CPA EPA Document No.
W EP A-822-P-23 -001
DRAFT FOR PUBLIC COMMENT
Economic Analysis for the Proposed Per- and
Polyfluoroalkyl Substances National Primary
Drinking Water Regulation
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Economic Analysis for the Proposed Per- and Polyfluoroalkyl Substances
National Primary Drinking Water Regulation
Prepared by:
U.S. Environmental Protection Agency
Office of Water
Office of Ground Water and Drinking Water
Standards and Risk Management Division
Washington, DC 20460
EPA Document Number: EPA-822-P-23-001
MARCH 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Disclaimer
This document has been reviewed in accordance with EPA policy and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
Proposed PFAS Rule Economic Analysis
i
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Contents
1 Executive Summary 1-1
2 Introduction 2-1
2.1 Summary of the Proposed PFAS Rule and Regulatory Alternatives 2-2
2.2 Economic Analysis Assumptions 2-3
2.2.1 Compliance Schedule and Period of Analysis for Proposed Rule 2-3
2.2.2 Dollar Year and Discount Rates 2-3
2.2.3 Annualization 2-3
2.2.4 Population 2-4
2.2.5 Valuation 2-4
2.3 Document Organization 2-4
2.4 Supporting Documentation 2-5
3 Need for the Rule 3-1
3.1 Previous EPA Nonregulatory and Regulatory Actions Potentially Affecting PFAS
Drinking Water Management 3-1
3.1.1 PFAS Strategic Roadmap and PFAS Council 3-1
3.1.2 EPA PFAS Health Advisories 3-1
3.1.3 Final Regulatory Determinations on the Fourth Drinking Water Contaminant
Candidate List 3-2
3.1.4 Unregulated Contaminant Monitoring Rule 3-3
3.2 Statutory Authority for Promulgating the Rule 3-3
3.3 Economic Rationale 3-4
4 Baseline Drinking Water System Conditions 4-1
4.1 Introduction 4-1
4.2 Data Sources 4-1
4.2.1 SDWIS/Fed 2021 4-2
4.2.2 Third Unregulated Contaminant Monitoring Rule 4-4
4.2.3 Independent State Sampling Programs 4-5
4.2.4 Six-Year Review Data 4-5
4.2.5 Geometries and Characteristics of Public Water Systems (2000) 4-5
4.2.6 Community Water System Survey (2006) 4-6
4.3 Drinking Water System Baseline/Industry Profile 4-6
4.3.1 Water System Inventory 4-6
4.3.2 Population and Households Served 4-9
4.3.3 Treatment Plant Characterization/Production Profile 4-11
4.3.4 Public Water System Labor Rates 4-15
4.3.5 Cost of Capital 4-17
4.4 Occurrence of PFAS 4-19
4.4.1 Overview of UCMR 3 Data 4-19
Proposed PFAS Rule Economic Analysis ii March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
4.4.2 Overview of State PFAS Data 4-20
4.4.3 Overview of PFAS Co-Occurrence 4-21
4.4.4 Summary of PFAS Occurrence Data Analysis 4-21
4.4.5 Summary of National PFAS Occurrence 4-24
4.5 Uncertainties in the Baseline and Compliance Characteristics of Systems 4-35
5 Cost Analysis 5-1
5.1 Introduction 5-1
5.1.1 Chapter Overview 5-1
5.1.2 Uncertainty Characterization 5-1
5.1.3 Summary of Quantified National Cost Estimates of the Proposed Rule 5-2
5.2 Overview of SafeWater Multi-Contaminant Benefit Cost Model (MCBC) 5-6
5.2.1 Modeling PWS Variability in SafeWater MCBC 5-7
5.3 Estimating Public Water System Costs 5-9
5.3.1 PWS Treatment Costs 5-9
5.3.2 Estimating PWS Sampling and Administrative Costs 5-30
5.4 Estimating Primacy Agency Costs 5-36
5.5 PWS Level Cost Estimates 5-38
5.6 Household-Level Cost Estimates 5-38
5.7 Discussion of Data Limitations and Uncertainty 5-39
6 Benefits Analysis 6-1
6.1 Introduction 6-1
6.1.1 Chapter Overview 6-1
6.1.2 Uncertainty Characterization 6-2
6.1.3 Summary of Quantified National Benefits Estimates of the Proposed Rule 6-3
6.1.4 Life Table Modeling Background 6-5
6.2 Overview of Benefit Categories 6-6
6.2.1 Availability of Pharmacokinetic (PK) Models 6-12
6.2.2 Benefits of PFOA and PFOS Exposure Reduction 6-12
6.2.3 Summary of Health Information Considered in the Economic Analysis 6-21
6.2.4 Nonquantifiable Benefits of PFAS in Proposed Rule and PFAS Expected to be Co-
Removed 6-22
6.2.5 Sensitive Populations 6-26
6.2.6 Co-Removal of Additional Contaminants 6-27
6.3 Blood Serum Concentration Modeling for PFAS 6-28
6.3.1 Introduction 6-28
6.3.2 Application of PK Models to Benefits Analyses 6-28
6.3.3 Contributions from Other Sources 6-30
6.4 Developmental Effects 6-31
6.4.1 Overview of the Birth Weight Risk Reduction Analysis 6-31
6.4.2 Estimation of Birth Weight Changes Between Baseline and Regulatory
Alternatives 6-34
Proposed PFAS Rule Economic Analysis
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.4.3 Estimation of Birth Weight Impacts 6-36
6.4.4 Valuation of Reduced Birth Weight Impacts 6-44
6.4.5 Results 6-48
6.5 Cardiovascular Disease 6-49
6.5.1 Overview of the Cardiovascular Disease Risk Analysis 6-49
6.5.2 Cardiovascular Disease Exposure-Response Analyses 6-52
6.5.3 Estimation of Cardiovascular Disease Risk Reductions 6-56
6.5.4 Valuation of Cardiovascular Disease Risk Reductions 6-66
6.5.5 Results 6-68
6.6 Renal Cell Carcinoma 6-70
6.6.1 Overview of the RCC Risk Reduction Analysis 6-70
6.6.3 RCC Exposure-Response Modeling 6-72
6.6.3 Estimation of RCC Risk Reductions 6-73
6.6.4 Valuation of RCC Risk Reductions 6-74
6.6.5 Results 6-76
6.7 Benefits from Co-Removal of Disinfection Byproducts 6-78
6.7.1 Overview of Reduced Disinfection Byproduct Formation 6-79
6.7.2 Estimation of Bladder Cancer Risk Reductions 6-100
6.7.3 Results 6-106
6.8 Limitations and Uncertainties of the Benefits Analysis 6-107
7 Comparison of Costs to Benefits 7-1
8 Environmental Justice Analysis 8-12
8.1 Introduction 8-12
8.2 Literature Review 8-13
8.2.1 Methods 8-13
8.2.2 Findings 8-13
8.2.3 Discussion and Limitations 8-16
8.3 EJ PFAS Exposure Analysis 8-17
8.3.1 Data Sources and Approach 8-18
8.3.2 EJ Exposure Analysis Results 8-26
8.4 SafeWater EJ Analysis of Proposed Regulatory Option and Alternatives 8-59
8.4.1 Methodology 8-59
8.4.2 SafeWater EJ Analysis Results 8-61
8.5 Conclusions 8-70
8.5.1 EJ PFAS Exposure Analysis 8-70
8.5.2 SafeWater EJ Analysis of Regulatory Options 8-72
8.5.3 Overall Environmental Justice Conclusion 8-72
9 Statutory and Administrative Requirements 9-1
9.1 Executive Order 12866: Regulatory Planning and Review and Executive Order
13563: Improving Regulation and Regulatory Review 9-1
Proposed PFAS Rule Economic Analysis iv March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
9.2 Paperwork Reduction Act 9-2
9.2.1 Primacy Agency Activities 9-2
9.2.2 Public Water System Activities 9-2
9.3 The Initial Regulatory Flexibility Analysis 9-4
9.3.1 Need for, Objectives, and Legal Basis of the Rule 9-5
9.3.2 Identification of Relevant Federal Rules 9-6
9.3.3 Summary of the SBAR Comments and Recommendations 9-6
9.3.4 Number and Description of Small Entities Affected 9-8
9.3.5 Description of Compliance Requirements of the Proposed Rule 9-9
9.3.6 Analysis of Impact of Regulatory Options on Small System Costs 9-10
9.3.7 Analysis of Significant Alternatives to the Proposed Rule 9-12
9.4 Unfunded Mandates Reform Act 9-13
9.5 Executive Order 13132: Federalism 9-14
9.6 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments 9-15
9.7 Executive Order 13045: Protection of Children from Environmental Health and
Safety Risks 9-16
9.8 Executive Order 13211: Actions That Significantly Affect Energy Supply,
Distribution, or Use 9-17
9.8.1 Energy Supply 9-17
9.8.2 Energy Distribution 9-17
9.8.3 Energy Use 9-18
9.9 National Technology Transfer and Advancement Act 9-18
9.10 Executive Order 12898: Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations, Executive Order 14008: Tackling
the Climate Crisis at Home and Abroad 9-18
9.11 Consultations with the Science Advisory Board, National Drinking Water Council,
and the Secretary of Health and Human Services 9-19
9.11.1 Science Advisory Board 9-19
9.11.2 National Drinking Water Advisory Council 9-19
9.11.3 Secretary of Health and Human Services 9-19
9.12 Affordability Analyses 9-19
9.12.1 National Small System Affordability Determination 9-20
9.12.2 Supplemental Affordability Analyses 9-23
10 References 10-1
Proposed PFAS Rule Economic Analysis
v
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Tables
Table 4-1: Data Sources Used to Develop the Water System Characteristics 4-1
Table 4-2: Inventory of CWSs 4-7
Table 4-3: Inventory of NTNCWSs 4-8
Table 4-4: Population and Number of Households Served by CWSs 4-10
Table 4-5: Population Served by NTNCWSs 4-10
Table 4-6: Frequency Distribution of Entry Point Inputs for CWSs 4-13
Table 4-7: Frequency Distribution of Entry Point Inputs for NTNCWSs 4-13
Table 4-8: Functions for Design and Average Daily Flow by System Types 4-14
Table 4-9: Design and Average Daily Flow for CWSs 4-15
Table 4-10: Design and Average Daily Flow for NTNCWSs 4-15
Table 4-11: Hourly Wage Rates Based on CWSS Data ($2007) 4-16
Table 4-12: Hourly Labor Costs Including Wages Plus Benefits ($2007) 4-17
Table 4-13: Hourly Labor Costs Escalated to $2021 4-17
Table 4-14: Weighted Average Cost of Capital by PWS Ownership and Size Category 4-18
Table 4-15: Non-Targeted State PFAS Finished Water Data - Summary of Samples with
Detections of PFAS Proposed for Regulation 4-20
Table 4-16: Non-Targeted State PFAS Finished Water Data - Summary of Systems with
Detections of Select PFAS 4-21
Table 4-17: State PFAS Regulations 4-23
Table 4-18: Total Systems Impacted, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt
and HI of 1.0) 4-25
Table 4-19: Total Systems Impacted, Option la (PFOA and PFOS MCLs of 4.0 ppt) 4-25
Table 4-20: Total Systems Impacted, Option lb (PFOA and PFOS MCLs of 5.0 ppt) 4-26
Table 4-21: Total Systems Impacted, Option lc (PFOA and PFOS MCLs of 10.0 ppt) 4-26
Table 4-22: Total Entry Points Impacted, Proposed Option (PFOA and PFOS MCLs of 4.0
ppt and HI of 1.0) 4-27
Table 4-23: Total Entry Points Impacted, Option la (PFOA and PFOS MCLs of 4.0 ppt) 4-28
Table 4-24: Total Entry Points Impacted, Option lb (PFOA and PFOS MCLs of 5.0 ppt) 4-28
Table 4-25: Total Entry Points Impacted, Option lc (PFOA and PFOS MCLs of 10.0 ppt).... 4-29
Proposed PFAS Rule Economic Analysis
vi
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-26: Total Population at PWSs Impacted, Proposed Option (PFOA and PFOS MCLs
of 4.0 ppt and HI of 1.0) 4-30
Table 4-27: Total Population at PWSs Impacted, Option la (PFOA and PFOS MCLs of 4.0
ppt) 4-30
Table 4-28: Total Population at PWSs Impacted, Option lb (PFOA and PFOS MCLs of 5.0
ppt) 4-31
Table 4-29: Total Population at PWSs Impacted, Option lc (PFOA and PFOS MCLs of
10.0 ppt) 4-32
Table 4-30: Total Population at Entry Points Impacted, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0) 4-32
Table 4-31: Total Population at Entry Points Impacted, Option la (PFOA and PFOS MCLs
of 4.0 ppt) 4-33
Table 4-32: Total Population at Entry Points Impacted, Option lb (PFOA and PFOS MCLs
of 5.0 ppt) 4-34
Table 4-33: Total Population at Entry Points Impacted, Option lc (PFOA and PFOS MCLs
of 10.0 ppt) 4-34
Table 4-34: Limitations and Uncertainties That Apply to the Baseline Analysis for the
Proposed PFAS Rule 4-35
Table 5-1: Quantified Sources of Uncertainty in Cost Estimates 5-1
Table 5-2: National Annualized Costs, Proposed Option (PFOA and PFOS MCLs of 4.0
ppt and HI of 1.0; Million $2021) 5-3
Table 5-3: National Annualized Costs, Option la (PFOA and PFOS MCLs of 4.0 ppt;
Million $2021) 5-4
Table 5-4: National Annualized Costs, Option lb (PFOA and PFOS MCLs of 5.0 ppt;
Million $2021) 5-5
Table 5-5: National Annualized Costs, Option lc (PFOA and PFOS MCLs of 10.0 ppt;
Million $2021) 5-6
Table 5-6: Model PWS Variability Characteristics and Data Sources 5-8
Table 5-7: Frequency Distribution to Estimate Influent TOC in mg/L 5-11
Table 5-8: Initial Compliance Forecast Including POU RO 5-12
Table 5-9: Initial Compliance Forecast Excluding POU RO 5-13
Table 5-10: Estimated Parameter Values for Technology-Specific Bed Life Equations 5-14
Table 5-11: Cost Elements Included in All WBS Models 5-19
Table 5-12: Technology-Specific Cost Elements Included in the GAC Model 5-21
Proposed PFAS Rule Economic Analysis
vii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-13: Technology-Specific Cost Elements Included in the PFAS-Selective IX Model.. 5-23
Table 5-14: Technology-Specific Cost Elements Included in the RO/NF Model 5-24
Table 5-15: Technology-Specific Cost Elements Included in the Non-Treatment Model 5-25
Table 5-16: Implementation Administration Startup Costs ($2021) 5-31
Table 5-17: Initial and Long-Term Sampling Frequencies Per System Entry Point 5-33
Table 5-18: Sampling Costs ($2021) 5-34
Table 5-19: Treatment Administration Costs ($2021) 5-35
Table 5-20: Public Notification Burden Estimate 5-36
Table 5-21: Primacy Agency Costs ($2021) 5-37
Table 5-22: Limitations that Apply to the Cost Analysis for the Proposed PFAS Rule 5-39
Table 6-1: Quantified Sources of Uncertainty in Benefits Estimates 6-2
Table 6-2: National Annualized Benefits, Proposed Option (PFOA and PFOS MCLs of 4.0
ppt and HI of 1.0; Million $2021) 6-3
Table 6-3: National Annualized Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt;
Million $2021) 6-4
Table 6-4: National Annualized Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt;
Million $2021) 6-4
Table 6-5: National Annualized Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt;
Million $2021) 6-5
Table 6-6: Overview of Health Benefits Categories Considered in the Analysis of Changes
in PFAS Drinking Water Levels 6-8
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on
Health Outcomes 6-10
Table 6-8: Summary of Studies Relating PFOA or PFOS to Birth Weight 6-34
Table 6-9: Serum Exposure-Birth Weight Response Estimates 6-35
Table 6-10: Race/Ethnicity- and Gestational Age-Specific Birth Weight Marginal Effects
and Odds Ratios from the Mortality Regression Models 6-39
Table 6-11: Simulated Cost Changes for Birth Weight Increases ($2021) (Based on Klein
and Lynch, 2018 Table 8) 6-46
Table 6-12: National Birth Weight Benefits, Proposed Option (PFOA and PFOS MCLs of
4.0 ppt and HI of 1.0) 6-48
Table 6-13: National Birth Weight Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt). 6-48
Table 6-14: National Birth Weight Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt). 6-49
Proposed PFAS Rule Economic Analysis
viii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-15: National Birth Weight Benefits, Option lc (PFOA and PFOS MCLs of 10.0
ppt) 6-49
Table 6-16: Studies Selected for Inclusion in the Meta-Analyses 6-54
Table 6-17: Estimated Shares of Fatal and Non-Fatal First Hard CVD Events Based on
MEPS and HCUP Data 6-63
Table 6-18: Estimated Risk of Post-Acute CVD Mortality Following the First Non-Fatal
Hard CVD Event 6-66
Table 6-19: Cost of Illness of Non-Fatal First CVD Event Used in Modeling 6-67
Table 6-20: National CVD Benefits, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt
and HI of 1.0) 6-68
Table 6-21: National CVD Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt) 6-68
Table 6-22: National CVD Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt) 6-69
Table 6-23: National CVD Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt) 6-69
Table 6-24: RCC Morbidity Valuation 6-76
Table 6-25: National RCC Benefits, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt
and HI of 1.0) 6-77
Table 6-26: National RCC Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt) 6-77
Table 6-27: National RCC Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt) 6-78
Table 6-28: National RCC Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt) 6-78
Table 6-29: Data Sources and How the Information Derived from each Source is Used in
the DBP Co-Removal Analysis 6-79
Table 6-30: DBP ICR (1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Summary of
Raw Water TOC Annual System Means for Ground Water Systems 6-84
Table 6-31: DBP ICR (1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Summary of
Raw Water TOC Annual System Means for Surface Water Systems 6-84
Table 6-32: SYR3 ICR (2011) and SYR4 ICR (2019) - Summary of Finished Water TOC
Annual System Means for Ground Water Systems 6-85
Table 6-33: SYR3 ICR (2011) and SYR4 ICR (2019) - Summary of Finished Water TOC
Annual System Means for Surface Water Systems 6-85
Table 6-34: DBP ICR (Aux 1; 1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Finished
Water Annual System Mean TOC; Common Surface Water Systems 6-86
Table 6-35: Summary of THM4 Baseline Comparing DBP ICR and SYR4 ICR 6-86
Table 6-36: DBP ICR (Aux 1) Summary of THM4 Concentrations Based on Disinfectant
and Source Water Type 6-88
Proposed PFAS Rule Economic Analysis
ix
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-37: TOC Reduction for All Waters (Both Surface Water and Ground Water) with
GAC EBCT of 20 Min and a 2-year Replacement Time 6-93
Table 6-38: Estimation of ATHM4 in Surface Water with a 20 Min EBCT, and a 2-year
GAC Replacement Time 6-94
Table 6-39: Estimation of ATHM4 in Ground Water with a 20 Min EBCT, and a 2-year
GAC Replacement Time 6-94
Table 6-40: Selected Distribution Systems from SYR4 Based on Outlined Criteria 6-96
Table 6-41: Information on Selected Distribution System and Corresponding ATHM4
Values 6-98
Table 6-42: Comparison Between ICR TSD Conservative ATHM4 and SYR4 ATHM4 for
Surface Water Systems 6-99
Table 6-43: Bladder Cancer Morbidity Valuation 6-105
Table 6-44: National Bladder Cancer Benefits, Proposed Option (PFOA and PFOS MCLs
of 4.0 ppt and HI of 1.0) 6-106
Table 6-45: National Bladder Cancer Benefits, Option la (PFOA and PFOS MCLs of 4.0
ppt) 6-106
Table 6-46: National Bladder Cancer Benefits, Option lb (PFOA and PFOS MCLs of 5.0
ppt) 6-107
Table 6-47: National Bladder Cancer Benefits, Option lc (PFOA and PFOS MCLs of 10.0
ppt) 6-107
Table 6-48: Limitations and Uncertainties that Apply to Benefits Analyses Considered for
the Proposed PFAS Rule 6-108
Table 6-49: Limitations and Uncertainties in the PK Model Application 6-112
Table 6-50: Limitations and Uncertainties in the Analysis of Birth Weight Benefits Under
the Proposed Rule 6-113
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule 6-116
Table 6-52: Limitations and Uncertainties in the Analysis of RCC Benefits Under the
Proposed Rule 6-121
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule 6-123
Table 7-1: Annualized Quantified National Costs and Benefits, Proposed Option (PFOA
and PFOS MCLs of 4.0 ppt and HI of 1.0; Million $2021) 7-2
Table 7-2: Annualized Quantified National Costs and Benefits, Option la (PFOA and
PFOS MCLs of 4.0 ppt; Million $2021) 7-3
Proposed PFAS Rule Economic Analysis
x
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-3: Annualized Quantified National Costs and Benefits, Option lb (PFOA and
PFOS MCLs of 5.0 ppt; Million $2021) 7-3
Table 7-4: Annualized Quantified National Costs and Benefits, Option lc (PFOA and
PFOS MCLs of 10.0 ppt; Million $2021) 7-4
Table 7-5: Summary of Quantified and Nonquantified Benefits and Costs 7-5
Table 7-6: Potential Impact of Nonquantifiable Benefits and Costs 7-7
Table 8-1: Categorizing of PWSs Based on Data Availability for PFAS Occurrence and
PWS Service Area Boundaries 8-19
Table 8-2: Data Sources for Predelineated PWS Service Areas 8-21
Table 8-3: Number of Category 1 and 2 PWSs and Populations Served by Size and State 8-28
Table 8-4: Population Served by Category 1 and 2 PWSs Compared to Percent of U.S.
Population by Demographic Group 8-30
Table 8-5: Baseline Scenario: Population Served by Category 1 and 2 PWS Service Areas
Above Baseline Thresholds and as a Percent of Total Population Served 8-34
Table 8-6: Modeled Average PFAS Concentrations (ppt) by Demographic Group in the
Baseline, Category 1 and 2 PWS Service Areas 8-35
Table 8-7: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above UCMR 5 MRLs and as
a Percent of Total Population Served 8-38
Table 8-8: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a
Hypothetical Regulatory Scenario with Maximum Contaminant Level at the
UCMR 5 MRLs, Category 1 and 2 PWS Service Areas 8-39
Table 8-9: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above 10.0 ppt and as a
Percent of Total Population Served 8-41
Table 8-10: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a
Hypothetical Regulatory Scenario with Maximum Contaminant Level at 10.0
ppt, Category 1 and 2 PWS Service Areas 8-42
Table 8-11: Population Served by Category 1 and 2 PWSs and Percent of U.S. Population
by Demographic Group, Large Systems 8-44
Table 8-12: Population Served by Category 1 and 2 PWSs and Percent of U.S. Population
by Demographic Group, Small Systems 8-44
Table 8-13: Baseline Scenario: Demographic Breakdown of Population Served by Category
1 and 2 PWS Service Areas Above Baseline Thresholds and as a Percent of
Total Population Served, Large Systems 8-47
Proposed PFAS Rule Economic Analysis
xi
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-14: Baseline Scenario: Demographic Breakdown of Population Served by Category
1 and 2 PWS Service Areas Above Baseline Thresholds and as a Percent of
Total Population Served, Small Systems 8-48
Table 8-15: Modeled Average PFAS Concentrations (ppt) by Demographic Group and
System Size in the Baseline, Category 1 and 2 PWS Service Areas 8-49
Table 8-16: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above UCMR 5 MRLs and as
a Percent of Total Population Served, Large Systems 8-52
Table 8-17: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above UCMR 5 MRLs and as
a Percent of Total Population Served, Small Systems 8-53
Table 8-18: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a
Hypothetical Regulatory Scenario with Maximum Contaminant Levels at the
UCMR 5 MRLs, Category 1 and 2 PWS Service Areas 8-54
Table 8-19: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above 10.0 ppt and as a
Percent of Total Population Served, Large Systems 8-56
Table 8-20: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population
Served by Category 1 and 2 PWS Service Areas Above 10.0 ppt and as a
Percent of Total Population Served, Small Systems 8-57
Table 8-21: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a
Hypothetical Regulatory Scenario with Maximum Contaminant Levels at 10.0
ppt, Category 1 and 2 PWS Service Areas 8-58
Table 8-22: Annualized Cases Avoided per 100,000 People by Race/Ethnicity Group,
Proposed Option (PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0) 8-63
Table 8-23: Annualized Cases Avoided per 100,000 People by Race/Ethnicity Group,
Option la (PFOA and PFOS MCLs of 4.0 ppt) 8-63
Table 8-24: Annualized Cases Avoided per 100,000 People by Race/Ethnicity Group,
Option lb (PFOA and PFOS MCLs of 5.0 ppt) 8-64
Table 8-25: Annualized Cases Avoided per 100,000 People by Race/Ethnicity Group,
Option lc (PFOA and PFOS MCLs of 10.0 ppt) 8-64
Table 8-26: Annualized Population Weighted Household Cost by PWS Size Category and
Race/Ethnicity Group ($2021) 8-67
Table 8-27: Annualized Population-Weighted Household Cost for Treating PWSs by Size
Category and Race/Ethnicity Group 8-69
Table 9-1: Average Annual Burden, Costs, and Responses for the Proposed Rule
Information Collection Request 9-3
Table 9-2: Total Burden, Costs, and Responses for Each Required Activity 9-4
Proposed PFAS Rule Economic Analysis
xii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 9-3: Inventory of Small CWSs 9-9
Table 9-4: Inventory of Small NTNCWSs 9-9
Table 9-5: Cost-Revenue Ratio for Small CWSs, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0; Commercial Cost of Capital) 9-11
Table 9-6: Annual Incremental Costs by PWS Size and Ownership, Proposed Option
(PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0; Million $2021, Commercial
Cost of Capital) 9-14
Table 9-7: SSCT Affordability Analysis Results - Technologies that Meet Effectiveness 9-21
Table 9-8: Expenditure Margins for SSCT Affordability Analysis 9-22
Table 9-9: Total Annual Cost per Household for Candidate Technologies 9-22
Table 9-10: Total Annual Cost per Household Assuming Hazardous Waste Disposal 9-23
Table 9-11: Potential Annual Expenditure Margins for SSCT Affordability Analysis 9-25
Table 9-12: Affordability Analysis Results Using a 1.0% of Annual Median Household
Income Expenditure Margin 9-25
Table 9-13: Affordability Analysis Results Using a 2.5% of Lowest Quintile of Annual
Household Income Expenditure Margin 9-26
Table 9-14: Annual Cost per Household for Candidate Technologies Assuming 100%
Financial Assistance for Technology Capital Costs 9-29
Table 9-15: Affordability Analysis Results Using a 2.5% of Annual Median Household
Income Minus the Baseline Median Annual Drinking Water Cost Expenditure
Margin and assuming 100% Financial Assistance for Technology Capital Costs.. 9-30
Table 9-16: Affordability Analysis Results Using a 1.0% of Annual Median Household
Income Expenditure Margin and assuming 100% Financial Assistance for
Technology Capital Costs 9-31
Table 9-17: Affordability Analysis Results Using a 2.5% of Lowest Quintile of Annual
Household Income Expenditure Margin and assuming 100% Financial
Assistance for Technology Capital Costs 9-31
Proposed PFAS Rule Economic Analysis
xiii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Figures
Figure 5-1: Approach Used by SafeWater MCBC to Model PWS Variability 5-9
Figure 6-1: Overview of Analysis of Birth Weight-Related Benefits 6-33
Figure 6-2: Comparison of Change in Incidence of Infant Death per 1 g Increase in birth
weight by Gestational Age Category and Race/Ethnicity (Deaths per 1,000
Births) 6-39
Figure 6-3: Weighted Mortality Odds Ratios Based on Populations of Infants Falling into
100 g Birth Weight Increments and Four Gestational Age Categories 6-41
Figure 6-4: Piecewise Medical Cost Function Calculated by Klein and Lynch (2018) for
Three Increments in Increased Birth Weight (18 g, 50 g, and 100 g) 6-45
Figure 6-5. Interpolated Cost of Illness at Baseline Average Birth Weights, by Estimated
Change in Birth Weight Under the Proposed Rule 6-47
Figure 6-6: Overview of the CVD Risk Model 6-51
Figure 6-7: Overview of Life Table Calculations in the CVD Model 6-58
Figure 6-8: CVD Model Calculations for Ages 40+ Tracking CVD 6-60
Figure 6-9: Overview of Analysis of Reduced RCC Risk 6-71
Figure 6-10: Overview of Analysis of Co-Removal Benefits 6-82
Figure 6-11: Estimated TOC Percent Removal in Ground Water Using GAC Based on
Logistic Equation Model 6-91
Figure 6-12: Estimated TOC Percent Removal in Surface Water Using GAC Based on
Logistic Equation Model 6-92
Proposed PFAS Rule Economic Analysis
xiv
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Acronyms and Abbreviations
ACS
American Community Survey
AFFF
Aqueous Film Forming Foam
AIX
Anion Exchange
ALT
Alanine Transaminase
ANSI
American National Standards Institute
AOC
Assimilable Organic Carbon
ARIC
Atherosclerosis Risk in Communities
ATSDR
Agency for Toxic Substances and Disease Registry
AWWA
American Water Works Association
BAT
Best Available Technology
BenMAP-CE
EPA's Enviromnental Benefits Mapping and Analysis Program - Community Edition
BIL
Bipartisan Infrastructure Law
BLS
Bureau of Labor Statistics
BP
Blood Pressure
BV
Bed Volumes
CARDIA
Coronary Artery Risk Development in Young Adults
CBX
SafeWater Cost Benefit Model
CCL
Contaminant Candidate List
CCR
Consumer Confidence Report
CCRs
Consumer Confidence Reports
CDC
Centers For Disease Control And Prevention
CFR
Code of Federal Regulations
CHMS
Canadian Health Measures Survey
COBRA
Co-Benefits Risk Assessment
COI
Cost of Illness
CVD
Cardiovascular Disease
CWSs
Community Water Systems
CWSS
Community Water System Survey
DBP
Disinfection Byproduct
DHS
Department of Homeland Security
DL
Detection Limits
DWSRF
Drinking Water State Revolving Fund
EBCT
Empty Bed Contact Time
ECEC
Employer Cost for Employee Compensation
ECI
Employment Cost Index
ECTT
Error Code Tracking Tool
EJ
Enviromnental Justice
EP
Entry Point
EPA
U.S. Enviromnental Protection Agency
Proposed PFAS Rule Economic Analysis
xv
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
GAC
Granular Activated Carbon
GDP
Gross Domestic Product
gfd
Gallons per Square Foot Per Day
gpm
Gallons per Minute
GWUDI
Ground Water Under the Direct Influence
HA
Health Advisory
HDLC
High-Density Lipoprotein Cholesterol
HECD
Health and Ecological Criteria Division
HESDs
Health Effects Support Documents
HFPO-DA
Hexafluoropropylene Oxide Dimer Acid
HHS
Department of Health and Human Services
HI
Hazard Index
HRL
Health Reference Level
HRRCA
Health Risk Reduction and Cost Analysis
HTN
Hypertension
HUC
Hydraulic Unit Code
ICR
Information Collection Request
ICR TSD
Information Collection Rule Treatment Study Database
IS
Ischemic Stroke
IX
Ion Exchange
LBW
Low Birth Weight
LDLC
Low-Density Lipoprotein Cholesterol
LRAA
Locational Running Annual Average
MCBC
Multi-Contaminant Benefit-Cost Model
MCLGs
Maximum Contaminant Level Goals
MCLs
Maximum Contaminant Levels
MCMC
Markov Chain Monte Carlo
MetS
Metabolic Syndrome
MGD
Million Gallons Per Day
MI
Myocardial Infarction
MRL
Minimum Reporting Level
NBW
Normal Birth Weight
NCHS
National Center for Health Statistics
NCWSs
Non-Community Water Systems
NDWAC
National Drinking Water Advisory Council
NF
Nanofiltration
NHANES
National Health and Nutrition Examination Survey
NOM
Natural Organic Matter
NPDWR
National Primary Drinking Water Regulation
NTNCWSs
Non-Transient Non-Community Water Systems
NTTAA
National Technology Transfer and Advancement Act
Proposed PFAS Rule Economic Analysis
xvi
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
O&M
Operation and Maintenance
OEHHA
California Environmental Protection Agency's Office of Environmental Health Hazard
Assessment
OES
Occupational Employment Survey
OMB
Office of Management and Budget
ORD
Office of Research and Development
OSHA
Occupational Safety and Health Administration
PAF
Population Attributable Fraction
PBPK
Physiological-Based Pharmacokinetic
PFAA
Perfhiorinated Alkyl Acids
PFAS
Per- And Polyfluoroalkyl Substances
PFBS
Perfluorobutanesulfonic Acid
PFHpA
Perfluoroheptanoic Acid
PFHxS
Perfluorohexanesulfonic Acid
PFNA
Perfhiorononanoic Acid
PFOA
Perfluorooctanoic Acid
PFOS
Perfluorooctane Sulfonate
PK
Pharmacokinetic
POURO
Point of Use Reverse Osmosis
PRA
Paperwork Reduction Act
PWS
Public Water System
PWSID
Public Water System Identifier
PWSS
Public Water System Supervision
Q
Design Flow
QA/QC
Quality Assurance/Quality Control
RCC
Renal Cell Carcinoma
RFA
Regulatory Flexibility Act
RfD
Chronic Oral Reference Dose
RO
Reverse Osmosis/ Nanofiltration
RO/NF
Reverse Osmosis
ROB
Risk of Bias
RSC
Relative Source Contribution
RSSCT
Rapid Small-Scale Column Test
SAB
Science Advisory Board
SBAR
Small Business Advocacy Review
SBREFA
Small Business Regulatory Enforcement Fairness Act
SDWA
Safe Drinking Water Act
SDWIS
Safe Drinking Water Information System
SEER
Surveillance, Epidemiology, And End Results
SER
Small Entity Representatives
SGA
Small for Gestational Age
SMF
Standardized Monitoring Framework
Proposed PFAS Rule Economic Analysis
xvii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
soc
Synthetic Organic Chemicals
SSCTs
Small System Compliance Technologies
T&C
Technologies and Costs
T3
T riiodo thy ro nine
T4
Thyroxine
TC
Total Cholesterol
TDP
Technology Design Panel
THM4
Four Regulated Trihalomethanes
TNCWSs
Transient Non-Community Water Systems
TOC
Total Organic Carbon
TRI
Volatile Organic Compounds
TSH
Thyroid Stimulating Hormone
UCMR
Unregulated Contaminant Monitoring Rule
UCMR3
Third Unregulated Contaminant Monitoring Rule
UCMR 4
Fourth Unregulated Contaminant Monitoring Rule
UMRA
Unfunded Mandates Reform Act
VOCs
Volatile Organic Compounds
VSL
Value of a Statistical Life
WBS
Work Breakdown Structure
WIFIA
Water Infrastructure Finance and Innovation
WIIN
Water Infrastructure Improvements for the Nation
Proposed PFAS Rule Economic Analysis
xviii
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
1 Executive Summary
Under the Safe Drinking Water Act (SDWA), U.S. Environmental Protection Agency (EPA or
"the Agency") has the authority to set enforceable National Primary Drinking Water Regulations
(NPDWRs) for drinking water contaminants and require monitoring of public water supplies.
EPA is proposing aNPDWR for per-and polyfluoroalkyl substances (PFAS) (EPA-HQ-OW-
2022-0114). The Agency initiated the process for developing a NPDWR for PFAS compounds in
March 2021, when EPA published the fourth regulatory determinations for contaminants on the
fourth Contaminant Candidate List (CCL), which included a final determination to regulate
perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) in drinking water.
Additionally, in EPA's final regulatory determination for PFOA and PFOS, as well as its PFAS
Strategic Roadmap, the Agency committed to evaluating additional PFAS beyond PFOA and
PFOS and considering actions to address groups of PFAS. The proposed NPDWR is one of
several actions consistent with the Agency's commitment to address these long-lasting "forever
chemicals" that occur in drinking water supplies and impact communities across the U.S.
The proposed PFAS NPDWR is a significant regulatory action that was submitted to the Office
of Management and Budget (OMB) for review. An economic analysis (EA) is required for all
significant rules under Executive Order (EO) 12866 (Regulatory Planning and Review). In
addition, Section 1412(b)(3)(C) of the 1996 Amendments to the SDWA requires EPA to prepare
a Health Risk Reduction and Cost Analysis (HRRCA) in support of any NPDWRs that include a
maximum containment level (MCL). This EA addresses these and other regulatory reporting
requirements, including those that direct EPA to conduct distributional and environmental justice
analysis. With respect to the SDWA HRRCA requirements, this document provides the
following:
• Quantifiable and nonquantifiable health risk reduction benefits for which there is a
factual basis in the rulemaking record to conclude that such benefits are likely to occur as
the result of compliance with each level of treatment (Chapter 6);
• Quantifiable and nonquantifiable health risk reduction benefits for which there is a
factual basis in the rulemaking record to conclude that such benefits are likely to occur
from reductions in co-occurring contaminants that may be attributed solely to compliance
with the MCL, excluding benefits resulting from compliance with other proposed or
promulgated regulations (Chapter 6);
• Quantifiable and nonquantifiable costs for which there is a factual basis in the rulemaking
record to conclude that such costs are likely to occur solely as a result of compliance with
the MCL, including monitoring, treatment, and other costs, and excluding costs resulting
from compliance with other proposed or promulgated regulations (Chapter 5);
• Incremental costs and benefits associated with each alternative MCL considered (Chapter
V);
• Effects of the contaminant on the general population and on groups within the general
population, such as sub-populations identified as likely to be at greater risk of adverse
Proposed PFAS Rule Economic Analysis
1-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
health effects due to exposure to contaminants in drinking water than the general
population (Chapters 6 and 8);
• Any increased health risk that may occur as the result of compliance, including risks
associated with co-occurring contaminants (Chapter 6); and
• Other relevant factors, including the quality and extent of the information, the
uncertainties in the analysis, and factors with respect to the degree and nature of the risk
(Chapters 5-7).
Upon final rule promulgation and implementation, the proposed NPDWR would reduce PFAS
concentrations in the drinking water distributed by public water systems (PWSs) from the current
baseline to drinking water concentrations that are in compliance with MCLs of 4 parts per trillion
(ppt; also expressed as ng/L) for PFOA, 4.0 ppt for PFOS, and a unitless hazard index (HI) of 1.0
for the group including perfluorononanoic acid (PFNA), HFPO-DA (hexafluoropropylene oxide
dimer acid) and its ammonium salt (also known as GenX chemicals)1, perfluorohexanesulfonic
acid (PFHxS), and perfluorobutanesulfonic acid (PFBS). These impacts are assessed in
comparison to the baseline scenario which is the PFAS occurrence and exposure conditions
expected in the absence of finalizing a PFAS drinking water regulation. The proposed rule is
referred to as the proposed option in presentation of EA results. This EA also presents the
incremental costs and benefits associated with three regulatory alternative MCLs for PFOA and
PFOS. The regulatory alternative MCLs are referred to as Option la (MCL of 4.0 ppt for PFOA
and 4.0 ppt for PFOS), Option lb (MCL of 5.0 ppt for PFOA and 5.0 ppt for PFOS), and Option
lc (MCL of 10.0 ppt for PFOA and 10.0 ppt for PFOS). The regulatory alternative MCLs for
PFOA and PFOS (Options la, lb, and lc) do not directly regulate additional PFAS, thereby
limiting public health protection and benefits relative to the proposed option.
In this EA, EPA presents the quantified and nonquantifiable health benefits expected from
reductions in PFAS exposures resulting from the proposed rule. Quantified benefits are assessed
as avoided cases of illness and deaths (or morbidity and mortality, respectively) associated with
exposure to PFAS contaminants. Adverse human health outcomes associated with PFAS
exposure but that cannot be quantified and valued are assessed as nonquantifiable benefits.
Additionally, this EA presents the costs associated with the proposed NPDWR. Costs presented
include those expenses incurred by PWSs to (1) monitor for PFAS, (2) inform consumers, (3)
install and operate treatment technologies, and (4) perform record-keeping and reporting to
comply with the PFAS NPDWR; and the costs incurred by states (or primacy agencies, i.e.,
states with authority to implement and enforce SDWA regulations) to implement the rule. EPA
presents annualized quantified benefits and costs discounted at 3 percent and 7 percent, which
are discount rates prescribed by the OMB (OMB Circular A-4, 2003).
Quantified economic benefits analyses consider the strength of evidence for each adverse health
effect and the availability of data to quantify the morbidity and mortality impacts associated with
that adverse health effect. To identify health effects that are associated with PFAS exposure and
can be monetized, EPA used the assessment of adverse health effects associated with PFOA and
1 EPA notes that the chemical HFPO-DA is used in a processing aid technology developed by DuPont to make fluoropolymers
without using PFOA. The chemicals associated with this process are commonly known as GenX Chemicals and the term is often
used interchangeably for HFPO-DA along with its ammonium salt.
Proposed PFAS Rule Economic Analysis
1-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
PFOS in the maximum contaminant level goal (MCLG) documents. EPA provides a quantitative
estimate of cardiovascular disease (CVD), birth weight, and renal cell carcinoma (RCC)-avoided
morbidity and mortality associated with reductions in PFOA and PFOS consistent with the
proposed rule. EPA provides a qualitative assessment of potential benefits for adverse health
effects that are associated with PFAS exposure but lack the economic or other information
needed for a quantitative analysis. In this EA, a qualitative discussion is provided for other
adverse health effects and potential avoided diseases associated with PFOA, PFOS, and the four
PFAS compounds included in the HI group (PFHxS, PFNA, PFBS, and HFPO-DA). The Agency
anticipates that the nonquantifiable human health benefits associated with reductions in drinking
water PFAS exposure are substantial and may reasonably exceed the benefits the Agency was
able to quantify for this regulatory proposal.
As part of its health risk reduction and cost analysis, EPA is directed by SDWA to evaluate
quantifiable and nonquantifiable health risk reduction benefits for which there is a factual basis
in the rulemaking record to conclude that such benefits are likely to occur from reductions in co-
occurring contaminants that may be attributed solely to compliance with the MCL (SDWA
1412(b)(3)(C)(II)). These co-occurring contaminants are expected to include additional PFAS
contaminants not directly regulated by the proposed PFAS NPDWR, co-occurring chemical
contaminants such as synthetic organic compounds (SOCs), volatile organic compounds (VOCs),
and disinfection byproduct (DBP) precursors.
The Agency anticipates that because of the PFAS NPDWR, some community and non-transient
non-community water systems will need to reduce their PFAS concentrations to comply with the
rule. This EA describes the costs associated with activities PWSs are expected to undertake to
comply with the proposed rule (e.g., installation of treatment technologies to remove PFAS), and
the costs associated with primacy agency implementation and administration of the proposed
rule. National quantified cost estimates are provided for PFOA, PFOS, and PFHxS treatment.
Due to occurrence data limitations, EPA has quantified the national treatment and monitoring
costs associated with the HI for PFHxS only and has not quantified the national cost impacts
associated with HI exceedances resulting from PFNA, PFBS, and HFPO-DA. Because these
costs are unquantified, national costs are underestimated. In instances where concentrations of
PFNA, PFBS, and HFPO-DA are high enough to cause or contribute to a HI exceedance when
the concentrations of PFOA, PFOS, and PFHxS would not have already otherwise triggered
treatment, the quantified costs may be underestimated. If these PFAS occur in isolation at levels
that affect treatment decisions, or if these PFAS occur in combination with PFHxS when PFHxS
concentrations are otherwise below the HI in isolation (i.e., <9.0 ppt) then the quantified costs
underestimate the impacts of the proposed rule. To characterize the costs associated with
treatment of other PFAS chemicals that are not included in the national cost estimates, EPA used
a model system approach to look at the potential differences in system level treatment costs that
could arise from the presence of PFNA, PFBS, and HFPO-DA which would cause HI
exceedances at systems precipitating additional systems to treat. EPA also use this model system
approach to estimate the incremental system level treatment expense resulting from co-
occurrence of PFNA, PFBS, and HFPO-DA at systems already required to treat because of
PFOA and/or PFOS MCLs and/or PFHxS HI exceedances. Additional discussion of the
methodology and results of this analysis can be found in Chapter 5, section 5.3.1.4 and Appendix
N.3.
Proposed PFAS Rule Economic Analysis
1-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA identified effective treatment technologies as part of the NPDWR, and consistent with
SDWA requirements found in Section 1412(b)(3)(C)(II) to consider benefits likely to occur from
reductions in co-occurring compounds, EPA estimated expected benefits from reductions in co-
occurring compounds as a result of PFAS treatment. Moreover, EPA developed a quantitative
analysis for reductions in bladder cancer morbidity and mortality that stem from removal of DBP
precursors. Disinfection byproducts, specifically trihalomethanes, are formed when disinfectants
interact with organic material in drinking water distribution systems. Since PFAS treatment has
been demonstrated to remove DBP precursors, the Agency anticipates that disinfection
byproducts, including trihalomethanes, will be reduced with PFAS treatment. EPA provides a
qualitative discussion of benefits for other potential water quality improvements that stem from
PFAS treatment, including those benefits associated with reductions in other co-occurring
contaminants besides DBPs.
In the tables below, quantified benefits and costs of the proposed NPDWR ("proposed option")
and alternative MCLs considered are presented. Table ES-1 presents the total estimated national
annualized benefits associated with the proposed option and regulatory alternatives considered.
Table ES-2 presents the total estimated national annualized costs associated with the proposed
option and regulatory alternatives considered. Quantitative estimates are presented using 3
percent and 7 percent discount rates. Throughout this EA, benefits and costs are presented using
mean (or "expected value"), 5th, and 95th percentile results to characterize key sources of
uncertainty, including but not limited to PFAS baseline occurrence and health effect slope factor
uncertainty, which is consistent with OMB and EPA guidance (OMB Circular A-4, 2003; U.S.
EPA, 2010a). All significant limitations and uncertainties of this economic analysis are described
in the pages that follow.
Table ES-1: Quantified Total National Annualized Benefits, All Options (Million $2021)
3% Discount Rate3 7% Discount Rate3
Option
5th
Percentileb
Expected
Value
95th
Percentileb
5th
Percentileb
Expected
Value
95th
Percentileb
Proposed Option0
$659.91
$1,232.98
$1,991.51
$477.69
$908.11
$1,462.43
Option lad
$651.19
$1,216.08
$1,971.01
$471.53
$895.36
$1,456.23
Option lbe
$553.37
$1,046.91
$1,706.81
$398.21
$773.33
$1,292.96
Option lcf
$280.42
$584.80
$1,030.56
$208.71
$436.24
$784.59
Notes: Detail may not add exactly to total due to independent rounding.
aSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
bThe 5tli and 95th percentile range is based on modeled variability and uncertainty described in Section 6.1.2 and Table 6-1 for
benefits. This range does not include the uncertainty described in Table 6-48 for benefits.
The proposed option sets PFOA and PFOS MCLs of 4.0 ppt and an HI of 1.0.
Option la sets PFOA and PFOS MCLs of 4.0 ppt.
eOption lb sets PFOA and PFOS MCLs of 5.0 ppt.
fOption lc sets PFOA and PFOS MCLs of 10.0 ppt.
Proposed PFAS Rule Economic Analysis
1-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table ES-2: Quantified Total National Annualized Costs, All Options (Million $2021)
3% Discount Rate£
i,b
7% Discount Rate£
i,b
Option
5th
Percentile0
Expected
Value
5th
Percentile0
5th
Percentile0
Expected
Value
95th
Percentile0
Proposed Option ' 0
$704.53
$771.77
$850.40
$1,106.01
$1,204.61
$1,321.01
Option laf
$688.09
$755.82
$833.48
$1,078.51
$1,177.31
$1,292.01
Option lbg
$558.71
$611.01
$674.32
$864.74
$942.28
$1,035.56
Option lch
$269.36
$292.57
$320.76
$396.22
$430.87
$472.20
Notes: Detail may not add exactly to total due to independent rounding.
aSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
bPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table do
not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns about
potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future, EPA
conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See Appendix
N, Section N.2 for additional detail.
The 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1 for
costs. This range does not include the uncertainty described in Table 5-22 for costs.
dTotal quantified national cost values do not include the incremental treatment costs associated with the cooccurrence of
HFPO-DA, PFBS, and PFNA at systems required to treat for PFOA, PFOS, and PFHxS. The total quantified national cost
values do not include treatment costs for systems that would be required to treat based on HI exceedances apart from systems
required to treat because of PFHxS occurrence alone. See Appendix N, Section N.3 for additional detail on cooccurrence
incremental treatment costs and additional treatment costs at systems with HI exceedances.
eThe proposed option sets PFOA and PFOS MCTs of 4.0 ppt and an HI of 1.0.
fOption la sets PFOA and PFOS MCTs of 4.0 ppt.
gOption lb sets PFOA and PFOS MCTs of 5.0 ppt.
''Option lc sets PFOA and PFOS MCTs of 10.0 ppt.
Proposed PFAS Rule Economic Analysis
1-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
2 Introduction
Per- and polyfluoroalkyl substances (PFAS) are a class of synthetic chemicals that have been
manufactured and in use since the 1940s (AAAS, 2020; U.S. EPA, 2022j). PFAS are most
commonly used to make products resistant to water, heat, and stains and are consequently found
in industrial and consumer products like clothing, food packaging, cookware, cosmetics,
carpeting, and fire-fighting foam (AAAS, 2020). PFAS manufacturing and processing facilities,
facilities using PFAS in the production of other products, airports, and military installations have
been associated with PFAS releases into the air, soil, and water (U.S. EPA, 2016b; U.S. EPA,
2016c). People may be exposed to PFAS by using certain consumer products, through
occupational exposure, and/or through consuming contaminated food or contaminated drinking
water (Domingo et al., 2019; Fromme et al., 2009).
Perfluorooctane sulfonate (PFOS) and perfluorooctanoic acid (PFOA) are part of a subset of
PFAS referred to as perfluorinated alkyl acids (PFAA) and are two of the most widely studied
and longest-used PFAS. Due to their widespread use and persistence in the environment, most
people have been exposed to PFAS, including PFOA and PFOS (U.S. EPA, 2016b; U.S. EPA,
2016c). PFOA and PFOS have been detected in up to 98 percent of blood serum samples taken in
biomonitoring studies that are representative of the U.S. general population (CDC, 2019).
Following the voluntary phase-out of PFOA by eight major chemical manufacturers and
processors in the U.S. under EPA's 2010/2015 PFOA Stewardship Program and reduced
manufacturing of PFOS (last reported in 2002 under Chemical Data Reporting), serum
concentrations have been declining. The National Health and Nutrition Examination Survey
(NHANES) data exhibited that 95th-percentile serum PFOS concentrations have decreased over
75 percent, from 75.7 (J,g/L in the 1999-2000 cycle to 18.3 (J,g/L in the 2015-2016 cycle (CDC,
2019; Jain, 2018; Calafat et al., 2007; Calafat et al., 2019).
Despite voluntary phase-outs and reduced exposure to some PFAS chemicals, PFAS are still
used in a wide range of consumer products and industrial applications. EPA's analysis of
drinking water monitoring data shows widespread occurrence of PFAS compounds in multiple
geographic locations. Most known exposures are relatively low, but some can be high,
particularly when people are exposed to a concentrated source over long periods of time. Studies
indicate that PFAS exposure above certain levels may result in adverse health effects, including
developmental effects to fetuses during pregnancy or to breast-fed infants, cancer, and other
immunologic-related effects.
Under the Safe Drinking Water Act (SDWA), the U.S. Environmental Protection Agency (EPA
or the Agency) is proposing to regulate PFAS in drinking water distributed by all community
water systems (CWSs)2 and non-transient non-community water systems (NTNCWSs). In 2021,
EPA determined that a NPDWR for PFAS would result in a meaningful opportunity to reduce
health risks (U.S. EPA, 2021b). Section 2.1 provides further detail on the proposed NPDWR for
PFAS.
2 Systems that supply water to the same population year-round.
Proposed PFAS Rule Economic Analysis
2-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
2.1 Summary of the Proposed PFAS Rule and Regulatory
Alternatives
EPA is proposing to regulate six PFAS in finished drinking water: (1) perfluorooctanesulfonic
acid (PFOS), (2) perfluorooctanoic acid (PFOA), (3) perfluorononanoic acid (PFNA), (4)
hexafluoropropylene oxide dimer acid (HFPO-DA or HFPO-DA), (5) perfluorohexanesulfonic
acid (PFHxS), and (6) perfluorobutanesulfonic acid (PFBS). The proposed regulation utilizes
compound-specific MCLs for PFOA and PFOS with a group MCL based on a hazard index (HI)
for PFNA, HFPO-DA, PFHxS, and PFBS. This regulatory approach utilizes the combined
toxicity framework peer reviewed by EPA's Science Advisory Board (SAB; U.S. EPA, 2022k)
and builds a framework for inclusion of additional PFAS through future rulemaking as new data
become available (U.S. EPA, 2023a). For more information on the HI approach, see EPA's Draft
Framework for Estimating Noncancer Health Risks Associated with Mixtures of Per- and
Polyfluoroalkyl Substances (PFAS) (U.S. EPA, 2023a).
Based on the best available scientific information on the health effects of PFOA and PFOS, EPA
is proposing maximum contaminant level goals (MCLGs) of 0 ppt for PFOA and 0 ppt for
PFOS. EPA has determined that it is feasible to set enforceable maximum contaminant levels
(MCLs) for PFOA and PFOS at 4.0 ppt each. Additionally, EPA has determined it is feasible to
set an MCL for four PFAS with a HI limit of 1.0. As such, EPA is proposing enforceable MCLs
of 4.0 ppt for PFOA, 4.0 ppt for PFOS, and a unitless HI of 1.0 for the group including PFNA,
HFPO-DA, PFHxS, and PFBS. For additional details about the MCLGs and MCLs in the
proposed rule, see the federal notice for this rulemaking. This proposed rule framework is
referred to as the "proposed option" within this EA.
Additionally, in this EA, EPA presents benefits and costs for the proposed rule as well as three
regulatory alternatives. The regulatory alternatives that EPA evaluated present individual MCLG
and enforceable MCL values for PFOA and PFOS. MCL values for PFOA and PFOS vary for
each alternative considered: 4.0 ppt in Option la, 5.0 ppt in Option lb, and 10.0 ppt in Option lc.
EPA evaluated benefits and costs for Option la to determine the difference in costs between
alternatives for PFOA and PFOS MCLs only versus MCLs for PFOA and PFOS and an HI for
four additional PFAS. EPA considered benefits and costs under Option lb—MCLs of 5.0 ppt for
PFOA and PFOS—because it is 25 percent above the compliance quantitation limit of 4.0 ppt
established for today's regulation. Lastly, EPA considered benefits and costs of Option lc—
MCLs of 10.0 ppt for PFOA and PFOS—to provide information on whether the Agency should
consider utilizing its authority under Section 1412(b)(6) to set an alternative MCL at the level at
which the benefits would justify the costs.
The Agency is also inviting comment on whether establishing a traditional MCLG and MCL for
PFHxS, HFPO-DA, PFNA, and PFBS instead of or in addition to the HI approach would change
public health protection, improve clarity for the rule, or change costs. EPA has not separately
presented changes in quantified costs and benefits for these approaches. If EPA adds individual
MCLs in addition to using the HI approach, EPA anticipates there will be no change in costs and
benefits relative to the proposed rule (i.e., the same number of systems will incur identical costs
to the proposed option and the same benefits will be realized). EPA has not separately quantified
the benefits and costs for the approach to regulate PFHxS, PFNA, PFBS, and HFPO-DA with
individual MCLs instead of the HI. However, EPA expects both the costs and benefits would be
Proposed PFAS Rule Economic Analysis
2-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
reduced under this approach as fewer systems may be triggered into treatment and its associated
costs. Additionally, systems that exceed one or more of the individual MCLs will treat to a less
stringent and public health-protective standard. Furthermore, under the proposed option, PWSs
are required to treat based on the combined occurrence of PFAS included in the HI which
considers the known and additive toxic effects and occurrence and likely co-occurrence of PFAS
compounds in the HI, providing more public health protection compared to an individual MCL
approach.
2.2 Economic Analysis Assumptions
2.2.1 Compliance Schedule and Period of Analysis for Proposed
Rule
For purposes of this EA, EPA assumes that the NPDWR will be promulgated by the end of 2023.
This analysis follows the standard NPDWR compliance schedule with regulatory requirements
taking effect three years after the date on which the regulation is promulgated. Therefore, EPA
assumes that actions to comply with the rule will begin taking place by 2026. In addition to this
initial time window, EPA's period of analysis includes the 80 years following the assumed
compliance date. This time span is based on an assumed median human lifespan of 80 years. In
this EA, EPA evaluates costs and benefits under the proposed rule for the period of analysis from
2023 through 2104. EPA selected this period of analysis to estimate human health risk reduction
to capture health effects from chronic illnesses that are typically experienced later in life (i.e.,
cardiovascular disease and cancer). Capital costs for installation of treatment technologies are
spread over the useful life of the technologies. EPA does not capture effects of compliance with
the proposed rule beyond the year 2104.
2.2.2 Dollar Year and Discount Rates
EPA presents estimated costs and benefits under the proposed rule in 2021 U.S. dollars.
Appendix J provides additional details on the price indices used for inflation adjustments.
The proposed rule analysis 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 be used
when evaluating a regulation that would mainly displace or alter the use of capital in the private
sector (OMB, 2003; updated 2009). OMB's Circular A-4 indicates that a 3 percent discount rate
represents the rate that an average saver uses to discount future consumption and is therefore
more appropriate for this rulemaking. EPA presents costs and benefits at both 3 and 7 percent.
The same discount rates are used for both benefits and costs. All future cost and benefit values
are discounted back to the initial year of the analysis, 2023, providing the present value of the
cost or benefit.
2.2.3 Annualization
Consistent with the timing of the proposed rule and associated reductions in PFAS levels, EPA
uses the following equation to annualize the future costs and benefits:
Proposed PFAS Rule Economic Analysis
2-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Equation 1:
r(PV)
AV = —
(1 + r)[l — (1 + r) n]
Where AV is the annualized value, PV is the present value,3 r is the discount rate (3% or 7%),
and n is the number of years (82 years).
2.2.4 Population
To determine the number of people expected to benefit from actions under the proposed rule,
EPA uses population data from the Safe Drinking Water Information System (SDWIS) 2021
Quarter 4 (Q4) database (U.S. EPA, 2021h). The SDWIS data provide the population served by
each PWS in the U.S. For analyses that rely on age-, sex-, and race/ethnicity-specific
populations, EPA uses county-level population proportions based on 2021 estimates from the
U.S. Census Bureau (2020a). EPA does not consider population growth during the period of
analysis (2023-2104). For more information on the SDWIS and U.S. Census Bureau (2020a)
data, see Appendix B.
2.2.5 Valuation
To estimate the economic value of avoided premature deaths, EPA uses Value of Statistical Life
(VSL) estimates. EPA follows Guidelines for Preparing Economic Analyses (U.S. EPA, 2010a)
and approximates Value of Statistical Life growth using a compound annual growth rate of
projected Value of Statistical Life values to obtain a Value of Statistical Life suitable for
valuation of mortality risk reductions during the period of analysis, 2023-2104. As the base
value, EPA used the Value of Statistical Life estimate of $4.8 million ($1990, 1990 income
year), which is the central tendency of the Value of Statistical Life distribution recommended for
use in EPA's regulatory impact analyses (U.S. EPA, 2010a). The base Value of Statistical Life
estimate is adjusted for inflation and income growth as described in Appendix J. The Value of
Statistical Life estimates employed in EPA's analysis range from $10.7 million ($2021) in 2023
to $17.7 million ($2021) in 2104 4
To estimate the economic value of avoided morbidity (i.e., non-fatal heart attacks and ischemic
strokes, birth weight decrements, and cancers), EPA used the cost of illness (COI) valuation
approach. The COI-based values used in this analysis reflect medical care expenditures and
opportunity costs associated with managing/treating the condition. The health endpoint-specific
morbidity valuation details are provided in Sections 6.4.4, 6.5.4, 6.6.4, and 6.7.2.5.
2.3 Document Organization
The remainder of this EA is organized into the following chapters:
• Chapter 2: Introduction summarizes the proposed PFAS rule and regulatory
alternatives, including the economic assumptions made in developing the rule.
3 The present value is the current value of a future sum of benefits given a specified discount rate. The present value represents
the expected value of benefits determined at the date of valuation.
4 Income growth projections from the U.S. Energy Information Administration (2021) are available through 2050.
Proposed PFAS Rule Economic Analysis
2-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Chapter 3: Need for the Rule summarizes the statutory requirements, regulatory
actions, and national EPA initiatives affecting PFAS in drinking water. It also explains
the contributors to the PFAS rule proposal, statutory authority, and the economic
rationale for the regulatory approach.
• Chapter 4: Baseline Drinking Water System Conditions describes the systems subject
to the proposed PFAS rule, PFAS water concentration levels, and data sources used to
characterize the baseline before the EPA models estimated changes that result from
complying with the proposed PFAS requirements.
• Chapter 5: Estimating Public Water System Costs provides a description of the
estimated costs for the proposed regulatory changes affecting systems and Primary
Agencies.
• Chapter 6: Benefits Analysis provides an estimate of the potential health benefits of the
proposed PFAS regulatory alternatives relative to the baseline, including quantification
and monetization where possible.
• Chapter 7: Comparison of Costs to Benefits provides a summary of costs and benefits
associated with the provisions of the proposed PFAS rule.
• Chapter 8: Environmental Justice Analysis provides a description of how the proposed
PFAS rule addresses Executive Order 12898: Federal Actions to Address Environmental
Justice in Minority Populations and Low-Income Populations.
• Chapter 9: Statutory and Administrative Requirements discusses analyses performed
to evaluate the effects of the proposed PFAS regulatory alternatives on different segments
of the population in accordance with 12 federal mandates and statutory reviews,
including but not limited to the Final Regulatory Flexibility Analysis (RFA/SBREFA),
Unfunded Mandates Reform Act (UMRA), and Executive Order 14008: Tackling the
Climate Crisis at Home and Abroad.
• Chapter 10: References includes a list of references cited throughout the proposed
PFAS rule economic analysis.
2.4 Supporting Documentation
This EA involves numerous detailed and complex analyses, and the following appendices are
provided to help the reader understand how those analyses were conducted and their underlying
data and assumptions:
• Appendix A: Framework of Bayesian Hierarchical Markov Chain Monte Carlo
Occurrence Model
• Appendix B: Affected Population
• Appendix C: Cost Analysis Results
• Appendix D: PFOA and PFOS Serum Concentration-Birth Weight Relationship
• Appendix E: Effects of Reduced Birth Weight on Infant Mortality
• Appendix F: Serum Cholesterol Dose-Response Functions
• Appendix G: CVD Benefits Model Details and Input Data
• Appendix H: Cancer Benefits Model Details and Input Data
• Appendix I: Trihalomethane Co-Removal Model Details and Analysis
• Appendix J: Value of a Statistical Life Updating
• Appendix K: Benefits Sensitivity Analyses
Proposed PFAS Rule Economic Analysis
2-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
• Appendix L: Uncertainty Characterization Details and Input Data
• Appendix M: Environmental Justice
• Appendix N: Supplemental Cost Analyses
• Appendix O: Appendix References
Proposed PFAS Rule Economic Analysis
2-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT
March 2023
3 Need for the Rule
This section provides the statutory and economic rationales for choosing a regulatory approach
to address the public health consequences of PFAS contamination in drinking water. EPA's
statutory requirements, regulatory actions, and Agency initiatives impacting PFAS in drinking
water are discussed.
3.1 Previous EPA Nonregulatory and Regulatory Actions
Potentially Affecting PFAS Drinking Water Management
This section provides a summary of actions and initiatives affecting PFAS in drinking water
prior to the publication of the proposed NPDWR for PFAS. Additionally, states have begun
proposing and promulgating their own regulatory and non-regulatory standards for PFAS in
drinking water. For more information on these state actions, see the Environmental Council of
the States' Processes & Considerations for Setting State PFAS Standards (ECOS, 2022).
3.1.1 PFAS Strategic Road map and PFAS Council
EPA Administrator Michael Regan established the EPA Council on PFAS in April 2021 and
charged it to develop a bold, strategic, whole-of-EPA strategy to protect public health and the
environment from the impacts of PFAS. The PFAS Council developed the PFAS Strategic
Roadmap to lay out EPA's whole-of-Agency approach to tackling PFAS and set timelines by
which the Agency plans to take concrete actions during the first term of the Biden-Harris
administration to deliver results for the American people. The Council comprises senior
technical and policy leaders from across EPA program offices and regions and is chaired by
Assistant Administrator for Water Radhika Fox and Acting Region 1 Administrator Deb Szaro
(U.S. EPA, 202le).
On October 18, 2021, Administrator Regan announced the Agency's PFAS Strategic
Roadmap—laying out a whole-of-agency approach to addressing PFAS. The PFAS Strategic
Roadmap sets timelines by which EPA plans to take specific actions and commits to bolder new
policies to safeguard public health, protect the environment, and hold polluters accountable.
Described in the Roadmap are key commitments the Agency made toward addressing these
contaminants in the environment. With this proposal, EPA is delivering on a key commitment in
the Roadmap to "establish a National Primary Drinking Water Regulation" for proposal and is
working toward promulgating the final NPDWR in Fall of 2023 (U.S. EPA, 202le).
3.1.2 EPA PFAS Health Advisories
In 2016, EPA published health assessments (Health Effects Support Documents or HESDs) for
PFOA and PFOS based on the Agency's evaluation of the peer reviewed science available at that
time. The lifetime Health Advisory (HA) of 70 ppt was used as the Health Reference Level
(HRL) for Regulatory Determination 4 and reflected the maximum combined concentration of
PFOA and PFOS in drinking water at which adverse health effects were not anticipated to occur
over a lifetime. Studies indicate that exposure to PFOA and/or PFOS above certain exposure
levels may result in adverse health effects, including developmental effects to fetuses during
pregnancy or to breast-fed infants (e.g., low birth weight, accelerated puberty, skeletal
variations), cancer (e.g., testicular, kidney), liver effects (e.g., tissue damage), immune effects
Proposed PFAS Rule Economic Analysis
3-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
March 2023
(e.g., antibody production and immunity), and other effects (e.g., cholesterol changes). Both
PFOA and PFOS are known to be transmitted to the fetus via the placenta and to the newborn,
infant, and child via breast milk. Both compounds were also associated with tumors in long-term
animal studies (U.S. EPA, 2016e; U.S. EPA, 2016f; NTP, 2020). For specific details on the
potential for adverse health effects and approaches used to identify and evaluate information on
hazard and dose-response, see Drinking Water Health Advisories for PFOA and PFOS and
Health Effects Support Documents for PFOA and PFOS (U.S. EPA, 2016b; U.S. EPA, 2016c;
U.S. EPA, 2016e; U.S. EPA, 2016f).
On June 15, 2022, EPA released four drinking water HAs for PFAS, including interim updated
HAs for PFOA and PFOS (U.S. EPA, 2022h). The HA levels for PFOA and PFOS are 0.004 ppt
and 0.02 ppt, respectively. These updated HA values are based on human studies in populations
exposed to PFOA and PFOS; studies have found associations between PFOA and/or PFOS
exposure and effects on the immune system, cardiovascular system, human development, and
cancer (U.S. EPA, 2022h).
Additionally, EPA issued final HAs for HFPO-DA and PFBS based on animal studies following
oral exposure to these chemicals. Exposure to HFPO-DA have been linked to health effects on
the liver, kidney, immune system, developmental effects, and cancer (U.S. EPA, 2022h).
Exposure to PFBS has been linked to health effects on the kidney, thyroid, reproductive system,
and developmental effects. The final HAs for HFPO-DA and PFBS are 10 ppt and 2,000 ppt,
respectively (U.S. EPA, 2022h).
3.1.3 Final Regulatory Determinations on the Fourth Drinking
Water Contaminant Candidate List
Section 1412(b)(l)(B)(i) of SDWA requires EPA to publish the CCL every five years after
public notice and an opportunity to comment. The CCL is a list of contaminants which are not
subject to any proposed or promulgated NPDWRs but are known or anticipated to occur in
PWSs and may require regulation under SDWA. SDWA section 1412(b)(l)(B)(ii) directs EPA to
determine, after public notice and an opportunity to comment, whether to regulate at least five
contaminants from the CCL every five years.
Under Section 1412(b)(1)(A) of SDWA, EPA will regulate a contaminant in drinking water if
the EPA Administrator determines that:
a) The contaminant may have an adverse effect on the health of persons;
b) The contaminant is known to occur or there is a substantial likelihood that the
contaminant will occur in PWSs with a frequency and at levels of public health concern;
and
c) In the sole judgment of the Administrator, regulation of such contaminant presents a
meaningful opportunity for health risk reduction for persons served by PWSs.
If after considering public comment on a preliminary determination, the decides to regulate a
contaminant, EPA will initiate the process to propose and promulgate a NPDWR. In that case,
the statutory time frame provides for Agency proposal of a regulation within 24 months and
action on a final regulation within 18 months of proposal.
Proposed PFAS Rule Economic Analysis
3-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
March 2023
On March 10, 2020, EPA published preliminary positive regulatory determinations for PFOS and
PFOA (85 FR 14098) (U.S. EPA, 2020a). On March 3, 2021, EPA published final regulatory
determinations for PFOS and PFOA (86 FR 12272) (U.S. EPA, 2021b). In doing so, EPA also
committed to evaluating a broader range of PFAS, including new monitoring and occurrence
data, and other information being developed by EPA, other federal agencies, state governments,
international organizations, industry groups, and other stakeholders (U.S. EPA, 2021b).
3.1.4 Unregulated Contaminant Monitoring Rule
As part of its responsibilities under the SDWA, EPA implements Section 1445(a)(2), Monitoring
Program for Unregulated Contaminants. This section requires that once every five years, EPA
issue a list of no more than 30 unregulated contaminants to be monitored by PWSs. This
monitoring is implemented through the Unregulated Contaminant Monitoring Rule (UCMR),
which collects data from community water systems and non-transient non-community water
systems. For each UCMR cycle, EPA establishes a new list of contaminants for monitoring,
specifies which systems are required to monitor, identifies the sampling locations, and defines
the analytical methods to be used.
The third Unregulated Contaminant Monitoring Rule (UCMR 3) was published on May 2, 2012.
UCMR 3 required monitoring for six PFAS: PFOA, PFOS, PFNA, PFHxS, PFBS, and PFHpA.
UCMR 3 data were used in the development of this economic analysis: see sections 4.2.2 and 4.4
for further discussion of these data.
On December 17, 2021, EPA Administrator Michael Regan signed the final Revisions to the
Unregulated Contaminant Monitoring Rule (UCMR 5) for Public Water Systems, and the rule
was subsequently published in the Federal Register on December 27, 2021 (86 FR 73131). The
five-year UCMR 5 cycle spans from 2022 to 2026, with preparations in 2022, sample collection
from 2023 to 2025, and completion of data reporting in 2026. UCMR 5 includes all 29 PFAS
that are within the scope of EPA Methods 533 and 537.1 (U.S. EPA, 2021b).
3.2 Statutory Authority for Promulgating the Rule
Section 1412(b)(1)(A) of SDWA authorizes EPA to establish NPDWRs for contaminants that
may have an adverse public health effect, that are known to occur or that present a substantial
likelihood of occurring in PWSs at a frequency and level of public health concern, and that
present a meaningful opportunity for health risk reduction for persons served by PWSs.
Section 1445(a) of SDWA authorizes the EPA Administrator to establish monitoring,
recordkeeping, and reporting regulations that the Administrator can use to establish regulations
under the SDWA, determine compliance with SDWA, and advise the public of the risks of
unregulated contaminants (42 U.S.C. § 300j-4(a)). In requiring a PWS to monitor under Section
1445(a), the Administrator may take into consideration the water system size and the
contaminants likely to be found in the system's drinking water (42 U.S.C. § 300j-4(a)). Section
1445(a)(1)(C) of the SDWA provides that "every person who is subject to a national primary
drinking water regulation" under section 1412 must provide such information as the
Administrator may reasonably require to assist the Administrator in establishing regulations
under section 1412 (42 U.S.C § 300j-4(a)(l)(C)).
Proposed PFAS Rule Economic Analysis
3-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
March 2023
Section 1413(a)(1) of the SDWA allows EPA to grant a state primary enforcement responsibility
("primacy") for NPDWRs when EPA has determined that the state has, among other things,
adopted regulations that are no less stringent than EPA's (42 U.S.C. § 300g-2(a)(l)). To obtain
primacy for this rule, states must adopt comparable regulations within two years of the EPA's
promulgation of the final rule, unless EPA grants the state a two-year extension (40 CFR
142.12(b)). State primacy requires, among other things, adequate enforcement (including
monitoring and inspections) and reporting. EPA must approve or deny state primacy applications
within 90 days of submission to EPA (42 U.S.C. § 300g-2(b)(2)). In some cases, a state
submitting revisions to adopt a NPDWR has interim primary enforcement authority for the new
regulation while EPA's decision on the revision is pending (42 U.S.C. § 300g-2(c)).
Section 1450 of the SDWA authorizes the Administrator to prescribe such regulations as are
necessary or appropriate to carry out his or her functions under the Act (42 U.S.C § 300j-9).
3.3 Economic Rationale
The OMB Circular A-4 (OMB, 2003) states that "in order to establish the need for the proposed
action, the analysis should discuss whether the problem constitutes a significant market failure."
This section describes the types of market failures that NPDWRs address.
In a perfectly competitive market, market forces guide buyers and sellers to attain the most
efficient social outcome. A perfectly competitive market occurs when both buyers and sellers are
price takers, usually when there are many producers and buyers of a product and both producers
and buyers have complete knowledge about that product. Also, there must not be any barriers to
entry into the industry, and existing producers in the industry must not have any advantage over
potential new producers. Several factors in the public water supply industry preclude it from
being a perfectly competitive market and lead to market failures that may require regulation.
First, it is not economically efficient to have multiple suppliers who would, for example,
compete by building multiple systems of pipelines, reservoirs, wells, and other facilities. Instead,
economic efficiency leads to a single firm or government entity performing these functions
generally under public control. Under these monopoly conditions, consumers are provided only
one level of service with respect to drinking water quality. If consumers do not believe that the
quality of tap water is adequate, they cannot simply switch to another water utility. Consumers
may purchase bottled water, but this option can be much more expensive due to the inefficiencies
of bottling and transporting bottled water. Consumers may also install and operate home
treatment systems, but this can also be considerably more expensive without the economies of
scale of large, centralized water systems. Additionally, home treatment systems potentially can
lead to increased health risks when not regularly maintained by the consumer.
Second, high information and transaction costs impede the public's understanding of health and
safety issues concerning drinking water quality. The health risks potentially posed by trace
quantities of drinking water contaminants requires EPA to analyze and distill complex
toxicological and health sciences data. EPA promulgated the Consumer Confidence Report
(CCR) rule to make water quality information more easily available to consumers. The CCR rule
requires CWSs to mail their customers an annual report on local drinking water quality.
The report provides customers with information on levels of detected contaminants in their
drinking water, limited health risk information associated with contaminant exposure when
Proposed PFAS Rule Economic Analysis
3-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
March 2023
levels exceed MCLs, and utility contact information. Even if informed consumers can engage
utilities regarding these health issues, the costs of such engagement, known as "transaction
costs" (in this case measured in personal time and commitment), can be a barrier to efficient
market outcomes.
SDWA regulations are intended to provide a level of protection from exposure to drinking water
contaminants that would not otherwise occur in the existing market environment of public water
supply. The regulations set minimum performance requirements for all public water supplies to
reduce the risk confronted by all consumers from exposure to drinking water contaminants.
SDWA regulations are not intended to restructure market mechanisms or establish competition in
supply; rather, SDWA standards establish the level of service needed to better reflect the public's
preference for safety. Federal regulations remove the high information and transaction costs by
acting on behalf of all consumers in balancing the risk reduction and social costs of achieving
this reduction.
Proposed PFAS Rule Economic Analysis
3-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
4 Baseline Drinking Water System Conditions
4.1 Introduction
In its Guidelines for Preparing Economic Analyses, EPA characterizes the baseline as a
reference point that reflects the world without the proposed regulation (U.S. EPA, 2010a); this
baseline is the starting point for estimating the potential benefits and costs of the proposed PFAS
NPDWR.
This chapter presents a characterization of PWSs and their current operations (i.e., the baseline)
before changes are made to meet the proposed PFAS NPDWR. Section 4.2 identifies each major
data source used to develop the baseline. Section 4.3 explains the derivation of each baseline
characteristic and presents results in detailed tables. Section 4.4 describes the Bayesian model
developed to estimate national PFAS occurrence in drinking water supplies. Section 4.5
summarizes limitations of the major data sources and uncertainties in the baseline
characterization (both quantified and nonquantifiable) in table format.
4.2 Data Sources
EPA used a variety of data sources to develop the baseline. Section 4.2.1 explains the relevant
information provided in the federal version of the Safe Drinking Water Information System
(SDWIS/Fed) and measures EPA has taken to verify the data. Section 4.2.2 describes the
purpose of the third Unregulated Contaminant Monitoring Rule (UCMR 3) data. Section 4.2.3
describes the independent state sampling program data. Sections 4.2.4 and 4.2.5 describe two
data sources used to develop key characteristics of system treatment plants. Section 4.2.6
explains the purpose of the 2006 Community Water System Survey (CWSS) and the
representativeness of the data. Table 4-1 identifies each major data source and the baseline data
element(s) derived from them.
Table 4-1: Data Sources Used to Develop the Water System Characteristics
Data Source
Baseline Data Derived from the Source
SDWIS/Fed fourth quarter
2021 Q4 "frozen" datasef
• Water System Inventory (Section 4.3.1): PWS inventory, including system
unique identifier, population served, number of service connections,
source water type, and system type.
• Population and Households Served (Section 4.3.2): PWS population
served.
• Treatment Plant Characterization (Section 4.3.3.1): Number of unique
treatment plant facilities per system, which are used as a proxy for entry
points when UCMR 3 sampling site data are not available.
UCMR 3 (U.S. EPA, 2017)
• Treatment Plant Characterization (Section 4.3.3): Number of unique entry
point sampling sites, which are used as a proxy for entry points.
• Treatment Plant Characterization (Section 4.3.3): PFAS concentration data
collected as part of UCMR 3.
Independent state sampling
programs
• Treatment Plant Characterization (Section 4.3.3): PFAS concentration data
collected by states. These data supplemented the occurrence modeling for
systems included in UCMR 3.
Proposed PFAS Rule Economic Analysis
4-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-1: Data Sources Used to Develop the Water System Characteristics
Data Source
Baseline Data Derived from the Source
SYR4 ICR Occurrence
Dataset (2012-2019)
• Treatment Plant Characterization (Section 4.3.3): TOC.
Geometries and
Characteristics of PWSs
(U.S. EPA, 2000)
• Treatment Plant Characterization (Section 4.3.3): Design and average
daily flow per system.
2006 CWSS (U.S. EPA,
2009)
• PWS Labor Rates (Section 4.3.4): PWS labor rates.
Abbreviations: CWSS - Community Water System Survey; ICR - Information Collection Request; PFAS - per- and
polyfluoroalkyl substances; PWS - public water system; SDWIS/Fed - Safe Drinking Water Information System/federal
version; SYR - Six-Year Review; TOC - total organic carbon; UCMR 3 - Third Unregulated Contaminant Monitoring Rule.
Note:
aContains information extracted on January 14, 2022.
4.2.1 SDWIS/Fed 2021
SDWIS/Fed (U.S. EPA, 2021h) is EPA's national regulatory compliance database for the
drinking water program. It contains system inventory, treatment facility, violation, and
enforcement information for PWSs as reported by primacy agencies, EPA regions, and EPA
headquarters personnel. Primacy agencies report data quarterly to EPA. The information
presented in the EA is based on the fourth quarter 2021 "frozen" dataset that was extracted on
January 14, 2022.
SDWIS/Fed contains information to characterize the inventory of PWSs, namely: system name
and location; retail population served, source water type, and PWS type.
4.2.1.1 PWS Type
EPA defines a PWS as a system that provides water for human consumption through pipes or
other constructed conveyances to at least 15 service connections or regularly serves an average
of at least 25 individuals per day for at least 60 days per year (U.S. EPA, 2021h). Systems are
categorized as follows:
• CWSs are systems that supply water to the same population year-round.
• Non-community water systems (NCWSs) are systems that supply water to a varying
population or one that is served less than year-round; these are sub-categorized as:
o Non-transient non-community water systems (NTNCWSs) are systems that are not
CWSs and that regularly supply water to at least 25 of the same people at least six
months per year (e.g., schools),
o Transient non-community water systems (TNCWSs) are NCWSs that do not meet the
non-transient criterion; they provide water in places such as gas stations or seasonal
campgrounds where people do not remain for long periods of time.
Proposed PFAS Rule Economic Analysis
4-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
A proposed rule to limit PFAS in drinking water would not apply to TNCWSs. Therefore,
system inventories in this analysis are classified into two categories: CWSs and NTNCWSs.
4.2.1.1.1 Population Served
Systems are also categorized by the number of people they serve.5 The following nine categories
of populations served by systems are used throughout this EA:
• < 100
• 101-500
• 501-1,000
• 1,001-3,300
• 3,301-10,000
• 10,001-50,000
• 50,001-100,000
• 100,001-1,000,000 (1M)
• >1M
EPA uses these system size categorizes based on distinctions in the way systems operate as the
amount of water supplied and number of service connections increases. Systems within each size
category can be expected to face similar implementation and cost challenges when complying
with the new regulatory requirements proposed for this regulatory effort.
4.2.1.1.2 Source Water Type
SDWIS/Fed classifies system by source water using the following six categories:
• Ground water (Ground Water)
• Ground water purchased
• Ground water under the direct influence (GWUDI)6
• Ground water under the direct influence purchased (purchased GWUDI)
• Surface water (Surface Water)
• Surface water purchased
For this analysis, EPA broadly categorized systems as Surface Water if any of their sources are
surface water, surface water purchased, GWUDI, or purchased GWUDI. Systems are classified
as Ground Water if they exclusively used Ground Water or purchased Ground Water.7
5 SDWIS/Fed classifies systems according to "retail" population that does not include the population served by other systems that
purchase water from them.
0 40 CFR section 141.2 defines ground water under the direct influence of surface water as "any water beneath the surface of the
ground with significant occurrence of insects or other macroorganisms, algae, or large-diameter pathogens such as Giardia
lamblia or Cryptosporidium, or significant and relatively rapid shifts in water characteristics such as turbidity, temperature,
conductivity, or pH which closely correlate to climatological or surface water conditions."
7 23 CWS and 11 NTNCWS have an unknown primary water source. For purposes of this analysis, EPA assigned these systems
to the source type ground water.
Proposed PFAS Rule Economic Analysis
4-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
4.2.1.1.3 Facilities
SDWIS/Fed provides additional information on system facilities, including the type of facility,
its activity status, and a unique facility identification number.
4.2.1.2 Verification of SDWIS/Fed Doto
EPA routinely conducts program reviews to verify whether information in the primacy agencies'
databases and files, such as inventory and violations for all regulations are correctly represented
in SDWIS/Fed. Between 2006 and 2016, EPA recorded the findings from these reviews in the
national Error Code Tracking Tool (ECTT) (U.S. EPA, 2007b). The ECTT contains, as
individual records, all actions assessed during each program review. EPA identifies records as
confirmed actions (correct compliance determinations and correct reporting to SDWIS/Fed),
compliance determination discrepancies (incorrect compliance determinations), or data flow
discrepancies (correct compliance determination but incorrect reporting). This section presents
data from the ECTT from program reviews conducted from 2006 to 2016 related to system
inventory.
It is important to note that treatment data (objective codes and process codes for plants in
SDWIS) are not evaluated during program reviews and therefore have more uncertainty
associated with the data as compared to inventory and compliance data.
4.2.1.2.1 System Inventory
From 2006 to 2016 EPA evaluated inventory data for a total of 2,180 systems. Prior to August
2007, the program reviews evaluated eight inventory fields: system type, system status, activity
status, source type, population, service connection, administrative contact, and administrative
address. After August 2007, the reviews did not include administrative contact or address. In
addition, in August 2007, the review policy changed so that discrepancies for inventory were
only identified if they affected monitoring requirements (e.g., a change in population that would
increase or decrease the minimum number of required samples).
Of the inventory fields evaluated from 2006 to 2016, only 82 (<1%) inventory discrepancies
were identified. Furthermore, some of these discrepancies, such as those related to administrative
contact and address, may not impact the PWS baseline characterization. The inventory data in
ECTT indicate a high degree of completeness and accuracy in SDWIS/Fed, and that the
information is largely representative of the regulated PWS.
4.2.2 Third Unregulated Contaminant Monitoring Rule
Every five years, EPA issues a new list of no more than 30 unregulated contaminants to be
monitored by PWSs. UCMR 3 was published in 2012 and required monitoring for six PFAS
from 2013-2015: PFOA, PFOS, PFBS, PFNA, PFHxS, and PFHpA.. The final UCMR 3 dataset
of analytical results was released in January 2017.
Under UCMR 3, all CWSs and NTNCWSs with more than 10,000 retail customers and a
representative sample of 800 systems serving 10,000 or fewer retail customers were required to
conduct assessment monitoring to collect occurrence data for the listed contaminants suspected
to be present in drinking water but that do not have health-based standards set under the SDWA.
Proposed PFAS Rule Economic Analysis
4-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Systems conducted assessment monitoring over one consecutive 12-month period between
January 2013 and December 2015. Ground Water systems were required to monitor twice during
that period, with sampling events occurring five to seven months apart. Surface Water systems
were required to monitor in four consecutive quarters, with sampling events occurring three
months apart. For the PFAS compounds, sampling was conducted at the entry point to the
distribution system post treatment.
4.2.3 Independent State Sampling Programs
EPA used state monitoring data from 12 states (Alabama, Colorado, Illinois, Kentucky,
Massachusetts, Michigan, New Hampshire, New Jersey, North Dakota, Ohio, South Carolina,
and Vermont). These states conducted non-targeted monitoring (i.e., random sampling) of
finished drinking water for one or more of the four PFAS in this analysis.
4.2.4 Six-Year Review Data
EPA used information from the fourth Six-Year Review Information Collection Request (ICR)
Dataset ("SYR4 ICR dataset") to characterize the total organic carbon (TOC) level for individual
systems. The SYR4 ICR dataset is the most comprehensive and current national drinking water
occurrence dataset, containing millions of records of water system compliance monitoring data
and treatment technique information for regulated chemical, radiological, and microbiological
contaminants collected from 2012 through 2019. The portion of the dataset containing the TOC
information was made publicly available in August 2022.8
4.2.5 Geometries and Characteristics of Public Water Systems
(2000)
An important factor in determining costs of treatment is average daily flow and design flow,
measured in gallons per day or million gallons per day (MGD), at a treatment plant. EPA
estimated the average daily flow and design flow for each entry point in the system based on the
relationship between retail population and flow as derived in EPA's Geometries and
Characteristics of Public Water Systems report (U.S. EPA, 2000).
Utilizing data from the 1995 CWSS, EPA conducted an extensive data-cleaning process9 to
develop a dataset of 1,734 records with paired responses for population and total average daily
flow. These data were then weighted to account for non-responses to individual questions from
the CWSS. EPA used this dataset to develop regression equations that predict average daily flow
based on retail population served (for both publicly-owned and privately-owned systems). The
data show a very good correlation as indicated by a high R value of 0.90. Additional information
and background data are provided in Chapter 4 of the Geometries and Characteristics of Public
Water Systems report (U.S. EPA, 2000).
8 Available at: https://www.epa.gov/dwsixvearreview/microbial-and-disinfection-bvproduct-data-files-2012-2019-epas-fourth-
six-vear
9 EPA adjusted the dataset to remove non-zero values; adjusted flow if needed to represent retail flow only removing wholesale
water flow; and adjusted for reporting discrepancies in population, flow, or service connections.
Proposed PFAS Rule Economic Analysis
4-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
4.2.6 Community Water System Survey (2006)
EPA periodically conducts the CWSS to obtain data to support the Agency's development and
evaluation of drinking water regulations. The 2006 CWSS is the most recent survey. For this EA,
EPA relied on the national average estimates of unit labor from the 2006 CWSS to derive the
unit labor rates.
EPA selected the CWSS as a data source because it is based on a nationally representative
sample of CWSs. The sample was drawn from SDWIS/Fed, which includes approximately
50,000 systems in the 50 states and the District of Columbia. The survey used a stratified random
sample design to ensure the sample was representative. EPA selected a survey sample of 2,210
systems, including all systems serving populations of 100,000 or more. In the 2006 CWSS, the
Agency took additional steps to improve response rates, ensure accurate responses, and reduce
the burden of the survey on systems, especially systems serving 3,300 or fewer persons. EPA
sent water system experts to collect data from systems serving 3,300 or fewer persons. For
systems serving more than 3,300 people, the Agency mailed the survey, made available a
spreadsheet and Web-based version of the questionnaire, and provided extensive assistance
through e-mail and a toll-free telephone hotline. The survey was designed to collect data for the
year 2006. Full-scale data collection occurred from June to December 2007. The overall
response rate was 59 percent with a total of 1,314 systems responding; 95 percent of selected
systems serving 3,300 or fewer persons (representing 571 of 600 systems sampled) participated
in the survey (U.S. EPA, 2009).
4.3 Drinking Water System Baseline/Industry Profile
This section presents the following baseline characterizations for the purposes of estimating costs
and benefits for the proposed rule. Section 4.3.1 provides a characterization of the inventory of
systems subject to the proposed rule (CWSs and NTNCWSs). Section 4.3.2 includes the
population served by CWSs and NTNCWSs and the number of households served by CWSs.
Section 4.3.3 provides treatment plant characteristics used to determine treatment costs. Section
4.3.4 describes the derivation of PWS labor rates. Finally, Section 4.3.5 describes the cost of
capital rates used to estimate household-level costs. Each section includes a characterization of
the baseline for CWSs, followed by NTNCWSs, if applicable, and a characterization of data
limitations and uncertainty. TNCWSs are not subject to the proposed rule.
4.3.1 Water System Inventory
A key component of the baseline is the inventory of systems—both CWSs and NTNCWSs—
subject to the proposed rule. As shown in Table 4-2, approximately 81 percent of all CWSs serve
3,300 or fewer people (39,746 of the total systems), and those serving 500 or fewer account for
about 54 percent of all CWSs (26,742 of the total systems). CWSs serving 3,301-50,000 people
represent about 17 percent of all CWSs (8,422 of the total systems), and those serving more than
50,000 people account for only about 2 percent (1,025 of the total systems). Most CWSs (about
77 percent or 37,733 systems) use Ground Water as their primary source. Most systems serving
more than 10,000 people, however, are classified as Surface Water systems (about 63 percent or
2,817 systems).
Proposed PFAS Rule Economic Analysis
4-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-2: Inventory of CWSs
CWSs'
System Size (Population
Served)
Ground Water
Surface Water
Total
A
B
C = A + B
< 100
10,654
739
11,393
101-500
13,037
2,042
15,079
501-1,000
4,132
1,179
5,311
1,001-3,300
5,503
2,460
7,963
3,301-10,000
2,784
2,223
5,007
10,001-50,000
1,385
2,030
3,415
50,001-100,000
162
417
579
100,001-1M
74
347
421
> 1M
2
23
25
TOTAL
37,733
11,460
49,193
Abbreviations: CWS - community water systems.
Notes:
includes 23 CWSs serving 10,000 or fewer people for which no primary source water type was reported to SDWIS/Fed.
EPA assigned these systems to the source type of Ground Water.
Source: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through January 14, 2022.
Includes all active CWSs.
As shown in Table 4-3, approximately 99 percent of all NTNCWSs serve 3,300 or fewer people
(17,135 of the total). NTNCWSs serving 3,301 - 50,000 people account for about 1 percent of all
NTNCWSs (200 of the total). Only two NTNCWSs serve more than 50,000 people, and none
serve more than 1 million people. Most NTNCWSs (about 95 percent or 16,531 systems) use
Ground Water as their primary source. Approximately 51 percent (21 systems) of those serving
10,001-100,000 people use Surface Water versus Ground Water and the one system serving
100,001-1 million people is classified as a Surface Water system.
Proposed PFAS Rule Economic Analysis
4-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-3: Inventory of NTNCWSs
NTNCWSs'
System Size (Population
Served)
Ground Water
Surface Water
Total
A
B
C=A+B
< 100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1M
> 1M
TOTAL
16,531
8,084
6,111
1,476
743
97
20
0
0
0
252
257
91
121
63
20
1
1
0
806
17,337
8,336
6,368
1,567
864
160
40
1
1
0
Abbreviations: NTNCWS - non-transient non-community water systems.
Notes:
includes 11 NTNCWSs serving 3,300 or fewer people for which no primary source type was reported to SDWIS/Fed. EPA
assigned these systems to the source water type of Ground Water.
Sources: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through January 14, 2022.
Includes all active NTNCWSs.
Proposed PFAS Rule Economic Analysis 4-8 March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
There is uncertainty in the approach used to assign source water type to the 23 CWSs and 11
NTNCWSs where no primary source type was reported to SDWIS/Fed. This analysis assumes
that these systems have Ground Water as their primary source based on the preponderance of
ground water systems in the inventory. This could result in an under- or overestimate of costs in
those instances where the cost model inputs vary by source type (e.g., number of entry points per
system); however, EPA expects the impact to be low because the systems without a source type
in SDWIS/Fed represent a small proportion of systems subject to the rule (23 of the total 49,193
CWSs and 11 of the total 17,337 NTNCWSs or 0.05 percent of all systems subject to the rule)
and all serve fewer than 10,000 people.
4.3.2 Population and Households Served
It is necessary to have an accurate characterization of population served by water systems when
assessing the potential benefits of a proposed regulation. Population is also an input for
estimating treated water volumes and associated granular activated carbon (GAC) or ion
exchange (IX) costs.
SDWIS/Fed tracks "retail" population served, meaning that it counts only the population that
purchases water directly from the water system, not the population of a system's wholesale
customers. The systems that purchase water appear in SDWIS/Fed as a separate system with a
unique PWS identification (PWSID) number.
Table 4-4 and Table 4-5 show the total population served and average population served per
system by size category for CWSs and NTNCWSs, respectively. Each exhibit is organized by
source water type (Surface Water or Ground Water) and is based on the SDWIS/Fed fourth
quarter 2021 "frozen" dataset that contains information reported by primacy agencies through
January 14, 2022.
Because systems often pass some or all of their costs onto customers in the form of rate
increases, the proposed rule cost analysis also includes analyses to assess the impact of the
proposed requirements on annual household expenditures. EPA estimated the number of
households served by affected CWSs by dividing the population for each system size category by
the average number of people per household. For CWSs, EPA assumed an average of 2.53
persons per household based on 2020 U.S. Census data (U.S. Census Bureau, 2020b). This
information is also included in Table 4-4 by system size and source type. NTNCWSs do not
serve households, thus, this information is not included in Table 4-5.
As shown in Table 4-4, although CWSs serving 3,300 or fewer people account for approximately
81 percent of all CWSs, they serve fewer than 8 percent of the population and households that
receive their water from a CWS. Although CWSs serving more than 50,000 people account for
only 2 percent of all CWSs, they serve more than half (59 percent) of the population and
households that receive their water from a CWS.
Proposed PFAS Rule Economic Analysis
4-9
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-4: Population and Number of Households Served by CWSs
Ground Water0
Surface Water
TOTAL
System Size
(Population
Served)
Population
Served
Average
Population
Per System
Number of
Households
Served
Population
Served
Average
Population
Per System
Number of
Households
Served
Population
Served
Average
Population
Per System
Number of
Households
Served
A
Ba
C = A/2.53b
D
Ea
F=D/2.53b
G
Ha
I=G/2.53b
< 100d
652,335
61
257,840
45,231
61
17,878
697,566
61
275,718
101-500
3,254,293
250
1,286,282
576,601
282
227,906
3,830,894
254
1,514,187
501-1,000
3,032,366
734
1,198,564
883,656
749
349,271
3,916,022
737
1,547,835
1,001-3,300
10,264,020
1,865
4,056,925
4,935,965
2,006
1,950,974
15,199,985
1,909
6,007,899
3,301-10,000
15,794,291
5,673
6,242,803
13,633,206
6,133
5,388,619
29,427,497
5,877
11,631,422
10,001-50,000
28,665,202
20,697
11,330,119
46,262,480
22,789
18,285,565
74,927,682
21,941
29,615,685
50,001-100,000
10,889,918
67,222
4,304,315
29,350,794
70,386
11,601,104
40,240,712
69,500
15,905,420
100,001-1M
15,082,760
203,821
5,961,565
84,675,709
244,022
33,468,660
99,758,469
236,956
39,430,225
> 1M
3,400,000
1,700,000
1,343,874
44,266,001
1,924,609
17,496,443
47,666,001
1,906,640
18,840,317
TOTAL6
91,035,185
2,413
35,982,287
224,629,643
19,601
88,786,420
315,664,828
6,417
124,768,707
Abbreviations: Ground Water - ground water; CWS - community water systems.
Notes:
aB, E, and H: Derived by dividing the population served by the number of systems presented in Table 4-2.
bC, F, and I: The average of 2.53 persons per household is from 2020 U.S. Census data (Table AVG1. Average Number of People per Household, by Race and Hispanic Origin/1,
Marital Status, Age, and Education of Householder: 2020).
cCWSs with unreported primary source were assumed to be Ground Water systems. Thus, the Ground Water column reflects an additional 23 CWSs with unreported primary
source type.
dEPA removed any CWS wholesaler serving less than 25 people from the analysis and assumed that any remaining CWS had a minimum possible population of 25.
eNumbers may not sum to total because of rounding.
Source for A, D, and G: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through January 14, 2022.
Proposed PFAS Rule Economic Analysis
4-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
As previously discussed, NTNCWSs serving 3,300 or fewer people account for approximately
99 percent of all NTNCWSs. As shown in Table 4-5, these systems serve approximately 70
percent of the population that receives their water from an NTNCWS. Those serving 3,301—
50,000 people and more than 50,000 people serve approximately 26 percent and 4 percent of the
population that receives water from an NTNCWS, respectively.
Table 4-5: Population Served by NTNCWSs
Ground Waterb Surface Water TOTAL
System Size
(Population Served)
Population
Served
Average
Population
Per System
Population
Served
Average
Population
Per
System
Population
Served
Average
Population
Per
System
A
Ba
D
Ea
F
Ga
< 100c
452,516
56
12,534
50
465,050
56
101-500
1,513,562
248
69,046
269
1,582,608
249
501-1,000
1,049,638
711
68,235
750
1,117,873
713
1,001-3,300
1,241,973
1,672
239,516
1,979
1,481,489
1,715
3,301-10,000
511,494
5,273
377,219
5,988
888,713
5,554
10,001-50,000
397,246
19,862
414,099
20,705
811,345
20,284
50,001-100,000
0
0
71,963
71,963
71,963
71,963
100,001-1M
0
0
203,375
203,375
203,375
203,375
> 1M
0
0
0
0
0
0
TOTALd
5,166,429
313
1,455,987
1,806
6,622,416
382
Abbreviations: NTNCWS - non-transient non-community water systems.
Notes:
aB, E, and G: Derived by dividing the population served by the number of systems presented in Table 4-3.
bNTCWSs with unreported primary source were assumed to be Ground Water systems. Thus, the "Ground Water" column
reflects an additional 11 NTCWSs with unreported primary source type.
CEPA assumed any non-wholesale NTNCWS had a minimum possible population of 25.
dNumbers may not sum to total because of rounding.
Source for A, D, and F: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through
January 14, 2022.
As noted previously, EPA consistently classifies systems in SDWIS/Fed according to the retail
population served by the system and does not include the population served by wholesale
customers. Wholesale customers who purchase water from another system and meet the PWS
definition have their own unique PWSID, retail population, and associated regulatory
requirements under SDWA. EPA uses retail population to estimate design and average daily flow
parameters, which are then used to estimate treatment costs associated with the rule. Use of retail
population may overestimate aggregate costs by assuming that each system will have an
individual treatment plant instead of the more common scenario of the seller having one large
plant and selling treated water to their wholesale customers. Because of returns to scale in
treatment capital costs, the cost of a single large plant will be less than the sum of the costs
across several small plants treating the same aggregate flow.
Proposed PFAS Rule Economic Analysis
4-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
In addition, given that some of the reported population values would create inconsistencies in the
analysis, EPA removed any CWS wholesaler serving less than 25 people from its analysis and
assumed that any remaining CWS had a minimum possible population of 25. EPA assumed any
non-wholesale NTNCWS had a minimum possible population of 25.
4.3.3 Treatment Plant Characterization/Production Profile
This section explains the baseline inputs for the following treatment-related PWS characteristics.
Section 4.3.3.1 discusses the entry points per system characterization. Section 4.3.3.2 discusses
EPA's TOC baseline assumptions and Section 4.3.3.3 presents the estimation method and the
computed average daily flows and design flows by system type and size.
4.3.3.1 Entry Points Per System
Entry points are the point of compliance for the proposed rule and systems can have multiple
entry points. EPA developed estimates of entry points per system using UCMR 3 unique
sampling points, SDWIS/Fed facility data, and a modeled frequency distribution.
UCMR 3 required a subset of CWSs and NTNCWSs to conduct assessment monitoring for six
PFAS compounds.10 The data record a unique identifying number for the entry point sample
location(s) for each system. Given the information provided, EPA assumes that the number of
unique sample point IDs per system approximates the total number of entry points per system.
For systems without UCMR 3 occurrence data, EPA developed estimates based on SDWIS/Fed
facilities data. The SDWIS/Fed data include unique identification numbers for system facilities,
as well as facility type and activity status. This analysis relies on active facilities identified as
treatment plants. Using the assumption that treatment plants are associated with one entry point,
the SDWIS/Fed facility data provide an approximation for the number of entry points per system
when a system does not have UCMR 3 occurrence data. EPA considers the UCMR 3 sampling
point data to be of higher quality than the SDWIS/Fed treatment facility data. If the SDWIS/Fed
treatment facility data value for a system exceeded the maximum number found for the
equivalent system size and source water combination in the UCMR 3 data, EPA limited the
system entry point value to the UCMR3 maximum number of entry points.
For systems without UCMR 3 occurrence data or SDWIS/Fed facility data, EPA relies on an
estimate of the number of entry points. The estimated value for each system with missing entry
point count data was imputed from known entry point counts for stratified SDWIS/Fed data.
Within each stratum, defined by a combination of system size and source water, EPA sampled
from systems with known entry point counts. Sampling was done with replacement after
truncating the entry point counts to the maximum recorded in UCMR 3. For reproducibility, EPA
performed this sample-based imputation in R using the 'base::sample' function (R Core Team,
2021).
Following this process, EPA relies on sample point values recorded in UCMR 3 for 5,419
systems, SDWIS/Fed facility data for 43,563 systems, and imputed entry point values for 17,523
10 UCMR 3 required all systems serving more than 10,000 people to collect and analyze samples for PFOA, PFOS, PFNA,
PFHxS, PFBS, and PFHpA at each distribution system entry point. EPA also identified a stratified random sample of 800 small
systems serving up to 10,000 people to collect samples for these six PFAS.
Proposed PFAS Rule Economic Analysis
4-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
systems. All systems have at least one entry point. Among CWSs, the maximum number of entry
points is 202, and the mean is 1.80. Among NTNCWSs, the maximum number of entry points is
22, and the mean is 1.31.
Table 4-6 summarizes the final frequency distribution of entry point input ranges for each CWS
stratum of size and source water combination. Table 4-7 summarizes the final frequency
distribution of entry point input ranges for each NTNCW stratum of size and source water
combination. These distributions are used to proportionally assign numbers of entry points to
systems in each system size and type category.11
11 The SDWIS/Fed data provide information on the PWS characteristics that typically define PWS categories, or strata, for
which EPA develops costs in rulemakings. These characteristics include system type (CWS, NTNCWS), number of people
served by the PWS, PWS's primary raw water source (ground water or surface water), PWS's ownership type (public or private),
and PWS state. For more information on the use of baseline and compliance characteristics to define model systems in EPA's
cost analysis, please see Section 5.2.
Proposed PFAS Rule Economic Analysis
4-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-6: Frequency Distribution of Entry Point Inputs for CWSs
Ground Water
Surface Water
1 EP
2-5
6-10
11-15
16-20
21-
>100
1 EP
2-5
6-10
11-15
16-20
21-
>100
System Size
EP
EP
EP
EP
100
EP
EP
EP
EP
EP
100
EP
EP
EP
< 100
90%
10%
0.1%
0
0
0
0
87%
13%
0
0
0
0
0
101-500
76%
24%
0
0
0
0
0
84%
16%
0
0
0
0
0
501-1,000
62%
38%
0.5%
0
0
0
0
76%
23%
0.8%
0
0
0
0
1,001-3,300
48%
50%
1%
0
0
0
0
70%
30%
0.7%
0
0
0
0
3,301-10,000
32%
59%
8%
0.9%
0.1%
0
0
54%
43%
3%
0.5%
0.04%
0
0
10,001-50,000
3%
58%
28%
7%
3%
1%
0.07%
3%
82%
10%
2%
1%
0.6%
0
50,001-100,000
0
51%
25%
8%
8%
9%
0
0.2%
74%
13%
6%
2%
4%
0
100,001-1M
0
34%
22%
11%
8%
24%
1%
0.3%
67%
13%
4%
9%
6%
0.3%
Abbreviations: CWS - community water systems; EP - entry point.
Table 4-7: Frequency Distribution of Entry Point Inputs for NTNCWSs
Ground Water
Surface Water
System Size
1 EP
2-5 EP
6-10 EP
11-20 EP
>20 EP
1 EP
2-5 EP
6-10 EP
11-20 EP
>20 EP
< 100
84%
16%
0.4%
0
0
82%
18%
0
0
0
101-500
81%
19%
0
0
0
74%
26%
0
0
0
501-1,000
0
0
0
0
0
0
0
0
0
0
1,001-3,300
68%
30%
2%
0
0
61%
31%
8%
0
0
3,301-10,000
53%
44%
2%
1%
0
35%
44%
14%
6%
0
10,001-50,000
10%
80%
0
10%
0
30%
40%
5%
20%
5%
50,001-100,000
0
0
0
0
0
0
100%
0
0
0
100,001-1M
0
0
0
0
0
0
100%
0
0
0
Abbreviations: NTNCWS - non-transient non-community water systems; EP - entry point.
Proposed PFAS Rule Economic Analysis
4-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
4.3.3.2 Total Organic Carbon
The effectiveness of the GAC treatment process varies with the level of TOC in the influent
water. There is no national dataset containing TOC values for every CWS or NTNCWS.
Therefore, EPA randomly assigned a TOC level to each system based on two distributions of
TOC in 'finished' water. The Agency developed distributions using TOC data voluntarily
submitted by states in response to the SYR4 ICR drinking water regulations. Because TOC
levels in Ground Water are lower on average than TOC levels in Surface Water, EPA separated
the data by system primary source water. TOC levels can also vary throughout a system. Source
water TOC measurements can be higher than finished water estimates if a treatment process
removes TOC. For each system, EPA identified TOC measurements that best represented
finished water quality. Using the resulting distribution of Ground Water or Surface Water
estimates, EPA identified decile midpoint values to randomly assign to each system.
4.3.3.3 Average Daily Production Flow and Design Flow
Average daily production flow and design flow per system are based on regression equations
from EPA's Geometries and Characteristics of Public Water Supplies report (U.S. EPA, 2000).
The average daily flow and design flow are functions of the population served, with different
equations for source water type (Surface Water or Ground Water). Table 4-8 presents these flow
equations. The flow was then divided by the number of entry points to calculate the flow per
treatment plant for the system (assuming each entry point has one treatment plant). EPA does not
have comparable flow-population regression equations for NTNCWSs and, therefore, used the
CWS relationships to estimate flow for NTNCWSs.
Table 4-8: Functions for Design and Average Daily Flow by System Types
Design Flow Functions (kgal)
Surface Water system
Ground Water system
Design Flow = 0.59028 x Population" 94573
(or 2 x Average Flow, whichever is greater)
Design Flow = 0.54992 x Population" 95538
(or 2 x Average Flow, whichever is greater)
Average Daily Flow Functions (kgal)
Surface Water system
Average Flow = 0.14004 x Population11997113
Ground Water system
Average Flow = 0.08575 x Population 1115839
Abbreviations: Ground Water - ground water; Surface Water - surface water, kgal - 1000 gallons.
As an example, Table 4-9 shows the design flow and average daily flow results when applying
the regression equations to the average population per system for each CWS system stratum. The
results for NTNCWSs are in Table 4-10. Note that these results are examples only. In practice,
EPA applied the regression equations to the population served of individual systems, instead of
the stratum average population. In addition, for systems serving more than 1 million people, EPA
obtained publicly available system-specific information on the average daily flow and design
flow for each entry point whenever possible (e.g., annual Consumer Confidence Reports).
Proposed PFAS Rule Economic Analysis
4-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-9: Design and Average Daily Flow for CWSs
Ground Water
Surface Water
System Size
Average
Design
Flow
Average
Flow
Average
Design
Flow
Average
Flow
Population
(MGD)
(MGD)
Population
(MGD)
(MGD)
< 100
61
0.028
0.007
61
0.029
0.008
101-500
250
0.107
0.030
282
0.123
0.039
501-1,000
734
0.301
0.093
749
0.309
0.103
1,001-3,300
1,865
0.733
0.248
2,006
0.784
0.275
3,301-10,000
5,673
2.121
0.806
6,133
2.255
0.837
10,001-50,000
20,697
7.305
3.171
22,789
7.804
3.098
50,001-100,000
67,222
22.512
11.031
70,386
22.671
9.535
100,001-1M
203,821
71.371
35.685
244,022
73.470
32.937
Abbreviations: CWS - community water systems; MGD - million gallons per day.
Table 4-10: Design and Average Daily Flow for NTNCWSs
Ground Water
Surface Water
System Size
Average
Design
Flow
Average
Flow
Average
Design
Flow
Average
Flow
Population
(MGD)
(MGD)
Population
(MGD)
(MGD)
< 100
56
0.026
0.006
50
0.024
0.007
101-500
248
0.107
0.029
269
0.117
0.037
501-1,000
711
0.292
0.089
750
0.309
0.103
1,001-3,300
1,672
0.660
0.221
1,979
0.774
0.271
3,301-10,000
5,273
1.978
0.746
5,988
2.205
0.817
10,001-50,000
19,862
7.023
3.035
20,705
7.127
2.815
50,001-100,000
Not
Not
Not
71,963
23.151
9.748
applicable
applicable
applicable
100,001-1M
Not
Not
Not
203,375
61.841
27.465
applicable
applicable
applicable
Abbreviations: NTNCWS - non-transient non-community water systems; MGD - million gallons per day.
4.3.4 Public Water System Labor Rates
EPA recognizes that there may be variation in labor rates across all systems. However, for
purposes of this EA, EPA used national average estimates of unit labor from the 2006 CWSS,
with a few modifications described below. Prior labor unit costs for managerial, technical, and
Proposed PFAS Rule Economic Analysis
4-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
clerical labor in EPA's work breakdown structure12 (WBS) were based on a review of data from
three sources:
• The Occupational Employment Survey (OES), a semi-annual Bureau of Labor Statistics
(BLS) survey that provides hourly wage estimates by occupation and industry (BLS,
2022).
• The Water Utility Compensation Survey, an annual American Water Works Association
(AWWA) survey that provides hourly wage estimates for the water and wastewater
industry by occupation. Data are in 2008 dollars.
