&EPA
United States
Environmental Protection
Agency
Economic Analysis for the
Proposed Revised Total
Coliform Rule
-------
Office of Water (4607-M) EPA 815-R-10-001 June 2010 www.epa.gov/safewater
-------
Table of Contents
Table of Contents i
Appendices v
Exhibits vi
List of Acronyms xii
Health Risk Reduction and Cost Analysis xv
Executive Summary ES-1
ES. 1 Need for the Rule ES-1
ES.2 Consideration of Regulatory Alternatives ES-4
ES.3 Summary of the Proposed RTCR Requirements ES-5
ES.4 Systems Subject to the Proposed RTCR ES-9
ES.5 National Benefits and Costs of the Proposed RTCR ES-9
ES.5.1 Derivation of Benefits ES-10
ES.5.2 Derivation of Costs ES-19
ES.6 Projected Impacts on Household Costs ES-20
ES.7 Comparison of Benefits and Costs, and Regulatory Alternatives of the Proposed
RTCR ES-22
ES.8 Conclusion ES-26
1. Introduction 1-1
1.1 Summary of the Proposed Revised Total Coliform Rule (RTCR) 1-1
1.2 Document Organization 1-4
1.3 Calculations and Citations 1-5
2. Statement of Need for the Rule 2-1
2.1 Introduction 2-1
2.1.1 Description of the Issue 2-1
2.2 Public Health Concerns, Fecal Contamination, and Waterborne Pathogens 2-3
2.2.1 Rule Objectives and Public Health Concerns 2-3
2.2.2 Total Coliforms as Indicators of Treatment Effectiveness and Integrity of the
Distribution System 2-8
2.2.3 Sanitary Defects 2-9
2.2.4 Occurrence of Fecal Contamination and/or Waterborne Pathogens 2-9
2.3 Statutory Authority for Promulgating the Rule 2-10
2.4 Economic Rationale 2-11
3. Consideration of Regulatory Options 3-1
3.1 Introduction 3-1
3.2 Total Coliform Rule/Distribution System Advisory Committee 3-1
3.3 Regulatory Options Considered 3-2
3.3.1 Comparative Summary of Regulatory Options 3-5
3.4 Proposed Rule Requirements 3-12
4. Baseline Conditions 4-1
4.1 Introduction 4-1
Economic Analysis for the Proposed RTCR i June 2010
-------
4.1.1 Background and Purpose 4-1
4.1.2 Chapter Organization 4-1
4.2 Data Sources 4-2
4.2.1 Background on SDWIS/FED Data 4-2
4.2.2 Background on 2005 Six-Year Review 2 Data 4-6
4.2.3 Background on Other Data and Information Used 4-9
4.3 Baseline Profile 4-11
4.3.1 Pre-GWR Baseline 4-12
4.3.2 Post-GWR Baseline 4-15
4.3.3 Baseline Population Served 4-16
4.3.4 Baseline Water Quality 4-18
4.4 Sensitive Sub-populations 4-22
4.5 Summary of Baseline Assumptions 4-23
5.0ccurrence and Predictive Model 5-1
5.1 Introducti on 5-1
5.2 Modeling of Current Total Coliform and E. coli Occurrence for Systems Serving up to
4,100 People 5-3
5.2.1 Distributional Model and Notation 5-3
5.2.2 Data Reduction 5-4
5.2.3 Basic Subsets of Systems 5-4
5.2.4 Estimation Methodology 5-6
5.2.5 Results 5-9
5.3 Predictive Modeling of Occurrence for Systems Serving Up to 4,100 People 5-13
5.3.1 Summary of GWR factors and timing affecting the TCR and Proposed RTCR 5-14
5.3.2 Summary of Predictive Model 5-16
5.3.3 Predictive Model Results 5-24
5.4 Occurrence Analysis for Systems Serving More Than 4,100 People 5-48
5.4.1 Model 5-49
5.4.2 Model Results 5-50
5.4.3 Model Uncertainty 5-50
5.5 Summary of Key Drivers for Benefit and Cost Analyses Output from the Predictive
Model 5-54
6. Benefits Analysis 6-1
6.1 Introduction 6-1
6.2 Qualitative Benefits Analyses 6-2
6.2.1 Implementation Activities 6-2
6.2.2 Routine Monitoring 6-2
6.2.3 Repeat Monitoring 6-5
6.2.4 Additional Routine Monitoring 6-6
6.2.5 Annual Site Inspections 6-7
6.2.6 Assessments 6-8
6.2.7 Corrective Actions 6-9
6.2.8 Public Notification 6-10
6.2.9 Summary of Qualitative Benefits Analyses 6-17
6.3 Assessment of Predictive Analysis Results 6-17
Economic Analysis for the Proposed RTCR ii June 2010
-------
6.3.1 Assessment of Predictive Analysis Results for Small (Serving <4,100 People)
and Large (Serving >4,100 People) Systems 6-19
6.4 Uncertainty and Sensitivity Analyses 6-28
6.5 Other Potential Benefits 6-33
6.5.1 Increased System Knowledge 6-33
6.5.2 Accelerated Infrastructure Repair/Replacement 6-34
6.5.3 Reduction in Averting Behavior 6-34
6.5.4 Reduction of Co-Occurring and Other Contaminants 6-34
6.5.5 Reduction in Outbreak Risk and Response Costs 6-35
7. Cost Analysis 7-1
7.1 Introduction 7-1
7.2 General Cost Assumptions and Methodology 7-1
7.2.1 Labor Rates 7-2
7.2.2 TCR Monitoring Costs per Sample 7-3
7.2.3 Technology Unit Costs and Compliance Forecasts 7-3
7.2.4 Cost Model 7-4
7.2.5 Modeled Variability and Uncertainty in National Costs 7-4
7.3 Projecting and Discounting National Costs 7-5
7.4 Derivation of Costs for PWSs and States 7-5
7.4.1 Rule Implementation and Annual Administration 7-6
7.4.2 Revising Sampling Plans 7-9
7.4.3 Monitoring 7-11
7.4.4 Annual Site Visits 7-17
7.4.5 Assessments 7-18
7.4.6 Corrective Actions 7-26
7.4.7 Public Notification 7-31
7.4.8 Uncertainty in Unit Costs 7-36
7.5 Household Costs 7-36
7.6 Nonquantified Costs 7-38
7.7 Uncertainty Analysis 7-39
7.8 Comparison of Total and Net Annualized Costs for All Regulatory Alternatives 7-42
8. Economic Impact Analysis 8-1
8.1 Introduction 8-1
8.2 Executive Order 12866: Regulatory Planning and Review 8-2
8.3 Paperwork Reduction Act 8-2
8.4 The Regulatory Flexibility Act 8-5
8.5 Unfunded Mandates Reform Act 8-9
8.6 Executive Order 13132: Federalism 8-10
8.7 Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments 8-11
8.8 Executive Order 13045: Protection of Children from Environmental Health Risks
and Safety Risks 8-12
8.9 Executive Order 13211: Action Concerning Regulations That Significantly
Affect Energy Supply, Distribution, or Use 8-12
8.10 National Technology Transfer and Advancement Act 8-13
Economic Analysis for the Proposed RTCR iii
June 2010
-------
8.11 Executive Order 12898: Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations 8-14
8.12 Consultations with the Science Advisory Board, National Drinking Water
Advisory Council, and the Secretary of Health and Human Services as Required by
Section 1412 (d) and (e) of the SDWA 8-14
8.13 Impacts on Sensitive Subpopulations as Required by Section 1412(b)(3)(c)(i)
of the 1996 Amendments to the Safe Drinking Water Act (SDWA) 8-15
8.14 Plain Language as Required by Executive Order 12866 8-16
9. Comparison of Benefits and Costs 9-1
9.1 National Benefits and Costs of the AIP Option Considered in Comparison to the
Current Total Coliform Rule and Alternative Option 9-1
9.1.1 National Benefits of the Regulatory Options Considered 9-3
9.1.2 National Cost Summary 9-19
9.1.3 Comparison of National Benefits and Costs of the Regulatory Options
Considered 9-27
9.2 Effect of Uncertainties and Nonquantified Benefit/Cost Estimates on the Estimation
of National Benefits and Costs 9-27
9.2.1 Summary of Major Uncertainties in EA Analyses 9-27
9.2.2 Summary of Nonquantified Costs and Benefits 9-29
9.3 Comparison of the Regulatory Options Considered 9-30
9.3.1 Incremental Costs and Benefits 9-30
9.3.2 Cost Effectiveness Measures 9-32
9.3.3 Break-Even Analysis 9-34
9.3.4 Summary of Conclusions 9-39
10. References 10-1
Economic Analysis for the Proposed RTCR iv
June 2010
-------
Appendices
Appendix A: Detailed Predictive Model Results
Appendix B: Graphs of Predicted Occurrence Over Time
Appendix C: Detailed Cost Model Results
Appendix D: Detailed Compliance Forecast and Unit Costs Estimates
Appendix E: Break-Even Analysis
Appendix F: Alternative Strategies for Combining Basic Subsets
Appendix G: Evaluation of Representativeness of Six-Year Review 2 Data
Appendix H: Analysis of Repeat Sample Records from the Six-Year Review 2 Data
Appendix I: Supporting Information for Regulatory Flexibility Act Screening Analysis
Economic Analysis for the Proposed RTCR
v
June 2010
-------
Exhibits
Executive Summary
Exhibit ES. 1 Implementation Schedule ES-8
Exhibit ES.2 GW Transient Noncommunity Water System Total Coliform
Occurrence ES-12
Exhibit ES.3 Potential Changes in Risk under the AIP and Alternative Options
Relative to the Current TCR ES-13
Exhibit ES.4 Predicted Average Annual Acute (current TCR) or E. coli MCL
(AIP and Alternative Options) Violations by Regulatory Option
and System Type ES-18
Exhibit ES.5 Comparison of Total and Net Change from Current TCR in Annualized
Present Value Costs ($Millions, 2007$) ES-20
Exhibit ES.6 Summary of Net Annual Per-Household Costs for the Proposed
RTCR (2007$) ES-22
Exhibit ES.7 Estimated Annual Breakeven Threshold for Avoided Cases of
STEC 0157 ES-24
Exhibit ES.8 Estimated Annual Breakeven Threshold for Avoided
Cases of Salmonella ES-24
Exhibit ES.9 Total Net Annual Cost Per Corrective Action Implemented
under AIP and Alternative Options, Annualized Using 3% and
7% Discount Rates ($2007) ES-25
Exhibit ES. 10 Incremental Rule Cost Per Corrective Action Implemented under
AIP and Alternative Options, Annualized Using 3 and 7 Percent
Discount Rates ($2007) ES-26
Exhibit ES. 11 Incremental Rule Cost Per Corrective Action for TNCWSs using
GW (Serving <100 People) Implemented under AIP and Alternative
options, Annualized Using 3 and 7 Percent Discount Rates ($2007) ES-26
1. Introduction
Exhibit 1.1 Proposed RTCR Monitoring Frequency Requirements (AIP Option) 1-3
2. Statement of Need for the Rule
3. Consideration of Regulatory Options
Exhibit 3.1 Comparison of Proposed RTCR Regulatory Options 3-6
4. Baseline Conditions
Exhibit 4.1 Pre-GWR Baseline Number of GW Systems 4-13
Exhibit 4.2 Baseline Number of SW Systems 4-13
Exhibit 4.3 Pre-GWR Baseline Population Served by GW Systems 4-14
Exhibit 4.4 Percent Distribution of Ground Water PWS Monitoring
Frequencies by PWS Size and Type for Current TCR 4-14
Exhibit 4.5 Post-GWR Baseline Number of GW Systems 4-16
Exhibit 4.6 Post-GWR Baseline Population Served by GW Systems 4-17
Exhibit 4.7 Baseline Population Served by SW Systems 4-17
Economic Analysis for the Proposed RTCR vi June 2010
-------
Exhibit 4.8 Household Baseline 4-18
Exhibit 4.9 Total Coliform and E. coli Percent Positive by System Size and Type 4-19
Exhibit 4.10 Baseline Number of TCR Violations by System Size and Type (2005) 4-21
Exhibit 4.11 Number of PWSs with Violations by System Type (2001 - 2007) 4-22
Exhibit 4.12 Estimates of Sensitive Subpopulations in the United States 4-23
Exhibit 4.13 Summary of Baseline Assumptions Influencing RTCR Estimates 4-24
5. Occurrence and Predictive Model
Exhibit 5.1 Basic Classifications of PWSs Used for Occurrence Modeling 5-5
Exhibit 5.2 Directed Graph of Model Used for RTTC Occurrence 5-7
Exhibit 5.3 Maximum Likelihood a and b Parameter Estimates for RTTC and RPTC ...5-10
Exhibit 5.4 Maximum Likelihood a and b Parameter Estimates for RTEC and RPEC ...5-11
Exhibit 5.5 a and P Parameter Estimates for RTTC and RPTC 5-12
Exhibit 5.6 a and P Parameter Estimates for RTEC and RPEC 5-13
Exhibit 5.7 Simulated Impacts of the AIP and Alternative Options on Systems
Serving <4,100 People - Surface Water Systems 5-20
Exhibit 5.8 Simulated Impacts of the AIP and Alternative Options on Systems
Serving <4,100 People - Ground Water Systems 5-21
Exhibit 5.9a Percent of GW Systems Assumed to be on Monthly, Quarterly, and
Annual (M/Q/A) Monitoring by System Category under Current
TCR (Baseline) - Initial Estimates (Post GWR Implementation) 5-22
Exhibit 5.9b Percent of GW Systems Predicted to be on M/Q/A Monitoring by
System Category under AIP Option - Adjusted Estimates
(Post RTCR Implementation) 5-23
Exhibit 5.9c Percent of GW Systems on M/Q/A Monitoring by System
Category under Alternative option - Adjusted Estimates
(Post RTCR Implementation) 5-24
Exhibit 5.10 Ground Water Community Water System Model Output Cumulative
Endpoints 5-28
Exhibit 5.11 Ground Water Nontransient Noncommunity Water System Model Output
Cumulative Endpoints 5-28
Exhibit 5.12 Ground Water Transient Noncommunity Water System Model Output
Cumulative Endpoints 5-29
Exhibit 5.13 Surface Water Community Water System Model Output Cumulative
Endpoints 5-29
Exhibit 5.14 Surface Water Nontransient Noncommunity Water System Model Output
Cumulative Endpoints 5-30
Exhibit 5.15 Surface Water Transient Noncommunity Water System Model Output
Cumulative Endpoints 5-30
Exhibit 5.16 Ground Water Community Water System TC Occurrence 5-31
Exhibit 5.17 Ground Water Nontransient Noncommunity Water System TC
Occurrence 5-32
Exhibit 5.18 Ground Water Transient Noncommunity Water System TC
Occurrence 5-33
Exhibit 5.19 Surface Water Community Water System TC Occurrence 5-34
Economic Analysis for the Proposed RTCR vii
June 2010
-------
Exhibit 5.20 Surface Water Nontransient Noncommunity Water System TC
Occurrence 5-35
Exhibit 5.21 Surface Water Transient Noncommunity Water System TC
Occurrence 5-36
Exhibit 5.22a Summary of Model Parameters Influencing RTCR Baseline and
Options 5-38
