Economic Analysis for the
Final Long Term 2 Enhanced
Surface Water Treatment Rule
-------
Office of Water (4606-M) EPA 815-R-06-001 December 2005 www.epa.gov/safewater
-------
Contents
Appendices vi
Exhibits vii
Acronyms xiii
Health Risk Reduction and Cost Analysis xvi
Executive Summary
ES.l Need for the Rule ES-1
ES.2 Consideration of Regulatory Alternatives ES-1
ES.3 Summary of the LT2ESWTR Requirements ES-3
ES.4 National Benefits and Costs of the LT2ESWTR ES-6
ES.4.1 Benefit Estimates ES-7
ES.4.2 National and Household Cost Estimates ES-10
ES.5 National Net Benefits and Summary of Comparison of Alternatives ES-12
Chapter 1: Introduction
1.1 Summary 1-1
.1.1 Monitoring and Treatment Requirements for Filtered Systems 1-1
.1.2 Monitoring and Treatment Requirements for Unfiltered Systems 1-7
.1.3 Requirements for Existing Uncovered Finished Water Reservoirs 1-7
.1.4 Disinfection Profiling and Benchmarking Requirements 1-8
.1.5 Implementation Timeline 1-8
1.2 Document Organization 1-9
Chapter 2: Statement of Need for the Rule
2.1 Introduction 2-1
2.2 Description of the Issue 2-1
2.3 Risk Balancing 2-2
2.4 Public Health Concerns to Be Addressed 2-3
2.4.1 Cryptosporidium 2-3
2.4.2 Uncovered Finished Water Reservoirs 2-6
2.5 Regulatory History 2-7
2.5.1 Statutory Authority for Promulgating the Rule 2-7
2.5.2 1979 Total Trihalomethane Rule 2-8
2.5.3 1989 Total Coliform Rule 2-8
2.5.4 1989 Surface Water Treatment Rule 2-8
2.5.5 1996 Information Collection Rule 2-9
2.5.6 1998 Interim Enhanced Surface Water Treatment Rule 2-9
2.5.7 1998 Stage 1 Disinfectants and Disinfection Byproducts Rule 2-10
2.5.8 2000 Proposed Ground Water Rule 2-10
2.5.9 2001 Filter Backwash Recycling Rule 2-11
Economic Analysis for the LT2ESWTR ~i December 2005
-------
2.5.10 2002 Long Term 1 Enhanced Surface Water Treatment Rule 2-11
2.5.11 2003 Proposed Stage 2 Disinfectants and Disinfection Byproducts Rule . . . 2-11
2.6 Economic Rationale for Regulation 2-11
Chapter 3: Consideration of Regulatory Alternatives
3.1 Introduction 3-1
3.2 Development Process for Regulatory Alternatives 3-1
3.3 Specific Regulatory Alternatives Considered in this EA 3-2
3.3.1 Summary of Bin Classification and Treatment Requirements for Regulatory
Alternatives 3-2
3.3.2 Additional Treatment for Direct Filtration Systems 3-4
3.4 Alternative Monitoring Approaches Considered 3-4
3.4.1 Indicators of Microbial Contamination 3-4
3.4.2 Cryptosporidium Monitoring Strategies for Bin Classification 3-5
Chapter 4: Baseline Conditions
4.1 Introduction 4-1
4.2 Data, Tools, and Processes Used in Baseline Development 4-2
4.2.1 ICRand ICRSS Observed Data 4-3
4.2.2 ICR and ICRSS Modeled Data and Method for Estimating Source Water
Occurrence 4-5
4.2.3 Surface Water Analytical Tool (SWAT) 4-8
4.3 Industry Profile 4-8
4.3.1 Public Water System Characterization 4-8
4.3.2 Systems, Plants, and Population Subject to the LT2ESWTR 4-10
4.3.3 Water Treatment Plant Design and Average Daily Flows 4-16
4.4 Baseline for Unfiltered Plants (Pre-LT2ESWTR) 4-19
4.4.1 Treatment Characterization for Unfiltered Plants 4-19
4.4.2 Number of Unfiltered Systems, Plants, and Population Served 4-20
4.4.3 Source Water Cryptosporidium Occurrence for Unfiltered Plants 4-21
4.4.4 Finished Water Cryptosporidium Occurrence for Unfiltered Plants 4-22
4.5 Baselines for Filtered Plants (Pre-LT2ESWTR) 4-24
4.5.1 Treatment Characterization for Filtered Plants 4-24
4.5.2 Number of Filtered Plants and Population Served 4-27
4.5.3 Source Water Cryptosporidium Occurrence for Filtered Plants 4-30
4.5.4 Finished Water Cryptosporidium Occurrence for Filtered Plants 4-35
4.5.5 Comparison of EPA Finished Water Cryptosporidium Estimates
with Aboytes et al. (2000) 4-40
4.5.6 Predicted Bin Classification for Filtered Plants 4-41
4.6 Baseline for Uncovered Finished Water Reservoirs 4-41
4.7 Households Incurring Costs Due to the LT2ESWTR 4-44
4.8 Summary of Uncertainties in Development of LT2ESWTR Baselines 4-47
Chapter 5: Benefits Analysis
5.1 Introduction 5-1
5.2 Quantified Health Benefits from Reduction in Exposure to Cryptosporidium 5-2
5.2.1 Overview of Risk Assessment Methodology 5-3
5.2.2 Hazard Identification 5-5
Economic Analysis for the LT2ESWTR ii December 2005
-------
5.2.3 Dose-Response Assessment 5-8
5.2.4 Exposure Assessment 5-23
5.2.4.1 Distribution of Infectious Cryptosporidium in Finished Water 5-23
5.2.4.2 Distribution of Individual Daily Drinking Water Consumption .... 5-30
5.2.4.3 Population Affected by the LT2ESWTR 5-32
5.2.5 Risk Model Structure 5-34
5.2.6 Individual Annual Risk Distributions 5-41
5.2.7 General Population Risk-Number of Cases Avoided 5-43
5.2.7.1 Unfiltered Systems 5-45
5.2.7.2 Filtered Systems 5-48
5.2.8 Reduction in Sensitive Subpopulation Risk 5-48
5.3 Monetized Benefits from Reduction in Exposure to Cryptosporidium Resulting f
rom the LT2ESWTR 5-48
5.3.1 Value of Reduction in Cryptosporidiosis Cases
5-49
5.3.
5.3.
5.3.
5.3.
5.3.
5.3.
.1 Value of Illnesses Avoided 5-49
.2 Value of Avoiding Fatal Cases of Cryptosporidiosis 5-58
.3 Measuring Benefits Over the LT2ESWTR Implementation
Schedule 5-58
.4 Adjustment for Income Elasticity 5-58
.5 Present Value of Future Benefits 5-60
.6 Summary of Quantified Benefits of LT2ESWTR 5-61
5.3.2 Monetization of Benefits to Sensitive Subpopulations 5-66
5.4 Summary of Uncertainties 5-67
5.6 Other Benefits of LT2ESWTR Provisions 5-73
5.6.1 Reduction in Outbreak Risk 5-73
5.6.2 Costs to Households to Avert Infection 5-74
5.6.3 Enhanced Aesthetic Water Quality 5-74
5.6.4 Risk Reduction from Co-occurring and Emerging Pathogens 5-74
5.6.5 Benefits from Reduction in Disinfection 5-75
5.6.6 Benefits from Other Rule Provisions 5-75
5.6.7 Summary of Nonqualified Benefits 5-76
Chapter 6: Cost Analysis
6.1 Introduction 6-1
6.1.1 Cost Description and Assumptions 6-2
6.2 Rule Implementation Costs 6-6
6.2.1 PWSs 6-6
6.2.2 States and Other Primacy Agencies 6-6
6.3 Source Water Monitoring for Initial Bin Classification Costs 6-7
6.3.1 PWSs 6-7
6.3.2 State and Other Primacy Agency Costs 6-8
6.4 Treatment Costs 6-9
6.4.1 Toolbox Options and Unit Costs 6-11
6.4.1.1 Toolbox Options Not Considered in the Cost Analysis 6-11
6 A.I.2 Technologies Considered for the LT2ESWTR Cost Analysis 6-12
6.4.2 Compliance Forecast and Technology Selection 6-15
6.4.3 Capital and Annual Costs 6-16
6.5 Costs of Treatment for Unfiltered Plants 6-17
Economic Analysis for the LT2ESWTR in December 2005
-------
6.6 Costs of Disinfection Profiling and Benchmarking and of Technology Reporting . . . 6-19
6.7 Costs of Treatment for Uncovered Finished Water Reservoirs 6-20
6.7.1 Unit Costs 6-20
6.7.2 Compliance Forecast and Technology Selection 6-21
6.7.3 Total Annual Treatment Costs 6-21
6.8 Future Source Water Monitoring 6-22
6.9 Summary of the National Costs of the LT2ESWTR 6-23
6.10 Household Costs 6-27
6.11 Summary of Uncertainties and Sensitivity Analyses 6-30
6.11.1 Cryptosporidium Occurrence Data Sets 6-31
6.11.2 Sensitivity Analysis of Influent Bromide Levels on Technology Selection
for Filtered Plants 6-33
6.12 Unqualified Costs 6-33
6.13 Comparison of Regulatory Alternatives 6-34
Chapter 7: Economic Impact Analysis
7.1 Introduction 7-1
7.2 Regulatory Flexibility Act and Small Business Regulatory Enforcement Fairness Act 7-1
7.3 Small-System Affordability 7-3
7.4 Feasible Treatment Technologies for All Systems 7-4
7.5 Effect of Compliance with the LT2ESWTR on the Technical, Managerial, and
Financial Capacity of Public Water Systems 7-4
7.5.1 Requirements ofthe LT2ESWTR 7-5
7.5.2 Systems Subjectto the LT2ESWTR 7-6
7.5.3 Impact ofthe LT2ESWTR on System Capacity 7-6
7.6 Paperwork Reduction Act 7-13
7.7 Unfunded Mandates Reform Act 7-14
7.7.1 Social Benefits and Costs 7-16
7.7.2 Disproportionate Budgetary Effects 7-18
7.7.3 Macroeconomic Effects 7-22
7.7.4 Consultation with Small Governments 7-23
7.7.5 Consultation with State, Local, and Tribal Governments 7-23
7.7.6 Regulatory Alternatives Considered 7-23
7.7.7 Impacts on Small Governments 7-24
7.8 Indian Tribal Governments 7-24
7.9 Impacts on Sensitive Subpopulations 7-29
7.9.1 Impacts on the Immunocompromised 7-30
7.9.2 Protection of Children from Environmental Health Risks and Safety Risks . 7-30
7.10 Environmental Justice 7-31
7.11 Federalism 7-32
7.12 Actions Concerning Regulations That Significantly Affect Energy Supply,
Distribution, or Use 7-33
Economic Analysis for the LT2ESWTR iv December 2005
-------
Chapter 8: Comparison of Benefits and Costs of the LT2ESWTR
8.1 Introduction 8-1
8.2 Summary of National Benefits, Costs, and Net Benefits of the Preferred Regulatory
Alternative 8-1
8.2.1 National Benefits Summary 8-4
8.2.2 National Cost Summary 8-7
8.2.3 National Net Benefits 8-9
8.3 Comparison of Regulatory Alternatives 8-11
8.3.1 Comparison of Benefits and Costs 8-12
8.3.2 Cost-Effectiveness Measures 8-29
8.4 Effect of Uncertainties on the Benefit-Cost Comparisons 8-35
8.5 Summary of Benefit and Cost Comparisons 8-37
Chapter 9: References
Economic Analysis for the LT2ESWTR v December 2005
-------
Appendices
Appendix A: Pre-LT2ESWTR Removal Credit
Appendix B: Characterizing Cryptosporidium Concentration and Methods for Predicting
Plant Binning
Appendix C: Benefits
Appendix D: National Costs for Rule Implementation and Monitoring
Appendix E: Unit Costs for Technologies Considered in the LT2ESWTR
Appendix F: Technology Selection Forecast Methodology
Appendix G: Technology Selection Results
Appendix H: Regulatory Flexibility Screening Analysis
Appendix I: Unit Costs for Uncovered Finished Water Reservoirs
Appendix J: Estimation of Household Costs
Appendix K: Additional Information on the Approach For Valuing Time Losses
Appendix L: Calculations Supporting The Cost of Illness (COI) Analysis
Appendix M: Small Community Surface Water and GWUDI Systems by State
Appendix N: Dose-Response Infectivity Analysis
Appendix O: Assigning LT2ESWTR Costs and Benefits
Appendix P: Sensitivity Analyses for Cost of Illness Values
Appendix Q: Cost Models
Appendix R: Sensitivity Analysis for AIDS-Related Mortality Rate
Appendix S: Analysis of Individual Risk by Initial Bin
Appendix T: Risk Assessment Model—Program and Data Files
Appendix U: Cost Effectiveness Analysis Using a Quality-Adjusted Life Years Approach
Economic Analysis for the LT2ESWTR vi December 2005
-------
Exhibits
Executive Summary
Exhibit ES.l: Summary of Treatment Requirements for Filtered Systems ES-3
Exhibit ES.2: Overview of Key LT2ESWTR Requirements ES-5
Exhibit ES.3: Implementation Timeline for LT2ESWTR ES-6
Exhibit ES.4: Summary of Nonquantified Benefits and Groups Affected ES-7
Exhibit ES.5: Summary of Annual Illnesses and Deaths Avoided ES-9
Exhibit ES.6a: Summary of Monetized Benefits ES-9
Exhibit ES.6b: Summary of Monetized Benefits—Traditional Cost of Illness ES-10
Exhibit ES.7: Summary of System Costs ($ Millions, 2003$) ES-11
Exhibit ES.8: Summary of Annual Household Cost Increases1 ($ per Year, 2003$) ES-12
Exhibit ES.9a: Comparison of Mean Net Benefits for All Regulatory Alternatives—Enhanced
Cost of Illness ($Millions, 2003$) ES-13
Exhibit ES.9b: Comparison of Mean Net Benefitsfor All Regulatory Alternatives—Traditional
Cost of Illness ES-13
Exhibit ES.lOa: Comparison of Mean ES-15
Exhibit ES.lOb: Comparison of Mean—Traditional Cost of Illness ES-16
Exhibit ES.l la: Cost Effectiveness Analysis Based on Low, Medium, and High Estimate Dose
Response Models Using the Enhanced COI Approach, by Data Set,
by Rule Alternative, 3% and 7% Discount Rates ES-18
Exhibit ES.l Ib: Cost Effectiveness Analysis Based on Low, Medium, and High Estimate Dose
Response Models Using the Traditional COI Approach, by Data Set, by Rule
Alternative, 3% and 7% Discount Rates ES-18
Chapter 1: Introduction
Exhibit 1.1: Bin Classifications and Treatment Requirements for Filtered Systems 1-4
Exhibit 1.2: Microbial Toolbox Components for the LT2ESWTR (To be used in addition to
existing treatment) 1-6
Exhibit 1.3: Implementation Time Line for LT2ESWTR for Filtered Systems 1-8
Chapter 2: Statement of Need for the Rule
Exhibit 2.1: Reported Cryptosporidiosis Outbreaks in U.S. Drinking Water Systems 2-5
Chapter 3: Consideration of Regulatory Alternatives
Exhibit 3.1: Summary of Bin Requirements for Filtered Systems 3-3
Exhibit 3.2: Probability of Misclassification for Monitoring and Binning Strategies Considered
forthe LT2ESWTR 3-6
Chapter 4: Baseline Conditions
Exhibit 4.1: Comparison of ICR and ICRSS Data Collection Methods 4-4
Exhibit 4.2: Methodology for "Linking" Consecutive Surface Water CWSs and NTNCWSs to
Their Selling Systems 4-12
Exhibit 4.3a: Inventory of Unlinked and Linked Surface Water and GWUDI CWSs 4-13
Exhibit 4.3b: Inventory of Unlinked and Linked Surface Water and GWUDI NTNCWSs 4-13
Exhibit 4.3c: Inventory of Surface Water and GWUDI TNCWSs 4-14
Economic Analysis for the LT2ESWTR
vn
December 2005
-------
Exhibit 4.4a: Average Daily and Design Flow by System Size for Filtered Plants 4-18
Exhibit 4.4b: Average Daily and Design Flow by System Size for Unfiltered Plants 4-19
Exhibit 4.5: Treatment Baseline for Unfiltered Plants by System Size 4-21
Exhibit 4.6: Observed ICR Total Oocyst Occurrence in Source Water for Unfiltered Plants 4-22
Exhibit 4.7: Modeled Cryptosporidium Occurrence in Source Water: ICR Data for Unfiltered
Systems 4-23
Exhibit 4.8: Predicted System Bin Assignments for Unfiltered Systems, Based on Central
Tendency of Cryptosporidium Occurrence 4-24
Exhibit 4.9: Percentage of Plants Qualifying for Pre-LT2ESWTR Cryptosporidium Log
Reduction Credits for Existing Technologies 4-25
Exhibit 4.10: Predicted Percentage of Plants Using Advanced Technologies Following
Implementation of the Stage 2 DBPR 4-26
Exhibit 4.11: Implementation and Monitoring Baseline for Filtered Systems 4-28
Exhibit 4.12: Treatment Baseline for Filtered Plants 4-30
Exhibit 4.13: Summary of Observed Cryptosporidium Total Oocyst Occurrence in Source
Water—Filtered Plant Data 4-31
Exhibit 4.14: Modeled Cryptosporidium Occurrence in Source Water—ICR Data for
Filtered Systems 4-32
Exhibit 4.15: Modeled Cryptosporidium Occurrence in Source Water—ICRSSM Data 4-33
Exhibit 4.16: Modeled Cryptosporidium Occurrence in Source Water—ICRSSL Data 4-33
Exhibit 4.17: Comparison of Modeled Cryptosporidium Occurrence in Source Water by
Data Set, Median Curves Only 4-34
Exhibit 4.18: Summary of Modeled Cryptosporidium Total Oocyst Occurrence in Source
Water—Median Curves Only 4-35
Exhibit 4.19: Predicted Ranges of Cryptosporidium Reduction Pre-LT2ESWTR 4-36
Exhibit 4.20b: Distribution of Cryptosporidium Reduction in Large Systems Pre-LT2ESWTR,
Standard Estimate 4-38
Exhibit 4.20d: Distribution of Cryptosporidium Reduction in Large Systems Pre-LT2ESWTR,
Estimate With 0.5 Log Reduction Credit 4-38
Exhibit 4.20c: Distribution of Cryptosporidium Reduction in Small Systems Pre-LT2ESWTR,
Estimate With 0.5 Log Reduction Credit 4-38
Exhibit 4.20a: Distribution of Cryptosporidium Reduction in Small Systems Pre-LT2ESWTR,
Standard Estimate 4-38
Exhibit 4.2la: Predicted Finished Water Cryptosporidium Occurrence Pre-LT2ESWTR,
Small Systems 4-39
Exhibit 4.2Ib: Predicted Finished Water Cryptosporidium Occurrence Pre-LT2ESWTR,
Large Systems 4-39
Exhibit 4.22: Predicted System Bin Assignments for Preferred Alternative 4-41
Exhibit 4.23: Baseline Numbers of Uncovered Finished Water Reservoirs 4-42
Exhibit 4.24: Surface Area and Flows for Uncovered Finished Water Reservoirs 4-43
Exhibit 4.25: Baseline Number of Uncovered Finished Water Reservoirs in Each System Size
Category 4-44
Exhibit 4.26: Universe of Households Affected by Rule Provisions 4-45
Exhibit 4.27: Baseline Numbers of Households Incurring Costs 4-46
Exhibit 4.28: Households Paying Treatment Costs for Uncovered Reservoirs 4-46
Exhibit 4.29: Mean Household Water Usage Rates by System Size 4-47
Exhibit 4.30: Summary of Uncertainties Affecting LT2ESWTR Baseline Estimates 4-48
Economic Analysis for the LT2ESWTR
vm
December 2005
-------
Chapter 5: Benefits Analysis
Exhibit 5.1: Risk Assessment Model Categories 5-3
Exhibit 5.2: Health Risk Assessment Framework for Cryptosporidium 5-5
Exhibit 5.3: Symptoms of 205 Patients with Confirmed Cases of Cryptosporidiosis During the
Milwaukee Outbreak 5-7
Exhibit 5.4 Characteristics of the Primary Model and Six Alternative Models 5-11
Exhibit 5.5: Comparison of Annual Illnesses Avoided Predicted by the Dose Response Models
Considered for Each Regulatory Alternative 5-18
Exhibit 5.6: Percent of Plants With Pre-LT2ESWTR Treatment Credit 5-27
Exhibit 5.7a: Predicted Log Removal Achieved for Systems without Credits 5-29
Exhibit 5.7b: Predicted Log Removal Achieved for Systems with Credits 5-30
Exhibit 5.8 Distribution of Individual Daily Drinking Water Consumption 5-32
Exhibit 5.9: Number of Systems, Population Served, and Annual National Exposure by
System Type 5-33
Exhibit 5.10: EPA Estimates for Exposure Days 5-34
Exhibit 5.11: Flowchart of Risk Model-Step 1: Computing Annual Individual
Risk of Illness 5-36
Exhibit 5.12: Overview of Risk Assessment Model Parameters 5-39
Exhibit 5.13: Secondary Spread and Secondary Attack Rates Associated with Cryptosporidiosis
Outbreaks 5-42
Exhibit 5.14: Annual Individual Risk Distributions Based Upon ICR Occurrence Data Set,
Filtered CWSs 5-44
Exhibit 5.15: Annual Individual Risk Distributions Based Upon ICR Occurrence Data Set,
Unfiltered CWSs 5-45
Exhibit 5.16: Annual Cases of Illness and Deaths Avoided for the LT2ESWTR, Preferred
Alternative, Unfiltered Systems, by Data Set 5-46
Exhibit 5.17: Annual Cases of Illness and Deaths Avoided Due to the LT2ESWTR, Preferred
Alternative, All Filtered Systems, by Data Set 5-47
Exhibit 5.18: Direct Medical Costs of a Case of Cryptosporidiosis 5-53
Exhibit 5.19: Days Lost and Days with Lost Productivity, by Severity of Illness 5-54
Exhibit 5.20: Weighted Average Days Lost for Particular Illness 5-54
Exhibit 5.21: Value of Time, 2003 5-55
Exhibit 5.22: Total Loss Per Case, Enhanced and Traditional COI, 2003$ 5-56
Exhibit 5.23: Yearly Total Loss Per Case, Enhanced and Traditional COI 5-57
Exhibit 5.24: Mean of Yearly Values for a Statistical Life 5-60
Exhibit 5.25a: Annualized Benefits of Illnesses and Deaths Avoided, Preferred Alternative,
Enhanced Cost of Illness 5-62
Exhibit 5.25b: Annualized Benefits of Illnesses and Deaths Avoided, Preferred Alternative,
Traditional Cost of Illness 5-63
Exhibit 5.26a: Annualized Benefits by Filtered and Unfiltered Systems, Preferred Alternative,
Enhanced Cost of Illness 5-64
Exhibit 5.26b: Annualized Benefits by Filtered and Unfiltered Systems, Preferred Alternative,
Traditional Cost of Illness 5-65
Exhibit 5.27a: Annualized Benefits by Illnesses and Deaths Avoided, Preferred Alternative,
Enhanced Cost of Illness 5-66
Exhibit 5.27b: Annualized Benefits by Illnesses and Deaths Avoided, Preferred Alternative,
Traditional Cost of Illness 5-66
Exhibit 5.28: Uncertainty and Variability in the Number of Illnesses Avoided 5-68
Exhibit 5.29: Summary of Uncertainties Affecting LT2ESWTR Benefits Estimates 5-69
Economic Analysis for the LT2ESWTR
IX
December 2005
-------
Exhibit 5.30a: Summary of Benefits of Annual Illnesses and Deaths Avoided from LT2ESWTR
for Regulatory Alternatives, Enhanced Cost of Illness 5-71
Exhibit 5.30b: Summary of Benefits of Annual Illnesses and Deaths Avoided from LT2ESWTR
for Regulatory Alternatives, Traditional Cost of Illness 5-72
Exhibit 5.31: Summary of Nonqualified Benefits 5-76
Chapter 6: Cost Analysis
Exhibit 6.1: Number of Systems and Plants Expected to Incur Costs, Preferred Alternative 6-3
Exhibit 6.2: Wage Rates by System Size 6-5
Exhibit 6.3: E. coll and Cryptosporidium Laboratory Costs 6-6
Exhibit 6.4: Estimated Costs of Implementation, Present Value— Discounted at 3 and 7 Percent,
Preferred Alternative 6-7
Exhibit 6.5: Cost Estimates for Initial Source Water Monitoring, Present Value— Discounted
at 3 and 7 Percent, Preferred Alternative (SMillions, 2003$) 6-9
Exhibit 6.6: Methodology for Estimating Treatment Costs 6-10
Exhibit 6.7: Toolbox Options Considered for the LT2ESWTR 6-14
Exhibit 6.8: Technology Selection Forecast for Filtered Plants 6-16
Exhibit 6.9: Treatment Costs for Filtered Systems, Discounted at 3 and 7 Percent,Preferred
Alternative (SMillions, 2003$) 6-17
Exhibit 6.10: Treatment Costs for Unfiltered Systems, Discounted at 3 and 7 Percent, Preferred
Alternative (SMillions, 2003$) 6-19
Exhibit 6.11: Disinfection Profiling and Benchmarking Estimated Costs, Present
Value-Discounted at 3 and 7 Percent, Preferred Alternative ($Millions, 2003$) 6-19
Exhibit 6.12: Technology Reporting Estimated Costs, Annualized at 3 and 7 Percent, Preferred
Alternative ($Millions, 2003$) 6-20
Exhibit 6.13a: Cost for Systems with Uncovered Finished Water Reservoirs, Annualized at 3 and 7
Percent ($Millions, 2003$) 6-22
Exhibit 6.13b: Reporting Cost for Systems with Uncovered Finished Water Reservoirs, Annualized
at 3 and 7 Percent ($Millions, 2003$) 6-22
Exhibit 6.14: Future Monitoring Cost Estimates, Present Value-Discounted at 3 and 7 Percent,
Preferred Alternative ($Millions, 2003$) 6-23
Exhibit 6.15: Initial Capital and One-Time Costs, Undiscounted, Preferred Alternative
($Millions, 2003$) 6-24
Exhibit 6.16: Annualized Total Costs-Discounted at 3 and 7 Percent, Preferred Alternative
($Millions, 2003$) 6-25
Exhibit 6.17a: Annualized Treatment Costs by System Size, Preferred Alternative, 3 Percent
Discount Rate ($Millions, 2003$) 6-26
Exhibit 6.17b: Annualized Treatment Costs by System Size, Preferred Alternative, 7 Percent
Discount Rate ($Millions, 2003$) 6-27
Exhibit 6.18: Summary of Annual Per-Household Cost1 Increases, Preferred Alternative ($/Year) . 6-29
Exhibit 6.19: Summary of Uncertainties Affecting LT2ESWTR Cost Estimates 6-30
Exhibit 6.20: Cost Model Estimates by Occurrence Distributions and Unit Cost Uncertainty 6-32
Exhibit 6.21: Sensitivity of Technology Selection to Influent Bromide Concentration for
Filtered Plants 6-33
Exhibit 6.22a: Comparison by Regulatory Alternative of Total Costs, Annualized at 3 Percent for
Filtered Plants ($Millions, 2003$) 6-34
Exhibit 6.22b: Comparison by Regulatory Alternative of Total Costs, Annualized at 7 Percent for
Filtered Plants ($Millions, 2003$) 6-35
Economic Analysis for the LT2ESWTR
December 2005
-------
Chapter 7: Economic Impact Analysis
Exhibit 7. la: Estimated Impacts of the LT2ESWTR on Small Systems' Technical,
Managerial, and Financial Capacity 7-8
Exhibit 7. Ib: Estimated Impacts of the LT2ESWTR on Large Systems' Technical,
Managerial, and Financial Capacity 7-9
Exhibit 7.2: Average Annual Burden Hours and Costs for the LT2ESWTR Information
Collection Request 7-14
Exhibit 7.3: Annualized Value of Public and Private Costs for the LT2ESWTR
(Annualized at 3 and 7 Percent) 7-15
Exhibit 7.4: Total Annualized Benefits and Costs of Regulatory Alternatives (SMillions, 2003$) 7-17
Exhibit 7.5: Percent of Population of CWSs Served by Small Surface and GWUDI
Systems by State 7-20
Exhibit 7.6: Number of Small Surface and GWUDI Systems by State 7-22
Exhibit 7.7: Numbers of Indian Tribal Public Water Systems Using Surface Water Sources .... 7-25
Exhibit 7.8: Number of Tribal Systems and Percent of Systems Nationally That Will Incur
Costs Due to the LT2ESWTR 7-27
Exhibit 7.9a: Estimates of the Total Annualized Costs Incurred by Indian Tribal Public
Water Systems Due to the LT2ESWTR (Annualized at 3 Percent) 7-28
Exhibit 7.9b: Estimates of the Total Annualized Costs Incurred by Indian Tribal Public
Water Systems Due to the LT2ESWTR (Annualized at 7 Percent) 7-29
Exhibit 7.10: Total Increased Annual National Energy Usage Attributable to the LT2ESWTR . . . 7-36
Exhibit 7.11: Sample Calculation for Determining Increase in Energy Usage
(Plants Predicted to Add UV) 7-37
Chapter 8: Comparison of Benefits and Costs of the LT2ESWTR
Exhibit 8.la: Summary of Undiscounted Benefit and Cost Estimates by Year Incurred, Preferred
Alternative, ICR Data Set, Enhanced COI 8-2
Exhibit 8.1b: Summary of Undiscounted Benefit and Cost Estimates by Year Incurred, Preferred
Alternative, ICR Data Set, Traditional COI 8-3
Exhibit 8.2: Summary of Nonqualified Benefits and Groups Affected 8-4
Exhibit 8.3: Summary of Annual Avoided Illnesses and Deaths, Preferred Alternative 8-5
Exhibit 8.4a: Summary of Quantified Benefits, Preferred Alternative—Enhanced Cost of Illness .. 8-6
Exhibit 8.4b: Summary of Quantified Benefits, Preferred Alternative—Traditional Cost of Illness . 8-6
Exhibit 8.5: Summary of the Costs for the LT2ESWTR Preferred Regulatory Alternative
(SMillions, 2003$) 8-8
Exhibit 8.6a: Mean Net Benefits, Preferred Alternative—Enhanced Cost of Illness
(SMillions, 2003$) 8-9
Exhibit 8.6b: Mean Net Benefits, Preferred Alternative—Traditional Cost of Illness 8-10
Exhibit 8.7a: Breakeven Points, Enhanced COI (Number of Avoided Illnesses Needed to Break
Even with Cost Estimates) 8-11
Exhibit 8.7b: Breakeven Points, Traditional COI (Number of Avoided Illnesses Needed to Break
Even with Cost Estimates) 8-11
Exhibit 8.8: Summary of Binning and Treatment Scenarios for Filtered Systems for All
Regulatory Alternatives 8-12
Exhibit 8.9: Comparison of Number of Illnesses and Deaths Avoided for All Regulatory
Alternatives 8-13
Exhibit 8. lOa: Comparison of Annualized Value of Illnesses and Deaths Avoided for
All Regulatory Alternatives 8-14
Economic Analysis for the LT2ESWTR
XI
December 2005
-------
Exhibit 8.10b: Comparison of Annualized Value of Illnesses and Deaths Avoided for
All Regulatory Alternatives, Traditional COI 8-15
Exhibit 8.11: Comparison of Costs for All Regulatory Alternatives ($ Millions, 2003$) 8-16
Exhibit 8.12a: Comparison of Mean Net Benefits for All Regulatory Alternatives—Enhanced COI 8-17
Exhibit 8.12b: Comparison of Mean Net Benefits for All Regulatory Alternatives—Traditional
COI (Million $/Year) 8-18
Exhibit 8.13a: Incremental Net Benefits for All Alternatives, By Data Set, Enhanced COI
($Millions, 2003$) 8-19
Exhibit 8.13b: Incremental Net Benefits for All Alternatives, By Data Set, Traditional COI
($Millions, 2003$) 8-20
Exhibit 8.14a: Upper End of 90 Percent Confidence Bound as a Percent of Mean Estimate of
Benefits, By Data Set, Annualized at 3 Percent 8-22
Exhibit 8.14b: Upper End of 90 Percent Confidence Bound as a Percent of Mean Estimate of
Benefits, By Data Set, Annualized at 3 Percent, Traditional COI ($Millions, 2003$) 8-23
Exhibit 8.15a: Comparison of Regulatory Alternatives Ranked by Net Benefits, 3 Percent Cost,
Enhanced COI 8-25
Exhibit 8.15b: Comparison of Regulatory Alternatives Ranked by Net Benefits, 7 Percent Cost,
Enhanced COI 8-26
Exhibit 8.16a: Comparison of Regulatory Alternatives Ranked by Net Benefits, 3 Percent Cost,
Traditional COI 8-27
Exhibit 8.16b: Comparison of Regulatory Alternatives Ranked by Net Benefits, 7 Percent Cost,
Traditional COI 8-28
Exhibit 8.17a: Range of Annualized Costs at Mean Benefit Level, All Regulatory
Alternatives—Enhanced Cost of Illness 8-30
Exhibit 8.17b: Range of Annualized Costs at Mean Benefit Level, All Regulatory
Alternatives—Traditional Cost of Illness 8-31
Exhibit 8.18: Incremental Net Cost per Discounted Illness Avoided, By Discount Rate,
Data Set, and Alternative 8-33
Exhibit 8.19: Incremental Net Cost per Discounted Death Avoided, By Discount Rate,
Data Set, and Alternative 8-34
Exhibit 8.20: Benefit-Cost Ratios for Each Regulatory Alternative 8-35
Exhibit 8.21: Effects of Uncertainties on the National Estimates of Benefits and Costs 8-36
Chapter 9: References
Economic Analysis for the LT2ESWTR xii December 2005
-------
Acronyms and Notations
AIDS
AIPC
AMWA
ASDWA
AWWA
CC-PCR
CCR
CDBG
CDC
CFE
CPU
CL2
CLM
CLO2
COI
CPI
CSFII
CWS
CWSS
DBFs
DWSRF
EA
EO
FACA
FBRR
FR
FS
FTE
GDP
GWR
GWUDI
HAAS
HPC
HRRCA
ICR
ICR
ICRSS
ICRSSM
ICRSSL
ICMA
IDSE
IESWTR
IMS
IRFA
IRIS
LCR
LRAA
LSP
Acquired Immunodeficiency Syndrome
All Indian Pueblo Council
Association of Metropolitan Water Agencies
Association of State Drinking Water Administrators
American Water Works Association
Cell culture and polymerase chain reaction
Consumer Confidence Report Rule (1998)
Community Development Block Grant
Centers for Disease Control and Prevention
Combined Filter Effluent
Colony forming unit
Chlorine
Chloramines
Chlorine Dioxide
Cost of Illness
Consumer Price Index
Continuing Survey of Food Intakes by Individuals
Community Water System
Community Water Systems Survey
Disinfection Byproducts
Drinking Water State Revolving Fund
Economic Analysis
Executive Order
Federal Advisory Committees Act
Filter Backwash Recycling Rule (May, 2001)
Federal Register
Flowing stream
Full-time Equivalent Employee
Gross Domestic Product
Ground Water Rule (proposed 2000)
Ground Water Under the Direct Influence of Surface Water
Haloacetic Acids [total of five]
Heterotrophic Plate Count
Health Risk Reduction and Cost Analysis
Information Collection Request
Information Collection Rule (1996)
Information Collection Rule Supplemental Survey
Information Collection Rule Supplemental Survey Medium Systems
Information Collection Rule Supplemental Survey Large Systems
International City/County Management Association
Initial Distribution System Evaluation
Interim Enhanced Surface Water Treatment Rule (1998)
Immunomagnetic separation
Initial Regulatory Flexibility Analysis
Integrated Risk Information System
Lead and Copper Rule (1992)
Locational Running Annual Average
Lab Spiking Program
Economic Analysis for the LT2ESWTR
xm
December 2005
-------
LT1ESWTR
LT2ESWTR
MCL
MCLG
MCMC
M-DBP
MF
MG
MOD
mg/L
MILY
MRDL
MRDLG
MRL
MWRA
NCSL
NCWS
NF
NGA
NIH
NLC
NODA
NPDWR
NRDC
NRWA
NTNCWS
NTU
O3
OGWDW
OMB
O&M
O&P
POE
POU
ppb
ppm
PWS
PWSS
QA/QC
QALY
RAA
RF
RFA
RIA
RL
RO
RUS
SAB
SBAR
SBREFA
SDWA
Long Term 1 Enhanced Surface Water Treatment Rule (January, 2002)
Long Term 2 Enhanced Surface Water Treatment Rule (under development)
Maximum Contaminant Level
Maximum Contaminant Level Goal
Markov Chain Monte Carlo [algorithm]
Microbial-Disinfectants/Disinfection Byproducts [Advisory Committee]
Microfiltration
Million gallons
Million Gallons per Day
Milligrams per Liter
Morbidity Inclusive Life Year
Maximum Residual Disinfectant Level
Maximum Residual Disinfectant Level Goal
Minimum Reporting Level
Massachusetts Water Resources Authority
National Conference of State Legislatures
Noncommunity water system
Nanofiltration
National Governors' Association
National Institutes of Health
National League of Cities
Notice of Data Availability
National Primary Drinking Water Regulations
Natural Resources Defense Council
National Rural Water Association
Nontransient Noncommunity Water System
Nephelometric Turbidity Unit
Ozone
Office of Ground Water and Drinking Water
Office of Management and Budget
Operations and Maintenance
Overhead and Profit
Point of entry
Point of Use
Parts per Billion
Parts per Million
Public Water System
Public Water Systems Supervision [Grants Program]
Quality Assurance/Quality Control
Quality Adjusted Life Year
Running Annual Average
Roughing Filter
Regulatory Flexibility Act
Regulatory Impact Analysis
Reservoir/Lake
Reverse Osmosis
Rural Utility Service
Science Advisory Board
Small Business Advocacy Review
Small Business Regulatory Enforcement Fairness Act
Safe Drinking Water Act (1974)
Economic Analysis for the LT2ESWTR
xiv
December 2005
-------
SDWIS
SER
SF
SW
SWAT
Stage 1 DBPR
Stage 2 DBPR
SWTR
T&C
TCR
TMF
TNCWS
TOC
TT
TTHMs
TTHMR
TWO
UF
UMRA
USDA
uv
VSL
WC
WTP
Safe Drinking Water Information System
Small Entity Representative
Secondary Filter
Surface Water
Surface Water Analytical Tool
Stage 1 Disinfectants and Disinfection Byproducts Rule (1998)
Stage 2 Disinfectants and Disinfection Byproducts Rule (under development)
Surface Water Treatment Rule (1989)
Technologies and Cost
Total Coliform Rule (1989)
Technical, Managerial, and Financial
Transient Noncommunity Water System
Total Organic Carbon
Treatment Technique
Total Trihalomethanes
Total Trihalomethane Rule (1979)
Technical Workgroup
Ultrafiltration
Unfunded Mandates Reform Act
United States Department of Agriculture
Ultraviolet [Light Disinfection]
Micrograms per Liter
Value of a Statistical Life
Watershed Control
Willingness to Pay
Economic Analysis for the LT2ESWTR
xv
December 2005
-------
Health Risk Reduction and Cost Analysis
Under the Safe Drinking Water Act (SDWA) Amendments of 1996, when proposing a national
primary drinking water regulation that includes a maximum contaminant level (MCL), the U.S.
Environmental Protection Agency (EPA or the Agency) must conduct a health risk reduction and cost
analysis (HRRCA). A HRRCA addresses seven requirements, all of which are addressed in this
Economic Analysis (EA) for the Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR).
HRRCA Crosswalk Summary
HRRCA Requirement
Quantifiable and nonquantifiable health risk
reduction benefits
Quantifiable and nonquantifiable health risk
reduction benefits from co-occurring contaminants
Quantifiable and nonquantifiable costs
Incremental costs and benefits associated with
regulatory alternatives
Effects of the contaminants on the general
population and sensitive subpopulations
Increased health risk that may occur as a result of
compliance
Other relevant factors (quality and uncertainty of
information)
Addressed in this Economic Analysis
Chapter 5 (All sections and exhibits)
Chapter 7 (Section 7.7; Exhibit 7.4)
Chapters (Sections 8.2.1, 8.2.3, 8.3; Exhibits
8.1-8.4, 8.6-8.10, 8.12-8.16, 8.19)
Chapter 5 (Section 5.6.4)
Chapter 6 (All sections and exhibits)
Chapter 7 (Sections 7.2, 7.6, 7.7, 7.8; Exhibits
7.2, 7.3, 7.4, 7.9)
Chapter 8 (Sections 8.2.2, 8.3; Exhibit 8.1 , 8.5,
8.11,8.17,8.18)
Chapter 5 (Section 5.5; Exhibit 5.30)
Chapter 6 (Sections 6.13; Exhibit 6.22)
Chapter 7 (Section 7.7; Exhibit 7.4)
Chapter 8 (Section 8.3; Exhibits 8.13, 8.18, 8.19)
Chapters (Sections 5.2.2, 5.2.8)
Chapter 7 (Section 7.9)
Chapter 2 (Section 2.3)
Chapter 4 (Section 4.8; Exhibit 4.30)
Chapter 5 (Section 5.4; Exhibits 5.28 and 5.29)
Chapter 6 (Section 6.11; Exhibits 6.19 and 6.20)
Chapters (Section 8.4; Exhibit 8.21)
Economic Analysis for the LT2ESWTR
xvi
December 2005
-------
Executive Summary
This document presents the Economic Analysis (EA), prepared by the U.S. Environmental
Protection Agency (EPA), of the benefits and costs of the Long Term 2 Enhanced Surface Water
Treatment Rule (LT2ESWTR). Executive Order 12866 requires Federal agencies to conduct an analysis
of the benefits and costs of proposed and final rules that cost more than $100 million annually.
ES.l Need for the Rule
More than 14,000 public water systems (PWSs), serving approximately 180 million people in the
United States and its Territories, use surface water, including ground water under the direct influence of
surface water (GWUDI), as their source. These sources often carry microbial pathogens, such as Giardia,
E. coli, and Cryptosporidium. Among pathogens in drinking water, Cryptosporidium is of particular
concern because it is resistant to standard drinking water disinfectants such as chlorine. Ingestion of
Cryptosporidium causes cryptosporidiosis, a gastrointestinal illness; health effects in sensitive
subpopulations may be severe, including death. There is no effective cure for cryptosporidiosis (Framm
and Soave 1997). The LT2ESWTR protects public health by requiring PWSs with the highest measured
source water levels of Cryptosporidium to provide treatment for this pathogen. These vulnerable systems
include both filtered systems with elevated levels of Cryptosporidium in their source water and all
unfiltered systems (i.e., systems meeting the filtration avoidance criteria of the Surface Water Treatment
Rule (40 CFR 141.71)), which currently provide little treatment for Cryptosporidium. The LT2ESWTR
will also reduce the public health risk associated with uncovered finished water reservoirs, which are
susceptible to microbial contamination, including Cryptosporidium.
EPA will also be promulgating the Stage 2 Disinfectants and Disinfection Byproducts Rule
(Stage 2 DBPR), which addresses disinfection byproduct (DBP) formation. The two rules are to be
promulgated concurrently to ensure that protection against microbial pathogens is not compromised by
efforts to reduce DBP formation. The Stage 2 DBPR and LT2ESWTR represent the final stage of a two-
stage strategy to reduce risk from microbial pathogens and DBFs that was developed in a regulatory
negotiation effort in 1992 and 1993.: These rules reflect recommendations presented by the Stage 2
Microbial and Disinfection Byproducts (M-DBP) Federal Advisory Committee Agreement in Principle,
signed in September 2000 (USEPA 2000e).
ES.2 Consideration of Regulatory Alternatives
The Stage 2 M-DBP Advisory Committee met from March 1999 to September 2000 to evaluate
whether and to what degree EPA should revise microbial standards to protect public health. The
committee reached consensus on an approach for addressing Cryptosporidium risk in unfiltered systems
lrThe key outcomes of the 1992-1993 regulatory negotiation effort were to proceed with rules addressing
DBFs and microbial pathogens in two stages and to collect relevant information from PWSs for use in the
development of these rules and the analysis of their impacts. This two-stage approach was subsequently
incorporated into the 1996 Safe Drinking Water Act (SDWA) Amendments. The first stage of the M-DBP
rulemaking process culminated with the joint promulgation of the Stage 1 DBPR and the Interim Enhanced Surface
Water Treatment Rule (IESWTR) by EPA in December 1998.
Economic Analysis for LT2ESWTR ES-1 December 2005
-------
and microbial risk in uncovered finished water reservoirs without formally identifying alternative
regulatory approaches. For filtered systems, however, several alternatives were considered. All involved
a compliance scheme whereby systems choose from a "toolbox" of treatment and managerial options for
meeting additional Cryptosporidium treatment requirements. The first would have required a single level
of additional treatment or removal for all systems regardless of source water Crytosporidium levels. The
other three alternatives (A2-A4) based their treatment requirements on the Cryptosporidium levels in a
plant's source water, as determined by a period of source water monitoring. The alternatives are shown in
Exhibit ES.l.
Economic Analysis for LT2ESWTR ES-2 December 2005
-------
Exhibit ES.1: Summary of Treatment Requirements for Filtered
Systems as a Function of Source Water Cryptosporidium
Concentrations
Source Water Cryptosporidium
Monitoring Results (oocysts/L)
Additional Treatment Requirements
Alternative A1
2.0 log treatment required for all systems
Alternative A2
<0.03
> 0.03 and < 0.1
> 0.1 and < 1.0
> 1.0
No action
0.5 log
1 .5 log
2.5 log
Alternative A3 - Preferred Alternative
< 0.075
> 0.075 and < 1.0
> 1.0 and < 3.0
> 3.0
No action
1 log
2 log
2.5 log
Alternative A4
<0.1
> 0.1 and < 1.0
>1.0
No action
0.5- log
1 .0 log
Notes: "Additional treatment requirements" are for systems with conventional treatment in full
compliance with existing rules (the IESWTR and LT1ESWTR). Requirements for systems with
other treatment types may differ, and are presented in chapter 1.
The term "log treatment" is used to express the expected percent reduction of a contaminant.
For example, 1 log treatment is expected to provide 90 percent reduction of a contaminant and 2
log treatment provides 99 percent reduction. Compliance with the log treatment requirements is
not based on quantifying the actual reduction; instead, other finished water quality or operational
conditions are used to determine compliance. This is consistent with previous SWTRs.
In this EA, EPA estimates and compares the costs and benefits of all regulatory alternatives. The
cost and benefit data presented in section ES.4 are for the Preferred Regulatory Alternative (A3), while
section ES.5 presents comparisons of the cost and benefit estimates for all of the alternatives.
ES.3 Summary of the LT2ESWTR Requirements
The LT2ESWTR applies to all PWSs that use surface water or GWUDI (excluding those that
purchase all their water). It builds on the SWTR, IESWTR, and the LTIESWTR by improving control of
microbial pathogens, specifically Cryptosporidium. Unlike the previous rules, the LT2ESWTR bases its
treatment requirements on a system's source water Cryptosporidium concentration and the type of
Economic Analysis for LT2ESWTR
ES-3
December 2005
-------
treatment already provided. It requires systems to monitor their source water for Cryptosporidium, and
based on the results, to meet one of four levels of treatment for Cryptosporidium (with the first level
requiring no additional treatment). The levels of treatment needed at each level will be reassessed in the
future based on a second round of source water monitoring under the current rule.
For those systems that do not already provide filtration, the LT2ESWTR has specific
requirements to inactivate two or three logs of Cryptosporidium, depending on source water monitoring
results. It also requires systems with uncovered finished water reservoirs either to cover the reservoirs or
to provide additional treatment to the reservoir effluent.
Exhibit ES.2 illustrates each rule activity for the Preferred Regulatory Alternative. The
implementation schedules for small and large systems differ; Exhibit ES.3 presents the implementation
time line.
Economic Analysis for LT2ESWTR ES-4 December 2005
-------
Exhibit ES.2: Overview of Key LT2ESWTR Requirements
Systems Subject to the LT2ESWTR
Public water systems using surface water or ground water under direct influence of surface water that are subject to
the SWTR. Systems with uncovered finished reservoirs.
Requirements for Uncovered Unfinished Reservoirs
Cover reservoirs or treat discharge to achieve 4 log virus, 3 log Giardia, and 2 log Cryptosporidium inactivation.
Initial Monitoring for Unfiltered
Systems
Monitor Cryptosporidium in source water to
determine treatment requirements. Exemptions are
available for systems that purchase all of their
water, provide at least 3 log of treatment, or have
2 years of historical Cryptosporidium data.
Large systems (serving at least 10,000 people):
sample Cryptosporidium monthly for 2 years.
Small systems: sample twice per month for 1 year.
Treatment Requirements for Unfiltered
Systems
Systems with Cryptosporidium levels <0.01
oocysts/L: provide 2 log inactivation. Systems
with Cryptosporidium levels >0.01 oocysts/L:
must provide 3 log inactivation.
Treatment Options for Unfiltered
Systems
Meet Cryptosporidium inactivation requirements
using ozone, chlorine dioxide, or ultraviolet light.
I
Initial Monitoring for Filtered Systems
Conduct initial Cryptosporidium monitoring to establish
bin classification. Exemptions are available for systems
that purchase all of their water, provide at least 5.5 log of
treatment, or have 2 years of historical Cryptosporidium
data.
Large systems (serving at least 10,000 people): must
monitor monthly for 2 years for Cryptosporidium, E. coli,
and turbidity. Small systems: monitors, coli biweekly for
1 year. If E. coli levels exceed trigger, monitor
Cryptosporidium twice per month for 1 year.
Bin Classification and Treatment
Requirements for Filtered Systems
Systems are placed in 1 of 4 "action bins" based on
monitoring results. Systems with levels <0.075 oocysts/L:
no action. Systems with levels >0.075 but <1.0 oocysts/L:
provide additional 1 log treatment. Systems with levels
>1.0 but<3.0 oocysts/L: provide an additional 2 log
treatment. Systems with levels >3.0 oocysts/L: provide an
additional 2.5 log treatment.
Treatment Options for Filtered Systems
Select from a "toolbox" of treatment or management
options.
Future Monitoring and Reassessment
Six years after initial bin classification, conduct a second round of monitoring. Methods, frequency, locations, etc.,
for this round may change based on stakeholder input, which could lead to a rule amendment. State/primacy agency
may also assess any changes in watershed/source water as part of the sanitary survey and determine appropriate
action, which could include requiring implementation of additional toolbox options.
Economic Analysis for LT2ESWTR
ES-5
December 2005
-------
Exhibit ES.3: Implementation Timeline for LT2ESWTR for Filtered Systems
Year!
System:
Year 2
Serving ;
YearS
: 100,000
Crypto Monitoring
System:
System:
System:
Year!
Serving
.0,000-99.
Crypto Monitoring
Serving
> Serving
Year 2
0,000-49,
Year 4
People
YearS
Year 6
Treatment Installation
999 Peop
e
Year?
YearS
Possible Extension
Year 9
Treatment Installation Possible Extension
999 Peop
Crypto Monitoring
< 10,000 F
'eople
£. co//Mon.
YearS
Year 4
e
Year 10
Year 11
2nd Round Crypto Mon.
Year 12
2nd Round Crypto Mon.
Treatment Installation Possible Extension
(Optional)
Crypto Monitoring
YearS
Year 6
Year 13
2nd Round Crypto Mon.
Treatment Installation Possible Extension
Year?
YearS
Year 9
Year 10
Year 11
2nd Round
Year 12
f. co// Mon.
Year 13
Note: This exhibit does not show the reassessment of treatment levels following the second round of source
monitoring.
ES.4 National Benefits and Costs of the LT2ESWTR
The benefits resulting from implementation of the LT2ESWTR are due to reductions in the
numbers of infectious Cryptosporidium oocysts and other pathogens reaching consumers. EPA has
quantified the benefits of this rule in terms of avoided endemic cryptosporidiosis illnesses and associated
deaths. The effects of reductions in other pathogens, if any, were not quantified.
Cryptosporidium can reach the consumer when there is a significant breakdown in the treatment
process, but also under normal operating conditions. This EA focuses on the benefits that result from
reducing only the continuous, relatively low levels of Cryptosporidium exposure that can occur even
under normal operating conditions. Requiring additional treatment is also expected to reduce the
frequency and severity of outbreaks; these benefits, however, are not quantified.
The costs incurred for LT2ESWTR activities are associated with rule implementation, source
water monitoring, and adding treatment.
Economic Analysis for LT2ESWTR
ES-6
December 2005
-------
ES.4.1 Benefit Estimates
The quantified benefits of the LT2ESWTR are the estimated reduction in the number of illnesses
and deaths associated with endemic Cryptosporidium. There are also benefits that cannot currently be
quantified but are likely to be substantial. These are summarized in Exhibit ES.4 and discussed in
Chapter 5.
Exhibit ES.4: Summary of Nonquantified Benefits and Groups Affected
Type of Benefit
Nonquantified Benefits
Group(s) Affected
Health benefits
Reduction in risk to sensitive
subpopulations (mortality for those with
AIDS and other sensitive subpopulations
has been quantified)
Immunocompromised individuals
served by systems that make
changes to or add treatment
Reduction in health risk during outbreaks
(and response costs)
Reduction in co-occurring or emerging
pathogen risk
All individuals served by systems that
make changes to or add treatment,
including those now served by
uncovered finished water reservoirs,
(between 46 and 66 million people)
Reduction in endemic morbidity and
mortality risk associated with uncovered
finished water reservoirs
All individuals receiving water from
uncovered finished water reservoirs
Reduction in health risks from certain
DBFs
All individuals served by systems that
install physical disinfection
technologies like membranes or UV1
Nonhealth
Benefits
Improved aesthetic water quality
All individuals served by systems that
make changes to or add treatment
that is likely to reduce taste and odor
problems (e.g., ozone)
Costs of consumers' attempts to avert
possible risks
Consumers in systems that cease
using uncovered finished water
reservoirs (through covering or taking
such reservoirs off-line) my have
greater confidence in water quality.
This may result in less averting
behavior that reduces both out-of-
pocket costs (e.g., purchase of bottled
water) and opportunity costs (e.g.,
time to boil water).
Systems that install chemical disinfection technologies like ozone may increase certain DBFs.
Economic Analysis for LT2ESWTR
ES-7
December 2005
-------
EPA developed a risk assessment model to predict the illnesses and deaths avoided using certain
variables, including the uncertainty in those variables. (Uncertainty means the degree of lack of
knowledge of those variables' values.) The variables are occurrence of Cryptosporidium in source water,
infectivity, water treatment used, average daily water consumption per capita, morbidity, and mortality.
To allow a comparison of benefits with the costs of implementing the rule, the estimates of morbidity and
mortality are multiplied by a calculated cost of illness (COI) and a standard value of a statistical life,
respectively.
There are three data sets for source water occurrence, designated as ICR, ICRSSL, and ICRSSM.
The ICR data set is the Information Collection Rule data that all large systems collect. The ICRSSL and
ICRSSM data sets represent 40 large systems and 40 medium-size systems from the Information
Collection Rule Supplemental Surveys. EPA judges each of these data sets to be equally likely to
represent the true distribution of Cryptosporidium in source waters for all systems. (Each set has
advantages and disadvantages that counteract those of other sets, as described in Chapter 4.) All benefit
and cost analyses are carried out using each data set to provide a range of possible benefits and costs. In
addition to using the three occurrence data sets, this EA monetizes benefits with two values of
COI—referred to as Enhanced and Traditional (see Exhibits ES.6a and ES.6b).
For Cryptosporidium infectivity (i.e., the likelihood of exposure to a particular dose of
Cryptosporidium resulting in infection), EPA considered results from human volunteer feeding studies.
Results from three studies were evaluated for the proposed LT2ESWTR, and results from three newer
studies were added in the analysis for the final LT2ESWTR. Further, EPA used six different model forms
to estimate dose-response relationships with these study results. This analysis and results are described in
Chapter 5 and Appendix N.
Variability in host susceptibility, response at very low oocyst doses typical of drinking water
ingestion, and the relative infectivity and occurrence of different Cryptosporidium isolates in the
environment are uncertain. To address this uncertainty, three sets of estimates are presented in this
Executive Summary: a "high" estimate based on the model which showed the highest mean baseline risk,
a "medium" estimate, based on the model and data used at proposal, which is in the middle of the range of
estimates produced by the six models using the newly available data, and a "low" estimate, based on the
model which showed the lowest mean baseline risk. These estimates are not upper and lower bounds on
illnesses avoided and benefits; for each model, a distribution of effects is estimated, and the "high" and
"low" estimates show only the means of these distributions for two different model choices.
Exhibit ES.5 summarizes the estimates of avoided illnesses and deaths resulting from the
LT2ESWTR. Exhibits ES.6a and ES.6b summarize the monetized value of those estimates for Enhanced
and Traditional COI values (annualized over a 25-year period and discounted at 3 percent and 7 percent)
The Traditional COI includes only values of medical costs and lost work time (including some portion of
nonmarket household production). The Enhanced COI also includes the values of lost personal (non-
work) time, such as child care and homemaking (to the extent not covered by the traditional COI), time
with family, and recreation, and lost productivity at work on days when workers are ill but go to work
anyway.
Economic Analysis for LT2ESWTR ES-8 December 2005
-------
Exhibit ES.5: Summary of Annual Illnesses and Deaths Avoided
Data Set
Annual Illnesses Avoided
Low
Medium
High
Annual Deaths Avoided
Low
Medium
High
Total after Full implementation
ICR
ICRSSL
ICRSSM
358,732
89,375
177,101
964,360
230,730
455,170
1,459,126
372,507
711,123
76
20
39
207
52
100
314
84
156
Annual Average over 25 years
ICR
ICRSSL
ICRSSM
264,980
66,187
130,918
712,732
170,977
336,652
1,078,796
276,078
438,203
57
15
29
154
39
74
232
62
116
Source: Chapter 8, Exhibit 8.3.
Note: High, medium and low estimates reflect the mean estimates for a range of dose-response modeling
assumptions. See Appendix N for more detail.
Exhibit ES.6a: Summary of Monetized Benefits—Enhanced Cost of Illness
Data Set
Value of Benefits
($ Millions, 2003$)
Low
Medium
High
Annualized Value (at 3%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 687
$ 177
$ 344
$ 1,853
$ 458
$ 886
$ 2,822
$ 744
$ 1,393
Annualized Value (at 7%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 556
$ 144
$ 279
$ 1,501
$ 371
$ 718
$ 2,286
$ 603
$ 1,128
Economic Analysis for LT2ESWTR
ES-9
December 2005
-------
Exhibit ES.6b: Summary of Monetized Benefits—Traditional Cost of Illness
Data Set
Value of Benefits
($ Millions, 2003$)
Low
Medium
High
Annualized Value (at 3%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 497
$ 130
$ 250
$ 1,341
$ 335
$ 644
$ 2,047
$ 546
$ 1,014
Annualized Value (at 7%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 403
$ 105
$ 203
$ 1,089
$ 272
$ 523
$ 1,662
$ 443
$ 824
Source: Chapter 8, Exhibits 8.4a and 8.4b.
Note: High, medium and low estimates reflect the
mean estimates for a range of dose-response modeling
assumptions. See Appendix N for more detail.
ES.4.2 National and Household Cost Estimates
The total national costs of the LT2ESWTR include costs to systems and States/Primacy Agencies
for implementation and compliance. This EA estimates costs for all rule-related activities:
implementation, source water monitoring, adding treatment, and compliance reporting. EPA assumes
nearly all surface water and GWUDI systems will incur rule implementation and initial source water
monitoring costs. Compliance reporting costs are estimated only for those systems predicted to add
treatment.
Approximately 90 percent2 of the estimated total national costs are for systems to meet additional
treatment requirements. EPA developed a least-cost approach to modeling treatment costs. The approach
is constrained to reflect site-specific conditions. The following series of steps was used to develop
treatment cost estimates for compliance with the rule.
1. Predict the percent of systems falling into each bin based on a model of source water
occurrence.
2. Model unit costs for each treatment technology.
3. Develop a technology forecast for each bin using the least-cost approach and estimates of the
maximum number of systems that will apply any one technology.
4. Calculate the number of plants selecting each technology.
5. Multiply the number of plants per technology by the technology' s unit cost.
Derived from Exhibit 6.3 using ICR data.
Economic Analysis for LT2ESWTR
ES-10
December 2005
-------
Exhibit ES.7 summarizes the system costs associated with the LT2ESWTR.
Exhibit ES.7: Summary of System Costs ($ Millions, 2003$)
Source: Exhibit 8.11.
Data
Set
ICR
ICRSSL
ICRSSM
Capital and One-Time
(Undiscounted at Full Implementation)
Mean
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Operation and Maintenance
(Undiscounted at Full Implementation)
Mean
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Nominal Costs
$ 2,104
$ 1,526
$ 1,719
$ 1,715
$ 1,164
$ 1,372
$ 2,425
$ 1,743
$ 1,941
$ 55
$ 33
$ 39
$ 48
$ 26
$ 33
$ 64
$ 39
$ 45
Data
Set
ICR
ICRSSL
ICRSSM
Total Annualized Costs at 3%
Mean
$ 133
$ 93
$ 106
90% Confidence Bound
Lower
(5th %ile)
$ 111
$ 72
$ 86
Upper
(95th %ile)
$ 160
$ 112
$ 126
Total Annualized Costs at 7%
Mean
$ 150
$ 107
$ 121
90% Confidence Bound
Lower
(5th %ile)
$ 125
$ 83
$ 99
Upper
(95th %ile)
$ 181
$ 129
$ 144
EPA assumes that systems will generally pass the costs of a new regulation on to their customers
in the form of rate increases. Household costs, which are in units of $ per household per year, are
estimated in order to provide a measure of the increase in water bills that could be expected to result from
the LT2ESWTR. Exhibit ES.8 summarizes household costs for those systems predicted to require
additional treatment.
Economic Analysis for LT2ESWTR
ES-11
December 2005
-------
Exhibit ES.8: Summary of Annual Household Cost Increases1
($ per Year, 2003$)
System Type/Size
Households
Mean
Median
90th
Percentile
95th
Percentile
Percent of
Systems with
Household
Cost Increase
<$12
Percent of
Systems with
Household
Cost Increase
<$120
ICR
All CWS
CWS<1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$2.59
$4.14
$13.09
$0.21
$0.56
$3.86
$6.43
$9.97
$28.66
$9.97
$14.79
$53.60
96.49%
91.19%
63.20%
99.99%
99.88%
98.87%
ICRSSL
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$1.67
$2.49
$8.58
$0.09
$0.36
$2.91
$6.37
$6.60
$17.44
$6.42
$9.37
$29.01
97.96%
96.46%
72.61%
100.00%
99.94%
99.50%
ICRSSM
All CWS
CWS < 10,000
CWS < 500
68,857,992
5,587,602
158,900
$1.97
$3.00
$10.10
$0.09
$0.49
$2.90
$6.37
$7.02
$26.24
$6.85
$11.39
$35.97
97.47%
95.19%
68.73%
99.99%
99.93%
99.31%
ICR -High
All CWS
CWS < 10,000
CWS < 500
68,857,992
5,587,602
158,900
$2.84
$4.58
$7.21
$0.21
$0.61
$2.91
$6.43
$11.50
$16.81
$9.97
$15.30
$26.25
96.09%
90.22%
75.79%
99.99%
99.86%
99.80%
ICRSSL - Low
All CWS
CWS < 10,000
CWS < 500
68,857,992
5,587,602
158,900
$1.42
$2.06
$14.42
$0.03
$0.23
$4.79
$5.65
$6.58
$30.00
$6.42
$7.47
$54.42
98.37%
97.21%
62.07%
100.00%
99.96%
98.58%
1Annualized at discount rates varied by system size and ownership (see Appendix J, Exhibit J.2).
Source: Exhibit 6.18.
ES.5 National Net Benefits and Summary of Comparison of Alternatives
The national net benefits (benefits remaining after costs are taken into account) for each of the
regulatory alternatives are shown in Exhibit ES.9. The cells outlined in bold show where net benefits are
highest for a particular combination of regulatory alternative, occurrence data set, cost of illness
assumptions, and discount rate. The annualized net national benefits for the Preferred Alternative (A3 in
Exhibit ES. 1.) range from $18 million to $2.67 billion, depending on occurrence, infectivity model, cost
of illness, and discount rates used.
From Exhibit ES.9, several important economic questions can be answered. First, the Preferred
Alternative (A3) has positive net benefits under all assumptions, a key threshold test of the reasonableness
of a regulation. This is also true for Alternatives A4 under all combinations of occurrence, cost of illness,
and discount rates, and true for Alternative Al and A2 for most of the combinations. In addition, this
exhibit shows that the Preferred Alternative (A3) produces the highest net benefits of each alternative
under 8 of 12 combinations of assumptions, and near the highest net benefits under 4 of 12 combinations.
Economic Analysis for LT2ESWTR
ES-12
December 2005
-------
Exhibit ES.9a: Comparison of Mean Net Benefits
for All Regulatory Alternatives—Enhanced Cost of Illness ($Millions, 2003$)
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Annualized Value
3%, 25 Years
Low
$ 260
$ 498
$ 527
$ 550
Medium
$ 1,492
$ 1,708
$ 1,720
$ 1,673
High
$ 2,447
$ 2,655
$ 2,662
$ 2,566
7%, 25 Years
Low
$ 126
$ 366
$ 396
$ 427
Medium
$ 1,098
$ 1,333
$ 1,351
$ 1,328
High
$ 1,897
$ 2,112
$ 2,126
$ 2,061
ICRSSL
A1
A2
A3 - Preferred
A4
$ (223)
$ 43
$ 65
$ 87
$ 156
$ 366
$ 365
$ 347
$ 466
$ 647
$ 632
$ 589
$ (265)
$ 6
$ 32
$ 58
$ 15
$ 257
$ 264
$ 261
$ 292
$ 496
$ 491
$ 465
ICRSSM
A1
A2
A3 - Preferred
A4
$ (58)
$ 198
$ 218
$ 230
$ 578
$ 782
$ 780
$ 731
$ 1,104
$ 1,285
$ 1,267
$ 1,171
$ (132)
$ 130
$ 153
$ 172
$ 358
$ 591
$ 597
$ 569
$ 809
$ 1,010
$ 1,002
$ 935
Exhibit ES.9b: Comparison of Mean Net Benefits
for All Regulatory Alternatives—Traditional Cost of Illness ($Millions, 2003$)
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Annualized Value
3%, 25 Years
Low
$ 64
$ 305
$ 337
$ 373
Medium
$ 967
$ 1,190
$ 1,208
$ 1,193
High
$ 1,649
$ 1,870
$ 1,887
$ 1,842
7%, 25 Years
Low
$ (31)
$ 211
$ 243
$ 285
Medium
$ 675
$ 917
$ 939
$ 941
High
$ 1,256
$ 1,481
$ 1,502
$ 1,478
ICRSSL
ICRSSM
A1
A2
A3 - Preferred
A4
A1
A2
A3 - Preferred
A4
$ (284)
$ (9)
$ 18
$ 46
$ (165)
$ 99
$ 124
$ 148
$ 0
$ 233
$ 242
$ 242
$ 306
$ 529
$ 538
$ 518
$ 214
$ 432
$ 433
$ 418
$ 676
$ 890
$ 889
$ 840
$ (315)
$ (35)
$ (7)
$ 25
$ (218)
$ 50
$ 77
$ 106
$ -109
$ 150
$ 166
$ 175
$ 138
$ 387
$ 402
$ 398
$ 90
$ 324
$ 331
$ 327
$ 465
$ 692
$ 698
$ 668
Notes: The traditional COI includes only values of medical costs and lost work time (including some portion of
nonmarket household production). The enhanced COI also includes the values of lost personal (non-work) time such
as child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
High, medium and low estimates reflect the mean estimates for a range of dose-response modeling, assumptions.
See Appendix N for more detail.
Source: Exhibits 8.12a and 8.12b.
Economic Analysis for LT2ESWTR
ES-13
December 2005
-------
Another key economic test is whether the proposed rule is cost-effective. Exhibits ES. lOa and
ES. lOb together show the annualized value of benefits and costs for the four alternatives, calculated for
each combination of occurrence data sets, COI values, and discount rates. For each alternative, the graph
plots the mean benefit versus its corresponding range of cost estimates (a 90 percent confidence bound
shown as a vertical bar). A trend line connects the mean estimate of costs for each alternative. These
graphs help to show the concept of cost-effectiveness and to compare the alternatives.
Cost-effectiveness can be defined simply as getting the greatest benefit for a given expenditure or
imposing the least cost for a given level of benefit. In Exhibits ES. lOa and ES. lOb, the test would be to
see if any regulatory alternative was to the right and completely below any other alternative on the graph.
If so, the alternative to the right and below would be more cost-effective and would dominate the
alternative that provided fewer benefits at higher costs.
In the strict sense, each of the regulatory alternatives is cost-effective—no regulatory alternative
provides more benefits at the same or a lower cost than another, and no alternative can achieve lower
costs for the same or a greater level of benefits than another. Thus, no alternative dominates any other or
is more cost-effective. Instead, the alternatives offer increasing levels of benefits at increasing levels of
cost. Chapter 8 provides additional analyses on cost effectiveness.
In addition to allowing a visual comparison of cost effectiveness, the exhibits show information
about the incremental benefits of each alternative. Compared to Alternative A4, the Preferred Alternative
achieves significant incremental benefits (the change in benefits from one alternative to another) at a
relatively low increase in costs. The step from the Preferred Alternative to Alternative A2 achieves more
benefits, but at a higher incremental rate. The step to Alternative Al achieves a similar increase in
benefits, but at a significantly higher cost. The Preferred Alternative, and perhaps Alternative A2, appear
to be good values; other alternatives have either significantly fewer benefits for similar costs, or greater
benefits at dramatically higher costs.
Economic Analysis for LT2ESWTR ES-14 December 2005
-------
Exhibit ES.10a: Comparison of Mean—Enhanced Cost of Illness1
3 Percent Discount Rate
7 Percent Discount Rate
(A
*- C
0) O
^ =
re 1=
re ^
oi w
0 ^
~ 0
O
•S «"
(O ^
re |
Q ^
W "~"
w £
oi w
O ^
~™ O
$500
$450
$400
$350
$300
$250
$200
$150
$100
$50
$0
Alt. 1 >
t
•
'
Alt 2 .'
\ '
AI" V
\ ...->'
\"""
$1,725 $1,750 $1,775 $1,800 $1,825 $1,850 $1,875 $1,900
$1,925
$450 -
$400 -
$350 -
$300 -
$250 -
$200 -
$150 -
$100 -
$50-
/
*
•
Alt. 2 •
Alt. 3 \ I,'
Alt. 4 ^ 1 . • 1
s '
$375 $400 $425 $450 $475 $500 $525
-^
.'
$550
$575
Alt. 1 *•
'
,
•
,
Alt. 2 .
Alt. 3 J
Alt 4 •••""" 1
[/---
\
\
$1,400 $1,425 $1,450 $1,475 $1,500 $1,525 $1,55
5
Alt. 1 ^
^^^
*
/
f
t
/
Alt. 2 ,'
Alt. 3 .'
I J- * * 1
\ ' "
$300 $320 $340 $360 $380 $400 $420 $440
$460
*- — •
-------
Exhibit ES.10b: Comparison of Mean—Traditional Cost of Illness1
3 Percent Discount Rate
7 Percent Discount Rate
I/)
+- c
Q) O
(/) |=
(Q p
13
o t^
o; u>
0 to
~ o
o
15 "w"
w -2
13 ~
Q S
• ^
If
C/5 ^
Q
13 ~
o i
w ?
Q; u>
25
$350
$300
$250
$200
$150
$100
$50
$0
'|
f
•
Alt. 2 /
*
Alt. 3 \ .'
Alt. 4 V ' t
1 • *• *
/ 1
r
$1,250 $1,275 $1,300 $1,325 $1,350 $1,375
$roo
$450
$400
$350
$300
$250
$200
$150
$100
$50
Alt. 1
"~ *•,
/
.'
/
•
Alt. 2 /
Alt. 4 Alt. 3 \ ,'
\ ' '
$275 $300 $325 $350 $375 $400 $425
$500 -,
$450
$400
$350
$300
$250
$200
$150
$100
$50
$0
Alt. 1 1
^""^•1
/'
/'
•
'
Alt. 2 •
Alt. 4 Alt. 3 X. , •'
\ >P>f
I'""
$550 $575 $600 $625 $650 $675 $700 $725
Mean of Benefits ($Millions)
t
t
•
,'
*
Alt. 2 ,'
Alt. 3 \ \
A"4 ^k I.- "I
/ """"""'I
J. --••"""""
$1,025 $1,050 $1,075 $1,100
$1,125
Alt. 1 .
^ e
'
.'
,'
/
•
Alt. 2 .'
Alt. 3 . »
Alt. 4 \. ^ 1'
» • • " l"
\ "
$225 $245 $265 $285 $305 $325 $345
Alt. 1
/
/'
Alt. 2 .
*
Alt. 3 "V ,'
Alt. 4 » ,
J >|, \
\""'
I
1
\
:\
$450 $470 $490 $510 $530 $550 $570 $590
Mean of Benefits ($Millions)
Economic Analysis for LT2ESWTR
ES-16
December 2005
-------
Notes on Exhibits ES.10a and 10b: The traditional COI includes only values of medical costs and lost work time
(including some portion of nonmarket household production). The enhanced COI also includes the values of lost
personal (non-work) time such as child care and homemaking (to the extent not covered by the traditional COI), time
with family, and recreation, and lost productivity at work on days when workers are ill but go to work anyway.
Source: Exhibit 8.7.
Cost effectiveness is generally used for determining which alternatives meet a certain criterion-a
threshold for costs or minimum cases avoided. A comparison of cost effectiveness is made in Exhibit ES
11 between the cost per MILY of alternative rules (based on the various combinations of dose-response
model, COI approach, Cryptosporidium occurrence data set, and discount rate) and cost thresholds found
in the literature.
As expected, Exhibit ES 11 shows that the cost per MILY saved increases with regulatory cost
and stringency: A4 has the lowest cost per MILY, the Preferred Alternative has the next lowest cost per
MILY, etc. Alternative A4 is cost saving under most combinations of assumptions (23 of 36
combinations). Cost saving simply means the regulation is saving more in avoided costs for medical and
lost time associated with avoided cases than the actual cost of the rule. The Preferred Alternative A3 is
also cost saving under many of the combinations (16 of 36), as is the next more stringent alternative, A2
(12 of 36 combinations).
These cost per MILY ratios are compared to prima facie cost per MILY thresholds, with the
understanding that the thresholds are arbitrary values, often derived by reference to the cost per QALY
(or MILY) for interventions that public health specialists agree are justified. The Harvard Cost Utility
Analysis database presents a median cost-utility ratio of $31,000 per QALY (or MILY) (2002$) for
respiratory and cardiovascular interventions, while Tengs et al. (1995) report a median cost per life-year
saved for life-saving interventions of $48,000 (1993$). The health economics literature often uses either
$50,000 or $100,000 per QALY (or MILY) as a threshold with ratios less than these values considered
prima facie cost effective. In general, EPA recommends that decisions as to whether a specific control
strategy is justified should be based on a complete comparison of benefits and costs.
In the majority of combinations of assumptions, Exhibit ES 9 shows larger net benefits for the
Preferred Alternative A3 than for Alternative A4 (21 of 36) and Alternative A2 (28 of 36). Therefore,
only the Preferred Alternative A3 is compared to the cost-utility thresholds in this summary, however,
complete results for all alternatives under all assumptions are shown in Exhibit ES 9.
The Preferred Alternative A3 is cost effective compared to the lowest of these thresholds in most
combinations of assumptions. In comparison to prima facie cost-utility ratios of $31,000 per MILY and
$50,000 per MILY, the Preferred Alternative A3 is cost effective in 30 of 36 and 33 of 36 possible
combinations of assumptions, respectively.
Economic Analysis for LT2ESWTR ES-17 December 2005
-------
Exhibit ES.11a: Cost Effectiveness Analysis Based on Low, Medium, and High
Estimate Dose Response Models Using the Enhanced COI Approach, by Data Set,
by Rule Alternative, 3% and 7% Discount Rates
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Cost per MILY1 Saved ($)
3%, 25 Years
Low
$ 62,646
$ 6,997
$ 226
cost saving
Medium
$ 4,789
cost saving
cost saving
cost saving
High
cost saving
cost saving
cost saving
cost saving
7%, 25 Years
Low
$ 87,814
$ 18,248
$ 9,887
cost saving
Medium
$ 14,392
cost saving
cost saving
cost saving
High
cost saving
cost saving
cost saving
cost saving
ICRSSL
A1
A2
A3 - Preferred
A4
$ 263,970
$ 69,348
$ 48,806
$ 22,705
$ 86,424
$ 9,299
$ 1,342
cost saving
$ 42,135
cost saving
cost saving
cost saving
$ 343,263
$ 100,080
$ 74,863
$ 41,241
$ 117,942
$ 21,479
$ 11,648
cost saving
$ 61,780
$ 2,463
cost saving
cost saving
ICRSSM
A1
A2
A3 - Preferred
A4
$ 138,642
$ 29,183
$ 16,993
$ 995
$ 36,566
cost saving
cost saving
cost saving
$ 13,753
cost saving
cost saving
cost saving
$ 184,254
$ 47,410
$ 32,383
$ 11,705
$ 54,711
$ 996
cost saving
cost saving
$ 26,455
cost saving
cost saving
cost saving
Footnote 1: MILYs (Morbidity Inclusive Life Years) are a combination of life years gained from avoided premature mortality plus
QALYs (Quality-Adjusted Life Years) gained from avoided morbidity.
Exhibit ES.11b: Cost Effectiveness Analysis Based on Low, Medium, and High
Estimate Dose Response Models Using the Traditional COI Approach, by Data
Set, by Rule Alternative, 3% and 7% Discount Rates
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Cost per MILY1 Saved ($)
3%, 25 Years
Low
$ 75,165
$ 19,491
$ 12,701
$ 1,260
Medium
$ 17,241
cost saving
cost saving
cost saving
High
$ 5,585
cost saving
cost saving
cost saving
7%, 25 Years
Low
$ 100,095
$ 30,505
$ 22,125
$ 7,501
Medium
$ 26,608
$ 795
cost saving
cost saving
High
$ 11,827
cost saving
cost saving
cost saving
ICRSSL
A1
A2
A3 - Preferred
A4
$ 276,517
$ 81,618
$ 60,927
$ 34,520
$ 98,897
$ 21,518
$ 13,420
$ 3,291
$ 54,616
$ 6,980
$ 2,050
cost saving
$ 355,572
$ 112,117
$ 86,753
$ 52,831
$ 130,177
$ 33,465
$ 23,496
$ 10,406
$ 74,023
$ 14,454
$ 8,369
$ 412
ICRSSM
A1
A2
A3 - Preferred
A4
$ 151,183
$ 41,597
$ 29,323
$ 13,084
$ 49,037
$ 5,981
$ 1,138
cost saving
$ 27,346
cost saving
cost saving
cost saving
$ 196,557
$ 59,587
$ 44,478
$ 23,564
$ 66,945
$ 13,112
$ 7,127
cost saving
$ 39,752
$ 2,581
cost saving
cost saving
Footnote 1: MILYs (Morbidity Inclusive Life Years) are a combination of life years gained from avoided premature mortality plus
QALYs (Quality-Adjusted Life Years) gained from avoided morbidity.
Economic Analysis for LT2ESWTR
ES-18
December 2005
-------
The Preferred Alternative (A3) was also evaluated against the other alternatives with respect to
key uncertainty variables. EPA conducted a series of sensitivity analyses to examine the effects of
exaggerating one uncertain variable while holding all others constant. These analyses tested the
following uncertain variables: the assumptions regarding loss of productivity and the value of nonmarket
work and leisure time in computing the cost of illness, the value of AIDS-related mortality rate, and the
overall value of benefits (presented in Appendix P, Appendix R, and Chapter 8, respectively). In addition
to the sensitivity analyses, the main body of the EA uses two possible values for the cost of illness.
Further, uncertainty in the occurrence of Cryptosporidium in source water is addressed by carrying out
separate analyses throughout the EA for the three possible occurrence distributions as well as confidence
bounds of the occurrence distributions. The results of all tests show that benefits still exceed costs for the
Preferred Alternative, it remains the favored alternative in the majority of conditions analyzed, and no
other alternative performs as well across the range of possible occurrence and values for benefits.
The Stage 2 Microbial Disinfectants and Disinfection Byproducts (Stage 2 M-DBP) Advisory
Committee recommended Alternative A3. Based on the recommendation, and supported by EPA's
evaluations of benefits and costs, EPA selected Alternative A3 as the proposed rule.
Economic Analysis for LT2ESWTR ES-19 December 2005
-------
1. Introduction
This document presents an analysis of the costs and benefits of the Long Term 2 Enhanced
Surface Water Treatment Rule (LT2ESWTR). The analysis is performed in compliance with Executive
Order 12866, Regulatory Planning and Review (58 Federal Register (FR) 51735), which requires that the
U.S. Environmental Protection Agency (EPA) estimate the economic impact of rules costing more than
$100 million annually and submit the analysis in conjunction with publishing the rule.
This chapter provides a summary of the LT2ESWTR in section 1.1 and describes the organization
of the Economic Analysis (EA) in section 1.2.
1.1 Summary
The LT2ESWTR builds on the Interim Enhanced Surface Water Treatment Rule (IESWTR) and
the Long Term 1 Enhanced Surface Water Treatment Rule (LTIESWTR) by improving control of
microbial pathogens, specifically the contaminant Cryptosporidium1. The LT2ESWTR also addresses the
tradeoff between competing risks that is posed by the simultaneous control of microbial pathogens and
disinfection byproducts (DBFs). The disinfectants commonly used to kill microorganisms react with
naturally occurring organic and inorganic matter in source water, forming DBFs that are known to have
adverse health effects, including cancer and developmental and reproductive effects. In order to balance
the risks posed by DBFs and microbial pathogens, the LT2ESWTR will be promulgated concurrently
with the Stage 2 Disinfection Byproducts Rule (DBPR). This will make it easier for water systems to
comply with both rules. The LT2ESWTR applies to all community water systems (CWSs) and
noncommunity water systems (NCWSs) that use surface water or ground water under the direct influence
of surface water (GWUDI) as a source.
The intent of the LT2ESWTR is to supplement existing microbial treatment requirements for
systems where additional public health protection is needed. The rule will require filtered systems to
monitor their source water for Cryptosporidium. Based on the results, filtered systems must meet one of
four levels of treatment for Cryptosporidium (with the first level requiring no additional treatment). All
unfiltered systems, which are not currently required to provide any treatment for Cryptosporidium, must
achieve 2 or 3 log Cryptosporidium inactivation, depending on their source water Cryptosporidium levels.
The rule also requires systems with uncovered finished water reservoirs either to cover the reservoirs or to
provide additional treatment to reservoir effluent. The rule's provisions are described in detail below.
1.1.1 Monitoring and Treatment Requirements for Filtered Systems
Systems must first monitor source water Cryptosporidium concentrations; based on those results,
they are assigned to different treatment bins. Within each bin, systems will choose technologies from a
toolbox of options for ensuring Cryptosporidium removal or inactivation from treated water. The bins for
source waters with higher concentrations of Cryptosporidium involve treatment options that provide
higher levels of inactivation and/or removal.
1 IESWTR (63 FR 69477 December 1998), LT IESWTR (67 FR 1811 January 2002)
Economic Analysis for the LT2ESWTR 1-1 December 2005
-------
Initial Monitoring for Bin Classification—Systems Serving at Least 10,000 People2
Medium and large filtered systems (those serving at least 10,000 people) will be required to
monitor their raw water sources for Cryptosporidium at each plant at least once per month for a minimum
of 2 years. Bin classification will be based on one of the following:
• The highest 12-month running annual average Cryptosporidium concentration (in oocysts per
liter) if samples are taken monthly (24 samples total), or
• The 2-year mean Cryptosporidium concentration. The facility may conduct monitoring twice
per month for 24 months (48 samples total) or perform additional sampling and include these
results in the calculation of the mean, but the additional samples must be evenly distributed
over the 2-year monitoring period.
Cryptosporidium analysis must be conducted in accordance with EPA Method 1622/23 using a
sample volume of at least 10 liters.3 Samples must also be analyzed for E. coll and turbidity. EPA and
stakeholders will use the E. coll and turbidity data to evaluate methods for predicting Cryptosporidium
occurrence.
Systems with at least 2 years of historical Cryptosporidium data that are equivalent in sample
number, frequency, and quality to data required under the LT2ESWTR may use these data to determine
bin classification in lieu of additional Cryptosporidium compliance monitoring, if the State approves the
use of these data.
Monitoring for systems serving at least 100,000 people starts no later than 6 months after
promulgation of the LT2ESWTR. Monitoring for systems serving at least 50,000 people and fewer than
100,000 people starts no later than 12 months after promulgation of the LT2ESWTR. Monitoring for
systems serving at least 10,000 people and fewer than 50,000 people starts no later than 24 months after
promulgation of the LT2ESWTR. Systems will submit monitoring data to EPA as they are generated;
they will then be entered into an EPA database. At the end of the 2-year monitoring period, EPA will
give the results to the States/Primacy Agencies, which will then work with their systems to determine
appropriate compliance steps.
Initial Monitoring for Bin Classification—Systems Serving Fewer than 10,000 People
Source water monitoring for small systems (those serving fewer than 10,000 people) will begin 2
years after the first large systems initiate source water Cryptosporidium monitoring. The required
monitoring is on a delayed schedule so EPA can incorporate information on E. coli and turbidity collected
by the medium and large systems into the monitoring requirements. (EPA will examine these data and
their use as indicators of Cryptosporidium.) In the absence of anew indicator, small systems will conduct
1 year of biweekly E. coli source water monitoring and will be required to conduct Cryptosporidium
monitoring only if E. coli concentrations exceed the following levels:
• An annual mean concentration greater than 10 E. coli per 100 mL for lake and reservoir
source waters; or
2The monitoring and treatment requirements for wholesale systems—i.e., those that sell water only to other
systems—are dependent on the population served by the largest system in the combined distribution system.
3 Systems must meet all requirements of the analytical methods for Cryptosporidium, which include
analysis of two matrix spiked samples.
Economic Analysis for the LT2ESWTR 1-2 December 2005
-------
• An annual mean concentration greater than 50 E. coll per 100 mL for flowing stream source
waters.
Systems that do not exceed these levels are assumed to have a Cryptosporidium concentration of
less than 0.075 oocysts/L and are placed in Bin 1 (see Exhibit 1.1). Small systems that exceed the E. coll
levels mentioned above will be required to conduct semimonthly Cryptosporidium monitoring for a 1-
year period or monthly for a 2-year period, beginning 6 months after the conclusion of E. coll monitoring.
Bin classification for small systems conducting Cryptosporidium monitoring is determined by the highest
12-month running annual average.
All filtered systems that provide or will provide 5.5 log treatment4 for Cryptosporidium by the
date they must comply with additional Cryptosporidium treatment requirements are exempt from
monitoring and subsequent bin classification. To meet the requirement for 5.5 log treatment, systems
using conventional treatment would be required to provide 2.5 log additional treatment, and systems
using direct filtration would be required to provide 3 log additional treatment.
Bins and Treatment Requirements—All System Sizes
Exhibit 1.1 presents the bins for filtered systems according to the type of treatment already in
place. Systems must meet Cryptosporidium treatment requirements by using one of the treatment options
in the "microbial toolbox," or by demonstrating performance equivalent to or exceeding the required
treatment. Systems have 3 years after first being assigned to a bin to meet the treatment requirements
associated with the bin. States/Primacy Agencies may grant systems a 2-year extension to comply if
capital investments are necessary.
4The term "log removal" is used when the contaminant is eliminated by way of filtration; "log inactivation"
is used when oocysts are killed by disinfection. The term "log treatment" encompasses both removal and
inactivation, and is used to reflect the fact that under the LT2ESWTR, treatment will be achieved using a
combination of filtration, disinfection, and other non-traditional methods.
Economic Analysis for the LT2ESWTR 1-3 December 2005
-------
Exhibit 1.1: Bin Classifications and Treatment Requirements for Filtered Systems
If your source
water
Cryptosporidium
concentration
(oocysts/L) is ...
< 0.075
> 0.075 and < 1.0
> 1.0 and < 3.0
>3.0
Your bin
classification
is ...
1
21
33
43
And if you use the following filtration treatment in full
compliance with existing regulations, then your additional
treatment requirements are . . .
Conventional
Filtration
No additional
treatment
1 log treatment
2 log treatment
2.5 log
treatment
Direct
Filtration
No
additional
treatment
1 .5 log
treatment
2.5 log
treatment
Slog
treatment
Slow Sand or
Diatomaceous
Earth
Filtration
No additional
treatment
1 log treatment
2 log treatment
2.5 log
treatment
Alternative
Filtration
Technologies
No additional
treatment
As determined
by the State2
As determined
by the State4
As determined
by the State5
1Systems may use any technology or combination of technologies from the microbial toolbox.
2Total Cryptosporidium treatment must be at least 4.0 log.
3Systems must achieve at least 1 log of the required treatment using ozone, chlorine dioxide, ultraviolet light (UV),
membranes, bag/cartridge filters, or bank filtration.
"Total Cryptosporidium treatment must be at least 5.0 log.
5Total Cryptosporidium treatment must be at least 5.5 log.
The total Cryptosporidium treatment required for Bins 2, 3, and 4 is 4.0 log, 5.0 log, and 5.5 log,
respectively. The additional treatment requirements in Exhibit 1.1 are based on a determination that
conventional, slow sand, and diatomaceous earth filtration plants in compliance with the IESWTR or
LT1ESWTR achieve an average of 3 log removal of Cryptosporidium. (The IESWTR and LT1ESWTR
require systems to achieve 2 log removal; this number is based on the minimum removal expected with
these types of filtration.) Therefore, conventional, slow sand, and diatomaceous earth filtration plants
will require an additional 1.0 to 2.5 log additional treatment to meet the total removal requirement,
depending on the bin in which they are placed.
EPA has determined that direct filtration plants achieve an average 2.5 log removal of
Cryptosporidium. (Their removal is less than in conventional filtration because they lack a sedimentation
process.) Consequently, under the LT2ESWTR, direct filtration plants in Bins 2-4 must provide 0.5 log
more in additional treatment than conventional plants to meet the total Cryptosporidium treatment
requirement.
Microbial Toolbox for Meeting Additional Treatment Requirements
To meet the Cryptosporidium treatment requirements for the bin in which they are classified,
filtered systems can select from a "toolbox" of treatment or management options. The technologies and
management strategies in the microbial toolbox are presented in Exhibit 1.2. Each option in the toolbox
is worth a certain amount of log treatment credit, which systems can apply toward their log treatment
requirements. Systems do not get the log credit automatically when they install these technologies; they
Economic Analysis for the LT2ESWTR
1-4
December 2005
-------
must show that they are meeting certain operational or other criteria specific to the technology. Log
treatment credit under existing rules (e.g., the IESWTR and LT1ESWTR) works much the same way.
Systems currently using ozone, chlorine dioxide, ultraviolet light (UV), or membranes in addition to
conventional treatment may receive credit for those technologies toward meeting bin requirements if they
meet the LT2ESWTR criteria for the chosen technology.
Economic Analysis for the LT2ESWTR 1-5 December 2005
-------
Exhibit 1.2: Microbial Toolbox Components for the LT2ESWTR
(To be used in addition to existing treatment)
Toolbox Option
Log Treatment Credit
Source Toolbox Components
Watershed control program
Alternative source/intake
management
0.5
None, but conduct source water monitoring
concurrently at both sources or under both
intake management plans and determine the
bin based on the lower mean concentration
Pre-Filtration Toolbox Components
Presedimentation basin with
coagulation
Two-stage lime softening
Bank filtration
0.5
0.5
0.5 or 1 .0, depending on setback
Treatment Performance Toolbox Components
Combined filter performance
Individual filter performance
Demonstration of performance
0.5
1.0
State approved1
Additional Filtration Toolbox Components
Bag filters
Cartridge filters
Membrane filtration
Second stage filtration
Slow sand filters
2.0 as individual and 2.5 for two in series
2.0 as individual and 2.5 for two in series
As demonstrated2
0.5
2.5
Inactivation Toolbox Components
Chlorine dioxide
Ozone
UV
As demonstrated3
As demonstrated3
As demonstrated2
1The State must approve the method used to demonstrate performance and must approve the log
credit claimed by the system.
2Credit for membrane filtration and UV is based on the results of equipment-specific testing.
3Credit for chlorine dioxide and ozone is based on CT values achieved (CT is the product of
disinfectant concentration and contact time).
Economic Analysis for the LT2ESWTR
1-6
December 2005
-------
Reassessment and Future Monitoring
Six years after initial bin classification, systems will be required to conduct a second round of
monitoring to reassess source water conditions for bin assignments. Two years before this reassessment
(4 years after initial binning), EPA plans to initiate a stakeholder process to review available analytical
methods for detecting Cryptosporidium. If there are new, improved methods, EPA, with stakeholder
input, will determine the appropriate analytical method, frequency, and locations for the second round of
national assessment monitoring. If no improved Cryptosporidium detection method is available,
monitoring will follow EPA Method 1622/23. Systems that provide a total of 5.5 log treatment for
Cryptosporidium are not subject to future monitoring.
In addition to the reassessment and re-binning described above, the State/Primacy Agency will
assess any significant changes in the watershed and source water as part of the sanitary survey process. It
will then determine what follow-up action is appropriate in response to any source water changes that
have taken place; responses could include actions from the microbial toolbox.
1.1.2 Monitoring and Treatment Requirements for Unfiltered Systems
Unfiltered systems that already have 3 log Cryptosporidium treatment in place prior to the date
they would have to comply with treatment requirements are exempt from monitoring and additional
Cryptosporidium inactivation requirements. Otherwise, large unfiltered systems must monitor
Cryptosporidium in their source water monthly for at least 2 years, and small unfiltered systems must
monitor semimonthly for 12 months or monthly for 24 months. All unfiltered systems must determine
their treatment requirements based on the arithmetic mean Cryptosporidium concentration. If their
average Cryptosporidium concentration is less than or equal to 0.01 oocysts/L, systems must provide 2
log Cryptosporidium inactivation. If their average concentration is greater than 0.01 oocysts/L, they must
provide 3 log inactivation.
Monitoring for unfiltered systems will be based on the same schedule as monitoring for filtered
systems, although unfiltered systems are not required to monitor E. coll or turbidity. As with the filtered
systems, unfiltered systems must conduct a second round of Cryptosporidium monitoring 6 years after the
initial bin assignment.
In addition to the new Cryptosporidium inactivation requirements, the LT2ESWTR will require
unfiltered systems to continue to meet the filtration avoidance criteria under the 1989 SWTR and to
continue to provide inactivation for Giardia and viruses. The overall inactivation requirements (i.e., 4 log
virus, 3 log Giardia, and 2 or 3 log Cryptosporidium) must be met using a minimum of two disinfectants.
Additionally, each of two disinfectants must meet the total inactivation for one of the three pathogens.
For example, a system could use UV to inactivate 2 log of Cryptosporidium and Giardia and use chlorine
to inactivate 4 log of viruses and 1 log of Giardia.
1.1.3 Requirements for Existing Uncovered Finished Water Reservoirs
The LT2ESWTR builds on the IESWTR and LT1ESWTR, which require covers only for new
finished water reservoirs. The LT2ESWTR will establish requirements for all systems with existing
uncovered finished water reservoirs. Systems must either cover the reservoir or treat reservoir discharge
to the distribution system to achieve 2 log Cryptosporidium, 3 log Giardia, and 4 log virus inactivation.
Economic Analysis for the LT2ESWTR 1-7 December 2005
-------
1.1.4 Disinfection Profiling and Benchmarking Requirements
The LT2ESWTR includes a disinfection profile and benchmark requirement to ensure that any
significant change in disinfection, whether for byproduct control under the Stage 2 DBPR, improved
Cryptosporidium control under the LT2ESWTR, or both, does not significantly compromise existing
Giardia and virus protection. A disinfection profile is a graphical representation of a system's level of
Giardia and viral inactivation measured during the course of 1 or more year(s). A benchmark is the
lowest monthly average of microbial inactivation during the disinfection profile period.
The profiling and benchmarking requirements under the LT2ESWTR are similar to those
promulgated under the IESWTR and LT1ESWTR and are applicable to systems making a significant
change to their disinfection practice. The LT2ESWTR requires these systems to prepare a disinfection
profile that characterizes current levels of Giardia lamblia and virus inactivation throughout the plant
over the course of 1 year. The profile may be developed using equivalent historical data. Prior to making
the change, the system must calculate a benchmark and consult with the State regarding how the proposed
change will affect the current disinfection level.
1.1.5 Implementation Timeline
Exhibit 1.3 shows the timeline of LT2ESWTR activities for filtered systems. The schedule for
monitoring and compliance with treatment requirements differs by population served.
Exhibit 1.3: Implementation Time Line for LT2ESWTR for Filtered Systems
YeaM
System:
Year 2
Serving ;
YearS
: 100,000
Crypto Monitoring
System!
System!
System!
