(NHD).86 The hydrography network represented in RSEI consists of approximately 2.5 million reaches
(unique COMIDs).87
86 RSEI utilizes the USGS's National Hydrology Dataset (NHD) which 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. The NHD
reaches in this analysis range from 0.003 miles to 9.11 miles in length.
87 Reaches represented in RSEI are those determined to be potentially fishable based on type and physical
characteristics. As documented in U.S. EPA (2012c): "Certain criteria were applied to the NHDPlus dataset to
select the reaches to be used in the model. Specifically, because RSEI calculates the movement of a chemical
release downstream using flow and velocity data, qualifying reaches must have at least one downstream or
upstream connecting reach and have a non-negative flow and velocity. RSEI will 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 pipelines, aqueducts, and certain types of reservoirs.
NHDPlus does not separate canals (presumably fishable) and ditches (presumably not fishable), so RSEI
excludes reaches in the canal/ditch category if the annual mean flow is less than 5 ft3/s. This is an arbitrary
minimum, and is intended primarily to exclude ditches at the point of the facility discharge. For reaches
designated as not fishable in NHDPlus, the chemical is still assumed to travel downstream to the next reach,
which may or may not be fishable."
September 29, 2015 5T
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix D: Water and Fish Tissue
The analysis involved the following key steps for the baseline and each of the five regulatory options:
> Summing plant-level loadings to the COMID. EPA summed the plant-level annual average loads
(see TDD) for each unique COMID receiving plant discharges from steam electric power plants in the
baseline. Chapter 4 in the EA report describes the approach EPA used to identify the receiving
waterbodies (U.S. EPA 2015a).
> Specifying loads in the water quality model. RSEI includes data on annual average pollutant
loadings to surface waters from facilities that reported to the Toxic Release Inventory (TRI) in 2012.
EPA replaced the loadings provided for Steam Electric plant dischargers in the TRI data set with
those obtained in Step 1. Loadings for other TRI reporters were left unchanged.
> Performing dilution and transport calculations. RSEFs water quality model uses a simple dilution
and first-order decay equation (where metals are treated as conservative substances) to estimate
average annual water concentrations for each individual reach. In the model, a plant is assumed to
release its annual load at a constant rate throughout the year. Each source-pollutant release is tracked
downstream throughout the NHD reach network until one of three conditions occurs: 1) the release
has traveled 300 km (186 miles) downstream; 2) the release has traveled a distance equivalent to one
week of travel time; or 3) the concentration reaches 1 x 10"9 mg/L. The 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 reach mean flow (provided in NHDPlus as an attribute of each COMID).
EPA used the approach above to estimate annual average concentrations often metals: arsenic, cadmium,
chromium VI, copper, lead, mercury, nickel, selenium, thallium, and zinc. The results show that, of the
2.5 million reaches represented in RSEI:
> 77,414 reaches (117,518 km) have non-zero concentrations, and
> 18,773 of these reaches (27,778 km) are affected by steam electric power plant discharges in the
baseline.
D.1.2 Estimating Fish Tissue Concentrations in each Reach
To support analysis of the human health benefits associated with water quality improvements (see Chapter 3),
EPA estimated concentrations of arsenic, lead, and mercury in fish tissue based on the RSEI model outputs
discussed above.
The methodology follows the same general approach described in the EA document for estimating fish tissue
concentrations for receiving reaches (U.S. EPA, 2015a), but applies the calculations to the larger set of
reaches modeled using RSEI, which include not only the receiving reaches analyzed in the EA, but also
downstream reaches. Further, the calculations use RSEI-estimated concentrations as inputs, which account not
only for the steam electric discharges, but also other major dischargers that report to TRI.
The analysis involved the following key steps for the baseline and each of the five regulatory options:
> Obtaining the relationship between water concentrations and fish tissue concentrations. EPA
used the results of the EA national model 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 metals.
> Calculating fish tissue data for affected reaches. For reaches for which RSEI provides non-zero
water concentrations (i.e., reaches affected by steam electric power plants or other TRI dischargers),
September 29, 2015 D-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix D: Water and Fish Tissue
EPA used the relationship obtained in Step 1 to calculate a preliminary fish tissue concentration for
each metal.
> Imputing the fish tissue concentrations for all other modeled reaches. For reaches for which
RSEI does not calculate water concentrations, EPA assigned background fish tissue concentrations
based on the 10th percentile of the distribution of reported concentrations in fish tissue samples in the
National Listing Fish Advisory (NLFA) data88 (see Table C-l). EPA found that the distribution of
these samples was consistent with values reported in Wathen et al (2014) and used the 10th percentile
as representative of background, "clean" reaches not affected by point source discharges.
> 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 EA national
model to ensure that the fish tissue concentrations calculated based on the RSEI 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. Section 5.1.2 of the EA
describes the methodology in greater detail.
The analysis provides background metal-specific composite fish fillet concentrations for each COMID
modeled in RSEI.
Table C-1: Assumed Background Fish Tissue Concentrations,
based on 10th percentile
Parameter
As
Hg
Pb
Pollutant Concentration (mg/kg)
0.039
0.058
0.039
D.2 Nutrients and Suspended Sediment
EPA used the USGS's SPARROW model to estimate nutrient and sediment concentrations in receiving and
downstream reaches. The calibrated, national models used for this analysis are the same as those used to
estimate in-stream concentrations of TN, TP and TSS in the Construction and Development Industry
Category ELGs (see U.S. EPA, 2009c). The approach involved the following steps:
> Referencing the receiving reaches to E2RF1 reaches. EPA overlaid the medium resolution NHD
and E2RF1 features in GIS to develop the crosswalk between the two hydrologic networks.
> Summing the loads for each E2RF1. EPA summed the plant-level loadings over each E2RF1 in the
baseline and under each of the five regulatory options.
> Calculating the change in loading for each E2RF1. EPA calculated the difference between the
baseline and post-compliance loadings under each of the five regulatory options.
> Specifying the change in loading in SPARROW. The national SPARROW models for nutrients do
not have an explicit explanatory variable for point source loadings in mass units. In the TN and TP
SPARROW models, point sources (e.g., wastewater treatment plants) are represented by a population
variable. The national calibrated models show contributions of 2.2514 kg TN/capita and
1 See http://mapl.epa.gov/.
September 29, 2015 D-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix D: Water and Fish Tissue
0.2319 kg TP/capita for point sources. EPA used these calibrated loading factors to express the load
reductions obtained under each of the regulatory options into population-equivalent in SPARROW.
This population-equivalent loading was subtracted from the baseline population value for each reach
when running the SPARROW model. For the suspended sediment model, EPA used the same
approach as used for the C&D ELG analysis, which involved adjusting the mass flux attributed to the
urban land explanatory variable in the model to subtract the change in loading achieved under each
option, under the assumption that steam electric power plant loadings are implicitly accounted for in
the urban land component of the model (see U.S. EPA, 2009c).
The model provides annual average post-compliance concentrations in each E2RF1, which EPA compared
with baseline conditions obtained directly from the national, calibrated model.
September 29, 2015 D-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix E: CVD Incidence and Mortality
Appendix E. Details on Modeling of Cardiovascular Disease Incidence
and Mortality
E.1 Benefits to Adults from Reduced Lead Exposure
E.1.1 Hazard Reduction under Final ELGs
For each sex, single-year age, and exposure cohort (hereafter cohort), the analysis characterizes two basic
survival analyses under the baseline and option scenarios: a death hazard function and a survival function. 89
A death hazard function (HF), h(x), for an individual surviving to age x years, is a function such that h(x)dx
gives the probability of death between ages x and x + dx (where dx -> 0). Specifically, h(x)dx =
P(x < Xd < x + dx\ Xd > x), where Xd is person's exact age at death. The death hazard function is also
commonly known as the hazard rate, and less commonly as the force of mortality. At any point in time, an
individual is at risk of death from competing causes, of which cardiovascular disease (CVD) is only one. As
discussed in Beyersmann et al. (2009), competing risks can be modeled using cause-specific death hazard
functions, which are additive (by the law of total probability):
Equation E-1 . h(x) = hCVD(x) + h°TH(x)
A survival function, S(x), is the probability that a person dies at some point after age x, specifically: S(x) =
P(Xd > x). The survival function and the hazard function are intrinsically linked, with the survival function
determined completely by the hazard function:
S(x) = exp{— / h(v)dv}, where v represents a time period typically different than x.
Thus, to estimate changes in mortality such as those resulting from implementation of the final ELGs, it is
necessary to characterize:
(1) baseline CVD death hazard function;
(2) baseline other cause (i.e., non-CVD) death hazard function; and
(3) option-specific CVD death hazard function.
The main source of data for hazard estimation in key simulation elements (1) and (2) above is a life table,
which is a collection of statistics that shows age-specific probabilities of survival and fecundity.90 Employed
heavily in actuarial science, demography, population biology, ecology and epidemiology, data from life tables
can be used to calculate probabilities of survival to a given age, age -specific life expectancy, population
growth rates, and many other demographic characteristics.
The statistics reported in a life table can be used to approximate the (baseline) hazard function. Of a particular
interest are the estimates of the initial age-specific mortality rate, qx: the proportion of people alive at exact
age x, who will die before attaining exact age x + 1. The initial mortality rate is closely related to the hazard
function:
Equation E-2. qx = 1 - S(x + l)/S(x) = 1 - exp {- J*+1 /i(i7)di?}.
89 Collett (2003), pp. 10-12.
90 An extensive discussion of life tables can be found in Shryock et al. (1980) Chapter 15.
September 29, 2015 Internal Draft - Deliberative, Predecisional - Do not Quote, Cite, or Distribute E-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix E: CVD Incidence and Mortality
rX+l r^H~l
Assuming a constant hazard function within each integer age, }x h(v~)dv = jx hxdv = hx, estimates of
the initial age-specific mortality rates qx reported in the life table can be used to obtain an estimate of the
baseline hazard function:
Equation E-3. h% = ln{l - q^}
At each integer age x, this overall hazard function can be further decomposed into CVD death hazard function
and other-cause death hazard function. To this end, annual age-specific CVD death rates (which are reported
by the Center for Disease Control and Prevention (CDC) can be used to approximate baseline qx ' and,
consequently, obtain an estimate of the baseline CVD hazard function:
Equation E-4. hcxVD'b = ln{l - qcxVD'b}
which is key simulation element (1) noted above.
Additionally, the estimate of the baseline other hazards function can be derived as:
Equation E-5. h°TH'b = hbx - hcxvD'b
providing key simulation element (2) from above.
EPA used a concentration-response function from a peer reviewed study, Menke et al. (2006). The study finds
a multivariate adjusted relative hazards of CVD mortality of 1.53 (1.21-1.94) per 3.4-fold increase in PbB,
based on a Cox proportional hazards model using the log of the blood lead level. The corresponding beta
estimate (£>BLL) f°r this Hazard Ratio (HR) is 0.35. Cox proportional hazards models, such as the one
estimated by Menke et al (2006) are designed to estimate a multiplicative relationship between an outcome
(i.e., the CVD death hazard function) and a set of predictors. Menke et al. (2006) assessed the HR in various
subgroups, and noted that subgroup interaction terms were not statistically significant, supporting the use of
the coefficient for the overall population in benefits estimation. A key assumption of this model, the
proportional hazards assumption [met in Menke et al (2006)], is that the HR is constant through time. Based
on this information, EPA derived an estimate for (3BLL and used it to compute HRs for changes of different
magnitudes. EPA calculated HR associated with a known PbB change (due to an ELG regulatory option) as:
BLL°
Equation E-6. HR° = exp]BBLL • In
Hazard functions accounting for reduced CVD mortality following ELG implementation were then calculated
as:
Equation E-7. hcxVD'° = hcxVD'b • HR° (eq. 7)
This equation provides key simulation element (3), above.
E.1.2 Estimating Premature Deaths Avoided Over Multiple Years
The value of a statistical life (VSL) is the marginal rate of substitution between wealth and mortality risk in a
defined time period, usually taken to be one year. Therefore, the product of VSL and the estimated reduction
in risk of premature death represents the affected population's aggregate willingness to pay (WTP) to reduce
its probability of death in one year. EPA estimated the benefits of multi-year mortality risk as the product of:
(1) The reduction in initial age-specific mortality rate (i.e., the proportion of people alive at exact age x,
who will die before attaining exact age x + 1; commonly represented as qx) in year t; and
September 29, 2015 Internal Draft - Deliberative, Predecisional - Do not Quote, Cite, or Distribute E-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix E: CVD Incidence and Mortality
(2) The number of individuals surviving to the beginning of year t. This value is calculated as the initial
cohort population size in 2014 multiplied by the probability that these individuals survive to age x
[commonly represented as 5(x)], and are alive at the beginning of year t to enjoy the benefits of the
year's mortality risk reduction. EPA used the survival probability under each ELG option (i.e., the
survival probability reflecting cumulative reductions in mortality risk prior to year t) in each
recursive step of this calculation.
Each pattern of annual mortality rates corresponds to a unique survival curve. Thus, each intervention-related
change in mortality rates (over multiple years) will generate a unique change in the cohort's survival curve.
Figure E-l illustrates such a change for an individual whose baseline initial mortality rate of 0.3 (i.e., a 30
percent chance of death in one year) was reduced to 0.2 (i.e., a 20 percent chance of death in one year) by an
intervention which begins in year 0 and continues indefinitely. The baseline survival curve is shown as a solid
line, while the dashed line shows the effects of intervention.
Baseline (qx = 0.3)
Option (qx = 0.2)
Figure E-1: Illustration of a Hypothetical Policy Effect on a Survival Curve.
Each intervention-related change in survival curve can be summarized by either the gain in life expectancy
(/'. e., the area between the survival curves) or by the number of avoided premature deaths at each point in time
(Hammitt 2007). The recursive calculation of the number of avoided premature deaths (described above) can
be seen as a series of shifts of the survival curve. Each shift represents the incremental effect of a mortality
risk reduction in year t (while keeping mortality rates in subsequent years at their baseline levels) on the
survival curve. Therefore, these shifts are consistent with the cumulative impact of mortality risk reductions in
years 0 to t — 1. Figure E-2 illustrates two such shifts: (1) the dashed survival curve reflects the impact of
reduced mortality rate in year 0 on the baseline survival curve (i.e., the incremental impact of mortality risk
reductions in year 0); (2) the dotted curve reflects the impact of reducing mortality rate in year 1 on the
dashed survival curve (i.e., the incremental impact of mortality risk reductions in year 1).
As noted earlier, the area between the survival curves represents the gain in life expectancy. Thus, each shift
of the survival curve reflects an incremental gain in life expectancy (e.g., in Figure E-2, the area between
solid and dashed lines represents the incremental gain in life expectancy because of reduced mortality in year
0; the area between dashed and dotted lines represent incremental gain in life expectancy because of reduced
mortality in year 1). By design, the sum of these incremental gains is equal to the original gain in life
expectancy from a multi-year reduction. This confirms that EPA's procedure generates an estimated reduction
September 29, 2015 Internal Draft - Deliberative, Predecisional - Do not Quote, Cite, or Distribute
E-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix E: CVD Incidence and Mortality
in the risk of premature death that is consistent with the estimated aggregate gain in life expectancy from the
final ELGs.
Baseline
— — Impact Year 0
Impact Year 1
20
25
Figure E-2: A Recursive Illustration of a Policy Effect on Survival Rates.
September 29, 2015 Internal Draft - Deliberative, Predecisional - Do not Quote, Cite, or Distribute
E-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix F: Human Health Sensitivity Analyses
Appendix F. Human Health Benefits Sensitivity Analysis
EPA conducted a series of sensitivity analyses to evaluate the impacts of varying assumptions related to a) the
distance travelled by recreational and subsistence anglers, b) the form of the concentration-response function
for lead, and c) the cancer slope function and cancer case valuation approach for the arsenic analysis. These
analyses are summarized below.
F.1 Alternative Fishing Distance Assumption
The set of reaches that may represent a source of contaminated fish for recreational anglers and subsistence
fishers in each CBG depends on the distance the typical angler travels to fish. In the human health benefits
analysis (Chapter 3), EPA assumed that anglers typically travel up to 50 miles to fish, using this distance to
estimate the relevant fishing sites for the population of anglers in each CBG.
Viscusi et al. (2008) found that 78 percent of anglers live within 100 miles of their fishing destinations. EPA
conducted a sensitivity analysis to assess the impact of the travel distance assumptions to the health analysis
results. Using a different travel distance assumption tends to increase the number of alternate fishing sites
visited by anglers within each CBG, but also increases the availability of substitute sites.
Table F-l shows the affected population using this alternative fishing travel distance, and Table F-2 shows
the results for the health benefit categories for which EPA conducted this sensitivity analysis.
Table F-1. Summary of Potentially Affected Population Living within 100
Miles of Affected Reaches (baseline)
Total population
Total angler population3
Angler population potentially exposed to
contaminated fishb
306,740,261
42,714,703
29,592,014
a. Total population living within 100 miles of an affected reach times the state-specific share of
the population who fishes based on U.S. FWS (2011; between 9% and 23%).
b. Total angler population adjusted to reflect lower fishing/consumption rates for reaches with
fish consumption advisories and catch-and-release practices.
Table F-2. Total Monetized Human Health Benefits for ELG Options Using a 100-mile Buffer Zone
(millions of 2013$)a'b
.
Regulatory
Option
Reduced lead
exposure for
children
Low
High
Reduced
lead
exposure for
adults
Reduced
mercury
exposure for
children
Low
High
Reduced
cancer
cases from
arsenic
Total
Low High
3% Discount Rate
Option A
Option B
Option C
Option D
Option E
$0.28
$0.28
$0.43
$0.57
$0.57
$0.40
$0.40
$0.60
$0.80
$0.80
$3.55
$3.55
$6.66
$8.98
$8.98
$1.37
$1.39
$2.36
$2.93
$3.30
$1.92
$1.95
$3.32
$4.11
$4.64
$0.00
$0.00
$0.00
$0.00
$0.00
$5.20
$5.22
$9.45
$12.49
$12.86
$5.87
$5.90
$10.58
$13.90
$14.42
September 29, 2015
F-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix F: Human Health Sensitivity Analyses
Table F-2. Total Monetized Human Health Benefits for ELG Options Using a 100-mile Buffer Zone
(millions of 2013$)a'b
Regulatory
Option
Reduced lead
exposure for
children
Low
High
Reduced
lead
exposure for
adults
Reduced
mercury
exposure for
children
Low
High
Reduced
cancer
cases from
arsenic
Total
Low High
7% Discount Rate
Option A
Option B
Option C
Option D
Option E
$0.05
$0.05
$0.07
$0.10
$0.10
$0.07
$0.07
$0.11
$0.14
$0.14
$2.96
$2.96
$5.56
$7.50
$7.50
$0.22
$0.23
$0.38
$0.48
$0.54
$0.33
$0.33
$0.57
$0.71
$0.80
$0.00
$0.00
$0.00
$0.00
$0.00
$O T3
j.2j
$3.24
$6.01
$8.07
$8.13
$3.36
$3.37
$6.24
$8.35
$8.44
Source: U.S. EPA Analysis, 2015
a. "-" indicates that a value was not estimated and "$0.00" indicates that annual benefits are less than $0.01 million.
b. Low bound assumes that the loss of one IQ point results in the loss of 1.76 percent of lifetime earnings (following Schwartz, 1994);
high bound assumes that the loss of one IQ point results in the loss of 2.38 percent of lifetime earnings (following Salkever, 1995).
og-Linear Concentration Response
Lead Exposure
:unction for IQ Impacts to Children from
In the analysis of benefits to children from reduced lead intake via fish consumption, EPA used a linear
concentration-response function to quantify the relationship between blood lead concentrations (PbB) and
intelligence quotient (IQ). This linear function is based on a concentration-response function based on
children with PbB below 7.5 |ag/dL since the average PbB among affected children is approximately 2.7
Hg/dL (see Section 3.3). EPA received several comments stating that this approach was inappropriate and
resulted in an overestimate of the benefits.
Given the uncertainty surrounding the lead concentration response relationship in children, EPA conducted a
sensitivity analysis applying a log-linear function used by EPA (2008b) in the RIA for the Lead National
Ambient Air Quality Standards (NAAQS). This function uses a log-linear function for PbB above 1.47 ug/dL
and a linear slope for PbB levels below that cut-point, as shown in Equation F-8 and Equation F-9,
respectively.
Equation F-8. IQ Loss = ft x ln(-
PbB
+ ft x outpoint
\cutpointj
Equation F-9. IQ Loss = ft x outpoint
Where:
outpoint = 1.47 |ag/dL
ft = -3.04 (log-linear regression coefficient)
ft = -2.1 (linear regression coefficient)
Table F-3 shows the results of the analysis of avoided IQ point losses among children from exposure to lead
using this alternative concentration-response function (corresponding to Table 3-4 in Section 3.3).
September 29, 2015
F-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix F: Human Health Sensitivity Analyses
Table F-3. Estimated Benefits from Avoided IQ Losses for Children Exposed to Lead using Log-
Linear Concentration Response Function (2013$)
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Average Annual
Number of
Affected
Children 0 to 7
3,326,127
3,326,127
3,326,127
3,326,127
3,326,127
Total Avoided
IQ Losses,
2021 to 2042
324
324
486
816
816
Annualized Value of Avoided IQ Point Losses" (Millions)
3% Discount Rate
Low Bound
$0.13
$0.13
$0.19
$0.32
$0.32
High Bound
$0.18
$0.18
$0.27
$0.46
$0.46
7% Discount Rate
Low Bound
$0.02
$0.02
$0.03
$0.05
$0.05
High Bound
$0.03
$0.03
$0.05
$0.08
$0.08
Source: U.S. EPA Analysis, 2015
a. Low bound assumes that the loss of one IQ point results in the loss of 1.76% of lifetime earnings (following Schwartz, 1994);
high bound assumes that the loss of one IQ point results in the loss of 2.38% of lifetime earnings (following Salkever, 1995).