Exhibit 5.22b Summary of Model Parameters Influencing Proposed RTCR
Options Only 5-39
Exhibit 5.23 Sensitivity Analysis Assumptions for Frequency and Effectiveness of
Corrective Actions following Level 1 or 2 Assessments 5-42
Exhibit 5.24a Cumulative Effect of Alternative Assumptions for Corrective Action
Effectiveness and Duration on AIP Option Model Results
(Nondisinfecting TNCWS Serving <500 People over 25 Years) 5-42
Exhibit 5.24b Cumulative Effect of 50% Assumption for Corrective Action
Implementation Rate on AIP Option Model Results (Nondisinfecting
TNCWS Serving <500 People over 25 Years) 5-44
Exhibit 5.25 Effect of Seasonality on Occurrence Analysis Endpoints 5-45
Exhibit 5.25 Comparison of TC+ Occurrence Predicted as a 25-Year Annual
Average under the Current TCR with 2005 Six-Year Review Data 5-47
Exhibit 5.26 Comparison of SDWIS Data for Nonacute and Acute Violations with
Predictive Model Annual Results for the Current TCR 5-48
Exhibit 5.27 Results for Systems Serving more than 4,100 People - Current TCR 5-51
Exhibit 5.28 Results for Systems Serving more than 4,100 People - AIP option 5-52
Exhibit 5.29 Results for Systems Serving more than 4,100 People -
Alternative option 5-53
6. Benefits Analysis
Exhibit 6.1 Potential Changes in Risk under the AIP and Alternative Option
Relative to the Current TCR 6-13
Exhibit 6.2 Predicted Outcomes (25-Year Period of Analysis) under Current TCR 6-23
Exhibit 6.3 Predicted Outcomes (25-Year Period of Analysis) for AIP 6-24
Exhibit 6.4 Predicted Outcomes (25-Year Analysis Period) for Alternative Option 6-25
Exhibit 6.5 Predicted Average Annual Acute Violations by Reg. Option and
System Type 6-26
Exhibit 6.6 Predicted Change in Average Annual Acute Violationsl by
Regulatory Option and System Type 6-27
Exhibit 6.7 Relative Impacts Analysis for TNCWS Serving <101 People
for the Range of Corrective Action and Sampling Regimens Predicted 6-31
7. Cost Analysis
Exhibit 7.1 Labor Rates by PWS Size (2007$) 7-2
Exhibit 7.2 Monitoring Costs per Sample (2007$) 7-3
Exhibit 7.3 Net Change in PWS Unit Burden and Cost Estimates for Rule
Implementation (2007$) 7-7
Exhibit 7.4 Net Change in State Unit Burden and Cost Estimates for Rule
Implementation (2007$) 7-8
Economic Analysis for the Proposed RTCR viii
June 2010
-------
Exhibit 7.5 Annualized Cost Estimates for Rule Implementation
(SVlillions. 2007$) 7-9
Exhibit 7.6 Net Change in PWS and State Burden and Cost Estimates for
Revising Sampling Plans (2007$) 7-10
Exhibit 7.7 Annualized Cost Estimates to Revise (PWSs) and Review (States)
Sampling Plans ($Millions, 2007$) 7-11
Exhibit 7.8 Summary of Monitoring Requirements Under the TCR, AIP, and
Alternative Options 7-12
Exhibit 7.9 Cumulative Number of Samples over 25-Year Period of Analysis 7-16
Exhibit 7.10 Annualized PWS and State Cost Estimates for Monitoring Costs
(SVlillions. 2007$) 7-17
Exhibit 7.11 PWS Unit Costs Estimates for Assessment Activities under
Current TCR (2007$) 7-20
Exhibit 7.12 PWS Unit Costs Estimates for Level 1 and Level 2 Assessments
(AIP and Alternative Option) (2007$) 7-21
Exhibit 7.13 Number of Level 1 and Level 2 Assessments over the 25-Year
Compliance Period 7-22
Exhibit 7.14 State Unit Cost Estimates for Review of Level 1 and Level 2
Assessments under TCR, AIP, and Alternative Options (2007$) 7-25
Exhibit 7.15 Annualized PWS and State Cost Estimates for Level 1 and Level 2
Assessments ($Millions, 2007$) 7-26
Exhibit 7.16a Compliance Forecast for Corrective Actions based on Level 1 and
Level 2 Assessments 7-28
Exhibit 7.16b Detailed PWS Compliance Forecast for Corrective Actions based on
Level 1 and Level 2 Assessments 7-29
Exhibit 7.17 Net Change in PWS and State Unit Costs Estimates for Reporting and
Recordkeeping for Corrective Actions (2007$) 7-30
Exhibit 7.18 Annualized PWS and State Cost Estimates for Corrective Actions
based on Level 1 and Level 2 Assessments ($Millions, 2007$) 7-31
Exhibit 7.19 PWS Unit Cost Estimates for Public Notification (2007$) 7-33
Exhibit 7.20 State Unit Costs Estimates for Public Notification (TCR, AIP,
Alternative Options) 7-34
Exhibit 7.21 Number of Tier 1 and Tier 2 Public Notifications over the 25-Year
Compliance Period 7-35
Exhibit 7.22 Annualized PWS and State Cost Estimates for Public Notification
(SVlillions. 2007$) 7-36
Exhibit 7.23 Summary of Net Annual Per-Household Costs for the Proposed
RTCR (2007$) 7-38
Exhibit 7.24 Sensitivity Analysis - Annualized Net Change in Costs based on
Changes in Compliance Forecast ($Millions, 2007$) 7-41
Exhibit 7.25 Cost Uncertainty Summary 7-42
Exhibit 7.26 Comparison of Total and Net Change from Current TCR in
Annualized Present Value Costs ($Millions, 2007$) 7-43
Exhibit 7.27 Comparison of Total and Net Change in Annualized Present
Value Costs by Rule Component ($Millions, 2007$) 7-44
Exhibit 7.28 Total and Net Change in Annualized Costs to PWSs by
PWS Size and Type (SVlillions. 2007$) 7-45
Economic Analysis for the Proposed RTCR ix
June 2010
-------
Exhibit 7.29 Total and Net Change in Annualized Per PWS Costs by
PWS Size and Type (2007$) 7-46
8. Economic Impact Analysis
Exhibit 8.1 Average Annual Net Change Burden and Costs for the Proposed RTCR
Information Collection Request 8-4
Exhibit 8.2 Average Annualized Revenue by System Size and Percent of
Systems with Costs Exceeding One Percent and Three Percent of
Revenue (Three Percent Discount Rate) 8-6
Exhibit 8.3 Average Costs per System and as Percentage of Revenue
(2007$) 8-7
Exhibit 8.4 Annualized Net Rule Costs Predicted for Small Entities
(PWSs serving <10,000) by Rule Component Using Three
Percent and Seven Percent Discount Rates (2007$) 8-8
9. Comparison of Benefits and Costs
Exhibit 9.1a Ground Water (GW) Transient Noncommunity Water System
TC Occurrence 9-5
Exhibit 9. lb Potential Changes in Risk under the AIP and Alternative Options
Relative to the Current TCR 9-6
Exhibit 9.2 Estimates of Non-Acute Violations (TCR) and Level 1 Assessment
Triggers (AIP and Alternative Options) 9-13
Exhibit 9.3 Estimates of Acute Violations (TCR) and E. coli MCL Violations
(AIP and Alternative Options) 9-14
Exhibit 9.4 Estimates of Corrective Actions 9-15
Exhibit 9.5 Discounted Estimates of Non-Acute Violations (TCR) and Level 1
Assessment Triggers (AIP and Alternative Options) (3% Discount Rate)....9-16
Exhibit 9.6 Discounted Estimates of Acute Violations (TCR) and E. coli
MCL Violations (AIP and Alternative Options) (3% Discount Rate) 9-17
Exhibit 9.7 Discounted Estimates of Corrective Actions (3% Discount Rate) 9-18
Exhibit 9.8 Net Costs to PWSs (2007$) 9-21
Exhibit 9.9 Discounted Net Costs to PWSs Over Time (3% Discount Rate) (2007$) 9-22
Exhibit 9.10 Discounted Net Costs to PWSs (7% Discount Rate) (2007$) 9-23
Exhibit 9.11 Comparison of Total and Net Change in Annualized Present
Value Costs (SVlillions. 2007$) 9-24
Exhibit 9.12 Comparison of Total and Net Change in Annualized Present Value
Costs by Rule Component ($Millions, 2007$) 9-25
Exhibit 9.13 Total and Net Change in Annualized Costs to PWSs by PWS Size
and Type ($Millions, 2007$) 9-26
Exhibit 9.14 Incremental Net Change in Annualized Present Value Costs
($Millions, 2007$) and Benefits (Number of Corrective Actions) 9-31
Exhibit 9.15 Total Net Annual Cost Per Corrective Action Implemented under
AIP option and Alternative Options, Annualized Using 3% and 7%
Discount Rates ($2007) 9-32
Exhibit 9.16 Incremental Rule Cost Per Corrective Action (CA) Implemented for
All PWSs under AIP and Alternative Options, Annualized Using
3% and 7% Discount Rates ($2007) 9-33
Economic Analysis for the Proposed RTCR x
June 2010
-------
Exhibit 9.17 Incremental Rule Cost Per CA Implemented for GW TNCWSs
under AIP and Alternative Options, Annualized Using 3% and
7% Discount Rates ($2007) 9-33
Exhibit 9.18 Average Estimated Value Per STEC 0157 Case Avoided (2007$) 9-36
Exhibit 9.19 Average Estimated Value Per Salmonella Case Avoided (2007$) 9-37
Exhibit 9.20 Estimated Annual Break-Even Threshold for Avoided Cases
of STEC 0157 9-37
Exhibit 9.21 Estimated Annual Break-Even Threshold for Avoided Cases
of Salmonella 9-38
10. References
Economic Analysis for the Proposed RTCR xi
June 2010
-------
List of Acronyms
A/CA
Assessment and corrective action if necessary
AGI
Acute Gastrointestinal Illness
AIDS
Acquired Immune Deficiency Syndrome
AIP
Agreement in Principle
ASDWA
Association of State Drinking Water Administrators
BMPs
Best Management Practices
CAs
Corrective Actions
CCR
Consumer Confidence Report
CDC
Centers for Disease Control and Prevention
COI
Cost of Illness
CWS
Community Water System
DBP
Disinfection Byproduct
DBPR
Disinfection Byproducts Rule
DisGW
disinfected ground water
DisinfSW
disinfected surface water
DV
Data Verification
EA
Economic Analysis
EC
E. coli bacteria
EC+
E. 6'o//-positive
ECI
Employee Cost Index
EO
Executive Order
EPA
United States Environmental Protection Agency
ERS
Economic Research Service
FAC
Federal Advisory Committee
FACA
Federal Advisory Committee Act
FC
Fecal coliform bacteria
FC+
Fecal Coliform-positive
FDA
U.S. Food and Drug Administration
FR
Federal Register
GW
Ground Water
GWR
Ground Water Rule
GWSs
Ground Water Systems
GWUDI
Ground Water Under the Direct Influence of Surface Water
HAA5
Five Haloacetic Acids
HAV
Hepatitis A Virus
HEV
Hepatitis E Virus
HUS
Hemolytic Uremic Syndrome
ICR
Information Collection Rule
ICR
Information Collection Request
IDSEs
Initial Distribution System Evaluations
IESWTR
Interim Enhanced Surface Water Treatment Rule
KRPEC
Number of repeat TC positive samples assayed that tested positive
for EC
KRPTC
Number positive repeat samples during year
Economic Analysis for the Proposed RTCR xii
June 2010
-------
KRTEC
Number of routine TCpositive samples assayed that tested positive
for EC
KRTTC
Number positive routine samples during year
LRAAs
Locational Running Annual Averages
LT1ESWTR
Long Term 1 Enhanced Surface Water Treatment Rule
LT2ESWTR
Long Term 2 Enhanced Surface Water Treatment Rule
M/DBP
Microbial/Disinfection By-Product
M/Q/A
Monthly, Quarterly, and Annual
MAC
Mycobacterium avium Complex
MCL
Maximum Contaminant Level
MCLG
Maximum Contaminant Level Goal
MCMC
Bayesian Markov Chain Monte Carlo
MLEs
Maximum Likelihood Estimates
MRDLGs
Maximum Residual Disinfectant Level Goals
MRDLs
Maximum Residual Disinfectant Levels
NAICS
North American Industry Classification System
NCWS
Non-Community Water System
NDPWR
National Primary Drinking Water Regulation
NondisGW
Nondisinfected ground water
NPDWRs
National Primary Drinking Water Regulations
NRPEC
Number of repeat TCpositive samples that were assayed for EC
NRPTC
Total number repeat samples assayed
NRTEC
Number of routine TCpositive samples that were assayed for EC
NRTTC
Number routine samples assayed
NTNC
Nontransient Noncommunity
NTNCWS
Nontransient Noncommunity Water Systems
NTTAA
National Technology Transfer and Advancement Act
NTUs
Nephelometric Turbidity Units
O&M
Operations and Maintenance
OMB
Office of Management and Budget
PAM
Primary Amoebic Meningioencephalitis
PN
Public Notification
POU
Point of Use
pRPEC
Repeat probability that a TCpositive repeat sample will also test
ECpositive
pRPTC
Probability that a repeat sample will test positive
pRTEC
Probability that a TCpositive routine sample will also test
ECpositive
pRTTC
Probability that a routine sample will test positive
PWSs
Public Water Systems
RAA
Running Annual Average
RFA
Regulatory Flexibility Act
RPEC
Repeat coli
RPTC
Repeat TC
RTCR
Revised Total Coliform Rule
RTEC
Routine coli
RTTC
Routine TC
Economic Analysis for the Proposed RTCR xiii
June 2010
-------
SAB Science Advisory Board
SBA Small Business Administration
SBREFA Small Business Regulatory Enforcement Fairness Act
SDWA Safe Drinking Water Act
SDWIS Safe Drinking Water Information System
SDWIS/FED Safe Drinking Water Information System Federal Version
Six Year Review Six Year Review of National Primary Drinking Water Regulations
SS Sanitary survey
STEC Shiga toxin-producing E. coli
SW Surface Water
SWTR Surface Water Treatment Rule
TC Total Coliform bacteria
TC+ Total Coliform-positive
TCR Total Coliform Rule
TCRDSAC Total Coliform Rule/Distribution System Advisory Committee
TMF Technical, Managerial, and Financial
TNC Transient Noncommunity
TNCWS Transient Noncommunity Water Systems
TT Treatment Technique
TTHM Total Trihalomethane
TWG Technical Work Group
UMRA Unfunded Mandates Reform Act
UV Ultraviolet light
VSL Value of a Statistical Life
Economic Analysis for the Proposed RTCR xiv
June 2010
-------
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) 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 proposed Revised Total Coliform Rule (RTCR).