YeaM
Serving
>0,000-99
Crypto Monitoring
Serving
; Serving
Year 2
10,000-49
Year 4
People
YearS
Year6
Treatment Installation
999 Peop
e
Year 7
YearS
Possible Extension
Year 9
Treatment Installation Possible Extension
999 Peop
Crypto Monitoring
<1 0,000 F
'eople
E. co// Mon.
YearS
Year 4
e
Year 10
Year 1 1
2nd Round Crypto Mon.
Year 12
2nd Round Crypto Mon.
Treatment Installation Possible Extension
(Optional)
Crypto Monitoring
YearS
Year6
Year 13
2nd Round Crypto Mon.
Treatment Installation Possible Extension
Year 7
YearS
Year 9
Year 10
Year 1 1
2nd Round
Year 12
- co// Mon.
Year 13
Economic Analysis for the LT2ESWTR
December 2005
-------
1.2 Document Organization
This EA is organized into the following chapters:
• Chapter 2 identifies the public health concerns addressed by the rule and provides a
20-year regulatory history that includes a description of relevant National Primary
Drinking Water Regulations (NPDWRs). It also explains the statutory authority for
promulgating the LT2ESWTR and the economic rationale for choosing a regulatory
approach.
• Chapter 3 describes the regulatory alternatives considered for the LT2ESWTR and
the process for developing them.
• Chapter 4 characterizes the baseline conditions that EPA expects to exist (including
system inventory, treatment, and water quality data) before systems complete the
monitoring and begin making treatment changes to meet the LT2ESWTR
requirements. Because of the timing of the IESWTR, LT1ESWTR, and Stage 1
DBPR5, EPA had to predict the changes in treatment and water quality made by
systems as a result of these rules to characterize pre-LT2ESWTR baseline conditions.
• Chapter 5 reviews available toxicological and epidemiological data related to
Cryptosporidium and presents the public health and economic benefits (both
quantifiable and unquantifiable) of this rule. It compares the benefits of the four
regulatory alternatives and presents several sensitivity analyses.
Chapter 6 presents an estimate of the costs of implementing the rule to the drinking
water industry, households, and States/Primacy Agencies. It also compares the costs
of the four regulatory alternatives.
• Chapter 7 discusses analyses performed to evaluate the effects of the rule on different
segments of the population. It also considers various executive orders and other
requirements, including the Regulatory Flexibility Act (RFA) and Unfunded
Mandates Reform Act (UMRA).
• Chapter 8 summarizes and analyzes the rule's benefit and cost estimates. It also
compares the results of the Preferred Regulatory Alternative to the other alternatives
considered.
5 The compliance deadlines for the IESWTR and Stage 1 DBPR were January 2002 for large and medium
surface water systems and January 2004 for small systems. The compliance deadline for the LT1ESWTR was
January 2005.
Economic Analysis for the LT2ESWTR 1-9 December 2005
-------
2. Statement of Need for the Rule
2.1 Introduction
This chapter presents the need for the LT2ESWTR by identifying the public health concerns that
this rule will address. Included is a discussion of related regulations that shows that these public health
concerns are not adequately addressed by other rules. This chapter concludes with a discussion of the
economic rationale for the rule. The remaining sections are organized as follows:
2.2 Description of the Issue
2.3 Risk Balancing
2.4 Public Health Concerns to be Addressed
2.4.1 Cryptosporidium
2.4.2 Uncovered Finished Water Reservoirs
2.5 Regulatory History
2.5.1 Statutory Authority for Promulgating the Rule
2.5.2 1979 Total Trihalomethane Rule
2.5.3 1989 Total Coliform Rule
2.5.4 1989 Surface Water Treatment Rule
2.5.5 1996 Information Collection Rule
2.5.6 1998 Interim Enhanced Surface Water Treatment Rule
2.5.7 1998 Stage 1 Disinfectants and Disinfection Byproducts Rule
2.5.8 2000 Proposed Ground Water Rule
2.5.9 2001 Filter Backwash Recycling Rule
2.5.10 2002 Long Term 1 Enhanced Surface Water Treatment Rule
2.5.11 2003 Proposed Stage 2 Disinfectants and Disinfection Byproducts Rule
2.6 Economic Rationale for Regulation
2.2 Description of the Issue
More than 14,000 public water systems (PWSs), serving approximately 180 million people in the
United States and its territories, use surface water or ground water under the direct influence of surface
water (GWUDI) as their source. These sources carry microbial contaminants, some of which pose
significant risks to public health. The U.S. Environmental Protection Agency (EPA or the Agency) is
particularly concerned about Cryptosporidium because it is resistant to many commonly used drinking
water disinfectants, such as chlorine, and it poses significant health risks, including death. Moreover,
there is no effective drug available to cure cryptosporidiosis, the health condition caused by
Cryptosporidium infection (Framm and Soave 1997). The primary issue of concern addressed by this rule
is the risk to public health in water supplies with inadequate Cryptosporidium treatment.
The 1989 SWTR requires most surface water and GWUDI systems to remove microbial
contaminants physically through filtration. (Exemptions to this filtration requirement are granted to
systems that meet specified avoidance criteria.) Types of filtration systems include the following:
• Conventional treatment—coagulation, flocculation, and sedimentation of particles, followed
by granular media filtration.
Economic Analysis for the LT2ESWTR 2-1 December 2005
-------
Direct filtration—coagulation and flocculation followed by rapid sand filtration, but no
sedimentation basin. This type of filtration system is designed for low-turbidity waters.
• Slow sand and diatomaceous earth filtration—filters that work at very low flow rates without
the use of a coagulant in pretreatment.
Alternative filtration—other technologies, including membranes and bag and cartridge filters.
Current regulations specify the performance of filtration systems in terms of filtered water
turbidity limits. Turbidity is a measure of the clarity of water and is quantified in nephelometric turbidity
units (NTU). The 1989 SWTR required all surface water and GWUDI systems using rapid sand filtration
technologies to meet combined filter effluent turbidity limits of 0.5 NTU 95 percent of the time. The
1998 Interim Enhanced Surface Water Treatment Rule (IESWTR) requires improved filtration
performance by lowering the turbidity standard to 0.3 NTU 95 percent of the time (with a maximum of 1
NTU at any time) for large systems using rapid sand filtration. The 2002 LT1ESWTR extended this
requirement to small systems. At this lower limit, EPA believes that systems are generally achieving a
minimum of 2 log (99 percent) removal ofCryptosporidium. Slow sand and diatomaceous earth filtration
systems can achieve at least 2 log removal at a higher effluent turbidity of 1 NTU 95 percent of the time
because of differences in their removal mechanisms. While the degree ofCryptosporidium reduction
achieved under these standards may provide adequate public health protection for some source waters,
EPA recognizes that some systems may have higher levels of contamination against which additional
protection is warranted. Methods such as additional filtration, the use of alternative disinfectants such as
ozone or ultraviolet light (UV), improved source water protection, or other treatment and management
initiatives can help systems achieve additional protection against Cryptosporidium.
The LT2ESWTR also addresses the risk of microbial pathogen contamination in unfiltered
systems, which lack the protective barriers from Cryptosporidium that filtered systems provide. The rule
requires unfiltered systems to provide at least 2 log inactivation ofCryptosporidium by disinfection; the
amount of disinfection will depend on the results of source water Cryptosporidium monitoring. The rule
also requires the use of two disinfectants to meet Cryptosporidium and existing inactivation requirements
(i.e., those for Giardia and viruses).
Lastly, the LT2ESWTR addresses health risks posed by uncovered finished water reservoirs.
There are approximately 81 uncovered reservoirs that hold finished water, not including those that are
scheduled to be covered or taken off-line, ranging in size from a few thousand to more than 3 billion
gallons. While the IESWTR and LTIESWTR require systems to cover all new finished water reservoirs,
the LT2ESWTR builds on these rules by addressing existing uncovered finished water reservoirs.
2.3 Risk Balancing
EPA expects some systems to change treatment practices in response to the Stage 2 Disinfectants
and Disinfection Byproducts Rule (DBPR) requirements. These changes have the potential to increase
the occurrence of microbial pathogens in drinking water as systems alter the use of disinfectants to
comply with the new disinfection byproduct (DBP) requirements. DBFs result from chemical reactions
between disinfectants and organic and inorganic compounds in the water. Some DBFs are associated
with health risks, including adverse developmental and reproductive health effects and cancer. The
LT2ESTWR, therefore, has additional disinfection profiling and benchmarking provisions to help ensure
that systems maintain control of microbial risks as they take steps to reduce the formation of DBFs.
Economic Analysis for the LT2ESWTR 2-2 December 2005
-------
EPA is making a concerted effort to understand and balance risks from DBFs and microbes, and
the costs and benefits of addressing those risks in its rulemaking efforts. To allow for simultaneous
compliance and balancing of risks between microbial pathogens and DBFs, EPA is promulgating the
LT2ESWTR concurrently with the Stage 2 DBPR. For detailed information regarding the Stage 2 DBPR,
see the draft Economic Analysis for the Stage 2 Disinfectants and Disinfection Byproducts Rule (USEPA
2003d).
2.4 Public Health Concerns to Be Addressed
In 1990, EPA's Science Advisory Board (SAB), an independent panel of experts established by
Congressional mandate, cited drinking water contamination as one of the most important environmental
risks and indicated that disease-causing microbial contaminants (i.e., bacteria, protozoa, and viruses) pose
a particularly high health risk due to the large populations that are directly exposed to them (SAB and
USEPA 1990). Information on waterborne disease outbreaks from the U.S. Centers for Disease Control
and Prevention (CDC) underscores this concern. Data collected by CDC indicate that between 1971 and
2002, 757 waterborne disease outbreaks, caused by various types of contamination, were reported (Craun
and Calderon 1996, Levy et al. 1998, Barwick et al. 2000, Lee et al. 2002, Blackburn et al. 2004).
The effects of waterborne disease are usually acute, resulting from a single or small number of
exposures. Most waterborne pathogens cause gastrointestinal illness with diarrhea, abdominal
discomfort, nausea, vomiting, or other symptoms. Most such cases involve a sudden onset and generally
are of short duration in healthy people. Some pathogens (e.g., Giardia and Cryptosporidium), however,
may cause extended illness, lasting weeks or longer in otherwise healthy individuals. The infection can
prove fatal for members of sensitive populations, such as the immunocompromised or the elderly. Other
waterborne pathogens cause, or at least are associated with, more serious disorders such as hepatitis,
particularly hepatitis A (Moore et al. 1993), peptic ulcers and gastric cancer (Helicobacterpylori) (Park et
al. 2001, Sepulveda and Graham 2002), myocarditis (group B coxsackievirus) (Kim et al. 2001),
meningitis (group B coxsackievirus and echoviruses) (Lee and Kim 2002, Amvrosieva et al. 2001), and
other diseases.
2.4.1 Cryptosporidium
Cryptosporidium is of particular concern to EPA because, unlike pathogens such as viruses and
bacteria, Cryptosporidium oocysts are resistant to inactivation by many common disinfection methods.
Since the oocyst is especially resistant to chlorine disinfection, simply increasing existing chlorination
dosage levels or contact time above those most commonly practiced in the United States is not effective.
Other emerging disinfectant-resistant pathogens, such asMicrosporidia, Cyclospora, and Toxoplasma, are
also a concern for similar reasons.
Cryptosporidiosis is a protozoal infection that usually causes 7 to 14 days of diarrhea, possibly
accompanied by low-grade fever, nausea, and abdominal cramps in individuals with healthy immune
systems (Juranek 1998). It is caused by the ingestion of infectious oocysts, which are readily carried in
water. The most common source of oocysts in water is the feces of infected hosts (Perz et al. 1998; Rose
1997). Although cryptosporidiosis often occurs through ingestion of contaminated food or water, it may
also result from direct or indirect contact with infected people or animals (Casemore 1990; Juranek 1998;
Rose 1997). Infected humans and other animals excrete oocysts, which can then be transmitted to others.
Okhuysen et al. (1998) and Dupont et al. (1995) found through human volunteer feeding studies that a
low dose of Cryptosporidiumparvum (or C. parvum) is sufficient to cause infection in healthy adults.
Economic Analysis for the LT2ESWTR 2-3 December 2005
-------
Some subpopulations are at greater risk of serious illness or death from waterborne disease than
the general population (Frost et al. 1997). These include children (especially the very young), the elderly,
pregnant women, and the immunocompromised. These sensitive groups account for almost 20 percent of
the population in the United States (Gerba et al. 1996; USEPA 1998a). The severity and duration of
illness are often greater in immunocompromised people than in healthy individuals, and death may result.
For instance, of the people who died in the 1993 Milwaukee cryptosporidiosis outbreak, 85 percent had
AIDS as the underlying cause of death (Hoxie et al. 1997).
Cryptosporidium has caused a number of documented waterborne disease outbreaks. However, it
is important to note that C. parvum was not identified as a human pathogen until 1976, and outbreaks
attributed to cryptosporidiosis were not reported in the United States prior to 1984. The first report of an
outbreak caused by Cryptosporidium was published during the development of the SWTR (D'Antonio et
al. 1985). EPA, CDC, and the Council of State and Territorial Epidemiologists have maintained a
collaborative surveillance program for collection and periodic reporting of data on waterborne disease
outbreaks since 1971. The CDC database and biennial CDC-EPA surveillance summaries include data
reported voluntarily by the States on the incidence and prevalence of waterborne illnesses.
Between 1991, the first year the SWTR and Total Coliform Rule were in effect, and 2002, the
most recent year for which data are available, 106 drinking water-related outbreaks associated with
confirmed or suspected microbiological causes occurred in PWSs. Twenty-one outbreaks occurred in
PWSs with surface water sources; the rest were in systems using wells or springs. The etiology of
outbreaks included Cryptosporidium; Giardia; bacteria such as Campylobacter jejuni, Shigella sonnei,
and E. coll O157:H7; Norwalk-like viruses and small round-structured viruses; and acute gastrointestinal
illness of unknown etiology (AGI). These outbreaks are listed individually in Appendix A of the
Occurrence and Exposure Assessment for the Long Term 2 Enhanced Surface Water Treatment Rule and
are based on CDC surveillance summaries (Moore et al. 1993, Kramer et al. 1996a, Levy et al. 1998,
Barwick et al. 2000, Lee et al. 2002, Blackburn et al. 2004).
From 1984 to 2002, there were 10 reported outbreaks of cryptosporidiosis associated with
drinking water in PWSs in the United States (Moore et al. 1993; Kramer et al. 1996a; Craun 1996; Levy
et al. 1998; Barwick et al. 2000, Lee et al. 2002, Blackburn et al. 2004). Two additional outbreaks
occurred in private wells, and another 52 outbreaks occurred in recreational waters. Exhibit 2.1
summarizes the cryptosporidiosis outbreaks associated with drinking water.
Economic Analysis for the LT2ESWTR 2-4 December 2005
-------
Exhibit 2.1: Reported Cryptosporidiosis Outbreaks in
U.S. Drinking Water Systems
Year
1984
1987
1991
1992
1993
1993
1994
1994
1998
2000
Location and
System Type
Braun Station, TX,
CWS
Carrollton, GA,
CWS
Berks County, PA,
NCWS
Medford (Jackson
County and
Talent), OR,
CWS
Milwaukee, Wl,
CWS
Cook County, MN,
NCWS
Clark County, NV,
CWS
Walla Walla, WA,
CWS
Williamson
County, TX,
CWS
Florida, CWS
Cases of Illness
117 (confirmed)
2,000 (estimated)
13,000 (estimated)
551 (estimated)
3,000 (estimated);
combined total for
Jackson County and
Talent
403,000 (estimated)
27 (confirmed)
103 (confirmed);
many were HIV
positive
134 (confirmed)
1,400 (confirmed)
5
Source
Water
Well
River
Well
Spring/River
Lake
Lake
River/Lake
Well
Well
Well
Treatment
Chlorination
Conventional filtration/
chlorination; inadequate
backwashing of some
filters
Chlorination
Chlorination/package
filtration plant
Conventional filtration
Filtered, chlorinated
Prechlorination,
filtration and post-
filtration chlorination
None reported
Chlorinated
Chlorinated
Suspected
Cause
Sewage-contaminated
well
Treatment
deficiencies
Ground water
under the influence of
contaminated surface
water
Source not identified
for Jackson County;
treatment deficiencies
at water treatment
plant in Talent
High source water
contamination and
treatment deficiencies
Possible sewage
backflow from
toilet/septic tank
Source not identified
Sewage contamination
Sewage contamination
Broken well, treatment
deficiencies
Source: Craun et al. (1998), Berwick et al. (2000), and Lee et al. (2002).
Five of the 10 outbreaks in Exhibit 2.1 originated from surface water or possibly GWUDI
supplied by PWSs serving fewer than 10,000 people. In total, the 10 outbreaks caused an estimated
421,337 cases of illness, the majority occurring in Milwaukee in 1993. These outbreaks demonstrate that
when treatment is not operating optimally or when source water is highly contaminated, Cryptosporidium
can be present in the finished drinking water and infect consumers, ultimately resulting in disease
outbreaks.
The National Research Council concluded that the number of identified and reported outbreaks in
the CDC database (both for surface and ground waters) represents a small percentage of actual waterborne
disease outbreaks (National Research Council 1997). Most outbreaks in CWSs are not recognized until a
sizable proportion of the population is ill (Perz et al. 1998, Craun 1996). In addition to the complications
involved in identifying waterborne disease outbreaks, some States do not have active outbreak
surveillance programs. Those that do exist are based on voluntary and confidential responses by State and
local public health officials. Even when outbreaks are recognized, few are successfully traced to the
Economic Analysis for the LT2ESWTR
2-5
December 2005
-------
drinking water source. Physicians, for instance, may not have sufficient community-wide information to
attribute gastrointestinal illness to any specific origin, such as a drinking water source. Many people who
experience gastrointestinal illness (predominantly manifested as diarrhea) do not seek medical attention,
and some healthy adults with cryptosporidiosis may not suffer severe symptoms from the disease. Even if
infected individuals consult a physician, Cryptosporidium is not identified by routine diagnostic tests for
gastroenteritis and, therefore, tends to be under-reported in the general population (Craun 1996).
The limited number of reported cases of waterborne disease such as cryptosporidiosis may be due
to the fact that a significant portion of these illnesses may be endemic (i.e., not associated with an
outbreak), and thus are even more difficult to recognize. One study, for example, found that 14 to 40
percent of the normal gastrointestinal illness in a community in Quebec was associated with treated
drinking water from a surface water source (Payment et al. 1997).
2.4.2 Uncovered Finished Water Reservoirs
Many PWSs store treated drinking water in some type of reservoir before delivering it to their
customers. Although good engineering practice dictates that such reservoirs be covered to prevent
recontamination, there are currently no regulations that require existing reservoirs to be covered. (The
IESWTR and LT1ESWTR require new reservoirs to be covered.) The use of uncovered finished water
reservoirs has been questioned since 1930 because of their susceptibility to contamination and subsequent
threats to public health. Many sources of contamination can lead to the degradation of water quality in
uncovered finished water reservoirs. These include, but are not limited to, surface water runoff, algal
growth, insects and fish, bird and animal waste, airborne deposition, and human activity. Algal blooms
are the most common problem in open reservoirs and can become a public health risk. Algae growth
leads to the formation of DBFs and causes taste and odor problems. Algae also provide a food source for
bacteria that decompose plant matter. Some blue-green algae (actually a type of bacterium called
cyanobacteria) contain toxins that can induce headaches, fever, diarrhea, abdominal pain, nausea, and
vomiting. Bird and animal wastes are other common and significant sources of contamination. These
wastes may carry microbial contaminants such as coliform bacteria, viruses, and human pathogens,
including Vibrio cholera, Salmonella, Mycobacteria, bacteria that cause typhoid fever, and Giardia, in
addition to Cryptosporidium (USEPA 1999c). Microbial pathogens can also be found in surface water
runoff along with agricultural chemicals, automotive wastes, turbidity, metals, and organic matter
(USEPA 1999c; LeChevallier et al. 1997b). In an effort to minimize contamination, systems have
implemented controls such as reservoir covers and liners, regular draining and washing, proper security
and monitoring, bird and insect control programs, and drainage designed to prevent surface runoff from
entering the reservoir (USEPA 1999c).
Few studies quantitatively evaluate the impacts of uncovered finished water reservoirs on public
health. LeChevallier et al. (1997b) compared the influent and the effluent water quality from six New
Jersey reservoirs for a 1-year period to determine the impact of uncovered finished water storage
reservoirs on water quality. There were significant increases in turbidity, particle counts, total coliform,
fecal coliform, and heterotrophic plate count bacteria in the effluent compared to the influent. There was
also a significant decrease in the chlorine residual in the effluent samples, meaning little chlorine would
be left to provide continued disinfection in the distribution system.
Economic Analysis for the LT2ESWTR 2-6 December 2005
-------
2.5 Regulatory History
The primary responsibility for regulating the quality of drinking water lies with EPA. The Safe
Drinking Water Act (SDWA) establishes this responsibility and defines the mechanisms at the Agency's
disposal to protect public health. EPA sets standards by identifying which contaminants should be
regulated and by establishing the maximum levels of the contaminants allowed in drinking water,
specifying treatment techniques to reduce contaminant levels.
2.5.1 Statutory Authority for Promulgating the Rule
Section 1412(b)(l) of the 1996 SDWA reauthorization mandated new drinking water
requirements. EPA's general authority to set Maximum Contaminant Level Goals (MCLGs) and develop
National Primary Drinking Water Regulations (NPDWRs) was modified to apply to contaminants that
"may have an adverse effect on the health of persons," are "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 for which, "in the sole judgment of the Administrator, regulation of such contaminant
presents a meaningful opportunity for health risk reductions for persons served by public water systems"
(SDWA 1412(b)(l)(A)).
To regulate a contaminant, EPA first sets an MCLG at a level at which no known or anticipated
adverse health effects occur. MCLGs are established solely on the basis of protecting public health and
are not enforceable. EPA simultaneously sets an enforceable Maximum Contaminant Level (MCL) as
close as technologically feasible to the MCLG, while taking costs into consideration. If it is not feasible
to measure the contaminant at levels presumed to have impacts on health, a treatment technique can be
specified in place of an MCL. Water systems comply with a drinking water regulation by not exceeding
the MCL or by meeting treatment technique requirements.
In addition to the general authorities cited above, SDWA Section 1412(b)(2)(C) requires
specifically that EPA promulgate the Final Enhanced Surface Water Treatment Rule (ESWTR).
The Administrator shall promulgate an Interim Enhanced Surface Water Treatment Rule,
a Final Enhanced Surface Water Treatment Rule, a Stage 1 Disinfectants and Disinfection
Byproducts Rule, and a Stage 2 Disinfectants and Disinfection Byproducts Rule in
accordance with the schedule published in volume 29, Federal Register, Page 6361
(February 10, 1994), in Table III. 13 of the proposed Information Collection Rule.
The promulgation of the IESWTR and LT1 ESWTR satisfied the statutory requirement for an
interim rule, and the LT2ESWTR satisfies the requirement for a final rule and the Congressional intent to
review and revise the IESWTR and LTIESWTR based on data available from the Information Collection
Rule (ICR) and Information Collection Rule Supplemental Survey (ICRSSs) (see section 2.5.5). Also, to
achieve the goals of the Stage 2 DBPR, the LT2ESWTR must be promulgated to achieve a balance
between the risks of microbial pathogens and DBFs.
The following sections summarize the development of NPDWRs over the past 20 years.
Economic Analysis for the LT2ESWTR 2-7 December 2005
-------
2.5.2 1979 Total Trihalomethane Rule
Under the Total Trihalomethane Rule (44 FR 68624 November 1979), EPA set an MCL for total
trihalomethanes (TTHM), the sum of chloroform, bromoform, bromodichloromethane,
dibromochloromethane, of 0.10 mg/L as a running annual average (RAA) of quarterly samples. This
standard applied to CWSs using surface or ground water that served at least 10,000 people and that added
a disinfectant to the drinking water during any part of the treatment process. This 1979 rule was
superseded by the 1998 Stage 1 DBPR(section2.5.7).
2.5.3 1989 Total Coliform Rule
The Total Coliform Rule (TCR) (54 FR 27544 June 1989) applies to all PWSs. Because
monitoring PWSs for every possible pathogenic organism is not feasible, coliform organisms are used as
indicators of possible distribution system contamination. Coliforms are easily detected in water and are
used to indicate a system's vulnerability to pathogens. In the TCR, EPA set an MCLG of zero for total
coliforms. EPA also set a monthly MCL for total coliforms and required testing of total-coliform-
positive cultures for the presence of E. coli or fecal coliforms. E. coli and fecal coliforms indicate more
immediate health risks from sewage or fecal contamination, and their presence is an acute MCL violation,
which requires immediate public notification. Coliform monitoring frequency is determined by the size
of the population served, the type of system (community or noncommunity) and the type of source water
(surface water, GWUDI, or ground water). In addition, the TCR required sanitary surveys every 5 years
(or 10 years for NCWSs using disinfected ground water) for systems that collect fewer than five routine
total coliform samples per month (typically serving fewer than 4,100 people).
2.5.4 1989 Surface Water Treatment Rule
Under the SWTR (54 FR 27486 June 1989), EPA set MCLGs of zero for Giardia lamblia,
viruses, and Legionella, and established treatment requirements for all PWSs using surface water or
GWUDI as a source. The SWTR includes treatment technique requirements for filtered and unfiltered
systems that are intended to protect against the adverse health effects associated with Giardia lamblia,
viruses, and Legionella, as well as many other pathogenic organisms. These requirements include:
Maintenance of a disinfectant residual in water entering and within the distribution system.
Removal/inactivation of at least 99.9 percent (3 log) of Giardia and 99.99 percent (4 log) of
viruses.
• Filtration, unless systems meet specified avoidance criteria.
• For filtered systems, a turbidity performance standard for the combined filter effluent
consisting of a 5-NTU maximum and 95 percent of measurements in 1 month not to exceed
0.5 NTU, based on 4-hour monitoring for treatment plants using conventional treatment or
direct filtration (with separate standards for other filtration technologies). The 1998 IESWTR
and the 2002 LTIESWTR superseded these particular requirements.
• Watershed control programs and water quality requirements for unfiltered systems.
Economic Analysis for the LT2ESWTR 2-8 December 2005
-------
2.5.5 1996 Information Collection Rule
The ICR (61 FR 24354 May 1996) applied to PWSs serving more than 100,000 people. A more
limited set of ICR requirements pertained to ground water systems serving 50,000 to 100,000 people.
The ICR authorized EPA to collect occurrence and treatment information to help evaluate the
need for possible changes to the microbial treatment practices and to help evaluate the need for future
regulation of disinfectants and DBFs. The ICR provided EPA with additional information on the national
occurrence in drinking water of (1) chemical byproducts that form when disinfectants used for microbial
control react with naturally occurring compounds present in source water; and (2) disease-causing
microorganisms, including Cryptosporidium, Giardia, viruses, and coliform bacteria. The ICR also
required water systems to collect plant configuration data showing the type of treatment provided. The
ICR monthly sampling data provided a total of 18 months of influent and treated water quality data
including pH, alkalinity, turbidity, temperature, calcium and total hardness, total organic carbon, UV254,
bromide, ammonia, and disinfectant residual. These data provided an indication of the "treatability" of
the water, the occurrence of contaminants, and the potential for DBF formation. The data collected under
the ICR were used in analyses supporting development of the LT2ESWTR and Stage 2 DBPR.
2.5.6 1998 Interim Enhanced Surface Water Treatment Rule
The IESWTR (63 FR 69478 December 1998) updated the 1989 SWTR for large systems. It
applies to PWSs serving at least 10,000 people and using surface water or GWUDI as a source. These
systems were to comply with the IESWTR by January 2002. The primary purpose of the IESWTR is to
improve control of Cryptosporidium and to address tradeoffs between the risks from microbial pathogens
and those from DBFs. The requirements and guidelines include:
• An MCLG of zero for Cryptosporidium;
• Removal of 99 percent (2 log) of Cryptosporidium for systems that provide filtration;
For treatment plants using conventional treatment or direct filtration, a turbidity performance
standard for the combined filter effluent consisting of a 1 NTU maximum and 95 percent of
measurements in 1 month not to exceed 0.3 NTU, based on 4-hour monitoring;
Continuous monitoring of individual filter effluent turbidity in conventional and direct
filtration plants and recording turbidity readings every 15 minutes when these filters are on-
line;
A disinfection benchmark to assess the level of microbial protection provided before facilities
change their disinfection practices to meet the requirements of the Stage 1 DBPR;
• Inclusion of Cryptosporidium in the definition of GWUDI and in the watershed control
requirements for unfiltered PWSs;
Covers for all new finished water storage facilities; and
A primacy provision that requires States to conduct sanitary surveys with a minimum
frequency for all surface water systems, including those serving fewer than 10,000 people.
Economic Analysis for the LT2ESWTR 2-9 December 2005
-------
EPA promulgated the IESWTR concurrently with the Stage 1 DBPR so that systems could
coordinate their response to the risks posed by DBFs and microbial pathogens.
2.5.7 1998 Stage 1 Disinfectants and Disinfection Byproducts Rule
The Stage 1 DBPR (63 FR 69390 December 1998) applies to all CWSs and NTNCWSs that add a
chemical disinfectant to their water. Certain requirements designed to provide protection against acute
health effects from chlorine dioxide also apply to transient noncommunity water systems (TNCWSs).
Surface water and GWUDI systems serving at least 10,000 people were required to comply with the rule
by January 2002. Surface water and GWUDI systems serving fewer than 10,000 people and all ground
water systems must comply by January 2004.
The Stage 1 DBPR sets maximum residual disinfectant level goals (MRDLGs) for chlorine (4
mg/L as C12), chloramines (4 mg/L as C12), and chlorine dioxide (0.8 mg/L as C1O2); and MCLGs for
bromodichloromethane (0 mg/L), bromoform (0 mg/L), dibromochloromethane (0.06 mg/L),
dichloroacetic acid (0 mg/L), trichloroacetic acid (0.3 mg/L), bromate (0 mg/L), and chlorite (0.8 mg/L).
The rule sets MRDLs for chlorine (4 mg/L as C12), chloramines (4 mg/L as C12), and chlorine dioxide (0.8
mg/L as C1O2); and MCLs for TTHM (0.080 mg/L), HAAS (0.060 mg/L), bromate (0.010 mg/L), and
chlorite (1.0 mg/L). The MRDLs and MCLs, except those for chlorite and chlorine dioxide, are
calculated as RAAs. For conventional surface water and GWUDI systems, a treatment
technique—enhanced coagulation/softening—is specified for the removal of DBP precursors.
As noted in section 2.5.6, the Stage 1 DBPR was promulgated concurrently with the IESWTR to
coordinate the control of DBFs and microbial contaminants.
2.5.8 2000 Proposed Ground Water Rule
The proposed Ground Water Rule (65 FR 30194 May 2000) addresses fecal contamination in
ground water systems. The rule also builds on the TCR through provisions based on further evaluation of
E. coli monitoring results measured under the TCR. Key components of the multibarrier approach for
protection of ground water included in the proposed rule are:
• Sanitary surveys for all ground water systems conducted at the same frequency as in surface
water systems;
Hydrogeologic sensitivity assessments to identify ground water sources that are susceptible to
fecal contamination;
• Source water monitoring for an indicator of fecal contamination for systems drawing from
susceptible ground water sources;
Correction of significant deficiencies and fecal contamination by eliminating the source of
contamination, correcting the deficiency, providing an alternative source of water, or
providing inactivation and/or removal of 99.99 percent (4 log) of viruses; and
Compliance monitoring to ensure that disinfection treatment is reliably operated when it is
used.
Economic Analysis for the LT2ESWTR 2-10 December 2005
-------
2.5.9 2001 Filter Backwash Recycling Rule
The Filter Backwash Recycling Rule (FBRR) (66 FR 31086 June 2001) regulates systems where
filter backwash is returned to the treatment process. The rule applies to surface water and GWUDI
systems that use direct or conventional filtration and recycle spent filter backwash water, sludge thickener
supernatant, or liquids from dewatering processes. The rule requires that these recycled liquids be
returned to a location such that all steps of a system's conventional or direct filtration are employed. The
rule also requires systems to notify the State that they practice recycling. Finally, systems must collect
and maintain information for review by the State.
2.5.10 2002 Long Term 1 Enhanced Surface Water Treatment Rule
The LT1ESWTR (67 FR 1812 January 2002) is an extension of the 1998 lESWTRto small
systems. LT1ESWTR extends control of Cryptosporidium and other disease-causing microbes to surface
water and GWUDI systems that serve fewer than 10,000 people. Key provisions in the LT1ESWTR are
very similar to those for the IESWTR, but provide additional flexibility for small systems.
2.5.11 2003 Proposed Stage 2 Disinfectants and Disinfection Byproducts Rule
The requirements of the Stage 2 DBPR apply to all CWSs and NTNCWSs that add a disinfectant
other than UV or that deliver water that has been treated with a disinfectant other than UV. The Stage 2
DBPR builds on the 1979 Total Trihalomethane Rule and the 1998 Stage 1 DBPR by requiring reduced
levels of DBFs in distribution systems. Each rule activity for the Preferred Regulatory Alternative and
the associated rule schedule are described below.
The Stage 2 DBPR is designed to reduce DBP occurrence peaks in the distribution system by
changing compliance monitoring requirements. Compliance monitoring will be preceded by an initial
distribution system evaluation (IDSE) to identify distribution system locations that represent high total
trihalomethane (TTHM) and haloacetic acids (F£AA5) levels. Systems may perform an IDSE by
completing either a system-specific study (SSS) or a standard monitoring program (SMP). NTNCWSs
serving fewer than 10,000 people are not required to conduct an IDSE, and other systems may receive
waivers from the IDSE requirement.
The Stage 2 DBPR changes the way sampling results are averaged to determine compliance. The
compliance determination for the Stage 2 DBPR is based on a locational running annual average (LRAA)
instead of the system-wide RAA used under the Stage 1 DBPR. LRAAs are RAAs calculated separately
for each sample location in the distribution system. With the Stage 2 LRAA requirement, the TTHM and
HAAS MCLs must be met at each monitoring location, while the Stage 1 RAA requires a system to
average results over all monitoring locations.
2.6 Economic Rationale for Regulation
This section addresses the economic rationale for choosing a regulatory approach. Such a
rationale is required by Executive Order Number 12866, Regulatory Planning and Review (The White
House 1993), which states the following:
Economic Analysis for the LT2ESWTR 2-11 December 2005
-------
...[E]ach agency shall identify the problem that it intends to address (including, where
applicable, the failures of the private markets or public institutions that warrant new
agency action) as well as assess the significance of that problem. (Section 1, b(l))
In addition, Office of Management and Budget (OMB) guidance dated January 11, 1996, states that "in
order to establish the need for the proposed action, the analysis should discuss whether the problem
constitutes a significant market failure" (USEPA 1996b).
In a perfectly competitive market, prices and quantities are determined solely by the aggregated
decisions of buyers and sellers. Such a market occurs when many producers of a product are selling to
many buyers, and both producers and consumers have perfect information on the characteristics and
prices of each firm's products. Barriers to entry in the industry cannot exist, and individual buyers and
sellers must be "price takers": i.e., their decisions cannot affect the price. Several properties of the public
water supply do not satisfy the conditions for a perfectly competitive market and, thus, lead to market
failures that require regulation.
First, many water systems are natural monopolies. A natural monopoly exists when it is
impossible for more than one firm in each area to recover the costs of production and survive. There are
high fixed costs associated with reservoirs and wells, transmission and distribution systems, treatment
plants, and other facilities. For other potential suppliers to enter the market, they would have to provide
the same extensive infrastructure to realize similar economies of scale and be competitive. A splitting of
the market with increased fixed costs (for example, two supplier networks in a single market) usually
makes this situation unprofitable for one or both suppliers. The result is a market suitable for a single
supplier and one that is hostile to alternative suppliers. In such natural monopolies, suppliers have fewer
incentives for providing high quality service or maintaining competitive prices. In these situations,
governments often intervene to help protect the public interest.
Because PWSs are legal as well as natural monopolies, they are often subject to price controls, if
not outright public ownership. While customers may demand improvements in water quality, the
regulatory regime may not transmit that demand to the water supplier or allow the supplier to raise its
price to recover the cost of the improvements. If consumers do not believe that their drinking water is
safe enough, they cannot simply switch to another water utility. Other options for obtaining safe drinking
water (e.g., buying bottled water or installing point-of-use filtration) often cost consumers more than
water purchased from public water supplies. Therefore, the water supplier may have little incentive to
improve water quality.
Second, the public may not understand the health and safety issues associated with drinking water
quality. Understanding the health risks posed by trace quantities of drinking water contaminants involves
analysis and synthesis of complex toxicological and health sciences data. Therefore, the public may not
be aware of the risks it faces. For example, cases of waterborne disease are likely to be under-reported
since a significant portion may be endemic, making them more difficult to recognize. There is, therefore,
a lack of occurrence data and related cost information on endemic waterborne disease available to the
public. EPA has implemented a Consumer Confidence Report (CCR) Rule (USEPA 1998) that makes
water quality information more readily available to consumers. This rule requires CWSs to publish an
annual report on Local drinking water quality. Consumers, however, still have to analyze this
information for its health risk implications. Even if informed consumers are able to engage water systems
in a dialogue about health issues, the costs of such interaction (measured in personal time and monetary
outlays) present a significant impediment to consumer expression of risk reduction preferences.
Moreover, these reports typically contain no information about the risk associated with Cryptosporidium
and most other microbial pathogens, because PWSs are not required to analyze for them.
Economic Analysis for the LT2ESWTR 2-12 December 2005
-------
SDWA regulations are intended to provide a level of protection from exposure to drinking water
contaminants by setting minimum performance requirements. These regulations are intended neither to
restructure market mechanisms nor to establish competition in supply; rather, they establish the level of
service to be provided that best reflects public preference for safety. The Federal regulations reduce the
high information and transaction costs by acting on behalf of consumers in balancing risk reduction and
the social costs of achieving this risk reduction.
Economic Analysis for the LT2ESWTR 2-13 December 2005
-------
3. Consideration of Regulatory Alternatives
3.1 Introduction
The U.S. Environmental Protection Agency (EPA or the Agency) evaluated a number of
regulatory alternatives that could mitigate the health concerns addressed by the Long Term 2 Enhanced
Surface Water Treatment Rule (LT2ESWTR). These evaluations took place during a regulatory
negotiation process that began in the Spring of 1999, and included consultation with the Stage 2
Microbial and Disinfection Byproducts (Stage 2 M-DBP) Advisory Committee that was convened under
the Federal Advisory Committees Act (FACA). This chapter summarizes the alternatives considered and
develops a context for the regulatory approach taken. The remainder of the chapter is organized as
follows:
3.2 Development Process for Regulatory Alternatives
3.3 Specific Regulatory Alternatives Considered in this EA
3.3.1 Summary of Bin Classification and Treatment Requirements for Regulatory
Alternatives
3.3.2 Additional Treatment for Direct Filtration Systems
3.4 Alternative Monitoring Approaches Considered
3.4.1 Indicators of Microbial Contamination
3.4.2 Cryptosporidium Monitoring Strategies for Bin Classification
3.2 Development Process for Regulatory Alternatives
Two efforts in the regulatory development process for the LT2ESWTR are particularly relevant
to evaluation of alternatives discussed in the Economic Analysis (EA): (1) the data synthesis and analysis
resulting from the Information Collection Rule (ICR) and ICR Supplementary Surveys (ICRSSs), and (2)
the deliberations and recommendations of the Stage 2 M-DBP Advisory Committee.1 EPA held 14
formal negotiation meetings of the Stage 2 M-DBP Advisory Committee between March 1999 and
September 2000. Before convening the committee, EPA also held three preparatory stakeholder meetings
on pathogen and disinfection byproduct (DBP) health effects, occurrence, and treatment. The objective of
the committee meetings was to reach a consensus regarding recommended provisions for the two rules
(Stage 2 Disinfectants and Disinfection Byproducts Rule (DBPR) and LT2ESWTR).
Technical support for the Stage 2 M-DBP negotiation meetings was provided by the Technical
Work Group (TWG), which the committee established at its first meeting and comprised EPA and
drinking water industry experts. The TWG's activities resulted in the collection, development,
evaluation, and presentation of key data related to the LT2ESWTR, including new data on pathogenicity,
occurrence, and treatment of microbial contaminants, specifically Cryptosporidium.
The ICR database provided much of the information evaluated for the LT2ESWTR. EPA
promulgated the ICR in 1996 pursuant to the Safe Drinking Water Act (SDWA) requirements. The ICR
required approximately 300 large public water systems (PWSs) with approximately 500 separate water
treatment plants to conduct 18 months of sampling for water quality and treatment parameters related to
1 The Stage 2 M-DBP Advisory Committee comprised representatives from a variety of stakeholder
organizations. A complete list of participating members (as well as a summary of committee findings) is included in
the docket for the LT2ESWTR.
Economic Analysis for the LT2ESWTR 3-1 December 2005
-------
DBF formation and the occurrence of microbial pathogens. After the ICR data collection, EPA obtained
additional data on pathogen occurrence through the ICRSSs. These involved 127 water treatment
systems, including 40 small systems. Large and medium systems collected semi-monthly samples for
Cryptosporidium, Giardia, and other water quality parameters for 1 year. Small systems (those serving
fewer than 10,000 people) collected monthly water quality data, but did not sample for protozoa.
EPA, in consultation with nationally recognized experts in statistics, evaluated ICR and ICRSS
data to generate estimates of national occurrence of Cryptosporidium in surface water and finished water.
These data were evaluated under various regulatory scenarios to estimate the reduction of
Cryptosporidium reaching consumers.
3.3 Specific Regulatory Alternatives Considered in this EA
The recommendations of the Advisory Committee are described in a document called the
Agreement in Principle (USEPA 2000e). The Advisory Committee reached consensus on the issues of
uncovered finished water reservoirs and treatment of unfiltered water without formally identifying
regulatory alternatives other than the proposed approaches. Consequently, no formal alternatives were
presented for these requirements. The committee's recommendations to address these issues are reflected
in the LT2ESWTR. For control of Cryptosporidium in filtered systems, however, several alternatives
were considered. The committee discussed, but quickly found impractical, alternatives based on
monitoring for Cryptosporidium in finished water. The occurrence of Cryptosporidium in finished water
is so low that the volume of water required for analysis would make monitoring costs prohibitive. Thus,
all the alternatives based on monitoring directed that monitoring be performed on the source water.
The following subsections detail the differences between the alternatives and explain the rationale
behind EPA's selection of the Preferred Alternative. Section 3.3.2 also describes the additional treatment
requirements proposed for direct filtration systems.
3.3.1 Summary of Bin Classification and Treatment Requirements for Regulatory Alternatives
In considering different approaches for filtered systems under the LT2ESWTR, the M-DBP
Advisory Committee focused on four regulatory alternatives (hereafter referred to as Alternatives Al
through A4). Alternative Al requires the same amount of reduction of Cryptosporidium for all systems,
while the other three base their treatment requirements on the amount of Cryptosporidium found in a
system's source water through monitoring. These measurements place a system in one of several bins,
ranging from "no action" to an additional 2.5 log Cryptosporidium treatment. Further, all alternatives
allow systems to select treatment technologies based on the amount of Cryptosporidium treatment needed
to meet requirements and the effectiveness of each technology.
In evaluating each binning scenario, the committee asked the following questions:
Do the treatment requirements adequately reduce Cryptosporidium concentrations in finished
water?
• How many systems would be required to add treatment?
• What is the likelihood of bin misclassification?
Economic Analysis for the LT2ESWTR 3-2 December 2005
-------
What are the chances that systems with high source water concentrations would be placed in
the bin requiring no action?
The predicted finished water Cryptosporidium concentrations and percentages of plants adding
treatment are shown for the Preferred Regulatory Alternative in Chapter 4. The likelihood of
classification in a certain bin for a given concentration and the predicted percentages of plants in each bin
for all the regulatory alternatives are shown in Appendix B.
Exhibit 3.1 summarizes binning and treatment scenarios for each regulatory alternative. These
alternatives were defined by two criteria: (1) bin boundaries, as defined by the results of Cryptosporidium
monitoring, and (2) the treatment scenarios (log reduction requirements for Cryptosporidium) required for
each bin. The alternatives were compared within the context of the economic analysis to assist EPA in
selecting a Preferred Regulatory Alternative.
On the basis of preliminary cost-benefit analyses, the Advisory Committee chose Alternative A3
as the Preferred Alternative. This EA continues to support Alternative A3 as the best choice and EPA is
promulgating Alternative A3 for this reason. Alternative A3 was shown to be the most cost effective and
to deliver the best value. Comparisons of the net benefits of each alternative are summarized in the
executive summary and described in more detail in Chapter 8.
Exhibit 3.1: Summary of Bin Requirements for Filtered Systems
Source Water Cryptosporidium
Monitoring Results (oocysts/L)
Additional Treatment Requirements
Alternative A1
2.0 log inactivation required for all systems
Alternative A2
<0.03
> 0.03 and < 0.1
> 0.1 and < 1.0
> 1.0
No action
0.5 log
1 .5 log
2.5 log
Alternative A3 - Preferred Alternative
< 0.075
> 0.075 and < 1.0
> 1.0 and < 3.0
> 3.0
No action
1 log
2 log
2.5 log
Alternative A4
<0.1
> 0.1 and < 1.0
>1.0
No action
0.5- log
1 .0 log
Note: "Additional treatment requirements" are for systems that have conventional treatment in
full compliance with existing rules (the IESWTR and LT1ESWTR).
Economic Analysis for the LT2ESWTR
3-3
December 2005
-------
3.3.2 Additional Treatment for Direct Filtration Systems
The Agreement in Principle (USEPA 2000e) recommended that EPA address direct filtration
systems in connection with Bins 2-4 of Alternative A3 in the LT2ESWTR. Direct filtration plants lack
sedimentation basins; their treatment processes move directly from addition of coagulant and mixing to
filtration. Conventional filtration plants use coagulation, sedimentation, and filtration. Sedimentation
reduces the Cryptosporidium load on the filters and helps plants respond to sudden changes in influent
water quality.
EPA considered the effectiveness of direct filtration in removing Cryptosporidium when
determining how to apply the Advisory Committee's treatment technique recommendations for
conventional plants to direct filtration plants. EPA has consistently recognized the value of employing
multiple barriers for pathogen removal to provide redundancy and reliability. Studies have shown that a
well-operated sedimentation basin can reduce Cryptosporidium levels by 0.5 log or more (Dugan et al.
1999; Edzwald and Kelley 1998; and Patania et al. 1995). The SWTR Guidance Manual (USEPA 1991)
also supports giving less credit to direct filtration systems; these systems are eligible for 0.5 log less credit
for Giardia than conventional filtration systems.
Based on these studies, EPA's prior consideration of the effectiveness of direct filtration systems
for Giardia, and the Agency's confidence in the multiple barrier approach, EPA concluded that direct
filtration plants should provide an additional 0.5 log treatment beyond that required for conventional
treatment plants. A more detailed discussion of Cryptosporidium removal by conventional and direct
filtration can be found in the EPA document, Occurrence and Exposure Assessment for the Long Term 2
Enhanced Surface Water Treatment Rule (USEPA 2003c).
3.4 Alternative Monitoring Approaches Considered
EPA considered a variety of monitoring approaches while developing LT2ESWTR regulatory
alternatives. These include evaluating other water quality parameters as surrogates for Cryptosporidium
and alternative monitoring strategies to minimize monitoring costs, especially for small drinking water
systems. These issues are described in the subsections below.
3.4.1 Indicators of Microbial Contamination
Due to the cost associated with Cryptosporidium monitoring, the Stage 2 M-DBP Advisory
Committee evaluated alternative source water quality parameters to determine if they could be used to
identify water sources with high Cryptosporidium levels. The committee assessed the 12-month means of
monitoring data for turbidity, total organic carbon (TOC), E. coll, fecal coliform, and total coliform
bacteria as surrogates for Cryptosporidium. Specifically, the committee evaluated whether these potential
surrogates could accurately assign plants to LT2ESWTR microbial framework bins. None of these
parameters correlated well with Cryptosporidium levels at all concentrations. Evidence indicated that E.
coli would be somewhat effective in identifying plants with Cryptosporidium levels above or below 0.075
oocysts/L in reservoirs, lakes, and flowing streams (USEPA 2003c). Under the Preferred Alternative, this
level is the cutoff between the no-action and the action bins (Bins 2, 3, and 4). Thus, the committee
recommended that E. coli be used as a screening test for small systems under the Preferred Alternative,
A3.
The selection of the E. coli level for determining when additional Cryptosporidium monitoring in
small systems should be conducted was based on limited data. Therefore, the Advisory Committee
Economic Analysis for the LT2ESWTR 3-4 December 2005
-------
agreed that additional data should be collected to evaluate the E. coli indicator criteria and to develop
alternative criteria, if appropriate. Accordingly, the Advisory Committee agreed that large systems would
measure E. coli and turbidity in their source water when they sample for Cryptosporidium. The
composite data will be submitted to EPA. This will give EPA time to develop possible alternative
indicator levels or indicator parameters (e.g., turbidity in combination with E. coli) prior to the date when
small systems are required to begin source water monitoring for E. coli. Following the completion of 1
year of monitoring under the LT2ESWTR, EPA will determine if alternative indicators (to the E. coli
levels prescribed in the rule) are appropriate for determining classification into bins. Depending upon its
findings, EPA will issue guidance for States to consider prescribing alternative indicator requirements for
small systems. Therefore, the LT2ESWTR allows for alternative indicators to be considered by the
Primacy Agency.
The use ofE. coli as a screen for Cryptosporidium monitoring is applicable only to Alternatives
A3 and A4. Using E. coli levels to predict a mean Cryptosporidium concentration of less than 0.03
oocysts per liter—the action cutoff level for Alternative A2—was less reliable (USEPA 2003c). Thus,
under Alternative A2, no screening test is available, and all small systems must monitor for
Cryptosporidium. Since Alternative Al requires the same level of treatment for all systems, no
monitoring provisions for large or small systems are included for Alternative A1.
3.4.2 Cryptosporidium Monitoring Strategies for Bin Classification
EPA and the Advisory Committee also evaluated alternative monitoring strategies to ensure that
levels of source water contamination would be adequately characterized, while minimizing the
monitoring burden. Approaches considered included taking 24, 48, or 72 source water samples to
determine bin classification using the bin boundaries in Alternative A3.
EPA chose to allow systems serving at least 10,000 people to collect and analyze 24 monthly
samples over a 2-year period and base the bin assignment on the maximum running annual average
(RAA). (The first RAA will be the average of the results of the first 12 months of monitoring; the second
RAA will be the average of results from months 2-13; the third will be the average of concentrations
from months 3-14; and so forth.) Alternatively, systems may collect two or more samples per month
over the 2-year period, at regular intervals, and use the simple average (the average of all 48 samples) to
determine bin placement. The following paragraphs discuss the methodology for choosing these
monitoring frequencies.
EPA knew that the measured amount of Cryptosporidium in each sample might be different from
the actual or "true" concentration because of error inherent in sampling and analytical methods. For
example, method error can be introduced by inefficiencies in oocyst recovery, false detections, and
analyst error. Sampling error is affected by the sample size and the fact that the concentration in a given
sample may misrepresent the concentration in the larger water body. EPA was primarily concerned that
high-occurrence sources could possibly be placed in either the no-action bin (Bin 1, for mean occurrence
below 0.075 oocysts/L) or bins that provided insufficient remedies. This could result in insufficient
protection of public health. A secondary concern was that systems could be assigned to a higher bin than
was warranted by their true concentration, resulting in unnecessary costs for systems.
EPA performed a Monte Carlo analysis to determine the probabilities of misclassification based
on different monitoring scenarios (see Appendix B for details). The analysis accounted for the volume
assayed, variation in source water Cryptosporidium occurrence, and variable method recovery.
Economic Analysis for the LT2ESWTR 3-5 December 2005
-------
The analysis specifically considered the likelihood that a system with a true mean
Cryptosporidium concentration a factor of 3.16 (0.5 log) above or below a bin boundary would be
assigned to the wrong bin. Probabilities were assessed for two cases:
Misclassification low: a system with a mean concentration of 0.24 oocysts/L (i.e., factor of
3.16 above the Bin 1 boundary of 0.075 oocysts/L) is misclassified low in Bin 1.
• Misclassification high: a system with a mean concentration of 0.024 oocysts/L (i.e., factor of
3.16 below the Bin 1 boundary of 0.075 oocysts/L) is misclassified high in Bin 2.
Exhibit 3.2 shows the error rates the model predicts at concentrations of 0.24 oocysts/L and 0.024
oocysts/L (afactor of 3.16 above and below the Bin 1 boundary of 0.075 oocysts/L) under different
monitoring scenarios.
Exhibit 3.2: Probability of Misclassification for Monitoring and Binning Strategies
Considered for the LT2ESWTR
Monitoring Strategy
48-sample simple mean
24-sample maximum RAA
24-sample simple mean
12-sample second highest value
8-sample maximum value
Probability of
Misclassification High
1 .7%
5.3%
2.8%
47%
66%
Probability of
Misclassification Low
1 .4%
1 .7%
6.2%
1.1%
1 .0%
Note: Probability of misclassification high into Bin 2 was calculated for systems with true Cryptosporidium
concentrations of 0.024 oocysts/L, or 0.5 log below the Bin 1 boundary of 0.075 oocysts/L. Probability of
misclassification low into Bin 1 was calculated for systems with Cryptosporidium concentrations of 0.24 oocysts/L, or
0.5 log above the Bin 1 boundary.
Source: Appendix B.
The first two approaches shown in Exhibit 3.2—the 48-sample simple mean and 24-sample
maximum RAA—were recommended by the Advisory Committee and are proposed for bin classification
under the LT2ESWTR because they have low misclassification rates. As shown in Exhibit 3.2, these
strategies have misclassification low rates of 1 to 2 percent, meaning there is a 98 to 99 percent likelihood
that a plant with an oocyst concentration 0.5 log above the Bin 1 boundary would be correctly assigned to
Bin 2. The misclassification high rate is near 2 percent for the 48-sample simple mean and 5 percent for
the 24-sample maximum RAA. These rates indicate that a plant with an oocyst concentration 0.5 log
below the Bin 1 boundary would have a 95 to 98 percent probability of being correctly assigned to Bin 1.
Bin misclassification rates across a wide range of concentrations are shown in Appendix B.
The 24-sample simple mean had a slightly lower misclassification high rate than the 24-sample
maximum RAA (2.8 vs. 5.3 percent) but the misclassification low rate of the simple mean was almost 4
times greater. Consequently, a plant with a mean Cryptosporidium level above the Bin 1 boundary would
be much more likely to be misclassified in Bin 1 using a 24-sample simple mean than with a 24-sample
maximum RAA. To increase the probability that systems with mean Cryptosporidium concentrations
above 0.075 oocysts/L will provide additional treatment, EPA is proposing that if only 24 samples are
taken, the maximum RAA be used to determine bin assignment.
Economic Analysis for the LT2ESWTR
3-6
December 2005
-------
EPA evaluated monitoring strategies involving only 12 and 8 samples to determine if less
frequent monitoring could provide satisfactory bin classification; these lower numbers of samples are not
adequate. For example, Exhibit 3.2 shows that if plants were classified in bins based on the second
highest concentration of 12 samples or the highest concentration of eight samples, then low
misclassification low rates could be achieved. A system with a mean Cryptosporidium level 0.5 log
above the Bin 1 boundary would have a 99 percent chance of being appropriately classified in a bin
requiring additional treatment under either strategy. However, a system with a mean oocyst concentration
0.5 log below the Bin 1 boundary would have a 47 percent chance of being incorrectly classified in Bin 2
using the second highest result among 12 samples, or a 66 percent likelihood of being misclassified in Bin
2 using the maximum result among 8 samples. Therefore, these strategies were not proposed.
Increasing the number of samples used to compute the maximum RAA above 24 also increased
the number of annual averages computed, so it did not reduce the likelihood of misclassification high.
Computing a simple mean based on more than 48 samples did reduce bin misclassification rates, but the
rates were already very small (1 to 2 percent for plants with levels 0.5 log above or below bin
boundaries). For sources with Cryptosporidium concentrations very near or at bin boundaries, increasing
the number of samples did not markedly improve the error rates, which remained near 50 percent at the
bin boundaries.
In summary, EPA believes that the prescribed sampling designs in the Preferred Alternative
perform well for the purpose of accurately assigning source waters to bins. More costly designs,
involving more frequent sampling and analysis, provide only marginally improved performance, while
placing a greater burden on limited laboratory capacity.
Economic Analysis for the LT2ESWTR 3-7 December 2005
-------
4. Baseline Conditions
4.1 Introduction
To estimate the impact of the LT2ESWTR regulatory alternatives on the water supply industry, it
is first necessary to establish the conditions that exist within the industry just before the regulatory
requirements take effect. This baseline allows a consistent comparison of public health impacts
(developed in Chapter 5) and the economic and financial impacts (developed in Chapters 6 and 7) across
regulatory alternatives.
Because the compliance deadlines of recently promulgated and proposed rules will occur after the
date EPA is required to promulgate the LT2ESWTR, many of the baseline conditions must be estimated
rather than directly measured. Thus, data on existing conditions are combined with projections of
changes to those conditions to estimate the baseline for the LT2ESWTR. The steps required to determine
the baseline conditions are as follows:
Compile an industry profile—identify and collect information on the segments of the water
supply industry subject to the rule.
• Characterize influent water quality—summarize the relevant characteristics of the raw water
treated by the industry.
Characterize treatment for other rules—predict what the industry will do to comply with the
provisions of rules that may precede the LT2ESWTR and that may generate changes relevant
to this rule, specifically the IESWTR, LT1ESWTR, Stage 1 DBPR, and Stage 2 DBPR.
Predict occurrence following implementation of other rules—estimate what the treated water
quality will be after the rules preceding the LT2ESWTR are implemented.
This chapter presents an analysis that is at a level of detail and precision appropriate to support
subsequent analyses and regulatory decisions for the LT2ESWTR. An exhaustive review of the water
supply industry, source waters, or industry practices was not needed to conduct the analysis. The
remainder of this chapter is organized as follows:
4.2 Data, Tools, and Processes Used in Baseline Development
4.2.1 ICR and ICRSS Observed Data
4.2.2 ICR and ICRSS Modeled Data and Method for Predicting Source Water
Occurrence
4.2.3 Surface Water Analytical Tool (SWAT)
4.3 Industry Profile
4.3.1 Public Water System Characterization
4.3.2 Systems, Plants, and Population Subject to the LT2ESWTR
4.3.3 Water Treatment Plant Design and Average Daily Flows
4.4 Baseline for Unfiltered Plants (Pre-LT2ESWTR)
4.4.1 Treatment Characterization for Unfiltered Plants
4.4.2 Number of Unfiltered Systems, Plants, and Population Served
4.4.3 Source Water Cryptosporidium Occurrence for Unfiltered Plants
4.4.4 Finished Water Cryptosporidium Occurrence for Unfiltered Plants
4.5 Baselines for Filtered Plants (Pre-LT2ESWTR)
4.5.1 Treatment Characterization for Filtered Plants
4.5.2 Number of Filtered Plants and Population Served
Economic Analysis for the LT2ESWTR 4-1 December 2005
-------
4.5.3 Source Water Cryptosporidium Occurrence for Filtered Plants
4.5.4 Finished Water Cryptosporidium Occurrence for Filtered Plants
4.5.5 Comparison of EPA Estimates with Aboytes et al. (2000)
4.5.6 Predicted Bin Classification for Filtered Plants
4.6 Baseline for Uncovered Finished Water Reservoirs
4.7 Households Incurring Costs Due to the LT2ESWTR
4.8 Summary of Uncertainties in Development of LT2ESWTR Baselines
4.2 Data, Tools, and Processes Used in Baseline Development
Several data sources were used to characterize the baseline and to predict changes in treatment
technologies and water quality for different regulatory alternatives. The Safe Drinking Water Information
System-Federal Version (SDWIS1) (4th Quarter Freeze Year 2003 data2) is used to create system and
population baselines (USEPA 2003e). SDWIS is EPA's national regulatory compliance database for the
drinking water program. It includes information on the nation's 170,000 public water systems (PWSs)
and on violations of drinking water regulations. EPA's Web site provides more information on SDWIS
(http://www.epa.gov/safewater/sdwisfed/sdwis.htm).
EPA also used Geometries and Characteristics of Water Systems (also called the Model Systems
Report) (USEPA 2000a). In this document, EPA used 1995 Community Water System Survey (CWSS)
data to develop equations to predict flows based on system populations. The 1995 CWSS was a mail
survey covering over 3,000 surface and ground water systems, to which 1,980 systems responded. The
data gathered included treatment practices, water demand, and financial information. See Community
Water System Survey (USEPA 1997c), for more information.
To characterize the influent water quality, treatment processes, and finished water quality, EPA
primarily used data from the 1996 Information Collection Rule (ICR), for which Cryptosporidium
monitoring requirements applied to all PWSs serving at least 100,000 people and using surface water or
ground water under the direct influence of surface water (GWUDI) as a source. The purpose of the ICR
was to collect DBP and microbial occurrence and treatment information to help evaluate the need for
further microbial and disinfection byproduct (DBP) rules. The ICR gathered plant-level data from about
300 water systems over 18 months (July 1997-December 1998). These data characterize the source water
occurrence of Cryptosporidium, Giardia, viruses, and indicators of microbial contamination, along with
types of treatment in place. The data used in this economic analysis (EA) are from the Auxiliary 1
(AUX1) database (USEPA 2000h).
The Information Collection Rule Supplemental Surveys were voluntary surveys, for which 40
medium (serving 10,000-99,999) and 40 large plants (serving 100,000 or more) collected
Cryptosporidium and Giardia source water occurrence data. These data are presented in the Occurrence
and Exposure Assessment for the Long Term 2 Enhanced Surface Water Treatment Rule (USEPA 2003c).
Treatment characterization and other information were obtained from industry and State sources.
Several analytical tools (models) also were used to estimate the following:
• Source water Cryptosporidium occurrence;
Throughout this document, the acronym "SDWIS" represents "SDWIS-Federal Version."
2 All industry baseline data reflect revisions to SDWIS 4th Quarter Year 2000 Freeze to account for
reporting errors in Massachusetts and Montana.
Economic Analysis for the LT2ESWTR 4-2 December 2005
-------
Treatment changes due to the Stage 1 and Stage 2 DBPRs;
Number of plants falling into various treatment bins, based on predicted source water
occurrence; and
Occurrence of Cryptosporidium in finished water under Pre-LT2ESWTR and LT2ESWTR
conditions.
4.2.1 ICR and ICRSS Observed Data
Three sets of monitoring data are used to characterize Cryptosporidium source water occurrence:
the ICR data set and the ICRSS data sets for medium and large systems. Microbial analyses for the ICR
were conducted according to the ICR Method (USEPA 1996a). The ICRSSs evaluated source water for
Cryptosporidium, Giardia, and coliforms at a sample of medium surface water systems (those serving
10,000 to 99,999 people) as well as a sample of large surface water systems. EPA Methods 1622
(USEPA 1999a) and 16233 (USEPA 1999b), summarized in the Occurrence and Exposure Assessment,
were used for Cryptosporidium and Giardia analyses. With Methods 1622 and 1623, the volume of water
analyzed was on the average larger than the volume analyzed with the ICR Method, yielding better
estimates of Cryptosporidium occurrence on a per-sample basis (the volume analyzed with the ICR
Method depended on the sample pellet volume after centrifugation and was based on the volume needed
to meet detection limits). The ICRSS data consist of semi-monthly observations taken over a 12-month
period at 40 randomly selected large plants and 40 randomly selected medium plants4. Exhibit 4.1
summarizes the differences between the ICR and ICRSS data collection methods.
3 Method 1622 was used for the first 4 months of data collection, at which time Method 1623 replaced
Method 1622. The primary difference between Method 1622 and Method 1623 is that the Method 1622
immunomagnetic separation (IMS) kit includes only reagents for Cryptosporidium purification, whereas Method
1623 uses the Giardia/Cryptosporidium-combmation kit, which includes reagents for both Cryptosporidium and
Giardia purification.
4 Forty small plants also were included in the survey, but they did not monitor protozoa concentrations.
Economic Analysis for the LT2ESWTR 4-3 December 2005
-------
Exhibit 4.1: Comparison of ICR and ICRSS Data Collection Methods
System size of plants participating
(population served)
Number of surface water plants
participating
Sample frequency
Sampling period
Required sample volume
Median sample volume analyzed
Average recovery rates for lab
method
Percentage of samples where
Cryptosporidium was not detected
Percentage of plants with at least one
positive Cryptosporidium sample
ICR
>100,000
350
Monthly
July 1997-
December1998
100 L
3.2 L
12%
93%
44%
ICRSSM
10,000-99,999
40
Semi-monthly
March 1999-
February 2000
10L
10L
ICRSSL
>1 00,000
40
Semi-monthly
March 1999-
February 2000
10L
10L
43%
83%
85%
87%
85%
Source: USEPA2003c, USEPA2000h, and USEPA 2000L
The ICR monitoring program resulted in nearly 6,000 Cryptosporidium measurements from 350
water sources (USEPA 2000h). The ICRSS monitoring produced approximately 1,900 measurements
from 80 source waters. Cryptosporidium was not detected in most of the samples (93 percent of ICR
samples and 86 percent of ICRSS samples). Approximately 44 percent of plants participating in the ICR
program had at least one positive sample, while the increased sensitivity of the methods under the ICRSS
led to a much higher percentage of plants (approximately 85 percent) having at least one positive sample.
The detection of Cryptosporidium oocysts is complex. Because of the low occurrence of
Cryptosporidium in source waters, a sample may not contain any oocysts even though the source water
does. Thus, a non-detection in a test volume is not definitive evidence against occurrence in the source.
In addition, the laboratory method is inefficient and may not recover all the oocysts that were in a sample.
While underestimation is much more likely, when detections do occur, sample concentrations also may
overestimate influent concentrations because of the small volume of sample involved (i.e., one oocyst
identified in a 10-liter sample may not represent the true proportion of oocysts in the much larger source
water volume).
During the ICR collection period, EPA implemented the ICR Laboratory Spiking Program (LSP)
to assess the recovery of Cryptosporidium oocysts from field samples analyzed with the ICR Method. At
the time of the ICR sample collection, duplicate 100-liter samples were collected and spiked with a
known quantity of Cryptosporidium oocysts. Recovery of the oocysts (i.e., the detection of known
oocysts) by laboratories was very low, with an average of 12 percent of the known quantity being
recovered per sample. The ICRSS Matrix Spike Program was used to assess the recovery of oocysts from
field samples using Methods 1622/1623 during the ICRSS. Duplicate samples were spiked with a known
quantity of Cryptosporidium oocysts, filtered, and analyzed using Methods 1622/1623. The average
recovery for the ICRSS was 43 percent. These spiking programs are further described in the LT2ESWTR
Occurrence and Exposure Assessment (USEPA 2003c).
Economic Analysis for the LT2ESWTR
4-4
December 2005
-------
Each set of data has its advantages; for instance, the ICR data set contains data from more plants
than the ICRSS data sets do, but the ICR data set does not include data for plants in medium-sized
systems. The ICRSS data set does not include enough data for unfiltered plants to be useful for modeling,
while the ICR set does. The ICRSS sets, on average, had higher recovery rates and larger sample
volumes than the ICR set. The ICR collected data for a longer time period, but the ICRSS data were
collected more frequently. In considering which data set best represents the national distribution of
Cryptosporidium occurrence, none was judged superior to the others; that is, no one set was considered to
have a greater likelihood of representing "true" occurrence. In view of this, each data set was kept
separate, and a weighting of their relative values to allow them to be combined was not attempted.
Cryptosporidium observations were characterized according to oocyst structure as observed under
a microscope, as defined below:
Empty: Oocyst-type walls, but not containing internal material.
• Non-Empty: Includes oocysts with internal structure and with amorphous structures.
Oocysts with amorphous structures have walls and internal material characteristic of
Cryptosporidium, but the material cannot be confirmed as Cryptosporidium.
• With Internal Structures: Oocysts that have a cell wall and recognizable internal
structures consistent with Cryptosporidium; subset of non-empty category.
With Amorphous Structures: Oocysts that have a cell wall and internal material but no
recognizable internal structures; subset of non-empty category.
Total: The combined count of empty oocysts and non-empty oocysts (those with either
internal or amorphous structures).
At meetings of the Technical Working Group of the Microbial-Disinfection Byproducts Advisory
Committee, participants agreed that the type of oocyst observed gives information about the level of
confidence that the oocyst is actually Cryptosporidium. The presence of internal structures may increase
the confidence that the observed object is indeed a Cryptosporidium oocyst, and not some other item or
organism with similar gross morphology. While oocysts that are empty are unlikely to be viable or
infectious at the time of the laboratory analysis, they are still indicators of the possible presence of
Cryptosporidium. Oocysts with amorphous structures give a level of confidence between that of empty
oocysts and those with internal structures. A detailed presentation of observed Cryptosporidium
occurrence and evaluation of results from the ICR and ICRSS is provided in the Occurrence and Exposure
Assessment (USEPA 2003c).
The analysis presented in this document assumes that total Cryptosporidium counts are the most
representative of the presence of Cryptosporidium in source waters. While some of these oocysts may not
have been infectious at the time of analysis, they likely represent Cryptosporidium that was present in the
water. The probability of oocyst infectivity is addressed in the risk assessment model in section 5.2.3.
To account for the limitations in observed data in the three data sets, modeled estimates of the
range of underlying total "true" Cryptosporidium occurrence, consistent with the observed occurrence,
were developed.
4.2.2 ICR and ICRSS Modeled Data and Method for Estimating Source Water Occurrence
An accurate representation of Cryptosporidium concentrations in source water is important for
estimating both costs and benefits. The LT2ESWTR treatment requirements are based on the results of
Economic Analysis for the LT2ESWTR 4-5 December 2005
-------
source water Cryptosporidium monitoring. Consequently, this EA predicts the number of plants requiring
treatment, as well as level of treatment, based on estimates of Cryptosporidium occurrence in the source
water. The cost associated with increased levels of treatment and estimated reduction in illness of the
affected population is also derived from those same occurrence estimates. This section provides the
rationale for using estimates modeled from observed data and summarizes the modeled results.
The ICR and ICRSS provide source water quality data for developing this estimated national
occurrence distribution. The raw data include the number of oocysts detected and the associated sample
volume analyzed. A straightforward approach to modeling is first to divide the counts by the volumes to
obtain concentrations and then to model the distribution of estimated concentrations. However, there are
limitations in the data that negate the usefulness of this approach.
First, sample volumes are low relative to the volume needed to calculate representative
concentrations of Cryptosporidium in source water consistently. The volumes collected also are
inconsistent from location to location or even month to month at a given location. For example, the
majority of sample counts in the ICR and ICRSS were zeros (no Cryptosporidium detected), but these
zero counts are based on widely varying sample volumes. Common sense suggests that a zero count from
a 10-liter sample should be weighted more heavily than a zero count from a 1-liter sample. Based on a
straightforward sample concentration calculation, however, both concentrations would be considered the
same.
Second, the majority of Cryptosporidium oocysts captured in samples likely were not detected in
testing, due to a lack of precision in analytical methods. Therefore, straightforward concentration
estimates would systematically under-estimate the true concentration of Cryptosporidium in national
source waters.
Finally, there are limitations in the counts or concentrations that can be reported due to the rare
occurrence of oocysts. Oocyst counts can only be whole numbers—it is impossible to detect half an
oocyst—and most counts were zeros or ones. Since the volume analyzed in each sample was generally
small, a limited number of concentrations could be calculated from the count and volume data. For
instance, one oocyst in 10 liters gives a concentration of 0.1 oocysts/L, while one oocyst in 5 liters gives a
concentration of 0.2 oocysts/L. The volume analyzed would have to be quite large or the number of
oocysts present greater to enable more precise calculations of concentration.
These limitations are examples of uncertainty, or lack of knowledge, about the true
Cryptosporidium concentrations. When the observed data are used to calculate concentrations, the data
limitations result in a large number of individual sample values that may each over- or under-represent the
true Cryptosporidium concentrations. To account for these limitations and other sources of variability in
the data and to be able to estimate national occurrence, model-based occurrence estimates are chosen over
observed data5. A more detailed discussion of this estimation procedure is provided in the Occurrence
and Exposure Assessment (USEPA 2003c).
The benefits of using model-based estimates are that a model can properly account for the
following conditions:
• Variability in the data, based on location, sampling technique, turbidity dependence, and
other factors;
5 Variability refers to observed differences attributable to true heterogeneity or diversity in a population or
parameter, as opposed to uncertainty, which refers to lack of accuracy or precision in the measurement method.
Economic Analysis for the LT2ESWTR 4-6 December 2005
-------
The low and variable recovery rates of the measurement methods;
The small volumes assayed and their adequacy in representing the much larger volume of
source water at the time of sampling; and
• The small number of samples assayed at each location and their ability to represent the
average concentration in that location's water during the 18 months of the survey.
EPA developed a probability model that links the survey data (sample volumes and laboratory
oocyst counts) to the unknown source water concentrations, which are the quantities of interest. The
model accomplishes two tasks. First, it adjusts concentration estimates to account for varying sample
volumes and test method recovery rates. Second, the assumed probability structure makes it possible to
quantify uncertainty in the concentration estimates.
To account for varying sample volumes and recovery rates, the model defines an expected count
for each survey sample. This is the average number of oocysts detected in repeated sampling from a
given survey location, assuming a particular source water concentration, sample volume, and test method
recovery rate:
Average Count = Concentration x Volume x Recovery
or, in terms of units:
oocysts detected = (oocysts present/liter) x (liters) x (oocysts detected/oocysts present)
Concentration is the unknown, and estimating it from data is the goal of the modeling. Count and
Volume are known. They are measured directly and reported for each survey sample. Recovery is the
ratio of oocysts detected in laboratory testing to oocysts present in the sample. Since Recovery cannot be
measured directly for an individual test, there are no sample-by-sample data available for it. For a
particular test method, however, the typical range of recovery values can be estimated from designed
experiments using "spiked" samples with known concentrations (see Chapter 3). This was done for each
of the Cryptosporidium lab methods used in the ICR and ICRSS, and recovery rate probability
distributions were fit to the results. Simulated Recovery values were drawn from these probability
distributions. Because Count, Volume, and Recovery are either known or simulated, fitting the data to the
above formula results in estimates for the unknown Concentration that account for the variation in these
other variables.
To estimate the uncertainty in the occurrence results derived from the above formula, we next
define a probability structure for the observed sample counts. Each observed count is assumed to come
from a Poisson probability distribution, a fundamental probability model for counting rare events (this is
the distribution that results from the fact that only zero or one oocyst is usually present in a sample). The
mean of each distribution is the average count as defined above. Incorporating these probability
distributions in the model allows for the calculation of uncertainty bounds for the concentration estimates.
A Markov Chain Monte Carlo (MCMC) approach was employed to fit this model to the data.
MCMC is an iterative technique for fitting statistical models to data. The Markov Chain is a sequence of
joint probability distributions that converges to a stable distribution for likely model parameter values.
Monte Carlo is a computational technique for solving intractable integrals through extensive, simulated
sampling.
The benefit and cost analyses use plant-mean estimates of source water concentrations. The
model not only predicts individual concentrations, but average concentrations for each plant over the time
Economic Analysis for the LT2ESWTR 4-7 December 2005
-------
period covered by the observed data. For each primary data set (ICR, ICRSSM, and ICRSSL) the model
generates multiple log-normal distributions of plant-mean estimates. (EPA believes the distribution of
plant-mean concentrations is log-normal; this is described in section 3.3.3.2 of the Occurrence and
Exposure Assessment.) Both the benefits and cost analyses draw samples from these distributions to
reflect both variability and uncertainty.
The results of the occurrence models employed in this EA are documented in sections 4.4.3 and
4.5.3 for unfiltered and filtered plants. As described above, there was no single occurrence distribution
that served as input but, instead, a collection of plausible occurrence distributions. Summary plots in
these sections, then, show both the mean occurrence distribution—which represents the "middle"
distribution—and confidence bounds that capture the range of occurrence distributions in a given
collection.
4.2.3 Surface Water Analytical Tool (SWAT)
SWAT uses source water and treatment data collected under the ICR to estimate the percentage of
large surface water plants that would require advanced technologies to meet Stage 1 and Stage 2 DBPR
limits for disinfection byproducts (DBFs). Several advanced technologies, namely chlorine dioxide,
ozone, ultraviolet light (UV) disinfection, and membrane processes, not only produce fewer byproducts
than chlorine, but also provide varying degrees of Cryptosporidium removal or inactivation. The
characteristics, costs, and effectiveness of these technologies are taken into account in developing the
baselines for this EA.
A more detailed description of SWAT and how it was used to predict changes in treatment
technologies under the Stage 2 DBPR is provided in the Economic Analysis for the Stage 2 Disinfectants
and Disinfection Byproducts Rule (USEPA 2003d). In addition, a detailed description of the SWAT
components and its operation can be found in the document Surface Water Analytical Tool (SWAT)
Version 1.1—Program Descriptions and Assumptions (USEPA 2000b).
4.3 Industry Profile
This section identifies the PWSs subject to the LT2ESWTR. Subsequent sections identify the
subset of systems subject to specific provisions of the rule (e.g., baselines for unfiltered and filtered
systems). The water system baseline characterizations presented here are key inputs to the cost and
benefit assessments described in this EA.
EPA's categorization scheme for water systems is summarized, followed by a presentation of
systems, plants, and populations subject to the LT2ESWTR. A summary of water treatment plant design
flows and average daily flows concludes the industry profile section.
4.3.1 Public Water System Characterization
Categorizing water systems allows EPA to determine the impacts of this rule on different types of
systems. For this EA, EPA sorted systems on the basis of size, ownership, and retail/wholesale
relationships provided in SDWIS. This section explains the classifications used.
PWS Type
As defined by the Safe Drinking Water Act (SDWA), a PWS is a water system that provides
piped water for human consumption and has at least 15 service connections or regularly serves an average
Economic Analysis for the LT2ESWTR 4-8 December 2005
-------
of at least 25 individuals per day for at least 60 days per year. EPA classifies PWSs into two broad
groups:
• Community Water Systems (CWSs) have at least 15 service connections used by year-round
residents or regularly serve at least 25 year-round residents.
• Noncommunity Water Systems (NCWSs) are PWSs that are not classified as CWSs.
EPA further classifies NCWSs into two types:
• Nontransient Noncommunity Water Systems (NTNCWSs) regularly serve at least 25 of the
same people more than 6 months per year.
• Transient Noncommunity Water Systems (TNCWSs) do not regularly serve at least 25 of the
same people more than 6 months per year.
Population Served
Under the LT2ESWTR, as with some previous rules, small systems (those serving fewer than
10,000 people, as defined in the rule) will face somewhat different requirements than larger systems.
System size is important in the regulatory analysis as well. Costs are estimated using the size of systems
as a factor, household costs are derived in part from the number of households in a system size category,
and separate technology decision trees are used for different sizes of systems.
Both the benefits and cost models use the following nine size categories:
• Small systems are broken down into five subcategories based on the number of people
served:
<25
25-<100
100-<500
500-<3,300
3,300-<10,000
• Medium systems are broken down into two subcategories:
10,000-<50,000
50,000-<100,000
• Large systems are broken down into two subcategories:
100,000-<1 million
• >1 million
In some parts of the benefits and cost models (e.g., applying assumptions about existing treatment
practices) and in other analyses, four sizes are used—very small (<500), small (500-<10,000), medium,
and large. Other parts use two categories—small (<10,000) and large (>10,000).
Source Water
Systems are classified by the source water from which they draw. Surface water systems
typically draw from reservoirs, natural lakes, or flowing streams. Ground water systems draw from wells.
Some ground water sources are under the direct influence of surface water sources. These systems, called
GWUDI systems, are considered directly influenced if surface water microorganisms are present. This
Economic Analysis for the LT2ESWTR 4-9 December 2005
-------
category is important to the extent that pathogens, such as Giardia cysts and Cryptosporidium oocysts,
can contaminate the ground water source. Some systems may have multiple source types and are referred
to as "mixed systems." In SDWIS and the Baseline Handbook, a mixed system is categorized as a surface
water system if it gets any portion of its flow from surface water. Based on the analysis in the Model
Systems Report (USEPA 2000a), it is estimated that 21 percent of surface water systems obtain some of
their water from ground water sources. Of these systems, one-third (8 percent of all surface water
systems) get the majority of their flow from ground water. Mixed systems may either be systems that
have some plants that are solely supplied by ground water and other plants that are solely supplied by
surface water, or they may have one or more plants in which both types of source waters are mixed.
Ownership
Systems are also categorized by ownership. Private systems are owned by private corporations,
associations, or individuals. Private systems are still PWSs. Public systems are owned by public entities,
such as a municipality, county, or special district. Ownership distinctions are important to the analysis
since differences exist between public and private systems in their access to capital and other means of
financing. This distinction becomes important in calculating household costs in Chapter 6 and the
Unfunded Mandates analysis presented in Chapter 7.
Purchased andNonpurchased systems
Systems are categorized according to whether they provide or treat water themselves or whether
they purchase it from other systems. Purchased systems are not expected to make treatment modifications
under the LT2ESWTR; instead, purchased systems will absorb (through rate increases) the costs of
additional treatment installed by the sellers. On the other hand, nonpurchased systems collect and treat
the water themselves and distribute it to their retail and wholesale customers. These systems are subject
to most of the provisions of the LT2ESWTR.
4.3.2 Systems, Plants, and Population Subject to the LT2ESWTR
This section estimates the baseline number of systems subject to the LT2ESWTR. The
LT2ESWTR applies to all PWSs, regardless of type or size, that use surface water or GWUDI.
The baseline presented in this section is used to estimate implementation costs, such as those for
training and becoming familiar with the rule (see Appendix D). Not all of the systems incurring
implementation costs will incur costs for other provisions of the rule. Depending on existing treatment
technology, size, and other factors, only subsets of the systems presented in this section are subject to
specific provisions of the rule. The numbers of unfiltered plants, filtered plants, and uncovered finished
water reservoirs subject to specific rule provisions are presented in sections 4.4 through 4.6.
Number of Systems
Systems in SDWIS are listed according to their retail population served. The advantage of
classifying them this way is that it appropriately accounts for both the total number of individual PWSs in
the United States and the total population served by all of those systems. However, a disadvantage of this
method (especially for surface water CWSs) when estimating national costs of regulations is that it does
not directly account for the fact that the water delivered by purchased systems to their retail customers is
actually treated by other, upstream systems. It is important to recognize that the total flow for systems
supplying surface water is actually treated by fewer than half of the surface water systems accounted for
in SDWIS. Because of economies of scale, the unit cost of treatment (in cents per gallon) is lower for
systems treating larger flows than it is for systems treating smaller flows. For example, it is typically
more expensive to build and operate two treatment plants serving 5,000 people than one treatment plant
Economic Analysis for the LT2ESWTR 4-10 December 2005
-------
serving 10,000 people. Failing to account for the fact that surface water is actually treated in larger
quantities at a smaller number of systems than SDWIS suggests could result in an upward bias in national
cost estimates of rules that affect a substantial portion of surface water systems.6
To compensate for this bias, an analysis was performed to link consecutive or purchased surface
water CWSs and NTNCWSs to their respective wholesale systems using SDWIS data. CWSs were only
linked to other CWSs and NTNCWSs were linked only to other NTNCWSs. TNCWSs were not linked,
since they usually do not purchase water from other TNCWSs. If a consecutive system could be linked to
a wholesaler, that system was removed from the system count and its population was added to the
population of the wholesale system. Consecutive systems that could not be definitively linked to their
wholesalers were considered stand-alone purchased systems. Although consecutive systems do not treat
their water, these systems are included in the treatment baseline because their populations and flows must
be accounted for in estimating treatment costs. EPA recognizes that including them as separate plants
overestimates treatment costs. The decision process for this analysis is summarized in Exhibit 4.2
The number of surface water systems (including mixed systems) and GWUDI systems per size
category in SDWIS (pre-linking) is shown on the left side of Exhibit 4.3. Systems whose ownership
category was listed as "other" in SDWIS were reallocated to private and public and purchased and
nonpurchased categories based on the existing proportion of each category to the total number of systems.
Note that the total number of nonpurchased systems in columns H and I is the same as the total number of
nonpurchased systems before linking in columns C and D. However, the numbers of public and private
nonpurchased systems changed slightly because of how the systems with "other" ownership types are
allocated within each size category. The inventory of nonpurchased systems, unlinked or pre-linking, is
used as the baseline for determining implementation costs. Purchased systems are not included in this
baseline because they do not have their own source water, so they will not be subject to monitoring or
treatment requirements and do not need to conduct implementation activities. This baseline is also used
with minor modifications to determine the number of systems (and plants) subject to monitoring costs.
6Many purchased systems do no treatment themselves; the supplying system treats their water. In fact,
fewer than half of the surface water systems in SDWIS treat the total flow from systems that supply surface water.
Economic Analysis for the LT2ESWTR 4-11 December 2005
-------
Exhibit 4.2: Methodology for "Linking" Consecutive Surface Water CWSs and
NTNCWSs to Their Selling Systems
If a system has multiple sources, (e.g., a system has a primary source of surface
water in addition to a purchased surface water source), it was assumed to be
adequately represented as a nonpurchased surface water system, and was not
linked to its seller (i.e., only 100-percent purchased surface water systems were
linked).
If a purchased surface water system (System P) purchases all of its water from
one nonpurchased surface water system (System S), its population was added to
that of System S, and it was removed from the inventory of purchased-water
systems.
If the purchased surface water system buys water from multiple nonpurchased-
water systems, it was assigned to the most directly related nonpurchased seller
with the largest population. For example, a purchased system (System C)
purchases from a nonpurchased system (System B1) and a purchased system
(System B2), which in turn purchases from a nonpurchased system (System A).
In this case, System C was linked to System B1, and the population of System C
was added to that of System B1.
• When the purchased system and its seller were not of the same type (e.g., a
CWS purchasing from a NTNCWS), they were not linked. Systems purchasing
from sellers of a different system type were counted as separate, unlinked
purchased-water systems.
If the PWS identification number of the seller did not correspond to an active
water system, the purchased system was counted as a separate, unlinked,
purchased system.
Some purchased-water systems have what is referred to as "cascading provider
relationships." For instance, a purchased system (System C) may purchase
water from another system (System B). This system (System B) does not treat its
own water, but instead purchases water from another system (System A). For
this analysis, the populations of both Systems B and C were added to the
population of System A, and Systems B and C were removed from the inventory
of purchased-water systems.
In a few cases, the seller could not be found, i.e., a purchased system (e.g.,
System C) could not be linked to a nonpurchased system (e.g., System A).
These purchased-water systems were counted as separate, unlinked, purchased-
water systems.
Economic Analysis for the LT2ESWTR 4-12 December 2005
-------
Exhibit 4.3a: Inventory of Unlinked and Linked Surface Water and GWUDI CWSs
System Inventory Before Linking
Population
Served
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
Total
Number of Systems
Purchased
Public
A
420
843
648
1,186
858
682
103
58
0
4,798
Private
B
302
641
391
303
140
91
18
9
0
1,895
Nonpurchased
Public
C
188
409
324
925
940
863
169
184
14
4,016
Private
D
235
314
102
158
85
99
31
25
2
1,051
Total No. of
Systems
E = A+B+C+D
1,145
2,207
1,465
2,572
2,023
1,735
321
276
16
11,760
Linked System Inventory
Number of Systems
Purchased
Public
F
23
40
21
53
48
32
7
7
0
231
Private
G
8
25
11
23
15
2
0
0
0
84
Nonpurchased
Public
H
109
391
301
846
934
955
218
234
17
4,007
Private
I
232
311
103
155
89
102
36
28
2
1,056
Total No. of
Systems
J=F+G+H+I
372
767
436
1,077
1,086
1,091
261
269
19
5,378
Plants per
System
K
1.0
1.0
1.1
1.0
1.0
1.1
1.2
1.4
3.4
Total Plants
L=J*K
376
767
459
1113
1128
1184
325
384
64
5,799
Population
Served
M
20,526
199,595
317,910
2,175,624
6,655,716
25,903,789
18,249,527
76,010,534
52,305,188
181,838,409
Percent of
Total
Population
N
0.01 %
0.11%
0.17%
1 .20%
3.66%
14.25%
10.04%
41 .80%
28.76%
100%
Exhibit 4.3b: Inventory of Unlinked and Linked Surface Water and GWUDI NTNCWSs
System Inventory Before Linking
Population
Served
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
Total
Number of Systems
Purchased
Public
A
14
29
10
17
8
2
0
1
0
81
Private
B
37
46
16
12
3
2
0
0
0
116
Nonpurchased
Public
C
67
84
23
18
1
0
0
0
0
194
Private
D
114
158
57
45
12
1
0
0
0
386
Total No. of
Systems
E = A+B+C+D
232
317
106
92
24
5
0
1
0
777
Linked System Inventory
Number of Systems
Purchased
Public
F
14
29
10
16
8
2
0
1
0
81
Private
G
32
44
14
12
3
2
0
0
0
106
Nonpurchased
Public
H
66
83
24
17
3
0
0
0
0
193
Private
I
114
156
58
46
11
1
0
0
0
386
Total No. of
Systems
J=F+G+H+I
226
312
106
91
25
5
0
1
0
766
Plants per
System
K
1.0
1.0
1.0
1.0
1.0
1.0
0.0
1.0
0.0
Total Plants
L=J*K
226
312
106
91
25
5
0
1
0
766
Population
Served
M
11,101
72,127
70,321
153,287
125,413
128,055
0
169,846
0
730,150
Percent of
Total
Population
N
1 .52%
9.88%
9.63%
20.99%
17.18%
17.54%
0.00%
23.26%
0.00%
100%
Economic Analysis for the LT2ESWTR Proposal
4-13
December 2005
-------
Exhibit 4.3c: Inventory of Surface Water and GWUDI TNCWSs
Population
Served
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
System Inventory
Number of Systems
Purchased
Public
F
83
25
11
14
3
3
0
0
2
141
Private
G
397
76
17
4
0
0
0
0
0
494
Nonpurchased
Public
H
224
175
48
36
14
7
0
1
0
504
Private
I
569
334
31
13
2
2
0
0
0
952
Total No. of
Systems
J=F+G+H+I
1,273
610
107
67
19
12
0
1
2
2,091
Plants per
System
K
1.0
1.0
1.0
1.0
1.0
1.0
0.0
1.0
1.0
Total Plants
L=J*K
1273
610
107
67
19
12
0
1
2
2,091
Population
Served
M
47,620
119,980
68,762
117,455
89,020
221 ,299
0
1 44,000
12,000,000
12,808,136
Percent of
Total
Population
N
0.37%
0.94%
0.54%
0.92%
0.70%
1 .73%
0.00%
1.12%
93.69%
100%
Notes: TNCWSs were not included in the linking exercise, and therefore only one inventory is presented in Exhibit 4.3c, compared to Exhibits 4.3a and 4.3b.
For TNCWSs, "population served" is actually the population served at a given time. These numbers are used for calculating treatment costs. The total number of
people served over 1 year (or whatever length of time the TNCWSs is in operation) is generally much larger. To calculate benefits, EPA adjusted the population to
account for the total number of people served per year and for the fact that each customer would be served by the system for a shorter period of time than 1 year.
These adjustments are described in Chapter 5.
The total number of nonpurchased systems remains the same before and after linking; however, the number of such systems in a size category may change. This
is because a nonpurchased system's population changes if a purchased system is linked to the nonpurchased system, and the population may change enough to
move the system to the next size category.
The number of purchased systems left after the linking process and the population associated with these systems are included in the baseline for plants subject to
treatment requirements because their population must be accounted for in determining treatment costs (EPA realizes that including the systems themselves does
result in an over-estimate of the number of systems requiring treatment under LT2ESWTR).
Sources: [A]-[D] SDWIS September 2003 (USEPA 2003e); excludes Massachusetts Regional Water Authority and their consecutive systems.
Systems not categorized as "public" or "private" in SDWIS were redistributed among public and private purchased and nonpurchased water systems according to
the proportions of systems in these categories.
[F]-[l] Data from Columns A-D modified using linking methodology described in Exhibit 4.2, except for TNCWSs, which were not linked.
[K] Derived from CWSS data (USEPA 1997c) and Model Systems Report (USEPA 2000a), modified to exclude ground water plants and weighted for
representativeness of each system to all CWSs.
[M] Includes SDWIS population served by surface water and GWUDI systems for nonpurchased and purchased systems in columns F-l based on SDWIS (USEPA
2003e). Original SDWIS population distribution was modified using linking methodology described in Exhibit 4.2, except for TNCWSs, which were not linked.
Economic Analysis for the LT2ESWTR Proposal
4-14
December 2005
-------
The right side of Exhibit 4.3 shows the baseline number of systems, after the linking process, that
are subject to the treatment requirements of the LT2ESWTR. (Actual treatment requirements will be
determined by results of Cryptosporidium monitoring.) This baseline includes both purchased and
nonpurchased systems. The baseline number of systems was converted to plants as described below.
Number of Plants
Water systems can have one treatment plant supplying all of the water distributed to the
population, or multiple plants treating water. The water may come from different sources, or even
different source types such as surface and ground water sources. Many of the costs in Chapter 6 are
developed by estimating the costs of installing additional treatment at existing plants. Therefore, for this
baseline analysis, EPA needed to calculate the number of plants represented by the systems reported in
existing databases.
EPA used the 2000 CWSS data to estimate the number of plants per system. The survey
requested information on the number of treatment plants per system, the source of water treated at each
plant, and the type of treatment in place. The analysis excluded systems whose responses to the treatment
questions were incomplete. (Some reported no treatment for surface water sources, which indicates the
source was likely purchased treated water.) The analysis also excluded systems whose flow rates per
person were unusually high or low. (These systems were identified as outliers in a separate analysis of
average daily flow and design capacity.) Ninety of the 1,246 systems in the sample were dropped from
the analysis, and of the remainder, 587 systems use surface water for at least a portion of their supply.
The number of plants that treat surface water or GWUDI was counted for each system; plants that treated
ground or purchased water only were not included in the count. The mean ratios by size category
incorporated sample weights for survey non-response and specific question non-response. (See The 2000
Community Water System Survey, Volume II for details on sample weight calculations.)
The ratios of plants to systems are shown in Exhibit 4.3. For surface water and GWUDI CWSs,
the average number of surface water treatment plants per system varies from 1.0 to 3.4.
Plant information is not available for noncommunity water systems. Because they typically serve
a single building or are located in a small area, this analysis assumes that the ratio of plants per system is
1:1 for all size categories. Exhibit 4.3 summarizes the total number of plants for CWSs, NTNCWSs, and
TNCWSs. The total number of plants displayed in Exhibit 4.3 is the baseline from which regulatory
impacts are estimated.
Population
The total population that the LT2ESWTR affects is derived from SDWIS data (USEPA 2003e).
As described in Exhibit 4.2, the linking process redistributes the population served by purchased systems
to the seller. This analysis also corrects for double-counted populations between wholesale and
consecutive systems. The breakdown of population by size category is shown in Exhibit 4.3a; this
breakdown includes adjustments made to populations during the linking process (see Exhibit 4.2).
One method for determining the impact of the rule involves a cost analysis on a per household
basis. Only the population served by CWSs is considered in the household cost analysis. People served
by NTNCWSs and TNCWSs when working, attending school, or traveling are also served on a regular
basis by another source, such as a private well or a CWS. Their consumption from a NCWS is an
incidental use in addition to their regular service. Adding the population served by NCWSs to that served
by CWSs would lead to double counting in cases where both types of systems served the same person. If
some people served by private wells receive water from NTNCWSs for part of the year, the error
introduced by ignoring this consumption from this source is less than it would be if the population served
by NCWSs were included in the baseline.
Economic Analysis for the LT2ESWTR 4-15 December 2005
-------
The population served is affected differently depending on the service type, in that people
experience different durations of exposure. This, in turn, affects water demand on the system, as well as
risk of exposure to contaminated water.
Uncertainty in Baseline Input Data
EPA recognizes that there is uncertainty in the data sources used to define the system inventory
for the LT2ESWTR. The uncertainty is not quantified in this EA; however, a qualitative discussion of the
identified uncertainties is provided below.
As noted above, SDWIS and the 2000 CWSS are the primary sources of system inventory data.
SDWIS is EPA's primary drinking water database. It stores State-reported information on each water
system, including name, ID number, number of people served, type of system (year-round or seasonal),
and source of water (ground or surface). These data are required fields of entry; additional data, such as
buying and selling information, are not required. In 1998, EPA began a major effort to assess the quality
of the data in SDWIS. The results, published in Data Reliability Analysis of the EPA SDWIS/FED, found
that the quality of the required inventory data was high (USEPA 2000c). Thus, EPA believes that
uncertainty in the system inventory data from SDWIS with respect to numbers of systems, source
information, and size classification is low, and need not be further accounted for in the analysis.
The 2000 CWSS was the primary data source used to develop estimates of the number of
treatment plants per system. It was developed to gather data on CWSs in the United States. Of the 1,870
systems statistically selected to receive the main survey questionnaire, 1,246 responded. These responses
were weighted and adjusted for item nonresponse to maintain statistical representation of the total
universe of CWSs (USEPA 2003f). Forthe surface-water-plant-per-system analysis, 587 systems were
included. This represents slightly more than 10 percent of the nonpurchased CWS systems in the
baseline. EPA believes that extrapolating mean estimates from these data to the CWS baseline is
appropriate, and that the error this procedure introduces is negligible.