F.3 Alternative Cancer Slope Factor and Case Valuation for Arsenic Analysis
The Integrated Risk Information System (IRIS) reports a cancer slope factor (CSF) of 1.5 cases per mg/kg
BW/day, which is based on incidences of skin cancer. EPA applied the 1.5 cases per mg/kg BW/day CSF to
estimate the benefits shown in Section 3.6. 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 oral
ingestion. The draft CSF is substantially higher - at 25.7 per mg/kg BW/day for women and 16.9 per mg/kg
BW/day for men (U.S. EPA, 2010b).
EPA conducted a sensitivity analysis using the more sensitive CSF, together with monetizing the avoided
cases using the value of a statistical life (VSL; $8.548 million), reflective of the higher mortality rates
associated with internal cancers. Table F-4 shows the results of this sensitivity analysis (corresponding to
Table 3-12 in Section 3.6).
Table F-4. Annual Benefits from Reduced Cancer Cases due to Arsenic Exposure, using Alternative
CSF and VSL
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Annual Affected
Population
35,972,005
35,972,005
35,972,005
35,972,005
35,972,005
Reduced Cancer Cases,
2019 to 2042
0.55
0.55
1.05
1.59
1.72
Benefits (Millions; 2013$)
3% Discount
$0.13
$0.13
$0.25
$0.38
$0.41
7% Discount
$0.07
$0.07
$0.14
$0.21
$0.23
Source: U.S. EPA Analysis, 2015
September 29, 2015
F-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Appendix G. WQI Regional Subindices
This appendix provides the ecoregion-specific parameters used in estimating the TSS, TN, or TP water
quality subindex, as follows:
> If [WQ Parameter] < WQ Parameter 100
> If WQ Parameter 100 < [WQ Parameter] < WQ Parameter 10
> If [WQ Parameter] > WQ Parameter 10
Subindex = 100
Subindex = a exp(b [WQ Parameter])
Subindex =10
> Where [WQ Parameter] is the measured concentration of either TSS, TN, or TP and WQ Parameter 10,
WQ Parameter 100, a, and b are specified in Table G-l for TSS, Table G-2 for TN, and Table G-3 for TP.
Table G-1: TSS Subindex Curve Parameters, by Ecoregion
ID
10.1.2
10.1.3
10.1.4
10.1.5
10.1.6
10.1.7
10.1.8
10.2.1
10.2.2
10.2.4
11.1.1
11.1.2
11.1.3
12.1.1
13.1.1
15.4.1
5.2.1
5.2.2
5.3.1
5.3.3
6.2.10
6.2.11
6.2.12
6.2.13
6.2.14
6.2.15
6.2.3
62A
623
62:1
6.2.8
6.2.9
7.1.7
7.1.8
7.1.9
Ecoregion Name
Columbia Plateau
Northern Basin and Range
Wyoming Basin
Central Basin and Range
Colorado Plateaus
Arizona/New Mexico Plateau
Snake River Plain
Mojave Basin and Range
Sonoran Desert
Chihuahuan Desert
California Coastal Sage, Chaparral, and Oak Woodlands
Central California Valley
Southern and Baja California Pine-Oak Mountains
Madrean Archipelago
Arizona/New Mexico Mountains
Southern Florida Coastal Plain
Northern Lakes and Forests
Northern Minnesota Wetlands
Northern Appalachian and Atlantic Maritime Highlands
North Central Appalachians
Middle Rockies
Klamath Mountains
Sierra Nevada
Wasatch and Uinta Mountains
Southern Rockies
Idaho Batholith
Columbia Mountains/Northern Rockies
Canadian Rockies
North Cascades
Cascades
Eastern Cascades Slopes and Foothills
Blue Mountains
Strait of Georgia/Puget Lowland
Coast Range
Willamette Valley
a
126.56
112.42
123.36
121.22
144.44
126.76
146.39
119.34
112.39
214.39
127.97
171.86
115.12
261.35
120.98
116.95
157.76
154.99
174.99
245.15
144.64
238.9
185.36
124.28
153.42
184.23
180.7
3%!62
24O95
192794
178.82
148.35
181.06
174.78
210.3
b
-0.0038
-0.0007
-0.001
-0.0018
-0.001
-0.0004
-0.0027
-0.0015
-0.0002
-0.0005
-0.0012
-0.0044
-0.0007
-0.0005
-0.0004
-0.0405
-0.0233
-0.0186
-0.0261
-0.0176
-0.0038
-0.0068
-0.0116
-0.0014
-0.0031
-0.0142
-0.0168
:oo308
:oo"f93
:oo"f§Y
-0.0145
-0.0037
-0.0224
-0.0114
-0.0114
TSSioo
63
160
220
109
363
668
142
121
567
1,419
205
122
197
2,053
477
4
20
24
21
51
98
129
53
160
140
43
35
45
46
36
40
107
27
49
65
TSS10
668
3,457
2,513
1,386
2,670
6,349
994
1,653
12,097
6,130
2,124
646
3,491
6,527
6,233
61
118
147
110
182
703
467
252
1,800
881
205
172
119
165
164
199
729
129
251
267
September 29, 2015
G-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Table G-1: TSS Subindex Curve Parameters, by Ecoregion
ID
8.1.1
8.1.2
8.1.3
8.1:4
8.1.5
8.1.6
8.1.7
8.1.8
8.1.10
8.2.1
8.2.2
8.2.3
8.2.4
8.3.1
8.3.2
8.3.3
8.3.4
8.3.5
8.3.6
8.3.7
8.3.8
8.4.1
8.4.2
8.4.3
8.4.4
8.4.5
8.4.6
8.4.7
8.4.8
8.4.9
8.5.1
8.5.2
8.5.3
8.5.4
9.2.1
9.2.2
9.2.3
9.2.4
9.3.1
9.3.3
9.3.4
9.4.1
9.4.2
9.4.3
9AA
9A5
9A6
9.4.7
Ecoregion Name
Eastern Great Lakes and Hudson Lowlands
Lake Erie Lowland
Northern Appalachian Plateau and Uplands
North Central Hardwood Forests
Driftless Area
S. Michigan/N. Indiana Drift Plains
Northeastern Coastal Zone
Maine/New Brunswick Plains and Hills
Erie Drift Plain
Southeastern Wisconsin Till Plains
Huron/Erie Lake Plains
Central Corn Belt Plains
Eastern Corn Belt Plains
Northern Piedmont
Interior River Valleys and Hills
Interior Plateau
Piedmont
Southeastern Plains
Mississippi Valley Loess Plains
South Central Plains
East Central Texas Plains
Ridge and Valley
Central Appalachians
Western Allegheny Plateau
Blue Ridge
Ozark Highlands
Boston Mountains
Arkansas Valley
Ouachita Mountains
Southwestern Appalachians
Middle Atlantic Coastal Plain
Mississippi Alluvial Plain
Southern Coastal Plain
Atlantic Coastal Pine Barrens
Aspen Parkland/Northern Glaciated Plains
Lake Manitoba and Lake Agassiz Plain
Western Corn Belt Plains
Central Irregular Plains
Northwestern Glaciated Plains
Northwestern Great Plains
Nebraska Sand Hills
High Plains
Central Great Plains
Southwestern Tablelands
_____
Cross Timbers
Edwards Plateau
Texas Blackland Prairies
a
144.62
112.79
322.68
148.68
117.97
191.44
158.48
156.02
133.08
121.34
145.17
187.95
235.18
175.82
149.68
220.47
224.11
205.3
492.49
184.36
162.32
186.83
166.76
183.67
216.16
175.16
329.77
283.25
212.77
207.09
182.17
131.35
138.62
283.76
136.43
174.13
135.01
201.19
133.98
130.6
289.85
125.61
156.84
137.77
27O93
L3T97
173/77
134.23
b
-0.0104
-0.0049
-0.0113
-0.0108
-0.0012
-0.0143
-0.0164
-0.025
-0.0037
-0.0042
-0.0058
-0.0033
-0.003
-0.0042
-0.0013
-0.0037
-0.0048
-0.0085
-0.0048
-0.0045
-0.0013
-0.0063
-0.0062
-0.0032
-0.0087
-0.0018
-0.0062
-0.004
-0.0048
-0.0071
-0.0178
-0.0029
-0.0144
-0.0463
-0.0005
-0.0042
-0.0009
-0.001
-0.0006
-0.0004
-0.0066
-0.0005
-0.0005
-0.0003
:0;"ooo9
:0;"ooo6
-oTooi
-0.0005
TSSioo
36
25
103
37
141
46
28
18
78
46
65
191
282
135
303
217
169
85
333
136
362
99
82
190
89
317
193
261
157
103
34
93
23
23
640
131
347
673
483
636
162
507
925
1,280
U>84
523
544
624
TSS10
257
494
307
250
2,057
206
168
110
700
594
461
889
1,053
683
2,081
836
648
356
812
648
2,144
465
454
910
353
1,591
564
836
637
427
163
888
183
72
5,226
680
2,892
3,002
4,325
6,424
510
5,061
5,505
8,743
3^666
4/337
27855
5,194
September 29, 2015
G-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Table G-1: TSS Subindex Curve Parameters, by Ecoregion
ID
9.5.1
9.6.1
Ecoregion Name
Western Gulf Coastal Plain
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
a
124.47
166.67
b
-0.0025
-0.0003
TSSioo
88
1,602
TSS10
1,009
9,378
Table G-2: TN Subindex Curve Parameters, by Ecoregion
ID
10.1.2
10.1.3
10.1.4
10.1.5
10.1.6
10.1.7
io".T.8
I0".2"."l
10"."2.2
10.2.4
11.1.1
11.1.2
11.1.3
12.1.1
13.1.1
15.4.1
5.2.1
5.2.2
5.3.1
5.3.3
6.2.10
6.2.11
6.2.12
6.2.13
6.2.14
6.2.15
6.2.3
6.2.4
6.2.5
6.2.7
6.2.8
6.2.9
7. .7
7. .8
7. .9
8. .1
8. .2
8. .3
8. .4
8. .5
8. .6
Ecoregion Name
Columbia Plateau
Northern Basin and Range
Wyoming Basin
Central Basin and Range
Colorado Plateaus
Arizona/New Mexico Plateau
Snake River Plain
Mojave Basin and Range
Sonoran Desert
Chihuahuan Desert
California Coastal Sage, Chaparral, and Oak Woodlands
Central California Valley
Southern and Baja California Pine-Oak Mountains
Madrean Archipelago
Arizona/New Mexico Mountains
Southern Florida Coastal Plain
Northern Lakes and Forests
Northern Minnesota Wetlands
Northern Appalachian and Atlantic Maritime Highlands
North Central Appalachians
Middle Rockies
Klamath Mountains
Sierra Nevada
Wasatch and Uinta Mountains
Southern Rockies
Idaho Batholith
Columbia Mountains/Northern Rockies
Canadian Rockies
North Cascades
Cascades
Eastern Cascades Slopes and Foothills
Blue Mountains
Strait of Georgia/Puget Lowland
Coast Range
Willamette Valley
Eastern Great Lakes and Hudson Lowlands
Lake Erie Lowland
Northern Appalachian Plateau and Uplands
North Central Hardwood Forests
Driftless Area
S. Michigan/N. Indiana Drift Plains
a
116.58
126.97
124.89
116.66
146.41
116.33
129.93
136769
117799
104.2
123.22
126.07
122.76
130.61
141.64
1000000
141.98
142.55
142.6
180.92
136.51
140.34
143.02
129.75
131.07
149.42
136.14
151.95
155.86
143.07
123.99
125.19
121.09
136.15
135.01
158.18
156.27
431.78
163.4
126.18
130.25
b
-0.663
-0.626
-0.445
-0.335
-0.588
-0.286
-0.594
-b"593
-bT495
-0.45
-0.889
-0.548
-0.564
-0.325
-0.541
-29.36
-0.985
-0.781
-0.854
-0.897
-0.991
-1.805
-1.424
-0.452
-0.66
-1.775
-1.599
-2.098
-1.231
-1.473
-1.07
-0.786
-0.723
-1.021
-0.809
-0.563
-0.38
-0.435
-0.599
-0.272
-0.149
TN100
0.23
0.38
0.50
0.46
0.65
0.53
0.44
6753
6733
0.09
0.23
0.42
0.36
0.82
0.64
0.33
0.36
0.45
0.42
0.66
0.31
0.19
0.25
0.58
0.41
0.23
0.19
0.20
0.36
0.24
0.20
0.29
0.26
0.30
0.37
0.81
1.18
3.36
0.82
0.85
1.78
TN10
3.70
4.06
5.67
7.33
4.56
8.58
4.32
4741
4799
5.21
2.82
4.62
4.45
7.91
4.90
0.39
2.69
3.40
3.11
3.23
2.64
1.46
1.87
5.67
3.90
1.52
1.63
1.30
2.23
1.81
2.35
3.22
3.45
2.56
3.22
4.90
7.23
8.66
4.66
9.32
17.23
September 29, 2015
G-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Table G-2: TN Subindex Curve Parameters, by Ecoregion
ID
8.1.7
8.1.8
8.1.10
8.2.1
8.2.2
8.2.3
8.2.4
8.3.1
8.3.2
8.3.3
8.3.4
sTis
873.6'
8.3.7
8.3.8
8A1
8A2
8.4.3
8.4.4
8.4.5
8.4.6
8.4.7
8.4.8
8.4.9
8.5.1
8.5.2
8.5.3
8.5.4
9.2.1
9.2.2
9.2.3
9.2.4
9.3.1
9.3.3
9.3.4
9.4.1
9.4.2
9.4.3
9.4.4
9.4.5
9.4.6
9.4.7
9.5.1
9.6.1
Ecoregion Name
Northeastern Coastal Zone
Maine/New Brunswick Plains and Hills
Erie Drift Plain
Southeastern Wisconsin Till Plains
Huron/Erie Lake Plains
Central Corn Belt Plains
Eastern Corn Belt Plains
Northern Piedmont
Interior River Valleys and Hills
Interior Plateau
Piedmont
Southeastern Plains
Mississippi Valley Loess Plains
South Central Plains
East Central Texas Plains
Ridge and Valley
Central Appalachians
Western Allegheny Plateau
Blue Ridge
Ozark Highlands
Boston Mountains
Arkansas Valley
Ouachita Mountains
Southwestern Appalachians
Middle Atlantic Coastal Plain
Mississippi Alluvial Plain
Southern Coastal Plain
Atlantic Coastal Pine Barrens
Aspen Parkland/Northern Glaciated Plains
Lake Manitoba and Lake Agassiz Plain
Western Corn Belt Plains
Central Irregular Plains
Northwestern Glaciated Plains
Northwestern Great Plains
Nebraska Sand Hills
High Plains
Central Great Plains
Southwestern Tablelands
Flint Hills
Cross Timbers
Edwards Plateau
Texas Blackland Prairies
Western Gulf Coastal Plain
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
a
125.75
139.55
148.99
134.85
119.06
135.57
149.12
146.34
120.48
146.39
148.67
138.73
123.15
149.84
136
158.11
161.22
125.25
158.16
145.69
168.59
135.4
162.34
143.42
123.43
119.57
118.73
110.04
141.62
119.49
129.28
142.81
120.91
125.65
113.81
121.41
129.36
136.03
142.74
130.87
141.98
133.84
106.22
102.35
b
-0.159
-0.553
-1.256
III:°jZ
-oTo9i
-0.087
-0.122
767314
-o'TJi
-0.446
-0.637
IIiO-Z^L
-0.379
-0.706
-0.344
-0.659
-0.907
-0.44
-0.777
-0.513
-1.108
-0.47
-0.942
-0.645
-0.444
-0.31
-0.701
-0.482
-0.086
-0.082
-0.074
-0.184
-0.386
-0.404
-0.324
-0.161
-0.178
-0.413
-0.343
-0.278
-0.588
-b"243
-0.301
-0.374
TN100
1.44
0.60
0.32
IIIl-§L
L91
3.50
3.28
L21
L43
0.85
0.62
III^Il.
0.55
0.57
0.89
0.70
0.53
6751
0.59
0.73
0.47
0.64
0.51
0.56
0.47
0.58
0.24
0.20
4.06
2.18
3.48
1.93
0.49
0.56
0.40
1.21
1.44
0.74
1.04
0.97
0.60
L20
0.20
0.06
TN10
15.92
4.77
2.15
IIj^leT
27722
29.96
22.15
8755
isTo'o
6.02
4.24
IIII-62~
6.62
3.83
7.59
4.19
3.07
5774
3.55
5.22
2.55
5.54
2.96
4.13
5.66
8.00
3.53
4.98
30.82
30.25
34.59
14.45
6.46
6.26
7.51
15.51
14.38
6.32
7.75
9.25
4.51
1O68
7.85
6.22
September 29, 2015
G-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Table G-3: TP Subindex Curve Parameters, by Ecoregion
ID
10.1.2
10.1.3
10.1.4
10.i-;5
10.1.6
10.1.7
10.1.8
10.2.1
10.2.2
10.2.4
11.1.1
11.1.2
11.1.3
12.1.1
13.1.1
15A1
s'H
"522
5.3.1
5.3.3
6.2.10
6.2.11
6.2.12
6.2.13
6.2.14
6.2.15
6.2.3
6.2.4
6.2.5
6.2.7
6.2.8
6.2.9
7. .7
7. .8
7. .9
8. .1
8. .2
8. .3
8. .4
8. .5
8. .6
8. .7
8. .8
8.1.10
8.2.1
8.2.2
8.2.3
8.2.4
Ecoregion Name
Columbia Plateau
Northern Basin and Range
Wyoming Basin
Central Basin and Range
Colorado Plateaus
Arizona/New Mexico Plateau
Snake River Plain
Mojave Basin and Range
Sonoran Desert
Chihuahuan Desert
California Coastal Sage, Chaparral, and Oak Woodlands
Central California Valley
Southern and Baja California Pine-Oak Mountains
Madrean Archipelago
Arizona/New Mexico Mountains
Southern Florida Coastal Plain
Northern Lakes and Forests
Northern Minnesota Wetlands
Northern Appalachian and Atlantic Maritime Highlands
North Central Appalachians
Middle Rockies
Klamath Mountains
Sierra Nevada
Wasatch and Uinta Mountains
Southern Rockies
Idaho Batholith
Columbia Mountains/Northern Rockies
Canadian Rockies
North Cascades
Cascades
Eastern Cascades Slopes and Foothills
Blue Mountains
Strait of Georgia/Puget Lowland
Coast Range
Willamette Valley
Eastern Great Lakes and Hudson Lowlands
Lake Erie Lowland
Northern Appalachian Plateau and Uplands
North Central Hardwood Forests
Driftless Area
S. Michigan/N. Indiana Drift Plains
Northeastern Coastal Zone
Maine/New Brunswick Plains and Hills
Erie Drift Plain
Southeastern Wisconsin Till Plains
Huron/Erie Lake Plains
Central Corn Belt Plains
Eastern Corn Belt Plains
a
147.39
165.9
143.83
167.24
123.74
168.68
140.75
139.89
122.92
132.89
125.05
126.32
212.01
140.62
555.88
15779
152/78
17L4
260.92
157.84
188.95
205.2
142.56
141.72
185.94
168.85
197.1
289.57
227.85
154.67
141.59
165.33
185.34
159.54
148.02
230.09
3440.2
317.21
132.65
141.49
184.34
174
174.73
151.79
141.21
247.17
223.41
196
b
-2.211
-2.78
-1.57
-2.541
-0.784
-3.39
-1.106
-0.978
-1.578
-3.737
-1.918
-2.138
-0.941
-1.331
-306
-26"64
-1617
-2L87
-21.53
-6.439
-15.04
-19.13
-2.752
-5.463
-21.89
-17.88
-27.87
-47.06
-26.77
-10.55
-3.31
-13.83
-14.77
-9.053
-7.95
-9.614
-8.887
-13.87
-4.905
-2.261
-5.59
-9.944
-28.94
-3.59
-1.577
-2.666
-3.555
-3.734
TPioo
0.18
0.18
0.23
0.20
0.27
0.15
0.31
0.34
0.13
0.08
0.12
0.11
0.80
0.26
0.01
67o2
6""6"3
6702
0.04
0.07
0.04
0.04
0.13
0.06
0.03
0.03
0.02
0.02
0.03
0.04
0.11
0.04
0.04
0.05
0.05
0.09
0.40
0.08
0.06
0.15
0.11
0.06
0.02
0.12
0.22
0.34
0.23
0.18
TP10
1.22
1.01
1.70
1.11
3.21
0.83
2.39
2.70
1.59
0.69
1.32
1.19
3.25
1.99
0.01
olb
O"l7
0/L3
0.15
0.43
0.20
0.16
0.97
0.49
0.13
0.16
0.11
0.07
0.12
0.26
0.80
0.20
0.20
0.31
0.34
0.33
0.66
0.25
0.53
1.17
0.52
0.29
0.10
0.76
1.68
1.20
0.87
0.80
September 29, 2015
G-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix G: Subindex Curves
Table G-3: TP Subindex Curve Parameters, by Ecoregion
ID
8.3.1
8.3.2
8.3.3
8.3.4
8.3.5
8.3.6
8O *1
.3.7
8.3.8
8.4.1
8.4.2
8.4.3
8.4.4
8.4.5
8.4.6
8.4.7
8.4.8
8.4.9
8.5.1
8.5.2
8.5.3
8.5.4
9.2.1
9.2.2
9.2.3
9.2.4
9.3.1
9.3.3
9.3.4
9.4.1
9.4.2
9.4.3
9.4.4
9.4.5
9.4.6
9.4.7
9.5.1
9.6.1
Ecoregion Name
Northern Piedmont
Interior River Valleys and Hills
Interior Plateau
Piedmont
Southeastern Plains
Mississippi Valley Loess Plains
South Central Plains
East Central Texas Plains
Ridge and Valley
Central Appalachians
Western Allegheny Plateau
Blue Ridge
Ozark Highlands
Boston Mountains
Arkansas Valley
Ouachita Mountains
Southwestern Appalachians
Middle Atlantic Coastal Plain
Mississippi Alluvial Plain
Southern Coastal Plain
Atlantic Coastal Pine Barrens
Aspen Parkland/Northern Glaciated Plains
Lake Manitoba and Lake Agassiz Plain
Western Corn Belt Plains
Central Irregular Plains
Northwestern Glaciated Plains
Northwestern Great Plains
Nebraska Sand Hills
High Plains
Central Great Plains
Southwestern Tablelands
Flint Hills
Cross Timbers
Edwards Plateau
Texas Blackland Prairies
Western Gulf Coastal Plain
Southern Texas Plains/Interior Plains and Hills with
Xerophytic Shrub and Oak Forest
a
160.97
156.71
197.72
223.39
177.2
168
166.39
178.13
225.67
187.73
174.12
152.7
204.88
287.21
158.46
169.72
153.95
141.25
144.74
126.84
156.13
132.19
197.19
200.96
134.12
143.32
185
153.07
188.57
139.57
218.92
131.7
159.98
149.63
127.17
104.21
147.39
b
-2.567
-3.616
-5.623
-9.266
-5.69
-4.659
-1.677
-6.407
-16.59
-8.367
-10.5
-2.889
-7.364
-5.786
-6.821
-7.296
-6.816
-3.807
-7.676
-8.388
-0.69
- .087
- .683
- .994
- .646
- .267
-3.788
-0.946
-1.178
-0.972
-2.351
-0.78
-1.384
-1.064
-1.863
-0.513
-2.211
TPioo
0.19
0.12
0.12
0.09
0.10
0.11
0.30
0.09
0.05
0.08
0.05
0.15
0.10
0.18
0.07
0.07
0.06
0.09
0.05
0.03
0.65
0.26
0.40
0.35
0.18
0.28
0.16
0.45
0.54
0.34
0.33
0.35
0.34
0.38
0.13
0.08
0.18
TP10
1.08
0.76
0.53
0.34
0.51
0.61
1.68
0.45
0.19
0.35
0.27
0.94
0.41
0.58
0.41
0.39
0.40
0.70
0.35
0.30
3.98
2.38
1.77
1.50
1.58
2.10
0.77
2.88
2.49
2.71
1.31
3.31
2.00
2.54
1.36
4.57
1.22
September 29, 2015
G-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Appendix H. Development of Meta-Regression Models of Willingness to
Pay for Water Quality Improvements
This meta-regression is a revised version of the 2009 meta-regression used in the benefit analysis of the
proposed ELGs to estimate the water quality improvement benefits of the proposed rule (U.S. EPA, 2013a).