• The 2006 CWSS, a periodic EPA survey that obtains employment information from a
sample of CWSs.
There are more recent wage data from the OES and AWWA surveys, but there has not been a
CWSS since 2006. A 2020 review of the WBS labor rates found that the WBS wage rates in
2019 dollars overstate labor costs for clerical labor hours as well as potentially overstate labor
costs for technical labor hours (Abt Associates, 2020). Following these findings, EPA adjusted
the labor costs used in the WBS to reflect occupation-specific escalation factors rather than the
seasonally adjusted employment cost index (ECI) for all civilian employees. The WBS labor
costs for managerial hours were not clearly over- or understated compared to OES data but were
consistently lower than the AWWA wage estimates (Abt Associates, 2020).
Table 4-11 presents the labor rate estimates used in the WBS in 2007 dollars. Labor rates were
calculated for three occupation categories: technical, managerial, and clerical. The rates do not
include benefits.
Table 4-11: Hourly Wage Rates Based on CWSS Data ($2007)
Occupation <500 501- 3,301- 10,001- 50,001- >100,000
3,300 10,000 50,000 100,000
Technical $16.97 $16.97 $18.10 $19.11 $19.95 $23.32
Managerial $24.06 $24.06 $27.52 $30.65 $35.76 $38.21
Clerical $16.21 $16.21 $16.21 $20.93 $20.93 $20.93
Abbreviations: CWSS - Community Water System Survey.
Source: Abt Associates, 2020
A review of updated BLS Employer Cost for Employee Compensation (ECEC) data indicated
that benefits account for a higher proportion of total compensation today than they did at the end
of 2006 (Abt Associates, 2020). The WBS assumes a benefit multiplier of 1.45, which is the
2020 multiplier for all civilians working in service-producing industries (Abt Associates, 2020).
The benefit-loaded wage rates are shown in Table 4-12.
12 To estimate treatment costs, EPA uses several engineering models using a bottom-up approach known as work breakdown
structure (WBS). Hie WBS models derive system-level costs and provide EPA with comprehensive, flexible and transparent
tools to help estimate treatment costs.
Proposed PFAS Rule Economic Analysis
4-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-12: Hourly Labor Costs Including Wages Plus Benefits ($2007)
Occupation <500 501- 3,301- 10,001- 50,001- >100,000
3,300 10,000 50,000 100,000
Technical $24.61 $24.61 $26.25 $27.71 $28.93 $33.81
Managerial $34.89 $34.89 $39.90 $44.44 $51.85 $55.40
Clerical $23.50 $23.50 $23.50 $30.35 $30.35 $30.35
Source: Abt Associates, 2020
Because the WBS relies on 2020 dollar values, EPA escalated the CWSS values using the OES
occupation-specific change in mean wage rate from 2007 to 2020 instead of the general civilian
ECI escalation rate. The escalation for the technical rate is 35.2 percent and the escalation for the
clerical rate is 36.3 percent. The WBS managerial wage rates are consistent with OES rates, but
slightly lower than AWWA rates (Abt Associates, 2020). At the time of the analysis in 2020, the
OES occupation-specific wage escalation rate for the managerial rate was comparable to the ECI
rate (Abt Associates, 2020). Therefore, the WBS retains the ECI escalated managerial labor
rates, which for 2020 is 41.4 percent. The national cost-benefit analysis method described in
Section 5.2 presents all values in 2021 dollars. The method uses the gross domestic product
(GDP) implicit price deflator to adjust values in other dollar years to 2021 dollars. Therefore, the
labor costs including wages and benefits in 2021 dollars shown in Table 4-13 reflect an
additional adjustment for dollar year. EPA applied the same system labor rates to both CWSs and
NTNCWSs.
Table 4-13: Hourly Labor Costs Escalated to $2021
Occupation
<500
501-
3,301-10,000
10,001-
50,001-
> 100,000
3,300
50,000
100,000
Technical
35.48
35.48
37.84
39.94
41.70
48.74
Managerial
52.60
52.60
60.16
67.02
78.19
83.55
Clerical
34.17
34.17
34.17
44.12
44.12
44.12
Note:
EPA escalated the 2020 labor costs in the WBS models to 2021 dollars for use in the national cost-benefit analysis. The
adjustment multiplier based on the GDP implicit price deflator was 1.066, equal to the October 2021 value of 121.188 divided
by the 2020 annual value of 113.633 (U.S. Bureau of Economic Analysis, 2022).
There is uncertainty in the derivation of labor rates that could result in an over- or underestimate
of national costs of the proposed rule. The mean labor rate is based on findings of the 2006
CWSS. The labor rate mix may have changed since the time of the survey. EPA accounted for
general changes in cost of labor by adjusting 2007 values to 2020 using occupation-specific
escalators and the ECI where appropriate. There is also uncertainty in assuming a 1.45 benefits
multiplier; this may cause an under- or overestimation of cost of the proposed rule.
4.3.5 Cost of Capital
For the social cost-benefit analysis, EPA uses two alternative social discount rates, 3 percent and
7 percent to discount future values and annualize discounted present value over the period of
analysis. These rates are in accordance with EPA policy and guidance from OMB.
When evaluating the economic impacts on PWSs and households, however, EPA uses estimated
cost of capital to discount future costs and annualize the discounted present value over the
analysis period. This rate best represents the actual costs of compliance that systems will incur
Proposed PFAS Rule Economic Analysis
4-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
over time. To estimate PWS cost of capital, EPA used data from the 2006 CWSS. The CWSS
defined the following categories of funding sources:
• Current revenue;
• Equity or other funds from private investors;
• Department of Homeland Security (DHS) grant;
• Other government grants;
• Drinking Water State Revolving Fund (DWSRF), including loans and Principal
Repayment Forgiveness;
• Other borrowing from public sector sources; and
• Borrowing from private sectors sources.
EPA calculated the overall weighted average cost of capital (across all funding sources and loan
periods) for each size/ownership category, weighted by the percentage of funding from each
source.13 Table 4-14 shows the cost of capital for each CWS size category and ownership type.
Similar cost of capital information is not available for NTNCWS. Therefore, EPA used the CWS
cost of capital when calculating the annualized cost per NTNCWS.
Table 4-14: Weighted Average Cost of Capital by PWS Ownership and Size Category
Size Category Publicly Owned CWS Privately Owned CWS
<100
3.8%
7.8%
101-500
5.5%
8.2%
501-1,000
4.0%
8.6%
1,001-3,300
4.7%
7.1%
3,301-10,000
5.8%
7.0%
10,001-50,000
6.1%
7.0%
50,001-100,000
4.9%
6.9%
100,001-500,000
4.7%
3.9%
Over 500,000
3.7%
7.8%
Abbreviations: PWS - public water system; CWS - community water system.
Since the CWSS data collection, Congress established new programs and expanded funding for
existing programs. These funding sources allow PWSs to lower their cost of capital. These
include the DWSRF, the Water Infrastructure Finance and Innovation (WIFIA) program, the
Water Infrastructure Improvements for the Nation Act of 2016 (WIIN Act), and the Bipartisan
Infrastructure Law of 2021 (BIL).
Through the DWSRF Program, the EPA allocates annual capitalization grants to states. The
grants, along with state matching monies, support a dedicated loan fund to finance eligible water
system infrastructure improvement projects. States are permitted to use funding from their
DWSRF to help PWS finance water treatment through low-interest loans. The WIFIA program
provides creditworthy PWSs access to low-interest direct federal loans for water treatment
investment. The WIIN Act established a grant program to help small, underserved, and
disadvantaged communities achieve compliance with drinking water standards. Additionally, the
13 See "Cost of Capital Approach.doc" in the docket for details of how the cost of capital estimates were developed.
Proposed PFAS Rule Economic Analysis
4-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Bipartisan Infrastructure Law (BIL) (P.L. 117-58) authorizes $5 billion as part of the Emerging
Contaminants in Small or Disadvantaged Communities grant program that can be used to reduce
PFAS in drinking water in communities facing disproportionate impacts. BIL funds will be
provided as grants and loan forgiveness associated with PFAS drinking water treatment capital
expenditures. Therefore, the actual cost of capital faced by some PWSs may be lower than those
used in this analysis.
4.4 Occurrence of PFAS
EPA's Technical Support Document for PFAS Occurrence provides estimates of the baseline
PFAS occurrence in PWSs (U.S. EPA, 2023g). After reviewing the available data on PFAS in
drinking water, EPA determined that the data from the UCMR 3 are the best available nationally
representative data to characterize the occurrence of multiple PFAS in drinking water. Consistent
with the Agency's commitment in the final regulatory determination for PFOA and PFOS and
EPA's PFAS Strategic Roadmap to present the best available occurrence information, the
Agency supplemented the UCMR 3 data with data collected by states that have made their data
publicly available (U.S. EPA, 2021b; U.S. EPA, 2021e).
This section summarizes the EPA's PFAS occurrence analysis (U.S. EPA, 2023g). Section 4.4.1
provides an overview of UCMR 3 and its PFAS occurrence data. Section 4.4.2 provides an
overview of state PFAS monitoring data. Section 4.4.4 summarizes EPA's analysis of PFAS
drinking water occurrence data. Section 4.4.5 summarizes the national PFAS occurrence
estimates used in the cost and benefit analyses.
4.4.1 Overview of UCMR 3 Data
The UCMR is a national drinking water monitoring program administered by EPA. The UCMR
3 monitoring cycle included a census of all large CWSs and NTNCWSs (i.e., those serving more
than 10,000 people) and a statistical sample of 800 small CWSs and NTNCWSs (i.e., those
serving 10,000 people or fewer). Monitoring under UCMR 3 occurred from 2013 to 2015. More
information on the UCMR 3 study design and data analysis can be found in U.S. EPA (2012) and
U.S. EPA (2019c).
EPA collected the UCMR 3 data from PWSs in all 50 states and seven additional primacy
agencies. UCMR 3 monitoring occurrence data are available for six PFAS: PFOS, PFOA, PFNA,
PFHxS, PFHpA, and PFBS. For the individual PFAS contaminants, EPA collected nearly 37,000
finished water samples from 4,920 PWSs.
Systems collected PFAS samples at each entry point to their customer distribution system. Entry
points are the point of compliance for the proposed rule, and systems can have multiple entry
points. The sampling frequency varied by source water: four quarterly samples in a one-year
period for surface water systems, and two samples at least six months apart for ground water
systems.
EPA's Technical Support Document for PFAS Occurrence (U.S. EPA, 2023g) describes the data
and analyses that EPA used to develop national estimates of PFAS occurrence in public drinking
water systems using UCMR 3 data.
Proposed PFAS Rule Economic Analysis
4-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
4.4.2 Overview of State PFAS Data
Outside of the UCMR 3 data collection, many states have undertaken individual efforts to
monitor for PFAS in both source and finished drinking water. EPA collected data from 23 states
that made their data publicly available as of August 2021; this action was in alignment with the
Agency's commitment in the final regulatory determination for PFOA and PFOS and its PFAS
Strategic Roadmap to present the best available information on sampling for PFAS in water
systems. EPA notes that this data collection cutoff was made to allow sufficient time for the
Agency to conduct analyses on the state information for the proposed NPDWR. Due to the
limitations in representation and reporting of some of the available data, EPA conducted
technical analyses using a subset of the available state data. These more recent state data,
collected using improved analytical methods that have lower reporting limits than under UCMR
3, show widespread occurrence of PFOA, PFOS, PFBS, PFNA, and PFHxS in multiple
geographic locations. These data also show that these PFAS occur with substantial frequency at
lower concentrations than were analyzed under UCMR 3, as demonstrated within EPA's
Technical Support Document for PFAS Occurrence (U.S. EPA, 2023g). Furthermore, these state
data include results for more PFAS than were included in the UCMR 3, including HFPO-DA.
EPA's analysis of state PFAS data shows occurrence in multiple geographic locations consistent
with what was observed during UCMR 3 monitoring. The Agency notes that the data vary in
terms of quantity and coverage; for example, some of these available data are from targeted
sampling efforts (i.e., monitoring in areas of known or potential contamination) and thus may not
be representative of levels found in all PWSs within the state. Summaries on the non-targeted
state PFAS finished water data are available in Table 4-15 and Table 4-16. Specifically, a
summary on the percent of samples in state datasets that were above detection limits for select
PFAS is provided in Table 4-15, and a summary on the number of systems in state datasets that
had detections for select PFAS is available in Table 4-16. Comprehensive summaries of state
data are available within EPA's Technical Support Document for PFAS Occurrence (U.S. EPA,
2023g).
Table 4-15: Non-Targeted State PFAS Finished Water Data - Summary of Samples with
Detections of PFAS Proposed for Regulation
State
PFHxS
PFNA
PFBS
HFPO-DA3
Colorado
10.8%
0.9%
11.0%
0.2%
Illinois
5.1%
0.2%
7.8%
0.0%
Kentucky
8.6%
2.5%
13.3%
13.6%
Massachusetts
31.9%
4.6%
35.5%
0.0%
Michigan
2.9%
0.1%
5.2%
0.04%
New Hampshire
12.9%
2.5%
22.7%
1.8%
New Jersey
24.7%
8.0%
24.9%
N/A
North Dakota
1.6%
0.0%
0.0%
0.0%
Ohio
5.8%
0.3%
4.7%
0.1%
South Carolina
13.5%
2.1%
38.3%
6.0%
Vermont
2.2%
1.7%
4.8%
0.2%
Abbreviations: PFAS - per- and polyfluoroalkyl substances.
Note:
aN/A indicates that no data are available. 0.0 % indicates that monitoring data were available for the compound/state but there
were no detections.
Proposed PFAS Rule Economic Analysis
4-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-16: Non-Targeted State PFAS Finished Water Data - Summary of Systems with
Detections of Select PFAS
State
PFHxS
PFNA
PFBS
HFPO-DA3
Colorado
13.4%
1.0%
13.4%
0.3%
Illinois
4.3%
0.2%
6.6%
0.0%
Kentucky
8.6%
2.5%
12.3%
13.6%
Massachusetts
30.2%
8.4%
39.4%
0.0%
Michigan
3.0%
0.2%
5.3%
0.1%
New Hampshire
17.6%
4.4%
26.1%
1.7%
New Jersey
32.6%
13.3%
34.0%
N/A
North Dakota
1.6%
0.0%
0.0%
0.0%
Ohio
2.2%
0.3%
2.4%
0.1%
South Carolina
20.0%
6.1%
56.0%
10.9%
Vermont
1.6%
1.3%
5.2%
0.5%
Abbreviations: PFAS - per-and polyfluoroalkyl substances.
Note:
aN/A indicates that no data are available. 0.0 % indicates that monitoring data were available for the compound/state but there
were no detections.
4.4.3 Overview of PFAS Co-Occurrence
Co-occurrence of multiple PFAS has been reported in drinking water, ambient surface waters,
aquatic organisms, biosolids (sewage sludge), and other environmental media. PFOA and PFOS
have historically been target analytes, which has partly contributed to their prevalence in
environmental monitoring studies, although some recent monitoring studies have begun to focus
on additional PFAS via advanced analytical instruments/methods and non-targeted analysis
(McCord et al., 2019; McCord et al., 2020).
EPA's analysis on PFAS co-occurrence using UCMR 3 data found that 4 percent of PWSs
reported results for which one or more of the six UCMR 3 PFAS were measured at or above their
respective minimum reporting levels. Additionally, several studies have demonstrated PFAS co-
occurrence in finished drinking water (Adamson et al., 2017; Cadwallader et al., 2022; Guelfo et
al., 2018). One study in particular used UCMR 3 data to demonstrate that two or more of the six
PFAS monitored under UCMR 3 co-occurred in 48 percent (285/598) of sampling events with
PFAS detected, and PFOA and PFOS co-occurred in 27 percent (164/598) of sampling events
with two or more PFAS detected (Guelfo et al., 2018).
For additional discussion and analysis on PFAS co-occurrence, reference EPA's Technical
Support Document for PFAS Occurrence (U.S. EPA, 2023g).
4.4.4 Summary of PFAS Occurrence Data Analysis
Identifying the systems and population exposed to PFAS exceeding the limits under the proposed
option and the three regulatory alternatives is a key step to estimating benefits and costs of the
proposed NPDWR. EPA used a Bayesian hierarchical Markov chain Monte Carlo (MCMC)
occurrence model to estimate national PFAS occurrence in PWSs. EPA used the MCMC
occurrence model output to estimate the PWSs and entry points with PFAS occurrence
exceeding the limits under the proposed option and regulatory alternatives. EPA assumed that the
Proposed PFAS Rule Economic Analysis
4-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
populations served by these PWSs were exposed to the PFAS concentration estimates generated
by the MCMC occurrence model.
This section summarizes the occurrence model and EPA's use of the model to identify the
systems and entry points with PFAS occurrence exceeding the regulatory alternatives considered
within the EA, as well as the corresponding populations exposed. Further details on the MCMC
model are available in Appendix A, Cadwallader et al. (2022), and U.S. EPA (2023g).
Data collected under UCMR 3 served as the primary dataset for the MCMC occurrence model
due to its nationally representative design. Additionally, EPA incorporated state PFAS
monitoring datasets to supplement UCMR 3 data in the occurrence model. These state datasets,
for which the monitoring has been conducted more recently than UCMR 3, generally have lower
reporting limits because the analytical methods have improved over the last 10 years, allowing
laboratories to reliably measure PFAS at concentrations approximately 5 to 20 times lower than
for UCMR 3. Thus, state datasets with lower reporting limits than those in UCMR 3 helped
inform the model by enabling observation of PFAS occurrence at lower concentrations. State
datasets also consist of more-recent samples than UCMR 3, which broadened the temporal range
of data used to fit the model. The supplemental state data were limited to samples collected from
systems that were also in UCMR 3 to prevent biasing the dataset toward states for which the data
from additional PWSs were available as well as maintain the nationally representative set of
systems selected for UCMR 3. Using these criteria, 17 states were identified as having some
state monitoring data to be included in fitting the national occurrence model.
The dataset used to fit the model included all data available in the final UCMR 3 dataset for
PFOS, PFOA, PFHpA, and PFHxS. This amounted to 36,972 samples each for PFOS, PFOA,
and PFHpA, and 36,971 samples for PFHxS. Of these four PFAS, 1,114 samples had results
reported at or above the UCMR 3 MRLs. The additional state datasets included to supplement
the UCMR 3 data contained 6,645 PFOS samples, 6,656 PFOA samples, 4,715 PFHpA samples,
and 5,114 PFHxS samples collected at systems that were included in UCMR 3. Of these samples,
2,200 (33%) were reported values for PFOS, 2,694 (40%) were reported values for PFOA, 932
(20%) were reported values for PFHpA, and 1,269 (25%) were reported values for PFHxS. The
remainder were listed as being below their respective reporting limits. A summary of the state
data used in the occurrence model, including system and sample counts, is available in Appendix
A.
Some states have promulgated drinking water standards for PFAS since the UCMR 3
monitoring. EPA reviewed state websites and identified states with enacted standards for the
PFAS compounds considered within the regulatory alternatives discussed in the EA. Table 4-17
summarizes state regulations on PFAS in drinking water, which are current as of July 2022. The
state PFAS regulation summary in Table 4-17 is reflective of only those states that have
promulgated PFAS drinking water regulations and does not include information from states that
have proposed PFAS drinking water regulations or issued guidance for PWSs.
Proposed PFAS Rule Economic Analysis
4-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-17: State PFAS Regulations
State
PFOA
PFOS PFBS
Regulated PFAS Levels (ppt)
PFHpA PFHxA PFHxS PFNA PFDA
HFPO- c.
_ . Sum
DA
New Jersey
14
13
13
Vermont1
*
*
* *
*
20
New
12
15
18
11
Hampshire
Massachusetts3
*
*
* *
* *
20
Michigan
8
16 420
400,000 51
6
370
New York
10
10
Abbreviations: PFAS - per-and polyfluoroalkyl substances.
Notes:
aAsterisks (*) indicate states that regulate PFAS compounds at an overall threshold value, as indicated in the Sum column.
Sources: State websites are as follows - New Jersey
(httvs://www.ni.zov/health/ceohs/documents/vfas drinkine%20water.vdf. Vermont (httvs://dec.Vermont.zov/water/drinkinz-
water/water-aualit\>-monitorinz/vfas). New Hampshire (httvs://wM'wl.nhwwa.orz/wv-content/iivloads/NHWWA-Water-is-
Essential-Seminar-Oct-20-202Q-PFAS-Arsenic-Ride-Uvdates.vdf). Massachusetts (httvs://www.mass, zov/lists/develovment-
of-a-vfas-drinkinz-water-standard-mcl#final-vfas-mcl-reziilations-). Michigan
(httvs://www.michizan.zov/vfasresvonse/drinkinz-water/mcl). New York
(httvs://www.health.nv.zov/environmental/water/drinkinz/docs/water suvvlier fact sheet new mcls.vdf).
To estimate the costs and benefits of the proposed rule, EPA assumed that all MCMC occurrence
model estimates exceeding state limits are equivalent to the state-enacted limit. For these states,
EPA assumed that the state MCL is the maximum baseline PFAS occurrence value for all entry
points in the state. This adjustment was made to the MCMC occurrence model PFAS estimates
for PFOA, PFOS, and PFHxS in this EA. Since the proposed rule standards are more stringent
than current state drinking water standards, systems in states with PFAS regulations are still
expected to incur incremental costs to comply with the proposed rule, although the estimated
compliance costs will be less compared to costs that do not adjust the MCMC occurrence data to
reflect the state MCLs. Similarly, populations served by PWSs in the states with PFAS
regulations are expected to benefit from further reductions in PFAS exposures, although the
incremental benefits for these populations will be less compared to benefits that do not adjust the
MCMC occurrence data to reflect the state MCLs.
EPA used system-level distributions, as described in Cadwallader et al. (2022), to simulate entry
point concentrations and estimate PFAS occurrence relative to the regulatory alternatives and
proposed option limits. EPA assumed entry point concentrations were constant. Simulated
sample data are composed of a set of 4,000 iterations with the number of simulated samples per
system within each iteration equal to the number of entry points. EPA estimated within system
variation from all available samples within each system as part of the model fitting process.
Although the data used to fit the model may have included multiple samples over time or entry
points, this simulation strategy assumes that all within-system variability is related to entry point.
For 4,920 systems with means fitted by the model (i.e., systems with PFAS data in UCMR 3),
EPA simulated system-specific samples based on the best-fit model. EPA simulated from the
high level multivariate normal distribution to produce means for each chemical at each non-
UCMR system and then used those distributions to simulate system-specific samples. The
Agency then generated random samples from the multivariate distribution and the value of the
fixed parameters for each iteration. The exception to this approach was systems serving more
Proposed PFAS Rule Economic Analysis
4-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
than 1 million people. For these systems, EPA used UCMR 3 and more recent monitoring data to
identify the entry points that might require PFAS removal. These relatively few very large
systems have the potential to affect aggregate costs and, therefore, require more precision in
baseline occurrence estimates.
4.4.5 Summary of National PFAS Occurrence
Using the MCMC occurrence model, EPA estimated baseline occurrence to understand changes
in occurrence and exposure for the proposed option and the regulatory alternative MCLs under
Options la - lc. These estimates vary across the 4,000 MCMC occurrence model iterations,
thereby characterizing baseline occurrence uncertainty. In addition, for PWSs in states with
existing MCLs for PFOA, PFOS, and PFHxS, EPA capped contaminant concentrations at the
state MCLs.
The estimated number of PWSs with at least one entry point above the MCL or HI are provided
in Table 4-18 through Table 4-21, while the total estimated number of entry points above the
MCL or HI are provided in Table 4-22 through Table 4-25. In Table 4-26 through Table 4-29,
EPA provides the population served by PWSs with at least one entry point above the MCL or HI.
The population served by entry points above the MCL or HI are provided in Table 4-30 through
Table 4-33. Each table provides expected value estimates as well as 5th percentile and 95th
percentile estimates that characterize the uncertainty of baseline PFAS occurrence.
Proposed PFAS Rule Economic Analysis
4-24
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-18: Total Systems Impacted, Proposed Option (PFOA and PFOS MCLs
of 4.0 ppt and HI of 1.0)
5th Percentile
Mean
Small Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs With Hazard Index Exceedance3
PWSs That Exceed One or More MCLs
Large Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs With Hazard Index Exceedance3
PWSs That Exceed One or More MCLs
All Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs With Hazard Index Exceedance3
PWSs That Exceed One or More Limits
61,463
1,801
836
145
2,115
4,433
721
803
178
978
65,896
2,522
1,639
323
3,093
61,463
2,905
1,520
320
3,259
4,433
791
878
207
1,062
65,896
3,696
2,399
528
4,321
95th
Percentile
61,463
4,260
2,422
563
4,699
4,433
868
959
239
1,150
65,896
5,128
3,381
802
5,849
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level; HI - hazard index.
Note:
aHazard Index exceedance is triggered by perfluorohexane sulfonate (PFHxS) occurrence estimates from the
Markov chain Monte Carlo (MCMC) occurrence model.
Table 4-19: Total Systems Impacted, Option la (PFOA and PFOS MCLs of 4.0
ppt)
Small Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
Large Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
All Systems
5th Percentile
61,463
1,801
836
2,111
4,433
721
803
975
Mean
61,463
2,905
1,520
3,251
4,433
791
878
1,060
95th
Percentile
61,463
4,260
2,422
4,676
4,433
868
959
1,145
Proposed PFAS Rule Economic Analysis
4-25
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-19: Total Systems Impacted, Option la (PFOA and PFOS MCLs of 4.0
ppt)
5th Percentile
Mean
95th
Percentile
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
65,896
2,522
1,639
3,086
65,896
3,696
2,399
4,310
65,896
5,128
3,381
5,821
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-20: Total Systems Impacted, Option lb (PFOA and PFOS MCLs of 5.0
ppt)
5th Percentile
Mean
95th
Percentile
Small Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
Large Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
All Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
PWSs That Exceed One or More MCLs
61,463
1,307
542
1,518
4,433
597
634
803
65,896
1,904
1,176
2,321
61,463
2,197
1,025
2,428
4,433
657
696
871
65,896
2,855
1,721
3,300
61,463
3,268
1,683
3,557
4,433
722
762
947
65,896
3,990
2,445
4,504
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-21: Total Systems Impacted, Option lc (PFOA and PFOS MCLs of 10.0
ppt)
5th Percentile
Small Systems
Total Number of PWSs
PWSs With PFOS Exceedance
PWSs With PFOA Exceedance
61,463
437
107
Mean
95th
Percentile
61,463
801
238
61,463
1,275
429
Proposed PFAS Rule Economic Analysis
4-26
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 4-21: Total Systems Impacted, Option lc (PFOA and PFOS MCLs of 10.0
ppt)
5th Percentile
Mean
95th
Percentile
PWSs That Exceed One or More MCLs
473
852
1,347
Large Systems
Total Number of PWSs
4,433
4,433
4,433
PWSs With PFOS Exceedance
293
330
369
PWSs With PFOA Exceedance
256
288
322
PWSs That Exceed One or More MCLs
382
422
464
All Systems
Total Number of PWSs
65,896
65,896
65,896
PWSs With PFOS Exceedance
730
1,130
1,644
PWSs With PFOA Exceedance
363
526
751
PWSs That Exceed One or More MCLs
855
1,274
1,811
Abbreviations: PFOA - perfluorooctanoic acid; PFOS ¦
- perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-22: Total Entry Points Impacted, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0)
5th Percentile
Mean
95th
Percentile
Small Systems
Total Number of Entry Points
87,895
87,895
87,895
Entry Points With PFOS Exceedance
2,294
3,768
5,520
Entry Points With PFOA Exceedance
1,051
1,913
3,040
Entry Points With Hazard Index Exceedance3
166
379
674
Entry Points That Exceed One or More MCLs
2,803
4,354
6,269
Large Systems
Total Number of Entry Points
22,441
22,441
22,441
Entry Points With PFOS Exceedance
1,812
1,981
2,156
Entry Points With PFOA Exceedance
1,932
2,107
2,296
Entry Points With Hazard Index Exceedance3
467
533
600
Entry Points That Exceed One or More MCLs
3,110
3,356
3,613
All Systems
Total Number of Entry Points
110,336
110,336
110,336
Entry Points With PFOS Exceedance
4,106
5,749
7,676
Entry Points With PFOA Exceedance
2,983
4,019
5,336
Entry Points With Hazard Index Exceedance3
633
912
1,274
Entry Points That Exceed One or More MCLs
5,913
7,710
9,882
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level; HI - hazard index.
Note:
Proposed PFAS Rule Economic Analysis
4-27
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-22: Total Entry Points Impacted, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0)
5th Percentile Mean 95th
Percentile
aHazard Index exceedance is triggered by perfluorohexane sulfonate (PFHxS) occurrence estimates from the
Markov chain Monte Carlo (MCMC) occurrence model.
Table 4-23: Total Entry Points Impacted, Option la (PFOA and PFOS MCLs
of 4.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Number of Entry Points 87,895 87,895 87,895
Entry Points With PFOS Exceedance 2,294 3,768 5,520
Entry Points With PFOA Exceedance 1,051 1,913 3,040
Entry Points That Exceed One or More MCLs 2,760 4,327 6,208
Large Systems
Total Number of Entry Points 22,441 22,441 22,441
Entry Points With PFOS Exceedance 1,812 1,981 2,156
Entry Points With PFOA Exceedance 1,932 2,107 2,296
Entry Points That Exceed One or More MCLs 3,004 3,238 3,487
All Systems
Total Number of Entry Points 110,336 110,336 110,336
Entry Points With PFOS Exceedance 4,106 5,749 7,676
Entry Points With PFOA Exceedance 2,983 4,019 5,336
Entry Points That Exceed One or More MCLs 5,764 7,564 9,695
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-24: Total Entry Points Impacted, Option lb (PFOA and PFOS MCLs
of 5.0 ppt)
95th
Percentile
5th Percentile
Mean
Small Systems
Total Number of Entry Points
Entry Points With PFOS Exceedance
Entry Points With PFOA Exceedance
Entry Points That Exceed One or More MCLs
Large Systems
Total Number of Entry Points
Entry Points With PFOS Exceedance
87,895
1,704
668
2,000
22,441
1,464
87,895
2,840
1,286
3,220
22,441
1,603
87,895
4,242
2,077
4,730
22,441
1,751
Proposed PFAS Rule Economic Analysis
4-28
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-24: Total Entry Points Impacted, Option lb (PFOA and PFOS MCLs
of 5.0 ppt)
5th Percentile Mean 95th
Percentile
Entry Points With PFOA Exceedance
Entry Points That Exceed One or More MCLs
All Systems
Total Number of Entry Points
Entry Points With PFOS Exceedance
Entry Points With PFOA Exceedance
Entry Points That Exceed One or More MCLs
1,467
2,386
110,336
3,168
2,135
4,386
1,603
2,579
110,336
4,443
2,889
5,799
1,748
2,777
110,336
5,993
3,825
7,507
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-25: Total Entry Points Impacted, Option lc (PFOA and PFOS MCLs of
10.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Number of Entry Points 87,895 87,895 87,895
Entry Points With PFOS Exceedance 547 1,026 1,662
Entry Points With PFOA Exceedance 136 299 541
Entry Points That Exceed One or More MCLs 638 1,119 1,762
Large Systems
Total Number of Entry Points 22,441 22,441 22,441
Entry Points With PFOS Exceedance 678 756 836
Entry Points With PFOA Exceedance 534 595 658
Entry Points That Exceed One or More MCLs 1,039 1,134 1,235
All Systems
Total Number of Entry Points 110,336 110,336 110,336
Entry Points With PFOS Exceedance 1,225 1,782 2,498
Entry Points With PFOA Exceedance 670 893 1,199
Entry Points That Exceed One or More MCLs 1,677 2,253 2,997
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Proposed PFAS Rule Economic Analysis
4-29
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-26: Total Population at PWSs Impacted, Proposed Option (PFOA and
PFOS MCLs of 4.0 ppt and HI of 1.0)
5th Percentile
Mean
Small Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
Population Impacted by Hazard Index
Exceedance3
Population Impacted by One or More MCL
Exceedances
Large Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
Population Impacted by Hazard Index
Exceedance3
Population Impacted by One or More MCL
Exceedances
All Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
Population Impacted by Hazard Index
Exceedance3
Population Impacted by One or More MCL
Exceedances
57,897,900
2,087,271
1,146,222
234,852
2,510,966
215,603,000
40,925,500
44,865,200
11,250,000
54,331,100
273,500,900
43,012,771
46,011,422
11,484,852
56,842,066
57,897,900
3,264,073
1,938,415
493,057
3,752,014
215,603,000
46,523,900
50,710,300
13,769,700
60,630,000
273,500,900
49,787,973
52,648,715
14,262,757
64,382,014
95th
Percentile
57,897,900
4,701,130
2,971,819
840,765
5,199,508
215,603,000
52,256,600
56,793,900
16,474,900
67,160,000
273,500,900
56,957,730
59,765,719
17,315,665
72,359,508
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level; HI - hazard index.
Note:
^Hazard Index exceedance is triggered by perfluorohexane sulfonate (PFHxS) occurrence estimates from the
Markov chain Monte Carlo (MCMC) occurrence model.
Table 4-27: Total Population at PWSs Impacted, Option la (PFOA and PFOS
MCLs of 4.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 2,087,271 3,264,073 4,701,130
Population Impacted by PFOA Exceedance 1,146,222 1,938,415 2,971,819
Population Impacted by One or More MCL 2,482,756 3,735,146 5,174,268
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Proposed PFAS Rule Economic Analysis
4-30
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-27: Total Population at PWSs Impacted, Option la (PFOA and PFOS
MCLs of 4.0 ppt)
5th Percentile Mean 95th
Percentile
Population Impacted by PFOS Exceedance 40,925,500 46,523,900 52,256,600
Population Impacted by PFOA Exceedance 44,865,200 50,710,300 56,793,900
Population Impacted by One or More MCL 54,219,800 60,480,200 67,106,000
Exceedances
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 43,012,771 49,787,973 56,957,730
Population Impacted by PFOA Exceedance 46,011,422 52,648,715 59,765,719
Population Impacted by One or More MCL 56,702,556 64,215,346 72,280,268
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-28: Total Population at PWSs Impacted, Option lb (PFOA and PFOS
MCLs of 5.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 1,547,309 2,495,969 3,630,458
Population Impacted by PFOA Exceedance 767,919 1,339,283 2,091,861
Population Impacted by One or More MCL 1,845,024 2,821,792 4,008,112
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Population Impacted by PFOS Exceedance 34,492,900 39,513,400 44,694,900
Population Impacted by PFOA Exceedance 36,129,300 41,217,600 46,370,500
Population Impacted by One or More MCL 45,034,700 50,937,000 56,823,600
Exceedances
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 36,040,209 42,009,369 48,325,358
Population Impacted by PFOA Exceedance 36,897,219 42,556,883 48,462,361
Population Impacted by One or More MCL 46,879,724 53,758,792 60,831,712
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Proposed PFAS Rule Economic Analysis
4-31
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-29: Total Population at PWSs Impacted, Option lc (PFOA and PFOS
MCLs of 10.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 540,037 950,658 1,469,750
Population Impacted by PFOA Exceedance 169,217 344,601 579,202
Population Impacted by One or More MCL 599,217 1,032,176 1,574,182
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Population Impacted by PFOS Exceedance 17,858,800 21,145,500 24,589,600
Population Impacted by PFOA Exceedance 15,387,800 18,369,100 21,638,200
Population Impacted by One or More MCL 23,155,800 26,728,800 30,481,500
Exceedances
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 18,398,837 22,096,158 26,059,350
Population Impacted by PFOA Exceedance 15,557,017 18,713,701 22,217,402
Population Impacted by One or More MCL 23,755,017 27,760,976 32,055,682
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
Table 4-30: Total Population at Entry Points Impacted, Proposed Option
(PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 1,522,862 2,491,841 3,659,561
Population Impacted by PFOA Exceedance 716,698 1,283,316 2,040,113
Population Impacted by Hazard Index 110,444 256,444 463,178
Exceedance3
Population Impacted by One or More MCL 1,898,416 2,905,970 4,124,296
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Population Impacted by PFOS Exceedance 15,309,500 17,333,700 19,400,000
Population Impacted by PFOA Exceedance 17,494,600 19,653,500 21,865,800
Population Impacted by Hazard Index 3,242,290 3,991,870 4,817,620
Exceedance3
Population Impacted by One or More MCL 26,877,800 29,883,500 32,989,200
Exceedances
Proposed PFAS Rule Economic Analysis
4-32
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-30: Total Population at Entry Points Impacted, Proposed Option
(PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0)
5th Percentile Mean 95th
Percentile
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 16,832,362 19,825,541 23,059,561
Population Impacted by PFOA Exceedance 18,211,298 20,936,816 23,905,913
Population Impacted by Hazard Index 3,352,734 4,248,314 5,280,798
Exceedance3
Population Impacted by One or More MCL 28,776,216 32,789,470 37,113,496
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; MCL - maximum
contaminant level; HI - hazard index.
Note:
aHazard Index exceedance is triggered by perfluorohexane sulfonate (PFHxS) occurrence estimates from the
Markov chain Monte Carlo (MCMC) occurrence model.
Table 4-31: Total Population at Entry Points Impacted, Option la (PFOA and
PFOS MCLs of 4.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 1,522,862 2,491,841 3,659,561
Population Impacted by PFOA Exceedance 716,698 1,283,316 2,040,113
Population Impacted by One or More MCL 1,876,207 2,885,852 4,135,782
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Population Impacted by PFOS Exceedance 15,309,500 17,333,700 19,400,000
Population Impacted by PFOA Exceedance 17,494,600 19,653,500 21,865,800
Population Impacted by One or More MCL 26,160,300 29,117,300 32,135,400
Exceedances
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 16,832,362 19,825,541 23,059,561
Population Impacted by PFOA Exceedance 18,211,298 20,936,816 23,905,913
Population Impacted by One or More MCL 28,036,507 32,003,152 36,271,182
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; MCL - maximum
contaminant level.
Proposed PFAS Rule Economic Analysis
4-33
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-32: Total Population at Entry Points Impacted, Option lb (PFOA and
PFOS MCLs of 5.0 ppt)
5th Percentile Mean 95th
Percentile
Small Systems
Total Population 57,897,900 57,897,900 57,897,900
Population Impacted by PFOS Exceedance 1,098,901 1,877,218 2,830,317
Population Impacted by PFOA Exceedance 456,340 862,265 1,409,382
Population Impacted by One or More MCL 1,332,730 2,145,682 3,143,289
Exceedances
Large Systems
Total Population 215,603,000 215,603,000 215,603,000
Population Impacted by PFOS Exceedance 12,230,900 13,904,100 15,676,000
Population Impacted by PFOA Exceedance 13,161,700 14,889,700 16,671,200
Population Impacted by One or More MCL 20,620,500 23,031,100 25,487,300
Exceedances
All Systems
Total Population 273,500,900 273,500,900 273,500,900
Population Impacted by PFOS Exceedance 13,329,801 15,781,318 18,506,317
Population Impacted by PFOA Exceedance 13,618,040 15,751,965 18,080,582
Population Impacted by One or More MCL 21,953,230 25,176,782 28,630,589
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; MCL - maximum
contaminant level.
Table 4-33: Total Population at Entry Points Impacted, Option lc (PFOA and
PFOS MCLs of 10.0 ppt)
5th Percentile
Mean
Small Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
Population Impacted by One or More MCL
Exceedances
Large Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
Population Impacted by One or More MCL
Exceedances
All Systems
Total Population
Population Impacted by PFOS Exceedance
Population Impacted by PFOA Exceedance
57,897,900
361,876
86,354
409,590
215,603,000
5,239,470
4,414,550
8,491,400
273,500,900
5,601,346
4,500,904
57,897,900
680,278
199,750
745,161
215,603,000
6,228,730
5,309,960
9,750,100
273,500,900
6,909,008
5,509,710
95th
Percentile
57,897,900
1,092,021
359,333
1,179,156
215,603,000
7,268,400
6,230,660
11,090,400
273,500,900
8,360,421
6,589,993
Proposed PFAS Rule Economic Analysis
4-34
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-33: Total Population at Entry Points Impacted, Option lc (PFOA and
PFOS MCLs of 10.0 ppt)
5th Percentile Mean 95th
Percentile
Population Impacted by One or More MCL 8,900,990 10,495,261 12,269,556
Exceedances
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; PWS - public water
system; MCL - maximum contaminant level.
4.5 Uncertainties in the Baseline and Compliance
Characteristics of Systems
This section summarizes limitations and uncertainties of the baseline analysis. In the chapter,
EPA described how the quantitative analysis incorporates some sources of uncertainty. The
agency also noted data limitations that introduce uncertainty because information is not available
for the baseline analysis. Table 4-34 provides a summary of sources that have quantifiable
uncertainty and data limitations.
EPA notes that in most cases it is not possible to determine the extent to which a particular
limitation or uncertainty can affect the magnitude of the baseline conditions. EPA notes the
potential direction of the impact on baseline inputs to the costs and/or benefits analysis when
possible, but the Agency does not prioritize the entries with respect to the impact magnitude.
Table 4-34: Limitations and Uncertainties That Apply to the Baseline Analysis for the
Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Quantitative
Analysis
Notes
The Agency assigned
Ground Water as the source
to systems missing source
water information.
Underestimate costs
The design and average flow equations for Ground Water
systems result in lower flow estimates than the equations
for Surface Water systems. If any of the systems assigned
Ground Water source are in fact Surface Water systems,
then the flow estimates used in the cost analysis will be
underestimated. In addition, initial monitoring costs will be
underestimated for small Surface Water systems that are
assigned as a Ground Water source.
SDWIS/Fed retail
populations used for
baseline analysis
Overestimate costs
EPA did not reallocate populations for purchased water
systems to the wholesale suppliers. All systems are in the
inventory with their respective retail populations. In
general, this will result in extra systems with small
populations in the analysis and smaller populations at the
wholesale systems. Both results will tend to increase cost
estimate because the cost curves reflect economies of
scale.
SDWIS/Fed data quality
Uncertain impact on
baseline number of systems
and entry points
EPA periodically reviews inventory information in
SDWIS/Fed (U.S. EPA, 2021h) and has generally found a
high level of completeness and accuracy. There is
uncertainty, however, in some of the population and
facility data reported per system. To address this, EPA
removed any CWS wholesaler serving fewer than 25
Proposed PFAS Rule Economic Analysis
4-35
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-34: Limitations and Uncertainties That Apply to the Baseline Analysis for the
Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Quantitative
Analysis
Notes
people from the analysis and assumed any remaining
CWSs had a minimum possible population of 25. EPA also
assumed any non-wholesale NTNCWSs had a minimum
possible population of 25. The maximum number of entry
points per system was limited to the maximum number
found for the equivalent system size and source water
combination in the UCMR 3 data.
Flow relationships for
CWS
Uncertain impact on flow
inputs to cost analysis
The equations used to estimate design and average daily
flow based on service population may over- or
underestimate actual system flows. In general, average per
capita household water consumption has declined since the
source data were collected because of increased water
efficiency.3 The change in nonresidential consumption is
unknown.
CWS flow curves applied
to NTNCWS
Uncertain impact on flow
inputs to cost analysis
EPA applied the CWS population-flow equations to
NTNCWSs. This approach may result in an over- or
underestimate of flow, and therefore cost for NTNCWSs.
Uniform entry point
population distribution
Uncertain impact on flow
inputs to cost analysis and
population inputs to benefits
analysis
EPA assumed a uniform distribution of system population
across system entry points. Actual entry point population
may be greater or lower than the modeled estimates.
System wage rates are
based on old survey data
Uncertain impact on cost
analysis
National average wage rates are based on CWSS data
finalized in 2006. EPA escalated the values to $2021 to
reflect current national industry averages, but actual wage
rates at affected systems may be greater or less than
national averages.
Baseline occurrence based
on MCMC occurrence
model outputs
Uncertain effect on
occurrence and exposure
The modeled occurrence values may over- or under-
estimate actual occurrence at individual entry points. The
4,000 iterations attempt to bound the range of uncertainty.
Baseline occurrence limited
to four PFAS
Underestimate occurrence
and exposure
Excluding occurrence estimates for PFNA, HFPO-DA, and
PFBS (three of the four HI contaminants) underestimates
the number of systems that would exceed the HI and
exposed population for the quantified SafeWater model
runs. In Appendix N, EPA evaluates the potential increase
in system level treatment costs for systems that exceed the
HI in addition to the PFOA and PFOS MCLs, and for
systems that do not exceed the PFOA and PFOS MCLs but
do exceed the HI.
Abbreviations: CWS - community water systems; CWSS- community water system survey; HI- hazard index; MCMC -
Markov chain Monte Carlo; NTNCWS - non-transient, non-community water systems; PFAS - per- and polyfluoroalkyl
substances; PFOA- perfluorooctanoic acid; PFOS- perfluorooctane sulfonate; SDWIS- safe drinking water information system.
Note:
aThere is uncertainty in using the equations from EPA's Geometries and Characteristics of Public Water Systems report (U.S.
EPA, 2000) to predict future average daily and design flow based on a system's retail population. Water use efficiency has
increased substantially since the 1980s, with a major improvement between 2005 and 2010 (Rockaway et al., 2011). A 2016
Water Research Foundation study reported a 22 percent decline in indoor water use (Water Research Foundation, 2016). Several
factors have contributed to increases in water efficiency. Technological changes, supported by policy, increased the efficiency of
water use. For example, the Energy Policy Act of 1992 required water efficiency standards for fixtures, including shower heads,
toilets, and washing machines. Water recycling and increased efficiency of power generation also reduces freshwater use. The
economic downturn of 2008 contributed to the drop in water use and the increase in use of water-efficient fixtures and
xeriscaping. Other demand-side management measures contributed to reduction in per capita use as well. The trend of lower
Proposed PFAS Rule Economic Analysis
4-36
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 4-34: Limitations and Uncertainties That Apply to the Baseline Analysis for the
Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Quantitative
Analysis
Notes
residential water use could result in lower flow per population and lower treatment costs as compared to predicted values in this
EA.
Proposed PFAS Rule Economic Analysis
4-37
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5 Cost Analysis
5.1 Introduction
In this chapter, EPA presents its cost analysis for the proposed PFAS National Primary Drinking
Water Regulation (the proposed rule) and other alternative rule options considered by the
Agency as part of the rulemaking process (Options la through lc). The contents include the
national cost estimates for the proposed rule as well as options and the approach EPA used to
derive those estimates. The estimates include the cost that PWSs, households, and primacy
agencies may incur in response to the proposed rule requirements.
5.1.1 Chapter Overview
This chapter has seven main sections including this introductory section. Section 5.2 provides an
overview of EPA's approach to estimate the cost of the proposed rule and options. In Section
5.3, EPA provides the data and algorithms used to calculate the cost of activities PWSs will
undertake to comply with the proposed rule. Section 5.4 provides the data and assumptions used
to calculate the cost activities primacy agencies will undertake to implement and administer the
proposed rule. Sections 5.1.3, 5.5, and 5.6 provide the cost estimates at the national, PWS, and
household level, respectively. As indicated below, some additional details on the approach and
data used to calculate the costs of the proposed rule are in Appendix C.
5.1.2 Uncertainty Characterization
Many of the input values used to calculate the costs of drinking water regulations are not known
with certainty. For example, estimated technology unit costs and contaminant occurrence values
are uncertain to some degree given imperfect information. EPA determined it does have enough
information about the level or distribution of uncertainty to conduct a Monte-Carlo based
uncertainty analysis as part of the SafeWater Multi-Contaminant Benefit-Cost Model (MCBC).
With respect to the cost analysis, EPA modeled the sources of uncertainty summarized in Table
5-1.
Table 5-1: Quantified Sources of Uncertainty in Cost Estimates
Source Description of Uncertainty
Total organic carbon concentration The TOC value assigned to each system is from a distribution derived
from the fourth Six-Year Review Information Collection Request database
(see Section 5.3.1.1)
Compliance technology unit cost Cost curve selection varies with baseline PFAS concentrations and also
curve selection includes a random selection from a distribution across feasible
technologies (see Section 5.3.1.1), and random selection from a triangular
distribution of low-, mid-, and high-cost equipment (25%, 50%, and 25%,
respectively).
Abbreviations: MCBC - Multi-Contaminant Benefit-Cost Model; PFAS - per- and polytluoroalkyl substances; TOC - total
organic carbon.
For each iteration, SafeWater MCBC assigned new values to the two sources of modeled
uncertainty as described in Table 5-1, and then calculated costs for each of the model PWSs.
This was repeated 4,000 times to reach an effective sample size for each parameter. At the end of
the 4,000 iterations, SafeWater MCBC outputs the expected value as well as the 90 percent
Proposed PFAS Rule Economic Analysis
5-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
confidence interval for each cost metric (i.e., bounded by the 5th and 95th percentile estimates for
each cost component). Detailed information on the data used to model uncertainty is provided in
Appendix L.
5.1.3 Summary of Quantified National Cost Estimates of the
Proposed Rule
In Table 5-2, EPA summarizes the total annualized cost of the proposed option at both a 3
percent and 7 percent discount rate. The first three rows show the annualized PWS sampling
costs, the annualized PWS implementation and administrative costs, and the annualized PWS
treatment costs. The fourth row shows the sum of the annualized PWS costs. At a 3 percent
discount rate, the expected annualized PWS costs are $764 million. The uncertainty range for
annualized PWS costs is $698 million to $842 million. Finally, annualized primacy agency
implementation and administrative costs are added to the annualized PWS costs to calculate the
total annualized cost of the proposed option. At a 3 percent discount rate, the expected total
annualized cost of the proposed option is $772 million with an uncertainty range of $705 million
to $850 million. At a 7 percent discount rate, the expected total annualized cost of the proposed
option is $1,205 billion, while the uncertainty range is $1,106 billion to $1,321 billion. As
discussed in Section 2.1, for purposes of this analysis, EPA is considering the cost analysis for
the proposed option to be representative of the alternate regulatory approach where PFHxS,
PFNA, PFBS, and HFPO-DA would be regulated by individual MCLs in addition to or instead
of using the HI approach.
Proposed PFAS Rule Economic Analysis
5-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-2: National Annualized Costs, Proposed Option (PFOA and PFOS MCLs of 4.0
ppt and HI of 1.0; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Annualized PWS Sampling
$76.33
$88.64
$102.15
$78.71
$91.27
$105.00
Costs
Annualized PWS
$1.71
$1.71
$1.71
$3.52
$3.52
$3.52
Implementation and
Administration Costs
Annualized PWS Treatment
$619.29
$673.59
$741.17
$1,012.54
$1,101.26
$1,206.49
Costs
Total Annualized PWS
$697.5-1
$763.93
$841.97
$1,098.59
$1,195.99
$1,311.59
Costs
Primacy Agency Rule
$6.91
$7.83
$8.86
$7.68
$8.64
$9.69
Implementation and
Administration Cost
Total Annualized Rule
$704.53
$771.77
$850.40
$1,106.01
$1,204.61
$1,321.01
Costsb'c'd
Abbreviations: PWS - public water system.
Notes: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because cost
components are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1.
This range does not include the uncertainty described in Table 5-22.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
Total quantified national cost values do not include the incremental treatment costs associated with the cooccurrence of
HFPO-DA, PFBS, and PFNA at systems required to treat for PFOA, PFOS, and PFHxS. The total quantified national cost
values do not include treatment costs for systems that would be required to treat based on HI exceedances apart from systems
required to treat because of PFHxS occurrence alone. See Appendix N, Section N.3 for additional detail on cooccurrence
incremental treatment costs and additional treatment costs at systems with HI exceedances.
dPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table do
not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns about
potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future, EPA
conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See Appendix
N, Section N.2 for additional detail.
In Table 5-3, Table 5-4, and Table 5-5 EPA summarizes the total annualized cost of Options la,
lb, and lc, respectively.
Proposed PFAS Rule Economic Analysis
5-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-3: National Annualized Costs, Option la (PFOA and PFOS MCLs of 4.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Annualized PWS Sampling
$75.70
$87.84
$101.27
$78.14
$90.45
$104.11
Costs
Annualized PWS
$1.71
$1.71
$1.71
$3.52
$3.52
$3.52
Implementation and
Administration Costs
Annualized PWS Treatment
$604.25
$658.51
$726.21
$985.22
$1,074.85
$1,176.48
Costs
Total Annualized PWS
$681.28
$748.05
$82-1.-1-1
$1,068.69
$1,168.79
$1,282.69
Costs
Primacy Agency Rule
$6.81
$7.77
$8.79
$7.59
$8.56
$9.61
Implementation and
Administration Cost
Total Annualized Rule
$688.09
$755.82
$833.48
$1,078.51
$1,177.31
$1,292.01
Costsb'c
Abbreviations: PWS - public water system.
Notes: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because cost
components are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1.
This range does not include the uncertainty described in Table 5-22.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table do
not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns about
potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future, EPA
conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See Appendix
N, Section N.2 for additional detail.
Proposed PFAS Rule Economic Analysis
5-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-4: National Annualized Costs, Option lb (PFOA and PFOS MCLs of 5.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Annualized PWS Sampling
$66.38
$77.03
$89.08
$68.71
$79.54
$91.74
Costs
Annualized PWS
$1.71
$1.71
$1.71
$3.52
$3.52
$3.52
Implementation and
Administration Costs
Annualized PWS Treatment
$481.16
$525.41
$577.23
$781.55
$851.63
$935.08
Costs
Total Annualized PWS
$550.-11
$604.16
$666.81
$857.-17
$934.69
$1,025.67
Costs
Primacy Agency Rule
$6.04
$6.84
$7.75
$6.76
$7.59
$8.47
Implementation and
Administration Cost
Total Annualized Rule
$558.71
$611.01
$674.32
$864.74
$942.28
$1,035.56
Costsb'c
Abbreviations: PWS - public water system.
Notes: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because cost
components are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1.
This range does not include the uncertainty described in Table 5-22.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table do
not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns about
potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future, EPA
conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See Appendix
N, Section N.2 for additional detail.
Proposed PFAS Rule Economic Analysis
5-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-5: National Annualized Costs, Option lc (PFOA and PFOS MCLs of 10.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Annualized PWS Sampling
$46.27
$52.21
$59.29
$48.37
$54.49
$61.57
Costs
Annualized PWS
$1.71
$1.71
$1.71
$3.52
$3.52
$3.52
Implementation and
Administration Costs
Annualized PWS Treatment
$215.41
$233.93
$256.36
$337.86
$367.50
$402.16
Costs
Total Annualized PWS
$265.05
$287.86
$315.-16
$391.00
$425.51
$-166.68
Costs
Primacy Agency Rule
$4.31
$4.72
$5.20
$4.93
$5.36
$5.85
Implementation and
Administration Cost
Total Annualized Rule
$269.36
$292.57
$320.76
$396.22
$430.87
$472.20
Costsb'c
Abbreviations: PWS - public water system.
Notes: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because cost
components are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1.
This range does not include the uncertainty described in Table 5-22.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table do
not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns about
potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future, EPA
conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See Appendix
N,, Section N.2 for additional detail.
5.2 Overview of SafeWater Multi-Contaminant Benefit Cost
Model (MCBC)
The SafeWater Cost Benefit Model (SafeWater CBX) was designed to calculate the costs and
benefits associated with setting a new or revised MCL. Since the proposed PFAS rule
simultaneously regulates multiple PFAS contaminants, EPA developed a new model version
called the SafeWater MCBC to estimate the costs and benefits associated with regulating more
than one contaminant. The following modifications were made to the SafeWater CBX model to
create the SafeWater MCBC model:
1. Instead of tracking a single contaminant's level and comparing to the proposed MCL to
determine if the PWS must take compliance actions, SafeWater MCBC tracks each
PWS's level of multiple PFAS contaminants and compares against proposed MCLs for
each contaminant (or group of contaminants).
Proposed PFAS Rule Economic Analysis
5-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
2. The structure of the occurrence data input to the model was updated to not only handle
multiple contaminants, but to incorporate all information from the PFAS occurrence
model on the predicted co-occurrence of contaminants.
3. The model structure allows for assignment of one or more compliance technologies that
achieve all regulatory requirements and estimates costs and benefits associated with
multiple PFAS contaminant reductions and calculates before and after treatment
concentrations of each contaminant for use in estimation of benefits.
5.2.1 Modeling PWS Variability in SafeWater MCBC
The costs incurred by a PWS depend on water system characteristics. The data describing some
of these characteristics for PWSs are in SDWIS/Fed. The SDWIS/Fed data provide information
on the PWS characteristics that typically define PWS categories, or strata, for which EPA
develops costs in rulemakings:
• System type (CWS, NTNCWS)
• Number of people served by the PWS
• PWS's primary raw water source (ground water or surface water)
• PWS's ownership type (public or private).
• PWS state.
Because EPA does not have complete PWS-specific data across the 49,193 CWSs and 17,337
NTNCWS in SDWIS/Fed for many of the baseline and compliance characteristics necessary to
estimate costs and benefits, such as design and average daily flow rates, water quality
characteristics, treatment in-place, and labor rates, EPA adopted a "model PWS" approach.
SafeWater MCBC creates model PWSs by combining the PWS-specific data available in
SDWIS/Fed with data on baseline and compliance characteristics available at the PWS category
level. In some cases, the categorical data are simple point estimates. In this case, every model
PWS in a category is assigned the same value. In other cases, where more robust data
representing system variability are available, the category-level data include a distribution of
potential values. In the case of distributional information, SafeWater MCBC assigns each model
PWS a value sampled from the distribution. These distributions are assumed to be independent.
Table 5-6 provides a list of all the PWS characteristics that impact model PWS compliance costs.
These data include inventory data specific to each system and categorical data for which
randomly assigned values are based on distributions that vary by category (e.g., ground water
and surface water TOC distributions or compliance forecast distributions that vary by system
size category).
Proposed PFAS Rule Economic Analysis
5-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-6: Model PWS Variability Characteristics and Data Sources
PWS Characteristic
Data Type and Description
System Type
Known SDWIS/Fed Inventory
Primary Source Water
Known: SDWIS/Fed Inventory
Ownership
Known: SDWIS/Fed Inventory
Population Served
Known: SDWIS/Fed Inventory
Number of Entry Points
Known: UCMR 3, SDWIS/Fed Inventory, and modeled from SDWIS/Fed
Inventory distribution (see Section 4.3.3.1)
PFAS Contaminant Concentration
Sampled from EPA Occurrence Model (see Section 4.3.3.2)
at each Entry Point
Influent TOC Level
Assigned from distribution derived from fourth Six-Year Review
Information Collection Request database (see Section 5.3.1.1)
Compliance Technology Forecast
Assigned from distribution derived from full-scale compliance actions
at each Entry Point
analyzed by EPA (see Section 5.3.1.1)
Abbreviations: EPA - U.S. Environmental Protection Agency; PFAS - per-and polyfluoroalkyl substances; SDWIS/Fed -
Safe Drinking Water Information System/Federal version; TOC - Total Organic Carbon; UCMR 4 - Fourth Unregulated
Contaminant Monitoring Rule.
As illustrated in Figure 5-1, once all the model PWSs are created and assigned baseline and
compliance characteristics, SafeWater MCBC estimates the quantified costs and benefits of
compliance for each model PWS under the proposed rule. Because of this model PWS approach,
SafeWater MCBC does not output any results at the PWS-level. Instead, the outputs are cost and
benefit estimates for 36 PWS categories, or strata. Each PWS category is defined by the system
type (CWS and NTNCWS), primary water source (ground or surface), and size category (there
are nine). Note EPA does not report state specific strata although state location is utilized in the
SafeWater MCBC model (e.g., current state level regulatory limits on PFAS in drinking water).
For each PWS category, the model then calculates summary statistics that describe the costs and
quantified benefits associated with the proposed rule compliance. These summary statistics
include total quantified costs of the proposed regulatory requirement, total quantified benefits of
the proposed regulatory requirement, the variability in PWS-level costs (i.e., 5th and 95th
percentile system costs), and the variability in household-level costs.
Proposed PFAS Rule Economic Analysis
5-8
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
r
k
r
r
A
I
Figure 5-1: Approach Used by SafeWater MCBC to Model PWS Variability
5.3 Estimating Public Water System Costs
EPA estimated PWS compliance activities that result in treatment costs, and administrative and
monitoring costs associated with the proposed rule. Each major regulatory component consists of
required activities, which EPA details here. EPA presents the costs associated with treatment
addition and non-treatment actions that could be taken in lieu of treatment in Section 5.3.1. EPA
presents the costs associated with the administrative and monitoring requirements associated
with the proposed rule in Section 5.3.2.
This section describes how EPA estimated costs associated with:
• Engineering, installing, operating, and maintaining PFAS removal treatment
technologies, including treatment media replacement and spent media destruction or
disposal
• Non-treatment actions that some PWSs might take in lieu of treatment, such as
constructing new wells in an uncontaminated aquifer or interconnecting with and
purchasing water from a neighboring PWS.
EPA used SafeWater MCBC to apply costs for one of these treatment technologies or non-
treatment alternatives at each entry point in a PWS estimated to be out of compliance with the
regulatory option under consideration. First, for each affected entry point, SafeWater MCBC
selected from among the compliance alternatives using the decision tree procedure described in
Proposed PFAS Rule Economic Analysis 5-9 March 2023
5.3.1 PWS Treatment Costs
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Section 5.3.1.1. Next, SafeWater MCBC estimated the cost of the chosen compliance alternative
using outputs from EPA's WBS cost estimating models. Specifically, SafeWater MCBC used
cost equations generated from the following models:
• The GAC WBS model
• The PFAS-selective IX WBS model
• The centralized reverse osmosis/nanofiltration (RO/NF) WBS model14
• The non-treatment WBS model.
The national cost analysis reflects the assumption that PFAS-contaminated wastes are not
considered hazardous wastes. As a general matter, EPA notes that such wastes are not currently
regulated under federal law as a hazardous waste. However, EPA anticipates proposing certain
PFAS be designated as Comprehensive Environmental Response, Compensation, and Liability
Act (CERCLA) hazardous substances to require reporting of PFOA and PFOS releases, enhance
the availability of data, and ensure agencies can recover cleanup costs.15 Stakeholders have
expressed concern to EPA that a hazardous substance designation for certain PFAS may limit
their disposal options for drinking water treatment residuals (e.g., spent media, concentrated
waste streams) and/or potentially increase costs. Although designating chemicals as hazardous
substances under CERCLA would not result in new requirements for disposal of PFAS drinking
water treatment residuals, to address stakeholder concerns, including those raised during the
SBREFA process, EPA conducted a sensitivity analysis with an assumption of hazardous waste
disposal for illustrative purposes only. EPA has estimated national costs, both assuming non-
hazardous disposal options and assuming hazardous waste disposal at 100 percent of systems
treating for PFAS to assess the effects of potential increased disposal costs. EPA acknowledges
that if federal authorities later determine that PFAS-contaminated wastes require handling as
hazardous wastes, the residuals management costs are expected to be higher. For a discussion of
the findings from this sensitivity analysis, see Appendix N.
Section 5.3.1.2 describes the WBS models. Section 5.3.1.2.2 describes the form of the resulting
cost equations and their application in SafeWater MCBC. The Technologies and Costs (T&C)
document (U.S. EPA, 2023h) provides a comprehensive discussion of each of the treatment
technologies, their effectiveness, and the WBS cost models. It also presents the cost equations
themselves in tabular form.
14 At this time, EPA is not including point-of-use (POU) RO in the national cost estimates because the regulatory options under
consideration require treatment to concentrations below 70 ng/L total of PFOA and PFOS, the current NSF/ANSI certification
standard for POU devices. However, POU treatment is reasonably anticipated to become a compliance option for small systems
in the future if NSF/ANSI or other independent third-party certification organizations develop a new certification standard that
mirrors EPA's proposed regulatory standard. Costs presented here reflect the costs of devices certified under the current testing
standard, not a future standard, which may change dependent on future device design. In the event POU treatment becomes a
valid compliance option, national costs could be lower than estimated in this application of the SafeWater MCBC.
15 The pre-publication in the Federal Register Notice version of the proposed rule entitled "Designation of Perfluorooctanoic
Acid (PFOA) and Perfluorooctanesulfonic Acid (PFOS) as CERCLA Hazardous Substances" is available at
https://www.epa.gOv/system/files/documents/2022-08/FRL%207204-02-
OLEM%20_%20Designating%20PFOA%20and%20PFOS%20as%20HSs%20_NPRM_20220823.pdf.
Proposed PFAS Rule Economic Analysis
5-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5.3.1.1 Decision Tree for Technology Selection
For entry points at which baseline PFAS concentrations exceed regulatory thresholds, the
decision tree selects a treatment technology or non-treatment alternative using a two-step process
that:
1. Determines whether to include or exclude each alternative from consideration given the
entry point's characteristics and the regulatory option selected
2. Selects from among the alternatives that remain viable based on percentage distributions
derived, in part, from data on recent PWS actions in response to PFAS contamination.
Inputs to the decision tree include the following:
• Influent concentrations of individual PFAS contaminants in ppt (ng/L);
• Entry point design flow in MGD; and
• TOC influent to the new treatment process in mg/L.
Section 4.4 describes EPA's method for estimating PFAS influent concentrations and Section
4.3.3.3 describes how EPA derived entry point flow estimates. SafeWater MCBC selects influent
TOC using the distribution shown in Table 5-7.
Table 5-7: Frequency Distribution to Estimate Influent TOC in mg/L
Percentile Surface Water Ground Water
0.05
0.65
0.35
0.15
1.1
0.48
0.25
1.38
0.5
0.35
1.6
0.5
0.45
1.85
0.58
0.5
1.97
0.69
0.55
2.14
0.75
0.65
2.54
1
0.75
3.04
1.39
0.85
3.63
2.01
0.95
4.81
3.8
Abbreviations: TOC - total organic carbon.
Source: EPA's analysis of total organic carbon concentrations in the fourth Six-Year Review Information Collection Request
database.
Step 1 of the decision tree uses these inputs to determine whether to include or exclude each
treatment alternative from consideration in the compliance forecast. For the treatment
technologies (GAC, IX, and RO/NF), this determination is based on estimates of each
technology's performance given available data about influent water quality and the regulatory
option under consideration. Section 5.3.1.1.1 describes this process for GAC and IX. Section
5.3.1.1.2 describes this process for RO.
EPA assumes a small number of PWSs may be able to take non-treatment actions in lieu of
treatment. The viability of non-treatment actions (interconnection with neighboring system or
new wells) is likely to depend on the quantity of water being replaced. Therefore, the decision
Proposed PFAS Rule Economic Analysis
5-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
tree considers non-treatment only for entry points with design flows less than or equal to 3.536
MGD. EPA's WBS model for non-treatment does not generate costs for flows greater than this
value, so the decision tree excludes non-treatment actions from consideration above this flow.
Step 2 of the decision tree selects a compliance alternative for each entry point from among the
alternatives that remain in consideration after Step 1. Table 5-8 shows the initial compliance
forecast that is the starting point for this step. The percentages in Table 5-8 consider data
presented in the T&C document (U.S. EPA, 2023h) on actions PWSs have taken in response to
PFAS contamination.
Table 5-8: Initial Compliance Forecast Including POU RO
Design flow less than 1 Design flow 1 to less than Design flow greater than
MGD 10 MGD or equal to 10 MGD
Compliance
Alternative
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
GAC
65%
50%
77%
50%
85%
50%
PFAS-selective IX
10%
25%
10%
37%
10%
45%
Central RO/NF
4%
4%
5%
5%
5%
5%
POURO
13%
13%
0%
0%
0%
0%
Interconnection
6%
6%
6%
6%
0%
0%
New Wells
2%
2%
2%
2%
0%
0%
Abbreviations: GAC - granular activated carbon; PFAS - per-and polyfluoroalkyl substances; MGD - million gallons per
day; IX - ion exchange; RO/NF - reverse osmosis/nanofiltration; POU - point of use; RO - reverse osmosis; TOC - total
organic carbon.
Source: EPA's analysis of total organic carbon concentrations in the fourth Six-Year Review Information Collection Request
database.
To date, the majority of PWSs for which data are available have installed GAC (U.S. EPA,
2023h). The first full-scale system treating drinking water using PFAS-selective IX began
operation in 2017 (WWSD, 2018). The data in the T&C document (U.S. EPA, 2023h) suggest
that an increasing share of PWSs have selected IX in response to PFAS since that first
installation. EPA expects this trend to continue, so the initial percentages include adjustments to
account for this expectation. In addition, as discussed in Section 5.3.1.1.1, the performance of
GAC is affected by the presence of TOC. Accordingly, the table includes adjusted distributions
for systems with higher influent TOC.
The initial percentages in Table 5-8 estimate that some small systems will choose POU RO as a
compliance alternative. At this time, EPA is not including POU RO in the national cost estimates
because the regulatory options under consideration require treatment to concentrations below 70
ppt PFOA and PFOS summed, the current certification standard for POU devices.16 Therefore,
the decision tree excludes POU RO from consideration and proportionally redistributes the
16POU treatment might become a compliance option for small systems in the future if NSF/ANSI develop a new certification
standard that mirrors EPA's proposed regulatory standard. In the event POU treatment becomes a valid compliance option,
national costs could be lower than estimated here.
Proposed PFAS Rule Economic Analysis
5-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
percentages among the other alternatives. Table 5-9 shows the final compliance forecast after
this redistribution.
Table 5-9: Initial Compliance Forecast Excluding POU RO
Design flow less than 1 Design flow 1 to less than Design flow greater than
MGD 10 MGD or equal to 10 MGD
Compliance
Alternative
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
TOC less
than or
equal to 1.5
mg/L
TOC
greater
than 1.5
mg/L
GAC
75%
57%
77%
50%
85%
50%
PFAS-selective IX
11%
29%
10%
37%
10%
45%
Central RO/NF
5%
5%
5%
5%
5%
5%
Interconnection
7%
7%
6%
6%
0%
0%
New Wells
2%
2%
2%
2%
0%
0%
Abbreviations: GAC - granular activated carbon; PFAS - per-and polyfluoroalkyl substances; MGD - million gallons per
day; IX - ion exchange; RO/NF - reverse osmosis/nanofiltration; POU - point of use; RO - reverse osmosis; TOC - total
organic carbon.
If all the compliance alternatives (other than POU RO) remain in consideration after Step 1, the
decision tree uses the forecast shown in Table 5-9. If Step 1 eliminated on one or more of the
alternatives, the decision tree proportionally redistributes the percentages among the remaining
alternatives and uses the redistributed percentages.
5.3.1.1.1 Estimating GAC and IX Performance
The viability of GAC and IX depends on bed life, which is the length of time the technology can
maintain a target removal percentage (e.g., 80 percent, 95 percent). Bed life can vary depending
on factors including type of media used (GAC or IX), specific PFAS contaminants targeted,
influent water quality, and removal performance required to meet regulatory option thresholds.
Bed life determines media replacement frequency and, therefore, affects both the practicality and
operation and maintenance (O&M) cost of these technologies. This analysis estimates bed life in
bed volumes (BV), which is a measure of throughput: the volume of water treated during the bed
life divided by the volume of the media bed.
The bed life estimates use linear equations derived as described in the T&C document (U.S.
EPA, 2023h). EPA estimated the equations based on pooled data from several studies of GAC as
well as IX performance and reflect central tendency results under varying water quality
conditions. As such, EPA believes they represent the best approach currently available for use in
a national cost estimation. However, they should not be used in lieu of site-specific engineering
analyses or pilot studies to guide the design or operation of specific treatment systems.
The bed life equations are technology-specific and shown below:
Proposed PFAS Rule Economic Analysis
5-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Equation 2:
BVcontam,GAC ~ ^TOC ^ TOC + Ar ^ °^°^contam Bcontain,GAC
BVcontam,IX ~ ^PFAS ^ P^^^total ^R,IX ^ °^°^contam Bcontam,IX
Where:
B Vcontam,tech = bed life of the given technology for a given PFAS contaminant in BV; tech =
GAC or IX
TOC = TOC influent to the new treatment process in mg/L
PFAStotai = total influent concentration of all PFAS contaminants (regulated or unregulated)
in ppt
%Rcontam = target percent removal of a given PFAS as a decimal (e.g., 0.8, 0.95)
Table 5-10 shows the estimated values of the parameter coefficients Atoc, Apfas, Adtech, and
intercepts BContam,tech
Table 5-10: Estimated Parameter Values for Technology-Specific Bed Life Equations
Parameter
GAC Model Value
IX Model Value
Atoc
-37,932
Not applicable3
Apfas
Not applicable3
-6.04
Ar
-36,309
-198,242
Bhfpo-da
113,034
Data not available
BpFHxA
113,967
212,867
Bpfbs
129,357
439,515
BpFHpA
129,357
319,511
BpFHxS
129,357
439,515
BpFOA
139,862
390,787
Bpfos
143,731
439,515
Note:
aTotal PFAS is not a significant parameter in GAC performance; TOC is not a significant parameter in IX performance.
Source: Technologies and Costs for Removing Per- and Polyfluoroalkyl Substances from Drinking Water (U.S. EPA, 2023h)
The bed life equations are only applicable over a specific range of water quality conditions (TOC
up to 3.2 mg/L for GAC; total PFAS up to 7,044 ppt for IX). Data are not available to estimate
performance beyond these limits. Therefore, the decision tree excludes GAC from consideration
if an entry point's influent TOC concentration is greater than 3.2 mg/L. It excludes IX if total
influent PFAS is greater than 7,044 ppt.
If GAC and/or IX remain in consideration, the decision tree calculates the percent removal
required for the regulatory option under consideration and uses the linear equations above to
estimate bed life. These calculations vary depending on the regulatory option. Section 5.3.1.1.1.1
Proposed PFAS Rule Economic Analysis
5-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
describes the calculations under Option 1 (individual MCLs for PFOS and PFOA). Section
5.3.1.1.1.2 describes the calculations under the proposed option (#2) (individual MCLs for PFOS
and PFOA plus group standard based on HI).
Based on data presented in the T&C document (U.S. EPA, 2023h), the decision tree assumes the
maximum PFAS removal achievable by GAC or IX is 99 percent. Therefore, if the relevant
regulatory option requires removal at an entry point greater than this maximum, the decision tree
removes GAC and IX from consideration, as described in the sections below. Additionally, the
decision tree assumes that bed lives less than 5,000 BV for GAC and less than 20,000 BV for IX
are impractical. These bed lives correspond to media replacement frequencies of two to five
months depending on the average flow of the entry point. If the relevant regulatory option results
in a final operating bed life below these limits, the decision tree removes the corresponding
technology from consideration. For entry points that ultimately select GAC or IX, the final
operating bed life is also an input to the cost estimates (see Section 5.3.1.3) and the calculation of
post-treatment PFAS concentrations used to estimate reduction in health risks).17
5.3.1.1.1.1 Bed Life Under Option 1
Under Option 1, PWSs must meet individual MCLs for PFOS and PFOA. For these options, the
decision tree calculates the percent removal required to meet each individual MCL:
Equation 3:
n/ n _ Co,contain ~ MCLCOntam ^ SF
'°^contam ~
,contam
Where:
%Rcontam = target percent removal of a given PFAS as a decimal (e.g., 0.8, 0.95)
Co,contain — influent concentration of the given PFAS in ppt
MCLcontam = MCL for the given PFAS in ppt
SF = 0.8, a safety factor that assumes PWSs will design and operate treatment processes to
achieve 80 percent of the MCL
The decision tree performs this calculation for each contaminant that occurs at an entry point and
has an MCL in the regulatory option, even if the contaminant occurs at a concentration below the
MCL. Including contaminants that are below their respective MCLs helps to account for
chromatographic peaking18; which is a concern in GAC along with IX and is discussed in greater
17 As shown in Equation 2, bed life and percent removal are directly related. SafeWater uses the same equation to back-calculate
final percent removal for each PFAS compound from final operating bed life. It then uses the filial removal efficiency to calculate
post-treatment concentrations.
18 Chromatographic peaking is a phenomenon in which less strongly sorbed contaminants are detached from sorbents by more
strongly bound sorbents and the less tightly bound sorbent re-enters drinking water. Direct competition with stronger sorbing
constituents can lead to effluent PFAS concentrations temporarily exceeding influent concentrations. Some PFAS species sorb
more strongly than other PFAS species which can cause more weakly sorbed species to re-enter drinking water.
Proposed PFAS Rule Economic Analysis
5-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
detail in the T&C document (U.S. EPA, 2023h). The calculations here are designed to account
for and avoid it.
If the percent removal required for any contaminant (%RContam) is greater than 0.99 (99 percent),
the decision tree removes GAC and IX from consideration. If the technologies remain in
consideration, the decision tree estimates the bed life for each contaminant using the linear
equations presented in Section 5.3.1.1.1. The final operating bed life is the minimum of the
individual contaminant-specific bed life estimates. If this final operating bed life is less than
5,000 BV for GAC or less than 20,000 BV for IX, the decision tree removes the corresponding
technology from consideration.
5.3.1.1.1.2 Bed Life Under the Proposed Option
Under the proposed rule, PWSs must meet a group standard based on HI, plus individual MCLs
for PFOS and PFOA. Due to limitations in occurrence data, the national cost estimates account
for only one of the contaminants included in the HI: PFHxS. Therefore, for this option, the
decision tree calculates the percent removal required to meet the individual health benchmark for
PFHxS:
Equation 4:
n/ n _ Q).pfhxs ~ HBPFHxS X SF
/oK PFHxS — T.
PFHxS
Where:
%Rpfhxs = target percent removal of PFHxS as a decimal (e.g., 0.8, 0.95)
Co,pfhxs = influent concentration of PFHxS in ppt
HBpfhxs = heath benchmark for PFHxS in ppt
SF = 0.8, a safety factor that assumes PWSs will design and operate treatment processes to
achieve 80 percent of the health benchmark
The decision tree performs this calculation even when PFHxS occurs at a concentration below its
health benchmark. Including contaminants that are below their respective MCLs prevents the
subsequent bed life calculations from selecting a bed life that results in a preferred PFAS
displacing a less preferred PFAS from the treatment media to the extent that the less preferred
PFAS periodically exceeds its MCL. This phenomenon is sometimes a concern in GAC as well
as IX design and operation and is discussed in greater detail in the T&C document (U.S. EPA,
2023h). The calculations here are designed to account for and avoid it.
If the percent removal required to meet the health benchmark for PFHxS is greater than 0.99 (99
percent), the decision tree removes GAC and IX from consideration. If the technologies remain
in consideration, the decision tree estimates the bed life for PFHxS using the linear equations
presented in Section 5.3.1.1.1. It also calculates the bed lives necessary to meet the individual
MCLs for PFOS and PFOA, as described in Section 5.3.1.1.1.1. The final operating bed life is
the minimum of all the bed life estimates resulting from the calculations for all three
contaminants (PFOS, PFOA, and PFHxS). If this final operating bed life is less than 5,000 BV
for GAC or less than 20,000 BV for IX, the decision tree removes the corresponding technology
from consideration.
Proposed PFAS Rule Economic Analysis
5-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5.3.1.1.2 Estimating the Performance of RO/NF
Designed and operated correctly, central RO/NF provides steady-state PFAS removal. The
technology's effectiveness does not vary substantially among PFAS compounds of similar
molecular size. There is no concept like bed life to consider for an RO membrane design. The
calculation of the required removal from RO/NF (%Rfmai.Ro) varies depending on the regulatory
option, as described in Sections 5.3.1.1.2.1 through 5.3.1.1.1.2 below. For entry points that
ultimately select RO, the required removal is also an input to the cost estimates (see Section
5.3.1.3) and the calculation of post-treatment PFAS concentrations.19
5.3.1.1.2.1 Required Removal Under Option 1
Under Option 1, PWSs must meet individual MCLs for PFOS and PFOA. For these options, the
decision tree calculates the percent removal required to meet each individual MCL20:
Equation 5:
n/ n _ Co,contain ~ MCLCOntam ^ SF
'°^contam ~
,contam
Where:
%Rcontam = target percent removal of a given PFAS as a decimal (e.g., 0.8, 0.95)
Co,contain = influent concentration of the given PFAS in ppt
MCLcontam = MCL for the given PFAS in ppt
SF = 0.8, a safety factor that assumes PWSs will design and operate treatment processes to
achieve 80 percent of the MCL
The final removal required from RO/NF (%Rfmai.Ro) is the maximum percent removal required
for any contaminant (%RContam) that exceeds its MCL.
5.3.1.1.2.2 Required Removal Under the Proposed Option
Under the proposed rule, PWSs must meet a group standard based on HI, plus individual MCLs
for PFOS and PFOA. The national SafeWater modelled cost estimates account for only one of
the contaminants included in the HI: PFHxS. Therefore, for this option, the decision tree
calculates the percent removal required to meet the individual health benchmark for PFHxS:
Equation 6:
n/ n _ Q).pfhxs ~ HBPFHxS X SF
/oK PFHxS — T.
PFHxS
Where:
%Rpfhxs = target percent removal of PFHxS as a decimal (e.g., 0.8, 0.95)
Co,pfhxs = influent concentration of PFHxS in ppt
HBpfhxs = heath benchmark for PFHxS in ppt
19 SafeWater uses Equations 5 and 6 to back-calculate final percent removal for each PFAS compound given the maximum
percent removal across the affected PFAS. It then uses the final removal efficiency to calculate post-treatment concentrations.
20 Equations 5 and 6 in this section are the same as Equations 3 and 4, respectively.
Proposed PFAS Rule Economic Analysis
5-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
SF = 0.8, a safety factor that assumes PWSs will design and operate treatment processes to
achieve 80 percent of the health benchmark
The decision tree also calculates the percent removal required to meet the individual MCLs for
PFOS and PFOA (%Rpfos and %Rpfoa), as described in Section 5.3.1.1.2.1. The final removal
required from RO/NF (%Rfmai.Ro) is the maximum of %Rpfhxs, %Rpfos, and %Rpfoa.
5.3.1.2 WBS Models
The WBS models are spreadsheet-based engineering models for individual treatment
technologies, linked to a central database of component unit costs. EPA developed the WBS
model approach as part of an effort to address recommendations made by the Technology Design
Panel (TDP), which convened in 1997 to review the Agency's methods for estimating drinking
water compliance costs (U.S. EPA, 1997). The TDP consisted of nationally recognized drinking
water experts from EPA, water treatment consulting companies, public as well as private water
utilities along with suppliers, equipment vendors, and Federal along with State regulators in
addition to cost estimating professionals.
In general, the WBS approach involves breaking a process down into discrete components for
the purpose of estimating unit costs. The WBS models represent improvements over past cost
estimating methods by increasing comprehensiveness, flexibility, and transparency. By adopting
a WBS-based approach to identify the components that should be included in a cost analysis, the
models produce a more comprehensive assessment of the capital and operating requirements for
a treatment system.
Section 5.3.1.2.1 is a brief overview of the common elements of all the WBS models. Section
5.3.1.2.2 provides information on the anticipated accuracy of the models. Sections 5.3.1.2.3
through 5.3.1.2.6 identify technology-specific cost elements included in each model and discuss
key inputs. The documentation for the individual WBS models (U.S. EPA, 2023i; U.S. EPA,
2023k; U.S. EPA, 20231; U.S. EPA, 2023j), provides more complete details on the structure,
content, and use of each model.
5.3.1.2.1 Common Model Components and Inputs
Each WBS model contains the work breakdown for a particular treatment process and
preprogrammed engineering criteria and equations that estimate equipment requirements for
user-specified design requirements (e.g., system size and influent water quality). Each model
also provides unit and total cost information by component (e.g., individual items of capital
equipment) and totals the individual component costs to obtain a direct capital cost. Additionally,
the models estimate add-on costs (e.g., permits and land acquisition), indirect capital costs, and
annual O&M costs, thereby producing a complete compliance cost estimate.
Primary inputs common to all the WBS models include design flow and average daily flow in
MGD. Each WBS model has default designs (input sets) that correspond to specified categories
of flow, but the models can generate designs for many other combinations of flows. To estimate
costs for PFAS compliance, EPA fit cost curves to the WBS estimates across a range of flow
rates, as described in Section 5.3.1.3.
Another input common to all the WBS models is "component level" or "cost level." This input
drives the selection of materials for items of equipment that can be constructed of different
Proposed PFAS Rule Economic Analysis
5-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
materials. For example, a low-cost system might include fiberglass pressure vessels and PVC
piping. A high-cost system might include stainless steel pressure vessels and stainless-steel
piping. The component level input also drives other model assumptions that can affect the total
cost of the system, such as building quality and heating and cooling. The component level input
has three possible values: low cost, mid cost, and high cost. To estimate costs for PFAS
treatment, EPA generated separate cost equations for each of the three component levels, thus
creating a range of cost estimates for use in national compliance cost estimates.
The third input common to all the WBS models is system automation, which allows the design of
treatment systems that are operated manually or with varying degrees of automation (i.e., with
control systems that reduce the need for operator intervention). The cost equations described in
Section 5.3.1.3 are for systems that are fully automated, minimizing the need for operator
intervention and reducing operator labor costs.
The WBS models generate cost estimates that include a consistent set of capital, add-on, indirect,
and O&M costs. Table 5-11 identifies these cost elements, which are common to all the WBS
models and included in the cost estimates below. Sections 5.3.1.2.3 through 5.3.1.2.6 identify the
technology-specific cost elements included in each model. The documentation for the WBS
models (U.S. EPA, 2023i; U.S. EPA, 2023k; U.S. EPA, 20231; U.S. EPA, 2023j) provide more
information on the methods and assumptions used in the WBS models to estimate the costs for
both the technology-specific and common cost elements.
Table 5-11: Cost Elements Included in All WBS Models
Cost Category Components Included
• Technology-specific equipment (e.g., vessels, basins, pumps, treatment media, piping,
valves)
• Instrumentation and system controls
• Buildings
• Residuals management equipment
Add-on Costs • Land
• Permits
• Pilot testing
• Mobilization and demobilization
• Architectural fees for treatment building
• Equipment delivery, installation, and contractor's overhead and profit
• Sitework
• Yard piping
• Geo technical
• Standby power
• Electrical infrastructure
• Process engineering
• Contingency
• Miscellaneous allowance
• Legal, fiscal, and administrative
• Sales tax
Direct Capital
Costs
Indirect Capital
Costs
Proposed PFAS Rule Economic Analysis
5-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-11: Cost Elements Included in All WBS Models
Cost Category
Components Included
• Financing during construction
• Construction management
O&M Costs:
• Operator labor for teclinology-specific tasks (e.g., managing backwash and media
Technology-
replacement)
specific
• Materials for O&M of teclinology-specific equipment
• Teclinology-specific chemical usage
• Replacement of teclinology-specific equipment that occurs on an annual basis (e.g..
treatment media)
• Energy for operation of technology-specific equipment (e.g., mixers)
O&M Costs: Labor
• Operator labor for O&M of process equipment
• Operator labor for building maintenance
• Managerial and clerical labor
O&M Costs:
• Materials for maintenance of booster or influent pumps
Materials
• Materials for building maintenance
O&M Costs:
• Energy for operation of booster or influent pumps
Energy
• Energy for lighting, ventilation, cooling, and heating
O&M Costs:
• Residuals management operator labor, materials, and energy
Residuals
• Residuals disposal and discharge costs
Abbreviations: O&M - operation & maintenance; WBS - work breakdown structure.
5.3.1.2.2 WBS Model Accuracy
Costs for a given system can vary depending on site-specific conditions (e.g., raw water quality,
climate, local labor rates, and location relative to equipment suppliers). The costs presented here
are based on national average assumptions and include a range (represented by low-, mid-, and
high-cost equations) intended to encompass the variation in costs that systems would incur to
remove PFAS. To validate the engineering design methods used by the WBS models and
increase the accuracy of the resulting cost estimates, EPA has subjected the individual models to
a process of external peer review by nationally recognized technology experts.
The GAC model underwent peer review in 2006. Two of the three reviewers expressed the
opinion that resulting cost estimates would be in the range of budget estimates (+30 to -15
percent). The other reviewer did not provide a precise estimate of the model's accuracy range but
commented that the resulting cost estimates were reasonable. EPA made substantial revisions to
the GAC model in response to the peer review.
The IX model underwent peer review in 2005, during an early stage of its development. One peer
reviewer responded that resulting cost estimates were in the range of budget estimates (+30 to -
15 percent). The other two reviewers thought the estimates were order of magnitude estimates
(+50 to -30 percent), with an emphasis on the estimates being high. The IX model has since
undergone extensive revision, both in response to the peer review and to adapt it for PFAS
treatment using selective resin.
Proposed PFAS Rule Economic Analysis
5-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
The RO/NF model underwent peer review in 2007. The majority of peer reviewers who
evaluated the model expressed the opinion that resulting cost estimates would be in the range of
budget estimates (+30 to -15 percent). The RO/NF model has since undergone substantial
revision in response to the peer review comments.
EPA received peer review comments on the non-treatment model in May 2012. The first
reviewer responded that cost estimates resulting from the non-treatment model were in the range
of budget estimates (+30 to -15 percent). The second reviewer thought the cost estimates were
order of magnitude estimates (+50 to -30 percent). The third reviewer felt the cost estimates were
definitive (+15 to -5 percent), except for land costs, which were difficult to assess due to regional
variations. EPA revised the nontreatment model in response to the peer review
recommendations.
5.3.1.2.3 GAC Model
Work Breakdown Structure-Based Cost Model for Granular Activated Carbon Drinking Water
Treatment provides a complete description of the engineering design process used by the WBS
model for GAC (U.S. EPA, 2023i). The model can generate costs for two types of design:
• Pressure designs where the GAC bed is contained in stainless steel, carbon steel, or
fiberglass pressure vessel
• Gravity designs where the GAC bed is contained in open concrete basins.
Table 5-12 shows the technology-specific capital equipment and O&M requirements included in
the GAC model. These items are in addition to the common WBS cost elements listed in Table
5-11.
Table 5-12: Technology-Specific Cost Elements Included in the GAC Model
Cost Category Major Components Included
Direct Capital
Costs
• Booster pumps for influent water
• Contactors (either pressure vessels or concrete basins) that contain the GAC
bed
• Tanks and pumps for backwashing the contactors
• GAC transfer and storage equipment
• Spent GAC reactivation facilities (if on-site reactivation is selected)
• Associated piping, valves and instrumentation
O&M Costs: Labor
• Operator labor for contactor maintenance (for gravity GAC designs)
• Operator labor for managing backwash events
• Operator labor for backwash pump maintenance (if backwash occurs weekly
or more frequently)
• Operator labor for GAC transfer and replacement
O&M Costs:
Materials
• Materials for contactor maintenance (accounts for vessel relining in pressure
designs, because GAC can be corrosive, and for concrete and underdrain
maintenance in gravity designs)
• Materials for backwash pump maintenance (if backwash occurs weekly or
more frequently)
• Replacement virgin GAC (loss replacement only if reactivation is selected)
Proposed PFAS Rule Economic Analysis
5-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-12: Technology-Specific Cost Elements Included in the GAC Model
Cost Category
Major Components Included
O&M Costs:
Energy
Operating energy for backwash pumps
O&M Costs:
Residuals
Discharge fees for spent backwash
Fees for reactivating spent GAC (if off-site reactivation is selected)
Labor, materials, energy, and natural gas for regeneration facility (if on-site
reactivation is selected)
Disposal of spent GAC (if disposal is selected)
Abbreviations: GAC - granular activated carbon; O&M - operation & maintenance; WBS - work breakdown structure.
For small systems (less than 1 MGD) using pressure designs, the GAC model assumes the use of
package treatment systems that are pre-assembled in a factory, mounted on a skid, and
transported to the site. The model estimates costs for package systems by costing all individual
equipment line items (e.g., vessels, interconnecting piping and valves, instrumentation, and
system controls) in the same manner as custom-engineered systems. This approach is based on
vendor practices of partially engineering these types of package plants for specific systems (e.g.,
selecting vessel size to meet flow and treatment criteria). The model applies a variant set of
design inputs and assumptions that are intended to simulate the use of a package plant and that
reduce the size and cost of the treatment system. U.S. EPA (2023i) provides complete details on
the variant design assumptions used for package plants.
To generate the cost equations discussed in Section 5.3.1.3, EPA used the following key inputs
in the GAC model:
• For pressure designs, two vessels in series with a minimum total empty bed contact time
(EBCT) of 20 minutes
• For gravity designs, contactors in parallel with a minimum total EBCT of 20 minutes
• Bed life varying over a range from 5,000 to 150,000 BV, estimated as discussed in
Section 5.3.1.1.1
EPA generated separate cost equations for two spent GAC management scenarios:
• Off-site reactivation under current RCRA non-hazardous waste regulations
• Off-site disposal as a hazardous waste and replacement with virgin GAC (i.e., single use
operation).
The T&C document (U.S. EPA, 2023h) provides a comprehensive discussion of these and other
key inputs and assumptions.
5.3.1.2.4 PFAS-selective IX Model
Work Breakdown Structure-Based Cost Model for Ion Exchange Treatment of Per- and
Polyflaoroalkyl Substances (PFAS) in Drinking Water provides a complete description of the
engineering design process used by the WBS model for PFAS-selective IX (U.S. EPA, 2023j).
Table 5-13 shows the technology-specific capital equipment and O&M requirements included in
the model. These items are in addition to the common WBS cost elements listed in Table 5-11.
Proposed PFAS Rule Economic Analysis
5-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-13: Technology-Specific Cost Elements Included in the PFAS-Selective IX Model
Cost Category
Major Components Included
Direct Capital
Costs
Booster pumps for influent water
Pre-treatment cartridge filters
Pressure vessels that contain the resin bed
Tanks and pumps for initial rinse and (optionally) backwash of the resin bed
Tanks (with secondary containment), pumps and mixers for delivering
sodium hydroxide for use in post-treatment corrosion control (optional)
Associated piping, valves, and instrumentation
O&M Costs: Labor
Operator labor for pre-treatment filters
Operator labor for managing backwash/rinse events
Operator labor for backwash pump maintenance (only if backwash occurs
weekly or more frequently)
Operator labor for resin replacement
O&M Costs:
Materials
Replacement cartridges for pre-treatment filters
Materials for backwash pump maintenance (only if backwash occurs weekly
or more frequently)
Chemical usage (if post-treatment corrosion control is selected)
Replacement virgin PFAS-selective resin
O&M Costs:
Energy
Operating energy for backwash/rinse pumps
O&M Costs:
Residuals
Disposal of spent cartridge filters
Discharge fees for spent backwash/rinse
Disposal of spent resin
Abbreviations: IX - ion exchange; O&M - operation & maintenance; PFAS - per-and polyfluoroalkyl substances.
For small systems (less than 1 MGD), the PFAS-selective IX model assumes the use of package
treatment systems that are pre-assembled in a factory, mounted on a skid, and transported to the
site. The IX model estimates costs for package systems using an approach similar to that
described for the GAC model, applying a variant set of inputs and assumptions that reduce the
size and cost of the treatment system (see Section 5.3.1.2.3). U.S. EPA (2023j) provides
complete details on the variant design assumptions used for IX package plants.
To generate the cost equations discussed in Section 5.3.1.3, EPA used the following key inputs
in the PFAS-selective IX model:
• Two vessels in series with a minimum total EBCT of 6 minutes
• Bed life varying over a range from 20,000 to 440,000 BV, estimated as discussed in
Section 5.3.1.1
EPA generated separate cost equations for two spent resin management scenarios:
• Spent resin managed as non-hazardous and sent off-site for incineration
Proposed PFAS Rule Economic Analysis
5-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Spent resin managed as hazardous and sent off-site for incineration.
The T&C document (U.S. EPA, 2023h) provides a comprehensive discussion of these and other
key inputs and assumptions.
5.3.1.2.5 RO/NF Model
Work Breakdown Structure-Based Cost Model for Reverse Osmosis Nanofiltration Drinking
Water Treatment provides a complete description of the engineering design process used by the
WBS model for RO/NF (U.S. EPA, 20231). Table 5-14 shows the technology-specific capital
equipment and O&M requirements included in the model. These items are in addition to the
common WBS cost elements listed in Table 5-11.
Table 5-14: Technology-Specific Cost Elements Included in the RO/NF Model
Cost Category
Major Components Included
Direct Capital
Costs
• High-pressure pumps for influent water and (optionally) interstage pressure
boost
Pre-treatment cartridge filters
Tanks, pumps, and mixers for pretreatment chemicals
Pressure vessels, membrane elements, piping, connectors, and steel structure
for the membrane racks
Valves for concentrate control and (optionally) per-stage throttle
Tanks, pumps, screens, cartridge filters, and heaters for membrane cleaning
Equipment, including dedicated concentrate discharge piping, for managing
RO/NF concentrate and spent cleaning chemicals
Associated pipes, valves, and instrumentation
O&M Costs: Labor
Operator labor for pre-treatment filters
Operator labor for routine O&M of membrane units
Operator labor to maintain membrane cleaning equipment
O&M Costs:
Materials
Replacement cartridges for pre-treatment filters
Chemical usage for pretreatment
Maintenance materials for pre-treatment, membrane process, and cleaning
equipment
Replacement membrane elements
Chemical usage for cleaning
O&M Costs:
Energy
Energy for high-pressure pumping
O&M Costs:
Residuals
Disposal costs for spent cartridge filters and membrane elements
Abbreviations: O&M - operation & maintenance; PFAS - per-and polytluoroalkyl substances; RO/NF - reverse
o smo sis/nano filtration.
The RO/NF model includes three default ground waters and three default surface waters, ranging
from high to low quality (i.e., from low to high total dissolved solids and scaling potential). To
generate the cost equations discussed in Section 5.3.1.3, EPA used the model's default high-
Proposed PFAS Rule Economic Analysis
5-24
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
quality influent water parameters to reflect the incremental cost of removing PFAS from
otherwise potable water. EPA used the following additional key inputs and assumptions:
• For systems larger than approximately 0.5 MGD, target recovery rates of 80 percent for
ground water and 85 percent for surface water21
• Target recovery rates of 70 to 75 percent for smaller systems
• Flux rates of 19 gallons per square foot per day (gfd) for ground water and 15 to 16 gfd
for surface water
• Direct discharge of RO/NF concentrate to a permitted outfall on a non-potable water
body (e.g., ocean or brackish estuary) via 10,000 feet of buried dedicated piping.
The T&C document (U.S. EPA, 2023h) provides a comprehensive discussion of these and other
key inputs and assumptions.
5.3.1.2.6 Non-treatment Model
Work Breakdown Structure-Based Cost Model for Nontreatment Options for Drinking Water
Compliance provides a complete description of the engineering design process used by the WBS
model for nontreatment actions (U.S. EPA, 2023k). The model can estimate costs for two
nontreatment alternatives: interconnection with another system and drilling new wells to replace
a contaminated source. Table 5-15 shows the technology-specific capital equipment and O&M
requirements included in the model for each alternative. The interconnection alternative does not
include any buildings. It includes all the indirect capital costs shown in Table 5-15 except for
yard piping, site work, and architectural fees. The new well alternative includes a small shed or
other low-cost building at the well site along with materials and labor for maintenance of this
building. It includes all the indirect capital costs shown in Table 5-15 except for yard piping.
Table 5-15: Technology-Specific Cost Elements Included in the Non-Treatment
Model
Cost Category
Major Components Included for
Interconnection
Major Components Included for
New Wells
Direct Capital
• Booster pumps or
•
Well casing, screens, and
Costs
pressure reducing valves
plugs
(depending on pressure at
•
Well installation costs
supply source)
including drilling.
• Concrete vaults (buried)
development, gravel pack.
for booster pumps or
and surface seals
pressure reducing valves
•
Well pumps
• Interconnecting piping
•
Piping (buried) and valves
(buried) and valves
to connect the new well to
the system
O&M Costs: Labor
• Operator labor for O&M
of booster pumps or
pressure reducing valves
(depending on pressure at
supply source) and
interconnecting valves
•
Operator labor for
operating and maintaining
well pumps and valves
21 Recovery rate is the percent of flow influent to RO that is recovered as useable treated water (permeate), as opposed to lost as
residual concentrate. It is not directly related to percent removal of PFAS.
Proposed PFAS Rule Economic Analysis
5-25
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-15: Technology-Specific Cost Elements Included in the Non-Treatment
Model
Cost Category Major Components Included for
Interconnection
Major Components Included for
New Wells
O&M Costs:
Materials
Cost of purchased water
Materials for maintaining
booster pumps (if
required by pressure at
supply source)
Materials for maintaining
well pumps
O&M Costs:
Energy
Energy for operating
booster pumps (if
required by pressure at
supply source)
• Energy for operating well
pumps
Abbreviations: O&M - operation & maintenance.
To generate the cost equations discussed in Section 5.3.1.3, EPA used the following key inputs
in the non-treatment model for interconnection:
• An interconnection distance of 10,000 feet
• Minimal differences in pressure between the supplier and the purchasing system, so that
neither booster pumps nor pressure reducing valves are needed
• An average cost of purchased water of $3.00 per thousand gallons in 2020 dollars.22
For new wells, EPA used the following key inputs:
• A maximum well capacity of 500 gallons per minute (gpm), such that one new well is
installed per 500 gpm of water production capacity required
• A well depth of 250 feet
• 500 feet of distance between the new wells and the distribution system.
The T&C document (U.S. EPA, 2023h) provides a comprehensive discussion of these and other
key inputs and assumptions.
5.3.1.3 WBS Cost Equations
EPA developed the cost estimates for PFAS treatment using outputs from the WBS models.
Outputs from these models are point estimates of total capital and O&M cost that correspond to a
given set of inputs that include design flow and average daily flow in MGD. Separately for total
capital and annual O&M cost, EPA fit cost equations to the WBS outputs for up to 49 different
flow rates. EPA choose from among several possible equation forms: linear, quadratic, cubic,
power, exponential, and logarithmic. For each equation, EPA selected the form that resulted in
the best correlation coefficient (R2), subject to the requirement that the equation must be
monotonically increasing over the appropriate range of flow rates (i.e., within the flow rate
category, the equation must always result in higher estimated costs for higher flow systems than
22 The WBS model presents costs in 2020 dollars, but the economic analysis is adjusted to present all costs and benefits in 2021
dollars.
Proposed PFAS Rule Economic Analysis
5-26
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
for lower flow systems). The resulting cost equations take one of the following forms, identified
by which coefficients (CI through CIO) are nonzero:
Equation 7:
Cost = CI QC2
or = C3 Ln(Q) + C4
or = C5 e(C<5 Q)
or =C7Q3 + C8Q2 + C9Q + CIO
In each case, Q is design flow in MGD for total capital costs, or average flow in MGD for annual
O&M costs. The resulting costs are in 2020 dollars.23
The equations are categorized by water source (surface water or ground water) and component
level (low, mid, or high cost). EPA developed separate equations for small, medium, or large
systems. These equations apply as follows:
• Small system equations apply where design flow (Q) is less than 1 MGD
• Medium system equations apply where design flow (Q) is 1 MGD or greater, but less
than 10 MGD
• Large system equations apply where design flow (Q) is 10 MGD or greater.
SafeWater MCBC selects from among the small, medium, and large equations and applies the
equations using the treated flow of the entry point. For GAC, IX, and non-treatment alternatives,
the treated flow is the entire flow of the entry point. Because RO/NF can continuously achieve
high removal efficiencies for PFAS, PWSs that require lower removals may be able to treat a
portion of their total flow and blend treated water and untreated water to meet regulatory
standards. EPA assumes systems using RO/NF will employ blending when they require less than
95 percent removal. Data presented in the T&C document (U.S. EPA, 2023h) show that RO/NF
can achieve greater than 95 percent removal efficiency for most PFAS compounds. Therefore,
this assumption errs on the side of higher costs. Accordingly, for entry points using RO/NF that
require less than 95 percent removal, SafeWater MCBC calculates a blending ratio and treated
design and average flow as follows:
23 The WBS model presents costs in 2020 dollars, but the economic analysis is adjusted to present all costs and benefits in 2021
dollars.
Proposed PFAS Rule Economic Analysis
5-27
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Equation 8:
D _ final,RO
B ~0.95
Qtreated,design — B X Qtotal,design
Qtreated,average — B X Qtotal,design
Where:
B = the blending ratio expressed as a decimal
%Rfmai,Ro = removal required from RO/NF expressed as a decimal and calculated as
described in Section 5.3.1.1.2
0.95 = the continuous removal achieved by RO/NF; an assumption based on data presented
in the T&C document (U.S. EPA, 2023h)
Qtreated = treated portion of entry point flow in MGD
Qtotai = total entry point flow in MGD
SafeWater MCBC assumes that entry points using RO/NF that require 95 percent removal or
greater will not employ blending and treat their entire flow.
For GAC and IX, EPA developed separate equations that vary according to the estimated bed
life. These equations are in increments of 5,000 BV for GAC and 20,000 BV for IX. Each bed
life increment corresponds to a change in media replacement frequency of two to five months,
depending on the average flow of the entry point. For entry points using GAC or IX, SafeWater
MCBC selects from among these equations based on the final operating bed life calculated as
described in Section 5.3.1.1.1, rounded down to the nearest increment of 5,000 BV for GAC and
20,000 BV for IX.
For GAC, there are separate equations for pressure designs and gravity designs. For ground
water entry points using GAC, EPA assumed PWSs would always use pressure designs to
maintain their existing pressure head. For surface water entry points using GAC, EPA assumed
PWSs would choose between pressure and gravity based on the design that results in the lower
annualized cost.
In total, there are almost 3,500 individual cost equations across the categories of capital and
O&M cost, water source, component level, flow, bed life (for GAC and IX), residuals
management scenario (for GAC and IX), and design type (for GAC). The T&C document (U.S.
EPA, 2023h) presents the equations in tabular form.
5.3.1.4 Incremental Treatment Costs of Other PFAS
EPA has estimated the national level costs of the proposed rule associated with PFOA, PFOS and
PFHxS. There are limitations with nationally representative occurrence information for the other
compounds in the proposed rule (PFNA, HFPO-DA and PFBS), therefore the additional
treatment cost, from co-occurrence of PFNA, HFPO-DA, PFBS or other PFAS, at systems
already required to treat because of PFOA, PFOS, or PFHxS MCL and HI exceedances are not
quantitatively assessed in the national cost estimates. Nor are treatment costs for systems that
Proposed PFAS Rule Economic Analysis
5-28
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
exceed the HI based on the combined occurrence of PFNA, HFPO-DA, PFBS, and PFHxS
(where PFHxS itself does not exceed its HBWC of 9.0 ppt) included in the national monetized
cost estimates. This section discusses EPA's model system approach for estimating potential
incremental treatment costs associated with co-occuring PFAS at systems already required to
treat in the national model framework and the potential per system costs for the set of systems
triggered into treatment as a result of HI exceedances not already captured in the national
analysis.
EPA's approach utilizes unit treatment cost information on three types of systems:
1. Baseline System: this model system has occurrence of PFAS included in the national
analysis (PFOA, PFOS, and PFHxS). It reflects the costs that are covered in the national
analysis and provides a basis for comparison.
2. System Type 1: this model system has no detections of PFOA, PFOS, or PFHxS.
However, it has occurrence of all the other PFAS considered in the HI. EPA considered
two scenarios for this system type: high occurrence of the other HI PFAS and medium
occurrence of the other HI PFAS. This system type represents additional systems that are
not currently captured in the national costs but would incur treatment costs because they
exceed the HI requirement under the proposed option.
3. System Type 2: this model system has occurrence of PFOA, PFOS, and/or PFHxS
identical to the baseline system. It also has occurrence of the other HI PFAS considered
in the proposed option. Like System Type 1, EPA considered two scenarios: high
occurrence of the other PFAS and medium occurrence of the other PFAS. This system
type illustrates a range of potential incremental treatment costs for systems that are
already treating in the national analysis.
Model System Type 1 cost estimate results characterize the system level costs that accrue as a
result of HI exceedances at locations that are not already treating for PFOA, PFOS, and/or
PFHxS in the national cost analysis. Model System Type 2 costs minus those of the Baseline
System provides the incremental system level cost for PWSs that are treating for PFOA, PFOS,
and/or PFHxS in the national model but also have significant concentrations of the other HI
PFAS that must be removed.
In this analysis, concentrations for PFOA, PFOS, and PFHxS correspond to the median for each
contaminant from the UCMR3 data, considering detected values only. Concentrations for the
other PFAS are 95th percentile and median values based on EPA's analysis of state-level
occurrence data. For more information on assumed baseline characteristics See Appendix N.3.
Given this occurrence information and basic system characteristics by system size category, EPA
estimated a range of costs for model systems in each size category for each of the three treatment
technologies (GAC, IX, and RO/NF). The range of costs reflects all combinations of two source
waters (ground and surface) and two cost levels (low and high). For GAC and IX, the range of
costs also incorporates two bed life scenarios corresponding to a range of influent TOC.
EPA has conducted additional occurrence modeling that indicates that 100-500 systems are
estimated to not exceed the PFOA and/or PFOS MCLs but are estimated to exceed the HI. In the
national model approximately 500 systems are estimated to exceed the HI based on PFHxS data
Proposed PFAS Rule Economic Analysis
5-29
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
alone. However, some of these systems are also estimated to exceed the PFOA and/or PFOS
MCLs. Therefore, a subset of the estimated 100-500 systems estimated to exceed the HI only
have already been captured in the national analysis because EPA includes an estimate of systems
where PFHxS exceeds 9.0 ppt in the national cost analysis. EPA does not capture HI related
treatment costs associated with HFPO-DA, PFNA, and PFBS in the national cost analysis.
Instead, EPA assesses Type 1 model systems, which represent additional systems that are not
currently captured in the national costs but would incur treatment costs because they exceed the
HI requirement. These systems are estimated to incur treatment costs in general ranging from
0.70 to 1.77 times the estimated baseline system costs. Type 1 systems with moderate occurrence
for HFPO-DA, PFNA, and PFBS have estimated costs that are the same as or somewhat lower
than systems captured in the national analysis (0.70 to 1.00 times baseline). Type 1 systems with
high occurrence (95th percentile) have estimated costs slightly lower to somewhat higher than
systems captured in the national analysis (0.92 to 1.77 times baseline).
EPA's national cost model estimated number of systems which exceed one or more limits
(MCLs for PFOA and/or PFOS and/or the HI for PFHxS alone) is approximately 4,300. Some
fraction of these systems may incur increased treatment costs because of the co-occurrence of
additional PFAS. As explained above, EPA used the UCMR3 median and 95th percentile HFPO-
DA, PFBS, and PFNA (the HI PFAS not already included in the national analysis) data to
characterize the potential change in treatment cost at the system level given co-occurrence. The
modeled Type 2 systems are designed to assess these impacts. Overall, the need to remove these
other HI compounds could increase treatment costs by 0 to 77 percent on a per-system basis. For
both IX and RO/NF there is no appreciable increase in the cost of treatment when the additional
PFAS are found, even when concentrations of HFPO-DA, PFBS, and PFNA are all present at the
95th percentile level. Only systems using GAC are expected to incur increased per system costs.
At the upper bound of the GAC cost range, the high TOC influent combined with the need to
remove the other HI compounds (particularly HFPO-DA) results in a shorter bed life and
increased costs of operation. Type 2 modeled systems with median co-occurrence for HFPO-DA,
PFNA, and PFBS experience increases in estimated GAC treatment costs that range from 0 to 9
percent. For Type 2 systems that with high co-occurrence (95th percentile of the additional HI
PFAS) GAC treatment costs increased from 0 to 77 percent. Based on EPA's national model
results, EPA estimates that of those 4,300 systems that are required to treat because of MCL
and/or HI exceedances GAC will be installed at approximately 50-85 percent of entry points,
depending on source water type and other factors (see Section 5.3.1.1).
For further detail on the assumptions and findings of EPA's analysis of incremental costs of
other PFAS, see Appendix N.3.