HRRCA Crosswalk Summary
HRRCA Requirement
Addressed in Economic Analysis
Quantifiable and nonquantifiable health risk
reduction benefits
Chapter 6 (all sections)
Chapter 8 (sections 8.8 and 8.12)
Chapter 9 (sections 9.1.2, 9.2.2, and 9.3)
Quantifiable and nonquantifiable health risk
reduction benefits from co-occurring
contaminants
Chapter 6 (section 6.5.1, 6.5.4)
Chapter 9 (section 9.2.2)
Quantifiable and nonquantifiable costs
Chapter 7 (all sections)
Chapter 8 (sections 8.3 - 8.6, 8.9, and 8.13 -
8.15)
Chapter 9 (sections 9.1.3, 9.2.2, and 9.3)
Incremental costs and benefits associated with
regulatory alternatives
Chapter 6 (sections 6.2 and 6.3)
Chapter 7 (sections 7.4, 7.5, and 7.8)
Chapter 9 (sections 9.1, 9.3.1, and 9.3.2)
Effects of the contaminants on the general
population and sensitive subpopulations
Chapter 2 (section 2.2)
Chapter 6 (all sections)
Chapter 8 (section 8.12)
Chapter 9 (sections 9.1.2, 9.2.2, and 9.3)
Increased health risk that may occur as a
result of compliance
Chapter 6 (section 6.2)
Chapter 9 (section 9.1.2)
Other relevant factors (quality and uncertainty
of information)
Chapter 4 (sections 4.2 and 4.3)
Chapter 5 (section 5.3.3.1)
Chapter 6 (section 6.4)
Chapter 7 (section 7.7)
Chapter 9 (section 9.2.1)
Economic Analysis for the Proposed RTCR xv
June 2010
-------
Executive Summary
This Economic Analysis (EA) presents the evaluation of the benefits and costs of the
proposed Revised Total Coliform Rule (RTCR). The analysis is performed in compliance with
Executive Order 12866, Regulatory Planning and Review (58 Federal Register (FR) 51735,
September 1993), which requires the United States Environmental Protection Agency (EPA or
Agency) to estimate the economic impact of rules that have an annual effect on the economy of
over $100 million. The Order further requires agencies to make the analysis for an
"economically significant" rule available to the public in conjunction with publication of the
final rule. Although EPA's analysis of the proposed RTCR has determined that its annual costs
are below this threshold, EPA has chosen to publish a complete EA for this rule.
In the Six-Year Review determination published in July 2003, EPA gave notice of its
intent to review the current Total Coliform Rule (current TCR). EPA has since developed the
proposed RTCR in collaboration with states, other interested stakeholders, and the Total
Coliform Rule/Distribution System Advisory Committee (TCRDSAC). The Agency's primary
reasons for revising the 1989 Total Coliform Rule are implementation-related issues. The
proposed revisions offer a meaningful opportunity for greater public health protection against
fecal contamination and waterborne pathogens in the distribution systems of public water
systems (PWSs) beyond the current Total Coliform Rule. These proposed revisions are in
accordance with the Safe Drinking Water Act (SDWA) as amended, which requires EPA to
review and revise, as appropriate, each national primary drinking water regulation promulgated
under the SDWA at least every six years. As with the current Total Coliform Rule, the proposed
Revised Total Coliform Rule applies to all PWSs.
ES.l Need for the Rule
EPA promulgated the 1989 TCR to decrease the risk of waterborne illness. Among all
SDWA rules promulgated for preventing waterborne illness, only the TCR applies to all PWSs,
making the rule an essential component of the multi-barrier approach in public health protection
against endemic (or epidemic) disease. In combination with the other SDWA rules, the proposed
RTCR will better address the TCR objectives and enhance the multi-barrier approach to
protecting public health, especially with respect to small ground water PWSs.
In recent years, the number of violations under the current TCR have remained relatively
steady, as shown in Exhibit 4.11 and discussed in Appendix G of the RTCR EA. EPA believes
that this is reflective of a steady state among PWSs complying with the current TCR;
improvements likely to occur under that rule have largely been achieved. In outlining
recommendations for further reductions in occurrence, EPA and the TCRDSAC developed the
Agreement in Principle (AIP), which has become the structure of the proposed RTCR.
The proposed RTCR maintains the objectives of the current TCR and is consistent with the
recommendations in the AIP. The objectives are: (1) to evaluate the effectiveness of treatment,
(2) to determine the integrity of the distribution system, and (3) to signal the possible presence of
fecal contamination. The proposed rule better addresses these objectives by requiring systems
that may be vulnerable to fecal contamination (as indicated by their monitoring results) to do an
assessment, to identify whether any sanitary defect(s) is (are) present, and to correct the defects.
Economic Analysis for the Proposed RTCR
ES-1
June 2010
-------
Therefore, greater public health protection is anticipated under the proposed RTCR compared to
the current TCR because of its more preventive approach to identifying and fixing problems that
affect or may affect public health. The key provisions of the proposed rule are summarized in
Section ES.3.
Public health concerns, fecal contamination, and waterborne pathogens
The proposed RTCR aims to increase public health protection through the reduction of
potential pathways of entry for fecal contamination into the distribution system. Since these
potential pathways represent vulnerabilities in the distribution system whereby fecal
contamination and/or waterborne pathogens, including bacteria, viruses, and parasitic protozoa
could possibly enter the system, the reduction of these pathways in general should lead to
reduced exposure and associated risk from the contaminants. Fecal contamination and
waterborne pathogens can cause a variety of illnesses, including acute gastrointestinal illness
(AGI) with diarrhea, abdominal discomfort, nausea, vomiting, and other symptoms. Most AGI
cases are of short duration and result in mild illness. Other more severe illnesses caused by
waterborne pathogens include hemolytic uremic syndrome (HUS) (kidney failure), hepatitis, and
bloody diarrhea (WHO 2004). Chronic disease such as irritable bowel syndrome, reduced kidney
function, hypertension and reactive arthritis can result from infection by a waterborne agent
(Clark et al. 2008).
When humans are exposed to and infected by an enteric pathogen, the pathogen becomes
capable of reproducing in the gastrointestinal tract. As a result, healthy humans shed pathogens
in their feces for a period ranging from days to weeks. This shedding of pathogens often occurs
in the absence of any signs of clinical illness. Regardless of whether a pathogen causes clinical
illness in the person who sheds it in his or her feces, the pathogen being shed may infect other
people directly by person-to-person spread, contact with contaminated surfaces, and other means
which are referred to as secondary spread. As a result, initially waterborne pathogens may
subsequently infect other people through a variety of routes (WHO 2004). Sensitive
subpopulations are at greater risk from waterborne disease than the general population (Gerba et
al. 1996a).
Indicators
Total coliforms are a group of closely related bacteria that, with few exceptions, are not
harmful to humans. Coliforms are abundant in the feces of warm-blooded animals, but can also
be found in aquatic environments, in soil, and on vegetation. Coliform bacteria may be
transported to surface water by run-off or to ground water by infiltration. Total coliforms are
common in ambient water and may be injured by environmental stresses such as lack of
nutrients, and water treatments such as chlorine disinfection, in a manner similar to most
bacterial pathogens and many viral enteric pathogens (including fecal pathogens). EPA considers
total coliforms to be a useful indicator that a potential pathway exists through which fecal
contamination can enter the distribution system. The absence (versus the presence) of total
coliforms in the distribution system indicates a reduced likelihood that fecal contamination
and/or waterborne pathogens are occurring in the distribution system.
Under the current TCR, each total coliform-positive sample is assayed for either fecal
coliforms oris. coli. Fecal coliform bacteria are a subgroup of total coliforms that traditionally
have been associated with fecal contamination. Since the promulgation of the TCR, more
Economic Analysis for the Proposed RTCR
ES-2
June 2010
-------
information and understanding of the suitability of fecal coliform and E. coli as indicators have
become available. Study has shown that the fecal coliform assay is imprecise and too often
captures bacteria that do not originate in the human or mammal gut (Edberg et al. 2000). On the
other hand, E. coli is a more restricted group of coliform bacteria that almost always originate in
the human or animal gut (Edberg et al. 2000). Thus, E. Coli is a better indicator of fecal
contamination than fecal coliforms.
Presence of fecal contamination
Fecal contamination is a very general term that includes all of the organisms found in feces,
both pathogenic and nonpathogenic. Fecal contamination can occur in drinking water both
through use of contaminated source water as well as direct intrusion of fecal contamination into
the drinking water distribution system. Biofilms in distribution systems may harbor waterborne
bacterial pathogens and accumulate enteric viruses and parasitic protozoa (Skraber et al. 2005;
Helmi et al. 2008). Waterborne pathogens in biofilms may have entered the distribution system
as fecal contamination from humans or animals.
Co-occurrence of indicators and waterborne pathogens is difficult to measure. The analytical
methods approved by EPA to assay for E. coli do not specifically identify most of the pathogenic
E. coli strains. There are at least 700 recognized E. coli strains (Kaper et al. 2004). About 10
percent of recognized E. coli strains are pathogenic to humans (Feng 1995; Hussein 2007; Kaper
et al. 2004). Pathogenic E. coli include E. coli 0157:H7, which is the primary cause of HUS in
the United States (Rangel et al. 2005). The US Centers for Disease Control and Prevention
(CDC) estimates that there are 73,000 cases of illness each year in the US due to E. coli
0157:H7 (Mead et al. 1999). The CDC estimates that about 15 percent of all reportedE. coli
0157:H7 cases are due to water contamination (Rangel et al. 2005). Active surveillance by CDC
shows that 6.3 percent of li. coli 0157:H7 cases progress to HUS (Griffin and Tauxe 1991; Gould et
al. 2009) and about 12 percent of HUS cases result in death within four years (Garg et al. 2003).
About 4 to 15 percent of cases are transmitted within households by secondary transmission
(Parry and Salmon 1998).
Because EPA-approved standard methods for E. coli do not typically identify the presence
of the pathogenic E. coli strains, an E. co/z'-positive monitoring result is an indicator of fecal
contamination but is not necessarily a measure of waterborne pathogen occurrence. Specialized
assays and methods are used to identify waterborne pathogens, including pathogenic E. coli.
One notable exception are the data reported by Cooley et al. (2007), which showed high
concentrations of pathogenic E. coli strains in samples containing high concentrations of fecal
indicator E. coli. These data are from streams and other poor quality surface waters surrounding
California spinach fields associated with the 2006 E. coli 0157:H7 foodborne outbreak. Data
equivalent to these samples are not available from drinking water samples collected under the
TCR.
Thus, based on the above discussion, EPA believes that E. coli is a meaningful indicator for
fecal contamination and the potential presence of associated pathogen occurrence.
Economic Analysis for the Proposed RTCR
ES-3
June 2010
-------
Waterborne disease outbreaks
The CDC defines a waterborne disease outbreak as occurring when at least two persons (or
one with amoebic meningoencephalitis) experience a similar illness after ingesting a specific
drinking water contaminated with pathogens (or chemicals) (Kramer et al. 1996). The CDC
maintains a database on waterborne disease outbreaks in the United States. The database is based
upon responses to a voluntary and confidential survey form that is completed by state and local
public health officials.
The National Research Council strongly suggests that the number of identified and reported
outbreaks in the CDC database for surface and ground waters represents a small percentage of
actual number of waterborne disease outbreaks (NRC 1997; Bennett et al. 1987; Hopkins et al.
1985 for Colorado data). Under-reporting occurs because most waterborne outbreaks in
community water systems are not recognized until a sizable proportion of the population is ill
(Perz et al. 1998; Craun 1996), perhaps 1 percent to 2 percent of the population (Craun 1996).
EPA drinking water regulations are designed to protect against endemic waterborne disease
and to minimize waterborne outbreaks. In contrast to epidemic, endemic refers to the persistent
low to moderate level or the unusual ongoing occurrence of illness in a given population or
geographic area (Craun et al. 2006).
ES.2 Consideration of Regulatory Alternatives
EPA evaluated the following three regulatory options as part of this revised rule proposal:
(1) the current TCR option, (2) the AIP option, and (3) an Alternative option. EPA discusses the
three regulatory options briefly in this executive summary and in greater detail in Chapter 3.
First, the current TCR option reflects EPA's understanding of how the current TCR
(USEPA 1989a, 54 FR 27544, June 29, 1989) is currently being implemented. That is, the
current TCR option is assumed to include "status quo" PWS and state implementation practices.
Next, the AIP option is a revised TCR based on the recommendations of the advisory committee.
The provisions of this proposed rule are based on the AIP option and are described in detail in
section III of the preamble of the proposed RTCR. Third, the Alternative option parallels the AIP
in most ways but includes variations of some of the provisions that were discussed by the
advisory committee before consensus was reached on the AIP. Under the Alternative option, at
the compliance date all PWSs are required to sample monthly for an initial period until they meet
the eligibility criteria for reduced monitoring. EPA assumes that eligibility for reduced
monitoring is determined during the next sanitary survey following the RTCR compliance date
(corresponding to year 11 of the model runs presented later on in this EA). This more stringent
approach differs from the AIP option that allows PWSs to continue to monitor at their current
frequencies (with an additional annual site visit or voluntary Level 2 assessment requirement for
PWSs wishing to remain on annual monitoring) until they are triggered into an increased
sampling frequency. Under the Alternative option, no PWSs are allowed to reduce monitoring to
an annual basis. EPA defined the Alternative option this way and included it in this EA to assess
the relative impacts of a more stringent rule and to better understand the balance between costs
and public health protection.
Economic Analysis for the Proposed RTCR
ES-4
June 2010
-------
To understand the relative impacts of the options, EPA gathered available data and
information to develop and provide input into an occurrence and predictive model. EPA
estimated both baseline conditions and changes to these conditions anticipated to occur over time
as a result of these revised rule options. The analysis is described in more detail in the remainder
of this EA.
ES.3 Summary of the Proposed RTCR Requirements
The proposed RTCR maintains and strengthens the objectives of the current TCR and is
consistent with the recommendations in the AIP. The objectives are: (1) to evaluate the
effectiveness of treatment, (2) to determine the integrity of the distribution system, and (3) to
signal the possible presence of fecal contamination. The proposed rule better addresses these
objectives by requiring systems that may be vulnerable to fecal contamination (as indicated by
their monitoring results) to do an assessment, to identify whether any sanitary defect(s) is (are)
present, and to correct the defects. Therefore, greater public health protection is anticipated under
the proposed RTCR compared to the current TCR because of its more preventive approach to
identifying and fixing problems that affect or may affect public health. The following is an
overview of the key provisions of the proposed RTCR:
MCLG and MCL for E. coli and coliform treatment technique for protection against potential
fecal contamination
The proposed RTCR establishes a maximum contaminant level goal (MCLG) and
maximum contaminant level (MCL) for E. coli. It takes a preventive approach to protecting
public health by establishing a coliform treatment technique for protection against potential fecal
contamination. The treatment technique uses both total coliforms and E. coli monitoring results
to start an evaluation process that, where necessary, will require the PWS to conduct follow-up
and corrective action that could prevent further incidences of contamination and exposure to
fecal contamination and/or waterborne pathogens. See section III. A.2 of the proposed RTCR
preamble (USEPA, 2010c) for a detailed discussion on the MCLG, MCL, and coliform treatment
technique requirements.