EPA made a number of improvements to the meta-regression model, including updating the set of studies and
introducing GIS explanatory variables into the analysis. In particular, the revised meta-model satisfies the
adding-up condition, a theoretically desirable property.91 This condition ensures that if the model were used to
estimate willingness to pay (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. EPA used the revised meta-analysis to estimate the
sum of use and non-use values for water quality improvements resulting from the final ELGs.
The following sections describe EPA's literature review to identify additional studies, meta-data development
and coding, model specification, regression results, and limitations and uncertainties associated with the
model.
"
.1 Literature review to identify additional studies
EPA used the 2009 meta-data (U.S. EPA, 2013a) as the starting point for the current revisions and conducted
a literature review for additional studies to include in the metadata. The Agency followed Stanley et al.
(2013)'s guidelines for meta-regression analyses in economics in documenting the literature search including
(a) the exact databases and other sources searched, (b) the precise combination of keywords, and (c) date
completed. To identify new studies, EPA relied on the following:
> Searches of general literature databases and search engines (EBSCO, Google Scholar, Google).
Search keywords were selected to be sufficiently broad so not to miss relevant studies:
- First terms: (1) water quality, (2) water clarity, (3) nutrient removal/improvement, (4) water
quality index/ladder, (5) clean water, (6) water pollution reduction, (7) water habitat
improvement, and (8) stream flow.
- Second terms: (1) willingness to pay, (2) stated preference, (3) contingent valuation, (4)
choice experiments, (5) contingent activity, and (6) conjoint analysis
> Searches of online reference and abstract databases (Environmental Valuation Resource Inventory
(EVRI), Benefits Use Valuation Database (BUVD), AgEcon Search, RePEc/IDEAs, and the Oregon
State University College of Forestry Recreation Use Values Database);
> Visits to webpages of authors and university programs known to publish stated preference studies
and/or water quality valuation research;92
92
If WQIO < WQI1 < WQI2, then for a WTP function WTP (WQIO, WQI2, YO) to satisfy the adding-up property,
it must meet the condition that WTP(WQIO, WQI1, YO) + WTP(WQI1, WQI2 , YO - WTP(WQIO, WQI1, YO))
= WTP( WQIO, WQI2 , YO) for all possible values of baseline water quality (WQIO), potential future water
quality levels (WQI1 and WQI2), and baseline income (YO).
This included G. Poe (Cornell), J. Bergstrom (University of Georgia), T. Haab (Ohio State), K. Viscusi
(Vanderbilt), R. Carson (U.C. San Diego), W. H. Desvousges, J. Whitehead (Appalachian State), K. Boyle
(Virginia Tech), R. Rosenberger (Oregon State), J. Loomis (Colorado State), J. Corrigan (Kenyon College), and
R.H. von Haefen (North Carolina State).
September 29, 2015
H-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
> Searches of web sites for Resources for the Future and EPA's National Center for Environmental
Economics, both known to conduct environmental and resource economics valuation research;
> Searches of key resource economics journals for 2005 to 2013 (Land Economics, Environmental and
Resource Economics, Marine Resource Economics, Journal of Environmental Economics and
Management, Water Resources Research, and Ecological Economics). EPA focused these journal-
specific searches on the recent period of 2005 to 2013 given the likelihood that older studies were
captured under the steps above or captured during literature reviews conducted for prior versions of
the MRM; and
> Review of bibliographies of other valuation meta-analyses (Van Houtven et al. 2007; Ge et al. 2013;
Brander and Brouwer 2011).
Studies identified during the search were screened according to the following criteria prior to inclusion in the
metadata to ensure validity, consistency, and applicability. The identification and review of studies was
completed and verified by multiple individuals following recommendations of Stanley et al. (2013):
> Commodity consistency - Study must value water quality changes affecting ecosystem services
provided by waterbodies, including recreational activities (such as fishing, boating, and swimming),
aquatic life support and other nonuse values.93 EPA did not include studies that estimate WTP for
improvements in surface waters used primarily for drinking water.
> Welfare consistency - The study must use general accepted stated preference approaches and report
theoretically comparable Hicksian welfare measures (Boyle et al. 2013).
> Amenity detail - As described by Johnston et al. (2005), "the study must provide sufficient
information regarding resource, context, and study attributes to warrant inclusion in the meta-data"
(p.223).
> Study location - The study must be conducted in the U.S.
> Research methods - The study must apply research methods that are supported by the literature and
that provide WTP estimates consistent with neoclassical welfare theory.
> Duplicative studies - Some studies may be released in multiple forms, such as working papers,
conference papers, journal articles, and book chapters. We included only the latest version of the
study (e.g., journal article) when there are multiple versions. Studies are screened to remove
duplicative analysis and analysis that were subsequently revised and published. Multiple values can
be included for a study if they address a different component of the study data or estimate values for
alternative resource changes.
Table H-l summarizes studies in the revised metadata after identifying and screening additional studies. In
total, the revised metadata includes 51 stated preference studies that estimated total WTP (use plus nonuse)
per household for water quality changes in U.S. waterbodies. The studies address various waterbody types
including, rivers, lakes, salt ponds/marshes, and estuaries. The fifteen studies added to the metadata during the
current revisions are shaded in Table H-l. Two of the new papers (Corrigan et al. 2009; Whitehead 2006)
replaced unpublished versions of the same studies (Azevedo et al. 2001; Whitehead et al. 2002). The revised
metadata excludes two studies by Viscusi et al. (2008) and Carson and Mitchell (1993). Carson and Mitchell
(1993) estimate WTP for water quality improvements nation-wide. It is excluded because its national scale
93 For example, a study that estimates WTP for recreational fishing improvements due to fish stocking would not
be included in the meta-data. However, a study that estimates WTP for recreational fishing improvements due
to nutrient reductions would be included in the meta-data, presuming it satisfies all other criteria.
September 29, 2015 R
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
means that the size of the affected resource area is substantially different from other studies in the meta-data,
which at most, assess improvements within a region. On the other hand, Viscusi et al. (2008) asked
respondents in a national survey to value water quality improvements within 100 miles of their home. To be
included in the metadata, a WTP observation must correspond to a specific affected resource or set of
resources that can be delineated in GIS. Viscusi et al. (2008) does not fit into EPA's data coding framework
because the affected resource varies across the sample, with each respondent asked to think about their own
area. The WTP estimates reported by Viscusi et al. (2008) reflect an average across many affected areas,
rather than WTP for a specific affected resource. Many studies contribute multiple observations due to in-
study variations in factors such as number or type affected waterbodies, magnitude of water quality
improvement, sampled market area, and elicitation methods. The inclusion of multiple observations per study
is standard in resource valuation metadata (Nelson and Kennedy 2009). In total, the 51 studies provide 140
WTP observations.
Table H-1: Primary Studies in the Metadata*3'0'0
Author(s) and
Publication Year
Aiken(1985)
Anderson and
Edwards (1986)
Banzhafetal. (2006)
Banzhafetal. (2011)
Bockstaeletal(1988)
Bockstael et al. (1989)
Borisova et al. (2008)
Cameron and Huppert
(1989)
Carson etal. (1994)
Clonts and Malone
(1990)
Collins and
Rosenberger (2007)
Collins et al. (2009)
Corrigan et al. (2009)
Croke etal. (1987)
De Zoysa (1995)
Desvousges et al.
(1987)
Downstream
Strategies (2008)
Farber and Griner
(2000)
Hayes etal. (1992)
Herriges and Shogren
(1996)
Hite (2002)
Huang etal. (1997)
Obs. in
Metadata
1
1
2
1
1
2
3
1
2
3
1
7
1
9
1
12
2
6
2
2
2
2
State(s)
CO
RI
NY
VA, WV,
TN, NC, GA
DC, MD, VA
MD
WV, VA
CA
CA
AL
WV
WV
IA
IL
OH
PA
PA
PA
RI
IA
MS
NC
Water Body
Type(s)
river and lake
salt pond/marshes
Lake
river/stream
Estuary
Estuary
river/stream
Estuary
estuary
river/stream
river/stream
river/stream
Lake
river/stream
river/stream
river/stream
river/stream
river/stream
estuary
Lake
river/stream
estuary
Willingness to Pay per Household
(2007$)
Mean
193.18
180.71
57.47
31.30
149.03
158.30
44.94
49.53
59.40
103.20
18.19
120.52
123.30
77.47
70.18
59.19
12.74
76.16
397.44
134.55
60.08
258.65
Min.
193.18
180.71
54.09
31.30
149.03
75.67
18.05
49.53
41.21
78.31
18.19
2.84
123.30
61.31
70.18
19.84
10.70
16.58
390.68
61.71
58.24
255.01
Max.
193.18
180.71
60.85
31.30
149.03
240.93
65.82
49.53
77.59
127.48
18.19
217.57
123.30
93.68
70.18
137.26
14.77
148.59
404.19
207.40
61.93
262.29
September 29, 2015
H-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Table H-1: Primary Studies in the Metadata*3'0'0
Author(s) and
Publication Year
Irvin et al. (2007)
Johnston etal. (1999)
Kaoru(1993)
Lant and Roberts
(1990)
Lant and Tobin( 1989)
Lichtkoppler and
Elaine (1999)
Lindsey (1994)
Lipton (2004)
Londono Cadavid and
Ando (2013)
Loomis (1996)
Lyke (1993)
Matthews etal. (1999)
Opaluch etal. (1998)
Roberts and Leitch
(1997)
Rowe et al. (1985)
Sanders etal. (1990)
Schulze etal. (1995)
Shrestha and
Alavalapati (2004)
Stumborg etal. (2001)
Sutherland and Walsh
(1985)
Takatsuka (2004)
Wattage (1993)
Welle (1986)
Welle and Hodgson
(2011)
Wey (1990)
Whitehead and
Groothuis (1992)
Whitehead (2006)
Obs. in
Metadata
4
1
1
3
9
1
8
1
2
1
2
2
1
1
1
4
2
2
2
1
4
3
6
3
2
3
3
State(s)
OH
RI
MA
IA, IL
IA, IL
OH
MD
MD
IL
WA
WI
MN
NY
MN, SD
CO
CO
MT
FL
WI
MT
TN
IA
MN
MN
RI
NC
NC
Water Body
Type(s)
all freshwater
river/stream
salt pond/marshes
river/stream
river/stream
river and lake
estuary
estuary
river/stream
river/stream
river and lake
river/stream
estuary
Lake
river/stream
river/stream
river/stream
river and lake
Lake
river and lake
river/stream
river/stream
Lake
Lake
salt pond/marshes
river/stream
river/stream
Willingness to Pay per Household
(2007$)
Mean
21.67
180.95
218.61
143.93
55.63
41.93
66.80
63.98
38.68
93.07
78.75
21.73
138.47
8.35
134.59
160.69
20.84
156.46
84.29
146.03
286.88
53.89
167.28
145.10
147.26
41.01
187.18
Min.
19.65
180.95
218.61
124.04
40.58
41.93
33.40
63.98
35.93
93.07
59.75
18.14
138.47
8.35
134.59
81.01
17.34
137.97
66.73
146.03
181.90
40.24
109.60
10.59
63.95
31.90
27.52
Max.
23.23
180.95
218.61
154.31
67.64
41.93
102.20
63.98
41.44
93.07
97.74
25.32
138.47
8.35
134.59
210.04
24.33
174.95
101.86
146.03
391.85
74.59
238.42
285.06
230.58
53.16
365.54
September 29, 2015
H-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Table H-1: Primary Studies in the Metadata*3'0'0
Author(s) and
Publication Year
Whitehead et al.
(1995)
Whittington et al.
(1994)
Obs. in
Metadata
2
1
State(s)
NC
TX
Water Body
Type(s)
estuary
estuary
Willingness to Pay per Household
(2007$)
Mean
95.44
194.72
Min.
78.29
194.72
Max.
112.59
194.72
Notes:
(A): Shading indicates studies added since the 2009 meta-analysis.
(B): Journal publications by Corrigan et al. (2009) and Whitehead (2006) replaced staff papers authored by Azevedo et al. (2001)
and Whitehead (2002), respectively.
(C): A study by Olsen et al. (1991) from the 2009 meta-analysis was excluded from the current metadata following robustness
testing and additional review of study details.
(D): The revised metadata excludes studies by Viscusi et al. (2008) and Carson and Mitchell (1993) that surveyed households
across the U.S.
"
.2 Variable Development and Coding
EPA entered values for all existing variables in the metadata for each new study identified during the
literature review. Next, EPA developed and coded new variables (for all studies) that address gaps in the 2009
metadata. The variables in the revised MRM fall into four general categories:
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.
Region and surveyed populations variables characterize such features as the US region in which the study was
conducted, the average income of respondent households and the representation of users and nonusers within
the survey sample.
Sampled market and affected resource variables characterize features such as the geospatial scale (or size) of
affected water bodies, the size of the market area over which populations were sampled, as well as land cover
and the quantity of substitute water bodies.
Water quality (baseline and change) variables characterize baseline conditions and the extent of the water
quality change.
Table H-2 presents and defines key variables from the metadata by category. Shading indicates that the
variable was developed during these revisions. The development of new variables focused primarily on
developing variables that capture geospatial factors including extent of sampled market, surveyed populations
and affected resources.94 The 2009 MRM largely relied on binary variables to distinguish broad categories of
affected resources (e.g., single vs multiple rivers, regional freshwater, etc.). Extensive GIS mapping is
required in order to quantitatively incorporate new spatial factors using continuous variables because many
primary studies give broad outlines but omit the detailed geospatial information regarding affected resources
and sampled populations. This type of supplementation has been used by others to include various types of
EPA also coded one new methodological variable, ce, to identify studies that used a choice experiment
approach.
September 29, 2015
H-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
information into valuation MRMs other than that necessary for the current metadata (e.g., Ghermandi et al.
2010; Ghermandi and Nunes 2013). By linking to publicly available GIS data layers, EPA provides a
consistent basis for the construction and coding of the new, spatial variables. The following subsections
describe both development and coding of water quality and spatial variables for surveyed populations, extent
of sampled market, study site, and affected resources. In accordance with Stanley et al. (2013) guidelines, all
variable coding was reviewed by multiple individuals and documented clearly in the metadata and data
dictionary.
Annual WTP values are typically desired for policy analysis to support a comparison of annualized costs and
benefits. The metadata includes total WTP as reported by the original study, adjusted to 2007$.95 For over 80
percent of observations (114 studies), total WTP reflects annual payments in perpetuity (113 studies) and one
observation reflects annual WTP for a 10-year payment period. For the remaining 26 observations, WTP
reflects a short-term payment period (11 observations use a one-time lump sum payment, 1 observation uses a
3-year payment period, and 14 observations use a 5-year payment period). EPA used the lump_sum binary
variable to account for short-term payment periods (lump_sum=\). EPA also tested models with WTP values
for short-term payment periods converted to perpetual streams but found that results were sensitive to
discount rate assumptions and that these models did not provide any practical advantages for benefit transfer
over models using the lump_sum variable.
Table H-2: Definition and Summary Statistics for Model Variables*
Variable
Definition
Units
Mean
St. Dev.
Study Methodology and Year
Ce
Thesis
Lnyear
Volunt
outliers trim
nonparam
non reviewed
lump sum
Binary variable indicating that the study is a choice
experiment
Binary variable indicating that the study is a thesis.
Natural log of the year in which the study was
conducted (i.e., data was collected), converted to an
index by subtracting 1980.
Binary variable indicating that WTP was estimated
using a payment vehicle described as voluntary as
opposed to, for example, property taxes.
Binary variable indicating that outlier bids were
excluded when estimating WTP.
Binary variable indicating that WTP was estimated
using non-parametric methods.
Binary variable indicating that the study was not
published in a peer-reviewed journal.
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.. This variable enables the benefit transfer
analyst to estimate annual WTP values by setting
lump sum=Q.
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Natural log of
years (year
ranges from
1981 to 2011).
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
0.1071
0.1143
2.2127
0.0857
0.1929
0.4286
0.2357
0.1857
0.3104
0.3193
0.9282
0.2809
0.3960
0.4966
0.4260
0.3903
EPA used $2007 because the a majority of observations were already converted to $2007 as part of the 2009
meta-analysis used ELGs (U.S. EPA, 2013a).
September 29, 2015
H-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Table H-2: Definition and Summary Statistics for Model Variables
Variable
wtp median
Definition
Binary variable indicating that the WTP measure
from the study is the median.
Units
Binary
(Range: 0 or 1)
Mean
0.0714
St. Dev.
0.2585
Region and Surveyed Populations
northeast
Binary variable indicating that the survey included
respondents from states within the Northeast U.S.,
defined as ME, NH, VT, MA, RI, CT, and NY. This
is equivalent to EPA Region 1 plus NY or the
Northeast USDA region.
Binary
(Range: 0 or 1)
0.0714
0.2585
Central
Binary variable indicating that the survey included
respondents from states within the Central U.S.,
defined as OH, MI, IN, IL, WI, MN, IA, MO, ND,
SD, ME, KS, MT, WY, UT, and CO. This is
equivalent to EPA regions 5, 7, and 8 or the
Midwest and Mountain Plains USDA regions.
Binary
(Range: 0 or 1)
0.3643
0.4830
southB,C
Binary variable indicating that the survey included
respondents from states within the Southern U.S,
defined as NC, SC, GA, FL, KY, TN, MS, AL, AR,
LA, OK, TX, and MM. This is equivalent EPA
regions 4 and 6, or the Southeast and Southwest
USDA regions.
Binary
(Range: 0 or 1)
0.1571
0.3652
nonusers
Binary variable indicating that the survey was
implemented over a population of recreational
nonusers (default category for this variable is a
survey of any population that includes both users
and nonusers).
Binary
(Range: 0 or 1)
0.0857
0.2809
Inincome
Natural log of the median income (in 2007$) for the
sample area of each study based on historical U.S.
Census data. It was designed to provide a consistent
income variable given differences in reporting of
respondent income across studies in the metadata
(i.e., mean vs. median). Also, some studies do not
report respondent income. This variable was
estimated for all studies in the metadata regardless
of whether the study reported summary statistics for
respondent income.
Natural log of
income
(2007$)
(sample area
income ranges
from $34,332
to $78,444)
10.7453
0.1731
Sampled Market and Affected Resource
mult bod
River
swim use
Binary variable indicating that the survey addressed
multiple waterbody types. The eight waterbody type
categories are (1) river/stream, (2) lake, (3) all
freshwater, (4) estuary, (5) wetlands, (6) river and
lake, (7) salt pond/marshes, and (8) multiple
(estuary and fresh water). Takes on a value of 1 if
the study is coded as category (3), (6), or (8).
Binary variable that takes on a value of 1 if the
study affects a river, such that river length>0, and
zero otherwise.
Binary variable indicating that the affected use(s)
stated in the study include swimming.
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
Binary
(Range: 0 or 1)
0.0786
0.6857
0.2643
0.2700
0.4659
0.4425
September 29, 2015
H-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Table H-2: Definition and Summary Statistics for Model Variables
Variable
Definition
Units
Mean
St. Dev.
gamefish
Binary variable indicating that the affected use
stated in the study is gamefishing.
Binary
(Range: 0 or 1)
0.0571
0.2329
boat use
Binary variable indicating that the affected use(s)
stated in the study includes boating.