5.3.2 Estimating PWS Sampling and Administrative Costs
This section details how EPA estimated the costs of compliance with the system sampling and
administrative activities associated with the proposed rule. In the subsections of 5.3.2, EPA
organizes and presents the cost information based on the series of activities that are required to
comply with the proposed PFAS NPDWR, with tables for each data element used to calculate the
proposed rule component costs. These tables include the data element name and a description of
the data variable, as well as any relevant sources for the data. EPA presents the costs categorized
as follows:
Proposed PFAS Rule Economic Analysis
5-30
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Administrative costs associated with implementation (Section 5.3.2.1);
• Sampling costs (Section 5.3.2.2); and
• Administrative costs associated with treatment (Section 5.3.2.3).
Consistent with standard Agency practice, EPA assumes compliance with the rule throughout the
economic analysis, and as a result, SafeWater MCBC does not accrue costs to any system for the
Tier 2 and 3 public notifications. Nevertheless, EPA presents a qualitative discussion of the
public notification costs potentially associated with the proposed rule in Section 5.3.2.4.
5.3.2.1 Implementation Administration Costs
Systems conduct the following one-time actions to begin implementation of the rule:
• Reading and understanding the rule; and
• Attending training provided by primacy agencies.
EPA assumes that systems will conduct these activities during years one through three of the
period of analysis. Table 5-16 lists the data elements and provides descriptions, values, and
sources for these costs. The cost per system for each activity is the product of the hourly labor
cost (labor sys rate) and the hours (hrs sys adopt rule and hrs sys initial ta), which vary by
system size. The total cost is the sum of per-system costs.
Table 5-16: Implementation Administration Startup Costs ($2021)
Data Element Name
Data Element
Description
Data Element Value
Data Element
Source
labor sys rate
The labor rate per
$35.48 (systems <3,300)
WBS Technical
hour for systems
$37.84 (systems 3,301-10,000)
$39.94 (systems 10,001-50,000)
$41.70 (systems 50,001-
100,000)'
$48.74 (systems >100,000)
Labor Cost
hrssysadoptrule
The average hours
per system to read
and adopt the rule
4 hours per system
Arsenic in Drinking
Water Rule
Economic Analysis
(EPA 815-R-00-
026)
hrssysinitialta
The average hours
16 hours per system (systems
Arsenic in Drinking
per system to attend
<3,300)
Water Rule
one-time training
32 hours per system (systems
Economic Analysis
provided by primacy
>3,300)
(EPA 815-R-00-
agencies
026)
Abbreviation: WBS - work breakdown structure.
Proposed PFAS Rule Economic Analysis
5-31
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5.3.2.2 Sampling Costs
EPA assumes that there will be initial and long-term monitoring for the proposed rule. As Table
5-17 shows, surface and ground water systems serving 10,000 or more people will collect one
sample each quarter, at each entry point, during the initial 12-month monitoring period. Surface
water systems serving 10,000 or fewer people are also required to collect a quarterly sample at
each entry point during the initial 12-month period. Ground water systems that serve 10,000 or
fewer people will be required to sample once at each entry point on a semi-annual basis for the
first 12-month monitoring period.
Long-term monitoring requirements differ based on two factors: (1) system size, and (2) whether
a system can demonstrate during the initial monitoring period that they are "reliably and
consistently" below the proposed MCLs for PFAS. EPA has set the PWS size threshold at
systems serving 3,300 or fewer people. The threshold for systems to demonstrate that they are
"reliably and consistently" below the proposed MCLs is set at a trigger level of one-third the
MCLs for PFOA or PFOS (1.3 ppt) or the HI (0.33). For systems below the trigger level values
during the initial 12-month monitoring period and in future long-term monitoring periods may
conduct triennial monitoring. Systems serving 3,300 or fewer people will collect one triennial
sample per entry point. Systems providing water for more than 3,300 people will take one
sample in two consecutive quarters at each entry point, totaling two samples in each triennial
period. For systems with concentration values at or above the trigger level regardless of system
size, a quarterly sample must be taken at each entry point.
For any samples that have a detection, the system will analyze the field reagent blank samples
collected at the same time as the monitoring sample. Systems that have an MCL exceedance will
collect one additional sample from the relevant entry point to confirm the results (i.e., a
confirmation sample) (U.S. EPA, 2004).
Proposed PFAS Rule Economic Analysis
5-32
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-17: Initial and Long-Term Sampling Frequencies Per System Entry Point
Long-Term
Initial
Monitoring
System
Size
Category
Initial 12-
Month
Monitoring
Period
Long-Term
Monitoring
System Size
Category
Monitoring3:
PFAS
Detection < 1.3
ppt (PFOA or
PFOS) or HI <
0.33
Long-Term
Monitoring3: PFAS
Detection > 1.3 ppt
(PFOA or PFOS) or
HI >0.33
Surface Water: 1
sample every
< 10,000
quarter
Groundwater: 1
sample every 6-
month period
Surface Water
< 3,300
1 triennial sample
2 triennial samples
1 sample every quarter
>10,000
and Ground
Water: 1 sample
every quarter
>3,300
(1 sample in two
consecutive
quarters)
1 sample every quarter
Abbreviations: HI - hazard index; PFAS - per-and polyfluoroalkyl substances.
Note:
aEPA used the following thresholds to distinguish whether PFAS concentrations are reliably and consistently below the
maximum contaminant level (MCL): PFOA and PFOS - one-third the MCL for each option; PFHxS - one-third the health
benchmark of 9 ng/L or 3 ng/L.
For the national cost analysis, EPA assumes that systems with either UCMR 5 data or monitoring
data in the State PFAS Database will not need to conduct the initial year of monitoring (See
Chapter 3.1.4). As a simplifying assumption for the cost analysis, EPA assumes all systems
serving a population of greater than 3,300 have UCMR 5 data and those with 3,300 or less do
not. For the State PFAS Database, EPA relied on the PWSIDs stored in the database and
exempted those systems from the first year of monitoring in the cost analysis.
EPA assumes that systems with an MCL exceedance will implement actions to comply with the
MCL by the compliance date. As indicated in 5.3.1, EPA assumes a treatment target, for systems
required to treat for PFAS, that includes a margin of safety so finished water PFAS levels at
these systems are 80 percent of the MCL or HI. This target is insufficient to meet the triennial
monitoring threshold. Therefore, systems implementing treatment will continue with quarterly
monitoring. All other systems that do not have PFAS concentrations at or below the trigger level
threshold will also continue quarterly monitoring.
For all systems, the activities associated with the sample collection in the initial 12-month
monitoring period are the labor burden and cost for the sample collection and analysis, as well as
a review of the sample results. Table 5-18 presents the data needs associated with the
implementation monitoring period. The cost per entry point for each sampling activity is the
product of the hourly labor cost and the hours plus the laboratory analysis cost. The laboratory
analysis cost will include the additional field blank cost when occurrence values exceed method
detection limits. The total cost is the sum of per-entry point costs.
Proposed PFAS Rule Economic Analysis
5-33
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-18: Sampling Costs ($2021)
Data Element Name
Data Element Description
Data Element Value
Data Element
Source
laborsysrate
The labor rate per hour for
$35.48 (systems <3,300)
WBS Technical
systems
$37.84 (systems 3,301-
10,000)
$39.94 (systems 10,001-
50,000)
$41.70 (systems 50,001-
100,000)
$48.74 (systems >100,000)
Labor Cost
numbintialsamples
The number of samples per
entry point per monitoring
round for the initial
monitoring in Year 1
2 samples (Ground Water
systems <10,000)
4 samples (all other
systems)3
Proposed rule
numb quarterly samples
The number of samples per
entry point per long-term
monitoring year for entry
points that exceed the
triennial monitoring
threshold
4 samples (all systems)
Proposed rule
numbtriennialsamples
The number of samples per
entry point per long-term
monitoring round for entry
points that meet the triennial
threshold
1 sample (systems <3,300)
2 samples (systems >3,300)
Proposed rule
hrssamp
The hours per sample to
travel to sampling locations,
collect samples, record any
additional information,
submit samples to a
laboratory, and review results
1 hour
UCMR5 ICR (EPA-
HQ-OW-2020-
0530-00141)
EPA533_cost
The laboratory analysis cost
per sample for EPA Method
533
$376
UCMR5 ICR (EPA-
HQ-OW-2020-
0530-0141)
EPA537_cost
The laboratory analysis cost
per sample for EPA Method
537.1
$302
UCMR5 ICR (EPA-
HQ-OW-2020-
0530-0141)
EPA53 3_fieldblank_cost
The laboratory analysis cost
per sample for field reagent
blank under EPA Method
533
$327b
EP A53 7_fieldblank_co st
The laboratory analysis cost
per sample for the field
reagent blank under EPA
Method 537.1
$266b
Abbreviations: EPA - U.S. Environmental Protection Agency; Ground Water - ground water ICR - Information Collection
Request; UCMR - Unregulated Contaminant Monitoring Rule; WBS - work breakdown structure.
Notes:
aSystems greater than 3,300 will rely on UCMR 5 data and a subset of other systems will rely on data in the State PFAS
Monitoring Database.
bThis incremental sample cost applies to all samples that exceed method detection limits. EPA used the Method 537.1
detection limits to apply this cost because Method 533 does not include detection limits.
Proposed PFAS Rule Economic Analysis
5-34
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5.3.2.3 Treatment Administration Costs
As described in Section 5.3.1, any system with an MCL exceedance adopts either a treatment or
non-treatment alternative to comply with proposed rule. The majority of systems are anticipated
to install treatment technologies while a subset, described in Section 5.3.1.1, will choose
alternative methods. EPA assumes that systems will have administrative costs associated with
obtaining permits for either the treatment or non-treatment methods. The costs vary depending
on whether the system installs treatment or selects a non-treatment method. For the economic
analysis, EPA assumes that systems install treatment in the fourth year of the period of analysis.
Table 5-19 presents the data elements and sources for these costs. The cost per entry point
requiring treatment or changing water source is the product of the hourly labor cost and the hours
per the relevant permit request. The total cost is the sum of per-entry point costs.
Table 5-19: Treatment Administration Costs ($2021)
Data Element Name
Data Element
Data Element Value
Data Element
Description
Source
laborsysrate
The labor rate per hour
$35.48 (systems <3,300)
WBS Technical
for systems
$37.84 (systems 3,301-
Labor Cost
10,000)
$39.94 (systems 10,001-
50,000)
$41.70 (systems 50,001-
100,000)'
$48.74 (systems
>100,000)
hrssystreat
The hours per entry
3 hours (systems <100)
Lead and Copper
point for a system to
5 hours (systems 101-
Rule Revisions
notify, consult, and
500)
Support Material
submit a permit request
7 hours (systems 501-
(EPA-HQ-OW-
for treatment
1,000)
2017-0300-1701)
installation3
12 hours (systems 1,001-
3,300)
22 hours (systems 3,301-
50,000)
42 hours (systems
>50,000)
hrssyssource
The hours per entry
6 hours
Lead and Copper
point for a system to
Rule Revisions
notify, consult, and
Support Material
submit a permit request
(EPA-HQ-OW-
for source water change
2017-0300-1700)
or alternative method3
Abbreviations: WBS - work breakdown structure.
Note:
aThe Lead and Copper Rule Revisions presents this burden per system, but EPA applied the cost per entry point for this
economic analysis because the notification, consultation, and permitting process occurs for individual entry points.
5.3.2.4 Public Notification Costs
While EPA assumes full compliance with the rule and does not include public notification costs
in the cost estimates, there are public notification requirements in the proposed rule for systems
with certain violations. The proposed rule designates MCL violations for PFAS as Tier 2, which
Proposed PFAS Rule Economic Analysis
5-35
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
requires systems to provide public notification as soon as practical, but no later than 30 days
after the system learns of the violation. The system must repeat notice every three months if the
violation or situation persists unless the primacy agency determines otherwise. At a minimum,
systems must give repeat notice at least once per year.
The proposed rule designates monitoring and testing procedure violations as Tier 3, which
requires systems to provide public notice not later than one year after the system learns of the
violation. The system must repeat the notice annually for as long as the violation persists. The
system may use an annual report detailing all violations that occurred during the previous year if
the timing requirements of the public notification are met.
To provide an approximate estimate of the burden associated with the Tier 2 and 3 violations,
EPA reviewed the ICR for the Public Water System Supervision (PWSS) Program, which
includes Tier 2 and 3 notifications. Table 5-20 presents the PWSS Program ICR burdens for the
preparation and delivery of the Tier 2 and 3 public notifications.
Table 5-20: Public Notification Burden Estimate
Data Element3
Data Element Value
Data Element Source
Preparation of initial Tier 2 notices
3.5 hours
PWSS Program ICR (EPA-HQ-
OW-2011-0433-0003)
Preparation of initial Tier 3 notices
3 hours (CWS)
3.5 hours (NTNCWS)
PWSS Program ICR (EPA-HQ-
OW-2011-0433-0003)
Delivery of initial Tier 2 notices
9 hours (CWS <500)
30 hours (CWS >500)
9 hours (NTNCWS)
PWSS Program ICR (EPA-HQ-
OW-2011-0433-0003)
Development and delivery of
repeated Tier 2 and 3 notices
3 hours
PWSS Program ICR (EPA-HQ-
OW-2011-0433-0003)
Abbreviations: CWS - community water system; PWSS - public water systems.
Note:
^Delivery of Tier 3 notices must occur not later than one year after the system learns of the violation. EPA assumes systems
will include this notice with the Consumer Confidence Reports sent to all customers annually, therefore Tier 3 delivery costs
are assumed to be zero.
5.4 Estimating Primacy Agency Costs
In addition to the PWS costs associated with the rule implementation, EPA assumes primacy
agencies will have upfront implementation costs as well as costs associated with the system
actions related to sampling and treatment. The activities associated with primacy agencies under
the proposed rule include:
• Reading and understanding the rule, as well as adopting regulatory requirements;
• Providing internal training for the rule implementation
• Providing systems with training and technical assistance during the rule implementation;
• Reporting to EPA on an ongoing basis any PFAS-specific information under 40 CFR
142.15 regarding violations as well as enforcement actions and general operations of
public water supply programs;
Proposed PFAS Rule Economic Analysis
5-36
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Reviewing the sample results during the implementation monitoring period and the SMF
monitoring period; and
• Reviewing and consulting with systems on the installation of treatment technology or
alternative methods, including source water change.
With the exception of the first four activities listed above, the primacy agency burdens are
incurred in response to an action taken by a system. For example, the cost to primacy agencies of
reviewing any sample result depends on the number of samples taken at each entry point by each
system under the jurisdiction of the primacy agency. Table 5-21 presents the data elements and
sources for all primacy agency costs. The data element descriptions indicate whether the cost is
per primacy agency, per sample, per system, or per entry point. In each instance, the primacy
agency labor rate is multiplied by the number of relevant hours and the activity frequency.
Table 5-21: Primacy Agency Costs ($2021)
Data Element
Name
Data Element Description
Data Element Value
Data Element Source
labor_pa_rate
hrs_pa_adopt_rule
hrspatrain
hrspainitialta
hrs sdwis
hrspareportep
hrspatreat
The labor rate per hour for primacy
agencies
The average hours per primacy
agency to read and understand the
rule, as well as adopt regulatory
requirements
The average hours per primacy
agency to provide initial training to
internal staff
The average hours per primacy
agency to provide initial training
and technical assistance to systems
The average hours per primacy
agency to report annually to EPA
information under 40 CFR 142.15
regarding violations, variances and
exemptions, enforcement actions
and general operations of State
public water supply programs
The hours per sample for a primacy
agency to review sample results
The hours per entry point for a
primacy agency to review and
consult on installation of a
treatment technique13
$58.14
416 hours per primacy
agency
250 hours per primacy
agency
2,080 hours per
primacy agency
0
1 hour
3 hours (systems
<100)
5 hours (systems 101-
500)
7 hours (systems 501-
1,000)
12 hours (systems
1,001-3,300)
22 hours (systems
3,301-50,000)
42 hours (systems
>50,000)
Loaded labor rate
(including the cost of
benefits) derived from the
Bureau of Labor Statistics3
Arsenic in Drinking Water
Rule Economic Analysis
(EPA 815-R-00-026)
Arsenic in Drinking Water
Rule Economic Analysis
(EPA 815-R-00-026)
Arsenic in Drinking Water
Rule Economic Analysis
(EPA 815-R-00-026)
EPA assumes that the
proposed PFAS rule will
have no discernable
incremental burden for
quarterly or annual reports
to SDWIS/Fed
Arsenic in Drinking Water
Rule Economic Analysis
(EPA 815-R-00-026)
Lead and Copper Rule
Revisions Support Material
(EPA-HQ-OW-2017-0300-
1701)
Proposed PFAS Rule Economic Analysis
5-37
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 5-21: Primacy Agency Costs ($2021)
Data Element Data Element Description Data Element Value
Data Element Source
Name
hrs_pa_source The hours per entry point for a 4 hours
primacy agency to review and
consult on a source water changeb
Lead and Copper Rule
Revisions Support Material
(EP A-HQ-0 W-2017-0300-
1700)
Notes:
aState employee wage rate of $33.91 from National Occupational Employment and Wage Estimates, United States, BLS SOC
Code 19-2041, "State Government, excluding schools and hospitals - Environmental Scientists and Specialists, Including
Health," hourly mean wage rate. May 2020 data (published in March 2021): https://www.bls.gov/oes/current/oesl92041.htm.
Wages are loaded using a factor of 62.2 from the BLS Employer Costs for Employee Compensation report, Table 3, March
2020. Percent of total compensation - Wages and Salaries - All Workers - State and Local Government Workers
(https://www.bls.gOv/news.release/archives/ecec_06182020.pdf). See worksheet BLS Table 3. The final loaded wage is
adjusted for inflation.
bThe Lead and Copper Rule Revisions present this burden per system, but EPA has applied the cost per entry point for this
economic analysis because the notification, consultation, and permitting process occurs for individual entry points.
In addition to the costs described above, a primacy agency may also have to review the
certification of any Tier 2 or 3 public notifications sent out by systems. EPA assumes full
compliance with the proposed rule but provides a brief discussion of the possible system costs
associated with this component in Section 5.3.2.4. The public notification burden associated with
primacy agencies is between 0.33 and 0.5 hours per system to review the system certification of
the public notification. The burden is derived from the LCRR estimates for a similar activity.
PWS-level cost estimates for the proposed rule (proposed option) and other regulatory options
are provided in Appendix C. PWS-level cost are provided for all PWSs by PWS-type, size
category, primary source water type, and ownership. In addition, a second set of PWS-level costs
are provided for PWSs that must take an action to comply with the rule (treat or change water
Household-level cost estimates for the proposed rule and other regulatory options are provided in
Appendix C. Household-level cost are provided for all CWSs by size category, primary source
water type, and ownership. In addition, a second set of household-level costs are provided for
households served by CWSs that must take an action to comply with the rule (treat or change
water source).24
24
Note that EPA does compute per household technology cost values in the separate national small system affordability
determination analysis. These household values are distinct from the values generated in the national cost estimates as they
include only small system compliance technology cost. For three small system size categories (systems serving 25-500, 501 -
3,300, and 3,301-10,000) EPA estimates a per household treatment technology cost range including the minimum and maximum
cost values. These cost estimates are based on system characteristics, contaminant reduction requirements, and technology
efficacy, across the set of small system compliance technology options. See Chapter 9.12 the EA for additional information on
the national small system affordability determination.
5.5 PWS-Level Cost Estimates
source).
5.6 Household-Level Cost Estimates
Proposed PFAS Rule Economic Analysis
5-38
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
5.7 Discussion of Data Limitations and Uncertainty
The preceding sections identify the nonquantifiable costs and the uncertainty information
incorporated in the quantitative cost analysis. There are also data limitations that could not be
incorporated in this analysis. Chapter 7 and Table 7-6 outline the nonquantifiable costs
associated with the regulatory requirements of the proposed option as well as Options la-c.
Table 5-22 lists the data limitations and characterizes the impact on the quantitative cost
analysis. EPA notes that in most cases it is not possible to judge the extent to which a particular
limitation or uncertainty could affect the cost analysis. EPA provides the potential direction of
the impact on the cost estimates when possible but does not prioritize the entries with respect to
the impact magnitude.
Table 5-22: Limitations that Apply to the Cost Analysis for the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Quantitative
Analysis
Notes
WBS engineering cost
model assumptions and
component costs
Uncertain
The WBS engineering cost models require many design
and operating assumptions to estimate treatment process
equipment and operating needs. Section 5.3.1 addressed
the bed life assumption. The Technologies and Costs
document (U.S. EPA, 2023h) and individual WBS models
in the rule docket provide additional information. The
component-level costs approximate national average costs,
which can over- or under-estimate costs at systems affected
by the proposed rule.
Compliance forecast
Uncertain
The forecast probabilities are based on historical full-scale
compliance actions. Site-specific water quality conditions,
changes in technology, and changes in market conditions
can result in future technology selections that differ from
the compliance forecast.
Total organic carbon
concentration
Uncertain
The randomly assigned values from the two national
distributions are based on a limited dataset. Actual TOC
concentrations at systems affected by the proposed rule can
be higher or lower than the assigned values.
National occurrence data
for HFPO-DA, PFBS, and
PFNA not available
Underestimate
The hazard index in the proposed option would regulate
PFBS, PFNA, and HFPO-DA in addition to the modeled
PFAS. In instances when concentrations of PFBS, PFNA,
and/or HFPO-DA are high enough to cause a hazard index
exceedance, the modeled costs may be underestimated. If
these PFAS occur in isolation at levels that affect treatment
decisions, or if they occur in sufficient concentration to
result in an exceedance when the concentration of PFHxS
alone would be below the HI, then costs would be
underestimated. Note that EPA has conducted an analysis
of the potential changes in system level treatment cost
associated with the occurrence of PFBS, PFNA, and
HFPO-DA using a model system approach which is
discussed in detail in Section 5.3.1.4 and Appendix N.3.
POU not included in
compliance forecast
Overestimate
If POU devices can be certified to meet concentrations that
satisfy the proposed rule, then small systems may be able
to reduce costs by using a POU compliance option instead
of centralized treatment or source water changes.
Process wastes not
classified as hazardous
Underestimate
The national cost analysis reflects the assumption that
PFAS-contaminated wastes are not considered hazardous
Proposed PFAS Rule Economic Analysis
5-39
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 5-22: Limitations that Apply to the Cost Analysis for the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Quantitative
Analysis
Notes
wastes. As a general matter, EPA notes that such wastes
are not currently regulated under federal law as a
hazardous waste. To address stakeholder concerns,
including those raised during the SBREFA process, EPA
conducted a sensitivity analysis with an assumption of
hazardous waste disposal for illustrative purposes only. As
part of this analysis, EPA generated a second full set of
unit cost curves that are identical to the curves used for the
national cost analysis with the exception that spent GAC
and spent IX resin are considered hazardous. EPA
acknowledges that if federal authorities later determine that
PFAS-contaminated wastes require handling as hazardous
wastes, the residuals management costs in the WBS
treatment cost models are expected to be higher. See
Appendix N for a sensitivity analysis describing the
potential increase in costs associated with hazardous waste
disposal at 100 percent of systems treating for PFAS. The
costs estimated in Appendix N are consistent with EPA
OLEM's "Interim Guidance on the Destruction and
Disposal of Perfluoroalkyl and Polyfluoroalkyl Substances
and Materials Containing Perfluoroalkyl and
Polyfluoroalkyl Substances" (U.S. EPA, 2020b).
Abbreviations: HFPO-DA - hexafluoropropylene oxide dimer acid; PFAS - per and polyfluoroalkyl substances; PFBS -
perfluorobutanesulfonic acid; PFNA - perfluorononanoic acid; POU - point-of-use; WBS - work breakdown structure.
Proposed PFAS Rule Economic Analysis
5-40
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6 Benefits Analysis
6.1 Introduction
This chapter discusses the potential quantified and nonquantifiable25 benefits to human health
resulting from changes in PFAS levels in drinking water due to implementation of the proposed
rule, as well as several regulatory alternatives. EPA's quantification of health benefits resulting
from reduced PFAS exposure in drinking water was driven by PFAS occurrence estimates,
pharmacokinetic (PK) model availability, information on exposure-response relationships, and
available information to monetize avoided cases of illness. EPA either quantitatively assesses or
qualitatively discusses health endpoints associated with exposure to PFAS. EPA assesses
potential benefits quantitatively if evidence of exposure and health effects is likely, it is possible
to link the outcome to risk of a health effect, and there is no overlap in effect with another
quantified endpoint in the same outcome group. Only a subset of the avoided morbidity and
mortality stemming from reduced PFAS levels in drinking water can be quantified and
monetized. The monetized benefits evaluated in the Economic Analysis for the proposed rule
include changes in human health risks associated with cardiovascular disease (CVD) and infant
birth weight from reduced exposure to PFOA and PFOS in drinking water and renal cell
carcinoma from reduced exposure to PFOA. EPA also quantified benefits from reducing bladder
cancer risk due to the co-removal of non-PFAS pollutants via the installation of drinking water
treatment, discussed in greater detail in Section 6.7. EPA was not able to quantify or monetize
other benefits, including those related to possible immune, hepatic, endocrine, metabolic,
reproductive, musculoskeletal, or other outcomes. EPA discusses these benefits qualitatively in
more detail below in Section 6.2 of the Economic Analysis.
EPA analyses the quantified costs and benefits of setting individual MCLs for PFOA and PFOS
at 4.0 ppt, 5.0 ppt, and 10.0 ppt, referred to as Options la through lc respectively. As discussed
in Section 2.1, the regulatory options include treatment thresholds that would reduce PFAS
levels in finished drinking water by various amounts. The change in PFAS levels at a particular
water system depends on baseline PFAS levels estimated using the occurrence model (Section
4.4) and the PFAS treatment threshold specified under each regulatory alternative.
EPA notes that the quantified benefits alone of this analysis are a significant underestimate of the
total benefits expected to result from this rule. Hence, as mandated by SDWA section
1412(b)(3)(C), EPA has considered both quantifiable and nonquantifiable benefits in informing
its decision making that the costs of this rule are clearly justified by the benefits.
6.1.1 Chapter Overview
Section 6.2 provides an overview of the health benefits categories considered in the analysis of
reductions of PFAS in drinking water. In addition to describing the benefits EPA is able to
quantify, this section includes a robust qualitative discussion of nonquantifiable benefits.
Because of the broad adverse health impacts of PFAS on many endpoints, the nonquantifiable
benefits of this proposed rule are likely substantial. Section 6.3 describes the application of
EPA's pharmacokinetic models for PFAS to estimate changes in blood serum concentrations
under each regulatory alternative. Section 6.4 presents the methodology and results of the
25 Nonquantifiable benefits are discussed qualitatively.
Proposed PFAS Rule Economic Analysis
6-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
impacts of the PFAS regulatory alternatives on a subset of developmental outcomes, namely
infant birth weight. Section 6.5 presents the methodology and results of the impacts of the PFAS
regulatory alternatives on cardiovascular disease (CVD) incidence. Section 6.6 presents the
methodology and results of the impacts of the PFAS regulatory alternatives on the incidence of
Renal Cell Carcinoma (RCC), one of the cancers with known association to PFAS exposure.
Section 6.7 presents the methodology and results of the impacts of the PFAS regulatory
alternatives on DBP formation and the associated incidence of bladder cancer. Finally, Section
6.8 describes limitations and uncertainties of the benefits analyses.
6.1.2 Uncertainty Chorocterizotion
EPA characterizes sources of uncertainty in its analysis of potential quantified benefits resulting
from changes in PFAS levels in drinking water. The analysis reports uncertainty bounds for
benefits estimated in each health endpoint category modeled for the proposed rule. Each lower
(upper) bound value is the 5th (95th) percentile of the category-specific benefits estimate
distribution represented by 4,000 Monte Carlo draws. Table 6-1 provides an overview of the
specific sources of uncertainty that EPA quantified in this benefits analysis. In addition to these
sources of uncertainty, reported uncertainty bounds also reflect the following upstream sources
of uncertainty: baseline PFAS occurrence (Section 4.4), affected population size and
demographic composition (Section 4.3), and the magnitude of PFAS concentration reductions
(Section 4.4). These analysis-specific sources of uncertainty are further described in Appendix L.
Table 6-1: Quantified Sources of Uncertainty in Benefits Estimates
Source Description of Uncertainty
Health effect-serum PFAS slope The slope factors that express the effects of serum PFOA and serum PFOS
factors on health outcomes (birth weight, CVD,a and RCC) are based either on
EPA meta-analyses or high-quality studies that provide a central estimate
and a confidence interval for the slope factors. To characterize uncertainty,
EPA assumed that these slope factors have a normal distribution with a
mean set at the central estimate and the standard deviation set at the
estimated standard error.
RCC risk reduction cap EPA implemented a cap on the cumulative RCC risk reductions due to
reductions in serum PFOA based on the PAF estimates for a range of
cancers and enviromnental contaminants. This parameter is treated as
uncertain; its uncertainty is characterized by a log-uniform distribution
with a minimum set at the smallest PAF estimate identified in the literature
and a maximum set at the largest PAF estimate identified in the literature.
The central estimate for the PAF is the mean of this log-uniform
distribution.
Abbreviations: PFAS - per and polyfluoroalkyl substances; PFOA - perfluorooctanoic acid; PFOS - perfluorooctane sulfonic
acid; RCC - renal cell carcinoma; PAF - population attributable fraction.
Note:
aThe slope factors contributing to the CVD benefits analysis include the relationship between total cholesterol and PFOA and
PFOS, and the relationship between blood pressure and PFOS.
EPA did not characterize the following sources of potentially quantifiable uncertainty: U.S.
population life tables (see Section 6.1.4), annual all-cause and health outcome-specific mortality
rates, coefficients of the CVD risk model linking total cholesterol (TC), high-density lipoprotein
cholesterol (HDLC), and blood pressure (BP) to cardiovascular event incidence (Goff et al.,
2014), CVD risk model predictors (e.g., share of smokers) estimated from health survey data,
Proposed PFAS Rule Economic Analysis
6-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
prevalence of CVD event history in the U.S. population, distribution of CVD events by type, the
estimated infant mortality-birth weight slope factor (See Section 6.4.3.1), state-level distributions
of infant births and infant deaths over discrete birth weight ranges, the 200-g cap on birth weight
changes estimated under the rule, cost of illness estimates for all modeled non-fatal health
outcomes, the Value of Statistical Life reference value, the Value of Statistical Life income
elasticity value used to approximate the Value of Statistical Life income growth adjustment, and
the gross domestic product per capita projection used for the Value of Statistical Life income
growth adjustment (see Appendix J). EPA expects that the sources listed in Table 6-1, in addition
to uncertainty surrounding the estimates of PFAS occurrence, affected population size, and the
magnitude of PFAS reduction, account for a substantial portion of the uncertainty in the benefits
analysis.
6.1.3 Summary of Quantified National Benefits Estimates of the
Proposed Rule
This section provides summary outputs for the benefits analysis of the proposed rule as well as
Options la-c. Total annual benefits include human health risk reduction benefits for the health
outcomes listed in Section 6.1.1. EPA annualized benefit values for each endpoint at two
discount rates, 3 percent and 7 percent. Both the expected value and the 90 percent confidence
interval is provided.
As discussed in Section 2.1, for purposes of this analysis, EPA is considering the benefits
analysis for the proposed option to be representative of the alternate regulatory approach where
PFHxS, PFNA, PFBS, and HFPO-DA would be regulated by individual MCLs in addition to or
instead of using the HI approach.
Table 6-2: National Annualized Benefits, Proposed Option (PFOA and PFOS MCLs of
4.0 ppt and HI of 1.0; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Annualized CVD Benefits
$111.78
$533.48
$1,051.00
$85.94
$421.10
$822.88
Annualized Birth Weight
Benefits
$97.36
$177.66
$279.49
$74.62
$139.01
$219.43
Annualized RCC Benefits
$54.23
$300.56
$758.03
$45.36
$217.37
$515.89
Annualized Bladder
Cancer Benefits
$173.09
$221.30
$273.62
$102.08
$130.63
$161.56
Total Annualized Rule
Benefitsb
$659.91
$1,232.98
$1,991.51
$477.69
$908.11
$1,462.43
Abbreviations: CVD - cardiovascular disease; RCC - renal cell carcinoma.
Note: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because health effects
are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-3: National Annualized Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Annualized CVD Benefits
$110.45
$525.05
$1,035.36
$86.32
$414.45
$817.79
Annualized Birth Weight
Benefits
$95.73
$175.05
$276.44
$74.66
$136.97
$217.02
Annualized RCC Benefits
$52.92
$295.53
$744.64
$45.09
$213.78
$508.56
Annualized Bladder
Cancer Benefits
$171.72
$220.48
$274.24
$101.34
$130.15
$161.56
Total Annualized Rule
Benefitsb
$651.19
$1,216.08
$1,971.01
$471.53
$895.36
$1,456.23
Abbreviations: C VD - cardiovascular disease; RCC - renal cell carcinoma.
Note: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because health effects
are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-4: National Annualized Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Annualized CVD
$99.73
$459.09
$908.82
$72.72
$362.42
$717.85
Benefits
Annualized Birth Weight
$83.27
$154.13
$246.43
$64.94
$120.59
$193.47
Benefits
Annualized RCC
$42.28
$250.60
$643.71
$36.32
$182.24
$446.80
Benefits
Annualized Bladder
$141.17
$183.10
$227.85
$83.31
$108.08
$135.37
Cancer Benefits
Total Annualized Rule
$553.37
$1,046.91
$1,706.81
$398.21
$773.33
$1,292.96
Benefitsb
Abbreviations: C VD - cardiovascular disease; RCC - renal cell carcinoma.
Note: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because health effects
are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-5: National Annualized Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt;
Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Annualized CVD Benefits
$51.00
$268.78
$571.32
$41.85
$212.18
$450.51
Annualized Birth Weight
Benefits
$43.22
$92.70
$164.19
$34.18
$72.51
$125.80
Annualized RCC Benefits
$18.58
$131.44
$367.38
$17.34
$97.30
$260.54
Annualized Bladder
Cancer Benefits
$68.26
$91.90
$118.64
$40.29
$54.25
$70.10
Total Annualized Rule
Benefitsb
$280.42
$584.80
$1,030.56
$208.71
$436.24
$784.59
Abbreviations: C VD - cardiovascular disease; RCC - renal cell carcinoma.
Note: Detail may not add exactly to total due to independent rounding. Percentiles cannot be summed because health effects
are not perfectly correlated.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
6.1.4 Life Table Modeling Background
EPA uses a life table modeling approach to evaluate reductions in CVD and cancer risk. This
approach allows for internally consistent estimation of the path-dependent health effects for
regulatory alternatives, including annual incidence of CVD events or cancers among those
without prior history of these conditions, which is dependent on the population prevalence of
these chronic conditions and survival over time.
The life table is a statistical tool used to analyze the mortality experience of a population over
time. Specifically, using data on the age-specific probability of death and the initial population
size (e.g., 100,000 persons), the life table computes the number of persons surviving to a specific
age, the number of deaths occurring at a given age, the number of person-years lived at a given
age, the number of person-years lived beyond a given age, and age-specific life expectancy. The
details of standard life table calculations can be found in Anderson (1999).
The life table modeling approach extends the standard life table calculations to characterize
populations with respect to their chronic condition status and estimate transitions into the
subpopulation affected by the chronic condition.26 EPA has previously used life table approaches
in regulatory analyses, including the analysis of lead-associated health effects in the 2015 Benefit
and Cost Analysis for the Effluent Limitations Guidelines, Standards for the Steam Electric
20 For example, a benefits model that evaluates the impact of contaminant exposure on incidence of cancer—a chronic
condition—would need to estimate the number of persons who are cancer free and, therefore, are eligible for the estimation of
new cancer risk (i.e., the risk of transition into the subpopulation affected by the chronic condition).
Proposed PFAS Rule Economic Analysis
6-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Power Generating Point Source Category (U.S. EPA, 2015), and PIVh.s-related health effects in
revisions to the National Ambient Air Quality Standards for ground-level ozone (U.S. EPA,
2008). Other examples of the use of a life table approach among federal agencies include EPA's
analysis of Benefits and Costs of the Clean Air Act from 1990 to 2020 (U.S. EPA, 201 la) and
the Occupational Safety and Health Administration (OSHA) assessment of lifetime excess lung
cancer, nonmalignant respiratory disease mortality, and silicosis risks from exposure to
respirable crystalline silica (OSHA, 2010; OSHA, 2016). Additionally, the Agency sought
advice from the EPA Science Advisory Board on the use of the life table in this application and
they supported this approach (U.S. EPA, 2022k). See Appendix G for details on application of
the life table for the CVD benefits analysis. See Appendix H for details on application of the life
table for cancer benefits analyses.
6.2 Overview of Benefit Categories
EPA notes that much of the information included in this section is based on draft MCLG
documents, which are expected to be finalized by the time of rule finalization. Therefore,
statements on evidence of associations between PFOA/PFOS and health effects may be updated.
Statements on evidence of associations between other PFAS compounds and health effects may
be updated as additional assessments are conducted and finalized. EPA's decision to quantify
health benefits resulting from reduced PFAS exposure in drinking water is driven by the
availability of PFAS related occurrence estimates, pharmacokinetic (PK) models, and
information on exposure-response relationships. In this benefits analysis, EPA either
quantitatively assesses or qualitatively discusses the health endpoints associated with exposure to
PFAS; EPA assesses potential benefits quantitatively if (1) there is indicative (likely) evidence of
a relationship between exposure and a health effect response, (2) it is possible to link the health
outcome (e.g., CVD) to risk of a health effect (e.g., increased total cholesterol), and (3) there is
no overlap in effect with another quantified endpoint in the same outcome group.
EPA describes occurrence modeling information in Section 4.4. Table 6-6 presents an overview
of the categories of health benefits expected to result from the implementation of treatment that
reduces PFAS levels in drinking water. The PFAS compounds that EPA identified as having
indicative evidence linking exposure to a particular health endpoint, as well as compounds
having reliable PK models estimating the distribution to PFAS compounds throughout the body,
include PFOA, PFOS, and PFNA.27
As seen in Table 6-6, only a small subset of the potential health effects of reduced PFAS levels
in drinking water can be quantified and monetized. The monetized benefits evaluated in this
proposed rulemaking include CVD, infant birth weight, and RCC. EPA also quantified benefits
from reducing bladder cancer risk due to the reduction of DBP formation as a result of the co-
removal of organic carbon via the installation of additional treatment for PFAS (Cantor et al.,
1998; Crittenden et al., 1993; Regli et al., 2015; Weisman et al., 2022). EPA notes that the
Agency anticipates additional benefits resulting from installing drinking water treatment for
PFAS chemicals and the subsequent removal of co-occurring non-PFAS contaminants, including
source water metals (e.g., chromium (VI)), organic regulated and unregulated contaminants,
(e.g., cyanotoxins (Foreman et al., 2021)), and certain pesticides. EPA was not able to quantify
27 EPA relies on the serum PFNA calculator from Lu et al. (2020). PFNA effects are described as part of a sensitivity analysis for
birth weight-related benefits in Appendix K.
Proposed PFAS Rule Economic Analysis
6-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
or monetize other benefits, including those related to possible immune, hepatic, endocrine,
metabolic, reproductive, musculoskeletal, many cancers, or other outcomes discussed in Section
6.1.2. EPA discusses these benefits qualitatively in Sections 6.2.2.2 and 6.2.4.
Proposed PFAS Rule Economic Analysis
6-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-6: Overview of Health Benefits Categories Considered in the Analysis of Changes in PFAS Drinking Water Levels
Health Outcome
PFAS Compounda'b
Benefits Analysis
Category
Endpoint
PFOA
PFOS PFNAd
Discussed
Quantitatively
Discussed
Qualitatively
Lipids
Total cholesterol (TC)
X
X X
X
High-density lipoprotein cholesterol (HDLC)
xc
xc
X
Low-density lipoprotein cholesterol (LDLC)
X
X
X
CVD
Blood pressure (BP)
X
X
Developmental
Birth weight
X
X X
X
Hepatic
Small for gestational age (SGA), non-birth
weight developmental
Alanine transaminase (ALT)
X
X
X
X
X
Immune
Antibody response (tetanus, diphtheria)
X
X
X
Metabolic
Leptin
X
X
Musculoskeletal
Osteoarthritis, bone mineral density
X
X
Cancer
Renal Cell Carcinoma (RCC)
Testicular
X
X
X
X
Abbreviations: PFAS - per- and polyfluoroalkyl substances.
Notes:
Tields marked with "X" indicate the PFAS compound for which there is evidence of an association with a given health outcome in humans.
bOutcomes with indicative (likely) evidence of an association between a PFAS compound and a health outcome are assessed quantitatively unless (1) there is an overlap within the same
outcome group (e.g., low density lipoprotein cholesterol overlaps with total cholesterol and small for gestational age overlaps with low birth weight), or (2) it is not possible to link the
outcome to the risk of the health effect (e.g., evidence is inconclusive regarding the relationship between PFOS exposure, leptin levels and associated health outcomes). Such health
outcomes are discussed qualitatively.
cAlthough evidence of associations between HDLC and PFOA and PFOS was mixed, certain individual studies reported robust associations in general adult populations (See Section
6.2.2.1.2 on Cardiovascular Effects). Based on comments and recommendations from the EPA SAB (U.S. EPA, 2022k), EPA assessed F1DLC in a sensitivity analysis (see Appendix K).
dNote that only PFOA and PFOS effects were modeled in the assessment of benefits under the proposed rule. PFNA was modeled only in sensitivity analyses of birth weight benefits
because some studies show a slight association between PFNA and birth weight effects, although the associations were not consistent (ATSDR, 2021; U.S. EPA, 2023d) and Lu et al.
(2020) provides an approach for estimating PFNA blood serum levels resulting from PFNA exposures in drinking water (see Appendix K).
Proposed PFAS Rule Economic Analysis
6-8
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
In Table 6-7, EPA presents an overview of the epidemiology and toxicology evidence regarding
the effects of exposure to PFAS compounds on health outcomes that were examined in various
EPA and ATSDR assessments. Health outcomes are classified as having:
• No evidence of an association28 (signified with a dot in the table);
• Evidence of an association noted as suggestive or slight (signified with an X in the table);
• Indicative (likely) evidence of an association (signified with a green-highlighted X in the
table);
• Health outcomes that are quantified in the benefits analysis for the proposed rule are
signified with a bold X*.
EPA further describes the associations, and supporting evidence of associations, in Section 6.2.2
for PFOA and PFOS and in Section 6.2.3 for additional PFAS compounds.
28 No evidence of an association is listed in instances where an absence of evidence precludes definitive conclusions about the
relationship between exposure and a given health effect and also when there is evidence demonstrating that exposure does not
result in a given health effect.
Proposed PFAS Rule Economic Analysis
March 2023
6-9
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on Health Outcomes
Health Outcomes
PFAS
0>
s.
H
¦a
V
W
u
H
HDLC
LDLC
BPa (human)Heart
histopathology (animal)
Birth weight
SGA, non-birth weight
developmental
ALT (human) Organ weight,
cell death (animal)
AbRa (tetanus, diphtheria)
(human) Various endpoints
(animal)
Thyroid hormone disruption
Leptin, body weight
(human)Body fat, body weight
(animal)
Uric acid (human) Organ
weight (animal)
Gestational hypertension/pre-
eclampsia (human)Various
endpoints (animal)
Osteoarthritis, bone mineral
density
Vitamin D levels, hemoglobin
levels, albumin levels
Other non-cancer
u
u
OS
Testicular
Other
PFOAg
Epi
X*
X
X
•
X*
X
X
X
X
X
•
X
X
•
•
X*
xb
•
Draft
MCLG
2021;
ATSDR
2021;
NASEM,
2022
Other non-cancer:
neurological
effects,
respiratory
effects,
gastrointestinal
Tox
•
•
•
•
X*
X
X
X
X
•
•
X
X
•
X
•
•
•
Draft
MCLG
2021;
ATSDR
2021
Other non-cancer:
neurological
effects,
respiratory
effects,
gastrointestinal
PFOSg
Epi
X*
X
X
X
x*c
•
X
X
X
•
•
xe
•
•
•
X
•
X
Draft
MCLG
2021;
ATSDR
2021;
NASEM,
2022
Other non-cancer:
neurological
effects,
gastrointestinal
Tox
•
•
•
•
X
X
X
X
X
•
•
X
•
•
X
•
•
X
Draft
MCLG
2021;
ATSDR
2021
Other non-cancer:
neurological
effects,
gastrointestinal
Proposed PFAS Rule Economic Analysis
6-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on Health Outcomes
Health Outcomes
PFAS
0>
s.
H
¦a
V
W
u
H
HDLC
LDLC
BPa (human)Heart
histopathology (animal)
Birth weight
SGA, non-birth weight
developmental
ALT (human) Organ weight,
cell death (animal)
AbRa (tetanus, diphtheria)
(human) Various endpoints
(animal)
Thyroid hormone disruption
Leptin, body weight
(human)Body fat, body weight
(animal)
Uric acid (human) Organ
weight (animal)
Gestational hypertension/pre-
eclampsia (human)Various
endpoints (animal)
Osteoarthritis, bone mineral
density
Vitamin D levels, hemoglobin
levels, albumin levels
Other non-cancer
u
u
OS
Testicular
Other
PFBAg
Epi
•
•
•
•
•
•
•
•
•
•
IRIS
Assessment
2022;
ATSDR
2021;
NASEM,
2022
No associations in
humans
Tox
•
•
X
X
•
X
•
•
•
•
•
•
IRIS
Assessment
2022;
ATSDR
2021
Other non-cancer:
ocular, respiratory
(ATSDR)
PFNAh
Epi
X
•
X
•
X
•
•
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
Other non-cancer:
respiratory effects
Tox
•
•
•
•
•
•
•
X
•
•
•
ATSDR
2021
Other non-cancer:
general toxicity
PFDAh
Epi
X
•
X
•
•
X
•
X
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
Tox
•
•
•
•
X
X
X
X
•
•
•
•
•
•
•
•
ATSDR
2021
Other non-cancer:
general toxicity
Proposed PFAS Rule Economic Analysis
6-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on Health Outcomes
PFAS
Evidence Type
Health Outcomes
Data Source(s)
Notes
Lipids
CVD
Developmental
Hepatic
Immune
Endocrine
Metabolic
Renal
Reproductive
Musculoskeletal
Hematologic
Other non-cancer
Cancer
u
H
HDLC
LDLC
BPa (human)Heart
histopathology (animal)
Birth weight
SGA, non-birth weight
developmental
ALT (human) Organ weight,
cell death (animal)
AbRa (tetanus, diphtheria)
(human) Various endpoints
(animal)
Thyroid hormone disruption
Leptin, body weight
(human)Body fat, body weight
(animal)
Uric acid (human) Organ
weight (animal)
Gestational hypertension/pre-
eclampsia (human)Various
endpoints (animal)
Osteoarthritis, bone mineral
density
Vitamin D levels, hemoglobin
levels, albumin levels
Other non-cancer
u
u
OS
Testicular
Other
PFHxS11
Epi
•
•
•
•
•
•
•
X
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
Tox
•
•
•
•
•
X
•
•
•
•
•
•
•
ATSDR
2021
Other non-cancer:
respiratory effects
PFHxAg
Epi
•
•
•
•
•
IRIS Draft
Assessment
2022;
ATSDR
2021;
NASEM,
2022
No associations in
humans
Tox
•
•
•
•
X
•
X
•
X
•
•
•
•
•
•
•
IRIS Draft
Assessment
2022;
ATSDR
2021
Other non-cancer:
nervous,
respiratory
(ATSDR)
PFBSf
Epi
•
•
•
•
•
•
•
EPA
Human
Health
Toxicity
Study 2021;
ATSDR
2021;
NASEM,
2022
No associations in
humans
Proposed PFAS Rule Economic Analysis
6-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on Health Outcomes
PFAS
Evidence Type
Health Outcomes
Data Source(s)
Notes
Lipids
CVD
Developmental
Hepatic
Immune
Endocrine
Metabolic
Renal
Reproductive
Musculoskeletal
Hematologic
Other non-cancer
Cancer
u
H
HDLC
LDLC
BPa (human)Heart
histopathology (animal)
Birth weight
SGA, non-birth weight
developmental
ALT (human) Organ weight,
cell death (animal)
AbRa (tetanus, diphtheria)
(human) Various endpoints
(animal)
Thyroid hormone disruption
Leptin, body weight
(human)Body fat, body weight
(animal)
Uric acid (human) Organ
weight (animal)
Gestational hypertension/pre-
eclampsia (human)Various
endpoints (animal)
Osteoarthritis, bone mineral
density
Vitamin D levels, hemoglobin
levels, albumin levels
Other non-cancer
u
u
OS
Testicular
Other
Tox
•
•
•
•
•
X
•
•
X
•
X
•
•
•
•
EPA
Human
Health
Toxicity
Study 2021;
ATSDR
2021
Other non-cancer:
respiratory effects
(ATSDR)
PFHpA
Epi
•
•
•
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
No associations in
humans
Tox
ATSDR
2021
PFUnA
Epi
•
•
•
•
•
•
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
No associations in
humans
Tox
X
•
•
•
ATSDR
2021
PFDoDA
Epi
•
•
•
•
•
•
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
No associations in
humans
Tox
•
•
•
•
•
•
•
•
•
ATSDR
2021
FOSA
Epi
•
•
•
•
•
•
•
ATSDR
2021;
NASEM,
2022
No associations in
humans
Proposed PFAS Rule Economic Analysis
6-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-7: Overview of Epidemiology and Toxicology Evidence of PFAS Effects on Health Outcomes
Health Outcomes
PFAS
a
H
O
Lipids
CVD
Developmental
Hepatic
Immune
Endocrine
Metabolic
Renal
Reproductive
Musculoskeletal
Hematologic
Other non-cancer
Cancer
Data Source(s)
Notes
fi
3
'>
u
H
HDLC
LDLC
BPa (human)Heart
histopathology (animal)
Birth weight
SGA, non-birth weight
developmental
ALT (human) Organ weight,
cell death (animal)
AbRa (tetanus, diphtheria)
(human) Various endpoints
(animal)
Thyroid hormone disruption
Leptin, body weight
(human)Body fat, body weight
(animal)
Uric acid (human) Organ
weight (animal)
Gestational hypertension/pre-
eclampsia (human)Various
endpoints (animal)
Osteoarthritis, bone mineral
density
Vitamin D levels, hemoglobin
levels, albumin levels
Other non-cancer
u
u
OS
Testicular
Other
Tox
•
•
ATSDR
2021
HFPO-
Epi
EPA
HFPO-DA
2021 final
toxicity
assessment
No data from
epidemiology
studies
DAf
Tox
•
•
X
X
X
X
•
X
X
EPA
HFPO-DA
2021 final
toxicity
assessment
Cancer: liver
tumors
Notes:
• Health outcomes examined, no evidence of associations (also noted as inadequate, or equivocal evidence).
X Health outcomes examined, slight or suggestive evidence of associations.
X Health outcomes examined, indicative (likely) evidence of associations (also noted as supports a hazard in IRIS assessments, evidence indicates, or evidence demonstrates).
X* Health outcomes quantified in benefits analyses, indicative (likely) evidence of associations.
[Blank cell] Health outcome was not examined.
aAbR: antibody response; BP: blood pressure; Epi: epidemiology; Tox: toxicology; RCC: renal cell carcinoma.
bSupported based onPFOA HESD (2016) andBartell et al. (2021) meta-analysis.
Supported by Dzierlenga et al. (2020) meta-analysis,
developmental delays: IRIS Draft Assessments (2021).
eAlso supported by recent meta-analysis from Gao et al. (2021) (PFOS and preeclampsia risk).
fPublished final EPA assessments.
gPublished draft EPA assessments.
hUnpublished draft EPA assessments.
Proposed PFAS Rule Economic Analysis
6-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.2.1 Availability of Pharmacokinetic (PK) Models
PK models describe the distribution of chemicals in the body and pharmacodynamic relation
between blood concentration and clinical effects. EPA evaluated existing PFOA and PFOS PK
models for their utility in predicting internal doses for use in both cancer and non-cancer dose-
response assessments (U.S. EPA, 2023d; U.S. EPA, 2023e). PFOA and PFOS PK models
typically take one of three forms:
• Classical compartment models, where modelers define the body as a one- or two-
compartment system with volumes and intercompartmental transfer fit specifically to the
PFAS PK dataset. The most common approach for prediction of serum PFAS levels is to
apply a simple single-compartment model.
• Modified compartment models, where modelers attempt to characterize absorption,
distribution, metabolism, and/or excretion through protein-binding, cardiac output, and
known renal elimination. These models also rely on fitting PFAS data to non-
physiological parameters.
• Physiological-based pharmacokinetic (PBPK) models, where tissues and organs of the
body are described as physiological-based compartments. In these models, transport
between compartments is informed by measures of blood flow and tissue perfusion.
These models are fit to time-course concentration data.
EPA's Toxicity Assessment and Proposed Maximum Contaminant Level Goal for PFOA in
Drinking Water (U.S. EPA, 2023 e) and Toxicity Assessment and Proposed Maximum
Contaminant Level Goal for PFOS in Drinking Water (U.S. EPA, 2023d)29 describe existing
PFOA and PFOS PK models. Briefly, EPA developed single-compartment PK models for adult
males and females to estimate blood serum PFOA and PFOS concentrations. These models are
described in U.S. EPA (2023e, 2023d), and the application of these models in health risk benefits
modeling is described in Section 6.3.
6.2.2 Benefits of PFOA and PFOS Exposure Reduction
This section provides an overview of the potential health benefits of reduced exposure to PFOA
and PFOS in drinking water. These benefits are expected to be realized as avoided adverse health
effects as a result of the proposed NPDWR, in addition to the benefits that EPA has quantified.
EPA identified a wide range of potential health effects associated with exposure to PFOA and
PFOS using five comprehensive federal government documents that summarize the recent
literature on PFAS (mainly PFOA and PFOS) exposure and its health impacts: EPA's Health
Effects Support Document for PFOA and Health Effects Support Document for PFOS, hereafter
referred to as the EPA HESDs (U.S. EPA, 2016e; U.S. EPA, 2016f); EPA's Toxicity Assessments
and Proposed Maximum Contaminant Level Goals for PFOA and PFOS in Drinking Water (U.S.
EPA, 2023d; U.S. EPA, 2023e); and the U.S. Department of Health and Human Services Agency
for Toxic Substances and Disease Registry's (ATSDR) Toxicological Profile for Perflnor oalkyls
(ATSDR, 2021). Each source presents comprehensive literature reviews on adverse health
effects associated with PFOA and PFOS.
29 For brevity, these documents are described throughout as EPA's Toxicity' Assessments and Proposed Maximum Contaminant
Level Goals for PFOA and PFOS in Drinking Water.
Proposed PFAS Rule Economic Analysis 6-12 March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
The most recent literature reviews on PFAS exposures and health impacts, which are included in
EPA's Toxicity Assessments and Proposed Maximum Contaminant Level Goals for PFOA and
PFOS in Drinking Water, discuss the weight of evidence supporting PFOA and PFOS
associations with health outcomes as indicative (likely), inadequate, or suggestive (U.S. EPA,
2023 d; U.S. EPA, 2023 e). For the purposes of the reviews conducted to develop the proposed
MCLGs, an association is deemed indicative when findings are consistent and supported by
substantial evidence. The association is inadequate if there is a lack of information or an inability
to interpret the available evidence (e.g., findings across studies). The association is suggestive if
findings are consistent but supported by a limited number of studies or analyses, or only
observed in certain populations or species. Note that these determinations are based on
information available as of February 2022. Section 6.2.2.1 discusses PFAS-related health effects
that were considered quantitatively (modeled and monetized) in the benefits analysis, while
Section 6.2.2.2 discusses PFAS-related health effects that were considered only qualitatively in
the benefits analysis. These sections specify whether evidence is based on animal (toxicology) or
human (epidemiology) studies, or both.
6.2.2.1 Quantitative benefits of PFOA and PFOS Exposure Reduction
In this section, EPA discusses some of the health benefits expected to result from reduced
exposure to PFOA and PFOS in drinking water. These benefits are expected to be realized as
avoided adverse health effects as a result of the proposed NPDWR and are quantified in Sections
6.4, 6.5, and 6.6 respectively.
6.2.2.1.1 Developmental Effects
Exposure to PFOA and PFOS is linked to developmental effects such as infant birth weight, birth
length, head circumference at birth, and other effects (Verner et al., 2015; U.S. EPA, 2016e; U.S.
EPA, 2016f; Negri et al., 2017; AT SDR, 2018; Waterfield et al., 2020; U.S. EPA, 2023d; U.S.
EPA, 2023e). Low birth weight (LBW) is an important health outcome affected by PFOA/PFOS
exposure because it is a significant factor in survival rates and medical care costs among infants
(ATSDR, 2021). Infants are exposed prenatally to PFOA and PFOS through maternal serum via
the placenta.
Because data on the cost of incremental changes in birth weight are available from Klein et al.
(2018), EPA selected birth weight as a key developmental health effect when assessing the
health impacts of reduced PFOA and PFOS exposures. Epidemiology studies on PFOA
supported an increased risk of LBW in infants with PFOA exposures (U.S. EPA, 2023e).
Similarly, epidemiology studies on PFOS showed an increased risk of LBW infants with PFOS
exposures. Overall, most epidemiology studies evaluating the association between maternal
serum PFOA/PFOS and birth weight reported negative relationships (i.e., increased exposure is
associated with decreased birth weight) (Darrow et al., 2013; Verner et al., 2015; Govarts et al.,
2016; Negri et al., 2017; Starling et al., 2017; Sagiv et al., 2018; Chu et al., 2020; Dzierlenga et
al., 2020; Wikstrom et al., 2020; Yao et al., 2021).30 Toxicology studies on PFOA further
supported an association between decreased offspring weight and PFOA exposure; several
studies conducted on rodents showed decreased fetal and pup weight with gestational PFOA
30 Recent evidence indicates that relationships between maternal serum PFOA/PFOS and birth weight may be impacted by
changes in pregnancy hemodynamics (Sagiv et al., 2018; Steenland et al., 2018).
Proposed PFAS Rule Economic Analysis
6-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
exposure (U.S. EPA, 2023e). Toxicology studies also reported that increased exposure to PFOS
was associated with decreased body weight in rodent fetuses and pups (U.S. EPA, 2023d). For
additional details on developmental effects studies and their individual outcomes, see Chapter
3.4.1 (Developmental) in U.S. EPA (2023d) and U.S. EPA (2023e). See Section 6.4 for EPA's
analysis of avoided infant birth weight impacts as a result of reduced PFOA and PFOS exposure
from the proposed rule.
6.2.2.1.2 Cardiovascular Effects
CVD is one of the leading causes of premature mortality in the U.S. (D'Agostino et al., 2008;
Goff et al., 2014; Lloyd-Jones et al., 2017). As discussed in EPA's Toxicity Assessments and
Proposed Maximum Contaminant Level Goals for PFOA and PFOS in Drinking Water, exposure
to PFOA and PFOS through drinking water contributes to increased serum PFOA and PFOS
concentrations and potentially elevated levels of TC, changes in levels of HDLC, and elevated
levels of systolic BP (U.S. EPA, 2023e; U.S. EPA, 2023d). Changes in TC, HDLC, and BP are
associated with changes in incidence of CVD events such as myocardial infarction (MI, i.e.,
heart attack), ischemic stroke (IS), and cardiovascular mortality occurring in populations without
prior CVD event experience (D'Agostino et al., 2008; Goff et al., 2014; Lloyd-Jones et al.,
2017).
Overall, epidemiology evidence suggested a positive association between PFOS/PFOA exposure
and TC levels (i.e., increased exposure is associated with increased TC levels) (ATSDR, 2021;
U.S. EPA, 2023d; U.S. EPA, 2023e). While most epidemiology studies reported positive
associations between exposure to PFOA and TC, some results were not statistically significant.
Epidemiology studies observed consistent positive associations between PFOA and LDLC (U.S.
EPA, 2023e). Most epidemiology studies on PFOS exposure pointed to a positive association
between exposure and TC levels (ATSDR, 2021). This association was observed in children as
well as in the general adult population and pregnant women (U.S. EPA, 2023d). Toxicology
studies generally reported decreases in serum lipids from oral exposure to PFOA and PFOS (U.S.
EPA, 2023e; U.S. EPA, 2023d). Although the biological significance of the decrease in various
serum lipid levels observed in animal models regardless of species, sex, or exposure paradigm is
unclear, these effects do indicate a disruption in lipid metabolism, which is coherent with effects
observed in humans. For additional details on the TC studies and their individual outcomes, see
Chapter 3.4.4 (Cardiovascular) in U.S. EPA (2023d) and U.S. EPA (2023e).
Existing epidemiology and toxicology studies provided inadequate evidence of associations
between PFOA and PFOS exposures and HDLC levels, with a mix of positive and some inverse
associations in adult populations (ATSDR, 2021; U.S. EPA, 2023d; U.S. EPA, 2023e). A single
study reported a statistically significant positive association between PFOA and HDLC in
pregnant women (Starling et al., 2017). In children, prenatal exposure was associated with lower
HDLC, especially in boys, whereas childhood exposure was associated with higher HDLC
(ATSDR, 2021; U.S. EPA, 2023e). Similarly, studies did not report consistent associations
between PFOS and HDLC levels (ATSDR, 2021; U.S. EPA, 2023d). Most of the evidence in
adults involved cross-sectional assessments, although associations between PFOS and lower
HDLC were also observed in the cohort study by P.-I. D. Lin et al. (2019). The available
evidence is currently limited to a single study that reported null associations between PFOS and
HDLC in pregnant women (Starling et al., 2017, U.S. EPA, 2023d). Toxicology studies of oral
exposure to PFOA and PFOS reported decreases in serum lipids levels, including HDLC, after
Proposed PFAS Rule Economic Analysis
6-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
exposure (U.S. EPA, 2023d; U.S. EPA, 2023e). Although evidence of associations between
PFOA and PFOS exposures and HDLC were mixed, certain individual studies reported robust
associations in general adult populations. Based on comments and recommendations from the
EPA SAB on EPA's analysis of CVD risk reductions resulting from changes in PFOA/PFOS
exposures (U.S. EPA, 2021a), EPA assessed HDLC in a sensitivity analysis (see Appendix K).
For additional details on the HDLC studies and their individual outcomes, see Chapter 3.4.4
(Cardiovascular) of U.S. EPA (2023 d) and U.S. EPA (2023 e).
Epidemiology studies observed inconsistent associations between PFOA exposure and BP
(ATSDR, 2021; U.S. EPA, 2023d; U.S. EPA, 2023e). Some epidemiology studies reported
positive associations between PFOA exposure and risk of hypertension (defined as elevated BP)
in adults, but the data were inconsistent (U.S. EPA, 2023e). Five studies in children, adolescents,
and pregnant women suggested no association between PFOA exposure and elevated BP (U.S.
EPA, 2023e). In adults, there was evidence of positive associations between PFOS exposure and
BP, although the results were not always consistent between systolic BP and diastolic BP, and
one study reported an inverse association (U.S. EPA, 2023d). However, there was overall
consistent evidence of an association between PFOS and BP in studies conducted in general
adult populations (U.S. EPA, 2023d). Evidence for associations between PFOS exposure and BP
in children and adolescents was limited and did not suggest an association with elevated BP
(U.S. EPA, 2023d). However, exposure duration was a limitation in these studies, and evidence
of an association between PFOS and increased risk of hypertension, specifically, was limited and
inconsistent (U.S. EPA, 2023d). ATSDR reported a single toxicology study that evaluated the
association between PFOS exposure and BP; systolic BP was significantly increased in female
and male offspring of exposed pregnant female rats (Rogers et al., 2014; ATSDR, 2021). For
additional details on the BP studies and their individual outcomes, see Chapter 3.4.4
(Cardiovascular) in U.S. EPA (2023d) and U.S. EPA (2023e).
Given the breadth of evidence linking PFOA and PFOS exposure to effects on TC and BP in
general adult populations, EPA quantified public health impacts of changes in these well-
established CVD risk biomarkers (D'Agostino et al., 2008; Goff et al., 2014; Lloyd-Jones et al.,
2017) by estimating changes in incidence of several CVD events. Specifically, EPA assumed
that PFOA/PFOS-related changes in TC and BP had the same effect on the CVD risk as the
changes unrelated to chemical exposure and used the Pooled Cohort Atherosclerotic
Cardiovascular Disease (ASCVD) model (Goff et al., 2014) to evaluate their impacts on the
incidence of MI, IS, and cardiovascular mortality occurring in populations without prior CVD
event experience (see Section 6.5). EPA observed that the direct evidence of associations
between PFOA/PFOS exposure and CVD risk was limited and inconsistent (U.S. EPA, 2023e;
U.S. EPA, 2023d), with mixed findings reported by one high-quality longitudinal epidemiology
study (Mattsson et al., 2015) and four medium-quality cross-sectional epidemiology studies
(Huang et al., 2018; Shankar et al., 2012; Hutcheson et al., 2019; Fry et al., 2017). However,
inconclusive evidence of the direct association between PFOA/PFOS exposure and CVD effects
from a limited collection of studies does not imply the absence of such an association. Future
analyses of CVD effects using large longitudinal studies, such as the ones used to develop the
ASCVD model (Goff et al., 2014), could help elucidate whether there is a consistent direct
association between PFOA/PFOS and CVD risk. See Section 6.5 for EPA's analysis of reduced
CVD impacts as a result of reduced PFOA and PFOS exposure from the proposed rule.
Proposed PFAS Rule Economic Analysis
6-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.2.2.1.3 Renal Cell Carcinoma
Data on the association between PFOA exposure and kidney cancer (i.e., RCC) suggest a
positive association between exposure and increased risk of RCC. Epidemiology studies
indicated that exposure to PFOA was associated with an increased risk of RCC (CalEPA, 2021;
U.S. EPA, 2016f; AT SDR, 2021 AT SDR, 2021; U.S. EPA, 2023e). In the HESD for PFOA
(U.S. EPA, 2016f), EPA determined that PFOA is likely to be carcinogenic to humans (U.S.
EPA, 2005c) based in part on evidence of associations between PFOA exposure and kidney
cancer in humans. PFOA exposure effects on RCC were shown in two occupational population
studies (Raleigh et al., 2014; Steenland et al., 2012) and two high-exposure community studies
(Vieira et al., 2013; Barry et al., 2013). A recent study of the relationship between PFOA and
RCC in the U.S. general population found strong evidence of a positive association between
exposure to PFOA and RCC in humans (Shearer et al., 2021). In EPA's Toxicity Assessment and
Proposed Maximum Contaminant Level Goal for PFOA in Drinking Water, the agency reviewed
the weight of the evidence and determined that PFOA is Likely to Be Carcinogenic to Humans,
as "the evidence is adequate to demonstrate carcinogenic potential to humans but does not reach
the weight of evidence for the descriptor Carcinogenic to HumansThis determination is based
on the evidence of kidney and testicular cancer in humans and Ley dig cell tumors (LCTs),
pancreatic acinar cell tumors (PACTs), and hepatocellular adenomas in rats. See Section 6.6 for
EPA's analysis of the benefits of reduced RCC as a result of reduced PFOA exposures from the
proposed rule.
Evidence of a positive association between PFOS exposure and kidney cancer was inconclusive;
the small number and limited scope of studies at the time were inadequate to make definitive
conclusions (U.S. EPA, 2016e; U.S. EPA, 2023d). One recent study observed an association
between PFOS and an increased risk of RCC in the highest exposed quartile and per doubling of
PFOS concentration (Shearer et al., 2021; U.S. EPA, 2023d). However, the association was no
longer statistically significant after adjusting for other PFAS (Shearer et al., 2021). EPA did not
report any PFOA or PFOS toxicology studies specifically relating to RCC, although there was
evidence of other cancer types in rodent models treated with PFOA (U.S. EPA, 2023d; U.S.
EPA, 2023e). For additional details on cancer studies and their individual outcomes, see Chapter
3.5 (Cancer) in U.S. EPA (2023d) and U.S. EPA (2023e).
6.2.2.2 Nonquontifioble Benefits of PFOA and PFOS Exposure Reduction
In this section, EPA qualitatively discusses the potential health benefits resulting from reduced
exposure to PFOA and PFOS in drinking water. These nonquantifiable benefits are expected to
be realized as avoided adverse health effects as a result of the proposed NPDWR, in addition to
the benefits that EPA has quantified. EPA anticipates additional benefits associated with
developmental, cardiovascular, hepatic, immune, endocrine, metabolic, reproductive,
musculoskeletal, and carcinogenic effects beyond those benefits that EPA has quantified. The
evidence for these adverse health effects is briefly summarized below.
6.2.2.2.1 Developmental Effects
In addition to the infant birth weight impacts that EPA has quantified (see Section 6.4), small for
gestational age (SGA) is a developmental health outcome of interest when studying potential
effects of PFOA/PFOS exposure, because infants who are SGA face increased health risks
during pregnancy and delivery as well as post-delivery (Osuchukwu et al., 2022). Epidemiology
Proposed PFAS Rule Economic Analysis
6-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
evidence related to PFOA/PFOS exposure was mixed; some studies indicated increased risk of
SGA with PFOA/PFOS exposure, while other studies observed null results (U.S. EPA, 2023e;
U.S. EPA, 2023d). For instance, some studies suggested a potentially positive association
between PFOA exposure and SGA (Govarts et al., 2018; Lauritzen et al., 2017; Y. Wang et al.,
2016; U.S. EPA, 2023e). For PFOS, few patterns were discernible, and overall confidence of an
association between the two factors was low (U.S. EPA, 2023d). Similarly, ATSDR found no
strong associations between PFOA or PFOS exposures and increases in risk of SGA infants
(ATSDR, 2021). Toxicology studies on PFOS exposures in rodents demonstrated relationships
with multiple developmental endpoints including increased offspring mortality, decreased
maternal body weight and body weight change, skeletal and soft tissue effects, and delayed eye-
opening (U.S. EPA, 2023d). For additional details on developmental studies and their individual
outcomes, see Chapter 3.4.1 (Developmental) in U.S. EPA (2023d) and U.S. EPA (2023e).
6.2.2.2.2 Cardiovascular Effects
In addition to the CVD effects that EPA quantified associated with changes in TC and BP from
exposure to PFOA and PFOS (see Section 6.5), available evidence suggests an association
between exposure to PFOA and PFOS and increased LDLC (ATSDR, 2021; U.S. EPA, 2023e;
U.S. EPA, 2023d). High levels of LDLC lead to the buildup of cholesterol in the arteries, which
can raise the risk of heart disease and stroke. Epidemiology studies showed a positive association
between PFOA and PFOS exposure and LDLC levels in children (U.S. EPA, 2023e; U.S. EPA,
2023d). In particular, the evidence suggested positive associations between serum PFOA and
PFOS levels and LDLC levels in adolescents ages 12-18, while positive associations between
serum levels and LDLC levels in younger children were observed only for PFOA (ATSDR,
2021). Studies conducted on PFOS showed evidence of an association between exposure and
LDLC levels in adults. For instance, all five epidemiology studies evaluated in EPA's Toxicity
Assessments and Proposed Maximum Contaminant Level Goals for PFOA and PFOS in
Drinking Water reported positive associations, although the association was only statistically
significant in obese women. Available evidence regarding the impact of PFOA and PFOS
exposure on pregnant women was too limited for EPA to determine an association (ATSDR,
2021; U.S. EPA, 2023e; U.S. EPA, 2023d). Toxicology studies generally reported alterations in
LDLC levels in mice and rats following oral exposure to PFOA (U.S. EPA, 2023e) or PFOS
(U.S. EPA, 2023d). Although the biological significance of the decrease in various serum lipid
levels observed in animal models regardless of species, sex, or exposure paradigm is unclear,
these effects do indicate a disruption in lipid metabolism, which is coherent with effects
observed in humans. For additional details on LDLC studies and their individual outcomes, see
Chapter 3.4.4 (Cardiovascular) in U.S. EPA (2023d) and U.S. EPA (2023e).
6.2.2.2.3 Hepatic Effects
Several biomarkers can be used clinically to diagnose liver diseases, including the alanine
aminotransferase (ALT). High levels of serum ALT may indicate liver damage. Epidemiology
data provides consistent evidence of a positive association between PFOS/PFOA exposure and
ALT levels in adults (ATSDR, 2021; U.S. EPA, 2023d; U.S. EPA, 2023e). Studies of adults
showed consistent evidence of a positive association between PFOA exposure and elevated ALT
levels at both high exposure levels and exposure levels typical of the general population (U.S.
EPA, 2023e). There is also consistent epidemiology evidence of associations between PFOS and
elevated ALT levels, although the associations observed were not large in magnitude. Study
Proposed PFAS Rule Economic Analysis
6-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
results showed inconsistent evidence on whether the observed changes led to changes in specific
liver disease (U.S. EPA, 2023d).
Associations between PFOS/PFOA exposure and ALT levels in children were less consistent
than in adults (U.S. EPA, 2023d; U.S. EPA, 2023e), and PFOA toxicology studies showed
increases in ALT and other liver enzymes across multiple species, sexes, and exposure
paradigms (U.S. EPA, 2023e). Toxicology studies on the impact of PFOS exposure also reported
increases in ALT and other liver enzyme levels in rodents, though these increases were modest
(U.S. EPA, 2023d). For additional details on the ALT studies and their individual outcomes, see
Chapter 3.4.2 (Hepatic) in U.S. EPA (2023d) and U.S. EPA (2023e).
6.2.2.2.4 Immune Effects
Proper antibody response helps maintain the immune system by recognizing and responding to
antigens. Evidence indicates a relationship between PFOA exposure and immunosuppression;
epidemiology studies showed suppression of at least one measure of the antibody response for
tetanus and diphtheria among people with higher prenatal, childhood, and adult serum
concentrations of PFOA (U.S. EPA, 2023e). It is less clear whether PFOA exposure impacts
antibody response to vaccinations other than tetanus and diphtheria (ATSDR, 2021; U.S. EPA,
2023e). Epidemiology evidence suggests that children with preexisting immunological
conditions are particularly susceptible to immunosuppression associated with PFOA exposure
(U.S. EPA, 2023e). Available studies supported an association between PFOS exposure and
immunosuppression in children, where increased PFOS serum levels were associated with
decreased antibody production (U.S. EPA, 2023d). However, an association between PFOS and
immunosuppression has not been observed to date in adults (U.S. EPA, 2023d).31 Other potential
associations with PFOS exposure with a high degree of uncertainty included asthma and
infectious diseases (e.g., the common cold, lower respiratory tract infections, pneumonia,
bronchitis, ear infections; U.S. EPA, 2023d). Toxicology evidence suggested that PFOA and
PFOS exposure results in effects similarly indicating immune suppression, such as reduced
response of immune cells (e.g., natural killer cell activity and immunoglobulin production) (U.S.
EPA, 2023d; U.S. EPA, 2023e). For additional details on antibody studies and their individual
outcomes, see Chapter 3.4.3 (Immune) in U.S. EPA (2023d) and U.S. EPA (2023e).
Because evidence indicates that PFOA and PFOS exposure results in immune effects, EPA
expects those impacts to potentially impact immune response to other diseases. For instance, the
coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome
coronavirus 2 (SARS-CoV-2) rapidly evolved into a global pandemic after its first report in
Wuhan, China, in December 2019. A few recent studies have considered the association between
PFOA and PFOS exposure and COVID-19 infection, severity, or mortality (Catelan et al., 2021;
Grandjean et al., 2020; Ji et al., 2021).
A case-control study in China (Ji et al., 2021) showed increased risks for COVID-19 infection
with high urinary PFOS, PFOA, and total PFASs after adjusting for potential confounding
factors including age, gender, diabetes, cardiovascular disease, and urine albumin-to-creatinine
ratio. Adjusted odds ratios (ORs) were 1.94 (95% CI: 1.39, 2.96) for PFOS and 2.73 (1.71, 4.55)
for PFOA. Using metabolome-wide association analysis, Ji et al. (2021) found that PFOA and
31 This may be due to the lack of high-quality data at present.
Proposed PFAS Rule Economic Analysis
6-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
PFOS exposure in COVID-19 patients was associated with metabolic disturbances in
biochemical pathways involved in mitochondria stress signaling and the regulation of immune
function, including fatty acid oxidation, tricarboxylic acid cycle, eicosanoid, and kynurenine
pathways. One cross-sectional study in Denmark (Grandjean et al., 2020) observed no
association between PFOA or PFOS concentrations and severity of COVID-19 development.32 In
a spatial ecological analysis, Catelan et al. (2021) showed higher mortality risk for COVID-19 in
a population heavily exposed to PFAS (including PFOA, PFOS, PFHxS, PFBS, PFBA, PFPeA,
PFHxA, and PFHpA) via drinking water in Veneto, Italy.
Although these studies provide a suggestion of possible associations, the body of evidence does
not permit any conclusions about the relationship between COVID-19 and exposures to PFAS.
6.2.2.2.5 Endocrine Effects
Elevated thyroid hormone levels can accelerate metabolism and cause irregular heartbeat; low
levels of thyroid hormone can cause neurodevelopmental effects, tiredness, weight gain, and
susceptibility to the common cold. There is suggestive evidence of a positive association
between PFOA/PFOS exposure and thyroid hormone disruption (ATSDR, 2021; U.S. EPA,
2023d; U.S. EPA, 2023e). Epidemiology studies reported inconsistent evidence regarding
associations between PFOA and PFOS exposure and general endocrine outcomes, such as
thyroid disease, hypothyroidism, and hypothyroxinemia (U.S. EPA, 2023d; U.S. EPA, 2023e).
However, studies reported suggestive evidence of positive associations for thyroid stimulating
hormone (TSH) in adults, and the thyroid hormone thyroxine (T4) in children (U.S. EPA, 2023d;
U.S. EPA, 2023e). Toxicology studies indicated that PFOA and PFOS exposure leads to
decreases in thyroid hormone levels33 and adverse effects to the endocrine system (ATSDR,
2021; U.S. EPA, 2023e; U.S. EPA, 2023d). Despite uncertainty around the applicability of
animal studies in this area, changes in thyroid hormone levels in animals did indicate PFOS and
PFOA toxicity relevant to humans (U.S. EPA, 2023e; U.S. EPA, 2023d). For additional details
on endocrine effects studies and their individual outcomes, see Chapter C.2 (Endocrine) in U.S.
EPA (2023b) and U.S. EPA (2023c).
6.2.2.2.6 Metabolic Effects
Leptin is a hormone that balances hunger, and high leptin levels are associated with obesity,
overeating, and inflammation (e.g., of adipose tissue, the hypothalamus, blood vessels, and other
areas). Evidence suggests a direct association between PFOA exposure and leptin levels in the
general adult population (ATSDR, 2021; U.S. EPA, 2023e). Based on a review of 69 human
epidemiology studies, evidence of associations between PFOS and metabolic outcomes appears
inconsistent, but in some studies, suggestive evidence was observed between PFOS exposure and
leptin levels (U.S. EPA, 2023d). Studies examining newborn leptin levels did not find
associations with maternal PFOA levels (ATSDR, 2021). Maternal PFOS levels were also not
associated with alterations in leptin levels (ATSDR, 2021). For additional details on metabolic
effect studies and their individual outcomes, see Chapter C.3 (Metabolic/Systemic) in U.S. EPA
(2023b) and U.S. EPA (2023c).
32 Note that the authors found that PFBA exposure was associated with increasing severity of COVID-19.
33 Decreased thyroid hormone levels are associated with effects such as changes in thyroid and adrenal gland weight, hormone
fluctuations, and organ histopathology (ATSDR, 2021; U.S. EPA, 2023d).
Proposed PFAS Rule Economic Analysis
6-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.2.2.2.7 Reproductive Effects
Studies of the reproductive effects from PFOA/PFOS exposure have focused on associations
between exposure to these contaminants and increased risk of gestational hypertension and
preeclampsia in pregnant women (ATSDR, 2021; U.S. EPA, 2023d; U.S. EPA, 2023e).
Gestational hypertension (high BP during pregnancy) can lead to fetal problems such as poor
growth and stillbirth. Preeclampsia—instances of gestational hypertension where the mother also
has increased levels of protein in her urine—can similarly lead to potentially fatal fetal problems
and maternal complications. The epidemiology evidence yields mixed (positive and non-
significant) associations, with some suggestive evidence supporting positive associations
between PFOA/PFOS exposure and both preeclampsia and gestational hypertension (ATSDR,
2021; U.S. EPA, 2023d; U.S. EPA, 2023e). For additional details on reproductive effects studies
and their individual outcomes, see Chapter C.l (Reproductive) in U.S. EPA (2023b) and U.S.
EPA (2023c).
6.2.2.2.8 Musculoskeletal Effects
Adverse musculoskeletal effects such as osteoarthritis and decreased bone mineral density
impact bone integrity and cause bones to become brittle and more prone to fracture. There is
limited evidence from studies pointing to effects of PFOS on skeletal size (height), lean body
mass, and osteoarthritis (U.S. EPA, 2023d). Epidemiology evidence suggested that PFOA
exposure may be linked to decreased bone mineral density, bone mineral density relative to bone
area, height in adolescence, osteoporosis, and osteoarthritis (ATSDR, 2021; U.S. EPA, 2023e).
Evidence from four PFOS studies suggested that PFOS exposure has a harmful effect on bone
health, particularly measures of bone mineral density, with more statistically significant effects
occurring among females (U.S. EPA, 2023d). Some studies found that PFOA/PFOS exposure
was linked to osteoarthritis, in particular among women under 50 years of age (ATSDR, 2021).
However, other reviews reported mixed findings on the effects of PFOS exposure including
decreased risk of osteoarthritis, increased risk for some demographic subgroups, or no
association (ATSDR, 2021). For additional details on musculoskeletal effects studies and their
individual outcomes, see Chapter C.8 (Musculoskeletal) in U.S. EPA (2023b) and U.S. EPA
(2023c).
6.2.2.2.9 Cancer Effects
In EPA's Toxicity Assessment and Proposed Maximum Contaminant Level Goal for PFOA in
Drinking Water the Agency evaluates the evidence for carcinogenicity of PFOA that has been
documented in both epidemiological and animal toxicity studies. The evidence in
epidemiological studies is primarily based on the incidence of kidney and testicular cancer, as
well as potential incidence of breast cancer in genetically susceptible subpopulations. Other
cancer types have been observed in humans, although the evidence for these is generally limited
to low confidence studies. The evidence of carcinogenicity in animal models is provided in three
chronic oral animal bioassays in Sprague-Dawley rats which identified neoplastic lesions of the
liver, pancreas, and testes (U.S. EPA, 2023e). EPA determined that PFOA is Likely to Be
Carcinogenic to Humans, as "the evidence is adequate to demonstrate carcinogenic potential to
humans but does not reach the weight of evidence for the descriptor Carcinogenic to Humans
This determination is based on the evidence of kidney and testicular cancer in humans and LCTs,
PACTs, and hepatocellular adenomas in rats (U.S. EPA, 2023e). EPA's benefits analysis for
avoided RCC cases from reduced PFOA exposure is detailed in Section 6.6.
Proposed PFAS Rule Economic Analysis
6-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
In EPA's Toxicity Assessment and Proposed Maximum Contaminant Level Goal for PFOS in
Drinking Water the Agency evaluates the evidence for carcinogenicity of PFOS and concluded
that several epidemiological studies and a single chronic cancer bioassay comprise the evidence
database for the carcinogenicity of PFOS (U.S. EPA, 2023d). The available epidemiology studies
report elevated risk of bladder, prostate, kidney, and breast cancers after chronic PFOS exposure.
However, in developing this proposal, EPA did not identify information to quantify the benefits
that reducing PFOS would have on reducing various cancers in humans. The sole animal chronic
cancer bioassay study provides support for multi-site tumorigenesis in male and female rats. EPA
reviewed the weight of the evidence and determined that PFOS is Likely to Be Carcinogenic to
Humans, as "the evidence is adequate to demonstrate carcinogenic potential to humans but does
not reach the weight of evidence for the descriptor Carcinogenic to Humans
EPA anticipates there are additional nonquantifiable benefits related to potential testicular,
bladder, prostate, kidney, and breast carcinogenic effects summarized above. For additional
details on cancer studies and their individual outcomes, see Chapter 3.5 (Cancer) in U.S. EPA
(2023e) and U.S. EPA (2023d).
6.2.3 Summary of Health Information Considered in the
Economic Analysis
After assessing available health and economic information, EPA was unable to quantify the
benefits of avoided health effects discussed above. The Agency prioritized health endpoints with
the strongest weight of evidence conclusions for this assessment and readily available data for
monetization, namely cardiovascular effects, developmental effects, and carcinogenic effects.
Several other health endpoints that had indicative evidence of associations with exposure to
PFOA and PFOS have not been selected for the economic analysis:
• While immune effects had indicative evidence of associations with exposure to PFOA
and PFOS, EPA did not identify the necessary information to connect the measured
biomarker responses (i.e., decrease in antibodies) to a clinical effect that could be valued
in the economic analysis;
• Evidence indicates associations between PFOA and PFOS exposure and hepatic effects,
such as increases in ALT. However, EPA is not able to model this health endpoint
because ALT is a non-specific biomarker.34 Similar challenges with non-specificity of the
biomarkers representing metabolic effects (i.e., leptin) and musculoskeletal effects (i.e.,
bone density) prevented economic analysis of these endpoints;
• There is indicative evidence of association with exposure to PFOA for testicular cancer;
however, the available slope factor implied small changes in the risk of this endpoint.
Furthermore, testicular cancer is rarely fatal which implies low expected economic value
34 Elevated ALT levels could be one of several contributors to the non-alcoholic fatty liver disease. Additionally, high ALT levels
can be associated with alcohol consumption, heart failure, hepatitis (A, B, and C), medication use (e.g., Tylenol and statins), and
obesity (Mayo Clinic, 2022) and this wide range of associations makes it difficult to model economic benefits of non-specific
ALT level changes in response to reduced exposures.
Proposed PFAS Rule Economic Analysis
6-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
of reducing this risk because Value of Statistical Life is the driver of economic benefits
evaluated in the EA;
• Finally, other health endpoints, such as small for gestational age and LDLC effects, were
not modeled in the EA because they overlap with effects that EPA did model. For
example, SGA infants are often born at low birth weight or receive similar care to infants
born at low birth weight. LDLC is a component of total cholesterol and could not be
modeled separately as EPA used total cholesterol as an input to the ASCVD model to
estimate CVD outcomes.
6.2.4 Nonquantifiable Benefits of PFAS in Proposed Rule and
PFAS Expected to be Co-Removed
EPA also qualitatively summarized the potential health benefits resulting from reduced exposure
to PFAS other than PFOA and PFOS in drinking water. The proposed option and all regulatory
alternatives are expected to result in additional benefits that have not been quantified. The
proposed option will reduce exposure to PFHxS, HFPO-DA, PFNA, and PFBS to below their
respective Health Based Water Concentrations (HBWCs). Benefits from avoided cases of the
adverse health effects discussed below are expected from the proposed rule due to co-occurrence
of these contaminants in source waters containing PFOA and/or PFOS, documented in detail in
the Technical Support Document - Per- and Polyflaoroalkyl Substances (PFAS) Occurrence &
Contaminant Background (U.S. EPA, 2023g). EPA also expects that compliance actions taken
under the proposed rule will remove additional unregulated co-occurring PFAS contaminants
where present because the best available technologies have been demonstrated to co-remove
additional PFAS. Treatment responses implemented to reduce PFOA and PFOS exposure under
the proposed option and Options la-c are likely to remove some amount of additional PFAS
contaminants where they co-occur.
Ion exchange (IX) and granulated activated carbon (GAC) are effective at removing PFAS; there
is generally a linear relationship between PFAS chain length and removal efficiency, shifted by
functional group (McCleaf et al., 2017; Sorengard, 2020). Perfluoroalkyl sulfonates (PFSAs),
such as PFOS, are removed with greater efficiency than corresponding perfluoroalkyl
carboxylates (PFCAs), such as PFOA, of the same carbon backbone length (Appleman et al.,
2014; Du, 2014; Eschauzier et al., 2012; Ochoa-Herrera et al., 2008; Zaggia et al., 2016).
Generally, for a given water type and concentration, PFSAs are removed approximately as
effectively as PFCAs, which have two additional fully perfluorinated carbons in the carbon
backbone. For example, PFHxS (i.e., sulfonic acid with a six-carbon backbone) is removed
approximately as well as PFOA (i.e., carboxylic acid with an eight-carbon backbone) and
PFHxA (i.e., carboxylic acid with a six-carbon backbone) is removed approximately as well as
PFBS (i.e., sulfonic acid with a four-carbon backbone). Further, PFAS compounds with longer
carbon chains display lower percentage decreases in average removal efficiency over time
(McCleaf etal., 2017).
In cases where the six PFAS included in the proposed rule occur at concentrations above their
respective regulatory standards, there is also an increased probability of co-occurrence of
additional unregulated PFAS. Further, as the same technologies also remove other long-chain
and higher carbon/higher molecular weight PFAS, EPA expects that treatment will provide
Proposed PFAS Rule Economic Analysis
6-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
additional public health protection and benefits due to co-removal of unregulated PFAS that may
have adverse health effects. While EPA has not quantified these additional benefits, the Agency
believes these important co-removal benefits further enhance public health protection.
EPA identified a wide range of potential health effects associated with exposure to PFAS
compounds other than PFOA and PFOS using documents that summarize the recent literature on
exposure and associated health impacts: ATSDR's Toxicology Profile for Perfluoroalkyls
(ATSDR, 2021); EPA's summary ofHFPO-DA toxicity (U.S. EPA, 2021c); publicly available
IRIS assessment for PFBA and draft IRIS assessments for PFDA, and PFHxA (; U.S. EPA,
2022f; U.S. EPA, 2022g); a human health assessment for PFBS (U.S. EPA, 2021d); and the
recent National Academies of Sciences, Engineering, and Medicine Guidance on PFAS
Exposure, Testing, and Clinical Follow-up (NASEM, 2022). Note that the determinations of
associations between PFAS compounds and associated health effects are based on information
available as of May 2022, and that the finalization of the IRIS assessments may result in slight
changes to the discussion of evidence.
Developmental effects: Toxicology and/or epidemiology studies observed evidence of
associations between birth weight and/or other developmental effects and exposure to PFBA,
PFDA, PFHxS, HFPO-DA, PFNA, and PFBS. Specifically, data from toxicology studies support
this association for PFBS, PFBA, and HFPO-DA, while both toxicology and epidemiology
studies support this association for PFDA and PFNA (ATSDR, 2021; U.S. EPA, 2021c; U.S.
EPA, 2022e; U.S. EPA, 2022f) although some mixed results have been found for birth
outcomes, particularly birth weight. In general, epidemiological studies did not find associations
between perfluoroalkyl exposure and adverse pregnancy outcomes (miscarriage, preterm birth, or
gestational age) for PFHxS, PFNA, PFDA, or PFUnA (ATSDR, 2021; NASEM, 2022).
Cardiovascular effects: Epidemiology and/or toxicology studies observed evidence of
associations between PFNA and PFDA exposures and total cholesterol, LDLC, and HDLC.
Evidence for associations between PFNA exposure and serum lipids levels in epidemiology
studies was mixed; associations have been observed between serum PFNA levels and total
cholesterol in general populations of adults but not in pregnant women, and evidence in children
is inconsistent (ATSDR, 2021). Most epidemiology studies did not observe associations between
PFNA and LDLC or HDLC. Similarly inconsistent evidence was observed for PFDA (ATSDR,
2021). Other PFAS for which lipid outcomes were examined in toxicology or epidemiology
studies observed limited to no evidence of associations. Studies have examined possible
associations between various PFAS and blood pressure in humans or heart histopathology in
animals. However, studies did not find suggestive or likely evidence for any PFAS in this
summary except for PFOS.
Hepatic effects: Toxicology studies reported associations between exposure to PFAS
compounds (PFBA, PFDA, PFHxA, PFHxS, HFPO-DA, and PFBS) and hepatotoxicity
following inhalation, oral, and dermal exposure in animals. The results of these studies provide
strong evidence that the liver is a sensitive target of PFHxS, PFNA, PFDA, PFUnA, PFBS,
PFBA, PFDoDA, and PFHxA toxicity. Observed effects in rodents include increases in liver
weight, hepatocellular hypertrophy, hyperplasia, and necrosis (ATSDR, 2021; U.S. EPA, 2021c;
U.S. EPA, 2022e; U.S. EPA, 2022f; U.S. EPA, 2022g). Increases in serum enzymes (such as
Proposed PFAS Rule Economic Analysis
6-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
ALT) and decreases in serum bilirubin were observed in one epidemiologic study of PFHxS, and
mixed effects were observed in epidemiologic studies for PFNA (ATSDR, 2021).
Immune effects: Epidemiology studies have reported evidence of associations between PFDA
and PFHxS exposure and antibody response to tetanus or diphtheria. There is also some limited
evidence for decreased antibody response for PFNA, PFUnA, and PFDoDA, although many of
the studies did not find associations for these compounds. There is limited evidence for
associations between PFHxS, PFNA, PFDA, PFBS, and PFDoDA and increased risk of asthma
due to the small number of studies evaluating the outcome and/or conflicting study results. The
small number of studies investigating immunotoxicity in humans following exposure to PFHpA
and PFHxA did not find associations (ATSDR, 2021; U.S. EPA, 2022g, NASEM, 2022).
Toxicology studies have reported evidence of associations between HFPO-DA exposure and
various immune-related endpoints in animals (ATSDR, 2021; U.S. EPA, 2021c). No laboratory
animal studies were identified for PFUnA, PFHpA, PFDoDA, or FOSA. A small number of
toxicology studies evaluated the immunotoxicity of other perfluoroalkyls and most did not
evaluate immune function. No alterations in spleen or thymus organ weights or morphology were
observed in studies on PFHxS, PFBA, and PFDA. A study on PFNA found decreases in spleen
and thymus weights and alterations in splenic lymphocyte phenotypes (ATSDR, 2021).
COVID-19: A cross-sectional study in Denmark (Grandjean et al., 2020) showed that PFBA
exposure was associated with increasing severity of COVID-19, with an OR of 1.77 [95%
Confidence Interval (CI): 1.09, 2.87] after adjustment for age, sex, sampling site, and interval
between blood sampling and diagnosis. However, the study design does not allow for causal
determinations. A case-control study showed increased risk of COVID-19 infection with high
urinary PFAS (including PFOA, PFOS, PFHxA, PFHpA, PFHxS, PFNA, PFBS, PFDA, PFUnA,
PFDoA, PFTrDA, PFTeDA) levels (Ji et al., 2021). Adjusted odds ratios were 1.94 (95% CI:
1.39, 2.96) for PFOS, 2.73 (95% CI: 1.71, 4.55) for PFOA, and 2.82 (95% CI: 1.97-3.51) for
sum PFAS, while other PFAS were not significantly associated with COVID-19 susceptibility
after adjusting for confounders. In a spatial ecological analysis, Catelan et al. (2021) showed
higher mortality risk for COVID-19 in a population heavily exposed to PFAS (including PFOA,
PFOS, PFHxS, PFBS, PFBA, PFPeA, PFHxA, and PFHpA) via drinking water. Overall, results
suggested a general immunosuppressive effect of PFAS and/or increased COVID-19 respiratory
toxicity due to a concentration of PFBA in the lungs; however, the study design precludes causal
determinations. Although these studies provide a suggestion of possible associations, the body of
evidence does not permit any conclusions about the relationship between COVID-19 infection,
severity, or mortality, and exposures to PFAS.
Endocrine effects: Epidemiology studies have observed associations between serum PFHxS,
PFNA, PFDA, and PFUnA and thyroid stimulating hormone (TSH), triiodothyronine (T3), or
thyroxine (T4) levels or thyroid disease, however the results are not consistent across studies and
a larger number of studies have not found associations (ATSDR, 2021; NASEM, 2022).
Toxicology studies have reported associations between thyroid hormone disruption in animals
and exposure to PFBA, PFHxA, and PFBS (U.S. EPA, 2021d; U.S. EPA, 2022e; U.S. EPA,
2022g).
Metabolic effects: Epidemiology and toxicology studies have examined possible associations
between various PFAS and metabolic effects, including leptin, body weight, or body fat in
Proposed PFAS Rule Economic Analysis
6-24
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
humans or animals (ATSDR, 2021). However, evidence of associations was not suggestive or
likely for any PFAS in this summary except for PFOA. Evidence did not include changes such as
body weight gain, pup body weight, or other developmentally focused weight outcomes
(ATSDR, 2021; NASEM, 2022).
Renal effects: A small number of epidemiology studies with inconsistent results evaluated
possible associations between PFHxS, PFNA, PFDA, PFBS, PFDoDA, or PFHxA and renal
functions (including estimated glomerular filtration rate and increases in uric acid levels)
(ATSDR, 2021; NASEM 2022). Toxicology studies have not observed impaired renal function
or morphological damage following exposure to PFHxS, PFDA, PFUnA, PFBS, PFBA,
PFDoDA, or PFHxA. Associations with kidney weight in animals were observed for PFBS and
HFPO-DA (ATSDR, 2021; U.S. EPA, 2021c; U.S. EPA, 202Id).
Reproductive effects: A small number of epidemiology studies with inconsistent results
evaluated possible associations between reproductive hormone levels and PFHxS, PFNA,
PFUnA, PFDoDA, or PFHxA. Some associations between PFAS (PFHxS, PFNA, PFDA)
exposures and sperm parameters have been observed, but often only one sperm parameter was
altered. While there is suggestive evidence of an association between PFHxS or PFNA exposure
and an increased risk of early menopause, this may be due to reverse causation since an earlier
onset of menopause would result in a decrease in the removal of PFAS in menstrual blood.
Epidemiological studies provide mixed evidence of impaired fertility (increased risks of longer
time to pregnancy and infertility), with some evidence for PFHxS PFNA, PFHpA, and PFBS but
the results are inconsistent across studies or were only based on one study (ATSDR, 2021).
Toxicology studies have evaluated the potential histological alterations in reproductive tissues,
alterations in reproductive hormones, and impaired reproductive functions. No effect on fertility
was observed for PFBS, PFHxS or PFDoDA, and no histological alterations were observed for
PFBS, PFHxS, and PFBA. One study found alterations in sperm parameters and decreases in
fertility in mice exposed to PFNA, and one study for PFDoDA observed ultrastructural
alterations in the testes (ATSDR, 2021).
Musculoskeletal effects: Epidemiology studies observed evidence of associations between
PFNA and PFHxS and musculoskeletal effects including osteoarthritis and bone mineral density,
but data are limited to two studies (ATSDR, 2021). Epidemiology studies reported limited to no
evidence of associations between exposure to PFDA and musculoskeletal effects. Toxicology
studies reported no morphological alterations in bone or skeletal muscle in animals exposed to
PFBA, PFHxA, PFHxS, or PFBS (ATSDR, 2021).
Hematological effects: A single epidemiologic study reported on blood counts in pregnant
women exposed to PFHxA (U.S. EPA, 2023d). Epidemiological data were not identified for the
other PFAS (ATSDR, 2021). A limited number of toxicology studies observed alterations in
hematological indices following exposure to higher doses of PFHxS, PFDA, PFUnA, PFBS,
PFBA, PFDoDA, or PFHxA (ATSDR, 2021). Toxicology studies observed evidence of
association between HFPO-DA exposure and hematological effects, including decreases in RBC
number, hemoglobin, and percentage of RBCs in the blood (U.S. EPA, 2021c).
Other non-cancer effects: A limited number of epidemiology and toxicology studies have
examined possible associations between various PFAS and dermal, ocular, and other non-cancer
effects. However, the evidence does not support associations for any PFAS compound in this
Proposed PFAS Rule Economic Analysis
6-25
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
summary except for PFOA and PFOS (ATSDR, 2021; U.S. EPA, 2021d; U.S. EPA, 2022e; U.S.
EPA, 2022f; U.S. EPA, 2022g).
Cancer effects: A small number of epidemiology studies reported limited associations between
multiple PFAS and cancer effects. No consistent associations were observed for breast cancer
risk for PFHxS, PFNA, PFHpA, or PFDoDA; increased breast cancer risks were observed for
PFDA and FOSA, but this was based on a single study (Bonefeld-Jorgensen et al., 2014). No
associations between PFHxS, PFNA, PFDA, or PFUnA and prostate cancer risk were observed.
However, among men with a first-degree relative with prostate cancer, associations were
observed for PFHxS, PFDA, and PFUnA, but not for PFNA (ATSDR, 2021). Epidemiological
studies examining potential cancer effects were not identified for PFBS, PFBA, or PFHxA
(ATSDR, 2021; U.S. EPA, 2022e). Aside from a study that suggested an increased incidence of
liver tumors in rats exposed to high doses of HFPO-DA, toxicology studies reported no evidence
of associations between exposure to other PFAS (i.e., PFDA and PFHxA) and risk of cancer
(ATSDR, 2021; U.S. EPA, 2021c).
6.2.5 Sensitive Populations
SDWA section 1412(b)(3)(C) establishes requirements for EPA to develop a health risk
reduction and cost analysis (HRRCA) that presents both quantifiable and nonquantifiable
benefits and costs likely to occur as a result of compliance with the NPDWR. In developing this
HRRCA, EPA considered adverse health effects to sensitive populations and subpopulations.
Adverse health effects of PFAS such as cancer, developmental, hepatic, immune, and serum lipid
effects (see Sections 6.2.2 and 6.2.3) have been observed in the general population, including
women of reproductive age. Effects have been observed in vulnerable populations of groups who
have relatively high exposures, for example workers and their families who worked at and/or
lived near facilities that used PFOA (such as the C8 Health Project35 populations). However, data
for the elucidation of differential susceptibility dependent on life stage (e.g., developing
embryo/fetus, or pregnant women) are very limited or not available. Children are frequently
more vulnerable to contaminants than the average adult because of the differences in their
behaviors and biology. These differences can result in greater exposure and/or unique windows
of developmental susceptibility during the prenatal and postnatal periods for both the pregnant
mother and the developing fetus.
In determining MCLGs, EPA considers the adverse health risks to infants/children, individuals
who are immunologically compromised, and the elderly to ensure the most sensitive population
groups are protected. In conducting risk analyses and assessments, other agencies and
organizations consider sensitive subpopulations to be pregnant women, infants/children,
individuals who are immunologically compromised, and the elderly (ATSDR, 2021; CalEPA,
2021; Minnesota Pollution Control Agency, 2021). CalEPA (2021) and the Minnesota Pollution
Control Agency (2021) also identify the timing of exposure to PFAS to be critical in the
development of adverse health effects. There is evidence of associations with birth weight effects
and exposure to PFDA, PFHxS, PFNA, PFOA, PFOS, or PFUnA (see Sections 6.2.2 and 6.2.3).
There is some sex-specific variation in the toxicokinetics of PFOA in hamsters, rabbits, and rats,
35 The C8 Health Project studied over 60,000 individuals who had lived, worked, or attended school for more than one year in
one of six water districts contaminated by PFOA between 1950 and 2004 (Frisbee et al., 2010).
Proposed PFAS Rule Economic Analysis
6-26
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
with females excreting PFOA faster than males (U.S. EPA, 2016c). Lactation and menstruation
were noted as important excretory routes in females; however, further research is needed to
determine whether those differences in toxicokinetics are relevant to toxicity of PFOA in humans
(U.S. EPA, 2016c).
Overall, given that evidence of exposure and adverse health effects of PFAS is observed in the
general population, not all potentially sensitive populations are quantified in developing this
HRRCA. However, the modeled endpoints, including birth weight (Section 6.4), CVD (Section
6.5), and renal cell carcinoma (Section 6.6), are prevalent in sensitive populations (i.e., infants
and the elderly).
6.2.6 Co-Removal of Additional Contaminants
Additional co-removal benefits can occur with the advanced treatment options for PFAS
removal. Advanced treatment technologies including granular activated carbon (GAC), ion
exchange (IX), as well as high-pressure membranes such as nanofiltration (NF) and reverse
osmosis (RO) can remove many contaminants in addition to those specifically targeted by the
Proposed PFAS Rule, including other contaminants that EPA may regulate in the future
(Chowdhury et al., 2013; de Abreu Domingos et al., 2018; McNamara et al., 2018; Pramanik et
al., 2015; Yu et al., 2012). For example, membrane technology (depending on pore size), can be
used to lower DBP formation by the removal of organic carbon, and can also remove many
microbial contaminants (e.g., bacteria and protozoans) of public health concern (Park et al.,
2019).
Organic matter can also be removed by IX and GAC (Crittenden et al., 1993; Kim et al., 1997;
Yapsakli et al., 2010; Dickenson et al., 2016; Yuan et al., 2022). Removing TOC, which
functions as a DBP precursor, may also help address DBP issues, including regulated and
nonregulated DBPs. Epidemiological studies have shown that increased exposure to chlorinated
DBPs is associated with higher risk of bladder cancer and other adverse health outcomes (Cantor
et al., 1998; Freeman et al., 2017). Weisman et al. (2022) found that approximately 8,000 of the
79,000 annual bladder cancer cases in the U.S. were potentially attributable to chlorinated DBPs
in drinking water systems.
TOC removal also lowers disinfectant demand and could lower disinfectant dose requirements
(Hooper et al., 2002). Membrane technology, IX, and GAC also lower nutrient availability for
bacterial growth, produce a more biologically stable finished water, and facilitate management of
water quality in the distribution system. Lower organic matter concentration is also associated
with lower assimilable organic carbon (AOC) and nutrient availability for biofilm growth,
helping maintain disinfectant residual in the distribution system, and reduce microbial risk (U.S.
EPA, 2005b).
A major concern for drinking water systems is biofilm control in reducing microbial risk. One
opportunistic pathogen of concern is Legionella, which can grow and multiply in amoeba that
live in biofilms and sediments (National Academies of Sciences, 2020). Certain conditions in the
distribution and plumbing systems can also support its proliferation, including low disinfectant
residual (U.S. EPA, 2016i; LeChevallier, 2020). Legionella exposure can lead to legionellosis,
Pontiac fever, or a form of pneumonia called Legionnaires' disease (National Academies of
Sciences, 2020). Collier et al. (2021) estimated that in 2014 there were 11,000 cases of
Proposed PFAS Rule Economic Analysis
6-27
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Legionnaires' disease due to waterborne exposure in the U.S., with an estimated one in 10 cases
leading to death.
Since membrane technology and GAC also remove SOCs, these advanced treatment options
provide additional protection from exposure to chemicals associated with accidental spills or
environmental runoff. EPA has previously used the term SOC to include volatile organic
carbons, herbicides, pesticides, and other anthropogenic organic compounds (U.S. EPA, 1998d).
One example of a volatile organic carbon that can be co-removed by GAC is dichloromethane,
which has been linked to liver, neurological, and blood cell damage in addition to various
cancers (U.S. EPA, 2014). EPA also identified alachlor as a herbicide that can be removed by
GAC and has been linked to liver, kidneys, and spleen damage (U.S. EPA, 1998a). Another SOC
example that can be removed by GAC treatment is atrazine, a pesticide that targets the endocrine
system and has been associated with adverse developmental reproductive effects (U.S. EPA,
2007a). Removal of any contaminants that may face current and/or future regulation could result
in additional public health protection and cost savings to a water system. As public water
systems move to advanced treatment, other non-health benefits are also anticipated including
better-tasting and smelling water.
6.3 Blood Serum Concentration Modeling for PFAS
6.3.1 Introduction
The U.S. EPA developed PK models to evaluate blood serum PFOA and PFOS levels in adults
resulting from exposure to PFAS via drinking water. This section discusses the application of the
PFOA and PFOS PK models in the context of the benefits analysis.
6.3.2 Application of PK Models to Benefits Analyses
EPA used baseline and regulatory alternative PFOA/PFOS drinking water concentrations as
inputs to its PK models to estimate blood serum PFOA/PFOS concentrations for adult males and
females. In this analysis, the Agency implemented an earlier version PFOA/PFOS PK model
version in SafeWater MCBC.36 See EPA's Toxicity Assessments and Proposed Maximum
Contaminant Level Goals for PFOA and PFOS in Drinking Water for further information on the
model (U.S. EPA, 2023d; U.S. EPA, 2023e) and https://github.com/USEPA/OW-PFOS-PFOA-
MCLG-support-PK-models. The PK models require total PFOA/PFOS dose in mg/kg of body
weight per day to be provided as an input. EPA multiplied PFOA/PFOS drinking water
concentrations in mg/L by a water intake of 0.013 L/kg of body weight per day based on EPA's
Exposure Factors Handbook (U.S. EPA, 201 lb) in order to compute the PFOA/PFOS dose from
drinking water sources.
To estimate the total daily dose, consistent with the 2016 PFOA and PFOS health advisories
(U.S. EPA, 2016e; U.S. EPA, 2016f) and EPA's Toxicity Assessments and Proposed Maximum
Contaminant Level Goals for PFOA and PFOS in Drinking Water (U.S. EPA, 2023 d; U.S. EPA,
2023e), EPA assumed that the dose from drinking water sources comprises 20 percent of the
total daily PFOA/PFOS dose under the baseline scenario (see Section 6.3.3 for discussion of
30 SafeWater MCBC was programmed for maximal computational efficiency. The implementation is mathematically consistent
with what is described in the SAB documentation and associated R code, however, SafeWater performs a series of pre-
calculations to reduce model runtime.
Proposed PFAS Rule Economic Analysis
6-28
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
contributions from other sources). EPA notes that the assumed baseline percent contribution
from drinking water sources does not affect the estimated changes in serum PFOA/PFOS, which
is the key quantity of interest to the benefits estimation. For the PK model in humans, EPA
selected a "linear" approach in which the rates in the model are all proportional to concentration.
In this type of model, predicted serum concentration is proportional to the dose, with a
proportionality constant that is dependent on time, but not dose. Holding the age, exposure
duration, and other features of a scenario constant, doubling the dose will double the predicted
serum concentration.37 This implies that the change in predicted serum concentration is
dependent only on the change in drinking water dose and independent of the dose from non-
drinking water sources. EPA additionally assumed that non-drinking water exposure is
independent of the drinking water PFOA/PFOS concentration and estimated the total regulatory
alternative dose as the sum of the baseline non-drinking water dose and the regulatory alternative
drinking water dose.38
EPA used the PK models to evaluate the following PWS entry point (EP)-specific exposure
scenarios in male and female subpopulations:
• Lifetime baseline exposure scenario: Lifetime exposure to baseline PFOA/PFOS
drinking-water dose for cohorts of all ages alive at the start of the evaluation period in
2023 and cohorts born after 2023;
• Lifetime regulatory alternative exposure scenario: Lifetime exposure to regulatory
alternative PFOA/PFOS drinking-water dose for cohorts born during or after 2026 (i.e.,
the year of full regulatory alternative implementation);
• Partial lifetime treatment exposure scenario: Exposure to baseline PFOA/PFOS
drinking-water dose until age A-l years and regulatory alternative PFOA/PFOS dose
thereafter for cohorts aged A > 0 years in 2026.
EPA selected the annual midpoint (the value on June 1 of each year) of the PK-modeled serum
PFOA/PFOS concentration time series to represent the annual average serum PFOA/PFOS
concentrations under the baseline and regulatory options. EPA estimated changes in annual
average serum PFOA/PFOS concentrations under the regulatory alternatives by subtracting
baseline cohort-specific serum PFOA/PFOS concentrations from either full or partial lifetime
cohort-specific serum PFOA/PFOS concentrations (as appropriate) under the regulatory
alternatives. EPA applied the PFOA/PFOS blood serum concentration time series estimated
37 Specifically, let C = a ¦ Dt, where C is serum concentration, a is a proportionality constant, and Dt is the total dose. This can
be expanded to C = a ¦ Dt = a ¦ (Ddw + D0 ), where the total dose is the sum of the dose from drinking water, Ddw, and from
other sources, D0. The change in concentration due to a change in dose from drinking water is then AC = a ¦ ADdw + a ¦ AD0 =
a ¦ ADdw, given that the dose from other sources is constant, AD0 = 0.