Monitoring
As with the current TCR, PWSs will continue to monitor for total coliforms and E. coli
according to a sample siting plan and schedule specific to the system.
Sample siting plans under the proposed RTCR must continue to be representative of the water
throughout the distribution system. Under the proposed RTCR, systems will have the flexibility
to propose repeat sample locations that will best verify and determine the extent of potential
contamination of the distribution system rather than having to sample within five connections
upstream and downstream of the total coliform-positive sample location. For some systems, most
likely those that have limited or no distribution systems, the repeat samples may satisfy the
requirement for source water samples. Because this represents a reduction in monitoring in
addition to costs, state approval is required. EPA believes that this reduction in monitoring is
appropriate only if the state determines that the dual purpose sample provides public health
protection equivalent to that provided by separate repeat and source water samples. EPA believes
that many ground water systems serving 1,000 or fewer, such as systems with extensive
Economic Analysis for the Proposed RTCR
ES-5
June 2010
-------
distribution systems, will not be able to show that this reduction in monitoring (i.e. a loss of
repeat sample that is near the total coliform-positive routine sample site, but not near the source
water sample site) will provide public health protection equivalent to separate samples.
As with the current TCR, the proposed RTCR allows reduced monitoring for some small
ground water systems. The proposed RTCR is expected to improve public health protection
compared to the current TCR by requiring small ground water systems that are on or wish to
move to reduced monitoring to meet certain eligibility criteria. Examples of the criteria include a
sanitary survey showing that the system is free of sanitary defects, a clean TCR compliance
history for 12 months, and a recurring annual site visit by the state and/or a voluntary Level 2
assessment by a party approved by the state or meeting criteria established by the state for
systems wishing to remain on or reduce to annual monitoring
For small ground water systems, the proposed RTCR requires increased monitoring for
higher-risk systems that meet certain criteria such as unacceptable compliance history under the
RTCR. The proposed RTCR specifies conditions under which systems will no longer be eligible
for reduced monitoring and be required to return to routine monitoring or to monitor at an
increased frequency.
The proposed RTCR requires systems on a quarterly or annual monitoring frequency
(applicable only to ground water systems serving 1,000 or fewer people) to conduct additional
routine monitoring the month following one or more total coliform-positive samples. Under the
proposed RTCR, systems must collect at least three routine samples during the next month,
unless the state waives the additional routine monitoring. This is a reduction in the required
number of additional routine samples from the current TCR, which requires at least five routine
samples in the month following a total coliform-positive sample for all systems serving 4,100 or
fewer people.
The current TCR requires all systems serving 1,000 or fewer people to collect at least
four repeat samples while PWSs serving 1,000 people or greater to collect three repeat samples.
The proposed rule requires three repeat samples after a routine total coliform-positive sample,
regardless of the system type and size.
Sections III.A.3 and III.A.4 of the proposed RTCR preamble provide detailed discussions
of the routine monitoring and repeat sampling requirements of the proposed RTCR.
Seasonal systems
The proposed RTCR establishes special monitoring requirements for seasonal systems for
the first time. Seasonal systems represent a special case in that the shutdown and start-up of these
water systems present additional opportunities for contamination to enter or spread through the
distribution system. Seasonal systems must demonstrate completion of a state-approved start-up
procedure. In addition, they must designate the time period(s) for monitoring based on site-
specific considerations (such as during periods of highest demand or highest vulnerability to
contamination) in their state-approved sample siting plan. Section III.A.3 of the proposed RTCR
preamble provides a detailed discussion of seasonal systems.
Economic Analysis for the Proposed RTCR
ES-6
June 2010
-------
Assessment and corrective action
As part of a treatment technique, all PWSs are required to assess their systems when
monitoring results show that the system may be vulnerable to contamination. Systems must
conduct a simple self-assessment (Level 1) or a more detailed assessment (Level 2) depending on
the severity and frequency of contamination. The system is responsible for correcting any
sanitary defect(s) found through either a Level 1 or Level 2 assessment. Section III. A. 5 of the
proposed RTCR preamble provides more discussion of the treatment technique requirement of
the proposed RTCR.
Violations and public notification
The proposed RTCR establishes an E. coli MCL violation, a treatment technique
violation, a monitoring violation, and a reporting violation. Public notification is required for
each type of violation, depending on the degree of potential public health concern consistent with
EPA's current public notification requirements under 40 CFR part 141, subpart Q. The proposed
RTCR also modifies the public notification and Consumer Confidence Report language to reflect
the construct of the proposed rule. Sections III.A.6 and III.A.7 of the proposed RTCR preamble
provide detailed discussions of violations and public notification under the proposed RTCR.
Transition to the RTCR
The proposed RTCR allows all systems to transition to the new rule at their current TCR
monitoring frequency, including systems on reduced monitoring under the current TCR. States
will then evaluate the monitoring frequency during each sanitary survey conducted after the
compliance effective date of the RTCR. This process reduces state burden by not requiring the
state to determine appropriate monitoring frequency at the same time as when the state is trying
to adopt primacy, develop policies, and train their own staff and the PWSs in the state.
Exhibit ES. 1 shows the implementation schedule, from year 1 through year 10, for the
proposed RTCR. PWSs will begin sampling under the RTCR in the fourth year after RTCR
promulgation1.
1 Chapter 4 of the RTCR EA describes the baseline schedule under which systems would begin sampling following
the RTCR effective date; Chapter 5 describes how the baseline schedule is adjusted, based on acute and non-acute
violations incurred during an initial assessment period, to determine the steady state distribution of systems on
monthly, quarterly, and annual sampling.
Economic Analysis for the Proposed RTCR ES-7 June 2010
-------
Exhibit ES.1 Implementation Schedule
Year
Year 1
Year 2
Year 3
Year 4
Year 5
Year 6
Year 7
Year 8
Year 9
Year 10
State
Implementation
Revising Sampling Plan
Routine monitoring, Additional routine monitoring, Repeat Monitoring
Site Inspections
Level 1 and Level 2 Site Assessments
Correction Actions Based on Level 1 and Level 2 Site Assessments
Public Notification
PWS
Implementation
Revising Sampling Plan
Routine monitoring, Additional routine monitoring, Repeat Monitoring
Site Inspections
Level 1 and Level 2 Site Assessments
Correction Actions Based on Level 1 and Level 2 Site Assessments
Public Notification
Source: Draft Information Collection Request for the Proposed Revised Total Coliform Rule, Figure 4.3.
Note: Activities occurring in Year 10 continue throughout the remaining years of analysis
Economic Analysis for the Proposed RTCR
ES-8
June 2010
-------
ES.4 Systems Subject to the Proposed RTCR
The proposed RTCR will apply to all PWSs in the United States. The baseline inventory
for the EA is derived from EPA's Safe Drinking Water Information System (SDWIS), fourth
quarter 2007 data2 (USEPA, 2007b), which contains information reported by primacy agencies
on the approximately 155,000 active PWSs. The systems are subdivided by water source (ground
water or surface water), type (community water systems (CWS), nontransient noncommunity
water systems (NTNCWS), and transient noncommunity water systems (TNCWS)), and size
(nine size categories ranging from those serving fewer than 100 people to those serving more
than 1 million people).
EPA developed the baseline water quality used in the RTCR EA from the Six-Year
Review 2 data collected between 1998 and 2005, which include coliform monitoring data
voluntarily provided to EPA by 37 primacy agencies (35 states and 2 tribes). The database
consists of over nine million TCR records collected between 1998 and 2005. In addition, EPA
incorporated information from the GWR, the U.S. Census, and the current TCR to develop
assumptions for the model.
EPA used the available data and the requirements of the regulatory options considered to
develop a model that predicts the number of systems in each category that experience TC+ and
EC+ assays over the 25-year period of analysis. The results of the occurrence model and cost
model indicate the number of systems that conduct the various activities under three regulatory
options (TCR and AIP and Alternative options). These results are discussed in terms of benefits
of the current rule (changes in acute violations incurred and corrective actions to be
implemented) and rule costs in Section ES.5 below.
ES.5 National Benefits and Costs of the Proposed RTCR
The consensus resulting from TCRDSAC deliberations was that a proposed RTCR that
was consistent with the AIP would achieve a net risk reduction compared to the current TCR.
The committee applied best professional judgment in determining that the increased protection
provided by the new requirements for implementing focused assessments and appropriate
corrective actions would more than offset any potential increase in risk introduced by the
reduction in samples or other changes resulting from the proposed RTCR.
Based on limitations in available data (described further in Chapter 6, Section 6.3), EPA
determined that benefits could not be calculated in terms of avoided cases of (or costs related to)
morbidity or mortality. Since E. coli is an indicator of fecal contamination, EPA assumed that a
decrease in E. coli occurrence in the distribution system would be associated with a decrease in
fecal contamination in the distribution system. In general, this decrease in fecal contamination
should reduce the potential risk to human health for PWS customers. Thus, any reduction in E.
coli occurrence is considered a benefit of the proposed RTCR. EPA's qualitative analysis in
Section 6.2 of this EA considers the overall changes in occurrence of contaminant indicators
under the proposed RTCR as compared to the current TCR, and predicts a net decrease in risk.
2
SDWIS-Federal Version (SDWIS/FED) is a database created by EPA containing data submitted by States and
regions regarding inventory as well as compliance with SDWA.
Economic Analysis for the Proposed RTCR
ES-9
June 2010
-------
As discussed in Chapter 6, this reduction in risk is an indicator of the benefits of the proposed
RTCR.
Sections ES.5.1 and ES.5.2 summarize the qualitative benefits analysis and the methods
used to derive the costs of the rule, respectively. Section ES.5.1 describes the qualitative analyses
of net changes in TC and EC occurrence anticipated to occur under the proposed RTCR
compared to the current TCR. As found in Section ES.5.2, EPA's national cost estimates include
cost to implement the rule; revise sampling plans; conduct routine monitoring, additional routine
monitoring, and repeat monitoring; perform Level 1 and Level 2 assessments; implement
corrective actions; and provide public notification in the case of violations. Estimates for present
value and annualized national costs are presented using a 3 percent and 7 percent discount rate.3
Chapters 6 (benefits), 7 (costs), and 9 (comparison of benefits and costs), as well as the
appendices, provide a more detailed discussion of all the analyses discussed in the sections
below.
Changes in risk associated with proposed RTCR activities are characterized by their
anticipated effects on potential pathways of contamination into PWS, as indicated by their effects
on such pathways for TC IE. coli. These activities are considered under each rule component
presented in Exhibit ES.3.
ES.5.1 Derivation of Benefits
In promulgating the RTCR, EPA expects to further reduce the risk of contamination of
public drinking water from the current baseline risk under the existing TCR. The options
considered during development of this proposed rule and analyzed as part of this EA (see
Chapters 6, 7, and 9) are designed to achieve this reduction while maintaining public health
protection in a cost-effective manner.
The EA examines the benefits in terms of trade-offs between compliance with the current
TCR and the other options considered (AIP and Alternative options). Because there are
insufficient data reporting the co-occurrence in a single sample of fecal indicator E. coli and
pathogenic organisms and because the available fecal indicator E. coli data from the Six-Year
Review 2 dataset were limited to presence-absence data, EPA was unable to quantify health
benefits for the proposed RTCR. EPA used several methods to qualitatively evaluate the benefits
of the proposed AIP and Alternative options. The qualitative evaluation uses both the judgment
of EPA as informed by the TCRDSAC deliberations as well as quantitative estimates of changes
in total coliform occurrence and counts of systems implementing corrective actions. The
evaluation characterizes, in relative terms, the reduction in risk for each regulatory scenario as
compared to baseline conditions.
3
There is much discussion among economists of the proper social discount rate to use for policy analysis.
For RTCR cost analyses, calculations are made using two social discount rates (3 and 7 percent) thought to best
represent current policy evaluation methodologies. Historically, the use of 3 percent is based on rates of return on
relatively risk-free investments, as described in the Guidelines for Preparing Economic Analyses (USEPA, 2000c).
The rate of 7 percent is a recommendation of the Office of Management and Budget (OMB) as an estimate of
"before-tax rate of return to incremental private investment" (USEPA, 1996b).
Economic Analysis for the Proposed RTCR ES-10 June 2010
-------
As presented in the description of the baseline (see Chapter 4), the percentages of
samples that are positive for total coliform and E. coli are generally higher for PWSs serving
4,100 or fewer people than those serving more than 4,100 people. PWSs with higher total
coliform and E. coli occurrence are more likely to be triggered into assessments and corrective
action. As discussed previously, EPA believes that the assessments and corrective action will
lead to a decrease in total coliform and E. coli occurrence. Because the PWSs serving 4,100
people or fewer have a higher initial E. coli occurrence and will be triggered into more
assessments and corrective actions than larger PWSs, the increase in benefits for these small
systems will likely be more evident as compared to the larger systems. In particular, model
results suggest that customers of small ground water TNCWSs serving 100 or fewer people,
which constitute approximately 40 percent of PWSs, experience the most improvement in water
quality under the proposed RTCR. That is, the occurrence of E. Coli is predicted to decrease
more for these systems that for other systems types.
When revising an existing drinking water regulation, one of the main concerns is to
ensure that backsliding on water quality and public health protection does not occur. SDWA
requires that EPA at least maintain or improve public health protection for any rule revision.
EPA believes that the proposed RTCR is more stringent than the current TCR with regard to
protecting public health. The basis for this perspective is provided in chapters 6 and 9 of this EA.
Risk reduction for the proposed RTCR is characterized by the activities performed that
are presumed to reduce risk of exposing the public to contaminated water. These activities are
considered under each rule component presented in Exhibit ES-3.
More frequent monitoring has the potential to decrease the risk of contamination in PWSs
based on an enhanced ability to diagnose and mitigate system issues in a more timely fashion.
Conversely, less frequent monitoring has the potential to increase risk. Real-time continuous
sampling would mitigate the most risk possible based on sampling schedule; however, it would
cost prohibitively more than the periodic sampling practiced under the current TCR and included
in the AIP and the Alternative options. EPA's objective in proposing the sampling schedules
included in the AIP and Alternative options was to find an appropriate balance between the
factors of risk mitigation and cost management.