Binary
(Range: 0 or 1)
0.1143
0.3193
ln_ar_agr
Natural log of the proportion of the affected
resource area which is agricultural based on NLCD,
reflecting the nature of development in the area
surrounding the resource. The affected resource
area is defined as all counties that intersect the
affected resource(s). EPA also tested a variable for
the fraction of lands that is developed land classes,
but it did not improve model fit.
Natural log of
proportion
(Range: 0 to 1)
-1.4329
0.9031
In ar ratio
A ratio of the sampled area, in km , relative to the
affected resource area. When not explicitly reported
in the study, the affected resource area is measured
as the total area of counties that intersect the
affected resource(s), to create the variable
ar_total_area. From here, ln_ar_ratio =
log(sa_area I ar_total_area), where sa_area is the
size of the sampled area in km2.
Natural log of
ratio
(km2/km2)
-1.1278
2.6067
sub proportion
The proportion of water bodies of the same
hydrological type affected by the water quality
change, within affected state(s). For rivers, this is
measured as the length of the affected river reaches
as a proportion of all reaches of the same order. For
lakes and ponds, this is defined as the area of the
affected water body as a proportion of all water
bodies of the same National Hydrography Dataset
classification. For bays and estuaries, this is defined
as the shoreline length of the water body as a
proportion of all analogous (e.g., coastal) shoreline
lengths. To account for observations where multiple
waterbody types are affected, the variable
sub proportion is defined as maximum of separate
substitute proportions for rivers, lakes, and
estuaries/bays. The affected resource appears in
both the numerator and denominator when
calculating sub_proportion.
Proportion
(Range: 0 to 1)
(km/km)
0.1880
0.2911
Water Quality Baseline and Change
Inquality ch
Lnbase
Natural log of the change in mean water quality
(quality _ch), specified on the WQI (McClelland
1974; Mitchell and Carson 1989).
Natural log of baseline water quality, specified on
the WQI (McClelland 1974; Mitchell and Carson
1989).
WQI units
WQI units
2.9070
3.5889
0.6039
0.6697
September 29, 2015
H-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
Table H-2: Definition and Summary Statistics for Model Variables*
Variable
Definition
Units
Mean
St. Dev.
Notes:
(A): Shading indicates that the variable was developed and coded during the current set of revisions.
(B): EPA merged studies from the Southeast and Southwest region into a single group because the metadata includes only one
observation from the Southwest region, Whittington et al. (1994) which studied estuaries in Texas
(C): The USDA regions omitted from the regional binary variables are the Mid-Atlantic region (NJ, DE, MD, DC, PA, WV, and
VA) and the Western region (WA, OR, ID, CA, NV, AZ, and AK).
H.2.1 Study Methodology and Year
As often found in meta-analyses within the valuation and benefit transfer literature (Navrud and Ready 2007),
a variety of study and methodology effects can be shown to influence WTP for water quality improvements.
The 2009 metadata included numerous variables characterizing factors including, but not limited to
publication type, elicitation method, response rate, treatment of outliers, payment period, and study year. EPA
coded values for these variables for the new primary studies following the variable definitions from the 2009
MRM. The majority of the methodological variables in the metadata are binary, with 1 indicating that a given
methodology was used in the primary study. Some others, such as study year, are continuous. Most of the
methodological variables available in the metadata did not improve model fit and were not included in the
revised MRM. Given that choice experiments are being used with increased frequency in the recent literature,
EPA coded a new binary variable, ce, to flag observations that are based on a choice experiment framework.
H.2.2 Region and Surveyed Populations
The 2009 metadata included variables describing the location of the study by state and region of the country
although these were used in a fairly simplistic way in the meta-analysis. EPA expanded the metadata by
delineating sampled area for each of the 140 WTP observations as GIS polygons. The sampled area is defined
as the geographic area over which the primary survey was fielded (i.e., the population it is meant to be
representative of). The Agency based its GIS delineation on the description of the study area from the study
documentation and matched to polygon boundaries from publicly available GIS datasets. The sampled area
can vary across observations from the same primary study if the WTP values are based on separate survey
samples (this is true for 5 of 51).
The sampled areas are typically defined by either jurisdictional boundaries or watershed boundaries:
For sampled areas defined by jurisdictional boundaries (e.g., states, counties, or cities), EPA defined
boundaries using the Census Topologically Integrated Geographic Encoding and Referencing (TIGER) state
and county shapefiles for 1990, 2000, and 2010. By using the shapefile vintage that most closely corresponds
to the year in which the study was fielded, EPA is able to accurately identify boundaries at the time of the
study.
For sampled areas defined by watersheds, EPA approximated sampled area boundaries by matching to the
U.S. Geologic Survey Hydrologic Unit Code (HUC) Watershed Boundary Dataset (WBD) to approximate the
boundaries of the study area.96 The Agency first matched the watershed polygon in the HUC dataset based on
96 The HUC WBD is a package of watershed shapefiles for the contiguous U.S. subdivided to seven levels of
resolution. Each resolution level is assigned a specific number of HUC digits, with the number of digits
increasing with higher resolutions. The lowest resolution is the HUC 2-digit, or HUC-2, watershed which divide
the country into 21 distinct drainage regions. The highest classification is the HUC12 dataset, consisting of
September 29, 2015 H^9~
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
the watershed name presented in the study. If the name could not be matched, EPA consulted the survey
materials or maps included in the study to either locate the corresponding HUC watershed or use the materials
to trace a watershed polygon manually. If maps are not provided, then EPA used any additional locational
information presented in the study to identify a HUC watershed to serve as a proxy for the study watershed.
This approach generally involved more interpretation than if the study defined the sampled area based on
jurisdictional boundaries (as in previous bullet).97
Income is inconsistently reported across primary studies, with some reporting median income, others
reporting mean income, and others not reporting any income statistics for the sample. After delineating the
sample area for all studies, EPA calculated income statistics for the population within the sampled area using
georeferenced historic U.S. Census income data to overcome reporting deficiencies and improve consistency
across observations. Historical median income data is available for 1980, 1990, 2000, and 2010. Each
sampled area was matched to the data vintage that is nearest to the year in which the study was fielded. EPA
derived sampled area income as a population-weighted average of median income across counties that
intersect the sampled are polygons. The population weights were adjusted based on the fraction of county area
that is within the sample area. Resulting income estimates were adjusted to 2007$ using the Consumer Price
Index (CPI).
The 2009 metadata included regional dummies for the study location based on the boundaries of U.S.
Department of Agriculture (USDA) regions.98 The USDA regions were used as the starting point for regional
assignments. The boundaries of the seven USDA regions are closely aligned with the ten EPA regions, EPA
generated two new binary variables that combine some regions for which effects were found to be similar.
Northeast identifies studies that sampled respondents in the Northeast U.S. (as far south as New York, central
identifies studies that sampled in the Midwest or Mountain Plains regions of the U.S, and south identifies
those that sampled in the Southeast and Southwest U.S. (as far west as New Mexico).99 The state boundaries
for each variable are listed in Table H-2.
EPA also calculated the continuous variables sa_area, defined as the size of the area sample in square
kilometers. Sa_area feeds into the index variable, ln_ar_ratio, described in detail in the following subsection.
H.2.3 Reconciliation of Water Quality Baseline and Change
An important component metadata development is the reconciliation of variables across observations
(Johnston et al. 2005; Smith and Pattanayak 2002; Smith et al. 2002; Van Houtven et al. 2007). Although the
calculation and reconciliation of most independent variables requires little explanation, there are some
variables for which additional detail is warranted. These include variables characterizing surface water quality
and its measurement. To reconcile measures of water quality across studies, EPA adapted the prior approach
approximately 98,000 watersheds. Additional information about the HUC watershed and delineation process
can be found at http://water.usgs.gov/GIS/huc.html.
97 EPA encourages authors of future studies to describe the sampled area clearly, to avoid the need for judgment
or interpretation when including studies in future meta-analyses.
98 Using ecoregion boundaries to define regional dummies is not feasible because it would result in a very small
number of observations per region. Each USDA region includes several ecoregions and thus is better suited for
grouping studies.
99 The default region corresponds to the Mid-Atlantic states (DE, MD, NJ, PA, VA, WV). It also includes the state
of California because only two studies (3 observations) correspond to California and because similarly to the
Mid-Atlantic states it is a coastal state. No meta-data observations correspond to other states in the Western
U.S. (i..e, OR, WA, NV, ID, HI, AZ, AK).
September 29, 2015 H-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
of Johnston et al. (2005) and van Houtven et al. (2007), mapping water quality changes to the 100-point
Water Quality Index (WQI).
A large number of the original studies in the metadata (10 studies and 31 percent of observations) include the
10-point water quality ladder (WQL) measures as a native component of the original primary studies either
using its numerical values (e.g., 2.5, 5.0, etc.) or descriptive levels (e.g., beatable, gamefishing).100 In other
cases, EPA used descriptive information from studies to map baseline and post-improvement to points on the
WQL. To calculate 100-point WQI values EPA multiplied the reported WQL values by ten. In most cases,
descriptions correspond closely to levels on WQL, rendering mapping straightforward. EPA used the
following guidelines for mapping:
WQL assignments are made based on consideration of information provided in the primary study
documentation and, if available, the survey materials. The assignments are not supplemented by external
water quality databases or reports. They should reflect the descriptions and metrics presented to respondents
within the valuation scenarios.
Identify recreational uses (e.g., boating, fishing, and swimming) provided in the description of baseline and
improved (or declined) conditions. These uses may be stated directly or embedded in the definition of water
quality metrics (e.g., low, moderate, high) and can include additional descriptive information such as effects
on sensitive aquatic species or indication of presence of or amount of specific pollutants.
Only consider those metrics that can be reasonably mapped to the WQL. Ancillary improvements from the
improvement plans, such as shoreline trash pickup or terrestrial bird species, are not considered.
Exclude observations that describe an improvement in recreational use that is not tied to a water quality
improvement (e.g., an increase in fish abundance not based on water quality improvement).
Start by assigning values corresponding to use thresholds defined on the WQL (e.g., 2.5, 4.5). Assign
intermediate values when the changes occur within a specific use category or changes extend beyond the
minimum for provision of use. EPA notes that the majority of assignments directly match values
corresponding to WQL use thresholds.
To the extent possible, assigned WQL values should reflect the "affected" portion of resource(s) described in
the survey, that is, the portion of the resource that is subject to water quality changes under the valuation
scenario.
In some cases, uses may be supported intermittently due to algae blooms, for example. If provided, use
information regarding the frequency of service provision to calculate intermediate WQL values by weighting
use threshold values.
For some types of environmental contamination, such as high acidity, the survey may state that the affected
resource supports swimming but has degraded fisheries. It supports a higher use on the WQL (swimming), but
not a lower use (gamefishing). . In these cases, we specify baseline WQL and WQL changes based on fishing
conditions, the use that is actually improved under the valuation scenario (e.g., a baseline of 4.5 for rough
fishing and a post-improvement value of 5.0 for game fishing).
For more detail on water quality description from the original studies and the assigned baseline and improved
water quality conditions for each study see Appendix B in Peer Review Package for Meta-analysis of the
Willingness-to-Pay for Water Quality Improvements (U.S. EPA, 2015e).
100 The WQL (Vaughan 1986) is expressed on a scale of 0 to 10 and can be mapped to the WQI by multiplying by
10. Refer to Chapter 10 and Appendix G of the benefits analysis for the C&D Effluent Guidelines (U.S. EPA
2009) for additional detail on the WQI and WQL.
September 29, 2015 H-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
EPA did not identify any systematic variation in results associated with studies for which the WQI was a
native component, versus those for which quality changes were mapped to the WQI. A binary variable
indicating that the observation used the WQL (CHNG WQL) was not individually significant and did not
improve model fit.
H.2.4 Delineation of the Affected Resource and Substitutes
WTP for WQI changes is likely better explained in concert with geospatial factors, given the geographic
heterogeneity in the metadata. In the context of water quality valuation, the geospatial scale of a water quality
change reflects the area over which the environmental change occurs. EPA expects, all else being equal, that
WTP for water quality improvements would be directly related to the size and number of affected
waterbodies. EPA defines an affected resource as a waterbody that experience a water quality change under
the valuation scenarios presented in the primary study. The number of affected waterbodies assessed in the
primary studies varies widely across the metadata from single waterbodies to all waterbodies throughout a
state or small region. The existing metadata identifies the geographic scale of the affected resources using
four categories: (1) single waterbody, (2) multiple waterbodies, (3) a small region, or (4) a large region. Table
H-3 summarizes the number and percent of observations in each of the four geographic scale categories. As
part of the revisions, EPA delineated the affected waterbodies by matching waterbody descriptions and maps
from the primary studies to publicly available georeferenced datasets. By delineating resources using GIS,
EPA is able to incorporate detailed spatial and hydrological features of the waterbodies. As with sampled
area, some primary studies provide separate WTP values for multiple affected resources or sets of affected
resources. These are always treated as separate observations within the metadata.
Table H-3: Summary of Scale of Resource Changes in the Meta-data.
Category
Single
Waterbody
Multiple
Waterbodies
Small Region
Large Region
Total
Variable
Single
Multi
Sm reg
Lg_reg
-
Description
Single, discrete waterbody such as a river
or lake.
Multiple discrete waterbodies (e.g. , 2
rivers)
Small region, includes watersheds and
state-wide analyses
Large region, includes one estuary
analysis that includes multiple states.
-
Observation
Count
49
13
77
1
140
%of
Observations
35.0%
9.3%
55.0%
0.7%
100.0%
The datasets used to map affected resources vary by affected waterbody type. For rivers and lakes, the
Agency opted to match to the NHDPlus dataset over other publicly available georeferenced hydrologic
datasets (e.g., E2RF1) because it is a comprehensive, national dataset of hydrologic features such as rivers,
lakes, ponds and catchments, and contains tabular data that can be linked to these hydrologic features. For
estuaries and coastlines, EPA used the National Oceanic and Atmospheric Administration (NOAA) Global
Self-Consistent Hierarchical, High-Resolution Geography Database (GSHHD). EPA attempted to match the
name of the affected resource to the datasets, and if a match could not be found, used narrative descriptions
and maps from the primary study to manually select the affected waterbodies in ArcGIS. Once delineated,
EPA calculated affected shoreline length for all waterbody types as proxy for the human/water interface,
although some uses actually occur in open water. For rivers, shoreline length is double the reach length to
reflect both shorelines. EPA also calculated the area of affected lakes.
September 29, 2015
H-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
Observed mean WTP in a primary study is also related to the extent of market analyzed (Loomis 2000;
Loomis and Rosenberger 2006). It is established in economic literature that mean WTP often declines as
distance increases to an affected resource (i.e., distance decay) (Bateman et al. 2006; J0rgensen et al. 2013;
Schaafsma et al. 2012). Because the scale of the sampled market area varies greatly across primary studies
and past work has shown that such differences can have important implications for welfare estimates, the
accuracy of benefit transfers depends on the ability to adjust for the distance between affected populations and
the resource (Bateman et al. 2006; Johnston and Duke 2009; Johnston and Ramachandran 2013; Loomis and
Rosenberger 2006). The size of the sampled area (sa_ared) in the metadata varies greatly across the metadata
mean of 49,065 mi2 with 5th and 95th percentiles of 13.3 and 104,094 mi2, respectively. Sa_area is not
correlated with the size of the affected resource because various factors affect researcher's choice of survey
region including budget and resource constraints and study objectives. As a result, the metadata does not
provide clear information regarding the extent of the market for the affected resources or distance effects.
Modeling results indicated that model performance was enhanced, in terms of model fit, variable significance,
and consistency with theoretical expectations, when resource size was specified as a function of sampled
market area (sa_area). The ln_ar_ratio variable reflects the size of the sampled market relative to the size of
the affected resource, defined as the (natural log of the) size of the sampled market area (sa_ared) divided by
the geographic area encompassing the affected waters in square kilometers (ar_total_area). Ar_total_area is
calculated based on the area of counties that insect the affected resource. For example, if a study valued all
freshwater in Iowa, then ar_total_area would equal the area of Iowa. On other hand, ar_total_area would
equal the area of a single Iowa county if the affected resource falls entirely within the county boundary. The
area ratio, ln_ar_ratio, is expected to have negative sign reflecting two effects. First, holding all else constant,
stated preference surveys over larger market areas imply greater distances between individual households and
affected waterbodies, and thus lower WTP per household. Secondly, improvements to more waters within the
sampled area should be associated with greater WTP, ceteris paribus.
About 55 percent of observations in the metadata estimated WTP for water quality changes in a single
watershed, group of watersheds, or a state. The other 45 percent focused on a single waterbody or small set of
waterbodies (e.g., one lake or 3 lakes). Specifying the affected resource area requires some assumptions if the
affected resource is defined as a river or lake (e.g., EPA assumed that counties intersecting the affected water
bodies represent the affected resource area). As noted above, EPA used intersecting counties to define all
affected resources, including discrete rivers, as geographic areas while avoiding the arbitrary selection of a
boundary distance.
Relative to the alternatives, ln_ar_ratio provides a more intuitive means to capture the scale of the affected
resource occurring throughout an area, such as watershed or region which is likely to be the case in the
context of national rulemakings. Model fit decreases when the affected area ratio is included in non-logged
form and the variable (ar_ratio) is no longer individually significant. Logging the ratio is also intuitive
because WTP is zero when the resource size is zero. EPA considered, but did not select, a variety of
alternative specifications of the geographic scale and extent of market variables (e.g., shoreline length).
The availability and quality of substitutes in the surrounding geographic area are also expected to influence
welfare estimates (Loomis and Rosenberger 2006; Schaafsma et al. 2012). All else equal, WTP should be
negatively related to substitute availability. EPA developed a continuous variable, sub proportion, defined as
the ratio of affected waterbodies to available use and nonuse substitutes within the state. EPA expects that the
relationship between distance and substitutes will differ for resource users and nonusers. For users, there is "a
positive cost of access which depends on distance" (Hanley et al. 2003, p.300), therefore, sites that are farther
away will generally be less attractive substitutes all other things being equal. On the other hand, "there seems
no theoretical reason to expect such a [distance decay] trend to be seen within the responses of those who are
September 29, 2015 H-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
present non-users" (Bateman et al. 2006, p. 453). There is some evidence that a sense of spatial and cultural
identity or "ownership" may be important for nonuse values for some environmental resources (Bateman et
al. 2006; Hanley et al. 2003). Hanley et al. (2003) provides the example that "I may have stronger non-use
values for Scottish wildlife sites if I am Scottish than for English wildlife sites" (p.300). In the U.S., this may
be manifested as affiliation with one's state of residence. Given these factors, EPA considers the state to be a
reasonable basis for the calculating the substitute proportion.
The substitute proportion variable (subproportion) represents a potential advance over less sophisticated
binary variables traditionally used to represent the size or scope of affected resources. For studies restricted to
rivers, the calculation of substitutes is restricted to rivers of the same stream order(s) as the affected resource
to ensure comparability. For non-river inland waterbodies (e.g., lakes and ponds), substitute proportion is
calculated based on area relative to area of lakes within the state(s).101 For estuaries and coasts, EPA
calculated the denominator as all coastline miles in the state(s). EPA also tested models with separate
substitute variables for each waterbody type, but model performance was not improved.
By delineating an affected area, in addition to the waterbodies themselves, EPA is also able to analyze the
development characteristics of areas bordering the waterbodies. In particular, EPA estimated the fraction of
the land in the study area used for agriculture (ln_ar_agr). EPA used the National Land Cover Dataset
(NLCD) to develop a land cover profile of the affected resource area. The NLCD is a high-resolution (30
meter) spatial dataset of land cover across the contiguous U.S. and available for the years 1992, 2001 and
2006. The NLCD has 16 or 21 categories of land cover depending on the year. For the purpose of consistency
and simplicity, EPA defined the agricultural land as the sum of "cultivated cropland" and "pasture/hay". The
Agency used the change in agricultural land between NLCD vintages to interpolate values for the actual year
of the study. The agricultural fraction is calculated as the area-weighted average of counties within the
affected resource area. Areas dominated by agricultural land uses have particular characteristics which
suggest that WTP for water quality improvements could be lower than other types of areas, ceteris paribus:
> First, unlike non-agricultural rural areas, heavily agricultural areas have generally been altered from
their natural ecosystem conditions, and do not tend to be highly prized for water-based recreation (or
characterized as pristine natural areas) - this would be expected to decrease WTP for improvements
in agricultural areas compared to many other rural areas.
> Second, unlike more heavily populated suburban or urban areas, agricultural lands do not have the
population base to support well-developed and used recreational areas. Improvements to water bodies
in suburban areas, for example, are often highly valued because these areas support extensive
recreational and other uses.
> Third, a greater proportion of the population in agricultural areas has employment linked to farm
activities that may be associated directly or indirectly with water pollution. Those whose employment
depends on farming activities may be hesitant to support programs to improve water quality (in stated
preference surveys or otherwise), for fear that these policies may lead to greater restrictions on farms
and farming activities.
For all of these reasons, one would expect areas dominated by agriculture to have lower WTP to improve
water quality, again holding all else constant. This intuition is strongly supported by model results, as
discussed below. EPA also developed and tested an analogous variable based on the fraction of land that is in
developed land use categories but it not improve model fit.
101
NHDPlus does not provide widths of river and streams.