38 EPA used the fraction of exposure from drinking water under baseline conditions to estimate the total daily dose of
PFOA/PFOS and the exposure from sources other than drinking water (i.e., 80 percent of the total daily dose), which did not
change upon implementation of the treatment scenario. While the total change in exposure is independent of the amount of
exposure from other sources, the relative change in exposure does depend on the relative amount of exposure from non-drinking
water sources. A greater fraction of exposure from drinking water sources will result in a greater relative change in total exposure
upon implementation of the treatment scenario. EPA also notes that, in reality, some portion of the non-drinking water exposure
will be related to drinking water concentration (e.g., water used for cooking). This portion is difficult to estimate, and, depending
on the relationship, there may be a time lag between the decrease in drinking water concentration and the decrease in the non-
drinking water exposure.
Proposed PFAS Rule Economic Analysis
6-29
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
using the PK models to all benefits analyses that considered changes in PFOA/PFOS drinking
water concentrations.
The birth weight analysis focuses only on women of childbearing age defined by the CDC as
those aged 15 to 44 (Ellington et al., 2020) and thus considers only maternal serum PFOA/PFOS
levels. As described above, the PK models provide estimates of changes in serum PFOA/PFOS
levels by PWS EP, age, and sex for each year during the period of analysis (2023 to 2104). The
birth weight analysis requires a single estimate of change in maternal serum levels for each
PFAS compound per year and location to evaluate potential changes in birth weight resulting
from the regulatory alternatives. Therefore, EPA used the race/ethnicity-specific distribution of
populations of women of childbearing age during the period of analysis to estimate average
annual race/ethnicity-specific change in PFOA/PFOS levels at each PWS EP and for each year.
EPA relied on the average age of race/ethnicity-specific women of childbearing age when
determining PFOA/PFOS serum levels to reflect differences in maternal age across these groups.
The population of women of childbearing age per PWS, race/ethnicity, age, and sex are based on
population estimates for women aged 15 to 44 based on county4evel data from the U.S. Census
(U.S. Census Bureau, 2020a; see Appendix B).39
6.3.3 Contributions from Other Sources
The regulatory alternatives considered in this economic analysis are based on potential
reductions in PFOA/PFOS levels in drinking water. However, human exposures to PFOA and
PFOS may also result from sources other than drinking water, including diet, ambient and indoor
air, incidental soil/dust ingestion, consumer products, and others (U.S. EPA, 2023d; U.S. EPA,
2023e). In development of an MCLG for PFOA and PFOS, EPA applies a relative source
contribution (RSC) to provide a margin of safety that ensures that an individual's total exposure
from PFOA or PFOS does not exceed the chronic oral reference dose (RfD) derived for each
contaminant's MCLG. EPA assumes that 20 percent of the exposure equal to the RfD is from
drinking water and that the remaining 80 percent is from other potential sources (U.S. EPA,
2023d; U.S. EPA, 2023e).
Following a systematic review of the PFOA and PFOS source contribution literature, EPA
identified ingestion of food as the dominant source of both PFOA and PFOS exposures (U.S.
EPA, 2023d; U.S. EPA, 2023e). This pathway is particularly dominant due to bioaccumulation
of PFOA and PFOS in food from environmental emissions, large amounts of foods being
consumed, and high gastrointestinal uptake. PFOA and PFOS may be present in food due to
contact with non-stick cookware or grease-proofing agents in food packaging. PFOA and PFOS
have also been shown to bioaccumulate in fish and shellfish. Consumer products, including
certain cosmetics, textiles, and other household goods, are also a source of PFOA and PFOS
exposure. While PFAS have been detected in ambient air globally, concentrations vary widely
depending on location. PFAS have been detected in soils and dust from carpets and upholstered
furniture. Incidental exposures from soils and dust are particularly important exposure routes for
small children, who have a higher level of hand-to-mouth behavior compared to adults. PFAS
39 Comity-level population estimates are linked to PWSs based on the "counties served" field provided by the SDWIS 2021 Q4
database.
Proposed PFAS Rule Economic Analysis 6-30 March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
levels in soils and surface water can also impact PFAS levels found in air particulates, fish, dairy
products, meat/poultry, and produce (ATSDR, 2021; U.S. EPA, 2023d; U.S. EPA, 2023e).
6.4 Developmental Effects
Research indicates that exposure to PFOA and PFOS is linked to developmental effects,
including infant birth weight (Verner et al., 2015; Negri et al., 2017; ATSDR, 2018; Waterfield
et al., 2020; U.S. EPA, 2016e; U.S. EPA, 2016f; U.S. EPA, 2023d; U.S. EPA, 2023e). The route
through which infants are exposed prenatally to PFOA and PFOS is maternal blood serum via
the placenta. Most studies of the association between maternal serum PFOA/PFOS and birth
weight report negative relationships (Verner et al., 2015; Negri et al., 2017; Dzierlenga et al.,
2020).40 This chapter outlines the overall methodology, assumptions, and data used for
estimating changes in birth weight among infants whose mothers were exposed to PFOA and
PFOS in drinking water during or prior to pregnancy.41
EPA also considered the potential benefits from reduced exposure to PFNA that may be realized
as a direct result of the proposed rule. The Agency explored the birth weight impacts of PFNA in
a sensitivity analysis, using a unit PFNA reduction scenario (i.e., 1 ppt change) and Lu et al.
(2020) to estimate PFNA blood serum levels resulting from PFNA exposures in drinking water.
To estimate blood serum PFNA based on its drinking water concentration, EPA used a first-order
single-compartment model whose behavior was previously demonstrated to be consistent with
PFOA pharmacokinetics in humans (Bartell et al., 2010). In addition to the PFOA-birth weight
and PFOS-birth weight effects analyzed in the EA, EPA examined the effect of inclusion of
PFNA-birth weight effects using estimates from two studies (Lenters et al., 2016; Valvi et al.,
2017). EPA found that inclusion of a 1 ppt PFNA reduction could increase annualized birth
weight benefits 5.4-7.7-fold, relative to the scenario that quantifies a 1 ppt reduction in PFOA
and a 1 ppt reduction in PFOS only. The range of estimated PFNA-related increases in benefits is
driven by the exposure-response, with smaller estimates produced using the slope factors from
Lenters et al. (2016), followed by Valvi et al. (2017). EPA notes that the PFNA slope factor
estimates are orders of magnitude larger than the slope factor estimates used to evaluate the
impacts of PFOA/PFOS reductions. EPA also notes that the PFNA slope factor estimates are not
precise, with 95% CIs covering wide ranges that include zero (i.e., serum PFNA slope factor
estimates are not statistically significant at 5% level). Caution should be exercised in making
judgements about the potential magnitude of change in the national benefits estimates based on
the results of these sensitivity analyses, although conclusions about the directionality of these
effects can be inferred. EPA did not include PFNA effects in the national benefits estimates for
the proposed rulemaking because of limitations associated with the UCMR 3 PFNA occurrence
data and the slope factor estimates are less precise. For EPA's PFNA sensitivity analysis, see
Appendix K.
6.4.1 Overview of the Birth Weight Risk Reduction Analysis
Figure 6-1 provides an overview of the approach used to quantify and value the changes in birth
weight-related risks associated with reductions in exposure to PFOA and PFOS via drinking
40 Note that recent evidence indicates that relationships between maternal serum PFOA/PFOS and birth weight may be impacted
by changes in pregnancy hemodynamics (Sagiv et al., 2018; Steenland et al., 2018).
41 The PK model assumes that mothers were exposed to PFOA/PFOS from birth to the year in which pregnancy occurred.
Proposed PFAS Rule Economic Analysis
6-31
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
water. Section 4.4 and Section 6.3 detail the PWS entry point (EP)-specific PFOA/PFOS
drinking water occurrence estimation and modeling of serum PFOA/PFOS concentrations,
respectively. EP-specific time series of the differences between serum PFOA/PFOS
concentrations under baseline and regulatory alternatives are inputs into this analysis. For each
EP, evaluation of the changes in birth weight impacts involves the following key steps:
1. Estimating the changes in birth weight based on modeled changes in serum PFOA/PFOS
levels and exposure-response functions for the effect of serum PFOA/PFOS on birth
weight;
2. Estimating the difference in infant mortality probability between the baseline42 and
regulatory alternatives based on changes in birth weight under the regulatory alternatives
and the association between birth weight and mortality;
3. Identifying the infant population affected by reduced exposure to PFOA/PFOS in
drinking water under the regulatory alternatives;
4. Estimating the changes in the expected number of infant deaths under the regulatory
alternatives based on the difference in infant mortality rates and the population of
surviving infants affected by increases in birth weight due to reduced PFOA/PFOS
exposure; and
5. Estimating the economic value of reducing infant mortality based on the Value of
Statistical Life and infant morbidity based on reductions in medical costs associated with
changes in birth weight for the surviving infants based on the cost of illness.
Section 6.4.2 discusses the exposure-response modeling for birth weight. Section 6.4.3 describes
estimation of birth weight-related mortality and morbidity impacts in the affected population.
Section 6.4.4 discusses EPA's valuation methodology for reductions in birth weight-related
mortality and morbidity. Section 6.4.5 presents the results of the analysis.
42 Based on mortality rates per state and 500 g birth weight increment from the Centers for Disease Control and Prevention
(CDC) from 2012 to 2018.
Proposed PFAS Rule Economic Analysis
March 2023
6-32
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Legend:
BW-infant mortality
relationship
Cost of illness
function per BW
change
Change in BW between
baseline and treatment
scenario
Change in infant
mortality rates between
baseline and treatment
Serum PFOA and PFOS
concentration difference
between baseline and
regulatory alternative
Result of upstrearrr
\ analysis
Data/Inputs
Model
# Analysis step
CDC data3
Valuation endpoint
Affected infant
population
SDWIS
population
served^
Census Bureau
population data
Change in BW-related
mortality
Value of reduced
BW-related infant
mortality
Value of a
statistical life
Total value of changes in BW
Notes:
SDWIS - Safe drinking water information system, CDC - Centers for Disease Control, BW - birth weight, VSL - value
of a statistical life
includes baseline state-level birth rate and average BW (varies by 100-gm BW increment) and infant mortality rate
(varies by 500-gm BW increment) data distributed based on national-level race/ethnicity-specific data.
Baseline infant mortality rates, along with the BW-infant mortality relationship, are used to determine the change in
infant mortality rate between the baseline and policy scenario. Birth rate and average BW data describe the
affected population of infants.
includes both large and small surface water and ground water systems.
cSmall public water system exposures are extrapolated to represent exposure at the stratum and national level
based on ratios of sampled to total populations served at small public water systems.
dMorbidity cases refer to the total affected population minus infant mortality cases under the regulatory
alternatives.
Figure 6-1: Overview of Analysis of Birth Weight-Related Benefits
Proposed PFAS Rule Economic Analysis
6-33
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.4.2 Estimation of Birth Weight Changes Between Baseline and
Regulatory Alternatives
To estimate changes in birth weight resulting from reduced exposure to PFOA and PFOS under
the regulatory alternatives, EPA relied on the estimated time series of changes in serum
PFOA/PFOS concentrations specific to women of childbearing age and serum-birth weight
exposure-response functions provided in recently published meta-analyses. The estimation of the
time series of changes in serum PFOA/PFOS concentrations is explained in Section 6.3.2. EPA
reviewed five recent meta-analyses of PFAS-birth weight relationships in detail. As described in
Table 6-8, two of the analyses used well-documented systematic review and risk of bias (ROB)
procedures to identify relevant studies in the literature (Johnson et al., 2014; Negri et al., 2017).
The three other studies did not document ROB protocols and study quality evaluation criteria,
however, EPA's Office of Science and Technology (EPA/OST) evaluated most of the studies
used in these meta-analyses for study quality (Verner et al., 2015; Dzierlenga et al., 2020;
Steenland et al., 2018). As discussed below, there was extensive overlap in the studies used in
the various meta-analyses. Two of the meta-analyses included exposure-response modeling for
both PFOS and PFOA (Verner et al., 2015; Negri et al., 2017), while one addressed only PFOS
(Dzierlenga et al., 2020) and the remaining two addressed only PFOA (Johnson et al., 2014;
Steenland et al., 2018).
Table 6-8: Summary of Studies Relating PFOA or PFOS to Birth Weight
Author
PFOA
PFOS
Documented ROB
Protocols
Johnson et al. (2014)
X
X
Verner et al. (2015)
X
X
Negri etal. (2017)
X
X
X
Steenland et al. (2018)
X
Dzierlenga et al. (2020)
X
Abbreviations: PFOS - periluorooctane sulfonic acid; PFOA - perfluorooctanoic acid; ROB - risk of bias.
EPA evaluated the applicability of these studies for use in the evaluation of birth weight changes
resulting from reduced PFOS and PFOA exposure based on the following criteria: number of
studies, homogeneity among studies, and sensitivity analyses. Based on these considerations, the
Agency selected results from Steenland et al. (2018) as the birth weight exposure-response
function for PFOA and results from Dzierlenga et al. (2020) as the birth weight exposure-
response function for PFOS.
Proposed PFAS Rule Economic Analysis
6-34
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Steenland et al. (2018) conducted a random effects meta-analysis based on 24 studies. The
authors estimated a slope of-10.5 g birth weight per ng PFOA/mL with significant
heterogeneity (I2 = 63%)43 (p-value for heterogeneity <0.0001). The Agency chose the results
from this study for use in the risk assessment from exposure to PFOA and benefits analysis of
reducing PFOA in drinking water because it is the most recent meta-analysis on PFOA-birth
weight, and it included a large number of studies.
Dzierlenga et al. (2020) conducted a random effects meta-analysis based on 32 results from 29
studies. An EPA reanalysis of this study44 estimated a slope of-3.0 g birth weight per ng
PFOS/mL with significant heterogeneity (I2 = 58%) (p-value for heterogeneity <0.001). The
Agency chose the results from this study for use in the risk assessment from exposure to PFOS
and benefits analysis of reducing PFOS in drinking water because it is the most recent meta-
analysis on PFOS-birth weight and included a large number of the most recent studies. While
sensitivity analyses suggested that results may be sensitive to the timing of blood draw, the
authors observed consistent inverse associations with birth weight among those with blood
measurements in early pregnancy and in later pregnancy.
Changes in serum PFOA and PFOS concentrations are calculated for each PWS EP during each
year in the analysis period. EPA assumes that, given long half-lives of PFOS and PFOA, any
one-time measurement during or near pregnancy is reflective of a critical window and not subject
to considerable error. The mean change in birth weight per increment in long-term PFOA and
PFOS exposure is calculated by multiplying each annual change in PFOA and PFOS serum
concentration (ng/mL serum) by the PFOA and PFOS serum-birth weight exposure-response
slope factors (g birth weight per ng/mL serum) provided in Table 6-9, respectively. The mean
annual change in birth weight attributable to changes in both PFOA and PFOS exposure is the
sum of the annual PFOA- and PFOS-birth weight change estimates. Appendix D provides
additional details on the derivation of the exposure-response functions. Appendix K presents an
analysis of birth weight risk reduction considering slope factors specific to the first trimester.
Table 6-9: Serum Exposure-Birth Weight Response Estimates
Compound
g Birth Weight/ng/mL Serum (95% CI)
PFOA3
-10.5 (-16.7, -4.4)
PFOSb
-3.0 (-4.9,-1.1)
Abbreviations: g - gram.
Notes:
aThe serum-birth weight slope factor for PFOA is based on the main random effects estimate from Negri et al. (2017);
Steenland et al. (2018).
bThe serum-birth weight slope factor for PFOS is based on an EPA reanalysis of Dzierlenga et al. (2020).
4312 represents the proportion of total variance in the estimated model due to inter-study variation.
44 In the original Dzierlenga et al. (2020) estimate, the authors duplicated an estimate from M. H. Chen et al. (2017) in the
pooled estimate. EPA reran the analysis excluding the duplicated estimate.
Proposed PFAS Rule Economic Analysis
6-35
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA places a cap on estimated birth weight changes in excess of 200 g based on existing studies
that found that changes to environmental exposures result in relatively modest birth weight
changes (Windham et al., 2008; Klein et al., 2018; Kamai et al., 2019).45 Modest changes in birth
weight even as a result of large changes in PFOA/PFOS serum concentrations may be due to
potential bias from studies only including live births (Liew et al., 2015). Additionally, the
magnitude of birth weight changes may be correlated with other developmental outcomes such
as preterm birth, gestational duration, fetal loss, birth defects, and developmental delays. As
described in Section 6.2, these developmental outcomes have limited epidemiology and
toxicology evidence showing associations with PFOA/PFOS exposure and due to this
uncertainty, these outcomes were not further assessed.
6.4.3 Estimation of Birth Weight Impacts
LBW is linked to a number of health effects that may be a source of economic burden to society
in the form of medical costs, infant mortality, parental and caregiver costs, labor market
productivity loss, and education costs (Chaikind et al., 1991; J. R. Behrman et al., 2004; R. E.
Behrman et al., 2007; Joyce et al., 2012; Kowlessar et al., 2013; Colaizy et al., 2016; Nicoletti et
al., 2018; Klein et al., 2018). Recent literature also linked LBW to educational attainment and
required remediation to improve student outcomes, childhood disability, and future earnings
(Jelenkovic et al., 2018; Temple et al., 2010; Elder et al., 2020; Hines et al., 2020 Chatterji et al.,
2014; Dobson et al., 2018). EPA's analysis focuses on two categories of birth weight impacts
that are amenable to monetization associated with incremental changes in birth weight: (1)
medical costs associated with changes in infant birth weight and (2) the value of avoiding infant
mortality at various birth weights.
The birth weight literature related to other sources of economic burden to society (e.g., parental
and caregiver costs and productivity losses) is limited in geographic coverage, population size,
and range of birth weights evaluated and therefore cannot be used in the economic analysis of
birth weight effects from exposure to PFOA/PFOS in drinking water (ICF, 2021). The following
sections summarize the relationship between infant mortality and birth weight as well as methods
used to estimate changes in the number of infant deaths and the number of surviving infants
whose birth weight is affected by reduced PFOA/PFOS exposures.
6.4.3.1 Impacts of Birth Weight on Infant Mortality
Infant mortality is defined as the deaths among infants who were delivered alive but passed
before their first birthday. Birth weight is a significant factor in infant survival (Jacob, 2016).
Epidemiology studies in the U.S. have reported relationships between birth weight and mortality.
Most of these studies typically evaluate relationships between infant mortality and birth weight
above or below various birth weight thresholds (e.g., Mclntire et al., 1999; Lau et al., 2013).
However, even small changes in birth weight could result in substantial avoided mortality
benefits.
45 Klein et al. (2018) indicate that birth weight changes in response to reduced environmental exposures are likely to be small and
simulated changes in birth weight up to 100 g. Kamai et al. (2019) found maximum changes in birth weight in response to
reduced exposures to cigarette smoke of 150 g, while Windham et al. (2008) found a maximum decrement in mean birth weight
of 200 g for infants of smokers.
Proposed PFAS Rule Economic Analysis
6-36
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Two studies showed statistically significant relationships between incremental changes in birth
weight and infant mortality: Almond et al. (2005) and Ma et al. (2010). Ma et al. (2010) used
2001 National Center for Health Statistics (NCHS) linked birth/infant death data for singleton
and multiple birth infants among subpopulations defined by sex and race/ethnicity to estimate a
regression model assessing the associations between 14 key birth outcome measures, including
birth weight, and infant mortality. They found notable variation in the relationship between birth
weight and mortality across race/ethnicity subpopulations, with odds ratios for best-fit birth
weight-mortality models ranging from 0.8-1 (per 100 g birth weight change). Almond et al.
(2005) used 1989-1991 NCHS linked birth/infant death data for multiple birth infants to analyze
relationships between birth weight and infant mortality within birth weight increment ranges. For
their preferred model, they reported coefficients in deaths per 1,000 births per 1 g increase in
birth weight that range from -0.420 to -0.002. However, the data used in these studies (Almond
et al., 2005 and Ma et al., 2010) are outdated (1989-1991 and 2001, respectively). Given the
significant decline in infant mortality over the last 30 years (ICF, 2020) and other maternal and
birth characteristics that are likely to influence infant mortality (e.g., average maternal age and
rates of maternal smoking), the birth weight-mortality relationship estimates from Almond et al.
(2005) and Ma et al. (2010) are likely to overestimate the benefits of birth weight changes.
Considering the discernible changes in infant mortality over the last 30 years, EPA developed a
regression analysis to estimate the relationship between birth weight and infant mortality using
the most recently available Period/Cohort Linked Birth-Infant Death Data Files published by
NCHS from the 2017 period/2016 cohort and the 2018 period/2017 cohort (CDC, 2017, 2018).
These data provide information on infants who are delivered alive and receive a birth
certificate.46 EPA selected variables of interest for the regression analysis, including maternal
demographic and socioeconomic characteristics, maternal risk and risk mitigation factors (e.g.,
number of prenatal care visits, smoker status), and infant birth characteristics. EPA included
several variables used in Ma et al. (2010) (maternal age, maternal education, marital status, and
others - see Appendix E for the complete list) as well as additional variables to augment the set
of covariates included in the analyses. In addition, EPA developed separate models for different
race/ethnicity categories (non-Hispanic Black, non-Hispanic White, and Hispanic) and interacted
birth weight with categories of gestational age, similar to Ma et al. (2010).47 Appendix E
provides details on model development and regression results.
Table 6-10 presents the resulting odds ratios and marginal effects (in terms of deaths per 1,000
births for every 1 g increase in birth weight) estimated for changes in birth weight among
different gestational age categories in the mortality regression models for non-Hispanic Black,
non-Hispanic White, and Hispanic race/ethnicity subpopulations. Marginal effects for birth
weight among different gestational age categories indicate the change in the incidence of infant
46 These data do not include information on miscarriages or stillbirths.
47 Note that Ma et al. (2010) developed a model for infants with Mexican heritage, rather than the Hispanic population, and
interacted birth weight with gestational age as a continuous interaction variable, rather than developing different birth weight
variables per gestational age category. Ma et al. (2010) did not consider the Hispanic paradox, a term for the epidemiological
finding that Hispanic and Latino Americans often have lower risk of poor health outcomes compared to race/ethnicity groups
with higher income and education levels. Note that Ma et al. (2010) developed a model for infants with Mexican heritage, rather
than the Hispanic population, and interacted birth weight with gestational age as a continuous interaction variable, rather than
developing different birth weight variables per gestational age category. Ma et al. (2010) did not consider the Hispanic paradox, a
term for the epidemiological finding that Hispanic and Latino Americans often have lower risk of poor health outcomes
compared to race/ethnicity groups with higher income and education levels.
Proposed PFAS Rule Economic Analysis
6-37
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
mortality per 1 g increase in birth weight.48 Marginal effects for birth weight among gestational
age categories vary across different race/ethnicity subpopulations. As shown in Figure 6-2, the
marginal effects for birth weight among different gestational age categories are higher in the
non-Hispanic Black model than in the non-Hispanic White and Hispanic models, particularly for
extremely and very preterm infants, indicating that LBW increases the probability of mortality
within the first year more so among non-Hispanic Black infants than among non-Hispanic White
and Hispanic infants.
EPA relies on odds ratios estimated using the birth weight-mortality regression model to assess
mortality outcomes of reduced exposures to PFOA/PFOS in drinking water under the regulatory
alternatives. To obtain odds ratios specific to each race/ethnicity and 100 g birth weight
increment considered in the birth weight benefits model,49 EPA averaged the estimated odds
ratios for 1 g increase in birth weight over the gestational age categories using the number of
infants (both singleton and multiple birth) that fall into each gestational age category as weights.
Separate gestational age category weights were computed for each 100 g birth weight increment
and race/ethnicity subpopulation within the 2017 period/2016 cohort and 2018 period/2017
cohort Linked Birth-Infant Death Data Files. The weighted birth weight odds ratios are then used
in conjunction with the estimated change in birth weight and baseline infant mortality rates to
determine the probability of infant death under the regulatory alternatives, as described further in
Section 6.4.3.1.
48 All marginal effect values for birth weight among different gestational age categories are negative and decrease in magnitude
with each higher gestational age category, indicating that the probability of mortality decreases as gestational age and birth
weight increase. For example, using marginal effects from the non-Hispanic Black model, for extremely preterm infants a 100 g
birth weight increase on average would translate to 20 fewer infant deaths per 1,000 births in this gestational age category or a
2% decrease in the probability of mortality within one year of birth. The same birth weight increase at a higher gestational age
would still decrease mortality risk but to a lesser extent.
49 The birth weight risk reduction model evaluates changes in birth weight in response to PFOA/PFOS drinking water level
reductions for infants who fall into 100 g birth weight increments (e.g., birth weight 0-99 g, 100-199 g, 200-299 g... 8,000-8,099
g, 8,100-8,165 g).
Proposed PFAS Rule Economic Analysis
6-38
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Non-Hispanic Black —±—Non-Hispanic White —•—Hispanic
0.00 a
-Q -0.05
CD
>
o
o
CD
in 5 -o.io
(1) CO
"O "3
+-> ^
Jj | -0-15
-- CUD
O ^
a;
u
c
a)
-a -0.20
-0.25
Extremely Preterm Very Preterm Moderately Preterm Term
Gestational Age Category
Figure 6-2: Comparison of Change in Incidence of Infant Death per 1 g Increase in birth
weight by Gestational Age Category and Race/Ethnicity (Deaths per 1,000 Births)
Notes: Gestational age categories defined as extremely preterm (<=28 weeks), very preterm (>28 weeks and <=32 weeks),
moderately preterm (>32 weeks and <=37 weeks), and term (>37 weeks). Data based on the 2016/17 and 2017/18 CDC Period
Cohort Linked Birth-Infant Death Data Files obtained from NCHS/NVSS. Marginal effects and odds ratios are estimated using a
regression model that also includes covariates representative of infant birth characteristics in addition to birth weight, maternal
demographic characteristics, and maternal risk factors. Details are included in Appendix E.
Table 6-10: Race/Ethnicity- and Gestational Age-Specific Birth Weight Marginal
Effects and Odds Ratios from the Mortality Regression Models
Race
Gestational Age
Categoryb
Marginal Effect per
1,000 births (95% CI)
Odds Ratio (95% CI)
Non-Hispanic Black
Extremely Preterm
-0.20400
(-0.21910, -0.18890)
0.99817
(0.99802, 0.99832)
Very Preterm
-0.04580
(-0.04820, -0.04340)
0.99816
(0.99804, 0.99827)
Moderately Preterm
-0.01030
(-0.01080, -0.009850)
0.99852
(0.99846, 0.99857)
Term
-0.00453
(-0.00472, -0.00434)
0.99856
(0.99851,0.9986)
Non-Hispanic White
Extremely Preterm
-0.12160
(-0.13080, -0.11240)
0.99866
(0.99855, 0.99878)
Very Preterm
-0.03290
0.9985
Proposed PFAS Rule Economic Analysis
6-39
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-10: Race/Ethnicity- and Gestational Age-Specific Birth Weight Marginal
Effects and Odds Ratios from the Mortality Regression Models
Race
Gestational Age
Categoryb
Marginal Effect per
1,000 births (95% CI)
Odds Ratio (95% CI)
(-0.03430, -0.03140)
(0.99842, 0.99858)
Moderately Preterm
-0.00677
(-0.00702, -0.00652)
0.99867
(0.99863, 0.99872)
Term
-0.00228
0.99865
(-0.00236, -0.00221)
(0.99861, 0.99868)
Hispanic
Extremely Preterm
-0.15260
(-0.16770, -0.13750)
0.99835
(0.99817, 0.99853)
Very Preterm
-0.03290
(-0.03510, -0.03070)
0.99846
(0.99835, 0.99858)
Moderately Preterm
-0.00626
(-0.00659, -0.00592)
0.99856
(0.99849, 0.99862)
Term
-0.00219
(-0.00229, -0.00208)
0.99849
(0.99844, 0.99855)
Notes:
aData based on the 2016/17 and 2017/18 CDC Period Cohort Linked Birth-Infant Death Data Files obtained from
NCHS/NVSS. Marginal effects and odds ratios are estimated using a regression model that also includes covariates
representative of infant birth characteristics in addition to birth weight, maternal demographic characteristics, and maternal
risk factors. All effects were statistically significant at the 5% level. Additional details are included in Appendix E.
bGestational age categories defined as extremely preterm (<=28 weeks), very preterm (>28 weeks and <=32 weeks),
moderately preterm (>32 weeks and <=37 weeks), and term (>37 weeks).
EPA weighted the race/ethnicity-specific mortality odds ratios in Table 6-10 by the proportions
of the infant populations who fell into each gestational age within a 100 g birth weight
increment, based on the 2016/17 and 2017/18 period cohort data, to obtain a weighted mortality
odds ratio estimate for each modeled race/ethnicity subpopulation and 100 g birth weight
increment. The weighted mortality odds ratios are shown in Figure 6-3. 50
50 Note that weighted mortality odds ratios for the Hispanic population at larger birth weight increments fluctuate between
0.99849 and 0.99856. Due to the small sample size of the Hispanic infant population within these birth weight increments, 100
percent of infants in a specific birth weight increment is associated with either moderately preterm or term gestational age
categories. For instance, all Hispanic infants included in the analysis who were between 7,800 and 7,899 g were full-term, while
all Hispanic infants who were between 7,900 and 7,999 g were moderately preterm. Therefore, the weighted mortality odds ratio
for Hispanic infants between 7,800 and 7,899 g is equal to the full-term mortality odds ratio estimated for the Hispanic infant
population, while the weighted mortality odds ratio for Hispanic infants between 7,900 and 7,999 g is equal to the moderately
preterm mortality odds ratio estimated for the Hispanic infant population.
Proposed PFAS Rule Economic Analysis
6-40
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
oo
tO
00
LO
00
*3"
00
on
oo
CM
00
00
Non-Hispanic Black Non-Hispanic White/Other
Hispanic
~
k r. * t " —v
xV /: • >.
ro
cr
oo
"O
"O
o
"O
CD
(D
5
o
o
O
o
o
O
O
o
o
o
o
o
o
o
o
o
o
o
O
o
o
O
O
o
o
o
o
o
o
o
o
o
o
LD
O
LO
o
LO
O
LO
o
LO
o
LO
o
LO
o
LO
o
1
T—1
r\j
r\j
m
m
LO
LO
to
to
r*>
r*>
00
BW Increment (g)
Figure 6-3: Weighted Mortality Odds Ratios Based on Populations of Infants Falling into 100 g Birth Weight Increments and
Four Gestational Age Categories
Note: Weighted mortality odds ratios refer to the exponentiation of the sum of odds ratios estimated for each gestational age category and race/ethnicity-specific infant population
multiplied by the proportions of the infant populations who fell into each gestational age within a 100 g birth weight increment, based on the 2016/17 and 2017/18 CDC Period
Cohort Linked Birth-Infant Death Data Files obtained from NCHS/NVSS, to obtain a weighted odds ratio estimate for each modeled race/ethnicity and 100 g birth weight
increment. EPA applies the weighted mortality odds ratios estimated for the non-Hispanic White subpopulation to the "other" race/ethnicity subpopulation because of similarities
in infant death rates from 2016 to 2018 among non-Hispanic White infants (4.75 deaths per 1,000) and non-Hispanic other infants (4.45 deaths per 1,000).
Proposed PFAS Rule Economic Analysis
6-41
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Note that EPA did not model the relationship between birth weight and infant mortality for other
race/ethnicity subpopulations because doing so for each individual race/ethnicity or combination
of all "other" races/ethnicities is precluded by very low sample sizes (i.e., imprecise coefficients
and imprecise marginal effects). EPA applies the weighted mortality odds ratios estimated for the
non-Hispanic White subpopulation to the "other" race/ethnicity subpopulation because of
similarities in infant death rates from 2016 to 2018 among non-Hispanic White infants (4.75
deaths per 1,000) and non-Hispanic other infants (4.45 deaths per 1,000).
6.4.3.2 Estimating the Number of Infants Affected by Birth Weight
Changes and Changes in Infant Mortality
Based on reduced serum PFOA/PFOS exposures under the regulatory alternatives and the
estimated relationship between birth weight and infant mortality, EPA estimates the subsequent
change in birth weight for those infants affected by decreases in PFOA/PFOS and changes in the
number of infant deaths. EPA evaluates these changes at each PWS EP affected by the regulatory
alternatives and the calculations are performed for each race/ethnicity group, 100 g birth weight
category, and year of the analysis.
6.4.3.2.1 Changes in Birth Weight
EPA combined estimated average annual changes in PFOA and PFOS serum levels for women
of childbearing age (15 to 44 years old) by analysis year, race/ethnicity group, and PWS EP (see
Section 6.3.2) with the serum PFOA/PFOS-birth weight exposure-response slope factors (see
Table 6-9) to compute average annual changes in birth weight per newborn as follows:
Equation 9:
ABWy r p max [CAP,SFbwppoa ' APF0A_S6VUTn,yrp + S F^wppos 1 FOSserumy r p)
Where ABW is the change in birth weight under the regulatory alternatives, y is the analysis
year, r is the race/ethnicity group, p is the PWS EP analyzed; APFOA_Serum is the change in
PFOA serum for women of childbearing age under the regulatory alternatives; APFOS_Serum is
the change in PFOS serum for women of childbearing age under the regulatory alternatives;
SFbw.pfoa and SFbwpfos are the serum-birth weight exposure-response slope factors for PFOA
and PFOS, respectively; and CAP is the 200 g cap placed on the birth weight changes.
6.4.3.2.2 Changes in Infant Death Rate
EPA used average annual changes in birth weight under the regulatory alternatives (Equation 9)
to estimate the associated infant mortality odds ratios, ORy i r p:
Equation 10:
ORy irp = exp(ABWyrp ¦ \n(ORir))
Where y is the analysis year, i is the 100 g birth weight increment, r is the race/ethnicity group,
p is the PWS EP analyzed, and ORi r is the weighted odds ratio for a 1 g birth weight increase
associated with each 100 g birth weight increment for a given race/ethnicity category (see
Section 6.4.3).
Proposed PFAS Rule Economic Analysis
6-42
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA combined the result of Equation 10 with the baseline infant death rate to estimate the infant
death rate under the regulatory alternatives, DRRegulatory Alternative y i r v.
Equation 11:
n n _ ORy i r p'DR^aseline y i r p
U^RegulatoryAlternative,y,i,r,p ~ T"j7_77/? no
y,i,r,p ' "Baseline,y,i,r,p
Where DRBaseUne y i r p is the baseline death rate per birth computed from 2012-2018 death rates
per 500 g birth weight increment (CDC, 2020a),51 y is the analysis year, i is 100 g birth weight
increment, r is the race/ethnicity group, p is the PWS EP analyzed, and ORy i r p is the mortality
odds ratio associated with the annual change in birth weight under the regulatory alternatives.
6.4.3.2.3 Affected Infant Population Size
The annual race/ethnicity- and PWS EP-specific number of infants affected by changes in
PFOA/PFOS drinking water levels is based on the 2021 retail population served at each PWS
from the SDWIS and 2021 race/ethnicity-specific population estimates from the U.S. Census
(U.S. Census Bureau, 2020a; see Appendix B). Because birth rates per race/ethnicity group and
100 g birth weight increment are often suppressed due to lack of data, EPA multiplied state-level
birth rates per race/ethni city group from the Centers for Disease Control and Prevention (CDC)
Linked Birth/Infant Death records from 2012 to 2018 (CDC, 2020a) by the ratio of infants falling
within each 100 g birth weight increment per state (not specific to race/ethni city) to the total
number of infants per state to distribute the number of affected infants in each state. EPA
imputed state-level data that was missing from the 2012-2018 CDC Linked Birth/Infant Death
records with data at the census region level. EPA used the same approach to assign average birth
weights per race/ethni city group over the 100 g birth weight increments for use in COI data
matching (See Section 6.4.4). Using the 2012-2018 imputed state-level birth rate data, EPA
computed the share of births that correspond to each 100 g birth weight increment (i),
race/ethni city (r), and PWS EP (p) as the ratio of race/ethni city- and state-specific (s) birth
rates52 in a particular birth weight increment to the sum of birth rates associated with all birth
weight increments:
Equation 12:
Share of Birthsirp = —( 2012 2018,1,>,s)
sum(BR2Ql2-2oia,i,r,s)
Next, EPA assumed that the share of births within each 100 g birth weight increment (from
Equation 12) would remain constant throughout the period of analysis and estimated the annual
51 EPA assumed that the same death rate applies to infants in all 100 g birth weight increments falling in the 500 g birth weight
range.
52 In this analysis, EPA applies state-specific birth rates that correspond to the state for which each PWS EP is located.
Proposed PFAS Rule Economic Analysis
6-43
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
affected infant population size for each future analysis year (y), 100 g birth weight increment (i),
race/ethnicity group (r), and PWS EP (p), Birthsyirp as follows:
Equation 13:
Birthsy i r p = Birthsy r v ¦ Share of Birthsirp
6.4.3.2.4 Infant Deaths Avoided and the Number of Surviving Infants
EPA used the estimated annual infant population size, Birthsy i r p, along with infant death rates,
D^Baseline,y,i,r,p &rid DRReguiatory Alternative,y,,i,,r.,P' compute the annual number of deaths
expected at baseline (Equation 14) and the annual number of deaths expected under the
regulatory alternatives (Equation 15):
Equation 14:
DeathsBaseiiney i r p BirthSyir p ¦ DR^asenney irp
Equation 15:
DeathSftggni^Qyy Alternative,y,i,r,P ~ BirthSyp 1 DRRegUiat:ory Alternative,y,i,r,P
EPA estimated the annual number of avoided infant deaths, Avoided Deathsy i r p, as:
Equation 16:
Avoided DeathSy i r p DeathSgaseiine y i r p De(XthSj^egUia^ory Alternative,y,i,r,V
EPA computed the population of surviving infants whose birth weight would be affected by
changes in PFOA/PFOS exposure {SurvivorsReguiatory Aiternative y i r p) as the number of births
less the number of deaths under the regulatory alternatives. EPA estimated the annual number of
avoided infant deaths, Avoided Deathsy i r p, as:
Equation 17:
SurvivorsReguiatory Alternative,y,i,r,p ~ BirthSy i r p 1 (1 — DRRegUiat:ory Alternative,y,i,r,p)
6.4.4 Valuation of Reduced Birth Weight Impacts
EPA uses the Value of Statistical Life to estimate the benefits of reducing infant mortality and
COI to estimate the economic value of increasing birth weight in the population of surviving
infants born to mothers exposed to PFOA and PFOS in drinking water. Value of Statistical Life
updating information is provided in Section 2.2.
EPA's approach to monetizing benefits associated with incremental increases in birth weight
resulting from reductions in drinking water PFOA/PFOS levels relies on avoided medical costs
associated with various ranges of birth weight. Although the economic burden of treating infants
at various birth weights also includes non-medical costs, very few studies to date have quantified
Proposed PFAS Rule Economic Analysis
6-44
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
such costs (Klein et al., 2018; ICF, 2021). EPA selected the medical cost function from Klein et
al. (2018) to monetize benefits associated with the estimated changes in infant birth weight
resulting from reduced maternal exposure to PFOA/PFOS.53 EPA selected the cost function from
Klein et al. (2018) because it is based on recent data on birth weight, healthcare utilization, and
healthcare costs that encompass a longer time period and a larger population than data used in
other studies (e.g., Almond et al., 2005). Additional studies that EPA reviewed provided only an
incremental cost for LBW infants compared to normal birth weight (NBW) infants (greater or
equal to 2,500 g; e.g., Almond et al., 2010 and Malits et al., 2018). Klein et al. (2018), on the
other hand, estimated incremental medical costs as a function of birth weight over the range from
900 to 4,500 g and used a continuous spline function (Figure 6-4), rather than allowing for a
discontinuity at the very low birth weight level (i.e., < 1,500 grams). Table 6-11 summarizes the
incremental cost changes associated with birth weight increases from Klein et al. (2018).
$80,000
o" $70,000
O
r\i
$60,000
L_
ro
>. $50,000
¦P
1/5
S $40,000
:§ $30,000
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-11: Simulated Cost Changes for Birth Weight Increases ($2021) (Based on Klein
and Lynch, 2018 Table 8)
Simulated Cost Changes for Birth Weight Increases, Dollars per Gram
Birth Weighta'b ($2021)°
+0.04 lb (+18 g)
+0.11 lb (+50 g)
+0.22 lb (+100 g)
2 lb (907 g)
-$126.53
-$112.87
-$109.39
2.5 lb (1,134 g)
-$94.88
-$84.64
-$82.03
3 lb (1,361 g)
-$71.15
-$63.47
-$61.51
3.3 lb (1,497 g)
-$59.86
-$53.40
-$51.75
4 lb (1,814 g)
-$40.00
-$35.69
-$34.59
4.5 lb (2,041 g)
-$30.00
-$26.76
-$25.93
5 lb (2,268 g)
-$22.49
-$20.07
-$19.45
5.5 lb (2,495 g)
-$0.93
-$0.84
-$0.84
6 lb (2,722 g)
-$0.91
-$0.83
-$0.83
7 lb (3,175 g)
-$0.88
-$0.80
-$0.80
8 lb (3,629 g)
-$0.85
-$0.77
-$0.77
9 lb (4,082 g)
$3.15
$2.87
$2.89
10 lb (4,536 g)
$3.54
$3.23
$3.26
Notes:
aValues for birth weight have been converted from lb to g.
bNote that simulated medical costs increase, rather than decrease, in response to increased birth weight changes among high
birth weight infants (those greater than 8 lb). Among HBW infants, there is a higher risk of birth trauma, metabolic issues, and
other health problems (Klein et al., 2018).
cValues scaled from $2010 to $2021 using the medical care Consumer Price Index (U.S. Bureau of Labor Statistics, 2021).
Using the incremental cost changes from Klein et al. (2018), EPA calculates the change in
medical costs resulting from changes in birth weight among infants in the affected population
who survived the first year following birth. To do so, EPA linearly interpolates between the birth
weight and cost values presented in Klein et al. (2018) to obtain a cost value for every 1 g birth
weight increment, as shown in Figure 6-5. EPA then matches this interpolated birth weight value
to the nearest baseline average birth weight value in each 100 g birth weight increment to obtain
the simulated cost change for birth weight increases that are estimated to be between zero and 18
g, between 19 and 50 g, and between 51 and 100 g or more.54
54 Note that EPA caps birth weight changes at 200 g, as described in earlier sections. EPA assumes that the cost of illness
estimates for birth weight increases between 51 and 100 g apply to birth weight increases greater than 100 g.
Proposed PFAS Rule Economic Analysis
6-46
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
$10
So
3 -$io
o
CM
-$20
(U
i/i
nj
0)
o
M
O
O
-$30
-$40
-$50
-$60
on
o
LT>
M -$70
oo
-$80
-$90
-$100
O -$no
-$120
-$130
/
/
q
m«"S
*
<•- *
7
// /
/.•• /
••• /
/.
/
/
/• -
/.• /
*
/: /
// /
// /
J /
# /
// »
// /
/; «
/
//
ff
i
*
*7
* :
i
»
i
i
+18g
+50g
+100g
500 1,000 1,500 2,000 2,500 3,000
Baseline Average Birth Weight (g)
3,500
4,000
4,500
Figure 6-5. Interpolated Cost of Illness at Baseline Average Birth Weights, by Estimated
Change in Birth Weight Under the Proposed Rule
Proposed PFAS Rule Economic Analysis
6-47
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.4.5 Results
Table 6-12 to Table 6-15 provide the health effects avoided and valuation associated with birth
weight impacts.
Table 6-12: National Birth Weight Benefits, Proposed Option (PFOA and PFOS MCLs
of 4.0 ppt and HI of 1.0)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Increase in Birth Weight
(millions of grams)
114.2
209.3
329.7
114.2
209.3
329.7
Number of Birth Weight-
Related Deaths Avoided
676.8
1,232.7
1,941.0
676.8
1,232.7
1,941.0
Total Annualized Birth
Weight Benefits (Million
$2021)b
$97.36
$177.66
$279.49
$74.62
$139.01
$219.43
Note: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-13: National Birth Weight Benefits, Option la (PFOA and PFOS MCLs of 4.0
ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Increase in Birth Weight
(millions of grams)
111.7
206.3
326.9
111.7
206.3
326.9
Number of Birth Weight-
Related Deaths Avoided
665.4
1,214.7
1,915.4
665.4
1,214.7
1,915.4
Total Annualized Birth
Weight Benefits (Million
$2021)b
$95.73
$175.05
$276.44
$74.66
$136.97
$217.02
Note: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-48
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-14: National Birth Weight Benefits, Option lb (PFOA and PFOS MCLs of 5.0
ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Expected
95th
5th
Expected
95th
Percentile3
Value
Percentile3
Percentile3
Value
Percentile3
Increase in Birth Weight
97.6
181.9
292.1
97.6
181.9
292.1
(millions of grams)
Number of Birth Weight-
578.9
1,069.5
1,707.3
578.9
1,069.5
1,707.3
Related Deaths Avoided
Total Annualized Birth
$83.27
$154.13
$246.43
$64.94
$120.59
$193.47
Weight Benefits (Million
$2021)b
Note: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-15: National Birth Weight Benefits, Option lc (PFOA and PFOS MCLs of 10.0
ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Increase in Birth Weight
(millions of grams)
51.0
109.2
195.3
51.0
109.2
195.3
Number of Birth Weight-
Related Deaths Avoided
299.5
643.3
1,140.5
299.5
643.3
1,140.5
Total Annualized Birth
Weight Benefits (Million
$2021)b
$43.22
$92.70
$164.19
$34.18
$72.51
$125.80
Note: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
6.5 Cardiovascular Disease
6.5.1 Overview of the Cardiovascular Disease Risk Analysis
Figure 6-6 provides an overview of the approach used to quantify and value the changes in CVD
risk associated with reductions in exposure to PFOA and PFOS via drinking water. Section 4.4
details the PWS EP-specific PFOA/PFOS drinking water occurrence estimation and Section 6.3
describes modeling of serum PFOA/PFOS concentrations. EP-specific time series of the
differences between serum PFOA/PFOS concentrations under baseline and regulatory
Proposed PFAS Rule Economic Analysis
6-49
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
alternatives are inputs into this analysis. For each EP, evaluation of the changes in CVD risk
involves the following key steps:
1. Estimation of annual changes in TC55 and BP levels using exposure-response functions
for the potential effects of serum PFOA/PFOS on these biomarkers;
2. Estimation of the annual incidence of fatal and non-fatal first hard CVD events, defined
as fatal and non-fatal myocardial infarction (MI; i.e., heart attack), fatal and non-fatal
ischemic stroke (IS), or other coronary heart disease (CHD) death occurring in
populations without prior CVD event experience (D'Agostino et al., 2008; Goff et al.,
2014; Lloyd-Jones et al., 2017), and post-acute CVD mortality corresponding to baseline
and regulatory alternative TC and BP levels in all populations alive during or born after
the start of the evaluation period; and
3. Estimation of the economic value of reducing CVD mortality and morbidity from
baseline to regulatory alternative levels, using the Value of Statistical Life and COI
measures, respectively.
Section 6.5.2 discusses the exposure-response models for TC and BP. Section 6.5.3 details the
estimated CVD risk reductions using the Pooled Cohort ASCVD risk model (Goff et al., 2014)
and the life table approach. Section 6.5.4 discusses EPA's valuation methodology for fatal and
non-fatal CVD events. Section 6.5.5 presents the results of the analysis.
55 EPA discusses the relationship between PFOA/PFOS exposure and other forms of cholesterol in Appendix F.
Proposed PFAS Rule Economic Analysis
6-50
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Legend:
Result of
upstream analysis
Data/Inputs
Model
# Analysis step
Serum PFOA and PFOS
concentration difference
between baseline and
regulatory alternative
i
Serum-TC and
-BP exposure-response
function
TC and BP difference
between baseline and
treatment scenario
Life table CVD model
Change in incidence
of non-fatal CVDb
ASCVD risk model
Location-specific
population size
Prevalence and
incidence of CVD
events
Annual cause-specific
mortality rates and life
table information3
Baseline total
cholesterol
Blood pressure level
and treatment status
Smoking and
diabetes status
Change in incidence
of fatal CVD*"
Medical costs of
CVD treatment
Value of reduced
CVD incidence
Value of avoided
excess mortality
Value of a
statistical life
Total value of reduced CVD
Abbreviations: PFOA - perfluorooctanoic acid, PFOS - perfluorooctanesulfonic acid, TC - total cholesterol, BP -
blood pressure, CVD - cardiovascular disease, ASCVD - atherosclerotic cardiovascular disease, MI - myocardial
infarction, IS - ischemic stroke, CFTD - coronary heart disease
Notes:
"Data from the Centers for Disease Control (CDC) and Prevention.
bNon-fatal CVD includes non-fatal first MI and non-fatal first IS
''Fatal CVD includes fatal first MI, fatal first IS, other fatal first CUD events, and post acute CVD mortality among
survivors of the first MI and the first IS.
Figure 6-6'. Overview of the CVD Risk Model
Proposed PFAS Rule Economic Analysis
6-51
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.5.2 Cardiovascular Disease Exposure-Response Analyses
6.5.2.1 Estimation of Cholesterol Changes
The ASCVD model includes TC as a predictor of first hard CVD events. EPA did not identify
any readily available relationships for PFOA or PFOS and TC that were specifically relevant to
the age group of interest (40-89 years, the years for which the ASCVD model estimates the
probability of a first hard CVD event). Therefore, the Agency developed a meta-analysis of
studies reporting associations between serum PFOA or PFOS and TC in general populations
(e.g., populations that are not a subset of workers or pregnant women). Statistical analyses that
combine the results of multiple studies, such as meta-analyses, are widely applied to investigate
the associations between contaminant levels and associated health effects. Such analyses are
suitable for economic assessments because they can improve precision and statistical power
(Engels et al., 2000; Deeks, 2002; Riicker et al., 2009). Appendix F provides details on the
studies selection criteria, meta-data development, meta-analysis results, and discussion of the
uncertainty and limitations inherent in EPA's exposure-response analysis.
EPA identified studies for inclusion in the meta-analysis using data from literature reviews,
including those performed by ATSDR in the development of their Toxicological Review Public
Comment Draft (ATSDR, 2018), which included literature through mid-2017, and those
performed for developing EPA's Toxicity Assessments and Proposed Maximum Contaminant
Level Goals for PFOA and PFOS in Drinking Water (U.S. EPA, 2023 d; U.S. EPA, 2023 e),
which included studies published from 2016 through September 2020. EPA included studies in
the meta-data if they reported quantitative estimates (e.g., regression coefficients) and measures
of uncertainty (e.g., standard errors, confidence intervals) of associations between serum PFOA
or PFOS and TC in general population adults aged 20 years and older. EPA included a total of 14
studies in the meta-analysis. Of these, 12 studies were used to develop exposure-response
relationships for serum PFOA or PFOS and TC (i.e., not all relevant studies report the effects for
both PFOA and PFOS). The unit in the meta-analysis was change in TC in mg/dL per increases
in serum PFOA or PFOS. EPA conducted four separate meta-analyses for each chemical (PFOA
or PFOS).
Table 6-16 summarizes the 14 studies that EPA identified from literature reviews and used to
derive slope estimates for PFOA and PFOS associations with serum TC levels.56 Six of the
studies that EPA retained for use in the meta-analysis were based on serum PFAS and serum TC
measurements from the U.S. general population (National Health and Nutrition Examination
Survey [NHANES]) (Dong et al., 2019; Fan et al., 2020; He et al., 2018; Jain et al., 2019; Liu et
al., 2018; Nelson et al., 2010); there were also general population studies from Canada (Fisher et
al., 2013), Sweden (Y. Li et al., 2020), Taiwan (Yang et al., 2018; C. Y. Lin et al., 2020), and
Henan Province, China (Fu et al., 2014). Chateau-Degat et al. (2010) reported on the association
between PFOS and TC in a Canadian Inuit population. EPA also retained the results from a study
of a highly exposed population in the U.S. (the C8 cohort) (Steenland et al., 2009) and from a
study using participants in a U.S. diabetes prevention program (P.-I. D. Lin et al., 2019). EPA
retained results from Steenland et al. (2009) because serum levels in the examined cohort were
only modestly elevated compared to less exposed populations (e.g., the median serum PFOA
56 For this effort, EPA focused on PFOA and PFOS, since these are by far the most well-studied perfluorinated
compounds.
Proposed PFAS Rule Economic Analysis
6-52
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
concentration in this cohort was 27 ng/mL, with an interquartile range of 13.1 to 67 ng/mL).
EPA retained results from P.-I. D. Lin et al. (2019) because the examined cohort included pre-
diabetic adults enrolled in a diabetes prevention program; thus, this cohort was representative of
a large portion of the U.S. adult population.
Proposed PFAS Rule Economic Analysis
6-53
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-16: Studies Selected for Inclusion in the Meta-Analyses
TC and Serum PFAS
Relationship Evaluated in
Author and Year Title Study
PFOA
PFOS
Steenland et al.,
2009a,d
Association of Perfluorooctanoic Acid and Perfluorooctane
Sulfonate With Serum Lipids Among Adults Living Near a
X
X
Chemical Plant
Chateau-Degat et
al.,2010ad
Effects of Perfluorooctanesulfonate Exposure on Plasma
Lipid Levels in the Inuit Population of Nunavik (Northern
X
Quebec)
Nelson et al.,
2010a,d
Exposure to Polyfluoroalkyl Chemicals and Cholesterol,
Body Weight, and Insulin Resistance in the General U.S.
X
X
Population
Fisher etal., 2013
a,d
Do Perfluoroalkyl Substances Affect Metabolic Function and
Plasma Lipids? —Analysis of the 2007-2009, Canadian
X
X
Health Measures Survey (CHMS) Cycle 1
Associations Between Serum Concentrations of
Fu et al., 2014a d
Perfluoroalkyl Acids and Serum Lipid Levels in a Chinese
Population
PFOA is Associated with Diabetes and Metabolic Alteration
X
X
He et al., 2018°
in US Men: National Health and Nutrition Examination
Survey 2003-2012
Association Among Total Serum Isomers of Perfluorinated
X
X
Liu et al., 2018°
Chemicals, Glucose Homeostasis, Lipid Profiles, Serum
X
X
Protein and Metabolic Syndrome in Adults: NHANES,
2013-2014
Using 2003-2014 U.S. NHANES Data to Determine the
Dong et al., 2019b
Associations Between Per- and Polyfluoroalkyl Substances
and Cholesterol: Trend and Implications
X
X
Jain et al., 2019b
Roles of Gender and Obesity in Defining Correlations
X
X
Between Perfluoroalkyl Substances and Lipid/Lipoproteins
P.-I. D. Lin et al.,
2019b
Per- and Polyfluoroalkyl Substances and Blood Lipid Levels
in Pre-Diabetic Adults—Longitudinal Analysis of the
X
X
Diabetes Prevention Program Outcomes Study
Fan etal., 2020b
Serum Albumin Mediates the Effect of Multiple Per- and
X
X
Polyfluoroalkyl Substances on Serum Lipid Levels
Associations Between Perfluoroalkyl Substances and Serum
Y. Li et al., 2020b
Lipids in a Swedish Adult Population With Contaminated
Drinking Water
X
X
Abbreviations: TC - total cholesterol; PFOS - perfluorooctane sulfonic acid; PFOA - perfluorooctanoic acid; PFAS - per and
polyfluoroalkyl substances; PFAS - per-and polyfluoroalkyl substances.
Notes:
aStudies identified based on ATSDR literature review.
bStudies identified based on EPA literature review.
cStudies available in both assessments.
dStudies available in PFOA and/or PFOS health effects support documents (U.S. EPA, 2016e; U.S. EPA, 2016f).
EPA developed exposure-response relationships between serum PFOA/PFOS and TC for use in
the CVD analysis using the meta-analyses restricted to studies of adults in the general population
reporting similar models. EPA used untransformed serum PFOA/PFOS to reduce bias due to
Proposed PFAS Rule Economic Analysis
6-54
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
back-transformations of effect estimates. For studies that provided results only for log-
transformed serum PFOA/PFOS (five studies) or log-transformed outcomes (two studies), or
both log-transformed serum PFOA/PFOS and outcomes (two studies), EPA approximated the
results for an untransformed analysis using the approach outlined by Rodriguez-Barranco et al.
(2017) and Dzierlenga et al. (2020). When using studies reporting linear associations between
TC and serum PFOA or PFOS, EPA estimated a positive increase in TC of 1.57 (95% CI: 0.02,
3.13) mg/dL per ng/mL serum PFOA (p-value=0.048), and of 0.08 (95% CI: -0.01, 0.16) mg/dL
per ng/mL serum PFOS (p-value=0.064). EPA selected the pooled slope estimate based on the
studies using linear models to ease interpretability and to reduce bias due to back-
transformations of effect estimates with log-transformed outcomes or exposures (see Appendix F
for details). While the association for PFOS and TC is not significant at the 0.05 confidence
level, it is significant at the 0.10 confidence level (p-value=0.064). Furthermore, the literature
provides sufficient support of a positive association (e.g., Chateau-Degat et al., 2010; Dong et
al., 2019; U.S. EPA, 2023d; U.S. EPA, 2023e). The studies are large with more than 700 and
8,900 participants, respectively (Chateau-Degat et al., 2010; Dong et al., 2019) and have low risk
of bias. In addition, the estimated values are supported by sensitivity analyses and by the
estimates from potential candidate studies from exposure-response modeling for ongoing Agency
efforts (Dong et al., 2019). Based on the systematic review conducted by EPA of 39
epidemiologic studies published between 2016 and September 2020 for developing EPA's
Toxicity Assessments and Proposed Maximum Contaminant Level Goals for PFOA and PFOS in
Drinking Water, the available evidence supports a positive association between PFOS and TC in
the general population (U.S. EPA, 2023d; U.S. EPA, 2023e). For more information on the
systematic review and results, see EPA's Toxicity Assessments and Proposed Maximum
Contaminant Level Goals for PFOA and PFOS in Drinking Water (U.S. EPA, 2023 d; U.S. EPA,
2023e).
Note that EPA sought comments from the EPA Science Advisory Board on the cardiovascular
disease exposure-response approach (U.S. EPA, 2022k). The Science Advisory Board
recommended that EPA evaluate how the inclusion of HDLC effects would influence results.
EPA evaluated the inclusion of HDLC effects in a sensitivity analysis, described in Appendix K.
6.5.2.2 Estimation of BP Changes
PFOS exposure has been linked to other cardiovascular outcomes, such as systolic BP and
hypertension (Liao et al., 2020; U.S. EPA, 2023d). Because systolic BP is another predictor used
by the ASCVD model, EPA included the estimated changes in BP from reduced exposure to
PFOS in the CVD analysis. EPA selected the slope from the Liao et al. (2020) study — a high
confidence study conducted based on U.S. general population data from NHANES cycles 2003-
2012. Liao et al. (2020) estimated an increase of 1.35 (95% CI: 0.18, 2.53) in mmHg systolic BP
per logl0(ng/mL) PFOS among those not using antihypertensive medications. For the purposes
of this analysis, EPA converted this slope to 0.044 (95% CI: 0.006, 0.083) mmHg per ng/mL.
The evidence on the associations between PFOA and BP is not as consistent as for PFOS (see
Section 6.2.2.1.2). Therefore, EPA is not including effect estimates for the serum PFOA-BP
associations in the CVD analysis.
Proposed PFAS Rule Economic Analysis
6-55
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.5.3 Estimation of Cardiovascular Disease Risk Reductions
EPA relies on the life table-based approach to estimate CVD risk reductions because (1) changes
in serum PFOA/PFOS in response to changes in drinking water PFOA/PFOS occur over multiple
years, (2) CVD risk, relying on the ASCVD model, can be modeled only for those older than 40
years without prior CVD history, and (3) individuals who have experienced non-fatal CVD
events have elevated mortality implications immediately and within at least five years of the first
occurrence.57 Recurrent life table calculations are used to estimate a PWS EP-specific annual
time series of CVD event incidence for a population cohort characterized by sex, race/ethnicity,
birth year, age at the start of the PFOA/PFOS evaluation period (i.e., 2023), and age- and sex-
specific time series of changes in TC and BP levels obtained by combining serum PFOA/PFOS
concentration time series (Section 6.3) with exposure-response information (Section 6.5.3).
Baseline and regulatory alternatives are evaluated separately, with regulatory alternative TC and
BP levels estimated using baseline information on these biomarkers from external statistical data
sources and modeled changes in TC and BP due to conditions under the regulatory alternatives
(see Appendix G for detailed information on data sources used in CVD modeling).
EPA estimated the incidence of first hard CVD events based on TC serum and BP levels using
the ASCVD model (Goff et al., 2014), which predicts the 10-year probability of a hard CVD
event to be experienced by a person without a prior CVD history (see Section 6.5.3.2).58 EPA
adjusted the modeled population cohort to exclude individuals with pre-existing conditions, as
the ASCVD risk model does not apply to these individuals. For BP effects estimation, EPA
further restricts the modeled population to those not using antihypertensive medications for
consistency with the exposure-response relationship (see Section 6.5.3.2 for detail). Modeled
first hard CVD events include fatal and non-fatal MI, fatal and non-fatal IS, and other CHD
mortality. EPA also has estimated the incidence of post-acute CVD mortality among survivors of
the first MI or IS within 6 years of the initial event (Section 6.5.3.3).
The estimated CVD risk reduction resulting from reducing serum PFOA and serum PFOS
concentrations is the difference in annual incidence of CVD events (i.e., mortality and morbidity
associated with first-time CVD events and post-acute CVD mortality) under the baseline and
regulatory alternatives. Appendix G provides detailed information on all CVD model
components, computations, and sources of data used in modeling.
6.5.3.1 Life Table Calculations
The CVD model integrates the ASCVD model predictions and post-acute CVD mortality rates in
the series of recurrent calculations that produce a life table estimate for the affected population
cohort (e.g., non-Hispanic White females aged 70 years at the beginning of the evaluation
period). For each PWS EP, EPA evaluates population cohorts defined by a combination of birth
year, age, sex (males and females), and race/ethnicity (non-Hispanic White, non-Hispanic Black,
Hispanic, Other). In addition to the key standard life table components (i.e., the number of
persons surviving to a specific age and the number of all-cause deaths occurring at a given age)
57 EPA notes that elevated mortality for hard CVD event survivors may persist beyond five years of the initial event. However,
EPA did not identify U.S. based studies with sufficiently long follow-up to quantify mortality impacts beyond five years of the
initial event.
58 EPA did not identify studies that found statistically significant associations between the modeled biomarkers (TC, BP) and
CVD events in populations with prior CVD history. Discussion of the relevant literature is provided in Appendix G.
Proposed PFAS Rule Economic Analysis
6-56
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
for ages 40 years or older, the CVD model estimates the number of surviving persons with and
without a history of hard CVD events, the number of persons experiencing hard CVD events at a
given age, and the deaths from CVD and non-CVD causes at a given age.
Figure 6-7 summarizes the CVD model calculations for a population cohort age 0 at the start of
the evaluation period.59 The CVD model calculations are identical across race/ethnicity and sex
demographic subgroups but use subgroup-specific parameters.60 For cohorts born prior to or in
2023, the CVD model is initialized using the PWS-specific number of persons estimated to be
alive at the beginning of 2023. For cohorts born after 2023 (i.e., 2024-2104), the CVD model is
initialized using the PWS EP-, race/ethnicity-, sex, and scenario-specific number of persons who
died in the previous calendar year of the analysis, thereby ensuring that the size of the modeled
population remains constant throughout the analysis period. Additional PWS EP- and sex,
race/ethnicity, and age-specific population estimation assumptions are provided in Section 2.2;
additional details are included in Appendix B.
Once the model is initialized, the following types of calculations occur for each year within the
simulation period:61
• Recurrent standard life table calculations that rely on the all-cause, age-specific
annual mortality rates to evaluate the number of deaths among persons of a specific
integer age and the number of survivors to the beginning of the next integer age.62 These
calculations are executed whenever the current cohort age is in the 0-39 range. They are
represented by the navy blue segment of the timeline shown in Figure 6-7.
• Recurrent life table calculations that separately track subpopulations with and
without a history of hard CVD events, including estimation of the number of annual
CVD and non-CVD deaths (in either subpopulation), as well as the number of annual
post-acute CVD deaths experienced by survivors of the first hard CVD events that
occurred, at most, 5 years ago. These calculations are executed whenever the current
cohort age is 40 years or older.63 These calculations are represented by the blue segment
of the timeline. Figure 6-7 and Figure 6-8 further illustrates the year-specific calculations
required for explicit tracking of subpopulations with and without a hard CVD event
history.
59 This initial population cohort age is chosen because it allows for illustration of the full set of calculation types used in the CVD
model.
60 There are different ASC VD model coefficients for non-Hispanic White and non-Hispanic Black males and females. The figure
shows the generalized approach of the CVD model.
61 EPA notes that the simulation period is the lifespan of individuals relevant to the analysis. The simulation period is distinct
from the period of analysis in that some parts of the simulation period may fall outside the period of analysis. For example, for a
person aged 40 years at the start of the analysis period, the period of analysis will not capture the first 40 years of simulation
results.
62 Tife table calculations are based on the present-day information about life expectancy, disease, environmental exposure, and
other factors.
63 People 85 years or older, are treated as a single cohort in the model. The mortality rate for this cohort is assumed to be the
average mortality rate for those age 85-100 years. This cohort also used serum PFOA/PFOS values at age 85.
Proposed PFAS Rule Economic Analysis
6-57
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Model initialization
with estimated
location- and
race/ethnicity-specific
male/female age 0
population Recurrent standard life table
calculations using age-,
race/ethnicity-, and
sex-specific all-cause
annual mortality rates
A
Modeled population is
split into CVD and non-CVD
subpopulations using
age-, race/ethnicity-, and sex-specific
CVD prevalence data
Recurrent life table calculations
with explicit treatment of CVD population,
using age- and sex-specific CVD prevalence,
cause-specific annual mortality rates,
ASCVD model-based CVD incidence,
and post-acute CVD event mortality
r
Ages 0-39
Ages 40+
Figure 6-7: Overview of Life Table Calculations in the CVD Model
Note: The figure illustrates the model for population cohort age 0 years at the beginning of the evaluation period (i.e., calendar year 2023). Hie model is initialized using the age 0
PWS IP-specific population (see Appendix B for PWS population estimation details).
Proposed PFAS Rule Economic Analysis
6-58
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Figure 6-8 provides additional information on the post-acute CVD mortality estimation. Each
person included in the surviving current age-specific incident CVD subpopulation64
(corresponding to the group F result in Figure 6-8) is tracked for 5 additional years to estimate
the number of CVD deaths occurring in that timeframe. The recurrent estimates rely on age-
specific non-CVD mortality rates, estimated based on the CDC's life table data and annual CVD
mortality rates, and on post-acute CVD mortality rates, estimated based on Thom et al. (2001)
and S. Li et al. (2019).
Further details of the life table calculations are provided in Appendix G. The outputs of the life
table calculations and application of the ASCVD model are the PWS EP-specific estimates of the
annual number of persons experiencing their first non-fatal MI or IS event and the number of
deaths among those who have experienced their first hard CVD event, at most, 6 years ago. Note
that the ASCVD model does not predict risks separately by type of first hard CVD event (i.e.,
non-fatal MI, non-fatal IS, and fatal CVD). The distribution of these events by type is estimated
using data publicly available on CVD prevalence, incidence, and hospital mortality statistics as
described in Section 6.5.3.2 and integrated into the overall CVD impacts modeling.
64 For example, persons who experienced their first non-fatal MI or IS at age 70 and survived through the first post-
event year.
Proposed PFAS Rule Economic Analysis
March 2023
6-59
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Non-CVD population estimated to be alive at the start of integer age (A), applicable to ages 40-89.
Non CVD deaths (B)
-
^^NorvCVDpopulation (A) is adjusted for non CVD deaths (B) anduse^^^
^^^^basis for estimating first hard CVD event incidence
~
Non-CVD population eligible for
the first hard CVD event in the
following year (A-B-C-D).
First hard CVD event deaths (C)
First hard CVD event survivors (D)
CVD event survivor population (D) is adjusted for post-acute
excess mortality in year 0 since the initial event (E)*
Post-acute excess deaths
among CVD survivors in
first post-event year (E)
First hard CVD event
survivors at the end of
first post-event year (F)
Living subpopulation without prior history
of CVD events.
Deaths occurring at the current integer age
Note:
* Estimated number of CVD events is an input to
the monetization step.
Living subpopulation that experienced first
hard CVD at the current integer age
Current-year calculations
• Calculations occurring in years 1-5
following the first hard CVD event
Figure 6-8: CVD Model Calculations for Ages 40+ Tracking CVD
6.5.3.2 Risk and Distribution of First Hard Cardiovascular Disease Event
The first hard CVD event incidence estimates are generated by the Pooled Cohort ASCVD
model (Goff et al., 2014). The ASCVD model is commonly used in clinical practice to estimate
Proposed PFAS Rule Economic Analysis
6-60
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
CVD risk for those between ages 40 and 80, as well as for overall population risk management
(Lloyd-Jones et al., 2017). The ASCVD model predicts the 10-year probability of a hard CVD
event—fatal and non-fatal MI, fatal and non-fatal IS, or CHD death—to be experienced by a
person without a prior history of MI, IS, congestive heart failure, percutaneous coronary
intervention, coronary bypass surgery, or atrial fibrillation. The ASCVD model is a survival
model that links predictor levels at the start of the 10-year follow-up period to the first hard CVD
event incidence during the follow-up period; the modeling does not account for changes in CVD
risk predictors over time.
Four large longitudinal community-based epidemiologic cohort studies were combined to
develop a geographically and racially diverse dataset used for the ASCVD model estimation.65
The predictors of the ASCVD model include age, TC and HDLC concentrations, systolic BP,
current smoking, diagnosed diabetes, and whether the participant is undergoing treatment for
high BP. The model was fit separately to four population subgroups: non-Hispanic White
females, non-Hispanic Black females, non-Hispanic White males, and non-Hispanic Black
males.
Several studies assessed predictive performance of the ASCVD risk model in racial and ethnic
groups other than other non-Hispanic White and non-Hispanic Black populations, as well as in
various sociodemographic subgroups in the U.S. Two studies concluded that the ASCVD risk
model overestimated CVD risk among Asian and Hispanic groups, while noting that these
groups were not included in the development and validation of the ASCVD model (Mongraw-
Chaffin et al., 2018; Rodriguez et al., 2019). Five studies acknowledged limitations for the
ASCVD risk model in terms of performance among individuals with high levels of CVD risk,
diabetes, older adults with frailty and multimorbidity, smokers, and women (Muntner et al.,
2014; Leigh et al., 2019; Mora et al., 2018; Q. D. Nguyen et al., 2020; Raghavan et al., 2020).
Overall, the literature across different sociodemographic subgroups concluded that the ASCVD
risk model tended to overestimate risk but suggested the model may improve through additional
input variables and recalibration given contemporary ASCVD prevalence, especially if the
prevalence differs significantly across geographic locations to which the model is applied (Mora
et al., 2018; (Muntner et al., 2014). Extended discussion of ASCVD risk model performance and
availability of alternative CVD risk prediction models for national analysis is provided in ICF
(2022a).
In light of these findings, EPA does not follow the Goff et al. (2014) recommendation that the
ASVCD risk model for non-Hispanic White populations be used for other race/ethnicity groups.
In the development and parameterization of the CVD model for Hispanic, Asian American, and
American Indian/Alaska Native populations, EPA applies the model for non-Hispanic Black
populations based on the ASCVD model validation relative to reported CVD prevalence and
mortality statistics (EPA analysis based on Medical Expenditure Panel Surveys from 2010-
2017), as described in Appendix G. The results of this validation exercise showed that the
ASCVD model coefficients for the non-Hispanic Black model are more consistent with data on
CVD prevalence and mortality for Hispanic and non-Hispanic other race subpopulations than the
ASCVD model coefficients for the non-Hispanic White model. The all-cause and CVD mortality
65 These studies include the Atherosclerosis Risk in Communities (ARIC) study (Williams, 1989) and the Cardiovascular Health
Study (Fried et al., 1991), along with applicable data from the Coronary Artery Risk Development in Young Adults (CARDIA)
study (Friedman et al., 1988) and the Framingham Original and Offspring cohort data (D'Agostino et al., 2008).
Proposed PFAS Rule Economic Analysis
6-61
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
was obtained from CDC's National Vital Statistics System, whereas CVD prevalence was
estimated using Agency for Healthcare Research and Quality survey data (see Appendix G for
details). As explained in Appendix G, race/ethnicity and sex-specific CVD incidence consistent
with these reported statistics was compared with the incidence estimated using the ASCVD
model, where the baseline race/ethnicity- and sex-specific values for the ASCVD model
predictors were obtained from CDC's public health surveys (see Appendix G for details).
The ASCVD model generates predictions of the 10-year probability of the first hard CVD event
without differentiation across CVD event types. The specifics of annual first hard CVD event
probability derivation, which is needed for the life table calculations in Section 6.5.3.1, are
provided in Appendix G. As is also detailed in Appendix G, EPA combined the Medical
Expenditure Panel Survey (MEPS) 2010-2017 data and the Healthcare Cost and Utilization
Project (HCUP) 2017 data to derive the ASCVD event distribution over the following event
types: non-fatal MI, non-fatal IS, and fatal CVD events. The fatal CVD events include fatal MI,
fatal IS, and other fatal CHD events. EPA used the MEPS data to identify the subpopulation of
persons without a prior CVD event history and estimate the rate of new CVD events by type (i.e.,
MI, IS, and other CHD) in this subpopulation. The probabilities of in-hospital death for MI, IS,
and other CHD were obtained from HCUP.
Table 6-17 shows the derived race/ethnicity-, sex-, and age group-specific shares of first hard
CVD events for the following event types: non-fatal MI, fatal MI, non-fatal IS, fatal IS, other
non-fatal CHD, and other fatal CHD. For males, looking across race/ethnicity and age categories,
the share of non-fatal MI events is 4.9 percent to 28 percent, the share of non-fatal IS events is
9.4 percent to 38 percent, and the share of other non-fatal CHD events is 44 percent to 78
percent. For females, across race/ethnicity and age categories, the share of non-fatal MI events is
6.4 percent to 19 percent, the share of non-fatal IS events is 8.7 percent to 29 percent, and the
share of other non-fatal CHD events is 51 percent to 76 percent. For both sexes, shares of all
fatal events increase with age. The share of fatal CVD events is largest for Hispanic and non-
Hispanic other race subpopulations of both sexes. Table 6-17 also shows derived race/ethnicity-,
sex-, and age group-specific shares of first hard CVD events over ASCVD event types (i.e., non-
fatal MI, non-fatal IS, and fatal CVD). Note that these shares were re-normalized to sum to 100
percent after exclusion of other non-fatal CHD not predicted by the ASCVD model. The CVD
model relies on the re-normalized shares to allocate the total number of first hard CVD events
predicted by the ASCVD model.
Proposed PFAS Rule Economic Analysis
6-62
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-17: Estimated Shares of Fatal and Non-Fatal First Hard CVD Events Based on
MEPS and HCUP Data
Non-Fatal CVD (%) Fatal CVD (%)
Sex Age (in years) Et^ Non.Fatal Non.Fatal Fatal MI Fatal IS p^HD
MI (%) IS (%) (%) (%)
Shares of First Hard CVD Events
Males
18-44
NH White
14
9.4
77
0.19
0.17
0
45-64
NH White
16
15
69
0.39
0.34
0.44
65-84
NH White
13
20
64
0.71
0.75
0.76
85 or older
NH White
13
20
63
1.3
1.4
1.9
18-44
NH Black
4.9
17
78
0.067
0.31
0
45-64
NH Black
11
38
50
0.28
0.88
0.32
65-84
NH Black
8.9
22
67
0.48
0.8
0.79
85 or older
NH Black
8.5
21
66
0.87
1.5
2
18-44
Hispanic
23
17
59
0.31
0.31
0
45-64
Hispanic
19
29
51
0.48
0.67
0.32
65-84
Hispanic
20
17
60
1.1
0.65
0.71
85 or older
Hispanic
19
17
59
2
1.2
1.8
18-44
NH Other
26
30
44
0.35
0.54
0
45-64
NH Other
28
19
52
0.71
0.43
0.33
65-84
NH Other
13
25
60
0.71
0.92
0.71
85 or older
NH Other
12
24
59
1.3
1.7
1.8
Females
18-44
NH White
8.1
19
72
0.13
0.41
0
45-64
NH White
6.9
20
72
0.2
0.55
0.54
65-84
NH White
11
28
58
0.68
1.2
0.82
85 or older
NH White
10
27
57
1.2
2.3
2.1
18-44
NH Black
15
8.7
76
0.23
0.18
0
45-64
NH Black
10
27
61
0.29
0.74
0.46
65-84
NH Black
6.7
29
62
0.42
1.2
0.87
85 or older
NH Black
6.4
28
61
0.76
2.3
2.2
18-44
Hispanic
00
00
18
73
0.14
0.38
0
45-64
Hispanic
13
27
59
0.37
0.73
0.45
65-84
Hispanic
19
26
52
1.2
1.1
0.73
85 or older
Hispanic
18
25
51
2.1
2.1
1.9
18-44
NH Other
11
13
75
0.17
0.27
0
45-64
NH Other
14
29
55
0.42
0.78
0.42
65-84
NH Other
12
28
58
0.74
1.2
0.81
85 or older
NH Other
11
27
56
1.3
2.3
2.1
Shares of First Hard CVD Event Categories Predicted by the ASCVD Model3
Males
18-44
NH White
58
40
-
1.5
45-64
NH White
50
47
-
3.7
65-84
NH White
37
57
-
6.2
85 or older
NH White
34
53
-
13
18-44
NH Black
22
77
-
1.7
45-64
NH Black
22
75
-
2.9
65-84
NH Black
27
66
-
6.4
85 or older
NH Black
25
62
-
13
18-44
Hispanic
56
42
-
1.5
45-64
Hispanic
38
59
-
3.0
65-84
Hispanic
50
44
-
6.1
85 or older
Hispanic
47
41
-
12
18-44
NH Other
46
53
-
1.6
Proposed PFAS Rule Economic Analysis
6-63
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-17: Estimated Shares of Fatal and Non-Fatal First Hard CVD Events Based on
MEPS and HCUP Data
Sex
Age (in years)
Race/
Ethnicity
Non-Fatal CVD (%)
Fatal CVD (%)
Non-Fatal
MI (%)
Non-Fatal
IS (%)
Other Non-
Fatal CHD
(%)
Fatal MI Fatal IS
(%) (%)
Other
Fatal CHD
(%)
45-64
NH Other
58
39
-
3.1
65-84
NH Other
33
62
-
5.8
85 or older
NH Other
30
58
-
12
Females
18-44
NH White
29
69
-
1.9
45-64
NH White
24
71
-
4.6
65-84
NH White
26
67
-
6.5
85 or older
NH White
24
63
-
13
18-44
NH Black
62
36
-
1.7
45-64
NH Black
26
70
-
3.9
65-84
NH Black
18
76
-
6.7
85 or older
NH Black
16
70
-
14
18-44
Hispanic
32
66
-
1.9
45-64
Hispanic
31
65
-
3.8
65-84
Hispanic
40
54
-
6.4
85 or older
Hispanic
37
51
-
12
18-44
NH Other
45
53
-
1.8
45-64
NH Other
32
64
-
3.6
65-84
NH Other
28
66
-
6.5
85 or older
NH Other
26
61
-
13
Abbreviations: CVD - cardiovascular disease; CHD - coronary heart disease; fatal CVD - includes fatal MI, fatal IS, and fatal
other coronary heart disease events; HCUP - Healthcare Cost and Utilization Project; IS - ischemic stroke; MEPS - Medical
Expenditure Panel Survey; MI - myocardial infarction; NH - non-Hispanic.
Note:
aThe distribution is derived by (1) excluding the other non-fatal CHD category; (2) aggregating fatal MI, fatal IS, and other fatal
CHD categories into the fatal CVD category; and (3) re-normalizing the data to sum to 100%.
6.5.3.3 Risk of Post-Acute Cordiovosculor Disease Mortality
Persons who have experienced non-fatal MI and non-fatal IS have an elevated risk of post-acute
CVD mortality and morbidity (Roger et al., 2012). Studies focusing on secondary hard CVD
events point to an elevated risk of these events among survivors of the first hard CVD event
(e.g., Beatty et al., 2015; S. Li et al., 2019; Thorn et al., 2001), but do not support the link
between these risks and TC/BP levels (Beatty et al., 2015). (See Appendix G for details.)
Therefore, the CVD model evaluates post-acute CVD mortality among survivors of the initial
MI/IS event under baseline and regulatory alternatives using the baseline post-acute mortality
rates that do not depend on the levels of modeled biomarkers. The CVD model does not
explicitly evaluate secondary CVD morbidity because available first non-fatal MI/IS valuation
measures (e.g., O'Sullivan et al., 2011) incorporate incidence of these secondary events.
For survivors of the first hard CVD event at ages 40-65, EPA uses estimates of sex- and
race/ethnicity-specific all-cause post-acute mortality for MI survivors at 1- and 5-year follow-up
from Thom et al. (2001). Because Thom et al. (2001) reports all-cause post-acute mortality rates,
EPA adjusted these rates to exclude deaths from non-CVD causes. To this end, EPA used
general population integer age- and sex-specific all-cause mortality from U.S. Life Tables, 2017
(Arias et al., 2019), U.S. CVD mortality rates (CDC, 2020b), and U.S. Life Tables Eliminating
Proposed PFAS Rule Economic Analysis
6-64
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Certain Causes of Death, 1999-2000 (Arias et al., 2013). Appendix G provides additional
estimation details. Although EPA was unable to identify comparable post-acute mortality
statistics for non-fatal IS, an analysis of the Medicare population by S. Li et al. (2019) suggests
that post-acute MI mortality is a reasonable approximation for post-acute IS mortality.66 Table
6-18 shows estimated post-acute CVD mortality rates for survivors of the first MI or IS at ages
40-65 that are used to parameterize the CVD model.
For survivors of the first hard CVD event at ages 66+, EPA uses the results in S. Li et al.
(2019) to estimate the number of post-acute CVD deaths within 6 years of the initial event.
Because S. Li et al. (2019) reports only all-cause post-acute mortality rates, EPA adjusted these
rates to exclude deaths from non-CVD causes. Integer age- and sex-specific probability of death
from non-CVD causes was derived from U.S. Life Tables, 2017 (Arias et al., 2019), U.S. CVD
mortality rates (CDC, 2020b), and U.S. Life Tables Eliminating Certain Causes of Death, 1999-
2000 (Arias et al., 2013). The sex-specific probabilities of death from non-CVD causes were
average using the demographic information for the cohorts analyzed by S. Li et al. (2019). See
Appendix G for additional estimation details. Table 6-18 shows estimated post-acute CVD
mortality rates for survivors of the first MI and survivors of the first IS at ages 66 years or older
that are used to parameterize the CVD model.67
66 For those age 65 or older, S. Li et al. (2019) have estimated the probability of death within 1 year after non-fatal IS to be 32.07
percent and the probability of death within 1 year after non-fatal MI to be 32.09 percent.
67 These rates are applied to all those aged 66+ in the SafeWater MCBC implementation of the model.
Proposed PFAS Rule Economic Analysis 6-65 March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-18: Estimated Risk of Post-Acute CVD Mortality Following the First Non-Fatal
Hard CVD Event
Type of
First Non- _ . . „
Fatal Hard Demographic Group
CVD Event
Source: Thom et al. (2001)
Non-Hispanic White b males aged
45-65 years
4,500
910
860
820
760
-
MI, ISa
Non-Hispanic Black males aged
45-65 years
12,000
1,200
1,100
1,100
1,000
-
Non-Hispanic White b females
aged 45-65 years
8,600
1,900
1,900
1,900
1,800
-
Non-Hispanic Black females aged
45-65 years
7,700
4,300
4,200
4,100
4,100
-
Source: S. Li et al. (2019)
MI
Persons aged 66+ years
27,000
11,000
9,600
9,040
8,600
8,040
IS
Persons aged 66+ years
28,000
9,900
10,000
9,800
8,900
8,030
Abbreviations: CVD - cardiovascular disease; IS - ischemic stroke (International Classification of Disease Ninth Revision
[ICD9]=433,434; International Classification of Disease Tenth Revision [ICD10]=I63), MI - myocardial infarction
(ICD9=410; ICD10=I21).
Notes:
aThom et al. (2001) reported data for the first MI survivors only for aged 45-64 years. The CVD model applies these rates to
both the first MI and first IS survivors.
bEstimates for non-Hispanic White populations are applied to other ethnic groups.
6.5.4 Valuation of Cardiovascular Disease Risk Reductions
EPA uses the Value of Statistical Life to estimate the benefits of reducing mortality associated
with hard CVD events in the population exposed to PFOA and PFOS in drinking water. Value of
Statistical Life updating information is provided in Section 2.2. EPA relies on COI-based
valuation that represents the medical costs of treating or mitigating non-fatal first hard CVD
events (MI, IS) during the three years following an event among those without prior CVD
history, adjusted for post-acute mortality.
The annual medical expenditure estimates for MI and IS are based on O'Sullivan et al. (2011).
The estimated expenditures do not include long-term institutional and home health care. For non-
fatal MI, O'Sullivan et al. (2011) estimated medical expenditures are $51,173 ($2021) 68 for the
initial event and then $31,871, $14,065, $12,569 annually within 1, 2, and 3 years after the initial
event, respectively. For non-fatal IS, O'Sullivan et al. (2011) estimated medical expenditures are
$15,861 ($2021) for the initial event and then $11,521, $748, $1,796 annually within 1, 2, and 3
years after the initial event, respectively. Annual estimates within 1, 2, and 3 years after the
initial event include the incidence of secondary CVD events among survivors of first MI and IS
events.
To estimate the present discounted value of medical expenditures within 3 years of the initial
non-fatal MI, EPA combined O'Sullivan et al. (2011) Mi-specific estimates with post-acute
08 Original values from the source were inflated to $2019 using the medical care Consumer Price Index (U.S. Bureau of Tabor
Statistics, 2021).
Post-Acute CVD Mortality Rate per 100,000 by Integer Year
Since the First Non-Fatal Hard CVD Event
0 1 2 3 4 5
Proposed PFAS Rule Economic Analysis
6-66
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
survival probabilities based on Thom et al. (2001) (for MI survivors aged 40-64) and S. Li et al.
(2019) (for MI survivors aged 65+). To estimate the present discounted value of medical
expenditures within 3 years of the initial non-fatal IS, EPA combined O'Sullivan et al. (2011) IS-
specific estimates with post-acute survival probabilities based on Thom et al. (2001) (for IS
survivors aged 40-64, assuming post-acute MI survival probabilities reasonably approximate
post-acute IS survival probabilities) and S. Li et al. (2019) (for IS survivors aged 65+). EPA did
not identify post-acute IS mortality information in this age group, but instead applied post-acute
MI mortality estimates for IS valuation.69 Table 6-19 presents the resulting MI and IS unit
values.
Table 6-19: Cost of Illness of Non-Fatal First CVD Event Used in Modeling
Present Discounted Value of 3-Year Medical Expenditures
Type of First Non fatal ($2021)a b Adjusted for Post-Acute Mortality0
Hard CVD Event Age Gr0up
3% discount rate 7% discount rate
MI 40-64 years $105,419 $104,155
65+years $92,658 $91,881
IS 40-64 years $29,154 $29,017
65+ years $26,844 $26,762
Abbreviations: CVD - cardiovascular disease; MI - myocardial infarction (ICD9=410; ICD10=I21), IS - ischemic stroke
(ICD9=433,434; ICD10=I63).
Notes:
Estimates of annual medical expenditures are from O'Sullivan et al. (2011).
bOriginal values from O'Sullivan et al. (2011) were inflated to $2021 using the medical care Consumer Price Index (U.S.
Bureau of Labor Statistics, 2021).
cPost-acute MI mortality data for those aged 40-64 years is from Thom et al. (2001); probabilities to survive 1 year, 2 years, and
3 years after the initial event are 0.93, 0.92, and 0.90, respectively. EPA applies these mortality values to derive the IS value in
this age group. Post-acute MI mortality data and post-acute IS mortality data for persons aged 65 years and older are from S. Li
et al. (2019). For MI, probabilities to survive 1 year, 2 years, and 3 years after the initial event are 0.68, 0.57, and 0.49,
respectively. For IS, probabilities to survive 1 year, 2 years, and 3 years after the initial event are 0.67, 0.57, and 0.48,
respectively.
69 Post-acute mortality estimates for IS and MI were very close in the Medicare population (S. Li et al., 2019). For those ages 65
years or older, Li et al. (2019) have estimated probability of death within 1 year after non-fatal IS to be 32.07 percent and
probability of death within 1 year after non-fatal MI to be 32.09 percent. Therefore, reliance on the post-acute mortality for MI to
approximate the same for stroke is reasonable.
Proposed PFAS Rule Economic Analysis
6-67
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.5.5 Results
Table 6-20 to Table 6-23 provide the health effects avoided and valuation associated with
cardiovascular disease.
Table 6-20: National CVD Benefits, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt
and HI of 1.0)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
MI Cases Avoided
1,251.5
6,081.0
11,738.7
1,251.5
6,081.0
11,738.7
Number of Non-Fatal
IS Cases Avoided
1,814.0
8,870.8
17,388.5
1,814.0
8,870.8
17,388.5
Number of CVD
Deaths Avoided
753.6
3,584.6
7,030.9
753.6
3,584.6
7,030.9
Total Annualized
CVD Benefits
(Million $2021)b
$111.78
$533.48
$1,051.00
$85.94
$421.10
$822.88
Abbreviations: CVD - cardiovascular disease, MI - myocardial infarction, IS - Ischemic Stroke.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-21: National CVD Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
MI Cases Avoided
1,248.7
5,983.8
11,614.9
1,248.7
5,983.8
11,614.9
Number of Non-Fatal
IS Cases Avoided
1,786.4
8,729.6
17,149.5
1,786.4
8,729.6
17,149.5
Number of CVD
Deaths Avoided
744.6
3,527.8
6,951.5
744.6
3,527.8
6,951.5
Total Annualized
CVD Benefits
(Million $2021)b
$110.45
$525.05
$1,035.36
$86.32
$414.45
$817.79
Abbreviations: CVD - cardiovascular disease, MI - myocardial infarction, IS - Ischemic Stroke.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-68
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-22: National CVD Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
MI Cases Avoided
1,105.9
5,220.7
10,215.4
1,105.9
5,220.7
10,215.4
Number of Non-Fatal
IS Cases Avoided
1,609.3
7,624.2
15,029.5
1,609.3
7,624.2
15,029.5
Number of CVD
Deaths Avoided
645.9
3,084.6
6,102.2
645.9
3,084.6
6,102.2
Total Annualized
CVD Benefits
(Million $2021)b
$99.73
$459.09
$908.82
$72.72
$362.42
$717.85
Abbreviations: CVD - cardiovascular disease, MI - myocardial infarction, IS - Ischemic Stroke.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
Table 6-23: National CVD Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
MI Cases Avoided
619.0
3,032.5
6,320.7
619.0
3,032.5
6,320.7
Number of Non-Fatal
IS Cases Avoided
878.1
4,445.9
9,439.4
878.1
4,445.9
9,439.4
Number of CVD
Deaths Avoided
343.8
1,806.7
3,835.8
343.8
1,806.7
3,835.8
Total Annualized
CVD Benefits
(Million $2021)b
$51.00
$268.78
$571.32
$41.85
$212.18
$450.51
Abbreviations: CVD - cardiovascular disease, MI - myocardial infarction, IS - Ischemic Stroke.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-69
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.6 Renal Cell Carcinoma
6.6.1 Overview of the RCC Risk Reduction Analysis
Figure 6-9 illustrates the approach used to quantify and value the changes in RCC risk associated
with lowered serum PFOA levels from reductions in drinking water PFOA concentrations under
the regulatory alternatives. Section 4.4 and Section 6.3 detail the PWS EP-specific PFOA
drinking water occurrence estimation and modeling of serum PFOA concentrations, respectively.
PWS EP-specific time series of the differences between serum PFOA concentrations under
baseline and regulatory alternatives are inputs into this analysis. For each PWS EP, evaluation of
the changes in RCC impacts involves the following key steps:
1. Estimating the changes in RCC risk based on modeled changes in serum PFOA levels
and the exposure-response function for the effect of serum PFOA on RCC;
2. Estimating the annual incidence of RCC cases and excess mortality among those with
RCC in all populations corresponding to baseline and regulatory alternative RCC risk
levels, as well as estimating the regulatory alternative-specific reduction in cases relative
to the baseline; and
3. Estimating the economic value of reducing RCC mortality from baseline to regulatory
alternative levels, using the Value of Statistical Life and COI measures, respectively.
Section 6.6.2 discusses the exposure-response modeling for RCC. Section 6.6.3 summarizes the
life table-based approach for estimation of RCC risk reductions. Section 6.6.4 discusses EPA's
valuation methodology for RCC mortality and morbidity. Section 6.6.5 presents the results of the
analysis.
Proposed PFAS Rule Economic Analysis
6-70
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Legend:
Result of
upstream analysis
Serum PFOA
concentration difference
between baseline and
regulatory alternative
Data/Inputs
Model
# Analysis step
Location-specific
population size
National Cancer
Institute SEER
program data
Annual cause-specific
mortality rates and life
table information3
Medical cost of RCC
treatment
Change in the
number of RCC
cases
\
3
r
Value of reduced
RCC cases
Change in RCC
population
mortality
Value of avoided
excess mortality
Value of a statistical
life
i
Total value of reduced RCC
Abbreviations:
PFOA - perfluorooctanoic acid, RCC - renal cell carcinoma, SEER - Surveillance, Epidemiology, and End Results program
Notes:
(a) Data from the Centers for Disease Control (CDC) and Prevention.
Figure 6-9: Overview of Analysis of Reduced RCC Risk
Proposed PFAS Rule Economic Analysis
6-71
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.6.3 RCC Exposure-Response Modeling
To identify an exposure-response function, EPA reviewed studies highlighted in the HESD for
PFOA (U.S. EPA, 2016f) and a recent study discussed in both the California Environmental
Protection Agency's Office of Environmental Health Hazard Assessment (OEHHA) PFOA
Public Health Goals report (CalEPA, 2021) and EPA's Toxicity Assessment and Proposed
Maximum Contaminant Level Goal for PFOA in Drinking Water (U.S. EPA, 2023e). Steenland
et al. (2015) observed an increase in kidney cancer deaths among workers with high exposures to
PFOA. Vieira et al. (2013) found that kidney cancer was positively associated with high and very
high PFOA exposures. Barry et al. (2013) found a slight trend in cumulative PFOA serum
exposures and kidney cancer among the C8 Health Project population.70 In a large case-control
general population study of the relationship between PFOA and kidney cancer in 10 locations
across the U.S., Shearer et al. (2021) found strong evidence that exposure to PFOA causes RCC,
the most common form of kidney cancer, in humans.
To evaluate changes between baseline and regulatory alternative RCC risk resulting from
reduced exposure to PFOA, EPA relied on the estimated time series of changes in serum PFOA
concentrations (Section 6.3) and the serum-RCC exposure-response function provided by
Shearer et al. (2021): 0.00178 (95% CI: 0.00005, 0.00352) per ng/mL. The analysis from Shearer
et al. (2021) was designed as a case-control study with population controls based on 10 sites
within the U.S. population. Shearer et al. (2021) included controls for age, sex, race, ethnicity,
study center, year of blood draw, smoking, and hypertension. Results showed a strong and
statistically significant association between PFOA and RCC. EPA selected the exposure-
response relationship from Shearer et al. (2021) because it included exposure levels typical in the
general population and was found to have a low risk of bias when assessed in EPA's Toxicity
Assessment and Proposed Maximum Contaminant Level Goal for PFOA in Drinking Water (U.S.
EPA, 2023e).
The linear slope factor based on Shearer et al. (2021) enables estimation of the changes in the
lifetime RCC risk associated with reduced lifetime serum PFOA levels:
Equation 18:
LR{x) = LR(z) + 0.00178 ¦ (x - z)
Where LR(x) is the probability of lifetime RCC incidence for an individual exposed to a lifetime
average serum PFOA concentration of x ng/mL, and LR(z) is the probability of lifetime RCC at
the baseline lifetime average serum PFOA concentration of z ng/mL.
Because baseline RCC incidence statistics are not readily available from the National Cancer
Institute public use data, EPA used kidney cancer statistics in conjunction with an assumption
that RCC comprises 90 percent of all kidney cancer cases to estimate baseline lifetime
probability of RCC (U.S. EPA, 2023e). EPA estimated the baseline lifetime RCC incidence for
males at 1.89 percent and the baseline lifetime RCC incidence for females at 1.05 percent.
Details of these calculations are provided in Appendix H. Because the Shearer et al. (2021) slope
70 The C8 Health Project collected data to ascertain the amount of C8 (otherwise known as PFOA) in blood among Mid-Ohio
Valley communities from 2005-2013. Mean PFOA at enrollment 24 ng/mL.
Proposed PFAS Rule Economic Analysis
6-72
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
factor is not sex-specific, EPA averaged sex-specific baseline lifetime RCC estimates to obtain
LR(z) = 0.0147 for use in the estimation of annual RCC risk changes.
To enable annual RCC risk estimation, EPA further assumed that the relative risk relationship
implied by Equation 18, i.e., RR(x,z) = LR(x)/LR(z) = 1 + 0.00178 ¦ (x — z)/L/?(z) = 1 +
0.00178 ¦ (x — z)/0.0147, also holds for the cumulative RCC risk and cumulative average
exposure to serum PFOA from birth to a specific age.
A person's cumulative serum PFOA exposure by age a—denoted by xa—is defined as:
EPA estimated the relative risk of RCC by a particular age from a change in average serum
PFOA experienced by this age as follows:
Where RR(xa, za) is the relative cumulative risk of RCC by age a associated with a change from
baseline cumulative exposure za to treatment cumulative exposure xa and PAF is the
environmental exposure-related population attributable fraction of RCC incidence set at 0.0394.
As such, this equation implies that EPA caps the magnitude of PFOA-related cumulative RCC
risk reduction at the PAF of 3.94 percent to ensure plausibility of the estimated RCC benefits
size. EPA developed this PAF estimate based on its review of literature on environmental
contaminant-attributable risk estimates for cancers (ICF, 2022b). In calculations of the annual
RCC risk changes, EPA continued to assume that RCC comprises 90 percent of annual kidney
cancer incidence.
EPA relies on the life-table approach to estimate RCC risk reductions because:
• Changes in serum PFOA in response to changes in drinking water PFOA occur over
multiple years;
• Annual risk of new RCC should be quantified only among those not already experiencing
this chronic condition;
• RCC has elevated mortality implications.
EPA used recurrent life table calculations to estimate PWS EP-specific time series of RCC
incidence for a population cohort characterized by sex, race/ethnicity, birth year, and age at the
beginning of the evaluation period (i.e., 2023) under the baseline scenario and the regulatory
alternatives. The life-table analysis accounts for the gradual changes in lifetime exposures to
Proposed PFAS Rule Economic Analysis 6-73 March 2023
Equation 19:
1 V-
Equation 20:
6.6.3 Estimation of RCC Risk Reductions
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
PFOA following implementation of treatment under the regulatory alternatives compared to the
baseline.71 Details of the life-table calculations are provided in Appendix H. The outputs of the
life-table calculations are the PWS EP-specific estimates of the annual change in the number of
RCC cases and the annual change in RCC population mortality.
Although the change in PFOA exposure likely affects the risk of developing RCC beyond the
end of the analysis period (the majority of RCC cases manifest during the latter half of the
average individual lifespan; see Appendix H), EPA does not capture effects after the end of the
period of analysis, 2104. Individuals alive after the end of the period of analysis likely benefit
from lower lifetime exposure to PFOA. Lifetime health risk model data sources include EPA
SDWIS; age-, sex-, and race/ethnicity-specific population estimates from the U.S. Census
Bureau (U.S. Census Bureau, 2020a); the Surveillance, Epidemiology, and End Results (SEER)
program database (National Cancer Institute),72 and the CDC National Center for Health
Statistics.73 Appendix H provides additional detail on the data sources and information used in
this analysis as well as baseline kidney cancer statistics. Appendix B describes estimation of the
affected population.
6.6.4 Valuation of RCC Risk Reductions
EPA uses the Value of Statistical Life to estimate the benefits of reducing mortality associated
with RCC in the population exposed to PFOA in drinking water. Section 2.2 provides
information on updating Value of Statistical Life for inflation and income growth. EPA uses the
COI-based valuation to estimate the benefits of reducing morbidity associated with RCC.
EPA used the medical cost information from a recent RCC cost-effectiveness study by
Ambavane et al. (2020) to develop COI estimates for RCC morbidity. Ambavane et al. (2020)
used a discrete event simulation model to estimate the lifetime treatment costs of several RCC
treatment sequences, which included first and second line treatment74 medication costs,
medication administration costs, adverse effect management costs, and disease management
costs on- and off-treatment. To this end, the authors combined RCC cohort data from a
CheckMate 214 clinical trial and recent US-based healthcare cost information assembled from
multiple sources (see supplementary information from Ambavane et al. (2020)).
Table 6-24 summarizes RCC morbidity COI estimates derived by EPA using Ambavane et al.
(2020)-reported disease management costs on- and off-treatment along with medication,
administration, and adverse effect management costs for the first line treatment that initiated the
most cost-effective treatment sequences as identified by Ambavane et al. (2020), i.e., the
nivolumab / ipilimumab drug combination. This is a forward-looking valuation approach in that
it assumes that the clinical practice would follow the treatment recommendations in Ambavane
et al. (2020) and other recent studies cited therein. EPA notes that the second line treatment costs
71 As described above, EPA models PFAS changes under the regulatory alternatives as being in effect for the years 2023 through
2104, with nonzero PFAS changes first occurring in 2026, the year when all PWSs are assumed to comply with PFAS treatment
requirements.
72 For cancer incidence and stage distribution data, EPA relies on SEER 21 (2009-2018); for cancer survival data, EPA relies on
SEER 18 (2000-2017).
73 CDC WONDER data on 1999-2019 all-cause and kidney cancer mortality by age and sex.
74 Second line cancer treatment is a treatment implemented after the failure of the initial treatment (i.e., first line treatment). The
first line treatment may fail because it stops working or has side effects that are not tolerated.
Proposed PFAS Rule Economic Analysis
6-74
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
are not reflected in EPA's COI estimates, because Ambavane et al. (2020) did not report
information on the expected durations of the treatment-free interval (between the first line
treatment discontinuation and the second line treatment initiation) and the second line treatment
phase, conditional on survival beyond discontinuation of the second line treatment. As such,
EPA valued RCC morbidity at $251,007 ($2021) during year 1 of the diagnosis, $190,969
($2021) during year 2 of the diagnosis, and $1,596 ($2021) starting from year 3 of the diagnosis.
Additionally, EPA assumed that for individuals with RCC who die during the specific year, the
entire year-specific cancer treatment regimen is applied prior to the death event. This may
overestimate benefits if a person does not survive the entire year.
Proposed PFAS Rule Economic Analysis
6-75
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-24: RCC Morbidity Valuation
^ tmm
.2 ^ « a S
-S ° £ 8 2
^ ^ S c S
• S- flJ £»~ S ^
^ fig *® ©X «' j2f '
T + . is 3 .2 ^ S? S s Total Total
Time Interval ^ ® +- -b J S3 5 *2 ®
Montlily cost, month 25+
from diagnosis8
Annual cost, year 3+ from
diagnosis
£
a
0©
0
QJ
O
fN
&
-
>
c
13
fi
O
<
e
s
*-
U
c
-
£
Urn
<
I g | a .S I | g g ($2018) ($2021)d
O
tc
w
a>
es
a>
Monthly cost, month 1-3 g5 M6 ?8 ?3 33 152 35,927
from diagnosis3'6
Monthly cost, month 4-24 8? 64? ?8 ?3 g5 15
from diagnosis0'1
123 123 133
Annual cost, year 1 from 222,438 7,371 934 878 231,621 251,007
diagnosis
Annual cost, year 2 from ^ 9M 8?8 Q
diagnosis
190,969
1,473 1,473 1,596
Abbreviations: RCC - renal cell carcinoma.
Notes:
aAmbavane et al. (2020) Table 1;
bAmbavane et al. (2020) p. 41, a maximum treatment duration assumption of 2 years;
The adverse effect management costs of $1,868 in Ambavane et al. (2020) Table 1 were reported for the treatment duration.
EPA used the treatment duration of 24 months (i.e., 2 years) to derive montlily costs of $77.83;
dTo adjust for inflation, EPA used U.S. Bureau of Tabor Statistics Consumer Price Index for All Urban Consumers: Medical
Care Services in U.S. (City Average).
eFirst line treatment induction
Tirst line treatment maintenance
gTreatment-free interval
6.6.5 Results
Table 6-25 to Table 6-28 provide the health effects avoided and valuation associated with renal
cell carcinoma.
Proposed PFAS Rule Economic Analysis
6-76
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-25: National RCC Benefits, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt
and HI of 1.0)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal RCC
Cases Avoided
1,313.6
6,872.0
17,387.8
1,313.6
6,872.0
17,387.8
Number of RCC-Related
Deaths Avoided
308.7
1,927.8
5,049.3
308.7
1,927.8
5,049.3
Total Annualized RCC
Benefits (Million $2021)b
$54.23
$300.56
$758.03
$45.36
$217.37
$515.89
Abbreviations: RCC - Renal Cell Carcinoma.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-26: National RCC Benefits, Option la (PFOA and PFOS MCLs of 4.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal RCC
Cases Avoided
1,289.6
6,753.3
17,147.8
1,289.6
6,753.3
17,147.8
Number of RCC-Related
Deaths Avoided
300.5
1,895.2
4,960.4
300.5
1,895.2
4,960.4
Total Annualized RCC
Benefits (Million $2021)b
$52.92
$295.53
$744.64
$45.09
$213.78
$508.56
Abbreviations: RCC - Renal Cell Carcinoma.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-77
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-27: National RCC Benefits, Option lb (PFOA and PFOS MCLs of 5.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal RCC
Cases Avoided
1,017.6
5,681.7
14,962.1
1,017.6
5,681.7
14,962.1
Number of RCC-Related
Deaths Avoided
235.9
1,602.1
4,317.6
235.9
1,602.1
4,317.6
Total Annualized RCC
Benefits (Million $2021)b
$42.28
$250.60
$643.71
$36.32
$182.24
$446.80
Abbreviations: RCC - Renal Cell Carcinoma.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-28: National RCC Benefits, Option lc (PFOA and PFOS MCLs of 10.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal RCC
Cases Avoided
433.5
2,903.0
8,205.4
433.5
2,903.0
8,205.4
Number of RCC-Related
Deaths Avoided
101.1
831.8
2,406.2
101.1
831.8
2,406.2
Total Annualized RCC
Benefits (Million $2021)b
$18.58
$131.44
$367.38
$17.34
$97.30
$260.54
Abbreviations: RCC - Renal Cell Carcinoma.
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
6.7 Benefits from Co-Removal of Disinfection Byproducts
As part of its health risk reduction and cost analysis, EPA is directed by SDWA to evaluate
quantifiable and nonquantifiable health risk reduction benefits for which there is a factual basis
in the rulemaking record to conclude that such benefits are likely to occur from reductions in co-
occurring contaminants that may be attributed solely to compliance with the maximum
contaminant level (SDWA 1412(b)(3)(C)(II)). These co-occurring contaminants are expected to
include additional PFAS contaminants not directly regulated by the proposed PFAS NPDWR,
co-occurring chemical contaminants such as SOCs, VOCs, and DBP precursors. In this section,
Proposed PFAS Rule Economic Analysis
6-78
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA presents a quantified estimate of the reductions in DBP formation potential that are likely to
occur as a result of compliance with the proposed PFAS NPDWR.75
6.7.1 Overview of Reduced Disinfection Byproduct Formation
DBPs are formed when disinfectants react with naturally occurring materials in water. Under the
Stage 2 Disinfectants and Disinfection Byproducts Rule (Stage 2 DBP Rule, U.S. EPA, 2006b),
EPA regulates 11 individual DBPs from three subgroups: four trihalomethanes, five haloacetic
acids, and two inorganic compounds (bromate and chlorite). Under the Stage 2 DBP Rule,
compliance is based on a locational running annual average (LRAA) calculation, where the
annual average at each sampling location in the distribution system is used to determine
compliance with the MCL of 0.08 mg/L for THM4 (regulated as TTHM, bromodichloromethane,
bromoform, chloroform, and dibromochloromethane). There is a substantial body of literature on
DBP precursor occurrence and THM4 formation mechanisms in drinking water treatment. The
formation of THM4 in a particular drinking water treatment plant is a function of several factors
including disinfectant type, disinfectant dose, bromide concentration, organic material type and
concentration, temperature, pH, and system residence times. Epidemiology studies have shown
that THM4 exposure, a surrogate for chlorinated drinking water, is associated with an increased
risk of bladder cancer, among other diseases (Cantor et al., 1998; Cantor et al., 2010; Costet et
al., 2011; Freeman et al., 2017; King et al., 1996; Regli et al., 2015; Villanueva et al., 2004;
Villanueva et al., 2006) and U.S. EPA (2019d). These studies considered THM4 as surrogate
measures for DBPs formed from the use of chlorination that may co-occur. Reductions in
exposure to THM4 is expected to yield significant public health benefits (Regli et al., 2015). In
what Richardson (2022) describes as the "largest risk assessment of DBPs in the U.S. to date,
focusing on bladder cancer cases associated with chlorinated drinking water", Weisman et al.
(2022) estimated that 8,000 of 79,000 national cases of bladder cancer are attributable to DBPs
in drinking water.
EPA used the following data sources for the DBP co-removal analysis (see Table 6-29).
Table 6-29: Data Sources and How the Information Derived from each Source is Used in
the DBP Co-Removal Analysis
Data Source Acronym How Specific Data was Used in Analysis
• Identify GAC treatment start date/year.
• Identify intended purpose for GAC treatment.
• Estimate baseline THM4 (four regulated
Consumer Confidence Reports CCR trihalomethanes) concentrations at systems when
SYR4 data were unavailable.
• Calculate THM4 reduction at systems when SYR4
data were unavailable.
75
The methodology detailed in Section 6.7.1 on estimated DBP reductions was externally peer reviewed by three experts in
GAC treatment for PFAS removal and DBP formation potential. Hie external peer reviewers supported EPA's approach and edits
based on their recommendations for clarity and completeness are reflected in the following analysis and discussion. Some peer
reviewer comments suggested EPA provide additional baseline data summaries for TOC and THM4 occurrence information.
EPA will include these additional summaries in the EA for the final rule.
Proposed PFAS Rule Economic Analysis
6-79
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-29: Data Sources and How the Information Derived from each Source is Used in
the DBP Co-Removal Analysis
Data Source
Acronym
How Specific Data was Used in Analysis
DBP Information Collection
Rule Treatment Study
Database
DBP ICR TSD
• Estimate changes in THM4 levels based on
implementing GAC treatment.
DBP ICR Aux 1 (1998)
Aux 1
• Evaluate changes in DBP precursor occurrence over
time by comparing TOC data to SYR3 TOC data.
Six-Year Review 3,
Information Collection Rule
(2011)
SYR3 ICR
• Evaluate raw water TOC data.