4.3.3 Water Treatment Plant Design and Average Daily Flows
Treatment technology costs are based on the volume of water treated per day. The cost analysis
described in Chapter 6 uses two types of treatment plant flow:
• Design flow—the maximum capacity at which the plant was intended to operate, expressed in
millions of gallons per day (mgd).
Average daily flow—the flow a treatment plant produces, averaged over 365 days, expressed
in mgd.
Design flows are used to estimate the capital costs of the technology that will be installed to meet
the requirements of the LT2ESWTR. Average daily flows are used to estimate the annual cost of ongoing
operations and maintenance (O&M). Average daily flows give a better indication of chemical usage and
operational costs than do design flows. The flows presented in this section are used to estimate costs for
both unfiltered and filtered plants.
To derive flow information for different-sized plants, EPA developed the following regression
equations relating design and average daily flow to population served for surface water systems using
data from the 1995 CWSS:
Design Flow (MGD) = 0.36971 ^97757/l,000
Average Daily Flow (MGD) = 0.10540 X102058/l,000
Economic Analysis for the LT2ESWTR 4-16 December 2005
-------
Where X= mean population served.
The derivation of these equations is presented in detail in the Model Systems Report (USEPA 2000a) and
summarized in the Baseline Handbook (USEPA 200Ic). EPA used these equations to estimate mean
flows per system, based on the population per system shown in column A of Exhibit 4.4, and then divided
by the number of plants per system to determine the flow per plant.
Exhibit 4.4a summarizes populations served and design and average daily flows for filtered plants
at CWSs. Exhibit 4.4b shows the flows and populations for unfiltered plants (all unfiltered plants are
CWSs except for one TNCWS, which was grouped with the CWSs). EPA recognizes that there is a range
of design and average daily flows within each category, but believes that using mean flow values is
adequate for the cost and benefit analyses in this EA.
An equivalent regression analysis relating NCWS flows to population served was not done in the
Model Systems Report. Therefore, average daily and design flows for NCWSs were estimated using
mean population served per plant for NCWSs substituted into the CWS regression equations. Flows are
summarized in Exhibit 4.4a for filtered NTNCWSs and TNCWSs. Plant flows for filtered CWSs,
NTNCWSs, and TNCWSs differ from each other because of the difference in mean population per plant
for each of the three categories, and the volume of water delivered to commercial and industrial
customers. Use of CWS equations to determine NCWS flows may result in an overestimation of flows
because NCWSs often serve people for only part of the day. This may lead to an overestimation of costs
for NCWSs. This overestimation is addressed as part of the uncertainties summarized in section 4.8.
For this rulemaking, EPA considered estimating flows for NCWSs according to service category
(e.g., schools, restaurants, hotels, industry), as has been done in some other rules, instead of size. EPA
decided against such an approach for the following reasons:
Service category flows are based on mean population served for all systems in that category,
regardless of source water type. EPA expects that surface water and GWUDI sources would
be more prevalent in larger noncommunity systems, but has no basis for developing revised
population estimates for each service category by source.
• More critical to the LT2ESWTR, the method used to predict technology selection in Chapter
6 is a function of population served, and does not directly apply to service categories that
may include a wide range of water system sizes (e.g., schools can be very small local
buildings or large universities).
Economic Analysis for the LT2ESWTR 4-17 December 2005
-------
Exhibit 4.4a: Average Daily and Design Flow by System Size for Filtered Plants
System Size
Population
per System
A
Plants per
System
B
Average Daily Flow (MGD)
per Plant
C = (0.10540*A1 02058)/(B*1,000)
Design Flow (MGD) per Plant
(0.36971*A° 97757)/(B*1 ,000)
CWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
55
259
723
2,000
6,027
23,481
68,891
277,100
2,320,282
1.0
1.0
1.1
1.0
1.0
1.1
1.2
1.4
3.4
0.01
0.03
0.08
0.24
0.73
2.81
7.34
26.49
98.62
0.02
0.08
0.22
0.60
1.76
6.38
15.95
54.21
184.16
NTNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
49
231
663
1,652
5,017
25,611
0
169,846
0
1.0
1.0
1.0
1.0
1.0
1.0
-
1.0
-
0.01
0.03
0.08
0.20
0.63
3.33
-
22.94
-
0.02
0.08
0.21
0.52
1.53
7.54
-
47.93
-
TNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
37
197
643
1,753
4,685
18,442
0
144,000
6,000,000
1.0
1.0
1.0
1.0
1.0
1.0
-
1.0
1.0
0.004
0.02
0.08
0.22
0.59
2.38
-
19.38
871 .95
0.01
0.06
0.21
0.55
1.43
5.47
-
40.79
1,563.07
Source: [A] Population served for each size category (Exhibit 4.12, Column E) divided by number of systems
in each category (Exhibit 4.12 Column C), based on the treatment baseline.
[B] 2000 CWSS.
[C] USEPA 2000a.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.4b: Average Daily and Design Flow by System Size for Unfiltered Plants
System Size
Population
per System
A
Plants per
System
B
Average Daily Flow (MGD) per
Plant
C = (0.10540*A1-02058)/(B*1,000)
Design Flow (MGD) per Plant
D = (0.36971*A°'97757)/(B*1,000)
CWS
<100
1 00-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1 ,000,000+
78
298
822
1,663
7,861
21,979
67,277
367,684
4,109,917
1.0
1.0
1.1
1.0
1.0
1.1
1.2
1.0
1.0
0.01
0.04
0.09
0.20
0.96
2.62
7.17
50.45
592.64
0.03
0.10
0.25
0.50
2.29
5.98
15.58
101.98
1,079.81
Source: [A] Population served for each size category (Exhibit 4.5, Column D) divided by number of systems in each
category (Exhibit 4.5, Column A), based on the treatment baseline.
[B], [C] USEPA 2000a; plant per system ratios for systems serving over one million based on system-specific data
supplied by States.
4.4 Baseline for Unfiltered Plants (Pre-LT2ESWTR)
Unfiltered plants are subject to different LT2ESWTR provisions than filtered plants. (For
example, all unfiltered plants must achieve some inactivation of Cryptosporidium; bin assignments only
determine the degree of inactivation required.) Therefore, the baselines for unfiltered and filtered plants
must be developed separately. The following sections summarize the existing treatment; system, plant,
and population data; source water Cryptosporidium occurrence; and predicted finished water
Cryptosporidium occurrence for unfiltered plants.
4.4.1 Treatment Characterization for Unfiltered Plants
EPA estimates that a number of plants that currently are not required to filter will need to install
an advanced disinfectant technology to meet the 2 log or 3 log Cryptosporidium inactivation requirement
for LT2ESWTR. Some treatment data collected over time by EPA regional offices are available on these
plants. Most of these plants are not predicted to add treatment to meet the Stage 1 or Stage 2 DBPR
requirements, because they generally have low-turbidity source water and, thus, low levels of precursors
for DBP formation. Unfiltered plants are not subject to IESWTR or LT1ESWTR filtration requirements.
EPA therefore used the existing regional data to develop the treatment characterization for these
unfiltered plants.
A review of these data reveals that unfiltered plants use a variety of treatments to disinfect or
control other water quality problems. Plants serving 3,300 or fewer people generally use chlorine as the
primary disinfectant, although at least one plant serving 501 to 1,000 people uses ozone. Some of these
plants may also employ corrosion control, as well as manganese and iron removal. Plants serving 3,301
or more people mainly use chlorination for disinfection, and at least one plant serving 3,301 to 10,000
people uses chlorine dioxide. Other treatment processes used in medium and large unfiltered plants
include corrosion control, softening, fluoridation, DBP control, taste and odor control, as well as organics
and iron removal. Some plants avoiding filtration may have already installed treatment equivalent to
filtration, and some systems may filter water from some but not all of their sources.
The unfiltered plant database contains treatment data for less than 50 percent of unfiltered plants.
The ICR database was also examined. Only one unfiltered plant participating in the ICR was found to use
Economic Analysis for the LT2ESWTR
December 2005
-------
ozone (USEPA 2000h), and its dose levels were not high enough to meet the LT2ESWTR requirements.
No ICR unfiltered plants used chlorine dioxide. Because of the high doses of ozone or chlorine dioxide
required to inactivate Cryptosporidium, it appears unlikely that any unfiltered plants are currently meeting
the requirements of the LT2ESWTR. Therefore, for this EA, EPA estimates with a high degree of
confidence that all unfiltered systems will have to add advanced disinfection to achieve the 2 log
Cryptosporidium inactivation minimum requirement.
4.4.2 Number of Unfiltered Systems, Plants, and Population Served
Systems that operate unfiltered plants can be placed in one of two categories:
• Those with plants that are now unfiltered but are required to filter under the 1989 Surface
Water Treatment Rule (SWTR); and
Those that meet the filtration avoidance criteria of the SWTR.
Three of the 12 unfiltered plants in the ICR are currently unfiltered, but will be changing to
filtration in the future. These systems are subject to requirements of the proposed LT2ESWTR for
filtered systems because they do not meet the avoidance criteria under the SWTR; they are therefore
included in the baseline for filtered systems (section 4.5). In addition, a fourth plant, the Massachusetts
Water Resources Authority, was omitted from the unfiltered baseline (and not moved to filtered baseline)
because of the uncertainties regarding its filtration avoidance status due to ongoing litigation at the time
this calculation was done. The purchased systems associated with this plant were also removed from the
baseline presented in Exhibit 4.3.
The baseline for unfiltered plants, therefore, includes only systems that meet the SWTR
avoidance criteria. The criteria include that the plants:
Disinfect to achieve 3 and 4 log reduction ofGiardia and viruses, respectively;
Have watershed control measures in place; and
• Are below source water limits on fecal coliform occurrence (20/100 ml) and turbidity (5
nephelometric turbidity units (NTU)).
Exhibit 4.5 presents the baseline for unfiltered systems and plants that is used for estimating costs
and benefits in this EA. Data on populations served and the number of systems are derived from SDWIS
and from the ICR for large systems (USEPA 2003e, 2000h). The number of plants is calculated using the
plant-per-system ratios given earlier. There is only one TNCWS unfiltered system, and there are no
unfiltered NTNCWSs. Maintaining a separate category in subsequent analyses for one system was
judged unlikely to add precision, so this TNCWS was grouped with CWSs in all subsequent analyses.
These baseline values, in conjunction with flows presented in section 4.3.3, are used to estimate costs for
unfiltered systems (see Chapter 6).
Economic Analysis for the LT2ESWTR 4-20 December 2005
-------
Exhibit 4.5: Treatment Baseline for Unfiltered Plants by System Size
System Size
(Population
Served)
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
Number of
Systems
A
1
4
3
15
14
13
4
4
2
60
Plants per
System
B
1.0
1.0
1.1
1.0
1.0
1.1
1.2
1.0
1.0
Number of
Plants
C=A*B
1
4
3
15
15
14
5
4
2
63
Population
Served
D
78
1,192
2,467
24,947
110,056
285,732
269,106
1 ,470,734
8,219,833
10,384,145
Percent ot Total
Population Served
by Surface and
GWUDI CWSs
E
0.38%
0.59%
0.77%
1.13%
1 .63%
1 .09%
1 .45%
1 .90%
13.58%
5.40%
Notes: All systems are CWSs except one TNCWS, which was grouped with the CWSs for analysis.
Sources: [A] SDWIS (USEPA 2003e) data adjusted by EPA to exclude systems that do not meet filtration
avoidance criteria.
[B] Exhibit 4.3, Column K.
[D] SDWIS data for the systems in Column A (2000h), modified to include populations added in the linking
process (see Exhibit 4.1).
[E] Population served (Column D) divided by total population served by surface water and GWUDI CWSs
(Exhibit 4.3a, Column M).
4.4.3 Source Water Cryptosporidium Occurrence for Unfiltered Plants
ICR data from 12 plants that are classified as unfiltered surface water are used to characterize
Cryptosporidium occurrence (USEPA 2000h). The results of the ICR Cryptosporidium monitoring were
evaluated using the model described in section 4.2.2. Because a few of these plants do not meet the
filtration avoidance criteria, a sensitivity analysis was performed to see if results would be significantly
different if they were excluded. The occurrence distributions with and without the affected systems were
nearly identical; therefore, the original results were used in the analysis.
Observed Cryptosporidium Occurrence
Observed results for ICR unfiltered plants are shown in Exhibit 4.6. A comparison with those for
filtered plants in the ICR (Exhibit 4.13) shows both a lower rate of positive samples and a lower average
concentration of Cryptosporidium for unfiltered plants. While too few unfiltered plants were sampled in
the ICRSS to carry out a meaningful analysis, the few that were sampled also were on the low side of the
ICR unfiltered occurrence distribution.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.6: Observed ICR Total Oocyst Occurrence in Source Water for
Unfiltered Plants
Total
Number of
Plants
12
Number of Plants with
at Least One Positive
Sample (Percent)
7 (58%)
Observed Plant-Mean Data
(oocysts/L)
Mean
0.002
Median
0.001
90th Percentile
0.005
Notes: Total Cryptosporidium includes non-empty and empty oocysts. Non-empty includes oocysts with
internal structures and with amorphous structures. For each plant, all monthly observations were averaged
over the sampling period (18 months) to produce plant-mean data. The mean, median, and 90th percentile
shown summarize plant-mean Cryptosporidium for all unfiltered plants.
Source: USEPA2000h.
Modeled Cryptosporidium Occurrence
The data from these 12 plants are used to fit the unfiltered plants occurrence model, which serves
as input to the EA. The modeling was carried out using the approach outlined in section 4.2.2. As
explained in that section, the modeling produces a collection of plausible occurrence distributions. Exhibit
4.7 summarizes this collection of distributions. The solid center curve represents the mean distribution
across the collection, and the dotted lines give a 90-percent confidence bound for the unfiltered
occurrence distribution.
4.4.4 Finished Water Cryptosporidium Occurrence for Unfiltered Plants
As mentioned in section 4.4.1, because unfiltered plants do not have advanced treatment
technologies in place that are capable of meeting 2.0 log Cryptosporidium removal or inactivation, they
are not expected to remove or inactivate any Cryptosporidium from their source water. Although most
unfiltered systems chlorinate their water (a few may use other disinfectants besides chlorine), chlorination
is ineffective for inactivation of Cryptosporidium. Therefore, the finished water occurrence of
Cryptosporidium for unfiltered plants is assumed to be the same as the source water occurrence shown in
Exhibit 4.7.
The occurrence distribution in Exhibit 4.7 is derived from the ICR data set (USEPA 2000h).
Although there were unfiltered plants in the ICRSS data sets, they were too few to use successfully in the
model to derive national distributions. Thus, there are no explicit estimated occurrence distributions for
the unfiltered ICRSS data sets. In order to develop national benefit estimates for the ICRSS data sets,
EPA estimated unfiltered results based on the ratios between ICR and ICRSSM and between ICR and
ICRSSL for filtered plants. Thus, although no explicit occurrence distributions were estimated for the
ICRSS data sets, the likely differences in Cryptosporidium occurrence between the ICR and ICRSS
unfiltered data sets are reflected in later analyses.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.7: Modeled Cryptosporidium Occurrence in Source Water:
ICR Data for Unfiltered Systems
100%
_ro
0_
80% -
60%-
D.
40% -
20%
0% -
This graph shows modeled variability
and uncertainty in source water
Cryptosporidium occurrence for unfiltered
systems. It summarizes 1,000 plausible
curves used in the Economic Analysis.
Ninety percent of the modeled curve values
fall within the dashed-line uncertainty
bounds. The heavy center line reflects the
central tendency across all 1 ,000 curves.
Among the 1 ,000 modeled curves, any single
curve describes variability in occurrence
across all filtered systems. Steeper curves
indicate less variability from system to system.
Example:
Modeled Percent of Plants with
Average Source Water
Cryptosporidium Concentration
Below 0.01 oocysts/L:
5th %tile Bound = 74%
Median Estimate = 59%
95 %tile Bound = 39%
5th/95th %tile
1e-005 0.0001 0.001 0.01 0.1 1
Plant Mean Cryptosporidium Concentration (Total oocysts/L)
10
Source: USEPA 2003c.
Under the LT2ESWTR, all unfiltered systems must provide treatment for Cryptosporidium.
Systems with Cryptosporidium concentrations less than or equal to 0.01 oocysts/L must provide 2 log
treatment, while systems with concentrations greater than 0.01 oocysts/L must provide 3 log treatment.
The predicted bin assignments for unfiltered systems, derived from the occurrence distribution in Exhibit
4.7, are shown in Exhibit 4.8. The percentages shown represent averages over 250 simulated assignments
of unfiltered systems to 2 or 3 log treatment bins. In the case of systems serving fewer than 100,000
people, the simulated assignments were based on 1,000 values drawn from the modeled unfiltered
occurrence distribution. For the small number of systems serving 100,000 or more, the simulation drew
Cryptosporidium concentrations directly from ICR survey results. In both cases, recovery was simulated
to match recovery rates expected in Cryptosporidium monitoring employing EPA Methods 1622/1623.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.8: Predicted System Bin Assignments for Unfiltered Systems, Based on
Central Tendency of Cryptosporidium Occurrence
System Size (Population Served)
<1 00,000
>100,000
2.0 Log Treatment
79.2%
81 .2%
3.0 Log Treatment
20.8%
18.8%
Source: Monte Carlo simulation (Appendix B). For systems serving fewer than 100,000, percentages based on
modeled occurrence distribution and lab method recovery rate distribution. For systems serving 100,000 or more,
percentages are based on actual ICR results and modeled lab method recovery distribution.
incurred.
The percentages in Exhibit 4.8 were used to determine treatment costs each unfiltered system
i
4.5 Baselines for Filtered Plants (Pre-LT2ESWTR)
This section presents the Pre-LT2ESWTR treatment characterization for filtered plants. It
includes estimates of the number of systems, the number of plants, and the population served for filtered
systems. It also contains Cryptosporidium occurrence in source water and finished water and predicted
bin classifications.
4.5.1 Treatment Characterization for Filtered Plants
The treatment characterization for the LT2ESWTR must take into account projected treatment
modifications made to comply with other existing and soon-to-be-promulgated rules. This adjustment
allows treatment changes attributable only to LT2ESWTR requirements to be isolated for the purpose of
benefit and cost analysis. In addition, systems with certain treatments already in place prior to the
implementation of the LT2ESWTR may qualify for pre-LT2ESWTR credit, meaning they can get credit
towards the log treatment requirements of the LT2ESWTR. The rest of this subsection explains how
treatment changes attributable to the IESWTR, LT1ESWTR, Stage 1 DBPR, and Stage 2 DBPRhave
been accounted for in developing the LT2ESWTR baseline. Although other rules have been promulgated
recently or are scheduled to be promulgated before this rule, they either do not affect surface water
supplies or do not involve installation of treatment that is expected to remove significantly or to inactivate
Cryptosporidium. Therefore, such rules were not considered further.
Post-IESWTR andLTlESWTR Treatment Characterization
For this EA, it is assumed that all medium and large systems (those serving at least 10,000
people) using conventional or direct filtration meet the combined effluent turbidity limit of 0.3 NTU 95
percent of the time, as required under IESWTR and LT1ESWTR. EPA assumed as part of the economic
analyses for the IESWTR and LT1ESWTR that systems would achieve less than 0.2 NTU 95 percent of
the time in order to operate within a margin of safety. Lower finished water turbidity, with a combined
effluent turbidity level of 0.15 NTU, is one of the approaches in the microbial toolbox (described in detail
in Chapter 6) available to systems for achieving an additional 0.5 log removal credit for Cryptosporidium.
There are several other toolbox technologies that plants may already have installed that would gain them
Pre-LT2ESWTR credit of 0.5 log toward Cryptosporidium removal requirements. These include
secondary filters and two-stage softening.
To determine the number of plants that might obtain credit for already using some of these
toolbox technologies, several data sources were reviewed. Data from the ICR, the American Water
Works Association (AWWA) (AWWA 2000), a survey of small systems conducted by the National Rural
Economic Analysis for the LT2ESWTR
December 2005
-------
Water Association (NRWA) (USEPA 200 Ib), and the 1995 CWSS were reviewed to determine the
percentage of plants that have these technologies that would qualify for removal credits. Exhibit 4.9
provides an estimate of the percentage of systems that would be capable of obtaining each of these Pre-
LT2ESWTR half-log removal credits with existing equipment and operations or with equipment and
operations predicted to be installed and in use prior to promulgation of this rule. Details on the derivation
of these numbers are provided in Appendix A. Some of these plants may be able to obtain credit for
having multiple technologies. The percentage of plants estimated to get 1.0 log removal credit is
presented in Exhibit 4.9 as well. The estimates of 1.0 log credits are derived from the individual credits
for plants receiving the various 0.5 log removal credits, assuming complete independence of the chance of
having any one of the 0.5 log credits.
Exhibit 4.9: Percentage of Plants Qualifying for Pre-LT2ESWTR Cryptosporidium
Log Reduction Credits for Existing Technologies
System
Size
(Population
Served)
Small
(< 10k)
Medium
(10k -100k)
Large
(>100k)
Plants with
0.15 NTU
Finished
Water
Turbidity
(0.5 Log)
A
34%
46%
46%
Plants with
Multiple Settling
Basins
(Conventional
and Softening)
(0.5 Log)
B
3%
5%
5%
Plants
with
Multiple
Filters
(0.5 Log)
C
0%
4%
7%
Plants with 1 .0 Log
Total Credit
E=(A*B)+(A*C)+(B*C)
1%
4%
6%
Plants with 0.5
Log Total
Credit
D=A + B + C - E
36%
51%
52%
Source: Appendix A, Exhibit A.7.
Post-Stage 2 DBPR Treatment Characterization
Under the LT2ESWTR, EPA exempts plants from monitoring, bin classification, and associated
treatment requirements if they are achieving 5.5 log reduction. EPA estimates that very few plants met
this requirement at the time of the ICR. A proportion of surface water plants, however, are expected to
implement advanced technologies to meet the DBP requirements of the Stage 1 DBPR and Stage 2
DBPR. Several of these technologies will provide additional logs of Cryptosporidium removal or
inactivation in addition to reducing DBFs. Advanced technologies that will provide both DBP and
Cryptosporidium control include the following:
• Chlorine dioxide
• UV
• Ozone
Microfiltration/ultrafiltration (MF/UF)
SWAT (see section 4.2.3) used source water and treatment data collected through the ICR to
predict the percentage of large systems that would have to add treatment to meet Stage 2 DBPR
Economic Analysis for the LT2ESWTR
December 2005
-------
requirements. SWAT was also able to predict plants' technology selection based on source water
characteristics, treatment plant configurations, and other factors. SWAT results were extrapolated to
medium and small systems using best professional judgment. EPA's Economic Analysis for the Stage 2
Disinfectants and Disinfection Byproducts Rule (USEPA 2003d) provides a detailed description of Stage
2 DBPR requirements, how SWAT modeled ICR systems, and how SWAT was used to develop the Stage
2 DBPR compliance forecast for plants of all sizes. Exhibit 4.10 presents predictions from SWAT of
technology use following Stage 2 DBPR implementation for the four technologies listed above.
Exhibit 4.10: Predicted Percentage of Plants Using Advanced Technologies
Following Implementation of the Stage 2 DBPR
System Size
(Population Served)
< 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 Million
> 1 Million
Chlorine
Dioxide
0%
2.4 %
2.4 %
5.1 %
5.1 %
7.0 %
7.0 %
7.0 %
7.0 %
uv
3.1 %
0.4 %
0.4 %
0.5 %
0.5 %
0.7 %
0.7 %
0.7 %
0.7 %
Ozone
0%
1 1 .9 %
1 1 .9 %
10.1 %
10.1 %
12.8%
12.8%
12.8%
12.8%
MF/UF
18.6%
9.9 %
9.9 %
5.4 %
5.4 %
1 .8 %
1 .8 %
1 .8 %
1 .8 %
Source: Economic Analysis for the Stage 2 Disinfectants/Disinfection Byproducts Rule
(USEPA 2003d), Exhibit 6.15a.
EPA did not adjust the baseline for plants that use chlorine dioxide or ozone. Chlorine dioxide
and ozone doses required for Cryptosporidium inactivation are higher than SWTR requirements for
inactivation ofGiardia and viruses. To evaluate the use of these technologies at doses that could
inactivate Cryptosporidium, both incremental costs for the increased dose and incremental benefits would
need to be evaluated. However, available studies do not allow the quantitative evaluation of inactivation
based on dose changes. In the absence of usable data, this analysis assumes that plants predicted to have
installed chlorine dioxide or ozone would not meet the monitoring and treatment exemption requirements
for the LT2ESWTR. The cost model for this EA assumes that these plants would install the entire
technology again for the LT2ESWTR rather than simply increase the dose, resulting in an overestimate of
costs. Uncertainties and biases affecting costs are summarized in section 4.8.
Currently, very few plants use UV. This is partly because of the need for higher doses required to
inactivate viruses (compared to that needed for protozoans) and the need to maintain a disinfectant
residual. These factors result in only about half a percent of plants nationwide using UV. Since this is
such a small number of plants, the LT2ESWTR baseline was not adjusted to reflect plants that already
have installed UV.
MF/UF is the only treatment process assumed to achieve 5.5 log removal. The LT2ESWTR
baseline is adjusted in several ways to account for the percentage of plants that may already be using
MF/UF. CWS and NTNCWS plants (both in small and large systems) that are predicted to have installed
MF/UF prior to the Stage 2 DBPR are removed entirely from the monitoring and treatment baselines.
Plants achieving greater than 5.5 log treatment of Cryptosporidium are meeting the highest level of
Economic Analysis for the LT2ESWTR
December 2005
-------
treatment that could be required based on source water monitoring. These plants would not have to add
treatment if they did monitor and were assigned to the most stringent treatment bin. EPA anticipates that
some plants will install MF/UF to meet the requirements of the Stage 2 DBPR. These plants are included
in the rule implementation and initial monitoring baselines, because they are still subject to the
LT2ESWTR and because the Stage 2 DBPR schedule is such that they will not have installed MF/UF
before initial Cryptosporidium monitoring begins. They are not, however, assigned to the future
monitoring and treatment baselines. TNCWSs are unlikely to have installed advanced technologies,
partly because Stage 1 and Stage 2 DBPRs do not address TNCWSs.
Treatment plants with different types of filtration systems have been regulated differently under
IESWTR and LT1ESWTR and, to a small extent, under LT2ESWTR. The types of filtration included in
this analysis are as follows:
Conventional filtration includes coagulation, flocculation, and sedimentation of particles,
followed by granular media filtration.
• Direct filtration involves coagulation and flocculation followed by rapid sand filtration, but
no sedimentation.
• Slow sand filtration works at very low filtration rates without the use of coagulant in
pretreatment.
• Diatomaceous earth filtration works at low filtration rates with the addition of diatomaceous
earth.
• Alternative filtration systems include membrane, bag, and cartridge filters.
Plants filtering by slow sand and diatomaceous earth are not required to meet new combined filter
effluent provisions under the IESWTR or LTIESWTR because they can generally achieve higher
Cryptosporidium removals at higher effluent turbidity levels. These treatment plants are still subject to
additional treatment requirements of the LT2ESWTR based on the results of source water monitoring
and, thus, are included in the filtered system baseline. Direct filtration plants also are included in this
baseline; however, they have slightly different Cryptosporidium reduction requirements under the
LT2ESWTR. The impacts of these additional requirements for direct filtration plants are addressed in
Chapter 6.
4.5.2 Number of Filtered Plants and Population Served
Because the LT2ESWTR requires treatment for only certain filtered systems based on the results
of their source water monitoring, two baselines are needed for filtered plants-monitoring and treatment.
Source water monitoring requirements for wholesale systems are determined by the population served by
the largest system in the combined distribution system of which the wholesale system is a part. During
the SDWIS linking analysis that linked buyers to sellers (described in section 4.3), EPA also determined
the population of the largest system in the combined distribution for use in estimating monitoring costs.
Exhibit 4.11 presents the filtered plant baseline for estimating implementation and source water
monitoring costs. Note that the number of plants does not include the purchased plants that could not be
linked to their sellers since no additional monitoring costs will be incurred (the treatment baseline
includes these systems as discussed below).
Economic Analysis for the LT2ESWTR 4-27 December 2005
-------
Exhibit 4.11: Implementation and Monitoring Baseline for Filtered Systems
System Size
Implementation Baseline
Number of Filtered Systems
A
Percent Avoiding
Monitoring and
Treatment
Requirements
B
Plants per
System
C
Monitoring Baseline
Number of Filtered
Systems
D=A*(1-B)
Number of Filtered
Plants
E=D*(1-C)
CWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
341
708
425
1,076
1,052
1,010
213
220
16
5,061
3.6%
3.6%
3.6%
3.6%
3.6%
0.4%
0.4%
0.4%
0.4%
1.01
1.00
1.05
1.03
1.04
1.08
1.24
1.43
3.35
329
683
410
1,037
1,014
1,006
212
219
16
4,926
333
683
432
1,072
1,054
1,092
264
313
53
5,294
NTNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
180
241
81
63
13
1
0
0
0
579
3.6%
3.6%
3.6%
3.6%
3.6%
0.4%
0.4%
0.4%
0.4%
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
174
232
78
61
13
1
0
0
0
558
174
232
78
61
13
1
0
0
0
558
TNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
793
509
79
49
16
9
0
1
0
1,456
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
0.0%
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
793
509
79
49
16
9
0
1
0
1,456
793
509
79
49
16
9
0
1
0
1,456
Sources:
[A] SDWIS September 2003 (USEPA 2003e) - Nonpurchased surface water and GWUDI systems.
[B] Percentage of plants predicted to have installed MF/UF to comply with Stage 1 DBPR, from SWAT technology
results for Pre-Stage 2 DBPR (USEPA 2003d).
[C] Derived from 2000 CWSS data.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.12 presents the treatment baseline for filtered plants. The plants estimated to have
MF/UF installed to comply with the Stage 2 DBPR are removed from the LT2ESWTR treatment baseline
as their costs and benefits are associated with the Stage 2 DBPR. The plants used to estimate benefits and
treatment costs (columns G and H) include nonpurchased plants and the purchased plants that could not
be linked with the systems from which they purchased their water. To capture the total flow that must be
treated and the population affected by the LT2ESWTR, EPA includes these purchased systems in the
analysis as though they are treating water themselves. This assumption places more plants in smaller size
categories since most are a part of small systems. In the process of linking purchased systems to sellers,
the population served by the purchased systems is added to that of the sellers. This can result in a
population change large enough to bump the seller up to a higher size category, increasing the flow and
consequently, unit costs for that system. Considering both effects, EPA believes that these purchased,
unlinked systems cause an overestimation of cost since a complete technology is allocated to the smaller
purchased system, instead of increasing the unit costs for the seller due to the increased population.
Uncertainties associated with categorization of some purchased systems as retail are summarized in
section 4.8.
Economic Analysis for the LT2ESWTR 4-29 December 2005
-------
Exhibit 4.12: Treatment Baseline for Filtered Plants
System Size
CWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
Number of
Systems
A
372
767
436
1,077
1,086
1,091
261
269
19
5,378
Percent
Avoiding
Treatment
Requirements
B
18.64%
9.93%
9.93%
5.44%
5.44%
1 .83%
1 .83%
1 .83%
1 .83%
Treatment Baseline
Number of
Filtered
Systems
C=A(1-B)
Number of
Filtered
Plants
D
Population
Served
E
303
691
393
1,018
1,027
1,071
256
264
19
5,041
306
691
414
1,052
1,067
1,162
319
377
63
5,450
16,636
178,702
284,120
2,036,517
6,189,576
25,148,862
17,651,109
73,174,602
43,277,931
167,958,055
NTNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
226
312
106
91
25
5
0
1
0
766
18.64%
9.93%
9.93%
5.44%
5.44%
1 .83%
0.00%
0.00%
0.00%
184
281
95
86
24
5
0
1
0
676
184
281
95
86
24
5
0
1
0
676
9,032
64,965
63,338
142,111
118,591
125,710
0
166,735
0
690,482
TNCWSs
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
1,273
610
107
67
19
12
0
1
2
2,091
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
1,273
610
107
67
19
12
0
1
2
2,091
1,273
610
107
67
19
12
0
1
2
2,091
47,620
119,980
68,762
117,455
89,020
221 ,299
0
1 44,000
12,000,000
12,808,136
Sources:
[A] Exhibit 4.3, Column J less unfiltered systems from Exhibit 4.5, Column A.
[B] Percentage of plants predicted to have installed MF/UF to comply with Stage 2 DBPR, from SWAT
technology results for Post-Stage 2 DBPR (USEPA 2003d).
[C] Number of filtered systems subject to treatment = linked nonpurchased and purchased systems
(Exhibit 4.3, Column J) * (1-Column F). This column includes nonpurchased systems that conducted
monitoring, but excludes systems that monitored but that are predicted to install MF/UF to comply with
the Stage 2 DBPR. Purchased systems are included so their populations can be used to determine
benefits of treatment installed under the LT2ESWTR (customers of purchased systems will incur
benefits of treatment installed at the associated nonpurchased system).
[D] Number of filtered plants = Column G * Exhibit 4.3, Column K.
[E] Population = (Pop. served by linked nonpurchased and purchased systems (Exhibit 4.3, Column M)
- pop. served by unfiltered systems (Exhibit 4.5, Column A)) * (1-F).
4.5.3 Source Water Cryptosporidium Occurrence for Filtered Plants
For filtered plants, the results of plant-specific source water monitoring for Cryptosporidium will
dictate the additional treatment required to meet provisions of the LT2ESWTR. This subsection
Economic Analysis for the LT2ESWTR
December 2005
-------
summarizes observed Cryptosporidium data and discusses statistical modeling used to estimate from the
observed data the underlying true Cryptosporidium concentration distributions. The true distributions are
then used to project bin classifications for filtered systems.
Observed Cryptosporidium Occurrence
Exhibit 4.13 summarizes the observed Cryptosporidium occurrence of the ICR and ICRSS
studies. See section 4.2.1 for a description of the study, laboratory methods, and method used to count
oocysts. Additional data on observed occurrence are available in the Occurrence and Exposure
Assessment (USEPA 2003c). The data shown for "total oocysts" were used to generate the modeled
Cryptosporidium distributions shown below.
Exhibit 4.13: Summary of Observed Cryptosporidium Total Oocyst Occurrence in
Source Water—Filtered Plant Data
Data Set
ICR
ICRSSL
ICRSSM
Total
Number of
Plants
350
40
40
Number of
Plants with at
Least One
Positive
Sample
(Percent)
154(44%)
34 (85%)
34 (85%)
Observed Plant-Mean Data
(oocysts/L)
Mean
0.066
0.040
0.080
Median
0.000
0.020
0.020
90th
Percentile
0.190
0.100
0.110
Notes: Total oocysts include non-empty and empty oocysts. Non-empty oocysts include oocysts with
internal structures and with amorphous structures. For each plant, all monthly observations were
averaged over the sampling period (12 months for ICRSSM and ICRSSL and 18 months for ICR) to
produce plant-mean data. The mean, median, and 90th percentile shown summarize plant-mean
Cryptosporidium for all plants.
Source: USEPA 2003c.
Modeled Cryptosporidium Occurrence
Each of the three data sets shown in Exhibit 4.13 was used to fit an occurrence model for filtered
plants using the approach outlined in section 4.2.2. As explained in that section, the modeling produces a
collection of plausible occurrence distributions from each of the three data sets. Exhibits 4.13 through
4.15 summarize the resulting collection of distributions from each of the three models. The solid center
curves represent the mean distribution across a given collection, and the dotted lines give a 90-percent
confidence bound for the true filtered occurrence distribution based on the particular data set.
As outlined in section 5.2.4.1, all three of these models were used, independently, as input to the
benefits modeling. Each model is also used, along with a distribution of lab method recovery rates, to
simulate results of initial LT2ESWTR Cryptosporidium monitoring. The assumed recovery rate
distribution is based on "spiked" sample evaluations of the lab method that will be used for initial
monitoring (EPA Methods 1622/23). The overall result of this Monte Carlo simulation is a predicted
distribution of systems assigned to each LT2ESWTR treatment bin; this distribution serves as input to the
cost model.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.14: Modeled Cryptosporidium Occurrence in Source Water—ICR Data
for Filtered Systems
100%-
80% -
§>
O
o>
a
60% -
40% -
20% -
This graph shows modeled variability
and uncertainty in source water
Cryptosporidium occurrence for filtered
systems. It summarizes 1,000 plausible
curves used in the Economic Analysis.
Ninety percent of the modeled curve values
fall within the dashed-line uncertainty
bounds. The heavy center line reflects the
central tendency across all 1,000 curves.
Among the 1,000 modeled curves, any single
curve describes variability in occurrence
across all filtered systems. Steeper curves
indicate less variability from system to system.
Uncertainty —•*,
• Variability •
Example:
Modeled Percent of Plants with
Average Source Water
Cryptosporidium Concentration
Below 0.01 oocysts/L:
- 5th %tile Bound = 28%
— Median Estimate = 23%
- 95th %tile Bound = 18%
5th/95th %tile
^^— Median
1e-005 0.0001 0.001 0.01 0.1 1
Plant Mean Cryptosporidium Concentration (Total oocysts/L)
10
Source: USEPA2003c.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.15: Modeled Cryptosporidium Occurrence in Source Water—ICRSSM
Data
100% -
80% -
g> 60% -
Q_
60% -
0)
B
ID 40%
0 20% -
0% -
1e-005 0.0001 0.001 0.01 0.1 1 10
Plant Mean Cryptosporidium Concentration (Total oocysts/L)
Source: USEPA2003c.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.17: Comparison of Modeled Cryptosporidium Occurrence in Source
Water by Data Set, Median Curves Only
100% -
80% -
g> 60% -
-------
Exhibit 4.18: Summary of Modeled Cryptosporidium Total Oocyst Occurrence in
Source Water—Median Curves Only
Data Set
Total
Number of
Plants
Modeled Plant-Mean Data
(oocysts/L)
Mean
Median
90th
Percentile
Filtered
ICR
ICRSSL
ICRSSM
350
40
40
0.57
0.09
0.19
0.048
0.045
0.05
1.3
0.24
0.33
Unfiltered
ICR
12
0.01
0.01
0.033
Notes: Total oocysts include non-empty and empty oocysts. Non-empty oocysts include
oocysts with internal structures and with amorphous structures.
Source: USEPA 2003c.
4.5.4 Finished Water Cryptosporidium Occurrence for Filtered Plants
Pre-LT2ESWTR finished water Cryptosporidium concentrations (presented in this section) will
be compared with predicted post-LT2ESWTR finished water levels (presented in Chapter 5) to assess the
benefits of the regulatory alternatives. As with the treatment characterization presented above, the
finished water Cryptosporidium occurrence estimates must account for improvements in finished water
concentrations predicted to result from implementation of other existing rules or rules under development
for implementation prior to LT2ESWTR. This section describes the methodology for predicting
Cryptosporidium finished water occurrence based on the treatment in place following the IESWTR,
LT1ESWTR, Stage 1 DBPR, and Stage 2 DPBR.
Compliance with the IESWTR, LT1ESWTR, and the Stage 1 and 2 DBPRs requires or will
require some systems to modify their treatment processes to improve the removal or control the formation
of DBFs. Other rules are not expected to have an appreciable impact because they affect mainly ground
water systems. EPA used the EAs from the IESWTR and the LT1ESWTR, along with more recent plant
performance data, to predict the number of systems in each size category that have made or will need to
make treatment modifications and the effectiveness of those modifications. (These rules have not been in
place long enough for all plants actually to have implemented the modifications or to have gathered
information on their effectiveness.) The IESWTR and the LTIESWTR establish filtration requirements
that EPA believes provide finished water having a minimum 2 log (99 percent) reduction in
Cryptosporidium concentrations relative to source water levels. Although systems are required to meet
only the 2 log removal target, it is recognized that most systems will achieve greater levels of
Cryptosporidium removal. For example, some systems have unit processes in addition to conventional
treatment that provide higher Cryptosporidium removals. To capture the different levels of treatment,
EPA estimated a range of Cryptosporidium removals in order to calculate finished water occurrence.7
7 The plants with MF/UF are assumed to achieve 5.5 log Cryptosporidium removal. Those predicted to
have it in place prior to the LT2ESWTR were subtracted from the monitoring and treatment baselines since they will
be exempt from those requirements. Consequently, their high removal capabilities were not included in the finished
water analysis.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.19 shows the range of Cryptosporidium log reduction that treatment plants are expected
to achieve just prior to implementation of the LT2ESWTR. To account for systems using conventional
treatment and minimally meeting the IESWTR and LT1ESWTR effluent turbidity requirements, the low
end of the range is set at 2.0 log. This reduction is based on the requirements of the IESWTR and
LT1ESWTR (which specify 2.0 log removal) and on several studies that show that plants can achieve 2
log Cryptosporidium removal even under stressed conditions. These studies are described in Chapter 7 of
theLT2ESWTR Occurrence and Exposure Assessment (USEPA 2003c). As described in section 4.5.1,
many plants are expected to be eligible for 0.5 log additional treatment credits for additional unit
processes or for achieving effluent turbidity below 0.15 NTU. For these systems, the range of
performance was shifted up by 0.5 log to reflect the improved performance. Exhibits 4.19 and 4.20 show
the triangular distributions for both the standard estimate (those minimally meeting IESWTR and
LT1ESWTR effluent and turbidity requirements) and the estimate with 0.5 log reduction credit.
Based on the studies cited in the Occurrence and Exposure Assessment, large systems under the
standard estimate are thought to achieve a maximum Cryptosporidium log reduction of 4.5. Many factors
can negatively affect the reductions that systems are able to achieve, and these factors are thought to
impact small systems more significantly. For example, smaller systems typically have fewer filters than
large ones. One consequence is that a single filter performing poorly is more likely to cause poorer
overall system performance. Backwashing a filter in a small system with few filters is more likely to
cause hydraulic fluctuations that could result in poorer performance of the other filters. Small systems
also tend to have less automated control and monitoring equipment, which makes controlling temporary
aberrations more difficult. For this reason, the maximum reduction that small systems are estimated to
achieve is smallerthan that achieved by large systems (for the standard estimate, 3.5 log vs. 4.5 log). For
systems eligible for an additional 0.5 log reduction credit, the top end of the range is increased to 4.0 log
for small systems and 5.0 log for large systems.
Exhibit 4.19: Predicted Ranges of Cryptosporidium Reduction Pre-LT2ESWTR
System Size
(Population Served)
Small (<1 0,000)
Standard Estimate
Large (> 10,000)
Standard Estimate
Small (<1 0,000)
Estimate w/ 0.5 Log
Removal Credit
Large (> 10,000)
Estimate w/ 0.5 Log
Removal Credit
See
Exhibit
4.20a
4.20b
4.20c
4.20d
Range of
Log
Reduction
2.0-3.5
2.0-4.5
2.5-4.0
2.5-5.0
Mode —
Lower End
2.25 log
2.5 log
2.75 log
3.0 log
Mode —
Higher End
2.75 log
3.0 log
3.25 log
3.5 log
Source: Chapter 7, LT2ESWTR Occurrence Assessment (USEPA 2003c).
The log reduction values at the low and high ends of the ranges are thought to be the exception
more than the rule. The studies cited in the Occurrence and Exposure Assessment noted median and/or
average log reduction results that fall in the middle of the ranges shown in Exhibit 4.19. Although little
specific information is available on how often these values occur, EPA believes, on a national scale, that
the values follow central tendency. Therefore, modes of 2.5 to 3.0 log were chosen for the distribution of
Economic Analysis for the LT2ESWTR
December 2005
-------
log reductions for large systems under the standard estimate (see second row of Exhibit 4.19). Values in
between 2.5 and 3.0 log are thought to have the same possibility of being modes.
As described above, small systems are more likely to have filtration performance problems. The
modes at small systems are, therefore, assumed to be lower than for large systems—2.25 and 2.75 logs
under the standard estimate (first row of Exhibit 4.19). The bottom half of Exhibit 4.19 shows the
estimated modes for large and small systems for systems getting a 0.5 log reduction credit, which was
also applied to the modes.
This EA assumes the distribution of log reduction is triangular. (When there is little information
available regarding the distribution of data, triangular distributions are commonly used.) In Exhibit 4.20,
the minimum and maximum of the range define the base of the triangle and the mode defines the top.
Each graph in Exhibit 4.20 corresponds to the range of one of the rows in Exhibit 4.19. Two triangles are
shown for each range, illustrating the lower and upper modes. The triangles defined by all the modes in
between are not shown, but are indicated by the two-way arrow on each graph. Each possible triangle, of
which there is an infinite number, is considered equally likely to represent the log reduction distribution
for a given system size category and treatment scenario. For example, Exhibit 4.20a shows two triangular
distributions for small systems under the standard treatment estimate, which, as described in Exhibit 4.19,
have a range of 2.0 to 3.5 log reduction and modes of 2.25 and 2.75 log. For the same group of systems,
there are an infinite number of triangles with the same range and with modes between 2.25 and 2.75 log.
Taken together, the collection of all the distributions for a given group of systems (e.g., small systems
without a 0.5 log reduction credit) reflects the uncertainty about the true Cryptosporidium log reduction.
These triangular distributions are used in benefit and cost modeling. For a given model iteration
for a given system size, the model first decides whether systems get the 0.5 log reduction credit (36
percent of small systems, 55 percent of medium systems, and 58 percent of large systems are assumed to
qualify). Then it randomly selects a mode from the appropriate set of triangular distributions. From the
individual distribution associated with that selected mode, the model randomly picks 100 log removal
values from that distribution and uses these log reduction values to predict finished water
Cryptosporidium concentrations. This process is repeated for 250 modes.
Exhibit 4.21 shows an example of the finished water concentration distributions generated with a
modeling process similar to that described above (the curves in 4.20 were generated choosing only 250
points from each triangular distribution). These distributions were generated using the same process as
the source water Cryptosporidium concentrations for different data sets in Exhibit 4.17 so that the
differences between them would be directly comparable. Because disinfection in unfiltered systems is
expected to have a negligible effect on Cryptosporidium concentrations, finished water concentrations in
unfiltered plants are assumed to be identical to source water concentrations (compare Exhibits 4.16 and
4.20). For filtered plants, however, modeled source water concentrations were higher than those for
unfiltered plants (see Exhibit 4.17); but modeled finished water concentrations were well below those
same unfiltered plant concentrations.
Economic Analysis for the LT2ESWTR 4-37 December 2005
-------
Exhibit 4.20a: Distribution of Cryptosporidium
Reduction in Small Systems Pre-LT2ESWTR,
Standard Estimate
Exhibit 4.20b: Distribution of Cryptosporidium
Reduction in Large Systems Pre-LT2ESWTR,
Standard Estimate
Exhibit 4.20c: Distribution of Cryptosporidium
Reduction in Small Systems Pre-LT2ESWTR,
Estimate With 0.5 Log Reduction Credit
Exhibit 4.20d: Distribution of
Cryptosporidium Reduction in Large
Systems Pre-LT2ESWTR, Estimate With 0.5
Log Reduction Credit
Log Removals
Source: USEPA 2003c.
Economic Analysis for the LT2ESWTR Proposal
4-38
December 2005
-------
Exhibit 4.21 a: Predicted Finished Water Cryptosporidium Occurrence
Pre-LT2ESWTR, Small Systems
1.00E-10 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01
Finished Water Cryptosporidium (Oocysts/L)
Exhibit 4.21 b: Predicted Finished Water Cryptosporidium Occurrence
Pre-LT2ESWTR, Large Systems
1.00E-10 1.00E-09 1.00E-08 1.00E-07 1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00 1.00E+01
Finished Water Cryptosporidium (Oocysts/L)
Source: USEPA2003c
Economic Analysis for the LT2ESWTR
December 2005
-------
4.5.5 Comparison of EPA Finished Water Cryptosporidium Estimates with Aboytes et al. (2000)
A study by Aboytes et al. (2004) provides an alternative perspective on Cryptosporidium
occurrence in finished drinking water and the efficacy of treatment. This study involved collecting 100-L
finished water samples monthly from 82 surface water utilities. Samples were analyzed for infectious
Cryptosporidium parvum with a cell culture-PCR method. The objective of the study was to "assess the
adequacy of treatment to protect against infectious Cryptosporidium in drinking water." All utilities in
the study were enrolled in the Partnership for Safe Water, a voluntary cooperative program that seeks to
optimize treatment plant performance. Most samples had turbidities below 0.1 NTU and all were below
0.3 NTU, the standard set by the IESWTR. Their sampling detected infectious Cryptosporidium in 22 of
the 82 plants; and in 24 of the 1,690 samples of 100 L. The authors determined an average CC-PCR
recovery efficiency of 32.3 percent. Hence, if it is assumed that one infectious oocyst accounted for each
positive sample, and the oocyst count is adjusted for average recovery, these results produce a mean
concentration of infectious oocysts of 4.4 x 10"4 oocysts/L, or 0.044 oocysts/100 L.
To compare results from Aboytes et al. with EPA's finished water Cryptosporidium estimates
based on results from the ICR and ICRSS, it is necessary to consider the fraction of oocysts that are
infectious. Because oocysts lose viability in the environment, it is expected that infectious oocysts are
only a small fraction of the total number of oocysts in a water sample. While the CC-PCR method
registers only infectious oocysts, the ICR Method and EPA Methods 1622/23 count total oocysts without
regard to whether they are viable and infectious. To estimate the fraction of oocysts that may be
infectious, EPA evaluated a study by LeChevallier et al. (2003) that analyzed several hundred source
water samples from six utilities using both the CC-PCR method and Method 1623. Oocysts were
detected in 60 of 593 samples (10.1 percent) by Method 1623 and infectious oocysts were detected in 22
of 560 samples (3.9 percent) by the CC-PCR procedure. Recovery efficiencies as determined with spiked
samples for the two methods were similar. According to the authors, these results suggest that
approximately 37 percent of the oocysts detected by Method 1623 were viable and infectious. Based on
these results, as well as consideration of the structure of counted oocysts, EPA assumes that the fraction
of oocysts that is infectious may average from 15 to 25 percent for the ICR and 30 to 50 percent for the
ICRSS (described in Section 5.2.4).
If the estimates of mean, large plant, finished water oocyst concentrations from Exhibit 4.21b are
multiplied by 40 percent to adjust for the fraction of oocysts that are infectious, the mean finished water
concentrations of infectious oocysts are as follows:
ICR Mean = 6.1 x IQ'5; ICRSSM Mean = 2.5 x 1Q'5; ICRSSL = 1.4 x 10'5 (oocysts/L)
Thus, the mean finished water infectious oocyst level reported in Aboytes et al. of 4.4 x 10~4
oocysts/L is a factor of 7 greater than the EPA mean estimate based on the ICR and a factor of 30 greater
than the ICRSSL estimate. While the reason for this significant discrepancy is unknown, it cannot be
fully attributed to potential error in factors such as the fraction of oocysts that are infectious. Rather, it
may indicate that the ICR and ICRSS underestimate source water Cryptosporidium occurrence and/or that
EPA has overestimated treatment efficacy, as discussed below.
Unfortunately, source water data are not available for plants in the Aboytes et al. study, so it is
not possible to determine directly their average treatment efficiency. However, removal efficiency may
be estimated based on other survey data. The ICRSS was conducted during a time frame similar to
Aboytes et al., and the filtration and separation steps used in the CC-PCR method of Aboytes et al. are
similar to those in Method 1622/23 used in the ICRSS. Consequently, the ICRSS source water data may
be somewhat comparable to the Aboytes et al. finished water data. The mean source water
Cryptosporidium concentration of plants in the ICRSS, adjusted for average recovery of 43 percent, was
0.14 oocysts/L. If this value is multiplied by 40 percent as an estimate of the fraction of oocysts that are
infectious, the result is a mean source water concentration of 0.056 infectious oocysts/L. If this were the
mean source water concentration in the Aboytes et al. study, then the plants in that survey achieved an
Economic Analysis for the LT2ESWTR 4-40 December 2005
-------
average removal of 1.7 log to produce the mean finished water concentration of 0.0011 oocysts/L that
was measured. For the plants in the Aboytes et al. survey to have achieved a mean oocyst removal of 2.5
log, which was the lowest mean removal assumed in EPA estimates, the mean source water oocyst
concentration would have to have been over twice that measured during the ICRSS. Either hypothesis
indicates that EPA may have underestimated the risk from Cryptosporidium in drinking water by
underestimating finished water oocyst concentrations.
4.5.6 Predicted Bin Classification for Filtered Plants
Under the LT2ESWTR, filtered plants will be assigned to a treatment bin based on the results of
source water monitoring for Cryptosporidium. Each bin is defined by a range of oocyst concentrations
and determines the amount of treatment each plant must provide. Using the modeled source water
occurrence distributions described in section 4.5.3, EPA predicted the percentage of filtered plants that
will fall into each bin. Appendix B presents the probability functions for the occurrence distributions
used to evaluate bin classification. Note that the tables in Appendix B estimate the percentage of systems
in each bin based on a prediction of the test measurement mean, which is the estimated underlying source
water concentration adjusted for sampling and laboratory analysis error. However, bin assignments in
this analysis are based on the predicted lab results, not on estimates of the "true" concentration. Exhibit
4.22 presents the plant bin assignments based on a Monte Carlo evaluation of the probability distributions
for the three data sets of Cryptosporidium occurrence for the Preferred Regulatory Alternative. Note that
the estimate based on the ICR data set is the most conservative of the three filtered plant data sets; in
other words, the predicted percentage of plants requiring treatment is largest for this data set.
The percentages shown in Exhibit 4.22 were used to determine costs of installing treatment for
each occurrence distribution. Plants were expected to incur different costs depending on the bin to which
they were assigned.
Exhibit 4.22: Predicted System Bin Assignments for Preferred Alternative
Occurrence
Dataset
ICR
ICRSSL
ICRSSM
Source Water
Distribution
5th Percentile
Mean
95th Percentile
5th Percentile
Mean
95th Percentile
5th Percentile
Mean
95th Percentile
Bins and Concentrations
No Action
<0.075 oocysts/L
67.6%
65.2%
61.1%
82.3%
77.6%
74.4%
76.0%
72.8%
70.1%
Log 1.0 Removal
(0.075-1 oocysts/L)
26.5%
27.2%
30.4%
17.5%
21.8%
24.7%
22.9%
25.4%
27.5%
Log 2.0 Removal
(1-3 oocysts/L)
3.7%
4.4%
4.8%
0.2%
0.5%
0.8%
0.9%
1 .4%
1.9%
Log 2.5 Removal
(>3.0 oocysts/L)
2.1%
3.2%
3.8%
0.0%
0.1%
0.1%
0.2%
0.4%
0.6%
Note: Bin assignment is based on the highest running annual average (RAA) of concentrations of 24 influent samples
taken over 2 years.
Source: Appendix B, Exhibits B.8 and B.12.
4.6 Baseline for Uncovered Finished Water Reservoirs
Number of Reservoirs
The LT2ESWTR baseline for uncovered finished water reservoirs is presented in Exhibit 4.23.
EPA regional offices provided data on these reservoirs based on information from States in their region;
the data do not include reservoirs that are scheduled to be covered or taken offline. Most such reservoirs
Economic Analysis for the LT2ESWTR
December 2005
-------
are located in only a few States or Territories: California, New York, New Jersey, Oregon, and Puerto
Rico. The largest is in Southern California. Although there are only 81 uncovered finished water
reservoirs, there is a limited amount of information on them. Most systems only reported reservoir
volume. Very few reported surface area or daily flows, forcing EPA to make assumptions to estimate
those parameters (both quantities are needed to estimate costs for covering the reservoirs or treating the
water they discharge). Populations served by each reservoir were also often poorly documented.
For this analysis, the reservoirs are categorized according to usable volume (in millions of
gallons, or MG). The mean volume, shown in the last column of Exhibit 4.23, is the average usable
volume of all reservoirs in a volume category and is used to estimate costs in Chapter 6.
Exhibit 4.23: Baseline Numbers of Uncovered Finished Water Reservoirs
Size Category (MG)
0-0.1
> 0.1 - 1
> 1 -5
>5- 10
> 10-20
> 20 - 40
> 40 - 60
> 60 - 80
>80- 100
> 100- 150
> 150-200
> 200 - 250
>250- 1000
> 1000
Total
Number of
Uncovered
Reservoirs
A
3
9
10
4
12
5
10
7
3
6
1
4
6
1
81
Mean Volume
(MG)
B
0.093
0.478
3.165
8.000
15.200
28.080
51.422
67.843
94.000
127.255
179.000
208.500
694.679
3,313.718
Source: EPA regions.
Surface Area, Average Daily Flow, and Design Flow of Uncovered Reservoirs
The cost of covering a reservoir is based largely on its surface area, while costs for disinfecting
the discharge are based on flow through the reservoir. As with technologies used for Cryptosporidium
removal, the design flow and average daily flow are used to estimate capital and O&M costs of reservoir
treatment, respectively. The purpose of this section is to derive mean surface area, average daily flow,
and design flow for each size category of uncovered finished water reservoir.
EPA regions provided the surface area for some individual reservoirs. Where this information
was not available, engineering assumptions were necessary. To calculate surface area, a representative
reservoir depth of 25 feet was assumed based on consultation with industry engineers. The mean surface
area for each size category presented in Exhibit 4.24 is the average surface area of all reservoirs in the
size category, whether based on actual data or based on volume with an assumed 25-foot depth.
Economic Analysis for the LT2ESWTR
December 2005
-------
Hydraulic residence time, defined as the time water spends inside a reservoir, can be used in
conjunction with volume to estimate average daily flow through the reservoir. Hydraulic residence times
in finished water storage reservoirs vary greatly across systems and seasons. The shortest times are often
in the summer, while the longest may be during lower-demand periods in the winter. Because water
systems strive to maintain a certain volume of storage in the distribution system for emergencies such as
fire, residence times can be as great as several weeks. Within the last several years water systems have
decreased average residence time to improve the quality of the water in the distribution system, striving to
turn over finished water in storage facilities on a regular basis. Considering these factors, typical
hydraulic residence times of 1 to 3 days were used in conjunction with volume to estimate average daily
flow for reservoirs up to 100 million gallons in size, as presented in Exhibit 4.24. For the very largest
reservoirs, longer hydraulic residence times are more likely; residence times as high as 21 days are
assumed. (Appendix I describes the available data and provides the rationale for assumptions.)
Design flow, used to estimate capital costs, represents the maximum possible flow exiting the
reservoir. Finished water storage facilities respond to daily water demand fluctuations in the distribution
system in order to maintain as constant as possible a flow at the treatment plant. Therefore, maximum or
design flow from a reservoir may be higher than the average flow at the treatment plant. For hydraulic
modeling purposes, peak hourly flows in a distribution system have been estimated as three times the
average flow (Lindeburg, 1997). Therefore, the design flow presented in Exhibit 4.24 is estimated to be
three times the average daily flow.
Exhibit 4.24: Surface Area and Flows for Uncovered Finished Water Reservoirs
Size Category (MG)
0-0.1
>0.1 -1
>1 -5
>5-10
>10-20
> 20 - 40
> 40 - 60
> 60 - 80
> 80 -100
> 100 -150
> 150 -200
> 200 - 250
> 250 -1000
>1000
Total
Number of
Uncovered
Reservoirs
A
3
9
10
4
12
5
10
7
3
6
1
4
6
1
81
Mean Volume
(MG)
B
0.093
0.478
3.165
8.000
15.200
28.080
51 .422
67.843
94.000
127.255
179.000
208.500
694.679
3,313.718
Mean Surface
Area (ft.2)
C
500
2,555
16,924
42,778
81 ,278
150,150
274,964
362,772
502,641
680,461
957,156
1,114,900
3,714,615
17,719,245
Estimated Average
Hydraulic Residence
Time (day)
D
1.00
1.00
1.00
2.00
3.00
3.00
3.00
3.00
3.00
4.00
4.00
4.00
14.00
21.00
Average
Flow (MGD)
E = B/D
0.09
0.48
3.17
4.00
5.07
9.36
17.14
22.61
31.33
31.81
44.75
52.13
49.62
157.80
Design
Flow (MGD)
F =3*E
0.28
1.43
9.50
12.00
15.20
28.08
51.42
67.84
94.00
95.44
134.25
156.38
148.86
473.39
Sources: [A] and [B] EPA regions.
[C] EPA regions for some reservoirs. Surface area, if not provided for an individual reservoir, is estimated
based on data on volume (from EPA regions) and an assumed depth of 25 feet.
[D] Professional judgment.
Economic Analysis for the LT2ESWTR
4-43
December 2005
-------
System Sizes for Uncovered Reservoirs
In order to apportion the costs to a specific size PWS for the costs to the rule and
household cost analyses, EPA needed to estimate a population served for each reservoir. Many
reservoirs are only one of several in a system, and in such cases, there is no basis to determine
what percentage of the full flow might pass through a single reservoir. Therefore, using the flow
equations described in section 4.3.3, population served by each reservoir was estimated and the
reservoirs were allocated to a system size based on those estimates. Exhibit 4.25 shows the
distribution of uncovered filtered water reservoirs by system size. To obtain the number of
households served by each reservoir for use in the household costs analysis, the number of
reservoirs is multiplied by the average population per system for that size category and divided by
2.59, the number of people per household (U.S. Census Bureau 2001). Appendix I details the
methodology for estimating reservoir modeling parameters.
Exhibit 4.25: Baseline Number of Uncovered Finished Water Reservoirs in
Each System Size Category
system Size
(Population Served)
<100
100-499
500 - 999
1,000-3,299
3,300-9,999
10,000-49,999
50,000 - 99,999
100,000-999,999
=1,000,000
Total
Number of
Reservoirs
3
0
0
0
9
26
5
37
1
81
Source: Appendix I
4.7 Households Incurring Costs Due to the LT2ESWTR
Estimating the number of households served by systems affected by the LT2ESWTR is necessary
to derive national per household annual costs. Exhibit 4.26 shows the possible combinations of the
various rule activities that systems may conduct, and thus households may incur additional cost. For
example, a filtered system may conduct monitoring and fall in a no-action bin for additional
Cryptosporidium treatment. The costs those households incur would reflect implementation and
monitoring activities. However, an filtered system with an uncovered finished water reservoir that falls
into an action bin would incur costs associated with implementation, monitoring, additional
Cryptosporidium treatment, and the uncovered reservoir.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.26: Universe of Households Affected by Rule Provisions
Implementation Costs
(All Households)
1
r
Implementation
V J
1
'
Monitoring Costs
i
r
X" "X
Monitoring
+ Implementation
v J
i
r
Uncovered Finished
Water Reservoir Cost
i
' i
/" 'N
Uncovered
Reservoir
+ Monitoring
+ Implementation
L J
\
• J
>
r T^ii i >
Filtered
Treatment
+ Uncovered
Reservoir
+ Monitoring
^f Implementation
i
1
Treatment Costs
/
V
r Unfiltered ^
Treatment
+ Uncovered
Reservoir
+ Monitoring
^ Implementation
i
' 1 '
c ^\ c ^\
Filtered Unfiltered
Treatment Treatment
+ Monitoring + Monitoring
+ Implementation + Implementation
V J V J
Exhibit 4.27 presents estimates of the number of households served by surface and GWUDI
systems, further subdivided by those subject to the various rule provisions. The estimates were derived
by dividing the total population served by CWSs subject to each rule provision within each size category
by an average number of people per household of 2.59 (U.S. Census Bureau 2001). Only the CWS
population is used to estimate households because it is assumed that only CWSs serve residential
customers. Exhibit 4.28 shows the number of households subject to uncovered reservoir costs.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.27: Baseline Numbers of Households Incurring Costs
System Size
(Population
Served)
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
Households
Paying
Implementation
Costs
A
7,925
77,064
122,745
840,009
2,569,775
10,001,463
7,046,149
29,347,697
20,195,053
70,207,880
Households
Paying
Monitoring
Costs
B
7,640
74,289
118,326
809,769
2,477,263
9,964,828
7,020,339
29,240,196
20,121,079
69,833,729
Households Paying
Treatment Costs for
Unfiltered Systems
C
30
460
953
9,632
42,493
110,321
103,902
567,851
3,173,681
4,009,322
Households Paying
Treatment Costs for
Filtered Systems
D
6,423
68,997
109,699
786,300
2,389,798
9,709,985
6,815,100
28,252,742
16,709,626
64,848,670
Notes: Includes households served by surface water or GWUDI systems only. All unfiltered systems will incur
treatment costs. Not all filtered systems shown above will incur treatment costs—the households shown in Column D
represent all systems that conducted monitoring less systems that are predicted to have installed MF/UF to comply
the the Stage 2 DBPR. Some of these systems will be assigned to the "no action" bin and will not incur treatment
costs.
Sources: [A] SDWIS population from unlinked inventory (USEPA 2003e) / 2.59 people per household (U.S. Census
Bureau 2001).
[B] Implementation baseline (Column A) * percentage of systems subject to monitoring requirements (1 - Exhibit 4.11,
Column C).
[C] Treatment baseline population for unfiltered systems (Exhibit 4.5, Column D) / 2.59 people per household.
[D] Treatment baseline population for filtered systems (Exhibit 4.12, Column E) 72.59 people per household.
Exhibit 4.28: Households Paying Treatment Costs for Uncovered Reservoirs
System Size
(Population Served)
<100
1 00 - 499
500 - 999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
=1,000,000
Total
Number of
Households
63
0
0
0
20,159
217,257
106,926
2,774,523
267,187
3,386,115
Source: Appendix I.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.29 presents the estimated mean water usage rates (in gallons per year) per household for
each system size category. These estimates are based on total residential consumption and the number of
residential connections from the Baseline Handbook. The consumption in the two smallest size
categories was adjusted based on analyses of 1995 CWSS data.
Exhibit 4.29: Mean Household Water Usage Rates by System Size
System Size
(Population
Served)
<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 Mil
>1 Mil
Mean
Household
Water Usage
Rate (gal/yr)
83,000
83,000
104,000
87,000
97,000
109,000
119,000
125,000
125,000
Source: USEPA2001c,
systems serving 500 or
on CWSS data.
modified for
fewer people based
4.8 Summary of Uncertainties in Development of LT2ESWTR Baselines
Uncertainties in this baseline analysis could result in either an overestimate or an underestimate
of the costs or benefits presented in Chapters 5 and 6. Exhibit 4.30 below presents an estimate of the
effects that each source of uncertainty may have on subsequent analyses. Note that, in many cases,
assumptions made in this baseline will overestimate both costs and benefits; however, costs are
overestimated in more cases than are benefits.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 4.30: Summary of Uncertainties Affecting LT2ESWTR Baseline Estimates
Assumption
Uncertainty in
baseline data
inputs (SDWIS,
ICR, and ICRSS
data)
Point estimates
instead of
distributions for
population and flow
CWS flow
equations for
NTNCWSs
No unfiltered plants
have advanced
disinfection
Filtered plants do
not get credit for
ozone and CIO2
from Stage 2
Predicted Pre-
LT2ESWTR
Cryptosporidium
removal using
triangular
distributions (with
uncertain modes)
and log reduction
achieved
Uncertainty in
baseline surface
area and flows for
uncovered finished
water reservoirs
Section with
Discussion
of
Uncertainty
4.3.2
4.4.2
4.3.3
4.3.3
4.4.1
4.5.1
4.5.1
4.6
Effect on Benefit Estimates
Under-
estimate
Over-
estimate
X
X
Under-
or Over-
Estimate
X
X
X
X
Effect on Cost Estimates
Under-
estimate
Over-
estimate
X
X
X
Under-
or Over-
Estimate
X
X
X
X
Note: The uncertainties associated with some assumptions are discussed in more detail in Chapters 5 and 6, and so
the summaries of those assumptions are reserved until the ends of those chapters. Those key assumptions include
the occurrence in source water and modeling of occurrence in finished water; risk parameters, such as infectivity and
the percent of viable oocysts; and binning assignments.
Economic Analysis for the LT2ESWTR
4-4
December 2005
-------
5. Benefits Analysis
5.1 Introduction
The LT2ESWTR will reduce the occurrence of viable waterborne pathogens, particularly
Cryptosporidium, in drinking water delivered by public water supplies that use surface water or ground
water under the direct influence of surface water (GWUDI). The quantified health benefits estimated for
this rule result from reducing the incidence of adverse health effects (illnesses and possible premature
death) caused by drinking water containing Cryptosporidium. This rule is also expected to reduce health
effects associated with other pathogens.
Section 5.2 describes the risk assessment used to estimate the number of illnesses and deaths
associated with endemic cryptosporidiosis that will be avoided because of the LT2ESWTR. Section 5.3
presents the methods used to monetize these benefits. Uncertainty and variability are inherent in any risk
assessment. In this EA, stochastic distributions and sensitivity analyses are used to account for
uncertainty and variability. The resulting estimates are shown as mean values with 90 percent confidence
bounds. Sections 5.2 and 5.3 also discuss supporting data for each variable used in the risk assessment
and monetization, with citations to additional information in appendices. Section 5.4 summarizes all
areas of uncertainty in the benefits analysis, noting how they are addressed and the likely effect on the
national estimate.
The quantified benefits presented in this chapter are not the only benefits expected from the
implementation of this rule. Other benefits, including those associated with reductions in sporadic
cryptosporidiosis outbreaks, reductions in other endemic illnesses and outbreaks from other pathogens,
and improved aesthetic water quality, are described in section 5.6, but are not quantified in this analysis.
In addition, the benefits realized from regulating uncovered finished water reservoirs, a provision of the
proposed rule, are not quantified.
The remainder of this chapter is organized as follows.
5.2 Quantified Health Benefits from Reduction in Exposure to Cryptosporidium
5.2.1 Overview of Risk Assessment Methodology
5.2.2 Hazard Identification
5.2.3 Dose-Response Assessment
5.2.4 Exposure Assessment
5.2.5 Risk Model Structure
5.2.6 Individual Annual Risk Distributions
5.2.7 General Population Risk—Number of Cases Avoided
5.2.8 Reduction in Sensitive Subpopulation Risk
5.3 Monetized Benefits from Reduction in Exposure to Cryptosporidium Resulting from the
LT2ESWTR
5.3.1 Value of Reduction in Cryptosporidiosis Cases
5.3.2 Monetization of Benefits to Sensitive Subpopulations
5.4 Summary of Uncertainties
5.5 Comparison of Regulatory Alternatives
5.6 Other Benefits of LT2ESWTR Provisions
5.6.1 Reduction in Outbreak Risk
5.6.2 Costs to Households to Avert Infection
5.6.3 Enhanced Aesthetic Water Quality
5.6.4 Risk Reduction from Co-occurring and Emerging Pathogens
Economic Analysis for the LT2ESWTR 5-1 December 2005
-------
5.6.5 Benefits from Other Rule Provisions
5.6.6 Summary of Nonqualified Benefits
5.2 Quantified Health Benefits from Reduction in Exposure to Cryptosporidium
This section describes the risk assessment methods and assumptions used to quantify the expected
health benefits of the LT2ESWTR associated with reduced exposure to Cryptosporidium. It also provides
the results of these calculations, expressed in terms of reduced cases of illness and avoided deaths.
The quantified health benefits presented in sections 5.2 and 5.3 are derived from estimates of the
Pre-LT2ESWTR annual levels of illness and death caused by endemic exposure to Cryptosporidium in
drinking water, and the reductions expected as a result of the LT2ESWTR. Annual endemic cases are
those occurring as a result of Cryptosporidium present in drinking water under normal operating
conditions. This endemic level does not include illnesses and deaths attributable to outbreaks of
cryptosporidiosis—those that are associated with events or conditions that are outside of normal treatment
plant operating conditions.
Endemic levels of cryptosporidiosis cannot be measured directly because symptoms are generally
underreported (relatively few seek medical attention), and because there are many potential causes of
gastrointestinal illness resembling cryptosporidiosis. Usually only in an outbreak will doctors test stool
samples for Cryptosporidium.
Even outbreaks are not always recognized, again because symptoms are underreported and not
always recognized as being due to cryptosporidiosis. Data on occurrence specifically related to outbreaks
were not available and dependable methods to model the future occurrence of outbreaks have not been
proven. Because of these difficulties, the incidence of illness is modeled (as opposed to directly
measured), and only for endemic illnesses. Thus, the illnesses estimated quantitatively in this Economic
Analysis (EA) should be thought of as representing a steady, underlying level of illness unadjusted for
outbreaks of the disease.
The risk assessment used to calculate the benefits of the LT2ESTWR involves a two-dimensional
Monte Carlo simulation model designed to explicitly consider probability distributions describing the
uncertainty in some of the model inputs, and the inherent variability in others. The structure of this
model, and the basis for the characterization of the uncertainty and variability distributions used in it, are
described in more detail in this chapter. The calculations for the model were carried out in SAS v8.2
(Appendix T provides details of the programming code, input data, and output results).
In addition to the use of probability distributions in the Monte Carlo simulation model, the risk
assessment is designed to compare benefit estimates (reductions in risk) across several key categories:
system type and size, treatment (filtered or unfiltered), and occurrence data sets of Cryptosporidium in
source water (Exhibit 5.1). Presenting the model results across these multiple categories provides
measures of variability beyond the uncertainty and variability derived through the use of probability
distributions within the model. Model runs produce total national benefit estimates as well as breakouts
for each of the categories listed in Exhibit 5.1.
Economic Analysis for the LT2ESWTR 5-2 December 2005
-------
Exhibit 5.1: Risk Assessment Model Categories
Data Set
ICR
ICRSSM
ICRSSL
System Size
(population served)
<100
100-<500
500 - <1 ,000
1,000-<3,300
3,300 -<1 0,000
1 0,000 -<50,000
50,000 -<1 00,000
1 00,000 -<1 Million
> 1 Million
<100
100-<500
500 - <1 ,000
1,000-<3,300
3,300 -<1 0,000
1 0,000 -<50,000
50,000 -<1 00,000
1 00,000 -<1 Million
> 1 Million
<100
100-<500
500 - <1,000
1,000 - <3,300
3,300 -<1 0,000
1 0,000 -<50,000
50,000 -<1 00,000
1 00,000 -<1 Million
> 1 Million
Filtered Plants
CWS
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
NTNCWS
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
TNCWS2
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Unfiltered Plants1
CWS
/
/
/
/
/
/
/
/
/
Note: 1The ICRSSM and ICRSSL have no unfiltered source water occurrence data. Section 5.2.7.1 describes how
unfiltered ICR data were adjusted to produce risk estimates for the ICR, ICRSSM, and ICRSSL data sets.
There is only one unfiltered noncommunity water systems (NCWS); it was grouped with community water
systems (CWSs) for data analysis.
2There are no nontransient and transient noncommunity water systems for the population categories with
shaded boxes.
Sections 5.2.1 through 5.2.5 present the risk assessment methodology. Model results for the
baseline and four regulatory alternatives considered in this Economic Analysis (EA) are presented and
discussed in sections 5.2.6 through 5.2.8, as well as in Appendices C and O.
5.2.1 Overview of Risk Assessment Methodology
Risk assessment is an analytical tool that is used to characterize the expected incidence of adverse
health effects associated with exposure to an environmental hazard, in this case Cryptosporidium. It is
also used to estimate the benefits of actions taken to reduce exposure to that hazard.
Economic Analysis for the LT2ESWTR
5-3
December 2005
-------
The risk assessment used to estimate the potential benefits of the LT2ESWTR comports with a
standard framework for risk assessment employed by the U.S. Environmental Protection Agency (EPA).
The framework is organized in accordance with the EPA Policy for Risk Characterization (USEPA
1995a), EPA's Guidance for Risk Characterization (USEPA 1995b), and EPA's Policy for Use of
Probabilistic Analysis in Risk Assessment (USEPA 1997a).
This standard framework requires the use of scientific data (or reasonable assumptions if data are
not available) to produce estimates of the nature, extent, and degree of a risk. When the risks posed are
not the same for all persons, that variability in risk should be described. Further, data are seldom known
with certainty, and therefore, that uncertainty must be described and its impact on the risk estimates
characterized. The risk assessment used here incorporates both types of information—the variability
associated with the distribution of risk levels within the affected population and the uncertainty expressed
by confidence bounds.
According to the 1995 EPA Policy for Risk Characterization (USEPA 1995a), health risk
assessments for environmental contaminants generally involve four components:
Hazard Identification addresses the nature of the potential adverse health effects associated
with exposure to the contaminant.
Dose-Response Assessment addresses information concerning the relationships, quantitative
where possible, between the magnitude of exposure to the contaminant and the extent and
severity of the adverse health effects.
• Exposure Assessment estimates both the number of people in the population exposed to the
contaminant and the distribution of levels of exposure within that population.
• Risk Characterization combines the hazard identification, dose-response assessment, and
exposure assessment information to describe overall risk to the exposed population, in terms
of both the distribution of risk levels in the population and the total number of cases of
adverse effects expected.
Exhibit 5.2 depicts these elements of the risk assessment for characterizing the endemic risk of
illness and death from exposure to Cryptosporidium in drinking water systems.
To derive benefit estimates of the LT2ESWTR using this risk assessment framework, the analysis
calculates the difference between illness and death estimates for the baseline (Pre-LT2ESWTR) condition
and illness and death estimates after implementation of the LT2ESWTR. Benefit estimates are the
number of illnesses and deaths avoided because of a regulatory requirement (i.e., Pre-LT2ESWTR and
the four regulatory alternatives for LT2ESWTR described in section 3.3).
Economic Analysis for the LT2ESWTR 5-4 December 2005
-------
Exhibit 5.2: Health Risk Assessment Framework for Cryptosporidium
Hazard Identification
• Cryptosporidium health
endpoints: Illness
(cryptosporidiosis) and
Mortality
Dose-Response Assessments
Relationships for probability of:
• Infection given exposure
• Illness given infection
• Death given illness
Exposure Assessment
• Number of people exposed to
Cryptosporidium in finished
drinking water
• Distribution of average daily
Cryptosporidium ingestion
levels across the exposed
population
Risk Characterization
• Estimated cases of illness and
death in the affected population
• Distribution of individual risk
levels of illness and mortality
5.2.2 Hazard Identification
This section presents summary information on the adverse health effects associated with
Cryptosporidium ingestion from drinking water, including a discussion of cryptosporidiosis and the
potential for illness. It also discusses the potential for mortality, particularly among the
immunocompromised. For further information on health effects associated with Cryptosporidium, see
Chapter 2 of this document as well as the Occurrence and Exposure Assessment for the Long Term 2
Enhanced Surface Water Treatment Rule (USEPA 2003c).
Ingesting Cryptosporidium oocysts can cause cryptosporidiosis, which typically is an acute, self-
limiting illness with symptoms that include diarrhea, abdominal cramping, nausea, vomiting, and fever
(Juranek 1995). Cryptosporidiosis can also cause non-gastrointestinal symptoms, such as eye and joint
pain, headaches, dizziness, and fatigue (Hunter et al. 2004). There is no treatment that can eliminate a
Cryptosporidium infection, and only a few antiparasite or antimicrobial agents have shown even a slight
ability to reduce a patient's parasite load (Guerrant 1997). In some occurrences, cryptosporidiosis can be
fatal, particularly among subpopulations such as Acquired Immunodeficiency Syndrome (AIDS) patients,
the elderly with other underlying illnesses, and other immuno-compromised individuals.
Economic Analysis for the LT2ESWTR
5-5
December 2005
-------
Limited information is available on the endemic incidence of cryptosporidiosis in the United
States. Mead et al. (1999) have estimated that there are approximately 15 million physician visits
annually for diarrhea, and that approximately 2 percent of these, or 300,000 cases, are due to
cryptosporidiosis. They also estimate that of these 300,000 cases, only about 10 percent are attributable
to food-borne transmission, with the remainder due to the consumption of contaminated water (from
drinking or recreational exposure) or person-to-person contact. The number of endemic cases estimated
by Mead et al. is probably low because only a fraction of people who experience diarrhea visit a
physician. For example, in the 1993 cryptosporidiosis outbreak in Milwaukee, Wisconsin, medical care
was sought in 12 percent of cases (Corso et al., 1993). Mead et al. estimate that there are approximately
211 million episodes of gastroenteritis in the United States each year, of which only about 38 million are
attributable to known pathogens. Published information does not provide estimates of cases due solely to
drinking water.
Another potential indicator of endemic cryptosporidiosis is the fraction of the human population
that has positive antibodies against Cryptosporidium. Studies of a variety of populations have found a
reactivity to C. parvum antigens in 25 to 35 percent of adults; this number is even higher in developing
countries (Chappell et al. 1999; Frost et al. 1998). While these positive reactions indicate past exposure
to Cryptosporidium, the number of exposures, durations of illness, and individual susceptibilities to re-
infection cannot be determined by simple testing. Levels of one antigen, IgG, have been observed to drop
over time, so a high level of IgG antibodies in an individual can be an indicator of recent exposure or
infection (van Herck et al. 2000; de Melker et al. 2000), but since the rate of decrease is not known and
possibly differs for each individual, it is difficult to estimate endemic rates through immunology.
Many, probably most, infected individuals do not seek medical treatment for their symptoms. If
they do seek medical treatment, primary care physicians may not be able to isolate Cryptosporidium as
the cause of the illness. If diagnosed, physicians may not report the information to the Centers for
Disease Control (CDC). These compounded effects could lead to gross underreporting and
underestimating of cryptosporidiosis cases (Okun et al. 1997). Additionally, individuals can be infected
with Cryptosporidium yet exhibit no symptoms of infection. This is seen more among people without
pre-existing immunity to Cryptosporidium, and can further hide the true incidence of cryptosporidiosis in
the United States.
Although the focus of the risk and benefits analysis conducted here is on endemic cases, most of
the information available on the health hazards from exposure to Cryptosporidium derives from studies
involving outbreaks. The 1993 Milwaukee outbreak was the largest recorded outbreak of waterborne
disease in the United States. Using standard epidemiological methods, CDC estimated that over 400,000
people became ill (Craun et al. 1998). Of those, 4,000 required hospitalization (approximately 1 percent
of those becoming ill), and there were 54 cryptosporidiosis-associated deaths, with at least 46 of these
being immunocompromised individuals (as reported on death certificates) (Mackenzie et al. 1994; Hoxie
etal. 1997).
Several subpopulations may be more sensitive to cryptosporidiosis, including the young, elderly
with other underlying illnesses, malnourished, disease-impaired (especially those with diabetes), and a
broad category of those with compromised immune systems (Rose 1997). There has been little research
in the United States on cryptosporidiosis in children and the disease-impaired. The extent to which the
elderly are more susceptible to cryptosporidiosis is not known; however, during the Milwaukee outbreak
(as well as prior to the outbreak), the rate of gastroenteritis-related emergency room visits among those 65
and over was shown to increase with age. It is not known whether these patients may have had
underlying illnesses or other risk factors that could have affected their immunity to Cryptosporidium.
The elderly were also shown to have a shorter incubation period during the outbreak than the general
adult population (Naumova et al. 2003). However, another study of the Milwaukee outbreak, a random
Economic Analysis for the LT2ESWTR 5-6 December 2005
-------
telephone survey, showed that the elderly had a lower illness rate than the general population (MacKenzie
et al. 1994). The possibility of increased risk of cryptosporidiosis among the elderly population is of some
concern because the elderly population is projected to increase in the coming decades. In the year 2000,
those 65 and over made up 12.4 percent of the U.S. population; that percentage is expected to reach 16.3
percent by 2020 (U.S. Census Bureau 2004).
Subpopulations with compromised immune systems include AIDS patients, those with lupus or
cystic fibrosis, transplant recipients, and those on chemotherapy (Rose 1997). Symptoms in the
immunocompromised Subpopulations are much more severe, including debilitating, voluminous diarrhea
that may be accompanied by severe abdominal cramps, weight loss, malaise, and low-grade fever
(Juranek 1995). Symptoms may also last longer than in those with healthy immune systems. Moreover,
mortality is a substantial threat to the immunocompromised infected with Cryptosporidium:
The duration and severity of the disease are significant: whereas 1 percent of the
immunocompetent population may be hospitalized with very little risk of mortality
(< 0.001), Cryptosporidium infections are associated with a high rate of mortality in the
immunocompromised (50 percent). (Rose 1997).
Exhibit 5.3 contains detailed information on some of the cryptosporidiosis symptoms observed
during the Milwaukee outbreak.
Exhibit 5.3: Symptoms of 205 Patients with Confirmed
Cases of Cryptosporidiosis During the Milwaukee Outbreak
Symptom
Diarrhea
Abdominal Cramps
Weight Loss
Fever
Vomiting
Percent of
Patients
93%
84%
75%
57%
48%
Mean
~12 days duration
N/A
10 pounds
100.9T
N/A
Range
1-55 days duration
N/A
1-40 pounds
99.0°-104.9°F
N/A
N/A: Not applicable.
Source: Mackenzie et al. 1994.
Although the Milwaukee outbreak represents the largest number of cases in a single
cryptosporidiosis outbreak in the United States, most identified outbreaks have occurred in small systems
serving fewer than 10,000 people. Between 1991 and 1996, 6 outbreaks caused by Cryptosporidium in
small water systems resulted in 271 reported cases of cryptosporidiosis and 3,822 estimated cases
(USEPA 2003c). Three of the six outbreaks were in small surface water systems, and three occurred in
GWUDI systems. During small-system outbreaks, the percent of the exposed population becoming ill
ranged from 8 to 70 percent.
Again, outbreak cases are believed to represent only a portion of the total incidence of
cryptosporidiosis. Only large outbreaks cases concentrated in a specific location are likely to be detected
and reported. Endemic cases (which are the focus of this analysis) and smaller outbreaks are less likely to
be identified.
Economic Analysis for the LT2ESWTR
5-7
December 2005
-------
5.2.3 Dose-Response Assessment
This section presents the dose-response model, which characterizes the relationship between
ingestion of Cryptosporidium oocysts and the likelihood of infection, illness, and mortality. This model
is the first of two steps in the overall risk assessment model. The second step, the exposure assessment, is
described in the following section. The specific variables of the dose-response model include the
following:
• Infectivity dose-response function
• Morbidity factor used to compute the risk of illness, given that an infection has occurred
• Mortality factor to compute the risk of death, given that an illness has occurred.
Previous Cryptosporidium risk assessments by Haas et al. (1996), Rose (1997), and Teunis et al.
(2002) have focused on assessing the dose-response relationship and exposure risks. The risk assessment
for waterborne cryptosporidiosis done by Haas et al. (1996) took existing data on Cryptosporidium
infectivity and used an exponential dose-response model to determine the median infectious dose (ID50).
This information was then used to determine a dose-morbidity ratio and the finished water concentration
of oocysts that would be acceptable given guidelines for annual risk of infection from any one type of
pathogen. The infectivity data used were from healthy subjects and were extrapolated to the broader
population.
Cryptosporidium occurs in different species and strains, with varying degrees of infectiousness
to humans. Cryptosporidium samples from several different sources (called isolates) have been collected
and cultivated for use in clinical trials. Two studies by Teunis et al. (2002) examined the dose-response
variation between different isolates of Cryptosporidium and between hosts with varying immune
responses. These studies concluded that for Cryptosporidium, illness and infectivity1 vary among
isolates, and that for individuals with elevated IgG levels, the probability of infection and illness
decreases. The assessment by Rose (1997) added to the previous work by documenting sources of
Cryptosporidium, those populations that are more susceptible to infection, historical cryptosporidosis
outbreaks, and the occurrence of Cryptosporidium in the environment.
Infectivity Dose-Response Model
The primary dose-response model used for the baseline risk assessment, as well as six alternative
dose-response models that were evaluated by EPA are all based upon the same model structure referred
to as the "Exponential Dose Response Model." The basic exponential model has the mathematical form
of:
p _ j _ e -dr
which describes the probability of an exposed individual becoming infected (P:) given an expected dose
(d). The parameter r in this model, as discussed in more detail below, characterizes the likelihood of a
host-organism interaction that will result in infection.
1 A person is considered ill due to Cryptosporidium if they have symptoms of cryptosporidiosis. A person
is considered infected if their stool contains oocysts.
Economic Analysis for the LT2ESWTR 5-8 December 2005
-------
The relatively simple mathematical form of the exponential dose response model can, to some
extent, obscure the fact that it is the result of the combination of two other probability functions: (1) a
Poisson distributed probability of ingesting one or more organisms capable of causing an infection given
the number of infectious organisms present in the water that is consumed, and (2) a Binomially
distributed probability of one or more of the ingested infectious organisms interacting with the host and
causing an infection given the number of organisms ingested.
The mathematical derivation of the exponential model equation as shown above from the
combination of the Poisson and binomial distributions is shown in detail in Haas et al. (1999).
At a somewhat simpler level, the 1 - e~d part of the equation describes the probability that an
individual will ingest at least one organism given an expected dose "d" of infectious organisms ingested,
given the concentration in water and the expected amount of water consumption for an individual. In this
context, d is effectively the product of the concentration of organisms in the water and the volume of
water consumed. As such, it can take on a non-integer value even though an individual can in fact only
ingest 0 or some integer number of organisms (not fractional numbers of organisms). Non-integer
"expected" doses can be used, however, when applying the relationship to populations rather than to a
single individual.
If the dose is 0, then 1 - ed = 0, indicating that the probability of ingesting infectious organisms is
0 (and therefore P: also becomes 0). If the dose is >0, 1 - ed is a value greater than 0 and approaches 1 as
the dose becomes large, indicating a probability approaching certainty that one or more infectious
organisms will be ingested when the expected dose is a high value.
This part of the equation reflects the underlying assumption that the organisms are Poisson
distributed in the water that is ingested. In that regard, it implies that even when the "expected dose" is
some non-zero value, the actual dose ingested by an individual can be some value other than the expected
dose. This can be seen by considering a situation where the "known" concentration is 1 oocyst per liter,
with an assumed ingestion of 1 liter of water. In this case, the "expected" ingestion for any individual is 1
organism. However, the value of 1 - el = 0.632, which is the cumulative Poisson probability that 1 or
more organisms will be ingested 63.2% of the time if the "expected" number ingested is exactly 1. This
also implies that the actual number ingested will be 0 approximately 36.8% of the time even when the
expected number is exactly 1.
As noted above, the parameter "r" in the Exponential Dose Response equation provides
information on the host-organism interactions that lead to infection. As discussed below, it is largely the
treatment of the "r" parameter in the Exponential Dose Response model that differentiates the 6 models in
Appendix N.
The "r" parameter in the Exponential Dose Response model can have values that lie only in the
range of 0 to 1. This is because "r" is effectively a measure of the probability that any infectious
organism that has been ingested will survive and cause an infection in the host, reflecting organism-host
interactions that work in combination to result in that infection. Values of "r" closer to 0 reflect a low
likelihood of infection occurring given ingestion of one or more infectious organism. Values of "r" closer
to 1 reflect a high likelihood of an infection occurring given ingestion of one or more infectious organism.
In the basic exponential model, the value of "r" is assumed to be the same for all host-organism
interactions.
It is important to note that the basic exponential dose response model is a non-threshold,
single-hit model. That is, this model allows for an individual to become infected by ingesting as few as
one infectious oocyst that survives the host's defenses. The non-threshold assumption is supported both
Economic Analysis for the LT2ESWTR 5-9 December 2005
-------
by what is known about Cryptosporidium's mechanism of infection and by some empirical data. The
mechanism of infection by Cryptosporidium involves the oocyst that has been ingested releasing four
sporozites which penetrate the cells lining the gastrointestinal tract where they reproduce. There is no
evidence to suggest that cooperativity among more than one oocyst is required to initiate an infection. In
addition, data show that mice have been infected following doses of 1 oocyst. Further, the basic
exponential model does not account for immunity in the population that would preclude some individuals
becoming infected regardless of the number of infectious organisms ingested.
The primary dose-response model and the six alternative dose response models included in
Appendix N are modifications to this basic exponential model. In all cases, there are modifications to the
model that address the assumption regarding "r" being the same for all host-organism interactions and the
underlying data sets used to parameterize the models.
The primary model used for the baseline risk and benefits analysis comprises two different
assumptions about the distribution of "r" values and two different sets of data for parameterizing those
distributions (that is, a total of 4 models that are all included in the baseline risk simulation).
The six alternative models use data from six studies to parameterize the r distributions. In
addition, two of the six alternative models considered also incorporate an additional adjustment to
account for potential immunity (host susceptibility) to infection by these organisms. Exhibit 5.4
summarizes the characteristics of the primary dose-response model and the six alternative models.
Economic Analysis for the LT2ESWTR 5-10 December 2005
-------
Exhibit 5.4 Characteristics of the Primary Model and Six Alternative Models
Model
Primary
Model
(random
selection
from 4
models)
Alternative
Model 1
Alternative
Model 2
Alternative
Model 3
Alternative
Model 4
Alternative
Model 5
Alternative
Model 6
Data Sets
Two
models use
2 studies;
two models
us 3
studies
6 studies
6 studies
6 studies
6 studies
6 studies
6 studies
Functional form
P, = 1 - e -dr
P, = 1 - e -" r
P, = 1 - e -" r
P, = 1 - e -dr
P, = Y * (1 - e -<")
P, = 1 - e -" r
P, = Y x (1 - e -<")
Distribution of
r
Two models use
normal (logit);
two models use
t(3)-distribution
Normal (logit)
t(3)-distribution
Beta distribution
Beta distribution
Beta distribution
Beta distribution
Assumptions
Assumes the 4
models are equally
plausible; r
distribution assumes
organism variability
r distribution
assumes organism
variability
r distribution
assumes organism
variability
r distribution
assumes organism
variability
r distribution
assumes organism
variability; additional
para meter for
information on host
immunity
r distribution
assumes host
variability
r distribution
assumes host
variability; additional
para meter for
information on host
immunity
Mean
risk1
0.082
0.036
0.046
0.052
0.137
0.140
0.105
1 Mean risk calculated from the distribution of r and g values used as inputs for the dose-response model (primary
model reflects combination of all four component models).
Primary Dose-Response Model and Alternative Models 1-3
The modifications made to the "r" value in the Primary dose-response model and in Alternative
Models 1 - 3 are conceptually very similar to one another and so are described here together. In fact,
Alternative Models 1 and 2 as described here are the two model forms that are used together in the
primary dose-response model.
In these models, the value of "r" that is assumed in the basic exponential model to be the same for
all host-organism interactions is replaced by a probability distribution of "r" values. These probability
distributions capture both the variability and the uncertainty in the underlying human challenge data used
to estimate the r parameter. Models 1-3 differ from one another only with respect to the assumed form
Economic Analysis for the LT2ESWTR
5-11
December 2005
-------
of the probability distribution for r. Furthermore, Models 1 - 3 are fit to the human challenge data
wherein it is assumed that the distribution in "r" is due to difference in the strains of Cryptosporidium
used, and not to any variability in host sensitivity of the individuals in the challenge studies2.
Model 1 uses a normal probability distribution and includes a logit transformation of the r values
(that is, it is assumed to be the values of logit(r), rather than the values of r, that are normally distributed).
The normal distribution is the basic bell-shaped curve probability distribution that is determined by two
parameters: the mean and standard deviation. However, the normal distribution has a domain of-8 to +8,
whereas the values for "r" as noted previously lie exclusively between 0 and 1. While it is possible to
constrain values to fall within the 0 to 1 range in the normal distribution, such truncations can be
computationally cumbersome and result in a distributional shape that is not truly a normal distribution.
For this reason, the logit transformation is widely used in circumstances such as this where a distribution
of a probability value is involved to convert the probability values that must fall in the range of 0 to 1 into
values on the real number line from -8 to +8 so that they can be handled in the normal distribution. The
logit transformation allows the modeler to both meet the requirement that the distribution of numbers
being evaluated be bounded by 0 and 1, and employ the normal distribution (and others as noted below)
which generally are not constrained to this range. Specifically, the logit transformation of "r" is logit(r) =
x = log[r/(l-r)]. The human challenge data are used to estimate the parameters for the normal distribution
of logit(r) values, which are then converted back to r values in the 0 to 1 range by the inverse of the logit
transformation, which is r = ex / [(1 + ex)] where x = logit(r) as indicated.
Model 2 differs from Model 1 in that it uses a t-distribution rather than the normal distribution for
the r values . It also includes the logit transformation of the r values for the same reasons as described for
the normal distribution. The t-distribution, which is similar to the normal distribution, has one additional
parameter besides the mean and the standard deviation - the degrees of freedom - that determines its
shape. The specific model used assumes 3 degrees of freedom (df) and is therefore referred to more
specifically as the t(3) distribution. The more degrees of freedom for a t-distribution, the closer it is to the
standard normal distribution (essentially the same for df > ~30). The fewer the number of df, the more
disperse the t-distribution is. Therefore, a t(3) distribution is more disperse than the normal distribution
of a data set having the same mean and standard deviation, so it reflects more variability. The use of the
t(3) distribution is based on a recommendation by the Science Advisory Board. The human challenge
data are used to estimate the mean and standard deviation of the logit(r) values to use in the t(3)
distribution of [logit(r)-mean / standard deviation]. As with Model 1, the logit(r) values obtained from
this distribution are then converted back to r values in the 0 to 1 range from the inverse of the logit
transformation noted above.
Again, Models 1 and 2 are used together as described in Appendix N as the primary dose-
response model for the baseline risk and benefits analysis.