Under the AIP and Alternative options, the reduction in the number of repeat samples and
additional routine samples for some PWSs has the potential to contribute to increased risk for
PWS customers. However, this increase in risk is expected to be more than offset by potential
decreases in risk from increased routine and the addition of the assessments and corrective action
provisions that find and fix problems indicated by monitoring. Exhibit ES-2 illustrates the
predicted reduced frequency at which total coliforms occur subsequent to the implementation of
the AIP and Alternative options. As discussed previously, the proposed RTCR uses total
coliform and E. coli occurrence as an indicator of potential pathways for possible contamination
to enter the distribution system. Exhibit ES-2 illustrates the combined effects on total coliform
occurrence resulting from changes in monitoring and the effects of assessments and corrective
actions for the different rule options illustrated. The relative trends indicated in Exhibit ES-2 for
transient noncommunity water systems also pertain to other PWS categories as illustrated in
Chapter 5. EPA chose to include the characterization for TNCWSs because they represent the
system category of largest influence on the national impacts.
Economic Analysis for the Proposed RTCR
ES-11
June 2010
-------
Exhibit ES.2 GW Transient Noncommunity Water System Total Coliform
Occurrence
0.045
0.04
0.035
£ 0.03
2
>
W 0.025
OJ
:1 0.02
O
Q-
H 0.015
nj
3
c
< 0.01
0.005
•TCR No GWR
¦TCR
AIP
Alt
10
15 20
Years
25
30
35
Note: 1) Six Year 2005 total coliform-positive occurrence is representative of all GW TNCWS. The rate presented
may underestimate the occurrence for systems serving 25-4,100 individuals. 2) Graph shows the 30-year modeled
period discussed in Ch. 5. Model years 3-27 represent the 25-year period of analysis for this EA. Model year 11
begins the steady state, during which systems that qualified for reduced monitoring are now sampling on their
reduced schedules. The criteria and timing of this monitoring adjustment is discussed in Section 5.3.2.2 of this EA.
3) The results represented by the curves for TCR and AIP and Alternative options all incorporate the effects of the
GWR.
The effect that the proposed changes to public notification requirements for monthly/non-
acute MCL violations have on risk is difficult to predict. Some factors, such as reduction in
available public information and possible PWS complacency, lead to a potential increase in risk
and other factors, such as less confusion (PN more in line with potential health risks) and PWSs
resources used more efficiently, lead to a potential decrease, as discussed in Chapter 9. This
change to PN is addressing a key concern expressed by various stakeholders in the advisory
committee and during the Six-Year Review 1 comment solicitation process. By eliminating the
requirement and replacing it with assessment and corrective action requirements, the Agency
expects less public confusion, more effective use of resources, and increased transparency. Other
proposed rule components are expected to have a negligible effect on risk. However, the overall
effect of the proposed RTCR is expected to be a further reduction in risk from the current
baseline risk under the current TCR. Chapter 6 presents a detailed discussion of the potential
influence on health risk for each proposed rule component.
Economic Analysis for the Proposed RTCR
ES-12
June 2010
-------
Exhibit ES.3 Potential Changes in Risk under the AIP and Alternative Options Relative to the Current TCR
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
AIP option
Alternative
option
AIP option
Alternative
option
AIP option
Alternative
option
Implementation
Activities
None
None
None
None
No change
No change
Routine
Monitoring
(Including
Reduced
Monitoring)
None
None
Increased stringency
in requirements to
qualify for reduced
monitoring along
with requirement to
return to baseline
monitoring upon loss
of these criteria is
expected to result in
decreased risk (i.e.,
fewer PWSs will
qualify and therefore
more will monitor
more frequently).
PWSs all monitor
monthly in the first few
years of implementation
of the RTCR, which is
an increase in sampling
frequency for systems
that monitor quarterly or
annually underthe
current TCR. After the
first few years, systems
may reduce to
quarterly, but none may
reduce to annual
monitoring, creating a
decrease in risk for
systems on annual
monitoring underthe
current TCR.
Decrease
Decrease
Repeat
Monitoring
Required repeat
samples reduced
from 4 to 3 for
systems serving
<1,000 people
Required
repeat samples
reduced from 4
to 3 for systems
serving <1,000
people
None
None
Increase
Increase
Additional
Additional routine
None
None
Increase
Increase
Economic Analysis for the Proposed RTCR
ES-13
June 2010
-------
Proposed Rule
Factors Leading to a Potential
Factors Leading to a Potential Decrease in
Overall Predicted Change in
Component
Increase in Risk
Risk
Risk
Routine
samples are no
Additional
Monitoring
longer required
for PWSs
monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional routine
samples from 5
to 3.
routine samples
are no longer
required for
PWSs
monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional
routine samples
from 5 to 3.
Based on
discussions
with
stakeholders,
states that
currently
None (only states
conduct annual
currently
site
None (only states
Annual
Inspections
performing
assessments
currently performing
annual
under the
annual inspections
None
No change
Increase
inspections are
expected to
continue)
current TCR
may no longer
have the
resources to
continue the
inspections and
conduct
quarterly
monitoring
under the
are expected to
continue)
Economic Analysis for the Proposed RTCR
ES-14
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
Alternative
option.
Assessments
None
None
Mandatory
assessments are a
new requirement.
Mandatory
assessments are a new
requirement.
Decrease
Decrease
Corrective
Actions
None
None
Mandatory
corrective actions
are a new
requirement.
Mandatory corrective
actions are a new
requirement.
Decrease
Decrease
Public
Notification -
Monthly/Non-
Acute MCL
Violations
Reduction in
available public
information
Possible PWS
complacency
Reduction in
available public
information
Possible PWS
complacency
Less confusion (PN
more in line with
potential health
risks)
PWS resources
used more efficiently
Less confusion (PN
more in line with
potential health risks)
PWS resources used
more efficiently
Unknown
Unknown
Public
Notification -
Monitoring and
Reporting
Violations
None
None
Increased stringency
of PNs motivates
PWSs to conduct
required sampling.
Increased stringency of
PNs motivates PWSs to
conduct required
sampling.
Decrease
Decrease
Overall
Decrease
Decrease
Note: Detailed discussion of the rationale for determinations of potential risk for each rule component is presented in Ch. 6 (Section 6.2) of this EA.
Implementation activities consist of administrative activities by PWSs and states to implement the rule.
Assessment of potential changes in risk for monitoring components is an overall assessment. Potential changes (or static state) of risk for particular system
sizes and types differ according to individual regulatory requirements and are discussed in Section 6.2. Chapter 3 provides a detailed description of the
regulatory components for all three regulatory scenarios, and the preamble to the RTCR provides additional discussion of the TCRDSAC process and the
rationale underlying the structure of the regulatory options considered.
Economic Analysis for the Proposed RTCR
ES-15
June 2010
-------
Although a qualitative analysis is the primary method employed for analyzing potential
changes in risk, some quantitative measures are considered in the overall assessment of the
proposed RTCR, as presented in Chapter 9 of the RTCR EA. In particular, EPA developed a
model to describe TC and EC occurrence in water systems for the current TCR (baseline for this
EA) and the proposed RTCR alternatives. The model generates estimates of reduced TC and EC
occurrence based on requirements of the AIP and Alternative options to perform assessments and
corrective actions not explicitly required in the current TCR. In addition, the model takes into
account in the baseline for the current TCR the reductions attributable to the implementation of
the GWR, which was effective as of December 2009. The predictive occurrence model and
analyses results are discussed further in Chapters 5, 6, and 9.
Exhibit ES.4 presents the number of acute violations (current TCR) or E. coli MCL (AIP
and Alternative options) violations predicted by the RTCR occurrence model to occur under each
regulatory option considered. These estimates represent the period of analysis beginning in year
six after promulgation, which begins the steady state following the initial five years during which
results are skewed by a strong influence from the "All Monthly" sampling under the Alternative
option. EPA did not quantify changes in violation or trigger rates for systems serving more than
4,100 people among the current TCR, proposed RTCR, and Alternative options because of: (1)
limited Six-Year Review data to characterize these systems, (2) the essentially unchanged
monitoring requirements across options for these systems, and (3) the level of effort already
occurring to implement the current TCR.
The estimates of acute violations have two major drivers: the total number of samples
taken over time (including routine, additional routine, and repeat) and the impact of corrective
actions taken. When looking at the comparisons between the current TCR option with the
proposed RTCR across all PWSs, the overall impact of the total numbers of samples taken is
negligible because the total number of samples predicted to be taken throughout the period of
analysis is almost the same (approximately 82 million samples) under both the current TCR
option and the AIP option. For the Alternative option, the analysis predicts that almost 88 million
total samples will be taken over the period of analysis.
The changes in the steady state estimates of annual acute violations from the current TCR
option to the AIP and Alternative options are shown in Exhibit 6.5. The steady state in the model
refers to the period beginning in years 7 (CWSs) and 9 (NCWSs) following promulgation, after
the proportions of systems sampling on monthly, quarterly, or annual regimens are adjusted
following a period of assessment4. Systems that qualify for reduced monitoring will begin their
new regimens in years 7 and 9 after promulgation, respectively, for CWSs and NCWSs. The
estimates shown in Exhibit ES.4 for small systems are from the predictive model and reflect the
average annual estimates for the 25-year period of analysis, which includes 3 years of
implementation followed by 22 years in which new rule requirements are in effect.
The steady state reductions in the number of annual acute violations found under the
proposed RTCR and Alternative option primarily reflect the benefits of corrective actions under
these two alternatives in preventing many of the acute violations that would otherwise occur over
this period. Under the proposed RTCR, reductions (or no increases) are predicted for all PWS
sizes and types, while under the Alternative option, reductions (or no increases) are predicted for
4 The effective date of the proposed RTCR occurs after 3 years of implementation, at the start of year 4 post
promulgation. For CWSs, years 4-6 post promulgation are the period of assessment for considering potential to
move to reduced monitoring for systems; for NCWSs, years 4 - 8 are the period of assessment.
Economic Analysis for the Proposed RTCR
ES-16
June 2010
-------
all PWSs except TNCWSs serving <1,000 people. For these small TNCWSs under the
Alternative option, increased monitoring is expected to lead to an overall increase in annual
acute violations (and is also the driver of the greater total number of annual acute violations
predicted). The interplay of the reducing effect on acute violations that required corrective
actions induce and the increasing effect from reductions in some monitoring requirements are
explored in Section 6.4 of this EA. The step-wise analysis presented in Section 6.4 shows how
acute violations could increase under the Alternative option and concludes that, on balance, more
acute events are prevented than are missed.
Economic Analysis for the Proposed RTCR
ES-17
June 2010
-------
Exhibit ES.4 Predicted Average Annual Acute (current TCR) or E. coli MCL (AIP and
Alternative Options) Violations by Regulatory Option and System Type1
PWS Size
(Population
Alternative
PWS Type
Served)
Number of Systems
TCR
AIP Option
Option
<100
1,170
7
5
5
101-500
2,150
8
5
5
501-1,000
1,173
5
3
3
1,001-4,100
2,938.0
7.4
6.6
6.6
cws sw
4,101-33,000
3,164
9
9
9
33,001-96,000
720
3
3
3
96,001-500,000
308
1
1
1
500,001-1 Million
31
0
0
0
> 1 Million
17
0
0
0
Totals
11,671
39
33
33
<100
11,938
52
39
42
101-500
13,892
47
32
34
501-1,000
4,467
13
9
9
1,001-4,100
6,443
22
16
16
CWS GW
4,101-33,000
3,156
12
12
12
33,001-96,000
335
2
2
2
96,001-500,000
63
0
0
0
500,001-1 Million
4
0
0
0
> 1 Million
3
0
0
0
Totals
40,301
148
109
116
<100
250
2
1
1
101-500
253
1
1
1
501-1,000
88
0
0
0
1,001-4,100
72
1
1
1
NTNC SW
4,101-33,000
22
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
1
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
0
0
0
0
Totals
688
4
3
3
<100
8,826
39
25
33
101-500
6,613
20.31
14.31
20.29
501-1,000
1,718
4.31
3.61
4.51
1,001-4,100
812
8
5
5
NTNC GW
4,101-33,000
70
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
0
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
0
0
0
0
Totals
18,041
72
49
63
<100
1,339
20
11
11
101-500
497
8
4
4
501-1,000
88.00
1.25
0.79
0.79
1,001-4,100
67.0
2.3
1.3
1.3
TNC SW
4,101-33,000
18
0
0
0
33,001-96,000
0
0
0
0
96,001-500,000
0
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
1
0
0
0
Totals
2,010
31
18
18
<100
60,200
302
249
354
101-500
19,275
94.93
73.11
110.96
501-1,000
1,963
10.07
8.04
10.47
1,001-4,100
617
8
5
5
TNC GW
4,101-33,000
67
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
1
0
0
0
500,001-1 Million
1
0
0
0
> 1 Million
0
0
0
0
Totals
82,126
416
336
481
All Systems
Grand Total
154,837
710
547
713
Source: Output from Proposed RTCR models (described in Sections 5.3
and 5.4 of this EA).
1 Average violation estimates are for the 22-year period following the
effective date of the RTCR and 3 years of implementation activities.
Economic Analysis for the Proposed RTCR
ES-18
June 2010
-------
ES.5.2 Derivation of Costs
Costs are estimated for different PWS types and size categories (nine size categories are
used based on population served) using unit costs developed in the Technology and Cost
Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d). Cost analyses for
PWSs include estimates to implement the rule; to revise sampling plans; to conduct routine
monitoring, additional routine monitoring, and repeat monitoring; to perform Level 1 and Level
2 assessments and implement corrective actions; and to provide public notification in the case of
violations. State cost analyses include estimates of the labor burdens that states would incur,
including staff training on proposed RTCR requirements, conducting annual administration,
reviewing monitoring reports, reviewing and approving corrective action plans, and for
recordkeeping. Chapter 7 of this EA provides detailed discussion on the underlying cost-buildup
for each rule component included in the cost model.
National costs are estimated using a cost model specifically developed for the proposed
RTCR. The model builds on the occurrence model described in Chapter 5. Within the modeling
structure, costs for PWSs serving 4,100 or more retail customers are analyzed differently from
smaller PWSs to capture differing baseline structures and to account for differences in available
occurrence data as described in Chapter 4 of this EA. The resulting national cost estimates for
the RTCR regulatory options considered are shown in Exhibit ES.5. To evaluate the impact of
costs under the proposed RTCR, emphasis in this EA is placed on the net changes in costs of the
AIP or Alternative options compared to the current TCR, also shown in Exhibit ES.5.
As noted throughout the RTCR EA, there is variability among many of the input
parameters to the proposed cost model and several rule compliance assumptions based on PWS
size and type (e.g., population served, labor rates, TC hit rates, and occurrence distributions are
different for different sizes and types of PWSs). However, there is insufficient information to
fully characterize the distribution of variability (i.e., calculating confidence bounds) within each
of these PWS classifications on a national scale; therefore, EPA uses mean values for these input
parameters.