Six-Year Review 4,
Information Collection Rule
SYR4 ICR
• Evaluate raw water TOC data.
• Estimate baseline THM4 concentrations.
(2019)
• Calculate THM4 reductions.
Unregulated Contaminant
Monitoring Rule 3
UCMR3
• Inform a Bayesian occurrence model to identify
PWSs expected to implement treatment under the
NPDWR.
• Identify PWSIDs that had a detectable level of
PFOA and/or PFOS to identify systems used in
trihalomethane reduction comparison.
Unregulated Contaminant
Monitoring Rule 4
UCMR4
• Identify plants that indicated GAC treatment.
• Inform disinfectant type.
Abbreviations: THM4 - Four Regulated Trihalomethanes; DBP - disinfection byproduct; NPDWR - National Primary
Drinking Water Regulation; PWS - public water system; PWSID - Public water system identifier; SYR - Six Year Review;
GAC - Granular Activated Carbon; PFOS - perfluorooctane sulfonic acid; PFOA - perfluorooctanoic acid.
6.7.1.1 Overview of PFAS Treatment with Disinfection Byproduct
Reduction
GAC adsorption has been used to remove synthetic organic chemicals, taste and odor
compounds, and natural organic matter (NOM) during drinking water treatment (Chowdhury et
al., 2013). Recently, many water utilities have installed or are considering installing GAC and/or
other advanced technologies as a protective or mitigation measure to remove various
contaminants of emerging concern, such as PFAS (Dickenson et al., 2016). Because NOM often
exists in a much higher concentration (in mg/L) than trace organics (in [j,g/L or ng/L) in water,
NOM, often measured as TOC, can interfere with the adsorption of trace organics by
outcompeting the contaminants for adsorption sites and by general fouling (blockage of
adsorption pores) of the GAC.
NOM and inorganic matter are precursors for the formation of THMs and other DBPs when
water is disinfected using chlorine and other disinfectants to control microbial contaminants in
finished drinking water. Removal of DBP precursors through adsorption onto GAC has been
included as a treatment technology for compliance with the existing DBP Rules and is a best
available technology (BAT) for the Stage 2 DBP Rule. Dissolved organic matter can be removed
Proposed PFAS Rule Economic Analysis
6-80
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
by GAC through adsorption and biodegradation (Crittenden et al., 1993; Kim et al., 1997;
Yapsakli et al., 2010). Upon startup, the initial removal is via adsorption of the DBP precursors;
GAC is well-established for removal of THM and HAA precursors (Dastgheib et al., 2004;
Cheng et al., 2005; Iriarte-Velasco et al., 2008; Summers et al., 2013; Cuthbertson et al., 2019;
L. Wang et al., 2019). However, biodegradation becomes the predominant mechanism over time
as adsorption capacity is exhausted and microbial growth within the GAC column establishes
itself (Speitel Jr et al., 1989; Velten et al., 2007). In addition to removal of organic DBPs, GAC
also exhibits some capacity for removal of inorganic DBPs such as bromate and chlorite (Kirisits
et al., 2000; Sorlini et al., 2005) and removal of preformed organic DBPs via adsorption and
biodegradation (Jiang et al., 2017; Terry et al., 2018). Further, GAC may offer limited removal
of dissolved organic nitrogen (Chili et al., 2012).
Based on an extensive review of published literature in sampling studies where both contaminant
groups (PFAS and DBPs) were sampled, few direct studies examined both PFAS and DBP co-
removal. To help inform its economic analysis, EPA relied on the DBP Information Collection
Rule Treatment Study Database and DBP formation studies to estimate reductions in THM4
(ATHM4) that may occur when GAC is used to remove PFAS. Subsequently, these results were
compared to THM4 data from PWSs that have detected PFAS and have indicated use of GAC.
The objective of the co-removal benefits analysis is to determine the reduction in bladder cancer
cases associated with the decrease of regulated THM4 in treatment plants due to the installation
of GAC for PFAS removal. Figure 6-10 illustrates EPA's approach for quantifying the human
health benefits of reducing THM4 levels in drinking water. The analysis entails:
1. Estimating the number of systems expected to install GAC treatment in compliance with
the proposed PFAS NPDWR and affected population size;
2. Estimating changes in THM4 levels that may occur when GAC is installed for PFAS
removal based on influent TOC levels;
3. Estimating changes in the cumulative risk of bladder cancer using an exposure-response
function linking lifetime risk of bladder cancer to THM4 concentrations in residential
water supply (Regli et al., 2015);
4. Estimating annual changes in the number of bladder cancer cases and mortality in the
bladder cancer population corresponding to changes in THM4 levels under the regulatory
alternative in all populations alive during or born after the start of the evaluation period;
5. Estimating the economic value of reducing bladder cancer morbidity and mortality from
baseline to regulatory alternative levels, using COI measures and the VSL, respectively.
Proposed PFAS Rule Economic Analysis
6-81
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Legend:
1 Public drinking
Abbreviations: THM4 = Four Regulated Species of Trihalomethanes; SEER = Surveillance, Epidemiology,
and End Results; TOC = Total Organic Carbon
Notes:
"Systems expected to be triggered into PFAS treatment using either granular activated carbon (GAC) or ion
exchange (IX) treatment technologies.
l'Based on median raw water TOC annual system-means for non-purchased water systems.
cBased on THM4 reductions due to GAC installation from the disinfection byproduct (DBP) information
collection rule (ICR) treatment studies. Reductions dependent on empty bed contact time (EBCT) and
source water type (surface water or groundwater).
Figure 6-10: Overview of Analysis of Co-Removal Benefits
Proposed PFAS Rule Economic Analysis
6-82
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.7.1.2 Baseline Information on DBP Precursors and Trihalomethane
Formation
DBP precursors are the chemical constituents that are reactants or intermediates in the formation
of DBPs. Precursors can be characterized by their origin and the nature of their chemistry
(inorganic vs. organic). Precursors include NOM and anthropogenic organic matter (i.e.,
wastewater) from watersheds, organic matter contaminants within treatment processes, and
biofilm growth within the distribution system. Additional precursors include inorganic matter
present in source water from anthropogenic and natural sources, or chemical additives introduced
during treatment. The presence of DBP precursors is site-specific and dependent on many factors
such as, but not limited to, environment, location, watershed, and treatment.
EPA evaluated raw water TOC data included in the SYR3 and SYR4 ICR datasets (U.S. EPA,
2016j; U.S. EPA, 2022i). The fourth Unregulated Contaminant Monitoring Rule (UCMR 4) TOC
data were not used since that dataset did not include THM4 information. In addition, EPA
compared the DBP ICR Aux 1 TOC data (pre-Stage 1 DBP Rule76) to the SYR3 ICR TOC data
to evaluate changes in DBP precursor occurrence over time. PWSs (specifically subpart H
systems77) are required to achieve a certain percentage of TOC removal; occurrence estimates for
TOC are typically evaluated at the plant-level. The SYR3 ICR dataset contains TOC data for 33
states and systems of all sizes. The SYR4 ICR dataset contains TOC data for 49 states/tribes and
systems of all sizes. To be consistent with SYR3 and SYR4 data management protocols, non-
detections of TOC were assigned a value of 0.0 mg/L for all plant-mean calculations (U.S. EPA,
2016a).
In U.S. EPA (2005b), EPA reviewed the raw water TOC levels for Ground Water plants included
in the DBP ICR Aux 1 data. The results shown in Table 6-30 represent the distribution of
Ground Water plant-mean data as calculated using ICR Aux 1 monthly data from the year 1998.
Only plants with reported data for at least 9 of the 12 months are included in this summary table.
Note that the table does not include results for blended, mixed, or purchased water plants. Table
6-31 shows the distribution of plant-mean TOC concentrations in raw water for non-purchased
Surface Water plants. Segmenting the plants with raw water TOC means provides some
indication of the percentage of plants that would be within each THM4 reduction category
outlined in Section 6.7.1.3. The levels in Ground Water plants tended to be lower compared to
concentrations in Surface Water plants (Table 6-30 and Table 6-32 compared to Table 6-31 and
Table 6-33). As mentioned above, TOC non-detections were assumed to be zero for plant-mean
calculations.
70 Stage 1 Disinfectants and Disinfection Byproducts Rule was promulgated by EPA in December 1998 (U.S. EPA, 1998e).
77 Subpart H systems are defined as public water systems using surface water or ground water under the direct influence of
surface water as a source that are subject to the requirements of subpart H of the National Primary Drinking Water Regulations
(U.S. EPA, 2006a).
Proposed PFAS Rule Economic Analysis
6-83
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-30: DBP ICR (1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Summary of Raw
Water TOC Annual System Means for Ground Water Systems
Data Source
(Year)3
Source Water
Type
Count of
Systems
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
Range of
System-
Means1'
DBP ICR (1998)
Ground Water
103
0.19
1.46
3.36
0.0-16.1
SYR3 ICR (2011)
Ground Water
68
2.19
3.33
5.85
0.42-17.0
SYR4 ICR (2019)
Ground Water
80
1.50
2.54
7.11
0.0-15.73
Notes:
Abbreviations: DBP - Disinfection Byproduct; ICR - Information Collection Rule; SYR - Six-Year Review; TOC - Total Organic
Carbon.
aUsing SYR3 cutoff values, values >100 mg/L were excluded from calculations.
bValues below the minimum reporting level (MRL) were converted to 0.0 mg/L to calculate system-means.
Source: ICRAUX1 database; table extracted from Exhibit 3.6 of U.S. EPA (2005b).
Table 6-31: DBP ICR (1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Summary of Raw
Water TOC Annual System Means for Surface Water Systems
Data Source
(Year)3
Source Water
Type
Count of
Systems
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
Range of
System-
Means1'
DBP ICR (1998)
Surface Water
307
2.71
3.14
5.29
0.0-21.4
SYR3 ICR (2011)
Surface Water
756
2.89
3.45
6.45
0.0-29.3
SYR4 ICR (2019)
Surface Water
802
3.29
3.88
6.93
0.0-38.9
Abbreviations: ICR - Information Collection Rule; SYR - Six-Year Review; TOC - Total Organic Carbon.
Notes:
aUsing SYR3 cutoff values, values >100 mg/L were excluded from calculations.
bValues below the MRL were converted to 0.0 mg/L to calculate system-means.
EPA reviewed the finished water TOC levels included in SYR3 ICR and SYR4 ICR data. The
results shown in Table 6-32 represent the distribution of TOC concentrations for Ground Water
plants. Note that Ground Water plants are not federally required to report finished water TOC
data. In addition, EPA reviewed finished water TOC levels for Surface Water plants included in
SYR3 and SYR4 ICR data. Table 6-33 displays the distribution of TOC levels in finished water
for Surface Water plants. Similar to the raw water comparison, TOC levels tended to be higher
among Surface Water plants compared to Ground Water plants.
Proposed PFAS Rule Economic Analysis
6-84
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-32: SYR3 ICR (2011) and SYR4 ICR (2019) - Summary of Finished Water TOC
Annual System Means for Ground Water Systems
Data Source
(Year)3
Source Water
Type
Count of
Systems
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
Range of
System-
Means1'
SYR3 ICR (2011)
SYR4 ICR (2019)
Ground Water
Ground Water
78
113
1.86
0.73
2.30
2.77
4.53
3.63
0.0-11.4
0.0-93.0
Notes:
Abbreviations: ICR - Information Collection Rule; SYR - Six-Year Review; TOC - Total Organic Carbon.
aUsing SYR3 cutoff values, values >100 mg/L were excluded from calculations.
bValues below the MRL were converted to 0.0 mg/L to calculate system-means.
Table 6-33: SYR3 ICR (2011) and SYR4 ICR (2019) - Summary of Finished Water TOC
Annual System Means for Surface Water Systems
Data Source
(Year)3
Source Water
Type
Count of
Systems
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
Range of
System-
Means'"
SYR3 ICR (2011)
SYR4 ICR (2019)
Surface Water
Surface Water
756
802
1.93
1.89
2.32
2.24
3.99
3.90
0.0-25.1
0.0-74.4
Abbreviations: ICR - Information Collection Rule; SYR - Six-Year Review; TOC - Total Organic Carbon.
Notes:
aUsing SYR3 cutoff values, values >100 mg/L were excluded from calculations.
bValues below the MRL were converted to 0.0 mg/L to calculate system-means.
EPA compared the levels of raw water TOC between the DBP ICR and SYR3 ICR to evaluate
the changes in TOC occurrence over time (U.S. EPA, 2016g). EPA used 1998 data from the DBP
ICR Aux 1 database and 2011 data from the SYR3 ICR dataset and included only the data from
systems that were found in both datasets (referred to as "common systems"). The evaluation of
TOC changes over time was limited to large Surface Water systems (>100,000 population
served) because the DBP ICR only covered large systems.
Table 6-34 below presents plant-level summary statistics for finished water TOC from common
systems in the Aux 1 database and SYR3 ICR. The common systems were distributed across 14
states (Alabama, Alaska, Illinois, Indiana, Iowa, Kentucky, Nevada, New Jersey, North Carolina,
Oklahoma, Pennsylvania, South Carolina, Virginia, and West Virginia). The comparison of data
for large Surface Water supplies between 1998 and 2011 shows a small decrease in treated water
TOC levels. The median finished water TOC concentrations at large systems were 1.76, 1.75,
and 1.51 mg/L in the Aux 1 database, SYR3 ICR dataset, and SYR4 ICR dataset, respectively.
Proposed PFAS Rule Economic Analysis
6-85
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-34: DBP ICR (Aux 1; 1998), SYR3 ICR (2011), and SYR4 ICR (2019) - Finished
Water Annual System Mean TOC; Common Surface Water Systems
Data Source
(Year)
Count of
Systems3
Median
(mg/L)
Mean
(mg/L)
90%ile
(mg/L)
95%ile
(mg/L)
% Means >
2 mg/L
% Means
> 3 mg/L
DBP ICR (1998)
1.76
1.77
2.90
3.23
34%
8%
SYR3 ICR (2011)
80
1.75
1.74
2.78
3.24
30%
8%
SYR4 ICR (2019)
80
1.51
1.49
2.44
2.81
21%
5%
Abbreviations: DBP - disinfection byproduct; ICR - Information Collection Rule; SYR - Six-Year Review; TOC - Total
Organic Carbon.
Note:
aSome systems included data for multiple plants.
Source: Table extracted from Exhibit 6.11 of U.S. EPA, 2016g
Table 6-35 summarizes THM4 baselines under DBP ICR, which represents pre-Stage 1 and
Stage 2 DBP Rules. Prior to evaluating the SYR4 ICR THM4 data, EPA removed values greater
than 10 times the MCL (800 |ig/L) due to potential data entry errors. Additionally, EPA
converted values below the MRL (10 |ig/L) to 0 |ig/L, which is consistent with previous SYR
data analysis (U.S. EPA, 2016a). Average THM4 values were higher for Surface Water plants
compared to Ground Water plants across the two datasets. Within the DBP ICR dataset,
representing PWSs serving populations >100,000, 82 Ground Water plants had a median THM4
concentration of 6.8 |ig/L with a range of 0-123 |ig/L. For the 213 Surface Water plants in the
DBP ICR, the median THM4 concentration was 40 |ig/L with a range of 0 to 117 |ig/L. In
comparison, post- Stage 1 and 2 DBP Rules SYR4 ICR data show median THM4 concentrations
of 5.0 |ig/L and 41.4 |ig/L and mean THM4 concentrations of 13.4 |ig/L and 41.1 |ig/L in
Ground Water and Surface Water, respectively. Plant means ranged from 0 to 371.4 |ig/L and
from 0 to 263.8 |ig/L for Ground Water and Surface Water, respectively. Note that the SYR4
dataset was from voluntary submissions and includes data from systems of all sizes. The SYR4
ICR reduced dataset, limited to PWSs serving populations >100,000, shows median THM4
concentrations of 24.4 and 36.1 |ig/L and mean THM4 concentrations of 25.0 and 35.1 |ig/L for
Ground Water and Surface Water, respectively. Plant means ranged from 0 to 66.6 |ig/L and
from 0 to 62.0 |ig/L for Ground Water and Surface Water, respectively.
Table 6-35: Summary of THM4 Baseline Comparing DBP ICR and SYR4 ICR
Data Source
Source Water
Type
Count of
Systems0
THM4
Median
(Mg/L)
THM4
Mean
(Mg/L)
90%ile
(Mg/L)
Range of
System-Means'
DBP ICR (1998)a
Ground Water
82
6.8
15.4
37
0-123
DBP ICR (1998)a
Surface Water
213
40
42
70
0-117
SYR4 ICR
84
24.4
25.0
53.1
0 - 66.6
Reduced (2012-
Ground Water
2019)b,e,f
SYR4 ICR
291
36.1
35.1
50.2
0 - 62.0
Reduced (2012-
Surface Water
2019)b,e,f
Proposed PFAS Rule Economic Analysis
6-86
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-35: Summary of THM4 Baseline Comparing DBP ICR and SYR4 ICR
Data Source
Source Water
Type
Count of
Systems0
THM4
Median
(Mg/L)
THM4
Mean
(Mg/L)
90%ile
(Mg/L)
Range of
System-Means'1
SYR4 ICR (2012-
2019)b,e
Ground Water
26,243
5.0
13.4
38.5
0-371.4
SYR4 ICR (2012-
2019)b,e
Surface Water
9,618
41.4
41.1
64.1
0 -263.8
Abbreviations: DBP - Disinfection Byproduct; ICR - Information Collection Rule; SYR - Six-Year Review; THM4 - Four Regulated
Trihalomethane s.
Notes:
aStage 2 DBP Rule Economic Analysis (U.S. EPA, 2005b), screened data from Exhibit 3.15 and 3.20
bUsing SYR3 cutoff values, values >10 times the MCL were excluded from calculations.
CNA values and blanks were removed prior to calculations.
dValues below the MRL were converted to 0.0 (ig/L to calculate system-means.
eSYR4 data collected from 2012 to 2019. All years were included in calculations.
fSYR4 reduced dataset included only PWSs serving populations >100,000
In the Economic Analysis (EA) for the Stage 2 DBP Rule, EPA estimated a combined average
THM4 reduction for all systems of 7.8 percent, with Surface Water systems ranging from 9.2
percent (systems serving >10,000) to 7.2 percent (systems serving <10,000), and Ground Water
systems ranging from 1.4 percent (systems serving >10,000) to 2.0 percent (systems serving
<10,000) (U.S. EPA, 2005b). Comparisons of the DBP ICR THM4 baseline data and the SYR4
data that reflects Stage 1 and Stage 2 DBP Rule changes indicate that the Stage 2 EA slightly
overestimated the ATHM4 for Surface Water systems (40 to 41.4 |ig/L, 3.5% increase) and
underestimated the ATHM4 for Ground Water systems (6.8 to 5.0 |ig/L, 26.5% reduction).
Comparing all systems (Surface Water and Ground Water) serving >100,000, no statistically
significant difference (P = 0.2) was observed between the DBP ICR and SYR4 dataset means.
Comparing Ground Water systems in the DBP ICR dataset to those in the reduced SYR4 dataset
showed a statistically significant difference (P = 0.0003) in THM4 means, with THM4
increasing in the more recent years (SYR4). Comparing Surface Water systems in the DBP ICR
dataset to those in the reduced SYR4 dataset showed no statistically significant difference (P =
0.3) in THM4 means. The lack of statistically significant differences in THM4 means between
the DBP ICR and SYR4 datasets for Surface Water systems indicates that TOC and THM4
trends support the use of the DBP ICR dataset to predict ATHM4 resulting from GAC treatment.
For large Ground Water systems (populations >100,000), reductions in THM4 mean
concentrations may be underestimated due to the increase in THM4 baseline concentrations
observed from data reported in the DBP ICR to the SYR4 ICR. Based on the TOC and THM4
trends over time and the percent differences observed between the DBP ICR and SYR4 dataset
means, EPA determined that using the DBP ICR Treatment Study Database results for ATHM4
to predict future ATHM4 resulting from GAC treatment was justified and reasonable.
Additionally, with this focus on GAC treatment and the reduction of THM4, it is important to
note that the DBP ICR treatment study required systems to conduct DBP precursor removal
studies (Treatment Study Database), which contains the most extensive amount of data on GAC
treatment and DBP formation potentials (U.S. EPA, 1996; L. Wang et al., 2019).
Larger datasets, such as SYR ICRs, have not included data on both disinfectant type and DBP
formation. The DBP ICR collected this information in addition to other source and water quality
Proposed PFAS Rule Economic Analysis
6-87
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
parameters. Table 6-36 shows mean THM4 concentrations in the DBP ICR per disinfectant type
and source water type.
Table 6-36: DBP ICR (Aux 1) Summary of THM4 Concentrations Based on Disinfectant
and Source Water Type
Disinfectant Type
Source Water Type Count of Plants / Facilities
Mean THM4 concentration (jig/L)
Ground Water
15
29.2
Chloramine
77
Surface Water
43.2
Ground Water
34
21.3
Free Chlorine
164
Surface Water
45.0
Free Chlorine +
Ground Water
1
18.7
Chloramine (DS)
Surface Water
20
53.2
Abbreviations: DBP - disinfection byproduct; THM4 - Four Regulated Trihalomethanes; DS - Distribution System.
Despite the significant public health improvements provided by EPA's Stage 2 Disinfectant and
Disinfection Byproducts Rule (D/DBPR), DBPs are still estimated to cause approximately 8,000
cases of drinking water-attributable bladder cancer cases every year (Weisman et al., 2022).
Hence, there are still public health benefits to be realized when DBPs are reduced when feasible.
Where systems install activated carbon, the PFAS rule will, for many systems, further reduce
DBP concentrations because of precursor removal. While the Stage 1 and Stage 2 D/DBPRs
were effective at reducing THM4, there are remaining risks associated with DBP exposure that
could be further reduced as shown in the baseline analysis above. The Stage 2 D/DBPR was
promulgated in 2006 and since the rule implementation there have been numerous peer-reviewed
studies that have shown an increased weight of evidence supporting a correlation between
chlorination DBPs and bladder cancer with updated estimates on attributable cases (Weisman et
al., 2022; Regli et al., 2015). Additionally, there is an increased understanding of the role of
genetically susceptible populations and exposure routes for THMs (i.e., oral, inhalation, and
dermal) that impact risk assessments. This comparison between the SYR4 ICR (2019) and DBP
ICR (1998) showed that the DBP ICR THM4 data were still relevant for the post Stage 2
D/DBPR baselines for both TOC (i.e., DBP precursors) and THM4. Because the baseline was
pre-Stage 1 (DBP ICR), EPA took the low-end estimate for THM4 reduction to reduce possible
overestimation. Further reduction in TOC concentrations in finished water could be achieved if
additional treatment is added (i.e., PFAS removal using GAC treatment).
6.7.1.3 Estimation ofTrihalomethane Reduction using Treatment
Models
6.7.1.3.1 DBP Information Collection Rule Treatment Study Database
The Information Collection Rule Treatment Study Database (ICR TSD) contains results of the
most extensive GAC study conducted on a national scale. The ICR TSD contains treatment study
data submitted by systems required to conduct DBP precursor removal studies under the DBP
ICR (U.S. EPA, 1996). The systems included in the ICR TSD were considered "challenged" in
their ability to achieve compliance with potential Stage 2 DBP rule revision MCLs. The
Proposed PFAS Rule Economic Analysis
6-88
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
participating systems included Surface Water systems (and Ground Water systems under the
direct influence of Surface Water) serving 100,000 or more people and having > 4 mg/L of TOC
in source water, and Ground Water systems serving 50,000 or more people and having > 2 mg/L
of TOC in finished water. Both free chlorine and chloramine systems were included in the
treatment study (U.S. EPA, 1996; L. Wang et al., 2019).
Data from the ICR TSD study from these "challenged systems" can be used to identify
conservative estimates of TOC reduction and associated ATHM4. Due to upstream pollution,
drought, and/or climate change, individual drinking water sources may be as challenged as when
the ICR TSD data were collected (Hashempour et al., 2020; McDonough et al., 2020). While the
GAC treatment dataset dates are from 1998, the physical/chemical relationships observed have
only improved with the current application of GAC being at least as effective for THM4 as was
observed in the ICR TSD (Yuan et al., 2022). While source water parameters and treatments at
individual plants have changed over time, as seen in the baseline characterization in Section
6.7.1.2, EPA determined the ICR TSD was still appropriate to inform estimates of ATHM4
formation potential given the lack of available data to directly inform ATHM4 from PFAS
adsorption studies and the low percent difference in TOC changes on a national scale between
the DBP ICR and SYR4 collection efforts.
From the 63 GAC systems included in the ICR TSD, a total of 182 pilot/rapid small-scale
column test (RSSCT) studies were conducted to develop breakthrough curves of TOC and DBP
formation changes. Two EBCTs, 10 and 20 min, were evaluated for treated water that had passed
through any full-scale treatment processes previously in place at each drinking water treatment
plant to remove precursors (i.e., coagulation/flocculation, sedimentation, filtration) but before
any disinfectant was added. To determine the effect of GAC treatment on DBP formation, these
studies evaluated TOC removal and THM4 formation potential for the treated water before and
after GAC treatment. Uniform formation condition procedures were standardized across each
study, with a reaction time of 24 ± 1 hours, temperature of 20.0 ± 1.0°C, buffered pH at 8.0 ±
0.2, and 24-hr chlorine residual of 1.0 ± 0.4 mg/L as Cb (U.S. EPA, 1996; Summers et al.,
1996).
The pilot/RSSCT studies were timed to account for seasonal changes and an "averaging"
approach was used to remove temporal variations. This approach was consistent with analysis
used to characterize different GAC options for compliance with the MCLs under the Stage 2
DBP Rule (Hooper et al., 2002). Additional details on the GAC study design specifications under
the ICR TSD are available in the "ICR Manual for Bench and Pilot Scale Treatment Studies"
(U.S. EPA, 1996).
For drinking water systems in the ICR TSD that used chloramines (n=123 pilots/RSSCTs) in
their distribution system, free chlorine was still used in the DBP formation tests, therefore the
pilot and RSSCT systems were not compared based on disinfectant type used by the individual
treatment system. For reference, a summary of the THM4 estimates by disinfectant type is
provided in Appendix I Table 1-1. Additionally, if the comparison categories were further parsed
by source water type, disinfectant type, and TOC concentrations, then the number of systems in
each bin would not provide sufficient studies to compare the ATHM4 estimates. Therefore, EPA
analyzed the THM4 reductions based on raw-water TOC.
Proposed PFAS Rule Economic Analysis
6-89
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
The TOC and THM4 formation potential reductions data from the ICR TSD were modeled with
a logistic equation using results from 182 pilot plant/RSSCT studies. EPA fit the logistic
function parameters for each EBCT and did not consider feed water quality parameters. Results
were categorized by TOC level and source water type. Further subdivision of these or additional
categories would have resulted in very small numbers of systems in bins and some bins not being
filled (see Appendix I Table 1-1 for example of "disinfection type" added as a category). The
model calculated individual system TOC removal for the EBCT and results were averaged for
each subset of systems for the GAC replacement interval. The model was not intended to
simulate the dynamics of TOC removal by GAC or the formation of THM4, but it simulated the
TOC ranges within the pilot/RSSCT studies and the changes in THM4 due to the reduction in
TOC observed in the ICR TSD. EPA used Python to individually fit data from each pilot or
RSSCT study in the ICR TSD to a logistic equation and the performance was then averaged.
Additional details on the data model are included in Appendix I.
To conservatively estimate national scale THM4 reduction due to GAC treatment to reduce
levels of PFAS compounds, EPA chose a 2-year GAC replacement time. EPA assumes that this
is the longest amount of time before replacement would be required and percent removals are
approximately at their long-term removal level with minimal further changes. The PFAS
NPDWR will likely result in some systems replacing GAC media more frequently than 2 years,
which EPA expects would result in a greater average TOC reduction since TOC removal
decreases over time with GAC treatment (see Figure 6-11 and Figure 6-12 for Ground Water and
Surface Water respectively). The overall trends seen in Figure 6-11 and Figure 6-12 show
greatest TOC removal in the first 200 days of use, after which the predicted TOC removal
becomes consistent for Ground Water with 26.9 percent (EBCT 20 min) and Surface Water with
37.5 percent (EBCT 20 min). EPA solicits public comment on whether the GAC replacement
interval of 2-years was too conservative an approach for estimating benefits in this DBP co-
removal analysis.
Proposed PFAS Rule Economic Analysis
6-90
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Replacement Interval (days)
Abbreviations: TOC - total organic carbon; GAC - granular activated carbon; EBCT - empty bed contact times.
Notes:
Pink shaded area represents ±1 standard deviation for Ground Water TOC with a GAC EBCT of 10 mill
Gray shaded area represents ±1 standard deviation for Ground Water TOC with a GAC EBCT of 20 mill
Figure 6-11: Estimated TOC Percent Removal in Ground Water Using GAC Based on
Logistic Equation Model
Proposed PFAS Rule Economic Analysis
6-91
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
100
Water Type: surface
\
surface • TOC • 10 mm tBCT
?8lt87% removed
suffice TOC 20 min EBCT
37.Si 13. J % removed
\ \
\ >»
\ N.
80 4
15
E 60
a:
£
u
o
40-
20-
100 200 300 400 500
Replacement Interval (days)
600
700
Abbreviations: TOC - total organic carbon; GAC - granular activated carbon; EBCT - empty bed contact times.
Notes:
Pink shaded area represents ± 1 standard deviation for Ground Water TOC with a GAC EBCT of 10 min
Gray shaded area represents ±1 standard deviation for Ground Water TOC with a GAC EBCT of 20 min
Figure 6-12: Estimated TOC Percent Removal in Surface Water Using GAC Based on
Logistic Equation Model
To estimate the TOC reduction, the ICR TSD pilot/RSSCT studies (n = 182) were partitioned
into five potential bins based on TOC concentrations in raw water (Very Low <1 mg/L, Low >1
to <2 mg/L TOC, Mid >2 to <3.5mg/L, High-Mid >3.5 to <5mg/L, High TOC >5mg/L TOC).
There were no systems in the ICR TSD that fell into very low TOC bin. Based on the logistic
equation for TOC reducti on, higher raw water TOC concentrations yield greater TOC reductions
(in absolute value) following to GAC treatment. Table 6-37 shows the TOC reduction for all
waters (both Ground Water and Surface Water) for a 20 min EBCT.
Proposed PFAS Rule Economic Analysis
6-92
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-37: TOC Reduction for All Waters (Both Surface Water and Ground Water)
with GAC EBCT of 20 Min and a 2-year Replacement Time
TOC Bin
Number of Studies in
GAC Treatment Dataset
TOC Reduction ± 1-
Standard Deviation (%)
TOC Reduction (mg/L)
TOC 1-2 mg/L
20
41.9 ±23.2
0.75 ±0.39
TOC 2-3.5 mg/L
103
37.1 ± 11.6
1.06 ±0.36
TOC 3.5-5 mg/L
44
32.0 ±9.6
1.31 ± 0.39
TOC above 5mg/L
15
26.3 ± 14.2
1.83 ±0.91
Abbreviations: TOC - Total Organic Carbon; GAC - Granular Activated Carbon; EBCT - Empty Bed Contact Time.
Notes: The model calculated individual system TOC removal and results were averaged for each influent TOC bin for a two-
year GAC replacement interval and the standard deviation was calculated for each subset average.
Using the same raw water TOC bins, EPA estimated national scale ATHM4 values resulting
from GAC treatment. The selected ATHM4 estimate was based on a conservative approach
(mean concentration minus one standard deviation), since the DBP ICR systems included in the
treatment studies were "challenged systems" (i.e., systems that had difficulty meeting regulatory
compliance requirements) that may experience increased TOC reduction due to GAC
installation.
The analysis here assumes operation of GAC with a replacement interval of 730 days (2 years).
Although some systems will operate with longer replacement intervals, after 730 day (2 years),
the modeled TOC reduction due to GAC shows consistent removal when further extending the
replacement interval (Figure 6-11 and Figure 6-12). While systems may replace GAC at shorter
time intervals, the 2-year replacement assumption also approximates blended systems (i.e.,
multiple GAC treatment trains in parallel with varying replacement intervals) and TOC long
term removal by adsorption since GAC treatment for PFAS uses adsorption rather than
biodegradation (Kempisty et al., 2022). Therefore, the estimated TOC reduction at 730 days (2
years) should also be representative for systems with longer replacement intervals or systems
with intermittent use. If GAC replacement occurred more frequently due to PFAS treatment
needs, higher average TOC removal would occur, resulting in greater THM4 reduction as shown
in the 6-month GAC replacement time-steps (1/2, 1, 1 V2, and 2 years) shown in Appendix I
(Tables 1-2,1-3,1-4, and 1-5). Using the longer replacement time of 2 years is consistent with
EPA's conservative approach to estimating ATHM4 even when the presence of PFAS
compounds and other source water conditions may affect GAC replacement frequency.
Based on common treatment designs, EPA expects GAC treatment parameters for PFAS removal
to be 20 min EBCTs (U.S. EPA, 2023h). Table 6-38 and Table 6-39 provide estimates of THM4
reductions in the modeled 182 pilot/RSSCT systems broken out by Surface Water vs Ground
Water and 20 min EBCT. The number of GAC systems included in each TOC bin is provided
along with the average ATHM4 and the "conservative" ATHM4 (defined as average ATHM4
minus 1-standard deviation) with a GAC replacement time of 2 years.
Proposed PFAS Rule Economic Analysis
6-93
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-38: Estimation of ATHM4 in Surface Water with a 20 Min EBCT, and a 2-year
GAC Replacement Time
Raw Water TOC Bin
Number of Studies in GAC
Treatment Dataset
Average ATHM4 ± 1-
Standard Deviation
(Mg/L)
Conservative ATHM4
(Mg/L)
TOC 1-2 mg/L
12
14.23 ±7.34
6.9
TOC 2-3.5 mg/L
89
31.54 ±24.02
7.5
TOC 3.5-5 mg/L
37
48.55 ±31.81
16.7
TOC above 5mg/L
7
67.2 ± 18.3
48.9
Abbreviations: EBCT - Empty Bed Contact Time; TOC - Total Organic Carbon; THM4 - Four Regulated Trihalomethanes;
GAC - Granular Activated Carbon.
Table 6-39: Estimation of ATHM4 in Ground Water with a 20 Min EBCT, and a 2-year
GAC Replacement Time
Raw Water TOC Bin
Number of Studies in GAC
Treatment Dataset
Average ATHM4 ± 1-
Standard Deviation (ju.g/L)
Conservative ATHM4
(Mg/L)
TOC 1-2 mg/L
8
15.14 ±8.98
6.2
TOC 2-3.5 mg/L
14
22.02 ± 17.48
4.5
TOC 3.5-5 mg/L
7
27.46 ±8.33
19.1
TOC above 5mg/L
8
56.66 ±38.69
18.0
Abbreviations: EBCT - Empty Bed Contact Time; TOC - Total Organic Carbon; THM4 - Four Regulated Trihalomethanes;
GAC - Granular Activated Carbon.
For the Low (l-2mg/L), Mid (2-3.5mg/L), and High-Mid (3.5-5mg/L) TOC bins, the
conservative ATHM4 estimates were reasonable compared to the mean concentrations reported
into the SYR4 baseline occurrence data. Conservative ATHM4 estimates in the High TOC bin
(>5mg/L) were higher due to the greater reduction in TOC. For the THM4 reduction observed in
the High TOC bin (>5mg/L), the conservative THM4 reduction estimates were higher due to the
greater reduction in TOC and may not be plausible based on the baseline occurrence information
and if it is assumed that all systems are currently in compliance (THM4 <80 |ig/L). However,
based on SDWIS violations, not all systems are currently in compliance with the Stage 2 DBP
Rule. EPA assumes that these larger ATHM4 estimates would be observed only in the 90th
percentile of TOC data. Ground Water systems in the High TOC bin may also be
mischaracterized during the ICR TSD and should be more accurately described as Ground Water
under the direct influence of Surface Water (GWUDI) (Brunke et al., 1997; Chin et al., 2000).
The GWUDI provisions of the 1989 Surface Water Treatment Rule instituted the concept of
Ground Water that is so closely connected to Surface Water that public water supply wells
should be regulated as Surface Water rather than as Ground Water (U.S. EPA, 1989). If one or
more Ground Water systems were mischaracterized, then this could overestimate the ATHM4
Proposed PFAS Rule Economic Analysis
6-94
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
estimate since these systems make act more like a Surface Water system in terms of TOC
removal.
Since these conservative ATHM4 estimates were based on the longest assumed time period for
GAC use (i.e., GAC replacement time) in the current regulatory options, EPA assumes that the
estimated ATHM4 values are conservative, considering that shorter replacement times would
increase the average TOC removal during that operation time.
Since these Surface Water and Ground Water systems have already been identified as
"challenged" in the ICR TSD (pre-Stage 1 and Stage 2 DBP Rules), this indicates the specific
advantages of using GAC to reduce THM4 precursors in comparison to conventional treatment
(i.e., coagulation/flocculation followed by sedimentation and filtration). When GAC treatment is
used for additional contaminant removal such as PFAS, TOC reduction benefits will also be
observed. Since there is a lack of PFAS and TOC co-removal data, the ICR TSD can provide the
largest dataset on TOC reduction and THM4 formation changes in drinking water to provide a
national estimate of ATHM4.
The limitations and uncertainties for using this method to quantify ATHM4 due to GAC
treatment for PFAS are listed in Section 6.8. One major limitation of using the ICR TSD was that
this dataset only used chlorine as a disinfectant and does not capture THM reduction in
chloraminating systems. This limitation may lead to an overestimate of THMs formed in systems
that used chloramines in the distribution system since THM4 can continue to form within the
distribution system and formation tends to be lower when chloramines are used in comparison to
free chlorine (Hua et al., 2008). Most chlorinating systems use free chlorine as a primary
disinfectant followed by the addition of ammonia to form chloramines for the secondary
disinfectant. Of the 9,838 Ground Water entry points to distribution systems included in UCMR
3, chlorine disinfection was used 8.8 times more often than chloramine (n=7,881 for chlorine
exclusively and n=896 for chloramines or both chlorine and chloramines) (U.S. EPA, 2016g).
For the 3,179 Surface Water entry points to distribution systems in UCMR 3, chlorine was used
1.9 times more than chloramine (n=l,648 for chlorine exclusively and n=879 for chloramines or
both chlorine and chloramines) (U.S. EPA, 2016g).
By assuming the use of free chlorine only, the estimates of ATHM4 from pilot/RSSCTs studies
may provide an overestimation when factoring in use of both free chlorine and chloramines.
Thus, using the conservative free chlorine THM4 formation potential (average ATHM4 minus 1-
standard deviation) rather than the average ATHM4, EPA attempted to address the
overestimation and provide a reasonable national estimate of ATHM4.
In a separate DBP formation study under the ICR TSD, individual DBP formation conditions
were selected to represent simulated distribution systems for each individual plant that accounted
for the disinfectant differences (i.e., chlorine versus chloramine) by using only chlorine as the
disinfectant and varying the reaction times. The simulated distribution system studies were not
included in the estimated ATHM4 provided in this document since including them would have
further increased the uncertainty error for systems using chloramine due to the longer reaction
times.
Proposed PFAS Rule Economic Analysis
6-95
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.7.1.3.2 Trihalomethane Reduction Comparison to Fourth Six Year Review PFAS
Plants with GAC Treatment
EPA compared ATHM4 estimates from the ICR TSD to the SYR4 data for PFAS-associated
plants that have installed GAC. The objective of this analysis was to compare the ICR TSD
modeled predictions of ATHM4 to the observed ATHM4 concentrations from PWSs that
installed GAC for PFAS treatment.
EPA identified systems that had detectable levels of PFOA and/or PFOS in UCMR 3.
Subsequently, EPA used UCMR 4 data to identify which systems indicated use of GAC
treatment. Finally, EPA used consumer confidence reports (CCRs) for all systems that detected
PFAS and specified GAC treatment for PFAS to approximate the year that GAC treatment was
installed and the purpose for installation. While this approach limited the number of systems
available for comparison (n = 7), it allowed EPA to pinpoint, approximately, which samples
were taken before and after GAC installation. EPA obtained THM4 compliance monitoring data
through the SYR4 ICR, based on data collected between 2012 and 2019. EPA calculated the
ATHM4 values based on observed THM4 levels before and after GAC installation.
EPA identified plants using the following criteria (see Table 6-40):
1. Detectable level of PFAS in UCMR 3 (i.e., detections of PFOA and/or PFOS above their
respective minimum reporting level (MRL) values).
2. GAC installed as indicated in UCMR 4 and confirmed for PFAS treatment by using CCR
information.
3. Ability to identify the year GAC was installed using CCR information.
4. THM4 data available from SYR4 (CCR THM4 data were used as an alternative when
SYR4 data were unavailable).
Table 6-40: Selected Distribution Systems from SYR4 Based on Outlined Criteria
PWSID
Source Water
Type
Disinfectant Type
Year GAC Began
AL0000577
Surface Water
Free Chlorine3
2018
AL0001092
Surface Water
Free Chlorine, Chlorine Dioxide
2016
AZ0407046
Ground Water
Free Chlorine
2017
MI0005370
Surface Water
Free Chlorine
2018
NY3503549
Surface Water
Free Chlorine
2018
OH2903412
Ground Water
Free Chlorine
2017
PA1090069
Ground Water
Free Chlorine
2017
Abbreviations: PWSID - Public water system identifier; SYR - Six Year Review; GAC - Granular Activated Carbon.
Note:
Tree chlorine includes gaseous chlorine, offsite generated hypochlorite, or onsite generated hypochlorite.
Proposed PFAS Rule Economic Analysis
6-96
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA chose sampling years to represent conditions before and after GAC treatment based on the
following criteria:
• If source water type was Surface Water, one year before and one year after the year in
which GAC treatment began was used.
• If source water type was Ground Water, two years before and two years after the year in
which GAC treatment began was used. Since Ground Water plants have fewer samples,
this was done to offset the lower sample number. (Note that Ground Water quality
typically has fewer fluctuations than does Surface Water quality, so EPA expects fewer
changes in year-to-year data for Ground Water systems.)
EPA extracted and matched sampling point IDs for the years that represent before and after GAC
treatment (see Appendix I). Only sampling point IDs with the same number of samples for
before and after GAC treatment were used to determine THM4 averages. The seasonality and
quantity of samples were considered, and EPA found that samples were taken consistently and
remained at the same frequency throughout the years selected to represent before and after GAC
treatment.
EPA calculated ATHM4 concentrations for each system at matched sampling point locations
using THM4 data collected before and after GAC installation. EPA also estimated ATHM4
concentrations at the broader plant level by aggregating all THM4 locational sampling data
collected before and after GAC installation (see Table 6-41).
Proposed PFAS Rule Economic Analysis
6-97
March 2023
-------
DRAFT FOR PUBLIC COMMENT
Table 6-41: Information on Selected Distribution System and Corresponding ATHM4 Values
MARCH 2023
PWSID
Source Water
Type
Disinfectant Type
Sampling Point IDb
Average THM4
(Before) (ju.g/L)
Average THM4
(After) (jig/L)
ATHM4 Qig/L)
Average
ATHM4
(Mg/L)c
AL0000577
Surface Water
Free Chlorine
12975
16.5
10.9
5.7
9.8
AL0001092
Surface Water
Free Chlorine,
Chlorine Dioxide
23592
16.6
6.4
10.2
15.7
AZ0407046
Ground Water
Free Chlorine
33997
28.8
21.6
7.3
4.8
MI0005370
Surface Water
Free Chlorine
CCR
84.9
66.4
18.5
18.5
NY3503549
Surface Water
Free Chlorine
334940
39.1
7.6
31.5
31.5
OH2903412
Ground Water
Free Chlorine
541452
8.9
7.0
1.9
-4.1
PA1090069
Ground Water
Free Chlorine
892902
21.0
21.3
-0.3
-10.7
Abbreviations: THM4 - Four Regulated Trihalomethanes.
Notes:
aATHM4= THM4 Average (Before) - THM4 Average (After).
bSampling point IDs that have a sampling point type of entry point (EP) were used when available. When unavailable, the first listed sampling point ID was used.
cAverage delta of pairwise changes in THM4 for each location in the entire distribution system.
Proposed PFAS Rule Economic Analysis
6-98
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Based on available data, the EBCT for the seven plants from SYR4 is unknown. EPA used TOC
values from SYR4 when available and used CCR TOC data as an alternative when TOC data
were missing from SYR4. TOC values for SYR4 Ground Water plants were missing from the
SYR4 dataset and corresponding CCRs, and due to this limitation, EPA did not use raw water
TOC bins, but instead used a range of ATHM4 values for comparison between SYR4 and ICR
TSD.
EPA compared ATHM4 values from the SYR4 to the ICR TSD dataset conservative approach
(see Table 6-42). Among SYR4 Ground Water plants, a THM4 change between -10.7 |ig/L to
4.8 |ig/L was observed. ICR TSD Ground Water ATHM4 values ranged from 3.5 |ig/L to 67.2
|ig/L. SYR4 Ground Water averages were between -7.2 |ig/L to 62.4 |ig/L lower than ICR TSD
Surface Water averages.
Table 6-42: Comparison Between ICR TSD Conservative ATHM4 and SYR4 ATHM4 for
Surface Water Systems
Raw Water
Surface Water
TOC Bin
ICR TSD Conservative ATHM4
(Mg/L)
PWSID
SYR4 ATHM4 Qig/L)
TOC 0-1 mg/L
No available data
AL0000577,
AL0001092
5.7,
15.7
TOC 1-2 mg/L
6.9
NY3503549
31.5
TOC 2-3.5 mg/L
7.5
MI0005370
18.5
TOC 3.5-5 mg/L
16.7
No available data
No available data
TOC >5 mg/L
48.9
No available data
No available data
Abbreviations: TOC - Total Organic Carbon; THM4 - Four Regulated Trihalomethanes; ICR - Information Collection Rule;
TSD - Treatment Study Database.
Notes:
aThree of the seven surface water PWSs had no TOC measurements.
bEBCTs were unknown.
c20 min EBCT was used to determine best-case and conservative ATHM4 values.
Two of the three Ground Water systems showed increased THM4 formation after the installation
of GAC. Possible reasons for increased formation may include source water changes (i.e.,
increased sediment runoff or spore concentration fluctuations in Ground Water), operational
challenges of the GAC treatment, changes to other treatments within the PWS, or changes in
retention time within the distribution system. The four Surface Water systems had ATHM4
values ranging from 5.7 to 31.5 |ig/L.
Three out of the seven plants had no available TOC data in SYR4 or CCRs. TOC data for the
SYR4 THM4 analysis were only available for Surface Water plants. SYR4 Surface Water plants
with influent TOC concentrations between 1-2 mg/L had an average ATHM4 of 31.5 |ig/L
compared to the ICR TSD conservative ATHM4 estimate of 6.9 |ig/L. For SYR4 Surface Water
plants with influent TOC concentrations between 2-3.5 mg/L, EPA observed an average ATHM4
of 18.5 |ig/L compared to the ICR TSD conservative ATHM4 estimate of 7.5 |ig/L. Both
comparisons of TOC bins for Surface Water show that the conservative estimates for THM4
Proposed PFAS Rule Economic Analysis
6-99
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
reduction are plausible. Note that this finding is based on a small subset of systems (n = 4) and
may not be representative of systems nationally.
Due to lack of TOC data for SYR4 Ground Water plants, EPA compared Ground Water plants to
the lowest TOC bin (1-2 mg/L) with ICR TSD data available. SYR4 Ground Water plants had
an average ATHM4 between ICR TSD Ground Water plants with influent TOC concentrations
between 1-2 mg/L had an average change in THM4 between -10.7 |ig/L to 4.8 |ig/L compared to
the ICR TSD conservative THM4 reduction estimate of 4.5 |ig/L. Limitations on the comparison
between the ICR TSD ATHM4 estimates and the SYR4 THM changes are described in Section
6.8.
6.7.2 Estimation of Bladder Cancer Risk Reductions
Evaluation of the expected reductions in bladder cancer risk resulting from treatment of PFAS in
drinking water involves five steps listed in Section 6.7.1.1. Section 6.7.1.3.2 provides details on
the estimation of changes in THM4, while Section 6.7.2.1 provides details on selecting the
changes in THM4 specific to the modeled scenarios.78
6.7.2.1 Application of Changes in THM4 to PFAS PWSs
EPA expects PWSs that exceed the PFAS regulatory threshold to consider both treatment and
non-treatment options to achieve compliance with the drinking water standard. EPA assumes that
the populations served by systems with entry points expected to install GAC based on the
compliance forecast detailed in Section 5.3 will receive the DBP exposure reduction benefits.
EPA notes that other compliance actions included in the compliance forecast could result in DBP
exposure reductions, including installation of RO. However, these compliance actions are not
included in the DBP benefits analysis because this DBP exposure reduction function is specific
to GAC. Switching water sources may or may not result in DBP exposure reductions, therefore
EPA assumed no additional DBP benefits for an estimated percentage of systems that elect this
compliance option. Lastly, EPA assumed no change in DBP exposure at water systems that
install IX, as that treatment technology is not expected to remove a substantial amount of DBP
precursors. Finally, EPA also assumed that PWSs included in this analysis use chlorine only for
disinfection and have conventional treatment in place prior to GAC installation.
As described in Section 6.7.1.3, EPA used the relationship between median raw water TOC
levels and changes in THM4 levels estimated in the 1998 DBP ICR to estimate changes in
THM4 concentrations in the finished water of PWSs fitted with GAC treatment. EPA applied
changes in THM4 levels to PWS treating for PFAS using the following steps:
1. Identifying the PWSs expected to be triggered into PFAS treatment under various
thresholds and the associated PWS populations served by source water type: Surface
Water and Ground Water;
78 The benefits analyses described herein relied on methodology implemented in R software (R Core Team, 2021) and differ
slightly from SafeWater MCBC methods. Specifically, SafeWater performs a set of pre-calculations to maximize computational
efficiency and, as such, the order of analytical steps across R and SafeWater models differs. However, results across models are
mathematically consistent.
Proposed PFAS Rule Economic Analysis
6-100
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
2. Estimating the TOC levels associated with each source water type, based on median raw
water TOC data collected among non-purchased Surface Water and Ground Water
systems from the 2019 SYR4 dataset; and
3. Identifying the associated THM4 reduction value based on relationships between raw
water TOC levels and changes in THM4 levels estimated in the 1998 DBP ICR.
As shown in the Section 6.7.1.3 tables, EPA estimated changes in THM4 levels that vary based
on the following characteristics:
• Replacement time: Assumed to be 730 days;
• EBCT:20min;
• Source water type: Surface Water, Ground Water;
• THM4 change scenario: Conservative (mean DBP ICR THM4 reduction minus one
standard deviation per TOC bin).
For the DBP risk reduction modeling, EPA focused on the following treatment scenario (See
Table 6-38 and Table 6-39):
• PWS treatment threshold: PFOA or PFOS mean concentration exceeds threshold
defined by regulatory alternatives;
• EBCT:20min;
• Source water type: Surface Water, Ground Water;
• THM4 change scenario: Conservative.
As described in Section 2.2.4, EPA models a scenario where reduced exposures to THM4 begin
in 2026. Therefore, EPA assumed that the population affected by reduced THM4 levels resulting
from implementation of GAC treatment is exposed to baseline THM4 levels prior to actions to
comply with the rule (i.e., prior to 2026) and to reduced THM4 levels from 2026 through 2104.
Proposed PFAS Rule Economic Analysis
6-101
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.7.2.2 Affected Population
Information on PWS attributes required for estimating changes in population-level bladder
cancer is obtained from EPA's 2021 Q4 SDWIS database (U.S. EPA, 2021h). This information
includes data on PWS primary sources of water (e.g., whether a PWS relies primarily on Ground
Water or Surface Water for their source water), operational status, and population served. Some
PWSs have multiple entry points delivering drinking water to the distribution network. As
discussed in Section 6.7.2.1, the analysis assumes that PWSs will reduce PFAS levels by fitting
individual entry points for either GAC or IX treatment and therefore changes in NOM and
THM4 will also be specific to entry points.
Rather than modeling individual locations (e.g., PWS), EPA evaluates changes in bladder cancer
cases among the aggregate population per treatment scenario and source water type that is
expected to install GAC treatment to reduce PFAS levels. Because of this aggregate modeling
approach, EPA used national-level population estimates to distribute the SDWIS populations
based on single-year age and sex and to grow the age- and sex-specific populations to future
years. Section 5.3 describes the decision tree for GAC technology selection. Appendix B
provides additional details on estimation of the affected population.
6.7.2.3 Bladder Cancer Exposure-Response Modeling
The relationship between exposure to DBPs, specifically trihalomethanes and other halogenated
compounds resulting from water chlorination, and bladder cancer has been the subject of
multiple epidemiology studies (Cantor et al., 2010; U.S. EPA, 2016g; NTP, 2018), meta-analyses
(Villanueva et al., 2003; Costet et al., 2011), and pooled analysis (Villanueva et al., 2004). EPA
used the relationship between THM4 levels and bladder cancer in the Villanueva et al. (2004)
study to support the benefits analysis for the Stage 2 DBP Rule79 which specifically aimed to
reduce the potential health risks from DBPs (U.S. EPA, 2005b).
Regli et al. (2015) analyzed the potential lifetime bladder cancer risks associated with increased
bromide levels in surface source water resulting in increased THM4 levels in finished water.80
To account for variable levels of uncertainty across the range of THM4 exposures from the
pooled analysis of Villanueva et al. (2004), they derived a weighted mean slope factor from the
odds ratios reported in Villanueva et al. (2004). They showed that, while the original analysis
deviated from linearity, particularly at low concentrations, the overall pooled exposure-response
relationship for THM4 could be well-approximated by a linear slope factor that predicted an
incremental lifetime cancer risk of 1 in ten thousand exposed individuals (10"4) per 1 |ig/L
increase in THM4. The linear slope factor developed by Regli et al. (2015) is 0.00427 per 1
|ig/L. Using a fixed effects meta-analysis model assumed by Regli et al. (2015), EPA estimated a
79 See DBP Rule documentation at https://www.epa.gov/dwreginfo/stage-l-and-stage-2-disinfectants-and-
disinfection-byproducts-rules
80
The Regli et al. (2015) slope factor was utilized in the recently peer-reviewed Weisman et al. (2022) study, which estimates
that 8,000 of 79,000 US bladder cancer cases are attributable to bladder cancer. Among other tilings, the authors found that there
is a stronger weight of evidence linking DBPs and bladder cancer since the promulgation of the 2006 Stage 2 DBP regulations
and even since publication of Regli et al. (2015).
Proposed PFAS Rule Economic Analysis
6-102
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
95% CI of 0.00331-0.00522 per 1 |ig/L. This slope enables estimation of the changes in the
lifetime bladder cancer risk associated with lifetime exposures to reduced THM4 levels:
Equation 21:
Odds(x) = Odds(0) ¦ exp(0.00427 * x)
Where Odds(x) are the odds of lifetime bladder cancer incidence for an individual exposed to a
lifetime average THM4 concentration in residential water supply of x |ig/L and Odds(0) are the
odds of lifetime bladder cancer in the absence of exposure to THM4 in residential water supply.
The relationship (Equation 21) has the advantage of being independent from the baseline THM4
exposure level, which is highly uncertain for most affected individuals due to lack of historical
data.
To enable annual bladder cancer risk estimation, EPA assumed that the relationship (Equation
21) also holds for the cumulative bladder cancer risk and cumulative average exposure to
residential water THM4 from birth to a specific age. A person's cumulative THM4 exposure
from drinking water by age a—denoted by xa—is defined as:
Equation 22:
1 va_1
xa ~ / THM4i, x0 — 0
a/—ii=0
EPA estimated the relative risk of bladder cancer by a particular age from a change in average
THM4 experienced by this age as follows:
Equation 23:
exp(0.00427 * [xa — za])
RR(xa,za) exp(0.00427 * [xa - zj) * LR{za) + 1 - LR(za)
Where RR(xa, za) is the relative cumulative risk of bladder cancer associated with a change
from baseline cumulative exposure za to treatment cumulative exposure xa. This calculation
requires an estimate of baseline cumulative bladder cancer risk LR(za) which is described in
Appendix H.
6.7.2.4 Estimation of Bladder Cancer Risk Reductions
EPA estimated changes in annual bladder cancer cases and annual mortality in the bladder cancer
population due to estimated reductions in lifetime THM4 exposure using a life table-based
approach. This approach was used because (1) annual risk of new bladder cancer should be
quantified only among those not already experiencing this chronic condition, and (2) bladder
cancer has elevated mortality implications.
EPA used recurrent life table calculations to estimate a water source type-specific time series of
bladder cancer incidence for a population cohort characterized by sex, birth year, and age at the
beginning of the PFOA/PFOS evaluation period under the baseline scenario and the GAC
regulatory alternative described in Section 6.7.2.1. The estimated risk reduction from lower
exposure to DBPs in drinking water is calculated based on changes in THM4 levels used as
Proposed PFAS Rule Economic Analysis
6-103
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
inputs to the Regli et al. (2015)-based health impact function, as shown in Section 6.7.2.3. The
life-table analysis accounts for the gradual changes in lifetime exposures to THM4 following
implementation of GAC treatment under the regulatory alternative compared to the baseline.81
Details of the life-table calculations are provided in Appendix H. The outputs of the life-table
calculations are the water source type-specific estimates of the annual change in the number of
bladder cancer cases and the annual change in bladder cancer population mortality.
Although the change in THM4 exposure likely affects the risk of developing bladder cancer
beyond the end of the analysis period (the majority of cancer cases manifest during the latter half
of the average individual life span; Hrudey et al., 2015), EPA does not capture effects after the
end of the period of analysis, 2104. Individuals alive after the end of the period of analysis likely
benefit from lower lifetime exposure to THM4. Lifetime health risk model data sources include;
EPA SDWIS; age- and sex-specific population estimates from the U.S. Census Bureau (U.S.
Census Bureau, 2020a); the Surveillance, Epidemiology, and End Results (SEER) program
database (National Cancer Institute),82 and the CDC National Center for Health Statistics.83
Appendix H provides additional detail on the data sources and information used in this analysis
as well as baseline bladder cancer statistics. Appendix B provides additional details on the
estimation of the affected population.
6.7.2.5 Valuation of Bladder Cancer Risk Reductions
EPA uses the Value of Statistical Life to estimate the benefits of reducing mortality associated
with bladder cancer in the affected population. Section 2.2 provides information on updating
Value of Statistical Life for inflation and income growth. EPA uses COI-based valuation to
estimate the benefits of reducing morbidity associated with bladder cancer. Specifically, EPA
used bladder cancer treatment-related medical care and opportunity cost84 estimates from Greco
et al. (2019). Table 6-43 shows the original COI estimates from Greco et al. (2019) which were
reported in $2010, along with the values updated to $2021 used in this analysis. EPA further
notes that the estimates for non-invasive bladder cancer subtype were used to value local,
regional, and unstaged bladder cancer morbidity reductions, while the estimates for the invasive
bladder cancer subtype were used to value distant bladder cancer morbidity reductions.85
81 As described above, EPA models THM4 changes under the treatment scenario as being in effect for the years 2023 through
2104, with nonzero THM4 changes first occurring in 2026, the year when all PWS are assumed to comply with PFAS treatment
requirements.
82 For cancer incidence and stage distribution data, EPA relies on SEER 21 (2009-2018); for cancer survival data, EPA relies on
SEER 18 (2000-2017).
83 CDC Wonder data on 1999-2019 all-cause and bladder cancer mortality by age and sex.
84 Opportunity (or indirect) costs modeled by this study were represented by the value of time needed to undergo the cancer
treatment, which could otherwise have been dedicated to work or leisure activities.
85 Local cancer is a malignant cancer confined entirely to the organ where the cancer began. Remote cancer refers to cancer that
has grown beyond the original (primary) tumor to nearby lymph nodes or organs and tissues. Distant cancer refers to cancer that
has spread from the original (primary) tumor to distant organs or distant lymph nodes; it is also called a distant metastasis.
Finally, unstaged cancer is a cancer whose subtype is unknown.
Proposed PFAS Rule Economic Analysis
6-104
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-43: Bladder Cancer Morbidity Valuation
Bladder Cancer
Subtype3
Type of Cost
Cost in First Year
($2010)b
Cost in Subsequent Years
($2010)b
Cost in First Year ($2021)°
Cost in Subsequent Years
($2021)c
Medical care
9,133
916
12,350
1,239
Non-invasive
Opportunity cost
4,572
24
5,921
31
Total cost
13,705
941
18,272
1,270
Medical care
26,951
2,455
36,445
3,320
Invasive
Opportunity cost
10,513
77
13,616
100
Total cost
37,463
2,532
50,061
3,420
Notes:
aThe estimates for non-invasive bladder cancer subtype were used to value local, regional, and unstaged bladder cancer morbidity reductions, while the estimates for the
invasive bladder cancer subtype were used to value distant bladder cancer morbidity reductions
bThe estimates come from Greco et al. (2019).
cTo adjust for inflation, EPA used U.S. Bureau of Labor Statistics Consumer Price Index for All Urban Consumers: Medical Care Services in U.S. (City Average).
Proposed PFAS Rule Economic Analysis
6-105
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
6.7.3 Results
Table 6-44 to Table 6-47 provide the health effects avoided and valuation associated with
bladder cancer.
Table 6-44: National Bladder Cancer Benefits, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
Bladder Cancer Cases
Avoided
4,079.1
5,238.6
6,475.3
4,079.1
5,238.6
6,475.3
Number of Bladder Cancer-
Related Deaths Avoided
1,436.0
1,844.4
2,280.0
1,436.0
1,844.4
2,280.0
Total Annualized Bladder
Cancer Benefits (Million
$2021)b
$173.09
$221.30
$273.62
$102.08
$130.63
$161.56
Notes:
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-45: National Bladder Cancer Benefits, Option la (PFOA and PFOS MCLs of 4.0
ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
Bladder Cancer Cases
Avoided
4,066.1
5,219.4
6,488.8
4,066.1
5,219.4
6,488.8
Number of Bladder Cancer-
Related Deaths Avoided
1,431.5
1,837.6
2,284.9
1,431.5
1,837.6
2,284.9
Total Annualized Bladder
Cancer Benefits (Million
$2021)b
$171.72
$220.48
$274.24
$101.34
$130.15
$161.56
Notes:
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Proposed PFAS Rule Economic Analysis
6-106
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-46: National Bladder Cancer Benefits, Option lb (PFOA and PFOS MCLs of 5.0
ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
Bladder Cancer Cases
Avoided
3,342.7
4,334.3
5,382.5
3,342.7
4,334.3
5,382.5
Number of Bladder Cancer-
Related Deaths Avoided
1,176.8
1,526.0
1,895.3
1,176.8
1,526.0
1,895.3
Total Annualized Bladder
Cancer Benefits (Million
$2021)b
$141.17
$183.10
$227.85
$83.31
$108.08
$135.37
Notes:
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable benefits, and the potential direction of impact these benefits would have on the
estimated monetized total annualized benefits in this table.
Table 6-47: National Bladder Cancer Benefits, Option lc (PFOA and PFOS MCLs of
10.0 ppt)
3% Discount Rate 7% Discount Rate
Benefits Category
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Number of Non-Fatal
Bladder Cancer Cases
Avoided
1,615.9
2,175.5
2,807.4
1,615.9
2,175.5
2,807.4
Number of Bladder Cancer-
Related Deaths Avoided
568.9
766.0
988.6
568.9
766.0
988.6
Total Annualized Bladder
Cancer Benefits (Million
$2021)b
$68.26
$91.90
$118.64
$40.29
$54.25
$70.10
Notes:
aThe 5th and 95th percentile range is based on modeled variability and uncertainty. This range does not include the uncertainty
described in Table 6-48.
bSee Table 7-6 for a list of the nonquantifiable costs, and the potential direction of impact these costs would have on the
estimated monetized total annualized costs in this table.
6.8 Limitations and Uncertainties of the Benefits Analysis
This section describes limitations of the quantified benefits analysis, along with uncertainties that
could not be modeled quantitatively as part of the national benefits analysis. The sources of
uncertainty characterized quantitatively are presented in Section 6.1.2. In the tables below, EPA
summarizes limitations and uncertainties that apply to:
• All quantitative benefits analyses implemented for the proposed PFAS rule (Table 6-48);
• Application of PK models for blood serum PFAS concentration estimation (Table 6-49);
Proposed PFAS Rule Economic Analysis
6-107
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Developmental effects (i.e., infant birth weight) modeling (Table 6-50);
• CVD impacts modeling (Table 6-51);
• RCC impacts modeling (Table 6-52); and
• Modeling of bladder cancer impacts from GAC treatment related THM4 reductions
(Table 6-53).
EPA notes that in most cases it is not possible to judge the extent to which a particular limitation
or uncertainty could affect the magnitude of the estimated benefits. Therefore, in each table
below, EPA notes the potential direction of the impact on the quantified benefits (e.g., a source
of uncertainty that tends to underestimate quantified benefits indicates expectation for larger
quantified benefits) but does not prioritize the entries with respect to the impact magnitude.
Table 6-48: Limitations and Uncertainties that Apply to Benefits Analyses Considered for
the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
EPA has quantified benefits
for three health endpoints
for PFOA and PFOS.
Underestimate
For various reasons, EPA has not quantified the benefit of
removing PFOA and PFOS from drinking water for most
of the health endpoints PFOA and PFOS are expected to
impact. See discussion in Section 6.2.2. for more
information about these nonquantifiable benefits.
EPA has quantified benefits
for one co-removed
contaminant group
(THM4).
Underestimate
Treatment technologies that remove PFAS can also
numerous other contaminants, including some other
PFAS compounds, additional regulated and unregulated
DBPs, heavy metals, organic contaminants, pesticides,
among others. These co-removal benefits may be
significant, depending on co-occurrence, how many
facilities install treatment and which treatment option
they select.
EPA has not quantified
benefits for any health
endpoint for PFHxS,
PFNA, PFBS, and HFPO-
DA.
Underestimate
PFHxS, PFNA, PFBS, and HFPO-DA each have
substantial health impacts on multiple health endpoints.
The analysis does not
explicitly consider changes
in PFOA/PFOS and THM4
concentrations for systems
that purchase their drinking
water from other PWSs.
Uncertain
Many PWSs purchase their primary source water from
PWSs that are likely to implement treatment under the
rule. The SDWIS/Fed inventory of PWSs includes these
systems with their retail populations instead of allocating
those populations to the wholesale systems. The MCMC
occurrence analysis outputs for the wholesale system and
purchasing system may vary from one another, resulting
in either an under- or over-estimate of affected population
in any iteration. The net effect on total benefits is
uncertain.
The analysis does not
account for populations that
consume bottled water as
their primary drinking
water source.
Uncertain
Studies indicate that between 13 percent and 33 percent
of the U.S. population consumes bottled water as their
primary drinking water source (Z. Hu et al., 2011;
Rosinger et al., 2018; Vieux et al., 2020). The benefits
models do not consider these populations. This could
result in an overestimate of avoided cases of health
effects and associated benefits. However, bottled water
Proposed PFAS Rule Economic Analysis
6-108
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-48: Limitations and Uncertainties that Apply to Benefits Analyses Considered for
the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
consumers can also be CWS customers and may still be
exposed to PFAS by using water for cooking etc.,
therefore, would benefit from PFAS removal. (U.S. FDA,
2022; Aquafina, 2022). Finally, the benefits may also be
underestimated because those using bottled water as a
primary drinking water source may switch to CWS supply
as a result of the proposed rule; EPA did not model this
behavioral response and hence the benefits do not account
for the potential cost savings to those consuming bottled
water at baseline.
The analysis considers
PFOA/PFOS
concentrations from
NTNCWSs.
Overestimate
Some SDWIS population served estimates for NTNCWSs
represent the both the population that has regular
exposure to the NTNCWS' drinking water (e.g., the
employees at a location) and the peak day transient
population (e.g., customers) who have infrequent
exposure to the NTNCWS' drinking water. Estimating the
demographic distribution and the share of daily drinking
water consumption for these two types of NTNCWS
populations would be difficult across many of the
industries which operate NTNCWSs. The inclusion of
NTNCWS results is an overestimate of benefits because
daily drinking water consumption for these populations is
also modeled at their residential CWS.
EPA assumes that the
effects of PFOA and PFOS
exposures are
independent.
Uncertain
The exposure-response functions used in benefits
analyses assume that the effects of serum PFOA/PFOS on
the health outcomes considered are independent and
therefore additive. Due to limited evidence, EPA does not
consider synergies or antagonisms in PFOA/PFOS
exposure-response.
The derivation of
PFOA/PFOS exposure-
response functions for the
relationship between
PFOA/PFOS serum and
associated health outcomes
assumes that there are no
threshold serum
concentrations below which
effects do not occur.
Overestimate
The new data and EPA's proposed MCLGs indicate that
the levels at which adverse health effects could occur are
much lower than previously understood when EPA issued
the 2016 health advisories for PFOA and PFOS (70 parts
per trillion or ppt) - including near zero for certain health
effects. Therefore, the exposure-response functions used
in benefits analyses assume that there are no threshold
serum concentrations below which effects do not
occur. This could result in a slight overestimate of
benefits for certain health endpoints.
The exposure-response
functions used to estimate
risk assume causality.
Overestimate
Analyses evaluating the evidence on the associations
between PFAS exposure and health outcomes are ongoing
and EPA has not conclusively determined causality. As
described in Section 6.2, EPA modeled health risks from
PFOA/PFOS exposure for endpoints for which the
evidence of association was found to be likely. These
endpoints include birth weight, TC, and RCC. While the
evidence supporting causality between DBP exposure and
bladder cancer has increased since EPA's Stage 2 DBP
Proposed PFAS Rule Economic Analysis
6-109
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-48: Limitations and Uncertainties that Apply to Benefits Analyses Considered for
the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Rule (NTP, 2021; Weisman et al., 2022), causality has
not yet been conclusively determined (Regli et al., 2015).
The analysis assumes that
quantified benefits
categories are additive.
Uncertain
EPA did not model birth weight, CVD, RCC, and bladder
cancer benefits jointly, in a competing risk framework.
Therefore, reductions in health risk in a specific benefits
category do not influence health risk reductions in another
benefits category. For example, lower risk of CVD and
associated mortality implies a larger population that could
benefit from cancer risk reductions, because cancer
incidence grows considerably later in life.
The scope of the analysis
does not include intra- or
international migration
throughout the evaluation
period.
Uncertain
Throughout the analysis period people may migrate from
one place to another. If persons migrate to locations with
larger decreases in PFOA/PFOS under the regulatory
alternative, EPA would be underestimating the impacts.
The opposite is true if persons migrate to locations with
smaller decreases in PFOA/PFOS under the regulatory
alternative.
The analysis does not take
into account population
growth and other changes
in long-term trends.
Underestimate
The benefits analysis does not reflect the effects of
growing population that may benefit from reduction in
PFOA/PFOS exposure. Furthermore, EPA uses present-
day information on life expectancy, disease,
environmental exposure, and other factors, which are
likely to change in the future.
There are two potential datasets that could inform
population growth under the final rule. EPA has described
these datasets below.
Population projections by year, county, single-year age,
sex, and race/ethnicity are available through 2050 from
the Woods & Poole Economics Inc. (2021) dataset and
could be used for the final rule. This dataset has been
used in prior rulemakings, such as the National Ambient
Air Quality Standards, the Steam Electric Effluent
Limitations Guidelines, and the Federal Recreational
Water Quality Criteria Applicable to Certain Waters in
New York (unpublished; currently on hold until January
2023 at the earliest). Woods & Poole Economics Inc.
(2021) population growth data are also used in EPA's air
quality benefits programs BenMAP-CE and COBRA.
EPA could project the county-, sex-, race/ethnicity-, and
age-specific distribution of Woods & Poole Economics
Inc. (2021) data from 2051 to 2104 using a transition ratio
approach with normalization to obtain population
projections throughout the period of analysis relevant to
the NPDWR.
Proposed PFAS Rule Economic Analysis
6-110
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-48: Limitations and Uncertainties that Apply to Benefits Analyses Considered for
the Proposed PFAS Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Additional population projection estimates are available
from the Socioeconomic Data and Applications Center
(SEDAC) by county, age, sex, and race/ethnicity in five-
year intervals through the year 2100. These projections
were used in EPA's recent Waters of the United States
rulemaking. If implemented in the PFAS NPD WR, EPA
would need to distribute population within five-year
intervals and project population estimates from 2101 to
2104.
The analysis does not
include the impacts of
COVID-19 on future
population health and
economic growth.
Uncertain
Impacts of the COVID-19 pandemic have had resulting
effects on conception, pregnancy, and birth rates (Aassve
et al., 2021; McLaren Jr et al., 2021; Ullah et al., 2020).
Some studies suggest that the economic recession caused
by the COVID-19 pandemic may impose long-term
impacts on fertility rates (McLaren Jr et al., 2021; Ullah
et al., 2020). Such impacts are not accounted for in this
benefits analysis.
For PWSs with multiple
entry points, the analysis
assumes a uniform
population distribution
across the entry points.
Uncertain
Data on the populations served by each entry point are not
available and EPA therefore uniformly distributes system
population across entry points. Effects of the regulatory
alternative may be greater or smaller than estimated,
depending on actual populations served by affected entry
points. For one large system serving more than one
million customers EPA has sufficient data on entry point
flow to proportionally assign effected populations.
Valuation of mortality risk
reductions assumes that per
capita income will grow at
the constant rate.
Uncertain
EPA uses Value of Statistical Life adjusted for income
growth to estimate economic value of the premature
mortality avoided in the future. Per capita income growth
projections were available through 2050. EPA estimated
the compound annual growth rate in per capita income
during 2023-2050 and applied it to project Value of
Statistical Life over the analysis period 2023-2104.
EPA does not characterize
uncertainty associated with
the Value of Statistical Life
reference value or Value of
Statistical Life elasticity.
Uncertain
EPA did not quantitatively characterize the uncertainty
for the Value of Statistical Life reference value and
income elasticity. Because the economic value of avoided
premature mortality comprises the majority of the overall
benefits estimate, not considering uncertainty surrounding
the Value of Statistical Life is a limitation.
Abbreviations: COVID-19 - coronavirus disease 2019; CVD - cardiovascular disease; CWS - community water system; DBP -
disinfection byproduct; MCLG - maximum contaminant level goal; PFOA - perfluorooctanoic acid; PFOS - perfluorooctane
sulfonic acid; PWS - public water system; RCC - renal cell carcinoma; RO - reverse osmosis; UCMR - Unregulated
Contaminant Monitoring Rule.
Proposed PFAS Rule Economic Analysis
6-111
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 6-49: Limitations and Uncertainties in the PK Model Application
Uncertainty/
Assumption
Effect on Benefits
Estimate
Notes
The benefits analysis
assumes that there are no
reductions in PFOA/PFOS
exposure from other
sources associated with
treatment-related
reductions in PFOA/PFOS
drinking water
concentrations.
Underestimate
Some portion of the non-drinking water PFOA/PFOS
exposure could be related to drinking water concentration
(e.g., food affected by water concentrations). This portion
is difficult to estimate, and, depending on the relationship,
there may be a time lag between the decrease in drinking
water concentration and the decrease in the non-drinking
water exposure.
The birth weight analysis
uses the adult female PK
model to estimate changes
in serum PFOA/PFOS
from changes in drinking
water PFOA/PFOS.
Overestimate
Evidence from epidemiology studies connects birth weight
to serum PFOA/PFOS levels throughout pregnancy:
The serum PFOS-birth weight slope factor in the birth
weight benefits module comes from the meta-analysis of
29 studies by Dzierlenga et al. (2020). Table 1 in
Dzierlenga et al. (2020) summarizes the timing of the
serum samples for the contributing studies, including pre-
pregnancy (2 studies), first trimester (6 studies), second
trimester (5 studies), third trimester (5 studies), and cord
blood samples/delivery (11 studies).3
The serum PFOA-birth weight slope factor comes from the
meta-analysis of 24 studies by Steenland et al. (2018).
Steenland et al. (2018) summarizes the timing of the serum
samples for the contributing studies, including pre-
pregnancy (2 studies), first trimester (4 studies), straddling
first and second trimester (1 study), second trimester (2
studies), straddling second and third trimester (2 studies),
third trimester (4 studies), and cord blood samples/delivery
(9 studies).13
Because the slope factors included epidemiological
evidence throughout pregnancy, a developmental version
of the PK model may be a more appropriate choice. A
developmental PK model would allow the observed
decrease in serum levels that occurs during pregnancy to
be captured by accounting for maternal physiological
changes. For example, Glynn et al. (2012) found a mean
decrease of 16 percent for PFOA and 11 percent for PFOS
between serial measurements taken in the 1st trimester and
3rd trimester of pregnancy. This decrease is associated
with increases in maternal plasma volume and transfer of
the chemicals to the placenta and fetus. EPA expects that
the use of the adult PK model overestimates the additive
difference in serum concentrations between baseline and
regulatory alternative (and, therefore, the birth weight
benefits of the regulatory alternative) because of the
expected larger volume of distribution for pregnant
females and, therefore, proportionally lower serum
concentrations.
Abbreviations: PFOA - perfluorooctanoic acid; PFOS - perfluorooctane sulfonic acid; PK - pharmacokinetic.
Notes:
Tor PFOS, EPA used 4 high confidence studies (Chu et al., 2020; Sagiv et al., 2018; Starling et al., 2017; and Wikstrom et al.,
2019) with a variety of PFOS exposure measures across the fetal and neonatal window. Sagiv et al. (2018) collected maternal
Proposed PFAS Rule Economic Analysis
6-112
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-49: Limitations and Uncertainties in the PK Model Application
Uncertainty/
Assumption
Effect on Benefits
Estimate
Notes
samples in trimester 1, while Wikstrom et al. (2020) collected them in trimesters 1 and 2. The samples from Starling et al. (2017)
were from trimesters 2 and 3, while Chu et al. (2020) collected exclusively in trimester 3. Of these studies, only Sagiv et al.
(2018) and Starling et al. (2017) were part of the Dzierlenga et al. (2020) meta-analysis.
bFor PFOA, EPA used 5 high confidence studies (Chu et al., 2020; Govarts et al., 2016; Sagiv et al., 2018; Starling et al., 2017;
and Wikstrom et al., 2020) with a variety of PFOA exposure measures across the fetal and neonatal window. Sagiv et al. (2018)
collected maternal samples in trimester 1, while Wikstrom et al. (2020) collected them in trimesters 1 and 2. The samples from
Starling et al. (2017) were from trimesters 2 and 3, while Chu et al. (2020) collected exclusively in trimester 3. The samples in
the Govarts et al. (2016) study were collected from umbilical cords. None of these studies were part of the Negri et al. (2017)
meta-analysis.
Table 6-50: Limitations and Uncertainties in the Analysis of Birth Weight Benefits Under
the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Characterizing the Exposed Population
The analysis does not
consider the effects of
PFOA/PFOS exposure on
fertility rates.
Uncertain
Studies have shown that exposure to PFAS may lead
to reduced fertility rates among women (Fei et al.,
2009; Waterfield et al., 2020), while the evidence
supporting PFAS effects on the male reproductive
system is inconclusive (C. C. Bach et al., 2016; U.S.
EPA, 2023d; U.S. EPA, 2023e). The birth weight
risk reduction analysis does not account for any
potential differences in birth rates among the
baseline and treatment scenario due to PFAS-related
changes in fertility.
EPA uses state-specific
birth rate data, distributed
based on census region-
level race/ethnicity-specific
birth rates, to determine the
share of infants born to
women of childbearing age
at each PWS and within
each 100 gram birth weight
increment.
Uncertain
County-level birth rates from CDC by 100 gram
birth weight increment are often tagged as
"unreliable" by CDC in cases where there are low
infant counts per birth weight increment. State-
specific 100 gram increment-specific birth rates may
not reflect the number of infants born in each 100
birth weight increment in PWS service area that is
affected by PFOA/PFOS through the pregnant
mother's ingestion of drinking water. Using state-
specific birth rates may over- or underestimate the
number of infants falling into each 100 gram birth
weight increment born to mothers who experience
PWS specific changes in drinking water
PFOA/PFOS levels. This in turn may over- or
underestimate benefits associated with changes in
PFOA/PFOS levels.
EPA uses state-specific
death rate data, distributed
based on national-level
race/ethnicity -specific
infant mortality rates, as
the baseline infant
mortality rate (i.e., number
Uncertain
State-specific death rates may not reflect the
baseline number of infants who die in each PWS that
is affected by PFOA/PFOS in mother's drinking
water. Using state-specific baseline death rates may
over- or underestimate the post-regulation death
rates determined using the birth weight-mortality
relationship and changes in birth weight, and result
Proposed PFAS Rule Economic Analysis
6-113
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-50: Limitations and Uncertainties in the Analysis of Birth Weight Benefits Under
the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
of deaths per 1,000 births)
of infants born to women
of childbearing age at each
PWS.
in an over- or underestimate of benefits associated
with changes in PFOA/PFOS levels.
Baseline infant death rates
per location are held
constant throughout the
years of the analysis.
Uncertain
Although changes in infant death rates may not be
consistent across race/ethnicity and location in the
US, medical advances in infant care will likely
reduce infant mortality in future years.
EPA uses county-specific
percentages of the
population that fall within
four race/ethnicity
categories (non-Hispanic
White, non-Hispanic Black,
Hispanic, and other) to
separate total PWS-specific
populations into race
categories for application
of the birth weight-
mortality marginal effects
estimates.
Uncertain
County-specific population percentages may not
accurately represent the race/ethnicity makeup of
PWS-level populations served. PWS populations
served may span multiple counties or may represent
a portion of a single county.
Modeling Changes in Health Risks
The analysis does not
model variability in
pregnancy stage-specific
serum PFOA/PFOS
concentrations and
exposure-response
relationships.
The studies estimating the link between maternal
serum PFOA/PFOS and infant birth weight use
serum PFOA/PFOS measurements from various
stages of pregnancy. EPA used a constant, adult PK
model-based estimate of serum PFOA/PFOS
concentration to represent exposure during
pregnancy, which is more consistent with early
pregnancy exposures and likely overestimates the
reduction in serum PFOA/PFOS exposure later in
pregnancy. In a sensitivity analysis (Appendix K),
EPA estimated birth weight benefits using exposure-
response functions that evaluated the association
between early pregnancy serum PFOA/PFOS and
birth weight. EPA found that using an early
pregnancy-based exposure-response function would
result in approximately a 60 percent reduction in
birth weight benefits.
Proposed PFAS Rule Economic Analysis
6-114
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-50: Limitations and Uncertainties in the Analysis of Birth Weight Benefits Under
the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
The analysis assumes that
birth weight changes
resulting from changes in
PFAS serum levels will not
exceed 200 g.
Underestimate
EPA places a cap on estimated birth weight changes
in excess of 200 g based on existing studies that
found that changes to environmental exposures
result in relatively modest birth weight changes
(Windham et al., 2008; Klein et al., 2018; Kamai et
al., 2019).
Economic Valuation of Changes in Health Risk
Some possible benefits
from increased birth weight
in infants are omitted from
the analysis.
Underestimate
Omitted benefit categories include reduction in IQ
loss, special education costs, early intervention
costs, and labor market productivity losses
associated with specific developmental diseases,
among others. EPA's analysis omitted these
categories because the available studies
documenting relationships between birth weight and
non-medical effects either did not identify methods
for determining the associated economic burden of
such effects or had other limitations such as older
(pre-2000s) data, limited geographical coverage,
small sample sizes, small ranges of birth weight
evaluated, performed outside of the U.S., or lack of
statistical significance. See ICF (2021) for additional
details.
The analysis does not
monetize medical treatment
costs for infants who die
within 1 year of birth.
Underestimate
This limitation likely results in an underestimate of
total benefits. The magnitude of this underestimate is
likely to be small because the number of infants who
do not survive represent a small percentage of the
total number of LB W infants. In addition, the
medical cost function is based on estimated
treatment expenses over a two-year period after birth
and thus EPA would have to scale down medical
costs to account for the distribution of infant death
timing within 1-year (e.g., within 28 days of birth or
3 months). Based on the 2016-2018 NCHS/NVSS
data, approximately 50 percent of LBW infant
deaths occur within the first 28 days of birth. Thus, it
is likely that only a small portion of medical costs
from Klein et al. (2018) is applicable to infants who
die within 1 year of birth.
Simulated medical cost
changes from Klein and
Lynch (2018) do not reflect
Uncertain
Preliminary modeling indicates that reductions in
PFOA/PFOS concentrations based on the regulatory
alternatives may lead to birth weight changes greater
Proposed PFAS Rule Economic Analysis
6-115
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-50: Limitations and Uncertainties in the Analysis of Birth Weight Benefits Under
the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
birth weight changes
greater than 100 g.
than 100 g. Although EPA caps birth weight change
estimates at 200 g, EPA uses the COI estimates
associated with a 100 g change in birth weight for all
birth weight changes between 100 and 200 g to
avoid extrapolation outside of the data range.
Abbreviations: birth weight - birth weight; CDC - Centers for Disease Control and Prevention; COI - cost of illness; g -
gram; LBW - low birth weight; NCHS - National Center for Health Statistics; NTNCWS - non-transient non-community
water system; NVSS - National Vital Statistics System; PFAS - per-and polyfluoroalkyl substances; PFOA -
perfluorooctanoic acid; PFOS - perfluorooctane sulfonic acid; PWS - public water system; SDWIS - Safe Drinking Water
Information System.
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Characterizing the Exposed Population
The analysis uses national-
level estimates of CVD
prevalence and incidence
rates, life tables, and
ASCVD model inputs (e.g.,
prevalence of treated and
untreated hypertension,
diabetes, smoking).
Uncertain
Using national-level baseline health data may over- or
underestimate the effects of regulatory alternatives on
CVD morbidity and mortality overall and in specific
PWSs.
The effects of statin use on
changes in CVD risk were
not modeled in this
analysis.
Uncertain
Because statin medications lower LDL cholesterol, statin
use may impact the relationship between serum
PFOA/PFOS levels and TC and, ultimately, the estimated
changes in CVD risk. EPA did not model population
variability with respect to this factor for two reasons.
First, as described in Appendix F, not all studies
modeling serum PFOA/PFOS levels and TC consider
and/or control for statin use. Exclusion of persons who
rely on statins for LDL control from the modeled
population would underestimate CVD benefits if serum
PFOA/PFOS-TC effect represents an average across
statin user and non-user groups. Second, there are
challenges in estimating statin use prevalence. Depending
on age, sex, race/ethnicity, and disease status,
approximately 20 percent-40 percent of the U.S.
population relies on statins (Robinson et al., 2010).
Factors such as overt CVD, healthcare, and demographics
are significantly associated with statin use (Leino et al.,
2020; Electricwala et al., 2020). While statin therapy is
intended to be permanent, many individuals who are
prescribed statins take them irregularly (Colantonio,
2019; Lewey et al., 2013; Ellis et al., 2004; Goldstein et
Proposed PFAS Rule Economic Analysis
6-116
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
al., 2016; Toth et al., 2019); Toth et al. (2019) found a
<25 percent rate of adherence 5 years after initiation of
therapy.
Modeling Changes in Health Risks
The analysis assumes that
there is no lag between
changes in serum
PFOA/PFOS
concentrations and changes
in TC and BP. Likewise,
the analysis assumes that
there is no lag between
changes in TC/BP and
changes in CVD risk.
Overestimate
The studies estimating the link between serum
PFOA/PFOS and TC/BP and the ASCVD model are not
dynamic, and hence do not provide insights into whether
TC/BP may respond gradually to changes in serum
PFOA/PFOS and/or if CVD risk may respond gradually
to changes in TC/BP. The analysis assumes immediate
adjustment, which may overestimate impacts to the
exposed population. Note, however, that reductions in
TC/BP and CVD risk do not instantaneously follow the
reductions in PFOA/PFOS drinking water concentrations,
because the reductions in serum PFOA/PFOS are gradual,
as predicted by the PK model.
The derivation of
PFOA/PFOS exposure-
response functions for the
relationship between
PFOA/PFOS serum and
TC levels assumes that the
studies used in the meta-
analysis represent the
PFOA/PFOS effects on
serum TC levels in general
population adults.
Uncertain
The exposure-response function was developed based on
six general population studies reporting linear serum
PFAS-TC level associations. Four of these studies were
high quality as reflected by the lower risk of bias
evaluations. These studies may not capture all possible
relationships between PFOA/PFOS and serum TC levels.
The analysis excludes
exposure-response
relationships between
serum PFOA/PFOS and
HDLC.
Uncertain
The relationship between serum PFOA/PFOS and HDLC
is uncertain. As shown in Section 6.5.2 and Appendix F,
the meta-analysis-based estimate of the effect of serum
PFOA/PFOS on HDLC concentration is positive but not
statistically significant. Single-study analyses of this
relationship have generated both positive (Dong et al.
(2019) serum PFOS-HDLC relationship) and negative
(Dong et al. (2019) serum PFOA-HDLC relationship, P.-
I. D. Lin et al. (2019) serum PFOA-HDLC and serum
PFOS-HDLC relationship) effect estimates that were not
statistically significant. To better understand the impact
of incorporating HDLC in the CVD risk model, EPA has
implemented a sensitivity analysis (see details in
Appendix K). EPA found that, using the meta-analysis
results, inclusion of HDLC would decrease benefits by
approximately 23-25%.
The analysis assumes that
the CVD risk impact of
changes in TC/BP from
Uncertain
While the CVD risk impacts of changes in TC/BP from
behavioral and medical interventions is well documented
(Lloyd-Jones et al., 2017), there is no information on
Proposed PFAS Rule Economic Analysis
6-117
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
reductions in serum
PFOA/PFOS is the same as
the CVD risk impact of
changes in these
biomarkers due to other
reasons such as behavioral
changes or medication.
whether changes in serum PFOS/PFOA leading to
changes in these biomarkers would result in similar
outcomes.
The CVD risk analysis
assumes that person's
TC/BP level history does
not have an impact on
changes in CVD risk due to
changes in the levels of
these biomarkers.
Uncertain
The ASCVD model links TC/BP levels at the start of the
10-year follow-up period to first hard CVD event
incidence during the follow-up period. The modeling
does not account for TC/BP changes over time, which
could have an impact on the CVD event risk.
The AS CVD model was
not recalibrated for the
contemporary CVD
incidence and prevalence.
Overestimate
Assessments of ASCVD risk model performance across
different sociodemographic subgroups (Asian
populations, Hispanic populations, persons with high
levels of CVD risk, diabetes, older adults with frailty and
multimorbidity, smokers, and women) indicated that the
model tended to overestimate risk but suggested that the
model may improve through additional input variables
and recalibration given contemporary CVD incidence and
prevalence (Mora et al., 2018; Muntner et al., 2014).
The analysis uses the
ASCVD model developed
for non-Hispanic Black
populations to assess
potential CVD risks for
race/ethnicity groups other
than non-Hispanic Black
and non-Hispanic White
populations.
Uncertain
The ASCVD model documentation encourages the use of
equations for non-Hispanic White populations for other
race/ethnicity categories, specifying that estimated risks
may be biased upward, especially for Hispanic and Asian
American populations. EPA's model validation analysis
detailed in Appendix G shows that the non-Hispanic
Black model is a better fit for these race/ethnicity groups.
However, the ultimate impact of this assumption is
uncertain.
EPA uses the fraction of
the population who smokes
and has diabetes as inputs
into the ASCVD model.
Underestimate
The ASCVD model uses binary values to indicate
whether a person is a current smoker or has diabetes.
EPA simplifies calculations by using the fraction of the
population who smokes and has diabetes as inputs to the
ASCVD model. EPA has implemented a targeted
evaluation of the effect of this assumption and confirmed
that this simplification likely underestimates impacts by
approximately 5 percent to 10 percent, depending on the
age group, due to the non-linearity of the estimated
model.
The analysis assumes that
the threshold for high BP is
a systolic/diastolic
measurement of 140/90.
Underestimate
In November 2017, the threshold defined for high BP was
reduced to 130/80. The analysis relies on high BP
prevalence data and treated, untreated, and normal BP
measurements that are based on NHANES surveys from
2011 to 2016. Therefore, EPA adheres to the pre-2017
threshold. Furthermore, the ASCVD model was
developed prior to the change in high BP definition.
Proposed PFAS Rule Economic Analysis
6-118
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Adhering to the pre-2017 threshold may affect the
number of people sorted into the high BP population
category, potentially underestimating CVD risk.
The analysis assumes
independence among the
prevalence of high BP,
smoking, and diabetes.
Overestimate
Smoking and high BP are often related, and smoking is a
risk factor for Type 2 diabetes. Assuming independence
among the prevalence of high BP, smoking, and diabetes
may result in overestimated CVD risk impacts.
The analysis assumes that
deaths from causes other
than hard CVD events
occur first.
Underestimate
By assuming that deaths from causes other than hard
CVD events occur first, EPA underestimates the eligible
population (e.g., population without CVD history)
evaluated for the first hard CVD event estimation.
The analysis does not
account for survivors of
first hard CVD events that
are neither MI nor IS. The
analysis does not account
for persons who were
younger than 40 years at
the time of their first hard
CVD event.
Underestimate
The ASCVD model captures risk of non-fatal MI, non-
fatal IS, and fatal CVD; however, it does not capture
other non-fatal CHD. The ASCVD model can be used to
predict the annual probability of a first hard CVD event
for persons aged 40-89 years; EPA applied this model to
populations aged 40 years and older. The prevalence of
CVD history before age 40 is low (<7% based on
estimates from the Medical Expenditure Panel Survey)
and likely includes persons whose CVD arises from
genetic factors (Zhang et al., 2019). Early life PFAS
exposures and TC are inconclusively associated for
PFOA and positively associated forPFOS (U.S. EPA,
2023d; U.S. EPA, 2023e). TC later in life is highly
positively correlated with early TC as seen in Pletcher et
al. (2016) and Zhang et al. (2019). This analysis does not
directly capture effects of early life increases in TC due to
PFAS exposures. The analysis does capture the effects of
early life TC indirectly to the extent that early and later in
life TC levels are correlated.
The analysis does not
capture post-acute CVD
mortality beyond 5 years of
the first MI or IS for those
ages 40-65 at the time of
the initial event nor does it
capture post-acute CVD
mortality beyond 6 years of
the first MI or IS for those
ages 66-89 at the time of
the initial event.
Underestimate
The risk of post-acute CVD mortality was estimated
based on Thom et al. (2001) for those aged 40-65 years
and on S. Li et al. (2019) for those older than 65 years.
Neither study reported post-acute mortality information
for a longer follow-up period. The reported information
does not support complete post-acute mortality risk
elimination beyond the longest follow-up period. EPA
did not identify U.S. population-based MI/IS survivor
studies that had a longer follow-up time and, thus, has no
reliable quantitative basis to estimate post-acute mortality
impacts beyond 6 years of the initial event.
The analysis assumes that
post-acute CVD mortality
for survivors of IS at ages
40-65 is the same as post-
acute CVD mortality for
survivors of MI at ages 40-
65.
Uncertain
Post-acute mortality estimates for IS and MI were very
close in the Medicare population (S. Li et al., 2019). For
those aged 65 years or older, S. Li et al. (2019) have
estimated the probability of death within 1 year after non-
fatal IS to be 32.07 percent and the probability of death
within 1 year after non-fatal MI to be 32.09 percent.
Proposed PFAS Rule Economic Analysis
6-119
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-51: Limitations and Uncertainties in the Analysis of CVD Benefits Under the
Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Therefore, reliance on the post-acute mortality for MI to
approximate the same for stroke is reasonable.
The analysis models the
85+ year old group jointly
and applies average
mortality rate for those
aged 85+ in this age group.
Uncertain
The effect of this modeling approximation on the CVD
benefits is not certain because the integer age-specific
mortality rates may be above or below the average
mortality rate.
The analysis models the
85+ year old group jointly
and uses serum
PFOA/PFOS estimates for
age 85 in initiate
calculations in this age
group.
Underestimate
Because the impacts of changes in PFOA/PFOS drinking
water concentrations on serum PFOA/PFOS levels
increase over time, the use of serum PFOA/PFOS
concentrations at 85 years to model the 85+ age group
will underestimate the CVD risk impacts in this group.
The analysis applies the
ASCVD model to those
older than 80 years.
Overestimate
The ASCVD model evaluates first hard CVD event risk
for adults aged 40-80. Applying the predicted hard CVD
event risk for those aged 80 years or older results in an
overestimate of benefits.
EPA does not characterize
uncertainty associated with
ASCVD model parameters.
Uncertain
EPA treats the coefficients of the ASCVD risk model as
certain. However, uncertainty surrounding race/ethnicity-
and sex-specific ASCVD model parameters could be
characterized by multivariate normal distribution using
the ASCVD model coefficient estimates, and the
variance-covariance matrix shared by the ASCVD model
authors. Assuming that ASCVD model parameters are
certain is a limitation of this analysis.
Economic Valuation of Changes in Health Risk
The analysis monetized
changes in non-fatal first
MI/IS risk using medical
expenditures that do not
cover long-term
institutional or at-home
care. Furthermore, the COI
estimates do not include
lost productivity. Finally,
the COI-based approach
does not account for the
pain and suffering
associated with non-fatal
CVD events.
Underestimate
This analysis likely understates morbidity benefits since
hard CVD events, particularly IS, require a longer
rehabilitation period. According to HCUP 2017 data, 65
percent of IS survivors and 33 percent of MI survivors
are discharged to a long-term care facility or to a home
healthcare setting. Lost productivity impacts are also
likely (Cropper et al., 2000; Skolarus et al., 2014). MI/IS
survivors also experience significant reductions in the
health-related quality of life (J. P. Bach et al., 2011;
Kirchberger et al., 2020; Martino Cinnera et al., 2020;
Mollonetal., 2017).
Abbreviations: ASCVD - Atherosclerotic cardiovascular disease; BP - blood pressure; CVD - cardiovascular disease; HDLC -
high-density lipoprotein; IS - ischemic stroke (ICD9=433, 434; ICD10=I63); MI - myocardial infarction (ICD9=410;
ICD 10=121); NHANES - National Health and Nutrition Examination Survey; PFOA - perfluorooctanoic acid; PFOS -
perfluorooctane sulfonic acid; TC - total cholesterol.
Proposed PFAS Rule Economic Analysis
6-120
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-52: Limitations and Uncertainties in the Analysis of RCC Benefits Under the
Proposed Rule
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
Characterizing the Exposed Population
The analysis uses national-level
estimates of kidney cancer
incidence, prevalence, stage
distribution, and relative survival
data, as well as national-level life
tables.
Uncertain
Using national-level baseline health data may
over- or underestimate the effects of regulatory
alternatives on RCC morbidity and mortality in
specific PWSs and well as overall.
EPA assumed that RCC comprises
90 percent of kidney cancer
incidence.
Uncertain
Because baseline RCC incidence statistics are
not readily available from the National Cancer
Institute public use data, EPA used kidney
cancer statistics in conjunction with an
assumption that RCC comprises 90 percent of
all kidney cancer cases to estimate baseline
lifetime probability of RCC. This assumption
was used in RCC exposure-response modeling
by U.S. EPA (2023e).
RCC risks are estimated for
populations for which reductions in
PFOA exposures relative to
baseline exposures start at different
ages, including children.
Uncertain
The relative cancer potency of PFOA in children
is unknown, which may bias benefits estimates
either upward or downward. Because RCC
incidence in children is very small, we assess
any bias to be negligible.
Modeling Changes in Health Risks
The analysis assumes that the
magnitude of RCC risk reductions
resulting from reductions in serum
PFOA levels will not exceed a PAF
of 3.94 percent.
Uncertain
EPA placed a cap of 3.94 percent on the
magnitude of the estimated cumulative RCC risk
reduction resulting from reductions in serum
PFOA levels, based on its analysis of PAF
values found in the literature on environmental
contaminants and cancers (ICF, 2022b). This
review found that changes in environmental
exposures result in relatively modest PAFs
(between 0.2 percent and 17.9%); however, few
of the studies provided PAFs related specifically
to RCC or kidney cancer. EPA characterized the
uncertainty surrounding this parameter using a
log-uniform distribution with a minimum of 0.2
percent and a maximum of 17.9 percent. For the
central estimate of RCC benefits, EPA used a
PAF of 3.94 percent, which is the mean of the
PAF uncertainty distribution. As such, EPA
assumed that RCC risk reduction estimates in
excess of the PAF are unreasonable even as a
result of large changes in serum PFOA
concentrations. Because this PAF cap is not
based on RCC studies specifically, it is
uncertain whether the RCC impacts are under-
or overestimated.
Proposed PFAS Rule Economic Analysis
6-121
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-52: Limitations and Uncertainties in the Analysis of RCC Benefits Under the
Proposed Rule
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
The analysis assumes that there is
no lag between changes in serum
PFOA concentrations and changes
in RCC incidence.
Overestimate
The studies estimating the link between serum
PFOA and RCC are not dynamic, and hence do
not provide insights into whether RCC incidence
may respond gradually to changes in serum
PFOA. The PK model estimates daily serum
levels, which are averaged annually for the
purposes of modeling gradual serum changes for
the RCC risk reduction analysis. The RCC risk
reduction analysis assumes immediate RCC
incidence adjustment within each year, which
may overestimate impacts to the exposed
population.
The analysis relies on public-access
SEER 18 10-year relative kidney
cancer survival data to model
mortality patterns in the kidney
cancer population.
Uncertain
Reliance on these data generates both a
downward and an upward bias. The downward
bias is due to the short, 10-year excess mortality
follow-up window. Survival rates beyond 10
years following the initial diagnosis are likely to
be lower. The upward bias comes from the
inability to determine how many of the excess
deaths were deaths from kidney cancer.
The analysis assumes that RCC
incidence patterns and survival are
reasonably approximated by the
kidney cancer statistics.
Uncertain
The exposure-response function provides
information on changes in RCC risk, while
detailed race/ethnicity-, sex-, and age-specific
cancer incidence, stage, and survival
information is available for kidney cancer only.
For consistency with the RCC exposure-
response modeling (U.S. EPA, 2023e), EPA
assumed that RCC comprises 90 percent of
kidney cancer cases. In absence of RCC-specific
detailed information, the model relies on
patterns based on kidney cancer statistics.
The analysis models the 85+ year
old group jointly and applies the
average mortality rate for those
aged 85+ in this age group.
Uncertain
The effect of this modeling approximation on
the RCC benefits is not certain because integer
age-specific mortality rates may be above or
below the average mortality rate.
The analysis models the 85+ year
old group jointly and uses serum
PFOA estimates for those aged 85
to initiate calculations in this age
group.
Underestimate
Because the impacts of changes in PFOA
drinking water concentrations on serum PFOA
levels increase over time, the use of serum
PFOA concentrations at 85 years to model the
85+ age group will underestimate the RCC risk
impacts in this group.
Proposed PFAS Rule Economic Analysis
6-122
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-52: Limitations and Uncertainties in the Analysis of RCC Benefits Under the
Proposed Rule
Uncertainty/Assumption
Effect on Benefits
Estimate
Notes
Economic Valuation of Changes in Health Risk
RCC morbidity valuation is based
on medical costs associated with
the first line treatment that resulted
in the most cost-effective treatment
sequences, as reported in
Ambavane et al. (2020).
Uncertain
The valuation is biased downward because it
does not account for (1) the second line
treatments that may also be applied; (2) lost
productivity by the person experiencing RCC
and family caregivers; and (3) the pain and
suffering associated with experiencing RCC
and/or adverse effects of RCC treatment. The
valuation is biased upward because (1) the full
year-specific cancer treatment is assumed to
occur prior to the year-specific cancer
population death; and (2) the first line treatment
may be discontinued prior to the assumed
maximum treatment duration of 2 years. The
effect of using costs associated with the most
cost-effective treatment from Ambavane et al.
(2020) rather than costs for treatments currently
prevalent in clinical practice is uncertain. EPA
could not assess the impact of this assumption
because EPA is not aware of publicly available
information on the frequency of various kidney
cancer treatments in the U.S. population.
Abbreviations: PFOA -
cell carcinoma.
perfluorooctanoic acid; PFOS - perfluorooctane sulfonic acid; PK - pharmacokinetic; RCC - renal
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Modeling Reduced THM4 in PWSs
Reductions in THM4 formation
depend only on the relationship
between raw water TOC levels and
THM4 levels as estimated in the
1998 DBP ICR. Other source water
quality parameters were not
modeled.
Uncertain
EPA assumes that PWSs affected by
implementation of PFAS treatment technologies
have similar characteristics as those evaluated in
the 1998 DBP ICR. Source water parameters
and treatments at individual plants may have
changed over time.
Proposed PFAS Rule Economic Analysis
6-123
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
EPA uses available TOC data to
estimate reduced THM4
concentration.
Uncertain
Due to the lack of site-specific information on
factors affecting THM4 formation at each
potentially affected drinking water treatment
plant, EPA uses relationships between TOC
levels and changes in THM4 levels among
GAC-treating systems from the 1998 DBP ICR
and median raw water TOC levels for each
source water type from the 2019 SYR4 dataset.
Actual changes in THM4 concentrations for a
given change in treatment at any specific PWS
could be higher or lower than that estimated
using EPA's approach.
EPA assigned TOC values at the
system level based on Ground
Water or Surface Water
distributions.
Uncertain
Because the TOC levels for all systems is not
available, EPA used TOC data provided by
states in response to the fourth Six-Year Review
to derive TOC probability distributions for
influent into a PFAS treatment process; one
distribution for Ground Water systems and
another for Surface Water systems. EPA
randomly assigned values from these
distributions to each Ground Water or Surface
Water system, respectively. The actual TOC
values may be higher or lower than the assigned
values. For systems using GAC for PFAS
removal, the corresponding impact would be
under-stating or over-stating costs.
EPA estimates THM4 reduction
based on free chlorine formation
potential but does not estimate the
reduction based on chloramine use.
Overestimate
The 1998 DBP ICR TSD provided information
for systems that only used free chlorine as a
disinfectant and did not capture THM4
reduction in chloraminating systems. This
limitation likely leads to an overestimate of
THM4 formed in systems that used chloramines
in the distribution system because THM4
formation within the distribution system is lower
when chloramines are used, compared to when
free chlorine is used (Hua et al., 2008). Based on
SYR3 data, 36 percent of surface water systems
and 4 percent of ground water systems use
chloramination(U.S. EPA, 2016j). Chloramines
may produce greater amounts of genotoxic and
carcinogenic DBPs, but a reduction in the TOC
prior to disinfection will also yield a reduction
in DBP formation (Cuthbertson et al., 2019).
Proposed PFAS Rule Economic Analysis
6-124
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
THM4 is assumed to be a surrogate
for other chlorination DBPs, some
of which are more genotoxic and
cytotoxic than THMs.
Uncertain
EPA's analysis relies on the slope factor from
Regli et al. (2015), which links lifetime risk of
bladder cancer to THM4 concentrations in
finished water. Regli et al. (2015) did not
explicitly account for brominated or nitrogenous
DBPs, but instead used THM4 as a surrogate for
the broad suite of chlorination DBPs. This is
consistent with the approach used in numerous
epidemiolocal studies (Costet et al., 2011;
Freeman et al., 2017) since insufficient data are
available to estimate the co-occurrence and co-
removal of specific genotoxic or cytotoxic
DBPs.
EPA estimates THM4 reduction
based on GAC use but does not
estimate the reduction in individual
THM4 species.
Uncertain
GAC has been shown to shift the speciation
among THM4 and can result in a relatively
larger fraction of brominated species (THM3)
compared to chloroform. However, studies show
that even as speciation shifts, the absolute
concentrations of each species is reduced
(Cuthbertson et al., 2019; L. Wang et al., 2019).
The logistic model uses
pilot/RSSCT results to predict
ATHM4.
Overestimate
RSSCTs may overpredict full-scale adsorption
capacity of GAC (Kempisty et al., 2022;
Zachman et al., 2010)
SYR4 Comparison
Estimates of reductions in THM4
formation assume that GAC
treatment is the only treatment
change in a distribution system.
Uncertain
Uncertainty exists if other changes (i.e., new
source water, chemical dosing, other treatments
added such as pre-chlorination, existing
treatments changed such as new filter media)
that could have been made in public water
systems beyond GAC treatment could
potentially over- or underestimate THM4
reduction.
EPA analyzed only systems that
were sampled under UCMR 3 and
indicated GAC treatment under
UCMR 4.
Uncertain
Assessing only UCMR GAC systems limited the
sample to PWS serving > 10,000 people.
Therefore, EPA was unable to compare THM4
reduction estimates to measured data for small
systems.
EPA relied on available CCRs to
estimate the GAC treatment start
date to determine before and after
treatment years.
Uncertain
Available CCRs were used to inform the GAC
start date. When CCRs were unavailable, EPA
searched the web to identify information about
the timeline of treatment for individual PWSs.
While installation dates were found, the exact
date for when the GAC systems went into full-
scale use was not always specified.
Proposed PFAS Rule Economic Analysis
6-125
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
EPA obtained THM4 values from
multiple data sources.
Uncertain
For PWSs that met criteria outlined in Section
6.7.1.3.2 but had no THM4 data available in
SYR4, EPA relied on CCR THM4 data.
Reporting on THM4 levels is inconsistent across
CCRs. If a CCR listed "Amount Detected"
instead of the THM4 average, then EPA used
the "Amount Detected" value to represent the
THM4 average.
Characterizing the Exposed Population
Analysis assumes that systems
implementing IX do not accrue
benefits associated with bladder
cancer risk reductions.
Underestimate
Systems using IX for PFAS removal will also
benefit from some TOC removal, but the
removal will be limited in comparison to GAC
treatment because PFAS-selective IX can show
preferential removal of PFAS over organic
matter (de Abreu Domingos et al., 2018).
The analysis does not model
location-specific demographics.
Uncertain
Because EPA models impacts to aggregate
populations based on systems triggered into
treatment under various scenarios, EPA relies on
national-level demographic and bladder cancer
data. The impact of this limitation is uncertain.
For instance, populations with a large portion of
elderly or male individuals will be more
sensitive to changes in THM4 levels due to the
high baseline bladder cancer incidence among
elderly and male populations, compared to
younger and female populations.
The analysis does not model
variability by race/ethnicity.
Uncertain
Because EPA models impacts based on a
national-level distribution of finished water
TOC levels, specific TOC levels at actual PWSs
are not available. Therefore, these impacts were
not included in EPA's DBP analysis.
Accordingly, EPA did not pursue race/ethnicity-
specific modeling of health risk because it
would not provide meaningful insight into
distributional effects.
Bladder cancer risks are estimated
for populations for which
reductions in THM4 exposures
relative to baseline exposures start
at different ages, including
children.
Uncertain
The relative cancer potency of THM4 in
children is unknown, which may bias estimates
either upward or downward. Past reviews found
no clear evidence that children are at greater risk
of adverse effects from bromoform or
dibromochloromethane exposure (U.S. EPA,
2005a), although certain modes of action and
health effects may be associated with exposure
to THM4 during childhood (U.S. EPA, 2016g).
Because bladder cancer incidence in children is
very small, EPA assesses any bias to be
negligible.
Proposed PFAS Rule Economic Analysis
6-126
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
Modeling Changes in Health Risks
Analysis assumes an immediate
and full reduction in bladder cancer
risk following THM4 exposure
reduction.
Overestimate
EPA did not model the transitional dynamics in
relative annual risk of bladder cancer following
the THM4 exposure reduction. Regli et al.
(2015) do not provide pertinent information; as
such, this is a cross-sectional analysis
quantifying the relationship between lifetime
cancer risk and lifetime average exposure.
Existing cancer risk cessation lag studies
focused on smoking and arsenic exposure (e.g.,
Hrubec et al., 1997, Hartge et al., 1987, and C.