Model 3 differs from Model 1 and Model 2 in that it assumes the r values follow a Beta
distribution. The Beta distribution is a standard, highly flexible distribution form that uses two
parameters (alpha and beta) and has the advantage over the normal and t-distributions in that it describes a
All fitting of these models (that is, estimating model parameters) to the empirical data was carried out using a
Bayesian approach employing a Markov Chain Monte Carlo (MCMC) procedure. Bayesian approaches to estimating model
parameters offer a number of advantages over classical statistical modeling methods. In particular, the outcome of Bayesian
analysis is a distribution of plausible values of the parameters of interest rather than a single, most likely estimate typically
provided by classical methods. MCMC procedures use "random walks" through the parameter space where each subsequent step
is dependent upon the preceding step such that over time the entire space is considered. MCMC methods provide a means to
solve complex multi-dimensional integrals that often arise in Bayesian analyses which can prove intractable to solution by
numerical methods.
Economic Analysis for the LT2ESWTR 5-12 December 2005
-------
distribution of values in the domain of 0 to 1, and therefore does not require the logit transformation as
used in Models 1 and 2.
In summary, Models 1-3 can be viewed as similar variations to the basic exponential dose
response model in that they replace the assumed constant value for r with a probability distribution. The
three models differ in that each uses a different underlying probability distribution assumption (normal,
t(3), and beta). In all three cases, the parameters for these models are fit using a MCMC method wherein
the distribution of r values is assumed to derive from differences in the infectivity of the various isolates
used in the challenge studies, and not to any differences in the susceptibility of the hosts in those studies.
Alternative Model 4
Model 4 is a variation of Model 3 that includes an additional parameter referred to here as gamma
(y) outside of the basic dose-response model:
The values for r in this model are, just as in Model 3, assumed to be Beta distributed where the alpha and
beta parameters of the Beta distribution are derived using the assumption that differences in r are due to
the isolates not to the hosts. The additional parameter y, is included to provide some information in the
dose response model concerning the hosts. Specifically, it provides an estimate of the portion of the
population that appears to have immunity to the organisms. The value of y falls between 0 and 1 and is
the probability that an exposed individual ingesting one or more infectious organisms is a non-immune
person. Because y represents the portion on the population that is not immune, the immune portion of
the population is, therefore, equal to 1 - y.
The value for the y parameter is estimated from the human challenge data in the Bayesian /
MCMC procedure in the same manner as the parameters for the distributions of r are estimated. That is, it
is an additional parameter that, together with the other parameters for that model, provides the best fit to
the observed data using the MCMC method.
Alternative Models 5 and 6
Model 5 and Model 6 are identical in form to Model 3 and Model 4, respectively. The difference
between each of them is the assumption used in the MCMC procedure to estimate the parameters of the
Beta distribution for the r values regarding the source of differences. As noted previously, Models 3 and
4 (and Models 1 and 2 as well) are parameterized from the human challenge data using a MCMC
procedure that assumes any differences in r are due to differences in the isolates used, and not to any
differences in the hosts in those studies. Models 5 and 6 are parameterized instead with the assumption
that all of the differences in the r values are due to differences in the hosts, while none is due to
differences in the isolates studied. Model 6 includes the additional host susceptibility factor of y which
implies that not only are there differences in the degree of susceptibility among hosts exposed to
infectious organisms but that some fraction of hosts (1 - y) are immune to infection.
It should be noted that the y value in these models indicates the portion of the population that
does not have complete immunity (again, 1 - y is the fully immune portion). The variability in r reflected
in the distributions derived in Models 5 and 6 describes differences in susceptibility among individuals to
becoming infected to a given dose of Cryptosporidium, which in turn can be viewed to also reflect
underlying differences in the resistance (partial immunity) to infection.
Economic Analysis for the LT2ESWTR 5-13 December 2005
-------
EPA has not attempted to estimate parameters for the distribution of r in the six model forms
considered to simultaneously capture both differences in the isolates and differences in the exposed
subjects. EPA considered efforts by Peter Teunis, who has attempted to do this, and determined that the
underlying data are probably not adequate to estimate two separate sets of parameters simultaneously for
hosts and isolates. In addition, the bulk of the total variability seen in the data appears to be explained by
the wide variability resulting from the consideration of differences among hosts.
What are the merits and limitations of each model?
One of the primary merits of these dose-response models is the basic exponential model form of
which they are all variations. A particular advantage of the exponential dose-response model form is that
combined within its relatively simple equation are the two key elements that determine whether an
individual consuming water containing infectious organisms will become infected.
The first of these elements is the probability that an individual who consumes contaminated water
will ingest one or more of those organisms. This element is based on a Poisson distribution, which is the
appropriate distributional form for describing this type of discrete "event" - namely, the presence of zero,
one, or more organisms in a given volume of water consumed taken randomly from a larger volume
containing a known concentration.
The second element addresses organism-host interactions in terms of the probability that the
infectious organism(s) will survive once ingested to reach a site where it can cause an infection.
As discussed in the preceding section, the basic dose-response model form assumes a constant "r"
value for all organism-host interactions. The use of a constant "r" value would imply that every
individual ingesting the same number of organisms would be expected to respond the same, and that there
would be no uncertainty around that response rate. Since this did not seem plausible or consistent with
the available human challenge data, EPA did not pursue models with a constant "r" value. Rather, in the
6 variants of the exponential dose-response model considered in Appendix N, EPA described "r" as a
distribution of values rather than a single value to capture the apparent variability and uncertainty in the
host-organism interaction. Three different distributional forms are used for this purpose.
The two forms (normal and t(3) distributions) used in the primary model and Models 1 and 2,
respectively, are standard distributions, and were recommended for use in these models by the SAB. Two
limitations in both of these distributions are (1) the distributional shape of both is the typical bell-shaped
(Gaussian) distribution and (2) these distributions are typically applied to data sets that can take on any
values, whereas the distribution of r (being a probability) is limited to the range of 0 to 1. This latter
limitation was overcome by using the logit transformation.
The limitation regarding the implied shape of the distribution of "r" values (more specifically, the
logit(r) values) by assuming the normal and t(3) distributions was addressed in part by the use of the Beta
distribution in Model 3. The Beta distribution has the advantage of not being constrained to a particular
shape but is sufficiently flexible to take on a variety of shapes depending upon the estimated parameters
alpha and beta. In addition, the Beta distribution considers values on the limited domain of 0 to 1 directly
without the logit or other transformations being necessary.
EPA is not currently aware of any biological basis for preferring any one of these three
distributional forms for "r" over the others.
Although the inclusion of "r" as a distribution of values in these models addresses variability and
uncertainty in the organism-host interactions, the approach to estimating the parameters for the
Economic Analysis for the LT2ESWTR 5-14 December 2005
-------
distributions of "r" values was limited to the assuming that the differences being due to either the
organism (Models 1 - 4) or to the host (Models 5 and 6). The available human challenge study data are
too limited to support developing models that would simultaneously account for the observed differences
due to hosts and differences due to isolates (this would require that two parameters be estimated each for
a distribution of "r" on isolate and a distribution of "r" on host). However, it appears that the models
based on differences among hosts explain the bulk of the total variability seen in the data.
Models 4 and 6 include an additional y parameter to account for the portion of the population that
has complete immunity. In this respect, Model 4 does to some extent for both isolate and host
differences, but the latter only in terms of complete immunity. Model 6 accounts for two aspects of host
differences - both the complete immunity by virtue of the parameter as well as differences in
susceptibility among hosts who are not completely immune.
Another area of potential limitation in these six models is that they all describe a non-threshold,
single-hit mechanism. Alternative models could include consideration that more than one infectious
organism would need to be ingested and survive to interact with the host to cause an infection. However,
as noted in Haas et al. (1999) where various threshold models are discussed along with the exponential
dose response model, there is strong evidence that the single-hit dose response models are consistent with
observation and expected biological mechanisms leading to infection. It is well recognized that a single
microorganism (including viruses, bacteria, and protozoa) can reproduce. Haas points out that in almost
all cases considered, the use of the exponential model forms, which assume no threshold, results in a
statistically significant improvement in fit over model forms that allow for a threshold.
How has each model been used in the literature for modeling the dose-response of
Cryptosporidium or similar organisms?
The basic exponential model, which underlies the primary model and all six alternative
dose-response models considered in Appendix N, has been discussed and applied extensively in the
literature on microbial risk assessments. Variations of the exponential model that address "r" as a
probability distribution, particularly the Beta distribution, rather than a fixed value have also been
extensively discussed. Some key early papers describing these models and applications include Furomoto
and Mickey (1967) and Regli et al. (1991), the latter of which considers these models specifically in the
context of modeling microbial risk in drinking water.
An extensive presentation of the basic exponential dose-response model, including its derivation
from the Poisson and Binomial distributions as discussed previously, is provided by Haas et al. (1999) in
their book "Quantitative Microbial Risk Assessment" in the chapter on dose-response functions. That
chapter of Haas et al. (1999) also provides a detailed discussion of the "Beta Poisson" model, which treats
r in the exponential model as a Beta distribution similar to the approach used in Models 3 through 6. The
Beta Poisson model is also discussed in detail by Teunis and Havelaar (2000).
The details of the Beta Poisson model discussed by Haas et al. (1999) and by Teunis and
Havelaar (2000) with respect to how the Beta probability distribution of r is integrated with the Poisson
component of the exponential function differs from that used in the Appendix N models. In these
discussions, both an "exact" form and an approximate form of the Beta Poisson model are described. The
exact form of the model involves an integration across all possible values of r (0 to 1) of the product of
the basic exponential model function and the Beta probability distribution of r values. As discussed by
both sets of authors, integrating these functions in the exact Beta Poisson does not have a closed form
Economic Analysis for the LT2ESWTR 5-15 December 2005
-------
solution and must be solved numerically3. Both sets of authors also discuss a simplified approximation to
the exact Beta Poisson that was first presented by Furomoto and Mickey (1967a, b) and which is also
referred to as a "Pareto" function:
Here, aand p are the equivalents of the alpha and beta parameters of the Beta distribution as
described previously for Model 3. Both sets of authors, and particularly Teunis and Havelaar (2000),
point out that this approximate form of the model gives results that agree closely with the exact form
under many conditions, but it does not agree with it in all cases and can in some circumstances imply a
risk of infection that exceeds the underlying risk of ingesting one or more infectious organisms.
The approach to incorporating the Beta probability distributions for r as used in Models 3 - 6 of
Appendix N, where the r parameter is treated directly as a distribution in the exponential portion of the
model where it appears, overcomes the complexity of the numerical integration of the exact Beta Poisson
model, without the potential shortcomings of the Pareto approximation form of the Beta Poisson model.
The use of the logit normal and logit t(3) distributions in Models 1 and 2 as alternatives to using
the Beta distribution for r was recommended to EPA by the SAB4. These model forms were not found
discussed specifically elsewhere in the literature, but both Haas et al. (1999) and Teunis and Havelaar
(2000) note the potential computational advantage of using probability distributions such as these for r
rather than the Beta distribution which does not have a closed form and requires numerical integration.
The latter authors specifically recognizing the potential need to include a logit transformation for certain
types of alternative distributions.
With respect to the inclusion of the immunity factor y in Models 4 and 6, Haas et al. (1999)
discuss the effects of immunity in the population when performing microbial risk assessments, although
they do not specifically include this factor in any of the dose-response models they present. However,
this approach is presented specifically by Holcomb et al. (1999) as one of six dose response models they
consider for use with food-borne pathogens. In their paper they refer to this as the "flexible exponential"
model and it is the basic exponential model (constant r value) with the additional parameter added that
sets a maximum probability of infection to a value less than 1 . (The other models they include in their
paper in addition to the simple exponential and this flexible exponential are a simple lognormal
distributions model, the approximate form of the Beta Poisson as described above, a Weibull-Gamma
model form that is similar to the approximate Beta Poisson, and a log-logistic model that is another
variation of the basic exponential form.)
Some recent examples of risk assessments of waterborne Cryptosporidium that use the
exponential dose-response model include Pouillot et al. (2004) and Makri et al. (2004), both of which
i
The "exact" form of the Beta Poisson model is obtained from 1 — 6 J \T)CIT where f(r) is the Beta
o
Distribution. With the Beta Distribution, this integral does not have a closed form solution and must be solve by numerical
methods.
4Prior to SAB's review and recommendations, EPA used a different form of the exponential model (1-e "d/k) The
parameter k (note that k = 1/r), can be roughly interpreted as the number of organisms that must be ingested to ensure that at least
1 survives to initiate an infection. In this model form, k was treated as a lognormal distribution to capture uncertainty
(constrained such that k > 1).
Economic Analysis for the LT2ESWTR 5-16 December 2005
-------
appeared in the February 2004 issue of Risk Analysis, which included a special issues section on
microbial risk assessment.
What criteria should be used to select a model for theLT2 risk assessment?
Alternative dose response models for consideration in the LT2 risk assessment should conform to
the basic requirements of biological plausibility as articulated by Haas et al. (1999). Those authors
identify the two key requirements of plausibility for dose-response models for microbial pathogens as (1)
the model should reflect that the population is exposed to a distribution of organisms, such that different
individuals consuming water with the same concentration can ingest different numbers of organisms, and
(2) the model should account for potential host-organism interactions whereby host barriers can mitigate
the potential for infection even when one or more infectious organisms have been ingested. As discussed
previously, the six dose-response models considered in the LT2 analysis are variants of the exponential
doe response model which, in turn, is derived from consideration of a Poisson distribution for exposure
and incorporates a specific parameter (r) for consideration of the host-organism interactions.
The selection of one or more specific models as the preferred alternative from among a set of
models fit to the same data set should include appropriate tests that compare how the models conform to
the underlying data used to estimate the model parameters. In the classical statistics (frequentist) context,
such model fit testing methods include consideration of measures of deviance (- 2 log-likelihood ratio) to
identify the models with the better fits (lower deviances) from among alternatives, as well as
goodness-of-fit methods (e.g., using a x2 distribution) where the acceptance or rejection of a model as
fitting the data is based on specified hypothesis tests.
The six models considered in Appendix N for the LT2 assessment were parameterized using a
Bayesian approach employing MCMC procedures (see earlier footnote). The methods for comparing
competing models in Bayesian analysis generally rest on estimates of the marginal likelihood. One
approach based on marginal likelihood estimates that is used for model assessment in Bayesian analyses
uses a measure referred to as the Deviance Information Criterion (DIG), also referred to as the Bayes
Information Criterion (BIC). The DIC provides a measure of the likelihood of observing the data at hand
given the estimated parameters for the models. In addition, the DIC includes a penalty for models that
have more parameters even if they provide improved fit, and as such also serves as a means to assess
model parsimony.
The DIC is generated as an output of the WinBUGS program used to estimate the parameters for
the models considered in Appendix N, and was included in Exhibit N.20. Lower values of the DIC
indicate better model fits. The DIC information for each of these models, as well as the estimated
illnesses avoided estimated from the six models for each regulatory alternative (these are based on mean
ICR occurrence data) are presented in Exhibit 5.5.
Overall, Models 4-6 appear to perform better by the DIC measure than Models 1-3. This
suggests that the observed differences in infectivity reflected in the dose response models are likely
driven more by the differences in susceptibility among hosts than differences in infectivity among
isolates. Model 6, which had the lowest DIC value (DIC = 37.0) includes consideration of host
differences both in terms of the portion of the population who are fully immune (the parameter) and the
susceptibility among those who are not fully immune in the parameters of the r distribution. Model 5,
which performs next best (DIC = 54.6) does not consider an immune subset of the population, but it does
use differences in susceptibility among hosts to parameterize the distribution of r values as done in Model
6. Model 4 is the third best fit (DIC = 66.3), and although it assumes that differences in infectivity are
due to isolates, it also includes a host component by virtue of the parameter to account for the portion of
the population who are fully immune. Models 1-3 (DIC values of 122.6, 122.4, 111.3), which do not fit
Economic Analysis for the LT2ESWTR 5-17 December 2005
-------
as well as Models 4-6, are all parameterized assuming that differences are due solely the isolates. While
these results do not necessarily imply that there are no differences in infectivity among isolates, they do
suggest that those differences are less influential to the overall observed differences than are the
differences in host susceptibility.
Exhibit 5.5 Comparison of Annual Illnesses Avoided Predicted by the Dose
Response Models Considered for Each Regulatory Alternative
Model
Primary
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Baseline
Illnesses
3,603,515
1 381,593
1,556,549
2,454,679
5,580,291
5,418,751
4,588,061
DIG1
NA
122.6
122.4
111.3
66.3
54.6
37.0
A1
989,954
369,328
413,857
689,748
1,501,445
1,495,997
1,283,450
A2
975,326
363,178
406,609
680,404
1,477,257
1,473,280
1,265,854
A3
964,360
358,732
401,401
673,445
1,459,126
1,456,257
1,252,707
A4
902,500
332,908
372,903
633,853
1,360,326
1,360,725
1,178,298
1 The Deviance Information Criteria (DIG) provides a measure of the likelihood of observing the data given the
estimated parameters for the model. See preceding text for more detail.
For comparison, the first row in Exhibit 5.5 shows the results using the model at proposal. The
primary model used both Models 1 and 2, and separately applied them both to three (Iowa, TAMU, UCP)
or only two (Iowa and TAMU) of the human challenge studies that were available at the time of proposal.
(rather than to the six that are currently available and which were used to evaluate Alternative Models 1 -
6 as shown above). Each of these four combinations of models and studies generated a set of r values. In
the primary analysis, the baseline risk and benefits were obtained using a Monte Carlo procedure to select
r values randomly from these four sets of r to include in the dose-response calculations.
As can be seen in Exhibit 5.5 , the choice of model can significantly influence the estimate of
illnesses avoided by the various rule alternatives. Because of the limitations of the human challenge
studies (small numbers of subjects, high dosage relative to drinking water exposure) it is not possible to
say with any confidence which, if any of these models, provides the "best" estimate of Cryptosporidium
infectivity and thus the risk reduction that will result from the rule. To address this limitation, EPA has
analyzed the human challenge data using a range of models, as recommended by the SAB and described
in this chapter.
The full analysis in the remainder of the EA is conducted using the primary model, which is the
same model used at proposal. This provides results that are roughly in the middle of the range of results
from the other models, and facilitates comparison with the analyses conducted for the proposed rule.
However, to more fully capture the range of possible results resulting from alternate model choices, EPA
has also conducted the analyses using a "high" estimate of baseline risk (Model 4) and a "low" estimate
of baseline risk (Model 1). These estimates should not be construed as upper and lower bounds on
illnesses avoided and benefits. For each model, a distribution of effects is estimated, and the "high" and
"low" estimates show only the means of these distributions for two different model choices. The detailed
distribution of effects is presented for the proposal model. Further, the six models analyzed do not cover
Economic Analysis for the LT2ESWTR
5-18
December 2005
-------
all the possible variations of models that might have been used to analyze the data, and it is possible that
estimates with other models would fall outside the range presented. However, as discussed in this
chapter, EPA believes that the models presented here provide a reasonable range of results based on
important dimensions of model choice (e.g., whether "r" is modeled as varying by isolate or by host).
For purposes of the summary tables presented in the preamble to the rule, EPA determined that
model choice was the most significant dimension of uncertainty affecting the analysis and is thus
presenting the tables in a format that shows "high," "medium" and "low" estimates of illnesses avoided
and benefits. These represent the mean estimates using the high model (Model 4), the proposed model,
and the low model (Model 1) respectively.
Morbidity Rate
The above elements of dose-response modeling relate to the prediction of an infection occurring
given various exposure circumstances. As noted at the outset of this section, the hazard identification for
Cryptosporidium includes not only the risk of infection, but also the risk of illness resulting from an
infection. Not all infections will result in illness and observable symptoms. The probability of becoming
ill given an infection is called the morbidity rate.
Some studies indicate that the morbidity rates increase at higher doses (DuPont et al. 1995).
However, for this risk assessment, the morbidity rate is independent of dose. After examining the
potential impact, EPA determined that a higher-morbidity-at-higher-dose effect was not significantly
relevant to this analysis. The fundamental effect being quantified is the endemic rate of illness and death
from persistent (and low) levels of Cryptosporidium, not the higher levels that might occur in an
outbreak. The underlying dose data, both as measured and modeled, reflect at most a few oocysts per day
for individuals. In the risk assessment model, the portion of the risk posed by the small portion of the
population ingesting even an expected two oocysts/L is negligible; the portion of the risk posed by people
ingesting three or more oocysts/L is virtually zero. Thus, the results of the analysis would not be affected
by using increased morbidity rates with significantly higher doses. (Although not quantified, the risks of
outbreaks are considered and are discussed in section 5.4.1.)
To develop an estimate of morbidity rate, EPA analyzed available literature and identified studies
with applicable data. Some of the preliminary human ingestion trials were conducted on healthy
individuals with no evidence of previous C. parvum infection (DuPont et al. 1995). Other studies
challenged individuals with existing antibodies or re-challenged those who had participated in earlier
studies. DuPont et al. (1995) found that 39 percent of those infected had clinical cryptosporidiosis. Haas
et al. (1996) provided information based on the same data also suggesting a morbidity rate of 39 percent,
but also computed 95 percent confidence limits of 19 and 62 percent. More recently, a study found that
after repeated exposure to C. parvum (IOWA strain), the morbidity rate was the same as for the initial
exposure in re-infected subjects (Okhuysen et al. 1998). Okhuysen et al. (1998) also found that 58
percent of their subjects who received doses of Cryptosporidium developed diarrhea, which is an
underestimate of morbidity since symptoms other than diarrhea contribute to the morbidity rate.
However, these subjects were given doses higher than those projected in water supplies. Chappell et al.
(1997) observed that the rate of diarrheal illness was higher for the TAMU or UCP isolates of C. parvum
than for the IOWA isolate first studied by DuPont et al. (1995) and Haas et al. (1996).
Given these results and the morbidity variability associated with C. parvum during reported
outbreaks, the actual morbidity rate may vary with the type of strain to which a population is exposed, as
well as with the immune status of the exposed population. However, the prevalence of strains and the
immune status of the population are unknown and therefore not quantified for this risk assessment. The
uncertainty around the value for morbidity, though, is considered in the risk assessment. The quality of
Economic Analysis for the LT2ESWTR 5-19 December 2005
-------
available data does not support making more than a generalized estimate of the range and nature of
uncertainty. The underlying data do support the use of a distribution with a central tendency and provide
information to establish reasonable ranges. As a result, morbidity was modeled as an uncertain variable
having a triangular distribution.
Analysis of the reviewed research resulted in a mode of 50 percent, lower bound of 30 percent,
and upper bound of 70 percent for the triangular uncertainty distribution. The following limitations in the
research were identified and considered in the derivation of the above values: the Okhuysen et al. (1998)
results based on diarrheal rates are probably an underestimate; Chappell et al. (1997) found that diarrheal
rates were higher for isolates other than IOWA; and the general population likely has a higher morbidity
rate than the healthy individuals used in the study groups.
The central tendency (mode) for the distribution used in the risk assessment model is 50 percent.
This is a bit below the Okhuysen et al. (1998) results (58 percent), but above the values estimated by Du
Pont et al. (1995) and Haas et al. (1996) (39 percent). These studies used the IOWA isolate, and a simple
average of them results in a value of 48.5 percent. The mode was rounded up to 50 percent to account for
the apparent underestimation of these studies, as noted above.
The upper bound for the distribution used in the risk assessment model is 70 percent. The upper
bound was set above the 95 percent confidence limit of 62 percent estimated by Haas et al. (1996). This
reflects that the absolute limit of the triangular distribution would reasonably be above that 95 percent
confidence limit and the apparent underestimation of these studies, as noted above. The difference in the
upper bound (70 percent) and the Haas et al. 95 percent confidence limit (62 percent) represents only 3
percent of the triangular distribution, indicating that the upper tail of the triangular distribution is
comparable to upper portion of Haas's distribution.
The lower bound for the distribution used in the risk assessment model is 30 percent. The lower
bound was set higher than the 19 percent estimated by Haas et al. (1996). While this bound does not
encompass the lower 95 percent confidence level in the distribution used in the risk assessment, it does
account for the apparent underestimation in the studies.
Mortality Rate
The third dose-response relationship used in this analysis is the probability of fatality given that
an illness has occurred. There are no general data on the rate of mortality from cryptosporidiosis. To
derive mortality estimates, data from the Milwaukee outbreak are used and adjusted to reflect changes in
rates of illnesses and advanced treatments that have lessened mortality among persons living with AIDS.
Further adjustments are used to reflect the differences between the populations of those living in areas
served by filtered and unfiltered systems. Since there is uncertainty around the mortality rate used in the
dose-response model, EPA conducted a sensitivity analysis that varied the AIDS mortality rate by +/- 50
percent. This analysis and its results are described in Appendix R.
The starting point is the mortality rates associated with the Milwaukee Cryptosporidium outbreak.
In that outbreak, 54 people died who had cryptosporidiosis listed on their death certificate. Of those, 46
also had AIDS listed as an underlying cause of death (Hoxie et al. 1997). The Milwaukee outbreak had
an estimated 403,000 cases of illness (Kramer et al. 1996b). The unadjusted mortality rate for AIDS-
related5 deaths is thus 46 deaths/403,000 illnesses, or 11.41 deaths/100,000 illnesses. The rate for the
5 The term "AIDS-related deaths" as used here and throughout this document refers only to deaths caused
by cryptosporidiosis (as listed on death certificates), but for which AIDS was also listed as an underlying cause of
death.
Economic Analysis for the LT2ESWTR 5-20 December 2005
-------
other, non-AIDS-related deaths is thus 8 deaths/403,000 illnesses, or 1.98 deaths/100,000 illnesses. (All
calculations in this section are rounded for ease of presentation, but unrounded data are used in the
analysis.)
There were no further adjustments made to the non-AIDS mortality rate. A review of available
statistics showed that data to compare the incidence of the other underlying illnesses (coccidiosis
(presumably cryptosporidiosis), viral hepatitis, brain tumor, heart failure, and alcoholic cirrhosis of the
liver) between Milwaukee in 1993 and the nation in 1999 or 2000 were generally unavailable. Even
comparison of proxy data (death rates rather than incidence) proved of little value. Data for Milwaukee
were, in general, inconclusive; too few cases were reported to make statistics meaningful. Only in the
case of alcoholic cirrhosis of the liver were data statistically significant, and in that case, the rate of deaths
per 100,000 population was comparable between Milwaukee (3.36) and the nation as a whole (3.03)
(CDC 2001a; Hoxie et al. 1997; CDC 1995). One factor that could affect the non-AIDS mortality rate is
age. Hoxie et al. do not provide data on the age of those who died in the outbreak. Although Naumova et
al. (2003) found that the rate of gastroenteritis during (and prior) to the outbreak increased with age, there
is no information on whether the elderly have a higher mortality rate from cryptosporidiosis. It is
conceivable that as the percentage of the population that is elderly increases in the coming decades, the
percentage with underlying non-AIDS related disease could also increase, affecting the mortality rate
indirectly. However, no data are yet available to support this hypothesis.
The Milwaukee AIDS-related mortality rate was adjusted to account for the decrease in the
mortality rate among people with AIDS from the time of the Milwaukee incident to 2001 (the most recent
year with comparable data), and the difference in the Milwaukee AIDS population in 1993 to the national
AIDS population in 2001. These adjustments are described below; the adjusted calculation is:
Deaths/100,000 illnesses in the Milwaukee outbreak (11.41) x factor to adjust for
lessened mortality overtime among persons with AIDS x factor to adjust for changes in
the prevalence of AIDS in the general population = AIDS-related deaths per 100,000
cryptosporidiosis illnesses.
The mortality rate for AIDS has declined greatly since 1993 due to the use of combination
retroviral therapies and other factors. Combination retroviral therapy raises the CD4+ cell count,
enabling people with AIDS to better fight off infection. Correlations have been shown between
cryptosporidiosis in AIDS patients and CD4+ counts (Inungu et al. 2000, Pozio et al. 1997). The AIDS
mortality rate in 2001 was 4,845 deaths per 100,000 AIDS population (17,402 deaths in a population of
359,141) (CDC 2002). In 1993, this rate was 25,963 (45,271 deaths in a population of 173,772) (CDC
2001). The ratio of these rates is 18.4% percent, that is, the rate of deaths among AIDS patients for all
reasons in 2001 was only 18.4 percent of what it was in 1993.
The second adjustment accounts for the difference in the percent of the national population that
was living with AIDS in 2001 and the percent of the Milwaukee population that was living with AIDS in
1993. This adjustment is calculated separately for areas served by unfiltered systems and filtered
systems. As an approximation of the value for populations served by unfiltered systems, the percentage
of the population living with AIDS, which is 0.196 percent (62,349 in apopulation of 31,859,141), was
used. As an approximation of the percentage in areas served by filtered systems, national estimates were
used, less what had been accounted for by unfiltered systems. (See Appendix C for details on these data
and calculations.) The rate for filtered systems is 0.118 percent (based on an AIDS population of 299,912
in a population base of 253,234,672) (CDC 2002; US Census 2001).
The percentages of people living with AIDS in 2001 served by filtered and unfiltered systems are
used separately to adjust and update the 1993 incidence rate of AIDS in Wisconsin. The data on AIDS
Economic Analysis for the LT2ESWTR 5-21 December 2005
-------
incidence and population should represent the same location; however, the areas for which data are
available do not match the exact geography of the areas served.6 The ratios that come from this approach
are still useful as approximations, and their use is an improvement over not including adjustments for
these factors at all. In Wisconsin in 1993, the percentage of the population that had AIDS was 0.017 (862
persons with AIDS in a population of 5,044,318). Extrapolating the Wisconsin data to all populations
served by unfiltered and filtered systems, gives a factor of 11.45 for unfiltered systems (0.196
percent/0.017 percent) and a factor of 6.93 (0.118 percent/0.017 percent) for filtered systems. That is, the
incidence of people living with AIDS in 2001 in areas served by unfiltered systems is 11.45 times the
incidence in Wisconsin in 1993. Similarly, there are 6.93 times as many people living with AIDS in 2001
and served by filtered systems as there were in Wisconsin in 1993.
Using the Milwaukee AIDS-related mortality rate and the adjustment factors described above, the
final mortality rate for unfiltered systems (expressed as deaths per 100,000 cryptosporidiosis illnesses) is
24.07 (11.41 x 18.4% x 11.45). Similarly, for filtered systems, the AIDS-related mortality rate is 14.56
deaths per 100,000 cryptosporidiosis illnesses.
The risk assessment model uses a combination of AIDS-related and non-AIDS-related mortality.
Thus, adding together these rates yields an overall mortality rate for unfiltered systems of 26.05 deaths
per 100,000 cryptosporidiosis illnesses (24.07 AIDS + 1.98 non-AIDS). For filtered systems, this figure
is 16.53 deaths per 100,000 cryptosporidiosis illnesses (14.65 AIDS + 1.98 non-AIDS). These mortality
factors are constants in the model (that is, no uncertainty is attributed to these parameters).
The mortality rate from the Milwaukee outbreak may not reflect the overall mortality rates from
low-level endemic exposure. The estimated levels of Cryptosporidium in the finished water supplies
during the Milwaukee outbreak were much higher than the levels expected in systems complying with the
SWTR, IESWTR, and LT1ESWTR. Thus, the higher level of Cryptosporidium in the water supply could
have resulted in a higher mortality rate than that expected from endemic exposure if responses increased
more than proportionately at higher dose levels.
No data are yet available, however, to support this hypothesis; data are available to indicate only
a higher probability of infection resulting from higher ingested doses. In an outbreak in Las Vegas,
similar mortality rates were observed in AIDS patients (52.6 percent among AIDS patients in Las Vegas
compared with 68 percent among AIDS patients in Milwaukee). These similar rates were observed
despite the hypothesis that the drinking water had been contaminated over an extended period of time
with intermittent low levels of oocysts, unlike Milwaukee's massive contamination (Rose 1997). A
recent study by Hunter et al. (2001) suggests that the level of endemic diarrhea from all sources was
underestimated in the Milwaukee incident, leading to an overestimation of the number of diarrheal
illnesses due to cryptosporidiosis. A lower estimate of illness would consequently raise the mortality rate
per case of illness by holding deaths constant as illnesses decreased. However, there is currently no
consensus on whether to accept the Hunter et al. conclusions, and responses to their analysis are being
prepared by other investigators of the Milwaukee outbreak. The model, therefore, uses the Hoxie et al.
1997 illness estimates (cited previously) for the Milwaukee outbreak.
6 Data on the AIDS population and on the population served by the water system for the area directly
affected by the Milwaukee Cryptosporidium outbreak are inconsistent in the sources used and the area covered, and
individual estimates varied. Data for the entire State of Wisconsin are used as the best consistent source of AIDS
data and population data. The State-level data are from U.S. Census and CDC sources. These data are comparable
to other data used in this analysis. See Appendix C for more details.
Economic Analysis for the LT2ESWTR 5-22 December 2005
-------
5.2.4 Exposure Assessment
This section discusses the three elements needed for characterizing human exposure to infectious
Cryptosporidium oocysts in drinking water.
• The distribution of total and infectious Cryptosporidium in finished water, reflecting source
water levels and treatment effectiveness (section 5.2.4.1)
• The distribution of individual daily drinking water consumption and number of days of
exposure (section 5.2.4.2)
The estimated population served by systems potentially affected by the LT2ESWTR (section
5.2.4.3)
5.2.4.1 Distribution of Infectious Cryptosporidium in Finished Water
The distribution of infectious Cryptosporidium in finished water to which the affected population
is exposed reflects three factors:
• The distribution of total Cryptosporidium concentrations in source water
The fraction of those oocysts that are considered to be infectious
• The removal and inactivation rates of the infectious Cryptosporidium predicted for Pre-
LT2ESWTR and predicted Post-LT2ESWTR treatment conditions
The National Distribution of Cryptosporidium Concentrations in Source Water
Simulated source water Cryptosporidium concentrations are drawn from the occurrence
distributions described in Chapter 4. Section 4.2.2 gives a general overview of the occurrence modeling
approach, and sections 4.4.3 and 4.5.3 provide more detail on the unfiltered and filtered occurrence
models, respectively.
At the broadest level, there are four basic occurrence models that provide input to this EA: three
based on filtered-system data from the ICR, ICRSSL and ICRSSM, and one based on unfiltered-system
data from the ICR. The separate unfiltered system model is motivated by fundamental differences
between the quality of source water in filtered and unfiltered systems. Differences among the three
filtered systems data sets arise from different survey sampling plans, lab methods, and sampling periods
(see section 4.2.1 for a detailed description of these survey differences).
National benefit estimates are derived from each of the three filtered-system data sets and
compared with one another, but they are not combined or weighted in any way. Each of the data sets has
strengths, but none was judged as superior in estimating national levels of filtered systems occurrence.
The fact that there are three different data sets for filtered systems, leading to three distinct occurrence
distributions, reflects significant uncertainty about the true national Cryptosporidium distribution and its
stability over time. Using all three models serves two purposes: it captures this uncertainty and portrays it
clearly, while at the same time providing three distinct, independently drawn, plausible representations of
the true national distribution.
As discussed in section 4.2.2, rather than fitting a single log-normal occurrence distribution to
each of the four data sets described above, the Cryptosporidium modeling approach generates a collection
Economic Analysis for the LT2ESWTR 5-23 December 2005
-------
of log-normal distributions from each data set. So, for example, 1,000 mean and standard deviation pairs
are drawn from the ICR filtered-systems model to serve as input to the risk assessment model. Each pair
defines a single log-normal distribution that could have plausibly generated the ICR survey results for
filtered systems. The other three models, corresponding to the other three data sets, are sampled in the
same way. The result is the four collections of log-normal distributions that are summarized in Exhibits
4.6, and 4.11 through 4.13.
These collections of occurrence distributions serve as inputs to the Monte Carlo risk simulation
(part of the risk assessment model) described in section 5.2.5. In this risk simulation, an outer loop
captures uncertainty about risk parameters, and an inner loop models variability in risk from water system
to water system. For each of these uncertainty loops, a single occurrence distribution is drawn from a
given collection of 1,000 distributions. Then, within the variability loop, the selected log-normal
distribution is used to simulate variability in occurrence—both system-to-system differences in average
Cryptosporidium concentration and sample-to-sample differences overtime. This process is repeated
until 250 uncertainty loops have been completed, yielding 250 national risk curves.
Again, the overall risk model is described in more detail in section 5.2.5. The key point here is
that the occurrence inputs to this risk model are carefully structured to separate uncertainty about the true
national distribution, on the one hand, from the estimated system-to-system variability in
Cryptosporidium concentration on the other. This approach provides a good match between the
occurrence inputs and the general structure of the broader risk model.
Infectious Oocyst Fraction
An important parameter when assessing the risk associated with a given concentration of
Cryptosporidium in a drinking water source is the percentage of oocysts that are infectious. The methods
used to analyze Cryptosporidium in the ICR and ICRSSs measured total oocyst counts without regard to
how many were actually infectious. Because oocysts degrade in the environment, it is expected that only
a fraction of the oocysts counted in these surveys would be capable of causing infection in a susceptible
host. Consequently, the distributions of total Cryptosporidium occurrence based on these surveys are
believed to overestimate the concentration of infectious oocysts.
Further, the parameter actually of concern is the ratio of the infectious oocyst fraction in the
environment to the same fraction in dose-response studies. Even in the best controlled laboratory studies,
the fraction of infectious oocysts is less than 100 percent (but was above 80 percent in the studies
considered here).
There is no direct way to assess the infectivity of oocysts counted with the ICR Method in the
ICR or with Methods 1622/23 in the ICRSSs. Rather, related information is gleaned from two sources:
the physical structure of observed oocysts and a comparison study where samples were analyzed with
both Method 1623 and a cell culture test for oocyst infectivity. From these two sources, an estimate was
made of the most likely proportion of counted oocysts (in the environment) that were infectious (at least
in a laboratory setting).
As discussed in section 4.5.3, Cryptosporidium oocysts counted with the ICR Method or Methods
1622/23 are characterized in one of three ways: (1) those with internal structures, i.e., those having
recognizable structure consistent with Cryptosporidium; (2) oocysts with amorphous structures, which
indicates that material is present in the oocyst, but it cannot be confirmed as Cryptosporidium; or (3)
empty oocysts. Assignment of these labels is dependent upon analyst judgment and none is a certain
indicator of whether an oocyst is truly infectious. Oocysts with internal structures are generally
considered to have the highest likelihood of being infectious, though laboratory studies have shown
Economic Analysis for the LT2ESWTR 5-24 December 2005
-------
oocysts can lose infectivity without loss of internal structures. Oocysts with amorphous structures may be
still infectious or, alternatively, may be some other microorganism that mimics the structure and
properties of Cryptosporidium. Oocysts that are empty of internal structures are assumed to be non-
infectious (LeChevallier et al. 1997a).
In the ICR data set, laboratories characterized 23 percent of the oocysts counted as having
internal structures, 39 percent having amorphous structures, and 38 percent as being empty. With the
ICRSSs, 37 percent of the oocysts had internal structures, 47 percent had amorphous structures, and 16
percent were empty. If it were assumed that the empty oocysts could not be infectious, then these data
suggest that the percentage of counted oocysts that were infectious were at most 62 percent in the ICR
and 84 percent in the ICRSS.
The lower percentage of empty oocysts in the ICRSS, versus the ICR, may be attributable to the
improved sample purification technique in Methods 1622/23. This technique, immunomagnetic
separation, prevents many non-Cryptosporidium particles from being transferred to the slide for
examination; some of these non-Cryptosporidium particles may have been incorrectly counted as empty
oocysts in the ICR (Connell et al. 2000). Moreover, the LT2ESWTR would require use of Methods
1622/23 for assigning systems to bins, so the ICRSS data may be more reflective of data that would be
generated under this rule.
A study by LeChevallier et al. (2003) provides another indication of the percentage of oocysts
counted by Method 1623 that are infectious. This study involved intensive sampling of six source waters
for Cryptosporidium and other microbiological and water quality parameters. Each Cryptosporidium
sample was analyzed by both Method 1623 and a method that used cell culture and polymerase chain
reaction (CC-PCR) to measure viability and infectivity. Cryptosporidium oocysts were detected in 60 of
593 samples (10.1 percent) by Method 1623 and infectious oocysts were detected in 22 of 560 samples
(3.9 percent) by the CC-PCR procedure. Recovery efficiencies for the two methods were similar.
According to the authors, these results suggest that approximately 37 percent of the Cryptosporidium
oocysts detected by Method 1623 were viable and infectious. Only one sample was positive by both
Method 1623 and CC-PCR, though this result is consistent statistically with the low oocyst concentration.
When using the data sets derived from the ICRSS, EPA characterized the percent of oocysts that
are infectious as an uncertain variable with a triangular distribution having a lower bound of 30 percent, a
mode of 40 percent, and an upper bound of 50 percent. The mode is consistent with results from the
LeChevallier et al. study (2003) where the number of samples with infectious oocysts was 37 percent of
the number with oocysts counted using EPA Method 1623. It is also consistent with the 37 percent of
oocysts counted during the ICRSS that had internal structures, which are considered the most likely to be
infectious. The bounds were set at +/- 25 percent of the mode, balancing the good quality of the
LeChevallier et al. (2003) data with the uncertainty in applying this result on a national basis. This
distribution also recognizes that an unknown fraction of the 47 percent of oocysts counted in the ICRSS
with amorphous structures were infectious, which could lead to a total fraction of infectious oocysts
greater than 40 percent; alternatively, a fraction of the oocysts with internal structures were likely not
infectious, which could lead to a total fraction of infectious oocysts less than 40 percent.
When using the data sets derived from the ICR, EPA characterized the percent of oocysts that are
infectious as an uncertain variable with a triangular distribution having a lower bound of 15 percent, a
mode of 20 percent, and an upper bound of 25 percent. The lower range for the ICR distribution reflects
the higher rate of empty oocysts, which are considered to be non-infectious, detected by the ICR Method
(38 percent in the ICR vs. 16 percent in the ICRSS), and the lower rate of oocysts with internal structures
(17 percent in the ICR vs. 39 percent in the ICRSS).
Economic Analysis for the LT2ESWTR 5-25 December 2005
-------
Pre-L T2ESWTR Removal/In activation of Cryptosporidium
Filtration is currently the primary treatment mechanism used in PWSs to remove Cryptosporidi-
um. Finished water Cryptosporidium concentrations developed for this risk characterization reflect
predicted filtration improvements to meet IESWTR and LT1ESWTR requirements. Chapter 4, section
4.5.4 presents the methodology used to estimate Pre-LT2ESWTR (Post-IESWTR and Post-LTIESWTR)
removal levels. To summarize, Pre-LT2ESWTR removal is modeled as triangular distributions as
follows:
• For small systems (serving fewer than 10,000 people): 2 to 4 log range of Cryptosporidium
removal to capture system-to-system variability; possible modes of the triangular
distributions between 2.25 and 3.25 to capture uncertainty in the "true" distribution of
removal.
• For large systems (those serving at least 10,000 people): 2 to 5 log range of removal to
capture system-to-system variability; possible modes of the triangular distributions between
2.5 and 3.5 to capture uncertainty in defining the "true" distribution of removal.
These distributions are intended to bound the uncertainty in the most likely removal values, as
defined by the variable modes of the triangular distributions. Uncertainty and variability in removal are
modeled independently of source water Cryptosporidium concentration. Also, removal is modeled
independently of filtration type (e.g., conventional or direct). This is primarily because the analysis
assumes that few systems designed their plants to account for Cryptosporidium concentrations. In
addition, reviews of the ICR data do not reveal treatment designs to be correlated to Cryptosporidium
levels. Further, few medium and small systems have collected data that they could have used for this
purpose.
As noted in Chapter 4, a small fraction of filtered plants7 are predicted to have added advanced
technologies that provide 5.5 logs of removal or inactivation of Cryptosporidium before the
implementation of LT2ESWTR. Although several technologies are capable of this level of performance,
the model only specifically takes account of those using microfiltration/ultrafiltration (MF/UF). Those
plants that had MF/UF in place are removed from the baseline for filtered plants because they are exempt
from additional monitoring and treatment requirements under the LT2ESWTR. Therefore, no
adjustments are made in the risk model to account for these plants. A small number of plants8 are
predicted to install these technologies to comply with the Stage 1 or Stage 2 Disinfection Byproduct Rule
(Stage 1 DBPR or Stage 2 DBPR). These plants are also excluded from the model for estimating risks,
although first-round monitoring costs are included. Benefits related to Cryptosporidium reduction for
these plants are attributed to the other rules and thus not captured in this EA.
Pre-LT2ESWTR finished water Cryptosporidium concentrations (mean and median values) are
summarized in Chapter 4, Exhibits 4.19a and b. Values are slightly lower for medium and large systems
because they are predicted to have better removal performance following the implementation of the
IESWTR and LTIESWTR
7 Percent of plants shown in Exhibit 4.11, column C.
8 Percent of plants shown in Exhibit 4.11, column F.
Economic Analysis for the LT2ESWTR 5-26 December 2005
-------
Post-LT2ESWTR Removal/Inactivation of Cryptosporidium
The additional Cryptosporidium reduction gained through the addition of advanced technologies
for the LT2ESWTR is estimated through a four-step process:
STEP 1—Predict source water occurrence for each plant
At the beginning of each uncertainty loop, the model defines an occurrence distribution by
randomly selecting a log-normal mean and standard deviation. At each iteration (at the variability level),
the model then randomly draws annual plant means from the distribution defined in the associated
uncertainty step, to simulate plant-to-plant variability in occurrence.
STEP 2—Predict bin classification for each plant
Section 4.5.6 and Appendix B summarize the predicted bin classification for the LT2ESWTR
Preferred Regulatory Alternative, as well as the two alternative bin classifications considered in this EA
(regulatory alternatives A2 and A4). In general, the risk model uses a probability function that takes a
"true" source water concentration and adjusts for test method recovery to classify a plant into a treatment
bin. Predicted binning has substantial impacts on estimated costs and benefits of the LT2ESWTR;
therefore, a sensitivity analysis was performed to evaluate effects of predicted bin assignment based on
alternative source water occurrence distributions (see Appendix B).
STEP 3—Adjust bin classification for plants with treatment credits prior to the LT2ESWTR
Some plants may have advanced treatment in place following the IESWTR and LT1ESWTR.
Because these plants installed this treatment prior to the LT2ESWTR, benefits and costs associated with
the associated toolbox options are not attributable to this rule. The risk model takes into account the
higher level of treatment achieved by these systems for both pre-LT2ESWTR finished water occurrence
and determining the required log credit. EPA estimated the percentage of plants that will achieve the
treatment requirements for combined filter performance, presedimentation basin, and secondary filtration
toolbox options, as described in Appendix A. Each of these toolbox options provides 0.5 log treatment
credit. Additionally, a few of these plants may have more than one of these options in place and thus
receive 1.0 log credit. Exhibit 5.6 shows the percent of plants estimated to receive 0.5 or 1.0 log credit.
Exhibit 5.6: Percent of Plants With Pre-LT2ESWTR Treatment Credit
System Size
(population served)
Very Small and Small (10,000 and fewer)
Medium (10,001 -100,000)
Large (more than 100,000)
Percent of Plants Achieving
37%
55%
58%
Source: Appendix A, Exhibit A.7.
Economic Analysis for the LT2ESWTR
5-27
December 2005
-------
STEP 4—Determine actual log reduction achieved
Systems have various treatment options available to meet Cryptosporidium reduction
requirements for a given bin. Some technologies (such as bag and cartridge filters) are projected to be
used only by small systems, and some technologies can only be used by larger systems. For example,
chlorine dioxide, ozone, and secondary filters are judged to be impractical for systems serving fewer than
500 people (Exhibit 6.9). Once those constraints are accounted for, the selection of technologies (detailed
in Chapter 6 and Appendix F) is performed using a "least-cost" approach, whereby EPA estimates (for the
purpose of estimating the national costs attributable to this rule) that systems will, for the most part, select
the least costly technology available to meet treatment requirements for that bin. In many cases, the least
costly technology results in higher levels of Cryptosporidium inactivation or removal than required for
that bin (this is always the case when UV is selected). Therefore, this risk analysis incorporates estimates
of "actual" reduction achieved beyond bin requirements.
Exhibit 5.7a and 5.7b present the predicted Cryptosporidium reduction achieved for systems in
five action levels for systems without and with Pre-LT2ESWTR credit, respectively. Different reduction
estimates are shown for very small, small, medium, and large systems because the least-cost decision tree
changes as system size changes, primarily because of different economies of scale among treatment
technologies (that is, as system size increases, some technologies will become less costly per unit of water
treated than others). Also, different technologies are available for different sizes of systems, usually
because of practical limitations such as very small systems not having 24-hour per day staffing. The
technology selections for each bin are developed independent of regulatory alternative and source water
occurrence distribution. For example, for any regulatory alternative that requires 2.0 log reduction, it is
estimated that large systems will actually select technologies that achieve a 3.0 log reduction 90 percent
of the time (the remaining 10 percent achieve 2.0 and 2.5 log reduction), regardless of source water
occurrence distribution.
Economic Analysis for the LT2ESWTR 5-28 December 2005
-------
Exhibit 5.7a: Predicted Log Removal Achieved for Systems without Credits1
Very Small
Systems
Serving < 500)
in o
£ s aT
*!!
W 0) Q.
.E a>
= > o>
ra £ o>
J si-
lo •sr
|S &
S,? S.
« 0.0,
c c o>
3 > °l
=5 S g
*SB
-------
Exhibit 5.7b: Predicted Log Removal Achieved for Systems with Credits1
Very Small
Systems
(Serving < 500)
Small Systems
(Serving 500 to
9,999 people)
Z o •HT
|8&
s,s a.
« 0.0,
c c o>
1 E 5-
•D 01 o>
I^S
, ra Q.
OT 0) 0
01 c O
S) > <=
5*1
Actual Log
Reduction
Achieved
0.5
1.0
1.5
2.0
2.5
3.0
Total
0.5
1.0
1.5
2.0
2.5
3.0
Total
0.5
1.0
1.5
2.0
2.5
3.0
Total
0.5
1.0
1.5
2.0
2.5
3.0
Total
Targeted Log Reduction^
0.5 Log
0.00%
90.00%
0.00%
0.00%
1 .00%
9.00%
100%
1%
90%
0%
0%
0%
9%
100%
11%
2%
0%
0%
0%
87%
100%
12%
2%
0%
0%
0%
86%
100%
1.0 Log
0.00%
90.00%
0.00%
0.00%
1 .00%
9.00%
100%
0%
90%
0%
0%
1%
9%
100%
0%
8%
0%
1%
1%
90%
100%
0%
8%
0%
1%
1%
90%
100%
1 .5 Log
0.00%
0.00%
0.00%
90.00%
1 .00%
9.00%
1 00%
0%
0%
0%
26%
1%
73%
1 00%
0%
0%
1%
5%
4%
90%
1 00%
0%
0%
2%
5%
4%
90%
1 00%
2.0 Log
0.00%
0.00%
0.00%
90.00%
1 .00%
9.00%
1 00%
0%
0%
0%
26%
1%
73%
1 00%
0%
0%
0%
5%
5%
90%
1 00%
0%
0%
0%
5%
5%
90%
1 00%
2.5 Log
0.00%
0.00%
0.00%
0.00%
10.00%
90.00%
100%
0%
0%
0%
0%
10%
90%
100%
0%
0%
0%
0%
10%
90%
100%
0%
0%
0%
0%
10%
90%
100%
Notes:
1Cells show percent of total number of systems assumed to achieve actual log reduction levels to
meet specific treatment bin requirements.
2Log reduction requirements associated with treatment bins for all regulatory alternatives.
Source: Appendix F, Exhibits F.3 through F.18, "Actual Log Credit" and "Percent of Plants
Selecting Technology by Bin" columns.
5.2.4.2 Distribution of Individual Daily Drinking Water Consumption
The second element of the exposure assessment is the characterization of drinking water
consumption in the exposed population. EPA bases its estimates of per-capita water ingestion on data
collected by the U.S. Department of Agriculture's (USDA) 1994-96 Continuing Survey of Food Intakes
by Individuals (CSFII). Data derived from this survey are presented in the report, "Estimated Per Capita
Water Ingestion in the United States" (USEPA 2000k).
The EPA water ingestion study reports water consumption data for two different aggregations of
the population: all respondents (which is used in this exposure assessment) and only those respondents
who report consuming water directly ("consumers"). The category of all respondents is more appropriate
to this exposure assessment because EPA assumes that all people consume or are exposed to tap water,
even if they reported no tap water consumption in the CSFII. That is, even people who report no
consumption of public water ingest water indirectly (for example, through washing vegetables and other
foods, and consuming foods prepared in restaurants) or are potentially exposed pathogens in tap water
Economic Analysis for the LT2ESWTR
5-30
December 2005
-------
(during showers and brushing teeth, for example). The survey also reports information by type of source
water (community water, bottled water, other sources, and non-reported source). The survey questions
categorized respondents based on their reported "main"source of direct water and indirect water. Thus,
many respondents who reported that their main source of drinking water was bottled water or another
source may still consume water from community sources at least some of the time. Likewise, respondents
who categorize their drinking water as being mainly from a community source may also consume bottled
water and water from other sources. More important, those who are not now served by community water
systems would report "other sources" or bottled water as their main source of water.
For estimating the impacts of the LT2ESWTR, EPA is most interested in the consumption of
water by those served by public water systems. The EPA study reports consumption data for source water
types and as national averages, where one group cannot be subtracted from the total without affecting the
value of the national average. Thus, the consumption of those who reported no source or "other sources"
as their main source of drinking water is included in the national average, because subtracting these
categories would lead to an underestimate of average consumption levels. Lacking the available data, it is
reasonable to assume their consumption patterns are similar to those served only by "community" systems
(analogous to public systems under SDWA). Therefore, no adjustments to average consumption were
made for those who reported no source or other sources as their main source of water.
Bottled water, however, is thought to replace of tap water, and thus an adjustment to average
consumption is more appropriate. In the CSFII study, 13 percent of all water was consumed by those
who categorized bottled water as their "main" source of water for direct or indirect ingestion. To reflect
this pattern in the exposure assessment, EPA uses the mean of water consumption from all sources, less
the mean of water consumed by those who identify bottled water as their main source of water for either
direct or indirect ingestion. Because the survey did not attempt to determine for each individual the
proportion of water from each source, the approach used in the exposure assessment may understate or
overstate actual consumption from public water systems, depending on the extent and direction of overlap
in drinking water sources. EPA believes, however, that making this adjustment produces an estimate of
drinking water consumption closer to actual practices.
The exposure assessment uses a mean national consumption of 1.071 liters per person per day.
This value is the mean consumption from all water sources (1.232 liters per person per day) less the mean
consumption of those who reported bottled water as their main source of drinking water (0.161 liters per
person per day).
The exposure assessment does not include variability in the mean (1.071 liters per person per
day). In the CSFII report, there is information about the 90 percent confidence interval of the mean for
both the total (All Sources) and for those relying mainly on bottled water. The 90 percent confidence
range for All Sources is 1.199 to 1.265 liters per person per day with a mean of 1.232, or less than +/-3
percent. The 90 percent confidence range for consumption by those relying mainly on bottled water is
0.147 to 0.176 liters per person per day with a mean of 0.161, or under +/-10 percent9. The statistics for
these ranges alone cannot be used to derive a confidence range for the mean estimate used in the exposure
assessment. Given the narrow boundaries in the confidence intervals, EPA judged it more important to
adjust the average consumption to reflect the use of bottled water (a 13 percent reduction) than to use the
data for All Sources only, which would allow for the use of confidence bounds. Variability in
individuals' consumption was reflected in calculating the individual risk reductions estimates (Exhibits
9 The difference in the variance of these two means might challenge the appropriateness of subtracting their
means; however, EPA has judged that the value of obtaining a more accurate estimate of tap water consumption is
more relevant to understanding the impacts of the rule alternatives than is additional information about the
variability.
Economic Analysis for the LT2ESWTR 5-31 December 2005
-------
5.14 and 5.15), but that variability is not reflected in the overall estimates of variability in illnesses, which
are based solely on the mean. Exhibit 5.8 shows the percentile values for individual consumption from all
sources and from all sources less bottled water.
Exhibit 5.8 Distribution of Individual Daily Drinking Water Consumption
(L/person/day)
Percentile
1
5
10
25
50
75
90
95
99
All Sources
0.01
0.16
0.28
0.57
1.04
1.63
2.34
2.91
4.81
All Sources
Less Bottled Water
(13 percent)
0.01
0.14
0.25
0.50
0.90
1.42
2.04
2.53
4.18
Source: All Sources from USEPA 2000k.
5.2.4.3 Population Affected by the LT2ESWTR and Exposure
The number of systems and the total population affected by the LT2ESWTR are discussed in
Chapter 4 for both unfiltered (section 4.4.2) and filtered (section 4.5.2) plants. Exhibit 5.9 summarizes
these numbers. Note that unfiltered plants are all within CWSs. Note also that more than 85 percent of
the population affected by the LT2ESWTR are served by medium and large CWSs.
Economic Analysis for the LT2ESWTR
5-32
December 2005
-------
Exhibit 5.9: Number of Systems, Population Served, and Annual National
Exposure by System Type
System Size
(Population Served)
Number of
Systems
A
Total Population
Served
B
Annual National
Exposure in
Person-Days
C
Unfiltered CWSs
<500
500-<10,000
10,000 -<100,000
> 100,000
Totals
Percent of All PWSs
5
32
17
6
60
0.8%
1,270
137,470
135,003
554,838
828,581
0.5%
444,500
48,114,500
47,251,050
194,193,300
290,003,350
0.5%
Filtered CWSs
<500
500-<10,000
10,000 -<100,000
> 100,000
Totals
Percent of All PWSs
993
2,438
1,327
283
5,041
64.1%
195,338
8,510,213
42,799,971
116,452,533
167,958,055
92.1%
68,368,326
2,978,574,402
14,979,989,877
40,758,386,590
58,785,319,195
95.5%
Filtered NTNCWSs
<500
500-<10,000
10,000 -<100,000
> 100,000
Totals
Percent of All PWSs
465
205
5
1
676
8.6%
73,997
324,040
125,710
166,735
690,482
0.4%
18,499,141
81,010,011
31,427,418
41,683,817
172,620,386
0.3%
Filtered TNCWSs
<500
500-<10,000
10,000 -<100,000
> 100,000
Totals
Percent of All PWSs
1,883
193
12
3
2,091
26.6%
167,600
275,237
221,299
12,144,000
12,808,136
7.0%
30,168,000
49,542,660
39,833,820
2,185,920,000
2,305,464,480
3.7%
Sources:
[A] Unfiltered CWSs: Exhibit 4.5, Column A; Filtered PWSs: Exhibit 4.12, Column C.
[B] Unfiltered CWSs: Exhibit 4.5, Column D; Filtered PWSs: Exhibit 4.12, Column E.
[C] Column B x Exposure Days Per Year from Exhibit 5.10.
The risk assessment model also accounts for the number of days per year that individuals in the
affected population consume water from the different types of systems. This is needed to calculate the
annual risk of infection and illness, as described above. The model assigns different exposure durations
to people served by community water systems (CWSs), nontransient noncommunity water systems
(NTNCWSs), and transient NCWSs (TNCWSs). Exhibit 5.10 shows EPA's estimates of the number of
days per year tap water is consumed by users of different types of water systems.
Economic Analysis for the LT2ESWTR
5-33
December 2005
-------
Exhibit 5.10: EPA Estimates for Exposure Days1
Type of System
cws
NTNCWS
TNCWS
Exposure Days Per Year
350
250
1802
1 Number of days in which tap water is consumed.
2 A triangular distribution 90-270 (mode 180) is used to represent "person days"
reflecting a large number of individuals, with fewer days exposure per individual
to more appropriately characterize consumption and risk among these transient
populations.
National exposure (days per year) is calculated as a function of exposure days and population
served. These components are addressed in two separate parts of the risk model. The first part of the
model incorporates average days of exposure, allowing individual annual risk to be estimated with
precision for each type of PWS. The second part of the model incorporates population served,
multiplying it by individual annual risk for each system type to obtain national risk.
Overall exposure is not directly proportional to population served due to the differences in
average days exposed for the three system types. For example, 100 people served by a CWS will have a
greater aggregate exposure than 100 people served by a TNCWS due to the greater number of days
exposed for CWS.
To account for the transient nature of the population served by TNCWSs, population is multiplied
by a factor between 90 and 270 (drawn from the triangular uncertainty distribution described in Exhibit
5.10). Without this adjustment, the reported TNCWS population would underestimate the true population
exposed, since each TNCWS system reports only a peak-season population and each individual exposure
is assumed to be just 10 days. The lower end of the triangular distribution represents a 3 month peak
season (9x10 days = 90 days) and the upper end a 9 month peak season (27 x 10 = 270 days). By
making this adjustment to the aggregate TNCWS population, rather than to the average number of
TNCWS days exposed, the analysis of individual risk remains distinct from the analysis of aggregate
national impact.
5.2.5 Risk Model Structure
The risk assessment model integrates the dose-response and exposure assessment components
discussed above into a Monte Carlo simulation structured to characterize (1) the distribution of individual
risk of illness and mortality and (2) the total number of national illnesses and deaths annually due to
Cryptosporidium in finished drinking water. Modeling is conducted for two treatment conditions: Pre-
LT2ESWTR and Post-LT2ESWTR, the former representing baseline (no action) conditions and the latter
incorporating assumptions of treatment improvements due to the LT2ESTWR. The difference between
the results obtained for the improved conditions and for Pre-LT2ESWTR conditions (in terms of cases of
illness and deaths avoided) constitute the quantified benefits of the rule.
The risk assessment modeling is carried out in two steps. The first step focuses on calculating the
annual risk of illness. The endpoints of this step are (1) the distribution of annual risks of illness
experienced by different individuals in the population, reflecting variability in exposure and infectivity
Economic Analysis for the LT2ESWTR 5-34 December 2005
-------
conditions, and (2) the estimated average annual risk for the population as a whole, with confidence
intervals on that estimate to reflect uncertainty. The flowchart in Exhibit 5.11 summarizes this first step
of risk mode ling.
The second step applies the estimated average annual risk of illness, including secondary spread,
to the overall population to estimate the annual cases of illness, the annual number of deaths, and the
regulatory benefits in terms of illnesses and deaths avoided due to each proposed rule. Discussed further
in section 5.3, the second step also integrates information about the monetized values of illnesses and
deaths avoided to provide a dollar value for the overall benefits of the rule.
Economic Analysis for the LT2ESWTR 5-35 December 2005
-------
Exhibit 5.11: Flowchart of Risk Model-Step 1: Computing Annual Individual
Risk of Illness
Select Occurrence
Distribution
Pre-LT2 Baseline
Select Plant-Mean Source Water
Concentration of Total Oocysts
LT2 Reduction
Select Pre-LT2 Log
Removal
Distribution
Calculate Pre-LT2 Log
Removal/Inactivation
Filtered
Simulate 24 Monthly Total
Oocyst Concentrations
Determine
Maximum
Running Annual
Average (RAA)
Apply Pre-LT2
Credit
Determine Rule
Alternative
Treatment Bin
(and Target Log-
Removal)
Determine
Average
Concentration
Calculate LT2 Log
Removal/Inactivation
/Select Treatment N.
^—^ Technology from y
NJ)ecision Matrices/
Select Infectious Oocyst Fraction
Calculate Finished Water
Concentration of Infectious Oocysts
Apply Daily Consumption of Drinking Water
Calculate Daily Dose of
Cryptosporidium
Apply Cryptosporidiosis Infectivity Distribution
Calculate Daily Risk of
Infection
Apply Days of Exposure per Year
Calculate Annual Risk of
Infection
Calculate Annual Risk of Illness
Apply Morbidity Factor
Risk distributions representing the different system types,
occurrence datasets, and regulatory conditions
Economic Analysis for the LT2ESWTR
5-36
December 2005
-------
Step 1 of the Risk Characterization
Step 1 is structured as a two-dimensional Monte Carlo simulation to appropriately address
variability and uncertainty in model inputs. The basic algorithm for Step 1 of the modeling is:
PM = M*(l-[exp((-C*v*I)*r)]n)
or more simply,
= M*(l-exp(-C*v*I*r*n))
Where:
PM is the individual annual risk of illness
M is the morbidity factor, or the probability of illness given an infection
C is the Cryptosporidium concentration in finished water (oocysts/Liter)
v is the ratio of the percentage of infectious oocysts in the environment to the percentage of
infectious oocysts in doses tested in clinical studies
I is mean individual daily drinking water consumption (Liters)
r is the infectivity dose-response parameter (the expected probability of a single ingested oocyst,
surviving long enough to reach an infection site in the body)
n is the average annual days per year of exposure
This formula is equivalent to that presented earlier in section 5.2.3 describing the dose-response
function for calculating the risk of illness. The difference is that the d variable (dose in infectious
oocysts/day) has been expanded to show its components, namely Cryptosporidium occurrence in finished
water (measured oocysts/liter), infectious Cryptosporidium rate (infectious oocysts/measured oocyst), and
drinking water consumption (liters/day).
As noted above, the first step of the risk assessment model was structured as a two-dimensional
Monte Carlo simulation. A two-dimensional simulation is used when the model includes both uncertainty
and variability components in the inputs, and where it is necessary to clearly distinguish the influences of
these elements on the model output. SAS v8.2 software was used to carry out the analysis (see Appendix
T for programming details).
In the risk formula shown above, the uncertainty and variability components are:
Uncertainty: Data set representing source water concentration (ICR, ICRSSL, ICRSSM)
True distribution of source water oocyst concentration
True distribution of Pre-LT2ESWTR removal
Morbidity factor (the probability of illness given an infection) (M)
Infectious oocyst fraction (per total oocyst's detected) (v)
True mean dose-response infectivity parameter (r)
Variability: Source water concentration (from a selected distribution)
Pre-LT2ESWTR Cryptosporidium removal (from a selected distribution)
Predicted binning
Earned log credit for enhanced filtration (0 or 0.5 log)
Actual log reduction achieved due to LT2ESWTR treatment
Variable daily ingestion (applied to step 1 results) (I)
Economic Analysis for the LT2ESWTR 5-37 December 2005
-------
The form and values for these variability and uncertainty distributions were discussed in the
preceding sections on hazard identification (section 5.2.2) and exposure assessment (section 5.2.4).
Exhibit 5.12 summarizes the model parameters.
In the two-dimensional simulation structure, a set value or a distribution of values is randomly
selected for each of the uncertainty parameters identified above. These uncertainty parameters are then
"frozen," and a specified number of iterations are performed, generating randomly selected values for the
variability parameters (referred to as "inner loops"). These results are stored, and a second set of
uncertainty parameters is chosen, for which the specified number of iterations are again run for the
variability factors. This process is repeated for some specified number of sets of uncertainty parameters
(referred to as "outer loops").
In the risk model used here, 250 sets of uncertainty parameters, or outer loops, were generated,
each with 1,000 variability iterations, or inner loops. For the model end-point of this step of the analysis
"PM, the annual individual risk of illness," the following key results were computed for each of the 250
uncertainty iterations based on the results of the 1,000 variability iterations performed within each of
those uncertainty loops:
• Mean
Standard Deviation
• Minimum
• Maximum
• Percentiles (every 5th percentile between the 5th and 95th)
• Percentage of population having individual risk levels exceeding 10"2, 10"3, 10"4, 10"5, 10"6,
and 10-8
There were 250 sets of these estimates for individual annual risk of illness produced, each
reflecting a different possible combination of the uncertainty factors. These 250 estimates of each of the
above statistics were then used to compute an overall mean value for each simulation group and
confidence bounds on those mean values.
By structuring the first step of the risk analysis in this way, it was possible to characterize both
the distribution of individual annual risk of illness in the affected population (reflecting variability in
Cryptosporidium occurrence levels and daily water consumption), and the overall population average
annual risk of illness. This latter value, and the associated uncertainty in it reflected by the 250
alternative values obtained, was then used in the second step of the modeling to compute the number of
cases of illness and death by applying these population average risks to the population exposed.
Economic Analysis for the LT2ESWTR 5-38 December 2005
-------
Exhibit 5.12: Overview of Risk Assessment Model Parameters
Variability or
Variable Units Values/Range Uncertainty Depends upon
Top Level Factors
Proposed regulatory alternative
PWS size
PWS type
PWS filtration status
Cryptosporidium occurrence dataset
A1 , A2, A3, A4
Nine size categories
CWS, NTNCWS, TNCWS
Filtered, Unfiltered
ICR, SSL, SSM
Variability
Variability
Variability
Uncertainty
Cryptosporidium Exposure
Source water Cryptosporidium concentration (C)
Infectious oocysts per oocyst detected (v)
Drinking water consumption (I)
Average annual daily exposures (n)
Population at risk (pop)
TNCWS peak population multiplier
plant mean oocysts/liter
rate/probability
L/person/day
per individual
per PWS size and type
multiplication factor
Est [5th, 95th] %tiles
[0.001 1,2.7657] ICR
[0.0033, 0.6765] SSM
[0.0059, 0.3460] SSL
[0.0004, 0.1 177] Unfiltered
[15%, 25%] ICR
[30%, 50%] SSM, SSL
Mean = 1 .07
10 (TNCWS)
250 (NTNCWS)
350 (CWS)
see Exhibit 5.8
[9, 27]
Both (V and U)
Uncertainty
Variability
Constant
Constant
Constant
Constants
Variability
Occurrence dataset
Filtration
Dataset
PWS type and size
PWS type (TNCWS only)
Dose-Response Model
Dose-response mean infectivity (r)
Morbidity rate (prob of illness given infection) (M)
Secondary illness factor
Mortality rate (prob of death given illness) (F)
rate/probability
rate/probability
rate/probability
rate/probability
[0,1] See Appendix N
[30%, 70%]
[10%, 40%]
1 .06E-04 (Filtered)
1 .66E-04 (Unfiltered)
Uncertainty
Uncertainty
Uncertainty
Constants
Filtration
(Population characterisitics)
Cryptosporidium Treatment Reductions
Pre-LT2ESWTR Cryptosporidium reduction
Pre-LT2ESWTR enhanced filtration credit
LT2 treatment bin removal requirement
Predicted LT2 technology selection
Actual Cryptosporidium reduction due to LT2 treatment
log(oocysts/liter)
log(oocysts/liter)
log(oocysts/liter)
log(oocysts/liter)
2.0 to 4.5 log
0 to 0.5 log
0 to 2.5 log
LT2 toolbox options
0 to 3 log
Both (V and U)
Variability
Variability
Variability
Variability
PWS size
PWS size
Regulatory alternative
Occurrence dataset
Infectious oocyst rate
PWS size
PWS type (including Filtration)
Predicted treatment bin
Predicted treatment bin
Predicted LT2 technology selection
Economic Analysis for the LT2ESWTR
5-39
December 2005
-------
Step 2 of the Risk Characterization
In Step 2 of the risk analysis, the key algorithms used were:
CM = PM x Pop
CF = CM x F
Where:
CM is the cases of illness in the affected population
PM is the distribution of individual annual probability of illness
Pop is the number of individuals in the affected population (a probability distribution for
TNCWS)
CF is the count of fatalities in the affected population
F is the probability of fatality given an illness (two values, see section 5.2.3)
Step 2 was conducted with a Monte Carlo simulation (separate from the simulation in Step 1) in
which the variable PM(Avg) is treated as an uncertainty variable, values for which were derived from a
custom distribution based on the 250 estimates of the mean individual annual risk obtained in Step 1 of
the analysis. From this process, estimates of the cases of illness and mortality were computed, as well as
the confidence bounds on those estimates, for the various baseline and Post-LT2ESWTR assumptions
regarding Cryptosporidium removal from source water.
Secondary Spread
The last step of Step 2 is to adjust the number of illness cases to account for secondary spread
(mortality is calculated from illness and this is also affected by secondary spread). Secondary spread in
this case is infection passed through contact with an individual initially infected by ingestion of
contaminated water. Secondary spread is quantified by the ratio of secondary cases to primary cases. The
ratio varies depending on a number of factors, such as whether infected persons are symptomatic (or
asymptomatic "carriers"), the age, health and immune status of the exposed, and sanitary conditions
within the household, office, or day care centers.
The secondary spread rate associated with endemic waterborne cryptosporidiosis is estimated
using data on secondary spread and household secondary attack rate compiled from past cryptosporidiosis
outbreaks (Exhibit 5.13). The outbreak data reported in the exhibit have secondary spread or household
secondary attack rates ranging from 4 percent to 46 percent. Most (8 of 11) values are in the range
between 15 percent and 37 percent.
In analyzing the available outbreak data, it is necessary to be aware of at least three potential
effects. First, infection by Cryptosporidium appears to confer limited immunity, so the secondary spread
rate may be affected by the immune status (previous infection history) of the potential secondarily
exposed population. Second, the number of secondary cases during a common source outbreak may be
limited because the outbreak is so large that most people are affected by the common source, so many
fewer people are available to be exposed through secondary spread. Third, secondary spread rates
associated with children (who often acquire infection in day-care centers) are high from the frequent
handling of soiled diapers and training pants and poor toddler hygiene habits.
The outbreak data in Exhibit 5.13 suggest a triangular distribution for the range of possible
secondary spread rates associated with endemic exposure. A preponderance of the rates are in the middle
of the distribution rather than at the margins, yet the unusual scenarios discussed above (and others) will
occasionally lead to extremes on both the high and low side of the typical range.
Economic Analysis for the LT2ESWTR 5-40 December 2005
-------
To capture uncertainty, a triangular distribution was used with a low at 10 percent, a high at 40
percent and a most likely value of 25 percent. The peak at 25 percent reflects the average value of
secondary spread and attack rates shown in Exhibit 5.13 The high value of 40 percent in the distribution
is below the highest reported rate in Exhibit 5.13, and the low value of 10 percent is above the lowest
reported rate. EPA chose to eliminate extreme values to minimize the impact of any hidden bias in
available data.
With the basic risk model described, we move on to a discussion of how the model is used to
estimate LT2ESWTR benefits. In short, the model is used to estimate baseline conditions (Pre-LT2) and
then expected conditions arising under each of the proposed rules. The differences-baseline to rule-are
the basis for obtaining the "cases avoided" (and confidence bounds on those estimates of cases avoided).
As further discussed in section 5.3, estimates of "cases avoided" generated in Step 2 of the risk modeling
are integrated with estimated costs of illness and mortality to produce the monetized benefit estimates for
the LT2ESWTR. The following subsections present these results.
5.2.6 Individual Annual Risk Distributions
The benefits of the LT2ESWTR regulatory alternatives have been explicitly estimated for two
health end-points: avoided illnesses and avoided deaths due to endemic cases of cryptosporidiosis. These
benefits are measured both in terms of the number of cases of illness and death avoided, and in terms of
the monetized value of those avoided cases. This section focuses on the reduction in risk as measured by
anticipated changes in the distribution of individual risks in the exposed population (Step 1 of the risk
characterization). Section 5.2.7 shows how those changes in the distribution of individual risks
aggregated across the exposed population translates into the reduction in cases of illness and death from
Cryptosporidium on a national level (Step 2). Results for the small population of unfiltered systems are
presented first, followed by filtered systems. Section 5.3 extends the benefits analysis to address the
monetization of those avoided cases.
The reader is reminded that this section and the next are narrowly focused on risks and benefits
related to endemic cases of cryptosporidiosis. Other recognized benefits from the LT2ESWTR that are
not explicitly captured in the analysis presented here are those associated with avoided Cryptosporidium
outbreaks, as well as benefits of avoided endemic and outbreak illnesses and deaths from waterborne
pathogens other than Cryptosporidium that might also be prevented or controlled by these regulations.
As summarized in the previous section, a key output of the risk model is the estimated
distribution of annual individual risks of endemic cryptosporidiosis. The variability in individual risks in
the exposed population reflects the differences in Cryptosporidium concentration from one location to the
next, treatment effectiveness, and individual average daily water consumption. As a result, the endemic
risk of cryptosporidiosis varies substantially across individuals in the population. Characterizing how
individual risks are distributed prior to the LT2ESWTR and how that distribution of risks is expected to
change after the regulation is an important component of the overall benefits analysis.
Economic Analysis for the LT2ESWTR 5-41 December 2005
-------
Exhibit 5.13: Secondary Spread and Secondary Attack Rates Associated with Cryptosporidiosis Outbreaks
Secondary
Spread 1
Household
Secondary
Attack Rate2
Number of
Confirmed
Cases (Adults
and Children)
Outbreak Type
Location and
Year
Tangerman
etal. 1991
31/101
31%
39
Day-care
Atlanta, 1989
Millard
etal.
1994
17/50
34%
53/353
15%
50
Food
Maine,
1993
Heijbel
etal.
1987
6/26
23%
77/204
38%
35
Day-care
Tulsa,
1984
Willocks
etal.
1998
17% -24%
345
Ground
Water
London,
UK, 1997
MacKenzie
etal. 1995a
6/1 183
5%
339
Surface
Water
Milwaukee,
1993
Brown et
al. 1989
32/69 46%
39
Unknown
Great
Yarmouth,
UK, 1986
Bridgman
etal. 1995
21%
-
47
Ground
Water
Warrington,
UK, 1992
Morgan
etal.
1995
37%
-
64
Ground
Water
UK, 1993
MacKenzie
etal. 1995b
3/69
4%
22
Recreational
Water
Oshkosh,
1993
Notes:
1Ratio of secondary cases to primary cases for those with laboratory-confirmed cryptosporidiosis.
2Ratio of the number of illness cases (not laboratory confirmed) to the total number of people potentially exposed in the household of laboratory-confirmed
cryptosporidiosis cases.
3Ratio of the number of illness cases (not laboratory confirmed) to the total number of people potentially exposed in the household of an ill visitor to Milwaukee
during the outbreak (two visitors had laboratory confirmed cryptosporidiosis and were associated with 44 household members).
Economic Analysis for the LT2ESWTR
5-42
December 2005
-------
Exhibits 5.14 and 5.15 present the distribution of individual annual endemic illness risk for
populations in CWSs that filter their water, and for populations in CWSs that do not filter their water,
using the ICR occurrence data set. These exhibits show the portion of the population that has individual
risk levels at or above specified values. This provides a means of focusing on the portion of population
having the highest individual risks, how large that portion of the population is, and how the upper tails of
the risk distribution change as a result of the regulatory alternatives. Appendix C presents similar
exhibits for the ICRSSM and ICRSSL data sets. Appendix S further analyzes the filtered risk
distributions according to bin requirements.
The filtered CWS individual risk distributions vary as a result of the regulatory alternatives
considered for this EA. Exhibit 5.14 shows the risk decreases from regulatory alternative A4 to Al.
In Appendix C, the filtered risk distributions are displayed again based on the ICRSS data sets.
Using different occurrence data sets produces different estimates of cases avoided and cases remaining.
The cases avoided using the ICR data set appear greater compared to using the ICRSS data sets (i.e., the
distance between the Pre-LT2ESWTR and regulatory alternatives' distributions is greater for the ICR data
set). This is because the ICR data predict higher concentrations of Cryptosporidium in source water,
which leads to more systems requiring treatment. Because many of the treatment options achieve more
treatment than is actually needed (see Exhibit 5.7), there is a greater reduction in exposure than if lower
levels of Cryptosporidium are assumed. Thus, analyses using higher predicted Cryptosporidium levels
will show that proportionally more cases are avoided (because of especially efficient technologies).
5.2.7 General Population Risk-Number of Cases Avoided
This analysis uses a two-dimensional Monte Carlo simulation to develop estimates of the range of
individual risks of infection and illness experienced in the general population, and of the number of
annual infections and illnesses resulting from those risks. The algorithms used for calculating individual
risk and the resulting number of infections and illnesses in the overall population at risk, as well as the
details on the forms of the distributions used in the Monte Carlo simulation employing those algorithms,
have been described previously in this chapter.
This section summarizes the reduction in general population risk for unfiltered systems followed
by filtered systems. Summary results presented in this chapter include all system categories. More
results are presented in Appendix C.
Economic Analysis for the LT2ESWTR 5-43 December 2005
-------
Exhibit 5.14: Annual Individual Risk Distributions Based Upon ICR Occurrence
Data Set, Filtered CWSs
100%
90%
80% -.7
.2 70%
OL
Example: Under Pre-LT2 conditions, about 46
percent of the population served by filtered
CWSs have annual individual risk greater than
- 0.0001 (one in a ten thousand), based upon the
ICR occurrence data set.
3
Q.
£
HI
Q.
Example: Under the Preferred Regulatory
Alternative (A3), an estimated 23 percent of
the population served by filtered CWSs have
annual individual risk greater than 0.0001
(one in ten thousand), based upon ICR
occurrence data.
1.00E-06 1.00E-05 1.00E-04 1.00E-03 1.00E-02
Individual Risk Level (Illness Rate)
1.00E-01
1.00E+00
Source: Risk Assessment Model.
Economic Analysis for the LT2ESWTR
5-44
December 2005
-------
Exhibit 5.15: Annual Individual Risk Distributions Based Upon ICR Occurrence
Data Set, Unfiltered CWSs
100%
90%
c
01
o
-------
Exhibit 5.16: Annual Cases of Illness and Deaths Avoided for the LT2ESWTR,
Preferred Alternative, Unfiltered Systems, by Data Set
Population at Risk
Mean
Serving
<1 0,000
738,740
A
Serving
> 10,000
10,245,405
B
All Systems
10,384,145
C
90% Confidence Bound
for All Systems
Lower
(5th %tile)
D
Upper
(95th %tile)
E
///nesses
ICR Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
5,492
28
5,464
496,214
1,386
494,828
501,706
1,414
500,291
101,303
154
101,149
986,331
4,215
982,116
ICRSSL Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
1,766
6
1,760
144,683
322
144,361
146,449
328
146,121
29,583
27
29,556
287,769
984
286,785
ICRSSM Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
3,059
11
3,048
254,284
587
253,696
257,342
598
256,744
51,984
52
51,932
505,670
1,755
503,915
Deaf/is
ICR Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
1
0
1
129
0
129
131
0
130
26
0
26
257
1
256
ICRSSL Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
0
0
0
38
0
38
38
0
38
8
0
8
75
0
75
ICRSSM Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
1
0
1
66
0
66
67
0
67
14
0
14
132
0
131
Note: Detail may not add due to independent rounding.
Sources: Population at risk from the risk assessment model.
[A] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row- Small Systems, ICR, ICRSSL,
and ICRSSM Unfiltered. Illnesses Avoided data from Appendix C, Exhibit C.7, Columns A and D, Row - Small
Systems, ICR, ICRSSL, and ICRSSM
[B] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row- Large Systems, ICR, ICRSSL,
and ICRSSM Unfiltered. Illnesses Avoided data from Appendix C, Exhibit C.7, Columns A and D, Row - Large
Systems, ICR, ICRSSL, and ICRSSM
[C] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row - All Systems, ICR, ICRSSL,
and ICRSSM Unfiltered. Illnesses Avoided data from Appendix C, Exhibit C.7, Columns A and D, Row - All
Systems, ICR, ICRSSL, and ICRSSM
[D] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns C and F, Row - All Systems, ICR, ICRSSL,
and ICRSSM Unfiltered. Illnesses Avoided data from Appendix C, Exhibit C.7, Columns B and E, Row - All
Systems, ICR, ICRSSL, and ICRSSM
[E] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns D and G, Row-All Systems, ICR, ICRSSL,
and ICRSSM Unfiltered. Illnesses Avoided data from Appendix C, Exhibit C.7, Columns C and F, Row - All
Systems, ICR, ICRSSL, and ICRSSM
Economic Analysis for the LT2ESWTR
5-46
December 2005
-------
Exhibit 5.17: Annual Cases of Illness and Deaths Avoided Due to the LT2ESWTR,
Preferred Alternative, All Filtered Systems, by Data Set
Population at Risk
Mean
Serving
<1 0,000
9,546,424
A
Serving
> 10,000
171,910,248
B
All Systems
181,456,672
C
90% Confidence Bound
for All Systems
Lower (5th
%tile)
D
Upper (95th
%tile)
E
///nesses
ICR Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
51,350
3,651
47,700
439,740
23,371
416,369
491,091
27,022
464,069
46,523
2,782
43,741
1,404,589
79,692
1,324,897
ICRSSL Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
16,432
7,416
9,016
130,753
55,159
75,594
147,185
62,575
84,609
15,445
7,668
7,778
426,739
172,224
254,515
ICRSSM Data
Pre-LT2ESWTR
Post-LT2ESWTR
Illnesses Avoided
28,481
7,383
21,098
229,504
52,176
177,328
257,985
59,559
198,426
24,193
6,370
17,823
802,927
171,165
631,762
Deaf/is
ICR Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
8
1
8
73
4
69
81
4
77
8
0
7
232
13
219
ICRSSL Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
3
1
1
22
9
13
24
10
14
3
1
1
71
29
42
ICRSSM Data
Pre-LT2ESWTR
Post-LT2ESWTR
Deaths Avoided
5
1
3
38
9
29
43
10
33
4
1
3
133
28
105
Note: Detail may not add due to independent rounding.
Sources: Population at risk from the risk assessment model.
[A] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row- Small Systems, ICR, ICRSSL,
and ICRSSM Filtered. Illnesses Avoided data from Appendix C, Exhibit C.6, Columns A and D, Row - Small
Systems, ICR, ICRSSL, and ICRSSM
[B] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row- Large Systems, ICR, ICRSSL,
and ICRSSM Filtered. Illnesses Avoided data from Appendix C, Exhibit C.6, Columns A and D, Row - Large
Systems, ICR, ICRSSL, and ICRSSM
[C] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns B and E, Row - All Systems, ICR, ICRSSL,
and ICRSSM Filtered. Illnesses Avoided data from Appendix C, Exhibit C.6, Columns A and D, Row - All
Systems, ICR, ICRSSL, and ICRSSM
[D] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns C and F, Row - All Systems, ICR, ICRSSL,
and ICRSSM Filtered. Illnesses Avoided data from Appendix C, Exhibit C.6, Columns B and E, Row - All
Systems, ICR, ICRSSL, and ICRSSM
[E] Pre-LT2ESWTR data from Appendix C, Exhibit C.3, Columns D and G, Row-All Systems, ICR, ICRSSL,
and ICRSSM Filtered. Illnesses Avoided data from Appendix C, Exhibit C.6, Columns C and F, Row - All
Systems, ICR, ICRSSL, and ICRSSM
Economic Analysis for the LT2ESWTR
5-47
December 2005
-------
This method is applicable to the Pre-LT2ESWTR estimates of cases avoided. However, when
estimating the number of illnesses avoided due to the LT2ESWTR, estimates are dependent on the
regulatory alternative. In order to carry out the calculation above, an ICRSSL/ICR illnesses-avoided
ratio for filtered systems is required, but which ratio is best? There are four regulatory alternatives for
filtered systems, and four corresponding ICRSSL/ICR ratios. (This question does not arise for unfiltered
systems because there is only one regulatory alternative.)
To address this issue, EPA evaluated all regulatory alternatives to identify the one with filtered
treatment requirements most similar to the unfiltered requirements. EPA concluded that alternative Al
was most similar, in terms of expected reductions in Cryptosporidium. Alternative A1 is an across-the-
board 2 log reduction that impacts all filtered systems regardless of initial monitoring results, as opposed
to alternatives A2 through A4 that are expected to impact fewer than half of filtered systems (see
Appendix B).
Therefore, in the equation above, the alternative Al ratio, ICRSSL to ICR, was used to compute
the expected number of illnesses avoided for unfiltered systems based on the ICRSSL data set. This
single estimate of ICRSSL unfiltered illnesses avoided then, was added to the each of four ICRSSL
estimates for filtered system illnesses avoided to obtain a national estimate for each regulatory
alternative.
5.2.7.2 Filtered Systems
Exhibit 5.17 summarizes the estimated Pre-LT2ESWTR cases of illness and deaths associated
with endemic Cryptosporidium occurrence, and estimated cases of illness and deaths avoided for
populations served by filtered water systems as a result of the LT2ESWTR. Results are presented for
small and large systems (those serving at least 10,000 people) separately and for the three occurrence
distribution data sets. Population figures are provided for reference.
5.2.8 Reduction in Sensitive Subpopulation Risk
Morbidity risk in this analysis is based on studies of infectivity and morbidity done on healthy
volunteers. No data currently exist that would give a differential infectivity or morbidity for the
immunocompromised and other sensitive subpopulations. Therefore, this analysis has not accounted for
possible elevated infectivity and morbidity in these populations. The mortality risk from Cryptosporidi-
um in this analysis is expressed as the probability of death given an illness, derived from the study of the
1993 Milwaukee outbreak. The majority of the fatalities due to cryptosporidiosis in that outbreak were
AIDS patients, and the remainder were elderly. Since all observed mortality has been in sensitive
subpopulations, all of the quantified deaths avoided due to the LT2EWSTR are presumed to be lives
saved in sensitive populations.
5.3 Monetized Benefits from Reduction in Exposure to Cryptosporidium Resulting from
the LT2ESWTR
Once the annual endemic illnesses and deaths avoided as a result of the LT2ESWTR are
estimated using the risk model described in the previous sections, monetary unit values can be applied to
these estimates to establish the monetary benefits attributable to the rule. Because the quantified
projection of illnesses avoided is underestimated (risks from outbreaks and risks from other pathogens
are not quantified), the projection of monetized benefits is similarly underestimated. Monetary benefits
are estimated using different methodologies for illnesses and deaths avoided. In addition, two alternative
values are used for estimates of the cost of illness (COI) avoided-the Enhanced COI and Traditional
Economic Analysis for the LT2ESWTR 5-48 December 2005
-------
COL The methodologies and the resulting monetary benefits estimates are presented in the subsections
that follow.
5.3.1 Value of Reduction in Cryptosporidiosis Cases
5.3.1.1 Value of Illnesses Avoided
The goal of this analysis is to provide as complete an accounting as possible of the social welfare
impacts of the regulatory options under consideration. In this context, based on the principles of welfare
economics, the preferred approach for valuing reductions in the risk of cryptosporidiosis-related
morbidity is to rely on estimates of willingness to pay for these risk reductions. However, a review of
the literature indicates that the available studies address illnesses having significantly different effects
from those associated with cryptosporidiosis; hence, estimates from this literature are inappropriate here.
This analysis instead estimates the value of averted morbidity risks based on (1) the avoided medical
costs and (2) the value of averted time losses. The rationale for, and limitations of, this approach are
introduced below and discussed in greater detail in Appendix K. Appendix L describes the calculations
used in the appendix. Appendix P contains a sensitivity analysis using alternative values for two key
inputs to the Enhanced COI.
The calculation of medical costs includes the costs of medical services and medications received
by ill individuals. The assumption behind using these costs as a benefit measure is that reduced
incidence of illness will yield benefits that are at a minimum equal to the costs saved. However, COI
estimates may significantly underestimate individual willingness to pay, for a variety of reasons. In
particular, these estimates: (1) may not fully address the value of avoiding pain and suffering; (2) do not
include costs that individuals incur to avoid the illness (i.e., defensive or averting expenditures);10 (3) do
not reflect aversion to risk (the fear of becoming ill); (4) do not consider ex ante values (they are based
on ex post costs, or costs determined after the fact); and (5) do not consider whether treatment returns
individuals to the original state of health (i.e., is equivalent to avoiding the illness entirely).
A number of researchers have explored the relationship between the COI and individual
willingness to pay to reduce risk from illnesses other than cryptosporidiosis. That research suggests that
the ratio of these two quantities varies greatly depending on the nature of the health effect, the
characteristics of the individuals studied, and the factors included in the construction of each estimate.
Comparison studies result in ratios of willingness to pay to cost of illness that range from
about a factor of 2 to as much as a factor of 79 (in one case); many of the ratios are between 3 and 6.11
In other words, the cost of illness estimates were typically one-third to one-sixth of the willingness to
pay estimates, but the ratio varied greatly.
In some cases, COI studies include indirect, as well as direct costs. These indirect costs usually
include lost earnings due to missed market work time (i.e., work time for which payment is received),
and may also include costs associated with reduced productivity while at work and/or lost nonmarket
work time (e.g., child care or housekeeping). Typically, these costs are estimated using the human
10 Although estimates of averting expenditures incurred during Cryptosporidium outbreaks are available,
the value of avoiding these expenditures is not quantified in this analysis. The risk assessment focuses on endemic
risk rather than risk from outbreaks, and risk-averting behaviors would be less likely to occur in the absence of a
publicized outbreak.
11 See Appendix B of EPA's Handbook for Non-Cancer Health Effects Valuation (USEPA 2000f) for a
review of these studies.
Economic Analysis for the LT2ESWTR 5-49 December 2005
-------
capital approach, which focuses on the value of goods and services that are bought and sold in the
marketplace and ignores other aspects of time use that affect individual well-being.
The analysis of cryptosporidiosis-related morbidity uses two measures of the COI-Traditional
and Enhanced. Both approaches include direct medical costs and the value of lost work time, but differ
in the assessment of the value of lost work time. Both consider the impact of time lost on foregone
market production, which affects the individual worker (e.g., in terms of lost income) as well as other
members of society (who benefit from the availability of the goods or services produced and the taxes
paid); and foregone nonmarket (household and volunteer) production, which affects the individual and
other household members and often has impacts outside the home. The Traditional COI values
nonmarket (unpaid) work time at its replacement cost. The other approach, the Enhanced COI, values
nonmarket work time based on opportunity costs. Both approaches include values for the nonmarket
time lost by friends or family members caring for those who are sick,12 but they use different values for
this lost time.
The Enhanced COI also includes the value of lost leisure time and lost productivity—the
reduced utility (or sense of well-being) associated with decreased enjoyment of time spent in both
market and nonmarket activities. The Enhanced COI is an attempt to more completely measure the loss
of welfare from an illness.
Regarding how best to value lost work time, a review of the literature suggests that researchers
have not attempted to directly estimate (e.g., through surveys) the difference between the value of time
in a well state compared to an ill state. Hence, this analysis relies on wage and compensation data to
estimate the opportunity costs of time usage. This approach recognizes that, because resources are
limited, any decision to use resources for one purpose means that they cannot be used for other
purposes. Therefore, the value of the unpriced resource can be determined based on the value of its
next best use.
The application of the opportunity cost approach to market work time is relatively clear, since
compensation can be used to estimate these costs. More precisely, lost market work is valued at the
median gross (pre-tax) wage rate plus benefits, also referred to as total compensation or employer's
costs. This approach most accurately represents the full social impact of lost work time because it
incorporates both the income loss to the individual and the loss to society in terms of reduced tax
revenue and decreased production of goods and services.
For unpaid time spent in nonmarket work and for leisure time, wage data are also used. The
Enhanced COI assumes that (at the margin) the wage represents the opportunity cost of engaging in
such activities. Lost nonmarket work and leisure time are valued at the median net (post-tax) wage rate.
This approach reflects the assumption that, at the margin, an individual will choose to engage in
nonmarket work or leisure activities only if the value of these activities exceeds the post-tax wage rate
that the individual would otherwise earn.13 These values are applied both to complete losses of time
(time spent in illness-related activities rather than normal activities) and to partial losses (time spent in
normal activities that are less productive or pleasurable than in the absence of illness). In the latter
case, however, the dollar value of the loss is prorated to reflect the fact that the individual does not
completely lose the productivity or utility associated with the activity. These values are applied to
12 Paid care is included in the medical cost component of the analysis and hence is not discussed in the
discussion of time losses.
13 A sensitivity analysis that uses alternative values for nonmarket work and leisure time in the calculation
of the Enhanced COI is included in Appendix P.
Economic Analysis for the LT2ESWTR 5-50 December 2005
-------
nonmarket caretakers whose normal activities are affected by illness as well as to time losses accruing
to the ill individual.14
For unpaid time spent in nonmarket work, the Traditional COI also uses wage data. The value
used is half of the after-tax wage. The use of 50 percent of the wage rate is consistent with the common
practice in the human capital literature of valuing nonmarket work time at the market rate for domestic
workers.15 This literature uses replacement cost as a measure of the productivity of nonmarket work,
rather than focusing on the opportunity cost (or utility loss) for the individual who chooses to engage in
nonmarket work. In support of the use of 50 percent of the after-tax wage rate, the median weekly
earnings of private household workers in the service industry were $276 per week in 2002, about 45
percent of the median weekly earnings of $609 for all workers (U.S. Census Bureau Table 641, 2003).
Private household workers include childcare workers, cleaners, and servants. The Traditional COI does
not include values for lost leisure time or lost productivity.
Sleep time presents special problems in this analysis, in part because data on the effect of
cryptosporidiosis-related morbidity on the amount or quality of sleep time is not available. Thus, this
analysis conservatively assumes that lost sleep time has zero value.
The use of medical costs and the opportunity cost of time to value cryptosporidiosis-related
morbidity may understate the value of these risk reductions for a variety of reasons.16 As noted earlier,
COI estimates generally understate willingness to pay for a variety of reasons, e.g., because they may
not fully consider the value of avoided pain and suffering or of risk aversion. In addition, the use of
wage and compensation data to value lost time may understate the utility of time spent in its preferred
use. The use of wage rates may understate the total utility associated with an activity even in the case
of market work, because individuals may derive intrinsic pleasure from the activity above and beyond
the income they receive. For nonmarket work and leisure, the value of the activity to the individual
may exceed the opportunity cost for similar reasons. In addition, nonmarket work and other activities
can provide benefits to other members of society that are not reflected in the individual wage rate.
Finally, neither the Tradtional or Enhanced COI approach includes the value of lost sleep time.
In addition, relying on wage data for valuing lost time presents difficulties in the case of
individuals for whom these data do not exist, such as children, the unemployed who are seeking
employment, and those out of the labor market. The approach taken in this analysis is to value all time
losses at the rates applicable to adult wage earners. It is unclear whether this approach under- or
overstates the value of times losses for the individuals in these other categories, given the lack of
information on these values.
14 A sensitivity analysis that uses alternative values for lost productivity in the calculation of the Enhanced
COI is included in Appendix P.