EPA also recognizes that there is uncertainty in the national cost estimates. Many of the
uncertainties have the same impact on both the TCR and AIP and Alternative options (e.g.,
baseline assumptions and effects of GWR implementation). Because the EA analyses focus on
net changes between the TCR and the AIP and Alternative options, these common sources of
uncertainty do not impact conclusions based on the net change analyses. For assumptions that are
major drivers of the analyses and differ between the TCR and AIP and Alternative options (e.g.,
corrective action compliance forecast), EPA has evaluated uncertainty and performed sensitivity
analyses to qualitatively and quantitatively characterize the potential impacts of alternative input
parameters. Chapters 4 and 5 present a comprehensive discussion of factors contributing
uncertainty to the occurrence analysis. Chapter 5 presents a sensitivity analysis pertaining to the
predictive occurrence model results, which also impact the cost calculations. Section 7.7 of the
EA discusses uncertainty and provides sensitivity analysis results as they specifically pertain to
the cost analyses.
Economic Analysis for the Proposed RTCR
ES-19
June 2010
-------
Exhibit ES.5 Comparison of Total and Net Change from Current TCR in
Annualized Present Value Costs ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR - Total
$ 185
$ 0.9
$ 186
$ 178
$ 0.9
$ 179
AIP-Total
$ 199
$ 1.1
$ 200
$ 191
$ 1.3
$ 192
AIP - Net Change
$ 14
$ 0.1
$ 14
$ 13
$ 0.4
$ 14
AIP - Percent Change
7%
16%
7%
7%
48%
8%
Alternative Option - Total
$ 212
$ 1.2
$ 213
$ 207
$ 1.5
$ 209
Alternative Option - Net Change
$ 27
$ 0.3
$ 27
$ 29
$ 0.6
$ 30
Alternative Option - Percent Change
15%
32%
15%
16%
67%
17%
Source: Proposed RTCR cost model.
Notes:
1) Detail may not add due to independent rounding.
2) Annualized costs are greater using the 7% than 3% discount rate for the Alternative option. This occurs because
the costs under the Alternative option are front-loaded in the 25-year time period due to the all-monthly sampling
requirement (which doesn't occur under the AIP option). Discounting a given stream using 3% and 7% will always
result in a higher present value using 3%; annualization of an identical value (any value) using 3% and 7% would
result in a higher value under 7%. Depending on how costs accrue over the period, and how long the period is, one
effect will be stronger than the other. Generally, the discounting effect is stronger, resulting in the pattern commonly
seen where the 3% annualized amount is greater than the 7% one. However, in cases like the Alternative option
where the costs accrue faster early in the time period, the annualization effect can more than compensate for the
discounting effect, resulting in a higher annualized value under 7% compared to 3%.
ES.6 Projected Impacts on Household Costs
The household cost analysis considers the impact that the costs incurred by CWSs have
on the households they serve. This analysis considers the potential increase in a household's
water bill if a CWS passed the entire cost increase resulting from the proposed rule on to their
customers. This analysis is a tool to gauge potential impacts and should not be construed as a
precise estimate of potential changes to household water bills. State costs and costs to TNCWSs
and NTNCWSs are not included in this analysis since their costs are not typically passed through
directly to households. Exhibit ES.6 presents the mean expected increases in annual household
costs for all CWSs, including those systems that do not have to take corrective action. Exhibit
ES.6 also presents the same information for CWSs that must take corrective action. Household
costs tend to decrease as system size increases, due mainly to the economies of scale for the
corrective actions.
The first section of Exhibit ES.6 presents net costs per household under the AIP and
Alternative options for all rule components spread across all CWSs. In this scenario, comparison
to the current TCR shows a cost savings for some households. For those households that are
expected to see a cost increase, the average annual water bill would be expected to increase by
less than ten cents on average.
While the average increase in annual household water bills to implement the proposed
RTCR is less than a dollar, customers served by a small CWS that have to take corrective actions
as a result of the proposed rule would incur slightly larger increases in their water bills. The
subsequent sections of the exhibit present net costs per household for three different subsets of
CWSs (e.g., CWSs that perform assessments but no corrective actions, CWSs that do perform
corrective actions, and CWSs that do not perform assessments or corrective actions). As shown
in the second section of Exhibit ES.6, approximately 67% of households are served by CWSs
that perform assessments but do not perform corrective actions over the 25-year period of
Economic Analysis for the Proposed RTCR
ES-20
June 2010
-------
analysis. The 9% of households belonging to CWSs that would perform corrective actions would
experience an increase in annual net household costs of less than $1 on average for CWSs
serving >4,100 people to approximately $23 for CWSs serving 100 people or fewer.
The final section of the exhibit presents the 24% of households belonging to CWSs that
would not perform assessments or corrective actions. Households of this category would
experience an increase in cost savings comparable to those performing corrective actions, and a
decrease in cost savings compared to those performing assessments but no corrective actions.
This decrease in costs savings is because no PN costs are associated with systems not performing
assessments. Overall, the main driver of additional household costs under the proposed RTCR is
whether or not additional corrective actions are performed.
Economic Analysis for the Proposed RTCR
ES-21
June 2010
-------
Exhibit ES.6 Summary of Net Annual Per-Household Costs for the Proposed
RTCR (2007$)
PWS Size
(Population
Served)
Number of
Households
(AIP)
Num ber of
Households
(Alternative
Option)
3% Discount Rate
7% Discount Rate
AIP Net
AIP Net Cost per
Houshold
Alternative
Option Net
Alternative
Option Net
Cost per
Household
AIP Net
AIP Net
Cost per
Houshold
Alternative
Option Net
Alternative
Option Net
Cost per
Household
A
B
C
D=C/A
E
F=BB
G
H=G/A
I
J=l/B
All Com m unity Water Systems (CWSs)
<100
307,243
307,243
$ 68,433
$ 0.223
$ 180,905
$ 0.589
$ 170,417
$ 0.555
$ 310,914
$ 1.012
101-500
1,589,510
1,589,510
$ 323,373
$ 0.203
$ 417,352
$ 0.263
$ 430,801
$ 0.271
$ 557,863
$ 0.351
501-1,000
1,624,853
1,624,853
$ 35,455
$ 0.022
$ 75,901
$ 0.047
$ 86,936
$ 0.054
$ 138,624
$ 0.085
1,001-4,100
7,816,592
7,816,592
$ 335,846
$ 0.043
$ 335,846
$ 0.043
$ 390,122
$ 0.050
$ 390,122
$ 0.050
4,101-33,000
27,997,647
27,997,647
$ 2,295,169
$ 0.082
$ 2,295,169
$ 0.082
$ 2,007,071
$ 0.072
$ 2,007,071
$ 0.072
33,001-96,000
21,933,438
21,933,438
$ 1,547,791
$ 0.071
$ 1,547,791
$ 0.071
$ 1,284,124
$ 0.059
$ 1,284,124
$ 0.059
96,001-500,000
26,770,609
26,770,609
$ 1,463,800
$ 0.055
$ 1,463,800
$ 0.055
$ 1,199,384
$ 0.045
$ 1,199,384
$ 0.045
500,001-1 Million
9,764,979
9,764,979
$ 153,975
$ 0.016
$ 153,975
$ 0.016
$ 122,635
$ 0.013
$ 122,635
$ 0.013
> 1 Million
16,309,853
16,309,853
$ (4,876)
$ (0.000)
$ (4,876)
$ (0.000)
$ (4,023)
$ (0.000)
$ (4,023)
$ (0.000)
Total
114,114,724
114,114,724
$ 6,218,966
$ 0.054
$ 6,465,863
$ 0.057
$ 5,687,467
$ 0.050
$ 6,006,714
$ 0.053
Com m unity Water Systems (CWSs) performing Level 1/Level 2 Assessments (and no Corrective Actions)
<100
125,340
124,920
$ (187,747)
$ (1.498)
$ (151,107)
$ (1.210)
$ (126,913)
$ (1.013)
$ (77,760)
$ (0.622)
101-500
460,577
464,568
$ (138,181)
$ (0.300)
$ (104,157)
$ (0.224)
$ (83,851)
$ (0.182)
$ (41,251)
$ (0.089)
501-1,000
394,643
401,009
$ (132,431)
$ (0.336)
$ (122,868)
$ (0.306)
$ (110,140)
$ (0.279)
$ (97,592)
$ (0.243)
1,001-4,100
2,341,578
2,341,578
$ (261,760)
$ (0.112)
$ (261,760)
$ (0.112)
$ (212,565)
$ (0.091)
$ (212,565)
$ (0.091)
4,101-33,000
24,827,588
24,827,588
$ (207,674)
$ (0.008)
$ (207,674)
$ (0.008)
$ (88,472)
$ (0.004)
$ (88,472)
$ (0.004)
33,001-96,000
19,232,570
19,232,570
$ (187,918)
$ (0.010)
$ (187,918)
$ (0.010)
$ (156,652)
$ (0.008)
$ (156,652)
$ (0.008)
96,001-500,000
23,912,325
23,912,325
$ (162,082)
$ (0.007)
$ (162,082)
$ (0.007)
$ (146,737)
$ (0.006)
$ (146,737)
$ (0.006)
500,001-1 Million
5,524,188
5,524,188
$ (45,790)
$ (0.008)
$ (45,790)
$ (0.008)
$ (42,392)
$ (0.008)
$ (42,392)
$ (0.008)
> 1 Million
-
-
$
$
$
$
$
$
$
$
Total
76,818,809
76,828,746
$ (1,323,584)
$ (0.017)
$ (1,243,356)
$ (0.016)
$ (967,721)
$ (0.013)
$ (863,420)
$ (0.011)
Community Water Systems (CWSs) performing Corrective Actions
<100
13,927
13,880
$ 325,679
$ 23.385
$ 347,280
$ 25.020
$ 296,928
$ 21.321
$ 318,549
$ 22.950
101-500
51,175
51,619
$ 469,264
$ 9.170
$ 462,084
$ 8.952
$ 412,574
$ 8.062
$ 410,784
$ 7.958
501-1,000
43,849
44,557
$ 164,754
$ 3.757
$ 167,578
$ 3.761
$ 143,922
$ 3.282
$ 146,752
$ 3.294
1,001-4,100
260,175
260,175
$ 556,364
$ 2.138
$ 556,364
$ 2.138
$ 484,508
$ 1.862
$ 484,508
$ 1.862
4,101-33,000
3,170,059
3,170,059
$ 2,526,599
$ 0.797
$ 2,526,599
$ 0.797
$ 2,132,792
$ 0.673
$ 2,132,792
$ 0.673
33,001-96,000
2,700,868
2,700,868
$ 1,736,797
$ 0.643
$ 1,736,797
$ 0.643
$ 1,446,819
$ 0.536
$ 1,446,819
$ 0.536
96,001-500,000
2,858,284
2,858,284
$ 1,624,577
$ 0.568
$ 1,624,577
$ 0.568
$ 1,345,503
$ 0.471
$ 1,345,503
$ 0.471
500,001-1 Million
613,799
613,799
$ 194,889
$ 0.318
$ 194,889
$ 0.318
$ 161,003
$ 0.262
$ 161,003
$ 0.262
> 1 Million
-
-
$
$
$
$
$
$
$
$
Total
9,712,136
9,713,240
$ 7,598,924
$ 0.782
$ 7,616,167
$ 0.784
$ 6,424,050
$ 0.661
$ 6,446,711
$ 0.664
Com m unity Water Systems (CWSs) not performing Level 1/Level 2 Assessments, or Corrective Actions
<100
167,976
168,442
(69,499)
$ (0.414)
(15,268)
$ (0.091)
402
$ 0.002
70,124
$ 0.416
101-500
1,077,758
1,073,324
(7,710)
$ (0.007)
59,425
$ 0.055
102,079
$ 0.095
188,331
$ 0.175
501-1,000
1,186,361
1,179,288
3,132
$ 0.003
31,191
$ 0.026
53,154
$ 0.045
89,464
$ 0.076
1,001-4,100
5,214,839
5,214,839
41,242
$ 0.008
41,242
$ 0.008
118,178
$ 0.023
118,178
$ 0.023
4,101-33,000
-
-
-
$
-
$
-
$
-
$
33,001-96,000
-
-
-
$
-
$
-
$
-
$
96,001-500,000
-
-
-
$
-
$
-
$
-
$
500,001-1 Million
3,626,992
3,626,992
-
$
-
$
-
$
-
$
> 1 Million
16,309,853
16,309,853
-
$
-
$
-
$
-
$
Total
27,583,779
27,572,738
$ (32,836)
$ (0.001)
$ 116,591
$ 0.004
$ 273,813
$ 0.010
$ 466,097
$ 0.017
Source:
(C), (E), (G), (I) Exhibit 7.28.
Source: Exhibit 7.23
ES.7 Comparison of Benefits and Costs, and Regulatory Alternatives of the
Proposed RTCR
As required by the SDWA, EPA has determined that the benefits of the proposed RTCR
justify the costs. In making this determination, EPA considered quantified and nonquantified
benefits and costs as well as the other components of the HRRCA outlined in section 1412
(b)(3)(C) of the SDWA.
Economic Analysis for the Proposed RTCR
ES-22
June 2010
-------
Additionally, EPA used several other techniques to compare benefits and costs including
a break-even analysis and a cost effectiveness analysis. The break-even analysis (see Chapter 9
of the proposed RTCR EA) was conducted using two example pathogens responsible for some
(unknown) proportion of waterborne illness in the United States. The analysis shows that a
relatively small number of cases (based on E. coli 0157:H7 or Salmonella) would need to be
avoided for the rule to break-even with the best estimates of net costs.
Under the AIP option, just two deaths would need to be avoided annually using a 3%
discount rate based on consideration of the bacterial pathogen STEC 0157. Alternatively,
approximately 3,000 or 8,000 non-fatal cases, using the enhanced or traditional benefits
valuations approaches5, respectively, would need to be avoided to break even with rule costs. As
expected based on its costs, the Alternative option would require that a higher number of cases
be avoided annually for that option to break even (approximately 60% to 100% more than the
AIP option under the traditional and enhanced approaches, respectively).
As Exhibit 9.21 shows, approximately 2 deaths would need to be avoided from a
Salmonella infection for the rule to break even. The estimated number of non-fatal Salmonella
cases that would need to be avoided to break even is approximately 10,000 or 65,000 cases under
the enhanced and traditional benefits valuations approaches, respectively.