W. Chen et al., 2003); show that, annual cancer
risk drops within the first 25 years after
exposure cessation, yet it may never reach the
annual cancer risk of persons who were always
exposed to the treatment contaminant levels. In
EPA's modeling this issue pertains to those alive
at the start of the evaluation period who have
been exposed to the pre-treatment THM4 levels
for a considerable amount of time, such as
persons older than 60 years at the start of the
evaluation period. This subpopulation comprises
approximately 20 percent of the affected
population alive in 2023.
The analysis relies on public-access
SEER 18 10-year relative bladder
cancer survival data to model
mortality patterns in the bladder
cancer population.
Uncertain
Reliance on these data generates both a
downward and an upward bias. The downward
bias is due to the short, 10-year excess mortality
follow-up window. Survival rates beyond 10
years following the initial diagnosis are likely to
be lower. The upward bias comes from the
inability to determine how many of the excess
deaths were deaths from bladder cancer.
The relationship from Regli et al.
(2015) is a linear approximation of
the odds ratios reported in
Villanueva et al. (2004).
Uncertain
Given the uncertainty about the historical,
location-specific THM4 baselines, Regli et al.
(2015) provides a reasonable approximation of
the risk. However, depending on the baseline
THM4 exposure level, the impact computed
based on Regli et al. (2015) may be larger or
smaller than the impact computed using the
Villanueva et al. (2004)-reported odds ratios
directly.
Proposed PFAS Rule Economic Analysis
6-127
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 6-53: Limitations and Uncertainties in the Analysis of DBP Quantified Benefits
Under the Proposed Rule
Uncertainty/ Assumption
Effect on Benefits
Estimate
Notes
The analysis does not apply a PAF-
based cap on the magnitude of
bladder cancer relative risk
reductions from reductions in
THM4 exposure.
Overestimate
While, for the RCC analysis, EPA placed a cap
of 3.94 percent on the magnitude of the
estimated cumulative RCC risk reduction
resulting from reductions in serum PFOA levels,
a similar cap was not implemented for the
bladder cancer model. This is because the
relative bladder cancer risk reductions from
reductions in THM4, estimated in this analysis,
have been modest, generally not exceeding 4
percent. Because the PAF cap developed by
EPA is not based on bladder cancer studies
specifically, it is uncertain to what extent the
bladder cancer impacts may have been
overestimated.
Economic Valuation of Changes in Health Risk
Bladder cancer morbidity valuation
is based on medical costs and
indirect/time costs (by cancer
stage), as reported in Greco et al.
(2019).
Uncertain
The valuation is biased downward because it
does not account for (1) lost productivity by the
family caregivers and volunteers; (2) broader
labor market participation effects for those
experiencing bladder cancer and/or providing
care; and (3) the pain and suffering associated
with experiencing bladder cancer and/or adverse
effects of bladder cancer treatment. The
valuation is biased upward because (1) the full
year-specific cancer treatment is assumed to
occur prior to the year-specific cancer
population death; and (2) the treatment may be
discontinued if it is no longer effective.
Abbreviations: CCR - consumer confidence reports; DBP - disinfection byproduct; GAC - Granular Activated Carbon; ICR -
Information Collection Request; PFAS - per-and polyfluoroalkyl substances; PFOA - perfluorooctanoic acid; PFOS -
perfluorooctane sulfonic acid; PWS - public water system; SYR - Six Year Review; THM4 - Four Regulated
Trihalomethanes; TOC - Total Organic Carbon; TSD - Treatment Study Database; UCMR - Unregulated Contaminant
Monitoring Rule, PAF - population attributable fraction.
Proposed PFAS Rule Economic Analysis
6-128
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
7 Comparison of Costs to Benefits
This chapter provides a comparison of the incremental costs and benefits of the proposed rule, as
described in Chapter 5 and Chapter 6.86 The incremental cost is the difference between costs that
will be incurred if the proposed rule is enacted over and above current baseline conditions.
Incremental benefits reflect the avoided future adverse health outcomes attributable to PFAS
reductions and co-removal of additional contaminants due to actions undertaken to comply with
the proposed rule. This chapter also provides benefits and costs for the alternatives to the
proposed option that EPA considered. Results for the proposed option precede estimates for the
alternatives.
Table 7-1 provides the incremental quantified costs and benefits of the proposed option at both a
3 percent and a 7 percent discount rate in 2021 dollars. The first row shows total monetized
annualized costs including total PWS costs and primacy agency costs. The second row shows
total monetized annualized benefits including all endpoints that could be quantified and valued.
For both discount rates, the estimates are the expected values, and the 5th percentile and 95th
percentile estimates are derived from the uncertainty distribution. These percentile estimates
come from the distributions of annualized costs and annualized benefits generated by the 4,000
iterations of SafeWater MCBC, as described in Sections 5.1.2 and 6.1.2. Therefore, these
distributions reflect the joint effect of the multiple sources of variability and uncertainty for costs
identified in Section 5.1.2 and for benefits identified in Section 6.1.2 as well as the baseline
uncertainties discussed throughout Chapter 4 such as baseline PFAS occurrence.
The third row shows net benefits (benefits minus costs). At a 3 percent discount rate, the net
annual incremental benefits are $461 million. The uncertainty range for net benefits is negative
$45 million to $1.14 billion. At a 7 percent discount rate, the net annual incremental benefits are
negative $297 million. The uncertainty range for net benefits is negative $628 million to $141
million.
80 The cost-benefit analysis results for each option reflect the variability and uncertainties that could be quantified given the best
available scientific data. There are many factors that EPA could not quantify because of data limitations. For example, benefits
will be underestimated if the PFOA and PFOS reductions result in avoided adverse health outcomes that cannot be quantified and
valued. Chapters 5 and 6 identify these limitations and the potential effect on the cost or benefit estimates, respectively.
Proposed PFAS Rule Economic Analysis
7-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-1: Annualized Quantified National Costs and Benefits, Proposed Option (PFOA
and PFOS MCLs of 4.0 ppt and HI of 1.0; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Total Annualized Rule
Costs
$704.53
$771.77
$850.40
$1,106.01
$1,204.61
$1,321.01
Total Annualized Rule
Benefits
$659.91
$1,232.98
$1,991.51
$477.69
$908.11
$1,462.43
Total Net Benefitsb c d
-$44.62
$461.21
$1,141.11
-$628.31
-$296.50
$141.42
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1 for
costs and Section 6.1.2 and Table 6-1 for benefits. This range does not include the uncertainty described in Table 5-22 for
costs and Table 6-48 for benefits.
bSee Table 7-6 for a list of the nonquantifiable benefits and costs, and the potential direction of impact these benefits and
costs would have on the estimated monetized total annualized benefits and costs in this table.
Total quantified national cost values do not include the incremental treatment costs associated with the cooccurrence of
HFPO-DA, PFBS, and PFNA at systems required to treat for PFOA, PFOS, and PFHxS. The total quantified national cost
values do not include treatment costs for systems that would be required to treat based on HI exceedances apart from systems
required to treat because of PFHxS occurrence alone. See Appendix N, Section N.3 for additional detail on cooccurrence
incremental treatment costs and additional treatment costs at systems with HI exceedances.
dPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table
do not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns
about potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future,
EPA conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See
Appendix N, Section N.2 for additional detail.
Table 7-2 to Table 7-4 summarize the monetized total annual costs and benefits for Options la,
lb, and lc, respectively.
Proposed PFAS Rule Economic Analysis
7-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-2: Annualized Quantified National Costs and Benefits, Option la (PFOA and
PFOS MCLs of 4.0 ppt; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Total Annualized Rule
Costs
$688.09
$755.82
$833.48
$1,078.51
$1,177.31
$1,292.01
Total Annualized Rule
Benefits
$651.19
$1,216.08
$1,971.01
$471.53
$895.36
$1,456.23
Total Net Benefitsb c
-$36.90
$460.26
$1,137.53
-$606.97
-$281.95
$164.22
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1 for
costs and Section 6.1.2 and Table 6-1 for benefits. This range does not include the uncertainty described in Table 5-22 for
costs and Table 6-48 for benefits.
bSee Table 7-6 for a list of the nonquantifiable benefits and costs, and the potential direction of impact these benefits and
costs would have on the estimated monetized total annualized benefits and costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table
do not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns
about potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future,
EPA conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See
Appendix N, Section N.2 for additional detail.
Table 7-3: Annualized Quantified National Costs and Benefits, Option lb (PFOA and
PFOS MCLs of 5.0 ppt; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Total Annualized Rule
Costs
$558.71
$611.01
$674.32
$864.74
$942.28
$1,035.56
Total Annualized Rule
Benefits
$553.37
$1,046.91
$1,706.81
$398.21
$773.33
$1,292.96
Total Net Benefitsb c
-$5.34
$435.90
$1,032.49
-$466.53
-$168.95
$257.40
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1 for
costs and Section 6.1.2 and Table 6-1 for benefits. This range does not include the uncertainty described in Table 5-22 for
costs and Table 6-48 for benefits.
bSee Table 7-6 for a list of the nonquantifiable benefits and costs, and the potential direction of impact these benefits and
costs would have on the estimated monetized total annualized benefits and costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table
do not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns
about potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future,
EPA conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See
Appendix N, Section N.2 for additional detail.
Proposed PFAS Rule Economic Analysis
7-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-4: Annualized Quantified National Costs and Benefits, Option lc (PFOA and
PFOS MCLs of 10.0 ppt; Million $2021)
3% Discount Rate 7% Discount Rate
5th
Percentile3
Expected
Value
95th
Percentile3
5th
Percentile3
Expected
Value
95th
Percentile3
Total Annualized Rule
Costs
$269.36
$292.57
$320.76
$396.22
$430.87
$472.20
Total Annualized Rule
Benefits
$280.42
$584.80
$1,030.56
$208.71
$436.24
$784.59
Total Net Benefitsb c
$11.06
$292.23
$709.80
-$187.51
$5.36
$312.39
Notes: Detail may not add exactly to total due to independent rounding.
aThe 5th and 95th percentile range is based on modeled variability and uncertainty described in Section 5.1.2 and Table 5-1 for
costs and Section 6.1.2 and Table 6-1 for benefits. This range does not include the uncertainty described in Table 5-22 for
costs and Table 6-48 for benefits.
bSee Table 7-6 for a list of the nonquantifiable benefits and costs, and the potential direction of impact these benefits and
costs would have on the estimated monetized total annualized benefits and costs in this table.
cPFAS-contaminated wastes are not considered hazardous wastes at this time and therefore total costs reported in this table
do not include costs associated with hazardous waste disposal of spent filtration materials. To address stakeholder concerns
about potential costs for disposing PFAS-contaminated wastes as hazardous should they be regulated as such in the future,
EPA conducted a sensitivity analysis with an assumption of hazardous waste disposal for illustrative purposes only. See
Appendix N, Section N.2 for additional detail.
The reported dollar figures in this benefit-cost analysis reflect benefits and costs that could be
quantified for each regulatory alternative given the best available scientific data. EPA notes that
the quantified benefit-cost results presented above are not representative of all benefits and costs
anticipated under the proposed NPDWR. Due to limited data on occurrence, health, and
economic information, there are several adverse health effects associated with PFAS exposure
and costs associated with treatment that EPA could not estimate in a quantitative manner.
PFAS are associated with a wide range of adverse health effects including reproductive issues
such as decreased fertility or increased high blood pressure in pregnant women; developmental
effects or delays in children, including low birth weight, accelerated puberty, bone variations, or
behavioral changes; increased risk of some cancers, including prostate, kidney, and testicular
cancers; reduced ability of the body's immune system to fight infections, including reduced
vaccine response; interference with the body's natural hormones; and increased cholesterol
levels and/or risk of obesity. EPA is only able to quantify three PFOA- and PFOS-related health
endpoints in this analysis. All regulatory alternatives are expected to produce substantial benefits
that have not been quantified. Treatment responses implemented to remove PFOA and PFOS
under Options la-c are likely to remove some amount of additional PFAS contaminants where
they co-occur. Co-occurrence among PFAS compounds has been observed frequently as
discussed in the PFAS Occurrence Technical Support Document (U.S. EPA, 2023g). The
proposed option is expected to produce the greatest reduction in exposure to PFAS compounds
because it includes PFHxS, HFPO-DA, PFNA, and PFBS in the regulation. Inclusion of the HI
will trigger more systems into treatment (as shown in Section 4.4.4) and provides enhanced
public health protection by ensuring reductions of these additional compounds when present
above the HI of 1.0. For further discussion of the quantitative and qualitative benefits associated
with the proposed rule, see Section 6.2.
Proposed PFAS Rule Economic Analysis
7-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA also expects that the proposed option will result in additional nonquantifiable costs in
comparison to Options la-c. As noted above, the HI is expected to trigger more systems into
more frequent monitoring and treatment. Due to occurrence data limitations, EPA has quantified
the national treatment and monitoring costs associated with the HI for PFHxS only and has not
quantified the cost impacts associated with HI exceedances resulting from HFPO-DA, PFNA,
and PFBS. In cases where these compounds co-occur at locations where PFAS treatment is
implemented because of nationally modeled PFOA, PFOS, and PFHxS MCLs or HI
exceedances, treatment costs are likely to be marginally higher as treatment media estimated
bed-life is shortened. In instances where concentrations of HFPO-DA, PFNA, and PFBS are high
enough to cause or contribute to an HI exceedance when the concentrations of PFOA, PFOS, and
PFHxS would not have already otherwise triggered treatment, the modeled costs may be
underestimated. If these PFAS occur in isolation at levels that affect treatment decisions, or if
these PFAS occur in combination with PFHxS when PFHxS concentrations were otherwise
below the HI in isolation (i.e., less than 9.0 ppt) then the quantified costs underestimate the
impacts of the proposed rule. As such, EPA conducted a semi-quantitative analysis of the
anticipated incremental costs associated with regulating HFPO-DA, PFNA, and PFBS (discussed
in Section 5.3.1.4 and Appendix N).
Another potential source of nonquantified cost comes from the fact that EPA has proposed
designating PFOA and PFOS as CERCLA hazardous substances (U.S. EPA, 2022b).
Stakeholders have expressed concern to EPA that a hazardous substance designation for certain
PFAS may limit their disposal options for drinking water treatment residuals (e.g., spent media,
concentrated waste streams) and/or potentially increase costs. In its estimated national costs,
EPA has maintained the assumption that disposal does not have to occur in accordance with
hazardous waste standards thus national costs may be underestimated. EPA has conducted a
sensitivity analysis that assumes hazardous waste disposal at all systems treating for PFAS to
assess the potential increase in costs (see Appendix N). Table 7-5 summarizes benefits and costs
that are quantified and nonquantified under the proposed NPDWR.
Table 7-5: Summary of Quantified and Nonquantified Benefits and Costs
Methods (Report
Category Quantified Non-quantified Section where
Analysis is Detailed)
Costs
PWS treatment costs3 ^
PWS sampling costs S
PWS implementation and ^
administration costs
Primacy agency rule
implementation and administration S
costs
Hazardous waste disposal for
treatment media
POU not in compliance forecast
Benefits
PFOA and PFOS birth weight ^
effects
PFOA and PFOS cardiovascular ^
effects
Section 5.3.1
Section 5.3.2.2
Section 5.3.2.1
Section 5.3.2
S Section 5.6
S Section 5.6
Section 6.4
Section 6.5
Proposed PFAS Rule Economic Analysis
7-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-5: Summary of Quantified and Nonquantified Benefits and Costs
Category
Quantified Non-quantified
Methods (Report
Section where
Analysis is Detailed)
PFOA and PFOS renal cell
carcinoma
V
Section 6.6
Health effects associated with
V
V
Section 6.7
Section 6.2.2.2
disinfection byproducts
Other PFOA and PFOS health
effects
Health effects associated with HI
compounds HFPO-DA, PFNA,
V
Section 6.2
PFBS, and PFHxS
Health effects associated with
other PFAS
V
Section 6.2
Abbreviations: HFPO-DA - hexafluoropropylene oxide dimer acid; PFAS - per and polyfluoroalkyl
substances; PFBS - perfluorobutanesulfonic acid; PFHxS - perfluorohexane sulfonate; PFNA -
perfluorononanoic acid; PFOA - Perfluorooctanoic Acid; PFOS- Perfluorooctane Sulfonate; POU - point of
use; PWS- public water system
Notes:
aDue to occurrence data limitations, EPA quantified the national treatment and monitoring costs associated
with the HI for PFHxS only and has not quantified the national cost impacts associated with HI exceedances
resulting from PFNA, PFBS, and HFPO-DA
Table 7-6 provides a summary of the potential impact of nonquantifiable benefit-cost categories.
In each case, EPA notes the potential direction of the impact on costs and/or benefits. For
example, benefits are underestimated if the PFOA and PFOS reductions result in avoided adverse
health outcomes that cannot be quantified and valued. Sections 5.7 and 6.8 identify the key
methodological limitations and the potential effect on the cost or benefit estimates, respectively.
Proposed PFAS Rule Economic Analysis
7-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 7-6: Potential Impact of Nonquantifiable Benefits and Costs
Source
Proposed Option
Option la
Option lb
Option lc
Nonquantifiable PFOA
and PFOS health
endpoints
Limitations with
nationally representative
HFPO-DA, PFNA, and
PFBS occurrence data
(HI)
B: underestimate B: underestimate B: underestimate B: underestimate
C: underestimate
N/A
N/A
N/A
Nonquantifiable HFPO-
DA, PFNA, PFHxS, and
PFBS health endpoints
(HI)
B: underestimate
N/A
N/A
N/A
Limitations with
nationally representative
occurrence data for
additional PFAS
compounds
Removal of co-occurring
non-PFAS contaminants
POU not in compliance
forecast
Unknown future
hazardous waste
management
requirements for PFAS
(HI)
B&C:
underestimate
B&C:
underestimate
C: overestimate
C: underestimate
B&C:
underestimate
B&C:
underestimate
C: overestimate
B&C:
underestimate
B&C:
underestimate
C: overestimate
C: underestimate C: underestimate
B&C:
underestimate
B&C:
underestimate
C: overestimate
C: underestimate
Abbreviations: B - benefits; C - costs; POU - point of use; PFAS - per-and polyfluoroalkyl substances
When proposing an NPDWR, the Administrator shall publish a determination as to whether the
benefits of the maximum contaminant level justify, or do not justify, the costs based on the
analysis conducted under paragraph 1412(b)(3)(C). With this proposed rule, the Administrator
has determined that the quantified and nonquantifiable benefits of the proposed PFAS NPDWR
justify the costs.
As indicated in Table 7-1, the monetized costs and benefits result in expected annualized
incremental benefits of $1,233 million at a 3 percent discount rate. At a 7 percent discount rate,
the expected annualized incremental benefits are $908 million. The Agency views the 3 to 7
percent range of costs and benefits as characterizing the significant portion of the uncertainty in
the discount rate and views the quantified endpoint values with equal weight.
Table 7-1 through Table 7-6 summarize the results of this proposed rule analysis. As indicated in
Section 2.2.2 of this EA, EPA discounted the estimated monetized cost and benefit values using
both 3 and 7 percent discount rates. In federal regulatory analyses, EPA follows OMB Circular
A-4 (OMB, 2003) guidance which recommends using both 3 percent and 7 percent to account for
Proposed PFAS Rule Economic Analysis
7-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
the different streams of monetized benefits and costs affected by regulation. The 7 percent
discount rate is intended to represent the estimated rate of return on capital in the U.S. economy,
to reflect the opportunity cost of capital when the main effect of a regulation is to displace or
alter the use of capital in the private sector. Regulatory effects, however, can fall on both capital
and private consumption.87 In 2003, Circular A-4 estimated the rate appropriate for discounting
consumption effects at 3 percent. The estimated monetized costs and benefits of this rulemaking
result in expected annual net benefits (total monetized annual benefits minus total monetized
annual costs) of $461.21 million at a 3 percent discount rate and $-296.50 at a 7 percent discount
rate. There are a variety of considerations with respect to the capital displacement in this
particular proposal. For example, a meaningful number of PWSs may not be managed as profit-
maximizing private sector investments, which could impact the degree to which the rate of return
on the use of capital in the private sector applies to PWS costs. Federal funding is expected to
defray many such PWS costs;88 where that occurs, such costs are transferred to the government.
Additionally, to the extent that the benefits extend over a long time period into the future,
including to future generations, Circular A-4 advises agencies to consider conducting sensitivity
analyses using lower discount rates. Regardless, the impacts in this rulemaking are such that
costs are expected to occur in the nearer term, and in particular that larger one-time capital
investments are expected to occur in the near term; and public health benefits are expected to
occur over the much longer term. Discounting across an appropriate range of rates can help
explore how sensitive net benefits are to assumptions about whether effects fall more to capital
or more to consumption.
EPA has followed Circular A-4's default recommendations to use 3 and 7 percent rates to
represent the range of potential impacts accounting for diversity in stakeholders' time
preferences. The Agency views the 3 to 7 percent range of costs and benefits as characterizing a
significant portion of the uncertainty in the discount rate and views the quantified endpoint
values as demonstrating a range of monetized costs and benefits which encompass a significant
portion of the uncertainty associated with discount rates. Material unquantified benefits expected
as a result of this proposed rulemaking are discussed in greater detail later in this section.
The quantified analysis is limited in its characterization of uncertainty. In Table 7-1, EPA
provides 5th and 95th percentile values associated with the 3 and 7 percent discounted expected
values for net benefits. These values represent the quantified, or modeled, potential range in the
expected net benefit values associated with the variability in system characteristics and the
uncertainty resulting from the following variables: the baseline PFAS occurrence; the affected
population size; the compliance technology unit cost curves, which are selected as a function of
baseline PFAS concentrations and population size, the distribution of feasible treatment
technologies, and the three alternative levels of treatment capital costs; the concentration of total
organic carbon in a system's source water, which impacts GAC O&M costs; the demographic
composition of the systems population; the magnitude of PFAS concentration reductions; the
health effect-serum PFOA and PFOS slope factors that quantify the relationship between
changes in PFAS serum level and health outcomes for birth weight, CVD, and renal cell
87 Private consumption is the consumption of goods and services by households for the direct satisfaction of individual needs
(rather than for investment).
88 As noted above in this preamble, "Infrastructure Investment and Jobs Act, also referred to as the Bipartisan Infrastructure Law
(BIL), invests over $11.7 billion in the Drinking Water State Revolving Fund (SRF); $4 billion to the Drinking Water SRF for
Emerging Contaminants; and $5 billion to Small, Underserved, and Disadvantaged Communities Grants."
Proposed PFAS Rule Economic Analysis
7-8
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
carcinoma; and the cap placed on the cumulative renal cell carcinoma risk reductions due to
reductions in serum PFOA. These modeled sources of uncertainty are discussed in more detail in
Sections 5.1.2 and 6.1.2. What the quantified 5th and 95th percentile values do not include are a
number of factors which impact both costs and benefits but for which the Agency did not have
sufficient data to include in the quantification of uncertainty. The factors influencing the
proposed rule cost estimates that are not quantified in the uncertainty analysis are detailed in
Table 5-22. These uncertainty sources include: the specific design and operating assumptions
used in developing treatment unit cost; the use of national average costs that may differ from the
geographic distribution of affected systems; the possible future deviation from the compliance
technology forecast; and the degree to which actual TOC source water values differ from EPA's
estimated distribution. EPA has no information to indicate a directional influence of the
estimated costs with regard to these uncertainty sources. To the degree that uncertainty exists
across the remaining factors it would most likely influence the estimated 5th and 95th percentile
range and not significantly impact the expected value estimate of costs.
Table 6-48 discusses the sources of uncertainty affecting the estimated benefits not captured in
the estimated 5th and 95th reported values. The modeled values do not capture the uncertainty in:
the exposure that results from daily population changes at NTNCWSs or routine population
shifting between PWSs, for example spending working hours at a NTNCWS or CWS and home
hours at a different CWS; the exposure-response functions used in benefits analyses assume that
the effects of serum PFOA/PFOS on the health outcomes considered are independent, additive,
and that there are no threshold serum concentrations below which effects do not occur; the
distribution of population by size and demographics across entry points within modeled systems
and future population size and demographic changes; and the Value of Statistical Life reference
value or income elasticity used to update the VSL. Given information available to the Agency,
four of the listed uncertainty sources would not affect the benefits expected value but the
dispersion around that estimate. They are the unmodeled movements of populations between
PWS which potentially differing PFAS concentrations; the independence and additivity
assumptions with regard to the effects of serum PFOA/PFOS on the health outcomes; the
uncertainty in the population and demographic distributions among entry points within
individual systems; and the VSL value and the income elasticity measures. Two of the areas of
uncertainty not captured in the analysis would tend to indicate that the quantified benefits
numbers are overestimates. First, the data available to EPA with regard to population size at
NTNCWs, while likely capturing peaks in populations utilizing the systems, does not account for
the variation in use and population and would tend to overestimate the exposed population. The
second uncertainty, which definitionally would indicate overestimates in the quantified benefits
values, is the assumption that there are no threshold serum concentrations below which health
effects do not occur. One factor not accounted for in the quantified analysis associated with the
underestimation of benefits is the impact of general population growth over the extended period
of analysis.
In addition to the quantified cost and benefit expected values, the modeled uncertainty associated
within the 5th and 95th percentile values, and the un-modeled uncertainty associated with a
number of factors listed above, there are also significant nonquantifiable costs and benefits,
which are important to the overall weighing of costs and benefits. Table 7-6 provides a summary
of these nonquantifiable cost and benefit categories along with an indication of the directional
Proposed PFAS Rule Economic Analysis
7-9
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
impact each category would have on total costs and benefit. Table 5-22 and Table 6-48 also
provide additional information on a number of these nonquantifiable categories.
On the nonquantifiable costs side of the equation, EPA had insufficient nationally representative
data to precisely characterize occurrence of HFPO-DA, PFNA, and PFBS at the national level
and therefore could not include complete treatment costs associated with: the co-occurrence of
these PFAS at systems already required to treat as a result of estimated PFOA, PFOS, or PFHxS
levels, which would shorten the filtration media life and therefore increase operation costs; and
the occurrence of HFPO-DA, PFNA, and/or PFBS at levels high enough to cause systems to
exceed the HI and have to install PFAS treatment. The quantified national costs are marginally
underestimated as a result of this lack of sufficient nationally representative occurrence data for
purposes of model integration. In an effort to better understand the costs associated with
treatment of potentially co-occurring HFPO-DA, PFNA, and PFBS at systems already required
to treat and the potential costs resulting from an HI exceedance associated with the same
chemicals, EPA estimated the potential unit treatment costs for model systems under both
scenarios for differing assumed HI PFAS concentrations. The analysis is discussed in Section
5.3.1.4 and Appendix N. Two additional nonquantifiable cost impacts stemming from
insufficient co-occurrence data could also potentially shorten filtration media life and increase
operation costs. The co-occurrence of other PFAS and other non-PFAS contaminants not
regulated in the proposed rule could both increase costs to the extent that they reduce media life.
EPA did not include POU treatment in the compliance technology forecast because current POU
units are not certified to remove PFAS to the standards required in the proposed rule. Once
certified, this technology may be a low cost treatment alternative for some subset of small
systems. Not including POU treatment in this analysis has resulted in a likely overestimate of
cost values. Appendix N contains a sensitivity analysis that estimates there may be a national
annual cost of $30 to $61 million, discounted at 3 and 7 percent, respectively, which would
accrue to systems if the waste filtration media from GAC and IX were handled as hazardous
waste. This sensitivity analysis includes only disposal costs and does not consider other potential
environmental costs associated with the disposal of the waste filtration.
There are significant nonquantifiable sources of benefits that were not captured in the quantified
benefits estimated for the proposed rule. While EPA was able to monetize some of the PFOA
and PFOS benefits related to cardiovascular disease, infant birth weight, and renal cell carcinoma
effects, the Agency was unable to quantify additional negative health impacts. EPA did not
quantify PFOA and PFOS benefits related to health endpoints including developmental,
cardiovascular, hepatic, immune, endocrine, metabolic, reproductive, musculoskeletal, and other
types of carcinogenic effects. Section 6.2.2 provides additional information on the
nonquantifiable impacts of PFOA and PFOS. Further, the Agency did not quantify any health
endpoint benefits associated with the potential reductions in HI PFAS, which include PFHxS,
HFPO-DA, PFNA, and PFBS, or other co-occurring non-regulated PFAS which would be
removed by the installation of required filtration technology at those systems with PFOA, PFOS,
or HI exceedances. The nonquantifiable benefits impact categories associated with PFHxS,
HFPO-DA, PFNA, and PFBS include developmental, cardiovascular, immune, hepatic,
endocrine, metabolic, reproductive, musculoskeletal, and carcinogenic effects. In addition, EPA
did not quantify the potential developmental, cardiovascular, immune, hepatic, endocrine,
metabolic, reproductive, musculoskeletal, and carcinogenic impacts related to the removal of
other co-occurring non-regulated PFAS. See Section 6.2.4 for additional information on the
Proposed PFAS Rule Economic Analysis
7-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
nonquantifiable impacts of PFHxS, HFPO-DA, PFNA, and PFBS, and other non-regulated co-
occurring PFAS.
The treatment technologies installed to remove PFAS can also remove numerous other non-
PFAS drinking water contaminants which have negative health impacts including additional
regulated and unregulated DBPs (the quantified benefits assessment does estimate benefits
associated with THM4), heavy metals, organic contaminants, and pesticides among others. The
removal of these co-occurring non-PFAS contaminants could have significant positive health
benefits. In total these nonquantifiable benefits are anticipated to be significant and are discussed
qualitatively in Section 6.2.
To fully weigh the costs and benefits of the action the Agency considered the totality of the
monetized values, the potential impacts of the unquantified uncertainties described above, and
the nonquantifiable costs and benefits. The Administrator has determined that the benefits of this
proposed regulation justify the costs.
Proposed PFAS Rule Economic Analysis
7-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8 Environmental Justice Analysis
8.1 Introduction
EPA defines environmental justice (EJ) as "the fair treatment and meaningful involvement of all
people regardless of race, color, national origin, or income with respect to the development,
implementation, and enforcement of environmental laws, regulations, and policies" (U.S. EPA,
2016h). The concept of fair treatment includes not just the distribution of burdens across
populations but also the distribution of risk reduction from EPA actions. EPA reviews potential
EJ concerns regarding minority populations, low-income populations, and/or indigenous peoples
(U.S. EPA, 2016h).
The framework used to evaluate the anticipated EJ impacts of the proposed rule for per- and
polyfluoroalkyl substances (PFAS) comes from the Technical Guidance for Assessing
Environmental Justice in Regulatory Analysis (U.S. EPA, 2016h), which provides the following
guiding questions:
• Are there potential EJ concerns associated with environmental stressors affected by the
regulatory action for population groups of concern in the baseline?
• Are there potential EJ concerns associated with environmental stressors affected by the
regulatory action for population groups of concern for the regulatory options under
consideration?
• For the regulatory options under consideration, are potential EJ concerns created or
mitigated compared to the baseline?
Contextualizing these questions for the proposed PFAS rule, EPA evaluated the following
questions:
• Are population groups of concern (i.e., people of color and low-income populations)
disproportionately exposed to PFAS compounds in drinking water delivered by PWSs?
• Are population groups of concern disproportionately affected by the proposed option and
regulatory alternatives under consideration for the proposed PFAS NPDWR?
• If any disproportionate impacts are identified, do they create or mitigate baseline EJ
concerns?
As part of the proposal process for the PFAS NPDWR, EPA conducted the EJ analyses in this
chapter to assess the demographic distribution of baseline PFAS drinking water exposure and
impacts that are anticipated to result from the proposed rule. EPA conducted two separate
analyses to address the research questions presented above. To inform the first question, EPA
conducted an analysis using EJScreen, the Agency's Environmental Justice Screening and
Mapping Tool (U.S. EPA, 2019a). To inform the second and third questions above, EPA
conducted an EJ analysis of EPA's proposed regulatory option and regulatory alternatives using
SafeWater MCBC.
Proposed PFAS Rule Economic Analysis
8-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Section 8.2 provides an overview of EPA's EJ literature review. Sections 8.3 and 8.4 describe
the EJ analyses EPA conducted. Section 8.5 presents the conclusions from EPA's EJ analyses.
8.2 Literature Review
EPA conducted a literature review to develop a broad understanding of current research at the
intersection of drinking water quality, PFAS exposure, and communities with related EJ
concerns. The literature covered a range of specific topics including the likelihood of exposure
based on proximity to sites of contamination, sociodemographic characteristics of communities
exposed to PFAS in region-specific studies and understanding the sociodemographic distribution
of health outcomes associated with exposure to PFAS. EPA's literature review also examined the
relationship between PFAS exposure via drinking water in vulnerable communities and a range
of health outcomes.
8.2.1 Methods
EPA conducted its literature review to evaluate and synthesize findings from studies that
explored associations between PFAS exposure via drinking water in vulnerable communities and
associated health outcomes, including those health endpoints EPA quantified as part of its
benefits analysis: changes in infant birth weight, CVD, and kidney cancer.
EPA applied a variety of search terms for the literature review, including: CVD; disparities;
disproportionate exposure; disproportionate impact; drinking water quality/contamination;
environmental justice; equity; forever chemicals; inequity; infant birth weight; kidney cancer;
low-income; minority; over-burdened; people of color; PFAS; PFAS interactions; PFC(s);
PFOA; PFOS; race differences in health effects after PFAS exposure; race disparities in health
effects, immune effects, and PFAS exposure; race ethnicity and health effects of PFAS exposure
and interactions; sociodemographic differences in health effects after PFAS exposure; social
justice; and tribal.
From the literature review, EPA found that there are a limited number of studies that focus on the
association between disproportionate exposure to PFAS via drinking water and health outcomes
for vulnerable communities on a national level. The Agency excluded studies that examined
exposure routes apart from drinking water and/or did not evaluate race/ethnicity within their
participant demographics. Of the studies that EPA identified as part of its literature review, all
but two studies were published in peer reviewed journals (with the remaining two studies
appearing in gray literature).
8.2.2 Findings
To contextualize its analysis of EJ impacts related to PFAS in drinking water, EPA reviewed
studies that evaluate overall EJ concerns related to environmental contamination. In 1987, EPA
reported in a nationwide study that roughly twice as many people of color resided in proximity to
a commercial hazardous waste facility compared to communities without a facility (U.S. EPA,
1994). Later research indicated that communities of low socioeconomic status are more likely to
reside in proximity to environmental hazardous facilities, thereby potentially facing a
disproportionate impact of exposure to toxic chemicals than communities of higher
socioeconomic status (Brown, 1995; Brulle et al., 2006). A 2010 study showed 63 percent of
Proposed PFAS Rule Economic Analysis
8-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
large polluters in a North Carolina county were operating in census tracts with per capita income
below $21,000, as identified in EPA's Toxics Release Inventory (TRI) (Banzhaf et al., 2019).
When specifically examining studies related to PFAS in drinking water, available literature
showed associations between PFAS contamination in drinking water and proximity to sites
including those critical for transportation infrastructure, industry, and national defense (Black et
al., 2021; X. C. Hu et al., 2016; Johnston et al., 2020 ; Sunderland et al., 2019). Researchers
noted that identifiable sources of PFAS are often prevalent at aforementioned locations and are
more frequently located in vulnerable communities (Black et al., 2021; X. C. Hu et al., 2016;
Stoiber et al., 2020).
PFAS have characteristics—namely high aqueous solubility and persistence within the
environment that allow them to travel readily between ecological zones (ATSDR, 2021; X. C.
Hu et al., 2016; Kotlarz et al., 2020). As such, PFAS contamination can negatively impact
drinking water sources downstream from an original contamination site, putting residents in
communities surrounding known sources of PFAS at a disproportionate risk of exposure. A 2019
study in Michigan by Desikan et al. (2019) evaluated the proportion of low-income households
and households with people of color in communities within five miles of PF AS-contaminated
sites compared to census projections for those areas. It found that 38,962 more low-income
households and 294,591 more households with people of color reside within five miles of a site
contaminated with PFAS than expected, based on U.S. Census data.
In California, Lee et al. (2021) demonstrated that vulnerable communities are more likely to be
served by PWSs with higher levels of PFAS. PFAS data were integrated with results from
CalEnviroScreen 3.0, a statewide EJ screening tool (OEHHA, 2016). Of the 7,896 PWSs in the
state, about 3 percent (n=248) had been monitored for PFAS, serving 42 percent of California's
total population. Results from the study showed that PFAS was detected in 160 of 248 PWSs, or
roughly 65 percent of systems monitored. Lee, Kar, et al. (2021) overlaid the upper 25 percent of
disadvantaged communities as identified by CalEnviroScreen 3.0 with water systems
experiencing the highest levels of PFAS contamination. Among the communities in the top
quartile for people of color and low-income demographic groups, 69 percent had PFAS detected
in their water system. Further, PWSs in 20 percent of vulnerable communities with PFAS
contamination fell within the highest quartile of PFAS concentration levels in the state of
California, suggesting that PFAS occurrence is disproportionately higher in drinking water
serving already vulnerable communities. Only 2 of the 10 water systems with the highest PFAS
concentrations fell below the state average for all relevant demographic indicators included in
the study (people of color, education level, unemployment, poverty, and housing burden).
At least two studies identified the use of aqueous film forming foam (AFFF) as a predictor of
PFAS concentrations in U.S. drinking water (Johnston et al., 2020; Sunderland et al., 2019).
Using nationally representative PFAS occurrence data from UCMR 3, a study from X. C. Hu et
al. (2016) found that the presence of a military fire training area using AFFF within a
watershed's eight-digit hydraulic unit code (HUC) increased the frequency of exposure to at least
one PFAS analyte in drinking water from 10.4 percent to 28.2 percent. For each additional
military site within a HUC, drinking water samples with detectable levels of PFAS found a 20
percent increase in PFHxS, a 10 percent increase in both PFHpA and PFOA, and a 35 percent
increase in PFOS.
Proposed PFAS Rule Economic Analysis
8-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
To remain consistent with the health endpoints associated with PFAS exposure that are
monetized as part of the proposed PFAS NPDWR's benefits analysis, the health outcomes of
focus in this literature review included CVD, kidney cancer, and impacts on infant birth weight.
For more information on EPA's quantified benefits analysis, see Chapter 6.
Literature showed that vulnerable communities experience relatively higher adverse health
outcomes compared to communities with fewer people of color (Driscoll et al., 2021; Fryar et al.,
2017; Pinheiro et al., 2021). Literature also showed that risk of CVD, kidney cancer, and changes
in infant birth weight are associated with PFAS exposure (Almond et al., 2005; Barry et al.,
2013; Goff et al., 2014; Ma et al., 2010; Raleigh et al., 2014; Steenland et al., 2012; Vieira et al.,
2013; U.S. EPA, 2016d; U.S. EPA, 2016h; U.S. EPA, 2021a; U.S. EPA, 2023d; U.S. EPA,
2023e), discussed in more detail in Chapter 6.
The Centers for Disease Control and Prevention (CDC) identified hypertension (HTN) as a
substantial risk factor for CVD (Fryar et al., 2017). Using the 140/90 mmHg threshold for HTN
diagnosis, the CDC reported that African American adults reported a higher burden of HTN
(40.3%) compared to White (27.8%), Asian (25.0%), or Hispanic (27.8%) adults (Fryar et al.,
2017). Additionally, a comprehensive narrative literature review by Graham (2015) found
disproportionate rates of CVD among minority subpopulations in the U.S., particularly the
African American population. African American subpopulations were found to have higher
incidence of myocardial infarction, heart failure, stroke, among other cardiovascular events and
experience the highest overall death rate from CVD among various minority population groups.
With regards to cancer, a study by Uche et al. (2021) showed statistically significantly greater
cumulative cancer risk was identified in communities in which small and large CWSs serve
higher proportions of Hispanic/Latino and Black/African American residents in Texas and
California89. In Texas, greater cumulative cancer risk was statistically significantly greater for
small and medium CWSs serving relatively higher proportions of Hispanic/Latino community
members. Additionally, small CWSs serving relatively higher proportions of Black/African
American residents had statistically significantly greater cumulative cancer risk. In California,
cumulative cancer risk was statistically significantly greater for very large CWSs serving
relatively higher proportions of Black/African American community members, followed by
small CWSs serving relatively higher proportions of Hispanic/Latino residents.
Pinheiro et al. (2021) studied kidney cancer rates in White, Black, Asian/Pacific Islander (API),
American Indian, all non-Hispanic, and Hispanic populations of any race by using reported
cancer deaths in California and Florida (2008-2018) and New York (2008-2017). This study's
methodology directly compared results for specific race/ethnicity groups to White populations.
Results indicated that American Indian individuals experience the highest mortality (54%), and
mortality is 53 percent higher for men and women when compared to mortality among White
participants (Pinheiro et al., 2021). Conversely, API populations showed significantly lower
mortality than White populations, with 45 percent lower mortality among males and 43 percent
lower mortality among females. Kidney cancer mortality among Black populations and all-
combined Hispanic populations (i.e., Cuban, Puerto Rican, and Mexican) was also significantly
lower than among White populations, but by smaller margins: mortality was 12 percent and 16
89 A CWS was defined as small if it served 501-3,300 people, medium if it served 3,301-10,000 people, large if it served 10,001-
100,000 people, and very large if it served more than 100,000 people.
Proposed PFAS Rule Economic Analysis
8-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
percent lower for Black males and females and 11 percent and 8 percent lower for Hispanic
males and females, respectively.
Additionally, the CDC's National Vital Statistics Reports used the 2020 birth file from the
National Vital Statistic System to display distributions in prepregnancy body mass index (BMI),
including three classes of obesity, by maternal race and Hispanic origin for women who gave
birth in 2020 (Driscoll et al., 2021). Infants born to non-Hispanic Black women had the highest
rate of low birth weight (14.19%), followed by infants of Hispanic women (7.40%). Infants of
non-Hispanic White women had the lowest rate of low birth weight (6.84%) (Driscoll et al.,
2021).
Furthermore, EPA reviewed studies that examine blood serum levels of PFAS across various
demographic groups. Studies analyzing biomarker data indicate some demographic disparities
that exist in blood serum levels across certain PFAS analytes (Boronow et al., 2019; Calafat et
al., 2007; Eick et al., 2021; C. Y. Lin et al., 2020; Nelson et al., 2012; V. K. Nguyen et al., 2020;
Park et al., 2019). Specifically, blood serum levels of PFNA and PFOS were found to be elevated
in Black adults (Boronow et al., 2019; Calafat et al., 2007; Eick et al., 2021; C. Y. Lin et al.,
2020; Nelson et al., 2012; Park et al., 2019). PFNA was also found to be elevated in Asian
American mothers, when compared to all other races (Eick et al., 2021). Additionally, PFDA
was found to be elevated in Asian American women, when compared to non-Hispanic White
populations (V. K. Nguyen et al., 2020). Finally, Me-FOSAA was found to be elevated in Black
women at some but not all study sites analyzed (Park et al., 2019).
However, many studies indicate lower average blood serum PFAS levels among people of color.
Three studies in particular demonstrated that non-Hispanic White populations had the highest
concentrations of PFAS across all analytes (Barton et al., 2020; Kato et al., 2014; Kingsley et al.,
2018). It should be noted, however, that the study design for Barton et al. (2020), Kato et al.
(2014), and Kingsley et al. (2018) each had majority non-Hispanic White participant
demographics of 75 percent, 63 percent, and 61 percent of study participants, respectively. The
literature also indicates that higher socioeconomic status (e.g., income) is associated with higher
PFAS blood serum levels (Buekers et al., 2018).
8.2.3 Discussion and Limitations
EPA's purpose in conducting its literature review was to examine the relationship between PFAS
exposure via drinking water in vulnerable communities and health outcomes related to CVD,
changes in infant birth weight, and kidney cancer. Presented studies indicate that higher
percentages of low-income and minority communities reside near a range of PF AS-contaminated
sites. Such contamination is also shown to occur at higher levels in low-income and minority
communities. Further, EPA's literature review analysis indicates that PFAS contamination
occurs more often and/or at higher levels in vulnerable communities.
It should be noted there are substantial gaps in current literature on PFAS exposure and health
outcomes in vulnerable communities. One substantial gap in the available literature is a dearth of
studies that examine differential impacts of health outcomes associated with PFAS exposure, as
reported by race or ethnicity. Potential gaps in understanding also relate to determining whether
the rate of developed risk for one or more of the aforementioned health endpoints is related to
exposure to PFAS contamination in drinking water rather than other exposure pathways.
Proposed PFAS Rule Economic Analysis
8-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
The blood serum PFAS studies evaluated as part of this literature review have their limitations in
extrapolating to the potential disproportionate impacts of PFAS drinking water exposure given
their focus on overall PFAS exposure across many exposure routes rather than drinking water-
specific exposures. Wilder et al. note that national average PFAS blood serum levels are
influenced by a variety of major exposure pathways, including diet and consumer products in
addition to exposure via drinking water (Wilder et al., 2017). As such, this limits conclusions
that can be drawn about the demographic breakdown of PFAS blood serum levels due to
drinking water exposure alone. Additional information on exposure via drinking water alone is
necessary to better understand the impacts of PFAS drinking water contamination on PFAS
blood serum levels within vulnerable communities.
Another limitation of these blood serum-based studies is their inequitable representation of study
participants by race. The participant demographic makeup of three published studies that
examined PFAS blood serum levels was highly biased toward the non-Hispanic White
population, resulting in an incomplete understanding of people of color's exposure to PFAS.
Statisticians can adjust the results if certain participant demographic groups are
disproportionately represented. However, these adjustments are based on assumptions about the
underlying demographic makeup of the study population.
8.3 EJ PFAS Exposure Analysis
This section describes the data sources and approach EPA used to characterize the demographic
distribution of PFAS exposure in drinking water. This analysis is designed to answer the
question posed in the beginning of the chapter: Are population groups of concern (i.e., people of
color and low-income populations) disproportionately exposed to PFAS compounds in drinking
water delivered by PWSs? This analysis estimates exposure rates above various PFAS
concentrations for four PFAS analytes, where occurrence of these is used as a proxy for co-
occurrence of many other PFAS compounds. In some cases, the thresholds that EPA uses in this
analysis overlap with regulatory alternatives considered by EPA in the proposed regulatory
action. This analysis does not evaluate the anticipated costs and benefits of the proposed option
and regulatory alternatives. EPA's analysis of the anticipated demographic distribution of costs
and benefits of the proposed option and regulatory alternatives can be found in Section 8.4.
EPA estimated the sociodemographic characteristics of populations that EPA anticipates are
exposed to levels higher than various threshold concentrations of four PFAS analytes (PFOA,
PFOS, PFHxS, and PFHpA). For this analysis, EPA had sufficient information on PFAS
occurrence and PWS service area boundaries in the sample population, which was a subset of
PWSs.90 PWSs were first categorized by available data (Section 8.3.1), using availability of
UCMR 3 sampling data, state sampling data, and availability of service area boundary
information (Table 8-1).
EPA used PWS service area data in conjunction with the EJSCREENbatch R package to obtain
sociodemographic characteristics of the populations served by PWSs (U.S. EPA, 2022a). The
EJSCREENbatch R package allows analysts to conduct EJ screening analyses for multiple
geographies using environmental and sociodemographic data from EJScreen and the American
Community Survey. EPA estimated the rate of exposure to PFAS across demographic groups
90 PWS service area boundaries are defined as the spatial extent of the geographic area served by a PWS.
Proposed PFAS Rule Economic Analysis
8-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
using PFAS occurrence data and the sociodemographic characteristics of populations served with
designated service area boundaries. EPA conducted this analysis using several thresholds:
Method 537.1 detection limits (also referred to as baseline occurrence level for this analysis),
UCMR 5 minimum reporting levels (MRLs), and 10.0 ppt. This analysis serves as an estimate of
possible exposure to PFAS levels over these thresholds, as EPA cannot confirm that these
populations consumed the water at the time of elevated PFAS occurrence at each PWS.
8.3.1 Data Sources and Approach
8.3.1.1 Cotegorizotion of Public Water Systems
EPA designated distinct categories for PWSs based on data availability for PFAS occurrence and
estimated PWS service area boundaries. The Agency used two types of PFAS occurrence data
sources in this analysis: (1) simulated PFAS occurrence data for PWSs with sampled PFAS
occurrence data under UCMR 3; and (2) state-collected PFAS occurrence data for PWSs not
sampled under UCMR 3 (U.S. EPA, 2017). PWS service area boundary data are distinguished by
three types: (1) those with predelineated PWS service area boundaries, (2) those where zip codes
served by PWSs were used as a proxy to approximate and delineate PWS service area
boundaries, and (3) those with no available PWS service area boundary information. Table 8-1
describes the characteristics of each of the six distinct PWS categories examined in this analysis.
For the EJ exposure analysis, EPA focused on reporting results for PWSs in categories 1 and 2,
which were sampled for PFAS under UCMR 3. The PWSs in categories 4 and 5 include systems
with state PFAS occurrence data, and EPA has summarized the results for these categories in
Appendix M. EPA used data from EJScreen (U.S. EPA, 2022a) and the American Community
Survey along with PWS service area boundary data to characterize the sociodemographic
characteristics of PWSs.
PWSs in categories 1 and 2 account for 239.6 million people served (n=4,723 PWSs), and PWSs
in categories 4 and 5 account for approximately 1.2 million people served (n=459 PWSs). PWSs
in categories 3 and 6 were not included in the exposure analysis, as PWS service area boundaries
or zip codes served by the PWS were unavailable.
Proposed PFAS Rule Economic Analysis
8-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-1: Categorizing of PWSs Based on Data Availability for PFAS Occurrence and
PWS Service Area Boundaries
PWS Included in UCMR 3 PWS State PFAS Occurrence Data
Available and Not Included in
UCMR 3
PWS Service Areas Available Category 1 Category 4
PWS Service Area Boundary Category 2 Category 5
Estimates from Zip Codes
No PWS Service Area Information Category 3 Category 6
Available
Abbreviations: PWS - public water system; UCMR - Unregulated Contaminant Monitoring Rule.
8.3.1.2 Data Sources
8.3.1.2.1 PFAS Occurrence
The two data source categories used to derive PFAS occurrence estimates for this analysis are
described in more detail below. All PFAS occurrence data are presented in parts per trillion
(PPt).
Generally, if a system was sampled for PFAS under UCMR 3, EPA used simulated occurrence
data that were based on system-specific results. For PWSs in categories 1 and 2 (n=4,723
PWSs), EPA simulated PFAS occurrence data using a hierarchical Bayesian model that was
optimized with PFAS occurrence data from UCMR 3 and, where available, state data (see
Cadwallader et al., 2022, and Section 4.4 for further description). EPA calculated the system-
level geometric mean occurrence value for each PWS from the simulated water sample
concentrations. All simulated values (i.e., simulated samples for PWSs in categories 1 and 2)
were above zero because the occurrence model assumes a log-normal distribution for water
concentration. The system-level geometric mean occurrence values for the category 1 and 2
PWSs ranged from 0.01 to 254.65 ppt.
For other systems, EPA used state sampling data. EPA used state monitoring data from 12
states91, which generally conducted nontargeted monitoring (i.e., random sampling) of finished
drinking water for one or more of the four PFAS in this analysis. PWSs that had state sampling
data but were not sampled under UCMR 3 fell into categories 4 and 5 (n=459). EPA calculated
the system-level geometric means of measured PFAS water sample concentrations to
characterize PFAS occurrence for each PWS. For this dataset, the Agency did not pursue
Bayesian estimation of non-detection concentrations due to a limited sample size and non-
standardized sampling regime. Instead, for these data, EPA set non-detections to a small
91 States include: Alabama, Colorado, Illinois, Kentucky, Massachusetts, Michigan, New Hampshire, New Jersey, North Dakota,
Ohio, South Carolina, and Vermont.
Proposed PFAS Rule Economic Analysis
8-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
constant, 10 percent of the lowest analyte sample value (i.e., 0.02 ppt for each analyte), before
calculating the system-level geometric mean.92
Among the 12 state occurrence datasets used in this analysis to characterize PFAS occurrence for
category 4 and 5 PWS service areas, EPA noted that different states utilized various reporting,
quantification, and/or detection limits when analyzing and presenting data, and for some states,
no clearly defined limits were publicly provided as part of the dataset. Further, the limits often
varied within the data for each state depending on the specific PFAS analyte. In some cases,
states reported detection, quantification, or reporting limits and/or presented data at
concentrations below EPA's proposed rule detection limits and/or practical quantitation limits
provided in the federal register notice for this proposed regulatory action. In addition to variable
reporting limits and PFAS analytes evaluated, sample collection routines across state datasets
also lacked uniformity. For more information on the collection and analysis of occurrence data,
see U.S. EPA (2023g).
For both simulated occurrence data and state-sampled occurrence data, system-level geometric
means were calculated to represent a typical concentration of a single sample for each PFAS
analyte in a system. The concentrations of samples are log-normally distributed for all four
PFAS analytes (PFOA, PFOS, PFHxS, PFHpA), meaning that while most samples have low
concentrations, some may have much higher concentrations.
8.3.1.2.2 PWS Service Area Boundaries
For CWSs and NTNCWSs that had PFAS occurrence data sampled under UCMR 3 or PFAS
occurrence data collected by states, EPA acquired or estimated service area boundaries. Since
transient noncommunity water systems (TNCWSs) have changing populations throughout the
year, they were not included in this analysis. Data were categorized by the availability of PWS
service areas, those with predelineated PWS service areas (categories 1 and 4), and those where
zip codes served by PWSs were used to approximate PWS service area boundaries (categories 2
and 5). When available, predelineated PWS service areas were prioritized over zip code-
approximated PWS service area boundaries. EPA used the federal version of the Safe Drinking
Water Information System (SDWIS/Fed) to inform the type of water system (e.g., CWS,
NTNCWS), population served, identify Native American-owned PWSs, and determine activity
status for PWSs included in the analysis. Only active systems, as identified in SDWIS/Fed fourth
quarter 2021, were included.
For predelineated PWS service area boundaries, EPA aggregated spatial data from a variety of
sources spanning multiple file formats into one ESRI file geodatabase.93 Data sources are
provided in Table 8-2.
92 EPA evaluated the difference between using 10 percent (0.02 ppt) and 50 percent (1 ppt) of the minimum reported sample
concentration for all analytes. The difference in population estimates from this change was less than 0.5 percent for all analytes.
10 percent of the minimum reported value was used in the analysis (0.02 ppt).
93 File formats included: ESRI ArcGIS Online (AGOL) layers, shapefiles, and Geo.TSON.
Proposed PFAS Rule Economic Analysis
8-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
Table 8-2: Data Sources for Predelineated PWS Service Areas
MARCH 2023
Accessed Through State Sources or EPA Correspondence
State
Source Name
Link
Date
CO
State of Colorado - Water District
Boundaries
httos ://data. Colorado. eov/W ater/W ater-District-
Boundaries/82ke-q8t2
Accessed
1/26/2022
CA
State of California - Division of
Drinking Water, California Water
Resources Control Board
https://gispublic.waterboards.ca.gov/portal/home/ite
m.html?id=fbba842bfl34497c9d611ad506ec48cc
Accessed
1/31/2022
NJ
EPA correspondence
EPA Office of Ground Water and Drinking Water
Accessed
1/31/2022
NM
State of New Mexico - water data
https://catalog.newmexicowaterdata.org/dataset/5d06
9bbb-lbfe-4c83-bbf7-
3 582a42fce6e/resource/03 7d915d-4a28-4c3 9-9922-
3556ec492698/download/mn pws areas.zip
Accessed
1/26/2022
NY
State of New York - Department
of Health
httos ://water. nv. ao\/doh2/aDDlinks/\\atcraual/assets/
PWS GeoJson3.ison
Accessed
1/31/2022
OK
State of Oklahoma - Water
Resources Board
https://www.owrb.ok.gov/maps/data/lay ers/Water%2
OSupply/wssy stem_service_areas.htm;
https://owrb.maps.arcgis.com/apps/webappviewer/in
dex.html?id=68c5f3fd492a43ee8386f39a80f88afb
Accessed
1/26/2022
PA
State of Pennsylvania -
Department of Enviromnental
Protection
https://newdata-padep-
1.opendata.arcgis.com/datasets/public-water-
systems-public-water-supplier-service-
areas/explore?location=40.917958%2C-
77.621150%2C8.24
Accessed
1/12/2022
RI
EPA correspondence
EPA Office of Ground Water and Drinking Water
Accessed
1/31/2022
Accessed through EPA ArcGIS Online Portal
State
Source
Link
Date
AR
EPA ArcGIS - Portal
AZ
EPA ArcGIS - Portal
CT
EPA ArcGIS - Portal
KS
MO
EPA ArcGIS - Portal
EPA ArcGIS - Portal
https://epa.maps.arcgis.com/home/item.html?id=59e
b7810caa044678f1e26e63 7b4fa7 9
Accessed
12/7/2021
MS
EPA ArcGIS - Portal
TX
EPA ArcGIS - Portal
UT
EPA ArcGIS - Portal
NC
EPA ArcGIS - Portal
https://www.nco nemap.gov/search?groupIds=9eb59a
7bdc8e4bdf8cbe2488c8584552
Accessed
1/10/2021
Proposed PFAS Rule Economic Analysis
8-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Under UCMR 3 and 4, PWSs sampled were asked to report U.S. Postal Service zip code(s) for
all areas being served water by a PWS. As such, when pre-delineated PWS service area
boundaries were unavailable, EPA used zip codes served by PWSs to delineate approximated
boundaries using the following steps:
• EPA joined zip codes served—as specified for PWSs in UCMR 3 (U.S. EPA, 2017) and
UCMR 4 (U.S. EPA, 2022c)—to a zip code polygon layer that represented postal service
delivery areas.
• EPA projected zip codes served by PWSs.
• In cases where zip codes did not have polygons (i.e., zip codes for post offices and large
volume mail customers), to map these zip codes as approximate service areas, EPA
selected and overlaid zip code points for each service area with zip code polygons to
select the polygon at that location. Then, EPA merged and dissolved all zip codes (both
point- and polygon-based) to map each service area.
• EPA aggregated all zip code polygons served by each PWS into one boundary
representative of PWS service area boundaries.
• In instances where one zip code was served by multiple PWSs, EPA included the zip
code boundary in all corresponding PWS service area boundaries. For example, if one zip
code was served by two PWSs, both PWS service area boundaries would contain the
same zip code region represented in their boundaries. In some cases, this resulted in EPA
double-counting population demographic characteristics; however, the populations were
not double-counted because population-served data were obtained from SDWIS/Fed and
were unique to each PWS.
• PWSs with pre-delineated PWS service areas (categories 1 and 4), account for 38.8
percent of all PWSs included in the analysis. PWSs with zip code delineated boundaries
(categories 2 and 5), account for 61.2 percent of all PWSs included in the analysis.
Because there is greater accuracy with the predelineated PWS service areas, and to reduce
double-counting of affected populations, EPA removed the portion of the zip code boundaries
that were already accounted for within the predelineated PWS service area boundaries.
For example, in rural areas, the zip code boundaries can be relatively large and therefore overlap
with predelineated PWS service area boundaries. To avoid redundancy and reduce double-
counting populations, EPA used the following approach:
• EPA used predelineated PWS service area boundaries (including overlap94) when
available.
• If predelineated PWS service areas were not available, EPA used zip code-approximated
PWS service area boundaries (as provided in UCMR 3 and UCMR 4).
94 For PWSs with predelineated PWS service area boundaries, EPA conducted a sensitivity analysis of the results of EPA's EJ
exposure analysis to evaluate the impact of retaining PWS boundaries including overlapping areas versus removing overlapping
boundaries. The impact on the results of EPA's EJ exposure analysis showed very few differences across the two approaches. As
such, EPA used service area boundaries with overlapping areas included.
Proposed PFAS Rule Economic Analysis
8-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA carved out or removed predelineated PWS service area boundaries from the zip code-
approximated PWS service area boundaries to reduce the risk of double-counting the
demographic composition of the populations served.
EPA used predelineated PWS service area boundaries and zip code-approximated PWS service
area boundaries as inputs to the EJSCREENbatch R package to estimate the sociodemographic
characteristics of PWS service areas included in the analysis (see Section 8.3.1.2.3 for more
detail on this process) (U.S. EPA, 2022a). The population served counts were obtained from
SDWIS/Fed for each PWS. Further description of the population-served data and
sociodemographic characteristics of the population served by PWS service areas is provided in
Section 8.3.2.1 and in Appendix M.
8.3.1.2.2.1 Categories 1 and 2
Categories 1 and 2 contained PWSs that had sampled PFAS occurrence data from UCMR 3.
Category 1 (n=l,699 PWSs) comprised PWSs that had predelineated PWS service area
boundaries, whereas category 2 (n=3,024 PWSs) comprised PWSs that had zip code-
approximated PWS service area boundaries.
The exposure analysis included service areas for 1,699 category 1 PWSs and 3,024 category 2
PWSs, for a total of 4,723 PWSs. There were 4,920 PWSs that conducted PFAS sampling under
UCMR 3, and categories 1 and 2 PWSs accounted for approximately 96 percent of all PWSs that
participated in UCMR 3. Of the 4,920 PWSs that participated in UCMR 3, 10 PWSs did not have
predelineated PWS service area boundaries or zip code-served data available to approximate
PWS service area boundaries. Systems were excluded from the analysis if they were classified as
"inactive" in SDWIS/Fed (67 PWSs). Additionally, PWSs could not be evaluated if there were
errors processing the EJSCREENbatch R package (120 PWSs). In such instances, the
EJSCREENbatch R package did not provide sociodemographic characteristics for a given PWS
service area.
Category 1 and 2 PWSs account for 239.6 million people served, or approximately 73 percent of
the U.S. population. However, the subset of category 1 and 2 PWSs captured in the analysis
represented roughly 3 percent of active PWSs.95
8.3.1.2.2.2 Categories 4 and 5
EPA used state PFAS occurrence data for PWSs in categories 4 and 5 because these systems did
not monitor for PFAS under UCMR 3. Category 4 (n=311 PWSs) included PWSs that had
predelineated PWS service areas, whereas category 5 (n=148 PWSs) included PWSs that had zip
code-approximated PWS service area boundaries.
The EJ exposure analysis includes PWS service areas for 311 category 4 PWSs and 148 category
5 PWSs. Category 4 and 5 PWSs account for approximately 7 percent of all PWSs with state
PFAS sample occurrence data. Ninety-three PWSs with state PFAS occurrence data have PFAS
occurrence data available in UCMR 3, and therefore are included in the analysis under categories
1 and 2. In addition, EPA included PWSs with state PFAS occurrence data in the analysis only if
finished water samples were available for at least one of the four PFAS analytes. The Agency
95 The number of active public water systems was retrieved from SDWIS Q4 2021 /Fed fourth quarter 2021.
Proposed PFAS Rule Economic Analysis
8-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
could not include many of the PWSs with state PFAS occurrence data because predelineated
PWS service areas or zip code approximated PWS service area boundaries were not available.
Category 4 and 5 PWSs account for 1.2 million people served, or approximately 0.4 percent of
the U.S. population. EPA summarized the results for these PWSs in Appendix M.
8.3.1.2.2.3 Categories 3 and 6
EPA did not include category 3 and 6 PWSs in the EJ exposure analysis because predelineated
PWS service areas and information containing zip codes served by PWSs were both unavailable.
8.3.1.2.3 Sociodemographic Data
EPA used the Agency's EJSCREENbatch R package to characterize the sociodemographic
makeup of populations living in PWS service areas, as described in Section 8.3.1.2.2 (U.S. EPA,
2022a). The EJSCREENbatch R package offers functions to extract and process Census block
group EJScreen data within user-provided geographies. EJScreen uses U.S. Census Bureau's
American Community Survey (ACS) 2015-2019 five-year estimates (U.S. EPA, 2022a).
EJScreen data are input into a function that spatially apportions (i.e., using areal apportionment)
data to service areas using a 1 km resolution raster population dataset from NASA's
Socioeconomic Data and Applications Center.
EPA used the following data outputted from the EJSCREENbatch package on the race, ethnicity,
and poverty status of populations served by the PWSs:
• Race: Percent American Indian or Alaska Native; percent Asian and Pacific Islander;
percent Black or African American; and percent non-Hispanic White.96
• Ethnicity: Percent Hispanic.
• Income: Percent of the population below twice the Federal poverty level; percent of the
population above twice the Federal poverty level.
In addition, the Agency identified PWSs that are Native American-owned and within EPA's
tribal primacy program using SDWIS/Fed data (U.S. EPA, 2021h).
Note that sociodemographic information used for EPA's EJ exposure analysis differs from that
used in EPA's benefits analysis, which relies on SDWIS/Fed and race/ethnicity-specific
population estimates from the U.S. Census Bureau (2020a). In particular, this analysis presents
race and ethnicity separately such that most race categories (Asian and Pacific Islander,
American Indian or Alaskan Native, and Black) include individuals who identify as Hispanic,
while the ethnicity category Hispanic includes individuals who identify as White or a race other
than White. Population estimates from the U.S. Census Bureau are available at the county level,
but more granular location-specific population data was needed for EPA's EJ exposure analysis.
For further information on the use of U.S. Census Bureau population proportions in EPA's
benefits analysis, see Appendix B.
90 In ail effort to avoid double counting populations, race/ethnicity categories reported here do not account for people who
selected "some other race alone" or "two or more races" in the ACS.
Proposed PFAS Rule Economic Analysis
8-24
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.1.3 EJ Exposure Analytic Approach
EPA conducted a baseline analysis of populations served by PWS service areas in categories 1
and 2 to evaluate the demographic characteristics of systems exposed to PFAS concentrations
above a baseline set of thresholds and two hypothetical regulatory thresholds.
For purposes of this baseline analysis, EPA assumed the following baseline thresholds, based on
Method 537.1 detection limits (U.S. EPA, 20 1 8):97-98"
• PFHpA: 0.71 ppt
• PFHxS: 1.4 ppt
• PFOS: 1.1 ppt
• PFOA: 0.53 ppt
EPA also evaluated the rate of exposure using two hypothetical regulatory thresholds: (1) the
UCMR 5 MRL values for each PFAS analyte, and (2) 10.0 ppt. For the purpose of this analysis,
these values are assumed to be individual regulatory thresholds for each contaminant. EPA notes
that while these thresholds are not exactly set at the proposed or alternate MCL values, EPA
began this analysis prior to refinement of those regulatory options. This analysis is not intended
to determine the demographic breakdown of costs and benefits expected to result from the
proposed regulatory option and alternatives; rather, this analysis determines whether vulnerable
communities are disproportionately exposed to PFAS over baseline conditions and these
hypothetical thresholds. The UCMR 5 MRL values for PFOA, PFOS, PFHpA, and PFHxS are as
follows:
• PFHpA: 3 ppt
• PFHxS: 3 ppt
• PFOS: 4 ppt
• PFOA: 4 ppt
EPA compared the estimated population served in each demographic group anticipated to
experience reductions in PFAS exposure under each hypothetical regulatory threshold to the total
97 There are no detection limits reported for Method 533 (U.S. EPA, 2019b).
98 EPA used these detection limits solely as baseline thresholds for purposes of its E.T analysis. EPA has defined the Rule
Detection Limit for purposes of consideration of monitoring data to determine monitoring schedules as 1/3 the MCL for PFOA
and PFOS, or 1.3 ppt. Refer to Sections VI, VIII, and IX of the federal register notice for this proposed regulatory action for
further discussion on EPA's analytical methods and the determination of practical quantitation limits (PQLs).
99 As noted in Section 8.3.1.2.1, different states utilized various reporting, quantification, and/or detection limits when analyzing
and presenting data, and for some states, no clearly defined limits were publicly provided as part of the dataset. Further, the limits
often varied within the data for each state depending on the specific PFAS analyte. In some cases, states reported detection,
quantification, or reporting limits and/or presented data at concentrations below EPA's proposed rule detection limits and/or
practical quantitation limits provided in the federal register notice for this proposed regulatory action. For more information on
the collection and analysis of occurrence data, see U.S. EPA (2022j).
Proposed PFAS Rule Economic Analysis
8-25
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
population served across all demographic groups. This analysis seeks to answer the following
question: When PFAS occurs in drinking water over a certain threshold, will vulnerable
communities be disproportionately exposed to PFAS compared to the total population that is
exposed to PFAS over the same threshold?
As described above, EPA's EJ exposure analysis for the proposed rule uses data from EJScreen
and the American Community Survey to examine anticipated exposure above set baseline and
theoretical regulatory thresholds using system-level mean occurrence data. As the literature
shows, the degree to which a community is above a specific PFAS threshold can vary. As such,
EPA also characterized population-weighted mean concentrations of PFAS to evaluate the extent
to which the levels of potential exposure correlate with community characteristics. EPA requests
comment on whether considering additional thresholds, metrics, or analyses would further
elucidate relative demographic disparities. In particular, EPA requests comment on whether
further investigation of pockets of concern, such as a detailed break-out analysis for one or more
demographic groups, would improve the analysis.
8.3.2 EJ Exposure Analysis Results
This section describes the demographic characterization of category 1 and 2 PWS service areas
in the baseline as well as the results of the analysis exploring the EJ implications of two
hypothetical regulatory thresholds. EPA focused on category 1 and 2 PWS service areas due to
the availability of spatial boundaries (from both predelineated PWS service area boundaries and
zip code-approximated PWS service area boundaries) and PFAS occurrence data from UCMR 3.
Results from categories 4 and 5 are reported in Appendix M.
8.3.2.1 Demographic Profile of PWS Service Areas
Table 8-3 summarizes the breakdown of category 1 and 2 PWS service areas by state and by
size, where small systems are those serving fewer than or equal to 10,000 people. In total, these
PWSs account for roughly 240 million people served, or approximately 73 percent of the U.S.
population. Category 1 and 2 PWSs span all states in the continental U.S. Category 1 and 2
PWSs included in this analysis capture roughly 3 percent of active PWSs. Among the 3 percent
of active PWSs captured by EPA's analysis (i.e., category 1 and 2 PWSs), there are 28 PWSs
within EPA's tribal primacy program, serving a population of approximately 314,182 people.
Additionally, approximately 19 percent of the systems are defined as small (serving fewer than
10,000 people), accounting for 1.4 percent of the total population served.
Table 8-4 summarizes the demographic profile for category 1 and 2 PWS service areas and
compares it to the demographic characteristics of the overall U.S. population. There are slight
differences in the demographic characteristics of the population served by PWS service areas
included in EPA's analysis compared to the overall U.S. population, with percent differences all
being less than +/- 3 percent. The population served by these PWSs has slightly higher
percentages of Asian and Pacific Islander (+0.9%) and Black (+1.8%) populations compared to
the overall U.S. population. The percentage of American Indian or Alaska Native populations is
consistent with the percent of these populations across the U.S. The Hispanic population served
by category 1 and 2 PWSs is slightly higher (+1.3%) and the non-Hispanic White population is
lower (-3%) than that of the overall U.S. population. When examining income demographics,
Table 8-4 shows that category 1 and 2 PWSs have a slightly higher percentage of populations
Proposed PFAS Rule Economic Analysis
8-26
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
with income below twice the poverty level (+2.2%) and a slightly lower percentage of population
with income above twice the poverty level (-2.2%) compared to the overall U.S. population.
Proposed PFAS Rule Economic Analysis
8-27
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 8-3: Number of Category 1 and 2 PWSs and Populations Served by Size and State
State
Number of Service
Areas
Percent Small
Service Areas
Total Population
Served3
Population Served in
Small Systems3
Population Served in
Medium and Large
Systems
Tribal Service Areas
28
50%
314,182
44,571
269,611
Alabama
124
15%
4,488,042
86,106
4,401,936
Arizona
68
19%
5,561,792
44,818
5,516,974
Arkansas
57
32%
1,449,872
81,217
1,368,655
California
412
9%
34,438,454
146,260
34,292,194
Colorado
200
61%
5,756,473
227,838
5,528,635
Connecticut
42
14%
2,457,248
13,799
2,443,449
Delaware
13
23%
642,261
13,535
628,726
District of Columbia
2
0%
648,013
-
648,013
Florida
258
11%
19,355,085
111,293
19,243,792
Georgia
126
16%
8,816,216
77,382
8,738,834
Idaho
26
23%
991,096
16,854
974,242
Illinois
252
13%
9,703,392
121,219
9,582,173
Indiana
100
19%
3,791,557
62,381
3,729,176
Iowa
57
26%
1,810,021
52,241
1,757,780
Kansas
40
33%
1,424,944
41,732
1,383,212
Kentucky
118
21%
3,572,262
169,375
3,402,887
Louisiana
87
28%
3,353,978
86,822
3,267,156
Maine
16
19%
411,385
16,456
394,929
Maryland
39
21%
4,980,513
20,084
4,960,429
Massachusetts
171
9%
6,236,022
74,117
6,161,905
Michigan
158
16%
5,895,618
122,403
5,773,215
Minnesota
98
14%
3,478,561
40,952
3,437,609
Mississippi
78
32%
1,400,826
88,145
1,312,681
Missouri
86
24%
3,879,698
87,393
3,792,305
Montana
15
40%
416,576
10,070
406,506
Nebraska
21
33%
1,136,091
12,642
1,123,449
Proposed PFAS Rule Economic Analysis
8-28
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 8-3: Number of Category 1 and 2 PWSs and Populations Served by Size and State
State
Number of Service
Areas
Percent Small
Service Areas
Total Population
Served3
Population Served in
Small Systems3
Population Served in
Medium and Large
Systems
Nevada
16
25%
2,826,471
10,200
2,816,271
New Hampshire
23
22%
570,449
10,907
559,542
New Jersey
163
9%
7,567,370
54,089
7,513,281
New Mexico
22
18%
590,288
6,862
583,426
New York
167
18%
15,963,267
96,915
15,866,352
North Carolina
145
14%
6,706,695
82,447
6,624,248
North Dakota
12
25%
425,637
4,903
420,734
Ohio
183
15%
8,969,887
112,278
8,857,609
Oklahoma
62
24%
2,482,622
49,472
2,433,150
Oregon
65
17%
2,875,275
33,730
2,841,545
Pennsylvania
170
20%
8,424,012
130,731
8,293,281
Rhode Island
17
12%
934,307
12,485
921,822
South Carolina
81
11%
3,487,233
46,773
3,440,460
South Dakota
18
28%
458,464
17,065
441,399
Tennessee
136
12%
6,114,639
86,951
6,027,688
Texas
329
26%
15,408,123
319,661
15,088,462
Utah
62
13%
2,595,756
32,847
2,562,909
Vermont
12
50%
142,888
23,438
119,450
Virginia
81
16%
6,291,660
48,692
6,242,968
Washington
132
15%
6,304,525
70,712
6,233,813
West Virginia
32
31%
818,159
35,605
782,554
Wisconsin
92
20%
2,920,851
82,496
2,838,355
Wyoming
11
18%
268,828
3,341
265,487
TOTAL
4,723
19%
239,557,584
3,242,305
236,315,279
Abbreviations: PWS - public water system.
Note:
Proposed PFAS Rule Economic Analysis
8-29
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 8-3: Number of Category 1 and 2 PWSs and Populations Served by Size and State
„ Number of Service Percent Small Total Population Population Served in 'J.°.')U1!at'0n ^l"vc(' 'n
State . c-i r, ' „ „ ^ Medium and Large
Areas Service Areas Served3 Small Systems3 „ ,
Systems
Population served by PWSs was obtained from SDWIS/Fed fourth quarter 2021. Small systems include those serving fewer than or equal to 10,000 people. Medium and large
systems serve populations more than 10,000 people.
Table 8-4: Population Served by Category 1 and 2 PWSs Compared to Percent of U.S. Population by Demographic Group
Results
Race and Ethnicity
Income
American
Indian or
Alaska Native
Asian and
Pacific
Islander
Black
Hispanic
Non-Hispanic
White
Below Above
Twice the Twice the
Poverty Poverty
Level Level
Total
Population
Served
Population Served
Percent of Total
Population Served
U.S. Population
Percent by
Demographic
Group3
Percent Difference
Between
Population Served
and U.S.
Population
1,518,369
0.6%
0.8%
16,087,571
6.7%
5.8%
34,583,262
14.4%
12.6%
46,573,256 136,673,130 76,716,883 162,840,701
19.4% 57.1% 32.0% 68.0%
18.2%
60.1%
29.8%
70.2%
-0.2%
0.9%
1.8%
1.3%
-3%
2.2%
-2.2%
239,557,584
100.0%
Note:
aU.S. population estimates were obtained from the U.S. Census Bureau's American Community Survey 2016-2020 five-year estimates.
Proposed PFAS Rule Economic Analysis
8-30
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.2 Exposure Analysis Results
8.3.2.2.1 Baseline Scenario
To evaluate impacts of the proposed rule on population groups of concern, the percent of a
specific demographic group with modeled PFAS above baseline thresholds needs to be presented
in relation to another group, typically referred to as a comparison group. The way in which the
comparison group is defined can have important implications for identifying differences in
potential exposure across population groups of concern in an EJ analysis. The Agency's
Technical Guidance for Assessing Environmental Justice in Regulatory Analysis notes that the
comparison group can be defined as individuals with similar socioeconomic characteristics
across different areas in the state, region or nation (i.e., within-group comparison) or as affected
individuals with different socioeconomic characteristics (i.e., across-group comparison) (U.S.
EPA, 2016h).
For this proposed regulatory action, EPA is examining individuals served by PWSs with
modeled PFAS exposure above the baseline concentration threshold or a specific hypothetical
alternative policy threshold. EPA presents the total affected population as a possible metric of
comparison, noting however that each affected demographic group is reflected also within the
total affected population. It is possible that EPA understates the magnitude of disproportionate
baseline exposure to PFAS for populations of concern by using the total affected population as
the basis of comparison.
As currently defined, race and ethnicity classifications are generally presented separately such
that the race categories include individuals who identify as Hispanic, while Hispanic ethnicity
includes individuals who identify as a race other than White. In aggregate, those who identify as
a race or ethnicity other than White and/or Hispanic are considered "people of color" when
considering potential EJ concerns. Thus, the disaggregated race and ethnicity categories in the
current analysis reflect some double counting among affected populations that ultimately
compose the aggregate category, people of color. EPA has therefore included the category non-
Hispanic White in the analysis, as this category does not include individuals who identify as a
race or ethnicity included within people of color. EPA requests comment on all aspects of the
environmental justice analysis, including its choice of comparison groups to help identify
potential demographic disparities in anticipated PFAS exposure.
The results of EPA's analysis of baseline exposure are shown in Table 8-5 and Table 8-6. Table
8-5 summarizes the population served by category 1 and 2 PWSs with modeled PFAS
occurrence above baseline thresholds based on the Method 537.1 detection limits. The second set
of rows in Table 8-5 summarizes the percentage of the total population served by demographic
group with modeled PFAS occurrence above these baseline thresholds by demographic group.
Table 8-7 shows average population-weighted PFAS concentrations across demographic groups.
In Table 8-5, percentages are bolded and italicized when the percentage of the population in a
specific demographic group with modeled PFAS above the baseline threshold is greater than the
percentage of the total population across all demographic groups exposed to modeled PFAS
above this threshold (right-hand column). In Table 8-5, the highlighted numbers represent where
percentages of the population served in a particular demographic group are more than 1 percent
greater than percentages of the total population. In Table 8-6, highlighted cells represent whether
the average concentration for a given demographic group is higher than the average for the total
Proposed PFAS Rule Economic Analysis
8-31
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
population served across all demographic groups (right-hand column). Higher percentages or
concentrations indicate higher PFAS exposure for a given demographic group compared to the
percentage of the population served across all demographic groups. Between 7.1 percent and
12.6 percent of the total population served for category 1 and 2 PWS service areas, depending on
the analyte, are exposed to modeled PFAS occurrence above baseline thresholds based on the
Method 537.1 detection limits.
The following are findings from EPA's baseline EJ exposure analysis:100
• The percentage of Asian and Pacific Islander and Hispanic populations served with
exposure to PFAS above baseline thresholds is higher across all four PFAS analytes
compared to the percentage of the total population served across all demographic groups
with anticipated PFAS exposure above the baseline thresholds. These percentages are
also higher than those of non-Hispanic White populations. Most percentages are more
than 1 percent greater than percentages exposed across the total population.
• The percentage of Asian and Pacific Islander populations served with exposure above
baseline thresholds is 0.7 percent to 2.2 percent points higher (depending on the analyte)
than the percentages of the population served across all demographic groups. When
compared to non-Hispanic White populations, the percentages are 1.8 to 3.4 percentage
points higher.
• The percentage of Hispanic populations served with exposure above baseline thresholds
is 3.1 percent to 3.6 percentage points higher (depending on the analyte) than the
percentage of the population served across all demographic groups. When compared to
non-Hispanic White populations, the percentages are 4.0 to 4.8 percentage points higher.
• While the percentage of American Indian or Alaska Native and Black populations served
have similar PFAS exposure above the baseline thresholds for PFAS compared to the
percentages of the population served across all demographic groups, they are somewhat
higher than those of non-Hispanic White populations for three of the four PFAS analytes.
• Other demographic groups, including those representing relative income status, are
anticipated to experience percentages of PFAS occurrence above baseline thresholds
similar to (within 0.5%) the percentage of the population served across all demographic
groups.
Table 8-6 characterizes population-weighted mean concentrations of PFAS by demographic
group. In addition to having a higher percentage of population served by PWSs with above
baseline concentrations of PFAS, Asian and Pacific Islander and Hispanic populations are also
exposed to higher mean concentrations than is typical for the total population served. Hispanic
populations are the most highly exposed across all four PFAS. On average, they are exposed to
0.2-0.3 ppt more of each of the four analyzed PFAS than non-Hispanic White populations
served. The results also suggest that Black, American Indian and Alaska Native, and low-income
individuals are exposed to higher average concentrations than the total population served for at
least two PFAS in each case. This finding suggests that, while these populations may not always
be more likely to be served by public water systems with above baseline concentrations of PFAS,
they may be exposed to higher average concentrations when exposure does occur. Collectively,
100 Although differences in anticipated exposure between a particular demographic group and the entire sample population are
<5%, all results are reported in EPA's summary of results regardless of magnitude.
Proposed PFAS Rule Economic Analysis
8-32
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
people of color are potentially exposed to 0.1 - 0.2 ppt more of each of the four PFAS analyzed
than non-Hispanic White populations served.
Proposed PFAS Rule Economic Analysis
8-33
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-5: Baseline Scenario: Population Served by Category 1 and 2 PWS Service Areas Above Baseline Thresholds and as
a Percent of Total Population Served
Race and Ethnicity
Income
Results
PFAS
American
Indian or
Alaska Native
Asian and
Pacific
Islander
Black
Hispanic
Non-Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Total
Population
Served
Population
Served Above
PFOS
PFHxS
148,381
114,653
1,958,054
1,492,095
3,469,513
2,454,674
6,612,120
4,817,842
12,941,535
8,024,430
7,879,907
5,400,599
17,530,418
11,675,163
25,410,325
17,075,762
Baseline
Threshold
PFHpA
PFOA
141,360
163,560
1,697,425
2,276,202
3,295,139
4,019,351
6,330,047
7,346,397
12,073,566
16,125,251
7,553,195
9,341,681
16,239,858
20,956,016
23,793,053
30,297,697
Population
Served Above
Baseline
Threshold as a
PFOS
PFHxS
9.8%
7.6%
12.2%
9.3%
10.0%
7.1%
14.2%
10.3%
9.5%
5.9%
10.3%
7.0%
10.8%
1.2%
10.6%
7.1%
PFHpA
9.3%
10.6%
9.5%
13.6%
8.8%
9.8%
10.0%
9.9%
Percent of
Total
Population
Served
PFOA
10.8%
14.1%
11.6%
15.8%
11.8%
12.2%
12.9%
12.6%
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - periluorohexanesulfonic acid; PFOA - periluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Proposed PFAS Rule Economic Analysis
8-34
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-6: Modeled Average PFAS Concentrations (ppt) by Demographic Group in the Baseline, Category 1 and 2 PWS
Service Areas
Race and Ethnicity
Income
PFAS
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
People of
Color3
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
loiai
Population
Served
PFOS
0.73
0.77
0.71
0.92
0.81
0.62
0.73
0.69
0.70
PFHxS
0.59
0.53
0.51
0.62
0.56
0.44
0.52
0.48
0.49
PFHpA
0.39
0.40
0.44
0.51
0.46
0.39
0.42
0.42
0.42
PFOA
0.79
0.93
0.89
1.03
0.95
0.83
0.87
0.88
0.88
Abbreviations: PFHpA - periluoroheptaiioic acid; PFHxS - periluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-35
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.2.2 Hypothetical Regulatory Scenario #1: UCMR5 MRLs
Table 8-7 and Table 8-8 summarize the results for population served by category 1 and 2 PWSs
with PFAS occurrence above UCMR 5 MRL values. For this hypothetical regulatory scenario,
EPA assumed that PWSs with PFAS system-level means above the MRL value will reduce
PFAS levels to comply with the proposed rule. The first set of rows in Table 8-7 summarizes
populations served by category 1 and 2 PWS service areas with modeled PFAS occurrence above
the UCMR 5 MRLs. The second set of rows provides these estimates as a percentage of the total
population served by PWSs included in EPA's analysis. Table 8-8 summarizes the population-
weighted average reductions in PFAS assuming all PWSs reduce their concentrations to UCMR
5 MRL levels.
In Table 8-7, percentages are bolded and italicized when the percentage of the population in a
specific demographic group with PFAS occurrence above the MRL value is greater than the
percentage of the total population across all demographic groups with PFAS occurrence above
the MRL (right-hand column). In Table 8-7, the highlighted numbers represent where
percentages of the population served in a particular demographic group are more than 1 percent
greater than percentages of the total population. In Table 8-8, highlighted cells represent whether
the average reduction in PFAS concentrations for a given demographic group is higher than the
average for the total populations served across all demographic groups (right-hand column). The
percentages that are bolded, italicized, or highlighted indicate higher PFAS exposure above the
MRL for a given demographic group; EPA anticipates that relatively higher reductions in PFAS
exposure will accrue to these demographic groups under this hypothetical regulatory scenario
compared to the percentage of the population across all demographic groups. EPA provides
additional details on anticipated exposure above UCMR 5 MRL values in Appendix M.
Between 2.7 percent and 4.8 percent of the population served by category 1 and 2 PWS service
areas, depending on the PFAS analyte, are exposed to modeled PFAS concentrations above the
MRL for PFOS, PFOA, PFHpA, and PFHxS. Under this hypothetical regulatory scenario, where
MCLs are assumed to be equal to UCMR 5 MRL values, EPA expects these populations to
experience reductions in PFAS exposure to below the hypothetical regulatory thresholds. EPA's
analysis of the demographic distribution of anticipated health benefits and household costs due to
reductions in PFAS exposure resulting from the proposed PFAS rule and regulatory alternatives
is discussed in Section 8.4.2.
Based on this analysis, American Indian or Alaska Native, Asian and Pacific Islander, Black,
Hispanic, and low-income populations are estimated to face higher rates of system-level mean
PFAS exposure above UCMR 5 MRL values compared to rates of exposure over these
thresholds for the population served across all demographic groups. The differences are even
greater when compared to the rates of exposure over these thresholds for non-Hispanic White
populations. Specifically, American Indian or Alaska Native populations served have higher
exposure above the UCMR 5 MRL values for PFOS, PFHxS, and PFHpA compared to the
percent of the population served across all demographic groups. These differences in exposure
are larger when compared to non-Hispanic White populations. Asian and Pacific Islander
populations served have higher exposure above the UCMR 5 MRL values for PFOS and PFOA
compared to the percent of the population served across all demographic groups. For PFOA and
PFOS, the percentage of Asian and Pacific Island populations exposed is over 1% greater than
for non-Hispanic White populations. Black populations served have higher exposure above the
Proposed PFAS Rule Economic Analysis
8-36
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
UCMR 5 MRL values for PFHpA and PFOA, compared to the percent of the population served
across all demographic groups, and they have higher exposure above the UCMR 5 MRLs for
PFOS and PFHxS compared to non-Hispanic White populations. Hispanic populations served
have higher exposure above the UCMR 5 MRL values across all four PFAS analytes compared
to the percent of the population served across all demographic groups. The percent of Hispanic
populations served with exposure above the UCMR 5 MRL values is generally at least double
the percent of non-Hispanic White populations with exposure above the UCMR 5 MRL values.
This is the most notable difference in exposure. The percent differences observed suggest that, in
this analysis, Hispanic populations are estimated to face nearly twice the level of exposure for all
four PFAS analytes compared to the entire sample population across all demographic groups. As
such, Hispanic populations could also be expected to experience the greatest reductions in PFAS
exposure under this hypothetical regulatory scenario. Populations served with income less than
twice the poverty level have higher PFAS exposure above the UCMR 5 MRL values across all
four PFAS analytes compared to the percent of the population served across all demographic
groups. Exposure percentages for populations served with income less than twice the poverty
level are at least 1% greater than exposure for non-Hispanic White populations. Table 8-8
displays population-weighted reductions in PFAS exposure in a hypothetical regulatory scenario
where system-level means are reduced to UCMR 5 MRLs to comply with the proposed rule.
Hispanic populations see the greatest reductions in concentrations for three PFAS in this
hypothetical regulatory scenario, which is consistent with Table 8-7. However, despite having
lower percentage of population affected than Hispanic populations, American Indian and Alaska
Native populations see the greatest reduction in PFHxS of any demographic group in this
hypothetical regulatory scenario. Black populations also see greater reductions in PFOS and
PFHxS than the average across the total population served, even though the percentage of Black
individuals with exposure above UCMR 5 MRLs is the same as the percentage of the population
served across all demographic groups. Collectively, people of color and those with income less
than twice the poverty level see greater reductions in PFAS exposure across all four analytes in
comparison to the total population served. These differences in PFAS reductions are larger when
compared to the non-Hispanic White population.
Proposed PFAS Rule Economic Analysis
8-37
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-7: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above UCMR 5 MRLs and as a Percent of Total Population Served
Race and Ethnicity
Income
Total
Results
PFAS
American
Indian or
Alaska Native
Asian and
Pacific
Islander
Black
Hispanic
Non-Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Population
Served
Population
Served
Above
PFOS
PFHxS
82,997
81,981
861,159
656,539
1,585,417
1,419,167
3,612,399
3,161,527
4,900,864
4,492,124
3,739,513
3,268,486
7,384,230
6,613,772
11,123,742
9,882,258
x VLy V_/ V ^
UCMR5
PFHpA
43,064
386,810
1,075,549
2,309,203
2,648,871
2,275,173
4,229,947
6,505,119
MRL
PFOA
60,023
867,254
1,809,985
3,500,258
5,245,421
3,737,007
7,846,216
11,583,222
Population
Served
PFOS
5.5%
5.4%
4.6%
7.8%
3.6%
4.9%
4.5%
4.6%
Above
UCMR 5
PFHxS
5.-1%
4.1%
4.1%
6.8%
3.3%
4.3%
4.1%
4.1%
MRL as a
Percent of
Total
PFHpA
2.8%
2.4%
3.1%
5.0%
1.9%
3.0%
2.6%
2.7%
Population
Served
PFOA
4.0%
5.4%
5.2%
7.5%
3.8%
4.9%
4.8%
4.8%
Abbreviations: MRL - minimum reporting level; PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS -
perfluorooctanesulfonic acid; UCMR - Unregulated Contaminant Monitoring Rule.
Proposed PFAS Rule Economic Analysis
8-38
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-8: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a Hypothetical Regulatory
Scenario with Maximum Contaminant Level at the UCMR 5 MRLs, Category 1 and 2 PWS Service Areas
PFAS
PFOS
PFHxS
PFHpA
PFOA
American
Indian or
Alaska
Native
0.26
0.25
0.04
0.16
Race and Ethnicity
Income
Asian and
Pacific
Islander
0.25
0.16
0.03
0.22
Black
0.25
0.18
0.06
0.19
Hispanic
0.33
0.19
0.07
0.25
People of
Color3
0.28
0.18
0.06
0.22
Non-
Hispanic
White
0.18
0.13
0.04
0.16
Below
Twice the
Poverty
Level
0.26
0.18
0.06
0.19
Above
Twice the
Poverty
Level
0.21
0.14
0.05
0.18
Total
Population
Served
0.22
0.15
0.05
0.18
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - periluorohexanesulfonic acid; PFOA - periluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-39
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.2.3 Hypothetical Regulatory Scenario #2: 10.0 ppt
Table 8-9 and Table 8-10 summarize the results of the population served by category 1 and 2
PWS service areas with modeled PFAS occurrence above 10.0 ppt. The first set of rows in Table
8-9 summarizes populations served by PWSs with PFAS occurrence above 10.0 ppt. The second
set of rows displays these estimates as a percent of the total population served for PWSs included
in EPA's analysis. Table 8-10 shows the population-weighted average reduction in PFAS
concentrations assuming all PWSs reduce their concentrations to 10.0 ppt.
In Table 8-9, percentages are bolded and italicized when the percentage of the population in a
specific demographic group with PFAS occurrence above 10.0 ppt is greater than the percentage
of the total population served across all demographic groups with PFAS occurrence above 10.0
ppt (right-hand column). In Table 8-10, highlighted cells represent whether the average reduction
in PFAS concentrations for a given demographic group is higher than the average for the total
populations served across all demographic groups (right-hand column). The percentages that are
bolded, italicized, or highlighted indicate greater PFAS exposure above 10.0 ppt for a given
demographic group compared to the total population served across all demographic groups; EPA
anticipates potentially relatively higher reductions in PFAS exposure to accrue to these
demographic groups under this hypothetical regulatory scenario compared to the percentage of
population across all demographic groups. Unlike the results from EPA's exposure analysis
where UCMR 5 MRLs are used as hypothetical MCL values, percentages in particular
demographic groups are less than 1 percent greater than percentages across the total population.
Between 0.2 percent and 1.1 percent of the population served by category 1 and 2 PWS service
areas, depending on the PFAS analyte, is exposed to PFAS occurrence above 10.0 ppt.
The following are findings from the EJ exposure analysis for PFAS occurrence above 10.0 ppt:
• American Indian or Alaska Native, Asian and Pacific Islander, Black, Hispanic, and low-
income populations have slightly higher PFAS exposure above 10.0 ppt for some PFAS
analytes compared to the population served across all demographic groups. These results
are essentially unchanged when comparing exposures above 10.0 ppt to non-Hispanic
White populations.
• The most notable difference is for PFHxS exposure for American Indian or Alaska
Native populations served, with 1.0 percent of American Indian and Alaska Native
populations served with PFAS exposure above 10.0 ppt compared to 0.4 percent of
population served across all demographic groups.
Table 8-10 characterizes population-weighted average reductions of PFAS by demographic
group in a hypothetical regulatory scenario where system-level means are reduced to 10.0 ppt.
This analysis reinforces the results for PFOS in Table 8-9, showing that people of color see
greater reductions in PFOS than the average for the total population served. Notably, for PFHxS,
Asian and Pacific Islander, Black, and Hispanic populations see greater reductions than the total
population served despite having similar percentages exposed above 10.0 ppt. Collectively,
people of color and populations with income below twice the poverty level see greater reductions
in PFOS and PFHxS than the total population served across all demographic groups.
Proposed PFAS Rule Economic Analysis
8-40
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-9: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above 10.0 ppt and as a Percent of Total Population Served
Race and Ethnicity Income
Total
Results
PFAS
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Population
Served
PFOS
18,295
189,271
431,476
602,304
1,275,424
931,413
1,623,668
2,555,081
Population
PFHxS
14,897
70,637
133,867
174,242
564,113
357,196
620,248
977,444
Served Above
10.0 ppt
PFHpA
2,535
10,837
82,138
52,059
220,554
131,151
244,513
375,664
PFOA
7,502
139,046
230,168
213,670
796,015
437,039
972,748
1,409,787
Population
PFOS
1.2%
1.2%
1.2%
1.3%
0.9%
1.2%
1.0%
1.1%
Served Above
PFHxS
1.0%
0.4%
0.4%
0.4%
0.4%
0.5%
0.4%
0.4%
10.0 ppt as a
Percent of
PFHpA
0.2%
0.1%
0.2%
0.1%
0.2%
0.2%
0.2%
0.2%
Total
Population
PFOA
0.5%
0.9%
0.7%
0.5%
0.6%
0.6%
0.6%
0.6%
Served
Abbreviations: PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; ppt - parts per
trillion.
Proposed PFAS Rule Economic Analysis
8-41
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-10: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a Hypothetical Regulatory
Scenario with Maximum Contaminant Level at 10.0 ppt, Category 1 and 2 PWS Service Areas
PFAS
PFOS
PFHxS
PFHpA
PFOA
American
Indian or
Alaska
Native
0.10
0.06
0.00
0.04
Race and Ethnicity
Income
Asian and
Pacific
Islander
0.11
0.07
0.00
0.04
Black
0.11
0.08
0.01
0.06
Hispanic
0.12
0.07
0.00
0.04
People of
Color3
0.11
0.07
0.00
0.05
Non-
Hispanic
White
0.08
0.05
0.01
0.05
Below
Twice the
Poverty
Level
0.11
0.08
0.01
0.05
Above
Twice the
Poverty
Level
0.08
0.05
0.00
0.04
Total
Population
Served
0.09
0.06
0.00
0.05
Abbreviations: PFHpA - periluoroheptaiioic acid; PFHxS - periluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-42
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.3 Comparison of Results by PWS Size
8.3.2.3.1Demographic Profile of PWS Service Areas
Table 8-11 and Table 8-12 summarize the demographic profile for category 1 and 2 PWS service
areas by system size for large and small PWS service areas, respectively. Small systems are
defined as systems serving fewer than or equal to 10,000 people while large systems serve more
than 10,000 people. Table 8-11 and Table 8-12 also provide a comparison to the demographic
characteristics of the overall U.S. population. Because category 4 and 5 PWS service areas make
up a relatively smaller proportion of the sample of PWS service areas included in EPA's
analysis, results for category 4 and 5 PWS service areas are not compared by size due to
inadequate sample size to conduct this analysis.
Table 8-11 shows that the population served by large category 1 and 2 PWS service areas has
slight differences in demographic characteristics compared to the overall U.S. population, with
percent differences all being less than +/- 3.3 percent. PWS service areas have higher
percentages of Black (+1.9%), Hispanic (+1.35%), and Asian and Pacific Islander populations
(+0.98%) populations and populations with income below twice the poverty level (+2.2%)
compared to the overall U.S. population. Additionally, the population served by large category 1
and 2 PWS service areas has lower percentages of non-Hispanic White (-3.3%) populations and
populations with income above twice the poverty level (-2.2%) compared to the overall U.S.
population. The percentage of American Indian or Alaska Native populations is relatively
consistent with the percent of these populations across the U.S.
Table 8-12 shows that the population served by small category 1 and 2 PWS service areas has
considerable differences in the demographic characteristics of the population served compared to
the overall U.S. population, with percent differences being generally greater than +/- 2.5 percent,
and the greatest difference being +13.09 percent. The population served by small category 1 and
2 PWS service areas has lower percentages of Asian and Pacific Islander (-3.7%), Black (-
2.79%), and Hispanic (-6.59%) populations and populations with income above twice the
poverty level (-4.07%) compared to the overall U.S. population. Additionally, the population
served by small category 1 and 2 PWS service areas has higher percentages of American Indian
or Alaska Native (+1%), non-Hispanic White (+13.09%) populations, and populations with
income below twice the poverty level (+4.07%) compared to the overall U.S. population.
Proposed PFAS Rule Economic Analysis
8-43
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-11: Population Served by Category 1 and 2 PWSs and Percent of U.S. Population by Demographic Group, Large
Systems
Race and Ethnicity
Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below Twice
the Poverty
Level
Above Twice
the Poverty
Level
Total
Population
Served
Population Served
Percent of Total Population
Served
U.S. Population Percent
Percent Difference Between
Population Served Percent
and U.S. Percent
1,460,058
0.62%
0.80%
-0.18%
16,019,564
6.78%
5.80%
0.98%
34,265,272
14.50%
12.60%
1.90%
46,196,824 134,300,144
19.55% 56.83%
18.20% 60.10%
1.35%
-3.27%
75,618,752 160,696,528 236,315,280
32.00% 68.00% 100.00%
29.80% 70.20%
2.20% -2.20%
Table 8-12: Population Served by Category 1 and 2 PWSs and Percent of U.S. Population by Demographic Group, Small
Systems
Race and Ethnicity
Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below Twice
the Poverty
Level
Above Twice
the Poverty
Level
Total
Population
Served
Population Served
Percent of Total
Population Served
U.S. Population Percent
Percent Difference
between Population
Served Percent and U.S.
Percent
58,311
1.80%
0.80%
1.00%
68,007
2.10%
5.80%
-3.70%
317,990
9.81%
12.60%
-2.79%
376,431
11.61%
18.20%
-6.59%
2,372,983
73.19%
60.10%
13.09%
1,098,134
33.87%
29.80%
4.07%
2,144,171
66.13%
70.20%
-4.07%
3,242,305
100.00%
Proposed PFAS Rule Economic Analysis
8-44
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.3.2 Baseline Scenario
Table 8-13 and Table 8-14 summarize the populations served by large and small category 1 and
2 PWS service areas with modeled PFAS occurrence above baseline thresholds based on the
Method 537.1 detection limits. The second set of rows in Table 8-13 and Table 8-14 summarize
the percentage of the total population served by demographic group with modeled PFAS
occurrence above baseline thresholds. Percentages are bolded when the percentage of the
population in a specific demographic group with modeled PFAS above baseline thresholds is
greater than the percentage of the total population across all demographic groups exposed to
modeled PFAS above the baseline thresholds. Additionally, percentages are highlighted when
the percentage of the population in a specific demographic group with modeled PFAS above the
baseline threshold represents greater than a 1 percent difference compared to the total population
across all demographic groups. Table 8-15 characterizes population-weighted average PFAS
concentrations across demographic groups in large and small category 1 and 2 PWSs.
Highlighted cells represent whether the average concentration for a given demographic group is
higher than the average concentration for the total population served across all demographic
groups (right-hand column).
Depending on the PFAS analyte, between 7.2 percent and 12.8 percent of the total population
served by large category 1 and 2 PWS service areas are exposed to modeled PFAS occurrence
above baseline thresholds based on the Method 537.1 detection limits. Depending on the PFAS
analyte, between 1.5 percent and 3.4 percent of the total population served by small category 1
and 2 PWS service areas is exposed to modeled PFAS occurrence above baseline thresholds
based on the Method 537.1 detection limits.
For large systems, the percentage of Asian and Pacific Islander and Hispanic populations served
by category 1 and 2 PWS service areas is higher across all four PFAS analytes compared to the
percentage of the total population served across all demographic groups with anticipated PFAS
exposure above baseline thresholds. Depending on the PFAS analyte, the percentage of Asian
and Pacific Islander populations served with exposure above baseline thresholds is 0.6 percent to
2.1 percentage points higher than percentages of the population served across all demographic
groups, or 1.6 to 3.3 percentage points higher than for non-Hispanic White populations.
Depending on the PFAS analyte, the percent of Hispanic populations served with exposure above
baseline thresholds is 3.1 percent to 4.7 percent higher than for the population served across all
demographic groups. This difference is 3.7 to 4.7 percentage points greater when compared to
the non-Hispanic White population.
For small systems, the percentage of Black populations served by category 1 and 2 PWS service
areas with PFAS above baseline thresholds is higher across all four PFAS analytes compared to
the percentage of the total population served across all demographic groups with anticipated
PFAS exposure above baseline thresholds. Depending on the PFAS analyte, the percent of Black
populations served with exposure above baseline thresholds is 0.2 percent to 2.2 percentage
points higher than percentages of the population served across all demographic groups. Given the
data gaps in occurrence information among small systems, extrapolating these results to small
systems across the country is not possible.
Table 8-15 provides detail on average concentrations across these demographic groups for large
and small water systems, respectively. The first panel of Table 8-15 supports the previous
Proposed PFAS Rule Economic Analysis
8-45
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
findings in Table 8-13 that, for large PWSs, Asian and Pacific Islander as well as Hispanic
populations served have greater exposure across at least three PFAS in comparison to exposure
for the total population served across all demographic groups. In addition, Table 8-15
demonstrates that Black and American Indian or Alaska Native populations have greater
exposure to PFOS and PFHxS in comparison to average exposure for the total population served
across all demographic groups for large PWSs, even though these groups have similar or even
lower percentages of exposure in comparison to the total population served. Collectively, people
of color and populations with income less than twice the poverty level have greater average
exposure to at least three PFAS in comparison to the total population served across all
demographic groups for large PWSs. These differences in potential exposure are greater when
compared to the non-Hispanic White population across all four PFAS.
The second panel of Table 8-15 shows that Black and non-Hispanic White populations have
greater potential exposure to PFOS and PFOA in comparison to the total population served
across all demographic groups served by small PWSs. Non-Hispanic White populations also see
greater exposure to PFHxS in comparison to the total population served in small PWSs. Of note,
Asian and Pacific Islander populations have somewhat higher average concentrations of PFOA
than the total population served across all demographic groups for small PWSs, despite the fact
that these populations have a 0.6 percentage point lower exposure rate in comparison to the total
population served across small PWSs.
Proposed PFAS Rule Economic Analysis
8-46
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-13: Baseline Scenario: Demographic Breakdown of Population Served by Category 1 and 2 PWS Service Areas
Above Baseline Thresholds and as a Percent of Total Population Served, Large Systems
Race and Ethnicity Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Total
Population
Served
System
Count
Population
Served Above
PFOS
148,105
1,956,415
3,459,772
6,608,365
12,885,593
7,863,869
17,475,594
25,339,463
306
PFHxS
114,411
1,491,117
2,449,433
4,816,112
7,984,226
5,390,515
11,637,142
17,027,657
200
Baseline
PFHpA
140,992
1,696,341
3,281,454
6,326,676
12,024,400
7,535,110
16,190,763
23,725,873
297
Threshold
PFOA
162,058
2,274,278
4,001,945
7,341,699
16,040,854
9,308,772
20,878,534
30,187,306
411
Population
PFOS
10.14%
12.21%
10.10%
14.30%
9.59%
10.40%
10.87%
10.72%
-
Served Above
Baseline
Threshold as a
PFHxS
7.84%
9.31%
7.15%
10.43%
5.95%
7.13%
7.24%
7.21%
-
PFHpA
9.66%
10.59%
9.58%
13.70%
8.95%
9.96%
10.08%
10.04%
-
Percentage of
Total
Population
Served
PFOA
11.10%
14.20%
11.68%
15.89%
11.94%
12.31%
12.99%
12.77%
-
Total Population Served
in Sampled Population
1,460,058
16,019,564
34,265,272
134,300,144
134,300,144
75,618,752
160,696,528
236,315,280
-
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - periluorohexanesulfonic acid; PFOA - periluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Proposed PFAS Rule Economic Analysis
8-47
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-14: Baseline Scenario: Demographic Breakdown of Population Served by Category 1 and 2 PWS Service Areas
Above Baseline Thresholds and as a Percent of Total Population Served, Small Systems
Race and Ethnicity
Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Total
Population
Served
System
Count
Population
PFOS
276
1,639
9,741
3,756
55,942
16,038
54,823
70,861
15
Served
Above
Baseline
Threshold
PFHxS
241
978
5,241
1,730
40,203
10,084
38,021
48,105
9
PFHpA
PFOA
367
1,502
1,084
1,924
13,685
17,406
3,371
4,699
49,166
84,397
18,085
32,909
49,095
77,481
67,180
110,390
12
21
Population
PFOS
0.47%
2.41%
3.06%
1.00%
2.36%
1.46%
2.56%
2.19%
-
Served
Above
PFHxS
PFHpA
0.41%
0.63%
1.44%
1.59%
1.65%
4.30%
0.46%
0.90%
1.69%
2.07%
0.92%
1.65%
1.77%
2.29%
1.48%
2.07%
-
Threshold
2.58%
2.83%
5.47%
1.25%
3.56%
3.00%
3.61%
3.40%
as a
Percentage pF0A
of Total
Population
Served
Total Population
Served in Sampled
Population
58,311
68,007
317,990
376,431
2,372,983
1,098,134
2,144,171
3,242,305
Abbreviations: PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic
acid.
Proposed PFAS Rule Economic Analysis
8-48
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-15: Modeled Average PFAS Concentrations (ppt) by Demographic Group and System Size in the Baseline, Category
1 and 2 PWS Service Areas
Race and Ethnicity
Income
Total
Population
Served
PFAS
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
People of
Color3
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Large Systems
PFOS
0.75
0.77
0.71
0.92
0.81
0.62
0.73
0.69
0.70
PFHxS
0.61
0.53
0.52
0.63
0.57
0.44
0.52
0.49
0.50
PFHpA
0.40
0.40
0.44
0.51
0.46
0.39
0.43
0.42
0.42
PFOA
0.81
0.94
0.89
1.03
0.96
0.83
0.88
0.89
0.89
Small Systems
PFOS
0.25
0.34
0.44
0.27
0.33
0.40
0.34
0.40
0.38
PFHxS
0.13
0.15
0.16
0.13
0.14
0.18
0.15
0.18
0.17
PFHpA
0.12
0.14
0.15
0.12
0.13
0.14
0.13
0.14
0.14
PFOA
0.30
0.39
0.41
0.32
0.35
0.38
0.34
0.39
0.37
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - periluorohexanesulfonic acid; PFOA - periluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-49
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.3.3 Hypothetical Regulatory Scenario #1: UCMR5 MRLs
Table 8-16, Table 8-17, and Table 8-18 summarize the results for populations served by large
and small category 1 and 2 PWS service areas with PFAS occurrence above UCMR 5 MRL
values, respectively. EPA assumed that PWS service areas with PFAS system-level means above
the UCMR 5 MRL value will reduce PFAS levels to comply with the proposed rule.
The first set of rows in Table 8-16 and Table 8-17 summarize populations served by large and
small category 1 and 2 PWSs with modeled PFAS occurrence above the UCMR 5 MRLs,
respectively. The second set of rows provides these estimates as a percentage of the total
population served by PWS service areas included in EPA's analysis. In Table 8-16 and Table
8-17, percentages are bolded and italicized when the percentage of the population in a specific
demographic group with PFAS occurrence above the MRL value is greater than the percentage
of the total population across all demographic groups with PFAS occurrence above the MRL.
Additionally, percentages are highlighted when the percentage of the population in a specific
demographic group with modeled PFAS above the MRL represents greater than a 1 percentage
point difference compared to the total population across all demographic groups exposed to
modeled PFAS above the MRL value. Table 8-17 characterizes population-weighted average
reductions in PFAS concentrations across demographic groups in large and small category 1 and
2 PWSs. Highlighted cells represent whether the average reduction for a given demographic
group is higher than the average reduction for the total population served across all demographic
groups (right-hand column). The percentages that are bolded, italicized, or highlighted indicate
more PFAS exposure above the MRL for a given demographic group; EPA anticipates relatively
higher reductions in PFAS exposure will accrue to these demographic groups under this
hypothetical regulatory scenario compared to the percentage of the population across all
demographic groups.
Depending on the PFAS analyte, between 2.8 percent and 4.9 percent of the total population
served by large category 1 and 2 PWS service areas are exposed to at least one of the modeled
four PFAS occurrence above UCMR 5 MRL values. For small category 1 and 2 PWS service
areas, depending on the PFAS analyte, between 0.3 percent and 1.2 percent of the total
population served is exposed to modeled PFAS occurrence above UCMR 5 MRL values.
Findings for large systems are as follows:
• American Indian or Alaska Native populations served have higher exposure above
UCMR 5 MRL values for PFOS, PFHpA, and PFHxS compared to the percent of the
population served across all demographic groups. American Indian or Alaska Native
populations also have higher PFOA exposure than the non-Hispanic White population.