15 A pioneering example of this approach is Rice 1966; a more recent example is Thamer et al. 1998.
16 A number of other simplifying assumptions inherent in this approach may lead it to under- or overstate
the value of time losses. These relate to factors such as the functioning of the labor market, the treatment of
individuals who are not labor force participants, the use of average or median (rather than marginal) earnings data,
and the possibility that substitute activities (e.g., watching TV instead of normal activities) have some positive value
if not offset by the utility losses from the discomfort and stress of being sick. It is unclear whether, in total, these
practical limitations serve to increase or decrease the bias that results from the sources discussed in this paragraph.
Economic Analysis for the LT2ESWTR 5-51 December 2005
-------
COI Calculations
The primary risk of illness that LT2ESWTR addresses is from endemic exposure to
Cryptosporidium and the resulting cases of cryptosporidiosis. Direct measurements of many elements
of the COI were made as part of an investigation of the 1993 Milwaukee Cryptosporidium outbreak.
Some of the data collected during that outbreak have been reported in Corso et al. (2003), whose data
sources include the original epidemiological investigation by CDC and State personnel, including
telephone surveys and a review of hospital records. The epidemiological investigation collected
information on the duration of illness, types of medication taken, medical care sought, if any, and the
costs associated with these services. The data from that report and other sources of information used in
the analysis that follows are shown in Appendix L.
The computation of COI involves two broad categories—direct medical costs and the value of
lost time (Exhibit 5.18). The components are updated to a common month and year (December 2003),
which is used as the starting point for projecting benefits into future time periods. For each of these
components, separate estimates are made based on the severity of the illness. Illnesses are sorted into
three severity categories as follows:
• Mild: the person did not seek professional medical care for the illness.
• Moderate: the person had one or more outpatient visits to a physician or emergency room,
but the person was ultimately not hospitalized.
Severe: the person was hospitalized one or more times.
The percentage of people in each of the severity classifications is used to derive an average
weighted cost per patient. The average loss per case of cryptosporidiosis incorporating all categories is
approximately $844 for the Enhanced COI and $274 for the Traditional COI. Of those totals,
approximately $107 is for direct medical costs (Exhibit 5.18). The details of the computations are
discussed in the next three subsections.
Direct Medical Costs
As Exhibit 5.18 shows, the weighted average of direct medical costs per case of illness is
$106.91. Costs for doctor visits, emergency room (ER) visits, hospital stays, ambulance costs, and
costs of medication comprise the direct medical costs. (As mentioned earlier, medications can relieve
some symptoms, but do not cure infection). All direct medical costs are obtained in December 1993
dollars and updated by a 1.47 cost per illness (CPI-U) update factor to December 2003 dollars.17 The
Corso et al. (2003) report states that those with mild cases did not visit a doctor (a non-ER physician)
but 95 percent of those with moderate cases and 29 percent of those with severe cases did. Conversely,
5 percent of moderate cases and 71 percent of severe cases went to the ER. Multiplying through by the
respective costs of doctor visits ($66.15) and ER visits ($329.28) yields average costs of $62.84 per
moderate case and $19.18 per severe case for doctor visits, and an average of $16.46 per moderate case
and $233.79 per severe case per ER visit.
17 Bureau of Labor Statistics, 302.1 (Dec2003$)/205.2 (Decl993$) = 1.47 CPI-U medical cost update
factor.
Economic Analysis for the LT2ESWTR 5-52 December 2005
-------
Exhibit 5.18: Direct Medical Costs of a Case of Cryptosporidiosis
Medical Cost
Doctor Visits
Emergency
Room Visits
Hospital Stays
Ambulance
Medication
Medication
after Health
Care
Medication
Taken upon
Recurrence
Totals
Average Cost1
1993$
Mild
NA
NA
NA
NA
$5.73
NA
$2.44
Moderate
$45.00
$224.00
NA
$228.00
$5.92
$8.91
$2.44
Severe
$45.00
$224.00
$6,152.96
$228.00
$6.74
$70.52
$2.44
December 2003$
Mild
NA
NA
NA
NA
$8.42
NA
$3.59
Moderate
$66.15
$329.28
NA
$335.16
$8.70
$13.10
$3.59
Severe
$66.15
$329.28
$9,044.85
$335.16
$9.91
$103.66
$3.59
Average Cost Per Patient
December 2003$
Mild
(88%)
NA
NA
NA
NA
30%*$8.42
=$2.53
NA
21%*$3.59
=$0.75
$3.28
Moderate
(11%)
95%*$66.15
=$62.84
5%*$329.28
=$16.46
NA
4.9%*5%*
$335.16
=$0.82
30%*$8.70
=$2.61
54%*$13.10
=$7.07
21%*$3.59
=$0.75
$90.56
Weighted Total
Severe (1%)
29%*$66.15
=$19.18
71%*$329.28
=$233.79
100%*$9,044.85
=$9,044.85
16.3%*$335.16
=$54.63
29%*$9.91
=$2.87
48%*$1 03.66
=$49.76
21%*$3.59
=$0.75
$9,405.84
$106.91
Notes: Detail may not add to totals due to independent rounding.
1AII direct medical costs are obtained in December 1993$ and updated by a 1.47 CPI-U update factor to December
2003$. CPI-U update factor = 302.1 (Dec2003$)/205.2 (Dec1993$) = (rounded to) 1.47.
Sources: Appendix L, 1993$ average cost data from Corso et al. (2003).
CPI-U data - Bureau of Labor Statistics.
The average cost of a hospital stay was $9,044.85. This figure is applied only to severe cases.
Hospital costs include all costs of hospitalization except those for consultations by specialists;
insufficient data were available on the latter costs.
Medication costs were the same for all levels of severity for medication taken upon recurrence
of Cryptosporidiosis. The average cost of medication for recurrence was $3.59 per person. Of the
people infected with Cryptosporidiosis, 21 percent had at least one recurrence. Otherwise, medication
costs varied depending on severity of illness. For medication used before receiving medical attention,
costs are similar across severity groups; $8.42 for mild, $8.70 for moderate and $9.91 for severe cases
of Cryptosporidiosis. Medication costs after health care varied. In addition, these costs only applied to
non-mild cases; costs amounted to $13.10 for moderate cases and $103.66 for severe cases, but are
weighted based on the percentage of people taking medication. Summing all of the above costs and
obtaining a weighted average by severity level yields an overall weighted average of $106.91 for direct
medical costs per illness.
Economic Analysis for the LT2ESWTR
5-53
December 2005
-------
Value of Lost Time Per Day
The value of lost time is derived through several steps shown in summary below and discussed
in detail in Appendix L. First, the number of days lost and days with lessened productivity (for the
Enhanced COI) must be calculated. Exhibit 5.19 shows days lost by severity of illness, and Exhibit 5.20
calculates average days lost weighted by percent of cases with each severity of illness. For the 21
percent of cases with a 2-day recurrence of the illness (Corso et al., 2003), the analysis assumes that
these have at least 2 days' reduced productivity (Appendix L).
Exhibit 5.19: Days Lost and Days with Lost Productivity, by Severity of Illness
Illness Severity
Mild
Moderate
Severe
Time Category
Mean
Duration
of Illness
(Days)
A
4.7
9.4
34.0
Days Lost to
Illness
B
1.3
3.8
5.6
Lost
Productivity
Days
C=A-B
3.4
5.6
20.5
Source: Exhibit L.5.
Exhibit 5.20: Weighted Average Days Lost for Particular Illness
Work Losses
(Patients)
Caregivers
Losses
Productivity
Losses
(Patients)
Severity
Mild
Moderate
Severe
Days Lost
1.3
3.8
13.5
Mild
Moderate
Severe
0.1
1.3
3.9
Mild
Moderate
Severe
Recurrence
3.4
5.6
20.5
2.0
Weight
88%
11%
1%
Total
88%
11%
1%
Total
88%
11%
1%
21%
Total
Weighted Average Days
1.144
0.418
0.135
1.697
0.088
0.143
0.039
0.270
2.992
06.616
0.205
0.420
4.233
Source: Exhibit L.6
Second, the value of time must be estimated. Exhibit 5.21 presents the hours lost per day of
illness. These values are based on estimates of the number of hours worked, adjusted by the percent of
the population engaged in market and nonmarket work, and assumes 8 hours of sleep per day per
person. Details of the sources, calculations, and assumptions to derive these values are provided in
Appendix L.5.
Economic Analysis for the LT2ESWTR
5-54
December 2005
-------
Exhibit 5.21 also presents the per-hour value of lost time. These values derive from reported
usual weekly earnings of $609 (U.S. Census Bureau Table 641, 2003). To value lost work time, this
figure is increased to reflect employers' costs (adding in benefits paid). Because employers are willing
to pay for workers' time at this level, it is the best measure of the value of that lost time ($20.82 per
hour). For the Enhanced COI, the value of lost nonmarket work time is the median aftertax wage. To
derive this figure, the weekly earnings estimate of $609 was adjusted downward to reflect after tax
wages (to $12.46). For the Traditional COI, half of this figure is used (or $6.23), as discussed above,
and in detail in Appendices K and L. Details of the sources, calculations, and assumptions to derive
these values are provided in Appendix K and Appendix L.6.
The two right-hand columns of Exhibit 5.21 multiply these dollar-per-hour values by the time
allocations to determine the weighted average value of time per hour and per day.
Exhibit 5.21: Value of Time, 2003
Enhanced COI
Traditional COI
Time Loss
Category
Market Work
Time
Non market
Work Time
Non market
Leisure Time
Caregiver Day
Market Work
Time
Non market
Work Time
Caregiver Day
Hours Per-Day
of Illness
3.4
2.3
10.3
Per-Hour Value
$20.82
$12.46
$12.46
Sum of per-day value of lost
market work, nonmarket work, and
leisure days
3.4
2.3
$20.82
$6.23
Sum of per-day value of lost
market and nonmarket work days
Per-Day Value
$70.79
$28.66
$128.34
$227.79
$70.79
$14.33
$85.12
Note: Detail may not add due to independent rounding.
Source: Exhibit L.9.
Total Morbidity Cost of Illness
There are two major components of the total value of the morbidity cost of avoiding a case of
cryptosporidiosis-direct medical costs and lost time. As discussed above, the total direct medical costs
are $106.91 per illness. Lost time estimates are derived from the estimate of the average days lost and
the value of each day lost. For the Enhanced COI, productivity losses are included. Because patients
are only fractionally as productive at work as well people, the loss associated with the less productive
days is a portion of the value of a full lost day, specifically 30 percent (rounded from Harrington 1991).
For example, this may be the result of frequent trips to the bathroom, reduced concentration on tasks, or
preparation of special meals. Summing the subcategories of total value of lost time yields a weighted
cost of $737.33 (Enhanced COI) and $167.43 (Traditional COI) per case of cryptosporidiosis in 2003.
Economic Analysis for the LT2ESWTR
5-55
December 2005
-------
Exhibit 5.22 shows these calculations. The total loss per case in 2003$ is $844.24 for the Enhanced COI
and about a third of that, or $274.34, for the Traditional COI.
Exhibit 5.22: Total Loss Per Case, Enhanced and Traditional COI, 2003$
Loss Category
Average
Days
Lost Per
Illness
Value Per Day
Enhanced
COI
Traditional
COI
Total Loss Per Case
Enhanced
COI
Traditional
COI
Direct Medical Costs
Lost Market Work Days
Lost Nonmarket Work
Days
Lost Leisure Time
1.697
B
$70.79
D=A*B
$106.91
$28.66
$128.34
$70.79
$14.33
$120.13
$48.64
$217.79
E=A*C
$106.91
$120.13
$24.32
Lost Caregiver Days
0.270
$227.79
$85.12
$61.50
$22.98
Lost Leisure Productivity
$128.34 x
30%
$162.98
4.233
Lost Productivity at Work
($70.79 +
$28.66) x
30%
$126.29
$737.33
$844.24
$167.43
$274.34
Lost Time Subtotal
Total
Notes: Detail may not calculate to totals due to independent rounding.
The Traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The Enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Source: Exhibit L.10.
The value of lost time can increase or decrease over time, depending on the change in real
income. Real income (after inflation) is projected to increase based on forecasts in growth in the gross
domestic product (GPD) and population. Real per capita income growth reflects the overall increase in
society's productivity. These changes in income growth and, therefore, the value of time, are shown in
Exhibit 5.23 and are derived in Appendix C. These changes in income growth mean that the loss due to
an illness would increase over time because lost time is recovered by wage rates or their equivalent. In
the benefits model, the cases avoided in each year are valued as shown in Exhibit 5.23 (the model uses
unrounded data). Benefits derived from medical costs are not adjusted for changes in income over time,
because medical costs do not necessarily have a direct or indirect link with income.
Economic Analysis for the LT2ESWTR
5-56
December 2005
-------
Exhibit 5.23: Yearly Total Loss Per Case, Enhanced and Traditional COI
Year
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
%Change in Income
(Real GDP per Capita)
A
Base Year
2.3%
3.9%
3.3%
2.3%
1 .9%
2.0%
2.0%
1 .8%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
1 .7%
Lost Time
Enhanced COI
B=(1+A) x previous year
$737.33
$754.34
$783.82
$809.88
$828.85
$844.23
$860.79
$877.73
$893.29
$908.22
$923.39
$938.83
$954.55
$970.57
$986.89
$1,003.55
$1,020.55
$1,037.91
$1,055.60
$1,073.60
$1,091.91
$1,110.54
$1,129.50
$1,148.81
$1,168.48
$1,188.48
$1,208.91
Traditional COI
C=(1+A) x previous year
$167.43
$171.29
$177.99
$183.90
$188.21
$191.70
$195.46
$199.31
$202.84
$206.23
$209.68
$213.19
$216.76
$220.39
$224.10
$227.88
$231 .74
$235.68
$239.70
$243.79
$247.95
$252.18
$256.48
$260.87
$265.33
$269.88
$274.51
Direct Medical
Costs
D
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
$106.91
Total Loses Per Case
Enhanced COI
E = B+D
$844.24
$861.25
$890.73
$916.79
$935.76
$951.14
$967.70
$984.64
$1,000.20
$1,015.13
$1,030.30
$1,045.74
$1,061.46
$1,077.48
$1,093.80
$1,110.46
$1,127.46
$1,144.82
$1,162.51
$1,180.51
$1,198.82
$1,217.45
$1,236.41
$1,255.72
$1,275.39
$1,295.39
$1,315.82
Traditional COI
F=C+D
$274.34
$278.20
$284.90
$290.81
$295.12
$298.61
$302.37
$306.22
$309.75
$313.14
$316.59
$320.10
$323.67
$327.30
$331.01
$334.79
$338.65
$342.59
$346.61
$350.70
$354.86
$359.09
$363.39
$367.78
$372.24
$376.79
$381 .42
Note: Full precision is used in model calculations. Rounded data are shown here.
The Traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The Enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Source: Exhibit L.11.
The sensitivity of different assumptions on the Enhanced COI was tested and is described in
Appendix P. Alternative values for several underlying variables were considered, but only the
following two were judged important to test in a sensitivity analysis:
The hourly value of nonmarket work and leisure time. The Enhanced COI uses a value of
$ 12.46 per hour (in 2003), but an alternative value of $6.23 per hour was used in the "low
estimate" sensitivity analysis, and a value of $18.69 was used in the "high estimate"
sensitivity analysis. These bounds represent the effect of (1) assuming that all nonmarket
work and leisure time is valued at a rate higher than the best estimate and (2) assuming that
some of the nonmarket work and leisure time lost is an incomplete loss of utility. The basis
for using these bounds at 50 percent and 150 percent of the best estimate is discussed in
detail in Appendix P.
The percent decrease in productivity for days in which work could be performed, but
effects of the illness prevented full productivity. The Enhanced COI uses 30 percent, and
the bounds for the sensitivity analysis are 20 percent and 40 percent. Related studies and
the basis for selecting these levels for a sensitivity analysis is discussed in Appendix P.
Economic Analysis for the LT2ESWTR
5-57
December 2005
-------
The total value of lost time using the Enhanced COI is $737.33 (in 2003$), and using these
alternative values in the derivation of that estimate would lower that value to $420.04 (57 percent of the
Enhanced COI) and raise that value to $1,121.08 (152 percent of the Enhanced COI). The overall effect
on total benefits is less pronounced because of the value of fatalities, and fixed direct medical costs.
Appendix P calculates the effect of using these alternatives on the total COI and total benefits.
5.3.1.2 Value of Avoiding Fatal Cases of Cryptosporidiosis
Benefits of the LT2ESWTR also derive from avoiding fatal cases of cryptosporidiosis. The
Value of a Statistical Life (VSL) is used to measure the value of these benefits. The VSL represents an
estimate of the monetary value of reducing risks of premature death. The VSL, therefore, is not an
estimate of the value of saving a particular individual's life. The value of a "statistical" life represents
the sum of the values placed on small individual risk reductions across an exposed population. For
example, if a regulation were to reduce the risk of premature death from cryptosporidiosis by
1/1,000,000 for 1 million exposed individuals, the regulation would "save" one statistical life
(1,000,000 x 1/1,000,000). If each of the 1,000,000 people were willing to pay $5 to achieve the
individual risk reduction anticipated from the regulation, the VSL would be $5 million ($5 x
1,000,000).
An EPA study characterized the range of possible VSL values as a Weibull distribution with a
mean of $4.8 million (1990 price level), based on 26 individual study estimates (USEPA 1997b). This
represents the value recommended for use in benefits analyses in EPA's Guidelines for Preparing
Economic Analyses (USEPA 2000d) and endorsed by the Science Advisory Board (SAB) Arsenic
review panel (USEPA 200 Id). For purposes of the LT2ESWTR benefits analysis, the VSL Weibull
distribution (with parameters of location = 0, scale = 5.32, shape = 1.51) was incorporated into the
benefits model Monte Carlo simulation. This enables quantification of the uncertainty surrounding
benefits estimates derived from the VSL. The VSL was also updated to a year 2003 price level using a
CPI adjustment factor (see Appendix C) and the distribution in 2004 has a mean of $7.5 million,
median of $6.5 million, a 5th percentile value of $1.1 million and a 95th percentile value of $17.2
million.
5.3.1.3 Measuring Benefits Over the LT2ESWTR Implementation Schedule
In order to extract benefits data from the model and present these benefits in comparable terms
to a similarly calculated stream of costs, it is necessary to calculate the present value of all benefits over
the lifetime of the implementation schedule. LT2ESWTR implementation occurs over several years as
States and PWSs learn the requirements, inform their staffs, and perform monitoring. A 25-year
horizon was chosen for this analysis because systems have several years to begin treatment associated
with LT2ESWTR, and many technologies in this analysis have a 20-year life-cycle. Calculations using
this time frame allows the analysis to capture all of the period when technologies would be installed and
avoids the complications that would be necessary to estimate rehabilitation or replacement costs for
installed equipment. This time frame also matches that used in other recent analyses such as the one for
the Stage 2 DBPR. A complete schedule of when costs are expected to be incurred and benefits
obtained is presented in Appendix O.
5.3.1.4 Adjustment for Income Elasticity
Although the price level (year 2003) is held constant across all benefits projections, values in
future years are adjusted to reflect changes in the valuation of avoiding health effects associated with
changes in income overtime. Estimates of how valuation varies with income growth (i.e, income
elasticities) are available from the economic literature, and in those cases income elasticities are
combined with estimates of income growth. Benefits based on potentially fatal health effects, which are
Economic Analysis for the LT2ESWTR 5-58 December 2005
-------
based on willingness to pay estimates that vary with income, are adjusted using estimates of income
elasticity and income growth. This section describes how this adjustment is carried out.
In the case of avoided-death benefits, income elasticity adjustments are applied to values in
future years. In general, income elasticity represents changes in valuation in relation to changes in real
income. For example, if, for every 1 percent increase in real income, a particular consumer's
willingness to pay for a particular item increases by 1 percent, this would be represented by an income
elasticity of 1. For most willingness to pay estimates, income elasticity values are less than 1, reflecting
slower growth in willingness to pay than in income.
In order to apply the income elasticity values in the benefits model, they must be combined
with projections of real income growth over the time frame for analysis. To accomplish this, population
and real gross domestic product (GDP) projections are combined to calculate per-capita real GDP
values18 (see Exhibit C. 14). Percent changes in these values over time can then be combined with
income elasticity figures to derive a single adjustment factor.19 Given any two points in time, this factor
is calculated as follows:
Income elasticity adjustment factor = (e^ - eI2 -12 -1]) / (eI2 - e^ -12 -1])
where: e = income elasticity
I] = real income (per-capita GDP) in the base year
I2 = real income (per-capita GDP) in the year of analysis
When applying this formula, income elasticity adjustment factors are calculated from the same
base year as the values subject to adjustment. In this case, income elasticity factors for fatal
cryptosporidiosis cases are calculated from a 1990 base year (Ij = 1990 in the above formula) because
that is the base year used in the study from which VSL estimates are derived.20
Kleckner and Neumann (2000) identified published studies from which elasticity values could
be derived for potentially fatal health effects. They suggest a triangular distribution with a mode of
0.40, and endpoints at 0.08 and 1.00. In the Monte Carlo simulation that assigns dollar values to
benefits, income elasticity values (e in the above equation) are drawn from this probability distribution.
Based on this formula and inputs, income elasticity factors are computed and applied to avoided-death
benefits in future years. At the average income elasticity value (0.49), the income elasticity factors
applied range from 1.224 (2009) to 1.445 (2029).
18 Ideally, income elasticity and income growth measurements would be calculated using real per capita
personal income growth. However, real per capita GDP is used as a proxy for real per capita personal income
growth due to lack of appropriate data projections for real personal income growth. Historical data suggests that
GDP and personal income grow at similar rates (i.e., Table B-31 of the 2002 Economic Report of the President
shows that both real per capita GDP and real per capita disposable personal income grew at an average annual rate of
2.3 percent between 1959 and 2000).
19 See Appendix A of Kleckner and Neumann (2000) for additional information on the derivation and
application of income elasticity adjustments.
20 The distribution of VSL values used in this EA is derived based on a meta-analysis of 26 different VSL
studies, all representing different year price levels. These price levels were updated to a common 1990 price level as
part of the analysis in "The Benefits and Costs of the Clean Air Act, 1970-1990" (USEPA 1997b), from which the
distribution used in this EA is taken.
Economic Analysis for the LT2ESWTR 5-59 December 2005
-------
Exhibit 5.24: Mean of Yearly Values for a Statistical Life ($Million)
Year
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
Mean of 10,000 Values
7.72
7.80
7.87
7.93
8.00
8.06
8.13
8.20
8.27
8.33
8.40
8.47
8.54
8.62
8.69
8.76
8.83
8.91
8.98
9.05
9.13
Source: Derived from Exhibit C.16.
The estimates for the value of a statistical life derive from a distribution of the value of
statistical life (discussed in section 5.3.1.2), and adjustments for income elasticity (discussed in this
section). In the benefits model, each year from 2009 to 2029 has a vector of 10,000 values for a
statistical life, and each value is used exactly once. The complete distributions (200,000 values) are
documented in the model, but to illustrate how these distributions are affected by the income elasticity
adjustments, the mean of each year's distribution of values is shown above in Exhibit 5.24. The values
are shown starting in the year 2009 because that is the first year with benefits. Appendix C has
additional discussion of the derivation of these data.
5.3.1.5 Present Value of Future Benefits
To allow comparison of future streams of costs and benefits, it is common practice to adjust
both streams to a present value (PV) using a social discount rate. This process takes into account the
time preference that society places on expenditures and benefits and allows comparison of cost and
benefit streams that vary over a given time period.21 A present value for any future period can be
calculated using the following equation:
PV = V(t)/(l+R)t
Where: t = The number of years from the reference period (year 0 of the benefits stream)
R = Social discount rate
V(t) = The benefits occurring t years from the reference period
The present values presented in this EA are the sum of the PVs for each year.
21 See EPA's Guidelines for Preparing Economic Analyses (USEPA 2000d) for a full discussion of the use
of social discount rates in the evaluation of policy decisions.
Economic Analysis for the LT2ESWTR
5-60
December 2005
-------
There is much discussion among economists of the proper social discount rate to use for policy
analysis. Therefore, for this EA, PV calculations are made using two social discount rates, 3 and 7
percent, which reflect OMB guidance to Federal agencies on the development of regulatory analyses
(OMB, 2003). The 3 percent rate is an estimate of how society discounts future consumption flows to
their present value, while the 7 percent rate is an estimate of the average before-tax rate of return to
private capital (OMB, 2003)22. To present results on an annual basis, the total PV of benefits are
annualized using the same social discount rates.
5.3.1.6 Summary of Quantified Benefits of LT2ESWTR
The risk assessment methodology described in this chapter estimates the quantified benefits of
reducing endemic cryptosporidiosis as applied to each of the regulatory alternatives considered for
LT2ESWTR. These alternatives—which are described in detail in Chapter 3—were evaluated to
provide EPA with information on different approaches for implementation of these regulations.
Exhibits 5.25 to 5.26 provide summaries of the cumulative monetary benefits estimated for the
Preferred Alternative using three occurrence distribution data sets and two COI estimates. Exhibit 5.25
presents benefits categorized by system size, Exhibit 5.26 presents benefits categorized by filtered and
unfiltered systems, and lastly, Exhibit 5.27 categorizes by illnesses and deaths avoided. All monetized
benefits presented in this chapter represent the value at full implementation, averaged over the 25-year
evaluation period in accordance with the schedule benefits are incurred.
See OMB's Circular A-4 (OMB, 2003) for a discussion of the use of social discount rates in the
evaluation of policy decisions.
Economic Analysis for the LT2ESWTR 5-61 December 2005
-------
Exhibit 5.25a: Annualized Benefits of Illnesses and Deaths Avoided, Preferred
Alternative, Enhanced Cost of Illness1
($ Millions/Year, 2003$)
System Size
(Population
Served)
ICR
Mean
A
5th %ile
B
95th % Me
C
ICRSSL
Mean
D
5th %ile
E
95th %ile
F
ICRSSM
Mean
G
5th %ile
H
95th %ile
I
3% Discount Rate
Illnesses and Deaths Avoided
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
> 1 Million
Total
$ 0.42
$ 2.29
$ 2.84
$ 17.20
$ 51.59
$ 121.21
$ 95.90
$ 461 .94
$1,100.12
$ 1 ,853.49
$ 0.03
$ 0.19
$ 0.26
$ 1.67
$ 5.30
$ 12.72
$ 10.57
$ 48.69
$ 140.49
$ 223.83
$ 1.36
$ 6.96
$ 8.52
$ 50.48
$ 149.09
$ 354.45
$ 275.92
$1,314.28
$2,856.23
$ 4,940.84
$ 0.07
$ 0.41
$ 0.54
$ 3.46
$ 10.83
$ 25.03
$ 20.40
$ 103.09
$ 294.17
$ 457.99
$ 0.00
$ 0.03
$ 0.04
$ 0.30
$ 1.04
$ 2.59
$ 2.22
$ 10.78
$ 37.69
$ 54.94
$ 0.22
$ 1.34
$ 1.73
$ 10.94
$ 33.50
$ 76.39
$ 60.83
$ 298.86
$ 761 .40
$1,242.24
$ 0.16
$ 0.95
$ 1.24
$ 7.78
$ 23.84
$ 54.77
$ 43.86
$ 213.39
$ 539.90
$ 885.89
$ 0.01
$ 0.07
$ 0.10
$ 0.67
$ 2.22
$ 5.36
$ 4.54
$ 21.43
$ 68.07
$ 103.31
$ 0.53
$ 3.09
$ 3.96
$ 24.35
$ 73.25
$ 167.86
$ 131.99
$ 621.30
$1,409.10
$2,419.77
7% Discount Rate
Illnesses and Deaths Avoided
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
> 1 Million
Total
$ 0.32
$ 1.73
$ 2.15
$ 13.02
$ 39.05
$ 95.35
$ 77.20
$ 376.11
$ 896.07
$1,501.01
$ 0.02
$ 0.15
$ 0.19
$ 1.26
$ 4.01
$ 10.01
$ 8.48
$ 39.61
$ 114.51
$ 181.38
$ 1.03
$ 5.27
$ 6.44
$ 38.25
$ 112.78
$ 278.75
$ 222.32
$1,069.81
$2,328.34
$3,997.81
$ 0.05
$ 0.31
$ 0.41
$ 2.62
$ 8.20
$ 19.69
$ 16.42
$ 83.94
$ 239.62
$ 371.26
$ 0.00
$ 0.02
$ 0.03
$ 0.23
$ 0.78
$ 2.03
$ 1.79
$ 8.76
$ 30.65
$ 44.56
$ 0.17
$ 1.02
$ 1.31
$ 8.28
$ 25.37
$ 60.13
$ 49.09
$ 243.19
$ 619.74
$1,005.17
$ 0.12
$ 0.72
$ 0.94
$ 5.89
$ 18.05
$ 43.09
$ 35.31
$ 173.75
$ 439.77
$ 717.64
$ 0.01
$ 0.05
$ 0.07
$ 0.51
$ 1.68
$ 4.22
$ 3.65
$ 17.44
$ 55.36
$ 83.58
$ 0.40
$ 2.34
$ 3.00
$ 18.44
$ 55.50
$ 132.04
$ 106.41
$ 506.40
$ 1,149.33
$ 1,961.08
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Sources:
[A] Appendix C, Exhibits C.4a and C.5a, Column M, Row - A3
[B] Appendix C, Exhibits C.4a and C.5a, Column N, Row - A3
[C] Appendix C, Exhibits C.4a and C.5a, Column O, Row - A3
[D] Appendix C, Exhibits C.4c and C.5c, Column M, Row - A3
[E] Appendix C, Exhibits C.4c and C.5c, Column N, Row - A3
[F] Appendix C, Exhibits C.4c and C.5c, Column O, Row - A3
[G] Appendix C, Exhibits C.4b and C.5b, Column M, Row - A3
[H] Appendix C, Exhibits C.4b and C.5b, Column N, Row - A3
[I] Appendix C, Exhibits C.4b and C.5b, Column O, Row - A3
Economic Analysis for the LT2ESWTR
5-62
December 2005
-------
Exhibit 5.25b: Annualized Benefits of Illnesses and Deaths Avoided, Preferred
Alternative, Traditional Cost of Illness1
($ Millions/Year, 2003$)
System Size
(Population
Served)
ICR
Mean
A
5th %ile
B
95th % Me
C
ICRSSL
Mean
D
5th %ile
E
95th %ile
F
ICRSSM
Mean
G
5th %ile
H
95th %ile
I
3% Discount Rate
Illnesses and Deaths Avoided
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
> 1 Million
Total
$ 0.28
$ 1.55
$ 1.93
$ 1 1 .75
$ 35.42
$ 83.77
$ 66.74
$ 325.54
$ 814.26
$1,341.24
$ 0.02
$ 0.11
$ 0.14
$ 0.93
$ 2.93
$ 7.02
$ 5.86
$ 27.71
$ 80.48
$ 127.85
$ 0.97
$ 5.06
$ 6.16
$ 36.76
$ 108.83
$ 263.43
$ 205.51
$ 991 .93
$2,324.94
$3,929.17
$ 0.04
$ 0.28
$ 0.37
$ 2.38
$ 7.50
$ 17.44
$ 14.34
$ 73.53
$ 219.20
$ 335.07
$ 0.00
$ 0.02
$ 0.02
$ 0.17
$ 0.58
$ 1.44
$ 1.25
$ 5.99
$ 21.35
$ 31.03
$ 0.16
$ 0.96
$ 1.26
$ 8.04
$ 24.70
$ 56.46
$ 45.68
$ 229.04
$ 622.33
$ 989.12
$ 0.11
$ 0.65
$ 0.84
$ 5.33
$ 16.42
$ 37.98
$ 30.65
$ 151.11
$ 400.83
$ 643.92
$ 0.01
$ 0.04
$ 0.05
$ 0.36
$ 1.22
$ 2.95
$ 2.54
$ 1 1 .90
$ 37.94
$ 57.80
$ 0.38
$ 2.24
$ 2.87
$ 17.90
$ 54.15
$ 124.22
$ 98.64
$ 476.01
$1,150.53
$1,919.36
7% Discount Rate
Illnesses and Deaths Avoided
<100
100-499
500-999
1 ,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
> 1 Million
Total
$ 0.22
$ 1.17
$ 1.46
$ 8.91
$ 26.87
$ 66.05
$ 53.87
$ 265.76
$ 664.80
$1,089.13
$ 0.01
$ 0.08
$ 0.11
$ 0.70
$ 2.22
$ 5.53
$ 4.74
$ 22.59
$ 65.75
$ 103.74
$ 0.73
$ 3.84
$ 4.68
$ 27.86
$ 82.55
$ 207.48
$ 166.08
$ 809.67
$ 1 ,897.94
$3,194.64
$ 0.03
$ 0.21
$ 0.28
$ 1.80
$ 5.69
$ 13.75
$ 1 1 .57
$ 60.03
$ 178.97
$ 272.33
$ 0.00
$ 0.01
$ 0.02
$ 0.13
$ 0.44
$ 1.13
$ 1.01
$ 4.89
$ 17.39
$ 25.28
$ 0.12
$ 0.73
$ 0.96
$ 6.09
$ 18.74
$ 44.43
$ 36.90
$ 186.41
$ 508.00
$ 802.43
$ 0.08
$ 0.49
$ 0.64
$ 4.04
$ 12.46
$ 29.95
$ 24.74
$ 123.37
$ 327.26
$ 523.02
$ 0.00
$ 0.03
$ 0.04
$ 0.27
$ 0.92
$ 2.33
$ 2.05
$ 9.71
$ 30.98
$ 46.98
$ 0.29
$ 1.70
$ 2.18
$ 13.58
$ 41.06
$ 98.09
$ 79.50
$ 387.92
$ 941 .56
$ 1 ,559.02
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Sources:
[A] Appendix C, Exhibits C.4d and C.5d, Column M, Row -A3
[B] Appendix C, Exhibits C.4d and C.5d, Column N, Row - A3
[C] Appendix C, Exhibits C.4d and C.5d, Column O, Row - A3
[D] Appendix C, Exhibits C.4f and C.5f, Column M, Row - A3
[E] Appendix C, Exhibits C.4f and C.5f, Column N, Row - A3
[F] Appendix C, Exhibits C.4f and C.5f, Column O, Row-A3
[G] Appendix C, Exhibits C.4e and C.5e, Column M, Row - A3
[H] Appendix C, Exhibits C.4e and C.5e, Column N, Row - A3
[I] Appendix C, Exhibits C.4e and C.5e, Column O, Row - A3
Economic Analysis for the LT2ESWTR
5-63
December 2005
-------
Exhibit 5.26a: Annualized Benefits by Filtered and Unfiltered Systems,
Preferred Alternative, Enhanced Cost of Illness1 ($ Millions/Year, 2003$)
System Type
(Population
Served)
ICR
Mean
A
5th % Me
B
95th % Me
C
ICRSSL
Mean
D
5th % Me
E
95th % Me
F
ICRSSM
Mean
G
5th % Me
H
95th % Me
1
3% Discount Rate
Illnesses and Deaths Avoided
Filtered
<1 0,000
>1 0,000
Total
$ 64.61
$ 688.49
$ 753.10
$ 5.54
$ 58.81
$ 64.63
$ 193.67
$2,125.35
$2,311.72
$ 12.17
$ 124.52
$ 136.69
$ 0.82
$ 9.11
$ 9.94
$ 40.78
$ 419.07
$ 459.51
$ 28.55
$ 292.78
$ 321.34
$ 1.98
$ 21.88
$ 23.91
$ 93.39
$ 958.66
$1,058.10
Unfiltered
<1 0,000
>1 0,000
Total
$ 9.71
$1,090.68
$1,100.40
$ 1.62
$ 148.85
$ 150.53
$ 24.14
$2,740.57
$2,762.39
$ 3.13
$ 318.17
$ 321.30
$ 0.52
$ 43.43
$ 44.01
$ 7.78
$ 799.39
$ 806.31
$ 5.42
$ 559.14
$ 564.56
$ 0.90
$ 76.33
$ 77.32
$ 13.47
$1 ,404.81
$1,416.80
7% Discount Rate
Illnesses and Deaths Avoided
Filtered
<1 0,000
>1 0,000
Total
$ 48.91
$ 557.03
$ 605.94
$ 4.19
$ 47.49
$ 51 .90
$ 146.56
$1,719.31
$1 ,861 .67
$ 9.21
$ 100.73
$ 109.94
$ 0.62
$ 7.35
$ 7.98
$ 30.87
$ 339.27
$ 370.12
$ 21.62
$ 236.85
$ 258.47
$ 1.49
$ 17.68
$ 19.20
$ 70.79
$ 776.48
$ 850.65
Unfiltered
<1 0,000
>1 0,000
Total
$ 7.36
$ 887.71
$ 895.06
$ 1.22
$ 120.81
$ 122.04
$ 18.29
$2,233.24
$2,252.00
$ 2.37
$ 258.95
$ 261.32
$ 0.39
$ 35.25
$ 35.66
$ 5.89
$ 651.57
$ 657.52
$ 4.10
$ 455.07
$ 459.17
$ 0.68
$ 61.95
$ 62.66
$ 10.20
$1,145.05
$1,155.35
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Sources:
[A] Appendix C, Exhibits C.6a, C.7a, C.8a, and C.9a, Column M, Row- A3, ICR.
[B] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column N, Row - A3, ICR.
[C] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column O, Row - A3, ICR.
[D] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column M, Row - A3, ICRSSL.
[E] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column N, Row - A3, ICRSSL.
[F] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column O, Row- A3, ICRSSL.
[G] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column M, Row - A3, ICRSSM.
[H] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column N, Row - A3, ICRSSM
[I] Appendix C, Exhibits C.6a, C.7a, C.8a and C.9a, Column O, Row - A3, ICRSSL..
Economic Analysis for the LT2ESWTR
5-64
December 2005
-------
Exhibit 5.26b: Annualized Benefits by Filtered and Unfiltered Systems,
Preferred Alternative, Traditional Cost of Illness1 ($ Millions/Year, 2003$)
System Type
(Population
Served)
ICR
Mean
A
5th % Me
B
95th % Me
C
ICRSSL
Mean
D
5th % Me
E
95th % Me
F
ICRSSM
Mean
G
5th % Me
H
95th % Me
1
3% Discount Rate
Illnesses and Deaths Avoided
Filtered
<1 0,000
>1 0,000
Total
$ 43.62
$ 466.86
$ 510.48
$ 3.06
$ 32.01
$ 35.28
$ 140.20
$1,543.95
$1,682.87
$ 8.20
$ 84.30
$ 92.50
$ 0.44
$ 4.86
$ 5.35
$ 29.34
$ 297.62
$ 328.11
$ 19.27
$ 198.43
$ 217.70
$ 1.06
$ 1 1 .45
$ 12.56
$ 67.41
$ 691.84
$ 758.56
Unfiltered
<1 0,000
>1 0,000
Total
$ 7.31
$ 823.45
$ 830.76
$ 0.93
$ 85.97
$ 86.81
$ 20.11
$2,297.23
$2,319.33
$ 2.35
$ 240.21
$ 242.57
$ 0.30
$ 25.08
$ 25.35
$ 6.48
$ 670.14
$ 677.24
$ 4.08
$ 422.14
$ 426.22
$ 0.52
$ 44.08
$ 44.53
$ 1 1 .22
$1,177.69
$1,190.00
7% Discount Rate
Illnesses and Deaths Avoided
Filtered
<1 0,000
>1 0,000
Total
$ 33.09
$ 378.78
$ 411.87
$ 2.33
$ 25.89
$ 28.45
$ 106.30
$1,251.74
$1,359.24
$ 6.22
$ 68.38
$ 74.61
$ 0.33
$ 3.95
$ 4.31
$ 22.23
$ 241.48
$ 264.21
$ 14.62
$ 160.97
$ 175.59
$ 0.80
$ 9.29
$ 10.14
$ 51.18
$ 561.54
$ 611.53
Unfiltered
<1 0,000
>1 0,000
Total
$ 5.55
$ 671.71
$ 677.25
$ 0.70
$ 70.03
$ 70.63
$ 15.24
$1,872.71
$1,889.78
$ 1.79
$ 195.94
$ 197.73
$ 0.23
$ 20.43
$ 20.62
$ 4.91
$ 546.34
$ 551.92
$ 3.09
$ 344.34
$ 347.43
$ 0.39
$ 35.90
$ 36.24
$ 8.50
$ 960.13
$ 969.79
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
[A] Appendix C, Exhibits C.6b C.7b C.8b and C.9b Column M, Row - A3, ICR.
[B] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column N, Row - A3, ICR.
[C] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column O, Row - A3, ICR.
[D] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column M, Row - A3, ICRSSL.
[E] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column N, Row - A3, ICRSSL.
[F] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column O, Row - A3, ICRSSL.
[G] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column M, Row - A3, ICRSSM.
[H] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column N, Row - A3, ICRSSM
[I] Appendix C, Exhibits C.6b C.7b C.8a and C.9b Column O, Row - A3, ICRSSL..
Economic Analysis for the LT2ESWTR
5-65
December 2005
-------
Exhibit 5.27a: Annualized Benefits by Illnesses and Deaths Avoided,
Preferred Alternative, Enhanced Cost of Illness1 ($ Millions/Year, 2003$)
Benefit Type
ICR
Mean
A
5th % Me
B
95th % Me
C
ICRSSL
Mean
D
5th % Me
E
95th % Me
F
ICRSSM
Mean
G
5th % Me
H
95th % Me
1
3% Discount Rate
Illnesses Avoided
Deaths Avoided
$ 729
$ 1,124
$ 113
$ 71
$ 1,718
$ 3,511
$ 175
$ 283
$ 29
$ 18
$ 395
$ 881
$ 345
$ 541
$ 55
$ 33
$ 838
$ 1,716
7% Discount Rate
Illnesses Avoided
Deaths Avoided
$ 587
$ 914
$ 91
$ 57
$ 1 ,382
$ 2,856
$ 141
$ 230
$ 24
$ 15
$ 318
$ 716
$ 278
$ 440
$ 44
$ 27
$ 675
$ 1 ,391
Exhibit 5.27b: Annualized Benefits by Illnesses and Deaths Avoided,
Preferred Alternative, Traditional Cost of Illness1 ($ Millions/Year, 2003$)
Benefit Type
ICR
Mean
A
5th % Me
B
95th % Me
C
ICRSSL
Mean
D
5th % Me
E
95th % Me
F
ICRSSM
Mean
G
5th % Me
H
95th % Me
1
3% Discount Rate
Illnesses Avoided
Deaths Avoided
$ 217
$ 1,124
$ 34
$ 71
$ 511
$ 3,511
$ 52
$ 283
$ 9
$ 18
$ 117
$ 881
$ 103
$ 541
$ 16
$ 33
$ 250
$ 1,716
7% Discount Rate
Illnesses Avoided
Deaths Avoided
$ 176
$ 914
$ 27
$ 57
$ 413
$ 2,856
$ 42
$ 230
$ 7
$ 15
$ 95
$ 716
$ 83
$ 440
$ 13
$ 27
$ 202
$ 1 ,391
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
5.3.2 Monetization of Benefits to Sensitive Subpopulations
The infectivity estimates used in this analysis are derived from clinical studies performed on
healthy adults and are applied to all populations. No separate monetization of benefits for sensitive
subpopulations was therefore performed. The morbidity estimates are based on a large
population—those affected by the Milwaukee outbreak—that included a mix of the general population
and sensitive subpopulations. By using morbidity factors from that outbreak, the monetization of
benefits to sensitive subpopulations is included, but could not be separately itemized. The mortality
rates used in this study are derived from data from the Milwaukee outbreak, where 46 of the 54 deaths
were persons with AIDS; the other fatalities were elderly and some had other illnesses. For normally
healthy adults, cryptosporidiosis is not considered a fatal disease. Therefore, all the mortality benefits
estimated for this rule are deaths avoided within sensitive subpopulations. Avoided illnesses and deaths
were not separately quantified for children, so monetization of these benefits is not shown separately.
Further discussion of the impact of the rule on sensitive populations is in Chapter 7, section 7.9.
Economic Analysis for the LT2ESWTR
5-66
December 2005
-------
5.4 Summary of Uncertainties
Throughout the section 5.2, the uncertainties and variability associated with each parameter of
the risk assessment model have been described. Exhibit 5.28 combines these discussions with a flow
chart illustrating each parameter and how the estimates are incorporated in the model. In the benefits
model, applying a distribution for one variable to a distribution of a second variable widens the range of
estimates, but has minimal effect on mean estimates. In the cases where the distributions are not
skewed (as is for most of the benefits variables) the mean estimate resulting from the two distributions
is similar to the result of calculating using two point estimates. Therefore, to demonstrate a range of
possible true values, all results presented in this chapter and other summary chapters show confidence
bounds along with the mean estimate.
While the assessment leading to the number of illnesses and deaths avoided incorporates
uncertainty and variability at numerous points, the monetization of these benefits also incorporates
uncertainty and variability. EPA uses two estimates for the cost of illness (enhanced and traditional)
and a distribution for values of a statistical life. EPA also conducted sensitivity analyses of the AIDS
mortality rate (see Appendix R). Exhibit 5.29 further assesses the uncertainties by indicating whether
uncertainties from either the risk assessment or monetization of benefits likely causes an under or over
estimation of benefits. In most cases, the direction of uncertainty is not known.
Economic Analysis for the LT2ESWTR 5-67 December 2005
-------
Exhibit 5.28: Uncertainty and Variability in the Number of Illnesses Avoided
Source Water Occurrence Sections 4.5.3
ICR ICRSSL ICRSSM and 5.2.4.1
/
Uncertainty addressed by sampling a distribution of possible occurrence in one loop of the model and
variability addressed by selecting an occurrence estimate from that distribution in a second inner loop.
uncertainty Treatment Secf/ons 4.5.4
_, Mode= Pre-LT2 and 5.2.4.1
Mode= ^(C — *-^ 2.75 log ^ ___ rlt; L ' ^
2.25 log — -* ""^./ 'i - A A
/ /^^^ '• / \ / \ Triangular distributions of log
/ ./ XO-. __--' small Plants removal for small and large plants
' ' &.""" accounts for variability and
, ; . A A
2 log 3 5 log /\ / \ uncertainty for those with existing
Variability — additional treatment (credit) and for
• . • Large Plants those without (no credit).
Post-LT2 - Point estimate - based on log credit achievable by each technology.
15% ,
25%
ICR
Percent of Cryptosporidium Infectious
Triangular distributions account for uncertainty in the
true mean estimate.
A Section
5.2.4.7
50%
ICRSS
Infectivity
^^X* \^^ Distribution modeled from human-challenge study
data, incorporating uncertainty and variability
Section 5.2.3
and N. 3
^
Consumption - Point Estimate
Section 5.2.4.2
Exposure Secf/on 5.2.4.3
Three different estimates by system type:
CWS - 350 person-days i=I> point estimate
NTNCWS- 250 person-days i—S point estimate
Triangular distribution of person-days to account for variability in
monthly operations periods among systems.
50% Morbidity Secf/on 5.2.3
Triangular distribution accounts for uncertainty in underlying
2Q0/o ^ ^ 70% c'a*a anc' variability of a persons response.
25% Secondary Spread Secf/on 525
/ \ Triangular distribution accounts for uncertainty in true
1 o% L. A 40% estimate
Economic Analysis for the LT2ESWTR 5-68 December 2005
-------
Exhibit 5.29: Summary of Uncertainties Affecting LT2ESWTR Benefits
Estimates
Uncertainty
Quantifying only cases of
endemic illness of
cryptosporidiosis
Infectivity for C. parvum
estimated from only three
isolates at higher dose levels
Morbidity based on triangular
distribution
Mortality based primarily on
deaths of patients with AIDS
Source water concentrations
estimated using three data
sets, calculation of central
tendencies and bounds
Proportion of measured
oocysts that were infectious,
represented by triangular
distribution
Binning assignments
Estimate of plant
implementation of enhanced
filtration
Pre-LT2 removal/1 nactivation
using triangular distributions
(with uncertain modes)
LT2 treatment log reduction
achieved
Morbidity benefits based on
COI data
Benefits from other rule
provisions EPA was not able
to quantify
Section with
Discussion
of
Uncertainty
5.2.3
5.2.3,
Appendix N
5.2.3
5.2.3,
Appendix R
5.2.4.1
5.2.4.1
4.5.6,
Appendix B
5.2.4.1
5.2.4.1
5.2.4.1
5.3.1,
Appendix
5.6.5
Effect on Benefits Estimates
Underestimate
X
X
X
Overestimate
Under or
Over
Estimate
X
X
X
X
X
X
X
X
X
Economic Analysis for the LT2ESWTR
5-69
December 2005
-------
5.5 Comparison of Regulatory Alternatives
Although the models were run separately for combinations of PWS type and size and different
Cryptosporidium occurrence data sets, all four regulatory alternatives (and the baseline) were always
computed within a given model run. This method effectively "blocked" the variability and uncertainty
from other sources so that direct, more precise comparisons could be made among all the regulatory
alternatives in a given simulation.
The five modeled regulatory conditions are (see Chapter 3 for descriptions of these):
Pre-LT2 Baseline (Regulatory Alternative AO)
Regulatory Alternative A1
Regulatory Alternative A2
Regulatory Alternative A3 (the Preferred Regulatory Alternative)
Regulatory Alternative A4
Exhibit 5.30 summarizes the quantified benefits for each regulatory alternative. Quantified
benefits do not vary substantially among alternatives for two reasons. First, roughly half of the benefits
are attributed to the unfiltered systems and the requirements for those systems are the same for each
alternative. Second, UV is the least expensive technology for most systems (thus the most selected
technology) and provides 3 log treatment regardless of the treatment requirements.
Economic Analysis for the LT2ESWTR 5-70 December 2005
-------
Exhibit 5.30a: Summary of Benefits of Annual Illnesses and Deaths Avoided
from LT2ESWTR for Regulatory Alternatives,
Enhanced Cost of Illness1 ($Millions, 2003$)
Data Set
Rule
Alternative
Estimated Value of Cases of
Illnesses Avoided ($ Millions)
Serving
< 10,000
A
Serving
> 10,000
B
All
Systems
C
Estimated Value of Deaths
Avoided ($ Millions)
Serving
< 10,000
D
Serving
> 10,000
E
All
Systems
F
Total
Value, All
Systems
G
3% Discount Rate
ICR
A1
A2
A3 - Preferred Alt.
A4
$ 35
$ 34
$ 33
$ 29
$ 713
$ 703
$ 696
$ 655
$ 748
$ 737
$ 729
$ 683
$ 4.1
$ 4.0
$ 3.9
$ 3.5
$ 110
$ 109
$ 108
$ 104
$ 1,146
$ 1,134
$ 1,124
$ 1 ,070
$ 1 ,895
$ 1,871
$ 1 ,853
$ 1 ,753
ICRSSL
A1
A2
A3 - Preferred Alt.
A4
$ 11
$ 9
$ 7
$ 5
$ 210
$ 181
$ 168
$ 145
$ 221
$ 189
$ 175
$ 150
$ 1.3
$ 1.0
$ 0.8
$ 0.6
$ 33
$ 30
$ 28
$ 26
$ 337
$ 300
$ 283
$ 254
$ 558
$ 489
$ 458
$ 405
ICRSSM
A1
A2
A3 - Preferred Alt.
A4
$ 20
$ 17
$ 15
$ 12
$ 369
$ 343
$ 329
$ 291
$ 388
$ 360
$ 345
$ 303
$ 2.0
$ 1.8
$ 1.6
$ 1.4
$ 57
$ 54
$ 52
$ 48
$ 593
$ 559
$ 541
$ 493
$ 981
$ 919
$ 886
$ 796
7% Discount Rate
ICR
A1
A2
A3 - Preferred Alt.
A4
$ 27
$ 26
$ 25
$ 22
$ 576
$ 568
$ 562
$ 529
$ 603
$ 594
$ 587
$ 551
$ 3.1
$ 3.0
$ 3.0
$ 2.7
$ 89
$ 88
$ 87
$ 84
$ 931
$ 921
$ 914
$ 870
$ 1 ,534
$ 1,515
$ 1,501
$ 1,421
ICRSSL
A1
A2
A3 - Preferred Alt.
A4
$ 9
$ 7
$ 5
$ 4
$ 170
$ 146
$ 136
$ 118
$ 178
$ 153
$ 141
$ 121
$ 1.0
$ 0.8
$ 0.6
$ 0.5
$ 27
$ 24
$ 23
$ 21
$ 274
$ 244
$ 230
$ 207
$ 452
$ 396
$ 371
$ 328
ICRSSM
A1
A2
A3 - Preferred Alt.
A4
$ 15
$ 13
$ 11
$ 9
$ 298
$ 277
$ 266
$ 235
$ 313
$ 290
$ 278
$ 244
$ 1.5
$ 1.3
$ 1.2
$ 1.0
$ 46
$ 44
$ 42
$ 39
$ 481
$ 455
$ 440
$ 401
$ 794
$ 744
$ 718
$ 645
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Sources:
[A] Appendix C, Exhibits C.4a-c and C.5a-c, Column G, Row- Small Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[B] Appendix C, Exhibits C.4a-c and C.5a-c, Column G, Row - Large Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[C] Appendix C, Exhibits C.4a-c and C.5a-c, Column G, Row-All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[D] Appendix C, Exhibits C.4a-c and C.5a-c, Column J, Row - Small Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[E] Appendix C, Exhibits C.4a-c and C.5a-c, Column J, Row - Large Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[F] Appendix C, Exhibits C.4a-c and C.5a-c, Column J, Row - All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[G] Appendix C, Exhibits C.4a-c and C.5a-c, Column M, Row - All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
Economic Analysis for the LT2ESWTR
5-71
December 2005
-------
Exhibit 5.30b: Summary of Benefits of Annual Illnesses and Deaths Avoided
from LT2ESWTR for Regulatory Alternatives,
Traditional Cost of Illness1 ($Millions, 2003$)
Data Set
Rule
Alternative
Estimated Value of Cases of
Illnesses Avoided ($ Millions)
Serving
< 10,000
A
Serving
> 10,000
B
All
Systems
C
Estimated Value of Deaths
Avoided ($ Millions)
Serving
< 10,000
D
Serving
> 10,000
E
All
Systems
F
Total
Value, All
Systems
G
3% Discount Rate
ICR
A1
A2
A3 - Preferred Alt.
A4
$ 10
$ 10
$ 10
$ 8
$ 212
$ 209
$ 207
$ 195
$ 223
$ 220
$ 217
$ 203
$ 1.2
$ 1.2
$ 1.2
$ 1.0
$ 33
$ 32
$ 32
$ 31
$ 1,146
$ 1,134
$ 1,124
$ 1 ,070
$ 1 ,369
$ 1 ,353
$ 1,341
$ 1 ,273
ICRSSL
A1
A2
A3 - Preferred Alt.
A4
$o
O
$o
O
$0
£-
$o
£-
$ 63
$ 54
$ 50
$ 43
$ 66
$ 56
$ 52
$ 45
$ 0.4
$ 0.3
$ 0.2
$ 0.2
$ 10
$ 9
$ 8
$ 8
$ 337
$ 300
$ 283
$ 254
$ 403
$ 356
$ 335
$ 299
ICRSSM
A1
A2
A3 - Preferred Alt.
A4
$ 6
$ 5
$ 4
$ 4
$ 110
$ 102
$ 98
$ 87
$ 116
$ 107
$ 103
$ 90
$ 0.6
$ 0.5
$ 0.5
$ 0.4
$ 17
$ 16
$ 16
$ 14
$ 593
$ 559
$ 541
$ 493
$ 708
$ 666
$ 644
$ 583
7% Discount Rate
ICR
A1
A2
A3 - Preferred Alt.
A4
$ 8
$ 8
$ 7
$ 6
$ 172
$ 170
$ 168
$ 158
$ 180
$ 178
$ 176
$ 165
$ 0.9
$ 0.9
$ 0.9
$ 0.8
$ 27
$ 26
$ 26
$ 25
$ 931
$ 921
$ 914
$ 870
$ 1,112
$ 1 ,099
$ 1 ,089
$ 1 ,034
ICRSSL
A1
A2
A3 - Preferred Alt.
A4
$o
O
$o
^
$ 2
$ 1
$ 51
$ 44
$ 41
$ 35
$ 53
$ 46
$ 42
$ 36
$ 0.3
$ 0.2
$ 0.2
$ 0.1
$ 8
$ 7
$ 7
$ 6
$ 274
$ 244
$ 230
$ 207
$ 327
$ 289
$ 272
$ 243
ICRSSM
A1
A2
A3 - Preferred Alt.
A4
$ 4
$ 4
$ 3
$ 3
$ 89
$ 83
$ 80
$ 70
$ 93
$ 87
$ 83
$ 73
$ 0.4
$ 0.4
$ 0.4
$ 0.3
$ 14
$ 13
$ 13
$ 12
$ 481
$ 455
$ 440
$ 401
$ 575
$ 541
$ 523
$ 474
Notes:
1The traditional COI only includes valuation for medical costs and lost work time (including some portion of unpaid
household production). The enhanced COI also factors in valuations for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the traditional COI), time with family, and recreation, and
lost productivity at work on days when workers are ill but go to work anyway.
Sources:
[A] Appendix C, Exhibits C.4d-f and C.5 d-f, Column G, Row- Small Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[B] Appendix C, Exhibits C.4d-f and C.5 d-f, Column G, Row - Large Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[C] Appendix C, Exhibits C.4d-f and C.5 d-f, Column G, Row - All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[D] Appendix C, Exhibits C.4d-f and C.5 d-f, Column J, Row - Small Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[E] Appendix C, Exhibits C.4d-f and C.5 d-f, Column J, Row - Large Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[F] Appendix C, Exhibits C.4d-f and C.5 d-f, Column J, Row - All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
[G] Appendix C, Exhibits C.4d-f and C.5 d-f, Column M, Row - All Systems, A1-A4, ICR, ICRSSL, and ICRSSM.
Economic Analysis for the LT2ESWTR
5-72
December 2005
-------
5.6 Other Benefits of LT2ESWTR Provisions
Section 5.6 describes qualitative benefits of the LT2ESWTR provisions. Although sufficient
information was not available to quantify these benefits of LT2ESWTR implementation, the
benefits—in terms of both health and monetary value—are thought to be significant.
5.6.1 Reduction in Outbreak Risk
Besides reducing the endemic risk of cryptosporidiosis, the LT2ESWTR will reduce the
likelihood of major outbreaks, such as occurred in Milwaukee. The economic value of reducing the risk
of outbreaks could be quite high when the magnitude of potential costs is considered. For example, if
the $745 COI per cryptosporidiosis infection estimate is applied to the estimated 2,000 cases attributed
to a sewage-contaminated well in Braun Station, Texas (Craun et al. 1998), health damages could reach
$1.5 million. Other costs associated with outbreaks include spending by Local, State, and national
public health agencies; emergency corrective actions by utilities; and possible legal costs if liability is a
factor. Affected water systems and local governments may incur costs through provision of alternative
water supplies and issuing customer water use warnings and health alerts. Commercial establishments
(e.g., restaurants) and their customers may incur costs due to interrupted and lost service. Local
businesses, institutions, and households may incur costs associated with undertaking averting and
defensive actions. To the extent that LT2ESWTR reduces the likelihood of waterborne disease
outbreaks, avoided response costs are potentially numerous and significant.
During outbreaks, consumers and businesses may use alternative water sources or may adopt
behaviors to reduce risk, such as boiling water. If the rule reduces the need for these risk-averting
behaviors, an economic benefit will accrue. To give a source of the possible scale, during an outbreak
of giardiasis, a disease with gastrointestinal symptoms similar to cryptosporidiosis, expenditures on
risk-averting behaviors, such as hauling in safe water, boiling water, and purchasing bottled water, were
estimated (in 2000$) between $1.74 to $5.53 per person per day during the Milwaukee outbreak
(Harrington et al. 1991). If these figures are applied to a small drinking water system serving 10,000
customers, total expenditures on risk-averting behavior during a Cryptosporidium outbreak could range
between $17,400 and $55,300 per day (2000$). Determining the reduction in outbreak risk and the
resulting benefits from avoiding risk-averting behaviors is not possible given current information, but
potential benefits are expected to be substantial.
Five studies were identified that used the averting cost approach to estimate household and
other costs attributable to short-term contamination of drinking water supplies (Abdalla 1990; Abdalla
et al. 1992; Harrington et al. 1991; Sun et al. 1992; Van Houtven et al. 1997). The most relevant of
these for the LT2ESWTR analysis is a study by Harrington et al. (1991), that analyzes the costs
associated with drinking water contamination by Giardia in Luzerne County, Pennsylvania. The
December 1983 outbreak resulted in 366 confirmed giardiasis cases resulting from sewage leaking into
the unfiltered source water. The total affected population was 75,000 individuals across Pittston
Borough and 17 other municipalities. The Harrington study also developed a theoretical and empirical
example of how outbreak costs are incurred, based on the Luzerne County example.
The four steps associated with a waterborne outbreak that may impose costs on society are
discovery, survey and testing, reaction, and aftermath. (Harrington et al. 1991). These are described
below:
Discovery. Health care providers or State, Local, or hospital laboratory technicians send
reports to State authorities notifying them of the need for further investigation when the
rate of new cases suddenly increases above the normal rate.
Economic Analysis for the LT2ESWTR 5-73 December 2005
-------
Survey and testing. Epidemiological surveys may be conducted, along with tests of the
water supply, once a few cases are confirmed.
Reaction. Local authorities and the water system may issue boil-water advisories or other
warnings to reduce exposure once a link is made between the drinking water supply and the
disease outbreak. Businesses, as well as households, may be affected by such action,
requiring government agencies to begin surveillance and enforcement activities and, in
some cases, provide alternative water sources.
• Aftermath. Long-term solutions to the problem are discussed, as well as how the costs of
the outbreak and prevention of future ones may be shared. These discussions can only take
place once the outbreak is contained by actions taken during the previous phase.
In the Luzerne County outbreak, individuals took actions to avoid exposure to the contaminated
water and those actions resulted in estimated losses between $20.8 million and $61.8 million (2000$).
The predominant cost was due to the need to boil water and the associated time lost. Losses due to
averting actions for restaurants, bars, schools and other businesses during the outbreak exceeded $1.0
million. The burden for government agencies was $230,000 and the outbreak cost the water supply
utility $1.8 million. These costs are in year 2000$ and do not include legal fees, adverse effects on
businesses (which were not investigated), leisure activities, or net losses due to substituting more
expensive beverages for tap water.
5.6.2 Costs to Households to Avert Infection
In addition to averting actions taken with regard to outbreaks, a reduction of everyday risk-
averting behaviors can be achieved. Many households may undertake on a daily basis the same
averting actions that they take during an outbreak (e.g., buying bottled water, boiling water, installing
point-of-use (POU) filtration). To the extent that the LT2ESWTR can be expected to reduce a
household's perceptions of the health risks associated with drinking water, regulatory action may
reduce the frequency of such averting actions and their costs.
5.6.3 Enhanced Aesthetic Water Quality
Some treatment improvements resulting from the implementation of the rule are likely to
improve the aesthetic quality of the drinking water. Consumers, presumably, would be willing to pay to
protect the aesthetic quality of drinking water, and therefore, these benefits should result in an
economic benefit. However, the benefits from such water quality improvements due to the rule are not
quantified for this analysis.
5.6.4 Risk Reduction from Co-occurring and Emerging Pathogens
While the benefits analysis for the LT2ESWTR only includes reductions in illness and
mortality attributable to Cryptosporidium, the rule is expected to reduce exposure to other pathogens
(e.g., Giardia or other waterborne bacterial or viral pathogens such as Cyclospora and
Microsporidium). For example, membrane processes that remove Cryptosporidium are also shown to
achieve equivalent log removal of Giardia under worst-case and normal operating conditions, and
nanofiltration also shows similar removal of Giardia as Cryptosporidium (USEPA 2003c). Goodrich
and Lykins (1995) evaluated bag filters and concluded that any microbe or object greater than 4.5
microns in size would be subject to 0.5 to 2.0 log removal. Strengthened regulatory requirements will
translate into increased removal of additional pathogens and a resulting reduction in risk. This may
prove valuable in reducing overall risk because the impact of emerging pathogens, although not well
established, could be significant.
Economic Analysis for the LT2ESWTR 5-74 December 2005
-------
5.6.5 Benefits from Reduction in Disinfection Byproducts
Treatment changes resulting from implementation of the rule may lead to decreased or
increased disinfection byproducts (DBFs). Systems that install chemical disinfection technologies like
ozone may increase certain DBFs but systems that install physical disinfection technologies like
membranes or UV and reduce their chemical disinfectant may reduce certain DBFs. A reduction in
DBFs could lead to reduction in DBF-related negative health effects.
5.6.6 Benefits from Other Rule Provisions
The benefit estimates discussed in this chapter result from increased treatment requirements that
improve the consumers' water quality. However, other provisions of the LT2ESWTR not directly
involving changes to treatment practices will also provide benefits to water consumers. Due to data
constraints, EPA was not able to quantify these benefits. Instead, a qualitative discussion of these
benefits is provided below.
Benefits of Source Water Monitoring
While source water monitoring does not provide any direct monetary benefits, the information
gained from turbidity, Cryptosporidium, and E. coli testing may provide benefits to the water systems
and ultimately to their customers. Although some large systems currently monitor their source water
for these contaminants, many do not. Most small systems do not monitor source water. Monitoring
allows systems to better understand variations in their source water and to adjust their operations
accordingly. For example, if a system discovers that pathogen levels are elevated in the spring, they
could plan to add more coagulant or bring another sedimentation basin online during that period.
Systems that find little or no Cryptosporidium will be able to boost consumer confidence in their water,
providing benefits through fewer home treatment devices and less time spent in dealing with customer
complaints. Systems that detect Cryptosporidium can use that information for public education about
source water protection and watershed management. These can help bring about changes in watershed
protection that will ultimately result in better source water quality. Improved source water quality can
produce cost savings for treatment.
Benefits of Covered Finished Water Reservoirs
The quality of water in uncovered finished water reservoirs is subject to similar environmental
influences as surface water, including deposition of airborne chemicals, surface water runoff, animal
carcasses, animal or bird droppings, and growth of algae and other aquatic organisms. In one study,
gulls contaminated a 10 million gallon reservoir and increased bacteriological growth; in another,
waterfowl were found to elevate coliform levels in small recreational lakes by 20 times their normal
levels (Morra 1979). Algal growth increases the biomass in the reservoir, which reduces dissolved
oxygen and thereby increases the release of iron, manganese, and nutrients from the sediments. This, in
turn, supports more algal growth (Cooke and Carlson 1989). Algae can cause taste and odor problems.
Further, uncovered finished water reservoirs may be subject to contamination by illegal swimming and
dumping. Documented water quality problems in open finished water reservoirs include increased algal
cells; heterotrophic plate count (HFC) bacteria; turbidity; color; particle counts; biomass; and decreased
chlorine residuals (Pluntze 1974; AWWA 1983; Silverman et al. 1983; LeChevallier et al. 1997b).
Finished water is usually not treated or tested again prior to consumption, so any contamination
in the uncovered reservoir may be passed directly to the customer. Therefore, requirements to cover all
finished water reservoirs or to treat the effluent will reduce the risk of contamination and result in
positive health benefits. Covering reservoirs or providing additional treatment of finished water will
Economic Analysis for the LT2ESWTR 5-75 December 2005
-------
also provide some additional protection from possible acts of terrorism. Data are not available,
however, to quantify the benefits associated with covering all finished water reservoirs.
5.6.7 Summary of Nonqualified Benefits
As explained above, several types of potential benefits were not included in the quantitative
analysis. Exhibit 5.31 shows how the rule provisions that have not been quantified would be expected
to affect the overall benefits derived from LT2ESWTR.
Exhibit 5.31: Summary of Nonquantified Benefits
Benefit Type
Potential Effect
on Benefits
Comments
Reducing outbreak risks and
response costs
Increase
Some outbreaks are caused by human or
equipment failures that may occur even with
the proposed new requirements; however, by
adding barriers of protection for some systems,
the rule will reduce the possibility of such
failures leading to outbreaks.
Reducing averting behavior (e.g.,
boiling tap water or purchasing
bottled water)
Increase / No
Change
Consumers in systems that cease using
uncovered finished water reservoirs (through
covering or taking such reservoirs off-line) my
have greater confidence in water quality. This
may result in less averting behavior that
reduces both out-of-pocket costs (e.g.,
purchase of bottled water) and opportunity
costs (e.g., time to boil water).
Improving aesthetic water quality
Increase
Some technologies installed for this rule
(e.g., ozone) are likely to reduce taste and
odor problems.
Reducing risk from co-occurring
and emerging pathogens
Increase
Although focused on removal of
Cryptosporidium from drinking water,
systems that change treatment processes
will also increase removal of pathogens
that the rule does not specifically regulate.
Additional benefits will accrue.
Increased source water
monitoring
Increase
The greater understanding of source water
quality that results from monitoring may
enhance the ability of plants to optimize
treatment operations in ways other than
those addressed in this rule.
Increased or decreased DBFs
Increase or
Decrease
Systems that install chemical disinfection
technologies like ozone may increase
certain DBPs. Systems that install
physical disinfection technologies like
membranes or UV and reduce chemical
disinfectant usage may reduce certain
DBPs.
Economic Analysis for the LT2ESWTR
5-76
December 2005
-------
Benefit Type
Potential Effect
on Benefits
Comments
Covering all finished water
reservoirs
Increase
Although insufficient data were available to
quantify benefits, the reduction of
contaminants introduced through
uncovered finished water reservoirs would
produce positive public health benefits.
Economic Analysis for the LT2ESWTR
5-77
December 2005
-------
6. Cost Analysis
6.1 Introduction
This chapter presents estimates of the total national costs for the four LT2ESWTR regulatory
alternatives. Total national costs include the costs of rule implementation, monitoring for bin
classification, additional treatment, and future monitoring. These costs are summarized first, and then
individual methodologies and cost details are provided. This chapter also summarizes per-household
costs for all systems covered by the rule.
The estimated costs of this rule, as presented in this chapter, are highly dependent on the
estimated occurrence of Cryptosporidium in source water. As discussed in Chapter 4, EPA uses three
different occurrence data sets, the ICR, ICRSSL, and ICRSSM. Each has uncertainties surrounding its
applicability to nationwide occurrence. As there is no data set that is clearly superior, separate cost
analyses were conducted using each of them. To further reflect the uncertainty and variability of the
occurrence data, this chapter calculates cost estimates corresponding to the 90 percent confidence bounds
of each data set. Section 6.11 provides a full discussion on the uncertainty and variability associated with
the cost estimates.
Section 6.1.1 describes the assumptions used to estimate national costs of the LT2ESWTR. The
remaining sections detailed below present the estimated costs of the LT2ESWTR.
6.2 Rule Implementation Costs
6.2.1 PWSs
6.2.2 States and Other Primacy Agencies
6.3 Source Water Monitoring for Initial Bin Classification Costs
6.3.1 PWSs
6.3.2 State and Other Primacy Agency Costs
6.4 Treatment Costs
6.4.1 Toolbox Options and Unit Costs
6.4.2 Compliance Forecast and Technology Selection
6.4.3 Capital and Annual Costs
6.5 Costs of Treatment for Unfiltered Plants
6.6 Costs for Benchmarking and Technology Reporting Requirements
6.7 Costs of Treatment for Uncovered Finished Water Reservoirs
6.7.1 Unit Costs
6.7.2 Compliance Forecast and Technology Selection
6.7.3 Total Annual Treatment Costs
6.8 Future Source Water Monitoring
6.9 Summary of the National Costs of the LT2ESWTR
6.10 Household Costs
6.11 Summary of Uncertainties and Sensitivity Analyses
6.11.1 Cryptosporidium Occurrence Data Sets
6.11.2 Sensitivity Analysis of Influent Bromide Levels on Technology Selection for
Filtered Plants
6.12 Unquantifiable Costs
6.13 Comparison of Regulatory Alternatives
Appendix D offers a comprehensive explanation of the laboratory and labor costs for rule
implementation, E. coll and Cryptosporidium monitoring for initial bin classification, and future
Economic Analysis for the LT2ESWTR 6-1 December 2005
-------
monitoring. Appendices E, F, and G support the cost estimates for filtered and unfiltered systems (unit
costs, methodology for technology selection, and summary of technology selection, respectively).
Appendix I presents unit costs for uncovered finished water reservoirs, and Appendix J summarizes the
per-household cost estimation methodology. Appendix Q describes all cost model programs and files.
6.1.1 Cost Description and Assumptions
To estimate the total national costs of the rule, EPA estimated costs to be incurred by public water
systems (PWSs) and by States or other Primacy Agencies. For PWSs, these include the costs of installing
treatment, the costs to administer the program and understand the rule, and monitoring costs. State and
Primacy Agency costs include estimates of the labor burdens that these agencies will face, such as
training employees on the requirements of the LT2ESWTR, reviewing PWS reports, responding to
inquiries, and record keeping.
EPA estimated costs for these activities using cost models, equipment price lists and quotes, wage
rates from the Bureau of Labor Statistics, stakeholder inputs, and assumptions used in economic analyses
performed for earlier drinking water rules. This section discusses the assumptions on discount rates,
wage rates, laboratory fees, and uncertainty parameters. More details on cost assumptions and results are
in Appendices D through J. EPA expresses costs as annualized values amortized over 25 years, based on
the present value of the stream of costs that occur overtime. All cost estimates are expressed in year
2003 dollars.
Systems Subject to Rule Provisions and Activities
The LT2ESWTR applies to all surface water systems and to systems that use ground water under
the direct influence of surface water (GWUDI). These are classified as community water systems
(CWSs), nontransient noncommunity water systems (NTNCWSs), or transient noncommunity water
systems (TNCWSs), as described in Chapter 4. Unfiltered and filtered systems are subject to different
rule provisions, as are systems with uncovered finished water reservoirs. It is estimated that all unfiltered
systems will add treatment to meet the rule's requirements, and all systems with uncovered finished water
reservoirs will cover their reservoirs or treat the discharge to inactivate viruses. See Chapter 4, sections
4.4 and 4.6 for baseline numbers of unfiltered systems and uncovered finished water reservoirs subject to
the rule.
Exhibit 6.1 shows the estimated number of systems and plants that are subject to rule
implementation, source water monitoring, treatment, and benchmarking costs. All nonpurchased systems
will incur rule implementation costs. Plants achieving 5.5 log treatment of Cryptosporidium will incur
neither treatment costs nor, possibly, source water monitoring costs, depending on when they meet the 5.5
log treatment with respect to LT2ESWTR promulgation.
EPA assumes no unfiltered plant currently achieves 2 log inactivation of Cryptosporidium as
required by the LT2ESWTR; therefore, all unfiltered plants will incur costs for adding treatment. Filtered
plants with source water monitoring results of 0.075 oocysts/L or greater will be required to provide
additional treatment for Cryptosporidium. All systems with uncovered finished reservoirs must cover
their reservoir or treat the effluent. Purchased systems do not have direct treatment costs and, thus, are
not included in Exhibit 6.1 (unless they have uncovered reservoirs); however EPA recognizes that they
will likely incur indirect costs through rate increases by the seller.
Economic Analysis for the LT2ESWTR 6-2 December 2005
-------
Exhibit 6.1: Number of Systems and Plants Expected to Incur Costs,
Preferred Alternative
Dataset
ICR
ICRSSL
ICRSSM
Nonpurchased Systems and Plants
System Size
(population
served)
< 10,000
> 10,000
Total
< 10,000
> 10,000
Total
< 10,000
> 10,000
Total
Systems
Incurring
Implementation
Costs
A
5,663
1,493
7,156
Source Water Monitoring - Plants
Initial E.
Co//
Monitoring
B
5,575
1,733
7,308
Same as ICR
Same as ICR
Initial
Crypto
Monitoring
C
1,978
1,762
3,741
1,285
1,762
3,047
1,555
1,762
3,317
Future
£. co//
Monitoring
D
4,977
1,184
6,161
5,237
1,379
6,615
5,181
1,306
6,487
Future
Crypto
Monitoring
E
1,732
1,184
2,916
1,171
1,379
2,550
1,409
1,306
2,715
Plants
Adding
Treatment
F
2,205
677
2,882
1,428
440
1,868
1,729
531
2,260
Systems
with
Uncovered
Reservoirs
G
12
69
81
Same as
ICR
Same as
ICR
Note: Detail may not add to totals due to independent rounding. Plants adding treatment in column F include
purchased unlinked plants (see section 4.3.2).
Sources:
[A] Appendix D, Exhibit D.4, column A.
[B] Appendix D, Exhibit D.4, column D.
[C] ICR—Appendix D, Exhibit D.4, column F; ICRSSL—Appendix D, Exhibit D.6, column F; ICRSSM—Appendix D,
Exhibit D.5, column F.
[D] ICR—Appendix D, Exhibit D.4, column I; ICRSSL—Appendix D, Exhibit D.6, column I; ICRSSM—Appendix D,
Exhibit D.5, column I.
[E] ICR—Appendix D, Exhibit D.4, column J; ICRSSL—Appendix D, Exhibit D.6, column J.
[F] ICR—Appendix G, Exhibits G.37-G.39, column A; ICRSSL—Appendix G, Exhibit G.43-G.45, column A;
ICRSSM—Appendix G, Exhibit G.49-G.51, column A. All include Exhibit 4.5, column C.
[G] Exhibit 4.23 and assuming one reservoir per system.
Scheduling and Discounting Assumptions for National Costs
Nominal costs for both non-treatment and treatment activities are of two kinds: (1) one-time costs
that occur near the beginning of the rule implementation period, and (2) annual "steady-state" costs that
systems and States/Primacy Agencies will incur after systems have made necessary changes to treatment
and/or monitoring to comply with the LT2ESWTR. For the purposes of this Economic Analysis (EA),
one-time and steady-state costs were projected over a 25-year time period to coincide with the estimated
life span of capital equipment (typically estimated as 20 years for most technologies) and an average time
lag of up to 5 years for technology installation after rule promulgation. Some portion of costs for the
capital costs of technology provide benefits beyond the 25-year time horizon of this analysis. For
example, for equipment installed in year 15, half of its useful life (and cost) is properly compared to
benefits generated within the 25-year period of analysis, and half of it useful life (and cost) will generate
benefits beyond this period. As a result, the proportion of costs that generate benefits after 25 years is
deducted from the capital costs before being discounted into present values. The projected schedules for
all rule activities are summarized in Appendix O.
Economic Analysis for the LT2ESWTR
6-3
December 2005
-------
As described previously in the Chapter 5 discussion of benefits, it is common practice to adjust
benefits and costs to a present value1 using a social discount rate so that they can be compared to one
another. This process takes into account the time preference that society places on expenditures and
allows comparison of cost and benefit streams that are variable over a given time period.2 Similar to
calculating the present value of benefits (see section 5.3.1.5), the present value of costs for any future
period can be calculated using the following equation:
PV = V(t)/(l+R)t
Where: t = The number of years from the reference period
R = Social discount rate
V(t) = The cost occurring t years from the reference period
The present values presented in this EA are the sum of the PVs for each year.
There is much discussion among economists of the proper social discount rate to use for policy
analysis. Therefore, for LT2ESWTR cost analyses, present value calculations are made using two social
discount rates thought to best represent current policy evaluation methodologies, 3 and 7 percent.
Historically, 3 percent is based on rates of return on relatively risk-free investments, as described in the
Guidelines for Preparing Economic Analyses (USEPA 2000d). The rate of 7 percent is recommended by
the Office of Management and Budget (OMB) as an estimate of "before-tax rate of return to incremental
private investment" (USEPA 1996b). For any future cost, the higher the discount rate, the lower the
present value. Specifically, a future cost (or stream of costs) evaluated at a 7 percent social discount rate
will always result in a lower total present value cost than the same future cost evaluated at a 3 percent
rate.
To allow evaluation on an annual basis, the total present value costs are annualized using the
same social discount rates (3 and 7 percent) over 25 years. Unlike the total present value, the relationship
between annualized costs at 3 and 7 percent is dependent on the time frame for annualization, as well as
when the costs are incurred (as set forth in the schedule of rule activities, Appendix O, Exhibits O.7-O.9).
When applying social discount rates to annualize costs, the higher (7 percent) discount rate will yield
lower annualized costs if costs occur early in a period and the present value is annualized over fewer
years (e.g., a 25-year stream of costs paid out or annualized over 5 years). Given a long enough time
frame, the 7 percent annualized value will eventually be greater than the 3 percent annualized value.
Thus, it is possible for the present value costs discounted at 7 percent to be lower than those discounted at
3 percent, and the annualized values to be opposite, with the value annualized at 7 percent to be higher
than those annualized at 3 percent.
Labor Rates
For costing purposes, EPA estimates the labor needs and hourly labor rates of systems and States
for two labor categories: managerial and technical. EPA recognizes that there may be significant
variation in labor rates across all PWSs. However, the best estimates available for use in national
estimates of the effect of drinking water rules are from Labor Costs for National Drinking Water Rules
1 For purposes of analyses in this EA, all present value figures are presented at a year 2003 price level.
Present value calculations are performed to the expected year of rule implementation (2005).
2 See EPA's Guidelines for Preparing Economic Analyses (USEPA 2000d) for a full discussion of the use
of social discount rates in the evaluation of policy decisions.
Economic Analysis for the LT2ESWTR 6-4 December 2005
-------
(USEPA, 2003b), which are used in this EA. The technical and managerial wage rates vary with system
size and include fringe benefits. The technical and managerial wage rates (2003$) are shown in
Exhibit 6.2.
Exhibit 6.2: Wage Rates by System Size
Loaded Wage Rate ($2003)
Technical labor rate
Managerial labor rate
Labor Cost (per hour)
System Size (Population Served)
<100
$ 21.44
$ 44.36
$ 21.44
100499
$ 23.09
$ 47.78
$ 23.09
500-3,299
$ 24.74
$ 51.20
$ 24.74
3,300-9,999
$ 25.34
$ 51.20
$ 30.51
10,000-99,999
$ 26.05
$ 51.20
$ 31.08
>100,000
$ 31.26
$ 51.20
$ 35.25
Notes: EPA estimates that systems with population greater than 3,300 use a combination of operators (technical) and
engineers (managerial), with an 80/20 ratio between the two, respectively.
Source: Labor Costs for National Drinking Water Rules (USEPA 2003b).
To represent the composition of staff at PWSs of smaller sizes (e.g., systems serving fewer than
3,300 people), EPA uses only the technical rate. For systems serving 3,300 or more people, EPA uses a
ratio of 80 percent technical labor to 20 percent managerial labor to arrive at a weighted labor rate of
$30.51 for systems serving 3,301-10,000 people, $31.08 for systems serving 10,001-100,000 people, and
$35.25 for systems serving greater than 100,000 people.
Labor costs attributable to States for administrative tasks are estimated using an average annual
full time equivalent (FTE) labor cost, including overhead and fringe benefits, of $65,255 (2001$). This
rate was established based on data from the 2001 State Drinking Water Needs Analysis (ASDWA 2001).
For use in the LT2ESWTR EA analyses, the $65,255 annual rate was updated to a year 2003 level
($70,132) and converted to an hourly basis (1 FTE = 2,080 hours) to establish a State rate of $33.60 per
hour.
Laboratory Fees
A laboratory fee, expressed as a cost per sample, is associated with E. coll and Cryptosporidium
monitoring and with future monitoring for bin reclassification. Exhibit 6.3 summarizes the range of fees
estimated from a survey of laboratories and EPA's experience during the ICR and ICRSSs. Cost
calculations for this EA use the average laboratory cost for water sample analyses. Costs are calculated
on a per-plant basis to be consistent with costs for treatment. Appendix D provides a more detailed
derivation of the laboratory costs.
Some of the factors that could cause the cost per sample to differ from one system to another are
regional variations in laboratory fees, the number of samples processed (quantity discounts) and
laboratory capacity for Cryptosporidium analysis.
Economic Analysis for the LT2ESWTR
6-5
December 2005
-------
Exhibit 6.3: E. coli and Cryptosporidium Laboratory Costs
Analyte
Cryptosporidium
£ coli
Utility Laboratory
Commercial Laboratory
Laboratory Cost Per Sample
Average
A
$ 530
21
70
Range
Min
B
$ 389
12
60
Max
C
$ 713
38
85
Total
Laboratory
Cost per Plant
D
$13,767
$546
$1,820
Source: Appendix D, described in sections D.4.1 and D.4.2.
[A] Cryptosporidium - Exhibit D.14a, column F; £ coli - Exhibit D.12, columns F and I.
[B] and [C] Cryptosporidium - DynCorp (2002); £ coli - DynCorp (2000).
[D] Column [A] multiplied by the number of samples (26) (biweekly samples for £. coli, and 24 regular samples
for Cryptosporidium plus 2 spiked samples).
6.2 Rule Implementation Costs
This section presents the estimated costs for PWSs and States/Primacy Agencies to perform
administrative activities associated with the LT2ESWTR. These cost estimates are the same for all
regulatory alternatives.
6.2.1 PWSs
All nonpurchased surface water and GWUDI systems subject to the LT2ESWTR (including
filtered and unfiltered systems) will incur one-time costs for implementation activities that include time
for staff to read the rule and become familiar with its provisions and for training employees on rule
requirements. The technical and managerial labor rates, as presented in section 6.1.1, were used along
with estimates of labor hours, to calculate rule implementation costs for all systems. Labor rates used to
estimate implementation costs vary by activity and system size. Costs for systems serving up to 3,300
people are based on only the technical rate. For those systems serving at least 3,300 people, costs are
based on an assumed 80/20 split between technical and managerial labor rates. Labor hour estimates are
based on EPA's experience with previous rules.
6.2.2 States and Other Primacy Agencies
State and other Primacy Agency implementation activities include:
Adopting the regulation and developing the program
Training State or other Primacy Agency staff
Training PWS staff and providing technical assistance
Updating the data management system
To estimate implementation costs to States/Primacy Agencies, the number of FTEs per activity is
multiplied by the number of labor hours per FTE, the cost per labor hour, and the number of
States/Primacy Agencies.
Economic Analysis for the LT2ESWTR
6-6
December 2005
-------
EPA estimates the number of FTEs required per activity based on previous experience with other
rules. In estimating State/Primacy Agency costs, a labor rate of $33.60 is assumed (see section 6.1.1).
The number of States and territories includes the 50 states, six territories (American Samoa,
Commonwealth of the Northern Marianas, Guam, Palau, Puerto Rico, and the Virgin Islands), and the
Navajo Nation, for a total of 57 entities.
Exhibit 6.4 summarizes the estimated implementation costs for PWSs and States to comply with
the LT2ESWTR. Implementation costs are the same for all rule options. In Appendix D, Exhibits D.10
and D.I 1 provide detailed estimates of hours and calculations for implementation costs by system size
(Appendix D costs are undiscounted; discounted costs are shown in Appendix O).
Exhibit 6.4: Estimated Costs of Implementation, Present Value— Discounted at 3
and 7 Percent, Preferred Alternative ($Millions, 2003$)
System / State
< 10,000
> 10,000
System Total
State Total
3 Percent
$ 1.12
$ 0.39
$ 1.51
$ 7.66
7 Percent
$ 1.04
$ 0.38
$ 1.42
$ 7.52
Note: Detail may not add to totals due to independent rounding.
Sources:
System costs: Appendix O, Exhibit O.11a and O.14a, Column A.
State costs: Appendix O, Exhibit O.10a and O.13a, Column A.
6.3 Source Water Monitoring for Initial Bin Classification Costs
6.3.1 PWSs
Source water monitoring costs are estimated on a per-plant basis. Purchased plants are assumed
not to treat source water and not to have any monitoring costs. There are three types of monitoring that
plants may be required to conduct—turbidity, E. coll, and Cryptosporidium. Source water turbidity is a
water quality parameter that most plants measure frequently for operational control. Also, to meet the
SWTR, IESWTR, and LT1ESWTR requirements, most water systems have turbidity analysis equipment
in house and their operators are experienced in its use. Thus, EPA assumes that the incremental turbidity
monitoring burden associated with the LT2ESWTR is negligible.
For large and medium systems, all are required to conduct monthly E. coll and Cryptosporidium
monitoring for 2 years to determine initial bin classification. If systems achieve, or will achieve at the
time treatment is required, 5.5 log Cryptosporidium removal/inactivation, they are not required to conduct
source water monitoring. (See Exhibits 4.5 and 4.11 for the baseline number of unfiltered and filtered
plants conducting monitoring, respectively.)
Economic Analysis for the LT2ESWTR
6-7
December 2005
-------
Small systems, except those achieving at least 5.5 log Cryptosporidium removal/inactivation,
must conduct bi-weekly E.coli monitoring for 1 year. Systems exceeding the following levels must
monitor semi-monthly for Cryptosporidium:
• Annual mean > 10 E.coli/100 mL for lakes and reservoirs
• Annual mean > 50 E.coli/100 mL for flowing streams
To estimate source water monitoring costs for small systems, this EA assumes that all systems
(except those achieving 5.5 log treatment) will conduct E. coll monitoring and only those predicted to
require additional treatment (from the binning assignment prediction discussed in section 4.5.4) will
monitor Cryptosporidium in their source water (see Appendix D, Exhibit D.4 for a calculation of plants
conducting E.coli and Cryptosporidium monitoring). EPA will continue to investigate the use of a
surrogate for determining source water microbial quality, in order to reduce the cost burden on these
systems. If a reliable indicator is identified, EPA will issue guidance for system monitoring. The costs in
this chapter may, therefore, be an overestimate of actual costs.
From the modeled Cryptosporidium occurrence distributions, EPA estimated the percentage of
plants that would fall into treatment bins for each rule option. Exhibit 6.5 presents the annualized
monitoring costs based on the modeled Cryptosporidium occurrence distributions. Appendix D provides
an explanation of how these costs are developed.
6.3.2 State and Other Primacy Agency Costs
Because EPA will directly manage the data collection for initial source water monitoring of large
and medium systems, States/Primacy Agencies are predicted not to incur any costs for these activities.
They will, however, incur costs from the small system initial monitoring requirement. The delayed start
of small system monitoring will allow some States to assume primacy for that effort. To estimate
State/Primacy Agency costs, the number of FTEs required per activity is multiplied by the number of
labor hours per FTE, the State/Primacy Agency labor hour cost, and the number of States or Primacy
Agencies. Exhibit 6.5 presents the estimated total cost incurred by States/Primacy Agency for initial
source water monitoring (see Appendix D, Exhibit D.I 7, for the derivation of costs).
Economic Analysis for the LT2ESWTR 6-8 December 2005
-------
Exhibit 6.5: Cost Estimates for Initial Source Water Monitoring, Present Value—
Discounted at 3 and 7 Percent, Preferred Alternative ($Millions, 2003$)
System
Size
< 10,000
> 10,000
Total
< 10,000
> 10,000
Total
ICR
Mean
$ 33.79
$ 25.22
$ 59.01
$ 29.05
$ 23.38
$ 52.42
Confidence Bounds
5th %ile
$ 32.11
$ 25.22
$ 57.33
$ 27.64
$ 23.38
$ 51.01
95th %ile
$ 36.63
$ 25.22
$ 61.86
$ 31.45
$ 23.38
$ 54.82
ICRSSL
Mean
$ 25.24
$ 25.22
$ 50.46
$ 21.85
$ 23.38
$ 45.22
Confidence Bounds
5th %ile | 95th %ile
3 Percent
$ 22.02
$ 25.22
$ 47.24
7 Percent
$ 19.13
$ 23.38
$ 42.50
$ 27.44
$ 25.22
$ 52.66
$ 23.70
$ 23.38
$ 47.07
ICRSSM
Mean
$ 28.57
$ 25.22
$ 53.79
$ 24.65
$ 23.38
$ 48.02
Confidence Bounds
5th %ile
$ 26.36
$ 25.22
$ 51.58
$ 22.79
$ 23.38
$ 46.17
95th %ile
$ 30.43
$ 25.22
$ 55.66
$ 26.22
$ 23.38
$ 49.60
Notes: Detail may not add to totals due to independent rounding.
Includes laboratory costs, labor costs, and reporting costs.
Sources: All data from Appendix O, and from Alternative A3.
At 3 percent:
Mean: Sum of columns B, C, and D from Exhibits O.11a (ICR), O.11d (ICRSSM), and O.11g (ICRSSL)
5th percentile: Sum of columns B, C, and D from Exhibits O.11b (ICR), O.11e (ICRSSM), and O.11h (ICRSSL)
95th percentile: Sum of columns B, C, and D from Exhibits O.11c(ICR), O.11f (ICRSSM), and O.11I (ICRSSL)
At 7 percent:
Mean: Sum of columns B, C, and D from Exhibits O.14a (ICR), O.14d (ICRSSM), and O.14g (ICRSSL)
5th percentile: Sum of columns B, C, and D from Exhibits O.14b (ICR), O.14e (ICRSSM), and O.14h (ICRSSL)
95th percentile: Sum of columns B, C, and D from Exhibits O.14c(ICR), O.14f (ICRSSM), and O.14I (ICRSSL)
6.4 Treatment Costs
As shown in Chapter 4, filtered plants make up the majority of surface water and GWUDI plants
subject to the LT2ESWTR. Costs of treatment associated with these plants make up most of the costs
associated with the LT2ESWTR. This section reviews the cost methodology and total costs for the
Preferred Regulatory Alternative as follows:
6.4.1 Discusses the technologies and describes how plant unit costs ($/plant) are derived.
6.4.2 Presents the compliance and technology selection forecasts.
6.4.3 Describes how total capital investment and annualized costs are estimated for each size
category and rule option.
Treatment costs are calculated by estimating the number of plants that will be adding treatment
technologies, then multiplying these estimates by the unit costs ($/plant) of the selected technologies.
Although individual information is not available on every plant, assumptions are made about the
percentages of plants nationwide that will select each technology. Capital improvement costs are
converted to present values and are then annualized using the discount rates presented in section 6.1.1.
Exhibit 6.6 provides an overview of the methodology used to generate national treatment costs.
Economic Analysis for the LT2ESWTR
6-9
December 2005
-------
Exhibit 6.6: Methodology for Estimating Treatment Costs
Calculations discussed in section 6.4.1
Design Flows
per Plant1
Technology
Cost Equations
(Capital)
Point Estimates
of Capital Unit
Costs ($/plant)2
Baseline Number of
Plants1
Calculations discussed in section 6.4.2
Compliance
Treatment Forecasts
Technologj
Selections
Number of
Plants Selecting
Technologies 2
Average Daily
Flows per
Plant1
Technology
Cost Equations
(O&M)
Point Estimates
of Annual O&M
Costs ($/plant)2
Calculations discussed in section 6.4.1
Cost
Uncertainty of
+/-15%by
Technology
Calculations discussed in section 6.4.3
Total Capital
Costs
Compliance
Schedule
Discount Rates
Total Annual
O&M Costs
Shading scheme:
Present Value
of Costs
Annualized Value
of Costs
| | Data produced from inputs
| | Results
1 Baseline information is discussed in Chapter 4.
2 Capital unit costs, annual O&M costs, and number of plants selecting technologies are also used in the derivation of household
cost distributions. See section 6.9 and Appendix J for further discussion.
Economic Analysis for the LT2ESWTR
6-10
December 2005
-------
6.4.1 Toolbox Options and Unit Costs
Under the LT2ESWTR, systems will have a "toolbox" of treatment and management options to
select from to meet their bin requirements (listed in Exhibit 1.2). This section describes those options that
plants will most likely select to meet the LT2ESWTR requirements and the typical conditions under
which the selected technologies may be installed. The derivation of per plant capital costs and annual
O&M costs for each technology is also discussed. Unit cost data for each technology for CWS plants are
presented in Appendix E.
Exhibit 6.7 summarizes the technologies considered for estimating benefits and costs of the
LT2ESWTR. Note, some toolbox options are not considered in this analysis and therefore not included in
the Exhibit 6.7. The following section briefly describes each of the options not included and provides
reasons why they were not considered. Section 6.5.1.2 discusses the technologies that were considered
for this EA.
6.4.1.1 Toolbox Options Not Considered in the Cost Analysis
Changing Sources
Plants may avoid adding treatment by connecting to a nearby water system or changing to a
source of water that has lower Cryptosporidium concentrations. EPA estimates that approximately 22
percent of small systems are located in metropolitan areas where distances between potential connecting
water systems may allow interconnection at reasonable costs (USEPA 2000f). Changing the water source
may also be possible, although EPA has no information on the number of systems that may have less
contaminated sources available to them. Even if EPA could estimate how many systems might choose
these options, the costs for both are highly variable. Costs for connecting to another system involve costs
for new pipes and facilities, and for those consolidating with another water system, as well as costs
associated with merging management capabilities. Costs for changing to another water source would
depend on the new source and the system. Costs could include such items as drilling a new well,
installation of new piping, and new treatment facilities or adaptation of existing ones to handle different
source quality. The costs for both of these options would be highly system-dependent and are difficult to
predict. For these reasons, these toolbox options were not considered in these cost estimates.
Intake Changes
Relocating the intake and managing the timing or level of withdrawal are all toolbox options.
The purpose of these options is to change the location or timing of withdrawal of water from the source in
order to draw from those locations and at those times when Cryptosporidium concentrations are lower.
These options may cost little compared to adding treatment, especially for systems drawing from
reservoirs. The costs would depend on the existing intake structures and the nature of the source. It is
unknown how many systems could likely take advantage of such strategies and how much reduction they
might achieve. Because of this uncertainty, these options were not considered in this LT2ESWTR
analysis.