September 29, 2015 H-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
EPA tested a number of meta-regression models based on WTP estimates for improvements in water
resources, derived from 5 1 original studies. However, only the marginal WTP model ("Model 1") and its
variant ("Model 2") are used in the analysis of benefits of the final ELG because these models satisfy adding-
up conditions (Diamond 1996). Model specification for the marginal WTP model, results, and interpretation
of the results are described in the following sections. EPA's Peer Review Package for Meta-analysis of the
Willingness-to-Pay for Water Quality Improvements (U.S. EPA, 2015e) and memorandum entitled
Accounting for Scope in the MRM2 Meta-Regression Model (Abt Associates 2015) provide additional detail
on model selection, testing and alternative specifications.102
H.3.1 Marginal Willingness to Pay (Model 1)
In the meta-regression model, marginal willingness -to -pay from observation i, MWTPi =
as follows:
dWTP
is modeled
ln(MWTP{) =
f(Qt;
(1)
In this equation, a0 is a constant, Xt is a vector of study and resource characteristics that act as demand
shifters, Qi is the absolute water quality index level at which marginal willingness-to-pay is being evaluated,
and 6t is a zero-mean normally -distributed error term. The function /(•) expresses the core relationship
between absolute water quality Qi and the log of marginal WTP, and /? is a vector of parameters that define
the functional form of /(•)• If /(') is assumed to be linear in Qi\
f(Qi; P) M • &
then marginal WTP can be expressed as follows:
(2)
(3)
However, MWTPi is not observed directly in the meta-data, and instead must be calculated from total
willingness-to-pay as part of the estimation routine. This would require a nonlinear least squares or maximum
likelihood estimation approach. As a simpler alternative, EPA assumed that marginal WTP can be
approximated by average WTP per unit of water quality:
with this approximation assumed to be valid at some point between Qi0 and Qtl, here approximated by the
midpoint of the water quality change valued for that meta-data observation:
Note that Model 1 here is the model referred to as MRM2 in the Peer Review Package, while Model 2 here is
the model referred to as MRM2-S in the Accounting for Scope memorandum.
September 29, 2015 H-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
Substituting Equations (4) and (5) into Equation (3) provides the version of the marginal WTP model that is
ultimately used as a regression model (Model 1):
The simplest version of this regression uses an unweighted OLS approach to generate a vector of estimated
parameter values (a0, a1; /?, <72), where a2 is the sum of squared residuals divided by N-K, where N is the
number of observations and K is the number of parameters. Model 1 then substitutes these estimated values
into Equation (3) above, to get an expression that can be used to predict the log of marginal WTP:
ln(MWTPi) = a0 + arfi + j§ • & + e£ (7)
Taking the exponent of both sides produces the following expression for marginal WTP evaluated for a
particular set of study and resource characteristics X and a particular water quality index level Q :
MWTP = exp(a0 + a^X + <72/2) • exp(/? • (?) (8)
Small changes in water quality (e.g., less than one unit) could be valued using an approximation that involves
multiplying predicted MWTP from equation (8) by the amount of the water quality change. Larger water
quality changes must account for the curvature of the marginal WTP function, and so must be calculated
using the integral:
WTP = J^Qo exp(a0 + a±X + «72/2) • exp(j§ • (?) dQ (9)
which is equal to:
WTP = exp(a0 + a±X + a2/2)
Because 98 percent of reach miles affected by the final ELG would experience WQI changes that range
between 0 and 1 EPA estimated WTP for water quality improvements by multiplying MWTP by the amount
of water quality changes expected from the final ELG, as described in Chapter 4.
H.3.2 Marginal Willingness to Pay (Model 2)
A key feature of Model 1, as described in Equation (1) above, is that marginal WTP does not depend on the
magnitude of the water quality change being valued. In other words, the model assumes that marginal WTP
for the one-unit change from 49 to 50 on the water quality index is the same, regardless of whether survey
respondents live in an area where baseline water quality is 35 or 49, and whether the expected total water
quality change is 30 points or 2 points, respectively.
September 29, 2015 H-16
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
The meta-data, however, do show a relationship between marginal WTP and the water quality change. This
relationship could occur for any number of reasons, including the transformations associated with studies that
did not use the WQI or WQL directly, or an omitted variable that is correlated with the water quality
improvement. More specifically, the meta-data show that survey respondents express high marginal WTP for
modest improvements in water quality, but then express much lower marginal willingness to pay for
additional improvements. For more detail see a memorandum entitled Accounting for Scope in the MRM2
Meta-Regression Model (Abt Associates 2015).
To address this relationship in the meta-data while satisfying the adding up condition, EPA developed a
variant of the marginal WTP model that includes water quality change as a study methodological
characteristic. Similar to other methodological parameters (e.g., the use of a choice experiment format, or a
lump sum payment), the water quality change parameter could be set to any appropriate value when the meta-
model is used for benefit transfer. Importantly, because water quality change would be treated as a
methodological characteristic instead of as a scenario characteristic, the value assigned to the water quality
change parameter need not correspond to the specific water quality change being valued in the benefit transfer
application. For the same reason, the resulting model will still satisfy the adding -up property.
The following equation presents an expanded version of marginal WTP that allows the water quality change
AQj to enter as a methodological characteristic (hereafter Model 2):
ln(MWTP{) = a0 + a^Xt + g(AQt; 0) + /(&; p) + et (11)
In this equation, AQj is the magnitude of the water quality change for meta-data observation i, and 0 is a
vector of parameters that specify the functional form for the function #(•). For simplicity, EPA modeled the
function $(•) as a linear function of the water quality change:
(12)
Substituting Equations (4), (5), and (12) into Equation (1 1) gives the following regression model:
* (13)
The estimated parameters from this equation can be used to generate an expression for both marginal WTP
and total WTP:
MWTP = exp(a0 + a^X + 0 • A(? + a2 /2) • exp(j§ • (?) (14)
WTP = exp(«0 + a±X + 0 • AQ +
Because AQ is treated as a study methodological characteristic, not as a characteristic of the benefit transfer
water quality improvement scenario, the parameter AQ need not be set equal to Qi — Qo m equations (14) and
(15).
H.4 Regression Result
Table H-4 presents regression results for Model 1 and Model 2. The models presented in Table H-4 were
selected after the estimation of numerous preliminary models with different specifications and groups of
September 29, 2015 H-17
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
independent variables. The selection was based on both statistical fit and correspondence with theoretical
expectations. EPA estimated the models using robust standard errors to account for study-level dependencies.
Measures of fit for the illustrated models are good (R2 = 0.641 and 0.773). These compare favorably to prior
meta-analyses in the published literature.
The model performs well, with intuitive results for virtually all statistically significant variables. The model
identifies numerous statistically significant coefficients for variables characterizing (1) study methodology,
(2) region and surveyed populations, (3) extent of the market, study site, and affected resources, and 4) water
quality. In total in Model 1,15 out of 23 non-intercept coefficient estimates are statistically significant at
p<0.05, and 12 of 23 are significant at p<0.01. The sub proportion geospatial variable is significant at p<0.01
with signs on coefficients that are consistent with expectations.
Given the trans-log functional form, these results for ln_ar_ratio imply diminishing marginal returns to scale
(e.g., WTP per unit of water quality improvement declines as the scope of water quality change increases).
Marginal WTP is lower among studies surveying only nonusers (nonusers), another expected result, because
users may hold both use and nonuse values, while nonusers hold only nonuse values, by definition.
Methodological variables also show expected patterns. For example, households are willing to pay a greater
nominal amount in a one-shot payment (lump_sum) than they would be willing to pay in repeated annual
payments.103 Marginal WTP also varies according to the type of uses potentially affected by the proposed
changes (particularly boatjise). In some cases, these variables may be correlated with water quality
condition. However, counterexamples exist: for example, waters may be swimmable but the study focuses on
improvements in game fishing (e.g., Banzhaf et al., 2006). EPA found that model fit was better with
swim_use, boatjise, and gamefish included.104 Thus, dropping these variables could result in omitted variable
bias.
As expected, sub proportion has a positive coefficient, indicating that WTP increases as the affected resource
constitutes a larger fraction of available substitutes. The coefficient for ln_ar_ratio is negative because the
sampled area (sa_area) is in the numerator indicating that marginal WTP decreases as the sampled market
area increases relative to the affected resource. This is consistent with the expectation that a larger sample
area will include a greater proportion of respondents who are resource nonusers, who are less familiar with
the resource, and/or who live at a greater distance from the affected waters.
A negative coefficient on the agricultural area variable (ln_ar_agr) suggests that areas dominated by
agriculture have lower WTP to improve water quality, holding all else constant. Because areas dominated by
agriculture may be significantly different in terms of both resource and population characteristics, as
discussed in the preceding section, this result is not surprising. EPA also notes that removing this variable
causes substantial changes elsewhere in the model, a sign that removing this variable could cause non-trivial
omitted variables bias.105
The absolute water quality variable (Qavg2) shows that water quality has a larger effect on marginal WTP
under Model 1, compared to Model 2. The coefficient is -0.017 under Model 1 and -0.004 under Model 2.
As discussed in Section 3, EPA also tested models with WTP values for short-term payment periods converted
to perpetual streams but found that results were sensitive to discount rate assumptions and that these models did
not provide any practical advantages for benefit transfer over models including the lump_sum variable.
104 Peer Review Package for Meta-analysis of the Willingness-to-Pay for Water Quality Improvements (U.S. EPA,
2015e)
105 EPA notes that some peer reviewers questioned inclusion of this variable in response to the peer review (U.S.
EPA, 2015e). In this analysis, the Agency followed the econometric guidance suggesting that if the inclusion of
a variable within a model is questionable, the default solution should be to include the variable (including an
irrelevant variable only reduces efficiency; omitting a relevant variable creates bias).
September 29, 2015 H-18
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
This pattern suggests that after accounting for the change in water quality, absolute water quality may have
less of an effect on marginal WTP. However, EPA notes that a t-test would fail to reject the hypothesis that
these coefficients are equal.
Model 2 also includes a water quality change variable: lnquality_ch. This variable is negative and highly
significant. This pattern indicates that marginal WTP is decreasing in the size of the water quality change
being valued. This is consistent with the patterns from the raw meta-data, as discussed in detail in a
memorandum entitled Accounting for Scope in the MRM2 Meta-Regression Model (Abt Associates 2015).
Overall, a comparison of the R2 values from the two models indicates Model 2 has greater explanatory power
than Model 1. Model lhas a R2 value of 0.641. In contrast, the R2 value for Model 2 is 0.773. Given that
Model 1 already has 23 variables and a constant, this substantial increase in R2 from one additional variable
suggests that log water quality change is an important variable to include in the model.
Table H-4: Regression Results
Parameter
ce
thesis
Inyear
volunt
OUTLIER _BIDS
nonparam
non reviewed
lump sum
WTP median
northeast
central
south
nonusers
Inincomel
mult bod
riverl
swim use
gamefish
boat use
In ar agr
In ar ratio 1
sub _proportion
Qavg2
Inquality ch
cons
Observations:
R2
Model 1
Parameter Estimate
0.3772
0.8664**
-0.4115**
-1.3915**
-0.3673
-0.4076**
-0.7094**
0.8427**
-0.1612
1.1785**
0.5607*
1.4028**
-0.5858**
0.3327
-0.8273**
-0.0789
-0.2342
0.2331
-0.7251**
-0.2713*
-0.0340
1.0983**
-0.0147*
-1.0388
Robust Standard
Error
0.3241
0.2721
0.1308
0.2405
0.2025
0.1215
0.2008
0.1895
0.3333
0.3389
0.2154
0.2361
0.1466
0.4756
0.2124
0.1748
0.1781
0.3894
0.2389
0.1106
0.0350
0.2738
0.0063
5.3335
140
0.641
Model 2
Parameter Estimate
0.4233*
0.7735**
-0.5000**
-1.1837**
-0.2912*
-0.3902**
-0.8708**
0.7732**
-0.1507
0.5932*
0.7262**
1.5625**
-0.5403**
0.9595**
-0.6300**
-0.1738
-0.2697*
-0.0103
-0.3204
-0.4134**
-0.0573
0.6066**
-0.0041
-0.7456**
-6.1401
Robust Standard
Error
0.1991
0.1828
0.0911
0.2040
0.1238
0.1322
0.1536
0.1338
0.1800
0.2366
0.1616
0.1698
0.1133
0.3570
0.1797
0.1194
0.1223
0.2454
0.1715
0.0875
0.0292
0.2003
0.0050
0.0947
3.9505
140
0.773
September 29, 2015
H-19
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix H: Meta-Regression Models
Table H-4: Regression Results
Parameter
Weights
Random Effects:
Model 1
Parameter Estimate
Robust Standard
Error
Yes
No
Model 2
Parameter Estimate
Robust Standard
Error
Yes
No
Notes: * denotes p<.05; ** denotes p<.01. All standard errors are clustered by study.
Application of Model 2 to estimating benefits of improved water quality resulting from the final ELG requires
selection of an appropriate value for the water quality change parameter. There are several potential
hypotheses that could explain why survey respondents' implied marginal WTP appears to depend on the
overall magnitude of the change in water quality conditions.106 Overall, the meta-data do not provide any
simple way of testing or ruling out these different possibilities. This makes choosing a single value for the
water quality change parameter in Model 2 difficult. Therefore, EPA conducted a sensitivity analysis, using
the range of water quality change scenarios found in the meta-dataset (e.g., water quality change equal to +5
units and +50 units). This analysis characterizes the range of values that would be generated if EPA were to
conduct a well-designed stated preference survey to elicit marginal WTP for the types of water quality
improvements expected under the final ELG and future CWA regulations. Note that a water quality
improvement of+5 is closer to the levels of water quality improvement in the benefits transfer application
here.
H.5 Limitations and Uncertainty
The validity and reliability of benefit transfer—including that based on meta-analysis—depends on a variety
of factors. While benefit transfer can provide valid measures of use and nonuse benefits, tests of its
performance have provided mixed results (e.g., Desvousges et al. 1998; Vandenberg et al. 2001; Smith et al.
2002; Shrestha et al. 2007). Nonetheless, benefit transfers are increasingly applied as a core component of
benefit cost analyses conducted by EPA and other government agencies (Bergstrom and De Civita 1999;
Rosenberger and Phipps 2007). Moreover, Smith et al. (2002, p. 134) argue that "nearly all benefit cost
analyses rely on benefit transfers, whether they acknowledge it or not." Given the increasing [or as Smith et
al. (2002) might argue, universal] use of benefit transfers, an increasing focus is on the empirical properties of
applied transfer methods and models.
Although the statistical performance of the models is good, EPA notes several limitations applicable to both
Model 1 and Model 2. These limitations stem largely from information available from the original studies, as
well as degrees of freedom and statistical significance. An important factor in any benefit transfer is the
ability of the study site or estimated valuation equation to approximate the resource and context under which
benefit estimates are desired. As is common, the meta-analysis model presented here provides a close but not
perfect match to the context in which values are desired. Although all of the studies used in the meta-analysis
valued changes in water quality improvements, many studies did not specify the cause of water quality
impairment in the baseline or focused on causes that are different from the pollutant of concern in the
regulation (i.e., heavy metals and nutrients). Preliminary models, however, suggest no systematic patterns in
WTP associated with such factors, at least in the present meta-data.
Accounting for Scope in the MRM2 Meta-Regression Model (Abt Associates 2015).
September 29, 2015
H-20
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix H: Meta-Regression Models
Additional limitations relate to the paucity of demographic variables available for inclusion in the model. The
only demographic variable incorporated in the analysis (Inincomel) was not statistically significant. Other
demographic variables are unavailable.
The estimated model produces statistically significant coefficients and allows estimation of WTP based on
study and site characteristics. However, strictly speaking, model findings are relative to the specific case
studies considered, and must be viewed within the context of the 140-observation data set, with all the
appropriate caveats. Although this represents a fairly standard-to-large sample size for a meta-analysis in this
context (the 51 studies in the analysis gather data from tens of thousands of respondents), it is relatively small
relative to other statistical applications in resource and environmental economics. Model results are also
subject to choices regarding functional form and statistical approach, although many of the primary model
effects are robust to reasonable changes in functional form and/or statistical methods. The rationale for the
specific functional form chosen here is detailed above.
As in all cases, results of the meta-analysis are dependent on the sample of studies available for the given
resource change (Navrud and Ready 2007), and may be subject to various selection biases if the available
literature does not provide a representative, unbiased perspective on welfare estimates associated with
resource changes (Rosenberger and Johnston 2007). In this case, however, the Agency took various steps to
ameliorate such potential biases, including the incorporation of both peer-reviewed and gray literature to
avoid possible publication biases (Rosenberger and Johnston 2007), and the use of a comprehensive literature
review in the attempt to avoid—as much as possible—other types of selection biases.
September 29, 2015 H-21
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Appendix I. Impacts of Steam Electric Pollutants on Aquatic Species
Table 1-1: Common Coal-combustion Wastewater Pollutants (adapted from EPA, 2009d)
Arsenic
BOD
Boron
Cadmium
Chlorides
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nitrogen
pH
Phosphorus
Selenium
Potential Environmental Concern
Frequently observed in high concentrations in coal combustion wastewater; causes poisoning
of the liver in fish and developmental abnormalities; is associated with an increased risk of
cancer in humans in the liver and bladder.
Can cause fish kills because of a lack of available oxygen; increases the toxicity of other
pollutants, such as mercury. Has been associated with FGD wastewaters that use organic
acids for enhanced SO2 removal in the scrubber.
Frequently observed in high concentrations in coal combustion wastewater; leachate into
groundwater has exceeded state drinking water standards; human exposure to high
concentrations can cause nausea, vomiting, and diarrhea. Can be toxic to vegetation.
Elevated levels are characteristic of coal combustion wastewater-impacted systems;
organisms with elevated levels have exhibited tissue damage and organ abnormalities.
Sometimes observed at high concentrations in coal combustion wastewater (dependent on
FGD system practices); elevated levels observed in fish with liver and blood abnormalities.
Elevated levels have been observed in groundwater receiving coal combustion wastewater
leachate; invertebrates with elevated levels require more energy to support their metabolism
and therefore exhibit diminished growth.
Coal combustion wastewater can contain high levels; invertebrates with elevated levels
require more energy to support their metabolism and therefore exhibit diminished growth.
Leachate from impoundments has caused elevated concentrations in nearby surface water;
biota with elevated levels have exhibited sublethal effects including metabolic changes and
abnormalities of the liver and kidneys.
Concentrations in coal combustion wastewater are elevated initially, but lead settles out
quickly; leachate has caused groundwater to exceed state drinking water standards. Human
exposure to high concentrations of lead in drinking water can cause serious damage to the
brain, kidneys, nervous system, and red blood cells.
Coal combustion wastewater leachate has caused elevated concentrations in nearby
groundwater and surface water; biota with elevated levels have exhibited sublethal effects
including metabolic changes and abnormalities of the liver and kidneys.
Biota with elevated levels have exhibited sublethal effects including metabolic changes and
abnormalities of the liver and kidneys; can convert into methylmercury, increasing the
potential for bioaccumulation; human exposure at levels above the MCL for relatively short
periods of time can result in kidney damage.
Frequently observed at elevated levels in coal combustion wastewater; may cause
eutrophication of aquatic environments.
Acidic conditions are often observed in coal combustion wastewater; acidic conditions may
cause other coal combustion wastewater constituents to dissolve, increasing the fate and
transport potential of pollutants and increasing the potential for bioaccumulation in aquatic
organisms.
Frequently observed at elevated levels in coal combustion wastewater; may cause
eutrophication of aquatic environments.
Frequently observed at high concentrations in coal combustion wastewater; readily
bioaccumulates; elevated concentrations have caused fish kills and numerous sublethal
effects (e.g. , increased metabolic rates, decreased growth rates, reproductive failure) to
aquatic and terrestrial organisms. Short term exposure at levels above the MCL can cause
hair and fingernail changes; damage to the peripheral nervous system; fatigue and irritability
in humans. Long term exposure can result in damage to the kidney, liver, and nervous and
circulatory systems.
September 29, 2015
1-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-1: Common Coal-combustion Wastewater Pollutants (adapted from EPA, 2009d)
Compound
Total Dissolved Solids
Potential Environmental Concern
High levels are frequently observed in coal combustion wastewater; elevated levels can be a
stress on aquatic organisms with potential toxic effects; elevated levels can have impacts on
agriculture & wetlands.
Zinc Frequently observed at elevated concentrations in coal combustion wastewater; biota with
elevated levels have exhibited sublethal effects such as requiring more energy to support
their metabolism and therefore exhibiting diminished growth, and abnormalities of the liver
and kidneys.