• Asian and Pacific Islander populations served have higher exposure above UCMR 5
MRL values for PFOS and PFOA compared to the percent of the population served
across all demographic groups. These populations also have higher exposure for PFHpA
and PFHxS in comparison to the non-Hispanic White population.
• Black populations served have higher exposure above the UCMR 5 MRL for PFHpA and
PFOA compared to the percent of the population served across all demographic groups.
Black populations also have higher exposures for PFOA, PFHxS, and PFHpA compared
to the non-Hispanic White population.
Proposed PFAS Rule Economic Analysis
8-50
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Hispanic populations served have higher exposure above the UCMR 5 MRL values for
all four PFAS analytes compared to the percent of the population served across all
demographic groups. Hispanic populations have at least double the exposure above
UCMR 5 MRL values in comparison to non-Hispanic White populations across all four
PFAS.
Findings for small systems are as follows:
• Asian and Pacific Islander populations served have higher exposure above UCMR 5
MRL values for PFHpA and PFOA compared to the percent of the population served
across all demographic groups.
• Black populations served have higher exposure above the UCMR 5 MRL for PFOS
compared to the percent of the population served across all demographic groups.
• Non-Hispanic White populations served have higher exposure above the UCMR 5 MRL
values across all four PFAS analytes compared to the percent of the population served
across all demographic groups.
• Populations with income above twice the poverty level have higher exposure above the
UCMR 5 MRL values for PFOS, PFHpA, and PFOA compared to the percent of the
population served across all demographic groups.
Table 8-18 characterizes population-weighted average reductions in PFAS exposures anticipated
to occur for large and small PWSs in a hypothetical regulatory scenario where system-level
means are reduced to UCMR 5 MRLs. As in Table 8-16, Hispanic populations have the greatest
exposures above UCMR 5 MRLs of any demographic group among large PWSs. The results also
show that Black populations have higher average exposures to PFOS and PFHxS than the total
population served across all demographic groups in large PWSs, although a lower percentage of
this population experiences exposure to PFAS above MRL levels in large PWSs. Collectively,
people of color served by large PWSs see larger reductions in exposure across all four PFAS in
this hypothetical regulatory scenario than the total population served across all demographic
groups. For small systems, Black and non-Hispanic White populations have larger reductions in
PFOS than the total population served across all demographic groups, and non-Hispanic White
populations also see somewhat larger reductions in PFOA. In general, however, differences in
PFAS reductions across demographic groups are slight for small systems.
It should be noted that the sample size of small PWS service areas included in categories 1 and 2
with PFAS exposure above UCMR 5 MRL values is limited and could meaningfully impact the
results presented herein. The population served by small category 1 and 2 PWS service areas
included in this analysis captures roughly 1 percent of the total U.S. population. Given that
approximately 20 percent of the U.S. population is served by small systems, this subset of
systems may not be representative of small systems across the U.S., and results from this
analysis cannot be extrapolated to be representative of small systems nationwide. Additionally,
the population served by the subset of small systems in categories 1 and 2 is disproportionately
non-Hispanic White, with 13.09 higher percentage point representation compared to the overall
U.S. population. The population served is also less Hispanic, with representation of this group
being 6.59 percentage points lower than the overall U.S. population. Further evaluation is needed
to demonstrate whether the sample population served by small category 1 and 2 PWS service
areas is representative of the demographic breakdown of all small systems nationwide.
Proposed PFAS Rule Economic Analysis
8-51
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 8-16: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above UCMR 5 MRLs and as a Percent of Total Population Served, Large Systems
Race and Ethnicity Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above Twice
the Poverty
Level
Total
Population
Served
System
Count
PFOS
82,776
860,437
1,581,076
3,610,906
4,867,952
3,731,804
7,352,841
11,084,645
306
Population
Served Above
UCMR 5 MRL
PFHxS
PFHpA
81,853
43,053
656,474
386,554
1,419,053
1,075,433
3,161,347
2,308,921
4,481,038
2,640,787
3,264,982
2,274,284
6,605,550
4,221,951
9,870,532
6,496,235
200
297
PFOA
59,931
866,578
1,809,639
3,498,968
5,225,904
3,734,032
7,827,235
11,561,267
411
Population
Served Above
UCMR 5 MRL as
PFOS
5.67%
5.37%
4.61%
7.82%
3.62%
4.94%
4.58%
4.69%
-
PFHxS
5.61%
4.10%
4.14%
6.84%
3.34%
4.32%
4.11%
4.18%
-
a Percentage of
Total Population
Served
PFHpA
PFOA
2.95%
4.10%
2.41%
5.41%
3.14%
5.28%
5.00%
7.57%
1.97%
3.89%
3.01%
4.94%
2.63%
4.87%
2.75%
4.89%
-
Total Population Served in
Sampled Population
1,460,058
16,019,564
34,265,272
46,196,824
134,300,144
75,618,752
160,696,528
236,315,280
-
Abbreviations: MRL - minimum reporting level; PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS -
perfluorooctanesulfonic acid; UCMR - Unregulated Contaminant Monitoring Rule.
Proposed PFAS Rule Economic Analysis
8-52
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-17: Hypothetical Regulatory Scenario #1: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above UCMR 5 MRLs and as a Percent of Total Population Served, Small Systems
Race and Ethnicity Income
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below Twice
the Poverty
Level
Above Twice
the Poverty
Level
lotai
Population
Served
System
Count
PFOS
221
722
4,341
1,494
32,911
7,708
31,389
39,097
15
Population
Served Above
UCMR 5 MRL
PFHxS
PFHpA
128
11
65
256
113
116
181
282
11,086
8,084
3,504
889
8,222
7,996
11,726
8,885
9
12
PFOA
93
676
346
1,290
19,517
2,974
18,981
21,955
21
Population
Served Above
UCMR 5 MRL
PFOS
0.38%
1.06%
1.37%
0.40%
1.39%
0.70%
1.46%
1.21%
-
PFHxS
0.22%
0.10%
0.04%
0.05%
0.47%
0.32%
0.38%
0.36%
-
as a Percentage
of Total
PFHpA
0.02%
0.38%
0.04%
0.07%
0.34%
0.08%
0.37%
0.27%
-
Population
Served
PFOA
0.16%
0.99%
0.11%
0.34%
0.82%
0.27%
0.89%
0.68%
-
Total Population Served in
Sampled Population
58,311
68,007
317,990
376,431
2,372,983
1,098,134
2,144,171
3,242,305
-
Abbreviations: MRL - minimum reporting level; PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS -
perfluorooctanesulfonic acid; UCMR - Unregulated Contaminant Monitoring Rule.
Proposed PFAS Rule Economic Analysis
8-53
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-18: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a Hypothetical Regulatory
Scenario with Maximum Contaminant Levels at the UCMR 5 MRLs, Category 1 and 2 PWS Service Areas
Race and Ethnicity Income
PFAS
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
People of
Color3
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Total
Population
Served
Large Systems
PFOS
0.27
0.25
0.25
0.33
0.28
0.18
0.26
0.21
0.22
PFHxS
0.25
0.16
0.18
0.20
0.19
0.13
0.18
0.14
0.16
PFHpA
0.04
0.03
0.06
0.07
0.06
0.04
0.06
0.05
0.05
PFOA
0.16
0.22
0.19
0.25
0.22
0.16
0.19
0.18
0.19
Small Systems
PFOS
0.03
0.06
0.13
0.05
0.07
0.12
0.09
0.11
0.10
PFHxS
0.02
0.02
0.01
0.03
0.02
0.04
0.03
0.04
0.04
PFHpA
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
PFOA
0.00
0.05
0.00
0.02
0.01
0.04
0.01
0.04
0.03
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-54
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.3.2.3.4 Hypothetical Regulatory Scenario #2: 10.0 ppt
Table 8-19 summarizes results for populations served by large category 1 and 2 PWS service
areas with PFAS occurrence above 10.0 ppt. Table 8-20 summarizes results for populations
served by small category 1 and 2 PWS service areas with PFAS occurrence above 10.0 ppt.
Table 8-21 characterizes population-weighted average reductions in PFAS exposures anticipated
to occur for large and small PWSs in a hypothetical regulatory scenario where system-level
means are reduced to 10.0 ppt. The first set of rows in Table 8-19 and Table 8-20 summarizes
populations served by large and small category 1 and 2 PWSs with modeled PFAS occurrence
above the 10.0 ppt, respectively. The second set of rows provides these estimates as a percentage
of the total population served by these PWS service areas.
In Table 8-19 and Table 8-20, percentages are bolded and italicized when the percent of the
population in a specific demographic group with PFAS occurrence above 10.0 ppt is greater than
the percentage of the total population across all demographic groups with PFAS occurrence
above 10.0 ppt. Table 8-21, highlighted cells represent whether the average reduction for a given
demographic group is higher than the reductions for the total population served across all
demographic groups in large and small PWSs (right-hand column). The percentages that are
bolded, italicized, or highlighted indicate more PFAS exposure above 10.0 ppt for a given
demographic group; EPA anticipates relatively higher reductions in PFAS exposure will accrue
to these demographic groups under this hypothetical regulatory scenario compared to the
percentage of the population across all demographic groups.
For large systems, American Indian or Alaska Native, Asian and Pacific Islander, Black, and
Hispanic populations as well as populations with income below twice the poverty level have
elevated exposure above 10.0 ppt for particular PFAS analytes compared to the population
served across all demographic groups.
For small systems, Asian and Pacific Islander, Black, and non-Hispanic White populations have
elevated exposure above 10.0 ppt for particular PFAS analytes compared to the population
served across all demographic groups. The average reductions by demographic group, shown in
Table 8-21, largely confirm the findings of Table 8-19 and Table 8-20, with greater average
reductions generally accruing to populations with a higher percentage of potentially exposed
individuals for large and small PWSs.
As previously noted, the sample size of small PWS service areas included in categories 1 and 2 is
limited, with population served capturing roughly 1 percent of the total U.S. population.
Additionally, the population served by the subset of small systems in categories 1 and 2 is
disproportionately White and non-Hispanic compared to the overall U.S. population, as
previously discussed. Further evaluation is needed to demonstrate whether the sample population
served by small category 1 and 2 PWS service areas is representative of the demographic
breakdown of all small systems nationwide.
Proposed PFAS Rule Economic Analysis
8-55
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-19: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above 10.0 ppt and as a Percent of Total Population Served, Large Systems
Race and Ethnicity Income
American
Indian or
Alaska
Asian and
Pacific
Islander
Black
Hispanic
Non-
Hispanic
White
Below
Twice the
Poverty
Above
Twice the
Poverty
Total
Population
Served
System
Count
Native
Level
Level
Population
PFOS
18,161
189,206
427,433
602,112
1,258,751
925,193
1,609,362
2,534,555
42
Served
PFHxS
14,774
70,599
133,793
174,106
556,601
355,138
614,293
969,431
25
Above 10.0
PFHpA
2,534
10,830
82,127
52,010
220,399
131,069
244,365
375,434
7
ppt
PFOA
7,491
138,790
230,052
213,388
787,930
436,150
964,752
1,400,902
34
Population
PFOS
1.24%
1.18%
1.25%
1.30%
0.94%
1.22%
1.00%
1.07%
-
Served
PFHxS
1.01%
0.44%
0.39%
0.38%
0.41%
0.47%
0.38%
0.41%
-
Above 10.0
PFHpA
0.17%
0.07%
0.24%
0.11%
0.16%
0.17%
0.15%
0.16%
-
ppt as a
Percentage
of Total
PFOA
0.51%
0.87%
0.67%
0.46%
0.59%
0.58%
0.60%
0.59%
-
Population
Served
Total Population
Served in Sampled
1,460,058
16,019,564
34,:265,:272
46,196,824
134,300,144
75,618,752
160,696,528
236,315,280
-
Population
Abbreviations: PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; ppt - parts per
trillion.
Proposed PFAS Rule Economic Analysis
8-56
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-20: Hypothetical Regulatory Scenario #2: Demographic Breakdown of Population Served by Category 1 and 2 PWS
Service Areas Above 10.0 ppt and as a Percent of Total Population Served, Small Systems
Race and Ethnicity
Income
Total
System
Count
American
Indian or
Alaska Native
Asian and
Pacific
Islander
Black
Hispanic
Non-Hispanic
White
Below Twice
the Poverty
Level
Above Twice
the Poverty
Level
Population
Served
Population
PFOS
134
65
4,043
192
16,673
6,219
14,307
20,526
4
Served
Above 10.0
ppt
PFHxS
PFHpA
123
1
39
7
74
11
136
48
7,513
155
2,058
81
5,955
149
8,013
230
2
1
PFOA
11
256
116
282
8,084
889
7,996
8,885
2
Population
PFOS
0.23%
0.10%
1.27%
0.05%
0.70%
0.57%
0.67%
0.63%
-
Served
Above 10.0
ppt as a
PFHxS
PFHpA
0.21%
0.00%
0.06%
0.01%
0.02%
0.00%
0.04%
0.01%
0.32%
0.01%
0.19%
0.01%
0.28%
0.01%
0.25%
0.01%
-
Percentage
of Total
Population
Served
PFOA
0.02%
0.38%
0.04%
0.07%
0.34%
0.08%
0.37%
0.27%
-
Total Population
Served in Sampled
58,311
68,007
317,990
376,431
2,372,983
1,098,134
2,144,171
3,242,305
Population
Abbreviations: PFHpA - perfluoroheptanoic acid; PFHxS - perfluorohexanesulfonic acid; PFOA - perfluorooctanoic acid; PFOS - perfluorooctanesulfonic acid; ppt - parts per
trillion.
Proposed PFAS Rule Economic Analysis
8-57
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-21: Reductions in Average PFAS Concentrations (ppt) by Demographic Group in a Hypothetical Regulatory
Scenario with Maximum Contaminant Levels at 10.0 ppt, Category 1 and 2 PWS Service Areas
Race and Ethnicity Income
PFAS
American
Indian or
Alaska
Native
Asian and
Pacific
Islander
Black
Hispanic
People of
Color8
Non-
Hispanic
White
Below
Twice the
Poverty
Level
Above
Twice the
Poverty
Level
Total
Population
Served
Large Systems
PFOS
0.10
0.11
0.11
0.12
0.11
0.08
0.12
0.08
0.09
PFHxS
0.07
0.07
0.08
0.07
0.07
0.05
0.08
0.05
0.06
PFHpA
0.00
0.00
0.01
0.00
0.00
0.01
0.01
0.00
0.00
PFOA
0.04
0.04
0.06
0.04
0.05
0.05
0.05
0.04
0.05
Small Systems
PFOS
0.01
0.03
0.05
0.03
0.04
0.06
0.05
0.05
0.05
PFHxS
0.01
0.02
0.01
0.02
0.02
0.02
0.02
0.02
0.02
PFHpA
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
PFOA
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.01
0.00
Abbreviations: PFHpA - periluoroheptanoic acid; PFHxS - periluorohexanesulfonic acid; PFOA - periluorooctanoic acid; PFOS - perfluorooctanesulfonic acid.
Note:
aThe demographic group people of color includes individuals who identify as Hispanic and/or a race other than White. It is calculated from E.TScreen's percent minority
indicator and is non-duplicative across race and ethnicity categories.
Proposed PFAS Rule Economic Analysis
8-58
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.4 SafeWater EJ Analysis of Proposed Regulatory Option and
Alternatives
8.4.1 Methodology
In addition to analyzing EJ exposure using the EJSCREENbatch R package, EPA also conducted
an EJ analysis of the proposed regulatory option and regulatory alternatives using the SafeWater
Multi-Contaminant Benefit-Cost Model (MCBC). EPA's proposed option sets MCLs of 4.0 ppt
for PFOA and PFOS and an HI of 1.0 for PFNA, HFPO-DA, PFHxS, and PFBS. Options la, lb,
and lc set MCL values for PFOA and PFOS at 4.0 ppt, 5.0 ppt, and 10.0 ppt, respectively.
SafeWater MCBC was used to analyze the distribution of anticipated health benefits and
household costs associated with the proposed PFAS NPDWR across race/ethnicity groups. For
more information on SafeWater MCMC and its application in EPA's analysis of national
quantified benefits and costs associated with the proposed PFAS NPDWR, see Section 5.2.
Using SafeWater MCBC, EPA estimated the quantified health benefits and household costs
expected to accrue to specific race/ethnicity groups for category 1 and 2 PWS service areas. As
previously described in Section 8.3.1, category 1 and 2 PWS service areas include systems that
have sampled PFAS occurrence data from UCMR 3 and have predelineated service area
boundaries or those estimated using zip code served information (n=4,723). The subset of
category 1 and 2 PWSs captured in the analysis represents roughly 3 percent of active PWSs.101
Results are presented across four race/ethnicity groups, consistent with the subpopulation
definitions used to estimate the national quantified benefits for the proposed PFAS NPDWR (see
Section 8.1). These race/ethnicity groups include: non-Hispanic Black, Hispanic, non-Hispanic
White, and Other.102 Race/ethnicity categories examined in EPA's analysis using SafeWater
MCBC differ from the demographic groups presented in the exposure analysis discussed
previously in this chapter due to the availability of demographic information utilized in EPA's
quantified benefits analysis. For more information on the selection of data inputs to EPA's
benefit analysis, see Chapter 6.
The total sample population captured by EPA's analysis using SafeWater MCBC is roughly 196
million people, with a breakdown by race/ethnicity group as follows:
• Non-Hispanic Black: 25.1 million (-13%)
• Hispanic: 32.6 million (-17%)
• Other: 12.2 million (-6%)
• Non-Hispanic White: 125.9 million (-64%)
When compared to the breakdown of the total U.S. population by these same race/ethnicity
groups, the makeup of the sample population in EPA's analysis is generally representative of the
101 The number of active PWSs was retrieved from SDWIS/Fed fourth quarter 2021.
102 The "Other" race/ethnicity category includes any race/ethnicity populations that are not non-Hispanic Black, Hispanic, or non-
Hispanic White.
Proposed PFAS Rule Economic Analysis
8-59
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
overall U.S. population. Non-Hispanic Black, Hispanic, and Other race/ethnicity groups (making
up -13 percent, -19 percent, and -8 percent of the U.S. population, respectively) are slightly
underrepresented, while the non-Hispanic White race/ethnicity group (making up -60% of the
U.S. population) is slightly overrepresented in EPA's analysis (U.S. Census Bureau, 2020a).
Because demographic proportion information utilized in EPA's benefits analysis was available at
the county level, EPA utilized the following step-by-step approach to identify the number of
people in each race/ethnicity group within a given PWS service area. Specifically, in this order,
EPA utilized the following stepwise approach:
1. Overlayed census block groups with PWS service area boundaries;
2. Calculated the area of each census block group and PWS service area boundary;
3. Calculated the percent of each census block group overlapping each PWS service area
boundary;
4. Multiplied the population of the census block group by the percent of each census block
overlapping each PWS service area boundary;
5. Summed across census block groups to calculate the population in each PWS service area
boundary that lives in each county;
6. Calculated the percent of the population in each county (PWS county weight) for each
PWS; and
7. Estimated the number of people served by a PWS for each subpopulation as follows:
SubPop = number of people in each subpopulation served by a PWS
PWS County weightc = the percentage of the PWS population in each county (c)
PWS Pop = Number of people served by PWS from SDWIS/Fed inventory
Subpop_sharec = The share of county (c) population consisting of the subpopulation from the
U.S. Census
As part of its national analysis of quantified benefits and costs using SafeWater MCBC, EPA
accounted for states that have enacted enforceable MCLs for PFAS contaminants. For these
states, EPA assumed that the state MCL is the maximum baseline PFAS occurrence value for all
entry points in the state. For more information on this assumption and on state-enacted MCLs,
see Section 4. EPA has applied this assumption as part of its EJ analysis conducted in SafeWater
Equation 24:
SubPop = PWS_county_weightc x PWS_Pop x Subpop_sharec
C
Where:
MCBC.
Proposed PFAS Rule Economic Analysis
8-60
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.4.2 SafeWater EJ Analysis Results
8.4.2.1 Health Benefits
To determine if there are disproportionate health impacts borne by any race/ethnicity
subpopulation under the proposed regulatory option or regulatory alternatives, EPA estimated the
annual avoided cases of mortality and morbidity per 100,000 people, as shown in Table 8-22
through Table 8-25.
For the analysis conducted in SafeWater MCBC, EPA reports the estimated avoided cases of
mortality and morbidity by race/ethnicity group for the following health endpoints:
• Cardiovascular disease (CVD): Non-fatal myocardial infarction (MI), non-fatal ischemic
stroke (IS), CVD deaths
• Renal cell carcinoma (RCC): Non-fatal RCC cases avoided, fatal RCC cases avoided
• Birth weight: Birth weight gain (total grams), birth weight-related deaths avoided
Baseline incidence associated with these health endpoints varies by race/ethnicity, and disparities
in underlying incidence by race/ethnicity likely influence the distribution of quantified health
benefits expected under the proposed PFAS NPDWR. For example, non-fatal MI incidence is
generally most prevalent among non-Hispanic White males, while non-fatal IS incidence is
generally most prevalent among non-Hispanic Black males. The demographic distribution of
quantified health benefits presented here incorporates differing prevalence in baseline health
outcomes by race/ethnicity. As such, the demographic distribution of quantified health benefits
that EPA reports here have not been adjusted for underlying disparities in death or disease
incidence by race/ethnicity and therefore provides a comprehensive evaluation of quantified
benefits across race/ethnicity groups. For a detailed breakdown of incidence associated with the
effects of reduced birth weight on infant mortality, CVD events, and RCC by race/ethnicity, see
Appendices E, G, and J, respectively.
EPA did not analyze the demographic breakdown of bladder cancer cases avoided that are
expected to result from the co-removal of PFAS and DBP precursors (discussed in Section 6.7).
EPA models bladder cancer impacts based on a national-level distribution of finished water TOC
levels; because specific TOC levels at actual PWSs are not available, EPA did not include these
impacts in the portion of its EJ analysis conducted in SafeWater MCBC.
Table 8-22 summarizes the number of avoided cases of morbidity and mortality per 100,000
people per year for all health endpoints evaluated under EPA's proposed regulatory option. Table
8-23 through Table 8-25 summarize the number of avoided cases of morbidity and mortality per
100,000 people per year for all health endpoints evaluated under EPA's regulatory alternatives.
Across the proposed option and all regulatory alternatives, benefits are anticipated to be realized
across all health endpoints and race/ethnicity groups evaluated. A summary of benefits
anticipated for each health endpoint is included below. In general, when comparing benefits
under the proposed option to those across regulatory alternatives, the distribution of quantified
health benefits for a given race/ethnicity group is relatively similar. Variation exists between the
proposed option and regulatory alternatives with respect to the total amount of health benefits
anticipated. Additionally, across all health endpoints evaluated and across all race/ethnicity
groups, the greatest benefits are anticipated under the proposed option.
Proposed PFAS Rule Economic Analysis
8-61
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Below is a summary of quantified health benefits categorized by endpoint, with results presented
across the proposed option and regulatory alternatives and across race/ethnicity groups.
Cardiovascular Disease
Non-Fatal MI Cases Avoided- Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 0.84 to 3.59 cases avoided per 100,000 people per year.
Under the proposed option and all alternatives, EPA anticipates the greatest benefit to accrue to
Other race/ethnicity groups and the lowest benefit to accrue to the non-Hispanic Black
race/ethnicity group.
Non-Fatal IS Cases Avoided - Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 1.73 to 5.97 cases avoided per 100,000 people per year.
Under the proposed option and all alternatives, EPA anticipates the greatest benefit to accrue to
the non-Hispanic Black race/ethnicity group and the lowest benefit to accrue to the non-Hispanic
White race/ethnicity group.103
CVD Deaths Avoided -Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 0.51 to 3.10 deaths avoided per 100,000 people per
year. Under the proposed option and all alternatives, EPA anticipates the greatest benefit to
accrue to the non-Hispanic Black race/ethnicity group and the lowest benefit to accrue to the
Hispanic race/ethnicity group.
Renal Cell Carcinoma
Non-Fatal RCC Cases Avoided - Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 1.01 to 3.16 cases avoided per 100,000 people per year.
Under the proposed option and all alternatives, EPA anticipates the greatest benefit to accrue to
Other race/ethnicity groups and the lowest benefit to accrue to the non-Hispanic Black
race/ethnicity group.
Fatal RCC Cases Avoided - Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 0.27 to 0.97 deaths avoided per 100,000 people per
year. Under the proposed option and all alternatives, EPA expects the greatest benefit to accrue
to the Hispanic race/ethnicity group and the lowest benefit to accrue to the non-Hispanic White
race/ethnicity group.
Birth Weight
Birth Weight Gain (total grams) - Under the proposed option and all alternatives and across all
race/ethnicity groups, values range from 34,024 grams to 115,689 grams of birth weight gain per
100,000 people per year. Under the proposed option and all alternatives, EPA expects the largest
benefit to accrue to the Hispanic race/ethnicity group and the lowest benefit to accrue to the non-
Hispanic White race/ethnicity group.
103 The non-Hispanic White race/ethnicity group is anticipated to experience the lowest benefit related to non-fatal IS cases
avoided, except under Option lc where both non-Hispanic White and Hispanic race/ethnicity groups are anticipated to experience
the lowest benefit (i.e., 1.73 cases avoided per 100,000 people).
Proposed PFAS Rule Economic Analysis
8-62
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Birth Weight-Related Deaths Avoided -Under the proposed option and all alternatives and
across all race/ethnicity groups, values range from 0.19 to 0.75 birth weight-related deaths
avoided per 100,000 people per year. Under the proposed option and all alternatives, EPA
anticipates the greatest benefit to accrue to the non-Hispanic Black race/ethnicity group and the
lowest benefit to accrue to the non-Hispanic White race/ethnicity group.
Table 8-22: Annualized Cases Avoided per 100,000 People by Race/Ethnicity
Group, Proposed Option (PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0)
Non-
Non-
Health Endpoint
Hispanic
Hispanic
Other
Hispanic
Black
White
Non-Fatal MI Cases Avoided
1.86
2.60
3.59
2.78
Non-Fatal IS Cases Avoided
5.97
3.67
3.95
3.62
CVD Deaths Avoided
3.10
1.08
1.32
1.21
Non-Fatal RCC Cases Avoided
2.61
2.78
3.16
2.72
Fatal RCC Cases Avoided
0.76
0.97
0.85
0.70
Birth Weight Gain (total grams)
92,441
115,689
105,872
67,668
Birth Weight-Related Deaths Avoided
0.75
0.64
0.47
0.38
Abbreviations: CVD - cardiovascular disease; MI ¦
- myocardial infarction; IS - ischemic stroke; RCC ¦
- renal cell
carcinoma.
Table 8-23: Annualized Cases Avoided per 100,000 People by Race/Ethnicity
Group, Option la (PFOA and PFOS MCLs of 4.0 ppt)
Non-
Non-
Health Endpoint
Hispanic
Hispanic
Other
Hispanic
Black
White
Non-Fatal MI Cases Avoided
1.83
2.56
3.54
2.74
Non-Fatal IS Cases Avoided
5.86
3.61
3.89
3.56
CVD Deaths Avoided
3.05
1.06
1.30
1.19
Non-Fatal RCC Cases Avoided
2.56
2.73
3.10
2.67
Fatal RCC Cases Avoided
0.75
0.95
0.84
0.68
Birth Weight Gain (total grams)
90,753
113,827
104,297
66,562
Birth Weight-Related Deaths Avoided
0.74
0.63
0.46
0.38
Abbreviations: CVD - cardiovascular disease; MI ¦
- myocardial infarction; IS - ischemic stroke; RCC ¦
- renal cell
carcinoma.
Proposed PFAS Rule Economic Analysis
8-63
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-24: Annualized Cases Avoided per 100,000 People by Race/Ethnicity
Group, Option lb (PFOA and PFOS MCLs of 5.0 ppt)
Non-
Non-
Health Endpoint
Hispanic
Hispanic
Other
Hispanic
Black
White
Non-Fatal MI Cases Avoided
1.58
2.22
3.08
2.37
Non-Fatal IS Cases Avoided
5.05
3.13
3.39
3.08
CVD Deaths Avoided
2.62
0.92
1.13
1.03
Non-Fatal RCC Cases Avoided
2.14
2.30
2.63
2.22
Fatal RCC Cases Avoided
0.62
0.80
0.71
0.57
Birth Weight Gain (total grams)
78,860
99,954
91,914
58,186
Birth Weight-Related Deaths Avoided
0.64
0.55
0.40
0.33
Abbreviations: CVD - cardiovascular disease; MI - myocardial infarction; IS - ischemic stroke; RCC - renal cell
carcinoma.
Table 8-25: Annualized Cases Avoided per 100,000 People by Race/Ethnicity
Group, Option lc (PFOA and PFOS MCLs of 10.0 ppt)
Non-
Non-
Health Endpoint
Hispanic
Hispanic
Other
Hispanic
Black
White
Non-Fatal MI Cases Avoided
0.84
1.23
1.70
1.33
Non-Fatal IS Cases Avoided
2.70
1.73
1.87
1.73
CVD Deaths Avoided
1.40
0.51
0.62
0.58
Non-Fatal RCC Cases Avoided
1.01
1.15
1.35
1.05
Fatal RCC Cases Avoided
0.29
0.40
0.37
0.27
Birth Weight Gain (total grams)
44,270
58,434
53,923
34,024
Birth Weight-Related Deaths Avoided
0.36
0.32
0.24
0.19
Abbreviations: CVD - cardiovascular disease; MI - myocardial infarction; IS - ischemic stroke; RCC - renal cell
carcinoma.
Proposed PFAS Rule Economic Analysis
8-64
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.4.2.2 Household Costs
For category 1 and 2 PWS service areas, EPA used SafeWater MCBC to estimate the distribution
of average annual incremental household costs across race/ethnicity groups. The results are
provided by system size category in Table 8-26 and Table 8-27. In addition to presenting average
incremental household costs for each race/ethnicity group, EPA also presents household costs
across "All" race/ethnicity groups to provide a basis for comparison.
In estimating annualized incremental household costs of the proposed PFAS NPDWR,
SafeWater MCBC first divided each PWS's total compliance costs by the PWS's average daily
flow to determine the cost of compliance per 1,000 gallons of daily flow. Next, this cost was
multiplied by the average household consumption from the Community Water System Survey
(CWSS) to calculate the average household cost of compliance for the PWS. To calculate the
average household cost for each race/ethnicity group by PWS system size strata, for each PWS
included in the subset of systems in EPA's EJ analysis, EPA calculated a weighted average
household cost by using the number of people in each race/ethnicity group served by each PWS
as the weight. In addition to estimating the demographic breakdown of annualized incremental
household costs of the proposed PFAS NPDWR for all systems included in EPA's EJ analysis,
EPA also estimated the demographic breakdown of annualized incremental household costs for
just the subset of PWSs that are anticipated to install treatment to comply with the rule.104
Below is a summary of the demographic distribution of incremental household costs, categorized
by system size, for the proposed option and regulatory alternatives. Results are presented both
for the entire subset of PWSs included in EPA's EJ analysis and just those anticipated to install
treatment under the rule. Note that an analysis of household costs served by systems serving
fewer than 3,300 people could not be completed due to limited sample size. Except in one case,
the proposed option is anticipated to have the largest associated costs, and Option lc is
anticipated to have the lowest associated costs.
8.4.2.2.1lncremental Household Costs for All PWSs
System size 3,300 to 10,000 - Average incremental household costs range from $5.20 to $17.79
per year across the proposed option and regulatory alternatives and across race/ethnicity groups.
When comparing household costs borne by particular race/ethnicity groups to those borne by the
overall population served by systems in this size category, the non-Hispanic Black race/ethnicity
group bears minimal elevated household costs under the proposed option and all regulatory
alternatives. Additionally, the Hispanic and Other race/ethnicity groups bear minimal elevated
household costs under the proposed option and Options la and lb. The magnitude of household
cost differences between each of these race/ethnicity groups and the overall population is small,
ranging from $0.39 to $2.53 per year across race/ethnicity groups and across the proposed option
and regulatory alternatives. The non-Hispanic Black race/ethnicity group bears the highest
household cost, while the Hispanic race/ethnicity group bears the lowest household cost.
System size 10,000 to 50,000 - Average incremental household costs range from $2.71 to $9.11
per year across the proposed option and regulatory alternatives and across race/ethnicity groups.
When comparing household costs borne by particular race/ethnicity groups to those borne by the
104 For additional detail on treatment technology selection among systems anticipated to install treatment under the proposed
rule, see Section 5.3.1.1.
Proposed PFAS Rule Economic Analysis
8-65
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
overall population served by systems in this size category, the Other race/ethnicity group bears
minimal elevated household costs under the proposed option and all regulatory alternatives. The
magnitude of these household cost differences is very small, ranging from $0.02 to $0.90 per
year across race/ethnicity groups and across the proposed option and regulatory alternatives. The
Other race/ethnicity group bears the highest household cost, while the non-Hispanic Black
race/ethnicity group bears the lowest household cost.
System size 50,000 to 100,000 - Average incremental household costs range from $1.51 to $5.97
per year across the proposed option and regulatory alternatives and across race/ethnicity groups.
When comparing household costs borne by particular race/ethnicity groups to those borne by the
overall population served by systems in this size category, the Hispanic and Other race/ethnicity
groups bear minimal elevated household costs under the proposed option and all regulatory
alternatives. The magnitude of these household cost differences is very small, ranging from
$0.21 to $0.74 per year across race/ethnicity groups and across the proposed option and
regulatory alternatives. The Other race/ethnicity group bears the highest household cost, while
the non-Hispanic Black race/ethnicity group bears the lowest household cost.
System size 100,000 to 1,000,000 - Average incremental household costs range from $2.53 to
$10.04 per year across the proposed option and regulatory alternatives and across race/ethnicity
groups. When comparing household costs borne by particular race/ethnicity groups to those
borne by the overall population served by systems in this size category, the Hispanic and Other
race/ethnicity groups bear minimal elevated household costs under the proposed option and all
regulatory alternatives. As in other system size categories, the magnitude of these household cost
differences is small, ranging from $0.53 to $1.87 per year across race/ethnicity groups and across
the proposed option and regulatory alternatives. The Hispanic race/ethnicity group bears the
highest household cost, while the non-Hispanic Black race/ethnicity group bears the lowest
household cost.
EPA's comparison of incremental household costs across system size categories reveals that, in
general, as system size increases, average incremental household costs decrease under the
proposed option and all regulatory alternatives and across all race/ethnicity groups. One
exception to this trend is among systems serving 100,000 to 1,000,000 people, where costs are
marginally higher than for systems serving 50,000 to 100,000 people.
The highest average incremental household costs under the proposed option and all regulatory
alternatives are realized for the smallest systems (i.e., systems serving 3,300 to 10,000 people).
The range of household costs within this system size category is $5.20 to $17.79 per year, and
EPA anticipates the highest cost ($17.79 per year) under the proposed option for the non-
Hispanic Black race/ethnicity group. The lowest average incremental household costs under the
proposed option and all regulatory alternatives are realized for systems serving 50,000 to
100,000 people. The range of household costs within this system size category is $1.51 to $6.48,
with the non-Hispanic Black race/ethnicity group having the lowest cost of $1.51 under Option
lc.
Comparing the magnitude of household costs anticipated across system size categories illustrates
the role that system size plays in household costs anticipated under the proposed PFAS rule. This
is an expected result due to economies of scale and the impact that a smaller customer and tax
base has on costs per household for funding and financing capital and operational infrastructure
Proposed PFAS Rule Economic Analysis
8-66
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
investments. Further, this analysis includes the estimated household costs for all systems
impacted by the rule, not just the systems expected to install and operate treatment after
exceeding the proposed MCLs. Households served by water systems triggered into treatment are
expected to face greater cost increases than those presented here. EPA presents the demographic
breakdown of estimated household costs for those systems anticipated to install treatment under
the proposed rule in Section 8.4.2.2.2. Additionally, EPA assesses the impact of treatment
technology costs specifically on small system households in the small system affordability
analysis. For more information, see EPA's assessment of small system affordability in Section
9.12.
Table 8-26: Annualized Population Weighted Household Cost by PWS Size Category and
Race/Ethnicity Group ($2021)
System Size3
Race/Ethnicity
Group
Proposed
Optionb
Option lac
Option lbd
Option lce
3,300 to 10,000
All
$15.25
$15.08
$11.93
$5.34
3,300 to 10,000
Non-Hispanic Black
$17.79
$17.61
$14.14
$6.40
3,300 to 10,000
Hispanic
$16.00
$15.83
$12.48
$5.20
3,300 to 10,000
Other
$15.88
$15.63
$12.32
$5.29
3,300 to 10,000
Non-Hispanic White
$14.75
$14.58
$11.51
$5.21
10,000 to 50,000
All
$8.20
$8.03
$6.38
$2.80
10,000 to 50,000
Non-Hispanic Black
$8.06
$7.90
$6.25
$2.71
10,000 to 50,000
Hispanic
$8.22
$8.03
$6.35
$2.77
10,000 to 50,000
Other
$9.11
$8.91
$7.15
$3.23
10,000 to 50,000
Non-Hispanic White
$8.16
$7.99
$6.35
$2.79
50,000 to 100,000
All
$5.74
$5.59
$4.31
$1.71
50,000 to 100,000
Non-Hispanic Black
$5.22
$5.10
$3.92
$1.51
50,000 to 100,000
Hispanic
$5.97
$5.80
$4.54
$1.99
50,000 to 100,000
Other
$6.48
$6.27
$4.93
$2.12
50,000 to 100,000
Non-Hispanic White
$5.69
$5.54
$4.25
$1.62
100,000 to 1,000,000
All
$8.25
$7.96
$6.41
$2.86
100,000 to 1,000,000
Non-Hispanic Black
$7.70
$7.42
$5.93
$2.53
100,000 to 1,000,000
Hispanic
$9.31
$8.96
$7.28
$3.39
100,000 to 1,000,000
Other
$10.04
$9.71
$8.08
$3.99
100,000 to 1,000,000
Non-Hispanic White
$7.84
$7.58
$6.07
$2.64
Notes:
aThe number of systems serving fewer than 3,300 people represented in the UMCR 3 occurrence data is too limited to accurately
estimate average population-weighted household costs by subpopulation. Therefore, results for these small systems are omitted.
Also, household costs in this exhibit are population-weighted and will not match average household costs by size category
shown in other exhibits in the economic analysis document that are not population-weighted.
bThe proposed option sets PFOA and PFOS MCLs of 4.0 ppt and an HI of 1.0.
cOption la sets PFOA and PFOS MCLs of 4.0 ppt.
Option lb sets PFOA and PFOS MCLs of 5.0 ppt.
eOption lc sets PFOA and PFOS MCLs of 10.0 ppt.
Proposed PFAS Rule Economic Analysis
8-67
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.4.2.2.2 Incremental Household Costs for Treating PWSs
System size 3,300 to 10,000 - Average incremental household costs for systems anticipated to
install treatment range from $91.40 to $122.25 per year across the proposed option and
regulatory alternatives and across race/ethnicity groups. When comparing household costs borne
by particular race/ethnicity groups to those borne by the overall population served by systems in
this size category, the non-Hispanic Black race/ethnicity group bears minimal elevated
household costs under the proposed option and all regulatory alternatives. Additionally, the
Hispanic race/ethnicity group bears minimal elevated household costs under the proposed option
and Option la, and the non-Hispanic White race/ethnicity group bears minimal elevated
household costs under Options lb and lc. The magnitude of household cost differences between
each of these race/ethnicity groups and the overall population ranges from $0.17 to $9.47 per
year across race/ethnicity groups and across the proposed option and regulatory alternatives. The
non-Hispanic Black race/ethnicity group bears the highest household cost, while the Hispanic
race/ethnicity group bears the lowest household cost.
System size 10,000 to 50,000 - Average incremental household costs for systems anticipated to
install treatment range from $24.58 to $33.31 per year across the proposed option and regulatory
alternatives and across race/ethnicity groups. When comparing household costs borne by
particular race/ethnicity groups to those borne by the overall population served by systems in this
size category, the non-Hispanic White race/ethnicity group bears minimal elevated household
costs under the proposed option and all regulatory alternatives. Additionally, the Other
race/ethnicity group bears minimal elevated household costs under the proposed option, and
Options la and lb and the non-Hispanic Black race/ethnicity group bears minimal elevated costs
under all regulatory alternatives (Options la-lc). The magnitude of household cost differences
between each race/ethnicity groups and the overall population is extremely small, ranging from
$0.01 to $0.43 per year across race/ethnicity groups and across the proposed option and
regulatory alternatives. The non-Hispanic White race/ethnicity group bears the highest household
cost, while the Hispanic race/ethnicity group bears the lowest household cost.
System size 50,000 to 100,000 - Average incremental household costs for systems anticipated to
install treatment range from $17.63 to $23.90 per year across the proposed option and regulatory
alternatives and across race/ethnicity groups. When comparing household costs borne by
particular race/ethnicity groups to those borne by the overall population served by systems in this
size category, the Other and non-Hispanic White race/ethnicity groups bear minimal elevated
costs under the proposed option and all regulatory alternatives. The magnitude of household cost
differences between each of these race/ethnicity groups and the overall population is very small,
ranging from $0.03 to $0.87 per year across race/ethnicity groups and across the proposed option
and regulatory alternatives. The Other race/ethnicity group bears the highest household cost,
while the Hispanic race/ethnicity group bears the lowest household cost.
System 100,000 to 1,000,000 - Average incremental household costs for systems anticipated to
install treatment range from $17.24 to $27.89 per year across the proposed option and regulatory
alternatives and across race/ethnicity groups. When comparing household costs borne by
particular race/ethnicity groups to those borne by the overall population served by systems in this
size category, the Other race/ethnicity group bears minimal elevated costs under the proposed
option and all regulatory alternatives. Additionally, the Hispanic race/ethnicity group bears
minimal elevated costs under the proposed option and Options la and lb. The magnitude of
Proposed PFAS Rule Economic Analysis
8-68
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
household cost differences between each of these race/ethnicity groups and the overall
population is small, ranging from $0.16 to $2.58 per year across race/ethnicity groups and across
the proposed option and regulatory alternatives. The Other race/ethnicity group bears the highest
household cost, while the non-Hispanic Black race/ethnicity group bears the lowest household
cost.
Consistent with EPA's findings for incremental household costs across all systems, EPA's
comparison of incremental household costs across system size categories for just treating
systems reveals that, in general, as system size increases, average incremental household costs
decrease under the proposed option and all regulatory alternatives and across all race/ethnicity
groups. One exception to this trend is among systems serving 100,000 to 1,000,000 people
where, in many cases, costs are marginally higher than for systems serving 50,000 to 100,000
people.
The highest average incremental household costs for treating systems under the proposed option
and all regulatory alternatives are realized for the smallest systems, with the range of incremental
household costs for systems serving 3,300 to 10,000 people ranging from $91.40 to $122.25 per
year. EPA anticipates the highest cost ($122.25 per year) would be for Option lc for the non-
Hispanic Black race/ethnicity group. Systems serving 10,000 to 50,000 people bear the lowest
average incremental household costs for treating systems under the proposed option and Options
la and lb, while systems serving 10,000 to 50,000 or 50,000 to 1,000,000 people, depending on
the race/ethnicity group, bear the lowest costs for treating systems under Options lc.
This analysis provides an opportunity to understand the demographic breakdown of incremental
household costs anticipated to be incurred due to treatment installation needed to comply with
the proposed PFAS NPDWR. Average incremental household costs for systems required to
install treatment are higher for all size categories and across all race/ethnicity groups compared
to average incremental household costs across all systems. These differences are expected, as
treatment installation costs are higher than other compliance costs (i.e., monitoring and
reporting). In some cases, such as for the smallest communities (i.e., systems serving 3,300 to
10,000 people), the annual average incremental household costs isolated among only systems
anticipated to install treatment are up to $100 higher than annual incremental household costs
averaged across all systems.
Table 8-27: Annualized Population-Weighted Household Cost for Treating PWSs by Size
Category and Race/Ethnicity Group
System Size"
Race/Ethnicity Group
Proposed
Optionb
Option
lac
Option
lbd
Option
lce
3,300 to 10,000
All
$118.50
$117.90
$117.18
$112.78
3,300 to 10,000
Non-Hispanic Black
$118.74
$118.06
$117.36
$122.25
3,300 to 10,000
Hispanic
$119.67
$118.84
$116.15
$91.43
3,300 to 10,000
Other
$116.35
$116.22
$113.09
$99.40
3,300 to 10,000
Non-Hispanic White
$118.39
$117.80
$117.50
$115.28
10,000 to 50,000
All
$32.88
$32.24
$30.83
$26.83
10,000 to 50,000
Non-Hispanic Black
$32.84
$32.25
$30.85
$27.23
10,000 to 50,000
Hispanic
$30.78
$30.15
$28.67
$24.58
10,000 to 50,000
Other
$33.07
$32.41
$30.99
$26.71
Proposed PFAS Rule Economic Analysis
8-69
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 8-27: Annualized Population-Weighted Household Cost for Treating PWSs by Size
Category and Race/Ethnicity Group
System Size3
Race/Ethnicity Group
Proposed
Optionb
Option
lac
Option
lbd
Option
lce
10,000 to 50,000
Non-Hispanic White
$33.31
$32.67
$31.26
$27.26
50,000 to 100,000
All
$23.03
$22.46
$21.43
$17.83
50,000 to 100,000
Non-Hispanic Black
$22.45
$21.96
$21.11
$17.65
50,000 to 100,000
Hispanic
$22.25
$21.65
$20.63
$17.63
50,000 to 100,000
Other
$23.90
$23.16
$22.07
$18.35
50,000 to 100,000
Non-Hispanic White
$23.26
$22.70
$21.65
$17.86
100,000 to 1,000,000
All
$25.45
$24.64
$23.02
$17.79
100,000 to 1,000,000
Non-Hispanic Black
$24.74
$23.89
$22.27
$17.24
100,000 to 1,000,000
Hispanic
$25.83
$24.94
$23.18
$17.38
100,000 to 1,000,000
Other
$27.89
$27.01
$25.60
$19.99
100,000 to 1,000,000
Non-Hispanic White
$25.16
$24.39
$22.79
$17.74
Notes:
aThe number of systems serving fewer than 3,300 people represented in the UMCR 3 occurrence data is too limited to
accurately estimate average population-weighted household costs by subpopulation. Therefore, results for these small systems
are omitted. Also, household costs in this exhibit are population-weighted and will not match average household costs by size
category shown in other exhibits in the economic analysis document that are not population-weighted.
bThe proposed option sets PFOA and PFOS MCLs of 4.0 ppt and an HI of 1.0.
cOption la sets PFOA and PFOS MCLs of 4.0 ppt.
Option lb sets PFOA and PFOS MCLs of 5.0 ppt.
eOption lc sets PFOA and PFOS MCLs of 10.0 ppt.
8.5 Conclusions
This section provides a summary of the EJ analyses for estimating baseline PFAS exposure and
exposure over several thresholds as well as the cost and benefits of the proposed PFAS NPDWR.
8.5.1 EJ PFAS Exposure Analysis
EPA's baseline analysis of demographic groups with PFAS exposure over baseline thresholds
based on Method 537.1 detection limits demonstrates that certain communities of color
experience elevated baseline PFAS drinking water exposures compared to the entire sample
population. For example, the percentage of Asian and Pacific Islander and Hispanic populations
with PFAS drinking water exposure above baseline thresholds is greater than the percentage of
the total population served across all demographic groups with PFAS drinking water exposure
above these levels. When these results are further filtered by system size, for large systems,
Asian and Pacific Islander and Hispanic populations have higher baseline PFAS drinking water
exposure compared to the percentage of the total population served across all demographic
groups. For small systems, Black populations served have higher baseline PFAS drinking water
exposure compared to the percentage of the total population served across all demographic
groups facing PFAS drinking water exposure over these thresholds.
Across all hypothetical regulatory thresholds, elevated exposure—and thus anticipated
reductions in exposure under the hypothetical regulatory scenarios—is anticipated to occur in
Proposed PFAS Rule Economic Analysis
8-70
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
communities of color and/or low-income populations. EPA estimates the most notable
differences in anticipated reductions in exposure are for Hispanic populations, specifically when
using UCMR 5 MRL values as hypothetical regulatory thresholds in the analysis. The results
from EPA's analysis indicate that Hispanic populations are estimated to face nearly twice the
rate of exposure for PFOA and PFOS. Hispanic populations are therefore also anticipated to have
greater reductions in exposure compared to the entire sample population.
These findings are supported by literature that indicates that communities of lower
socioeconomic status are more likely to live near environmentally hazardous facilities and face
disproportionate impacts of exposure to toxic chemicals than are communities of relatively
higher socioeconomic status (Brown, 1995; Brulle et al., 2006; Banzhaf et al., 2019; U.S. EPA,
1994). The literature also indicates that people of color and low-income populations are more
likely to be served by water systems with higher PFAS occurrence or reside in proximity to a
PFAS contamination site, thereby increasing baseline exposure (Black et al., 2021; Lee, Kang, et
al., 2021; Desikan et al., 2019).
Proposed PFAS Rule Economic Analysis
8-71
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
8.5.2 SafeWater EJ Analysis of Regulatory Options
EPA's analysis of the demographic distribution of health benefits and household costs
anticipated to result from the proposed PFAS NPDWR demonstrate that for all race/ethnicity
groups, EPA's proposed option offers the greatest quantified benefits. Additionally, in all but one
instance, EPA's proposed option will result in the highest household costs.
Under the proposed option, quantified health benefits are highest in every evaluated
race/ethnicity group and health endpoint compared to the other regulatory alternatives.
Additionally, across all health endpoints, communities of color (i.e., Hispanic, non-Hispanic
Black, and/or Other race/ethnicity groups) are anticipated to experience the greatest quantified
benefits associated with the proposed option. This finding could be driven by disparities in
baseline exposure to PFAS and underlying disparities in death and/or disease incidence by
race/ethnicity. This potential explanation is supported by literature demonstrating that vulnerable
communities continue to experience elevated rates of morbidity and mortality (Uche et al., 2021;
Driscoll et al., 2021; Fryar et al., 2017). Additionally, evidence in the literature indicates that
people of color and low-income populations are more likely to be served by water systems with
higher PFAS occurrence or reside in close proximity to a PFAS contamination site, which also
supports this finding (Black et al., 2021; Lee, Kang, et al., 2021; Desikan et al., 2019).
While some race/ethnicity groups (i.e., non-Hispanic Black, Hispanic, and Other) are anticipated
to bear elevated costs compared to incremental household costs for the overall population across
race/ethnicity groups, relative differences in these household costs are small. The minimal
differences in household costs anticipated to result from the proposed PFAS NPDWR can likely
be attributed to disparities in baseline PFAS exposure among different race/ethnicity groups.
Additionally, incremental household costs to all race/ethnicity groups generally decrease as
system size increases, which is expected due to economies of scale. Due to the overlap in
vulnerabilities demonstrated by slightly elevated household costs anticipated for particular
race/ethnicity groups and consistently elevated household costs for households served by small
systems, communities of color served by small systems are anticipated to face compounding
burdens. This is especially true if systems serving these communities are required to install
treatment to comply with the PFAS NPDWR.
8.5.3 Overall Environmental Justice Conclusion
EPA conducted the EJ analyses presented in this chapter on populations served by a subset of
PWSs to assess the demographic distribution of exposure to PFAS and the EJ impacts that are
anticipated to result from the proposed PFAS NPDWR. EPA conducted two separate analyses to
address the following questions:
(1) Are population groups of concern (i.e., people of color and low-income populations)
disproportionately exposed to PFAS compounds in drinking water delivered by PWSs?
(2) Are population groups of concern disproportionately affected by the proposed rule?
(3) If any disproportionate impacts are identified, do they create or mitigate baseline EJ
concerns?
Proposed PFAS Rule Economic Analysis
8-72
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
When examined collectively, results from these analyses identify communities of color and low-
income communities as being disproportionately exposed to PFAS in drinking water under
current baseline conditions. In one hypothetical regulatory scenario, communities of color are
currently exposed to twice the rate of PFAS exposure in drinking water compared to exposure
faced by the entire sample population. When quantifying the race/ethnicity distribution of
quantified health benefits anticipated to result from the proposed PFAS NPDWR, EPA found
that of the race/ethnicity groups evaluated, communities of color are anticipated to experience
the greatest health benefits under the proposed option and all regulatory alternatives. For
instance, for the non-Hispanic Black race/ethnicity group, under the proposed option, it is
anticipated that 3.10 deaths from CVD will be avoided per 100,000 people per year.
When comparing benefits across the proposed option and regulatory alternatives, quantified
health benefits were the highest for communities of color under the proposed option. This
finding could be influenced by the fact that elevated baseline exposure rates for these populations
translate to higher benefits associated with the proposed option, as greater reductions in exposure
are anticipated to occur as a result of implementing the proposed PFAS NPDWR.
To alleviate potential cost disparities identified by EPA's analysis, there may be an opportunity
for some communities to utilize BIL (P.L. 117-58) funding to provide financial assistance for
addressing emerging contaminants. BIL funding has specific allocations for both disadvantaged
and/or small communities and emerging contaminants, including PFAS.
Proposed PFAS Rule Economic Analysis
8-73
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
9 Statutory and Administrative Requirements
As part of the rulemaking process, EPA is required to address the burden that the proposed rule
may place on certain types of governments, businesses, and populations. This chapter presents
analyses performed by EPA in accordance with the following federal mandates and statutory
requirements:
1. Executive Order 12866: Regulatory Planning and Review and Executive Order 13563
(2011): Improving Regulation and Regulatory Review.
2. Paperwork Reduction Act (PRA) (U.S. EPA, 2010b).
3. The Regulatory Flexibility Act (RFA) of 1980, as amended by the Small Business
Regulatory Enforcement Fairness Act (SBREFA) of 1996.
4. Unfunded Mandates Reform Act (UMRA) of 1995.
5. Executive Order 13132: Federali sm
6. Executive Order 13175: Consultation and Coordination with Indian Tribal Governments.
7. Executive Order 13045: Protection of Children from Environmental Health and Safety
Risks.
8. Executive Order 13211: Actions That Significantly Affect Energy Supply, Distribution,
or Use.
9. National Technology Transfer and Advancement Act of 1995 (NTTAA).
10. Executive Order 12898: Federal Action to Address Environmental Justice in Minority
Populations and Low-Income Populations.
11. Consultations with the Science Advisory Board (SAB), National Drinking Water
Advisory Council (NDWAC), and the Department of Health and Human Services.
12. SDWA Section 1412(b)(4)(E) National Small System Affordability Determination.
Many of the statutory requirements and executive orders listed above call for an explanation of
why the proposed requirements are necessary, the statutory authority for the proposed
requirements, and the primary objectives that the proposed requirements are intended to achieve
(see Chapter 3 for additional information regarding the need for the proposed rule). Others are
designed to assess the financial and health effects of the proposed regulatory requirements on
sensitive, low-income, and tribal populations as well as on small systems and governments.
9.1 Executive Order 12866: Regulatory Planning and Review
and Executive Order 13563: Improving Regulation and
Regulatory Review
Executive Order 12866, 1993 (58 FR 51735, October 4, 1993) gives OMB the authority to
review regulatory actions that are categorized as "significant" under section 3(f) of Executive
Order 12866. The Order defines "significant regulatory action" as one that is likely to result in a
rule that may:
1. Have an annual effect on the economy of $100 million or more or adversely affect in a
material way the economy, a sector of the economy, productivity, competition, jobs, the
environment, public health or safety, or state, local, or tribal governments or
communities.
Proposed PFAS Rule Economic Analysis
9-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
2. Create a serious inconsistency or otherwise interfere with an action taken or planned by
another agency.
3. Materially alter the budgetary impact of entitlements, grants, user fees, or loan programs
or the rights and obligations of recipients thereof.
4. Raise novel legal or policy issues arising out of legal mandates, the President's priorities,
or the principles set forth in the Executive Order.
This action is an economically significant regulatory action that was submitted to the OMB for
review. Any changes made in response to OMB recommendations have been documented in the
docket. The analysis in Chapter 7 compares the annual estimated incremental costs and the
annual incremental benefits of the proposed rule. In addition to the monetized costs and benefits
of the proposed regulation, a number of non-monetized impacts exist. See Sections 5.7, 6.2.2,
and 6.2.3 of this EA for greater detail on the non-monetized impacts of the proposed regulation.
9.2 Paperwork Reduction Act
The information collection requirements for the proposed rule will be submitted for approval to
OMB under the Paperwork Reduction Act (PRA), 44 U.S.C. 3501 et seq. The ICR supporting
statement prepared by the EPA has been assigned the EPA ICR number 2732.01 and is available
in the docket at https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
The PRA requires the EPA to estimate the burden, as defined in 5 CFR 1320.3(b), on PWSs and
Primacy Agencies of complying with the rule. The information collected as a result of the
proposed rule should allow Primacy Agencies and EPA to determine appropriate requirements
for specific systems and evaluate compliance with the proposed rule. Burden is defined at 5 CFR
1320.3(b) and means the total time, effort, and financial resources required to generate, maintain,
retain, disclose, or provide information to or for a federal agency. The burden includes the time
needed to conduct Primacy Agency and system activities during the first three years after
promulgation, as described below.
9.2.1 Primacy Agency Activities
EPA anticipates Primacy Agencies will be involved in the following activities for the first three
years after publication of the final rule:
• Startup activities - read and understand the rule, adopt regulatory change, and provide
internal and system staff with training and technical assistance;
• Review the initial monitoring event results, including confirmation sample results for
MCL exceedances; and
• Review the results of quarterly monitoring from systems.
9.2.2 Public Water System Activities
EPA anticipates systems will be involved in the following activities for the first three years after
publication of the final rule:
• Startup activities - read and understand the rule and attend initial training from the
primacy agency;
• Conduct initial monitoring including confirmation sampling for MCL exceedances; and
Proposed PFAS Rule Economic Analysis
9-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
• Conduct quarterly monitoring, as needed; EPA assumed that sampling for triennial
monitoring would not occur until after the three-year ICR period.
For the first three years after publication of the rule in the Federal Register, information
requirements apply to an average of 38,089 respondents annually, including 38,033 PWSs and 56
Primacy Agencies. The burden associated with the proposed rule over the three years covered by
the ICR is 3.8 million hours, for an average of 1.3 million hours per year. The total costs over the
three-year period is $142.6 million, for an average of $47.5 million per year (simple average over
three years). The average burden per response (i.e., the amount of time needed for each activity
that requires a collection of information) is 6.6 hours for PWSs and 1.1 hours for primacy
agencies; the average cost per response is $234.41 for PWSs and $60.89 for primacy agencies.
The collection requirements are mandatory under SDWA (42 U.S.C. 300g-7). Details on the
calculation of the proposed rule information collection burden and costs can be found in the ICR
for the proposed rule and Chapter 5 of this EA. A summary of the average annual burden and
costs of the collection is presented in Table 9-1. The burdens and costs reflect labor and
laboratory analysis costs.
Table 9-1: Average Annual Burden, Costs, and Responses for the Proposed Rule
Information Collection Request
Burden (hours in Costs (Million Responses
thousands)3 $2021)a
Systems
1,189
$42.3
180,630
Primacy agencies
91
$5.2
85,225
Totalb
1,281
$47.5
265,855
Average per response - systems (hours
or dollars)
6.6
$234
Not applicable
Average per response - primacy agencies
(hours or dollars)
1.1
$61
Not applicable
Notes:
different units indicated for the estimates of burden and cost average per response.
bDetail may not add to totals because of independent rounding.
Source: ICR Supporting Statement, available in the docket at https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
The estimates of total responses, burden, and cost for system and primacy agency startup
activities are provided in Table 9-2.
Proposed PFAS Rule Economic Analysis
9-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 9-2: Total Burden, Costs, and Responses for Each Required Activity
Burden
Item
(thousand
Costs (Million
$2021)
Responses
hours)
System startup activities
1,485
$52.8
133,060
Systems collect initial samples
1,127
$39.0
218,557
Systems collect quarterly samples
956
$35.2
190,274
System subtotal
3,568
$127.0
541,891
Primacy agency startup activities
154
$8.7
168
Primacy agency review initial monitoring data
73
$4.1
160,370
Primacy agency review quarterly samples
48
$2.7
95,137
Primacy agency subtotal3
274
15.6
255,675
Combined systems and primacy agency3
3,842
142.6
797,566
Note:
aDetail may not add to totals because of independent rounding.
Source: ICR Supporting Statement, available in the docket at https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
An agency may not conduct or sponsor, and a person is not required to respond to, a collection of
information unless it displays a currently valid OMB control number. The OMB control numbers
for the EPA's regulations in 40 CFR are listed in 40 CFR part 9.
As part of the Federal Register notice on the proposed rule, EPA will solicit comments on this
information collection and the estimates in this ICR. EPA will solicit comments on specific
aspects of the proposed information collection, as described below:
1. EPA's need for this information.
2. The accuracy of the provided burden estimates.
3. Any suggested methods for minimizing respondent burden.
Comments should be directed to Docket ID Number EPA-HQ-OW-2022-0114.
In compliance with the PRA (44 USC 3501 et seq.), EPA will submit the ICR for the proposed
rule to OMB for review. EPA will summarize any comments received from OMB on the ICR at
that time.
9.3 The Initial Regulatory Flexibility Analysis
The Regulatory Flexibility Act (RFA) of 1980, amended by the Small Business Regulatory
Enforcement Fairness Act (SBREFA) of 1996, requires regulators to assess the effects of
regulations on small entities including businesses, nonprofit organizations, and governments.
RFA/SBREFA generally requires an agency to prepare an initial regulatory flexibility analysis
(IRFA) of any rule subject to notice and comment rulemaking requirements under the
Administrative Procedure Act or any other statute unless the agency certifies that the rule will
not have a significant economic impact on a substantial number of small entities (SISNOSE).
Small entities include small businesses, small organizations, and small governmental
jurisdictions. Under the RFA, the IRFA must include a description of the reasons why action by
the agency is being considered, a succinct statement of the objectives and legal basis for the
Proposed PFAS Rule Economic Analysis
9-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
proposed rule. It must also include a description of and, where feasible, an estimate of the
number of small entities that will be affected and it must describe the projected reporting,
recordkeeping, and other compliance requirements of the proposed rule and must identify any
relevant federal rules that may duplicate, overlap, or conflict with the proposed rule. Finally, the
IRFA must describe any significant regulatory alternatives to the rule that would accomplish the
stated objectives of the applicable statutes and would minimize any significant economic impacts
of the rule on small entities.
The RFA provides default definitions for each type of small entity. Small entities are defined as:
(1) a small business as defined by the Small Business Administration's (SBA) regulations at 13
CFR 121.201; (2) a small governmental jurisdiction that is a government of a city, county, town,
school district, or special district with a population of less than 50,000; and (3) a small
organization that is any "not-for-profit enterprise which is independently owned and operated
and is not dominant in its field." The RFA also authorizes an agency to use alternative
definitions for each category of small entity, "which are appropriate to the activities of the
agency" after proposing the alternative definition(s) in the Federal Register and taking comment
(5 USC 601(3)-(5)). In addition, to establish an alternative small business definition, agencies
must consult with SBA's Chief Counsel for Advocacy.
For purposes of assessing the impacts of the proposed rule on small entities, EPA considered
small entities to be systems serving 10,000 people or fewer. This is the threshold specified by
Congress in the SDWA 1996 Amendments for small system flexibility provisions. As required
by the RFA, EPA proposed using this alternative definition in the Federal Register (FR) (63 FR
7620, February 13, 1998), requested public comment, consulted with the SBA, and finalized the
alternative definition in the Agency's Consumer Confidence Reports regulation (U.S. EPA,
1998c, 63 FR 44524, August 19, 1998). As stated in that Final Rule, the alternative definition
would be applied for all future drinking water regulations.
EPA notes that the Infrastructure Investment and Jobs Act (also known as the Bipartisan
Infrastructure Law (BIL), P.L. 117-58) invests over $11.7 billion in the Drinking Water State
Revolving Fund (SRF) General Supplemental fund; $4 billion in the Drinking Water SRF
Emerging Contaminants fund; and $5 billion in the Emerging Contaminants in Small or
Disadvantaged Communities grant program. Together, these funds will reduce people's exposure
to perfluoroalkyl and polyfluoroalkyl substances (PFAS) and other emerging contaminants
through their drinking water. The BIL funding will prioritize investment in local communities
that are on the frontlines of PFAS contamination and that have few options to finance solutions
through traditional programs and, help them meet their obligations under this proposed
regulation.
9.3.1 Need for, Objectives, and Legal Basis of the Rule
The need for the rule, the objectives of the rulemaking, the stakeholder outreach conducted, and
the statutory authority EPA is utilizing to finalize the rule are described in detail in Chapter 3.
See Section 3.1 for detailed information on the need for the rule, Section 9 for information on
stakeholder outreach during the rulemaking process, and Section 3.2 for additional detail on the
statutory authority for the promulgation of the PFAS regulation. In summary, SDWA authorizes
EPA to establish NPDWRs for contaminants that may have an adverse public health effect, that
are known to occur or that present a substantial likelihood of occurring in PWSs at a frequency
Proposed PFAS Rule Economic Analysis
9-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
and level of public health concern, and that present a meaningful opportunity for health risk
reduction for persons served by PWSs. As a result, EPA is proposing an NPDWR for six PFAS
including PFOA, PFOS, PFNA, PFHxS, HFPO-DA, and PFBS. Additionally, under the SDWA,
the EPA Administrator is authorized to establish monitoring, recordkeeping, and reporting
regulations that the Administrator can use to establish regulations under the SDWA, determine
compliance with SDWA, and advise the public of the risks of unregulated contaminants.
EPA is also addressing PFAS through several of its statutory authorities other than SDWA,
including the CERCLA, RCRA, Toxic Substances Control Act (TSCA), Clean Water Act, Clean
Air Act, and Emergency Planning and Community Right-to-Know Act. For example, as part of
the EPA PFAS Strategic Roadmap, in 2022, EPA anticipates proposing to designate certain
PFAS as CERCLA hazardous substances to require reporting of PFOA and PFOS releases,
enhance the availability of data, and ensure agencies can recover cleanup costs. EPA recognizes
that future actions under some of these statutes may have direct or indirect impacts for drinking
water treatment facilities and could impact the compliance requirements related to disposal of
PFAS treatment residuals that are generated by water systems. EPA has also committed to
restrict PFAS discharges from industrial sources through a multi-faceted Effluent Limitations
Guidelines program to proactively establish national technology-based regulatory limits.
Additionally, EPA is seeking to proactively use National Pollutant Discharge Elimination
System (NPDES) authorities to reduce discharges of PFAS at the source and obtain more
comprehensive information through monitoring on the sources of PFAS discharges and quantity
of PFAS discharged by these sources. EPA notes that these actions may prevent or reduce PFAS
entering into sources of drinking water in the future. More information on these statutory
authorities and PFAS-related EPA activities can be found in the Roadmap.
9.3.2 Identification of Relevant Federal Rules
The proposed rule is not anticipated to duplicate, overlap, or conflict with any other federal rules.
There are NPDWRs for over 90 contaminants and when developing drinking water regulations,
the agency factors in the water quality impacts of compliance with a new regulation on the
system's compliance with existing drinking water regulations (e.g., Lead and Copper Rule,
Interim Enhanced Surface Water Treatment Rule, Stage 1 and Stage 2 Disinfectants and
Disinfection Byproducts Rules). EPA will continue to consider and evaluate how water systems
will need to manage simultaneous compliance with the proposed PFAS NPDWR requirements
and these other EPA drinking water regulations. Further, while the proposed PFAS NPDWR is
not anticipated to duplicate, overlap, or conflict with any other federal rules, EPA notes that
monitoring under the UCMR 5 may also support monitoring requirements associated with the
proposed PFAS NPDWR.
9.3.3 Summary of the SBAR Comments and Recommendations
A Small Business Advocacy Review Panel (SBAR Panel or Panel) was convened to review the
planned proposed rulemaking on the Proposed PFAS NPDWR. In addition to EPA's Small
Business Advocacy Chairperson, the Panel consists of the Director of the Standards and Risk
Management Division of the EPA Office of Ground Water and Drinking Water, the
Administrator of the Office of Information and Regulatory Affairs within the Office of
Management and Budget, and the Chief Counsel for Advocacy of the Small Business
Administration. The panel consulted with and reported on the comments of small entity
Proposed PFAS Rule Economic Analysis
9-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
representatives (SERs) and made findings on issues related to elements of an IRFA under section
603 of the RFA. The SERs were presented with information related to PFAS background (such
as health and occurrence, the SDWA regulatory development process and EPA's actions to
address PFAS in drinking water potential monitoring and reporting rule compliance
considerations, treatment and feasibility considerations, potential public notification and
education rule compliance considerations, and preliminary economic impacts to small systems.
EPA also provided to SERS that the agency's final regulatory determination for PFOA and
PFOS outlined avenues that the agency considered to further evaluate additional PFAS
chemicals, other than PFOA and PFOS, and consider groups of PFAS as supported by use of the
best available science. Additionally, as part of EPA's PFAS Strategic Roadmap, EPA reaffirmed
its commitment to evaluate additional PFAS and consider regulatory actions to address
additional PFAS or groups of PFAS as it develops the NDPWR. Further, EPA provided to SERs
that as EPA considers whether to include additional PFAS as part of this proposed regulation, the
agency would consider several factors, including whether the same treatment approaches co-
remove certain PFAS contaminants and how different PFAS are anticipated to be removed as
part of the treatment process, the likelihood that the PFAS co-occur, the similarity of health
effects and chemical structures, the environmental persistence characteristics, and the availability
of accepted and approved analytical methods or indicators with comparable costs to those
currently identified by EPA to evaluate PFAS removal from drinking water, among other
considerations.
In light of the SERs' comments, the Panel considered the regulatory flexibility issues and
elements of the IRFA specified by RFA/SBREFA and developed the findings and discussion
summarized in the SBAR report. For example, the SBAR Panel recommended several
flexibilities in monitoring requirements for small systems, including the use of existing
monitoring data (such as the UCMR 5) for initial monitoring purposes; as well as reduced
compliance monitoring requirements specifically for small ground water systems. Regarding
public comment requests, the Panel recommended that EPA request this for a few areas, such as
laboratory capacity for monitoring, additional treatment technologies other than those identified
in the proposed rule that have been shown to reduce levels of PFAS to the proposed regulatory
standards, additional monitoring flexibilities, and PFAS disposal considerations. Moreover,
specific to PFAS disposal, the Panel recommended that EPA continue to evaluate the potential
impacts related to the disposal of PFAS treatment residuals and potential implications from other
EPA statutory authorities. This recommendation included presenting the costs of both non-
hazardous and hazardous waste disposal of treatment residuals as a part of the proposed rule. As
a general matter, EPA notes that such wastes are not currently regulated under federal law as a
hazardous waste. To address stakeholder concerns, including those raised during the SBREFA
process, EPA conducted a sensitivity analysis with an assumption of hazardous waste disposal
for illustrative purposes only. As part of this analysis, EPA generated a second full set of unit
cost curves that are identical to the curves used for the national cost analysis with the exception
that spent GAC and spent IX resin are considered hazardous. EPA acknowledges that if federal
authorities later determine that PFAS-contaminated wastes require handling as hazardous wastes,
the residuals management costs are expected to be higher. EPA incorporated all of these Panel
recommendations, as well as others, in the proposed rule.
The Panel also recommended EPA to consider rule implementation delays for potential
laboratory capacity-related challenges if those challenges potentially impact the ability of water
Proposed PFAS Rule Economic Analysis
9-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
systems to monitor for PFAS and reasonably comply with the NPDWR. As described in the
preamble (section XII.D.), in accordance with SDWA 1412(b)(10), a state or EPA may grant an
extension of up to two additional years to comply with an NPDWR's MCL if the state or EPA
determines a system needs additional time for capital improvements. At this time, EPA does not
intend to provide a two-year extension nationwide. However, under SDWA 1412(b)(10) or 1416
States may provide such as extension on an individual system basis which may address
compliance issues associated with treatment, laboratory, and disposal capacity. Additionally,
EPA notes that in the preamble (section IX.F.) the agency is seeking public comment on the
proposed initial monitoring timeframe, particularly for NTNCWS or all systems serving 3,300 or
fewer.
The report includes a number of other observations and recommendations to meet the statutory
obligations for achieving small-system compliance through flexible regulatory compliance
options. The report was finalized on August 1, 2022 and transmitted to EPA Administrator for
consideration. Detailed information on the overall panel process, including the comprehensive
comments of the SERs and full description of Panel recommendations, can be found in the panel
report titled, Final Report of the Small Business Advocacy Review Panel on EPA's Planned
Proposed Ride Per- and Polyflaoroalkyl Substances National Primary Drinking Water
Regulation and can be found in the rulemaking docket at:
https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
9.3.4 Number and Description of Small Entities Affected
EPA used SDWIS/Fed data from the fourth quarter of 2021 to identify 62,048 small PWSs that
may be impacted by the proposed PFAS regulation. A small PWS serves between 25 and 10,000
people. These water systems include 44,753 CWSs that serve year-round residents and 17,295
NTNCWSs that serve the same persons over six months per year (e.g., a PWS that is an office
park or church). EPA does not anticipate that the proposed NPDWR will affect TNCWSs as
those systems will likely not be subject to the rule requirements. Additional information on the
characteristics of these small drinking water systems along with a discussion of uncertainty in the
dataset used to derive the estimated number of small systems impacted by the proposed PFAS
regulation can be found in 4.2.1.
Table 9-3 and Table 9-4 show the number of affected small CWSS and NTNCWs respectively.
Proposed PFAS Rule Economic Analysis
9-8
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 9-3: Inventory of Small CWSs
System Size (Population
Ground Water
Surface Water
Total
Served)
A
B
C = A + B
< 100
10,654
739
11,393
101-500
13,037
2,042
15,079
501-1,000
4,132
1,179
5,311
1,001-3,300
5,503
2,460
7,963
3,301-10,000
2,784
2,223
5,007
TOTAL
36,110
6,601
44,753
Abbreviations: CWS - community water systems.
Note:
includes 23 CWSs serving 10,000 or fewer people for which no primary source water type was reported to SDWIS/Fed.
EPA assigned these systems to the source type of Ground Water.
Source: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through January 14, 2022.
Includes all active CWSs.
Table 9-4: Inventory of Small NTNCWSs
NTNCWSs'
System Size (Population
Served)
Ground Water
SW
Total
A
B
C=A+B
< 100
8,084
252
8,336
101-500
6,111
257
6,368
501-1,000
1,476
91
1,567
1,001-3,300
743
121
864
3,301-10,000
97
63
160
TOTAL
16,551
784
17,295
Abbreviations: NTNCWS - non-transient non-community water systems.
Note:
includes 11 NTNCWSs serving 3,300 or fewer people for which no primary source type was reported to SDWIS/Fed. EPA
assigned these systems to the source water type of Ground Water.
Sources: SDWIS/Fed fourth quarter 2021 "frozen " dataset that contains information reported through January 14, 2022.
Includes all active NTNCWSs.
9.3.5 Description of Compliance Requirements of the Proposed
Rule
For a detailed description of the regulatory requirements under the proposed PFAS regulation see
Section 2.1. Under the proposed rule requirements, PWSs subject to the rule are required to
conduct initial monitoring or demonstrate that recent, previously collected monitoring data can
be used to determine the level of PFAS in their water system. The proposed NPDWR includes a
provision, made available to PWSs of all sizes, including CWSs and NTNCWs serving 10,000 or
fewer people, to use qualified previously collected monitoring data. EPA assessed the extent to
which this significant alternative minimizes the economic impact on small PWSs specifically in
Section 9.3.7.1 below.
Proposed PFAS Rule Economic Analysis
9-9
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Based on initial monitoring results, systems will be required to conduct ongoing monitoring at
least every three years or as often as four times per year. Details on the monitoring frequency
requirements of the proposed NPDWR can be found in Section IX of the Federal Register Notice
for the proposed rule. EPA has included a provision in the proposed NPDWR where ground
water systems serving a population of 10,000 or fewer may collect two quarterly samples over a
one-year period for the purpose of initial monitoring, rather than collecting four quarterly
samples. EPA assessed the extent to which this regulatory flexibility minimizes the economic
impact on small PWSs in Section 9.3.7.2 below.
PWSs that exceed the drinking water standard are required to choose between treatment and non-
treatment compliance options. EPA identified the following Small System Compliance
Technologies (SSCTs) GAC, Anion Exchange (ADC), and High-pressure membranes (Reverse
Osmosis [RO] and Nanofiltration [NF], POU RO is not currently listed as a compliance option
because the regulatory options under consideration require treatment to concentrations below the
current NSF International/American National Standards Institute (NSF/ANSI) certification
standard for POU device removal of PFAS. However, POU treatment is reasonably anticipated
to become a compliance option for small systems in the future if NSF/ANSI or other independent
third-party certification organizations develop a new certification standard that mirrors EPA's
proposed regulatory standard. Details on SSCTs and costs can be found in Section 5.3.1 and Best
Available Technologies and Small System Compliance Technologies for Per- and
Polyfluoroalkyl Substances (PFAS) in Drinking Water (U.S. EPA, 2023f).
9.3.6 Analysis of Impact of Regulatory Options on Small System
Costs
EPA limited the quantitative cost impact analysis to small CWSs because small NTNCWSs
operate in numerous industries and EPA does not have information on NTNCWSs' revenues.
EPA's decision to limit its cost impact analysis to CWSs is supported by EPA's Assessment of
the Vulnerability of Noncommunity Water Systems to SDWA Cost Increases (2008). In this
study, EPA examined the burden of SDWA rule costs in comparison to the average revenues of
various categories of NTNCWSs. All the NTNCWS categories reviewed were less vulnerable to
SDWA-related increases than a typical CWS. The report notes that in some categories of
businesses, costs are more easily passed on to the customer base than in others. In each
NTNCWS category, however, total expenditures on water were found to be a relatively small
percentage of total revenues. Water expenditures (including expenditures for sewer service and
miscellaneous other utilities) totaled less than one percent of total revenues in nearly all cases
and were not more than 1.3 percent of total revenues for any category. The implication is that an
increase in water costs would similarly be less than one percent of revenue. This report included
several caveats such as one that considered the potential for underestimating the impact to golf
courses, which were grouped in with other recreational entities whose use of water was less
significant to the core business than the golf courses. EPA notes, however, that irrigation water
for golf courses would not need to meet the proposed rule; only water used for human
consumption would need to be treated. Despite the significant caveats listed, the report strongly
suggests that NTNCWSs should not be considered particularly vulnerable to operating cost
increases resulting from SDWA rulemakings.
Proposed PFAS Rule Economic Analysis
9-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
To indicate the potential economic impact on small CWSs, EPA divided annual costs by annual
revenues and converted the decimal values to percentages and identified the number and percent
of CWSs for which the impact percentages exceeded thresholds of one percent and three percent.
For each system, EPA estimated annual revenue using each system's average daily flow and the
average revenue per thousand gallons delivered from the CWSS (U.S. EPA, 2009). For annual
costs, EPA estimated annual average monitoring costs based on system size and baseline PFAS
occurrence. Annual costs also included annual treatment costs when baseline PFAS
concentrations exceeded the PFAS limits of the proposed rule or options. Annual treatment costs
are the sum of annual operating and maintenance costs and annualized capital costs, which vary
for 3 percent and 7 percent discount rate assumptions.
Table 9-5 shows the number and proportion of CWSs incurring annual costs that exceed 1
percent and 3 percent of annual revenue at the commercial rate of capital for the proposed
option. Under the proposed option, 17,726 small CWSs (39 percent of small CWSs) could incur
annual costs greater than 1 percent of annual revenue and 9,233 small CWSs (21 percent of small
CWSs) could incur annual costs greater than 3 percent of revenue. These potential impacts are
high enough to preclude a finding of no SISNOSE. For systems that install treatment to reduce
PFAS, annual costs can range from approximately $8,000 to $315,000. Details on treatment
costs curves can be found in Section 5.3.1 and Best Available Technologies and Small System
Compliance Technologies for Per- and Polyfluoroalkyl Substances (PFAS) in Drinking Water
(U.S. EPA, 2023f). For EPA's estimates of treatment costs by system size, see Appendix C.l.
For information on federal financial assistance available to small systems for the installation of
PFAS treatment technology, see Section 9.12.2.2.
Table 9-5: Cost-Revenue Ratio for Small CWSs, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0; Commercial Cost of Capital)
Ownership
Source
Water
Population
Served Size
Category
Number of
CWSs
Number of
CWSs
with Cost
Revenue
Ratio >
1%
Number of
CWSs
with Cost
Revenue
Ratio >
3%
Percent of
CWS with
Cost
Revenue
Ratio >
1%
Percent of
CWS with
Cost
Revenue
Ratio >
3%
Private
Ground
Less than 100
9,250
9,250
4,800
100%
52%
Private
Ground
100 to 500
8,223
2,758
1,468
34%
18%
Private
Ground
500 to 1,000
1,313
250
108
19%
8%
Private
Ground
1,000 to 3,300
1,046
154
72
15%
7%
Private
Ground
3,300 to 10,000
347
30
24
9%
7%
Private
Surface
Less than 100
399
398
196
100%
49%
Private
Surface
100 to 500
769
207
129
27%
17%
Private
Surface
500 to 1,000
244
44
16
18%
7%
Private
Surface
1,000 to 3,300
278
27
15
10%
6%
Private
Surface
3,300 to 10,000
184
14
12
7%
6%
Public
Ground
Less than 100
1,308
697
250
53%
19%
Public
Ground
100 to 500
4,684
1,142
806
24%
17%
Public
Ground
500 to 1,000
2,767
521
255
19%
9%
Proposed PFAS Rule Economic Analysis
9-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 9-5: Cost-Revenue Ratio for Small CWSs, Proposed Option (PFOA and PFOS
MCLs of 4.0 ppt and HI of 1.0; Commercial Cost of Capital)
Ownership
Source
Water
Population
Served Size
Category
Number of
CWSs
Number of
CWSs
with Cost
Revenue
Ratio >
1%
Number of
CWSs
with Cost
Revenue
Ratio >
3%
Percent of
CWS with
Cost
Revenue
Ratio >
1%
Percent of
CWS with
Cost
Revenue
Ratio >
3%
Public
Ground
1,000 to 3,300
4,385
681
315
16%
7%
Public
Ground
3,300 to 10,000
2,401
239
186
10%
8%
Public
Surface
Less than 100
330
170
71
51%
21%
Public
Surface
100 to 500
1,241
277
194
22%
16%
Public
Surface
500 to 1,000
925
164
65
18%
7%
Public
Surface
1,000 to 3,300
2,160
257
123
12%
6%
Public
Surface
3,300 to 10,000
2,026
146
128
7%
6%
Total
44,280
17,426
9,233
39%
21%
Abbreviations: CWS - community water system
9.3.7 Analysis of Significant Alternatives to the Proposed Rule
Significant alternatives presented by the SBAR panel report are described below. EPA evaluated
the minimized economic impact for small systems for each of these alternatives.
9.3.7.1 Use of Previously Collected PFAS Monitoring Doto
EPA has included a provision in the proposed NPDWR where PWSs of all sizes may use
previously collected monitoring data if it meets stated criteria in lieu of initial monitoring. This
significant alternative is expected to offer a substantial costs savings to small PWSs, particularly
those serving between 3,301 and 10,000 that participate in UCMR 5. For the national cost
analysis, EPA assumes that systems with either UCMR 5 data or monitoring data in the State
PFAS Database (U.S. EPA, 2023g) will not need to conduct the initial year of monitoring. As a
simplifying assumption for the cost analysis, EPA assumes all systems serving a population of
greater than 3,300 have UCMR 5 data and those serving 3,300 or less do not. EPA notes that this
assumption is conservative and will likely overestimate costs for systems less than 3,300 as
many state monitoring programs and other efforts will have collected monitoring data that can be
used as initial monitoring data for these systems, thus offsetting those costs. Under these
assumptions, EPA estimates that this provision will reduce the economic burden on small
systems nationally by $39 million dollars.
9.3.7.2 Reduced Monitoring for Small Ground Water Systems
EPA has included a provision in the proposed NPDWR where ground water systems serving a
population of 10,000 or fewer may collect two quarterly samples over a one-year period for the
purpose of initial monitoring, rather than collecting four quarterly samples. EPA estimates that
this provision will reduce the economic burden on small systems nationally by $3 million per
year.
Proposed PFAS Rule Economic Analysis
9-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
9.3.7.3 Point of Use (POU) Technologies os Small System Compliance
Technologies (SSCTs)
In the Best Available Technologies and Small System Compliance Technologies for Per- and
Polyfluoroalkyl Substances (PFAS) in Drinking Water (U.S. EPA, 2023f), EPA discusses POUs
and notes that the current certification standard is 70 ppt, which would not ensure these devices
are able to meet the MCLs of the proposed rule. EPA notes that based on the technologies used
in many POU devices (e.g., RO), the Agency anticipates devices are or will be capable of
meeting the MCLs in this proposed rulemaking. If POU certifications are updated and do meet
the SSCT criteria in the final NPDWR, this could minimize the economic impact of the final
regulation on small PWSs, particularly on water systems in the smallest size category (e.g., those
serving between 25 and 500 people). In particular, NTNCWS that control all of their potable taps
(e.g., schools, gas stations, churches) may find use of POUs to be a particularly attractive option.
POUs tend to be most cost effective for the smallest water systems. Costs for POUs may range
between $317 to $326 per year per system. These costs are lower than the costs for implementing
centralized treatment options such as GAC or IX in the smallest system size category, which can
range from $376 to $698 per year per system (U.S. EPA, 2023f). See Table 9-9 for more
information costs by system size and treatment technology. EPA has not estimated the potential
national economic impact reduction because the current certification prevents POUs from
meeting the SSCT criteria for the proposed NPDWR. However, EPA notes there is a potential
for significant burden reduction particularly for very small water systems if POU certifications
are updated and POUs meet the SSCT criteria for the final NPDWR.
9.4 Unfunded Mandates Reform Act
The Unfunded Mandates Reform Act (UMRA) (1995) seeks to protect state, local, and tribal
governments from the imposition of unfunded federal mandates. In addition, the Act seeks to
strengthen the partnership among the federal government and state, local, and tribal
governments.
Title II of UMRA establishes requirements for federal agencies to assess the effects of their
regulatory actions on state, local, and tribal governments and the private sector. Under Section
202 of UMRA, EPA generally must prepare a written statement, including a cost-benefit
analysis, for proposed and Final Rules with "federal mandates" that may result in expenditures
by state, local, and tribal governments, in the aggregate, or by the private sector, of $100 million
or more in any one year, adjusted for inflation. EPA has calculated the cost of the rule in 2021
dollars, therefore, the UMRA requirements are triggered if expenditures exceed $168 million in
one year (escalation based on GDP deflator).
Section 205 of UMRA generally requires EPA to identify and consider a reasonable number of
regulatory alternatives and adopt the least costly, most cost-effective or least burdensome option
that achieves the objectives of the rule. The provisions of Section 205 do not apply when they are
inconsistent with applicable law. Moreover, Section 205 allows EPA to adopt an alternative
other than the least costly, most cost-effective or least burdensome alternative if the
Administrator publishes with the rule an explanation of why that alternative was not adopted.
Before EPA establishes any regulatory requirements that may significantly or uniquely affect
small governments, including tribal governments, it must have developed under Section 203 of
Proposed PFAS Rule Economic Analysis
9-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
UMRA a small government agency plan. The plan must provide for notifying potentially
affected small governments, enabling officials of affected small governments to have meaningful
and timely input in the development of EPA regulatory proposals with significant federal
intergovernmental mandates, and informing, educating, and advising small governments on
compliance with the regulatory requirements. Options being considered for the proposed rule
also met the consultation requirements of Federalism, therefore EPA elected to engage the
UMRA and Federalism stakeholders in the same consultation as there are overlapping interests,
and a discussion of potential options for the development of the proposed rule was more
effectively communicated simultaneously. For more information on the consultation, refer to the
Summary Report on Federalism and Unfunded Mandates Reform Act Consultation for the
Development of the Proposed PFAS NPDWR in the public docket at
https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
The proposed rule does contain a federal mandate that may result in expenditures to state, local,
and tribal governments, in the aggregate, or to the private sector, of $168 million or more in any
one year. For the proposed rule, the highest annual incremental cost over the analysis period
occurs in the 4th year after rule promulgation. In this year publicly owned PWSs are expected to
have undiscounted incremental costs of $8.0 billion, privately owned PWSs are expected to have
undiscounted incremental costs of $1.8 billion, and Primacy Agencies will have undiscounted
incremental costs of $18 million. Therefore, the proposed rule has costs in a single year of $9.8
billion and, therefore, is subject to the requirements of Sections 202 and 205 of UMRA.
The annualized incremental costs of the proposed rule, that are borne by public, private, and
tribal PWSs are provided in Table 9-6. As the exhibit shows, public entities bear most of the
costs. As discussed in Chapter 5, in addition to these PWS costs primacy agencies will incur
annualized incremental administrative costs of $8 million (3 percent discount rate) or $9 million
(7 percent discount rate) under the proposed rule.
Table 9-6: Annual Incremental Costs by PWS Size and Ownership, Proposed Option
(PFOA and PFOS MCLs of 4.0 ppt and HI of 1.0; Million $2021, Commercial Cost of
Capital)
Publicly-Owned Public Water
Systems
Privately-Owned Public Water
Systems
Tribal-Owned Public Water
Systems
Public Water Systems Serving <
10,000 People
$127
$71
All Public Water Systems
$712
$183
$5
9.5 Executive Order 13132: Federalism
Executive Order 13132 (1999), entitled "Federalism" (64 FR 43255, August 10, 1999), requires
EPA to develop an accountable process to ensure "meaningful and timely input by state and local
officials in the development of regulatory policies that have federalism implications." "Policies
that have federalism implications" are defined in the Executive Order to include regulations that
have "substantial direct effects on the states, on the relationship between the national government
Proposed PFAS Rule Economic Analysis
9-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
and the states, or on the distribution of power and responsibilities among the various levels of
government."
This action has federalism implications due to the substantial direct compliance costs on state or
local governments. The net change in Primacy Agency related cost for state, local, and tribal
governments in the aggregate is estimated to be $8 million (3 percent discount rate) or $9 million
(7 percent discount rate).
To fulfill requirements of Executive Order 13132 section 6, EPA held a Federalism consultation
with state and local government officials as well as their representative associations to solicit
input on key areas to inform the development of the proposed rule. Options being considered for
the proposed rule also met the consultation requirements of UMRA, therefore EPA elected to
engage the UMRA stakeholders in the same consultation because there are overlapping interests,
and a discussion of potential options for the development of the proposed rule was more
effectively communicated simultaneously. For more information on the consultation, refer to the
Summary Report on Federalism and Unfunded Mandates Reform Act Consultation for the
Development of the Proposed PFAS NPDWR in the public docket at
https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
9.6 Executive Order 13175: Consultation and Coordination with
Indian Tribal Governments
Executive Order 13175 (2000), entitled "Consultation and Coordination with Indian Tribal
Governments" (65 FR 67249, November 9, 2000), requires EPA to develop an accountable
process to ensure "meaningful and timely input by tribal officials in the development of
regulatory policies that have tribal implications." The Executive Order defines "policies that
have tribal implications to include regulations that have "substantial direct effects on one or more
Indian tribes, on the relationship between the federal government and the Indian tribes, or on the
distribution of power and responsibilities between the federal government and Indian tribes."
Under Executive Order 13175, EPA may not issue a regulation that has tribal implications, that
imposes substantial direct compliance costs, and that is not required by statute, unless the federal
government provides the funds necessary to pay the direct compliance costs incurred by tribal
governments, or EPA consults with tribal officials early in the process of developing the
proposed regulation and develops a tribal summary impact statement.
EPA has identified 998 public water systems serving tribal communities, 84 of which are
federally owned. EPA estimates that tribal governments will incur public water system
compliance costs of $5 million per year attributable to monitoring, treatment or non-treatment
actions to reduce PFAS in drinking water, and administrative costs, and that these estimated
impacts will not fall evenly across all tribal systems. The proposed PFAS NPDWR does offer
regulatory relief by providing flexibilities for all water systems to potentially utilize pre-existing
monitoring data in lieu of initial monitoring requirements and for ground water CWSs and
NTNCWSs serving 10,000 or fewer to reduce initial monitoring from quarterly monitoring
during a consecutive 12-month period to only monitoring twice during a consecutive 12-month
period. These flexibilities may result in implementation cost savings for many tribal systems
since 98 percent of tribal CWSs and 94 percent of NTNCWs serve 10,000 or fewer people.
Proposed PFAS Rule Economic Analysis
9-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
EPA has concluded that this proposed rule has Tribal implications, because it will impose direct
compliance costs on Tribal governments, and the federal government will not provide funds
necessary to pay those direct compliance costs. However, EPA notes that the federal government
will provide a potential source of funds necessary to offset some of those direct compliance
costs. The Infrastructure Investment and Jobs Act (also known as the Bipartisan Infrastructure
Law (BIL), P.L. 117-58) invests over $11.7 billion in the Drinking Water State Revolving Fund
(SRF) General Supplemental fund; $4 billion in the Drinking Water SRF Emerging
Contaminants fund; and $5 billion in the Emerging Contaminants in Small or Disadvantaged
Communities grant program. Together, these funds will reduce people's exposure to
perfluoroalkyl and polyfluoroalkyl substances (PFAS) and other emerging contaminants through
their drinking water.
Consistent with the EPA's Policy on Consultation and Coordination with Indian Tribes (May 4,
2011), EPA consulted with tribal officials early in the process of developing this proposed
regulation to gain an understanding of tribal views on key areas of the proposed PFAS NPDWR
and provide tribal officials an opportunity to have meaningful and timely input on its
development. For more information on the consultation with tribes, refer to the Summary Report
on Tribal Consultation: Development of the Proposed PFAS NPDWR in the public docket at
https://www.regulations.gov/docket/EPA-HQ-OW-2022-0114.
9.7 Executive Order 13045: Protection of Children from
Environmental Health and Safety Risks
Executive Order 13045 (1997), entitled "Protection of Children from Environmental Health and
Safety Risks" (62 FR 19885; April 23, 1997) applies to any rule initiated after April 21, 1998,
that (1) is determined to be "economically significant" as defined under Executive Order 12866;
and (2) concerns an environmental, health, or safety risk that the EPA has reason to believe may
have a disproportionate effect on children. If the regulatory action meets both criteria, EPA must
evaluate the environmental, health, or safety effects of the planned rule on children, and explain
why the planned regulation is preferable to other potentially effective and reasonably feasible
options considered by EPA.
The proposed rule is subject to Executive Order 13045 because it is economically significant as
defined in Executive Order 12866. This action's health and risk assessments are contained in
Section 6.2.2 of the EA, and the associated appendices. EPA expects that the proposed rule
would provide additional protection to both children and adults who consume drinking water
supplied by the affected systems. EPA also expects that the benefits of the proposed rule,
including reduced health risk, will provide significant benefits to infants and children. As
detailed in Toxicity Assessments and Proposed Maximum Contaminant Level Goals for PFOA
andPFOS in Drinking Water (U.S. EPA, 2023d; U.S. EPA, 2023e), there is evidence for adverse
effects of PFOA and PFOS for several developmental and reproductive endpoints, as well as
evidence for adverse cardiovascular, endocrine, immune, and metabolic effects in infants or
children. EPA discusses the qualitative benefits from avoided adverse health effects of PFOA,
PFOS, and other PFAS, including effects on infants and children in Section 6.2.2.2 of the EA. In
Section 6.2.2.1.1 of the EA, EPA quantifies the avoided morbidity and mortality associated with
reductions in infant birth weight from reduced maternal PFOA and PFOS exposure in drinking
Proposed PFAS Rule Economic Analysis
9-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
water. EPA also assesses the potential benefits of reduced PFNA on infant birth weight in a
sensitivity analysis found in Appendix K.
Additionally, for chemicals exhibiting a threshold for toxic effects, EPA establishes the MCLGs
based on an oral reference dose (RfD). The chronic RfD discussed in the Toxicity Assessments
and Proposed Maximum Contaminant Level Goals for PFOA and PFOS in Drinking Water (U.S.
EPA, 2023d; U.S. EPA, 2023e) provides an estimate of a daily oral exposure to the human
population (including sensitive subpopulations) that is likely to be without an appreciable risk of
deleterious non-cancer effects during a lifetime.
9.8 Executive Order 13211: Actions That Significantly Affect
Energy Supply, Distribution, or Use
Executive Order 13211 (2001), "Actions Concerning Regulations That Significantly Affect
Energy Supply Distribution, or Use," provides that agencies shall prepare and submit to the
Administrator of the Office of Information and Regulatory Affairs, OMB, a Statement of Energy
Effects for certain actions identified as "significant energy actions." Section 4(b) of Executive
Order 13211 defines "significant energy actions" as "any action by an agency (normally
published in the Federal Register) that promulgates or is expected to lead to the promulgation of
a final rule or regulation, including notices of inquiry, advance notices of proposed rulemaking,
and notices of proposed rulemaking: (l)(i) that is a significant regulatory action under Executive
Order 12866 or any successor order, and (ii) is likely to have a significant adverse effect on the
supply, distribution, or use of energy; or (2) that is designated by the Administrator of the Office
of Information and Regulatory Affairs as a significant energy action."
The proposed rule is not a "significant energy action" as defined in Executive Order 13211. This
rule is a significant regulatory action under Executive Order 12866; however, it is not likely to
have a significant adverse effect on the supply, distribution, or use of energy, for the reasons
described as follows.
9.8.1 Energy Supply
The proposed rule does not regulate power generation, either directly or indirectly, and public
and private systems subject to the proposed rule does not, as a general rule, generate power.
Further, the energy cost increases borne by customers of systems as a result of the proposed rule
is a low percentage of the total cost of water. Therefore, power generation utilities that purchase
water as part of their operations are unlikely to face any significant effects as a result of the
proposed rule.
9.8.2 Energy Distribution
The proposed rule does not regulate any aspect of energy distribution and systems that are
regulated by the proposed rule already have electrical service. The rule is not expected to
increase peak electricity demand at systems. Therefore, EPA assumes that the existing
connections are adequate and that the proposed rule has no discernible adverse effect on energy
distribution.
Proposed PFAS Rule Economic Analysis
9-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
9.8.3 Energy Use
EPA has determined that the incremental energy used to implement water treatment at drinking
water systems in response to the proposed regulatory requirements is minimal. Therefore, EPA
does not expect any noticeable effect on the national levels of power generation in terms of
average and peak loads.
9.9 National Technology Transfer and Advancement Act
Section 12(d) of the National Technology Transfer and Advancement Act (NTTAA) of 1995
directs EPA to use voluntary consensus standards in its regulatory activities unless to do so
would be inconsistent with applicable law or otherwise impractical. Voluntary consensus
standards are technical standards (e.g., materials specifications, test methods, sampling
procedures, and business practices) that are developed or adopted by voluntary consensus
standards bodies. NTTAA directs EPA to provide Congress, through OMB, explanations when
EPA decides not to use available and applicable voluntary consensus standards.
EPA's approved monitoring and sampling protocols generally include voluntary consensus
standards developed by agencies such as the American National Standards Institute (ANSI) and
other such bodies wherever EPA deems these methodologies appropriate for compliance
monitoring.
9.10 Executive Order 12898: Federal Actions to Address
Environmental Justice in Minority Populations and Low-Income
Populations, Executive Order 14008: Tackling the Climate Crisis
at Home and Abroad
Executive Order 12898 (1994), "Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations" (59 FR 7629, February 16, 1994) establishes federal
executive policy on environmental justice. Its main provision directs federal agencies, to the
greatest extent practicable and permitted by law, to make environmental justice part of their
mission. Agencies must do this by identifying and addressing as appropriate any
disproportionately high and adverse human health or environmental effects of their programs,
policies, and activities on minority populations and low-income populations in the U.S. For
information on EPA's Environmental Justice Analysis, see Chapter 8.
On March 2, 2022 and April 5, 2022, EPA held public stakeholder meetings related to EJ and the
development of the proposed NPDWR. The meetings provided an opportunity for EPA to share
information and for communities to offer input on EJ considerations related to the development
of the proposed rule. EPA received public comment on topics including establishing an MCL for
PFAS and regulating PFAS as a class, affordability of PFAS abatement options and
responsibility for remediation, limiting industrial discharge of PFAS, and EPA's relationship
with community groups. For more information on the EJ stakeholder meetings, refer to the EJ
Considerations for the Development of the Proposed PFAS Drinking Water Regulation Public
Meeting Summaries in the public docket at https://www.regulations.gov/docket/EPA-HQ-OW-
2022-0114.
Proposed PFAS Rule Economic Analysis
9-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
9.11 Consultations with the Science Advisory Board, National
Drinking Water Council, and the Secretary of Health and Human
Services
9.11.1 Science Advisory Board
As required by Section 1412(e) of the SDWA, in 2021-2022, EPA asked SAB to evaluate the
current scientific data on the following: EPA's Proposed Approaches to the Derivation of a
Draft Maximum Contaminant Level Goal for PFOA and PFOS in Drinking Water (U.S. EPA,
2021f; U.S. EPA, 2021g); a draft framework for estimated noncancer health risks associated with
mixtures of PFAS; and EPA's methodology for evaluating reduced cardiovascular disease risks.
EPA sought SAB comment on whether the analyses provided in these documents are
scientifically supported, clearly described, and informative toward supporting EPA's proposed
National Primary Drinking Water Rulemaking effort.105 The SAB PFAS Review Panel
deliberated and sought input from public meetings held in December 2021, January 2022, and
May 2022. The SAB Chartered Body conducted a quality review of the draft panel report July
2022. The SAB's final report, titled "EPA's Analyses to Support EPA's National Primary
Drinking Water Rulemaking for PFAS" was transmitted to the EPA Administrator on August 22,
2022. For information on EPA responses to SAB's review, see U.S. EPA (2022k).
9.11.2 National Drinking Water Advisory Council
In accordance with Section 1412 (d) of the SDWA, EPA consulted with NDWAC, on the
proposed rule. EPA consulted with NDWAC in a public meeting on April 19, 2022, on key areas
of the proposed rule including monitoring, treatment, public notification, and PFAS mixtures.
For more information on the consultation with the NDWAC, refer to the NDWAC Virtual Public
Meeting Summary in the public docket at https://www.regulations.gov/docket/EPA-HQ-OW-
2022-0114.
9.11.3 Secretary of Health and Human Services
In accordance with Section 1412 (d) of the SDWA, on September 28, 2022, EPA consulted with
the Department of Health and Human Services (HHS). EPA provided information to HHS
officials on the draft proposed NPDWR and considered HHS input as part of the interagency
review. A summary of this meeting is available in the docket at EPA-HQ-OW-2022-0114 at
www.regulations.gov.
9.12 Affordability Analyses
The SDWA, as amended in 1996, requires that EPA list technologies for small systems [Section
1412(b)(4)(E)(ii)]:
The Administrator shall include in the list any technology, treatment technique, or other
means that is affordable, as determined by the Administrator in consultation with the
States, for small public water systems serving -
(I) a population of 10,000 or fewer but more than 3,300;
105 For specific charge questions, visit the SAB website at
https://sab.epa.gov/ords/sab/f?p=100:19:7661444876021 :::19:P19_ID:963#charge.
Proposed PFAS Rule Economic Analysis
9-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
(II) a population of 3,300 or fewer but more than 500; and
(III) a population of 500 or fewer but more than 25;
and that achieves compliance with the maximum contaminant level (MCL) or treatment
technique, including packaged or modular systems and point-of-entry or point-of-use
treatment units (POU).
EPA's long-standing methodology for determining whether there are affordable compliance
technologies for a new drinking water standard for small systems compares the cumulative cost
of providing drinking water that complies with the new standard to an affordability threshold
equal to 2.5 percent of median household income (63 FR 42032). Should EPA determine there
are no affordable SSCTs, the SDWA Section 1412(b)(15)(B) requires EPA to identify variance
technologies that may not achieve compliance with the drinking water standard but achieve the
maximum reduction or inactivation efficiency that is affordable considering the size of the
system and the quality of the source water.
In addition to the required analysis for small system affordability, EPA is exploring the use of
alternative expenditure margins and other potential changes to the national level affordability
methodology to better understand the cost impacts of new standards on low income and
disadvantaged households served by small drinking water systems. As part of this analysis, EPA
is utilizing a number of recommendations from the SAB, NDWAC, and other stakeholders such
as the American Water Works Association (AWWA). The Agency conducted supplemental
affordability analyses using alternative metrics suggested to EPA by these advisory bodies and
stakeholders to demonstrate the potential affordability implications of the proposed NPDWR on
the determination of affordable technologies for small systems at the national level of analysis.
EPA is seeking public comment on the national level analysis of affordability of small system
compliance technologies and specifically on the potential methodologies presented.
EPA's national small system affordability determination can be found in Section 9.12.1. EPA's
supplementary affordability analyses can be found in Section 9.12.2.
9.12.1 National Small System Affordability Determination
EPA determined that there are several affordable treatment technologies for small systems. The
determination, documented in Best Available Technologies and Small System Compliance
Technologies for Per- and Polyflaoroalkyl Substances (PFAS) in Drinking Water (U.S. EPA,
2023f), compared the estimated incremental treatment costs per household with a baseline
expenditure margin that equals 2.5 percent of median household income minus baseline drinking
water utility cost per household. Table 9-7 shows which technologies satisfy the affordability
criterion for three small system size categories. For systems serving between 25-500 and 501-
3,300 people, GAC, ion exchange and point-of-use reverse osmosis are affordable technologies,
but centralized reverse osmosis is not. For systems serving 3,301-10,000 people GAC, ion
exchange and centralized RO are affordable technologies, and POU RO is not applicable to
systems of that size category.106
100 Note, the results shown in Table 9-7 and discussed in this section are dependent on the estimated annual household technology
costs reported in Table 9-9 which assumes costs associated with standard waste management of spent GAC and spent IX resin
using current typical management practices (reactivation for GAC and incineration for resin). Future changes to regulations
Proposed PFAS Rule Economic Analysis
9-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
A technology must be both effective and affordable to be designated as an SSCT. Technologies
that meet the effectiveness criterion include those designated as BATs for the proposed rule:
GAC, PFAS-selective IX, and RO. This section also presents preliminary affordability results for
Table 9-7: SSCT Affordability Analysis Results - Technologies that Meet Effectiveness
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Yes
Yes
No
Yes
501 to 3,300
Yes
Yes
Nob
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable0
Abbreviations: GAC - granular activated carbon; POU RO - Point-of-use reverse osmosis; RO - reverse osmosis; SSCT - small
system compliance technology.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered preliminary
because they reflect the costs of devices certified under the current standard, not a future standard.
bUpper bound estimated annual household treatment costs exceed expenditure margin. Tower bound estimated annual household
treatment costs do not exceed the expenditure margin.
cEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
POU RO. POU RO is not currently evaluated as a compliance option because the regulatory
options under consideration require treatment to concentrations below 70 ppt total of PFOA and
PFOS, the current certification standard for POU devices. However, POU treatment is
anticipated to become a compliance option for small systems in the future should NSF/ANSI or
another accredited third-party certification entity develop a new certification standard that
mirrors EPA's proposed regulatory standard. NSF has an update in progress today, and has noted
it will consider future updates as needed and appropriate. EPA does not anticipate additional
costs for water systems associated with the certification updating process. To evaluate
affordability, EPA compared incremental costs per household for each technology against an
expenditure margin. Table 9-8 shows the expenditure margins for each system size category. It
also shows how EPA derived the expenditure margins, beginning with estimates of MHI, which
vary by system size category. The annual affordability threshold for household expenditures on
drinking water is 2.5 percent of MHI. EPA deducted estimates of baseline or current water bills
from the affordability threshold to obtain the expenditure margin estimates.
might result in classification of spent GAC or spent IX resin as hazardous waste. EPA estimated annual cost per household if
systems are required to dispose of these residuals as hazardous waste and conducted the same national level affordability analysis
using the higher hazardous waste handling treatment costs. The Agency found the increased treatment costs for both GAC and IX
did not change the affordability conclusions. See Table 9-10 for annualized cost per household assuming hazardous waste
disposal and U.S. EPA (2023f) for the complete analysis.
Proposed PFAS Rule Economic Analysis
9-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 9-8: Expenditure Margins for SSCT Affordability Analysis
System Size
(Population
Median Household
Income3
Affordability
Thresholdb
Baseline Water
Cost0
Expenditure
Margin
Served)
A
B = 2.5% x A
C
D = B - C
25 to 500
$55,377
$1,384
$507
$877
501 to 3,300
$53,596
$1,340
$587
$753
3,301 to 10,000
$58,717
$1,468
$613
$855
Abbreviations: SSCT - small system compliance technology.
Notes:
aMHI based on U.S. Census Bureau's American Community Survey five-year estimates (U.S. Census Bureau, 2010) stated in
2010 dollars, adjusted to 2020 dollars using the CPI (for all items) for areas under 2.5 million persons.
bAffordability threshold equals 2.5 percent of MHI.
cHousehold water costs derived from 2006 Community Water System Survey (U.S. EPA, 2009), based on residential revenue
per connection within each size category, adjusted to 2020 dollars based on the Consumer Price Index for All Urban
Consumers: Water and Sewer and Trash Collection Services in U.S. City Average.
Table 9-9 provides ranges of per-household costs for each technology and system size category.
The ranges indicate minimum and maximum costs, for further information on SSCT costs, see
U.S. EPA (2023f).
Table 9-9: Total Annual Cost per Household for Candidate Technologies
System Size
(Population Served)
GAC
IX
RO
POU ROa
25 to 500
$395 to $727
$376 to $645
$3,711 to $4,676
$317 to $326
501 to 3,300
$139 to $332
$133 to $235
$608 to $1,169
$299 to $300
3,301 to 10,000
$136 to $329
$121 to $218
$326 to $462
Not applicable13
Abbreviations: GAC - granular activated carbon; IX - ion exchange; POU RO - point-of-use reverse osmosis; RO -
reverse osmosis; SSCT - small system compliance technology.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However,
POU treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or
another accredited third-party certification entity develop a new certification standard that mirrors (or is demonstrated
to treat to concentrations lower than) EPA's proposed regulatory standard. Costs presented here should be considered
preliminary estimates because they reflect the costs of devices certified under the current testing standard, not a future
standard.
bEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD
design flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
The results discussed above assume management of spent GAC and spent IX resin using current
typical management practices (reactivation for GAC and incineration for resin). EPA is in the
process of proposing some PFAS be designated as hazardous substances under CERCLA and
listed as hazardous substances constituents under the RCRA. If finalized, neither of these actions
would result in new requirements as to how PFAS containing waste, including spent GAC or
resin, is required to be managed. However, waste management facilities may, at their own
discretion, refuse to accept PFAS-containing materials or drinking water treatment operations
may choose to send spent GAC and resin containing PFAS to facilities permitted to treat and/or
Proposed PFAS Rule Economic Analysis
9-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
dispose of hazardous wastes. To consider the implications of this possibility, EPA has developed
an assessment of the current unit costs for disposing spent treatment materials and the costs
associated with their disposal as hazardous waste. Table 9-10 shows the resulting cost per
household if systems dispose of these residuals as hazardous waste. Although costs increase in
this scenario, the increases are not significant enough to change the conclusions about
affordability.
Table 9-10: Total Annual Cost per Household Assuming Hazardous Waste Disposal
System Size (Population Served)
GAC
IX
25 to 500
$417 to $827
$397 to $678
501 to 3,300
$149 to $368
$138 to $243
3,301 to 10,000
$146 to $360
$124 to $222
Abbreviations: GAC - granular activated carbon; IX - ion exchange.