Chlorine Dioxide
Cryptosporidium can be inactivated by chlorine dioxide; however, the operational and water
quality conditions are challenging to meet. The disinfection capability of chlorine dioxide rapidly
diminishes as water temperature decreases, as reflected by the high contact time (CT) requirements below
10 degrees Celsius. Chlorine dioxide when added to water forms chlorite, a regulated disinfection
Economic Analysis for the LT2ESWTR 6-11 December 2005
-------
byproduct, thus limiting the dose. It also degrades relatively rapidly, which limits the allowable contact
time.
EPA used SWAT (see Section 4.2.3) to predict the percent of plants that could achieve 0.5 log
removal or treatment while not exceeding the chlorite MCL. The results showed only a few percent could
achieve this credit. Therefore, this EA assumes no systems will be able to use chlorine dioxide to comply
with Cryptosporidium inactivation requirements.
The rule does provide chlorine dioxide and ozone CT requirements for 0.25 log credit; and thus,
allowing the combination of the two disinfectants to receive 0.5 log credit (no other toolbox options have
presumed credits of 0.25 log or an increment of 0.25 log). Due to the high cost of installing two
technologies and the limitations on their use, EPA did not evaluate this option for receiving 0.5 log credit,
but recognizes some systems may elect this option.
Performance Studies
Some plants may have filtration processes that achieve a higher level of Cryptosporidium
removal across their treatment train than assumed by the LT2ESWTR. Plants may conduct performance
studies of their treatment process to demonstrate to the State that the required level of Cryptosporidium
removal or inactivation is being achieved. The number of systems is unknown, but likely very low, that
could demonstrate a higher level of treatment on a consistent basis. Also, the cost of such studies could
be higher than implementing another toolbox option.
The toolbox options that were omitted from the costs analysis may be less expensive than the
technologies considered and, therefore, the costs presented here may be overestimated. Cost uncertainties
for this analysis are summarized in section 6.12.
6.4.1.2 Technologies Considered for the LT2ESWTR Cost Analysis
Exhibit 6.7 shows technologies that were included in the cost analysis for the LT2ESWTR. The
second column summarizes the condition(s) under which the technology use is constrained in this EA.
Plants may be constrained from installing a technology for various reasons. During the Small Surface
Water Delphi process, industry experts and the Technical Work Group (TWG) identified limitations in
the use of several technologies. A more extensive explanation of these groups and their conclusions can
be found in the Stage 2 DBPR Economic Analysis (USEPA 2003b).
The third column in Exhibit 6.7 identifies the conditions for which cost estimates are developed.
To capture the variation in costs, many technologies are evaluated over a range of possible influent water
qualities and operating conditions. For the purposes of estimating the costs of the LT2ESWTR, the TWG
selected the water quality and operating parameters to capture the "normal" circumstances under which
plants may use the technology. EPA does not assume that all systems would be designed or operated
under these conditions, but that the cost equations generate capital and O&M costs that are typical for the
range of system types and sizes. While these assumptions simplify the true variety of operating
conditions, EPA believes they are adequate to enable reasonable estimates of national costs.
EPA recognizes that similar systems may experience different capital or O&M costs due to
site-specific factors. Inputs to unit costs such as water quality conditions, labor rates, and land costs can
be highly variable and increase the system-to-system variability in unit costs. In developing the unit cost
estimates, there is insufficient information to fully characterize what the distributions of this variability
will be on a national scale for all of the treatment technologies and all possible conditions.
Instead, the unit costs are developed as average or representative estimates of what these unit
costs will be on a national basis. That is, in developing unit costs, design criteria for the technologies are
Economic Analysis for the LT2ESWTR 6-12 December 2005
-------
selected to represent typical, or average, conditions for the universe of systems. As a result, there is
uncertainty inherent in these unit cost estimates reflecting the fact that they are based on independent
assumptions with supporting data and vendor quotes, where available, rather than on a detailed
aggregation of State, regional, or local estimates based on actual field conditions. In this EA, uncertainty
in these national average unit costs factors is characterized as follows:
Capital costs: triangular distribution of +/- 30%
O&M costs: triangular distribution of+/-15%
These estimates were developed by EPA and reflect information presented by the National
Drinking Water Advisory Council (2001) in its review of the national cost estimation methodology for
the Arsenic Rule. EPA believes that the uncertainties in capital and O&M costs for a given technology
are independent of one another, and that uncertainties across all technologies are independent.
The final column in Exhibit 6.7 provides the source for the unit costs. The unit costs are derived
from equations and other information in the Technology and Cost document (USEPA 2003a), and are
revised to incorporate labor rates from Labor Costs for National Drinking Water Rules (USEPA, 2003b).
Unit costs presented in Appendix E are based on labor rates presented in Exhibit 6.2, and are in 2003$.
For some technologies, unit costs were provided by the FACA committee and its TWG in June 2000 as
part of the Stage 2 M-DBP negotiation process or are estimated from earlier EPA rules.
In most cases, technology unit costs are specified for single average daily and design flows for
nine system size categories. In reality, there will be a range of unit costs for a category, relative to the
range of flows. EPA believes, however, that using a mean unit cost ($/plant) in each category, derived
from mean flow data, provides an accurate representation of total national costs. For technologies that are
a combination of two or more unit processes (e.g., cartridge filters with ozone), the technology unit costs
are simply assumed to be the sum of the costs for each unit process. This may result in the overestimation
of "combined" technology options since some economies of scale are expected when such combined
technologies are implemented.
Economic Analysis for the LT2ESWTR 6-13 December 2005
-------
Exhibit 6.7: Toolbox Options Considered for the LT2ESWTR
Technology
Bag Filtration
Cartridge Filtration
Combined Filter Performance
Bank Filtration
Microfiltration/
Ultrafiltration (MF/UF)
Ozone (O3)
Secondary Filter (SF)
UV[3]
Watershed Control (WC)
Constraints1
Not practical for systems serving
more than 10,000 people
Not practical for systems serving
more than 10,000 people
None
Not practical for systems serving
fewer than 10,000 people
None
Not practical for systems serving
500 people or fewer
Not practical for systems serving
500 people or fewer
None
Not practical for systems serving
10,000 people or fewer
Water Quality and Operational
Parameters
Bag replacement four times per year
Cartridge replacement twice per year
None
None
0.3 NTU, 10°C, disposal of reject stream to
sewer
Concentration for 0.5, 1 .0, and 2.0 log of
Cryptosporidium inactivation determined
by SWAT analyses.
None
Median water quality parameters:
UV254 = 0.051cm-1, turbidity = 0.1 NTU,
alkalinity = 60 mg/L as CaCO3,
hardness = 100 mg/L as CaCO3
None
Source of Unit Cost2 in
T&C Document
Figures D-19and D-20
Figures D-21 and D-22
Figures D-30 and D-31
Figure D-23
Figures D-17and D-18
Figure D-11 through Figure
D-16
Figures D-24 and D-25
Figures D-7 and D-8
Figures D-28 and D-29
Constraints identified by the TWG and the Small Surface Water Delphi Group (USEPA 2003a).
2Unit costs ($/plant) for CWSs provided in Appendix E for each technology shown.
3Patent fee of $0.015/1,000 gallons included in household costs, see section 6.10 for rationale.
Economic Analysis for the LT2ESWTR
6-14
December 2005
-------
6.4.2 Compliance Forecast and Technology Selection
Three key inputs are required to estimate the technologies plants will add to comply with
LT2ESWTR:
1. The percent of plants that must make a treatment change to meet the LT2ESWTR
requirements.
2. The treatment technologies these plants have in place prior to implementation of the rule.
3. The treatment technologies these plants are predicted to select to comply with the rule.
These inputs, coupled with baseline data presented in Chapter 4, provide an estimate of the number of
filtered plants that will use each technology to meet the requirements of the LT2ESWTR. Input 1 above
is largely a result of classifying systems into treatment "bins" based on initial source-water
Cryptosporidium monitoring (see section 3.3.1). The Cryptosporidium occurrence distributions derived
from the ICR, ICRSSL, and ICRSSM (see section 4.2) are used to determine the bin assignments for all
sizes of systems, yielding three sets of bin assignments for each rule alternative.
The second input accounts for plants that already have some of the toolbox technologies in place
and will be able to obtain credit for Cryptosporidium treatment without any additional costs or obtain a
portion of their required treatment credit. Therefore, the existing technologies cannot be considered as a
possible technology selection for these plants. To accommodate the two treatment baseline scenarios, two
separate technology selection forecasts are used for plants that already have toolbox technologies in place
and those that do not. These two forecasts follow the same methodology as described below, with the
former omitting the existing technologies from the selection process.
The overall methodology used to develop the technology selection forecast in the third input is
based on the "least-cost principle." EPA assumes that the cost of the rule is best estimated by assuming
that drinking water plants will select the least expensive technology or combination of technologies
available to meet the treatment requirements of a given action bin. EPA recognizes not all plants may not
be able to implement certain technologies because of site-specific conditions. Therefore, the technology
selections are limited by maximum use percentages. Appendix F details the methodology for technology
selection forecast, including assumptions for estimating maximum use percentages.
Exhibit 6.8 shows the selections for the Preferred Alternative, derived from the mean ICR,
ICRSSL, and ICRSSM occurrence distributions. Appendix G provides results for all regulatory
alternatives and sensitivity analyses. In many cases, the least costly technology results in higher levels of
Cryptosporidium inactivation or removal than required for that bin (this is always the case when UV is
selected). Although direct filtration plants have 0.5 log higher bin requirements than conventional and
other filtration plants, no additional treatment is estimated for them. The 0.5 log higher requirement for
those plants is adequately addressed by the higher levels of Cryptosporidium treatment achieved by the
selected technologies.
Economic Analysis for the LT2ESWTR 6-15 December 2005
-------
Exhibit 6.8: Technology Selection Forecast for Filtered Plants
Technology
Selections1
Bag Filter
1.0 Log
Cartridge Filter
2.0 Log
Combined Filter
Performance
0.5 Log
In-bank Filtration
1.0 Log
MF/UF
2. 5 Log
Data Set'
ICR
1,523
209
16
6
37
ICRSSL
1,219
20
12
5
13
ICRSSM
1,421
58
14
6
18
Technology
Selections1
Ozone
0.5 Log
Ozone
1.0 Log
Ozone
2.0 Log
Secondary Filter
1.0 Log
UV
2.5 Log
WS Control
0.5 Log
Data Set'
ICR
27
18
10
0
979
0
ICRSSL
21
14
3
0
503
0
ICRSSM
25
16
4
0
641
0
Notes:
Selection includes non-purchased plants in CWSs, NTNCWSs, and TNCWSs adding treatment; and purchased
plants that could not be linked with their sellers. Some plants select more than one technology to meet the bin
requirements.
forecasts from the mean occurrence distributions. Forecasts from 95th and 5th percentile distributions presented in
Appendix G.
Source: Appendix G; ICR data from Exhibits G.37-39; ICRSSL data from Exhibits G.43-45; ICRSSM data from
Exhibits G.49-51.
6.4.3 Capital and Annual Costs
To estimate the treatment costs for filtered plants, the technology unit costs (capital and O&M)
are multiplied by the number of plants within each size category predicted to install each technology. The
O&M costs are costs that systems incur yearly to maintain system performance. The capital costs are
adjusted to account for the life-span of the capital equipment, assumed to be 20 years. The methodology
for projecting and discounting costs is as follows:
Project all undiscounted costs (capital and O&M) over a 25-year time horizon based on the
rule implementation schedule in Appendix O.
• Calculate total present value costs using social discount rates of 3 and 7 percent (see section
6.1.1 for a discussion of these rates).
Annualize the costs over 25 years using the same social discount rates.
Appendix O, Exhibit O.2 shows the schedule of implementation used to determine when systems will
install the equipment.
The treatment costs for CWS, NTNCWS, and TNCWS plants are calculated separately since each
type has different population-per-system averages, producing different mean flows (see Exhibit 4.4a) and,
thus, different mean plant unit costs. These costs are summed across technologies and size categories to
estimate the total treatment costs for LT2ESWTR. Exhibit 6.9 shows the mean and 90 percent confidence
bound estimates for the Preferred Alternative based on the ICR, ICRSSL, and ICRSSM occurrence
distributions.
Economic Analysis for the LT2ESWTR
6-16
December 2005
-------
Exhibit 6.9: Treatment Costs for Filtered Systems, Discounted at 3 and 7 Percent,
Preferred Alternative ($Millions, 2003$)
System size
(Population
Served)
Capital - Present Value
Mean
A
5th %ile
B
95th %ile
c
O & M - Annualized
Mean
D
5th %ile
E
95th %ile
F
Total - Annualized
Mean
G
5th %ile
H
95th %ile
i
ICR
3 Percent
<10K
>10k
Total
$ 82.64
$ 930.88
$ 1,013.52
$ 64.16
$ 734.43
$ 798.59
$ 104.23
$ 1,181.51
$ 1,285.74
$ 4.41
$ 26.95
$ 31.35
$ 3.68
$ 22.82
$ 26.51
$ 5.25
$ 31.80
$ 37.05
$ 9.15
$ 80.40
$ 89.56
$ 7.37
$ 65.00
$ 72.37
$ 11.24
$ 99.65
$ 110.88
7 Percent
<10K
>10k
Total
$ 61.03
$ 756.28
$ 817.30
$ 47.38
$ 596.62
$ 644.00
$ 76.97
$ 959.92
$ 1,036.90
$ 3.63
$ 23.72
$ 27.35
$ 3.04
$ 20.09
$ 23.13
$ 4.33
$ 27.99
$ 32.32
$ 8.87
$ 88.62
$ 97.48
$ 7.10
$ 71.28
$ 78.39
$ 10.94
$ 110.36
$ 121.29
ICRSSL
3 Percent
<10K
>10k
Total
$ 42.16
$ 543.39
$ 585.56
$ 28.04
$ 365.33
$ 393.37
$ 55.67
$ 712.22
$ 767.90
$ 2.32
$ 14.09
$ 16.41
$ 1.67
$ 10.35
$ 12.02
$ 2.87
$ 17.20
$ 20.06
$ 4.74
$ 45.30
$ 50.04
$ 3.28
$ 31.33
$ 34.61
$ 6.06
$ 58.10
$ 64.16
7 Percent
<10K
>10k
Total
$ 31.14
$ 441.34
$ 472.48
$ 20.70
$ 296.69
$ 317.40
$ 41.12
$ 578.50
$ 619.61
$ 1.91
$ 12.39
$ 14.30
$ 1.38
$ 9.10
$ 10.48
$ 2.36
$ 15.13
$ 17.49
$ 4.58
$ 50.27
$ 54.85
$ 7.10
$ 71.28
$ 78.39
$ 10.94
$ 110.36
$ 121.29
ICRSSM
3 Percent
<10K
>10k
Total
$ 53.77
$ 674.22
$ 727.99
$ 39.88
$ 505.63
$ 545.51
$ 68.22
$ 850.51
$ 918.73
$ 2.94
$ 17.94
$ 20.87
$ 2.36
$ 14.62
$ 16.98
$ 3.50
$ 21.13
$ 24.63
$ 6.02
$ 56.66
$ 62.68
$ 4.65
$ 43.66
$ 48.31
$ 7.41
$ 69.97
$ 77.39
7 Percent
<10K
>10k
Total
$ 39.71
$ 547.63
$ 587.34
$ 29.45
$ 410.66
$ 440.11
$ 50.38
$ 690.86
$ 741.24
$ 2.42
$ 15.78
$ 18.20
$ 1.94
$ 12.86
$ 14.81
$ 2.88
$ 18.59
$ 21.47
$ 5.83
$ 62.77
$ 68.60
$ 7.10
$ 71.28
$ 78.39
$ 10.94
$ 110.36
$ 121.29
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O.
[A]-[C] Exhibit O.12 (3%) and O.15 (7%); Columns A-C.
[D]-[F] Exhibit O.18 (3%) AND O.21 (7%); Columns D-F.
[G]-[l] Exhibit O.18 (3%) AND O.21 (7%); Sum of columns A and D, Band E, and C and F.
6.5 Costs of Treatment for Unfiltered Plants
As summarized in Chapter 1, the LT2ESWTR requires all unfiltered plants to achieve 2 log
Cryptosporidium inactivation if their source water concentration is less than or equal to 1 oocyst per 100
liters, and 3 log inactivation if it is greater than 1 oocyst per 100 liters. UV is the least expensive
technology that can achieve the required log inactivation of Cryptosporidium and, therefore, the most
likely to be installed. However, as with filtered systems, EPA estimated that a small percentage of the
plants would elect to install a technology more expensive than UV due to the configuration of existing
equipment or other factors. Ozone is the next cheapest technology that will meet the inactivation
requirements, and therefore is projected to be used by plants that cannot use UV. Due to the high
concentrations of ozone necessary to inactivate Cryptosporidium, EPA estimated that ozone can achieve
only 2.0 log inactivation and all systems that are required to obtain 3.0 log inactivation are assumed to use
UV. Although the toolbox lists UV as achieving 2.5 log Cryptosporidium inactivation, studies have
shown that it can achieve much greater inactivation at the standard drinking water dose of 40 mJ/cm2
(Clancy et al. 2000; Craik et al. 2001). MF/UF was considered as a substitute for UV in this analysis, but
is much more expensive than these technologies, especially for larger systems.
Economic Analysis for the LT2ESWTR
6-17
December 2005
-------
Unit costs for UV and ozone and the conditions under which they can be used are the same as for
the filtered plants3 (section 6.5). Ozone unit costs are for an ozone concentration that provides 2.0 log of
Cryptosporidium inactivation.
All unfiltered plants must meet the requirements of the LT2ESWTR; therefore, 100 percent of
such plants will add technology. The following assumptions were used in developing the technology
selection forecast for plants needing 2.0 log Cryptosporidium inactivation:
100 percent of very small systems will use UV.
90 percent of other plants will be able to install UV (the least expensive of the two
technologies).
• 10 percent of other plants will add ozone.
Consistent with assumptions for filtered systems, very small plants were also assumed to be
unable to use ozone. This is because of the high level of operator attention and training required to
operate and maintain an ozone system. Because of the high costs for alternatives like MF/UF for very
small plants, the analysis assumes that all these plants would find a way to use UV. EPA believes this to
be a reasonable assumption because small plants have less existing infrastructure that might limit their
technology selection. Exhibit 6.10 summarizes the treatment costs at 3 and 7 percent discount rates.
3EPA incorporated project costs for New York City's planned UV system due to the extraordinarily large
size of their water system in comparison to all other systems, which are adequately represented by the unit costs.
Economic Analysis for the LT2ESWTR 6-18 December 2005
-------
Exhibit 6.10: Treatment Costs for Unfiltered Systems, Discounted at 3 and 7
Percent, Preferred Alternative ($Millions, 2003$)
System Size
(Population
Served)
Capital - Present Value
Mean
A
5th %ile
B
95th %ile
c
O & M -Annualized
Mean
D
5th %ile
E
95th %ile
F
Total -Annualized
Mean
G
5th %ile
H
95th %ile
I
3 Percent
<10K
>10k
Total
$ 5.69
$ 407.27
$ 412.96
$ 4.82
$ 325.03
$ 329.86
$ 6.57
$ 488.60
$ 495.16
$ 0.28
$ 2.09
$ 2.36
$ 0.25
$ 1.90
$ 2.16
$ 0.30
$ 2.27
$ 2.56
$ 0.60
$ 25.47
$ 26.08
$ 0.53
$ 20.57
$ 21.10
$ 0.67
$ 30.33
$ 31.00
7 Percent
<10K
>10k
Total
$ 4.20
$ 335.56
$ 339.76
$ 3.56
$ 267.78
$ 271.35
$ 4.85
$ 402.58
$ 407.43
$ 0.23
$ 1.85
$ 2.07
$ 0.21
$ 1.68
$ 1.89
$ 0.24
$ 2.01
$ 2.25
$ 0.59
$ 30.64
$ 31.23
$ 0.52
$ 24.66
$ 25.18
$ 0.66
$ 36.55
$ 37.21
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O.
[A]-[C] Exhibit O.12 (3%) and O.15 (7%); Columns G-l.
[D]-[F] Exhibit O.18 (3%) and O.21 (7%); Columns J-L.
[G]-[l] Exhibit O.18 (3%) and O.21 (7%); Sum of columns G and J, H and K, and I and L.
6.6 Costs of Disinfection Profiling and Benchmarking and of Technology Reporting
The disinfection profiling and benchmarking requirement was introduced in the IESWTR and
LT1ESWTR. In the LT2ESWTR, it requires any system proposing a change to its disinfection process to
complete an evaluation of the existing process and consult with the State/Primacy Agency about how the
proposed change will affect disinfection performance. To estimate costs for this provision of the rule,
EPA assumes the systems selecting UV and ozone will require time to evaluate data and consult with the
State. Appendix D provides further detail on the derivation of costs for systems and for States. Exhibit
6.11 presents the estimated costs for systems to develop a disinfection profile, calculate a benchmark, and
consult with the State.
Exhibit 6.11: Disinfection Profiling and Benchmarking Estimated Costs, Present
Value-Discounted at 3 and 7 Percent, Preferred Alternative ($Millions, 2003$)
System
Size
ICR
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile
95th %ile
3 Percent
< 10,000
> 10,000
Total
$ 0.06
$ 0.07
$ 0.13
$ 0.05
$ 0.07
$ 0.12
$ 0.06
$ 0.08
$ 0.14
$ 0.03
$ 0.05
$ 0.08
$ 0.03
$ 0.04
$ 0.07
$ 0.04
$ 0.05
$ 0.09
$ 0.04
$ 0.06
$ 0.09
$ 0.04
$ 0.05
$ 0.08
$ 0.04
$ 0.06
$ 0.10
7 Percent
< 10,000
> 10,000
Total
$ 0.04
$ 0.06
$ 0.10
$ 0.04
$ 0.05
$ 0.09
$ 0.05
$ 0.06
$ 0.11
$ 0.02
$ 0.04
$ 0.06
$ 0.02
$ 0.03
$ 0.05
$ 0.03
$ 0.04
$ 0.07
$ 0.03
$ 0.04
$ 0.07
$ 0.03
$ 0.04
$ 0.07
$ 0.03
$ 0.05
$ 0.08
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O; Exhibit O.11 (3%) and O.14 (7%); Column H.
Systems required to meet additional Cryptosporidium treatment will also have to report
compliance monitoring data, depending on the toolbox option implemented. EPA assumes UV, ozone,
bank filtration, and microfiltration technologies will impose additional burden for reporting compliance.
Exhibit 6.12 presents the estimated costs for systems to conduct the reporting activities. Appendix D,
Exhibits D.21-D.23 for systems and Exhibits D.18-D.20 for States/Primacy Agencies, show the
Economic Analysis for the LT2ESWTR
6-19
December 2005
-------
derivation of system and States/Primacy Agency costs and total costs for technology compliance
reporting.
Exhibit 6.12: Technology Reporting Estimated Costs, Annualized at 3 and 7
Percent, Preferred Alternative ($Millions, 2003$)
System
Size
ICR
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile
95th %ile
3 Percent
< 10,000
> 10,000
Total
$ 0.36
$ 0.48
$ 0.83
$ 0.32
$ 0.44
$ 0.76
$ 0.39
$ 0.53
$ 0.92
$ 0.21
$ 0.31
$ 0.52
$ 0.19
$ 0.25
$ 0.44
$ 0.22
$ 0.35
$ 0.57
$ 0.24
$ 0.37
$ 0.61
$ 0.22
$ 0.33
$ 0.55
$ 0.26
$ 0.41
$ 0.66
7 Percent
< 10,000
> 10,000
Total
$ 0.30
$ 0.41
$ 0.71
$ 0.26
$ 0.38
$ 0.64
$ 0.32
$ 0.45
$ 0.78
$ 0.17
$ 0.27
$ 0.44
$ 0.15
$ 0.22
$ 0.37
$ 0.18
$ 0.30
$ 0.49
$ 0.20
$ 0.32
$ 0.52
$ 0.18
$ 0.29
$ 0.47
$ 0.21
$ 0.35
$ 0.56
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O; Exhibit O.17 (3%) and O.20 (7%); Column I.
6.7 Costs of Treatment for Uncovered Finished Water Reservoirs
As part of the LT2ESWTR, systems with uncovered finished water reservoirs have the options to
cover the reservoirs or to provide disinfection downstream of the reservoir. Disinfection alternatives must
achieve at least 2 log of Cryptosporidium, 3 log Giardia, and 4 log of virus inactivation. To develop
national cost estimates for systems to comply with this provision of the LT2ESWTR, unit costs for each
treatment alternative and the percentage of systems selecting each alternative are estimated for the
inventory of systems having uncovered finished water reservoirs (presented in Chapter 4). This section
summarizes the methodology for developing the unit costs of reservoir covers and disinfection. Appendix
I provides further details on the derivation of unit costs. The basis for estimating the number of systems
that select each alternative is also discussed in this section, followed by the presentation of national cost
estimates for this provision of the rule.
6.7.1 Unit Costs
There are two types of reservoir covers—fixed and floating. Fixed covers are commonly
constructed of concrete, steel, or aluminum. Floating covers are flexible membrane structures generally
made of polypropylene or similar material. The unit costs for fixed covers are estimated from information
provided in the Uncovered Finished Water Reservoirs Guidance Manual (USEPA 1999c). The unit costs
for floating covers were estimated by obtaining vendor quotes (detailed in Appendix I). They are both
based on the estimated surface area of a reservoir and the average cost of materials.
Economic Analysis for the LT2ESWTR
6-20
December 2005
-------
Systems have the option to disinfect the water leaving the reservoir instead of installing a cover.
A review of available disinfection options showed that the combination of UV and gaseous chlorine is the
least-cost option to achieve the required Cryptosporidium, Giardia, and virus inactivation. Unit costs of
gas chlorination presented in Appendix I are a function of flow and include the costs of typical process
equipment and the chemical building. Unit costs for UV are the same as those used for filtered and
unfiltered treatment plants.
6.7.2 Compliance Forecast and Technology Selection
The technology selection methodology for the uncovered finished water reservoirs also uses a
least-cost approach. For systems with reservoir capacities of 10 million gallons (MG) or less, covering is
the least expensive alternative and all of this size are assumed to install covers. The technology selection
for the remaining reservoirs is split between installing covers and disinfecting the effluent. To meet the
disinfection requirements, a combination of UV and chlorine is the least expensive option for achieving
virus, Giardia, and Cryptosporidium inactivation requirements. However, the ability of a system to use
booster chlorination depends on its current residual disinfectant type. Approximately 50 percent of all
surface water systems are predicted to use chloramination following implementation of the Stage 2
DBPR. Adding chlorine to water treated with chloramines can cause quality problems; therefore, a
maximum of 50 percent of systems were assumed to add booster chlorination after the reservoir.
Because the technology selection is based on least costs, and fixed-cover costs are the most
expensive treatment option considered, no systems were assumed to install fixed covers. EPA recognizes
that some systems may select fixed covers for other reasons, but the incremental costs are not attributable
to this rule.
Systems will also incur costs for reporting the presence of an uncovered finished water reservoir
to the Primacy Agency and for submitting a plan to either cover or treat the reservoir. Costs for reporting
are covered in detail in Appendix D.
6.7.3 Total Annual Treatment Costs
Total annual treatment costs are calculated by multiplying the number of reservoirs in a category
by the percent selecting a technology and the unit cost for that technology. Exhibit 6.13a summarizes the
costs for all systems with uncovered finished water reservoirs to comply with the rule. Exhibit 6.13b
shows the reporting costs for systems to report to the Primacy Agency.
Economic Analysis for the LT2ESWTR 6-21 December 2005
-------
Exhibit 6.13a: Cost for Systems with Uncovered Finished Water Reservoirs,
Annualized at 3 and 7 Percent ($Millions, 2003$)
System Size
(Population
Served)
<1 0,000
>1 0,000
Total
Annualized Cost at 3%
Capital
$ 0.01
$ 6.52
$ 6.53
O&M
$ 0.00
$ 3.73
$ 3.73
Total
$ 0.01
$ 10.24
$ 10.26
Annualized Cost at 7%
Capital
$ 0.01
$ 9.39
$ 9.40
O&M
$ 0.00
$ 3.68
$ 3.68
Total
$ 0.02
$ 13.07
$ 13.08
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O, Exhibits O.18a (3%) and O.21a (7%), Columns N and Q.
Exhibit 6.13b: Reporting Cost for Systems with Uncovered Finished Water
Reservoirs, Annualized at 3 and 7 Percent ($Millions, 2003$)
System Size
(Population
Served)
<1 0,000
>1 0,000
Total
Annualized
Cost at 3%
$ 0.0002
$ 0.0011
$ 0.0012
Annualized
Cost at 7%
$ 0.0002
$ 0.0016
$ 0.0018
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O, Exhibits O.17a (3%) and O.20a (7%), Column J.
6.8 Future Source Water Monitoring
Six years after initial bin classification, filtered plants will be required to conduct a second round
of monitoring to reassess source water conditions for bin assignments. EPA will evaluate new analytical
methods and surrogate indicators of Cryptosporidium in the interim. While the costs of monitoring are
likely to change in the 6 years following rule promulgation, it is difficult to predict how they will change.
In the absence of other information, it was assumed that the laboratory costs would be the same as for the
initial monitoring. All plants that conduct initial monitoring are assumed to conduct future monitoring as
well, except for those systems that achieve 5.5 log Cryptosporidium treatment credit. Exhibit 6.14 shows
the costs for future monitoring. Costs vary among the Cryptosporidium occurrence data sets because the
numbers of plants that add technologies and achieve 5.5 log treatment credit differ and the number of
small plants triggered into Cryptosporidium monitoring differ. Confidence bounds represent the low (5th
percentile) and high (95th percentile) occurrence distributions for each data set. Appendix D, Exhibits
D.30-D.35 show the calculations for the cost estimates.
Economic Analysis for the LT2ESWTR
6-1
December 2005
-------
Exhibit 6.14: Future Monitoring Cost Estimates, Present Value-Discounted at 3
and 7 Percent, Preferred Alternative ($Millions, 2003$)
System
Size
ICR
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile | 95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile
95th %ile
3 Percent
< 10,000
> 10,000
Total
$ 22.71
$ 13.20
$ 35.91
$ 21.86
$ 13.63
$ 35.49
$ 24.40
$ 12.50
$ 36.89
$ 17.74
$ 15.37
$ 33.11
$ 15.52
$ 16.14
$ 31.66
$ 19.23
$ 14.84
$ 34.07
$ 19.92
$ 14.55
$ 34.48
$ 18.47
$ 15.09
$ 33.57
$ 21.13
$ 14.10
$ 35.23
7 Percent
< 10,000
> 10,000
Total
$ 13.79
$ 8.85
$ 22.63
$ 13.28
$ 9.14
$ 22.42
$ 14.79
$ 8.38
$ 23.17
$ 10.82
$ 10.31
$ 21.13
$ 9.49
$ 10.82
$ 20.32
$ 11.72
$ 9.95
$ 21.67
$ 12.13
$ 9.76
$ 21.89
$ 11.26
$ 10.12
$ 21.38
$ 12.85
$ 9.45
$ 22.30
Note: Detail may not add to totals due to independent rounding.
Includes laboratory costs, labor costs, and reporting costs.
Sources: Appendix O, Exhibits O.11 (3%) and O.14 (7%), Sum of columns E, F, and G.
6.9 Summary of the National Costs of the LT2ESWTR
This section combines the cost estimates described in the previous sections to summarize the total
national cost of the rule. Costs are broken out in several ways: by type of system or plant subject to rule
provisions, by system size, and by nature of the cost (one-time or annual). All costs presented in this
section are for the Preferred Regulatory Alternative for the LT2ESWTR. Cost estimates for the other
regulatory alternatives are presented in section 6.13.
Exhibit 6.15 summarizes the estimated total initial capital and one-time costs of the LT2ESWTR,
which include costs of rule implementation and all monitoring activities. Costs are presented for the
means of the three occurrence distributions. Exhibit 6.16 follows with annualized costs and includes
confidence bounds for each occurrence data set to show the full range of estimates. As can be seen in
Exhibit 6.15, treatment costs comprise most of the costs of the rule. Exhibit 6.17a and 6.17b, therefore,
provide further detail of treatment costs-a breakdown by system size for all three occurrence distributions
(ICR, ICRSSL, and ICRSSM), including confidence bounds, discounted at 3 percent and 7 percent.
Economic Analysis for the LT2ESWTR
6-23
December 2005
-------
Exhibit 6.15: Initial Capital and One-Time Costs, Undiscounted,
Preferred Alternative ($Millions, 2003$)
Type of Cost
Serving < 10,000 People
ICR | ICRSSL
ICRSSM
Serving > 10,000 People
ICR [ ICRSSL
ICRSSM
All Systems
ICR
ICRSSL
ICRSSM
Total
National (System + State)
$2,104.32
$ 1,526.27
$ 1,719.41
System
System Total
Treatment
Implementation
Initial Monitoring
Second Monitoring
Benchmarking
Tech Reporting
Uncovered Reservoirs
$ 214.30
$ 140.74
$ 1.19
$ 38.03
$ 33.47
$ 0.07
$ 0.65
$ 0.14
$ 132.32
$ 76.26
$ 1.19
$ 28.27
$ 26.05
$ 0.04
$ 0.37
$ 0.14
$ 157.93
$ 94.75
$ 1.19
$ 32.07
$ 29.30
$ 0.05
$ 0.43
$ 0.14
$ 1,869.74
$ 1,706.86
$ 0.39
$ 26.77
$ 18.01
$ 0.08
$ 0.74
$ 116.88
$1,373.81
$ 1 ,208.24
$ 0.39
$ 26.77
$ 20.97
$ 0.06
$ 0.49
$ 116.88
$ 1 ,541 .30
$1,376.73
$ 0.39
$ 26.77
$ 19.86
$ 0.07
$ 0.58
$ 116.88
$2,084.04
$ 1 ,847.60
$ 1.59
$ 64.80
$ 51 .48
$ 0.16
$ 1.39
$ 117.03
$ 1,506.13
$ 1 ,284.50
$ 1.59
$ 55.04
$ 47.02
$ 0.10
$ 0.86
$ 117.03
$ 1 ,699.23
$ 1,471.48
$ 1.59
$ 58.84
$ 49.16
$ 0.11
$ 1.01
$ 117.03
State
State Total
Implementation
Initial Monitoring
Second Monitoring
Benchmarking
Tech Reporting
Uncovered Reservoirs
$ 20.28
$ 7.77
$ 5.98
$ 6.18
$ 0.09
$ 0.27
$ 0.00
$ 20.15
$ 7.77
$ 5.98
$ 6.18
$ 0.06
$ 0.17
$ 0.00
$ 20.19
$ 7.77
$ 5.98
$ 6.18
$ 0.07
$ 0.19
$ 0.00
Notes: Detail may not add to totals due to independent rounding.
Sources: All data from Appendix O, and from Alternative A3.
System: 8
Treatment: Sum of columns B and H from Exhibits O.6a (ICR), O.6b (ICRSSM), and O.6c (ICRSSL)
Implementation: Column A from Exhibits O.5a (ICR), O.5d (ICRSSM), and O.5g (ICRSSL)
Initial Monitoring: Sum of columns B, C, and D from Exhibits O.5a (ICR), O.5d (ICRSSM), and O.5g (ICRSSL)
Second Monitoring: Sum of columns E, F, and G from Exhibits O.5a (ICR), O.5d (ICRSSM), and O.5g (ICRSSL)
Benchmarking: Column H from Exhibits O.5a (ICR), O.5d (ICRSSM), and O.5g (ICRSSL)
Tech Reporting: Column I from Exhibits O.5a (ICR), O.5d (ICRSSM), and O.5g (ICRSSL)
Uncovered Reservoirs: Column N from Exhibits O.6a (ICR), O.6b (ICRSSM), and O.6c (ICRSSL)
State:
Implementation: Column A from Exhibits O.4a (ICR), O.4d (ICRSSM), and O.4g (ICRSSL)
Initial Monitoring: Sum of columns B and C from Exhibits O.4a (ICR), O.4d (ICRSSM), and O.4g (ICRSSL)
Second Monitoring: Sum of columns D and E from Exhibits O.4a (ICR), O.4d (ICRSSM), and O.4g (ICRSSL)
Benchmarking: Column F from Exhibits O.4a (ICR), O.4d (ICRSSM), and O.4g (ICRSSL)
Tech Reporting: Column G from Exhibits O.4a (ICR), O.4d (ICRSSM), and O.4g (ICRSSL)
Economic Analysis for the LT2ESWTR
6-24
December 2005
-------
Exhibit 6.16: Annualized Total Costs-Discounted at 3 and 7 Percent, Preferred
Alternative ($Millions, 2003$)
National
System Total
State Total
National
System Total
State Total
ICR
Mean
Confidence Bounds
5th %ile |95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile 1 95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile |95th %ile
3%
$ 133.42
$ 132.27
$ 1.15
$ 111.05
$ 109.91
$ 1.14
$ 160.00
$ 158.83
$ 1.17
$ 92.88
$ 91.78
$ 1.09
$ 72.11
$ 71.03
$ 1.08
$ 112.17
$ 111.07
$ 1.10
$ 105.90
$ 104.79
$ 1.11
$ 86.30
$ 85.20
$ 1.10
$ 125.74
$ 124.62
$ 1.12
7%
$ 150.48
$ 149.07
$ 1.41
$ 125.12
$ 123.72
$ 1.39
$ 180.61
$ 179.19
$ 1.42
$ 106.77
$ 105.42
$ 1.35
$ 83.21
$ 81.87
$ 1.34
$ 128.83
$ 127.47
$ 1.36
$ 120.93
$ 119.56
$ 1.37
$ 98.58
$ 97.22
$ 1.36
$ 143.61
$ 142.23
$ 1.38
Notes: Detail may not add to totals due to independent rounding.
Sources: All data from Appendix O, and from Alternative A3.
System (at 3 percent):
Mean: Sum of column J from Exhibits O.17a (ICR), O.17d (ICRSSM), and O.17g (ICRSSL) and column T from
Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
5th percentile: Sum of column J from Exhibits O.17b (ICR), O.17e (ICRSSM), and O.17h (ICRSSL) and column S
from Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
95th percentile: Sum of column J from Exhibits O.17c(ICR), O.17f (ICRSSM), and O.17I (ICRSSL) and column U
from Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
System (at 7 percent):
Mean: Sum of column J from Exhibits O.20a (ICR), O.20d (ICRSSM), and O.20g (ICRSSL) and column T from
Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
5th percentile: Sum of column J from Exhibits O.20b (ICR), O.20e (ICRSSM), and O.20h (ICRSSL) and column S
from Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
95th percentile: Sum of column J from Exhibits O.20c (ICR), O.20f (ICRSSM), and O.20I (ICRSSL) and column U
from Exhibits O.18a (ICR), O.18b (ICRSSM), and O.18c (ICRSSL)
State (at 3 percent):
Mean: Column H from Exhibits O.16a (ICR), O.16d (ICRSSM), and O.16g (ICRSSL)
5th percentile: Column H from Exhibits O.16b (ICR), O.16e (ICRSSM), and O.16h (ICRSSL)
95th percentile: Column H from Exhibits O.16c (ICR), O.16f (ICRSSM), and O.16I (ICRSSL)
State (at 7 percent):
Mean: Column H from Exhibits O.19a (ICR), O.19d (ICRSSM), and O.19g (ICRSSL)
5th percentile: Column H from Exhibits O.19b (ICR), O.19e (ICRSSM), and O.19h (ICRSSL)
95th percentile: Column H from Exhibits O.19c (ICR), O.19f (ICRSSM), and O.19I (ICRSSL)
Economic Analysis for the LT2ESWTR
6-25
December 2005
-------
Exhibit 6.17a: Annualized Treatment Costs by System Size, Preferred Alternative,
3 Percent Discount Rate ($Millions, 2003$)
System Size
(Population
Served)
Capital - Present Value
Mean
A
5th %ile
B
95th %ile
c
O & M -Annualized
Mean
D
5th %ile
E
95th %ile
F
Total - Annualized
Mean
G
5th %ile
H
95th %ile
i
ICR
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
$ 5.0
$ 7.0
$ 4.9
$ 19.1
$ 52.4
$ 221.4
$ 145.8
$ 419.3
$ 551.6
$ 1,426.5
$ 3.9
$ 5.6
$ 3.9
$ 14.9
$ 40.7
$ 176.5
$ 114.7
$ 331.4
$ 436.9
$ 1,128.4
$ 6.3
$ 8.8
$ 6.1
$ 23.8
$ 65.7
$ 279.3
$ 185.5
$ 530.2
$ 675.2
$1,780.9
$ 0.3
$ 0.5
$ 0.3
$ 1.2
$ 2.3
$ 7.3
$ 3.6
$ 11.3
$ 6.8
$ 33.7
$ 0.3
$ 0.4
$ 0.3
$ 1.0
$ 1.9
$ 6.3
$ 3.0
$ 9.6
$ 5.8
$ 28.7
$ 0.4
$ 0.6
$ 0.4
$ 1.5
$ 2.7
$ 8.6
$ 4.2
$ 13.3
$ 8.0
$ 39.6
$ 0.6
$ 0.9
$ 0.6
$ 2.3
$ 5.3
$ 20.0
$ 11.9
$ 35.4
$ 38.5
$ 115.6
$ 0.5
$ 0.7
$ 0.5
$ 1.9
$ 4.3
$ 16.4
$ 9.6
$ 28.7
$ 30.9
$ 93.5
$ 0.8
$ 1.1
$ 0.7
$ 2.8
$ 6.5
$ 24.7
$ 14.8
$ 43.7
$ 46.7
$ 141.9
ICRSSL
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
$ 3.0
$ 4.0
$ 2.8
$ 10.4
$ 27.8
$ 134.5
$ 88.8
$ 250.4
$ 476.8
Total $ 998.5
$ 2.0
$ 2.7
$ 1.9
$ 7.2
$ 19.1
$ 92.3
$ 60.3
$ 171.3
$ 366.4
$ 723.2
$ 3.9
$ 5.2
$ 3.6
$ 13.4
$ 36.1
$ 174.4
$ 116.1
$ 325.7
$ 584.7
$1,263.1
$ 0.2
$ 0.2
$ 0.2
$ 0.7
$ 1.3
$ 4.3
$ 2.0
$ 6.1
$ 3.8
$ 18.8
$ 0.1
$ 0.2
$ 0.1
$ 0.5
$ 1.0
$ 3.2
$ 1.5
$ 4.6
$ 3.0
$ 14.2
$ 0.2
$ 0.3
$ 0.2
$ 0.8
$ 1.6
$ 5.2
$ 2.4
$ 7.3
$ 4.6
$ 22.6
$ 0.3
$ 0.5
$ 0.3
$ 1.3
$ 2.9
$ 12.0
$ 7.1
$ 20.5
$ 31.2
$ 76.1
$ 0.2
$ 0.3
$ 0.2
$ 0.9
$ 2.1
$ 8.5
$ 5.0
$ 14.4
$ 24.0
$ 55.7
$ 0.4
$ 0.6
$ 0.4
$ 1.6
$ 3.7
$ 15.2
$ 9.1
$ 26.0
$ 38.1
$ 95.2
ICRSSM
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
$ 3.7
$ 4.9
$ 3.4
$ 12.9
$ 34.6
$ 164.7
$ 108.5
$ 306.9
$ 501.5
Total $ 1,141.0
$ 2.7
$ 3.7
$ 2.6
$ 9.7
$ 26.0
$ 125.1
$ 81.4
$ 231.7
$ 392.6
$ 875.4
$ 4.6
$ 6.2
$ 4.2
$ 16.1
$ 43.7
$ 205.9
$ 136.8
$ 385.5
$ 610.8
$1,413.9
$ 0.2 $ 0.2
$ 0.3 $ 0.3
$ 0.2 $ 0.2
$ 0.8 $ 0.7
$ 1.6 $ 1.3
$ 5.3 $ 4.4
$ 2.5 $ 2.0
$ 7.6 $ 6.3
$ 4.7 $ 3.9
$ 23.2
$ 19.1
$ 0.3
$ 0.4
$ 0.2
$ 1.0
$ 1.9
$ 6.2
$ 2.9
$ 8.9
$ 5.4
$ 27.2
$ 0.4
$ 0.6
$ 0.4
$ 1.6
$ 3.6
$ 14.7
$ 8.7
$ 25.2
$ 33.5
$ 88.8
$ 0.3
$ 0.5
$ 0.3
$ 1.2
$ 2.8
$ 11.5
$ 6.7
$ 19.6
$ 26.4
$ 69.4
$ 0.5
$ 0.7
$ 0.5
$ 1.9
$ 4.5
$ 18.0
$ 10.7
$ 31.0
$ 40.5
$ 108.4
Note: Detail may not add to totals due to independent rounding.
Sources: Appendix O.
[A]-[C] Exhibit O.12, Columns G-l.
[D]-[F] Exhibit O.18, Columns J-L.
[G]-[l] Exhibit O.18, Sum of columns G and J, H and K, and I and
Economic Analysis for the LT2ESWTR
6-26
December 2005
-------
Exhibit 6.17b: Annualized Treatment Costs by System Size, Preferred Alternative,
7 Percent Discount Rate ($Millions, 2003$)
System Size
(population
served)
Capital - Present Value
Mean
A
5th %ile
B
95th %ile
c
O& M -Annualized
Mean
D
5th %ile
E
95th %ile
F
Total -Annualized
Mean
G
5th %ile
H
95th %ile
I
ICR
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
$ 3.7
$ 5.2
$ 3.6
$ 14.1
$ 38.7
$ 172.8
$ 118.1
$ 345.9
$ 455.0
$1,157.1
$ 2.9
$ 4.1
$ 2.9
$ 11.0
$ 30.0
$ 137.7
$ 93.0
$ 273.3
$ 360.4
$ 915.3
$ 4.6
$ 6.5
$ 4.5
$ 17.6
$ 48.6
$ 218.0
$ 150.3
$ 437.3
$ 556.9
$1,444.3
$ 0.3
$ 0.4
$ 0.3
$ 1.0
$ 1.9
$ 6.3
$ 3.1
$ 10.1
$ 6.1
$ 29.4
$ 0.2
$ 0.3
$ 0.2
$ 0.9
$ 1.6
$ 5.4
$ 2.7
$ 8.6
$ 5.1
$ 25.0
$ 0.3
$ 0.5
$ 0.3
$ 1.2
$ 2.2
$ 7.4
$ 3.7
$ 11.8
$ 7.1
$ 34.6
$ 0.6
$ 0.8
$ 0.6
$ 2.2
$ 5.2
$ 21.1
$ 13.3
$ 39.8
$ 45.1
$ 128.7
$ 0.5
$ 0.7
$ 0.5
$ 1.8
$ 4.2
$ 17.2
$ 10.7
$ 32.0
$ 36.1
$ 103.6
$ 0.7
$ 1.0
$ 0.7
$ 2.7
$ 6.4
$ 26.1
$ 16.6
$ 49.4
$ 54.9
$ 158.5
ICRSSL
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
$ 2.2
$ 2.9
$ 2.0
$ 7.7
$ 20.5
$ 105.0
$ 72.0
$ 206.6
$ 393.3
$ 812.2
$ 1.5
$ 2.0
$ 1.4
$ 5.3
$ 14.1
$ 72.0
$ 48.9
$ 141.3
$ 302.2
$ 588.7
$ 2.9
$ 3.8
$ 2.6
$ 9.9
$ 26.7
$ 136.1
$ 94.1
$ 268.7
$ 482.3
$1,027.0
$ 0.1
$ 0.2
$ 0.1
$ 0.5
$ 1.1
$ 3.7
$ 1.7
$ 5.4
$ 3.4
$ 16.4
$ 0.1
$ 0.1
$ 0.1
$ 0.4
$ 0.8
$ 2.8
$ 1.3
$ 4.1
$ 2.6
$ 12.4
$ 0.2
$ 0.2
$ 0.2
$ 0.7
$ 1.4
$ 4.5
$ 2.1
$ 6.5
$ 4.1
$ 19.7
$ 0.3
$ 0.5
$ 0.3
$ 1.2
$ 2.9
$ 12.7
$ 7.9
$ 23.1
$ 37.2
$ 86.1
$ 0.2
$ 0.3
$ 0.2
$ 0.9
$ 2.0
$ 8.9
$ 5.5
$ 16.2
$ 28.6
$ 62.9
$ 0.4
$ 0.6
$ 0.4
$ 1.5
$ 3.6
$ 16.1
$ 10.2
$ 29.6
$ 45.4
$ 107.9
ICRSSM
<100
100-499
500-999
1,000-3,299
3,300-9,999
10,000-49,999
50,000-99,999
100,000-999,999
1,000,000+
Total
$ 2.7
$ 3.6
$ 2.5
$ 9.5
$ 25.6
$ 128.5
$ 87.9
$ 253.2
$ 413.6
$ 927.1
$ 2.0
$ 2.7
$ 1.9
$ 7.2
$ 19.2
$ 97.6
$ 65.9
$ 191.1
$ 323.8
$ 711.5
$ 3.4
$ 4.6
$ 3.1
$ 11.9
$ 32.3
$ 160.7
$ 110.9
$ 318.0
$ 503.8
$1,148.7
$ 0.2 $ 0.2
$ 0.3 $ 0.2
$ 0.2 $ 0.1
$ 0.7 $ 0.6
$ 1.4 $ 1.1
$ 4.5 $ 3.7
$ 2.2 $ 1.8
$ 6.8 $ 5.6
$ 4.2 $ 3.4
$ 20.3
$ 16.7
$ 0.2
$ 0.3
$ 0.2
$ 0.8
$ 1.6
$ 5.3
$ 2.5
$ 7.9
$ 4.8
$ 23.7
$ 0.4
$ 0.6
$ 0.4
$ 1.5
$ 3.6
$ 15.6
$ 9.7
$ 28.5
$ 39.7
$ 99.8
$ 0.3
$ 0.4
$ 0.3
$ 1.2
$ 2.8
$ 12.1
$ 7.4
$ 22.0
$ 31.2
$ 77.7
$ 0.5
$ 0.7
$ 0.5
$ 1.8
$ 4.4
$ 19.1
$ 12.1
$ 35.2
$ 48.1
$ 122.3
Notes: Detail may not add to totals due to independent rounding.
Sources: Appendix O.
[A]-[C] Exhibit O.15, Columns G-l.
[D]-[F] Exhibit O.21; Columns J-L.
[G]-[l] Exhibit O.21; Sum of columns G and J, H and K, and I and L.
6.10 Household Costs
EPA assumes that systems will pass some or all of the costs of a new regulation on to their
customers in the form of rate increases. These rate increases will include interest costs and patent costs
that are not included in cost estimates shown in other sections. Household costs, which are in units of $
per household per year, are estimated in this chapter to provide a measure of the increase in water bills
that may result from the LT2ESWTR. Exhibit 6.18 presents the mean expected increases in yearly
household costs by system size, system type, and occurrence data set, for those systems subject to the
Economic Analysis for the LT2ESWTR
6-27
December 2005
-------
rule. (Appendix J, Exhibit J.4 presents household cost estimates for those systems predicted to make
treatment changes.)
These costs incorporate the expenses of rule implementation (e.g., reading and understanding the
rule), initial and future monitoring for bin classification, covering or treating effluent from uncovered
finished water reservoirs, treatment changes, benchmarking, and compliance reporting. A detailed
description of the derivation of per-household costs is in Appendix J. Per-household costs for uncovered
finished water reservoirs are determined by taking the costs for fixing the reservoirs from section 6.8 and
assigning them to systems as described in section 4.6.
To annualize capital costs for the purposes of determining the costs to households, EPA uses
different discount rates for private and public systems and for systems of different sizes. The discount
rate differences between systems represent the different borrowing sources each type of system has
available to it, differences in risk, and expectations regarding inflation. The rates vary from 5.20 to 6.27
percent depending on system size and ownership, and are summarized in Appendix J, Exhibit J.I. Per-
household costs also include costs for royalty payments on the use of UV light (described below).
For each system size category, the unit costs for treatment in dollars per thousand gallons is then
multiplied by the annual per-household usage rate to obtain their contribution to per-household costs.
Although rule implementation and monitoring represent relatively small, one-time costs, they have been
included in the analysis to provide a complete distribution of the potential per-household cost increase.
Calgon Carbon Corporation holds a patent (No. 6,129,893) for "A method for prevention of
Cryptosporidium oocysts and similar organisms in water by irradiating the water with ultraviolet light in a
range of 200 to 300 nm in concentrations of about 10 mJ/cm2 to about 175 mJ/cm2." This patent applies
to systems using medium-pressure mercury vapor lamps to inactivate Cryptosporidium. EPA also
understands that Calgon has applied for a continuation in part to extend coverage of the patent to low-
pressure mercury vapor lamps and to lower UV concentrations. Calgon is charging a license fee of
$0.015/1,000 gallons treated to water producers using UV under conditions covered by the patent. This
cost was added to the unit cost of UV in the per-household cost calculations. It was not used in the
national cost estimates because it represents a transfer payment involving no net use of resources.
For purchased systems that are linked to larger, nonpurchased systems, the per-household costs
are calculated using the unit costs of the larger system; however, they are reported within the size
category distributions for the purchased system. Household costs for these purchased systems are based
on the per-household usage rates appropriate for the retail system and not for the system selling that
wholesales water. This reflects the fact that although purchased systems will not face increased costs
from adding their own treatment, whatever costs the wholesale utility incurs would likely be passed on as
higher water costs.
Economic Analysis for the LT2ESWTR 6-28 December 2005
-------
Exhibit 6.18: Summary of Annual Per-Household Cost1 Increases, Preferred
Alternative ($/Year)
System Type/Size
Households
Mean
Median
90th
Percentile
95th
Percentile
Percent of
Systems with
Household
Cost Increase
<$12
Percent of
Systems with
Household
Cost Increase <
$120
ICR
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$2.59
$4.14
$13.09
$0.21
$0.56
$3.86
$6.43
$9.97
$28.66
$9.97
$14.79
$53.60
96.49%
91.19%
63.20%
99.99%
99.88%
98.87%
ICRSSL
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$1.67
$2.49
$8.58
$0.09
$0.36
$2.91
$6.37
$6.60
$17.44
$6.42
$9.37
$29.01
97.96%
96.46%
72.61%
100.00%
99.94%
99.50%
ICRSSM
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$1.97
$3.00
$10.10
$0.09
$0.49
$2.90
$6.37
$7.02
$26.24
$6.85
$11.39
$35.97
97.47%
95.19%
68.73%
99.99%
99.93%
99.31%
ICR -High
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$2.84
$4.58
$7.21
$0.21
$0.61
$2.91
$6.43
$11.50
$16.81
$9.97
$15.30
$26.25
96.09%
90.22%
75.79%
99.99%
99.86%
99.80%
ICRSSL - Low
All CWS
CWS <1 0,000
CWS < 500
68,857,992
5,587,602
158,900
$1.42
$2.06
$14.42
$0.03
$0.23
$4.79
$5.65
$6.58
$30.00
$6.42
$7.47
$54.42
98.37%
97.21%
62.07%
100.00%
99.96%
98.58%
Note:
1Annualized at discount rates varied by system size and ownership (see Appendix J, Exhibit J.2).
Households served by systems subject to the LT2ESWTR.
Source: Appendix J, Exhibit J.3.
EPA estimates that all households served by surface and GWUDI sources will face some increase
in costs due to implementation of the LT2ESWTR (except for those few that have already installed 5.5
log of treatment for Cryptosporidium; see Chapter 4 for a summary of households served by systems
subject to various LT2ESWTR provisions). Of all the households subject to the rule, 24 to 35 percent are
projected to incur costs for adding treatment, depending on the Cryptosporidium occurrence data set used.
Approximately 95 percent of the households potentially subject to the rule are connected to systems
serving at least 10,000 people; these systems will experience the smallest increases in costs, due to
economies of scale. Over 90 percent of all households will face an annual cost increase of less than 7
dollars.
Economic Analysis for the LT2ESWTR
6-29
December 2005
-------
6.11 Summary of Uncertainties and Sensitivity Analyses
As described throughout this chapter, uncertainty and variability are inherent in developing
national cost estimates. EPA addresses these issues by using probability distributions of key variables,
conducting the analysis for multiple data sets, and analyzing specific variables of the model to determine
how sensitive is their effect on national estimates. This section first discusses the uncertainties in the
national cost estimates and then provides sensitivity analyses to demonstrate how the national estimates
can fluctuate as certain parameters are changed.
Exhibit 6.19 below presents a summary of the uncertainties, references to the section or appendix
where the issue is discussed, and estimates the effects that each may have on national costs.
Exhibit 6.19: Summary of Uncertainties Affecting LT2ESWTR Cost Estimates
Source of Uncertainty
Occurrence data used to predict plant
bin assignments
All systems are charged the same
laboratory fee for Cryptosporidium
monitoring
Unit costs developed for typical system
conditions, where site-specific factors
often drive costs and treatment
selections1
Single flow rate used to evaluate unit
costs within each of 9 size categories
Potentially lower-cost treatments or
toolbox options not considered
Economies of scale not considered for
combined treatment technologies
Inability to link all purchased systems
with their suppliers
Number of systems achieving credit for
technologies in place
Section with Full
Discussion of
Uncertainty
4.5.3
Appendix O
6.1.1
6.4.1
6.4.1
6.4.1
6.4.1
4.3.2
4.5.1,
Appendix A
Effect on Cost Estimates
Under-
estimate
Over-
estimate
X
X
X
X
Under or
Over
Estimate
X
X
X
X
1Source water quality and plant building or existing infrastructure may inhibit the use of a technology or cause
decreased or increased capital and O&M costs.
Of the uncertainties in Exhibit 6.19, those affecting (1) the number of plants predicted to add
treatment, and (2) the technology selection will have the greatest effect on the LT2ESWTR national cost
estimates. Section 6.11.1 discusses the effect of the occurrence estimates; and Section 6.11.2 provides
sensitivity analyses on source water quality.
The uncertainty surrounding unit costs is significant. EPA addresses this uncertainty by
assigning a range of unit costs—±30 percent for capital costs and ±15 percent for O&M costs. Although
Economic Analysis for the LT2ESWTR
6-30
December 2005
-------
actual costs incurred by a plant implementing a given technology may be substantially higher or lower
than implied by this uncertainty, this analysis is concerned with national estimates. The unit costs
represent mean estimates for a given size and technology, which is appropriate to use for estimating
national costs, as one would expect the variability imposed by site-specific factors to have similar effects
in both directions of lower or higher costs. For example, some metropolitan systems may incur
extraordinarily high costs due to land cost, while some others may incur extraordinarily low costs because
they have the space available to install the required infrastructure in their existing facility.
6.11.1 Cryptosporidium Occurrence Data Sets
The source water occurrence of Cryptosporidium has much uncertainty due to the nonuniform
distribution of the oocysts in a body of water, limitations in analytical methods, and atypical weather or
source water contamination events. EPA recognized these issues with the ICR and ICRSS data and
modeled likely distributions of national Cryptosporidium occurrence. However, the model results still
present a significant amount of uncertainty. To show how the national estimate may vary depending on
the source water Cryptosporidium concentrations, all costs were estimated for mean values and 90 percent
confidence bounds of each occurrence data set, resulting in nine occurrence estimates.
The results presented throughout this chapter show mean, 5th percentile, and 95th percentile values
for each data set. The cost model also incorporates uncertainty in capital and O&M unit costs by
estimating a mean, 5th percentile, and 95th percentile value for each occurrence distribution. The result is
27 estimates of treatment costs (implementation, monitoring, and reporting costs do not incorporate
uncertainty of the unit costs). The confidence bounds of treatment and total national costs presented in
Exhibits 6.15, 6.16, and 6.17 are the 5th percentile cost of the 5th percentile occurrence distribution and the
95th percentile cost of the 95th percentile occurrence distribution, or the low of the low and high of the
high, for each of the ICR, ICRSSL, and ICRSSM data sets (see Exhibit 6.20).
The 95th percentile of the ICR data set represents the highest occurrence estimate and the 5th
percentile of the ICRSSL data set represents the lowest occurrence estimate. From the national total
annualized costs presented in Exhibit 6.16, the ICR 95th percentile is $159 million and the ICRSSL 5th
percentile is $71 million (discounted at 3 percent). The sensitivity of the cost estimate to the uncertainty
of Cryptosporidium occurrence is shown by the comparison of cost estimates between ICR and ICRSSL
data sets. The mean national cost estimate between these data sets varies by roughly a factor of two.
Economic Analysis for the LT2ESWTR 6-31 December 2005
-------
Exhibit 6.20: Cost Model Estimates by Occurrence Distributions
and Unit Cost Uncertainty
ICR
Source Water Occurrence
ICRSSL
ICRSSM
5th Mean 95th
Percentile Percentile
5th Mean 95th
Percentile Percentile
5th
Percentile
Mean 95th
Percentile
o
-30%/ \ +30%
0
-15%,
Uncertainty
Capital Unit Cost
O&M Unit Cost
H15%
ICR
Cost Estimates
ICRSSL
ICRSSM
' U 1
Mean 95th 5th | 95th 5th Mean
15th] | Mean | [95th
iTr i "[> iTr
i_Mean 95th 5th I 95th 5th Mean I
5th
* / X
vvv Results presented in the EA as mean and 90 percent ,,' *' *,*'
confidence bounds.
Economic Analysis for the LT2ESWTR
6-32
December 2005
-------
6.11.2 Sensitivity Analysis of Influent Bromide Levels on Technology Selection for Filtered Plants
Bromide in the treatment plant influent can limit ozone use due to the byproduct formation of
bromate. In the LT2ESWTR least cost-modeling approach, ozone is selected after UV disinfection and a
few other technologies, depending on system size and log treatment credit required. EPA conducted a
sensitivity analysis to evaluate how the technology selection, and thus national cost estimate, would
change if more plants had source water bromide concentrations that restricted the use of ozone.
The ICR database includes plant influent bromide concentrations for July 1997 through
December 1998. Bromide levels vary from year to year and are highest during drought periods. There is
concern that ICR bromide data were not collected during a drought period and, thus, do not accurately
reflect the maximum influent levels that plants would use when designing their ozonation systems. For
the standard conditions used in the main cost analysis, maximum use percentages for ozone reflect the
SWAT analysis using influent bromide equivalent to the values reported in the ICR (USEPA 2003b).
EPA conducted another SWAT analysis in which the influent bromide concentrations for each plant were
increased by 50 parts per billion (ppb). Those results provided a maximum use percent of ozone for the
LT2ESWTR decision tree, under high influent bromide levels. Exhibit 6.21 compares the number of
plants selecting UV and ozone and the filtered plant treatment cost estimates for each technology
selection (standard and influent bromide increased by 50 ppb). Technology selection constraints on
ozone use have little impact on annual costs. Appendix G presents technology selection forecasts that
reflect the three occurrence data sets and an increased source water bromide concentration for all
regulatory alternatives.
Exhibit 6.21: Sensitivity of Technology Selection to Influent Bromide
Concentration for Filtered Plants
Number of Plants
Converting to Ozone
Standard Condition
(Influent Bromide as reported in ICR)
ICR
A
54
ICRSSL
B
38
ICRSSM
C
45
Influent Bromide Increased by 50 ppb
ICR
D
41
ICRSSL
E
32
ICRSSM
F
37
Total Annual Treament Cost ($ Millions, 2003$)
3 Percent
7 Percent
$90
$97
$50
$55
$63
$69
$103
$112
$56
$62
$71
$77
Sources:
"Number of Plants Converting to Ozone" Appendix G, Exhibits G.37-48; Row - Total Plants; Columns H-J.
"Treatment Cost (Annual)" Appendix O, Exhibits O.18 and O.21; Column B and E, Total rows for Rule Alternative A3
and Rule Alternative A3 UV90-10B.
6.12 Unquantified Costs
EPA has quantified all of the major costs for this rule and has provided uncertainty analyses to
bound the over- or underestimates in the mean cost values. Some cost effects are unquantifiable,
however, because of a lack of information. One such cost effect on systems that must comply with
several rules at the same time. This analysis took into consideration compliance with the Stage 2 DBPR,
the LT1ESWTR, and the IESWTR. It did not, however take into account other rules that are or will be
promulgated before this rule. These include the Arsenic Rule, the Ground Water Rule, and the Filter
Backwash Recycling Rule. Although most of these will not affect surface water sources, they may limit
Economic Analysis for the LT2ESWTR
6-33
December 2005
-------
the use of alternative sources. The rules affecting ground water could affect the GWUDI systems in the
LT2ESWTR. There could be lower-costs for some systems if technologies they install for this rule also
achieve reductions in other contaminants. There could be unquantifiable savings in monitoring and
implementation costs for complying with several rules at once. There are also unquantifiable savings
associated with some of the treatment and management strategies listed in section 6.5.1.1 that were not
included in this analysis, but that may be less expensive than the treatment technologies which were
evaluated.
Another cost not quantified is that of systems merging to comply with this rule. Although
mergers could make compliance easier or treatment costs lower (due to economies of scale) for many
small systems, it is difficult to tell how many mergers would result from this rule and how many would
occur because of other factors. Costs would also be difficult to quantify. There could be savings because
of economies of scale, but there could also be increases because of additional capital costs to connect the
systems.
Other toolbox options that were not quantified included source water intake management and
performance studies. These measures may prove cheaper than the technologies considered in the
analyses, so their inclusion would result in lower-costs. The cost savings are difficult to quantify,
however, because the effectiveness and applicability of these options is unknown.
6.13 Comparison of Regulatory Alternatives
Exhibits 6.22a-b provide a summary of the annualized present value of filtered plant costs for
each regulatory alternative, for each data set, using 3 and 7 percent discount rates, based on a 25-year
period of analysis. They do not include costs to unfiltered plants and uncovered finished water reservoirs
because regulatory requirements to these entities do not change among regulatory alternatives.
Exhibit 6.22a: Comparison by Regulatory Alternative of Total Costs, Annualized
at 3 Percent for Filtered Plants ($Millions, 2003$)
System Size
(Population
Served)
ICR
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile
95th %ile
Preferred Alternative
<10K
>10k
Total
$ 9.15
$ 80.40
$ 89.56
$ 7.37
$ 65.00
$ 72.37
$ 11.24
$ 99.65
$ 110.88
$ 4.74
$ 45.30
$ 50.04
$ 3.28
$ 31.33
$ 34.61
$ 6.06
$ 58.10
$ 64.16
$ 6.02
$ 56.66
$ 62.68
$ 4.65
$ 43.66
$ 48.31
$ 7.41
$ 69.97
$ 77.39
Alternative 1
<10K
>10k
Total
$ 38.47
$ 323.67
$ 362.14
$ 33.40
$ 292.12
$ 325.51
$ 43.51
$ 355.30
$ 398.81
$ 38.47
$ 323.67
$ 362.14
$ 33.40
$ 292.12
$ 325.51
$ 43.51
$ 355.30
$ 398.81
$ 38.47
$ 323.67
$ 362.14
$ 33.40
$ 292.12
$ 325.51
$ 43.51
$ 355.30
$ 398.81
Alternative 2
<10K
>10k
Total
$ 13.84
$ 101.19
$ 115.03
$ 1 1 .59
$ 85.02
$ 96.61
$ 17.26
$ 126.04
$ 143.30
$ 9.09
$ 65.86
$ 74.96
$ 6.76
$ 47.90
$ 54.66
$ 11.41
$ 83.33
$ 94.74
$ 10.62
$ 78.24
$ 88.86
$ 8.50
$ 62.15
$ 70.65
$ 12.93
$ 95.34
$ 108.26
Alternative 4
<10K
>10k
Total
$ 4.53
$ 32.78
$ 37.31
$ 3.63
$ 25.61
$ 29.24
$ 5.66
$ 41.17
$ 46.83
$ 2.12
$ 13.00
$ 15.11
$ 1.44
$ 8.76
$ 10.21
$ 2.87
$ 17.94
$ 20.81
$ 2.94
$ 18.99
$ 21.93
$ 2.26
$ 14.26
$ 16.51
$ 3.68
$ 24.10
$ 27.77
Sources: Appendix O, Exhibits O.18, Column B and E.
Economic Analysis for the LT2ESWTR
6-34
December 2005
-------
Exhibit 6.22b: Comparison by Regulatory Alternative of Total Costs, Annualized
at 7 Percent for Filtered Plants ($Millions, 2003$)
System Size
(Population
Served)
ICR
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSL
Mean
Confidence Bounds
5th %ile
95th %ile
ICRSSM
Mean
Confidence Bounds
5th %ile
95th %ile
Preferred Alternative
<10K
>10k
Total
$ 8.87
$ 88.62
$ 97.48
$ 7.10
$ 71 .28
$ 78.39
$ 10.94
$ 110.36
$ 121.29
$ 4.58
$ 50.27
$ 54.85
$ 3.16
$ 34.56
$ 37.71
$ 5.89
$ 64.77
$ 70.66
$ 5.83
$ 62.77
$ 68.60
$ 4.47
$ 48.10
$ 52.57
$ 7.21
$ 77.87
$ 85.08
Alternative 1
<10K
>10k
Total
$ 37.35
$ 351.05
$ 388.41
$ 32.18
$ 315.08
$ 347.26
$ 42.51
$ 387.08
$ 429.59
$ 37.35
$ 351.05
$ 388.41
$ 32.18
$ 315.08
$ 347.26
$ 42.51
$ 387.08
$ 429.59
$ 37.35
$ 351.05
$ 388.41
$ 32.18
$ 315.08
$ 347.26
$ 42.51
$ 387.08
$ 429.59
Alternative 2
<10K
>10k
Total
$ 13.43
$ 110.64
$ 124.07
$ 11.19
$ 92.44
$ 103.63
$ 16.83
$ 138.46
$ 155.28
$ 8.81
$ 72.20
$ 81.01
$ 6.51
$ 52.18
$ 58.69
$ 11.10
$ 91.81
$ 102.91
$ 10.29
$ 85.73
$ 96.03
$ 8.19
$ 67.68
$ 75.87
$ 12.59
$ 104.96
$ 117.55
Alternative 4
<10K
>10k
Total
$ 4.37
$ 36.15
$ 40.52
$ 3.49
$ 28.02
$ 31.51
$ 5.49
$ 45.64
$ 51.13
$ 2.04
$ 14.26
$ 16.30
$ 1.38
$ 9.54
$ 10.93
$ 2.78
$ 19.79
$ 22.58
$ 2.84
$ 20.87
$ 23.71
$ 2.17
$ 15.54
$ 17.71
$ 3.57
$ 26.63
$ 30.20
Sources: Appendix O, Exhibits O.21, Column B and E.
Economic Analysis for the LT2ESWTR
6-35
December 2005
-------
7. Economic Impact Analysis
7.1 Introduction
As part of the rulemaking process, EPA is required to address the direct and indirect burdens that
the LT2ESWTR may place on certain types of governments, businesses, and populations. This chapter
presents the analyses performed by EPA in accordance with the following 12 Federal mandates:
1. The Regulatory Flexibility Act (RFA) of 1980, as amended by the Small Business Regulatory
Enforcement Fairness Act (SBREFA) of 1996.
2. Analysis of small system affordability to determine variance technologies in accordance with
Section 1415(e)(l) of the 1996 Safe Drinking Water Act (SDWA) Amendments.
3. Feasible technologies available to all systems as required by Section 1412(b)(4)(E) of the
1996 SDWA Amendments.
4. Technical, financial, and managerial capacity assessment as required by Section 1420(d)(3)
of the 1996 Amendments to SDWA.
5. Paperwork Reduction Act (a separate Information Collection Request document contains the
complete analysis).
6. Unfunded Mandates Reform Act (UMRA) of 1995.
7. Executive Order 13175 (Consultation and Coordination with Indian Tribal Governments).
8. Impacts on sensitive subpopulations as required by Section 1412(b)(3)(c)(i) of the 1996
SDWA Amendments.
9. Executive Order 13045 (Protection of Children from Environmental Health Risks and Safety
Risks).
10. Executive Order 12898 (Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations).
11. Executive Order 13132 (Federalism).
12. Executive Order 13211 (Actions Concerning Regulations That Significantly Affect Energy
Supply, Distribution, or Use).
Many of the requirements and executive orders listed above call for an explanation of why the
rule is necessary, the statutory authority for the rule, and the primary objectives that the rule is intended to
achieve (refer to Chapter 2 for more information regarding the objectives of the rule). More specifically,
they are designed to assess the financial and health effects of the rule on sensitive, low-income, and Tribal
populations as well as on small systems. The chapter also examines how much additional capacity
systems will need to meet LT2ESWTR requirements and whether there are existing, feasible technologies
and treatment techniques available to meet rule requirements.
7.2 Regulatory Flexibility Act and Small Business Regulatory Enforcement Fairness
Act
The RFA generally requires an agency to prepare a regulatory flexibility analysis for any rule
subject to notice and comment rulemaking requirements under the Administrative Procedure Act or other
statute, unless the Agency certifies that the rule will not have a significant economic impact on a
substantial number of small entities (5 U.S.C. 603(a)). Small entities include small businesses, small
organizations, and small governmental jurisdictions.
Economic Analysis for the LT2ESWTR 7-1 December 2005
-------
The RFA provides default definitions for each type of small entity, discussed in more detail in
Appendix H. 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 that is independently owned an
operated and is not dominant in the field." However, the RFA also authorizes an agency to use
alternative definitions for each category of small entity that "are appropriate to the activities of the
Agency after proposing the alternative definition(s) in the Federal Register and taking comment" (5
U.S.C. §601(3)-(5)). In addition, to establish an alternative small business definition, agencies must
consult with SBA's Chief Counsel for Advocacy. In assessing the impacts of the LT2ESWTR on small
entities, EPA considered small entities to be PWSs serving 10,000 or fewer persons, which is the cut-off
level specified by Congress in the 1996 Amendments to SDWA for small system flexibility provisions.
EPA conducted a screening analysis to determine if the LT2ESWTR would have a significant
economic impact on a substantial number of small entities (see Appendix H). In this analysis, EPA
evaluated the potential economic impact of the rule on small entities by comparing annualized compliance
costs as a percentage of annual revenues1 for different small-entity classifications. Chapter 4 of this EA
provides data on small entities potentially subject to the LT2ESWTR, and Chapter 6 discusses actions
systems would need to take to comply with the rule and their associated costs. Using information from
these two chapters, along with additional information from the Safe Drinking Water Information System
(SDWIS), the Community Water System Survey (CWSS), and the U.S. Census, EPA conducted a
quantitative analysis of small system impacts resulting from the rule.
After considering the economic impacts of the LT2ESWTR on small entities based on the
information presented in Appendix H, EPA certifies that the LT2ESWTR will not have a significant
economic impact on a substantial number of small entities. The small entities directly regulated by the
LT2ESWTRare small businesses, small organizations, and small governmental jurisdictions.
EPA has determined that 152 small entities, which are 2.3 percent of all small entities affected by
the LT2ESWTR, will experience an impact of 1 percent or greater of average annual revenues. Majority
of those systems, 105 of 152, will experience an impact between 1.0% and 1.5% of average annual
revenues. Twenty eight systems will experience an impact between 1.5 % and 2.0% of average revenues.
The Agency has determined that the remaining 18 small entities, which are 0.3 percent of all small entities
subject to the LT2ESWTR, will experience an impact of 3 percent or greater of average annual revenues.
EPA is certifying that the LT2ESWTR will not lead to significant economic impacts for a
substantial number of small entities, and, therefore is not required by the RFA, as amended by SBREFA,
to conduct a final regulatory flexibility analysis (FRFA). Nevertheless, EPA has tried to reduce the
impact of this rule on small systems.
Summary of the SBREFA Process
The RFA, as amended by SBREFA, and Section 203 of UMRA require EPA to provide small
governments with an opportunity for timely and meaningful participation in the regulatory development
process. EPA provided stakeholders, including small governments, with several opportunities to provide
input on the LT2ESWTR. For example, EPA conducted three conference calls to solicit feedback and
1 Revenue information was used whenever available. When it was not available, different measures, such
as sales or annual operating expenditures, were used.
Economic Analysis for the LT2ESWTR 7-2 December 2005
-------
information from the Small Entity Representatives (SERs) on issues regarding LT2ESWTR impacts on
small systems. SERs included small system operators, local government officials, and small nonprofit
organizations.
During the first call, held on January 28, 2000, EPA presented an overview of the SDWA, as
amended in 1996 and SBREFA. Issues and schedules for the LT2ESWTR rules were also discussed. The
second call was held on February 25, 2000. EPA presented the stakeholders with an overview of the EPA
regulatory development process and background on the development of the Stage 2 Microbial -
Disinfectants/Disinfection Byproduct (M-DBP) Rules, particularly regarding health risks, issues and
options identified by the Federal Advisory Committees Act (FACA) Committee, and Disinfection
Byproduct (DBP) and microbial occurrence in small systems. The third meeting was held on April 7,
2000. EPA presented SERs with a cost estimate and an impact analysis for selected regulatory options. In
addition, EPA presented SERs with schedules for the FACA and SBREFA processes.
These three conference calls generated a wide range of information, issues, and technical input
from SERs. To provide SERs with a foundation for commenting on these rules, EPA gave them
extensive background information. In general, the SERs were concerned about the impact of these rules
on small water systems (because of their small staff and limited budgets); small systems' ability to
acquire the technical and financial capability to implement this rule's requirements; maintaining the
flexibility to tailor requirements to their needs; and general limitations of small systems. The Agency
used the feedback received during these meetings in developing the LT2ESWTR. EPA also mailed a
draft version of the preamble to the attendees of these meetings.
The Agency convened a Small Business Advocacy Review (SBAR) Panel, in accordance with the
RFA as amended by SBREFA, to address small entity concerns, including those of small local
governments. EPA convened the SBAR Panel after completing the consultation meetings with SERs on
the LT2ESWTR. Eight of the small entities represented small governments. SERs' concerns were
provided to the SBAR Panel when the panel convened on April 25, 2000.
7.3 Small-System Affordability
Section 1415(e)(l) of SDWA applies to most rules and allows States to grant variances to small
water systems in lieu of complying with a maximum contaminant level (MCL) if EPA determines that no
nationally affordable compliance technologies exist for that system size/water quality combination. The
system must then install an EPA-listed variance treatment technology (Section 1412(b)(15)) that makes
progress toward the MCL, if not necessarily reaching it. Section 1415(e)(6)(B) of SDWA, however,
applies to the LT2ESWTR and states that a variance shall not be available under the above-noted
subsection for a "national primary drinking water regulation for a microbial contaminant (including a
bacterium, virus, or other organism) or an indicator or treatment technique for a microbial contaminant."
This EA does not identify affordable compliance technologies or variance treatment technologies because
the LT2ESWTR is a regulation to control a microbial contaminant.
7.4 Feasible Treatment Technologies for All Systems
In accordance with Section 1412(b)(4)(E) of the 1996 SDWA Amendments, EPA examined
whether there were existing, feasible technologies and treatment techniques available that would allow
systems to meet the LT2ESWTR requirements. EPA determined that filtered systems of all sizes could
meet the LT2ESWTR requirements using ultraviolet light (UV). In addition, small systems could
potentially meet the requirements using bag or cartridge filtration, while medium and large systems could
Economic Analysis for the LT2ESWTR 7-3 December 2005
-------
use ozone. According to the toolbox of treatment techniques (described in Chapter 6), UV can achieve 3
log reduction of Cryptosporidium, bag and cartridge filtration may achieve up to a 2 log reduction, and
ozone can achieve 2 log reduction. In fact, many small systems are predicted to choose cartridge
filtration or UV as treatment techniques.
The LT2ESWTR requires unfiltered systems to use two disinfectants to meet Cryptosporidium,
Giardia, and virus inactivation requirements, in which one disinfectant meets the full inactivation
requirement of at least one of the three pathogens. Considering studies that show UV can achieve greater
reduction of Cryptosporidium and Giardia at relatively low doses and chlorine can easily meet the virus
inactivation requirement, it is feasible for unfiltered systems to achieve the two disinfectant requirement.
All uncovered finished water reservoirs can meet the LT2ESWTR requirements by covering their
reservoirs or treating the effluent.
7.5 Effect of Compliance with the LT2ESWTR on the Technical, Managerial, and
Financial Capacity of Public Water Systems
Section 1420(d)(3) of SDWA, as amended, requires that, in promulgating a National Primary
Drinking Water Regulation (NPDWR), the Administrator shall include an analysis of the likely effect of
compliance with the regulation on the technical, managerial, and financial (TMF) capacity of PWSs. The
following analysis fulfills this statutory obligation by identifying the incremental impact that the
LT2ESWTR will have on the TMF of regulated water systems. Analyses presented in this document
reflect only the impact of new or revised requirements, as established by the LT2ESWTR; the impacts of
previously established requirements on system capacity are not considered.
Overall water system capacity is defined in Guidance on Implementing the Capacity Development
Provisions of the Safe Drinking Water Act Amendments of 1996 (USEPA 1998c) as the ability to plan for,
achieve, and maintain compliance with applicable drinking water standards. Capacity encompasses three
components: technical, managerial, and financial. Technical capacity is the operational ability of a water
system to meet SDWA requirements. Key issues of technical capacity include:
• Source water adequacy—Does the system have a reliable source of water with adequate
quantity? Is the source generally of good quality and adequately protected?
Infrastructure adequacy—Can the system provide water that meets SDWA standards? What
is the condition of its infrastructure, including wells or source water intakes, treatment and
storage facilities, and distribution systems? What is the infrastructure's life expectancy?
Does the system have a capital improvement plan?
• Technical knowledge and implementation—Are the system's operators certified? Do the
operators have sufficient knowledge of applicable standards? Can the operators effectively
implement this technical knowledge? Do the operators understand the system's technical and
operational characteristics? Does the system have an effective O&M program?
Managerial capacity is the ability of a water system's managers to make financial, operating, and
staffing decisions that enable the system to achieve and maintain compliance with SDWA requirements.
Key issues include:
Ownership accountability—Are the owners clearly identified? Can they be held accountable
for the system?
Economic Analysis for the LT2ESWTR 7-4 December 2005
-------
• Staffing and organization—Are the operators and managers clearly identified? Is the system
properly organized and staffed? Do personnel understand the management aspects of
regulatory requirements and system operations? Do they have adequate expertise to manage
water system operations (i.e., to conduct implementation, monitor for E. coli and
Cryptosporidium, install treatment, and cover or disinfect reservoir discharge to meet the
LT2ESWTR requirements)? Do personnel have the necessary licenses and certifications?
Effective external linkages—Does the system interact well with customers, regulators, and
other entities? Is the system aware of available external resources, such as technical and
financial assistance?
Financial capacity is a water system's ability to acquire and manage sufficient financial resources
to allow the system to achieve and maintain compliance with SDWA requirements. Key issues include:
• Revenue sufficiency—Do revenues cover costs?
• Creditworthiness—Is the system financially healthy? Does it have access to capital through
public or private sources?
Fiscal management and controls—Are adequate books and records maintained? Are
appropriate budgeting, accounting, and financial planning methods used? Does the system
manage its revenues effectively?
7.5.1 Requirements of the LT2ESWTR
This capacity analysis is presented only for the Preferred Alternative, although EPA took similar
considerations into account in the selection of the Preferred Alternative over the other alternatives. This
process led to the incorporation of less expensive rule features for systems having fewer capabilities. For
example, there is a possibility that a better, less burdensome indicator test for Cryptosporidium can be
developed based on the results of the source water monitoring conducted by large systems. If that is the
case, the burden on small systems will be less than that estimated here. Further, the schedule for small
systems to begin monitoring is 2 years after large systems. This time extension may increase familiarity
with these tests and perhaps lower the costs of laboratory analysis. Beyond the design of the rule, the
options available for small systems to comply were factored into the decision tree of available
technologies. The decision tree is discussed in detail in Chapter 6 and Appendix F.
This capacity analysis is based on the ICR occurrence data set. Analysis of the ICR data set
predicts the highest level of Cryptosporidium occurrence and, therefore, the greatest challenges that water
systems may face. Although two other data sets are available (ICR Supplemental Survey data for medium
systems (ICRSSM) and ICR Supplemental Survey data for large systems (ICRSSL)), these project that
fewer plants will need additional treatment and project fewer technical, financial, and managerial
challenges. EPA used the ICR data set to develop the most conservative capacity impact analysis.
The LT2ESWTR establishes five new requirements that may affect the TMF capacity of affected
PWSs:
1. Monitoring for E. coli (first or second round)
2. Monitoring for Cryptosporidium (first or second round)
Economic Analysis for the LT2ESWTR 7-5 December 2005
-------
3. Installing treatment (filtered systems)
4. Installing treatment (unfiltered systems)
5. Covering or disinfecting reservoir discharge
In addition, personnel from systems regulated under the LT2ESWTR will need to familiarize
themselves with the rule and its requirements.
7.5.2 Systems Subject to the LT2ESWTR
The LT2ESWTR will apply to all PWSs that treat surface water or GWUDI. However, because
systems purchasing surface water or GWUDI may incur costs through rate increases, EPA estimates that
the LT2ESWTRmay affect 5,378 CWSs, 766 NTNCWSs and 2,091 TNCWSs—8,235 filtered systems
and 60 unfiltered systems in all (see Exhibits 4.3 and 4.5). While most will not, some systems may
require increased TMF capacity to comply with the new requirements, or will need to tailor their
compliance approaches to match their capacities. Refer to section 7.5.4 for a detailed discussion of the
changes in TMF capacity for small and large systems.
7.5.3 Impact of the LT2ESWTR on System Capacity
The estimates presented in Exhibit 7.1 reflect the anticipated impact of the LT2ESWTR on
system capacity based on the expected measures that systems will be required to adopt. The extent of the
expected impact of a particular requirement on system capacity is estimated using a scale of 0-5, where 0
represents a requirement that is not expected to have any impact, 1 represents a requirement that is
expected to have a minimal impact, and 5 represents a requirement that is expected to have a very
significant impact on system capacity. Criteria used to develop the scores and associated impacts are
discussed further in section 7.5.4.
Impacts are assessed separately for small systems (Exhibit 7. la) and for large systems (Exhibit
7.1b). This distinction is necessary because most large systems will face fewer challenges in
implementing the rule than most small systems. For both large and small systems, EPA evaluated the
capacity impact of each requirement on those systems affected by that particular requirement. For
example, EPA only evaluated the impact of the Cryptosporidium monitoring requirement on small
systems that are required to monitor for Cryptosporidium as a consequence of the results of their E. coli
monitoring. In many cases, the requirements only affect a small percentage of systems/plants. The
exhibits, therefore, also display the number of systems and percent of systems/plants (of the subset of
small or large systems/plants) estimated to be affected by each specific requirement.
Economic Analysis for the LT2ESWTR 7-6 December 2005
-------
Exhibit 7.1a: Estimated Impacts of the LT2ESWTR on Small Systems' Technical,
Managerial, and Financial Capacity
(0 = no impact, 1 = minimal impact, and 5 = very significant impact)
Requirement
Familiarization with requirements of the rule
Monitoring for E. coli (first and second round)
Monitoring for Cryptosporidium
(first and second round)
Installation of treatment (filtered plants) (Bins 2
and 3)
Installation of treatment (filtered plants)
(Bin 4)
Installation of treatment (unfiltered plants)
Cover or disinfect reservoir discharge1
Number and
Percent of
Small Plants
5,663 (100.0%)1
5,575 (97%)2
1,978(34%)2
2,069 (36%)
136(2%)
38 (0.6%)
12(0.2%)
Technical Capacity
CD
£ >•
> 0
CD OT
Q g-
=! CD
O T3
0
0
0
0
0
2
0
CD
1 ra
ro 1"
^<
0
0
0
2
4
4
4
oa J
sfi
O O ~Q_
1
1
2
4
4
4
3
Managerial Capacity
>,
.9-^
en "c
CD =!
r- O
a o
0
<
0
1
1
2
2
3
2
g
D).N
c c
it ro
JS ^
OT O
1
0
0
2
3
3
3
CO
CD
X
LU
~ D)
O CO
CD ^
8= !=
LU _l
0
0
1
2
3
3
3
Financial Capacity
CD £
=5 CD
CD =5
OL 0)
0
1
3
4
5
5
5
y>
o
^
o
0
0
0
3
4
4
4
o3
D)
lo -t
0 £=
in o
LL O
0
0
2
2
3
5
3
1Number and percent of small systems.
2This cell contains only the number and percentage of plants expected to be affected by first-round monitoring requirements. The number of plants participating in
the second round of monitoring is expected to be smaller. The rankings for capacity for monitoring include both the first and second rounds.
Note: To analyze the impact of these requirements on system capacity, the requirements believed to have the most and least impact on affected systems (i.e., the
installation of treatment by plants placed into Bin 4, and familiarization with the requirements of the rule, respectively), were analyzed first, as described in section
7.5.4. These initial analyses were then used as the basis against which the relative impacts of the remaining requirements were assessed. The impact estimates
developed for each requirement were also compared to the Stage 2 DBPR to ensure cross-rule consistency and enable cross-rule comparisons. Analysis is based
on data modeled from the ICR data set, because that data set predicts the highest level of occurrence and, therefore, the greatest challenges systems will face.
Source: Number and percent of plants/systems impacted by each requirement are derived from systems serving • 10,000 (Exhibit 6.2). Number and percent of
small plants making treatment changes from Appendix G, Exhibits G.37-G.39 and unfiltered systems are derived from Exhibit 4.5. Number of systems required to
cover or disinfect reservoir discharge are derived from Exhibit 4.25. Impact on capacity is determined relative to previous regulations based on the cost and
number of systems/plants that require additional capacity to comply with each requirement, as described in section 7.5.4.
Economic Analysis for the LT2ESWTR
7-7
December 2005
-------
Exhibit 7.1 b: Estimated Impacts of the LT2ESWTR on Large Systems'
Technical, Managerial, and Financial Capacity
(0 = no impact, 1 = minimal impact, and 5 = very significant impact)
Requirement
Familiarization with requirements of the rule1
Monitoring for E. coli (first and second round)
Monitoring for Cryptosporidium
(first and second round)
Installation of treatment (filtered plants)
(Bins 2-4)
Installation of treatment (unfiltered plants)
Cover or disinfect reservoir discharge1
Number and
Percent of
Large Plants
1,493(100%)
1,733(98%)2
1,762(99.6%)2
654 (33%)
25 (0.4%)
69 (5%)
Technical Capacity
i_
-2
1 t?
CD CD
a §•
=3 CD
O T3
0
0
0
0
2
0
i-
"o >*
1 %
« 1
0
0
0
3
3
3
D)
T3
CD d
o ID
^ c
0 CD
£= Q.
-g E
1- o3
1
1
2
3
4
4
Managerial Capacity
>,
i=
!E CD
<5 ^
CD ^
r- O
a 0
O
0
1
1
2
3
2
c
D).N
"m O)
W O
1
0
0
3
3
1
CD
CD
"X
LJJ
^ D)
O CD
LJJ _l
0
0
1
2
3
2
Financial Capacity
>,
QJ
O
!t
^
QJ
QJ
QJ
cc
0
0
1
2
3
3
£=
_C
|
^
CD
il
O
0
0
0
2
2
2
o3
-g
D)
"CD J?
0 £=
LL O
0
0
1
2
3
3
1Number and percent of large systems.
2This cell contains only the number and percentage of plants expected to be affected by first-round monitoring requirements. The number of plants participating in
the second round of monitoring is expected to be smaller. The rankings for capacity for monitoring include both the first and second rounds.
Note: To analyze the impact of these requirements on system capacity, the requirements believed to have the most and least impact on affected systems (i.e., the
installation of treatment by plants placed into Bin 4, and familiarization with the requirements of the rule, respectively), were analyzed first, as described in section
7.5.4. These initial analyses were then used as the basis against which the relative impacts of the remaining requirements were assessed. The impact estimates
developed for each requirement were also compared to the Stage 2 DBPR to ensure cross-rule consistency and enable cross-rule comparisons. Analysis is based
on data modeled from the ICR data set, because that data set predicts the highest level of occurrence and, therefore, the greatest challenges systems will face.
Source: Number and percent of plants/systems impacted by each requirement are derived from systems serving • 10,000 (Exhibit 6.2). Number and percent of
large filtered plants making treatment changes from Appendix G, Exhibits G.37-G.39 and unfiltered systems are derived from Exhibit 4.5. Number of systems
required to cover or disinfect reservoir discharge are derived from Exhibit 4.25. Impact on capacity is determined relative to previous regulations based on the cost
and number of systems/plants that require additional capacity to comply with each requirement, as described in section 7.5.4.
Economic Analysis for the LT2ESWTR
7-&
December 2005
-------
7.5.4 Derivation of the LT2ESWTR Scores
EPA developed a 5-point scoring system to analyze the impact compliance with all new
regulations will have on the technical, managerial, and financial capacity of PWSs. For each regulation,
it is necessary to complete the following steps:
1. Determine the type and number of PWSs to which the regulation applies
2. List all of the requirements of the regulation
3. Determine the type and number of PWSs to which each requirement applies
4. Evaluate the impact of each requirement on the capacity of affected PWSs
The determination of the universe of affected systems and the evaluation of the capacity impact
of individual requirements requires the use of the cost and technical information contained in SDWIS,
EAs developed for other rules, information collection requests, and other supporting documentation for
the rule. These data sources are also used to develop a qualitative description of the expected response of
affected systems to each requirement.
The overall evaluation of the impact of a requirement on the affected systems, presented in
Exhibit 7.1, is determined by the impact of each requirement on nine sub-categories of capacity—three
sub-categories under each of the broader divisions of technical, managerial, and financial capacity.
Within these sub-categories, a professional engineer with extensive water system experience reviewed the
costs, number of systems affected, and complexity of each requirement. After estimating the technical,
managerial, and financial impacts within each sub-category, the professional engineer assigned the scores
using best professional judgment. Costs were considered cumulatively for each requirement for small and
large systems. This score reflects the additional capacity that systems will need to develop to comply
with each requirement. Due to a lack of available information on operating budgets, this analysis does
not include a quantitative component.
To ensure the ability to make cross-rule comparisons, to standardize the assignment of numerical
scores, and to minimize the subjectivity of the scoring system, the requirements made on systems by the
regulation in question are compared to the requirements of those regulations for which capacity impact
analyses have already been conducted (e.g., Ground Water Rule, and LT1ESWTR). Similar requirements
were assigned similar impact scores.
These group assignments are reviewed by the EPA Rule Manager and other EPA staff cognizant
of small system issues to ensure that they accurately reflect the cumulative impact of the rule
requirements on system capacity. Any disagreements over the assignments are discussed. The EPA Rule
Manager and other EPA staff discuss the rationale for the disagreement and evaluate whether the
assignments need to be adjusted. EPA adjusts the assignments only after review of the rule support
documents and an analysis of the expected system response to the rule requirements.
Small Water Systems (Those Serving 10,000 or Fewer People)
Most small systems will likely face only a minimal challenge to their technical and managerial
capacity as a result of efforts to familiarize themselves with LT2ESWTR and comply with the
requirements for monitoring of E. coll (Exhibit 7. la). Systems monitoring for Cryptosporidium,
however, will require additional assistance. In addition, systems with source waters that trigger them to
monitor for Cryptosporidium will need to pay higher sampling costs since these systems have not
previously performed the strict sampling protocols that are part of EPA's approved analytical method. To
Economic Analysis for the LT2ESWTR 7-9 December 2005
-------
meet these challenges, it is likely that systems will need to develop or enhance linkages with technical and
financial assistance providers (including State extension agents).
Exhibit 7. la indicates that installation of some new treatment technologies will pose significant
challenges to small systems' technical, managerial, and financial capacity; the less the current level of
treatment, the greater the impact will be. As with Cryptosporidium monitoring, the development and
enhancement of external linkages will be very important to systems that must install new equipment.
Technical and financial assistance providers can help systems analyze their needs as well as the trade-offs
between cost and health protection. In addition, they may be able to assist systems in finding the funding
necessary to install and operate new equipment.
The requirement to obtain additional log-removal credits will likely challenge, to a significant
degree, the financial capacity of some of the small systems affected. This is especially true since those
systems may not have included the costs associated with complying with the LT2ESWTR in their long-
term financial plans. Incurring these costs will tend to reduce a system's credit rating since it will be
forced to direct more of its revenue to new equipment, and systems (or entities owning the systems) may
be required to issue bonds or obtain loans. It is also evident from Exhibit 7. la that the impacts of the
LT2ESWTR on the capacity of systems assigned to Bins 2 to 4 will be similar.
The scores presented in Exhibit 7.1 apply to those (relatively few) systems most heavily affected.
For example, of the very small systems (serving 500 or fewer people) that are required to add treatment
technologies to comply with the requirements of Bins 2 and 3,10 percent or fewer are predicted to install
UV (see Appendix F). Of those very small systems that are predicted to require an additional 2.5 log
treatment (as required by Bin 4), all must install UV. UV requires little involvement for the operator since
UV can be monitored on-line. Only systems with the capability to handle this technology are expected to
use it. The rest will be able to rely on low-cost, uncomplicated bag and cartridge filters.
Small plants serving between 500 and 10,000 people may install microfiltration or ultrafiltration
(MF/UF) to comply; however, fewer than one-third of 1 percent of those requiring treatment are expected
to choose MF/UF. MF/UF requires a daily integrity test during which finished water production must be
interrupted, so only those few systems that have the capability would make such a choice. For those in
Bin 4, only 1 percent are projected to use MF/UF, with 90 percent using UV and the rest using
combinations of technologies.
Those few systems that do not now filter their water will be required to provide an additional 2 or
3 log disinfection under LT2ESWTR. In many cases, the financial capacity of these systems will be
affected since they may need to revise their budgeting process to account for new capital and O&M
expenses.