Table I-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Acipenser brevirostrum
Acipenser medirostris
Acipenser oxyrinchus desotoi
Acipenser oxyrinchus oxyrinchus
Alasmidonta heterodon
Alasmidonta raveneliana
Amblema neislerii
Amblyopsis rosae
Ambystoma bishopi
Ambystoma cingulatum
Ambystoma macrodactylum
Ambystoma tigrinum
Ammodramus savannarum floridanus
Anguispira picta
Antrobia culveri
Antrolana lira
Aphelocoma coerulescens
Athearnia anthonyi
Batrisodes texanus
Batrisodes venyivi
Boloria acrocnema
Brachylagus idahoensis
Brachyramphus marmoratus
Brychius hungerfordi
Bufo houstonensis
Cambarus aculabrum
Campeloma decampi
Campephilus principalis
Canis lupus
Canis rufus
Charadrius melodus
Cicindela dorsalis dorsalis
Cicindela nevadica lincolniana
Cicindela puritana
Species Group
Fishes
Fishes
Fishes
Fishes
Clams
Clams
Clams
Fishes
Amphibians
Amphibians
Amphibians
Amphibians
Birds
Snails
Snails
Crustaceans
Birds
Snails
Insects
Insects
Insects
Mammals
Birds
Insects
Amphibians
Crustaceans
Snails
Birds
Mammals
Mammals
Birds
Insects
Insects
Insects
Vulnerability
High
High
High
High
High
High
High
High
Moderate
Moderate
Moderate
Moderate
Low
Low
High
High
Low
High
Low
Low
Low
Low
Moderate
High
Moderate
High
High
Low
Low
Moderate
Moderate
Moderate
Moderate
Moderate
September 29, 2015
I-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Cicurina baronia
Cicurina madia
Cicurrna"venii
Cicurina vespera
Corynorhinus (=Plecotus) townsendii ingens
Corynorhinus (=Plecotus) townsendii virginianus
Cryptobranchus alleganiensis
Cyprinella caerulea
Cyprogenia stegaria
Dendroica chrysoparia
Dendroica kirtlandii
Discus macclintocki
Dromus dramas
Drymarchon corais couperi
Elimia crenatella
Elliptic chipolaensis
Elliptic spinosa
Elliptic steinstansana
Elliptoideus sloatianus
Enhydra lutris nereis
Epioblasma brevidens
Epioblasma capsaeformis
Epioblasma florentina curtisii
Epioblasma florentina florentina
Epioblasma florentina walkeri
Epioblasma florentina walkeri (=E. walkeri)
Epioblasma metastriata
Epioblasma obliquata obliquata
Epioblasma obliquata perobliqua
Epioblasma othcaloogensis
Epioblasma torulosa rangiana
Epioblasma torulosa torulosa
Epioblasma turgidula
Etheostoma chermocki
Etheostoma chienense
Etheostoma etowahae
Etheostoma fonticola
Etheostoma nianguae
Etheostoma nuchale
Etheostoma phytophilum
Etheostoma scotti
Etheostoma sellare
Eubalaena glacialis
Eurycea nana
Species Group
Arachnids
Arachnids
Arachnids
Arachnids
Mammals
Mammals
Amphibians
Fishes
Clams
Birds
Birds
Snails
Clams
Reptiles
Snails
Clams
Clams
Clams
Clams
Mammals
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Fishes
Fishes
Fishes
Fishes
Fishes
Fishes
Fishes
Fishes
Fishes
Mammals
Amphibians
Vulnerability
Low
Low
Low
Low
Low
Low
High
High
High
Low
Low
Low
High
Low
High
High
High
High
High
Moderate
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
High
Low
High
September 29, 2015
I-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Fusconaia cor
Fusconaia cuneolus
Gambusia georgei
Gammarus acherondytes
Glaucomys sabrinus coloratus
Glaucomys sabrinus fuscus
Gopherus polyphemus
Graptemys flavimaculata
Graptemys oculifera
Gras americana
Gras canadensis pulla
Hemistena lata
Heraclides aristodemus ponceanus
Herpailuras (=Felis) yagouaroundi cacomitli
Hesperia leonardus montana
Heterelmis comalensis
Juturnia kosteri
Lampsilis abrupta
Lampsilis altilis
Lampsilis higginsii
Lampsilis perovalis
Lampsilis subangulata
Lampsilis virescens
Lasmigona decorata
Leopardus (=Felis) pardalis
Leopardus (=Felis) wiedii
Leptodea leptodon
Leptonycteris nivalis
Leptoxis ampla
Leptoxis foremani
Leptoxis plicata
Leptoxis taeniata
Lepyrium showalteri
Lioplax cyclostomaformis
Lycaeides melissa samuelis
Lynx canadensis
Margaritifera hembeli
Medionidus acutissimus
Medionidus parvulus
Medionidus penicillatus
Medionidus simpsonianus
Mesodon clarki nantahala
Mesodon magazinensis
Microhexura montivaga
Species Group
Clams
Clams
Fishes
Crustaceans
Mammals
Mammals
Reptiles
Reptiles
Reptiles
Birds
Birds
Clams
Insects
Mammals
Insects
Insects
Snails
Clams
Clams
Clams
Clams
Clams
Clams
Clams
Mammals
Mammals
Clams
Mammals
Snails
Snails
Snails
Snails
Snails
Snails
Insects
Mammals
Clams
Clams
Clams
Clams
Clams
Snails
Snails
Arachnids
Vulnerability
High
High
High
Moderate
Low
Low
Low
High
High
Moderate
Moderate
High
Low
Low
Low
High
Low
High
High
High
High
High
High
High
Low
Low
High
Low
High
High
High
High
High
High
Low
Low
High
High
High
High
High
Low
Low
Low
September 29, 2015
I-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Microtus pennsylvanicus dukecampbelli
Mustela nigripes
Mycteria-americana
Myotis grisescens
Neoleptoneta microps
Neonympha mitchellii francisci
Neonympha mitchellii mitchellii
Neotoma floridana smalli
Nicrophoras americanus
Notropis cahabae
Noturas crypticus
Noturas placidus
Obovaria retusa
Odocoileus virginianus clavium
Odocoileus virginianus leucurus
Oncorhynchus clarkii stomias
Oncorhynchus clarkii stomias
Oncorhynchus kisutch
Orcinus orca
Orthalicus reses (not incl. nesodryas)
Palaemonias alabamae
Palaemonias ganteri
Panthera onca
Pegias fabula
Percina antesella
Percina aurolineata
Percina rex
Percina tanasi
Peromyscus gossypinus allapaticola
Peromyscus polionotus ammobates
Peromyscus polionotus niveiventris
Peromyscus polionotus phasma
Phoebastria (=Diomedea) albatrus
Phoxinus cumberlandensis
Picoides borealis
Plethobasus cicatricosus
Plethobasus cooperianus
Plethodon nettingi
Pleurobema clava
Pleurobema collina
Pleurobema curtum
Pleurobema decisum
Pleurobema furvum
Pleurobema georgianum
Species Group
Mammals
Mammals
Birds
Mammals
Arachnids
Insects
Insects
Mammals
Insects
Fishes
Fishes
Fishes
Clams
Mammals
Mammals
Clams
Fishes
Fishes
Mammals
Snails
Crustaceans
Crustaceans
Mammals
Clams
Fishes
Fishes
Fishes
Fishes
Mammals
Mammals
Mammals
Mammals
Birds
Fishes
Birds
Clams
Clams
Amphibians
Clams
Clams
Clams
Clams
Clams
Clams
Vulnerability
Moderate
Low
Moderate
Moderate
Low
Low
Low
Low
Low
High
High
High
High
Moderate
Moderate
High
High
High
Low
Low
Moderate
Moderate
Low
High
High
High
High
High
Low
Low
Low
Low
Low
High
Low
High
High
Low
High
High
High
High
High
High
September 29, 2015
I-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Pleurobema hanleyianum
Pleurobema marshalli
Pleurobema perovatum
Pleurobema plenum
Pleurobema pyriforme
Pleurobema taitianum
Pleurocera foreman!
Polyborus plancus audubonii
Polygyriscus virginianus
Potamilus capax
Potamilus inflatus
Pseudemys alabamensis
Ptychobranchus greenii
Ptychocheilus lucius
Puma (=Felis) concolor coryi
Pyrgulopsis (=Marstonia) pachyta
Pyrgulopsis neomexicana
Pyrgulopsis ogmorhaphe
Pyrgulopsis roswellensis
Quadrula fragosa
Quadrula intermedia
Quadrula sparsa
Quadrula stapes
Rangifer tarandus caribou
Rhadine exilis
Rhadine infernalis
Rhadine persephone
Rhinichthys osculus thermalis
Salvelinus confluentus
Scaphirhynchus albus
Scaphirhynchus suttkusi
Sciurus niger cinereus
Somatochlora hineana
Sterna antillarum
Sternotherus depressus
Strix occidentalis caurina
Strix occidentalis lucida
Stygobromus (=Stygonectes) pecki
Stygoparnus comalensis
Succinea chittenangoensis
Sylvilagus palustris hefneri
Tartarocreagris texana
Texamaurops reddelli
Texella cokendolpheri
Species Group
Clams
Clams
Clams
Clams
Clams
Clams
Snails
Birds
Snails
Clams
Clams
Reptiles
Clams
Fishes
Mammals
Snails
Snails
Snails
Snails
Clams
Clams
Clams
Clams
Mammals
Insects
Insects
Insects
Clams
Fishes
Fishes
Fishes
Mammals
Insects
Birds
Reptiles
Birds
Birds
Crustaceans
Insects
Snails
Mammals
Arachnids
Insects
Arachnids
Vulnerability
High
High
High
High
High
High
High
Low
Low
High
High
High
High
High
Low
High
High
High
Low
High
High
High
High
Low
Low
Low
Low
High
High
High
High
Low
High
High
High
Low
Low
High
Low
Low
High
Low
Low
Low
September 29, 2015
I-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix I: Impacts on Aquatic Species
Table 1-2: T&E Species with Habitat Overlapping Waterbodies Affected by Steam Electric Power Plants
Species
Texella reddelli
Texella reyesi
Toxolasma cylindrellus
Trichechus manatus
Triodopsis platysayoides
Tryonia alamosae
Tulotoma magnifica
Tympanuchus cupido attwateri
Typhlomolge rathbuni
Ursus americanus luteolus
Vermivora bachmanii
Villosa trabalis
Xyrauchen texanus
Zapus hudsonius preblei
Species Group
Arachnids
Arachnids
Clams
Mammals
Snails
Snails
Snails
Birds
Amphibians
Mammals
Birds
Clams
Fishes
Mammals
Vulnerability
Low
Low
High
Low
Low
High
High
Low
High
Low
Moderate
High
High
Low
September 29, 2015
I-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Appendix J. Supporting Data for Impoundment Release Analysis
Table J-1: Historical Releases from EPA Survey (U.S. EPA, 2012d)
Facility
Meredosia Power
Station
Chesterfield Power
Station
Cliffside Power
Station
Bowen Power Station
Bowen Power Station
Harllee Branch
Power Station
R M Schahfer
Power Station
Sherburne County
Power Station
Sherburne County
Power Station
State
IL
VA
NC
GA
GA
GA
IN
MN
MN
Release
Year
2006
2005
2005
2002
2008
2000
1998
2007
2008
Unit
Height
(feet)
24
19
38
45
45
83
4
57
41
Storage
Capacity
(ac-ft)
700
740
3676
3676
1050
48
6198
620
Category
Big
Big
Big
Big
Big
Big
Small
Big
Big
Gallons
Released
500
35000
600
8000
Capacity
Factor
0%
U
u
U
u
0%
u
0%
0%
Release
other
other
other
other
other
other
other
other
other
Description
2006 - less than 500 gallons spilled from
the Fly Ash Pond
2005 unusual discharge
10/7/2005 the Cliffside Steam Station
experienced a significant localized flood
event. The floodwaters from the Suck
Creek entered into the retired Units 1-4
ash basin, topped the top of the dam and
washed away part of the basin's dike.
Notifications were
7/28/02 sink hole; 11 cubic yards of ash
sediment reached creek.
9/9/08 Stack Failure. Approximately 40
tons of ash left plant property and flowed
to near by residential properties and
approximately 2 tons of ash reached
Euharlee Creek
discharge of slurry 12/1/00
no details March 1998
600 Gal spill May 2007
Spring 2008, the piping used to transmit
the fine fraction of the bottom ash from
hydraulic dredging of the bottom ash pond
broke and approximately 8000 gallons of
water and as was discharged over the
Bottom Ash Pond embankment to the gro
September 29, 2015
J-1
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-1: Historical Releases from EPA Survey (U.S. EPA, 2012d)
Facility
Dave Johnston Power
Station
Naughton Power
Station
PPL Martins Creek
Power Station
Colstrip Steam
Electric Station
Colstrip Steam
Electric Station
Colstrip Steam
Electric Station
Colstrip Steam
Electric Station
Roxboro Power
Station
W. H. Weatherspoon
Power Station
Winyah Power
Station
Johnsonville Power
Station
Kingston Power
Station
Kingston Power
Station
State
WY
WY
PA
MT
MT
MT
MT
NC
NC
sc
TN
TN
TN
Release
Year
2009
2005
2003
1995
2000
2006
2008
2001
2008
2004
2003
2006
Unit
Height
(feet)
0
56
43
25
88
88
88
37.5
28
30
30
50
50
Storage
Capacity
(ac-ft)
3
3370
1085
245
4370
4370
4370
53
1375
1700
2702
8907
8907
Category
Small
Big
Big
Big
Big
Big
Big
Big
Big
Big
Big
Big
Big
Gallons
Released
14400
9249802
1E+08
2700
100
50
2000
Capacity
Factor
1%
1%
28%
0%
0%
0%
0%
U
U
U
U
U
U
Release
other
other
other
other
other
other
other
wall breach
other
other
other
other
other
Description
Jan 2009 - 14,400 gallons of process water
overflowed the canal
55,000 cu. Yds of fly ash was spilled
outside the pond boundries
2005 - Fly Ash spill (100 Million Gal)
estimated 10 acres
3/18/03 spill +/- 2700
10/11/95 small spill +/- 100 gallons C
Cell
8/29/00 small spill +/- 50 gallons C Cell
2/1/06 hole in HOPE liner C Cell +/- 2000
gallons
2008 Pond Breach
Breach of an internal dike 200 1
2/14/08 Release of wastewater - seal
failure
Reported release of small quantity of
ceneospheres on 03/27/2004 when
discharge structure was disturbed during
maintenance.
1 1/07/2003 an ash release occurred to land
from slough in the Dredge Cell
embankment.
1 1/01/2006 an ash release occurred to land
from slough in the Dredge Cell
embankment.
September 29, 2015
J-2
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-1: Historical Releases from EPA Survey (U.S. EPA, 2012d)
Facility
Kingston Power
Station
Widows Creek Power
Station
Widows Creek Power
Station
Widows Creek Power
Station
Eagle Valley
Generating Station
Eagle Valley
Generating Station
Limestone Electric
Generating Station
LaCygne Generating
Station
LaCygne Generating
Station
LaCygne Generating
Station
Riverside Generating
Station
State
TN
AL
AL
AL
IN
IN
TX
KS
KS
KS
IA
Release
Year
2008
2008
2004
2009
2008
2007
2000
2007
2007
2009
2002
Unit
Height
(feet)
50
115
115
150
38
28
0
45
45
45
15
Storage
Capacity
(ac-ft)
8907
11709
11709
10961
415
415
50
9298
9298
9298
140
Category
Big
Big
Big
Big
Big
Big
Small
Big
Big
Big
Big
Gallons
Released
1.1E+09
30000000
30000000
500
Capacity
Factor
38%
U
U
U
22%
22%
0%
U
U
U
U
Release
wall breach
other
other
other
wall breach
wall breach
other
other
other
other
other
Description
A release into the Emory River occurred
on 12/22/2008 from the Dredge Cell
embankment failure. No reports found of
releases from the Main Ash Pond or S
Reported release of small quantity of
ceneospheres 01/30/2008.
Reported release of small quantity of
ceneosperes 12/10/2004 due to intense
precipitation.
An abandoned decant weir in Pond 2B
failed on 01/09/2009
1 levee breache (1/30/2008)- State
notified. 30 million gallons spill.
1 levee breache (2/14/2007)- State
notified. 30 million gallons spill.
May 19, 2000: Approximately 500 gallons
of water was discharged resulting from a
severe rainfall event. The pH of the
discharge was 8.5 su, TSS was 88 mg/1
and Selenium was <0.010 mg/1. The
discharge ultimately made its way to Lynn
Creek. Event was repor
July 2007 Due to unusual rainfall events
Sept 2007 Due to unusual rainfall events
May 2009- Due to unusual rainfall events
April 14, 2002- Caused by Mississippi
River flooding, fixed using drilled grout
September 29, 2015
J-3
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-1: Historical Releases from EPA Survey (U.S. EPA, 2012d)
Facility
Interstate Power &
Light Co - M.L.
Kapp Generating
Station
PSCo Comanche
Station
PSCo Valmont
Station
State
IA
CO
CO
Release
Year
2009
2007
2008
Unit
Height
(feet)
10
0
0
Storage
Capacity
(ac-ft)
2
12
16
Category
Small
Small
Small
Gallons
Released
2500
4204.455
Capacity
Factor
U
0%
0%
Release
other
other
other
Description
3/13/2009, IPL reported to the Iowa DNR
an unpermitted release of water from this
pond at a location which sealed the pond
from a previous discharge channel. This
water leakage was immediately repaired
on 3/13/2009.
April 9, 2007- between 2,000-3,000
gallons spilled from a broken pipe
Feb 14, 2008 - 25 cuyd spill of bottom
ash slurry
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
Row
1
2
3
4
5
6
7
8
State
AL
AL
AK
AK
AZ
AZ
AR
CA
Original NRD
Amount
$491,976
$1,000,000,000
$600,000,000
$644,017
$1,000,000,000
$7,000,000
$6,800,000
$2,000,000
$16,300,000
Year
2006
2011
2013
1997
1989
2013*
2012
1987
1990
Updated NRD
Amount
$596,000
$1,081,352,000
$600,000,000
$1,257,000
$2,968,963,000
$7,000,000
$7,031,000
$6,898,000
$45,789,000
Case
Shelby County Train
Derailment
Deepwater Horizon
M/V Kuroshima
Exxon Valdez
ASARCO LLC
Freeport-McMoRan Corp
Morenci Mine
Vertac Chemical
Corporation Superfund
Site
American Trader Oil
Spill
Notes
soybean spill leading to damage of
aquatic life (fish, mussels, and
snails)
oil spill in Gulf of Mexico
oil spill along coastline
Exxon Valdez
three historic mining sites
hazardous substance release
dioxin release from
herbicide/pesticide plant
biological injury (effects on fish
and birds) from a 416,598 gallon
crude oil spill
Included
or
Excluded
Included
Excluded
Excluded
Excluded
Excluded
Excluded
Included
Excluded
Excluded
Reason for
Exclusion
Ocean
Ocean
Ocean
Ocean
CERCLA
CERCLA
Ocean
September 29, 2015
J-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
9
10
11
12
13
14
15
16
17
18
19
20
21
State
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
Original NRD
Amount
$14,000,000
$30,000,000
$32,300,000
$5,400,000
$10,800,000
$2,700,000
$1,400,000
$1,400,000
$7,100,000
$3,625,000
$4,820,000
$1,900,000
$3,900,000
Year
2001
2005
2007
1986
1988
1991
1992
1993
1994
1996
1997
1997
1998
Updated NRD
Amount
$22,133,000
$38,481,000
$37,469,000
$19,761,000
$34,538,000
$7,346,000
$3,596,000
$3,419,000
$16,318,000
$7,517,000
$9,405,000
$3,707,000
$7,208,000
Case
Cantara Loop/ Dunsmuir
Chemical Spill
Montrose Chemical
Cosco Busan Oil Spill
Apex Houston Oil Spill
Shell/Martinez Oil Spill
Exxon Mobil/Santa Clara
River Oil Spill
Avila I Oil Spill
McGrath Oil Spill
ARCO/Santa Clara River
Oil Spill
SS Cape Mohican Oil
Spill
M/V Kure/Humboldt Bay
Oil Spill
Torch/Platform Irene Oil
Spill
T/V Command Oil Spill
Notes
spill of 19,000 gallons of an
herbicide which damaged habitat
and fish
long-time accidental and
purposeful discharges of DDT
spill of 53,000 gallons of bunker
fuel
spill of 25,800 gallons of crude oil,
killing 10,577 birds
release of 400,000 gallons of crude
oil into creek
spill of 74,000 gallons of crude oil
spill of 24,200 gallons of crude oil
into Pacific Ocean
release of 87,150 gallons of crude
oil into McGrath Lake and the
Pacific Ocean
release of 190,000 gallons of crude
oil into 16 miles of river
96,000 gallons of intermediate fuel
oil released, with 40,000 gallons
spilling into San Francisco Bay
bunker fuel oil spill killing birds
and affecting saltmarsh, mudflats,
kayaking, camping, surfing
oil release into ocean, killing 700
birds, affecting sandy and rocky
shoreline habitat
3,000 gallons of intermediate
bunker oil discharged, damaging
birds, shoreline habitat,
recreational beach use
Included
or
Excluded
Included
Excluded
Excluded
Excluded
Included
Included
Excluded
Excluded
Included
Excluded
Excluded
Excluded
Excluded
Reason for
Exclusion
CERCLA
Ocean
Ocean
Ocean
Ocean
Ocean bay
Ocean bay
Ocean
Ocean
September 29, 2015
J-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
22
23
24
25
26
27
28
29
30
31
32
33
State
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CA
CO
Original NRD
Amount
$6,710,000
$358,000
$6,000,000
$1,150,000
$950,000
$9,000,000
$9,000,000
$6,750,000
$2,850,000
$850,000
$22,700,000
$194,000,000
Year
1999
2000
2000
2004
2013*
2013*
2013*
2013*
2013*
2009
2007
2009
Updated NRD
Amount
$11,661,000
$584,000
$9,795,000
$1,573,000
$950,000
$9,000,000
$9,000,000
$6,750,000
$2,850,000
$990,000
$26,333,000
$226,019,000
Case
M/V
Stuvyesant/Humboldt Oil
Spill
East Walker River Oil
Spill
Avila II Oil
Contamination
Kinder-Morgan/Suisun
Marsh Oil Spill
Searles Valley
Minerals/Searles Lake
(Trona)