As the discussion presented in Chapter 2 of this EA describes, disease and deaths are
attributable to drinking water contamination across the country. The CDC has found that 15% of
outbreak cases of STEC 0157:H7 are waterborne (Rangel et al., 2005); if that rate applies as well
to endemic cases, then approximately 11,000 cases would be due to waterborne exposure to
STEC 0157:H7 and could be mitigated. Based on that assumption, up to 11,000 cases annually
may be avoided by measures that mitigate pathways of contamination into PWSs for just this one
potential contaminant. That amount of reduction in terms of STEC 0157:H7 would be enough for
the rule to break even based on either the enhanced or traditional cost of illness approaches. If
more than one contaminant was reduced or prevented from occurring in PWSs, the rule would be
that much more likely to break even. Avoided illness and death from secondary transmission of
infection could also be significant. Additionally, fewer cases would need to be avoided if all the
benefits of the rule as described in chapter 6 were included in the analysis. If increased
assessments and corrective actions result in a level of knowledge of the system that enables
earlier mitigation of potential pathways of contamination, then additional illnesses or deaths may
be avoided beyond those suggested by the decrease in acute events that the model predicts. In the
case of the outbreak of Salmonella infections in Alamosa, Colorado, as described in Chapter 9 of
the EA, some of the key factors that contributed to the outbreak include significant levels of
sediment and the presence of breaches to the integrity of the storage tank.
Chapter 9 of the proposed RTCR EA has a complete discussion of the break-even
analysis and how costs per case were calculated.
5 Both traditional and enhanced COI approaches count the value of the direct medical costs and of time lost that
would been spent working for a wage, but differ in their assessment of the value of time lost that would be spent in
nonmarket work (e.g. housework, yardwork, and raising children) and leisure (e.g. recreation, family time, and
sleep). They also differ in their valuation of (other) disutility, which encompasses a range of factors of well being,
including both inconvenience and any pain and suffering. A complete discussion of the traditional and enhanced
COI approaches can be found in Appendix E.
Economic Analysis for the Proposed RTCR
ES-23
June 2010
-------
Exhibit ES.7 Estimated Annual Breakeven Threshold for Avoided Cases of STEC
0157
COI
Discount
AIP Option
Alternative Option
Methodology
Rate
Non-fatal
Fatal cases
Non-fatal
Fatal cases
cases only
o
3
<
cases only
o
3
<
Traditional
3%
8,000
1.6
16,000
3.1
COI
7%
8,000
1.5
17,000
3.4
Enhanced COI
3%
3,000
1.6
5,000
3.1
7%
3,000
1.5
6,000
3.4
Calculations for fatal cases include the non-fatal COI component for the underlying
illness prior to death.
Notes: The number of cases needed to reach break-even threshold calculated by dividing
the net change in costs for the RTCR (Exhibit 9.5) by the average estimated value of
avoided cases (Exhibit 9.18). Threshold estimates based on a weighted average of the
cost of "all cases" that includes both fatal and non-fatal cases is shown in Appendix E.
STEC 0157 is only an example of a pathogenic endpoint that could have been used for
this analysis. Use of additional pathogenic contaminants in addition to this single endpoint
would result in lower threshold values. Detail may not add due to independent rounding.
The break-even threshold is higher using a 7% discount rate than a 3% discount rate
under the Alternative option. This result is consistent with the annualized costs of the
Alternative option being higher using the 7% discount rate, which is caused by the
frontloading of costs in the period of analysis, as explained further in Chapter 7 of the EA.
Exhibit ES.8 Estimated Annual Breakeven Threshold for Avoided Cases of
Salmonella
COI
Discount
AIP Option
Alternative Option
Methodology
Rate
Non-fatal
Fatal cases
Non-fatal
Fatal cases
cases only
o
3
<
cases only
o
3
<
Traditional
3%
65,000
1.6
130,000
3.1
COI
7%
65,000
1.6
141,000
3.4
Enhanced COI
3%
10,000
1.6
20,000
3.1
7%
10,000
1.6
21,000
3.4
Note: Calculations for fatal cases include the non-fatal COI component for the underlying
illness prior to death.
Note: The number of cases needed to reach break-even threshold calculated by dividing
the net change in costs for the RTCR (Exhibit 9.5) by the average estimated value of
avoided cases (Exhibit 9.19). Threshold estimates based on a weighted average of the
cost of "all cases" that includes both fatal and non-fatal cases is shown in Appendix E.
Salmonella is only an example of a pathogenic endpoint that could have been used for
this analysis. Use of additional pathogenic contaminants in addition to this single endpoint
would result in lower threshold values. Detail may not add due to independent rounding.
The break-even threshold is higher using a 7% discount rate than a 3% discount rate
under the Alternative option. This result is consistent with the annualized costs of the
Alternative option being higher using the 7% discount rate, which is caused by the
frontloading of costs in the period of analysis, as explained further in Chapter 7 of the EA.
Economic Analysis for the Proposed RTCR
ES-24
June 2010
-------
Cost effectiveness is another way of examining the benefits and costs of the proposed
rule. ES.9 shows the cost of the rule per corrective action implemented. The cost effectiveness
analysis, as with the net benefits, is limited because EPA was able to only partially quantify and
monetize the benefits of the proposed RTCR. The AIP option achieves the lowest cost per
corrective action implemented among the options considered.
Exhibit ES.9 Total Net Annual Cost Per Corrective Action Implemented under AIP
and Alternative Options, Annualized Using 3% and 7% Discount Rates ($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Net Cost
$ 13.7
$ 13.7
AIP Net Corrective Actions (L1 & L2)
598
555
AIP CEA(net rule cost/CA)
$ 22,899
$ 24,610
Alternative Option Net Cost
$ 27.2
$ 29.7
Alternative Option Net Corrective Actions (L1 & L2)
785
765
Alternative Option CEA(net rule cost/CA)
$ 34,718
$ 38,812
Notes: Exhibit includes only the number of corrective actions predicted by the RTCR occurrence model to be
implemented in addition to those implemented under the current TCR. Includes both L1 & L2 corrective actions. Total
net costs for each option and total CAs (not incremental) are shown in Exhibit 9.15. Detailed benefits and cost
information is provided in Appendices A and C, respectively, of this EA.
EPA also considered the incremental cost effectiveness of the AIP option as compared to
the Alternative option to determine the additional benefit associated with that portion of cost for
the Alternative option that exceeds the cost of the proposed RTCR. Exhibit ES.9 shows that in
incremental terms for all PWSs, the AIP option has a far lower unit cost per unit benefit than the
Alternative option. EPA further considered the group of 60,200 TNCWSs using GW, which are
the largest subset of systems by size and type, and which are expected to bear the highest burden
under the AIP option. ES. 10 shows that in incremental terms, the cost effectiveness of the AIP
far exceeds that of the Alternative options using either a 3% or 7% discount rate. The two
incremental analyses (ES.9 and ES.10) together indicate that the AIP option is significantly more
cost effective than the Alternative option for the most burdened subset of systems and for the all
PWSs together. Additional information about this analysis and other methods used to compare
benefits and costs can be found in Chapter 9 of the proposed RTCR EA.
Economic Analysis for the Proposed RTCR
ES-25
June 2010
-------
Exhibit ES.10 Incremental Rule Cost Per Corrective Action Implemented under
AIP and Alternative Options, Annualized Using 3 and 7 Percent Discount Rates
($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Incremental Net Costs ($ millions)1
$ 13.7
$ 13.7
AIP Incremental Net Corrective Actions (L1 & L2)1
598
555
AIP Incremental Cost per CA ($)
$ 22,902
$ 24,613
Alternative Option Incremental Net Costs ($ millions)2
$ 13.5
$ 16.0
Alternative Option Incremental Net Corrective Actions (L1 & L2)2
187
210
Alternative Option Incremental Cost per CA ($)
$ 72,582
$ 76,299
Notes: Exhibit includes only the number of corrective actions predictec
mplemented in addition to those implemented under the current TCR.
net costs for each option and total CAs (not incremental) are shown in
by the RTCR occurrence model to be
ncludes both L1 & L2 corrective actions. Total
Exhibit 9.15. Detailed benefits and cost
information is provided in Appendices A and C, respectively, of this EA.
1 Represents the incremental increase of the AIP option over the current TCR.
2 Represents the incremental increase of the Alternative option over AIP option. Add incremental net values for
Alternative option to incremental net values for AIP option to calculate total net values of Alternative option over
current TCR.
Exhibit ES.11 Incremental Rule Cost Per Corrective Action for TNCWSs using
GW (Serving <100 People) Implemented under AIP and Alternative options,
Annualized Using 3 and 7 Percent Discount Rates ($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Incremental Net Costs ($ millions)1
$ 5.1
$ 5.1
AIP Incremental Corrective Actions (L1 & L2) (TNCWS <101 only)1
279
257
AIP Incremental Cost per CA($)
$ 18,219
$ 19,965
Alternative Option Incremental Net Costs ($ millions)2
CO
CO
$ 9.8
Alternative Option Incremental Corrective Actions (L1 & L2) (TNCWS <101 only)2
128
145
Alternative Option Incremental Cost per CA ($)
$ 64,731
$ 67,762
Notes: Detail may not add due to independent rounding. Incremental Net Costs are based on TNCWS serving <101
people. Detailed benefits and cost information is provided in Appendices A and C, respectively, of this EA.
1 Represents the incremental increase of the AIP option over the current TCR.
2 Represents the incremental increase of the Alternative option over AIP option. Add incremental net values for
Alternative option to incremental net values for AIP option to calculate total net values of Alternative option over
current TCR.
ES.8 Conclusion
EPA is proposing the RTCR as described by the AIP option to further reduce occurrence
of TC and EC (and any waterborne pathogens that may co-occur) beyond the reductions
achieved under the TCR. EPA's goal is to increase public health protection from potential fecal
contamination and/or waterborne pathogen exposure, while continuing to pursue the objectives
of the current TCR. Although EPA's analysis of the proposed RTCR has determined that its
annual costs are most likely below the threshold stated in Executive Order 12866 of $100
million, EPA has chosen to publish a complete EA for this rule.
EPA has developed the proposed RTCR based on the AIP because it will not allow for
backsliding from current levels of protection of public health and will likely reduce risk from
waterborne disease further while costing relatively little on a net basis. This conclusion is
Economic Analysis for the Proposed RTCR
ES-26
June 2010
-------
consistent with recommendations of the TCRDSAC. As required by the SDWA, EPA has
determined that the benefits of the proposed RTCR justify the costs. In making this
determination, EPA considered quantified and nonquantified benefits and costs as well as the
other components of the HRRCA outlined in section 1412 (b)(3)(C) of the SDWA.
As shown in the uncertainty analysis presented in Chapter 6 (section 6.4), reducing repeat
samples and additional samples both have the potential to increase risk for some PWSs (see also
sections III.A.3 and III.A.4 of the proposed RTCR preamble for discussions on the repeat sample
and additional routine sample provisions respectively). However, this increase in risk is expected
to be more than offset by potential decreases in risk from increased routine monitoring and the
addition of the assessments and corrective action provisions that find and fix problems indicated
by monitoring. That is, based on the analyses presented in this EA, EPA concludes that the effect
of the corrective actions outweighs that of reduced sampling; both the AIP and the Alternative
options are expected to result in a net decrease in risk as compared to the current TCR. Although
the Alternative option decreases risk to a greater degree than the AIP option (proposed RTCR),
this additional reduction is achieved at a higher cost, both in absolute terms and in terms of cost-
effectiveness. The estimated net cost of the AIP option is small (approximately $14 million
annually using either a 3% or 7% discount rate) compared to the Alternative option ($27 - $30
million using a 3% and 7% discount rate, respectively). This cost differential is especially
important considering the potential concentration of impacts on the smallest TNCWSs and the
potential front loading of costs under the Alternative option. In addition, the net benefits are
expected to be positive under the AIP option and no backsliding in overall risk is predicted.
While the number of corrective actions under the Alternative option would be greater than under
the AIP option, the achievement of these benefits is not as cost effective as under the AIP option.
The analyses performed as part of this EA support the collective judgment and consensus
of the TCRDSAC that the proposed RTCR requirements provide for effective and efficient
revisions to current TCR regulatory requirements.
Economic Analysis for the Proposed RTCR
ES-27
June 2010
-------
1. Introduction
This Economic Analysis (EA) presents the evaluation of the benefits and costs of the
proposed Revised Total Coliform Rule (RTCR). The analysis is performed in compliance with
Executive Order 12866, Regulatory Planning and Review (58 FR 51735), which requires the
United States Environmental Protection Agency (EPA) to estimate the economic impact of rules
that have an annual effect on the economy of over $100 million and make that analysis available
to the public in conjunction with publication of the final rule. Although EPA's analysis of the
proposed RTCR has determined that its annual costs are most likely below this threshold, EPA
has chosen to publish a complete EA for this rule.
EPA developed the proposed RTCR in collaboration with states, other interested
stakeholders, and the Total Coliform Rule/Distribution System Advisory Committee
(TCRDSAC). In the Six-Year Review determination published in July 2003, EPA gave notice of
its intent to review the Total Coliform Rule (TCR). The Agency's primary reasons for revising
the 1989 Total Coliform Rule are implementation-related issues. The proposed revisions offer a
meaningful opportunity for greater public health protection against waterborne pathogens in the
distribution systems of public water systems beyond the current Total Coliform Rule.
This chapter provides a summary of the proposed RTCR in Section 1.1. Section 1.2
outlines the organization of this EA, and Section 1.3 provides information regarding supporting
calculations and citations.
1.1 Summary of the Proposed Revised Total Coliform Rule (RTCR)
The proposed RTCR applies to all community and noncommunity Public Water Systems
(PWSs). The proposed RTCR takes a proactive approach to protect public health, maintaining a
Maximum Contaminant Level Goal (MCLG) and Maximum Contaminant Level (MCL) for E.
coli and using both E. coli and Total Coliform (TC) monitoring to establish a framework for
PWSs to assess for sanitary defects and to correct them as appropriate.
Like the requirements of the current TCR, the requirements of the proposed RTCR ensure
that PWSs address the following objectives:
• Evaluate the effectiveness of treatment
• Determine the integrity of the distribution system; and
• Signal the possible presence of fecal contamination.
Economic Analysis for the Proposed RTCR
1-1
June 2010
-------
To accomplish the above objectives, the proposed RTCR will use TC as an indicator to
start an evaluation process that, where necessary, will require the PWS to correct sanitary
defects, defined in the proposed RTCR as:
A defect that could provide a pathway of entry for microbial contamination into the
distribution system or that is indicative of a failure or imminent failure in a barrier that is
already in place.
Under the proposed RTCR, E. coli will remain a regulated contaminant with an MCLG of
zero and an MCL of zero. All fecal coliform provisions (including the MCLG and MCL) are
removed in the proposed RTCR. There will no longer be an MCL for TC: instead, TC will be
used as an indicator as part of a coliform treatment technique.
For all PWSs serving >4,100 people, monitoring requirements under the proposed RTCR
remain essentially unchanged from existing TCR requirements. However, PWSs serving <4,100
people may face changes in their required monitoring schemes depending on specific size, type,
and source water categorizations. Changes for these PWSs may be for routine, additional routine
or repeat monitoring as summarized in Exhibit 1.1 below.
Under the proposed RTCR, PWSs must complete Level 1 or Level 2 assessments to
identify the presence of "sanitary defects" and defects in distribution system coliform monitoring
practices if sampling results in one of the following triggers listed below.