Systems that rely entirely on purchased water will experience negligible technical and managerial
impacts, if any. The responsibility for implementing the necessary changes inherent in LT2ESWTR will
fall upon those systems that sell water to other systems, not those that purchase it. The latter will not be
responsible for implementing any technical changes and will only experience economic effects associated
with their supplier's compliance with LT2ESWTR. Issues of sufficient revenue may arise; however, any
additional costs that result will eventually be passed on to consumers, resulting in little effect on small
purchased systems.
Finally, the TMF capacity of small systems required to cover a reservoir or to provide post-
reservoir disinfection will be affected to approximately the same degree as the capacity of those filtered
systems that must install additional treatment based on their action bin. In both cases, system staff will
need to learn to operate and maintain new equipment. System management will need to ensure the
presence of adequately trained staff and will need to explain the need for the additional equipment to
Economic Analysis for the LT2ESWTR 7-10 December 2005
-------
customers and rate boards. The systems also will face significant new costs, potentially requiring
adjustments to the rate structure, billing practices, and capital planning practices.
The overall impacts on small systems' technical, managerial, and financial capacity will vary.
Monitoring and familiarization with new rules will have no significant effects on small systems, with the
exception of moderate revenue constraints on those systems that need to implement monitoring for
Cryptosporidium. It should be noted that all second-round monitoring will impose reduced impacts. The
largest impacts will occur as a result of attaining 2.5 log treatment levels, covering uncovered reservoirs,
or disinfecting reservoir discharge.
Large Water Systems (Those Serving at Least 10,000 People)
Large regulated systems will likely not face more than a minimal challenge to their technical and
managerial capacity as a result of efforts to familiarize themselves with the LT2ESWTR and monitor for
E. coli (Exhibit 7.1b). The three monitoring requirements established under the LT2ESWTR, however,
vary substantially in their impact on system capacity. While measuring turbidity levels and monitoring
coliforms will have a minimal effect, monitoring for Cryptosporidium may require increased system
management, technical assistance, and cost of sampling. However, even the largest monitoring impacts
will not be very significant. Those systems serving more than 100,000 people have already upgraded to
meet Cryptosporidium monitoring needs (and collected such samples under the ICR); therefore, systems
serving between 10,000 and 100,000 people will experience the majority of monitoring impacts.
Exhibit 7. Ib shows that the installation of new treatment technology will pose moderate
challenges to large systems' technical, managerial, and financial capacity. As with Cryptosporidium
monitoring, the development and enhancement of external linkages will be very important to systems that
must install new equipment.
The requirement to obtain additional log-removal credits will likely challenge the financial
capacity of systems only to a small degree. Incurring costs might reduce a system's credit rating since it
will be forced to direct more of its revenue to new equipment, and may be required to issue bonds or
obtain loans. The LT2ESWTR will have essentially similar impacts on the capacity of large systems
assigned to different action bins because similar technologies are expected to be installed in spite of
varying requirements.
Large systems that rely on purchased water will experience only minimal impacts since the
responsibility of meeting the requirements of LT2ESWTR will fall primarily on systems selling water.
Systems purchasing water will experience only those economic effects associated with their supplier's
compliance with LT2ESWTR and any additional costs will ultimately be applied to consumers.
Therefore, the LT2ESWTR will have little effect on large systems purchasing water.
The capacity of systems required to install a cover or to provide post-reservoir disinfection will
be affected to a greater degree than the capacity of those filtered systems that must install additional
treatment. In both cases, system staff will need to learn how to operate and maintain new equipment.
System management will need to ensure the presence of adequately trained staff and will need to explain
the need for the additional equipment to customers and rate boards. Additionally, the systems will face
significant new costs that could require adjustments to rate structures, billing practices, and capital
planning methods.
Overall, EPA assumed that large systems will have the technical, financial, and managerial
capacity to implement LT2ESWTR requirements based on the scale and complexity of their operations.
The nature of their operations generally assures that they have access to the technical and managerial
expertise to carry out all activities required by the LT2ESWTR. It is also generally easier for large
Economic Analysis for the LT2ESWTR 7-11 December 2005
-------
systems to fund capital improvements than small systems, since costs can be spread over a larger
customer base, making them smaller on a per-household basis.
7.6 Paperwork Reduction Act
The information collection requirements for the LT2ESWTR have been submitted for approval to
the Office of Management and Budget (OMB) under the Paperwork Reduction Act, 44 U.S.C. 3501 et
seq. The information collected as a result of this rule will allow the States/Primacy Agencies and EPA to
determine appropriate requirements for specific systems and evaluate compliance with the rule.
The Paperwork Reduction Act requires EPA to estimate the burden on PWSs and States/Primacy
Agencies of complying with the rule. Burden means the total time, effort, and financial resources
required to generate, maintain, retain, disclose, or provide information to or for a Federal agency. This
burden includes the time needed to conduct these activities:
• Review instructions
• Develop, acquire, install, and employ technology and systems for the purposes of collecting,
validating, verifying, processing, maintaining, and disclosing information
Adjust the existing ways to comply with any previously applicable instructions and
requirements
• Train personnel to respond to information collected
• Search data sources
• Complete and review the collection of information
• Transmit or otherwise disclose the information
For the first 3 years after publication of the final LT2ESWTR in the Federal Register, the major
information requirements pertain to implementation activities for States/Primacy Agencies and PWSs,
covering uncovered finished water reservoirs, monitoring activities for large systems, and preparation for
monitoring activities by small systems. The information collection requirements are mandatory under
Part 141 of the NPDWRs. The calculation of LT2ESWTR information collection burden and costs can be
found in the Information Collection Request for the Long Term 2 Enhanced Surface Water Treatment
Rule (USEPA 2004b).
The total burden associated with LT2ESWTR requirements over the 3 years covered by the
Information Collection Request is 423,886 hours, an average of 141,295 hours per year. The total cost
over the 3-year clearance period is $34.1 million, an average of $ 11.4 million per year (simple average
over 3 years). (These estimates are based on modeled results of the ICR Cryptosporidium occurrence data
set.) EPA assumes that the systems affected by the LT2ESWTR have already purchased the basic
equipment required for monitoring and reporting. Therefore, there are no capital start-up costs associated
with information collection under this rule. The average burden per response (i.e., the amount of time
needed for each activity that requires a collection of information) is 0.63 hours; the average cost per
response is $50.35. Exhibit 7.2 provides a summary of the results of the Information Collection Request
calculations.
Economic Analysis for the LT2ESWTR 7-12 December 2005
-------
Exhibit 7.2: Average Annual Burden Hours and Costs for the LT2ESWTR
Information Collection Request
Burden
Hours
Labor Cost
Capital
Cost
Non-Labor
Cost
Average
Annual Cost
Water Systems
Implementation
£. co// monitoring
Cryptosporidium monitoring
Reporting
19,236
15,337
5,216
7,390
$ 529,252
$ 317,941
$ 144,370
$ 209,252
$
$
$
$
$
$ 1,492,548
$ 5,523,681
$
$ 529,252
$ 1,810,489
$ 5,668,051
$ 209,252
States and Territories
Implementation
Reporting
Total
77,064
16,796
141,295
$ 2,589,722
$ 564,427
$ 4,363,363
$
$
$
$
$
$ 7,016,230
$ 2,589,722
$ 564,427
$11,379,592
Note: Data represent burden and cost for only the 3-year ICR clearance period. Data are based on nominal (or
undiscounted) values. Detail may not add due to independent rounding.
Source: Information Collection Request for the Long Term 2 Enhanced Surface Water Treatment Rule (USEPA
2004a).
7.7 Unfunded Mandates Reform Act
The UMRA of 1995, Public Law 104-4, consists of four Titles and numerous sections. Sections
201 through 205 of Title II, entitled "Regulatory Accountability and Reform,"are relevant to the
LT2ESWTRand are discussed in this section. Title II, Section 201 of the UMRA, requires Federal
agencies to assess the effects of their regulatory actions on State, Local, and Tribal governments, and the
private sector. Under UMRA Section 202, 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 1 year. Section 203 requires the Agency to establish a small government agency plan before
establishing any regulatory requirements that may significantly or uniquely affect small governments.
Section 204 of the UMRA requires the Agency to develop an effective process to permit elected
officers of State, Local, and Tribal governments to provide meaningful and timely input in the
development of regulatory proposals that contain significant Federal intergovernmental mandates.
Finally, Section 205 generally requires EPA to identify and consider a reasonable number of regulatory
alternatives and adopt the least costly, most cost-effective, or least burdensome alternative that achieves
the objectives of the rule before promulgating a rule for which a written statement is needed under
Section 202. 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 final rule an explanation why that
alternative was not adopted.
EPA has determined that this rule contains a Federal mandate that may result in expenditures of
$100 million or more for State, Local, and Tribal governments, in the aggregate or the private sector in
any one year, as shown in Exhibit 7.3.
Economic Analysis for the LT2ESWTR
7-13
December 2005
-------
Exhibit 7.3: Annualized Value of Public and Private Costs for the
LT2ESWTR (Annualized at 3 and 7 Percent)
Publicly Owned PWS
Costs
State Costs
Tribal Costs
Total Public Costs
Total Private Costs
Total Costs
Range of Annualized Costs
(Million$, 2003$)
3% Discount Rate
$57.4 - $82.7
$1.1 - $1.2
$0.1 - $0.2
$58.6 - 84.1
$34.3 - 49.4
$92.9 - $133.4
7% Discount Rate
$65.9 - $88.6
$1.4 - 1.4
$0.2 - $0.3
$67.4 - 90.3
$39.3 - 60.2
$106.8 - 150.5
Percent of Total Cost
61.8% - 62.0%
1 .2% - 0.9%
0.2% - 0.2%
63.1% - 63.0%
36.9% - 37.0%
100.0% - 100.0%
Note: The ranges reflect the difference between the ICRSSL (lowest) and ICR (highest) modeled Cryptosporidium
occurrence distributions. The percentages of total cost in the last column were calculated only for the 3 percent
discount rate. Detail may not add due to independent rounding.
Source: "Publicly owned PWS costs" are from the total system cost in Exhibit 6.4, multiplied by the proportion of
PWSs that are publicly owned according to SDWIS. "State costs" are from Exhibit 6.4. "Tribal costs" are from Exhibit
7.9 for the high end of each range in this exhibit and are assumed to represent the same proportion of total costs in
the low end of the range. "Total private costs" include costs for all privately owned PWSs and are calculated by
subtracting total public costs from total costs.
Thus, the LT2ESWTR is subject to the requirements of Sections 202 and 205 of UMRA, and
EPA is obligated to prepare a written statement addressing the following items:
• The authorizing legislation
• Benefit-cost analysis, including an analysis of the extent to which the costs of State, Local
and Tribal governments will be paid for by the Federal government
Estimates of future compliance costs and disproportionate budgetary effects
• Macroeconomic effects
A summary of EPA's consultation with State, Local, and Tribal governments and their
concerns, including a summary of the Agency's evaluation of those comments and concerns
• Identification and consideration of regulatory alternatives and the selection of the least costly,
most cost-effective, or least burdensome alternative that achieves the objectives of the rule
The legislative authority for the LT2ESWTR is discussed in Chapter 2. The remaining items are
discussed below, but are also addressed in other chapters of this EA, such as Chapters 3 and 6.
7.7.1 Social Benefits and Costs
The social benefits are those that accrue primarily to the public through increased protection from
illness and potential death caused by exposure to microbial pathogens in drinking water. To assign a
monetary value to the reductions in illness, EPA used a cost-of-illness measure. This is considered to be a
lower-bound estimate of actual benefits because it does not include the pain and discomfort associated
with the illness. Mortalities were valued using a value of statistical life estimate consistent with EPA
Economic Analysis for the LT2ESWTR 7-14 December 2005
-------
policy. Chapter 5 presents the benefits analysis, which includes both qualitative and monetized benefits
of improvements to health and safety. The potential nonquantifiable benefits may include reduced risks
to sensitive subpopulations, reduced outbreak risks and response costs, reduced risk-averting behavior
(e.g., boiling tap water or purchasing bottled water), reduced risk from co-occurring pathogens, increased
source water monitoring, increased regulation of unfiltered systems, and covering or treating finished
water reservoirs. In addition, certain nonhealth-related benefits may exist, such as enhanced aesthetic
water quality. The estimated annualized quantified benefit for the traditional cost of illness (COI) for the
Preferred Alternative of the LT2ESWTR using a 3-percent discount rate ranges from $335 to $1,341
million (or $272 to $1,089 million using a 7-percent discount rate) (Exhibit 7.4). Similarly, the estimated
annualized quantified benefit using the enhanced COI ranges from $458 to $1,853 million (or $371 to
$1,501 million using a 7-percent discount rate).
Measuring the social costs of the rule requires identifying affected entities by ownership (public
or private), considering regulatory alternatives, calculating regulatory compliance costs, and estimating
any disproportionate impacts. Chapter 6 of this document details the cost analysis performed for the
LT2ESWTR. Under the Preferred Alternative, the likely compliance scenario is expected to result in total
annualized costs of approximately $93 to $133 million using a 3-percent discount rate (or $107 to $150
million using a 7-percent discount rate). Exhibit 7.4 summarizes the range of annualized costs and
benefits for each regulatory alternative.
Economic Analysis for the LT2ESWTR 7-15 December 2005
-------
Exhibit 7.4: Total Annualized Benefits and Costs of Regulatory Alternatives
(SMillions, 2003$)
Regulatory
Alternative
Alternative A1
Alternative A2
Alternative A3
(Preferred
Alternative)
Alternative A4
Enhanced COI
Range of
Annualized
Benefits (3%)
$558 - $1 ,895
$489 - $1 ,871
$458 - $1 ,853
$405 - $1 ,753
Traditional COI
Range of
Annualized
Benefits (3%)
$403 - $1 ,369
$356 - $1 ,353
$335 - $1 ,341
$299 - $1 ,273
Enhanced COI
Range of
Annualized
Benefits (7%)
$452 - $1 ,534
$396 -$1,51 5
$371 -$1,501
$328 - $1 ,421
Traditional COI
Range of
Annualized
Benefits (7%)
$327 -$1,1 12
$289 - $1 ,099
$272 - $1 ,089
$243 - $1 ,034
Range of
Annualized
Costs (3%)
$403
$123 -$163
$93 -$133
$57 - $81
Range of
Annualized
Costs (7%)
$436
$139 -$182
$107 -$150
$68 - $93
Source: Benefits from Exhibits 5.28a-b. Costs from Exhibit 6.16.
Various Federal programs exist to provide financial assistance to State, Local, and Tribal
governments in complying with this rule. The Federal government provides funding to States that have
primary enforcement responsibility for their drinking water programs through the Public Water Systems
Supervision (PWSS) Grants Program. States may use these funds to develop primacy programs or to
contract with other State agencies to assist in the development or implementation of their primacy
programs. However, they may not use these funds to contract with regulated entities (i.e., water systems).
States may use PWSS Grants to set up and administer a State program that includes such activities as
public education, testing, training, technical assistance, development and administration of a remediation
grant and loan or incentive program (excluding the actual grant or loan funds), or other regulatory or
nonregulatory measures.
Additional funding is available from other programs administered by EPA or other Federal
agencies. These include EPA's Drinking Water State Revolving Fund (DWSRF), the U.S. Department of
Agriculture's Rural Utilities' Loan and Grant Program, and the Department of Housing and Urban
Development's Community Development Block Grant (CDBG) Program.
SDWA authorizes the EPA Administrator to award capitalization grants to States, which in turn
can provide low-cost loans and other types of assistance to eligible PWSs. The DWSRF assists PWSs
with financing the costs of infrastructure needed to achieve or maintain compliance with SDWA
requirements. Each State has considerable flexibility to determine the design of its DWSRF Program and
to direct funding toward its most pressing compliance and public health protection needs. States may
also, on a matching basis, use up to 10 percent of their DWSRF allotments for each fiscal year to assist in
running the State drinking water program. In addition, States have the flexibility to transfer a portion of
funds from their Clean Water State Revolving Fund accounts to their DWSRF accounts.
A State can use the financial resources of the DWSRF to assist small systems. In fact, a
minimum of 15 percent of a State's DWSRF grant must be used to provide infrastructure loans to systems
serving 10,000 or fewer people. Two percent of the State's grant is set-aside funding that can only be
used to provide technical assistance to small systems. In addition, up to 14 percent of the State's grant
may be used to provide technical, managerial, and financial assistance to all system sizes. For small
systems that are disadvantaged, up to 30 percent of a State's DWSRF may be used for increased loan
subsidies. Tribes have separate set-aside funding that they can use under the DWSRF.
In addition to the DWSRF, money is available from the Department of Agriculture's Rural Utility
Service (RUS) and Housing and Urban Development's CDBG Program. RUS provides loans, loan
guarantees, and grants to improve, repair, or construct water supply and distribution systems in rural areas
and towns with a population of up to 10,000 people. In fiscal year 2003, RUS had over $1.5 billion of
Economic Analysis for the LT2ESWTR
7-16
December 2005
-------
available funds for water and environmental programs. Also, three sources of funding exist under the
CDBG program to finance building and improvements of public facilities such as water systems. These
include: 1) direct grants to communities with populations over 200,000; 2) direct grants to States, which
in turn are awarded to smaller communities, rural areas, and colonas in Arizona, California, New Mexico,
and Texas; and 3) direct grants to U.S. territories and trusts. The CDBG budget for fiscal year 2003
totaled over $4.4 billion.
7.7.2 Disproportionate Budgetary Effects
UMRA is intended to reduce the burden on State, Local, and Tribal governments of Federal
mandates that are not accompanied by adequate Federal funding. Section 202 of UMRA requires an
analysis of possible disproportionate budgetary effects of certain classes of rules, in which LT2ESWTR
falls.2 Such an analysis is required if EPA determines that accurate estimates are reasonably feasible.
The specific concern is disproportionate budgetary effects of the LT2ESWTRupon certain areas or
industries:
• Any particular regions of the United States
• Any particular State, Local, or Tribal government
• Urban or rural or other types of communities
• Any segment of the private sector
This EA has considered how best to interpret and comply with these requirements. The
remainder of this section describes ways to consider these requirements, whether meaningful data can be
provided, and whether accurate estimates are possible. The general conclusion for this section, however,
is that there are little basis and insufficient data to make accurate estimates of budgetary impacts that
differ among groups, regions, governments, types of communities, or segments of the private sector.
Further, from all of the Agency's analysis and consultations, the Agency believes the rule will treat
similarly situated systems (in terms of size, water quality, available data, installed technology, and
presence of uncovered finished reservoirs) in similar (proportionate) ways, without regard to geographic
location, type of community, or segment of industry. The LT2ESWTR is a rule whose requirements are
proportional to risk. In this analysis, the estimates of occurrence are different only for filtered and
unfiltered systems, reflecting the judgment that differentiation along these other specific characteristics
(regions, type of government, and so forth) is not possible with the available data. Although some groups
may have differing budgetary effects as a result of the LT2ESWTR, those costs are proportional to the
need for additional monitoring and the risk posed by Cryptosporidium occurrence.
Most of the following analyses begin with national data and then disaggregate those data, when
possible, using other measures. Because the data and estimates are national in scope, breakouts by region
or other parameters tend to be merely proportional extensions based on a characteristic, not "bottom-up"
estimates of actual differences among types of systems, communities, or economic sectors. Thus, the
analyses may not reveal true differences attributable to the impacts of the rule alternatives on various
regions. Local conditions at each regulated entity will drive the actual cost impacts of the rule (e.g., the
level of Cryptosporidium in the source water) and these data are not available. In fact, the LT2ESWTR
2 "...[T]he agency shall prepare a written statement containing. . . (3) estimates by the agency, if and to the
extent that the agency determines that accurate estimates are reasonably feasible, of.. . (B) any disproportionate
budgetary effects of the Federal mandate upon any particular regions of the nation or particular State, Local, or
Tribal government, urban or rural or other types of communities, or particular segments of the private sector..."
Economic Analysis for the LT2ESWTR 7-17 December 2005
-------
requires most systems to test their water to determine the impact of the rule on their system, a process
that, in time, could generate that information.
When considering disproportionate impacts, it is necessary to consider whom the LT2ESWTR
affects. The rule, by definition, covers some communities and a segment of the private sector. Most of
those communities and PWSs that use surface water or GWUDI and those having uncovered finished
water reservoirs will incur some costs. Communities and PWSs that use only ground water or have no
uncovered finished water reservoirs will avoid these costs, although they may have to comply with other
rules to which surface water systems may not be subject. In an economic sense, these differences
between communities and utilities do not disadvantage one group over the other because the systems are
not in a national market that allows for direct competition for customers. In general, those systems are
better considered local natural monopolies.
Regions
There are no specific data available that suggest that the compliance costs and other effects of
LT2ESWTR will cause disproportionate budgetary effects by region. LT2ESWTR is a national mandate
and applies uniformly to all States. These effects will be felt at the system level, with most systems
primarily facing monitoring, rather than treatment, costs. Some contaminants in drinking water are
distributed unevenly across regions. The data available on the occurrence of Cryptosporidium, however,
indicate that it is prevalent nationwide. More data on Cryptosporidium occurrence will become available
after the monitoring required by the rule is completed.
Although Cryptosporidium levels do not show strong regional patterns, it is still possible that
budgetary effects could differ between regions. There is no direct measure of potential budgetary impacts
by regions, but proxy measures are considered. One possible proxy for potential regional impacts is the
projection that smaller systems will be subject to greater impacts on their financial capacity (including
revenue sufficiency) (Exhibit 7. la) and that small systems will face greater budgetary pressures,
particularly if installing treatment, because of economies of scale (this effect is seen in higher average
household costs for those served by small PWSs). Regions have varying proportions of small, medium,
and large systems that supply public water. To the extent that some regions are more dependent on small
systems, the regions as a whole could be considered more likely to face greater impacts on budgets of
many small entities, even if there is no "regional budget."
To show what proxy measures based on dependency on small systems might reveal, two
measures are used. The first is the percent of the population of a State that is served by small, rather than
large, systems. Exhibit 7.5 indicates that the States that have more dependence on small water systems
tend to be lower-population States such as Vermont, West Virginia, and Alaska. Most relevant to this
analysis, no regional patterns are apparent.
Economic Analysis for the LT2ESWTR 7-18 December 2005
-------
Exhibit 7.5: Percent of Population of CWSs Served by Small Surface and
GWUDI Systems by State
I American Samoa
| 2% | Guam
| 5% | N. Mariana Islands
|i4% | Palau | 6% | Puerto Rico
I I Virgin Islands
Small Community Surface
Water and GWUDI Systems
over the United States
fj >20% (7)
fj 11-20% (11)
in 0-10% (39)
o% D.C.
Source: Appendix M.
The second proxy measure used is the absolute number of small systems. Not surprisingly, this
tends to correlate with high total population, with New York, California, and Texas among the largest.
Again, no regional patterns are evident.
This analysis concludes that accurate estimates are not reasonably feasible, but significant
regional impacts are not expected. Further, tests with proxy measures support the conclusion of no
regional disproportionate budgetary impacts.
State, Local, or Tribal Governments
There is no expectation that there will be disproportionate budgetary effects upon State, Local, or
Tribal governments. Costs are expected to be proportional to the risk Cryptosporidium occurrence poses,
even if unevenly distributed among systems and perhaps types of systems. Furthermore, there are no
accurate estimates to address the differing budgetary effects of LT2ESWTR on State, Local or Tribal
governments.
There are few data available that bear on this issue. Exhibit 7.3 breaks out national-level costs for
public PWSs, Tribal costs, and State costs, but only allocates costs to these categories rather than
revealing any disproportionate impacts on the budgets of these groups. Exhibits 7.5 and 7.6 imply that
State impacts will be larger to the extent that States contain a greater proportion of small systems. Even
using dependence on small systems as a measure, an accurate distribution of potential impacts will not be
available until after the monitoring phase is complete.
Economic Analysis for the LT2ESWTR
7-19
December 2005
-------
Urban and Rural
There are no data that distinguish between the budgetary effects of LT2ESWTR on urban versus
rural areas, and there is no expectation that one kind of area will pay disproportionately higher costs than
the other.
The only data available that may be relevant are the cost differences that are expected to exist
between small and larger systems. The analyses in this EA, based on the Technologies and Costs
Document (USEPA 2003a), estimate that there are economies of scale: as the population served increases,
average costs of service decrease. If it is assumed that rural systems are smaller than urban systems, the
latter may face smaller per-household costs than rural systems. However, this variation does not imply an
effect that is disproportionate to risk. The variations in costs will not predominantly fall along the lines of
population served (although population served can be used as a proxy), but will specifically depend on
which treatment bin a water system is assigned to—depending on the level of treatment that systems'
water needs. Further, many large systems are suburban systems, and many systems sell water to other,
sometimes rural, systems. Again, there is no expectation of disproportionate budgetary impact, based on
the design of the rule and known patterns of occurrence, and accurate estimates are not now feasible.
Segments of the Private Sector
Only one segment of the private economy is directly affected by this rule—drinking water
providers. Section 7.5 discusses the budgetary impacts on this sector. An indirect indicator of the
reasonableness of the cost of the LT2ESWTR is the agreement achieved by the Stage 2 M-DBP FACA
Committee on major rule elements. Based on this agreement, the budgetary impact could not
disproportionately affect drinking water providers since it is explicitly proportionate to risk.
Economic Analysis for the LT2ESWTR 7-20 December 2005
-------
Exhibit 7.6: Number of Small Surface and GWUDI Systems by State
D.C.
TT Palau
|lOI American Samoa
[~n Guam
| 1 | N. Mariana Islands
|i8i] Puerto Rico
Owl Virgin Islands
Number of Small CWSS
301 or more (0)
201 to 300 (4)
101 to 200 (8)
0 to 100 (45)
Source: Appendix M.
7.7.3 Macroeconomic Effects
Under UMRA Section 202, EPA is required to estimate the potential macroeconomic effects of
the regulation. These include effects on productivity, economic growth, full employment, and creation of
Gross Domestic Product (GDP) (USEPA 2000d). Macroeconomic effects tend to be measurable in
nationwide econometric models only if the economic impact of the regulation reaches 0.25 percent to 0.5
percent of GDP. In 2003, real GDP was $10,321 billion (U.S. Department of Commerce BEA 2004a);
thus, a rule would have to cost at least $26 billion annually to have a measurable effect. A regulation
with a smaller aggregate effect is unlikely to have any measurable impact, unless it is highly focused on a
particular geographic region or economic sector. The LT2ESWTR should not have a measurable effect
on the national economy; the total annualized costs for the Preferred Regulatory Alternative range from
$93 to $133 million to $107 to $150 million using a 3 and 7 percent discount rate, respectively. Using
these annualized figures as a measure, the annual cost of the LT2ESWTR is an insignificant fraction of a
$26 billion annual cost that would be considered a measurable macroeconomic impact. Thus, annualized
LT2ESWTR costs measured as a percentage of the national GDP will only decline overtime as GDP
grows.
7.7.4 Consultation with Small Governments
Before the Agency establishes any regulatory requirements that may significantly or uniquely
affect small governments, including Tribal governments, it must have developed, under Section 203 of
UMRA, a small government agency plan. The plan must provide for the notification of potentially
Economic Analysis for the LT2ESWTR
7-21
December 2005
-------
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. EPA consulted with small governments to address impacts of regulatory requirements in
the LT2ESWTR that might significantly or uniquely affect small governments. A variety of stakeholders,
including small governments, were provided with several opportunities to participate early in the
regulatory development process, as described in section 7.2.
7.7.5 Consultation with State, Local, and Tribal Governments
Section 204 of UMRA requires the Agency to develop an effective process to permit elected
officers of State, Local, and Tribal governments (or their designated authorized employees) to provide
meaningful and timely input in the development of regulatory proposals that contain significant Federal
intergovernmental mandates. Consistent with these provisions, EPA held consultations with affected
governmental entities prior to proposal of the rule, as described in sections 7.2 and 7.8. EPA conducted
four outreach conference calls, discussed in section 7.2, and contacted each of the 12 Native American
Drinking Water State Revolving Fund Advisors to invite them, and representatives of their organizations,
to participate in the meetings. In addition to the conference calls, EPA presented the LT2ESWTR at
several health, environmental, and Native American conferences.
Representatives from State, Local, and Tribal governments were also involved in the
development of the Agreement in Principle, which was created early in the regulatory process. EPA
provided the Association of State Drinking Water Administrators (ASDWA) with an opportunity to
comment before officially proposing the LT2ESWTR. EPA accepted comments from ASDWA and other
Federal Advisory Committee Act (FACA) members, such as the National League of Cities (NLC), on a
draft of the LT2ESWTR posted on its web site and, to the extent possible, comments were incorporated
into the rule.
In addition to these efforts, EPA will educate, inform, and advise small systems, including those
run by small governments, about the LT2ESWTR requirements. The Agency is developing plain-English
guidance that will explain what actions a small entity must take to comply with the rule. Also, the
Agency has developed fact sheets that concisely describe various aspects and requirements of the
LT2ESWTR. Additional details on Tribal involvement in the rulemaking process can be found in section
7.8.
7.7.6 Regulatory Alternatives Considered
As required under Section 205 of UMRA, EPA considered several regulatory alternatives and
numerous methods to identify systems most at risk to microbial contamination. Chapter 3 provides a
detailed discussion of these alternatives. EPA chose the Preferred Regulatory Alternative because it
provided substantial benefits at an acceptable level of costs. In addition, the FACA Committee
recommended the Preferred Regulatory Alternative in the Stage 2 M-DBP Agreement in Principle.
7.7.7 Impacts on Small Governments
EPA has determined that this rule contains no regulatory requirements that might significantly or
uniquely affect small governments. As described in section 7.2, EPA has certified that this rule will not
have a significant economic impact on a substantial number of small entities. Estimated annual
expenditures by small systems for the LT2ESWTR range from $4.7 to $9.2 million at a 3 percent discount
rate. While the treatment requirements of the LT2ESWTR apply uniformly to both small and large
Economic Analysis for the LT2ESWTR 7-22 December 2005
-------
PWSs, large systems will bear most of the total costs of compliance with the rule. This is because large
systems treat a majority of the drinking water that originates from surface water sources.
7.8 Indian Tribal Governments
Executive Order 13175, 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 concluded that this rule may have Tribal implications, because it may impose substantial
direct compliance costs on Tribal governments. There are 93 Tribal water systems serving a population
of 82,216 (see Exhibit 7.7). As presented in Exhibit 7.9a, they will bear an annualized cost of $227,365,
at a 3 percent discount rate, to implement this rule ($334,265 at a 7 percent discount rate). Accordingly,
EPA provides a Tribal summary impact statement as required by Section 5(b) of Executive Order 13175.
The Federal government will not specifically provide the funds necessary to pay costs for Tribal systems
associated with the LT2ESWTR because EPA consulted with Tribal officials early in the process of
developing the regulation.
Economic Analysis for the LT2ESWTR 7-23 December 2005
-------
Exhibit 7.7: Numbers of Indian Tribal Public Water
Systems Using Surface Water Sources
System Size
(Population
Served)
Number of
Systems
Community Water Systems
< 100
101 -500
501 - 1000
1,001 -3,300
3,301 -10,000
16
40
11
13
4
Nontransient Noncommunity
Water Systems
< 100
101 -500
1,001 -3,300
2
1
1
Transient Noncommunity Water
Systems
< 100
101 -500
1,001 -3,300
10,001 -50,000
1
2
1
1
Source: EPA SDWIS Database, September 2004.
EPA consulted with Tribal officials in a variety of ways to permit them to have meaningful and
timely input into development of the LT2ESWTR. Tribes were able to have long-term input in the rule
by participating in the Federal Advisory Committee. During the Las Vegas EPA/Inter-Tribal Council of
Arizona in February 1999, a number of Tribal representatives requested that the All Indian Pueblo
Council (AIPC) representative be the FACA representative for Federal Tribes, given his knowledge of
drinking water systems. Approximately 20 Tribes are associated with the AIPC.
In addition to obtaining FACA Tribal input, EPA presented the LT2ESWTR at three conferences:
the 16th Annual Consumer Conference of the National Indian Health Board, the National Tribal
Environmental Council's Annual Conference in April 2000, and the EPA/Inter-Tribal Council of Arizona,
Inc. Tribal consultation meeting. Over 900 attendees representing Tribes from across the country
attended the National Indian Health Board's Consumer Conference, and representatives from over 100
Tribes attended the annual conference of the National Tribal Environmental Council. Finally,
representatives from 15 Tribes participated at the EPA/Inter-Tribal Council of Arizona meeting. At the
first two conferences, an EPA representative conducted two workshops on their drinking water program
Economic Analysis for the LT2ESWTR
7-24
December 2005
-------
and upcoming regulations, including the LT2ESWTR. The presentation materials and meeting summary
were sent to over 500 Tribes and Tribal organizations.
EPA distributed fact sheets describing the requirements of the LT2ESWTR and requested Tribal
input at an annual EPA Tribal meeting in San Francisco and a Native American Water Works Association
meeting in Scottsdale, Arizona. EPA also worked through its Regional Indian Coordinators and the
National Tribal Operations Committee to mail fact sheets on the LT2ESWTRto all of the Federally
recognized Tribes in November 2000.
After reviewing the fact sheets, a few Tribes requested more information and expressed concern
about having to implement too many regulations. They were also concerned about infrastructure costs
and the lack of funding attached to the rule. In response to a Tribal representative's comments, EPA
explained the health protection benefits associated with the LT2ESWTR, which some members of the
Tribal Caucus also noted. EPA directed Tribes to the Agreement in Principle on the EPA Web site for
more information.
On January 24, 2002, EPA held a teleconference for Tribal representatives as another step in
Tribal consultation. Prior to the teleconference, EPA send invitations to all Federally-recognized Tribes,
along with fact sheets explaining the rule. Twelve Tribal representatives and four regional Tribal
Program Coordinators attended the teleconference, requested further explanation of the rule, and
expressed concerns about funding sources. EPA explained that capital projects related to the rule would
rank high on lists of current funding sources due to the health risks associated with Cryptosporidium.
Tribes also called EPA after the teleconference to provide additional feedback.
Tribal Summary Impact Statement
EPA performed an analysis to estimate the impact of the LT2ESWTR on Tribal systems. EPA
has identified 93 Indian Tribal systems that might be subject to the LT2ESWTR. As seen in Exhibit 7.7,
all but one Tribal system is classified as small (serving 10,000 or fewer people).
EPA has estimated the costs for Indian Tribal systems to comply with the LT2ESWTR, based on
the assumption that the percentages of systems expected to incur costs for each size category will be the
same for Tribal systems as for systems nationwide. The costs for Tribal systems are calculated in two
steps. First, the number of Indian Tribal systems in each size category is multiplied by the percentage of
systems nationally in each size category expected to incur costs for various rule activities (e.g., E. coll
monitoring, Cryptosporidium monitoring, additional treatment). Second, the average cost of each rule
requirement is multiplied by the number of Tribal systems expected to incur costs.
Exhibit 7.8 shows the percentage of systems expected to incur costs in various categories based
on the ICR Cryptosporidium occurrence data set. For example, EPA projects that among systems serving
10,000 or fewer people, 34.3 percent would be triggered into monitoring for Cryptosporidium and 86.3
percent would conduct future E. coll monitoring. Among systems serving more than 10,000 people, 99.6
percent of plants would monitor for Cryptosporidium and 67.0 percent would conduct future E. coll
monitoring.
Exhibits 7.9a and b show annualized costs per system for various compliance activities including
implementation, monitoring, and treatment based on the ICR occurrence data set. Each cost has been
annualized at 3 percent over 25 years. Costs for individual systems would vary around these averages,
depending on the circumstances of the particular system. For example, E. coll monitoring costs would be
lower for those systems that could do the analysis onsite, as opposed to shipping samples to a commercial
laboratory, and only a fraction of small systems would be required to monitor for Cryptosporidium.
Economic Analysis for the LT2ESWTR 7-25 December 2005
-------
Many systems would incur no additional treatment costs, while a fraction would be required to provide
additional treatment as a result of their bin assignment.
EPA estimates that total mean annualized costs per system range from $1,956 to $3,521 for
CWSs, $1,947 to $2,390 for NTNCWSs, and $1,944 to $7,068 for TNCWSs. These costs result in an
estimated total annualized cost to Indian Tribes, including Primacy Agency costs, of $227,365 for the
LT2ESWTR, as shown in Exhibit 7.9a. (All costs in this paragraph are based on a 3-percent discount
rate.)
Exhibit 7.8: Number of Tribal Systems and Percent of Systems Nationally That
Will Incur Costs Due to the LT2ESWTR
System Size
(Population
Served)
<1 0,000
>10,000
Number of
Tribal
Systems
92
1
Percent of Systems Nationally Incurring Cost
Implemen-
tation
100.00%
100.00%
Initial
E.coli
Monitoring1
96.69%
98.10%
Initial
Crypto
Monitoring2
34.30%
99.64%
Future
£. co//
Monitoring3
86.34%
66.99%
Future
Crypto
Monitoring
30.05%
66.99%
Adding
Treatment4
34.92%
34.90%
1AII systems would be required to conduct £ co// monitoring, unless a system currently provides at least 5.5 log
Cryptosporidium treatment.
2Small systems are required to monitor for Cryptosporidium only if source water £ co// concentration exceeds the
trigger value. Based on the ICR occurrence data set, EPA estimates a maximum of 35 percent of systems that
monitor for £ co//will be triggered into Cryptosporidium monitoring.
3Systems would be required to conduct another round of monitoring 6 years after the initial bin assignment. This
monitoring would not be required for those systems that provide at least 5.5 log Cryptosporidium treatment.
4EPA estimates that 5.3 to 30.4 percent of all plants (including plants that purchase water) would incur costs for
additional treatment as a result of being assigned to Bins 2-4.
Note: For systems serving more than 10,000 people, the percentages represent the probability that one system/plant
will incur costs as a result of rule requirements.
Sources: Percentages derived from Exhibits 6.1 and 4.11. Percentages are assumed to be the same for Tribal
plants/systems as for those nationwide.
Economic Analysis for the LT2ESWTR
7-26
December 2005
-------
Exhibit 7.9a: Estimates of the Total Annualized Costs Incurred by Indian Tribal
Public Water Systems Due to the LT2ESWTR (Annualized at 3 Percent)
System Size
(Population Served)
Total CWS
^100
101-500
501-1,000
1,001-3,300
3,301-10,000(1]
Total NTNCWS
^100
101-500
1,001-3,300
Total TNCWS
^100
101-500
501-1,000
10,001-50,000
Number
of Tribal
Systems
A
84
16
40
11
13
4
4
2
1
1
5
1
2
1
1
Implementation
Cost per System
B
Monitoring
Cost per
System
C
Future
Monitoring
Cost per
System
D
Mean Cost
per System -
Treatment
E
Mean Cost
per System -
Total
F = B+C+D+E
$10
$11
$14
$14
$14
$909
$911
$916
$916
$917
$909
$911
$916
$916
$917
$129
$196
$363
$677
$1,674
$1 ,956
$2,029
$2,208
$2,522
$3,521
$10
$11
$14
$909
$911
$916
$909
$911
$916
$120
$182
$545
$1 ,947
$2,015
$2,390
$10
$11
$14
$14
$909
$911
$916
$876
$909
$911
$916
$876
$117
$174
$309
$5,303
$1 ,944
$2,007
$2,154
$7,068
Total Annualized Costs for All System Types
Annualized Costs for One Primacy Agency [1]
Total Annualized Tribal Costs
Estimated
Total Tribal
Costs
G = A*F
$183,626
$31 ,297
$81,166
$24,292
$32,785
$14,086
$8,300
$3,894
$2,015
$2,390
$15,180
$1 ,944
$4,014
$2,154
$7,068
$207,105
$20,260
$227,365
Economic Analysis for the LT2ESWTR
7-27
December 2005
-------
Exhibit 7.9b: Estimates of the Total Annualized Costs Incurred by Indian Tribal
Public Water Systems Due to the LT2ESWTR (Annualized at 7 Percent)
System Size
(Population Served)
Total CWS
^100
100-499
500-999
1 ,000-3,299
3,300-9,999(1]
Total NTNCWS
-,100
100-499
1 ,000-3,299
Total TNCWS
-,100
100-499
500-999
10,000-49,999
Number
of Tribal
Systems
A
84
16
40
11
13
4
4
2
1
1
5
1
2
1
1
Implementation
Cost per System
B
Monitoring
Cost per
System
C
Future
Monitoring
Cost per
System
D
Mean Cost
per System -
Treatment
E
Mean Cost
per System -
Total
F = B+C+D+E
$15
$16
$21
$21
$21
$1 ,358
$1 ,362
$1 ,368
$1 ,368
$1 ,370
$1 ,358
$1 ,362
$1 ,368
$1 ,368
$1 ,370
$192
$293
$542
$1,011
$2,502
$2,923
$3,032
$3,300
$3,768
$5,262
$15
$16
$21
$1 ,358
$1 ,362
$1 ,368
$1 ,358
$1 ,362
$1 ,368
$179
$272
$814
$2,910
$3,011
$3,571
$15
$16
$21
$21
$1 ,358
$1 ,362
$1 ,368
$1 ,368
$1 ,358
$1 ,362
$1 ,368
$1 ,368
$174
$260
$461
$7,923
$2,905
$2,999
$3,218
$10,681
Total Annualized Costs for All System Types
Annualized Costs for One Primacy Agency [1]
Total Annualized Tribal Costs
Estimated
Total Tribal
Costs
G = A*F
$274,379
$46,765
$121,280
$36,298
$48,989
$21 ,047
$12,402
$5,819
$3,011
$3,571
$22,802
$2,905
$5,998
$3,218
$10,681
$309,583
$24,682
$334,265
1 The Primacy Agency cost for the one Tribe that has primacy is based on the average costs to State primacy
agencies (total costs for the Preferred Alternative in Exhibits O.16a and O.19a divided by total number of primacy
agencies (57)).
Note: Detail may not add to total due to independent rounding.
Sources:
[A] Number and categories of Tribal systems taken from SDWIS Database, September 2004.
[B] Implementation costs taken from Appendix D, Exhibit D.10, annualized, then divided by the total number of
systems to calculate costs per system.
[C] Monitoring costs taken from Appendix D, Exhibits D.12, D.14, and D.16, annualized, then divided by the total
number of systems to calculate costs per system.
[D] Future monitoring costs taken from Appendix D, Exhibits D.32, D.35, and D.38, annualized, then divided by the
total number of systems to calculate costs per system.
[E] System costs taken from Appendix H, Exhibit H. 1, annualized, then divided by the total number of systems to
calculate costs per system.
7.9 Impacts on Sensitive Subpopulations
EPA's Office of Water has historically considered risks to sensitive subpopulations, including
children) in establishing drinking water assessments, advisories or other guidance, and standards.
Maximizing health protection for sensitive subpopulations requires balancing risks to achieve the
recognized benefits of controlling waterborne pathogens while minimizing risk of potential DBF toxicity.
A primary purpose of LT2ESWTR is to improve control of microbial pathogens, specifically the
protozoan Cryptosporidium. The health effect of cryptosporidiosis on sensitive subpopulations is much
more severe and debilitating than on the general population. These sensitive subpopulations include the
young, the elderly (especially those weakened by other conditions), the malnourished and disease-
impaired (especially those with diabetes), and a broad category of those with compromised immune
systems, such as Acquired Immune Deficiency Syndrome (AIDS) patients, people with lupus or cystic
fibrosis, transplant recipients, and individuals undergoing chemotherapy (Rose 1997). The Agency has
evaluated several regulatory alternatives and selected the alternative that balances cost with providing
significant benefits. It should be noted that the Stage 2 DBPR, which is being promulgated concurrently
Economic Analysis for the LT2ESWTR
7-28
December 2005
-------
with this rule, reduces DBF concentrations in drinking water and achieves the goal of increasing the
protection of children.
Research outlining the potential health benefits of the LT2ESWTRto both sensitive subpopu-
lations and the general public is discussed in greater detail in the Occurrence and Exposure Assessment
for the Long Term 2 Enhanced Surface Water Treatment Rule (USEPA 2003c) as well as in Chapter 5 of
this EA.
7.9.1 Impacts on the Immunocompromised
As stated in Chapter 5, mortality as a result of cryptosporidiosis is a much greater risk for
sensitive subpopulations, particularly for immunocompromised individuals, than it is for the general
population. The duration and severity of the disease are significant: whereas the disease may hospitalize
1 percent of the immunocompetent population with very little risk of mortality among those hospitalized
(<0.001), Cryptosporidium infections are associated with a high rate of mortality in the
immunocompromised (50 percent) (Rose 1997).
The duration of cryptosporidiosis in those with compromised immune systems is considerably
longer than in those with competent immune systems, with more severe symptoms often requiring lengthy
hospital stays. For subpopulations that are immunocompromised, the cost of illness (COI) from
cryptosporidiosis would be much greater than that for the general population. During a 1993 outbreak in
Milwaukee, 33 AIDS patients with cryptosporidiosis accounted for 400 hospital-days at a cost of nearly
$760,000 (Rose 1997). COI due to these hospital-days alone is estimated at $23,000 per patient
($760,000/33 patients). In addition, of the 54 deaths in the Milwaukee outbreak, 46 were AIDS patients.
Based on the severity of illness and the high costs of treatment experienced by the
immunocompromised as a result of Cryptosporidium infection, the Agency expects LT2ESWTR to have a
disproportionately positive impact on all sensitive subpopulations mentioned earlier.
7.9.2 Protection of Children from Environmental Health Risks and Safety Risks
Executive Order 13045 (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 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 alternatives considered by
the Agency.
This final rule subject to the Executive Order because it is economically significant as defined in
Executive Order 12866. As a matter of policy, EPA has examined the environmental health effects of
Cryptosporidium on children. The risk of illness and death due to cryptosporidiosis depends on several
factors including age, nutrition, exposure, genetic variability, disease, and the immune status of the
individual.
Young children are of particular concern since they are more susceptible than adults to
cryptosporidiosis, and the risk of mortality resulting from diarrhea is greatest in the very young and
elderly (Rose 1997; Gerba et al. 1996; Payer and Ungar 1986). Based on data presented in the
Occurrence and Exposure Assessment for the Long Term 2 Enhanced Surface Water Treatment Rule
Economic Analysis for the LT2ESWTR 7-29 December 2005
-------
(USEPA 2003c), children under 5 years of age make up approximately 6.9 percent of the total population
served by surface water and GWUDI sources. An infected child can also spread the disease to other
children or family members, and evidence of such secondary transmission of cryptosporidiosis has been
found in a number of outbreak investigations (Casemore 1990; Cordell et al. 1997; Frost et al. 1997).
During the 1993 Milwaukee drinking water outbreak, associated mortalities in children were
reported. Also, children with laboratory-confirmed cryptosporidiosis were more likely to have an
underlying disease that altered their immune status (Cicirello et al., 1997). In that study, the observed
association between increasing age of children and increased numbers of laboratory-confirmed
cryptosporidiosis suggested to the authors that the data are consistent with increased tap water
consumption of older children. However, due to data limitations, this observation could not be
adequately analyzed.
Chapell et al. (1999) found that prior exposure to Cryptosporidium through the ingestion of a low
oocyst dose provides limited protection from infection and illness. It is not known, however, whether this
immunity is life-long or temporary. Data also indicate that either mothers confer short-term immunity to
their children or that babies have reduced exposure to Cryptosporidium, resulting in a decreased incidence
of infection during the first year of life. For example, in a survey of over 30,000 stool sample analyses
from different patients in the United Kingdom, the 1-5 year age group suffered a much higher infection
rate than individuals less than 1 year of age. For children under 1 year of age, those older than 6 months
showed a higher rate of infection than individuals aged fewer than 6 months (Casemore 1990). Children
may be in contact with environmentally contaminated surfaces and may not have established immunity
against the disease (DuPont et al. 1995). Although children consume less water than adults and would,
therefore, be expected to have less exposure to Cryptosporidium oocycsts, studies show that
cryptosporidiosis occurrence is greater in children (Casemore 1990).
EPA has not been able to quantify the health effects for children as a result of Cryptosporidium-
contaminated drinking water. However, the result of the LT2ESWTR will be a reduction in the risk of
illness for the entire population, including children. Because available evidence indicates that children
may be more vulnerable to cryptosporidiosis than the rest of the population, the LT2ESWTR would,
therefore, result in greater risk reduction for children than for the general population.
7.10 Environmental Justice
Executive Order 12898 (59 FR 7629) establishes a Federal policy for incorporating
environmental justice into Federal agency missions by directing agencies to identify and address
disproportionately high adverse human health or environmental effects of its programs, policies, and
activities on minority and low-income populations. The Agency has considered environmental justice-
related issues concerning the potential impacts of this action and consulted with minority and low-income
stakeholders.
Two aspects of the LT2ESWTR comply with the order that requires the Agency to consider
environmental justice issues in the rulemaking and to consult with stakeholders representing a variety of
economic and ethnic backgrounds. These are: (1) the overall nature of the rule, and (2) the convening of
a stakeholder meeting specifically to address environmental justice issues.
The Agency built on the efforts conducted during the development of the Interim Enhanced
Surface Water Treatment Rule (IESWTR) to comply with Executive Order 12898. On March 12, 1998,
the Agency held a stakeholder meeting to address various components of pending drinking water
regulations and how they might impact sensitive subpopulations, minority populations, and low-income
Economic Analysis for the LT2ESWTR 7-30 December 2005
-------
populations. This meeting was a continuation of stakeholder meetings that started in 1995 to obtain input
on the Agency's Drinking Water Programs. Topics discussed included treatment techniques, costs and
benefits, data quality, health effects, and the regulatory process. Participants were national, State, Tribal,
municipal, and individual stakeholders. EPA conducted the meeting by video conference call among 11
cities. The major objectives for the March 12, 1998, meeting were the following:
• To solicit ideas from stakeholders on known issues concerning current drinking water
regulatory efforts.
• To identify key areas of concern to stakeholders.
• To receive suggestions from stakeholders concerning ways to increase representation of
communities in OGWDW regulatory efforts.
In addition, EPA developed a plain-English guide for this meeting to assist stakeholders in
understanding the multiple and sometimes complex issues surrounding drinking water regulations.
The LT2ESWTR and other drinking water regulations promulgated or under development are
expected to have a positive effect on human health regardless of the social or economic status of a
specific population. The LT2ESWTR serves to provide a similar level of drinking water protection to all
groups. Where water systems have high Cryptosporidium levels, they must treat their water to achieve a
given level of protection. Further, to the extent that levels of Cryptosporidium in drinking water might be
disproportionately high now among minority or low-income populations (which is unknown), the
LT2ESWTR will work to remove those differences. Thus, the LT2ESWTR meets the intent of Federal
policy requiring incorporation of environmental justice into Federal agency missions.
The LT2ESWTR applies uniformly to CWSs, NTNCWSs, and TNCWSs that use surface water
or GWUDI as their source. Consequently, this rule provides health protection from pathogen exposure
equally to all income and minority groups served by surface water and GWUDI systems.
7.11 Federalism
Executive Order 13132, "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 and the States, or on the distribution of
power and responsibilities among the various levels of government."
Under Section 6(b) Executive Order 13132, EPA may not issue a regulation that has Federalism
implications, imposes substantial direct compliance costs, and is not required by statute, unless the
Federal government provides the funds necessary to pay the direct compliance costs incurred by State and
Local governments, or consults with State and Local officials early in the process of developing the
regulation.
EPA has concluded that the LT2ESWTR may have Federalism implications because it will
impose substantial direct compliance costs on State or Local governments. It contains a significant
intergovernmental mandate under UMRA Section 202, i.e., it is likely to result in expenditure by State,
Local, and Tribal governments in the aggregate of $100 million or more in any one year. The aggregate
cost to these entities ranges from $93 to $133 million on average annually at a 3 percent discount rate.
Economic Analysis for the LT2ESWTR 7-31 December 2005
-------
Accordingly, EPA provides the following Federalism summary impact statement, as required by Section
6(b) of Executive Order 13132.
EPA consulted with State and Local officials early in the process of developing the LT2ESWTR
to permit them to have meaningful and timely input into its development. On February 20, 2001, EPA
held a dialogue with representatives of State and Local governmental organizations including those that
represent elected officials. Representatives from the following organizations attended the consultation
meeting: ASDWA, the National League of Cities (NLC), the National Governors' Association (NGA),
the National Conference of State Legislatures (NCSL), the International City/County Management
Association (ICMA), NLC, the County Executives of America, and health departments. Attendees posed
questions ranging from a basic inquiry into how Cryptosporidium gets into water to more detailed queries
about anticipated implementation guidance, procedures, and schedules. Some of the State and Local
organizations that attended this meeting were also participants in the Stage 2 M-DBP Federal Advisory
Committee and signed the Agreement in Principle. In addition, EPA consulted with a mayor in the
SBREFA consultation. EPA considered all input from these consultations in the development of the
LT2ESWTR.
7.12 Actions Concerning Regulations That Significantly Affect Energy Supply,
Distribution, or Use
Executive Order 13211, "Actions Concerning Regulations That Significantly Affect Energy
Supply, Distribution, or Use" (66 FR 28355; May 22, 2001), 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 LT2ESWTR has not been designated by the Administrator of the Office of Information and
Regulatory Affairs as a significant energy action because it is not likely to have a significant adverse
effect on the supply, distribution, or use of energy. This determination is based on the analysis presented
below.
Energy Supply
The first consideration is whether the LT2ESWTR would adversely affect the supply of energy.
The LT2ESWTR does not regulate power generation, either directly or indirectly, and the public and
private PWSs that the LT2ESWTR regulates do not, as a rule, generate power. Further, the cost increases
borne by customers of PWSs as a result of the LT2ESWTR are a low percentage of the total cost of water,
except for a very few small systems that will need to spread the cost of installing advanced technologies
over a narrow customer base. Therefore, those customers that are power generation utilities are unlikely
to face any significant effects as a result of the LT2ESWTR. In summary, the LT2ESWTR does not
regulate the supply of energy, does not generally regulate the utilities that supply energy, and is unlikely
significantly to affect the customer base of energy suppliers. Thus, the LT2ESWTR would not would not
adversely affect the supply of energy.
Economic Analysis for the LT2ESWTR 7-32 December 2005
-------
In response to the LT2ESWTR, some water utilities are expected to increase their energy use, and
those impacts are discussed later in this section.
Energy Distribution
The second consideration is whether the LT2ESWTR would adversely affect the distribution of
energy. The LT2ESWTR does not regulate any aspect of energy distribution. PWSs that are regulated by
the LT2ESWTR already have electrical service. The rule is projected to increase peak electricity demand
at PWSs by only 0.024 percent (see below). Therefore, EPA assumes that the existing connections are
adequate and that the LT2ESWTR has no discernable adverse effect on energy distribution.
Energy Use
The third consideration is whether the LT2ESWTR would adversely affect the use of energy.
Because some PWSs are expected to add treatment technologies that use electrical power, this potential
impact of the LT2ESWTR on the use of energy requires further evaluation. The analyses that underlay
the estimation of costs in Chapter 6 are national in scope and do not identify specific plants or systems
that may install treatment in response to the LT2ESWTR. As a result, no analysis of the effect on specific
energy suppliers is possible with the available data. The approach used to estimate the impact of energy
use, therefore, also focuses on national-level impacts. It estimates the additional energy use due to the
LT2ESWTR and compares that to the national levels of power generation in terms of average and peak
loads.
The first step is to estimate the energy used by the technologies expected to be installed as a result
of the LT2ESWTR. Energy use is not directly estimated in Technologies and Costs Document (USEPA
2003a), but the annual cost of energy for each technology addition or upgrade necessitated by the
LT2ESWTR is provided. An estimate of plant-level energy use is derived by dividing the total energy
cost per plant for a range of flows by an average national cost of electricity of $0.076 per kilowatt hour
per year (kWh/y) (U.S. DOE EIA 2004a3). The energy use per plant for each flow range and technology
is then multiplied by the number of plants predicted to install each technology in a given flow range. The
energy requirements for each flow range are then added to produce a national total. No electricity use is
subtracted to account for the technologies that may be replaced by new technologies, resulting in a
conservative estimate of the increase in energy use. The results of the analysis are shown in Exhibit 7.10
for the ICR, ICRSSL, and ICRSSM Cryptosporidium occurrence data sets. The incremental national
annual energy usage is estimated at 165,552 megawatt hours (MWh) for the ICR occurrence data sets.
Exhibit 7.11 provides a sample calculation for UV showing the increase in energy usage as a
result of the LT2ESWTR.
To determine if the additional energy required for systems to comply with the rule would have a
significant adverse effect on the use of energy, the numbers in Exhibit 7.10 are compared to the national
production figures for electricity. According to the U.S. Department of Energy's Information
Administration, electricity producers generated 3,848 million MWh of electricity in 2003 (USDOE EIA
3 EPA is aware that DOE has updated its 2003 average national cost of electricity per kilowatt hour per year
from $0.076 to $0.074. However, EPA continues to use the $0.076 value to maintain consistency with the
Technologies and Cost Document.
Economic Analysis for the LT2ESWTR 7-33 December 2005
-------
2004b4). Therefore, even using the highest assumed energy use for the LT2ESWTR (165,551,898
kWh/y), the rule would result in only a 0.004 percent increase in annual average energy use when fully
implemented. This calculation is shown below:
1. 165,551,898 kWh/y * (mWh/1,000 kWh) = 165,552 mWh/y
2. 165,552 MWh/y - 3,848,000,000 MWh/y * 100 = 0.004%
Exhibit 7.10: Total Increased Annual National Energy Usage Attributable to
the LT2ESWTR
Technology
UV
03 (0.5 log)
03 (1.0 log)
03 (2.0 log)
ME/UF
Bag Filters
Cartridge Filters
Total
Plants Selecting
Technology
A
1,038
27
18
14
37
1,523
209
2,867
Total Annual
Energy Required
(kWh/yr)
B
100,829,791
20,617,993
18,827,749
16,245,643
7,343,320
1,605,380
82,022
165,551,898
Sources: [A] Plants selecting technology taken from Appendix G and Appendix I for uncovered finished
water reservoirs predicted to disinfect instead of covering.
[B] Energy costs derived from the Technologies and Costs Document (USEPA 2003a). Energy costs
were converted to energy usage by dividing the costs by the unit costs for energy listed in Table 4-3 of
the Technologies and Costs Document. Energy usage is different for different size categories; the
average per plant is the weighted average for all plants selecting the technology.
4 EPA is aware that DOE has updated its estimate of total electricity produced in 2003 from 3,848 million
to 3,883 million. However, EPA continues to use the 3,848 million estimate to maintain consistency with related
electricity estimates used in this EA and the Technologies and Cost Document.
Economic Analysis for the LT2ESWTR
7-34
December 2005
-------
Exhibit 7.11: Sample Calculation for Determining Increase in Energy Usage
(Plants Predicted to Add UV)
System Size
(population served)
Average
Daily Flow
Flow per
Plant (MGD)
A
Total
Number of
Plants
B
Number of
Plants
Selecting
C
Annual Energy
Cost per Plant
($/plant/yr)
D
Annual Energy
Requirement
(kWhr/plant/yr)
E = D/$0.076 per
kWhr
Total Energy Usage
for Plants Selecting
(kWhr/year)
F=C*E
CWSs
<100
1 01 - 500
501 -1,000
1,001 -3,300
3,301 -10,000
10,001 -50,000
50,001 -100,000
100,001 - 1 Million
> 1 Million
0.01
0.03
0.08
0.24
0.73
2.81
7.34
26.49
98.62
377
771
462
1,128
1,143
1,198
330
388
66
16
37
23
111
112
361
100
114
20
$ 82
$ 207
$ 416
$ 991
$ 2,304
$ 4,772
$ 9,521
$ 23,618
$ 82,270
1,082
2,722
5,478
13,046
30,322
62,789
125,273
310,761
1,082,497
16,801
100,032
123,363
1,452,608
3,391 ,442
22,644,385
12,517,834
35,455,004
22,004,695
NTNCWSs
<100
1 01 - 500
501 -1,000
1,001 -3,300
3,301 -10,000
10,001 -50,000
50,001 -100,000
100,001 -1 Million
> 1 Million
0.01
0.03
0.08
0.20
0.63
3.33
-
22.94
-
226
312
106
91
25
5
0
1
0
9
13
5
8
2
1
0
0
0
$ 81
$ 191
$ 406
$ 861
$ 2,054
$ 5,317
$
$ 21,027
$
1,060
2,512
5,348
1 1 ,330
27,020
69,959
0
276,669
0
9,240
33,459
24,203
89,963
58,944
102,735
0
79,596
0
TNCWSs
<100
1 01 - 500
501 -1,000
1,001 -3,300
3,301 -10,000
10,001 -50,000
50,001 -100,000
100,001 -1 Million
> 1 Million
TOTALS
0.00
0.02
0.08
0.22
0.59
2.38
-
19.38
871 .95
1,273
610
107
67
19
12
0
1
2
8,720
60
29
5
6
2
4
0
0
1
1,038
$ 71
$ 169
$ 398
$ 907
$ 1 ,948
$ 4,325
$
$ 18,436
$ 283,182
936
2,226
5,236
1 1 ,938
25,638
56,909
0
242,581
3,726,077
6,099,407
56,471
64,366
26,557
73,807
44,951
204,313
0
71,091
2,183,930
100,829,791
Notes: Detail may not add due to independent rounding.
Sources: [A] The flows are taken from Exhibit 4.4a.
[B] The baseline numbers of filtered plants are taken from Exhibit 4.3, and added to the number of
unfiltered plants, taken from Exhibit 4.5.
[C] Numbers of plants selecting UV are taken from Appendix G (Exhibits G.31-G.33).
[D] The electricity cost per plant is taken from the Technologies and Costs Document (USEPA 2003a).
[E] Electricity cost is $0.076/KWh, as presented in the Technologies and Costs Document (USEPA
2003a).
In addition to average energy use, the impact at times of peak demand is important. To examine
whether increased energy usage might significantly affect the capacity margins of energy suppliers, their
peak season generating capacity reserve was compared to an estimate of peak incremental power demand
by water utilities. Both energy use and water use peak in the summer months, so the most significant
effects on supply would be seen then. In the summer of 2003, U.S. generation capacity exceeded
Economic Analysis for the LT2ESWTR
7-35
December 2005
-------
consumption by 15 percent, or approximately 160,000 MW (USDOE EIA 2004b5). Assuming around-
the-clock operation of water treatment plants, the total energy requirement can be divided by 8,760 hours
per year to obtain an average power demand of 19 MW for the modeled ICR occurrence distribution.
Assuming that power demand is proportional to water flow through the plant and that peak flow can be as
high as twice the average daily flow during the summer months, about 38 MW could be needed to operate
treatment technologies installed to comply with the LT2ESWTR. This is only 0.024 percent of the
capacity margin available at peak use. This calculation is presented below:
1. 165,551,898 kWh/y * (y/8,760 hr) * (MW/1,000 kW)*2 = 38 MW
2. 38 MW - 160,000 MW * 100 = 0.024%
Although EPA recognizes that not all regions have a 15 percent capacity margin and that this
margin varies across regions and over time, this analysis reflects the effect of the rule on national energy
supply, distribution, and use. While certain areas, notably California, have experienced shortfalls in
generating capacity in the recent past, a peak incremental power requirement of 38 MW nationwide is not
likely to significantly change the energy supply, distribution, or use in any given area.
Conclusion
The LT2ESWTR is not a "significant energy action" as defined in Executive Order 13211,
"Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution, or Use" (66 FR
28355 May 22, 2001) because it is not likely to have a significant adverse effect on the supply,
distribution, or use of energy (based on annual average use and conditions of peak power demand).
Therefore, a statement of Energy Effects for the LT2ESWTR has not been prepared.
The total increase in energy usage by water systems as a result of the LT2ESWTR is predicted to
be approximately 166 million kWh/y, which is less than five one-thousandths of 1 percent of the total
energy produced in 2003. While the rule may have some adverse energy effects, EPA does not believe
that this constitutes a significant adverse effect on the energy supply.
5 EPA is aware that DOE has updated its estimate of capacity exceeding consumption in the summer of
2003 from 160,000 to 159,000 MW. However, EPA continues to use the estimate of 160,000 MW to maintain
consistency with related electricity estimates used in this EA and the Technologies and Cost Document.
Economic Analysis for the LT2ESWTR 7-36 December 2005
-------
8. Comparison of Benefits and Costs of the LT2ESWTR
8.1 Introduction
This chapter presents several comparisons of the benefits and costs of the LT2ESWTR. Section
8.2 focuses on the benefits and cost comparisons of the Preferred Regulatory Alternative. Section 8.3
presents several analyses comparing the benefits and costs of all four regulatory alternatives, including
multiple measures of cost-effectiveness in Section 8.3.2. The effect of uncertainties on these estimates is
discussed in section 8.4. Finally, section 8.5 presents a summary of the conclusions from these analyses.
For comparison purposes, this chapter sometimes presents only mean estimates of benefits and
costs. These estimates are discussed in Chapters 5 and 6, respectively. To avoid repetition, the following
discussion assumes the reader is familiar with those chapters, the data sets used, and the analyses
performed. The remaining sections of this chapter are organized as follows.
8.2 Summary of National Benefits, Costs, and Net Benefits of the Preferred Regulatory
Alternative
8.2.1 National Benefits Summary
8.2.2 National Cost Summary
8.2.3 National Net Benefits
8.3 Comparison of Regulatory Alternatives
8.3.1 Comparison of Benefits and Costs
8.3.2 Cost-Effectiveness Measures
8.4 Effect of Uncertainties on the Benefit-Cost Comparisons
8.5 Summary of Benefit and Cost Comparisons
8.2 Summary of National Benefits, Costs, and Net Benefits of the Preferred Regulatory
Alternative
This section summarizes national benefits, costs, and net benefits of the Preferred Regulatory
Alternative for the LT2ESWTR.
The rule will be implemented over time, not instantaneously, and therefore, the treatment costs
incurred and benefits realized by the affected systems and population they serve will vary by year.
Exhibits 8.la and 8.1b summarize the undiscounted benefit and cost estimates incurred by systems,
according to the implementation schedule (presented in Appendix 0, Exhibit 0.7-0.9), over the 25 year
period analyzed in this EA. Exhibit 8.la shows benefits calculated using the Enhanced cost of illness
(COI) and 8.1b shows benefits calculated using the Traditional COI (section 5.3 explains the two COI
values).
The analyses in this EA assume that implementation of this rule will begin in 2005. If
implementation of the rule began a year or two later, the fundamental conclusions of the analysis would
not be significantly changed. In the first several years, before systems have installed treatment, no
benefits are realized, but costs are incurred for rule implementation and monitoring. Once systems begin
to install treatment, illnesses and deaths are projected to be avoided in the year following installation of
treatment and for each year thereafter. By 2015, all treatment is projected to have been installed;
Economic Analysis for the LT2ESWTR 8-1 December 2005
-------
therefore, starting in 2016, the level of illnesses and deaths avoided annually is estimated to be constant
for the rest of the period of analysis (through 2029).
Exhibit 8.1a: Summary of Undiscounted Benefit and Cost Estimates by Year
Incurred, Preferred Alternative, ICR Data Set, Enhanced COI1
(Millions$, 2003$)
Year
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
Systems <1 0,000
Benefits
A
$
$
$
$
$
$
$
$ 13
$ 39
$ 66
$ 106
$ 121
$ 136
$ 138
$ 139
$ 141
$ 142
$ 144
$ 146
$ 147
$ 149
$ 151
$ 153
$ 154
$ 156
Cost
B
$ 0.2
$ 0.2
$ 6
$ 8
$ 14
$ 14
$ 16
$ 29
$ 45
$ 47
$ 21
$ 8
$ 21
$ 22
$ 23
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
Systems > 10,000
Benefits
c
$
$
$
$
$ 218
$ 677
$ 1,394
$ 1,927
$ 2,464
$ 2,543
$ 2,590
$ 2,618
$ 2,646
$ 2,675
$ 2,705
$ 2,735
$ 2,765
$ 2,796
$ 2,827
$ 2,859
$ 2,891
$ 2,923
$ 2,956
$ 2,989
$ 3,023
Cost
D
$ 44
$ 49
$ 54
$ 143
$ 292
$ 470
$ 391
$ 391
$ 124
$ 80
$ 55
$ 54
$ 51
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
All systems
Benefits
E=A+C
$
$
$
$
$ 218
$ 677
$ 1 ,394
$ 1 ,940
$ 2,503
$ 2,609
$ 2,696
$ 2,739
$ 2,782
$ 2,813
$ 2,844
$ 2,876
$ 2,908
$ 2,940
$ 2,973
$ 3,006
$ 3,040
$ 3,074
$ 3,109
$ 3,144
$ 3,179
Cost
F=B+D
$ 44
$ 49
$ 59
$ 151
$ 307
$ 484
$ 407
$ 420
$ 170
$ 127
$ 76
$ 62
$ 71
$ 69
$ 71
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
Note: Benefits increase each year due to the phasing in of installed treatment and the increases in
real income growth, which directly affects COI and, though adjustments for income elasticity, also
affects the VSL.
Economic Analysis for the LT2ESWTR
8-2
December 2005
-------
Exhibit 8.1 b: Summary of Undiscounted Benefit and Cost Estimates by Year
Incurred, Preferred Alternative, ICR Data Set, Traditional COI1
(Millions$, 2003$)
Year
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
Systems <1 0,000
Benefits
A
$
$
$
$
$
$
$
$ 9
$ 27
$ 46
$ 74
$ 84
$ 94
$ 95
$ 96
$ 97
$ 98
$ 98
$ 99
$ 100
$ 101
$ 102
$ 103
$ 104
$ 105
Cost
B
$ 0.2
$ 0.2
$ 6
$ 8
$ 14
$ 14
$ 16
$ 29
$ 45
$ 47
$ 21
$ 8
$ 21
$ 22
$ 23
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
$ 9
Systems
Benefits
c
$
$
$
$
$ 162
$ 503
$ 1,032
$ 1,422
$ 1,814
$ 1,867
$ 1,896
$ 1,913
$ 1,930
$ 1,947
$ 1,964
$ 1,981
$ 1,999
$ 2,016
$ 2,034
$ 2,052
$ 2,071
$ 2,089
$ 2,108
$ 2,126
$ 2,145
> 10,000
Cost
D
$ 44
$ 49
$ 54
$ 143
$ 292
$ 470
$ 391
$ 391
$ 124
$ 80
$ 55
$ 54
$ 51
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
$ 48
All systems
Benefits
E=A+C
$
$
$
$
$ 162
$ 503
$ 1 ,032
$ 1,431
$ 1,841
$ 1,913
$ 1,970
$ 1,997
$ 2,024
$ 2,042
$ 2,060
$ 2,078
$ 2,096
$ 2,115
$ 2,134
$ 2,153
$ 2,172
$ 2,191
$ 2,211
$ 2,230
$ 2,250
Cost
F=B+D
$ 44
$ 49
$ 59
$ 151
$ 307
$ 484
$ 407
$ 420
$ 170
$ 127
$ 76
$ 62
$ 71
$ 69
$ 71
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
$ 57
Note: 1The Traditional COI only includes valuation for medical costs and lost work time (including
some portion of unpaid household production). The Enhanced COI also factors in valuations for lost
personal time (non-work time) such as childcare and homemaking (to the extent not covered by the
traditional COI), time with family, and recreation, and lost productivity at work on days when workers
are ill but go to work anyway.
Sources:
[A] and [C] Exhibit C.17a (Enhanced) and C.17b (Traditional)
[B] and [D] Undiscounted State (O.4a), implementation and monitoring (O.5a), and treatment (O.6a)
multiplied by Schedules (O.7-O.9)
Economic Analysis for the LT2ESWTR
8-3
December 2005
-------
8.2.1 National Benefits Summary
The quantified benefits of the LT2ESWTR derive from the reduction in the incidence of adverse
health effects, specifically the endemic morbidity and mortality from cryptosporidiosis, attributable to
consumption of drinking water from the PWSs affected by the rule. However, the value of other
additional benefits that cannot be quantified (and therefore cannot be reflected explicitly in quantitative
benefit-cost comparisons) are likely to be substantial. Exhibit 8.2 summarizes the nonqualified benefits
and Chapter 5 (section 5.6.6) describes them in more detail. In every comparison of benefits and costs,
these real, but nonquantified, benefits must be considered in addition to those that are quantified.
Exhibit 8.2: Summary of Nonquantified Benefits and Groups Affected
Type of Benefit
Nonquantified Benefits
Group(s) Affected
Reduction in risk of illness to sensitive
subpopulations (although mortality for
those with AIDS and other although
sensitive subpopulations has been
included)
Immunocompromised individuals
served by systems that make
changes to or add treatment.
Reduction in health risk during outbreaks
(and in related response costs)
Health related
Reduction in co-occurring/emerging
pathogen risk
All individuals served by systems that
make changes to or add treatment,
including those now served by
uncovered finished water reservoirs,
(between 34 and 55 million people).
Reduction in endemic morbidity and
mortality risk associated with uncovered
finished water reservoirs
All individuals receiving water from
uncovered finished water reservoirs.
Reduction in health risks from certain
DBFs
All individuals served by systems that
install physical disinfection
technologies like membranes or UV1
Improved aesthetic water quality
All individuals served by systems that
make changes to or add treatment
that is likely to reduce taste and odor
problems (e.g., ozone).
Nonhealth related
Reduced costs of risk-averting behaviors
Consumers in systems that cease
using uncovered finished water
reservoirs (through covering or taking
such reservoirs off-line) my have
greater confidence in water quality.
This may result in less averting
behavior that reduces both out-of-
pocket costs (e.g., purchase of bottled
water) and opportunity costs (e.g.,
time to boil water).
1 Systems that install chemical disinfection technologies like ozone may increase certain DBPs.
Economic Analysis for the LT2ESWTR
December 2005
-------
Before reviewing benefits, some reminders are necessary to understand the graphs and tables
that follow in this chapter. First, quantified benefits have been calculated separately for each of the three
occurrence data sets, two COI values, and two discount rates. Second, the underlying results are in the
form of distributions of numbers. These values are obtained from a two-dimensional Monte Carlo
simulation, the dimensions being variability and uncertainty. To give a sense of the extent of the
distributions, data are summarized using central tendency estimates and usually some bounding
information. For example, in many exhibits, data are expressed using mean values, which is based on the
5th and 95th percentiles of a distribution of possible values resulting from the Monte Carlo simulation. The
mean value represents the best estimate of benefits, and the 5th and 95th percentiles capture the 90
percent confidence interval of the distribution reflecting uncertainty in that best estimate.
Exhibit 8.3 shows two kinds of estimates. One is the average number of cases avoided annually
once the rule is fully implemented (i.e., when all treatment changes are operational) and the second is the
annual average over the 25 year period. The second set of estimates is lower because the rule is
implemented over time, not instantly. This figure is the simple average of the illnesses and deaths avoided
over the first 25 years, including years in which no treatment is yet installed, years in which varying
percentages of treatment have been installed, and years at full implementation.
Exhibit 8.3: Summary of Annual Avoided Illnesses and Deaths,
Preferred Alternative
Data Set
Annual Illnesses
Avoided
Mean
90 % Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annual Deaths
Avoided
Mean
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annual Total after Full Implementation
ICR
ICRSSL
ICRSSM
964,360
230,730
455,170
149,241
38,281
72,128
2,277,367
521,925
1,112,374
207
52
100
34
9
17
468
113
230
Annual Average over 25 years
ICR
ICRSSL
ICRSSM
712,732
170,977
336,652
109,486
28,314
52,763
1,685,176
392,979
826,004
154
39
74
25
7
12
348
85
172
Sources: Appendix C, Exhibit C.4 and C.5, Columns A-F. Annual Average derived from
Exhibits C.4 (and C.5) and Exhibit O.9 (O&M schedules).
Exhibits 8.4a and 8.4b monetize these avoided illness and death estimates and present their
annualized values using 3 and 7 percent discount rates and Enhanced and Traditional COI values. The
calculation also includes factors for income growth and income elasticity that vary by year. The COI
values are applied to cases of illness avoided, and a distribution of undiscounted estimates for Value of a
Statistical Life (VSL) (represent uncertainty in the estimate) is applied to deaths avoided. The initial value
for COI is a constant, but its value of lost time in any year is adjusted to reflect growth in real income as
described in Chapter 5. The VSL also considers the long-term growth in real income and includes a
portion of that growth using an income elasticity analysis, as described in Chapter 5. The data in Exhibit
8.4 represent the monetized values from each year, discounted to the year 2003 (to obtain present values)
Economic Analysis for the LT2ESWTR
8-5
December 2005
-------
and then annualized over the 25 year period. These figures represent the annualized value of the
estimated annual number of illnesses and deaths avoided according to the rule schedule. (That is, they are
the annualized values of the undiscounted benefit data presented in Exhibit 8.1.)
Exhibit 8.4a: Summary of Quantified Benefits,
Preferred Alternative—Enhanced Cost of Illness1
Data Set
Value of
Mean
Benefits — Enhanced COI1
$ Millions, 2003$)
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annualized Value (at 3%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 1,853
$ 458
$ 886
$ 224
$ 55
$ 103
$ 4,941
$ 1 ,242
$ 2,420
Annualized Value (at 7%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 1,501
$ 371
$ 718
$ 181
$ 45
$ 84
$ 3,998
$ 1,005
$ 1,961
Exhibit 8.4b: Summary of Quantified Benefits,
Preferred Alternative—Traditional Cost of lllness[1]
Data Set
Value of I
Mean
Benefits— Traditional COI1
$ Millions, 2003$)
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annualized Value (at 3%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 1,341
$ 335
$ 644
$ 128
$ 31
$ 58
$ 3,929
$ 989
$ 1,919
Annualized Value (at 7%, 25 Years)
ICR
ICRSSL
ICRSSM
$ 1,089
$ 272
$ 523
$ 104
$ 25
$ 47
$ 3,195
$ 802
$ 1,559
Notes:1 The Traditional COI includes values for medical costs and lost work time (including
a portion of unpaid household production). The Enhanced COI also includes values for lost
personal time (non-work time) such as child care and homemaking (to the extent not covered
by the Traditional COI), time with family, and recreation, and lost productivity on days when
workers are ill but go to work anyway.
Sources: Exhibit C.4 (3%) and C.5 (7%), Columns M-O.
Economic Analysis for the LT2ESWTR
December 2005
-------
8.2.2 National Cost Summary
Exhibit 8.5 presents a national cost summary that reflects all costs estimated for promulgating the
LT2ESWTR. The costs shown are based on detailed information presented in Chapter 6 for the
Preferred Regulatory Alternative. The total national costs of the LT2ESWTR include those associated
with implementation, monitoring for bin classification, covering reservoirs, and additional treatment.
As with the benefit estimates, the estimates of the national annualized costs of the LT2ESWTR
are not point estimates, but are best characterized as distributions. Uncertainty in the occurrence
estimates, together with uncertainty in the estimates of capital and operation and maintenance (O&M)
unit costs, contributed to the uncertainty estimated for the national costs. (See section 6.11 for further
explanation of the distribution of cost estimates.)
Exhibit 8.5 summarizes cost information in four ways. The first block of estimates is the total
undiscounted capital and one-time costs in 2003 dollars. The second block is for total annual O&M costs
once all treatment has been installed (again in 2003 dollars). Because both capital/one-time costs and
O&M costs occur over time, these figures are also discounted to present values (Year 2003 dollars), and
then annualized over 25 years. The annualized costs (at both 3- and 7-percent discount rates) are shown
in the last two blocks of information.
Economic Analysis for the LT2ESWTR 8- 7 December 2005
-------
Exhibit 8.5: Summary of the Costs for the LT2ESWTR Preferred Regulatory Alternative ($Millions, 2003$)
Data Set
ICR
Total Capital and One-Time Costs
(At Full Implementation)
Mean
$ 2,104
90 Percent
Confidence Bound
Lower
(5th %ile)
$ 1,715
Upper
(95th %ile)
$ 2,425
Annual Operations and
Maintenance Costs
(At Full Implementation)
Mean
$ 55
90 Percent
Confidence Bound
Lower
(5th %ile)
$ 48
Upper
(95th %ile)
$ 64
Annualized Cost
3 percent, 25 Years
Mean
$ 133
90 Percent
Confidence Bound
Lower
(5th %ile)
$ 111
Upper
(95th %ile)
$ 160
7 Percent, 25 Years
Mean
$ 150
90 Percent
Confidence Bound
Lower
(5th %ile)
$ 125
Upper
(95th %ile)
$ 181
ICRSSL
$ 1,526
$ 1,164
$ 1,743
$ 33
$ 26
$ 39
$ 93
$ 72
$ 112
$ 107
$ 83
$ 129
ICRSSM
$ 1,719
$ 1,372
$ 1,941
$ 39
$ 33
$ 45
$ 106
$ 86
$ 126
$ 121
$ 99
$ 144
Sources: Exhibit 8.11 (AS-Preferred).
Economic Analysis for the LT2ESWTR
December 2005
-------
8.2.3 National Net Benefits
Net benefits are the difference between the estimated value of human health benefits from the
LT2ESWTR and the estimated costs of complying with the rule. The net benefit calculations use the
estimate of quantified benefits only and do not include nonqualified benefits. All of the comparisons that
follow, therefore, should be considered in the light of the additional benefits that are likely attributable to
this rule but were not quantified.
The overall conclusion from the analyses of net national benefits is that the LT2ESWTR meets a
basic economic threshold condition: benefits are very likely to exceed costs. The first set of tables shows
the net benefits for the Preferred Alternative based on the mean estimate of annualized benefits, less the
mean estimate of annualized costs (Exhibits 8.6a and 8.6b).
Benefits and costs were not computed within the same model, and the confidence bounds
characterizing the uncertainty in the mean benefits and cost estimates cannot be compared directly in a
meaningful way. Consequently, uncertainty in net benefits based on their bounding estimates has not been
explicitly quantified. The uncertainty assessments in the benefits and costs do indicate that the
uncertainty in the value of the benefits is greater than the uncertainty in the costs.
Exhibit 8.6a: Mean Net Benefits,
Preferred Alternative—Enhanced Cost of Illness1 ($Millions, 2003$)
Data
Set
Mean
Benefits
Mean
Costs
Mean Net
Benefits
3 Percent, 25 Years
ICR
ICRSSL
ICRSSM
$ 1 ,853
$ 458
$ 886
$ 133
$ 93
$ 106
$ 1 ,720
$ 365
$ 780
7 Percent, 25 Years
ICR
ICRSSL
ICRSSM
$ 1,501
$ 371
$ 718
$ 150
$ 107
$ 121
$ 1,351
$ 264
$ 597
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.6b: Mean Net Benefits,
Preferred Alternative—Traditional Cost of Illness1 ($Millions, 2003$)
Data
Set
Mean
Benefits
Mean
Costs
Mean Net
Benefits
3 Percent, 25 Years
ICR
ICRSSL
ICRSSM
$ 1,341
$ 335
$ 644
$ 133
$ 93
$ 106
$ 1,208
$ 242
$ 538
7 Percent, 25 Years
ICR
ICRSSL
ICRSSM
$ 1,089
$ 272
$ 523
$ 150
$ 107
$ 121
$ 939
$ 166
$ 402
Note:1 The Traditional COI includes values for medical costs and lost work time (including
a portion of unpaid household production). The Enhanced COI also includes values for lost
personal time (non-work time) such as child care and homemaking (to the extent not
covered by the Traditional COI), time with family, and recreation, and lost productivity on
days when workers are ill but go to work anyway.
Sources: Exhibit 8.10, 8.11, 8.12, Preferred Alternative.
One approach to evaluating net benefits when there is a significant discrepancy between the
magnitudes of the uncertainties for costs and benefits is a "Breakeven Analysis." Using the estimate with
less uncertainty (in this case, costs), a calculation is made of the minimum levels of benefits which, if
achieved, would cause the rule to break even. Comparing this breakeven level of benefits with the actual
benefit estimate provides a measure of the likelihood that the Preferred Regulatory Alternative will have
positive net benefits.
In this analysis, monetized values of the two health endpoints—illnesses and deaths—are
compared against the mean cost estimate. This method of comparison does not take into account the
timing of illnesses and deaths avoided, nor does it incorporate income elasticities into the value of cases
avoided. The cost estimates are based on a stream of costs, discounted into year 2003, and annualized
over 25 years. Thus, this analysis compares the annualized compliance costs to the costs of illness and
deaths in the same year (both expressed in year 2003 dollars). The breakeven point is where the benefits
value, calculated from avoided illnesses and deaths, equals the cost estimate.
Exhibits 8.7a and 8.7b present the number of illnesses avoided at the breakeven point using
Enhanced COI and Traditional COI. The number of deaths avoided are not included in the exhibit
because deaths are assumed to result from a fixed percentage of illnesses. Deaths increase or decrease
proportionally with illnesses but at a much lower rate, and are included proportionally in these estimates.
The exhibits also present the mean estimates of illnesses avoided to show that the estimated benefits are
well above the breakeven point for all occurrence estimates and discount rates using the mean number of
avoided illness. This is true for all conditions except when using the ICRSSL data set and using the 5th
percentile of estimated avoided illness.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.7a: Breakeven Points, Enhanced COI1'2
(Number of Avoided Illnesses Needed to Break Even with Cost Estimates)
Data Set
ICR
ICRSSL
ICRSSM
Annual Avoided Illnesses
Mean
964,360
230.730
455,170
90 Percent
Confidence Bound
Lower
(5th % Me)
149,241
38.281
72,128
Upper
(95th %ile)
2,277,367
521.925
1,112,374
Breakeven Cases
(at 3 Percent, 25 Years)
Mean
56,616
38,166
44,365
90 Percent
Confidence Bound
Lower
(5th %ile)
47,122
29,632
36,153
Upper
(95th %ile)
67,891
46,095
52,677
Breakeven Cases
(at 7 Percent, 25 Years)
Mean
63,853
43,878
50,660
90 Percent
Confidence Bound
Lower
(5th %ile)
53,091
34,193
41 ,300
Upper
(95th %ile)
76,639
52,941
60,163
Exhibit 8.7b: Breakeven Points, Traditional COI1'2
(Number of Avoided Illnesses Needed to Break Even with Cost Estimates)
Data Set
ICR
ICRSSL
ICRSSM
Annual Avoided Illnesses
Mean
964.360
230.730
455,170
90 Percent
Confidence Bound
(5th % Me)
149.241
38.281
72,128
(95th %ile)
2.277.367
521.925
1,112,374
Breakeven Cases
(at 3 Percent, 25 Years)
Mean
74,674
49.838
58,279
90 Percent
Confidence Bound
(5th %ile)
62,153
38.694
47,491
(95th %ile)
89,546
60.191
69,198
Breakeven Cases
(at 7 Percent, 25 Years)
Mean
84,220
57.296
66,549
90 Percent
Confidence Bound
(5th %ile)
70,025
44.650
54,252
(95th %ile)
101,083
69.131
79,032
Notes: 1 Breakeven points include benefits from both avoided illness and deaths. The number of illnesses
avoided implies a proportioned number of avoided deaths. The 90 percent confidence bounds for breakeven
cases results from the 90 percent confidence bounds of the cost estimates.
2 The Traditional COI includes values for medical costs and lost work time (including a portion of unpaid
household production). The Enhanced COI also includes values for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the Traditional COI), time with family, and recreation, and
lost productivity on days when workers are ill but go to work anyway.
Sources: Annual Avoided Illnesses from Exhibit 8.3.
Breakeven Cases derived from Exhibit 8.6, Annualized Cost and Appendix C, Exhibits C.6-C.9, Columns A-C.
If the value of nonquantified benefits could be included in the breakeven analysis, the number of
cases that would need to be avoided in order to break even would be even lower. This is additional
confirmation that benefits are very likely to exceed costs.
8.3 Comparison of Regulatory Alternatives
As discussed in Chapter 3, EPA and the Stage 2 Microbial and Disinfection Byproduct (M-DBP)
Federal Advisory Committee considered numerous regulatory alternatives for the LT2ESWTR. Through
this process, the potential regulatory alternatives were narrowed to four major ones for further evaluation,
one of which was agreed upon by the Advisory Committee as the preferred approach. As previously
mentioned, the rule provisions for the filtered systems were the only ones with alternatives (that is, there
were no alternatives identified for the promises of the rule governing unfiltered systems and uncovered
finished-water reservoirs). Detailed benefit and cost analyses for each of the alternatives are presented
in Chapters 5 and 6, respectively, and are summarized in this chapter. The following sections present
several analyses that compare the Preferred Regulatory Alternative to the other three.
Economic Analysis for the LT2ESWTR
8-11
December 2005
-------
Exhibit 8.8: Summary of Binning and Treatment Scenarios for
Filtered Systems for All Regulatory Alternatives
Source Water Cryptosporidium
Monitoring Results (oocysts/L)
Additional Cryptosporidium
Treatment Requirements
Alternative A1
2.0 log inactivation required for all systems
Alternative A2
<0.03
> 0.03 and < 0.1
>0.1 and < 1.0
>1.0
No action
0.5-log
1.5-log
2.5-log
Alternative A3 - Preferred Alternative
< 0.075
> 0.075 and < 1.0
> 1.0 and < 3.0
>3.0
No action
1-log
2-log
2.5-log
Alternative A4
<0.10
>0.1 and < 1.0
> 1.0
No action
0.5-log
1-log
Note: Additional treatment requirements are in addition to levels already required
under existing rules (Interim Enhanced Surface Water Treatment Rule and Long Term
1 Enhanced Surface Water Treatment Rule).
8.3.1 Comparison of Benefits and Costs
Exhibit 8.9 presents a summary of quantified benefits in terms of illnesses and deaths avoided.
Two sets of numbers are shown—annual after full implementation and annual average over 25 years.
The first set represents the number of cases avoided once all systems have the required treatment
installed (year 2015). The second set takes the number of cases avoided each year over a 25 year period
(using the implementation schedule in Exhibit 0.9) and calculates an annual average of those 25 values.
Exhibits 8.10a and 8.1 Ob present the annualized value of the benefits for each regulatory alternative using
Enhanced and Traditional COI values. The benefits derive only from treatment improvements made as a
result of being in an Action Bin (for filtered systems) or from the 2 and 3 log treatment improvements
made at unfiltered plants. No benefits are included for improvements made by systems with uncovered
finished water reservoirs. Further, only reductions in endemic illness and mortality associated with
cryptosporidiosis (as opposed to those that occur as outbreaks) are included. As stated previously,
unqualified benefits are likely significant.
Economic Analysis for the LT2ESWTR
8-12
December 2005
-------
Exhibit 8.11 follows with a presentation of the quantified costs. The annualized total costs of the
rule include costs for filtered systems, unfiltered systems, uncovered reservoirs, and implementation costs
for systems and for States/Primacy Agencies.
Exhibit 8.9: Comparison of Number of Illnesses and Deaths Avoided
for All Regulatory Alternatives
Data Set
Rule
Alternative
Annual Illnesses Avoided
Mean
90 Percent
Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annual Deaths Avoided
Mean
90 Percent
Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annual Average after Full implementation 1
ICR
ICRSSL
ICRSSM
A1
A2
AS-Preferred
A4
A1
A2
AS-Preferred
A4
A1
A2
AS-Preferred
A4
989,954
975,326
964,360
902.500
292,992
250,153
230,730
197,892
514,150
476,008
455,170
399.799
151,965
150,295
149,241
143.231
45,926
40,612
38,281
34,572
78,383
74,499
72,128
66.646
2,347,055
2,307,247
2,277,367
2.088.993
702,369
578,528
521 ,925
431 ,263
1 ,285,283
1,173,586
1,112,374
949.132
211
209
207
197
62
55
52
47
109
103
100
91
35
35
34
33
10
10
9
9
18
17
17
16
480
474
468
437
143
123
113
98
259
241
230
204
Annual Average over 25 years
ICR
ICRSSL
ICRSSM
A1
A2
AS-Preferred
A4
A1
A2
AS-Preferred
A4
A1
A2
AS-Preferred
A4
731 ,304
720,635
712,732
667,732
216,232
185,009
170.977
146,951
379,491
351,691
336,652
296,138
1 1 1 ,442
110,288
109,486
104,766
33,661
29,918
28.314
25,629
57,314
54,474
52,763
48,695
1 ,737,522
1,707,991
1,685,176
1 ,549,243
521 ,823
433,840
392.979
325,686
950,498
871 ,678
826,004
706,762
157
155
154
146
46
41
39
35
81
76
74
67
26
25
25
25
8
7
7
6
13
13
12
12
357
352
348
325
107
92
85
74
193
179
172
152
Note:1 Full implementation occurs 7 years after rule promulgation.
Source: Appendix C, Exhibit C.4, Columns A-F. Annual Average derived from Exhibits C.4 and Exhibit O.9.
Economic Analysis for the LT2ESWTR
8-13
December 2005
-------
Exhibit 8.1 Oa: Comparison of Annualized Value of Illnesses and Deaths Avoided
for All Regulatory Alternatives, Enhanced COI1
Data Set
Rule
Alternative
Value of Benefits ($ Millions, 2003$)
Mean
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annualized Value (at 3%, 25 Years)
ICR
A1
A2
A3 - Preferred
A4
$ 1,895
$ 1,871
$ 1,853
$ 1,753
$ 227
$ 225
$ 224
$ 216
$ 5,079
$ 4,992
$ 4,941
$ 4,642
ICRSSL
A1
A2
A3 - Preferred
A4
$ 558
$ 489
$ 458
$ 405
$ 63
$ 58
$ 55
$ 51
$ 1,553
$ 1,341
$ 1,242
$ 1,070
ICRSSM
A1
A2
A3 - Preferred
A4
$ 981
$ 919
$ 886
$ 796
$ 110
$ 106
$ 103
$ 96
$ 2,713
$ 2,533
$ 2,420
$ 2,134
Annualized Value (at 7%, 25 Years)
ICR
A1
A2
A3 - Preferred
A4
$ 1,534
$ 1,515
$ 1,501
$ 1,421
$ 184
$ 182
$ 181
$ 175
$ 4,115
$ 4,042
$ 3,998
$ 3,766
ICRSSL
A1
A2
A3 - Preferred
A4
$ 452
$ 396
$ 371
$ 328
$ 51
$ 47
$ 45
$ 41
$ 1,258
$ 1,086
$ 1,005
$ 867
ICRSSM
A1
A2
A3 - Preferred
A4
$ 794
$ 744
$ 718
$ 645
$ 89
$ 86
$ 84
$ 78
$ 2,199
$ 2,051
$ 1,961
$ 1,730
Economic Analysis for the LT2ESWTR
8-14
December 2005
-------
Exhibit 8.1 Ob: Comparison of Annualized Value of Illnesses and Deaths Avoided
for All Regulatory Alternatives, Traditional COI1
Data Set
Rule
Alternative
Value of Benefits ($ Millions, 2003$)
Mean
90% Confidence Bound
Lower
(5th %ile)
Upper
(95th %ile)
Annualized Value (at 3%, 25 Years)
ICR
A1
A2
A3 - Preferred
A4
$ 1,369
$ 1,353
$ 1,341
$ 1,273
$ 130
$ 129
$ 128
$ 124
$ 4,017
$ 3,969
$ 3,929
$ 3,692
ICRSSL
A1
A2
A3 - Preferred
A4
$ 403
$ 356
$ 335
$ 299
$ 35
$ 32
$ 31
$ 29
$ 1,219
$ 1,065
$ 989
$ 861
ICRSSM
A1
A2
A3 - Preferred
A4
$ 708
$ 666
$ 644
$ 583
$ 62
$ 59
$ 58
$ 54
$ 2,140
$ 1,995
$ 1,919
$ 1,706
Annualized Value (at 7%, 25 Years)
ICR
A1
A2
A3 - Preferred
A4
$ 1,112
$ 1,099
$ 1,089
$ 1,034
$ 105
$ 104
$ 104
$ 100
$ 3,257
$ 3,220
$ 3,195
$ 3,001
ICRSSL
A1
A2
A3 - Preferred
A4
$ 327
$ 289
$ 272
$ 243
$ 29
$ 26
$ 25
$ 23
$ 990
$ 864
$ 802
$ 700
ICRSSM
A1
A2
A3 - Preferred
A4
$ 575
$ 541
$ 523
$ 474
$ 50
$ 48
$ 47
$ 44
$ 1,739
$ 1,623
$ 1,559
$ 1,386
Note:1 The Traditional COI includes values for medical costs and lost work time (including
a portion of unpaid household production). The Enhanced COI also includes values for lost
personal time (non-work time) such as child care and homemaking (to the extent not
covered by the Traditional COI), time with family, and recreation, and lost productivity on
days when workers are ill but go to work anyway.