Guadalupe Oil Field
Contamination
Iron Mountain Mine
CERCLA Site
New Almaden Mine
CERCLA Site
Chevron/Castro
Dubai Star/San Francisco
Bay Spill
S.S. Jacob Luckenbach
Oil Spill
California Gulch Site
Notes
2,000 gallons of intermediate fuel
oil spilled into Pacific Ocean,
killing 2,405 birds, damaging
coastal beaches, shrimp, fish, and
beach use
3,600 gallons of fuel oil spilled,
impacting 15 miles of river
numerous pipeline leaks caused
underground plume of oil products
spill of 70,000 gallons of diesel
into marsh
Hypersaline industrial wastewater
discharged into large ponds
pipeline leaks causing 80 plumes
of diluent
acid mine drainage over many
decades
release of mercury in Guadalupe
River watershed and south San
Francisco Bay
oil and mercury dischargers
contaminating intertidal and sub-
tidal mudflats
400 gallons of intermediate fuel oil
spilled, affecting shoreline, birds,
human uses
457,000 gallons of fuel released
from a sunken ship, killing 51,569
birds between 1990 to 2003 and
sea otters
superfund site with surface water
and habitat remediation needs
Included
or
Excluded
Excluded
Included
Excluded
Excluded
Included
Excluded
Excluded
Excluded
Included
Excluded
Excluded
Excluded
Reason for
Exclusion
Ocean
Groundwater
only
Ocean
Ocean
CERCLA
CERCLA
Ocean
Ocean
CERCLA
September 29, 2015
J-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
34
35
36
37
38
39
40
41
42
43
44
45
State
CO
CO
CT
CT
DE
FL
FL
FL
FL
GA
HI
ID
Original NRD
Amount
$27,500,000
$10,000,000
$15,000,000
$2,700,000
$27,500,000
$3,100,000
$1,200,000
$77,000
$3,600,000
$11,800,000
$5,800,000
$498,500,000
Year
1992
1994
1977
1967
2010
1999
2006*
2013*
1997
2006
2013
2009
Updated NRD
Amount
$70,639,000
$22,983,000
$120,787,000
$52,632,000
$30,881,000
$5,387,000
$1,455,000
$77,000
$7,025,000
$14,303,000
$5,800,000
$580,775,000
Case
Rocky Mountain Arsenal
Rocky Flats
Housatonic and
Connecticut Rivers
Quinnipiac River Basin
Athos I
Tampa Bay Oil Spill
Sapp Battery
Whitehouse Oil Pits
Superfund Site
Alafia River
RJ. Schlumberger
M/V Cape Flattery
Bunker Hill Mining
Superfund Site
Notes
former weapons and chemicals
manufacturing site
former nuclear weapons
manufacturer with plutonium,
uranium, volatile organic
compounds, metals, radionuclide
materials, nitrates, and asbestos
contamination
PCB contamination from longtime
discharges from a power plant
settlement funds to replace lost
drinking water source from a
landfill
discharge of 265,000 gallons into
Delaware River and tributaries
362,000 gallons of fuel oil and
other petroleum products,
damaging beaches of shellfish beds
metal contaminations in soil,
surface water, groundwater, and
wetlands from battery salvage
assessment costs and restoration of
a wetland
discharge of 50 million gallons of
process water from a
phosphogypsum stack breach
PCB contamination
oil spill damaging 20 acres of coral
reef, fish, algae, sea urchins, and
other reef animals
100 million tons of mining wastes
in a river system; wildlife habitat
damage, poisoning of birds and
other wildlife
Included
or
Excluded
Excluded
Excluded
Excluded
Excluded
Included
Excluded
Excluded
Excluded
Included
Excluded
Excluded
Excluded
Reason for
Exclusion
CERCLA
CERCLA
CERCLA
Ocean
CERCLA
CERCLA
Ocean
CERCLA
September 29, 2015
J-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
46
47
48
49
50
51
52
53
54
55
56
57
State
ID
IL
IL
IL
IL
IL
IN
IN
IN
IN
IN
IN
Original NRD
Amount
$60,000,000
$1,843,000
$450,000
$263,000
$154,648
$105,000
$86,000,000
$6,250,000
$600,000
$31,309
$200,000
$597,000
Year
1995
2011
2007
2001
2001
2013*
1998
2000
2000
1985
2009
1999
Updated NRD
Amount
$131,504,000
$1,993,000
$522,000
$416,000
$244,000
$105,000
$158,936,000
$10,203,000
$979,000
$121,000
$233,000
$1,037,000
Case
Blackbird Mine
Superfund Site
Former Indian Refinery
Saline Branch and Salt
Fork River
Marathon Oil Company
Vesuvius USA
Corporation
Williams Pipeline
Company
Grand Calumet
White River
American Chemical
Services
I. Jones Recycling, Inc.
Lakeland Disposal
Landfill
Waste Inc. Landfill
Notes
mining damage to surface water
and wildlife (Chinook salmon)
damages from various
contaminants to groundwater,
surface water, soils, and adjacent
properties
2 fish kills from sudden ammonia
releases
numerous spills of crude oil and
refined petroleum products
impacting 29 counties
spill of an industrial chemical into
3 tributaries of the Embarras River
pipeline leak of 10,000 gallons of
gasoline and diesel oil into
tributary of Salt Creek
dredging after PCB, oil, benzene,
cyanide, and heavy metal
contamination from Stell
manufacturer, plus restoration of
habitat
excessive chemical discharges
killing 4.6 million fish; note that
the settlement was for $14 million,
$6.25 million of which was for
NRD
discharge of chemicals from
storage drums
hazardous waste release
waste leakage into groundwater,
surface water, and sediments
liquid waste drainage into aquifer,
creek, and surrounding wetlands
Included
or
Excluded
Excluded
Included
Included
Included
Included
Included
Excluded
Included
Included
Included
Included
Included
Reason for
Exclusion
CERCLA
CERCLA
September 29, 2015
J-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
58
59
60
61
62
63
64
65
66
67
68
69
70
State
KS
KY
LA
ME
ME
ME
ME
MD
MD
MA
MA
MA
MA
Original NRD
Amount
$1,200,000
$2,500,000
$750,000
$1,000,000
$125,000
$930,000
$160,440
$2,700,000
$507,300
$20,200,000
$1,353,000
$157,000
$3,100,000
Year
2008
2007
2011
1996
1994
1997
2013*
2000
2007
1992
1993
1995
1998
Updated NRD
Amount
$1,369,000
$2,900,000
$811,000
$2,074,000
$287,000
$1,815,000
$160,000
$4,408,000
$588,000
$51,888,000
$3,304,000
$344,000
$5,729,000
Case
Cherokee County
Superfund Site
Russellville Plant
Calcasieu Estuary and
Bayou Verdine
Julie N Oil Spill
F. O'Connor Superfund
Site
Maine Yankee
S.D. Warren Facility
PEPCO Spill
Spectron, Inc. Superfund
Site
AVX/New Bedford
Harbor
Charles George Landfill
PSC Resources
Nyanza/ Sudbury River
Notes
runoff from zinc and lead mine
tailings that entered local streams
and contaminated groundwater
PCB pollution in groundwater,
streams, and rivers
release of hazardous substances to
soil and water, impacting assorted
benthos and other marine resources
Oil spill; 130 acres of habitat
enhancement/acquisition
groundwater damage from PCBs
and solvents
petroleum, solvents, and
radiological materials from nuclear
power plant
waste from manufacture of coated
paper which contaminated
groundwater
oil spill; restoration of wetlands,
oyster beds, waterfowl nesting
areas, and terrapin habitat
contamination and oil from site
which migrated to Little Elk Creek
water column, sediments, shellfish,
birds, anadromous fish,
recreational fishing, beach usage
landfill pollution, gases, leachate,
contamination, migratory birds,
fish
groundwater and wetlands
surface water (riverine habitat),
wetlands, fisheries, other wildlife,
recreational use
Included
or
Excluded
Excluded
Included
Excluded
Included
Excluded
Included
Excluded
Included
Excluded
Excluded
Included
Included
Included
Reason for
Exclusion
CERCLA
CERCLA
CERCLA
Groundwater
only
CERCLA
CERCLA
September 29, 2015
J-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
71
72
73
74
75
76
77
78
79
80
81
82
83
State
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
MA
Original NRD
Amount
$30,000
$19,700,000
$30,000
$500,000
$142,000
$1,300,000
$312,500
$747,000
$1,650,000
$1,094,000
$6,076,000
$875,000
$4,250,000
Year
1999
2000
2003
2004
2004
2007
2008
2010
2010
2011
2011
2012
2012
Updated NRD
Amount
$52,000
$32,159,000
$44,000
$684,000
$194,000
$1,508,000
$357,000
$839,000
$1,853,000
$1,183,000
$6,570,000
$905,000
$4,395,000
Case
Hallmark/ Mystic River
General Electric/
Housatonic River
Sulfuric Acid Spill/
North River
Coal Tar Deposits/ CT
River
Posavina Oil Spill/
Chelsea Creek
Textron Systems
Corporation/Mass
Military Reservation
Superfund Site
Global/Irving Chelsea
Creek Oil Spill
Rubchinuk Landfill Site
Sutton Brook Disposal
Area Superfund Site
Blackburn and Union
Privileges Superfund Site
Bouchard B-120
Buzzards Bay Oil Spill
GM Assembly Plant in
Framingham
Pharmacia Corp./Bayer
CropScience Superfund
Site
Notes
surface water (riverine habitat),
recreational use
ground and surface water, nesting
habitats, recreational fishing and
boating, various aquatic organisms
and birds
Various aquatic resources, aquatic
fish, amphibians, invertebrates, and
plant species
various aquatic resources,
endangered species
coastal land and habitat, salt water
vegetation, migratory birds, fish
groundwater
surface water, shoreline, wetlands,
salt marsh
community use
groundwater, biological resources
and their habitats
groundwater, biological resources
and their habitats
aquatic and shoreline, ram island
shoreline, recreation and shellfish,
and piping plovers
streambed, banks, and surrounding
wetlands; birds, wildlife, and
benthic macroinvertebrates
wetland, river, and lake habitat;
fish, turtles, amphibians, and
migratory birds
Included
or
Excluded
Included
Included
Included
Included
Excluded
Excluded
Included
Included
Excluded
Excluded
Excluded
Included
Excluded
Reason for
Exclusion
Ocean
CERCLA
CERCLA
CERCLA
Ocean bay
CERCLA
September 29, 2015
J-10
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
84
85
86
87
88
89
90
91
92
93
94
State
MI
MI
MS
MO
MO
MO
MT
MT
MT
NV
NH
Original NRD
Amount
$807,490
$26,200,000
$3,000,000
$49,000
$20,100,000
$41,200,000
$138,000,000
$169,000,000
$5,900,000
$37,000,000
$1,600,000
$859,528
$1,500,000
Year
1989
1998
1999
2013*
2009
2009
1999
2008
2009
2008
2012
2013
2005
Updated NRD
Amount
$2,397,000
$48,420,000
$5,214,000
$49,000
$23,417,000
$48,000,000
$239,823,000
$192,848,000
$6,874,000
$42,221,000
$1,654,000
$860,000
$1,924,000
Case
Verona Well Field
Saginaw River and Bay
Genesis Pipeline Spill
Cominco/Halliburton
Newton County Mine
Tailings Superfund Site
Southeast Missouri Lead
Mining District
Atlantic Richfield
Company
Mike Horse Dam
Silvertip Pipeline Oil
Spill
Rio Tinto Mine
Coakley Landfille
Superfund Site
Notes
groundwater contamination from
leaking solvents
PCB release into Saginaw River
crude oil spill; surface water,
sediments, shoreline habitat,
wildlife
seven lead and copper metal
concentrate spill sites
releases of cadmium, lead, and
zinc to groundwater, surface water,
sediments, aquatic and terrestrial
plants and organisms, aquatic
mammals, fish, aquatic and
terrestrial invertebrates, and
migratory birds
four mine sites impacting surface
water, geological resources,
groundwater, and aquatic and
terrestrial biota
decades of mining and mineral
processing releasing hazardous
substances
Dam failure in 1975 due to heavy
rains. Contaminated tailings were
washed into the Beartrap Creek
and the Upper Blackfoot River
63,000 gallons of oil spilled into
Yellowstone River
abandoned copper mine waste
disposal in Mill Creek
contamination of wetlands with
various pollutants; restoration of
338 acres of degraded saltmarsh
Included
or
Excluded
Excluded
Included
Included
Included
Excluded
Excluded
Included
Included
Included
Included
Included
Included
Excluded
Reason for
Exclusion
Groundwater
only
CERCLA
CERCLA
CERCLA
September 29, 2015
J-11
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
95
96
97
98
99
100
101
102
103
104
105
106
107
108
State
NJ
NJ
NJ
NJ
NM
NM
NM
NM
NM
NY
NY
OH
OH
OH
Original NRD
Amount
$20,000
$40,462,000
$150,000
$3,218,700
$1,000,000
$1,100,000
$13,000,000
$5,500,000
$30,000
$7,500,000
$20,300,000
$12,000,000
$13,750,000
$5,500,000
$2,040,000
Year
2013*
2013*
2009
2009
2000
2006*
2010
2012
2004
2006
2013
2006
2008
2012
2006
Updated NRD
Amount
$20,000
$40,462,000
$175,000
$3,750,000
$1,632,000
$1,333,000
$14,598,000
$5,687,000
$41,000
$9,091,000
$20,300,000
$14,545,000
$15,690,000
$5,687,000
$2,473,000
Case
In re Former Owens-
Illinois Closure Site
In re Phelps Dodge Site
In re Jimmie's Raceway
Service Station
Combe Fill South
Landfill Superfund Site
Spartan Technology Site
Albuquerque ATSF Site
Freeport McMoRan
SOHIO L-Bar Facility
State of New Mexico v
General Electric
Company et al.
St Lawrence River in
Massena
Lake Ontario System
Fernald Uranium
Products
Ashtabula River
Ohio River
Notes
contamination of groundwater
hazardous waste discharge into
groundwater
contamination of soil and
groundwater from gas station
contamination of groundwater and
nearby brook
damage to groundwater from
discarded solvents and plating
wastes
railroad tie treating plant; damages
to groundwater and wildlife habitat
damages to groundwater, wildlife,
and wildlife habitat
uranium tailings which
contaminated groundwater
contaminated groundwater from a
Superfund site
PCB contamination in a river
release of dangerous chemicals
into Lake Ontario System
uranium products (over 1,000
acres) damaging groundwater
remediation of contaminated
sediment
restoration of damage to mussels,
fish, and snails from contamination
from a metals company
Included
or
Excluded
Excluded
Excluded
Excluded
Excluded
Excluded
Included
Excluded
Included
Excluded
Excluded
Excluded
Included
Excluded
Excluded
Included
Reason for
Exclusion
Groundwater
only
Groundwater
only
Groundwater
only
CERCLA
Groundwater
only
Groundwater
only
Groundwater
only
CERCLA
CERCLA
CERCLA
September 29, 2015
J-12
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
109
110
111
112
113
114
115
116
117
118
119
State
OK
OR
OR
OR
PA
PA
RI
RI
RI
RI
SC
Original NRD
Amount
$300,000
$50,000
$100,000
$300,000
$21,000,000
$7,350,000
$8,000,000
$1,415,000
$6,000,000
$750,000
$121,000
Year
2007
2013*
2013*
2009
2009
2006
2013*
2005
2011
2013
2012
Updated NRD
Amount
$348,000
$50,000
$100,000
$350,000
$24,466,000
$8,909,000
$8,000,000
$1,815,000
$6,488,000
$750,000
$125,000
Case
Double Eagle Superfund
Site
Whitaker Slough Cleanup
Johnson Lake
Union Carbide Site
Palmerton Zinc
Superfund Site
Sinnemahoning Creek
Watershed
North Cape Oil Spill
Calf Pasture Point and
Allen Harbor Landfill
Buzzard's Bay
Davis Liquid Waste
Superfund Site
Cooper River/Charleston
Harbo
Notes
contamination of groundwater by
hazardous waste
contamination of Whitaker Slough
by electroplating wastewater
overflow from settling ponds and
stormwater discharges
Claims related to waste from
carbide and ferroallow processing
contaminating the Columbia
Slough.
injuries to aquatic and terrestrial
resources from zinc and other
metals
spill of sodium hydroxide from
train derailment, damaging several
waterbodies
828,000 gallons of home heating
oil spilled, killing at least 9 million
lobsters, thousands of birds, and
millions of clams, crabs, and fish
discharge of chemical wastes
(chlorinated hydrocarbons and
VOCs) into groundwater
oil spill injuring shoreline and
aquatic resources, piping plovers,
and recreational uses
contamination of groundwater
release of 12,500 gallons of fuel oil
affecting shoreline habitats,
sediments, migratory birds,
shellfish beds, and recreational
shrimp baiting
Included
or
Excluded
Excluded
Included
Included
Included
Excluded
Included
Excluded
Excluded
Excluded
Excluded
Included
Reason for
Exclusion
CERCLA
CERCLA
Ocean
Groundwater
only
Ocean bay
CERCLA
September 29, 2015
J-13
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
120
121
122
123
124
125
126
127
128
129
130
State
SD
TN
TN
TX
UT
UT
UT
VA
VA
WA
WA
Original NRD
Amount
$4,000,000
$543,203
$50,000
$3,120,000
$37,000,000
$2,580,000
$3,500,000
$3,700,000
$2,500,000
$5,000,000
$5,500,000
Year
2006*
2002
2010
2012
1995
2007
2011
2003
2001
2006*
2005
Updated NRD
Amount
$4,848,000
$831,000
$56,000
$3,226,000
$81,094,000
$2,993,000
$3,785,000
$5,399,000
$3,952,000
$6,061,000
$7,055,000
Case
South Dakota v
Homestake Mining
Company
Obed Wild and Scenic
River Site
U.S. Department of
Energy's Oak Ridge
Reservation
Malone Service Co.
Disposal Facility
Southwest Jordan Valley
Ensign-Bickford Trojan
Facility
Red Butte Creek Oil Spill
Tazewell County Spill
Powell River
Elliot Bay/ Duwamish
River
Skykomish Facility
Notes
damage to surface and
groundwater from metals (from
mining)
oil spill and fire in tributaries of
the Obed Wild and Scenic River
damages to fishing from release of
hazardous substances and
radioactive compounds
hazardous waste material that
contaminated groundwater and
migrated to Galveston Bay
$28 million in restoration of
groundwater, damaged by historic
mining, $9 in compensation
discharges from explosives
manufacturing facility into
groundwater (3 mile plume)
pipeline rupture releasing oil into
creek and other waters
1,300 gallons of a rubber
accelerant, which damaged
endangered species of freshwater
mussels
six million gallons of coal slurry
release to river watershed (20
miles downstream), impacting fish,
endangered mussels, aquatic
habitat, bats, and migratory birds
habitat development and
restoration after damage from
sewer discharges
discharge of diesel fueil which
leaked to water table
Included
or
Excluded
Included
Included
Included
Excluded
Included
Excluded
Included
Included
Included
Included
Included
Reason for
Exclusion
CERC LA
Groundwater
only
September 29, 2015
J-14
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs
Appendix J: Impoundment Release Analysis Supporting Data
Table J-2: 137 NRD Settlements Documented by Israel for 1967 to 2013 (Israel 2006; 2013)
_
Row
131
132
133
134
135
136
137
State
WA
WA
WV
WV
WI
WI
WY
Original NRD
Amount
$9,000,000
$512,857
$2,040,000
$500,000
$35,000,000
$1,900,000
$50,000
Year
1994
2008
2006
2009
2013*
2012
2013*
Updated NRD
Amount
$20,685,000
$585,000
$2,473,000
$583,000
$35,000,000
$1,965,000
$50,000
Case
Tenyo Mara
Crystal Mountan
Emergency Generation
Facility
Ohio River
Consol Energy /Dunkard
Creek
Fox River/Green Bay
Ashland Lakefront
Gasoline Spill
Notes
release of 354,800 gallones of fuel
oil and 97,800 gallons of diesel
fuel, affecting coastal waters and
shorelines and killing 4,300
seabirds
release of 18,200 gallons of diesel
fuel into creek
restoration of native freshwater
mussels, snails, and fish in Ohio
River
discharge of mining wastewater
containing chloride, resulting in an
algae bloom that killed thousands
offish, mussels, and amphibians
contamination from PCB and other
discharges from paper mills
contaminated sediments
restocking a surface water after a
fish kills from a gasoline spill
Included
or
Excluded
Excluded
Included
Included
Included
Excluded
Included
Included
Reason for
Exclusion
Ocean
CERCLA
September 29, 2015
J-15
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
Appendix K. Methodology for Benefits from Avoided Dredging Costs
The following subsections describe EPA's methodology for estimating sediment dredging benefits by
applying the unit dredging costs to changes in the volume of sediment deposited to navigable waterways and
reservoirs under the five regulatory options. The section is organized as follows:
> Review and analysis of historical dredging data including dredging locations, intervals, and costs;
> Estimation of sediment deposition to dredged waterways and reservoirs using SPARROW; and
> Estimation of dredging costs savings under regulatory options based on changes in sediment
deposition and unit cost of dredging.
EPA notes that the methodology presented here focuses solely on avoided dredging costs based on the volume
of sediment. Due to data limitations, the analysis does not quantify environmental benefits of reducing the
frequency of maintenance dredging, nor does it account for potentially lower costs associated with disposing
of dredged material spoils due to reduced discharge of toxic metals.
K.1 Review and Analysis of Historical Dredging
EPA used data from the United States Army Corps of Engineers (USAGE) Dredging Information System
(USAGE, 2013) to analyze baseline navigational dredging activity and to estimate the cost per cubic yard of
sediment dredged. The system catalogs all USAGE dredging contracts and USAGE-conducted dredging jobs
from 1998 to 2012, including the location of dredging activities, start and stop dates, the volume of sediment
dredged, and the cost of the dredging job. The system does not report separate cost components, but costs
typically include (Sohngen & Rausch, 1998):
> Cost of dredging sediment from the waterway's channel bed and loading onto a boat,
> Cost of transporting dredged material to a disposal facility, and
> Cost of confining or disposing of the dredged material.
EPA reviewed available information about sediment removal methods at existing U.S. reservoirs. The review
indicated that dredging is a practical and common approach to sediment removal in existing reservoirs. With
the exception of draining a reservoir and excavating settled sediment (which is typically more expensive and
less common than dredging), dredging is the only feasible option for sediment removal in many reservoirs
(Morris and Fan, 1997). Other methods to counteract sedimentation or reclaim capacity at existing reservoirs
include:
> Sediment routing - a group of techniques that allow sediment-laden water to pass around or through a
reservoir without allowing the sediment to settle, optimized to address sediment-laden flows from
events such as storms and floods (Morris and Fan, 1997). EPA was not able to find any information
on the prevalence or effectiveness of these techniques.107
> Flushing - the scouring of accumulated sediment from a reservoir by partially or fully draining it and
allowing the erosive force of the draining water to carry sediment through and downstream of the
reservoir. The practice does not appear to be very common in the United States as it creates
Including sediment pools during initial construction and maintaining these pools can also reduce sediment
deposition (Crowder 1987).
September 29, 2015 RT
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
extraordinarily high sediment concentrations downstream of the reservoir,108 and the reservoir may
still require dredging (Morris and Fan, 1997).