9.12.2 Supplemental Affordability Analyses
In 2002, Congress required EPA to re-evaluate small system variance policy because of the
concern with the high cost of arsenic treatment in small communities. In response, in 2003, EPA
consulted with NDWAC and SAB. The SAB and NDWAC made a number of recommendations
regarding the method by which EPA evaluates the affordability of compliance with drinking
water standards.
Some key recommendations made by both the SAB and the NDWAC include:
• EPA should consider the household cost of each new regulation on an incremental basis
rather than a total cost of all water treatment regulations, and
• EPA should consider reducing the current affordability threshold, and
• financial assistance should be incorporated in the affordability calculations if the
financial support is generally available to all systems (nationwide).
In addition to the SAB and NDWAC recommendations, several additional reports by
stakeholders have offered recommendations on the improvement of EPA's affordability
methodology, including:
• The National Academy of Public Administration (NAPA) report, Developing a New
Framework for Community Affordability of Clean Water Services (NAPA, 2017),
• The National Association of Clean Water Agencies, American Water Works Association,
and Water Environment Federation report, Developing a New Framework for Household
Affordability and Financial Capability Assessment in the Water Sector (Raucher et al.,
2019), and
• The American Water Works Association expert panel report, Improving the Evaluation of
Household-Level Affordability in SDWA Rulemaking: New Approaches (AWWA,
2021)
Proposed PFAS Rule Economic Analysis
9-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
In large part, the recommendations in these reports point to the need to further assess the impacts
of new regulatory costs across income groups with a particular focus on low income and
disadvantaged communities and individuals within water systems. In particular, the American
Water Works Association (2021) expert panel report stressed that the Agency also assess the
affordability impacts to low-income households by setting the per household expenditure margin
based on the lowest quintile (20th percentile) of the income distribution.
EPA is considering the development of additional metrics to assess the impact of new regulations
on communities served by small drinking water systems and is requesting comment of the
methods which should be employed for this type of analysis. To provide commenters additional
information the Agency has estimated the impact of some potential changes to National Level
Affordability Criteria and analysis based on suggested changed from the SAB, NDWAC, and
AWWA's expert panel. In the following sub-sections, EPA estimated small system affordability
based on; (1) an incremental approach with expenditure margins of 1.0 percent of annual MHI
and 2.5 percent of the lowest quintile of annual household income, and no additional adjustment
for total current annual water expenditures, and (2) taking into account nationally available
financial assistance when assessing affordability.
9.12.2.1 Small System Affordability Analysis with Potential Additional
Expenditure Margins
As part of EPA's consideration of potential additional annual expenditure margins to improve
the assessment of affordability impacts to low income and disadvantaged communities this sub-
section provides PFAS example analyses to inform public comment. Specifically, two
incremental cost analyses are conducted utilizing alternative potential expenditure margins.
Given the recommendations from the NDWAC, the first expenditure margin threshold is based
on 1.0 percent of annual MHI. The second expenditure margin threshold is set equal to 2.5
percent of the lowest quintile of annual household income and is based on the American Water
Works Association (2021) expert panel report. These expenditure margins are estimated for each
of the small system size categories: 25 to 500, 501 to 3,300, and 3,301 to 10,000 people served.
As this is an incremental analysis no additional adjustments are made to the values to account for
current annual drinking water cost. Table 9-11 shows the calculated annual expenditure margins
by system size.
Proposed PFAS Rule Economic Analysis
9-24
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
Table 9-11: Potential Annual Expenditure Margins for SSCT Affordability Analysis
System Size (Population Served)
1.0% of Median Household
Income3
2.5% of Lowest Quintile
Incomeb
A
B
25 to 500
$554
$643
501 to 3,300
$536
$628
3,301 to 10,000
$587
$681
Abbreviations: SSCT - small system compliance technology.
Notes:
aMHI based on U.S. Census Bureau's American Community Survey five-year estimates (U.S. Census Bureau, 2010) stated in
2010 dollars, adjusted to 2020 dollars using the CPI (for all items) for areas under 2.5 million persons.
bLowest quintile (20th percentile) household income based on U.S. Census 2010 American Community Survey 5-year
estimates (U.S. Census Bureau, 2010) stated in 2010 dollars, adjusted to 2020 dollars using the CPI (for all items) for areas
under 2.5 million persons.
Given these alternative annual expenditure margins the remainder of the assessment process is
the same as EPA's current small system affordability methodology. The estimated total annual
household costs for each of the deemed efficient treatment technologies presented in Table 9-9
are compared against the estimated annual expenditure margin thresholds from Table 9-11 for
each system size category. Table 9-12 presents the affordability results using the 1.0 percent of
annual MHI expenditure margin and Table 9-13 provides the information when the 2.5 percent
of the lowest quintile of annual household income is used as the threshold.
Table 9-12: Affordability Analysis Results Using a 1.0% of Annual Median Household
Income Expenditure Margin
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Nob
Nob
No
Yes
501 to 3,300
Yes
Yes
No
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable0
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
bUpper bound estimated annual household treatment costs exceed expenditure margin. Tower bound estimated annual
household treatment costs do not exceed the expenditure margin.
cEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
Proposed PFAS Rule Economic Analysis
9-25
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 9-13: Affordability Analysis Results Using a 2.5% of Lowest Quintile of Annual
Household Income Expenditure Margin
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Nob
Nob
No
Yes
501 to 3,300
Yes
Yes
Nob
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable0
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
The results in both Table 9-12 and Table 9-13, which utilize the potential additional expenditure
margins, of 1.0 percent of annual MHI and 2.5 percent of the lowest quintile of annual household
income, and the results of EPA's national level affordability analysis in Table 9-3, which utilizes
a household expenditure margin estimated by adjusting 2.5 percent of median household income
minus baseline median annual drinking water costs, differ in the case of GAC and IX for systems
serving 25 to 500 people. As indicated by the "Nob" reported in Table 9-12 and Table 9-13 for
GAC and IX the upper bound annual household treatment cost for both these technologies
exceed both the 1.0 percent of annual MHI and 2.5 percent of the lowest quintile of annual
household income expenditure margins, however, the estimated lower bound annual household
treatment costs do not exceed the expenditure margins. The alternative expenditure margins also
changed the affordability results for RO in the 501-3,300 system size category. In the national
affordability analysis using the 2.5 percent of MHI with baseline adjustment upper bound RO
annual household cost estimates exceed the expenditure margin but the lower bound costs do not.
When using both the 1.0 percent of annual MHI and 2.5 percent of the lowest quintile of annual
household income potential criteria both the high and low bound estimated annual household
treatment costs exceed the expenditure margins.107
9.12.2.2 Small System Affordability Analysis When Accounting for
Financial Assistance
The SAB and NDWAC recommended to EPA that the national level affordability analysis
should include the impact of financial assistance if the financial support is generally available to
all systems (nationwide). EPA is considering including this recommendation in the national
affordability calculations. The recommendations themselves indicate a two-step process; (1)
107 Note, the results shown in Table 9-12 and Table 9-13 and discussed in this section are dependent on the estimated annual
household technology costs reported in Table 9-9 which assumes costs associated with standard waste management of spent
GAC and spent IX resin using current typical management practices (reactivation for GAC and incineration for resin). Future
changes to regulations might result in classification of spent GAC or spent IX resin as hazardous waste. EPA estimated annual
cost per household if systems are required to dispose of these residuals as hazardous waste and conducted the same national level
affordability analyses with the 1.0 percent of MHI and 2.5 percent of the lowest quintile of annual household income expenditure
margins and using the higher hazardous waste handling treatment costs. The Agency found the increased treatment costs for both
GAC and IX did not change the affordability conclusions.
Proposed PFAS Rule Economic Analysis
9-26
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
determine if and how much financial assistance is available to small systems on a national level
for compliance with a specific rule, in this case the DW PFAS rule, and (2) calculate the
potential impact of the financial assistance on the estimated per household treatment costs for
each of the small system size categories.
On the national level, significant financial assistance is available to small systems for the
installation of PFAS treatment technology. One critical and long-established source of this
assistance is available through EPA's Drinking Water State Revolving Fund (DWSRF) Program
that was authorized by Congress as part of the 1996 Amendments to the Safe Drinking Water
Act. The DWSRF's purpose is to provide a source of financial assistance to water systems and
states to help them achieve the public health protection objectives of SDWA. A unique feature of
the DWSRF Program is that it is state based. EPA awards capitalization grants to states who
provide a 20 percent match, creating a dedicated fund from which loans are made to water
systems and into which the loan repayments (and interest) are deposited so they can be loaned
out again. Within some broad statutory constraints contained in SDWA, the states have
considerable flexibility to tailor the DWSRF Program to their own unique needs and
circumstances.
The SDWA established three criteria at the core of the process used by states in ranking projects
in priority order to receive funding. States are required, to the maximum extent practicable, to
give priority for the use of DWSRF funds to projects that:
1. Address the most serious risk to human health;
2. Are necessary to ensure compliance with SDWA requirements; and
3. Assist systems most in need on a per household basis according to state affordability
criteria.
Thus, system level affordability, according to state affordability criteria, is a central
consideration in ranking projects eligible to receive DWSRF assistance. Each state has
developed, and EPA has approved, a project priority ranking procedure. The specific weight
given to affordability considerations vis-a-vis public health and SDWA compliance
considerations varies from state to state. States are required to include their project priority
ranking system as part of the Intended Use Plan they are required to develop in support of their
application for each capitalization grant. The Intended Use Plan must contain both the project
priority ranking system and the priority list of projects eligible for DWSRF assistance. The state
must provide notice and opportunity for public comment on the priority list of projects.
Under the core DWSRF Program, the state may establish an interest rate between zero percent
and the market rate. The lower the interest rate, the greater the subsidy provided to the borrower.
SDWA requires states to establish a Disadvantaged Communities Program within their DWSRF
under which communities considered disadvantaged according to state developed affordability
criteria could receive additional subsidies beyond a zero percent loan. These additional subsidies
often take the form of principal forgiveness (i.e., loan forgiveness) or grants. There is no limit to
the amount of additional subsidy that can be provided to a particular project except for an overall
limit on the total amount of additional subsidy of 35 percent of the state's annual capitalization
grant.
This additional subsidization could be directed entirely to a few projects, essentially making the
assistance those projects receive equivalent to a 100 percent grant; or the additional subsidization
Proposed PFAS Rule Economic Analysis
9-27
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
could be distributed among a larger number of projects and combined with zero or low-interest
loans. States may also offer communities they consider disadvantaged108 a loan term of 40 years
rather than the base period of 20 to 30 years. Notably, the loan term cannot extend beyond the
design life of the capital improvement constructed via the DWSRF loan.
The SDWA provided EPA with the authority to publish information to assist states in
establishing affordability criteria for purposes of a disadvantaged community program. The
Agency worked with a group of expert stakeholders and published "Information for States on
Developing Affordability Criteria for Drinking Water" (document number 816-R-98-002) in
February 1998 (U.S. EPA, 1998b). The Agency provided additional information to assist states'
affordability criteria development in the "Implementation of the Clean Water and Drinking
Water State Revolving Fund Provisions of the Bipartisan Infrastructure Law" memorandum in
March 2022 (U.S. EPA, 2022d).
PFAS drinking water treatment loans and grants have been and will continue to be available to
systems of all sizes under the traditional DWSRF program funding and allocation structure. In
addition to these funding sources, on November 15, 2021, the Bipartisan Infrastructure Law
(BIL) (P.L. 117-58), also known as the "Infrastructure Investment and Jobs Act of 2021" (IIJA)
appropriated $4 billion over 5 years ($800,000,000 per year) for projects that are DWSRF
eligible whose primary purpose must be to address emerging contaminants, with a focus on
PFAS. EPA expects to establish a NPDWR for PFOA and PFOS. The Agency is also evaluating
additional PFAS and groups of PFAS. Given stated Congressional intent of this appropriation,
PFAS-focused projects will be eligible for funding under this appropriation regardless of whether
EPA has established a NPDWR for that particular PFAS or group of PFAS. These BIL funds
must be distributed to communities entirely as forgivable loans or grants, and states are not
required to provide matching funds as with most DWSRF projects. 25 percent of this BIL
funding is targeted toward disadvantaged communities and/or communities fewer than or equal
to 25,000 people.
In addition to the DWSRF BIL funds, as part of a government-wide effort to confront PFAS
pollution, the BIL authorizes $5 billion as part of the Emerging Contaminants in Small or
Disadvantaged Communities grant program that can be used to reduce PFAS in drinking water in
communities facing disproportionate impacts. The goal of the Emerging Contaminants in Small
or Disadvantaged Communities grant program is for states to provide grants to public water
systems in small or disadvantaged communities to address emerging contaminants, including
PFAS. Funding will be provided to participating states and territories to benefit small or
disadvantaged communities in scoping, planning, testing and remediating emerging
contaminants in drinking and source water.
These funds can be used in small or disadvantaged communities to address emerging
contaminants like PFAS in drinking water through actions such as technical assistance, water
quality testing, contractor training, and installation of centralized treatment technologies and
systems. On June 15, 2022, EPA announced that it is making $1 billion available in FY2022 of a
total of $5 billion for fiscal years 2022-2026.
108 Disadvantaged community is defined as the service area of a public water system that meets affordability criteria established
after public review and comment by the State in which the public water system is located.
Proposed PFAS Rule Economic Analysis
9-28
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Given the BIL emerging contaminant funding being made available through the DWSRF and the
Emerging Contaminants in Small or Disadvantaged Communities grant program, EPA expects
that most small systems will have access to financial assistance for PFAS related capital
expenditures. EPA estimates that the total amount of capital treatment technology expenditures
for small systems nationally ranges between approximately $1.1 and $2.5 billion. EPA expects
funding from BIL to be more than sufficient to cover the capital costs for small systems. Hence,
it seems reasonable to consider these funds for the purposes of illustrating the potential impact of
including financial assistance in the calculation of the national level affordability assessment for
small system compliance technologies. Because BIL funds are limited to providing grants and
loan forgiveness associated with PFAS drinking water treatment capital expenditures, EPA in
this example zeroed out only the capital cost of the candidate effective technologies. The annual
per household treatment cost ranges presented in Table 9-9 represent operations and maintenance
costs for the technologies by small system size category. Comparing the cost ranges in Table
9-14 with unadjusted cost ranges in Table 9-9 demonstrates the potential large decrease in
technology cost when financial assistance is considered. The decreases across technologies and
system size categories range from 28 percent to 76 percent.
Table 9-14: Annual Cost per Household for Candidate Technologies Assuming 100%
Financial Assistance for Technology Capital Costs
System Size (Population
Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
$125 to $198
$133 to $153
$1,081 to $1,153
$225 to $234
501 to 3,300
$51 to $122
$57 to $76
$252 to $309
$209 to $210
3,301 to 10,000
$57 to $129
$58 to $84
$167 to $193
Not applicable13
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
bEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
Table 9-15, Table 9-16, and Table 9-17 below show the affordability results utilizing the 2.5
percent of annual MHI minus the baseline median annual drinking water cost, the incremental
1.0 percent of annual MHI, and using the 2.5 percent of the lowest quintile of annual household
income expenditure margins, respectively. Given the significant reduction in estimated per
household annual treatment costs for GAC and ion exchange, the technologies were found to
satisfy the national level affordability criterion for the three statutorily mandated small system
size categories.109 Centralized RO with high per household operations and maintenance costs, of
109 Note, the results shown in Table 9-15, Table 9-16 and Table 9-17 and discussed in this section are dependent on the estimated
annual household technology costs reported in Table 9-14 which assumes operations and maintenance costs associated with
standard waste management of spent GAC and spent IX resin using current typical management practices (reactivation for GAC
and incineration for resin). Future changes to regulations might result in classification of spent GAC or spent IX resin as
Proposed PFAS Rule Economic Analysis
9-29
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
$1,081 to $1,153, in the system size category of 25-500 people served was found to be
unaffordable in that system size category across all alternative expenditure margins, but
economies of scale reduce per household costs in systems serving between 501 and 10,000
people sufficiently to approve the technology as affordable under the three alternative
expenditure margins. POU RO was also found to be affordable at the national level of analysis
for systems serving 25 to 500 and 501 to 3,300 people across the three presented expenditure
margins. POU RO is not applicable to systems serving more than 3,300 people given the
increasing complexity of managing POU programs at such large scales.
Table 9-15: Affordability Analysis Results Using a 2.5% of Annual Median Household
Income Minus the Baseline Median Annual Drinking Water Cost Expenditure Margin
and assuming 100% Financial Assistance for Technology Capital Costs
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Yes
Yes
No
Yes
501 to 3,300
Yes
Yes
Yes
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable13
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
bEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
hazardous waste. EPA estimated annual operations and maintenance cost per household if systems are required to dispose of
these residuals as hazardous waste and conducted the same national level affordability analyses using the three alternative
expenditure margins using the higher hazardous waste handling treatment costs. The Agency found the increased treatment costs
for both GAC and IX did not change the affordability conclusions.
Proposed PFAS Rule Economic Analysis
9-30
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Table 9-16: Affordability Analysis Results Using a 1.0% of Annual Median Household
Income Expenditure Margin and assuming 100% Financial Assistance for Technology
Capital Costs
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Yes
Yes
No
Yes
501 to 3,300
Yes
Yes
Yes
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable13
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
bEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
Table 9-17: Affordability Analysis Results Using a 2.5% of Lowest Quintile of Annual
Household Income Expenditure Margin and assuming 100% Financial Assistance for
Technology Capital Costs
System Size (Population Served)
GAC
Ion Exchange
RO
POU ROa
25 to 500
Yes
Yes
No
Yes
501 to 3,300
Yes
Yes
Yes
Yes
3,301 to 10,000
Yes
Yes
Yes
Not applicable13
Abbreviations: GAC - granular activated carbon; POU RO - point-of-use reverse osmosis; and RO - reverse osmosis.
Notes:
aPOU RO is not currently a compliance option because the regulatory options under consideration require treatment to
concentrations below 70 ppt total of PFOA and PFOS, the current certification standard for POU devices. However, POU
treatment is anticipated to become a compliance option for small systems in the future should NSF/ANSI or another accredited
third-party certification entity develop a new certification standard that mirrors (or is demonstrated to treat to concentrations
lower than) EPA's proposed regulatory standard. The affordability conclusions presented here should be considered
preliminary because they reflect the costs of devices certified under the current standard, not a future standard.
bEPA's WBS model for POU treatment does not cover systems serving more than 3,300 people (greater than 1 MGD design
flow), because implementing and maintaining a large-scale POU program is likely to be impractical.
Proposed PFAS Rule Economic Analysis
9-31
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
10 References
Aassve, A., Cavalli, N., Mencarini, L., Plach, S., & Sanders, S. (2021). Early assessment of the
relationship between the COVID-19 pandemic and births in high-income countries.
Proceedings of the National Academy of Sciences, 118(36).
Abt Associates. (2020). Labor Costs for National Drinking Water Rides. Submitted to Rajiv
Khera, U.S. Environmental Protection Agency, Office of Ground Water and Drinking
Water. EPA Contract No. EP-B16C-0001. Retrieved from
Adamson, D. T., Pina, E. A., Cartwright, A. E., Rauch, S. R., Anderson, R. H., Mohr, T., &
Connor, J. A. (2017). 1, 4-Dioxane drinking water occurrence data from the third
unregulated contaminant monitoring rule. Science of the total environment, 596, 236-245.
Agency for Toxic Substances and Disease Registry. (2018). Toxicological Profile for
Perflaoroalkyls.
Agency for Toxic Substances and Disease Registry. (2021). Toxicological Profile for
Perflaoroalkyls.
Almond, D., Chay, K. Y., & Lee, D. S. (2005). The costs of low birth weight. The Quarterly
journal of economics, 120(3), 1031-1083. doi: 10.1093/qj e/120.3.1031
Almond, D., Doyle Jr, J. J., Kowalski, A. E., & Williams, H. (2010). Estimating marginal returns
to medical care: Evidence from at-risk newborns. The Ouarterly journal of economics,
125(2), 591-634.
Ambavane, A., Yang, S., Atkins, M. B., Rao, S., Shah, A., Regan, M. M., . . . Michaelson, M. D.
(2020). Clinical and economic outcomes of treatment sequences for intermediate- to
poor-risk advanced renal cell carcinoma. Immunotherapy, 12(1), 37-51. doi:10.2217/imt-
2019-0199
American Association for the Advancement of Science. (2020). Per- and Polyfluoroalkyl
Substances (PFAS) in Drinking Water. Retrieved from https://www.aaas.org/epi-
center/pfas
American Water Works Association. (2021). Improving the Evaluation of Household-Level
Affordability in SDWA Rulemaking: New Approaches.
Anderson, R. N. (1999). Methodfor constructing complete annual US life tables (Vol. 129):
National Ctr for Health Statistics.
Appleman, T. D., Higgins, C. P., Quinones, O., Vanderford, B. J., Kolstad, C., Zeigler-Holady, J.
C., & Dickenson, E. R. (2014). Treatment of poly-and perfluoroalkyl substances in US
full-scale water treatment systems. Water research, 51, 246-255.
Aquafina. (2022). Aquafina FAQ. Retrieved from https://www.aquafina.com/en-
US/faq.html#:~:text=Aquafma%20originates%20from%20public%20water.can%20affect
%20a%20water' s%20taste
Arias, E., Heron, M., & Tejada-Vera, B. (2013). United States life tables eliminating certain
causes of death, 1999-2001.
Arias, E., & Xu, J. (2019). United States Life Tables, 2017.
Bach, C. C., Vested, A., Jorgensen, K. T., Bonde, J. P. E., Henriksen, T. B., & Toft, G. (2016).
Perfluoroalkyl and polyfluoroalkyl substances and measures of human fertility: a
systematic review. Critical reviews in toxicology, 46(9), 735-755.
Bach, J. P., Riedel, O., Pieper, L., Klotsche, J., Dodel, R., & Wittchen, H. U. (2011). Health-
related quality of life in patients with a history of myocardial infarction and stroke.
Cerebrovascular Diseases. 31(1), 68-76.
Proposed PFAS Rule Economic Analysis
10-1
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Banzhaf, S., Ma, L., & Timmins, C. (2019). Environmental Justice: The Economics of Race,
Place, and Pollution. The Journal of Economic Perspectives: A Journal of the American
Economic Association, 33(1), 185-208.
Barry, V., Winquist, A., & Steenland, K. (2013). Perfluorooctanoic Acid (PFOA) Exposures and
Incident Cancers Among Adults Living Near A Chemical Plant. Environmental health
perspectives, 727(11-12), 1313-1318. doi:https://doi.org/10.1289/ehp. 1306615
Bartell, S. M., Calafat, A. M., Lyu, C., Kato, K., Ryan, P. B., & Steenland, K. (2010). Rate of
decline in serum PFOA concentrations after granular activated carbon filtration at two
public water systems in Ohio and West Virginia. Environmental health perspectives,
775(2), 222-228.
Bartell, S. M., & Vieira, V. M. (2021). Critical review on PFOA, kidney cancer, and testicular
cancer. J Air Waste Manag Assoc, 77(6), 663-679. doi: 10.1080/10962247.2021.1909668
Barton, K. E., Starling, A. P., Higgins, C. P., McDonough, C. A., Calafat, A. M., & Adgate, J. L.
(2020). Sociodemographic and Behavioral Determinants of Serum Concentrations of Per-
and Polyfluoroalkyl Substances in A Community Highly Exposed to Aqueous Film-
Forming Foam Contaminants in Drinking Water. Int JHyg Environ Health, 223(1), 256-
266. doi: 10.1016/j.ijheh.2019.07.012
Beatty, A. L., Ku, I. A., Bibbins-Domingo, K., Christenson, R. H., DeFillippi, C. R., Ganz, P., . .
. Whooley, M. A. (2015). Traditional risk factors versus biomarkers for prediction of
secondary events in patients with stable Coronary Heart Disease: from the Heart and Soul
Study. Journal of the American Heart Association, 4. doi: 10.1161/JAHA. 114.001646
Behrman, J. R., & Rosenzweig, M. R. (2004). Returns to birthweight. Review of Economics and
Statistics, 86(2), 586-601.
Behrman, R. E., & Butler, A. S. (2007). Preterm birth: causes, consequences, and prevention.
Institute of Medicine of the National Academies.
Black, J., Moreland, A., McNeill Ransom, M., & Sanchez, E. (2021). Perfluoroalkyl and
Polyfluoroalkyl Substances: Using Law and Policy to Address These Environmental
Health Hazards in the United States. Health Matrix, 31, 340.
Bonefeld-Jorgensen, E. C., Long, M., Fredslund, S. O., Bossi, R., & Olsen, J. (2014). Breast
cancer risk after exposure to perfluorinated compounds in Danish women: A case-control
study nested in the Danish National Birth Cohort. Cancer Causes Control, 25(11), 1439-
1448.
Boronow, K. E., Brody, J. G., Schaider, L. A., Peaslee, G. F., Havas, L., & Cohn, B. A. (2019).
Serum concentrations of PFASs and exposure-related behaviors in African American and
non-Hispanic white women. Journal of exposure science & environmental epidemiology,
29(2), 206-217.
Brown, P. (1995). Race, class, and environmental health: A review and systematization of the
literature. Environmental research, 69(1), 15-30.
doi:https://doi.org/10.1006/enrs. 1995.1021
Brulle, R. J., & Pellow, D. N. (2006). Environmental justice: Human health and environmental
inequalities. Annual Review of Public Health, 27, 103-124.
doi: http s: //doi. or g/10.1146/annurev. publhealth .27.021405.102124
Brunke, M., & Gonser, T. (1997). The ecological significance of exchange processes between
rivers and groundwater. Freshwater biology, 57(1), 1-33.
Buekers, J., Colles, A., Cornelis, C., Morrens, B., Govarts, E., & Schoeters, G. (2018). Socio-
Economic Status and Health: Evaluation of Human Biomonitored Chemical Exposure to
Proposed PFAS Rule Economic Analysis
10-2
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Per- and Polyfluorinated Substances across Status. International journal of environmental
research and public health, 75(12). doi:10.3390/ijerphl5122818
Cadwallader, A., A. Greene, H. Holsinger, A. Lan, M. Messner, M. Simic, & Albert, R. (2022).
A Bayesian hierarchical model for estimating national PFAS drinking water occurrence.
A WW A Water Science, el284.
Calafat, A. M., Kato, K., Hubbard, K., Jia, T., Botelho, J. C., & Wong, L.-Y. (2019). Legacy and
alternative per- and polyfluoroalkyl substances in the U.S. general population: Paired
serum-urine data from the 2013-2014 National Health and Nutrition Examination
Survey. Environment international, 131, 105048.
Calafat, A. M., Wong, L.-Y., Kuklenyik, Z., Reidy, J. A., & Needham, L. L. (2007).
Polyfluoroalkyl Chemicals in the U.S. population: Data from the National Health and
Nutrition Examination Survey (NHANES) 2003-2004 and Comparisons with NHANES
1999-2000. Environmental health perspectives, 775(11), 1596-1602.
California Environmental Protection Agency. (2021). Proposed Public Health Goals for
Perflaorooctanoic Acid and Perflaorooctane Sulfonic Acid in Drinking Water.
Cantor, K. P., Lynch, C. F., Hildesheim, M. E., Dosemeci, M., Lubin, J., Alavanja, M., & Craun,
G. (1998). Drinking water source and chlorination byproducts. I. Risk of bladder cancer.
Epidemiology (Cambridge, Mass.), 9(1), 21-28. doi:http://dx.doi.org/10.1097/00001648-
199801000-00007
Cantor, K. P., Villanueva, C. M., Silverman, D. T., Figueroa, J. D., Real, F. X., Garcia-Closas,
M., . . . Tardon, A. (2010). Polymorphisms in GSTT1, GSTZ1, and CYP2E1, disinfection
by-products, and risk of bladder cancer in Spain. Environmental health perspectives,
775(11), 1545-1550.
Catelan, D., Biggeri, A., Russo, F., Gregori, D., Pitter, G., Da Re, F., . . . Canova, C. (2021).
Exposure to perfluoroalkyl substances and mortality for COVID-19: A spatial ecological
analysis in the Veneto region (Italy). International journal of environmental research and
public health, 75(5), 2734.
Centers for Disease Control and Prevention. (2017). User Guide to the 2017 Period 2016 Cohort
Linked Birth Infant Death Public Use File.
Centers for Disease Control and Prevention. (2018). User Guide to the 2018 Period 2017 Cohort
Linked Birth Infant Death Public Use File.
Centers for Disease Control and Prevention. (2019). Fourth National Report on Human
Exposure to Environmental Chemicals: Updated Tables, January 2019, Volume One.
Retrieved from cdc_75822_DSl.pdf
Centers for Disease Control and Prevention. (2020a). National Vital Statistics System, Linked
Birth Infant Death Records 1995-2019 on CDC WONDER Online Database. Data are
fi'om the Linked Birth Infant Deaths Records 1995-2019, as compiledfi'om data
provided by the 57 vital statistics jurisdictions through the Vital Statistics Cooperative
Program. Retrieved from: https://wonder.cdc.gov/lbd-current.html
Centers for Disease Control and Prevention. (2020b). Underlying Cause of Death, 1999-2020 on
CDC WONDER Online Database, released in 2020. Data are fi'om the Multiple Cause of
Death Files, 1999-2019, as compiled from data provided by the 57 vital statistics
jurisdictions through the Vital Statistics Cooperative Program. Retrieved from:
http://wonder.cdc.gov/ucd-icdlO.html
Chaikind, S., & Corman, H. (1991). The impact of low birthweight on special education costs.
Journal of Health Economics, 70(3), 291-311.
Proposed PFAS Rule Economic Analysis
10-3
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Chateau-Degat, M.-L., Pereg, D., Dallaire, R., Ayotte, P., Dery, S., & Dewailly, E. (2010).
Effects of perfluorooctanesulfonate exposure on plasma lipid levels in the Inuit
population of Nunavik (Northern Quebec). Environmental research, 110(7), 710-717.
Chatteiji, P., Kim, D., & Lahiri, K. (2014). Birth weight and academic achievement in childhood.
Health Economics, 23(9), 1013-1035.
Chen, C. W., & Gibb, H. (2003). Procedures for calculating cessation lag. Regulatory Toxicology
and Pharmacology, 38(2), 157-165.
Chen, M. H., Ng, S., Hsieh, C. J., Lin, C. C., Hsieh, W. S., & Chen, P. C. (2017). The impact of
prenatal perfluoroalkyl substances exposure on neonatal and child growth. Science of the
total environment, 607-608, 669-675. doi:10.1016/j.scitotenv.2017.06.273
Cheng, W., Dastgheib, S. A., & Karanfil, T. (2005). Adsorption of dissolved natural organic
matter by modified activated carbons. Water research, 59(11), 2281-2290.
Chili, C. A., Westerhoff, P., & Ghosh, A. (2012). GAC removal of organic nitrogen and other
DBP precursors. Journal-American Water Works Association, 104(1), E406-E415.
Chin, D. A., & Qi, X. (2000). Ground water under direct influence of surface water. Journal of
Environmental Engineering, 126(6), 501-508.
Chowdhury, Z. Z., Zain, S. M., Rashid, A. K., Rafique, R. F., & Khalid, K. (2013). Breakthrough
Curve Analysis for Column Dynamics Sorption of Mn(II) Ions from Wastewater by
Using Mangostana garcinia Peel-Based Granular-Activated Carbon. Journal of
Chemistry, 2013, 959761. doi: 10.1155/2013/959761
Chu, C., Zhou, Y., Li, Q.-Q., Bloom, M. S., Lin, S., Yu, Y.-J., . . . Yang, B.-Y. (2020). Are
perfluorooctane sulfonate alternatives safer? New insights from a birth cohort study.
Environment international, 135, 105365.
Colaizy, T. T., Bartick, M. C., Jegier, B. J., Green, B. D., Reinhold, A. G., Schaefer, A. J., . . .
Jobe, A. H. (2016). Impact of optimized breastfeeding on the costs of necrotizing
enterocolitis in extremely low birthweight infants. The Journal of Pediatrics, 175, 100-
105.el02.
Colantonio, L. D. M., PhD; Rosenson, Robert S. MD; Deng, Luqin PhD; Monda, Keri L. PhD;
Dai, Yuling MS, MSPH; Farkouh, Michael E. MD, MSc; Safford, Monika M. MD;
Philip, Kiran MD, MBA; Mues, Katherine E. PhD, MPH; Muntner, Paul PhD. (2019).
Adherence to Statin Therapy Among US Adults Between 2007 and 2014. Journal of the
American Heart Association, 5(1). doi:https://doi.org/10.1161/JAHA.l 18.010376
Collier, S. A., Deng, L., Adam, E. A., Benedict, K. M., Beshearse, E. M., Blackstock, A. J., . . .
Fullerton, K. E. (2021). Estimate of burden and direct healthcare cost of infectious
waterborne disease in the United States. Emerging infectious diseases, 27(1), 140.
Costet, N, Villanueva, C. M., Jaakkola, J. J. K., Kogevinas, M., Cantor, K. P., King, W. D., . . .
Cordier, S. (2011). Water disinfection by-products and bladder cancer: is there a
European specificity? A pooled and meta-analysis of European case-control studies.
Occupational and environmental medicine, 68(5), 379-385.
Crittenden, J. C., Vaitheeswaran, K, Hand, D. W., Howe, E. W., Aieta, E. M., Tate, C. H., . . .
Davis, M. K. (1993). Removal of dissolved organic carbon using granular activated
carbon. Water research, 27(4), 715-721. doi:https://doi.org/10.1016/0043-
1354(93)90181-G
Cropper, M. L., & Krupnick, A. J. (2000). Social costs of chronic heart and lung disease.
Cuthbertson, A. A., Kimura, S. Y., Liberatore, H. K, Summers, R. S., Knappe, D. R., Stanford,
B. D., . . . Richardson, S. D. (2019). Does granular activated carbon with chlorination
Proposed PFAS Rule Economic Analysis
10-4
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
produce safer drinking water? From disinfection byproducts and total organic halogen to
calculated toxicity. Environmental Science & Technology, 53(10), 5987-5999.
D'Agostino, R. B., Vasan, R. S., Pencina, M. J., Wolf, P. A., Cobain, M., Massaro, J. M., &
Kannel, W. B. (2008). General cardiovascular risk profile for use in primary care.
Circulation, 777(6), 743-753.
Darrow, L. A., Stein, C. R., & Steenland, K. (2013). Serum Perfluorooctanoic Acid and
Perfluorooctane Sulfonate Concentrations in Relation to Birth Outcomes in the Mid-Ohio
Valley, 2005-2010. Environ. Health Persped., 727(10), 1207-1213.
doi: 10.1289/ehp. 1206372
Dastgheib, S. A., Karanfil, T., & Cheng, W. (2004). Tailoring activated carbons for enhanced
removal of natural organic matter from natural waters. Carbon, 42(3), 547-557.
de Abreu Domingos, R., & da Fonseca, F. V. (2018). Evaluation of adsorbent and ion exchange
resins for removal of organic matter from petroleum refinery wastewaters aiming to
increase water reuse. Journal of environmental management, 214, 362-369.
Deeks, J. J. (2002). Issues in the selection of a summary statistic for meta-analysis of clinical
trials with binary outcomes. Statistics in medicine, 21(11), 1575-1600.
Desikan, A., Carter, J., Kinser, S., & Goldman, G. (2019). Abandoned Science, Broken
Promises: How the Trump Administration's Neglect of Science Is Leaving Marginalized
Communities Further Behind.
Dickenson, E., & Higgins, C. (2016). Treatment Mitigation Strategies for Poly- and
Perfluorinated Chemicals. Water Research Foundation.
Dobson, K. G., Ferro, M. A., Boyle, M. H., Schmidt, L. A., Saigal, S., & Van Lieshout, R. J.
(2018). How do childhood intelligence and early psychosocial adversity influence income
attainment among adult extremely low birth weight survivors? A test of the cognitive
reserve hypothesis. Development andpsychopathology, 30(4), 1421-1434.
Domingo, J. L., & Nadal, M. (2019). Human exposure to per-and polyfluoroalkyl substances
(PFAS) through drinking water: A review of the recent scientific literature.
Environmental research, 177, 108648.
Dong, Z., Wang, H., Yu, Y. Y., Li, Y. B., Naidu, R., & Liu, Y. (2019). Using 2003-2014 US
NHANES data to determine the associations between per-and polyfluoroalkyl substances
and cholesterol: Trend and implications. Ecotoxicology and environmental safety, 173,
461-468.
Driscoll, A. K., & Gregory, E. C. W. (2021). Prepregnancy Body Mass Index and Infant
Outcomes by Race and Hispanic Origin: United States, 2020. Natl Vital Stat Rep, 70(16),
1-8.
Du, Z., Deng, S., Bei, Y., Huang, Q., Wang, B., Huang, J., & Yu, G. 2014. . (2014). Adsorption
behavior and mechanism of perfluorinated compounds on various adsorbents—A review.
Journal of hazardous materials, 274, 443-454.
Dzierlenga, M., Crawford, L., & Longnecker, M. (2020). Birth weight and perfluorooctane
sulfonic acid: a random-effects meta-regression analysis. Environmental Epidemiology,
4(3).
Eick, S. M., Enright, E. A., Geiger, S. D., Dzwilewski, K. L., DeMicco, E., Smith, S., . . .
Morello-Frosch, R. (2021). Associations of maternal stress, prenatal exposure to per-and
polyfluoroalkyl substances (PFAS), and demographic risk factors with birth outcomes
and offspring neurodevelopment: an overview of the ECHO. CA. IL prospective birth
cohorts. International journal of environmental research and public health, 18(2), 742.
Proposed PFAS Rule Economic Analysis
10-5
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Elder, T., Figlio, D., Imberman, S., & Persico, C. (2020). The role of neonatal health in the
incidence of childhood disability. American Journal of Health Economics, 6(2), 216-250.
Electricwala, B., Wassel, C. L., Amari, D. T., & Cristino, J. (2020). PCV73 Risk Factors
Associated with Statin Use Among Adult Nhanes Participants with Ascvd. Value in
Health, 23, S499. doi:10.1016/j.jval.2020.08.564
Ellington, S., Strid, P., Tong, V. T., Woodworth, K., Galang, R. R., Zambrano, L. D., . . . Gilboa,
S. M. (2020). Characteristics of women of reproductive age with laboratory-confirmed
SARS-CoV-2 infection by pregnancy status—United States, January 22-June 7, 2020.
Morbidity and Mortality Weekly Report, 69(25), 769.
Ellis, J. J., Erickson, S. R., Stevenson, J. G., Bernstein, S. J., Stiles, R. A., & Fendrick, A. M.
(2004). Suboptimal statin adherence and discontinuation in primary and secondary
prevention populations. Journal of General Internal Medicine, 19(6), 638-645.
doi :10.1111/j.l 525-1497.2004.30516.x
Engels, E. A., Schmid, C. Ft., Terrin, N., Olkin, I, & Lau, J. (2000). Heterogeneity and statistical
significance in meta-analysis: an empirical study of 125 meta-analyses. Statistics in
medicine, 79(13), 1707-1728.
Environmental Council of the States. (2022). Processes and Considerations for Setting State
PFAS Standards.
Eschauzier, C., Beerendonk, E., Scholte-Veenendaal, P., & De Voogt, P. (2012). Impact of
treatment processes on the removal of perfluoroalkyl acids from the drinking water
production chain. Environmental Science & Technology, -16(3), 1708-1715.
Executive Order 12866. 1993. Regulatory Planning and Review. 58 FR 51735, October 4, 1993.
Available at https://www.reginfo.gov/public/isp/Utilities/EO 12866.pdf.. (1993).
Executive Order 12898. (1994). Federal Actions to Address Environmental Justice in Minority
Populations and Low -Income Populations. 59 FR 7629, February 16, 1994. Available at
https: www.svo.govfdsyspkg FR-1994-02-16 html' 94-3685.htm. Retrieved from
Executive Order 13045. 1997. Protection of Children from Environmental Health Risks and
Safety Risks. 62 FR 19885, April 23, 1997. Available at
https://www.gpo.gov/fdsvs/pkg/FR-1997-04-23/pdf/97-10695.pdf.. (1997).
Executive Order 13132. 1999. Federalism. 64 FR 43255, August 10, 1999. Available at
https://www.gpo.gov/fdsvs/pkg/FR-1999-08-10/pdf/99-20729.pdf (1999).
Executive Order 13175. 2000. Consultation and Coordination with Indian Tribal Governments.
65 FR 67249, November 9, 2000. Available at https://www.gpo.gov/fdsvs/pkg/FR-200Q-
1 l-09/pdf/00-29003.pdf.. (2000).
Executive Order 13211. 2001. Actions Concerning Regulations That Significantly Affect Energy
Supply, Distribution, or Use. 66 FR 28355, May 22, 2001. Available at
https://www.gpo.gov/fdsvs/pkg/FR-2001-Q5-22/pdf/01-13116.pdf.. (2001).
Executive Order 13563. 2011. Improving Regulation and Regulatory Review. 76 FR 3821,
January 21, 2011. Available at https://www.gpo.gov/fdsvs/pkg/FR-201 l-01-21/pdf/2011-
1385.pdf.. (2011).
Fan, Y., Li, X., Xu, Q., Zhang, Y., Yang, X., Han, X., . . . Lu, C. (2020). Serum albumin
mediates the effect of multiple per-and polyfluoroalkyl substances on serum lipid levels.
Environmental pollution, 266, 115138.
Fei, C., McLaughlin, J. K., Lipworth, L., & Olsen, J. (2009). Maternal levels of perfluorinated
chemicals and subfecundity. Human reproduction, 24(5), 1200-1205.
doi: 10.1093/humrep/den490
Proposed PFAS Rule Economic Analysis
10-6
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Fisher, M., Arbuckle, T. E., Wade, M., & Haines, D. A. (2013). Do perfluoroalkyl substances
affect metabolic function and plasma lipids?—Analysis of the 2007-2009, Canadian
Health Measures Survey (CHMS) Cycle 1. Environmental research, 121, 95-103.
Foreman, K., Vacs Renwick, D., McCabe, M., Cadwallader, A., Holsinger, H., Kormondy, C., &
Albert, R. (2021). Effects of harmful algal blooms on regulated disinfection byproducts:
Findings from five utility case studies. AWWA Water Science, 3(3), el223.
Freeman, L. E., Cantor, K. P., Baris, D., Nuckols, J. R., Johnson, A., Colt, J. S., . . . Silverman,
D. T. (2017). Bladder Cancer and Water Disinfection By-product Exposures through
Multiple Routes: A Population-Based Case-Control Study (New England, USA).
Environmental health perspectives, 125(6).
Fried, L. P., Borhani, N. O., Enright, P., Furberg, C. D., Gardin, J. M., Kronmal, R. A., . . .
Newman, A. (1991). The cardiovascular health study: design and rationale. Annals of
epidemiology, 7(3), 263-276.
Friedman, G. D., Cutter, G. R., Donahue, R. P., Hughes, G. H., Hulley, S. B., Jacobs Jr, D. R., . .
. Savage, P. J. (1988). CARDIA: study design, recruitment, and some characteristics of
the examined subjects. Journal of clinical epidemiology, ¥7(11), 1105-1116.
Frisbee, S. J., Shankar, A., Knox, S. S., Steenland, K., Savitz, D. A., Fletcher, T., & Ducatman,
A. M. (2010). Perfluorooctanoic acid, perfluorooctanesulfonate, and serum lipids in
children and adolescents: results from the C8 Health Project. Archives of pediatrics &
adolescent medicine, 164(9), 860-869.
Fromme, H., Tittlemier, S. A., Volkel, W., Wilhelm, M., & Twardella, D. (2009). Perfluorinated
compounds-exposure assessment for the general population in Western countries.
International journal of hygiene and environmental health, 212(3), 239-270.
Fry, K., & Power, M. C. (2017). Persistent organic pollutants and mortalityin the United States,
NHANES 1999-2011. Environmental Health, 16(105), 1-12. doi:10.1186/sl2940-017-
0313-6
Fryar, C. D., Ostchega, Y., Hales, C. M., Zhang, G., & Kruszon-Moran, D. (2017). Hypertension
Prevalence and Control Among Adults: United States, 2015-2016. NCHSData
Brief(289), 1-8.
Fu, Y., Wang, T., Fu, Q., Wang, P., & Lu, Y. (2014). Associations between serum concentrations
of perfluoroalkyl acids and serum lipid levels in a Chinese population. Ecotoxicology and
environmental safety, 106, 246-252.
Gao, X., Ni, W., Zhu, S., Wu, Y., Cui, Y., Ma, J., . . . Yang, P. (2021). Per- and polyfluoroalkyl
substances exposure during pregnancy and adverse pregnancy and birth outcomes: A
systematic review and meta-analysis. Environmental research, 201, 111632.
Glynn, A., Berger, U., Bignert, A., Ullah, S., Aune, M., Lignell, S., & Darnerud, P. O. (2012).
Perfluorinated Alkyl Acids in Blood Serum from Primiparous Women in Sweden: Serial
Sampling during Pregnancy and Nursing, And Temporal Trends 1996-2010.
Environmental Science & Technology, 46(16), 9071-9079. doi:10.1021/es301168c
Goff, D. C., Lloyd-Jones, D. M., Bennett, G., Coady, S., D'Agostino, R. B., Gibbons, R., . . .
Wilson, P. W. F. (2014). 2013 ACC/AHA guideline on the assessment of cardiovascular
risk: A report of the American College of Cardiology/American Heart Association task
force on practice guidelines. Circulation, 129(25 suppl 2), 49-73.
Goldstein, K. M., Zullig, L. L., Bastian, L. A., & Bosworth, H. B. (2016). Statin Adherence:
Does Gender Matter? Current Atherosclerosis Reports, 18(11), 63. doi:10.1007/sll883-
016-0619-9
Proposed PFAS Rule Economic Analysis
10-7
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Govarts, E., Iszatt, N., Trnovec, T., De Cock, M., Eggesbo, M., Murinova, L. P., . . . Koppen, G.
(2018). Prenatal exposure to endocrine disrupting chemicals and risk of being born small
for gestational age: Pooled analysisof seven European birth cohorts. Environment
international, 115, 267-278. doi:10.1016/j.envint.2018.03.01
Govarts, E., Remy, S., Bruckners, L., Hond, E. D., Sioen, I., Nelen, V., . . . Schoeters, G. (2016).
Combined Effects of Prenatal Exposures to Environmental Chemicals on Birth Weight.
International journal of environmental research and public health, 75(495), 1-19.
doi:10.3390/ijerphl3050495
Graham, G. (2015). Disparities in cardiovascular disease risk in the United States. Current
cardiology reviews, 11(3), 238-245.
Grandjean, P., Timmermann, C. A. G., Kruse, M., Nielsen, F., Vinholt, P. J., Boding, L., . . .
Molbak, K. (2020). Severity of COVID-19 at elevated exposure to perfluorinated
alkylates. PLoS One, 75(12), e0244815.
Greco, S. L., Belova, A., Haskell, J., & Backer, L. (2019). Estimated burden of disease from
arsenic in drinking water supplied by domestic wells in the United States. Journal of
water and health, 77(5), 801-812.
Guelfo, J. L., & Adamson, D. T. (2018). Evaluation of a National Data set for Insights into
Sources, Composition, and Concentrations of Per- and Polyfluoroalkyl Substances
(PFASs) in U.S. Drinking Water. Environmental pollution, 236, 505-513.
doi:https://doi.org/10.1016/i.envpol.2018.01.066
Hartge, P., Silverman, D., Hoover, R., Schairer, C., Altman, R., Austin, D., . . . Marrett, L. D.
(1987). Changing cigarette habits and bladder cancer risk: a case-control study. Journal
of the National Cancer Institute, 78(6), 1119-1125.
Hashempour, Y., Nasseri, M., Mohseni-Bandpei, A., Motesaddi, S., & Eslamizadeh, M. (2020).
Assessing vulnerability to climate change for total organic carbon in a system of drinking
water supply. Sustainable Cities and Society, 53, 101904.
doi:https://doi.org/10.1016/i.scs.2019.101904
He, X., Liu, Y., Xu, B., Gu, L., & Tang, W. (2018). PFOA is associated with diabetes and
metabolic alteration in US men: National Health and Nutrition Examination Survey
2003-2012. Science of the total environment, 625, 566-574.
Hines, C. T., Padilla, C. M., & Ryan, R. M. (2020). The effect of birth weight on child
development prior to school entry. Child development, 97(3), 724-732.
Hooper, S. M., & Allgeier, S. C. (2002). GAC and membrane treatment studies Chapter 21 in
Information Collection Rule Data Analysis. Water Research Foundation andAWWA.
Hrubec, Z., & McLaughlin, J. K. (1997). Former cigarette smoking and mortality among US
veterans: a 26-year follow-up, 1954-1980. Changes in cigarette-related disease risks and
their implication for prevention and control. Bethesda, MD: US Government Printing
Office, 501-530.
Hrudey, S. E., Backer, L. C., Humpage, A. R., Krasner, S. W., Michaud, D. S., Moore, L. E., . . .
Stanford, B. D. (2015). Evaluating evidence for association of human bladder cancer with
drinking-water chlorination disinfection by-products. Journal of Toxicology and
Environmental Health, Part B, 18(5), 213-241.
Hu, X. C., Andrews, D. Q., Lindstrom, A. B., Bruton, T. A., Schaider, L. A., Grandjean, P., . . .
Sunderland, E. M. (2016). Detection of Poly- and Perfluoroalkyl Substances (PFASs) in
U.S. Drinking Water Linked to Industrial Sites, Military Fire Training Areas, and
Proposed PFAS Rule Economic Analysis
10-8
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Wastewater Treatment Plants. Environmental Science & Technology Letters, 5(10), 344-
350. doi:10.1021/acs.estlett.6b00260
Hu, Z., Morton, L. W., & Mahler, R. L. (2011). Bottled water: United States consumers and their
perceptions of water quality. International journal of environmental research and public
health, 8(2), 565-578. doi:10.3390/ijerph8020565
Hua, G., & Reckhow, D. A. (2008). DBP formation during chlorination and chloramination:
effect of reaction time, pH, dosage, and temperature. Journal-American Water Works
Association, 700(8), 82-95.
Huang, M., Jiao, J., Zhuang, P., Chen, X., Wang, J., & Zhang, Y. (2018). Serum polyfluoroalkyl
chemicals are associated with risk of cardiovascular diseases in national US population.
Environment international, 119, 37-46.
Hutcheson, R., Innes, K., & Conway, B. (2019). Perfluoroalkyl substances and likelihood of
stroke in persons with and without diabetes. Diabetes and Vascular Disease Research,
77(1), 1479164119892223. doi: 10.1177/1479164119892223
ICF. (2020). Memorandum re: Trends in Infant Mortality by Birth Weight Categories, EPA
Contract No. 68HE0C18D0001, TO 68HERC20F0107. Submitted to U.S. Environmental
Protection Agency Office of Groundwater and Drinking Water July 30, 2020. Retrieved
from https://www.regulations.gov/docket/EPA-HQ-OW-2022-Q114
ICF. (2021). Memorandum re: Summary of Literature Review of Non-Medical Effects and
Economic Burden Related to Low Birth Weight Infants. Submitted to U.S. Environmental
Protection Agency Office of Groundwater and Drinking Water October 5, 2021.
Retrieved from https://www.regulations.gov/docket/EPA-HQ-QW-2022-0114
ICF. (2022a). Memorandum re: Literature Review ofASCVD Risk Models and Limitations.
Submitted to U.S. Environmental Protection Agency Office of Groundwater and Drinking
Water April 4, 2022. Retrieved from https://www.regulations.gov/docket/EPA-HQ-OW-
2022-0114
ICF. (2022b). Memorandum re: Literature Review of Population Attributable Fraction Estimates
and Next Steps for Renal Cell Carcinoma PFAS Benefits Modeling, EPA Contract No.
68HE0C18D0001, TO 68HERC22F0262. Submitted to U.S. Environmental Protection
Agency Office of Groundwater and Drinking Water July 28, 2022. Retrieved from
https://www.regulations.gov/docket/EPA-HQ-QW-2022-0114
Iriarte-Velasco, U., I. Alvarez-Uriarte, J., Chimeno-Alanis, N., & R. Gonzalez-Velasco, J.
(2008). Natural organic matter adsorption onto granular activated carbons: implications
in the molecular weight and disinfection byproducts formation. Industrial & engineering
chemistry research, 47 (20), 7868-7876.
Jacob, J. A. (2016). U.S. infant mortality rate declines but still exceeds other developed
countries. Journal of the American Medical Association, 315(5), 451-452.
Jain, R. B. (2018). Time trends over 2003-2014 in the concentrations of selected perfluoroalkyl
substances among U.S. adults aged >20 years: Interpretational issues. Science of the total
environment, 645, 946-957. doi:10.1016/j.scitotenv.2018.07.198
Jain, R. B., & Ducatman, A. (2019). Roles of gender and obesity in defining correlations
between perfluoroalkyl substances and lipid/lipoproteins. Science of the total
environment, 653, 74-81.
Jelenkovic, A., Mikkonen, J., Martikainen, P., Latvala, A., Yokoyama, Y., Sund, R., . . . Kim, J.
(2018). Association between birth weight and educational attainment: an individual-based
pooled analysis of nine twin cohorts. J Epidemiol Community Health, 72(9), 832-837.
Proposed PFAS Rule Economic Analysis
10-9
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Ji, J., Song, L., Wang, J., Yang, Z., Yan, H., Li, T., . . . Li, J. (2021). Association between
urinary per-and poly-fluoroalkyl substances and COVID-19 susceptibility. Environment
international, 153, 106524.
Jiang, J., Zhang, X., Zhu, X., & Li, Y. (2017). Removal of intermediate aromatic halogenated
DBPs by activated carbon adsorption: a new approach to controlling halogenated DBPs
in chlorinated drinking water. Environmental Science & Technology, 57(6), 3435-3444.
Johnson, P. I., Sutton, P., Atchley, D. S., Koustas, E., Lam, J., Sen, S., . . . Woodruff, T. J.
(2014). The Navigation Guide—Evidence-Based Medicine Meets Environmental Health:
Systematic Review of Human Evidence for PFOA Effects on Fetal Growth.
Environmental health perspectives, 722(10), 1028-1039. doi: 10.1289/ehp. 1307893
Johnston, J., & Cushing, L. (2020). Chemical Exposures, Health, and Environmental Justice in
Communities Living on the Fenceline of Industry. Current Environmental Health
Reports, 7(1), 48-57. doi: 10.1007/s40572-020-00263-8
Joyce, C., Goodman-Bryan, M., & Hardin, A. (2012). Preterm Birth and Low Birth Weight.
Retrieved from The Urban Child Institute:
http ://www.urbanchildinstitute. org/ sites/all/files/2010-10-01 -PTB-and-LBW.pdf
Kamai, E. M., McElrath, T. F., & Ferguson, K. K. (2019). Fetal growth in environmental
epidemiology: mechanisms, limitations, and a review of associations with biomarkers of
non-persistent chemical exposures during pregnancy. Environmental Health, 75(1), 43.
Kato, K., Wong, L.-Y., Chen, A., Dunbar, C., Webster, G. M., Lanphear, B. P., & Calafat, A. M.
(2014). Changes in serum concentrations of maternal poly-and perfluoroalkyl substances
over the course of pregnancy and predictors of exposure in a multiethnic cohort of
Cincinnati, Ohio pregnant women during 2003-2006. Environmental Science &
Technology, 48( 16), 9600-9608.
Kempisty, D., Arevalo, E., Spinelli, A., Edeback, V., Dickenson, E., Husted, C., . . . Knappe, D.
(2022). Granular activated carbon adsorption of perfluoroalkyl acids from ground and
surface water. AWWA Water Science, 4(l)el269. doi:https://doi.org/10.1002/aws2.1269
Kim, W. H., Nishijima, W., Shoto, E., & Okada, M. (1997). Competitive removal of dissolved
organic carbon by adsorption and biodegradation on biological activated carbon. Water
Science and Technology, 35(7), 147-153. doi:https://doi.org/10.1016/S0273-
1223(97)00125-X
King, W. D., & Marrett, L. D. (1996). Case-control study of bladder cancer and chlorination by-
products in treated water (Ontario, Canada). Cancer causes & control: CCC, 7(6), 596-
604. doi:https://doi.org/10.1007/BF00051702
Kingsley, S. L., Eliot, M. N., Kelsey, K. T., Calafat, A. M., Ehrlich, S., Lanphear, B. P., . . .
Braun, J. M. (2018). Variability and predictors of serum perfluoroalkyl substance
concentrations during pregnancy and early childhood. Environmental research, 165, 247-
257. doi:10.1016/j.envres.2018.04.033
Kirchberger, I., Burkhardt, K, Heier, M., Thilo, C., & Meisinger, C. (2020). Resilience is
strongly associated with health-related quality of life but does not buffer work-related
stress in employed persons 1 year after acute myocardial infarction. Quality of Life
Research, 29(2), 391-401. doi: 10.1007/sl 1136-019-02306-6
Kirisits, M. J., Snoeyink, V. L., & Kruithof, J. C. (2000). The reduction of bromate by granular
activated carbon. Water research, 34(17), 4250-4260.
Klein, R., & Lynch, M. (2018). Development of Medical Cost Estimates for Adverse Birth
Outcomes. Prepared for U.S. EPA National Center for Environmental Economics
Proposed PFAS Rule Economic Analysis
10-10
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Kotlarz, N., McCord, J., Collier, D., Lea, C. S., Strynar, M., Lindstrom, A. B., . . . Hoppin, J. A.
(2020). Measurement of Novel, Drinking Water-Associated PFAS in Blood from Adults
and Children in Wilmington, North Carolina. Environ Health Perspect, 128(7), 77005.
doi: 10.1289/ehp683 7
Kowlessar, N. M., Jiang, H. J., & Steiner, C. (2013). Hospital stays for newborns, 2011:
statistical brief# 163.
Lau, C., Ambalavanan, N, Chakraborty, H., Wingate, M. S., & Carlo, W. A. (2013). Extremely
low birth weight and infant mortality rates in the United States. Pediatrics, 131(5), 855-
860.
Lauritzen, H. B., Larose, T. L., 0ien, T., Sandanger, T. M., Odland, J. O., van der Bor, M., &
Jacobsen, G. W. (2017). Maternal serum levels of perfluoroalkyl substances and
organochlorines and indices of fetal growth: a Scandinavian case-cohort study. Pediatric
research, 57(1), 33-42. doi:10.1038/pr.2016.187
LeChevallier, M. W. (2020). Managing Legionella pneumophila in Water Systems. Journal:
American Water Works Association, 112(2).
Lee, S., Kang, D., Kim, Y., Kim, Y. J., & Kim, S. Y. (2021). Activity-Based Exposure Levels
and Cancer Risk Assessment Due to Naturally Occurring Asbestos for the Residents Near
Abandoned Asbestos Mines in South Korea. Int J Environ Res Public Health, 75(10).
doi:10.3390/ijerphl8105225
Lee, S., Kar, A., & Reade, A. (2021). Dirty Water: Toxic "Forever" PFAS Chemicals Are
Prevalent in the Drinking Water of Environmental Justice Communities. Retrieved from
https://www.nrdc.org/sites/default/files/dirtv-water-pfas-ei-communities-report.pdf.
Leigh, A., McEvoy, J. W., Garg, P., Carr, J. J., Sandfort, V., Oelsner, E. C., . . . Yeboah, J.
(2019). Coronary artery calcium scores and atherosclerotic cardiovascular disease risk
stratification in smokers. JACC: Cardiovascular Imaging, 12(5), 852-861.
Leino, A. D., Dorsch, M. P., & Lester, C. A. (2020). Changes in Statin Use Among U.S. Adults
With Diabetes: A Population-Based Analysis of NHANES 2011-2018. Diabetes Care,
43( 12), 3110-3112. doi: 10.2337/dc20-1481
Lenters, V., Portengen, L., Rignell-Hydbom, A., Jonsson, B. A. G., Lindh, C. H., Piersma, A. H.,
. . . Vermeulen, R. (2016). Prenatal Phthalate, Perfluoroalkyl Acid, and Organochlorine
Exposures and Term Birth Weight in Three Birth Cohorts: Multi-Pollutant Models Based
on Elastic Net Regression. Environmental health perspectives, 124, 365-372.
doi: 10.1289/ehp. 1408933
Lewey, J., Shrank, W. H., Bowry, A. D. K., Kilabuk, E., Brennan, T. A., & Choudhry, N. K.
(2013). Gender and racial disparities in adherence to statin therapy: A meta-analysis.
American Heart Journal, 165(5), 665-678. doi:https://doi.org/10.1016/i.ahi.2013.02.011
Li, S., Peng, Y., Wang, X., Qian, Y., Xiang, P., Wade, S. W., . . . Handelsman, Y. (2019).
Cardiovascular events and death after myocardial infarction or ischemic stroke in an
older Medicare population. Clinical cardiology, 42(3), 391-399.
Li, Y., Barregard, L., Xu, Y., Scott, K, Pineda, D., Lindh, C. H., . . . Fletcher, T. (2020).
Associations between perfluoroalkyl substances and serum lipids in a Swedish adult
population with contaminated drinking water. Environmental Health, 19(1), 1-11.
Liao, S., Yao, W., Cheang, I., Tang, X., Yin, T., Lu, X., . . . Li, X. (2020). Association between
perfluoroalkyl acids and the prevalence of hypertension among US adults. Ecotoxicology
and environmental safety, 196, 110589.
Proposed PFAS Rule Economic Analysis
10-11
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Liew, Z., Olsen, J., Cui, X., Ritz, B., & Arah, O. A. (2015). Bias from conditioning on live birth
in pregnancy cohorts: an illustration based on neurodevelopment in children after
prenatal exposure to organic pollutants. International Journal of Epidemiology, 44( 1),
345-354.
Lin, C. Y., Lee, H. L., Hwang, Y. T., & Su, T. C. (2020). The association between total serum
isomers of per-and polyfluoroalkyl substances, lipid profiles, and the DNA
oxidative/nitrative stress biomarkers in middle-aged Taiwanese adults. Environmental
research, 182, 109064.
Lin, P.-I. D., Cardenas, A., Hauser, R., Gold, D. R., Kleinman, K. P., Hivert, M.-F., . . . Horton,
E. S. (2019). Per-and polyfluoroalkyl substances and blood lipid levels in pre-diabetic
adults—longitudinal analysis of the diabetes prevention program outcomes study.
Environment international, 129, 343-353.
Liu, H.-S., Wen, L.-L., Chu, P.-L., & Lin, C.-Y. (2018). Association among total serum isomers
of perfluorinated chemicals, glucose homeostasis, lipid profiles, serum protein and
metabolic syndrome in adults: NHANES, 2013-2014. Environmental pollution, 232, 73-
79.
Lloyd-Jones, D. M., Huffman, M. D., Karmali, K. N., Sanghavi, D. M., Wright, J. S., Pelser, C., .
. . Goff Jr, D. C. (2017). Estimating longitudinal risks and benefits from cardiovascular
preventive therapies among Medicare patients: the million hearts longitudinal ASCVD
risk assessment tool: a special report from the American heart association and American
College of cardiology. Circulation, 755(13), e793-e813.
Lu, S., & Bartell, S. M. (2020). Serum PFAS Calculator for Adults. Retrieved from:
www.ics.uci.edu/~sbartell/pfascalc.html
Ma, S., & Finch, B. K. (2010). Birth Outcome Measures and Infant Mortality. Population
Research and Policy Review, 29(6), 865-891.
Malits, J., Blustein, J., Trasande, L., & Attina, T. M. (2018). Perfluorooctanoic acid and low
birth weight: Estimates of US attributable burden and economic costs from 2003 through
2014. International journal of hygiene and environmental health, 221(2), 269-275.
doi: 10.1016/j.ijheh.2017.11.004
Martino Cinnera, A., Bonni, S., Pellicciari, M. C., Giorgi, F., Caltagirone, C., & Koch, G.
(2020). Health-related quality of life (HRQoL) after stroke: Positive relationship between
lower extremity and balance recovery. Topics in Stroke Rehabilitation, 27(1), 534-540.
doi: 10.1080/10749357.2020.1726070
Mattsson, K., Rignell-Hydbom, A., Holmberg, S., Thelin, A., Jonsson, B. A., Lindh, C. H., . . .
Rylander, L. (2015). Levels of perfluoroalkyl substances and risk of coronary heart
disease: Findings from a population-based longitudinal study. Environmental research,
142, 148-154.
Mayo Clinic. (2022, January 6). Elevated Liver Enzymes - Causes. Retrieved from
https://www.mavoclinic.org/svmptoms/elevated-liver-enzymes/basics/causes/svm-
20050830
McCleaf, P., Englund, S., Ostlund, A., Lindegren, K., Wiberg, K., & Ahrens, L. (2017). Removal
efficiency of multiple poly-and perfluoroalkyl substances (PFASs) in drinking water
using granular activated carbon (GAC) and anion exchange (AE) column tests. Water
research(\20), 77-87.
Proposed PFAS Rule Economic Analysis
10-12
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
McCord, J., & Strynar, M. (2019). Identification of per-and polyfluoroalkyl substances in the
Cape Fear River by high resolution mass spectrometry and nontargeted screening.
Environmental Science & Technology, 55(9), 4717-4727.
McCord, J., Strynar, M., Washington, J., Bergman, E., & Goodrow, S. (2020). Emerging
chlorinated polyfluorinated polyether compounds impacting the waters of southwestern
New Jersey identified by use of nontargeted analysis. Environmental Science &
Technology Letters, 7(12), 903-908.
McDonough, L. K., Santos, I. R., Andersen, M. S., O'Carroll, D. M., Rutlidge, H., Meredith, K.,
. . . Lapworth, D. J. (2020). Changes in global groundwater organic carbon driven by
climate change and urbanization. Nature communications. 11(1), 1-10.
Mclntire, D. D., Bloom, S. L., Casey, B. M., & Leveno, K. J. (1999). Birth weight in relation to
morbidity and mortality among newborn infants. New England Journal of Medicine,
340(16), 1234-1238.
McLaren Jr, R. A., Trejo, F. E., Blitz, M. J., Bianco, A., Limaye, M., Brustman, L., . . . Minkoff,
H. (2021). COVID-related "lockdowns" and birth rates in New York. American journal
of obstetrics & gynecology MFM, 5(6), 100476.
McNamara, J. D., Franco, R., Mimna, R., & Zappa, L. (2018). Comparison of activated carbons
for removal of perfluorinated compounds from drinking water. Journal-American Water
Works Association, 11()( \), E2-E14.
Minnesota Pollution Control Agency. (2021). Minnesota's PFAS Blueprint, (p-genl-22).
Retrieved from https://www.pca.state.mn.us/sites/default/files/p-genl-22.pdf
Moll on, L., & Bhattacharjee, S. (2017). Health related quality of life among myocardial
infarction survivors in the United States: a propensity score matched analysis. Health and
Quality of Life Outcomes, 75(1), 235. doi:10.1186/sl2955-017-0809-3
Mongraw-Chaffin, M., Gujral, U. P., Kanaya, A. M., Kandula, N. R., Carr, J. J., & Anderson, C.
A. (2018). Relation of ectopic fat with atherosclerotic cardiovascular disease risk score in
South Asians living in the United States (from the Mediators of Atherosclerosis in South
Asians Living in America [MASALA] Study). The American journal of cardiology,
727(3), 315-321.
Mora, S., Wenger, N. K., Cook, N. R., Liu, J., Howard, B. V., Limacher, M. C., . . . Paynter, N.
P. (2018). Evaluation of the pooled cohort risk equations for cardiovascular risk
prediction in a multiethnic cohort from the Women's Health Initiative. JAMA Internal
Medicine, 775(9), 1231-1240.
Muntner, P., Colantonio, L. D., Cushman, M., Goff, D. C., Jr., Howard, G., Howard, V. J., . . .
Safford, M. M. (2014). Validation of the atherosclerotic cardiovascular disease Pooled
Cohort risk equations. Jama, 577(14), 1406-1415. doi:10.1001/jama.2014.2630
National Academies of Sciences, E., and Medicine,. (2020). Management of Legionella in Water
Systems. Washington, DC.: The National Academies Press
National Academies of Sciences, E., and Medicine,. (2022). Guidance on PFASExposure,
Testing, and Clinical1 Follow-Up (2022). Washington, DC: The National Academies
Press.
National Academy of Public Administration. (2017). Developing a New Framework for
Community Affordability of Clean Water Services.
National Technology Transfer and Advancement Act. 1995. Public Law No. 104-113, 110 Stat.
783. 104th Congress. Available at https://www.gpo.gov/fdsvs/pkg/PLAW-
104publ 113/pdf/PLAW-104publ 113 .pdf.. (1995).
Proposed PFAS Rule Economic Analysis
10-13
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
National Toxicology Program. (2018). Report on Carcinogens: Monograph on Haloacetic Acids
Found as Water Disinfection By-Products. Research Triangle Park, NC.
National Toxicology Program. (2020). NTP Technical Report on the Toxicology and
Carcinogenesis Studies of Perflaorooctanoic Acid (CASRN 335-67-1) Administered in
Feed to Sprague Dawley (Hsd: Sprague Daw ley® SD®) Rats.
National Toxicology Program. (2021). Report on Carcinogens, Fifteenth Edition.; Research
Triangle Park, NC: U.S. Department of Health and Human Services, Public Health
Service.
Negri, E., Metruccio, F., Guercio, V., Tosti, L., Benfenati, E., Bonzi, R., . . . Moretto, A. (2017).
Exposure to PFOA and PFOS and fetal growth: a critical merging of toxicological and
epidemiological data. Critical reviews in toxicology, 10.1080 10408444.2016.1271972,
489-515. doi: 10.1080/10408444.2016.1271972
Nelson, J. W., Hatch, E. E., & Webster, T. F. (2010). Exposure to polyfluoroalkyl chemicals and
cholesterol, body weight, and insulin resistance in the general US population.
Environmental health perspectives, 118(2), 197-202.
Nelson, J. W., Scammell, M. K., Hatch, E. E., & Webster, T. F. (2012). Social disparities in
exposures to bisphenol A and polyfluoroalkyl chemicals: a cross-sectional study within
NHANES 2003-2006. Environmental Health, 77(1), 1-15.
Nguyen, Q. D., Odden, M. C., Peralta, C. A., & Kim, D. H. (2020). Predicting risk of
atherosclerotic cardiovascular disease using pooled cohort equations in older adults with
frailty, multimorbidity, and competing risks. Journal of the American Heart Association,
9(18), eO 16003.
Nguyen, V. K., Kahana, A., Heidt, J., Polemi, K., Kvasnicka, J., Jolliet, O., & Colacino, J. A.
(2020). A comprehensive analysis of racial disparities in chemical biomarker
concentrations in United States women, 1999-2014. Environment international, 137,
105496.
Nicoletti, C., Salvanes, K. G., & Tominey, E. (2018). Response of parental investments to child's
health endowment at birth. In Health Econometrics: Emerald Publishing Limited.
O'Sullivan, A. K., Rubin, J., Nyambose, J., Kuznik, A., Cohen, D. J., & Thompson, D. (2011).
Cost estimation of cardiovascular disease events in the US. Pharmacoeconomics, 29(8),
693-704.
Occupational Safety and Health Administration. (2010). Occupational Exposure to Respirable
Crystalline Silica — Review of Health Effects Literature and Preliminary Quantitative
Risk Assessment.
Occupational Safety and Health Administration. (2016). Occupational Exposure to Respirable
Crystalline Silica. (81 FR 16285 2016-04800). Federal Register Retrieved from
https://www.federalregister.gov/documents/2016/03/25/2016-0480Q/occupational-
exposure-to-respirable-crystalline-silica
Ochoa-Herrera, V., & Sierra-Alvarez, R. (2008). Removal of perfluorinated surfactants by
sorption onto granular activated carbon, zeolite and sludge. Chemosphere, 72(10), 1588-
1593.
Office of Environmental Health Hazard Assessment. (2016). CalEnviroScreen 3.0.
Osuchukwu, O., & Reed, D. (2022). Small for Gestational Age. StatPearls [Internet],
Park, S. K., Peng, Q., Ding, N, Mukheijee, B., & Harlow, S. D. (2019). Determinants of per-and
polyfluoroalkyl substances (PFAS) in midlife women: evidence of racial/ethnic and
geographic differences in PFAS exposure. Environmental research, 175, 186-199.
Proposed PFAS Rule Economic Analysis
10-14
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Pinheiro, P. S., Medina, H. N., Callahan, K. E., Koru-Sengul, T., Sharma, J., Kobetz, E. N., &
Penedo, F. J. (2021). Kidney cancer mortality disparities among Hispanics in the US.
Cancer Epidemiology, 72, 101938. doi:https://doi.org/10.1016/i.canep.2021.101938
Pletcher, M. J., Vittinghoff, E., Thanataveerat, A., Bibbins-Domingo, K., & Moran, A. E. (2016).
Young adult exposure to cardiovascular risk factors and risk of events later in life: The
Framingham Offspring Study. PLoS One, 77(5), e0154288.
Pramanik, B. K., Pramanik, S. K., & Suja, F. (2015). A comparative study of coagulation,
granular-and powdered-activated carbon for the removal of perfluorooctane sulfonate and
perfluorooctanoate in drinking water treatment. Environmental technology, 36(20), 2610-
2617.
R Core Team. (2021). R: A language and environment for statistical computing. Vienna, Austria:
R Foundation for Statistical Computing.
Raghavan, S., Ho, Y.-L., Vassy, J. L., Posner, D., Honerlaw, J., Costa, L., . . . Cho, K. (2020).
Optimizing Atherosclerotic Cardiovascular Disease Risk Estimation for Veterans With
Diabetes Mellitus. Circulation. Cardiovascular quality and outcomes, 13(9), 120006528.
doi: 10.1161/circoutcomes. 120.006528
Raleigh, K. K., Alexander, B. H., Olsen, G. W., Ramachandran, G., Morey, S. Z., Church, T. R.,
. . . Allen, E. M. (2014). Mortality and cancer incidence in ammonium perfluorooctanoate
production workers. Occupational Environmental Medicine, 71, 500-506.
Raucher, R., Clements, J., Rothstein, E., Mastracchio, J., & Green, Z. (2019). Developing a New
Framework for Household Affordability and Financial Capability Assessment in the
Water Sector.
Regli, S., Chen, J., Messner, M., Elovitz, M. S., Letkiewicz, F. J., Pegram, R. A., . . . Wright, J.
M. (2015). Estimating potential increased bladder cancer risk due to increased bromide
concentrations in sources of disinfected drinking waters. Environmental Science &
Technology, 49(22), 13094-13102.
Richardson, S. D. (2022). Invited Perspective: Existing Rules for Disinfection By-Products Are
Good, but They Are Not Enough. Environmental health perspectives, 750(8), 081302.
Robinson, J. G., & Booth, B. (2010). Statin use and lipid levels in older adults: National Health
and Nutrition Examination Survey, 2001-2006. Journal of Clinical Lipidology, 4(6),
483-490. doi: 10.1016/j.jacl.2010.10.002
Rockaway, T. D., Coomes, P. A., Rivard, J., & Kornstein, B. (2011). Residential water use
trends in North America. Journal-American Water Works Association, 103(2), 76-89.
Rodriguez-Barranco, M., Tobias, A., Redondo, D., Molina-Portillo, E., & Sanchez, M. J. (2017).
Standardizing effect size from linear regression models with log-transformed variables
for meta-analysis. BMC medical research methodology, 77(1), 1-9.
Rodriguez, F., Chung, S., Blum, M. R., Coulet, A., Basu, S., & Palaniappan, L. P. (2019).
Atherosclerotic cardiovascular disease risk prediction in disaggregated Asian and
Hispanic subgroups using electronic health records. Journal of the American Heart
Association, 5(14), e011874.
Roger, V. L., Go, A. S., Lloyd-Jones, D. M., Benjamin, E. J., Berry, J. D., Borden, W. B., . . .
Ford, E. S. (2012). Heart disease and stroke statistics—2012 update: a report from the
American Heart Association. Gradation, 125(1), e2.
Rogers, J., Ellis-Hutchings, R., Grey, B., Zucker, R., Norwood, J., Jr, Grace, C., . . . Lau, C.
(2014). Elevated blood pressure in offspring of rats exposed to diverse chemicals during
pregnancy. Toxicological Sciences. doi:10.1093/toxsci/kft248
Proposed PFAS Rule Economic Analysis
10-15
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Rosinger, A. Y., Herrick, K. A., Wutich, A. Y., Yoder, J. S., & Ogden, C. L. (2018). Disparities
in plain, tap and bottled water consumption among US adults: National Health and
Nutrition Examination Survey (NHANES) 2007-2014. Public health nutrition, 21{8),
1455-1464. doi: 10.1017/S1368980017004050
Riicker, G., Schwarzer, G., Carpenter, J., & Olkin, I. (2009). Why add anything to nothing? The
arcsine difference as a measure of treatment effect in meta-analysis with zero cells.
Statistics in medicine, 25(5), 721-738.
Sagiv, S. K., Rifas-Shiman, S. L., Fleisch, A. F., Webster, T. F., Calafat, A. M., Ye, X., . . .
Oken, E. (2018). Early-pregnancy plasma concentrations of perfluoroalkyl substances
and birth outcomes in Project Viva: confounded by pregnancy hemodynamics? American
journal of epidemiology, 757(4), 793-802.
Shankar, A., Xiao, J., & Ducatman, A. (2012). Perfluorooctanoic acid and cardiovascular disease
in US adults. Archives of Internal Medicine, 772(18), 1397-1403.
Shearer, J. J., Callahan, C. L., Calafat, A. M., Huang, W. Y., Jones, R. R., Sabbisetti, V. S., . . .
Hofmann, J. N. (2021). Serum concentrations of per- and polyfluoroalkyl substances and
risk of renal cell carcinoma. Journal of the National Cancer Institute, 113(5), 580-587.
doi: 10.1093/j nci / dj aa 143
Skolarus, L. E., Burke, J. F., Brown, D. L., & Freedman, V. A. (2014). Understanding stroke
survivorship: expanding the concept of poststroke disability. Stroke, -/5(1), 224-230.
Sorengard, M., Ostblom, E., Kohler, S., & Ahrens, L. (2020). Adsorption behavior of per-and
polyfluoralkyl substances (PFASs) to 44 inorganic and organic sorbents and use of dyes
as proxies for PFAS sorption. Journal of Environmental Chemical Engineering, 5(3),
103744.
Sorlini, S., & Collivignarelli, C. (2005). Chlorite removal with granular activated carbon.
Desalination, 77(5(1-3), 255-265.
Speitel Jr, G. E., Turakhia, M. H., & Lu, C.-J. (1989). Initiation of micropollutant biodegradation
in virgin GAC columns. Journal (American Water Works Association), 168-176.
Starling, A. P., Adgate, J. L., Hamman, R. F., Kechris, K., Calafat, A. M., Ye, X., & Dabelea, D.
(2017). Perfluoroalkyl substances during pregnancy and offspring weight and adiposity at
birth: examining mediation by maternal fasting glucose in the Healthy Start Study.
Environ Health Perspect, 125, 067016-067011. doi:10.1289/EHP641
Steenland, K., Barry, V., & Savitz, D. (2018). An Updated Meta-Analysis of the Association of
Serum PFOA and Birthweight, with an Evaluation of Potential Biases. Paper presented at
the ISEE Conference Abstracts.
Steenland, K., Tinker, S., Frisbee, S., Ducatman, A., & Vaccarino, V. (2009). Association of
perfluorooctanoic acid and perfluorooctane sulfonate with serum lipids among adults
living near a chemical plant. American journal of epidemiology, 770(10), 1268-1278.
Steenland, K., & Woskie, S. (2012). Cohort mortality study of workers exposed to
perfluorooctanoic acid. American journal of epidemiology, 176, 909-917.
Steenland, K., Zhao, L., & Winquist, A. (2015). A cohort incidence study of workers exposedto
perfluorooctanoic acid (PFOA). Occup Environ Med, 72, 373-380. doi: 10.1136/oemed-
2014-1023 643 73Workpl ace
Stoiber, T., Evans, S., & Naidenko, O. V. (2020). Disposal of Products and Materials Containing
Per- and Polyfluoroalkyl Substances (PFAS): A Cyclical Problem. Chemosphere, 260,
127659. doi:https://doi.org/10.1016/i.chemosphere.2020.127659
Proposed PFAS Rule Economic Analysis
10-16
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Summers, R. S., Hooper, S., Shukairy, H., Solarik, G., & Owen, D. M. (1996). Assessing DBP
Yield: Uniform Formation Conditions. J. Amer. Water Works Assoc., 88(6), 80-93.
Summers, R. S., Kim, S. M., Shimabuku, K., Chae, S. H., & Corwin, C. J. (2013). Granular
activated carbon adsorption of MIB in the presence of dissolved organic matter. Water
research, 47(10), 3507-3513.
Sunderland, E. M., Hu, X. C., Dassuncao, C., Tokranov, A. K., Wagner, C. C., & Allen, J. G.
(2019). A review of the pathways of human exposure to poly- and perfluoroalkyl
substances (PFASs) and present understanding of health effects. J Expo Sci Environ
Epidemiol, 29(2), 131-147. doi:10.1038/s41370-018-0094-l
Temple, J. A., Reynolds, A. J., & Arteaga, I. (2010). Low birth weight, preschool education, and
school remediation. Education and urban society, 42(6), 705-729.
Terry, L. G., & R.S., S. (2018). Biodegradable organic matter and rapid-rate biofilter
performance: A review. Water research, 128, 234-245.
Thom, T., Kannel, W., Silbershatz, H., & D'Agostino, R. (2001). Cardiovascular diseases in the
United States and prevention approaches. Hurst 's the heart, 1, 3-17.
Toth, P. P., Granowitz, C., Hull, M., Anderson, A., & Philip, S. (2019). Long-term statin
persistence is poor among high-risk patients with dyslipidemia: a real-world
administrative claims analysis. Lipids in Health and Disease, 18(1). doi:10.1186/sl2944-
019-1099-z
U.S. Bureau of Economic Analysis. (2022). Gross Domestic Product: Implicit Price Deflator
[GDPDEF], retrievedfrom FRED, Federal Reserve Bank of St. Louis;
https: fred.stlouisfed.ors series GDPDEI'', September 14, 2022.
U.S. Bureau of Labor Statistics. (2021). Medical care in U.S. city average, all urban consumers,
not seasonally adjusted, 1980-2021 CUUR0000SAM.
U.S. Bureau of Labor Statistics. (2022). Table 1. Employment Cost Index for total compensation,
by occupational group and industry. Employment Cost Index Historical Listing - Volume
III. Retrieved from https://www.bls.gov/web/eci/eci-current-nominal-.dollar.pdf
U.S. Census Bureau. (2010). American Community Survey Five-Year Estimates. American
Community Survey Demographic and Housing Estimates. Retrieved from
https://www.census.gov/programs-survevs/acs/technical-documentation/table-and-
geography-changes.2010.html
U.S. Census Bureau. (2020a). 2019 Population Estimates by Age, Sex, Race and Hispanic
Origin. Retrieved from https://www.census.gov/newsroom/press-kits/2020/population-
estimates-detailed.html
U.S. Census Bureau. (2020b). Average Number of People per Household, by Race and Hispanic
Origin, Marital Status, Age, and Education of Householder. Retrieved from
https://www2.census.gov/programs-survevs/demo/tables/families/2020/cps-
2020/tabavgl.xls
U.S. Energy Information Administration. (2021). Annual Energy Outlook 2021. Retrieved
August 10, 2021 https://www.eia.gov/outlooks/aeo/
U.S. Environmental Protection Agency. (1989). Surface Water Treatment Ride.
U.S. Environmental Protection Agency. (1994). EPA09: Establish a Blueprint for Environmental
Justice Throughout EPA 's Operations. Retrieved from
https://govinfo.librarv.unt.edu/npr/library/reports/EPA9.html
U.S. Environmental Protection Agency. (1996). DBP ICR Analytical Methods Manual.
Cincinnati, Ohio
Proposed PFAS Rule Economic Analysis
10-17
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
U.S. Environmental Protection Agency. (1997). Discussion Summary: EPA Technology Design
Workshop. Retrieved from https://downloads.regulations.gOv/EPA-HO-OW-2018-0780-
0197/content.pdf
U.S. Environmental Protection Agency. (1998a). Alachlor Technical Fact Sheet.
https://www3.epa.sov/pesticides/chem search res actions reregistration fsPC-
090501 l-Dec-98.pdf.
U.S. Environmental Protection Agency. (1998b). Information for States on Developing
Affordability Criteria for Drinking Water. EPA 816-R-98-002.
U.S. Environmental Protection Agency. (1998c). National Primary Drinking Water Regulations:
Regulatory Flexibility Act. Federal Register 63(30):7620. February 13, 2008.
U.S. Environmental Protection Agency. (1998d). Primary Drinking Water Regulations, Code of
Federal Regulations Title 40 Chapter I Subchapter D Part 141
httDs://www.ecfr.gov/current/tiile-40/chaDter-I/subchaDter-D/vart-141.
U.S. Environmental Protection Agency. (1998e). Stage 1 Disinfectants and Disinfection
Byproducts Ride. Retrieved from https://www.govinfo.gov/content/pkg/FR-1998-12-
16/pdf/98-32887.pdf
U.S. Environmental Protection Agency. (2000). Geometries and Characteristics of Public Water
Systems. (EPA-815-R-00-24).
U.S. Environmental Protection Agency. (2004). The Standardized Monitoring Framework: A
Quick Reference Guide. (EPA-816-F-04-010). Retrieved from
https://www.epa.gov/sites/default/files/2020-05/documents/smf 2020 final 508.pdf
U.S. Environmental Protection Agency. (2005a). Drinking Water Criteria Document for
Brominated Trihalomethanes. (EPA-822-R-05-011).
U.S. Environmental Protection Agency. (2005b). Economic Analysis for the Final Stage 2
Disinfectants and Disinfection Byproducts Ride. (EPA 815-R-05-010).
U.S. Environmental Protection Agency. (2005c). Guidelines for Carcinogen Risk Assessment.
Washington, D.C. Retrieved from
https://www3.epa.gov/airtoxics/cancer guidelines final 3-25-05.pdf
U.S. Environmental Protection Agency. (2006a). National Primary Drinking Water Regulations:
Long Term 2 Enhanced Surface Water Treatment Rule; Final Rule, EPA 815-Z-06-001.
(EPA 815-Z-06-001).
U.S. Environmental Protection Agency. (2006b). National Primary Drinking Water Regulations:
Stage 2 Disinfectants and Disinfection Byproducts Rule; Final Rule. 71FR 388. .
Retrieved from https://www.federalregister.gov/documents/2006/01/04/Q6-3/national-
primarv-drinking-water-regulations-stage-2-disinfectants-and-disinfection-bvproducts-
rule
U.S. Environmental Protection Agency. (2007a). Atrazine Chemical Summary.
https://archive.epa.gov region5 teach web pdf atrazine summary.pdf.
U.S. Environmental Protection Agency. (2007b). Office Ground Water and Drinking Water's
Error Code Tracking Tool [for SDWIS/Fed], Developed 2007.
U.S. Environmental Protection Agency. (2008). Final Ozone NAAOSRegulatory Impact
Analysis. (EPA-452/R-08-003). Research Triangle Park, North Carolina Retrieved from
https://www3.epa.gov/ttn/ecas/docs/ria/naaqs-o3 ria final 2008-03.pdf
U.S. Environmental Protection Agency. (2009). Community Water System Survey, Volume II:
Detailed Tables and Survey Methodology. EPA 815-R-09-002. Retrieved from
http://nepis. epa.gov/Exe/ZyPDF. cgi?Dockev=P 1009USA.txt
Proposed PFAS Rule Economic Analysis
10-18
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
U.S. Environmental Protection Agency. (2010a). Guidelines for Preparing Economic Analyses.
Paperwork Reduction Act, as amended. 44 USC 3501-44 USC 3521. Available at
https://www.gpo.gov/fdsvs/pkg/USCODE-2015-title44/pdf/USCODE-2015-title44-
chap35-subchapl-sec3501.pdf (2010b).
U.S. Environmental Protection Agency. (2011a). The Benefits and Costs of the Clean Air Act
fi'om 1990 to 2020. Final Report - Rev. A. Retrieved from
https://www.epa.gov/sites/default/files/2015-07/documents/fullreport rev a.pdf
U.S. Environmental Protection Agency. (201 lb). Exposure Factors Handbook, 2011 Edition
(Final). (EPA-600-R-09-025F). U.S. Environmental Protection Agency, Washington, DC
U.S. Environmental Protection Agency. (2012). Revisions to the Unregulated Contaminant
Monitoring Regulation (UCMR 3) for Public Water Systems; Final Ride. Federal Register
Retrieved from https://www.federalregister.gov/documents/2012/05/02/2012-
9978/revisions-to-the-unregulated-contaminant-monitoring-regulation-ucmr-3-for-public-
water-svstems
U.S. Environmental Protection Agency. (2014). Dichloromethane Technical Fact Sheet.
https: archive, epa.sov water archive web pdf archived-technical-fact-sheet-on-
dichloromethane.pdf.
U.S. Environmental Protection Agency. (2015). Benefit and Cost Analysis for the Effluent
Limitations Guidelines and Standards for the Steam Electric Power Generating Point
Source Category. (EPA-821-R-15-005).
U.S. Environmental Protection Agency. (2016a). The Data Management and Quality
Assurance Quality Control Process for the Third Six-Year Review Information Collection
Rule Dataset. (EPA 810-R-16-015).
U.S. Environmental Protection Agency. (2016b). Drinking Water Health Advisory for
Perfluorooctane Sulfonate (PFOS). (EPA822-R-16-004).
U.S. Environmental Protection Agency. (2016c). Drinking Water Health Advisory for
Perfluorooctanoic Acid (PFOA). (EPA 822-R-16-005).
U.S. Environmental Protection Agency. (2016d). Fact Sheet: PFOA& PFOS Drinking Water
Health Advisories. (EPA 800-F-16-003). Retrieved from
https://www.epa.gov/sites/default/files/2Q16-
06/documents/drinkingwaterhealthadvisories pfoa pfos updated 5.31.16.pdf
U.S. Environmental Protection Agency. (2016e). Health effects support document for
Perfluorooctane Sulfonate (PFOS). (EPA 822-R-l6-002). Retrieved from
https://www.epa.gov/sites/production/files/2016-05/documents/pfos hesd final 508.pdf
U.S. Environmental Protection Agency. (2016f). Health effects support document for
Perfluorooctanoic Acid (PFOA). (EPA 822-R-16-003). Retrieved from
https://www.epa.gov/sites/production/files/2016-Q5/documents/pfoa hesd final-plain.pdf
U.S. Environmental Protection Agency. (2016g). Six-Year Review 3 Technical Support
Document for DisinfectantsDisinfection Byproducts Rides. (EPA-810-R-16-012).
Retrieved from https://www.epa.gov/sites/production/files/2016-
12/documents/810rl6012.pdf
U.S. Environmental Protection Agency. (2016h). Technical Guidance for Assessing
Environmental Justice in Regulatory Analysis. Retrieved from
https://www.epa.gov/sites/default/files/2016-Q6/documents/eitg 5 6 16 v5.1.pdf
U.S. Environmental Protection Agency. (2016i). Technologies for Legionella Control in Premise
Plumbing Systems: Scientific Literature Review. EPA 810-R-16-001.
Proposed PFAS Rule Economic Analysis
10-19
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
U.S. Environmental Protection Agency. (2016j). Third Six-Year Review of National Primary
Drinking Water Standards. Retrieved from https://www.epa. gov/sites/default/files/2016-
12/documents/six year 3 review frn 12-21-16.pdf
U.S. Environmental Protection Agency. (2017). Third Unregulated Contaminant Monitoring
Rule: Occurrence Data. Retrieved from https://www.epa.gov/dwucmr/occurrence-data-
unregulated-contaminant-monitoring-rule#3
U.S. Environmental Protection Agency. (2018). Method 537.1: Determination of Selected Per-
and Polyfluorinated Alky J Substances in Drinking Water by Solid Phase Extraction and
Liquid Chromatography Tandem Mass Spectrometry (IX" MS MS), Version 1.0.
(EPA/600/R-18/352). Retrieved from
https://cfpub.epa.gov/si/si public record Report.cfm?dirEntrvId=343042&Lab=NERL
U.S. Environmental Protection Agency. (2019a). EJSCREEN Technical Documentation.
U.S. Environmental Protection Agency. (2019b). Method 533: Determination of Per- and
Polyfluoroalkyl Substances in Drinking Water by Isotope Dilution Anion Exchange Solid
Phase Extraction and Liquid Chromatography Tandem Mass Spectrometry. (EPA 815-B-
19-020). Retrieved from https://www.epa.gov/sites/default/files/2019-
12/documents/method-5 3 3-815bl9020.pdf
U.S. Environmental Protection Agency. (2019c). Occurrence Datafi'om the Third Unregulated
Contaminant Monitoring Ride (UCMR 3). Retrieved from
https://www.epa.gov/dwucmr/occurrence-data-unregulated-contaminant-monitoring-
rule#3
U.S. Environmental Protection Agency. (2019d). Supplemental Environmental Assessment for
Proposed Revisions to Effluent Limitations Guidelines and Standards for the Steam
Electric Power Generating Point Source Category (EA). EPA-821-R-19-010. .
U.S. Environmental Protection Agency. (2020a). Announcement of Preliminary Regulatory
Determinations for Contaminants on the Fourth Drinking Water Contaminant Candidate
List. EPA-HO-OW-2019-0583 FRL-10005-88-OW. 85 FR 14098. Retrieved from
https://www.federalregister.gOv/documents/2020/03/10/2020-04145/announcement-of-
preliminarv-regulatorv-determinations-for-contaminants-on-the-fourth-drinking-water
U.S. Environmental Protection Agency. (2020b). Interim Guidance on the Destruction and
Disposal of Perfluoroalkyl and Polyflaoroalkyl Substances and Materials Containing
Perfluoroalkyl and Polyflaoroalkyl Substances. Washington, D.C. Retrieved from
https://www.epa.gov/svstem/files/documents/2021-ll/epa-hq-olem-2020-0527-
0002 content.pdf
U.S. Environmental Protection Agency. (2021a). Analysis of Cardiovascular Disease Risk
Reduction as a Result of Reduced PFOA and PFOS Exposure in Drinking Water. (EPA
815-D-21-001). Retrieved from
https://sab.epa.gov/ords/sab/f?p=100:19:9687145615924:::RP.19:P19 ID:963
U.S. Environmental Protection Agency. (2021b). Announcement of Final Regidatory
Determinations for Contaminants on the Fourth Drinking Water Contaminant Candidate
List. (86 FR 12272). Retrieved from
https://www.federalregister.gov/documents/2021/03/03/2021-Q4184/announcement-of-
final-regulatorv-determinations-for-contaminants-on-the-fourth-drinking-water
U.S. Environmental Protection Agency. (2021c). Human Health Toxicity Values for
Hexafluoropropylene Oxide (HFPO) Dimer Acid and Its Ammonium Salt (CASRN13252-
13-6 and CASRN 62037-80-3) Also Known as "GenXChemicals". EPA-8232-PR-1821-
Proposed PFAS Rule Economic Analysis
10-20
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
001010. Retrieved from https://www.epa. gov/system/files/documents/2021-10/genx-
chemicals-toxicitv-assessment tech-edited oct-21-508.pdf
U.S. Environmental Protection Agency. (2021d). Raman Health Toxicity Values for
Perfluorobutane Sulfonic Acid (CASRN 375-73-5) and Related Compound Potassium
Perfluorobutane Sulfonate (CASRN 29420-49-3). EPA 600 R-20 345F. Washington, DC
Retrieved from
file:///C:/Users/50367/Downloads/EPA_PFBS_HH%20Tox%20Study_FINAL_APR202
l.PDF
U.S. Environmental Protection Agency. (2021e). PFASStrategicRoadmap: EPA's Commitments
to Action 2021-2024. (EPA-100-K-21-002).
U.S. Environmental Protection Agency. (2021f). Proposed Approaches to the Derivation of a
Draft Maximum Contaminant Level Goal for Perflaorooctane Sulfonic Acid (PFOS)
(CASRN 1763-23-1) in Drinking Water. (EPA 822D21002).
U.S. Environmental Protection Agency. (2021g). Proposed Approaches to the Derivation of a
Draft Maximum Contaminant Level Goal for Perfluorooctanoic Acid (PFOA) (CASRN
335-67-1) in Drinking Water. (EPA 822D21001). Washington, DC
U.S. Environmental Protection Agency. (2021h). Safe Drinking Water Information System 2021
Quarter 4 Database. Retrieved from https://www.epa.gov/ground-water-and-drinking-
water/safe-drinking-water-information-svstem-sdwis-federal-reporting
U.S. Environmental Protection Agency. (2022a). 2022 EJSCREENBatch Rpackage. Retrieved
from: https: xithub.com CS/J'A l-JSCREIi\Hatch.
U.S. Environmental Protection Agency. (2022b). Comprehensive Environmental Response,
Compensation, and Liability Act Hazardous Substances: Designation of
Perfluorooctanoic Acid and Perfluorooctanesulfonic Acid. EPA-HO-OLEM-2019-0341-
0001. Washington, D.C. Retrieved from https://www.regulations.gov/document/EPA-
HO-QLEM-2019-0341 -0001
U.S. Environmental Protection Agency. (2022c). The Fourth Unregulated Contaminant
Monitoring Ride Occurrence Data. Office of Ground Water and Drinking Water. January
2022. Retrieved from https://www.epa.gov/dwucmr/occurrence-data-unregulated-
contaminant-monitoring-rule#4
U.S. Environmental Protection Agency. (2022d). Implementation of the Clean Water and
Drinking Water State Revolving Fund Provisions of the Bipartisan Infrastructure Law.
Retrieved from https://www.epa.gov/dwsrf/bipartisan-infrastructure-law-srf-
memorandum
U.S. Environmental Protection Agency. (2022e). IRIS ToxicologicalReview of
Perfluorobutanoic Acid (PFBA) and Related Salts (Final Report, 2022). EPA 635 R-
22/277F. Retrieved from https://www.epa.gov/chemical-research/iris-toxicological-
review-perfluorobutanoic-acid-pfba-and-related-salts-final
U.S. Environmental Protection Agency. (2022f). IRIS Toxicological Review of
Perfluorodecanoic Acid [CASRN 335-76-2J and Related Salts (Public Comment and
External Review Draft).
U.S. Environmental Protection Agency. (2022g). IRIS Toxicological Review of
Perfluorohexanoic Acid (PFHxA) and Related Salts (Public Comment and External
Review Draft). (EPA/635/R-21/312a). Retrieved from
https://cfpub.epa.gov/ncea/iris drafts/recordisplav.cfm?deid=352767
Proposed PFAS Rule Economic Analysis
10-21
March 2023
-------
DRAFT FOR PUBLIC COMMENT MARCH 2023
U.S. Environmental Protection Agency. (2022h). Lifetime Drinking Water Health Advisories for
Four Perflaoroalkyl Substances. Retrieved from
https://www.govinfo.gov/content/pkg/FR-2022-06-21/pdf/2022-13158.pdf
U.S. Environmental Protection Agency. (2022i). Microbial and Disinfection Byproduct Data
Files (2012-2019) fi'om EPA's Fourth Six-Year Review Information Collection Request.
Retrieved from https://www.epa.gov/dwsixyearreview/microbial-and-disinfection-
bvproduct-data-files-2012-2019-epas-fourth-six-vear
U.S. Environmental Protection Agency. (2022j). PFAS Explained. Retrieved from
https://www.epa.gov/pfas/pfas-explained
U.S. Environmental Protection Agency. (2022k). Review of EPA's Analyses to Support EPA's
National Primary Drinking Water Ridemaking for PFAS. EPA-SAB-22-008. Washington,
D.C. Retrieved from
https://sab.epa.gov/ords/sab/f?p=l00:18:16490947993:::RP.18:P18 ID:2601#report
U.S. Environmental Protection Agency. (2023a). Framework for Estimating Noncancer Health
Risks Associated with Mixtures of Per- and Polyflaoroalkyl Substances (PFAS). EPA-
822-P-23-003. Washington, D.C.
U.S. Environmental Protection Agency. (2023b). Public Comment Draft - Appendix: Toxicity
Assessment and Proposed Maximum Contaminant Level Goal for Perflaorooctane
Sulfonic Acid (PFOS) in Drinking Water. EPA-822-P-23-008. Washington, D.C.
U.S. Environmental Protection Agency. (2023c). Public Comment Draft - Appendix: Toxicity
Assessment and Proposed Maximum Contaminant Level Goal for Perflaorooctanoic Acid
(PFOA) in Drinking Water. EPA-822-P-23-006. Washington, D.C.
U.S. Environmental Protection Agency. (2023d). Public Comment Draft - Toxicity Assessment
and Proposed Maximum Contaminant Level Goal for Perflaorooctane Sulfonic Acid
(PFOS) in Drinking Water. EPA-822-P-23-007. Washington, D.C.
U.S. Environmental Protection Agency. (2023e). Public Comment Draft - Toxicity Assessment
and Proposed Maximum Contaminant Level Goal for Perflaorooctanoic Acid (PFOA) in
Drinking Water. EPA-822-P-23-005. Washington, D.C.
U.S. Environmental Protection Agency. (2023f). Technical Support Document - Best Available
Technologies and Small System Compliance Technologies for Per- and Polyflaoroalkyl
Substances (PFAS) in Drinking Water. EPA-822-P-23-009. Retrieved from
U.S. Environmental Protection Agency. (2023g). Technical Support Document - Per- and
Polyfluoroalkyl Substances (PFAS) Occurrence & Contaminant Background. EPA-822-
P-23-010.
U.S. Environmental Protection Agency. (2023h). Technical Support Document - Technologies
and Cost for Removing Per- and Polyflaoroalkyl Substances (PFAS) fi'om Drinking
Water. EPA-822-P-23-011.
U.S. Environmental Protection Agency. (2023i). Work Breakdown Structure-Based Cost Model
for Granular Activated Carbon Drinking Water Treatment. EPA-822-P-23-012.
Retrieved from https://www.epa.gov/svstem/files/documents/2022-03/gac-
documentation-.pdf O.pdf
U.S. Environmental Protection Agency. (2023j). Work Breakdown Structure-Based Cost Model
for Ion Exchange Treatment of Per- and Polyflaoroalkyl Substances (PFAS) in Drinking
Water. EPA-822-P-23-013.
U.S. Environmental Protection Agency. (2023k). Work Breakdown Structure-Based Cost Model
for Nontreatment Options for Drinking Water Compliance. EPA-822-P-23-014.
Proposed PFAS Rule Economic Analysis
10-22
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Retrieved from https://www.epa.gov/sites/default/files/2019-07/documents/wbs-
nontreatm ent-docum entati on-i une-2019. pdf
U.S. Environmental Protection Agency. (20231). Work Breakdown Structure-Based Cost Model
for Reverse Osmosis Nanofiltration Drinking Water Treatment. EPA-822-P-23-015.
U.S. Food and Drug Administration. (2022). Bottled Water Everywhere: Keeping it Safe.
Retrieved from https://www.fda.gov/consumers/consumer-updates/bottled-water-
everywhere-keeping-it-safe
U.S. Office of Management and Budget. (2003). Circular A-4: Regulatory Analysis.
Washington, DC: OMB
Uche, U. I., Evans, S., Rundquist, S., Campbell, C., & Naidenko, O. V. (2021). Community-
Level Analysis of Drinking Water Data Highlights the Importance of Drinking Water
Metrics for the State, Federal Environmental Health Justice Priorities in the United
States. International journal of environmental research and public health, 75(19), 10401.
Ullah, M., Moin, A. T., Araf, Y., Bhuiyan, A. R., Griffiths, M. D., & Gozal, D. (2020). Potential
effects of the COVID-19 pandemic on future birth rate. Frontiers in public health, 8, 893.
Unfunded Mandates Reform Act of 1995. Public Law 104-4. 110 Stat. 48. 104th Congress.
Available at https://www.gpo.gov/fdsvs/pkg/PLAW-104publ4/pdf/PLAW-104publ4.pdf..
(1995).
Valvi, D., Oulhote, Y., Weihe, P., Dalgard, C., Bjerve, K. S., Steuerwald, U., & Grandjean, P.
(2017). Gestational diabetes and offspring birth size at elevated environmental pollutant
exposures. Environment international, 107, 205-215.
Velten, S., Hammes, F., Boiler, M., & Egli, T. (2007). Rapid and direct estimation of active
biomass on granular activated carbon through adenosine tri-phosphate (ATP)
determination. Water research, -(1(9), 1973-1983.
Verner, M. A., Loccisano, A. E., Morken, N. H, Yoon, M., Wu, H., McDougall, R., . . .
Longnecker, M. P. (2015). Associations of Perfluoroalkyl Substances (PFAS) with Lower
Birth Weight: An Evaluation of Potential Confounding by Glomerular Filtration Rate
Using a Physiologically Based Pharmacokinetic Model (PBPK). Environmental health
perspectives, 725(12), 1317-1324. doi: 10.1289/ehp. 1408837
Vieira, V. M., Hoffman, K., Shin, H., Weinberg, J. M., Webster, T. F., & Fletcher, T. (2013).
Perfluorooctanoic acid exposure and cancer outcomes in a contaminated community: A
geographic analysis. Environmental health perspectives, 727(3), 318-323.
Vieux, F., Maillot, M., Rehm, C. D., Barrios, P., & Drewnowski, A. (2020). Trends in tap and
bottled water consumption among children and adults in the United States: analyses of
NHANES 2011-16 data. Nutrition Journal, 79(1), 10. doi:10.1186/sl2937-020-0523-6
Villanueva, C. M., Cantor, K. P., Cordier, S., Jaakkola, J. J. K., King, W. D., Lynch, C. F., . . .
Kogevinas, M. (2004). Disinfection byproducts and bladder cancer: a pooled analysis.
Epidemiology, 75(3), 357-367.
Villanueva, C. M., Cantor, K. P., King, W. D., Jaakkola, J. J., Cordier, S., Lynch, C. F., . . .
Kogevinas, M. (2006). Total and specific fluid consumption as determinants of bladder
cancer risk. International Journal of Cancer, 775(8), 2040-2047.
doi:https://doi.org/10.1002/iic.21587
Villanueva, C. M., Fernandez, F., Malats, N., Grimalt, J. O., & Kogevinas, M. (2003). Meta-
analysis of studies on individual consumption of chlorinated drinking water and bladder
cancer. J Epidemiol'Community Health, 57(3), 166-173. doi:10.1136/jech.57.3.166
Proposed PFAS Rule Economic Analysis
10-23
March 2023
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Wang, L., Vacs Renwick, D., & Regli, S. (2019). Re-assessing effects of bromide and granular
activated carbon on disinfection byproduct formation. AWWA Water Science, 7(4),
el 147. doi:https: //doi. or g/10.1002/aws2.114 7
Wang, Y., Adgent, M., Su, P.-H., Chen, H.-Y., Chen, P.-C., Hsiung, C. A., & Wang, S.-L.
(2016). Prenatal exposure to perfluorocarboxylic acids (PFCAs) and fetal and postnatal
growth in the Taiwan Maternal and Infant Cohort Study. Environmental Health
Perspectives, 124(11), 1794-1800.
Water Research Foundation. (2016). Residential End Uses of Water, Version 2: Executive
Report. Retrieved from https://www.circleofblue.org/wp-
content/uploads/2016/04/WRF REU2016.pdf
Waterfield, G., Rogers, M., Grandjean, P., Auffhammer, M., & Sunding, D. (2020). Reducing
exposure to high levels of perfluorinated compounds in drinking water improves
reproductive outcomes: evidence from an intervention in Minnesota. Environmental
Health, 19, 1-11.
Weisman, R. J., Heinrich, A., Letkiewicz, F., Messner, M., Studer, K., Wang, L., & Regli, S.
(2022). Estimating national exposures and potential bladder cancer cases associated with
chlorination DBPs in US drinking water. Environmental health perspectives, 750(8),
087002. doi:https://doi.org/10.1289/EHP9985
Widefield Water and Sanitation District. (2018). From Pilot to Full-Scale: A Case Study for the
Treatment of Perfluorinated Compounds (PFCs) with Ion Exchange. Paper presented at
the AWWA Water Quality Technology Conference, Toronto, Ontario, Canada.
Wikstrom, S., Lin, P.-I., Lindh, C. H., Shu, H., & Bornehag, C.-G. (2020). Maternal serum levels
of perfluoroalkyl substances in early pregnancy and offspring birth weight. Pediatric
research, 87(6), 1093-1099.
Wikstrom, S., Lindh, C. H., Shu, H., & Bornehag, C.-G. (2019). Early pregnancy serum levels of
perfluoroalkyl substances and risk of preeclampsia in Swedish women. Scientific reports,
9(1), 1-7.
Wilder, L., Worley, R., & Breysse, P. (2017). DIRECT FROM ATSDR: Community Exposures
to Per-and Polyfluoroalkyl Substances in Drinking Water: A National Issue. Journal of
Environmental Health, 80(2), 38-41.
Williams, O. (1989). The Atherosclerosis Risk in Communities (ARIC) study-design and
objectives. American journal of epidemiology, 129(4), 687-702.
Windham, G., & Fenster, L. (2008). Environmental contaminants and pregnancy outcomes.
Fertility and sterility, 89(2), el 11-el 16.
Woods & Poole Economics Inc. (2021). Complete Demographic Database. Retrieved from:
https://www.woodsandpoole.com/our-databases/united-states/all-geographies/
Yang, Q., Guo, X., Sun, P., Chen, Y., Zhang, W., & Gao, A. (2018). Association of serum levels
of perfluoroalkyl substances (PFASs) with the metabolic syndrome (MetS) in Chinese
male adults: A cross-sectional study. Science of the total environment, 621, 1542-1549.
doi:10.1016/j.scitotenv.2017.10.074
Yao, Q., Gao, Y., Zhang, Y., Qin, K., Liew, Z., & Tian, Y. (2021). Associations of paternal and
maternal per-and polyfluoroalkyl substances exposure with cord serum reproductive
hormones, placental steroidogenic enzyme and birth weight. Chemosphere, 285, 131521.
Yapsakli, K., & Ce
-------
DRAFT FOR PUBLIC COMMENT
MARCH 2023
Yu, J., Lv, L., Lan, P., Zhang, S., Pan, B., & Zhang, W. (2012). Effect of effluent organic matter
on the adsorption of perfluorinated compounds onto activated carbon. Journal of
hazardous materials, 225, 99-106.
Yuan, J., Passeport, E., & Hofmann, R. (2022). Understanding adsorption and biodegradation in
granular activated carbon for drinking water treatment: A critical review. Water research,
210, 118026. doi:https://doi.org/10.1016/i.watres.2021.118026
Zachman, B. A., & Summers, R. S. (2010). Modeling TOC breakthrough in granular activated
carbon adsorbers. Journal of Environmental Engineering, 136(2), 204-210.
Zaggia, A., Conte, L., Falletti, L., Fant, M., & Chiorboli, A. (2016). Use of strong anion
exchange resins for the removal of perfluoroalkylated substances from contaminated
drinking water in batch and continuous pilot plants. Water research, 91, 137-146.
Zhang, Y., Vittinghoff, E., Pletcher, M. J., Allen, N. B., Zeki A1 Hazzouri, A., Yaffe, K., . . .
Ives, D. G. (2019). Associations of Blood Pressure and Cholesterol Levels During Young
Adulthood with Later Cardiovascular Events. Journal of the American College of
Cardiology, 74(3), 330-341.
Proposed PFAS Rule Economic Analysis
10-25
March 2023
------- |