Sources: Appendix C, Exhibits C.4 (Enhanced COI) and C.5 (Traditional COI), Columns M-O
Economic Analysis for the LT2ESWTR
8-15
December 2005
-------
Exhibit 8.11: Comparison of Costs for All Regulatory Alternatives ($ Millions, 2003$)1
Data
Set
ICR
Rule
Alternative
A1
A2
AS-Preferred
A4
Capital and One-Time
(Undiscounted, at Full Implementation)
Mean
$ 5,631
$ 2,511
$ 2.104
$ 1 ,328
90%
Confidence Bound
Lower
(5th %ile)
$ 4,879
$ 2,104
$ 1.715
$ 1 ,085
Upper
(95th %ile)
$ 6,364
$ 2,840
$ 2.425
$ 1,454
Operations and Maintenance
(Undiscounted, at Full Implementation)
Mean
$ 226
$ 74
$ 55
$ 28
9
Confiden
Lower
(5th %ile)
$ 213
$ 65
$ 48
$ 24
0%
ce Bound
Upper
(95th %ile)
$ 240
$ 87
$ 64
$ 32
Annualized Value
(Discounted at 3%, 25 Years)*
Mean
$ 403
$ 163
$ 133
$ 81
90%
Confidence Bound
Lower
(5th %ile)
$ 361
$ 140
$ 111
$ 67
Upper
(95th %ile)
$ 444
$ 197
$ 160
$ 95
Annualized Value
(Discounted at 7%, 25 Years)*
Mean
$ 436
$ 182
$ 150
$ 93
90%
Confidence Bound
Lower
(5th %ile)
$ 389
$ 155
$ 125
$ 78
Upper
(95th %ile)
$ 484
$ 219
$ 181
$ 110
ICRSSL
A1
A2
AS-Preferred
A4
$ 5.631
$ 1 .959
$ 1 ,526
$ 973
$ 4.879
$ 1 .538
$ 1,164
$ 786
$ 6.364
$ 2.147
$ 1,743
$ 1.049
$ 226
$ 50
$ 33
$ 16
$ 213
$ 40
$ 26
$ 13
$ 240
$ 59
$ 39
$ 19
$ 403
$ 123
$ 93
$ 57
$ 361
$ 98
$ 72
$ 47
$ 444
$ 148
$ 112
$ 68
$ 437
$ 139
$ 107
$ 68
$ 389
$ 110
$ 83
$ 56
$ 484
$ 167
$ 129
$ 80
ICRSSM
A1
A2
AS-Preferred
A4
$ 5,631
$ 2.155
$ 1.719
$ 1 ,082
$ 4,879
$ 1 .757
$ 1 .372
$ 886
$ 6,364
$ 2.342
$ 1.941
$ 1,156
$ 226
$ 58
$ 39
$ 20
$ 213
$ 49
$ 33
$ 17
$ 240
$ 66
$ 45
$ 23
$ 403
$ 137
$ 106
$ 65
$ 361
$ 114
$ 86
$ 54
$ 444
$ 161
$ 126
$ 76
$ 437
$ 154
$ 121
$ 76
$ 389
$ 128
$ 99
$ 63
$ 484
$ 181
$ 144
$ 88
Note:1 Operation and maintenance costs are annual costs incurred by all systems upon full implementation of the rule.
Sources: Appendix O.
Capital and One-time: Undiscounted-Exhibit O.4 (State total), Exhibit O.5 (system implementation and monitoring total), and O.6 (sum of filtered and unfiltered
treatment costs)
Annual Operations and Maintenance Cost: Undiscounted-Exhibit O.6 (sum of filtered and unfiltered O&M costs)
Annualized Cost at 3 Percent: Exhibits O.16 + O.17 + O.18.
Annualized Cost at 7 Percent: Exhibits 0.19 + O.20 + O.21.
Economic Analysis for the LT2ESWTR
December 2005
-------
Net Benefits
As described previously, net benefits are calculated as the difference between the monetized
benefits and cost estimates. Exhibit 8.12a and 8.12b present net benefits based on the annualized present
value of quantified benefits at 3 and 7 percent discount rates for both enhanced and traditional cost of
illness. Adding nonquantified benefits would raise the overall net benefits. The data are based on the
mean benefits less the mean values for costs. Using net benefits as a threshold measure shows that
almost all alternatives provide benefits that exceed their costs, based on their average values.
Maximum Net Benefits
Identifying the maximum net benefits among the regulatory alternatives is a first step in a
comparative analysis of regulatory alternatives. The bold numbers in heavily outlined boxes in Exhibits
8.12a and 8.12b indicate the maximum net benefit of the four rule alternatives. For most combinations of
occurrence data sets, COI values, and discount rates, the Preferred Regulatory Alternative (A3) had the
maximum net benefits among the alternatives (8 of 12 scenarios). However, the differences are often
slight among three of the regulatory alternatives (A2, A3, and A4). The range from the high to the low of
A2, A3, and A4 is 1 to 10 percent for all but one of the combinations.
Exhibit 8.12a: Comparison of Mean Net Benefits for All Regulatory
Alternatives—Enhanced COI1 (Million $/Year)
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Annualized Value
3%, 25 Years
$ 1 .492
$ 1,708
$ 1,720
$ 1.673
7%, 25 Years
$ 1.098
$ 1,333
$ 1,351
$ 1 .328
ICRSSL
A1
A2
A3 - Preferred
A4
$ 156
$ 366
$ 365
$ 347
$ 15
$ 257
$ 264
$ 261
ICRSSM
A1
A2
A3 - Preferred
A4
$ 578
$ 782
$ 780
$ 731
$ 358
$ 591
$ 597
$ 569
Economic Analysis for the LT2ESWTR
8-17
December 2005
-------
Exhibit 8.12b: Comparison of Mean Net Benefits for All Regulatory
Alternatives—Traditional COI1 (Million $/Year)
Data
Set
ICR
Rule
Alternative
A1
A2
A3 - Preferred
A4
Annualized Value
3%, 25 Years
$ 967
$ 1,190
$ 1,208
$ 1,193
7%, 25 Years
$ 675
$ 917
$ 939
$ 941
ICRSSL
A1
A2
A3 - Preferred
A4
$
$ 233
$ 242
$ 242
$ -109
$ 150
$ 166
$ 175
ICRSSM
A1
A2
A3 - Preferred
A4
$ 306
$ 529
$ 538
$ 518
$ 138
$ 387
$ 402
$ 398
Note:1 The Traditional COI includes values for medical costs and lost work time (including a
portion of unpaid household production). The Enhanced COI also includes values for lost
personal time (non-work time) such as child care and homemaking (to the extent not covered by
the Traditional COI), time with family, and recreation, and lost productivity on days when workers
are ill but go to work anyway.
Sources: 3 percent data: Exhibit 8.10, "Mean" column - Exhibit 8.11, Annualized Value at 3 percent,
"Mean" column.
7 percent data: Exhibit 8.10, "Mean" column - Exhibit 8.11, Annualized Value at 7 percent, "Mean"
column.
Incremental Net Benefits
Rule alternatives can also be compared on the basis of their incremental net benefits. Generally,
the goal of an incremental analysis is to identify the last regulatory option with positive net incremental
benefits. Each additional regulatory alternative costs more per unit of protection than its predecessor.
The result is a declining net benefit as more stringent alternatives are considered. When the next more
stringent alternative costs more than it returns in absolute benefits, net benefit becomes negative and the
alternative is not worth pursuing. However, the usefulness of this analysis is limited because many
benefits from the rule are nonqualified and not monetized.
Exhibits 8.13a and 8.13b present the incremental net benefits calculated from Enhanced and
Traditional COI values. Usually an incremental analysis implies increasingly stringent control along a
single parameter, with each alternative providing all the protection of the previous alternative, plus
additional protection. However, the regulatory alternatives for this rule base their treatment requirements
on a system's source water quality (except for Al). As a result, the number of systems requiring
additional treatment, and thus the population affected by the rule, differ among the alternatives. With net
benefits calculated using different occurrence estimates, COI values, and discount rates, the last
alternative with positive net benefit is not always the same. The sensitivity of the mean net benefit
estimates to various input assumptions in this analysis is illustrated by the ranking analysis presented in
section 8.4.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.13a: Incremental Net Benefits[1] for All Alternatives, By Data Set,
Enhanced COI[2] ($Millions, 2003$)
Data
Set
Rule
Alternative
Annual
Costs
A
Annual
Benefits
B
Incremental
Costs1
C
Incremental
Benefits
D
Incremental Net
Benefits
E=D-C
3 Percent Discount Rate
ICR
A4
A3 - Preferred
A2
A1
$ 81
$ 133
$ 163
$ 403
$ 1,753
$ 1,853
$ 1,871
$ 1,895
$ 81
$ 53
$ 30
$ 239
$ 1 ,753
$ 100
$ 18
$ 24
$ 1 ,673
$ 47
$ -12
$ -216
ICRSSL
A4
A3 - Preferred
A2
A1
ICRSSM
A4
A3 - Preferred
A2
A1
$ 57
$ 93
$ 123
$ 403
$ 405
$ 458
$ 489
$ 558
$ 57
$ 35
$ 30
$ 280
$ 65
$ 106
$ 137
$ 403
$ 796
$ 886
$ 919
$ 981
$ 65
$ 41
$ 31
$ 266
$ 405
$ 53
$ 31
$ 69
$ 796
$ 90
$ 33
$ 62
$ 347
$ 18
$ 1
$ -210
$ 731
$ 49
$ 2
$ -204
7 Percent Discount Rate
ICR
A4
A3 - Preferred
A2
A1
$ 93
$ 150
$ 182
$ 436
$ 1,421
$ 1,501
$ 1.515
$ 1,534
$ 93
$ 57
$ 31
$ 255
$ 1,421
$ 80
$ 14
$ 19
$ 1 ,328
$ 23
$ -17
$ -236
ICRSSL
A4
A3 - Preferred
A2
A1
$ 68
$ 107
$ 139
$ 437
$ 328
$ 371
$ 396
$ 452
$ 68
$ 39
$ 32
$ 298
$ 328
$ 43
$ 25
$ 56
$ 261
$ 4
$ -7
$ -242
ICRSSM
A4
A3 - Preferred
A2
A1
$ 76
$ 121
$ 154
$ 437
$ 645
$ 718
$ 744
$ 794
$ 76
$ 45
$ 33
$ 283
$ 645
$ 73
$ 27
$ 50
$ 569
$ 27
$ -6
$ -233
Economic Analysis for the LT2ESWTR
8-19
December 2005
-------
Exhibit 8.13b: Incremental Net Benefits1 for All Alternatives, By Data Set,
Traditional COI2 ($Millions, 2003$)
Data
Set
Rule
Alternative
Annual
Costs
A
Annual
Benefits
B
Incremental
Costs1
C
Incremental
Benefits
D
Incremental Net
Benefits
E=D-C
3 Percent Discount Rate
ICR
A4
A3 - Preferred
A2
A1
$ 81
$ 133
$ 163
$ 403
$ 1,273
$ 1,341
$ 1,353
$ 1,369
$ 81
$ 53
$ 30
$ 239
$ 1 ,273
$ 68
$ 12
$ 16
$ 1,193
$ 15
$ -18
$ -223
ICRSSL
A4
A3 - Preferred
A2
A1
$ 57
$ 93
$ 123
$ 403
$ 299
$ 335
$ 356
$ 403
$ 57
$ 35
$ 30
$ 280
$ 299
$ 36
$ 21
$ 47
$ 242
$ 1
$ -9
$ -233
ICRSSM
A4
A3 - Preferred
A2
A1
$ 65
$ 106
$ 137
$ 403
$ 583
$ 644
$ 666
$ 708
$ 65
$ 41
$ 31
$ 266
$ 583
$ 61
$ 23
$ 42
$ 518
$ 20
$ -9
$ -224
7 Percent Discount Rate
ICR
A4
A3 - Preferred
A2
A1
$ 93
$ 150
$ 182
$ 436
$ 1,034
$ 1,089
$ 1,099
$ 1,112
$ 93
$ 57
$ 31
$ 255
$ 1 ,034
$ 55
$ 10
$ 13
$ 941
$ -3
$ -22
$ -242
ICRSSL
A4
A3 - Preferred
A2
A1
$ 68
$ 107
$ 139
$ 437
$ 243
$ 272
$ 289
$ 327
$ 68
$ 39
$ 32
$ 298
$ 243
$ 29
$ 17
$ 38
$ 175
$ -10
$ -15
$ -260
ICRSSM
A4
A3 - Preferred
A2
A1
$ 76
$ 121
$ 154
$ 437
$ 474
$ 523
$ 541
$ 575
$ 76
$ 45
$ 33
$ 283
$ 474
$ 49
$ 18
$ 34
$ 398
$ 4
$ -15
$ -249
Notes: 1
Derivation: Incremental costs are the cost of each regulatory alternative minus the cost of the next least expensive
alternative (which is zero for Alternative A4, the least expensive alternative). The derivation for incremental benefits
is analogous to the derivation for incremental costs.
2 The Traditional COI includes values for medical costs and lost work time (including a portion of unpaid
household production). The Enhanced COI also includes values for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the Traditional COI), time with family, and recreation, and
lost productivity on days when workers are ill but go to work anyway.
Sources: [A] Exhibit 8.11, Annualized Value at 3 percent, "Mean" column and Annualized Value at 7 percent, "Mean"
column.
[B] Exhibit 8.10, "Mean" column.
Consideration of Uncertainty in the Ranking of Alternatives by Net Benefits
Cost estimates for a given occurrence level contain less uncertainty than the corresponding
estimates of benefits. Benefit estimates have significant uncertainty due to inputs such as infectivity,
morbidity, income elasticity, and even the value of a statistical life. In addition, quantified benefits
constitute only a portion of the likely total benefits from this rule. Under these conditions, it is important to
assess how the ranking of alternatives will change as estimates of benefits vary. Using two COI values,
Economic Analysis for the LT2ESWTR
December 2005
-------
as has been done so far, gives a snapshot of how the ranking may change. This analysis goes further by
analyzing how the ranking changes on a continuum of changing benefits.
As estimates of possible benefits proportionately increase while cost estimates are held constant,
the alternative with the highest net benefits shifts from Alternative A4 to Alternative A3 (the Preferred
Alternative), then to A2, and finally to Al (A4 being the least protective through to Al being the most
protective). Because the level of benefits dictates which alternative performs best and how well each
performs relative to the other alternatives, the selection of the Preferred Alternative requires examining
where these transition points are, and how the relative closeness of the alternatives to each other varies.
The selection of an occurrence data set and a discount rate affects this pattern of relationships between
alternatives. Graphs, therefore, were constructed for all combinations of occurrence and discount rates
using Enhanced and Traditional COI values.
Exhibits 8.14a and 8.14b show the value of the upper 90 percent confidence bound benefit
estimates for each alternative and how much greater these values are than the mean benefit estimates (as
a percent of the mean). Exhibits 8.15 and 8.16 show these rankings and transition points graphically.
Each graph has four lines representing the four alternatives. On the horizontal axis, the graphs show a
range of benefits, expressed as multiples of the calculated benefits shown in Exhibit 8.10. The entire
scale shows a range of benefits from zero to as much as five times the calculated benefits. The graphs
display this large range because (1) the range of quantified benefits at the 90th percentile is about 247 to
291 percent of the mean benefits (Exhibit 8.14), and (2) the consideration of unquantified benefits dictates
that the analysis consider a range of benefits beyond the range of calculated quantified benefits.
Consequently, a five-fold range is reasonable to display. The vertical axis of the graphs shows the relative
ranking of the four alternatives. Specifically, for any given level of benefits, one alternative will yield the
highest net benefits (shown as 100 percent), and one will give the lowest (shown as zero percent). The
relative benefits of the remaining two alternatives fall somewhere in between. Thus, a vertical reading of
the graphs shows the ranking of the alternatives at that level of benefits, and the closeness of the relative
rankings.
Economic Analysis for the LT2ESWTR 8-21 December 2005
-------
Exhibit 8.14a: Upper End of 90 Percent Confidence Bound as a Percent of Mean
Estimate of Benefits, By Data Set, Annualized at 3 Percent, Enhanced COI
($Millions, 2003$)
Data
Set
ICR
Rule
Alternative
A1
A2
A3-P referred
A4
Benefits ($Millions, 2000$)
Mean
$ 1,895
$ 1,871
$ 1,853
$ 1,753
Upper End of
90%
Confidence
$ 5,079
$ 4,992
$ 4,941
$ 4,642
Upper End of
90%
Confidence
Bound as a %
268%
267%
267%
265%
ICRSSL
A1
A2
A3-P referred
A4
$ 558
$ 489
$ 458
$ 405
$ 1 ,553
$ 1.341
$ 1 .242
$ 1 .070
278%
274%
271%
264%
ICRSSM
A1
A2
A3-P referred
A4
$ 981
$ 919
$ 886
$ 796
$ 2,713
$ 2,533
$ 2,420
$ 2,134
277%
276%
273%
268%
Economic Analysis for the LT2ESWTR
8-22
December 2005
-------
Exhibit 8.14b: Upper End of 90 Percent Confidence Bound as a Percent of Mean
Estimate of Benefits, By Data Set, Annualized at 3 Percent, Traditional COI
($Millions, 2003$)
Data
Set
ICR
Rule
Alternative
A1
A2
A3-P referred
A4
Benefits ($Millions, 2000$)
Mean
$ 1 .369
$ 1,353
$ 1,341
$ 1 ,273
Upper End of
90%
Confidence
Bound
$ 4.017
$ 3,969
$ 3,929
$ 3,692
Upper End of
90%
Confidence
Bound as a %
of Mean
293%
293%
293%
290%
ICRSSL
A1
A2
A3-P referred
A4
$ 403
$ 356
$ 335
$ 299
$ 1,219
$ 1 ,065
$ 989
$ 861
302%
299%
295%
288%
ICRSSM
A1
A2
A3-P referred
A4
$ 708
$ 666
$ 644
$ 583
$ 2,140
$ 1 ,995
$ 1.919
$ 1 ,706
302%
299%
298%
293%
Notes:
The Traditional COI includes values for medical costs and lost work time (including a portion of
unpaid household production). The Enhanced COI also includes values for lost personal time
(non-work time) such as child care and homemaking (to the extent not covered by the Traditional
COI), time with family, and recreation, and lost productivity on days when workers are ill but go to
work anyway.
Upper end of 90 percent confidence bound is the 95th percentile.
Source: Exhibit 8.10.
What conclusions can be derived from the graphs of the actual data? Consider the first graph in
Exhibit 8.15a, based on occurrence distributions modeled from the ICR data set. The expected pattern is
depicted: at low levels of benefits, Alternative A4 ranks highest, but as estimates of benefits increase,
Alternative A3 and then A2 have the highest net benefits. Alternative Al will eventually become the
alternative with highest net benefits, but at a level of benefits beyond the scale of the graph. More
important, the graph shows that A4 has the highest net benefits only for low estimates of benefits: from
zero benefits to about 50 percent of quantified benefits. Alternative A3, the Preferred Alternative, has
the highest net benefits throughout the range that lies nearest to the mean quantified benefits: from about
50 percent to 150 percent of quantified benefits (ICR data set).
Under most combinations of occurrence data sets, discount rates, and COI values, the Preferred
Alternative ranks highest near the mean estimate of quantified benefits. The relative strength of this
alternative across all scenarios is also shown by the fact that over a wide range, from zero benefits to
approximately 350 percent of quantified benefits, the Preferred Alternative has either the highest or
second highest level of net benefits. So even under a range of uncertainty, the absolute value of the
projected benefits does not affect the relative ranking of the Preferred Alterative.
Economic Analysis for the LT2ESWTR
8-23
December 2005
-------
Alternative A4 appears to be the best alternative only when benefits are less than the mean
estimated values of the various data sets, and for a few of the data sets using Traditional COI values in
the range of best estimates. However, Alternative A4's relative ranking falls rapidly as benefits increase
and taking into account the nonqualified benefits, A4 would likely not be the best alternative for any of
the estimates.
Alternative A2 is also a strong alternative over a wide range of benefits above those discussed
for the Preferred Alternative. If total benefits are substantially greater than the value of mean quantified
benefits, then Alternative A2 would be a consistently strong choice.
Alternative Al, the most stringent alternative, is not a strong contender unless true benefits are
many multiples of the mean quantified benefits. The lowest level of benefits at which Alternative Al has
the highest net benefits is at about 450 percent of mean quantified benefits (ICRSSL, Enhanced COI,
3 percent).
Overall, the Preferred Alternative and Alternative A2 are strong alternatives considering both the
range of quantified benefits (with an upper 90 percent confidence bound of 267 to 302 percent of mean
benefits) and the significant benefits that have not been quantified. The M-DBP Federal Advisory
Committee and EPA selected the Preferred Alternative after also considering other factors such as
technical, managerial, and financial capacities of water systems.
Economic Analysis for the LT2ESWTR 8-24 December 2005
-------
Exhibit 8.15a
Comparison of Regulatory Alternatives Ranked by Net Benefits, 3 Percent Cost
Enhanced COI
£ £
"s "*
c g
m 5
11
Rankings of Regulatory Alternatives, ICR Data Set
(by Net Benefits)
0%
1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
I!
11
Rankings of Regulatory Alternatives, ICRSSL Data Set
(by Net Benefits)
20%
0%
1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Rankings of Regulatory Alternatives, ICRSSM Data Set
(by Net Benefits)
0.5
1.5 2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Economic Analysis for the LT2ESWTR
8-25
December 2005
-------
Exhibit 8.15b
Comparison of Regulatory Alternatives Ranked by Net Benefits, 7 Percent Cost
Enhanced COI
-2
»=
is K
Rankings of Regulatory Alternatives, ICR Data Set
(by Net Benefits)
0 0.5 1 1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
Rankings of Regulatory Alternatives, ICRSSL Data Set
(by Net Benefits)
1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Rankings of Regulatory Alternatives, ICRSSM Data Set
(by Net Benefits)
o%
0.5 1
2 25 3 35 4 45 5
Multiple of Calculated Benefits
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.16a
Comparison of Regulatory Alternatives Ranked by Net Benefits, 3 Percent Cost
Traditional COI
*- 01
i 2
a
Rankings of Regulatory Alternatives, ICR Data Set
(by Net Benefits)
20%
0%
1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
I §
1!
= 0
HI
•sJ
r s1
5 -i-
Rankings of Regulatory Alternatives, ICRSSL Data Set
(by Net Benefits)
100%
0.5
1.5 2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Rankings of Regulatory Alternatives, ICRSSM Data Set
(by Net Benefits)
1.5
2 2.5 3 3.5 4 4.5
Multiple of Calculated Benefits
A2
'A3
•A4
Economic Analysis for the LT2ESWTR
8-27
December 2005
-------
Exhibit 8.16b
Comparison of Regulatory Alternatives Ranked by Net Benefits, 7 Percent Cost
Traditional COI
a =
Rankings of Regulatory Alternatives, ICR Data Set
(by Net Benefits)
1.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Rankings of Regulatory Alternatives, ICRSSL Data Set
(by Net Benefits)
o%
0.5
1.5 2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Rankings of Regulatory Alternatives, ICRSSM Data Set
(by Net Benefits)
0.5
2 2.5 3 3.5 4
Multiple of Calculated Benefits
4.5
Economic Analysis for the LT2ESWTR
December 2005
-------
8.3.2 Cost-Effectiveness Measures
Cost-effectiveness analysis is a policy evaluation tool that allows comparisons of regulatory
alternatives. Evaluating cost-effectiveness for actions that reduce both illness and deaths is difficult and
there is no universally accepted approach. In this EA, EPA presents several approaches to assess cost
effectiveness: a "traditional approach" in which benefits and costs are graphically compared across
alternatives, and comparisons of cost per illness avoided, cost per death avoided, incremental cost per
illness and death avoided, and benefit-cost ratios. Additionally, a Quality-Adjusted Life Years approach,
historically used in policy decisions regarding medical interventions, is applied to this analysis in
Appendix U as an experimental approach to evaluating environmental policy costs and benefits.
Cost-effectiveness-Traditional Approach
The concept of cost-effectiveness can be defined simply as getting the greatest level of benefits
for a given expenditure or imposing the least cost for a given level of benefits. Exhibits 8.17a and 8.17b
show the annualized value of benefits and costs for the four alternatives, calculated using the various
combinations of occurrence data sets, COI values, and discount rates. For each alternative, the graph
plots the mean benefit versus its corresponding range of cost estimates (a 90 percent confidence bound
shown as a vertical bar). A trend line connects the mean estimate of costs for each alternative. These
graphs help to visually show the concept of cost-effectiveness and to compare the alternatives. In Exhibit
8.17, the test would be to see if any alternative was to the right and completely below any other
alternative. If so, the alternative to the right and below would be more cost-effective and would
"dominate" the alternative that provided fewer benefits at higher costs.
In Exhibit 8.17, some graphs show the lower range of the cost estimate for Alternative A2
extending below the top of the cost range for Alternative A3. Does this mean that Alternative A2 in some
cases "dominates" A3? The answer is no, because the modeling approach that generates the higher
portion of the cost range for Alternative A3 also generates the higher portions of the range for Alternative
A2. Thus, it is most appropriate to compare corresponding values from each range to determine the
possibility or extent of overlap. In the cases shown, therefore, Alternative A2 cannot be said to be more
cost effective than Alternative A3.
In the strict sense, each of the regulatory alternatives is cost effective—no regulatory alternative
provides more benefits at the same or a lower cost than another, and no alternative can achieve lower
costs for the same or a greater level of benefits than another. Thus, no alternative dominates any other or
is more cost effective. Instead, the alternatives offer increasing levels of benefits at increasing levels of
cost, as seen in Exhibit 8.17.
The alternatives shown in Exhibit 8.17 map boundaries of cost-effective alternatives. In addition
to allowing a visual comparison of cost-effectiveness, the exhibit shows information graphically about the
incremental benefits of each alternative. Compared to Alternative A4, the Preferred Alternative achieves
significant incremental benefits at a relatively low increase in costs. The step to Alternative A2 achieves
more benefits, but at higher incremental rate. The step to Alternative Al achieves a similar increase in
benefits, but at a significantly higher cost. The Preferred Alternative, and perhaps Alternative A2, appear
to be good values; other alternatives have either significantly fewer benefits for similar costs or greater
benefits, but at dramatically higher costs.
Economic Analysis for the LT2ESWTR 8-29 December 2005
-------
Exhibit 8.17a: Range of Annualized Costs at Mean Benefit Level, All Regulatory
Alternatives—Enhanced Cost of Illness1
3 Percent Discount Rate
7 Percent Discount Rate
in
*- c
r
r--'
$1,725 $1,750 $1,775 $1,800 $1,825 $1,850 $1,875 $1,900 $1,925
$450 •
$400 •
$350 •
$300 •
$250 •
$200 •
$150 •
$100 •
$50 •
Alt. 1
'
/
*
,'
Alt. 2
Alt3 V J-
Alt 4 Nil-' 1
v- • '
$375 $400 $425 $450 $475 $500 $525
*
$550 $575
$450 •
$400 •
$350 •
$300 •
$250 •
$200 •
$150 •
$100 •
$50 •
Alt. 1 ^.^
*-
,
,
'
/
,'
Alt. 2 /
Alt. 3 ^V^ L'
Alt. 4 ^V 1^ ~ 1
V ,.-••*'!
*'
$750 $800 $850 $900 $950
$1,000
Mean of Benefits (SMillions)
Alt. 1 ^^
*
w
'
,
„ r.^
f
$1 ,400 $1 ,425 $1 ,450 $1 ,475 $1 ,500 $1 ,525
$1,55C
.
.
_
-
_
-
•
Alt. 1 f^,,^
,'
t*
i
*
/
•'
*
Alt. 2
T3,x
$300 $320 $340 $360 $380 $400 $420 $440
$460
A«. 1 ..^
'
,'l
•
*
t
»
Alt. 2
Alt. 3 ^^ - '
Alt. 4 ^^^ V
\ f . . | ' '
I- -""'
$600 $640 $680 $720 $760 $800
Mean of Benefits (SMillions)
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.17b: Range of Annualized Costs at Mean Benefit Level,
All Regulatory Alternatives—Traditional Cost of Illness1
3 Percent Discount Rate
7 Percent Discount Rate
I/)
4-i C
w .2
3 i
13
Q <*
M "
O t/)
— o
o
1?
o
IT "~
ro —
Q 5
i (ft
C£ "5
o w
CO ^
nj ._
ij* —
Q 5
P «A
Q? "K
ft O
^ o
$500
$450
$400
$350
$300
$250
$200
$150
$100
$50
Alt. 1 l
/
*
t
Alt. 2 *
*
Alt. 3 ^^ .'
Alt. 4 ^^ * |
/ ..... -V1
f"
$1 ,250 $1 ,275 $1 ,300 $1 ,325 $1 ,350 $1 ,37
Alt. 1 —*
•
,
•
,
•
Alt. 2 /
Alt. 3 \ T
r
$1,025 $1,050 $1,075 $1,100 $1,125
$450 .
$400 •
$350 .
$300 .
$250 •
$200 .
$150 .
$100 .
$50.
.
t
*
f
»
Alt. 2 *
Alt. 4 Alt. 3 ^ ,*
\ \-'\
t--'"
$275 $300 $325 $350 $375 $400 $42
$450 .
$400 .
$350 .
$300 .
$250 .
$200 .
$150 .
$100 .
$50 •
$5
Alt. 1 —
^^"*
*
,
*
r
*
*
Alt. 2 '
Alt. 4 Alt. 3 \. f
\ .... >r>f
'
Alt 1 ^
^^^^^ ,
f'
•
t
t
^*
*
Alt. 2 *
Alt. 3 . *
Alt 4 S^ % ]'
*"
$225 $245 $265 $285 $305 $325 $34i
Alt. 1 [
^^ *
*'
*
*
te
Alt. 2
}
Alt. 3 \. /
Alt. 4 ^ .
i .>i--r
\- ••""
$450 $470 $490 $510 $530 $550 $570 $59t
Mean of Benefits (SMillions)
Mean of Benefits (SMillions)
Note(from8.17aand8.17b):
[1 ]The Traditional COI includes values for medical costs and lost work time (including a portion of unpaid
household production). The Enhanced COI also includes values for lost personal time (non-work time) such as
child care and homemaking (to the extent not covered by the Traditional COI), time with family, and recreation, and
lost productivity on days when workers are ill but go to work anyway.
Sources: Exhibits 8.10 and 8.11.
Economic Analysis for the LT2ESWTR
December 2005
-------
Other Measures-Cost Per Illness Avoided and Cost Per Death Avoided
Other measures related to cost-effectiveness not only analyze the performance of the regulatory
alternatives, but also allow for comparisons across rules. The cost-effectiveness measures presented in
Exhibits 8.18 and 8.19 include the cost, net cost, and incremental net cost for each illness and death
avoided. All regulatory alternatives for the LT2ESWTR reduce the risk of both illness and deaths.
Cost per illness or death avoided assigns the total cost of a regulatory alternative entirely to
avoided illness or entirely to avoided deaths. Net cost adjusts the cost of regulation for avoided illnesses
and deaths, so that cost of illnesses is considered alone (by removing the benefits of deaths) and vice
versa. Incremental net cost considers only the increase in cost and in avoided illnesses or deaths from
one regulatory alternative to the next (in the direction of the least to the most costly alternative).
In addition to the net cost concept, this cost-effectiveness analysis provides a different
comparison for cost to cases of illnesses and deaths by discounting the number of cases avoided.
Because estimates of future costs are discounted, the number of illnesses and deaths avoided are
discounted to make the cost and benefit data comparable. The practice of discounting outcomes (such as
illnesses and deaths) is less common than discounting the valuation of illnesses and deaths, which is
carried out in Chapter 5 and used in the all other analyses of this chapter.
The numbers of avoided illnesses and deaths are discounted by incorporating the schedule of
benefits incurred, discount rates, value of a statistical life yearly values (for deaths), and cost of illness
yearly values (for illness). Exhibit 8.18 shows the cost per illness avoided, net cost per discounted illness,
and incremental net cost per discounted illness. Exhibit 8.19 follows with cost per death avoided analyses.
Both exhibits show these data by regulatory alternative, Cryptosporidium occurrence data set, cost of
illness method (Enhanced COI (ECOI) or Traditional COI (TCOI)), and discount rate.
To compare against the cost per illness avoided values in Exhibits 8.18, EPA estimated the
average cost of illness over the 20-year evaluation period for both ECOI and TCOI (presented below
Exhibit 8.18). These weighted average values adjust for changes in income growth and the four
implementation schedules used for systems of different sizes. The cost of illness grows due to income
growth because the value of time increases with real income. The four implementation schedules (for
different system size categories) produce four weighted average costs of illness; thus, the results are
expressed here as a range between the highest and lowest COIs (the large and small systems bracket the
range). All costs are expressed in 2003$.
The results of the comparison for net cost per illness show that all regulatory alternatives, except
Al, have costs per avoided illness below the ECOI and TCOI under every discount and occurrence data
set combination. When considering incremental net costs per illness avoided, the Preferred Alternative is
below the ECOI in four of the six combinations, and below the TCOI in two of the combinations. Only
Alternative A4 consistently is below both measures.
Economic Analysis for the LT2ESWTR 8-32 December 2005
-------
Exhibit 8.18: Incremental Net Cost per Discounted Illness Avoided,
By Discount Rate, Data Set, and Alternative
Data
Set
ICR
ICRSSL
ICRSSM
Rule
Alternative
A4
A3 - Preferred
A2
A1
A4
A3 - Preferred
A2
A1
A4
A3 - Preferred
A2
A1
Cost Per
Discounted Illness
Avoided ($)
3%
$ 147
$ 227
$ 275
$ 668
$ 476
$ 661
$ 808
$ 2,258
$ 265
$ 382
$ 473
$ 1,287
7%
$ 309
$ 468
$ 559
$1.322
$1,022
$1,385
$1,663
$4,472
$ 565
$ 796
$ 969
$2,548
Net Cost Per Discounted
Illness Avoided ($)
3%
$ (1,708.3)
$ (1,599.2)
$ (1,546.8)
$ (1.147.8)
$ (1,527.3)
$ (1,257.9)
$ (1,071.0)
$ 447.2
$ (1,660.5)
$ (1,480.9)
$ (1,370.2)
$ (524.4)
7%
$ (1,548.4)
$ (1,360.9)
$ (1,264.8)
$ (495.6)
$ (983.3)
$ (535.8)
$ (218.8)
$ 2,658.8
$ (1,362.7)
$ (1,069.1)
$ (876.4)
$ 734.7
Incremental Net Cost
Per Discounted
Illness Avoided ($)
3%
$ 147
$ 1,408
$ 4,506
$ 26.961
$ 476
$ 1,780
$ 2,574
$ 10,758
$ 265
$ 1,228
$ 2,473
$ 11,480
7%
$ 309
$ 2,816
$ 8,745
$ 52.665
$ 1,022
$ 3,597
$ 5,022
$21,045
$ 565
$ 2,477
$ 4,813
$ 22,442
Comparison Data:
Enchanced COI
Traditional COI
Range of Weighted
Average, 2003$
$ 1,150.74
$ 343.94
$ 1,177.37
$ 349.99
Notes: Cost per Discounted Illness Avoided: cost represents full cost of the rule (i.e., no subtraction for deaths
avoided). Net Cost per Discounted Illness Avoided: cost represents only illnesses avoided (i.e., cost attributed to
avoiding deaths subtracted to produce net cost).
Comparison data: The Traditional COI includes values for medical costs and lost work time (including a portion of
unpaid household production). The Enhanced COI also includes values for lost personal time (non-work time)
such as child care and homemaking (to the extent not covered by the Traditional COI), time with family, and
recreation, and lost productivity on days when workers are ill but go to work anyway.
Sources: Derived from Exhibits C.4 and C.5 (number of illnesses and deaths avoided); Exhibit O.9 (schedule);
and Exhibit 8.11 (total cost of the rule).
Economic Analysis for the LT2ESWTR
December 2005
-------
To compare against the net incremental cost per death avoided values presented in Exhibits 8.19,
EPA has estimated the quantified benefits of preventing a fatality from cryptosporidiosis as $5.0-$5.2
million. When considering the values of deaths avoided, EPA uses a distribution to represent the value of
a statistical life. For this analysis, a weighted average is calculated that incorporates the four
implementation schedules used for systems of different sizes and adjusts for changes in income growth
and income elasticity.
Similar to the cost per illness analyses, Alternatives A3 and A4 consistently have net cost per
death avoided values below the range of VSL estimates under all combinations of discount rates and
occurrence datasets. The incremental net cost analysis shows, as expected, that incremental costs
increase with rule stringency.
Exhibit 8.19: Incremental Net Cost per Discounted Death Avoided,
By Discount Rate, Data Set, and Alternative
Data
Set
ICR
ICRSSL
ICRSSM
Rule
Alternative
A4
A3 - Preferred
A2
A1
A4
A3 - Preferred
A2
A1
A4
A3 - Preferred
A2
A1
Cost Per Discounted
Death Avoided
(SMillions)
3%
$ 0.7
$ 1.1
$ 1.3
$ 3.1
$ 2.0
$ 2.9
$ 3.7
$ 10.6
$ 1.2
$ 1.7
$ 2.2
$ 60
7%
$ 1.4
$ 2.2
$ 2.6
$ 6.2
$ 4.3
$ 6.1
$ 7.5
$ 21.0
$ 2.5
$ 3.6
$ 4.5
$ 119
Net Cost Per Discounted Death Avoided
(SMil ions)
EC
3%
$ (2.5)
$ (2.2)
$ (2.0)
$ (0.2)
$ (1 .0)
$ (0.2)
$ 0.5
$ 7.3
$ (1 .9)
$ (1.5)
$ (1.1)
$ 27
Ol
7%
$ (0.3)
$ 0.4
$ 0.8
$ 4.4
$ 2.7
$ 4.4
$ 5.8
$ 19.2
$ 0.8
$ 1.9
$ 2.7
$ 10 1
TCOI
3%
$ (0.3)
$ 0.1
$ 0.3
$ 2.1
$ 1.1
$ 2.0
$ 2.7
$ 9.6
$ 0.2
$ 0.8
$ 1.2
$ 5 1
7%
$ 0.9
$ 1.6
$ 2.1
$ 5.6
$ 3.8
$ 5.6
$ 7.0
$ 20.4
$ 2.0
$ 3.1
$ 3.9
$ 114
Incremental Net Cost Per Discounted
Death Avoided (SMillions)
EC
3%
$ (2.5)
$ 4.3
$ 23.0
$ 158.7
$ (1.0)
$ 6.5
$ 11.3
$ 60.8
$ (1 .9)
$ 3.2
$ 10.7
$ 65 1
Ol
7%
$ (0.3)
$ 14.7
$ 50.5
$ 315.9
$ 2.7
$ 19.4
$ 28.0
$ 124.9
$ 0.8
$ 12.7
$ 26.8
$ 133 3
TCOI
3%
$ (0.3)
$ 7.2
$ 26.0
$ 161.6
$ 1.1
$ 9.5
$ 14.3
$ 63.7
$ 0.2
$ 6.2
$ 13.7
$ 68 1
7%
$ 0.9
$ 16.3
$ 52.2
$ 317.5
$ 3.8
$ 21.0
$ 29.7
$ 126.5
$ 2.0
$ 14.3
$ 28.4
$ 1349
Comparison Data:
Range of VSL Weighted
Average, $Millions, 2003$
$
4.98 | $
5.16
Notes: Cost per Discounted Death Avoided: cost represents full cost of the rule (i.e., no subtraction for illnesses
avoided). Net Cost per Discounted Death Avoided: cost represents only illnesses avoided (i.e., cost attributed to
avoiding illnesses subtracted to produce net cost).
The Traditional COI includes values for medical costs and lost work time (including a portion of unpaid household
production). The Enhanced COI also includes values for lost personal time (non-work time) such as child care
and homemaking (to the extent not covered by the Traditional COI), time with family, and recreation, and lost
productivity on days when workers are ill but go to work anyway.
Sources: Derived from Exhibits C.4 and C.5 (number of illnesses and deaths avoided); Exhibit O.9 (schedule);
and Exhibit 8.11 (total cost of the rule).
Economic Analysis for the LT2ESWTR
8-34
December 2005
-------
Benefit-Cost Ratios
In addition to an evaluation of cost per illness avoided, EPA has evaluated the benefit/cost ratio
for each alternative. This measure compares the ratio of the overall value of benefits to the overall costs
(including costs to the States). The benefit-cost ratio is used as a threshold measure of cost-
effectiveness. Benefit/cost ratios should exceed 1, that is, benefits should exceed costs. All but one of
the ratios shown in Exhibit 8.20 are above this cost-effectiveness threshold based on mean values of
benefit and cost estimates. This is not surprising given that this proportion is simply another way of
expressing the results of section 8.2.3, National Net Benefits. If the nonqualified benefits were included,
the ratios would all be larger.
Exhibit 8.20: Benefit-Cost Ratios for Each Regulatory Alternative
Data
Set
ICR
ICRSSL
ICRSSM
Rule
Alternative
A1
A2
A3 - Preferred
A4
A1
A2
A3 - Preferred
A4
A1
A2
A3 - Preferred
A4
Benefit/Cost
Ratio (Enhanced
COI)
3%
4.7
11.5
13.9
21.7
1.4
4.0
4.9
7.1
2.4
6.7
8.4
12.3
7%
3.5
8.3
10.0
15.3
1.0
2.9
3.5
4.9
1.8
4.8
5.9
8.5
Benefit/Cost
Ratio (Traditional
COI)
3%
3.4
8.3
10.1
15.8
1.0
2.9
3.6
5.2
1.8
4.9
6.1
9.0
7%
2.5
6.0
7.2
11.1
0.7
2.1
2.6
3.6
1.3
4.3
4.3
6.3
8.4 Effect of Uncertainties on the Benefit-Cost Comparisons
Detailed discussions of the uncertainties and assumptions associated with the national benefits
and costs are contained in Chapters 5 and 6, respectively. Exhibit 8.21 is a summary of the most
important assumptions and their effects on the estimates. It is EPA's judgment that the overall
uncertainties regarding the occurrence of Cryptosporidium in drinking water have been greatly reduced
over the past several years through many data collection efforts, the most significant being the
Information Collection Rule and the ICR Supplemental Surveys. The result is that many uncertainties
have been identified, researched, and where possible, resolved. As can be seen in Exhibit 8.20, the
remaining uncertainties are inherent in the data and represent assumptions where the direction of the
biases is unknown. It is EPA's judgment that the largest uncertainty that affects the conclusions in this
EA is that surrounding the infectivity of C. parvum.
Economic Analysis for the LT2ESWTR
December 2005
-------
Exhibit 8.21: Effects of Uncertainties on the National Estimates of
Benefits and Costs
Assumptions for Which
There Is Uncertainty
Section with
Full
Discussion of
Uncertainty
Effect on Estimates
Under-estimate
Over-
estimate
Under- or
Over-
estimate
Benefits
Not all benefits are quantified
Infectivity for C. pan/urn estimated
from only three known isolates
Source water concentrations
estimated using three data sets,
calculation of central tendencies,
and bounds
Fraction of oocysts that are
infectious (represented by
triangular distribution)
Pre-LT2 removal/inactivation using
triangular distributions (with
uncertain modes)
Value of illnesses avoided based
on COI data rather than WTP data
8.2.1,5.2.3
5.2.3
5.2.4.1
5.2.4.1
5.2.4.1
5.3.1.1
X
X
X
X
X
X
Costs
Using ICR, ICRSSL, and ICRSSM
occurrence distribution data to
predict plant bin assignments
Single flow rate used to evaluate
unit costs within each of 9 size
categories
Potentially lower-cost treatment
options not considered
Typical water quality and operating
parameters used to estimate unit
costs
4.5.3
6.5.1
6.5.1
6.5.1
X
X
X
X
Economic Analysis for the LT2ESWTR
December 2005
-------
8.5 Summary of Benefit and Cost Comparisons
The following is a summary of the important points regarding the potential net benefits of the
LT2ESWTR.
The Preferred Alternative passes key threshold economic criteria:
• The Preferred Alternative (A3) has positive net benefits (Exhibit 8.12). (In fact, this is also
true for Alternatives A2 and A4, and true for Alternative Al under 11 of the 12 possible
scenarios of occurrence, COI value, and discount rate.) In other words, for these
alternatives, benefits are very likely to exceed costs (Exhibit 8.6) and have a benefit-cost ratio
greater than 1.0 (Exhibit 8.19). This conclusion is especially strong because the benefit
estimates do not include the value of the unqualified benefits and are therefore artificially
low.
• The number of illnesses that would have to be prevented in order to break even relative to
costs is well below the mean estimated number of avoided illnesses and deaths (Exhibits 8.3,
8.7a and 8.7b).
• The Preferred Alternative is cost-effective: no other alternative achieves greater benefits at
the same cost or the same benefits at lower cost (Exhibit 8.17).
• The Preferred Alternative is cost-effective based on the cost of the rule and the number of
illnesses and deaths avoided (net cost per discounted death or illness), but only Alternative A4
is below the comparison threshold when considering incremental net cost per illness or death
(Exhibits 8.18 and 8.19).
The Preferred Alternative is the superior alternative across a wide variety of measures
when considering all combinations of occurrence data sets, discount rates, and COI values:
• The Preferred Alternative shows the highest mean net benefits under more conditions than
any other one alternative (Exhibits 8.12, 8.15, and 8.16).
• In the analysis of incremental benefits, the Preferred Alternative is the last alternative having
positive net benefits in 8 of the 12 combinations of cost of illness, discount rate, and data set
(Exhibit 8.13).
The Preferred Alternative is the superior alternative when benefits are near the average
values estimated. At substantially higher levels of estimated benefits, Alternative A2 becomes
economically superior:
• The net benefits of the Preferred Alternative are most often highest near the mean value of
benefits and over much of the range that describes uncertainty in benefits estimates (Exhibits
8.15 and 8.16). If unqualified benefits are determined to be substantially higher, Alternative
A2 would become an increasingly better choice from an economic perspective.
Alternative A3 was recommended by the Stage 2 Microbial Disinfectants and Disinfection
Byproducts (Stage 2 M-DBP) Advisory Committee. Based on this recommendation, and supported by
the evaluations presented above, EPA selected Alternative A3 as the Preferred Alternative.
Economic Analysis for the LT2ESWTR 8-3 7 December 2005
-------
9. References
Abdalla, C.W. 1990. Measuring economic losses from ground water contamination: an investigation of
household avoidance costs. Water Resources Bulletin 26:451-463.
Abdalla, C.W., B.A. Roach, and D.J. Epp. 1992. Valuing environmental quality changes using averting
expenditures: application to groundwater contamination. Land Economics 68:163-69.
Aboytes, Ramon, G. Di Giovanni, F. Abrams, C. Rheinecker, W. McElroy, N. Shaw, and M.W.
LeChevallier. 2004. Detection of Infectious Cryptosporidium in Filtered Drinking Water .
Journal of AWWA 96(9):88-97.
Akiyoshi, D.E., S. Mor, S. Tzipori. 2003. Rapid Displacement of Cryptosporidium parvum Type 1 by
Type 2 in Mixed Infections in Piglets. Infection and Immunity. 71(10):5765-5771.
AWWA (American Waterworks Association) Committee Report. 1983. Deterioration of water quality
in large distribution reservoirs (open reservoirs). AWWA Committee on Control of Water
Quality in Transmission and Distribution Systems. Journal of AWWA 75(4): 191-195.
AWWA. 2000. WATER:\STATS. Database containing 1996 AWWA survey of water systems. Denver:
American Water Works Association. June, 2000.
Amvrosieva, T.V., L.P. Titov, M. Mulders, T. Hovi, O.V. Dyakonova, V.I. Votyakov, Z.B. Kvacheva,
V.F. Eremin, RM. Sharko, S.V. Orlova, O.N. Kazinets, and Z.F. Bogush. 2001. Viral water
contamination as the cause of aseptic meningitis outbreak in Belarus. Central European JPublic
Health 9(3): 154-157.
Barwick, R.S., D.A. Levy, G.F. Craun, M.J. Beach, and R.L. Calderon. 2000. Surveillance for
waterborne-disease outbreaks - United States, 1997-1998. Morbidity and Mortality Weekly
Report 49(SS04):l-35.
Beesley, M. 1965. The value of time spent traveling: some new evidence. Economica 32(2): 174-185.
Berger, Mark C., Glenn C. Blomquist, Don Kenkel, and George S. Tolley. 1987. Valuing changes in
health risks: a comparison of alternative measures. The Southern Economics Journal 53:977-984.
Blackburn, E.G., G.F. Craun, J.S. Yoder, V. Hill, R.L. Calderon, N. Chen, S. H. Lee, D.A. Levy, and M.
J. Beach. 2004. Surveillance for waterborne-disease outbreaks associated with drinking water —
United States, 2001-2002. Morbidity and Mortality Weekly Report 53(SS08):23-45.
Boardman, Anthony E., David H. Greenberg, Aidan R. Vining and David L. Weimer, Cost-Benefit
Analysis Concepts and Practice, 1996, p. 405.
Boutin, B., M. B. Brown, R. Clark, J. C. Lipscomb, R. Miltner, P. Murphy, L. R. Papa, D. Reasoner, and
R. Rheingans. Comparative Risk Framework Methodology and Case Study. U.S. Environmental
Protection Agency, Office of Research and Development, National Center for Environmental
Assessment, Washington, DC, 1998.
Economic Analysis for the LT2ESWTR 9-1 December 2005
-------
Boyle, Brian. 2002. HAART has significantly causes of death in HIV-infected persons.
http://www.hivandhepatitis.eom/2002conf/9thcroi/3 7 .html
Bridgman S.A., Robertson R.M., Syed Q., Speed N., Andrews N., Hunter P.R 1995. Outbreak of
cryptosporidiosis associated with a disinfected groundwater supply. Epidemiol. Infect. 115:555-
566.
Brock, D.W. 2002. "Chapter 14.3: Fairness and Health." In Summary Measures of Population Health:
Concepts, Ethics, Measurement and Applications, C.J.L. Murray, J.A. Salomon, C.D. Mathers,
and A.D. Lopez (eds.). Geneva: World Health Organization.
Brown, D. 2001. Study Finds Drug-Resistant HIV in Half of Infected Patients. Washington Post,
December 19, Page A02.
Brown, E.A.E., Casemore D.P., Gerken A., Greatorex I.F. 1989. Cryptosporidiosis in Great Yarmouth,
the investigation of an outbreak. Public Health 103:3-9.
Bruzelius, Nils. 1979. The value of travel time. London: Croon Helm.
Buzby, J.C., T. Roberts, C.T. Jordan Lin, and J.M. MacDonald. 1996. Bacterial foodborne illness:
medical costs and productivity losses. Economic Research Service, U.S. Department of
Agriculture, Agricultural Economic Report No. 741. August 1996.
Carollo. 1999. Report on water quality issues at Sheffield Reservoir for the City of Santa Barbara.
Casemore, D.P. 1990. Epidemiological aspects of human cryptosporidiosis. Epidemiol. Infect.
104:1-28.
CDC (Centers for Disease Control and Prevention). 2002. HIV/AIDS Surveillance Report vol. 14.
http://www.cdc.gov/hiv
CDC. 2001a. Wonder Database, http://wonder.cdc.gov
CDC. 2001b. HIV/AIDS Surveillance Supplemental Report l(l):l. http://www.cdc.gov/hiv
CDC. 2000. HIV/AIDS Surveillance Report 12(2):7,30. http://www.cdc.gov/hiv
CDC. 1999. HIV/AIDS Surveillance Report 11(2):7,31,37. http://www.cdc.gov/hiv
CDC. 1995. Vital and Health Statistics: Detailed Diagnoses and Procedures, National Hospital
Discharge Survey, 1993. October, 13(122). http://www.cdc.gov/nchs/
CDC. 1993. HIV/AIDS Surveillance Report 5(4):26. http://www.cdc.gov/hiv
Economic Analysis for the LT2ESWTR 9-2 December 2005
-------
CDC. 2004 Fact Sheet on Cryptosporidiosis (accessed January 18, 2005):
www.cdc.gov/ncidod/dpd/parasites/cryptosporidiosis/2004_PDF Cryptosporidiosis.pdf
Cesario, F.J. 1976. Value of time in recreation benefit studies. Land Economics 52:32-41.
Chappell, C.L..et al. 2004. Experimental Challenge of Healthy Adults with the Human Genotype (Type 1)
of Cryptosporidium.
Chappell C.L., P.C. Okhuysen, C.R. Sterling, C. Wang, W. Jakubowski, and H.L. DuPont. 1999.
Infectivity of Cryptosporidium parvum in healthy adults with pre-existing anti-C parvum serum
immunoglobulin G. AmJTropMedHyg. 60:157-64.
Chappell, C.L., H.L. DuPont, and P.C. Okhuysen. 1997. Virulence factors in Cryptosporidium and
infective dose in humans. Progress Report for Grant R82-4759, U.S. EPA, to the University of
Texas-Houston School of Public Health, Center for Infectious Diseases.
Cicirello, H., K. Kehl, D. Addiss, M. Chusid, R. Glass, J. Davis, and P. Havens. 1997. Cryptosporidiosis
in children during a massive waterborne outbreak in Milwaukee, Wisconsin: Clinical, laboratory
and epidemiologic findings. Epidemiol. Infect. 119(1):53-60.
Clancy, J.L., Bukhari, Z., Hargy, T.M., Bolton, J.R., Dussert, B.W., Marshall, M.M.. 2000. Using UV
to inactivate Cryptosporidium. JournalAWWA 92(9):97-104.
Connell, K., C.C. Rodgers, H.L. Shank-Givens, J. Scheller, M.L. Pope, and K. Miller. 2000. Building a
better protozoa data set. Journal of AWWA 92:10:30.
Cooke, G.D. and R.E. Carlson. 1989. Reservoir Management for Water Quality and THM Precursor
Control. AWWARF. Denver, CO.
Cooper, B.S. and D.P. Rice. 1976. The economic cost of illness revisited. Social Security Bulletin
39(2):21-36.
Cordell, R.L., P.M. Thor, D.G. Addiss, J. Theurer, R. Lichterman, S.R Ziliak, D.D. Juranek, and J.P.
Davis. 1997. Impact of a massive waterborne Cryptosporidiosis outbreak on child care facilities
in metropolitan Milwaukee, Wisconsin. Pediatr. Infect. Dis. J. 16(7):639-644.
Corso, P.S., M.H. Kramer, K.A. Blair, D.G. Addiss, J.P. Davis, and A.C. Haddix 2003. Cost of illness in
the 1993 waterborne Cryptosporidium outbreak, Milwaukee, Wisconsin. Emerging Infectious
Diseases 9(4):426-43\.
Craik, S.A., D. Weldon, G.R. Finch, J.R. Bolton, and M Belosevic. 2001. Inactivation of
Cryptosporidium parvum oocysts using medium- and low-pressure ultraviolet radiation. Water
Research 35(6): 1387-1398.
Economic Analysis for the LT2ESWTR 9-3 December 2005
-------
Craun, G.F. 1996. Task 1. Cases of illness required to initiate a waterborne outbreak investigation.
September 18. Task 2. Epidemiological and water supply information from waterborne outbreaks
of giardiasis and levels of cysts detected in water during outbreaks. September 23. Task 3.
Epidemiological and water supply information from waterborne outbreaks of cryptosporidiosis
and levels of oocysts detected in water during outbreak investigations. September 20. Task 4.
Waterborne disease in the USA: reported waterborne outbreaks, 1993-94. September 27. 68-C3-
0368. Wade Miller Associates, Inc.
Craun, G.F. and R. Calderon. 1996. Microbial risks in groundwater systems: epidemiology of
waterborne outbreaks. In: Under the Microscope: Examining Microbes in Groundwater. Boston:
Proceedings of the Groundwater Foundation's 12th Annual Fall Symposium.
Craun, G.F., S. Hubbs, F.J. Frost, R. Calderon, and S. Via. 1998. Waterborne outbreaks of
cryptosporidiosis. Journal of A WWA 90(9):81-91.
Crystal-Peters, J., W.H. Crown, R.Z. Goetzel, B.C. Schutt. 2000. The cost of productivity losses
associated with allergic rhinitis. The American Journal of Managed Care 6(3):373-378.
CWC (Culp/Wesner/Culp Engineering). 1994. W/W Costs and design criteria guidelines 2.0, computer
software for estimating water and wastewater technology costs. San Clemente, CA: CWC
Engineering Software.
de Melker, H.E., F.G.A. Versteegh, M.A.E. Conyn-van Spaendonck, L.H. Elvers, G.A.M. Berbers, A. van
der Zee. 2000. Specificity and sensitivity of high levels of IgG antibodies against pertussis toxin
in a single serum for diagnosis of infection with Bordetella pertussis. J. Clin. Microbiol. 38:800-
806.
Douglas, J., G. Kenny, and T.R. Miller. 1990. Which estimates of household production are best?
Journal of Forensic Economics 4:25-45.
Dugan, N.R, K.R. Fox, R.J. Miltner, D.A. Lytle, D.J. Williams, C.J. Parrett, C.M. Feld, and J.H. Owens.
1999. Control ofCryptosporidium oocysts by steady-state conventional treatment. Cincinnati:
EPA Water Supply and Water Resources Division. Annual Conference Proceedings of the
American Water Works Association. #ACE99567.
DuPont, H.L., C. Chappell, C. Sterling, P. Okhuysen, J. Rose, and W. Jakubowski. 1995. The infectivity
of Cryptosporidium parvum in healthy volunteers. New England J. Med. 332:855-859.
DynCorp. 2002. E-mail dated May 22 from Kevin Connell to Curtis Haymore at The Cadmus Group,
Inc., Alexandria, Virginia.
DynCorp I&ET. 2000. E-mail to EPA. July 6, 2000.
D'Antonio, R.G., R.E. Winn, J.P. Taylor, T.L. Gustafson, W.L. Current, M.M. Rhodes, G.W. Gary, and
R.A. Zajac. 1985. A waterborne outbreak of cryptosporidiosis in normal hosts. Ann. Intern.
Med. 103:888.
Economic Analysis for the LT2ESWTR 9-4 December 2005
-------
Edzwald, J.K. and M.B. Kelley. 1998. Control of Cryptosporidium from reservoirs to clarifiers to filters.
Water Si. Technol. 3 7(2): 1-8.
Egerov, A., F. Frost, T. Muller, E. Naumova, A. Tereschenko, T. Ford. 2003. Serological Evidence of
Cryptosporidium Infections in a Russian City and Evaluation of Risk Factors for Infections.
AEP. 14(2): 129-136.
Payer, R. and B.L.P. Ungar. 1986. Cryptosporidium spp. and cryptosporidiosis. Microbiol. Rev.
50(4):458-483.
Framm, S.R and R. Soave. 1997. Agents of diarrhea. Med. Clin. N. Amer. 81(2):427-447.
Freeman, A.M. 1979. The Benefits of Environmental Improvement: Theory and Practice. Baltimore:
Johns Hopkins University Press.
Freeman, A.M. III. 1993. The measurement of environmental and resource values: theory and methods.
p.6. Washington, DC: Resources for the Future.
Frisby H.R, D.G. Addiss, W.J. Reiser, B. Hancock, J.M. Vergeront, N.J. Hoxie, J.P. Davis. 1997.
Clinical and epidemiologic features of a massive waterborne outbreak of cryptosporidioisis in
persons with HIV infection. J. Acq. Immun. Def. Synd. 16(5):367-373.
Frost, F.J., RL. Calderon, T.B. Muller, M. Curry, J.S. Rodman, D.M. Moss, A.A. de la Cruz. 1998. A
two-year follow-up survey of antibody to Cryptosporidium in Jackson County, Oregon following
an outbreak of waterborne disease. Epidemiological Infection. 121: 213-217.
Frost, F., G.F. Craun, R. Calderon, and S.A. Hubbs. 1997. So many oocysts, so few outbreaks. Journal
ofAWWA 89(12):8-10.
Furomoto, W.A. and Mickey, R. 1967a. A mathematical model for the infectivity-dilution curve of
tobacco mosaic virus: theoretical considerations. Virology 32: 216.
Furomoto, W.A. and Mickey, R. 1967b. A mathematical model for the infectivity-dilution curve of
tobacco mosaic virus: experimental tests. Virology 32: 224.
Gerba, C.P, J.B. Rose, and C.N. Haas. 1996. Sensitive populations: who is at the greatest risk? Int. J.
Food Microbiol 30(1-2): 113-123.
Gold, M. R., J.E. Siegel, L.B. Russell, and M.C. Weinstein (eds.). 1996. Cost-effectiveness in health and
medicine. New York: Oxford University Press.
Economic Analysis for the LT2ESWTR 9-5 December 2005
-------
Gold, M.R., D. Stevenson, and D.G. Fryback. 2002. "HALYS and QALYS and DALYS, Oh My:
Similarities and Differences in Summary Measures of Population Health." Annual Review of
Public Health 23:115-34
Goldschmidt-Clermont, L. 1982. Unpaid work in the household, p. 4. Prepared for the International
Labor Office, United Nations.
Goodrich, J.A., SY. Li, and B.W. Lykins Jr. 1995. Cost and Performance of Alternative Filtration
Technologies for Small Communities. Proceedings of the 1995 AWWA Annual Conference.
Griffith, C. 1988. Floating Reservoir Cover Controls Algae Growth, Allows Lowered THM's. Journal
ofAWWA 80(10):66,68-69.
Griffiths, J.K. 2002. Personal Communication. October 23, 2002.
Gronau, R. 1986. Home production - a survey. Handbook of Labor Economics Volume 1. Orley
Ashenfelter and Richard Layard, eds. New York: North-Holland.
Grunberg, S.M., Boutin, N., Ireland, A., Miner, S., Silvera, J., and Ashikaga, T., "Impact of
Nausea/Vomiting on Quality of Life as a Visual Analogue Scale-Derived Utility Score," Support
Care Cancer, November 1996, 4(6): 435-439.
Grunberg, S.M., Srivastava, A., Grunberg, K.J., Weeks, J., "Intensity of Chemotherapy-Induced Emesis
and Overall Survival as Determinants of a Global Utility Score," Support Care Cancer, November
2002, 10(8): 624-629.
Guerrant, R.L. 1997. Cryptosporidiosis: An emerging, highly infectious threat. Emerg. Infect. Dis.
3(1): 1-9. January-March. www.cdc.gov/ncidod/EID/vol3nol/guerrant.htm
Haas, C.N., Rose, J.B. and Gerba C.P. 1999. Quantitative Microbial Risk Assessment. John Wiley and
Sons.
Haas, C.N., C.S. Crockett, J.B. Rose, C.P. Gerba and A.M. Fazil. 1996. Assessing the Risk Posed by
Oocysts in Drinking Water. Journal ofAWWA 88(9): 131-136.
Haas, C.N., and J.B. Rose. 1994. Reconciliation of Microbial Risk Models and Outbreak Epidemiology:
The Case of the Milwaukee Outbreak. Proceedings of the A WWA Annual Conference, New York:
517-521.
Holcomb, DL, Smith, MA, Ware, GO, Hung, Y-C, Brackett, RE and Doyle, MP. 1999. Comparison of
six dose-response models for use with food-borne pathogens. Risk Analysis. 19(6): 1091-1100.
Harrington, W., A. J. Krupnick, and W.O. Spofford. 1991. Economics and Episodic Disease: The
Benefits of Preventing a Giardiasis Outbreak. Washington, DC: Resources for the Future.
Economic Analysis for the LT2ESWTR 9-6 December 2005
-------
Hartunian, N.S., C.N. Smart, and M.S. Thompson. 1981. The incidence and economic costs of major
health impairments: a comparative analysis of cancer, motor vehicle injuries, coronary heart
disease, and stroke. An Institute for Highway Safety Book. p. 45. Lexington, MA: Lexington
Books.
Havelaar, A.H., de Hollander, A.E.M., Teunis, P.F.M., Evers, E.G., van Kranen, H.J., Versteegh, P.M.,
van Koten, J.E.M., and Slob, W., "Balancing the Risks and Benefits of Drinking Water
Disinfection: Disability Adjusted Life-Years on the Scale," Environmental Health Perspectives,
108(4): 315-321, April 2000.
Heijbel H, K. Slaine, B. Seigel, P. Wall, S.J.N McNabb, W. Gibbons, G.R. Istre. 1987. Outbreak of
diarrhea in a day care center with spread to household members: the role of Cryptosporidium.
PediatrInfectDis J. 6(6):532-535.
Highleyman, L. 2000. Adverse Effects Associated with Antiretroviral Therapy. Bulletin of
Experimental Treatments for AIDS. April.
http://www.aegis.com/pubs/beta/2000/BE000405.html
Hoxie, N.J., J.P. Davis, J.M. Vergeront, R.D. Nashold, and K.A. Blair. 1997. Cryptosporidiosis-
associated mortality following a massive waterborne outbreak in Milwaukee, Wisconsin. Amer.
J. Publ. Health 87(12):2032-2035.
Hunter, P.R., S. Hughes, S. Woodhouse, N. Raj, Q. Syed, R. Chalmers, N. Verlander, J. Goodacre. 2004.
Health Sequelae of Human Cryptosporidiosis in Immunocompetent Patients. CID. 15:504-510.
Hunter, P.R. and Q. Syed. 2001. Community Surveys of self-reported diarrhoea can dramatically
overestimate the size of outbreaks of waterborne Cryptosporidiosis. Water Science and
Technology 43(12):27-30
Industrial Economics, Incorporated. 2001. Assessing the Benefits of Drinking Water Regulations (Draft
Final Report). Prepared for the U.S. Environmental Protection Agency. June 2001.
Industrial Economics, Incorporated. 2002. Valuing Time Losses Due to Illness under the 1996
Amendments to the Safe Drinking Water Act (Draft Final). Prepared for the Office of Ground
Water and Drinking Water, U.S. Environmental Protection Agency. October 2002.
Inungu, J.N., A.A. Morse, and C. Gordon. 2000. Risk factors, seasonality, and trends of cryptosporidisois
among patients infected with human immunodeficiency virus. Am. J. Trop. Med. Hyg. 62(3):384-
387.
Juranek, D. 1995. Cryptosporidiosis: Sources of infection and guidelines for prevention. Clin. Infect.
Dis. 21(Suppl. 1):S57-S61.
Economic Analysis for the LT2ESWTR 9-7 December 2005
-------
Karon, J.M., P.L. Fleming, R.W. Steketee, and K.M. De Cock. 2001. HIV in the United States at the
Turn of the Century: An Epidemic in Transition. Am. J. Pub. Health 91(7): 1060-68.
Keech, M., A.J. Scott, and P.J.J. Ryan. 1998. The impact of influenza and influenza-like illness on
productivity and healthcare resource utilization in a working population. Occupational Medicine
48(2):85-90.
Kessler, R.C., C. Barber, H.G. Birnbaum, R.G. Frank, P.E. Greenberg, R.M. Rose, G.E. Simon, and P.
Wang. 1999. Depression in the workplace: effects on short-term disability. Health Affairs
18(5):163-171.
Kim, K.S., G. Hufnagel, N.M. Chapman, and S. Tracy. 2001. The group B coxsackieviruses and
myocarditis. Rev. Med. Virol. ll(6):355-68.
Kleckner, N. and J. Neumann. 2000. Memorandum: Update to Recommended Approach to Adjusting
WTP Estimates to Reflect Changes in Real Income. September 30, 2000.
Kocagil P., N. Demarteau, A. Fisher, and J.S. Shortle. 1998. The Value of Preventing Cryptosporidium
Contamination Risk. Health, Safety, & Environment 175(Spring 1998).
Kramer, M.H., B.L. Herwaldt, R.L. Calderon, and D.D. Juranek. 1996a. Surveillance for
Waterborne-Disease Outbreaks—United States, 1993-1994. MMWR 45(SS-l):l-33.
Kramer, M.H., B.L. Herwaldt, G.F. Craun, RL. Calderon, and D.D. Juranek. 1996b. Waterborne
Disease: 1993 and 1994. Journal of A WWA 88(3):66-80.
Larson, D.M. 1993. Joint Recreation choices and implied values of time. Land Economics 69(3):270-
286.
LeChevallier, M.W., G.D. Di Giovanni, J.L. Clancy, Z. Bukhari, S. Bukhari, J.S. Rosen, J. Sobrinho, and
M.M. Frey. 2003. Comparison of method 1623 and cell culture-PCR for detection of
Cryptosporidium spp. in source waters. Appl. Environ. Microbiol 69(2):971-979.
LeChevallier, M.W., J.L. Clancy, Z. Bukhari, S. Bukhari, T. Hargy, J.S. Rosen, J. Sobrinho, and M.M.
Frey. 2000. Source Water Assessment: Variability of Pathogen Concentrations (Paper presented
at 2000 WQTC).
LeChevallier, M.W., W. North, M. Abbaszadegan, T. Atherhold, and J. Rosen. 1997a. Variations in
Giardia and Cryptosporidium in Source water: Statistical Approaches to Analyzing ICR Data.
(Paper presented at 1997 WQTC).
LeChevallier, M.W., W.D. Norton, and T.B. Atherholt. 1997b. Protozoa in open reservoirs. Journal of
AWWA. 89(9):84-96.
Economic Analysis for the LT2ESWTR 9-8 December 2005
-------
Labutiuk, C.W., F.W. Schaefer, III, G.R. Finch, and M. Belosevic. 1991. Comparison of animal
infectivity, excystation and fluorogenic dye as measure of Giardia muris cyst inactivation by
ozone. Appl. Environ. Microbiol. 57:3187-3192.
Lee, S.H. and S.J. Kim. 2002. Detection of infectious enteroviruses and adenoviruses in tap water in
urban areas in Korea. Water Research 36(l):248-256.
Lee, S.H., D.A. Levy, G.F. Craun, M.J. Beach, and R.L. Calderon. 2002. Surveillance for waterborne
disease outbreaks, United States, 1999-2000. Morbidity and Mortality Weekly Report,
Surveillance Summaries 51(SS08): 1-28.
Levy, D.A., M.S. Bens, G.F. Craun, R.L. Calderon, and B.L. Herwaldt. 1998. Surveillance for
Waterborne Disease Outbreaks - United States, 1995-1996. MMWR 47(55-5): 1-34.
Lindeburg, M.R. 1997. Civil engineering reference manual for the PE exam. 6th ed. p. 7-17. Belmont,
CA: Professional Publications, Inc.
Louie, J., L.C. Hsu, D.H. Osmond, M.H. Katz, S.K. Schwarcz. 2002. Trends in causes of death in
persons with AIDS in the era of High Active Antiretrovial Therapy, San Francisco: The
importance of risk factor reduction. UCSF AIDS Research Institute.
http://ari.ucsf.edu/pdf/Posters/barcelona/louie.pdf (10/30/02).
MacKenzie, W.R, W.L. Schell, KA. Blair, D.G. Addiss, D.E. Peterson, N.J. Hoxie, J.J. Kazmierczak, J.P.
Davis. 1995a. Massive outbreak of waterborne Cryptosporidium infection in Milwaukee,
Wisconsin: Recurrence of illness and risk of secondary transmission. Clinical Infect. Diseases
21:57-62.
MacKenzie, W.R., J.J. Kazmierczak, J.P. Davis. 1995b. An outbreak of cryptosporidiosis associated
with a resort swimming pool. Epidemiol. Infect. 115:545-553.
MacKenzie, W.R., N.J. Hoxie, M.E. Proctor, M.S. Gradus, K.A. Blair, D.E. Peterson, J.J. Kazmierczak,
D.G. Addiss, K.R. Fox, J.B. Rose, and J.P. Davis. 1994. A massive outbreak in Milwaukee of
Cryptosporidium infection transmitted through the public water supply. New England Jour. Med.
331(3):161-167.
Makri, A, Modarres, R, Parkin, R. 2004. Cryptosporidiosis susceptibility and risk: a case study. Risk
Analysis. 24(1): 209-220.
Marble, M. 2000. AIDS Therapies: Relapse is Rapid after Cessation of Therapy. AIDSWeekly Plus,
January 24th. http://www.aegis.com/pubs/aidswkly/2000/AW000109.html (10/30/02).
Mauskopf, J.A. and French, M.T., "Estimating the Value of Avoiding Morbidity and Mortality from
Foodborne Illnesses," Risk Analysis, 11 (4):619-631, December 1991.
Economic Analysis for the LT2ESWTR 9-9 December 2005
-------
McConnell, K.E., and I. Strand. 1981. Measuring the cost of time in recreation demand analysis: an
application to sportfishing. American Journal of Agricultural Economics 63:153-156.
McDonald, A., W. MacKenzie, D. Addiss, M. Gradus, G. Linke, E. Zembrowski, M. Hurd, M. Arroweed,
P. Lammie, and J. Priest. 2001. Cryptosporidium parvum-specific antibody responses among
children residing in Milwaukee during the 1993 waterborne outbreak. Jour. Infectious Diseases
183:1373-1379.
McDowell, I. and Newell, C., Measuring Health: A Guide to Rating Scales and Questionnaires, Oxford
University Press: New York, 1996.
McMenamin, P. 1994. Costs of hay fever in the United States in 1990. Annals of Allergy 73:35-39.
Mead, P.S., Slutsker, L., Dietz, V., McCaig, L.F., Bresee, J.S., Shapiro, C., Griffin, P.M., and Tauxe,
R.V. 1999. Food-Related Illness and Death in the United States. Center for Disease Control
and Prevention 5(5):607-625.
Medscape. 2002. Studies Involving Complete Treatment Withdrawal or Interruption.
http://www.medscape.com/viewarticle/408261_3 (10/30/02).
Millard, P.S., K.F. Gensheimer, D.G. Addiss, D.M. Sosin, G.A. Beckett, A. Houckjankoski, A. Hudson.
1994. An outbreak of cryptosporidiosis from fresh-pressed apple cider. JAMA 272:1592-1596.
Ministry of Women's Affairs, Statistics New Zealand. 1999. New Zealand Time Use Survey, p. 22-23.
December 1999.
Moore, A.C., B.L. Herwaldt, G.F. Craun, R.L. Calderon, A.K. Highsmith, and D.D. Juranek. 1993.
Surveillance for waterborne disease outbreaks - United States, 1991-1992. MMWR. 42(SS-5):1-
22 (November 19).
Morgan, D., M. Allaby, S. Crook, D. Casemore, T.D. Healing, N. Soltanpoor, S. Hill, and W. Hooper.
1995. Waterborne cryptosporidiosis associated with a borehole supply. Communicable Disease
Report/CDR Review 5(7):R93-R97.
Morgan-Ryan, U., A. Fall, L. Ward, N. Hijjawi, I. Sulaiman, R. Payer, R. Thompson, M. Olson, A. Lai,
and L. Xiao. 2002. Cryptosporidium hominis n. sp. (Apicomplexa: Cryptosporidiidae) from
Homo sapiens. Jounal of Eukaryotic Microbiology. 49:6:443-440.
Morra, J.J. 1979. A review of water quality problems caused by various open distribution storage
reservoirs, p. 316-21.
Muller, T.B., F.J. Frost, G.F. Craun, and R.L. Calderon. 2001. Serological Responses to
Cryptosporidium Infection. Letters to the Editor. Infection and Immunity. 69:3.
Economic Analysis for the LT2ESWTR 9-10 December 2005
-------
National Drinking Water Advisory Council. 2001. Report of the Arsenic Cost Working Group to the
National Drinking Water Council. August 14, 2001.
National Research Council, and Commission on Behavioral and Social Sciences and Education. 2000.
Time use measurement and research: report of a workshop. Committee on National Statistics,
Michele Ver Ploeg, Joseph Altonji, Norman Bradburn, Julie DaVanzo, William Nordhaus, and
Francisco Samaniego (eds.). Washington, D.C: National Academy Press.
National Research Council. 1997. Safe water from every tap. National Academy Press, Washington,
DC.
Naumova, E.N, A.I. Egorov, R.D. Morris, and J.K. Griffiths. 2003. The elderly and waterborne
Cryptosporidium infection: gastroenteritis hospitalizations before and during the 1993 Milwaukee
outbreak. Emerging Infectious Diseases 9(4):418-425.
NJMS (New Jersey Medical School), NJMS National Tuberculosis Center. 1996. Brief History of TB.
http://www.umdnj.edu/~ntbcweb/history.htm (July 23).
NYSDHAI (New York State Department of Health AIDS Institute). 2002. Criteria for the Medical Care
of Adults with HIV Infection: Antiretroviral Therapy for Adults.
http://www.hivguidelines.org/public_html/CENTER/clinical-guidelines/adult_hiv_guidelines/sup
plemental_pages/adult_arv/adults_arv_page3.htm. February 2002.
Nord, Erik, "Cost-Value Analysis in Health Care: Making Sense of QALYS," Cambridge University
Press, 1999, p. 50.
Office of Management and Budget. 2003. Circular A-4, (September).
Okhuysen, P.C., C.L. Chappell, J.H. Crabb, C.R. Sterling, H.L. DuPont. 1999. Virulence of Three
Distinct Cryptosporidium parvum Isolates for Healthy Adults. The Journal of Infectious Diseases.
180:1275-81.
Okhuysen, P.C., C.L. Chappell, C.R. Sterling, W. Jakubowski, and H. L. DuPont. 1998. Susceptibility
and serologic response of health adults to reinfection with Cryptosporidium parvum. Infect.
Immun. 66(2): 441-443.
Okun, D., G. Craun, J. Edzwald, J. Gilbert, and J. Rose. 1997. New York City: To filter or not to filter?
Journal of AWWA 89(3):62-74.
Osewe P., D.G. Addiss, K.A. Blair, A. Hightower, M.L. Kamb, J.P. Davis. 1996. Cryptosporidiosis in
Wisconsin: a case-control study of post-outbreak transmission. Epidemiol Infec. 117(2):297-304.
Economic Analysis for the LT2ESWTR 9-11 December 2005
-------
Park, S.R., W.G. Mackay, and B.C. Reid. 2001. Helicobacter sp. recovered from drinking water biofilm
sampled from a water distribution system. Water Research 35(6): 1624-1626.
Partnership for Safe Water. 2002. Partnership for safe water annual data summary report. January
2002.
Patania, N.L., Jacangelo, J.G., Cummings, L., and Wilczak, A. 1995. Removal of Cryptosporidium,
Giardia, and Particles by Granular Media Filtration in Pilot Studies. Proceedings from 1995
AWWA Annual Conference.
Pauley, M.V., S. Nicholson, J. Xu, D. Polsky, P.M. Danzon, J.F. Murray, M.L. Berger. 2002. A general
model of the impact of absenteeism on employers and employees. Health Economics 11(3):221-
231.
Payment, P., J. Siemiatycki, L. Richardson, R. Gilles, E. Franco, and M. Prevost. 1997. A Prospective
epidemiological Study of Gastrointestinal Health Effects Due to the Consumption of Drinking
Water. International J. Environmental Health Research. 7:5-31.
Perz J., F. Ennever, and S. Le Blanq. 1998. Cryptosporidium in tap water; Comparison of predicted risks
with observed levels of disease. Amer. J. Epidem. 147:289-301.
Pluntze, J.D. 1974. Health Aspects of Uncovered Reservoirs. Journal of AWWA. 66(8):432-37.
Posnett, J and S. Jan. 1996. Indirect cost in economic evaluation: the opportunity cost of unpaid inputs.
Health Economics 5:13-23.
Pouillot, R, Beaudeau, P, Denis, J-B, and Drouin, F. 2004. A quantitative risk assessment of waterborne
Cryptosporidiosis in France using second-order Monte Carlo simulation. Risk Analysis. 24(1):
1-17.
Pozio, E., G. Rezza, A. Boschini, P. Pezzotti, A. Tamburrini, P. Rossi, M. Di Fine, C. Smacchia, A.
Schiesari, E. Gattei, R. Zucconi, and P. Ballarini. 1997. Clinical Cryptosporidiosis and human
immunodeficiency virus (HlV)-induced immunosuppression findings from a longitudinal study of
HIV-positive and HIV-negative former injection drug users. J. Infect. Dis. 176:969-975.
Ray, N.F., M. Thamer, E. Gardner, J.K. Chan, R. Kahn. 1998. Economic consequences of Diabetes
Mellitus in the U.S. in 1997. Diabetes Care 21(2):296-309.
Regli, S, Rose, JB, Haas, CN and Gerba CP. 1991. Modeling the risk from Giardia and Viruses in
Drinking Water. Journal AWWA. November. 76-84.
Reid, M. 1934. Economics of Household Production. New York: John and Wiley Sons, Inc.
Economic Analysis for the LT2ESWTR 9-12 December 2005
-------
Rice, D.P. 1966. Estimating the cost of illness. Health Economic Series No. 6. PHS Publication No.
947-6. Washington, DC: U.S. Government Printing Office. May 1966.
Rice, D.P. and W. Max. 1992. The cost of smoking in California, 1989. Sacramento, California:
California State Department of Health Services.
Rochelle, P.A., M.M. Marshall, J.R. Mead, A.M. Johnson, D.G. Korich, J.S. Rosen, R. De Leon. 2002.
Comparison of In Vitro Cell Culture and a Mouse Assay for Measuring Infectivity of
Cryptosporidium parvum. Applied and Environmental Microbiology. 68(8):3809-3817.
Rose, J.B. 1997. Environmental ecology of Cryptosporidium and public health implications. Annual
Rev. Public Health 18:135-161.
Rowe, R.D. and L.G. Chestnut. 1986. Valuing changes in morbidity: WTP vs. COI measures. Presented
at the American Economics Association annual meeting. December 1986.
SAB. 2003. SAB Executive Committee Review Draft (Jan 14-15, 2003 Meeting).
SAB and USEPA. 1990. Reducing risk: setting priorities and strategies for environmental protection.
Washington, DC: U.S. EPA Science Advisory Board. SAB-EC-90-021.
Sepulveda, A.R. and D.Y. Graham. 2002. Role of Helicobacter pylori in gastric carcinogenesis.
Gastroenterol. Clin. North America 31(2):517-535.
Shaw, W.D. and P. Feather. 1999. Possibilities for including the opportunity cost of time in recreation
demand systems. Land Economics 75(4):592-602.
Silverman, G.S., L.A. Nagy, and B.H. Olson. 1983. Variations in Particulate Matter, Algae, and Bacteria
in Uncovered Finished Drinking Water Reservoirs. Journal ofAWWA 75(4): 191-195.
Small, K.A. 1992. Urban transportation economics. Chur, Switzerland: Harwood Academic Publishers.
Statistics Canada. 1999. "Average time spent on activities, total population and participants, by sex,"
General Social Survey, 1998. http://www.statcan.ca/english/Pgdb/People/Families /famil36a.htm
as viewed August 2002.
Statistics New Zealand. 1999. "Table 19: Average minutes per day spent on each individual activity
counting only primary activities, with subtotals giving the four major types of time, by sex, cross-
classified by family type, for all people aged 12 and over living in private households." New
Zealand Time Use Survey, Ministry of Women's Affairs.
Economic Analysis for the LT2ESWTR 9-13 December 2005
-------
Sun, H., J.C. Bergstrom, and J.H. Dorfman. 1992. Estimating the Benefits of Groundwater
Contamination Control. Southern Journal of Agricultural Economics 63-71.
Tangerman, R.H., S. Gordon, P. Wiesner, and L. Kreckman. 1991. An outbreak of cryptosporidiosis in a
day-care center in Georgia. American Jour, of Epidemiology 133:471-476.
Teunis P.F.M., C.L. Chappell, and P.C. Okhuysen. 2002. Cryptosporidium dose response studies:
variation between isolates. Risk Analysis 22(1): 175-183.
Teunis, PFM and Havelaar, AH. 2002. The Beta Poisson dose-response model is not a single hit model.
Risk Analysis. 20(4): 513-520.
Tolley, G., D.Kenkel, and R. Fabian (eds.). 1994. Valuing health for policy: an economic approach.
University of Chicago Press.
Tolley, G.S. and L. Babcock. 1986. Valuation of reductions in human health symptoms and risks.
University of Chicago. Final report for the U.S. Environmental Protection Agency. January.
Torrance, G.W., Zhang, Y., Feeny, D., Furlong, W., Barr, R.. 1992. Multi-attribute preference functions
for a comprehensive health status classification system. CHEPA Working Paper Series No.
92-18, McMaster University.
Trewin, D. 2000. Unpaid work and the Australian economy: 1997. Australian Bureau of Statistics.
October 2000.
Ungar W.J. and P.C. Coyte. 2000. Measuring productivity loss days in asthma patients. Health
Economics 9:37-46.
U.S. Bureau of Labor Statistics. Undated. American Time Use Survey: frequently asked questions.
http://www.bls.gov/tus/atusfaqs.htnrfQA4 (August 2002).
U.S. Bureau of Labor Statistics. Undated. "Table 2: Income before taxes: Average annual expenditures
and characteristics." Consumer Expenditure Survey, 2000.
U.S. Bureau of Labor Statistics. Undated. "Table 39: Median weekly earnings of full-time wage and
salary workers by detailed occupation and sex." Current Population Survey.
U.S. Bureau of Labor Statistics. 2001. 2000 National Occupational Employment and Wage Estimates.
"17-2051 Civil Engineers" and "51-8031 Water and Liquid Waste Treatment Plant and System
Operators." http://stats.bls.gov/oes/2000/oes_nat.htm.
Economic Analysis for the LT2ESWTR 9-14 December 2005
-------
U.S. Bureau of Labor Statistics. 2002. "Table 626: Employer costs for employee compensation per hour
worked: 2001." Statistical Abstract of the United States: 2001.
U.S. Bureau of Labor Statistics. 2001. Employment cost index, http://www.bls.gov.
U.S. Bureau of Labor Statistics. 2004. Consumer Price Index for All Urban Consumers, Not Seasonally
Adjusted, http://www.bls.gov.
U.S. Census Bureau. 2004. No. 11. Resident population by age and sex: 1930 to 2002; No. 12. Resident
population projections by sex and age: 2005 to 2050. Statistical Abstract of the United States:
2003. pp. 13-14. http://www.census.gov/prod/2004pubs/03statab/pop.pdf
U.S. Census Bureau. 2002. Table RDM. "Household income before and after taxes: 1979 to 2000."
March 2002. http://www.census.gov/hhes/income/histinc/rdi01xl.html.
U.S. Census Bureau. 2001a. Households and Families: 2000. Census 2000 Brief. C2KBR/01-8.
U.S. Census Bureau. 2001. Statistical Abstract of the United States: 2001. Washington, DC: U.S.
Department of Commerce. "Table No. 567. Employment Status of the Civilian Population: 1960
to 2000;" "Table No. 582. Persons at Work by Hours Worked: 2000;" "Table No. 621. Full-Time
Wage and Salary Workers—Number and Earnings: 1985 to 2000." November 2001.
U.S. Census Bureau. 2000. Statistical Abstract of the United States: 2000. Washington, DC: U.S.
Department of Commerce. "Table 701. Employer Costs for Employee Compensation Per Hour
Worked: 2000." December 2000.
U.S. Council of Economic Advisors. 2002. Economic Report of the President: 2002. Washington, DC
"Table B-31 Total and per capita disposable personal income and personal consumption
expenditures, and per capital gross domestic product, in current and real dollars, 1959-2001."
U.S. Department of Agriculture. 1994-96 Continuing Survey of Food Intakes by Individuals (CSFII).
http://www.usda.gov
U.S. Department of Commerce, Bureau of Economic Analysis. 2004a. Table 1.1.6. Real Gross Domestic
Product, Chained Dollars (Billions of chained (2000) dollars), http://www.bea.gov
U.S. Department of Commerce, Bureau of Economic Analysis. 2004b. Table 3.9.4 Price Indexes for
Government Consumption Expenditures and Gross Investment, http://www.bea.doc.gov
U.S. Department of Commerce. 1992. 7992 Census of Governments, GC92(4)-4: Finances of Municipal
and Township Governments. U.S. Dept. of Commerce, Bureau of the Census.
Economic Analysis for the LT2ESWTR 9-15 December 2005
-------
U.S. Department of Energy, Energy Information Administration. 2004a. Table ES1.A. Total Electric
Power Industry Summary Statistics, 2004 and 2003.
http://www.eia.doe.gov/cneaf/electricity/epm/tableesla.html
U.S. Department of Energy, Energy Information Administration. 2004b. Table 7.1. Electricity Overview
(Billion Kilowatthours). http://www.eia.doe.gov/emeu/mer/txt/mer7-l
U.S. Department of Labor. Undated. "Fact sheet #23: Overtime pay requirements of the FLSA."
http://www.dol.gov/esa /regs/compliance/whd/whdfs23.htm as viewed August 2002.
USEPA. 2005. Regulatory Impact Analysis for the Final Clean Air Interstate Rule, EPA-452/R-05-002,
Appendix G, pp. 369 - 420 (March).
USEPA. 2004a. Information Collection Request for the Long Term 2 Enhanced Surface Water
Treatment Rule. Prepared by The Cadmus Group, Inc. U.S. EPA Contract 68-C-02-026.
October 2004.
USEPA. 2004b. Cost and Technology Document for the Ground Water Rule. Prepared by The Cadmus
Group, Inc., Arlington, VA. U.S. EPA Contract 68-C-99-206. 2004.
USEPA 2003a. Technologies and Costs for Control ofMicrobial Contaminants and Disinfection By-
products. Office of Ground Water and Drinking Water. Standards and Risk Reduction Branch.
Standards and Risk Management Division. U.S. EPA Contract 68-C-02-026. May 2003.
USEPA. 2003b. Labor Costs for National Drinking Water Rules.
USEPA. 2003c. Draft Occurrence and Exposure Assessment for the Long Term 2 Enhanced Surface
Water Treatment Rule. Prepared by The Cadmus Group, Inc. Arlington, VA. U.S. EPA Contract
68-C-02-026. June 2003.
USEPA. 2003d. Draft Economic Analysis for the Stage 2 Disinfectants and Disinfection Byproducts
Rule. Prepared by The Cadmus Group, Inc. Arlington, VA. U.S. EPA Contract 68-C-02-026
July 2003.
USEPA. 2003e. SDWISDatabase. September 2003 SDWIS freeze data, adjusted for population errors.
USEPA. 2002. Community Water System Survey 2000. December 2002.
USEPA. 2001a. Assessing the Benefits of Drinking Water Regulations. June 2001.
USEPA. 200Ib. National Rural Water Association Survey User Database. Prepared by The Cadmus
Group, Inc. U.S. EPA Contract 68-C-99-206.
Economic Analysis for the LT2ESWTR 9-16 December 2005
-------
USEPA. 2001c. Drinking Water Baseline Handbook, Third Edition. Draft. Prepared by International
Consultants, Inc. Contract 68-C6-0039. May, 2001.
USEPA. 2001d. Arsenic Rule Benefits Analysis: An SAB Review. Science Advisory Board. EPA-SAB-
RSAC-01-005. May 2001.
USEPA. 2000a. Geometries and characterization of public water systems. Office of Ground and
Drinking Water. U.S. EPA Contract 815-R-00-024. December 2000.
USEPA. 2000b. Surface Water Analytical Tool (SWAT) Version 1.1 -program design and assumptions.
Prepared by Malcolm Pirnie, Inc.
USEPA. 2000c. Data Reliability Analysis of the EPA Safe Drinking Water Information System/Federal
Version (SDWIS/FED). Office of Water. U.S. EPA Contract 816-R-00-020. October, 2000.
USEPA. 2000d. Guidelines for Preparing Economic Analyses. U.S. EPA Office of the Administrator.
U.S. EPA Contract 240-R-00-003. September 2000.
USEPA. 2000e. Stage 2M-DBP Agreement in Principle. (65 FR 83015, December 29, 2000).
USEPA. 2000f. Handbook for Non-cancer Health Effects Valuation. Prepared by the Non-Cancer
Health Effects Valuation Subcommittee of the EPA Social Science Discussion Group, with
assistance from Industrial Economics, Incorporated. The literature review in this Handbook was
updated and extended to include cancers in: Industrial Economics, Incorporated. Unpublished.
Appendix B: Monetary values for health effects. Prepared for the U.S. Environmental Protection
Agency.
USEPA. 2000g. Final Heavy Duty Engine/Diesel Fuel Rule: Air Quality Estimation, Selected Health
and Welfare Effects Methods, and Benefits Results. December 2000.
USEPA. 2000h. A UX1 Database CD version 5.0. EPA Office of Water. Data extracted from ICRFED
database.
USEPA. 2000i. ICR Supplemental Survey Database. Prepared by DynCorp, Inc.
USEPA. 2000J. Regulatory Impact Analysis for the LT1ESWTR. U.S. EPA Contract 815-R-00-005.
February 2000.
USEPA. 2000k. Estimated Per Capital Water Ingestion in the United States. U.S. EPA Office of Water.
Washington, DC. EPA-822-R-00-008. April 2000.
Economic Analysis for the LT2ESWTR 9-17 December 2005
-------
USEPA. 1999a. Method 1622: Cryptosporidium in Water by Filtration/IMS/FA. Office of Water,
Washington, DC. U.S. EPA Contract 821-R-99-001. January 1999.
USEPA. 1999b. Method 1623: Cryptosporidium and Giardia in Water by Filtration/IMS/FA. Office of
Water, Washington, DC. U.S. EPA Contract 821-R-99-006. April 1999.
USEPA. 1999c. Uncovered finished water reservoirs guidance manual. Washington, DC: U.S. EPA
Office of Research and Development. U.S. EPA Contract 815-R-99-011. April 1999.
USEPA. 1998. Consumer Confidence Reports'.Final Rule. Fed. Reg. 63:44511-44536.
USEPA. 1998a. Demographic distribution of sensitive population groups. Final Report. Prepared by
SRA Technologies, Inc., Falls Church, VA. Work Assignment No. B-ll/22 (SRA 557-05/14).
February 24, 1998.
USEPA. 1998b. Regulatory impact analysis for the Interim Enhanced Surface Water Treatment Rule
(IESWTR). Prepared by The Cadmus Group and by SAIC. U.S. EPA Contract 68-C6-0059.
November 12, 1998.
USEPA. 1998c. Guidance on implementing the capacity development provisions of the Safe Drinking
Water Act Amendments of 1996. Office of Water. U.S. EPA Contract 816-R-98-006. July 1998.
USEPA. 1997a. Policy for use of probabilistic analysis in risk assessment. Office of Research and
Development. May 15, 1997.
USEPA. 1997b. The Benefits and Costs of the Clean Air Act, 1970-1990. Prepared for U.S. Congress.
USEPA. 1997c. 1995 Community Water Systems Survey. U.S. EPA Contract 815-R-97-001b.
USEPA. 1997d. Exposure Factors Handbook (Final Report). Washington, D.C.: Office of Research and
Development. "Chapter 15 - Activity Factors and Table 15A-2".
USEPA. 1996a. ICR Microbial Laboratory Manual. U.S. EPA Office of Research and Development.
Washington, DC. EPA/600/R-95/178. April 1996.
USEPA. 1996b. Economic analysis of Federal regulations under Executive Order 12866. Office of
Management and Budget. January 11, 1996.
USEPA. 1995a. Policy for risk characterization at the U.S. Environmental Protection Agency.
Memorandum. March 21, 1995.
USEPA. 1995b. Guidance for Risk Characterization. Science Policy Council. February 1995.
Economic Analysis for the LT2ESWTR 9-18 December 2005
-------
USEPA. 1991. Guidance Manual for Compliance with the Filtration and Disinfection Requirements for
Public Water Systems Using Surface Water Sources. AWWA. Prepared for USEPA.
USEPA. 1984. Estimation of Small System Water Treatment Costs (Water Model). USEPA Contract
No. 68-03-3093. Drinking Water Research Division, Cincinnati, OH.
US FDA. 2001. 21 CFR Part 120. Hazard Analysis and Critical Control Point (HAACP); Procedures
for the Safe and Sanitary Processing and Importing of Juice; Final Rule, 66(13), 6137-6202,
January 19,2001.
US FDA. 2004. 21 CFR Parts 1 and 11. Establishment and Maintenance of Records Under the Public
Health and Security and Bioterrorism Preparedness and Response Act of 2002; Establishment and
Maintenance of Records for Foods; Notice of Public Meeting; Availability of Draft Guidance for
Records Access Authority; Final Rules and Notice. 69(236), 71561-71655, December 9, 2004.
US FDA. 1998. 21 CFR Parts 101 and 120. Preliminary Regulatory Impact Analysis and Initial
Regulatory Flexibility Analysis of the Proposed Rules to Ensure the Safety of Juice and Juice
Products; Proposed Rule, 63(84), 24253-24302, May 1, 1998.
U.S. House of Representatives. 1996. House Report No. 104-632 (Commerce Committee). June 24,
1996 in U.S. Code Congressional and Administrative News (USCCAAN), 1996, 4, pp. 1373,
1401 and 1409, discussing §§132(b) and 1418(a) of the House bill.
Van Houtven, G.L., J.C. Whitehead, T.H. Bingham and B. Depro. 1997. Valuing Drinking Water
Benefits: Theory, Methods, and Research Needs. Draft Report.
Waitzman, N.J., R.M. Scheffler, and P.S. Romano. 1996. The Cost of Birth Defects: Estimates of the
Value of Prevention. Lanham: University Press of America, Incorporated.
Waters, II, W.G. 1992. Values of travel time savings and the link with income. Paper presented at the
Annual Meeting of the Canadian Transportation Research Forum. Banff, Alberta.
Weinstein, M.C., Siegel, J.E., Garber, A.M., Lipscomb, J., Luce, B.R., Manning, Jr., W.G., and Torrance,
G.W. 1997. Productivity Costs, Time Costs and Health-related Quality of Life: A Response to
the Erasmus Group. Health Economics, 6:505-510.
Whynes, O.K. and Neilson, A.R. 1993. Convergent Validity of Two Measures of the Quality of Life.
Health Economics, 2: 229-235.
The White House. 1993. Executive Order 12866. Regulatory Planning and Review. Signed September
30, 1993. Fed. Reg. 58:190:51735-51744. October 4, 1993.
Economic Analysis for the LT2ESWTR 9-19 December 2005
-------
Willocks, L., A. Crampin, L. Milne, C. Seng, M. Susman, R. Gair, M. Moulsdale, S. Shafi, R. Wall, R.
Wiggins, N. Lightfoot. 1998. A large outbreak of cryptosporidiosis associated with a public
water supply from a deep chalk borehole. Communicable Disease and Public Health 1(4): 23 9-
243.
Economic Analysis for the LT2ESWTR 9-20 December 2005
------- |