Theoretically, reductions in sediment loadings could delay the construction of new reservoir capacity and
associated construction costs, including land for the reservoir itself, embankments, mitigation lands,
appurtenances to the dam (if applicable), such as pump stations, and pipelines to connect the reservoir to the
treatment plant or raw water users. In practice, it may not be feasible to construct new capacity in many cases
due to space constraints or adverse ecological effects associated with reservoir construction. Furthermore, the
ELG regulatory options result in only marginal changes in sediment deposition to existing reservoirs that are
unlikely to be of sufficient magnitude to influence decision-making regarding capacity expansion.
Based on this review, EPA focused its benefits analysis for reservoirs on dredging over other capacity
reclamation approaches due to the frequent use and broad feasibility of dredging. Given likely similarities in
cost components for navigational dredging and reservoir dredging, EPA used regional estimates of cost per
cubic yard of sediment dredged based on the USAGE Dredging Information System as inputs for its analysis
of avoided reservoir dredging costs. The Agency recognizes that dredging costs are highly variable and driven
my factors including costs for dewatering sites, weather, topography, and characteristics of bottom sediments,
and that at some sites dredging costs could exceed the costs of other alternatives, potentially including the unit
cost of constructing new capacity.109 EPA notes that the volume of sediment that can be dredged from
reservoirs may also affected by the availability of funds or nearby disposal sites (Morris & Fan, 1997).
K.1.1 Dredging Location, Recurrence Interval, and Dredging Volumes
Each observation in the Dredging Information System corresponds to a given date range and location at which
dredging occurred, referred to here as "dredging occurrence." Many locations have multiple dredging
occurrences because recurrent dredging is may be necessary to maintain navigability. EPA uses the term
"dredging job" to refer to multiple dredging occurrences at the same location. For each dredging job, EPA
identified:
> The number of occurrences from 1998 to 2012. EPA merged dredging occurrences at the same
location that were less than 30 days apart because these may be continuations of the same dredging
occurrence, rather than a new dredging occurrence. If determined to be a continuation of prior
occurrence, EPA summed the volume of sediment dredged and costs of the records to generate a
single observation in the database.
> Average dredging volume from 1998 to 2012. EPA divided the total quantity of sediment dredged for
each job (single location) over the past 15 years by the number of occurrences of that job to calculate
an average quantity of sediment dredged for an occurrence of that job. EPA assumed that this quantity
of sediment would be dredged each time the job occurs in the future under the baseline scenario, and
that it would be reduced due to the final ELGs
Concentration can sometimes exceed 1,000,000 mg/L and thus may require special permissions (Morris and
Fan, 1997)
109 Alan Plummer Associates (2005) compared the cost of dredging to the costs of constructing additional
reservoir capacity for reservoirs in Texas and found that dredging unit costs are at least twice that of securing
storage in new reservoirs. New reservoir costs are approximately $1 for each cubic yard of water stored in the
conservation pool (2011$, updated based on the construction cost index), whereas dredging costs for large-sized
dredging projects are approximately $2 per cubic yard (2011$, updated based on the construction cost index)
and, depending on variability, could cost more than two times that amount.
September 29, 2015 K^2~
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
> The time elapsed between dredging occurrences. EPA calculated an average frequency of recurrence
for each dredging job in years by dividing 15 (the number of data years, 1998-2012) by the number of
occurrences of that job.
The number of dredging jobs in the system varies by region, and is likely to be influenced by the size of the
region, the number of navigable waterways, and their economic importance.
Where the USAGE Dredging Information System did not contain cost or quantity of sediment dredged for
a listed dredging occurrence, EPA estimated the missing information from other jobs. The Agency
calculated 10th, 50th, and 90thpercentile costs and quantity boundaries and used these to fill in all
incomplete records in each EPA region. EPA used the 10th and 90th percentile cost and amount dredged
estimates in the low and high total dredging cost estimates, respectively. The cost data were adjusted to 2013$
using the construction cost index.
K.1.2 Determining Affected Navigational Dredging Jobs and Unit Costs
For this analysis, EPA mapped each navigational dredging job with latitude-longitude coordinates to the
nearest reach modeled in SPARROW (E2RF1 reaches), within a one-mile radius. For jobs for which latitude-
longitude coordinates were not provided, EPA used alternate job location information such as the name of the
job (usually the waterway dredged) and the USAGE district that performed the job.110 EPA excluded dredging
jobs that are greater than 1 mile from a SPARROW reach from the analysis.111
This approach matched 29 unique dredging jobs and 98 dredging occurrences to analyzed E2RF1 reaches,
equivalent to 0.8 percent of the dredging occurrences with coordinates reported in the Dredging Information
System. The remaining 99 percent of occurrences were either greater than 1 mile from a modeled reach. Table
K-l summarizes dredging jobs and recurrence intervals in the affected reaches. The recurrence interval ranges
from 1 to 15 years across all affected reaches, with an average of 9.6 years.
Table K-2 summarizes the average cost of dredging. Costs vary considerably across affected reaches, from
approximately $$1.59 per cubic yard at Establishment Bar in North Carolina to $28.08 per cubic yard at
Bonum Creek in Virginia. The average unit cost of dredging for the entire coterminous United States is
$$5.56 per cubic yard.
110 EPA reviewed to the data to identify cases where latitude-longitude coordinates appear to have been
misreported based on district and reach information. Where possible, EPA populated missing coordinates by
interpolating from other occurrences of the same dredging job.
111 To identify the nearest reach segment, EPA used an unprojected version of the ERF 1.2 (Enhanced Reach File)
from USGS. For this analysis, EPA researched latitude-longitude coordinates for jobs where they were not
provided and linked them to Reach File Version 1.0 (RF1) reaches. Each latitude/longitude of interest was
matched to the nearest point in the ERF 1.2 universe of points using a spherical model of the earth and a
standard haversine distance formula. No reach types were excluded from consideration in the nearest reach
calculation
September 29, 2015 K^
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
table K-1: Navigational Dredging Jobs and Recurrence Intervals from 1998 to 2012 within Affected
eaches
Number of Affected
Dredging Jobs
29
Number of
Recurring
Jobs
15
Number of Single
Occurrence Jobs
15
Average Interval for
Single Occurrence and
Recurring Jobs (years)
9.6
Average Interval
for Recurring
Jobs (years)
3.7
Note: Includes dredging jobs with latitude-longitude coordinates that could be mapped to affected E2RF1 reaches.
Source: Dredging Information System (USAGE, 2013).
Table K-2: Historic Dredging Costs from 1998 to 2012 within Affected Reaches
Number of
Affected Reaches
29
Total Sediment
Removed (millions of
cubic yards)
72.0
Total Cost (millions of
2013$)
$754.0
Average Cost per cubic
yard (2013$)
$10.5
Note: Only includes jobs EPA was able to map to affected reaches.
Source: Dredging Information System (USAGE, 2013).
K.1.3 Determining Reservoir Dredging Locations
EPA relied upon the "reservoir" flags within the E2RF1 dataset to identify reservoir locations. SPARROW
models reservoirs as impoundments located on the main reach network. For the purposes of this analysis,
EPA assumed, consistent with Crowder (1987), that all sediment entering reservoirs must be removed in order
to maintain current water storage capacity.
K.2 Sediment Deposition in Navigable Waterways and Reservoirs
EPA estimated annual sediment deposition to waterways using sedimentation outputs from SPARROW. EPA
assumed that all sediment deposited within historically dredged waterways will be dredged at some future
point. Likewise, EPA assumes that all sediment deposited in the affected reaches with reservoirs will also be
dredged.
As described in Chapter 4, the SPARROW outputs reflect changes in annual sediment deposition based on
the changes in sediment loadings under the final ELGs. EPA assumed that reduced deposition starts in 2021,
which is the midpoint of the period during which steam electric plants are expected to implement changes to
meet the final rule limitations and standards.
K.3 Estimating Dredging Costs under Baseline and Regulatory Options
Benefits under the final ELGs are calculated for each year from 2021 to 2042 as the difference between
baseline dredging costs and post-compliance dredging costs. Each waterway or reservoir will have unique
benefits based on sediment deposition, dredging frequency, and the unit cost of dredging. Dredging costs only
occur in years when the waterbody or reservoir is projected to be dredged.
K.3.1 Estimating Annualized Dredging Costs for Navigational Waterbodies
For each dredging job location, EPA assumes that future dredging occurrences will happen at the same
frequency as in the past. The volume of sediment removed in each dredging occurrence is equal to the amount
September 29, 2015
K-4
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
of sediment deposited in that reach since the prior occurrence at that job location. For example, if dredging
occurs every two years at given job location, then the first analyzed occurrence project for the second year of
the analysis period (2022). The volume of sediment dredged in that occurrence is the sum of sediment
deposited in the current year and prior year (2021 and 2022). The dredging cost incurred in 2022 is the
product of the cubic yards dredged and unit cost per cubic yard. By the end of the 2042, the reach will have
been dredged ten times.
Equation J-l presents the calculation of annual dredging costs for navigational waterways for the baseline and
regulatory options. EPA discounted the costs following the analytic framework described in Chapter 1 .
17 T 1 /IT/ V/ /<2&tX(l-fl)XC\ „ , / dX(l + d)" \
Equation J-l AV = Z ( (1+d)t_2015 ) X ((1+d)n+1_J
Where:
AV = Annualized value
/ = The index for the dredging occurrence. The number of occurrences is
based on the total number of periods (22) and the recurrence interval for
the dredging job. The recurrence interval is calculated by dividing 15
years by the number of occurrences of the job in the USAGE dredging
data.
Qb = Cubic yards of dredged materials under the baseline scenario in a given
yeart
T = Year of dredging occurrence
R = Percentage of cubic yards of dredge material remaining under the
regulatory scenario relative to baseline
C = Cost per cubic yard of sediment dredged for the dredging job location
(2013$)
D = Annual discount rate. EPA used both 3 percent and 1 percent, in
accordance with OMB guidance (Office of Management and Budget,
2003)
N = Number of periods for annualization (24 years for this benefits analysis)
EPA conducted a sensitivity analysis for navigational dredging by varying assumptions for projected future
dredging occurrence, generating low, medium, and high estimates for navigational dredging:
> For medium and high estimates, EPA assumed that single occurrence dredging jobs occur once every
15 years (i.e., the number of data years). Single occurrence jobs are dropped from future projections
for the low estimate.
> For low and medium estimates, dredging is interpreted as occurring at the end of each interval (i.e.,
2022 in the example described above). For the high estimate, EPA assumes that dredging occurs at
the beginning of the interval, rather than the end (e.g., 2021). This second approach tends to generate
greater discounted benefits because jobs occur sooner and are discounted less.
> EPA varied costs for jobs lacking actual cost values and dredging volumes in the Dredging
Information System. EPA assigned 10th, 50th, and 90th percentile costs for low, mean, and high
estimates, respectively.
> For the mean estimate, EPA assumed a minimum recurrence interval of 90 days. Occurrences that
began less than 90 days from the end of the prior occurrence at the same job location were considered
September 29, 2015 K-5
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
to be continuation, and combined in the data. EPA used intervals of 30 and 180 days for the low and
high estimates, respectively.
These assumptions are summarized in Table K-3.
Table K-3: Parameters Used in Sensitivity Analysis of Avoided Dredging Costs
Parameter
Value
Low
Mean
High
Treatment of Single
Occurrences
Excluded
Included
Included
Treatment of Interval
Start
End
End
Percentile of Cost
Estimates Used for
Jobs Lacking Cost
Data
10th Percentile
50th Percentile
90th Percentile
Minimum Recurrence
Interval
30 days
90 days
180 days
K.3.2 Estimating Annualized Dredging Costs for Reservoirs
The frequency of reservoir dredging is highly site-specific, depending on many factors including the average
sediment concentration of the influent river or stream, the flow regime, the size of the reservoir and excess
storage capacity, and any sediment routing practices. For this analysis, EPA chose a general frequency of
reservoir dredging based on information presented by the USAGE in a Final Dredged Material Management
Plan and Environmental Impact Statement for reservoirs in Washington (USAGE, 2002). The report states
that "dredging cycles may vary from 2 to 10 years" (USAGE, 2002, p. 66). EPA used these frequencies as
high and low estimates and 6 years as a mean estimate. This approach provides a range of benefits estimates
to account for uncertainty in the frequency of reservoir dredging.
EPA was unable to identify a comprehensive source of cost data for reservoir dredging.112 The Agency used
the average unit cost of dredging from the analysis of USAGE Dredging Information System Data grouped by
EPA region. Because this cost is given per cubic yard, the sediment attenuation in reservoirs given by
SPARROW will be converted from kilograms to cubic yards using a sediment density of 1.5 g/cms (Hargrove
2007). This translates to a conversion of 1,147 kilograms per cubic yard. Equation K-2 summarizes the
calculation of annualized avoided costs for an affected reservoir. The total annualized avoided cost is the sum
of annualized cost savings at all affected reservoirs.
Equation K-2: ACr =
ixC
dx(i+d)"
Where:
AC
r
I
Qb
Q*
1,147
Annualized avoided cost of dredging all sediment settling in
reservoir r
Reservoir reach ID number
The assumed interval in years of reservoir dredging; varied
between 2, 6, and 10
Quantity of sediment present at baseline (kg)
Quantity of sediment present under the regulatory option (kg)
Number of kilograms per cubic yard of sediment
112 Some limited information on the costs of reservoir dredging is available in the literature. For example, Crowder
(1987) provides a unit cost but provides not empirical basis for the estimate.
September 29, 2015
K-6
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
C
r
t
d
n
Regional average of historic dredging job cost per cubic yard,
1998-2012.
ELG regulatory option (Option A - Option E)
Year of the dredging occurrence
Annual discount rate. EPA used both 3 percent and 7 percent.
Number of periods for annualization (24 years for the benefits
analysis)
Benefits under each regulatory option are equal to avoided costs, calculated as the difference in total
annualized dredging costs at baseline and under regulatory options. The subsections below summarize
navigational dredging and reservoir dredging benefits.
K.4.1 Navigational Dredging Benefits
Table K-4 presents estimates of baseline sediment dredging from 2021 to 2042 and low, mean, and high cost
estimates. Total baseline navigational dredging costs range from $ 36 thousands to $48 thousands per year,
using a 3 percent discount rate, and between $29 to $42 thousands using a 7 percent discount rate. Table K-5
presents estimated benefits for navigational dredging for the five regulatory options. Annualized benefits for
Option D are less than one thousand, with both 3 and 7 percent discount rates.
Table K-4: Annualized Dredging Costs at Affected Reaches under the Baseline (Thousands of 2013$)
Total Sediment Dredged
(millions cubic yards)
Low
96.4
Mean
97.3
High
125.5
Costs at 3% discount rate
(thousands of 2013$ per year)
Low
36.2
Mean
36.5
High
47.9
Costs at 7% discount rate
(thousands of 2013$ per year)
Low
29.2
Mean
29.4
High
42.1
Source: EPA analysis, 2015.
Table K-5: Annualized Benefits from Reduced Dredging Costs (Thousands of 2013$)
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Total Reduction in Sediment
Dredged (Baseline - Option;
cubic yards)
Low
0.0
0.0
763.8
1,496.7
3,122.7
Mean
0.0
0.0
819.6
1,735.0
3,361.5
High
0.0
0.0
2,606.4
6,648.3
8,530.1
3% discount rate
(thousands of 2013$ per year)
Low
0.0
0.0
0.4
0.8
1.5
Mean
0.0
0.0
0.4
0.8
1.5
High
0.0
0.0
0.9
2.2
3.0
7% discount rate
(thousands of 2013$ per year)
Low
0.0
0.0
0.3
0.6
1.1
Mean
0.0
0.0
0.3
0.6
1.2
High
0.0
0.0
0.9
2.0
2.7
Source: EPA analysis, 2015.
K.4.2 Reservoir Dredging Benefits
Table K-6 presents the total amount of sediment that is estimated to be dredged in 2021 from reservoirs, and
the estimated annualized cost of dredging under the baseline scenario, including low, mean, and high
estimates. Estimated dredging costs for the reservoirs range between $477.6 million and $577.8 million with a
September 29, 2015
K-7
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
three percent discount rate and $344.4 million and $480.0 million with a seven percent discount rate under the
baseline scenario.
Table K-6: Annualized Reservoir Dredging Costs under the Baseline (Millions 2013$)
Total Sediment Dredged
(millions cubic yards) (2021-
2042)
Low
2,934.6
Mean
2,641.2
High
3,228.1
3% Discount Rate ($/year)
Low
477.6
Mean
469.5
High
577.8
7% Discount Rate ($/year)
Low
344.4
Mean
378.9
High
480.0
Source: EPA analysis, 2015.
The difference between the anticipated dredging costs under the baseline and a particular regulatory option
represents the avoided costs of that regulatory option. Table K-7 presents reductions in sedimentation and
subsequent avoided costs from reduced reservoir dredging for each regulatory option, including low, mean,
and high estimates under these regulatory options. Because the range of estimates is relatively small between
the low and high estimates, the values presented below in the discussion below are mean estimates unless
otherwise stated.
Avoided costs from a reduction in reservoir sedimentation vary depending on the regulatory option, the
assumed frequency of reservoir dredging, and the discount rate. Annualized benefits for Option D are less
than two thousand dollars with both 3 and 7 percent discount rates.
Table K-7: Total Annualized Benefits of Reduced Reservoir Dredging (2013$)
Regulatory
Option
Option A
Option B
Option C
Option D
Option E
Total Reduction in Sediment
Dredged (cubic yards) (2021-
2040)a
Low
0.0
0.0
8,106.0
9,624.1
11,114.0
Mean
0.0
0.0
7,295.4
8,661.7
10,002.6
High
0.0
0.0
8,916.6
10,586.5
12,225.4
3% Discount Rate (thousands
of 2013$ per year)
Low
0.0
0.0
1.6
1.9
2.2
Mean
0.0
0.0
1.6
1.9
2.2
High
0.0
0.0
2.0
2.3
2.7
7% Discount Rate thousands of
2013$ per year)
Low
0.0
0.0
1.2
1.4
1.6
Mean
0.0
0.0
1.3
1.5
1.8
High
0.0
0.0
1.6
1.9
2.2
Source: EPA analysis, 2015.
K.5 Limitations and Uncertainties
Table K-8 summarizes key uncertainties and limitations for the analysis of sediment dredging benefits. Note
that the SPARROW model used to estimate sediment deposition also has a number of limitations, described
in Environmental Impact and Benefits Assessment for Final Effluent Guidelines and Standards for the
Construction and Development Category (U.S. EPA 2009c).
September 29, 2015
K-8
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
Table K-8. Limitations and Uncertainties in the Analysis of Avoided Costs of Navigational Waterway
Dredging
Issue
Lack' of standardized job
names in USAGE Database
Analysis restricted to jobs
reported in USAGE
Database for 1998 to 2012
Lack of latitude and
longitude data in USAGE
Database
Do not estimate benefits to
waterways that are not
dredged
Omission of reservoirs
located off E2RF1 channels
Lack of site specific data on
sediment density
Lack of data on the
frequency of reservoir
dredging and the amount of
sediment dredged
Assumption that excess
sediment is removed from
reservoirs by dredging
rather than building new
reservoir capacity
Effect on Benefits
Estimate
Underestimate
Underestimate
Underestimate
Underestimate
Underestimate
Uncertain
Uncertain
Uncertain
Notes
The USAGE dredging database identifies dredging jobs by
name, usually the name of the dredged waterbody, but lacks
standardized naming conventions. It is possible that the same
waterbody is dredged under different job names. For the low
cost estimates, this may result in the exclusion of dredging job
names that only appear once in the database, but in fact were
carried out in the same waterbodies as a differently named job.
This effect would tend to underestimate benefits in EPA's low
estimates.
The USAGE database is limited to USAGE dredging contracts
from 1998 to 2012. It does not capture dredging jobs contracted
by other organizations or jobs occurring before 1998 or after
2012.
Many dredging occurrences lack or have incomplete latitude
coordinates. As a result, EPA was only able to map about 71
percent of all dredging occurrences with records in the data.
EPA did not attempt to use other methods, such as Google
Earth, to map dredging locations to the E2RF1 reaches due to
resource constraints. The result is a downward bias in benefits
estimates because the analysis excludes some dredging jobs that
may benefit under the final rule.
EPA's dredging analysis is limited to navigable waterways that
have been dredged in the past and reservoirs flagged within the
E2RF1 dataset. Other waterbodies not identified by these data
could require dredging in the future and benefit from sediment
reductions under the proposed rule. Thus, EPA's estimates may
be underestimated.
The benefits analysis for modeled watersheds explicitly omits
any reservoirs that are not located on the E2RF1 network. The
omission of other reservoirs is likely to bias estimated reservoir
dredging benefits downward.
EPA used a single sediment density estimate to convert between
sediment mass and volume. This may reduce the accuracy of
resulting benefits estimates because sediment density is related
soil type in the area and is not uniform. The direction of this
potential bias is uncertain.
There is significant uncertainty as to the types of reservoirs that
are dredged and the unit cost of dredging. The appropriateness
of benefits estimates for reservoirs is conditional on assumption
that water storage volume will be maintained. Actual benefits
may diverge from estimates presented here if reservoirs take
response actions other than dredging to address sediment loads.
EPA's analysis of reservoir benefits assumes that all excess
sediment is removed by dredging. Reservoir capacity could be
replaced with new reservoirs in some cases. The unit cost of
constructing new reservoirs may be higher or lower than
dredging depending on site-specific characteristics.
September 29, 2015
K-9
-------
Benefit and Cost Analysis for Steam Electric Power Generating ELGs Appendix K: Avoided Dredging Methodology
Table K-8. Limitations and Uncertainties in the Analysis of Avoided Costs of Navigational Waterway
Dredging
Issue
Omission of natural water
storage facilities from the
analysis
Effect on Benefits
Estimate
Underestimate
Notes
The SPARROW model does not take into account sediment
build-up in natural water storage facilities such as glacial lakes
and ponds. Any activity to mitigate sedimentation in these
waterbodies is not included in this benefits analysis. This may
bias benefits estimates downward.
September 29, 2015
K-10
------- |