Level 1 Assessment Triggers
• For PWSs taking 40 or more samples per month, the PWS exceeds 5.0% TC-positive
samples for the month; or
• For PWSs taking fewer than 40 samples per month, the PWS has two or more TC
positive samples in the same month; or
• Failure to take every required repeat sample after a single routine TC-positive sample.
Level 2 Assessment Triggers
• The PWS has an E. coli violation (See Section III.A.6 of the proposed RTCR
preamble for a description of what constitutes an E. coli MCL violation.); or
• The PWS has a second Level 1 treatment trigger within a rolling 12-month period,
unless the first Level 1 treatment trigger was based on exceeding the allowable
number of total coliform-positive samples, the State has determined a likely reason
for the total coliform-positive samples that caused the initial Level 1 treatment
trigger, and the State establishes that the system has fully corrected the problem;
• For PWSs with approved reduced annual monitoring, a Level 1 treatment technique
trigger in two consecutive years.
Economic Analysis for the Proposed RTCR
1-2
June 2010
-------
Exhibit 1.1 Proposed RTCR Monitoring Frequency Requirements (AIP Option)
PWS Type
Source
Population Served
Monitoring Requirements
Routine Monitoring1
Noncommunity
(NCWS)
Ground water
< 1,000 people
Quarterly monitoring for TC2Additional
requirements to qualify for reduced monitoring
apply3
Community
(CWS)
Ground water
< 1,000 people
Monthly monitoring for TC
Additional requirements to qualify for reduced
monitoring apply4
All
(CWS &
NCWS)
Surface water
All
Same requirements as under current TCR
(Monthly monitoring forTC)
Ground water
> 1,000 people
Same requirements as under current TCR
(Monthly monitoring forTC)
Additional Routine Monitoring
Ground water
< 1,000 people
Additional routine monitoring no longer
required if monitoring monthly
Number of additional routine samples required
reduced from five to three if monitoring
quarterly or annually
All
(CWS &
NCWS)
Ground water
1,101 -4,100
people
Additional routine monitoring no longer
required
Surface water
< 4,100 people
Additional routine monitoring no longer
required
Ground and
surface water
> 4,100 people
Same requirements as under current TCR
(Additional routine monitoring not required)
Repeat Monitoring
All
(CWS &
NCWS)
Ground and
surface water
< 1,000 people
Number of repeat samples required reduced
from four to three
Ground and
surface water
>1,000 people
Same requirements as under current TCR
(Three repeat samples required)
monitoring.
2 PWSs may be increased to monthly monitoring based on compliance with rule. Examples include system with Level
2 assessment violation, an E. coli MCL violation, an RTCR treatment technique violation, and /or two routine
monitoring violations in a rolling 12-month period.
3 PWSs may qualify for annual monitoring based on compliance record, the existence of enhancements to water
system barriers to contamination, and completion of an initial and recurring annual site visit conducted by the state or
an annual voluntary Level 2 assessment by a party approved by the state.
4 PWSs may qualify for quarterly monitoring based on compliance record and either the existence of enhancements
to water system barriers to contamination or the completion of an annual site visit conducted by the state or an
annual voluntary Level 2 assessment by a party approved by the state.
Economic Analysis for the Proposed RTCR
1-3
June 2010
-------
The PWS shall be responsible for correcting sanitary defects found through either Level 1
or Level 2 assessments as defined above by implementing appropriate corrective actions. Failure
to complete a required assessment or take necessary corrective action constitutes a Treatment
Technique violation under the proposed RTCR.
The proposed RTCR requires public notification for three events that are tiered as
follows:
Tier 1. E. coli MCL violation
Tier 2. Treatment technique (TT) violation
Tier 3. Routine monitoring violation
The proposed RTCR also provides the primacy agency with the discretion to reduce the
monitoring frequency for well operated ground water PWSs serving <1,000 people, if the PWS
can demonstrate that it meets the criteria for reduced monitoring stipulated in the rule. The
specific criteria to qualify for reduced monitoring are described in Chapter 3.
Ground Water Rule (GWR) implementation will occur prior to the transition from the
TCR to the proposed RTCR. Compliance with the GWR requirements, including sanitary
surveys and site visits, can be used to help determine the level of monitoring required for ground
water PWSs serving <1,000 people.
1.2 Document Organization
The remainder of this EA is organized into the following chapters:
• Chapter 2 summarizes the technical, regulatory, and public health issues addressed by
the proposed rule. It also explains the statutory authority for the proposed RTCR and
the economic rationale for the regulatory approach.
• Chapter 3 reviews various regulatory options that EPA considered during the
development of the rule and presents the rationale for selecting the proposed rule
requirements.
• Chapter 4 characterizes baseline conditions that exist (including PWS inventory,
treatment, and water quality data) before PWSs make changes to meet the proposed
RTCR requirements.
• Chapter 5 summarizes TC and EC occurrence analysis, providing a description of the
occurrence model, the sources used, and the limitations or constraints to the analysis
based on the nature of the data provided by those sources.
• Chapter 6 presents the relative risk evaluation performed to estimate the potential
benefits of the proposed RTCR relative to the baseline and other regulatory options
considered.
Economic Analysis for the Proposed RTCR
1-4
June 2010
-------
• Chapter 7 presents an estimate of the net costs of implementing the proposed RTCR
to industry, households, and states. The net costs of all regulatory options considered
are compared.
• Chapter 8 discusses distributional analyses performed to evaluate the effects of the
rule on different segments of the population, and considers various executive orders
and requirements, including the Regulatory Flexibility Act (RFA) and Unfunded
Mandates Reform Act (UMRA).
• Chapter 9 compares the benefits and costs of the proposed RTCR to evaluate the
potential net benefits and cost-effectiveness. The results are discussed and compared
to other regulatory options considered.
• Chapter 10 includes a detailed list of references.
1.3 Calculations and Citations
This EA involves detailed and complex analyses, and the following are provided to help
the reader:
• The detailed reference section provided in Chapter 10.
• Appendices containing supporting spreadsheets and analyses:
- Appendix A - Detailed Predictive Model Results,
- Appendix B - Graphs of Predicted Hit Rates Over Time,
- Appendix C - Detailed Cost Model Results,
- Appendix D - Detailed Compliance Forecast and Unit Costs Estimates,
- Appendix E - Supporting information for Value of Statistical Life and Cost of
Illness estimates used in break even analyses,
- Appendix F - Detailed analysis of potential uncertainty and variability in
occurrence estimates,
- Appendix G - Evaluation of representativeness of Six Year Review data, and
- Appendix H - Analysis of potential impacts of reduced repeat sampling.
- Appendix I - Supporting Information for Regulatory Flexibility Act Screening
Analysis
• Exhibits. Most tabular exhibits include a row that provides the formulas used to
compute the contents of each column.
• Sources for values used if they were not calculated within the exhibits.
• Supporting electronic file outputs (i.e., RTCR occurrence and cost model output).
• Flowcharts that illustrate methodologies of analyses as well as proposed rule
requirements.
Economic Analysis for the Proposed RTCR
1-5
June 2010
-------
2. Statement of Need for the Rule
2.1 Introduction
The United States Environmental Protection Agency (EPA) is proposing to revise the
1989 Total Coliform Rule (TCR) to better protect public health. EPA noticed its intent to revise
the TCR as part of its Six-Year Review determination published in 2003 (USEPA 2003b, 68 FR
42907, July 18, 2003). In July 2007 EPA convened a Total Coliform Rule/Distribution System
Federal Advisory Committee (TCRDSAC or advisory committee) charged with evaluating how
well the objectives of the TCR are met and recommending possible revisions to the rule. The
TCRDSAC completed its analysis and provided recommendations for revising the TCR in an
Agreement in Principle (AIP) (USEPA, 2009) that was signed by advisory committee members
in September 2008. EPA committed to proposing a rule that was consistent with the
recommendations in the AIP (see Chapter 3 of this Economic Analysis (EA) for more details on
the TCRDSAC).
The proposed RTCR maintains and strengthens the objectives of the current TCR and is
consistent with the recommendations in the AIP. The objectives are: (1) to evaluate the
effectiveness of treatment, (2) to determine the integrity of the distribution system, and (3) to
signal the possible presence of fecal contamination. The proposed rule better addresses these
objectives by requiring systems that may be vulnerable to fecal contamination (as indicated by
their monitoring results) to do an assessment, to identify whether any sanitary defect(s) is (are)
present, and to correct the defects. Therefore, greater public health protection is anticipated under
the proposed RTCR compared to the current TCR because of its more preventive approach to
identifying and fixing problems that affect or may affect public health (see Chapter 3 of this EA
for more details on the AIP and the provisions of the proposed RTCR).
This chapter summarizes the technical, regulatory, and public health issues addressed by
the proposed rule. It also explains the statutory authority for the proposed RTCR and the
economic rationale for choosing a proposed regulatory approach rather than non-regulatory
alternatives.
2.1.1 Description of the Issue
EPA is required to review each existing national primary drinking water regulation
(NPDWR) every six years. In 2003, EPA completed its review of the TCR and 68 NPDWRs for
chemicals that were established prior to 1997. The purpose of the review was to identify current
health risk assessments, changes in technology, and other factors that would provide a health or
technological basis to support a regulatory revision that will maintain or improve public health
protection. In the 2003 announcement of the completion of the Six Year Review, EPA provided
public notice of its intent to revise the TCR (USEPA 2003b, 68 FR 42907, July 18, 2003).
Implementation-related issues are the primary reason for EPA's decision to revise the TCR.
Since promulgation of the TCR, EPA has received comments from a number of stakeholders
suggesting modifications to reduce the burden of implementing the TCR. These comments
included, among others, suggestions to modify the TCR's monitoring requirements (e.g., with
regards to sampling locations and repeat samples) (USEPA 2002b, 67 FR 19030, April 17, 2002
and USEPA 2003b, 68 FR 42907, July 18, 2003).
Economic Analysis for the Proposed RTCR 2-1
June 2010
-------
The 1989 TCR was promulgated to decrease the risk of waterborne illness. Among all
rules promulgated for preventing waterborne illness, only the TCR applies to all PWSs, so the
rule is an essential component of the multiple barrier approach in public health protection against
endemic and epidemic disease. However, improvements to the current TCR construct and
requirements can better protect public health as shown in the following discussions.
In recent years, violation rates under the TCR have remained relatively steady (see
Chapter 4, Exhibit 4.11 of this Economic Analysis (EA)). EPA believes that this is reflective of a
steady state among PWSs complying with the current TCR, suggesting the possibility that
improvements likely to occur under the current TCR have largely been achieved. Potential
exposure to waterborne pathogens continues to be a health risk. Within the United States, disease
is caused by a relatively small variety of bacterial, viral, and parasitic protozoan pathogens in
drinking water. Nevertheless, the number of potential waterborne pathogens and the number of
potential waterborne illnesses are significant. Humans can be exposed via fecal-oral, inhalation
or dermal exposure, with fecal-oral exposure being the most significant and prevalent pathway.
The current TCR established a maximum contaminant level goal (MCLG) and a
maximum contaminant level (MCL) for total coliforms (including fecal coliforms (FC) and E.
coli (EC)). A TC MCL violation requires public notification. However, many of the organisms
detected by total coliform and fecal coliform methods are not of fecal origin and do not have any
direct public health implication. Information has also become available since promulgation of the
current TCR in 1989 that indicates that measurement of fecal coliforms sometimes detects
organisms that may not have any connection to fecal contamination (Edberg et al. 2000).
Attributing greater public health significance to the presence of total coliforms or fecal coliforms
could result in public confusion. On the other hand, EPA believes that E. coli is a meaningful
indicator of fecal contamination and of the potential presence of associated pathogens (see
Section 2.2.4.1 of this chapter).
The proposed RTCR aims to maintain and strengthen the objectives of the current TCR
more effectively and efficiently. E. coli remains a regulated contaminant with a defined
maximum contaminant level goal (MCLG) and a maximum contaminant level (MCL) while TC
(including fecal coliforms) no longer has an MCLG and an MCL. Instead, the proposed RTCR
uses TC as an indicator as part of a coliform treatment technique. The treatment technique uses
both total coliforms and E. coli monitoring results to start an evaluation process that, where
necessary, will require the PWS to conduct assessments and corrective actions. The proposed
rule also takes into account the capacity of small systems (especially for PWSs serving 1,000 or
fewer people) and primacy agencies to effectively implement the proposed rule requirements.
Modeling results show that the assessment and corrective action provisions of the RTCR
result in a decrease in E. coli occurrence (see Exhibit 6.7) which is anticipated to lead to a
decrease in the potential exposure of the public to fecal contamination and its associated
pathogens. In general, this decrease in fecal contamination should reduce the potential risk to
human health of PWS customers. Thus, any reduction in E. coli occurrence is considered a
benefit of the proposed RTCR. Also, since fecal contamination can contain waterborne
pathogens including bacteria, viruses, and parasitic protozoa, in general a reduction in fecal
contamination should also reduce the potential risk from these other contaminants.
Economic Analysis for the Proposed RTCR 2-2
June 2010
-------
2.2 Public Health Concerns, Fecal Contamination, and Waterborne Pathogens
2.2.1 Rule Objectives and Public Health Concerns
The proposed RTCR aims to increase public health protection through a reduction of
potential pathways of entry for fecal contamination into the distribution system. Since these
potential pathways represent vulnerabilities in the distribution system whereby fecal
contamination and/or waterborne pathogens, including bacteria, viruses, and parasitic protozoa
could possibly enter the system, the reduction of these pathways in general should lead to
reduced exposure and associated risk from these contaminants. Waterborne pathogens that can
be associated with fecal contamination may cause a variety of illnesses, including Acute
Gastrointestinal Illness (AGI) with diarrhea, abdominal discomfort, nausea, vomiting, and other
symptoms. Most AGI cases are of short duration and result in mild illness. Consuming water
contaminated with waterborne pathogens can also result in acute illnesses such as hepatitis,
hemolytic uremic syndrome (kidney failure), and bloody diarrhea; milder, acute illnesses such as
conjunctivitis; and severe chronic illnesses such as diabetes and dilated cardiomyopathy. Other
examples of potential chronic diseases resulting from infection by a waterborne agent include
irritable bowel syndrome, hypertension, and reactive arthritis; chronic illnesses are generally
costly to treat.1
Sensitive subpopulations are at greater risk from waterborne disease than the general
population. These sensitive subpopulations include children (especially the very young); the
elderly; the malnourished; pregnant women; chronically ill patients (e.g., those with diabetes or
cystic fibrosis); and a broad category of those with compromised immune systems, such as
Acquired Immunodeficiency Syndrome (AIDS) patients, those with autoimmune disorders (e.g.,
rheumatoid arthritis, lupus erythematosus, and multiple sclerosis), organ transplant recipients,
and those receiving chemotherapy (Rose, 1997). Sensitive subpopulations represent almost 20
percent of the population in the United States (Gerba etal. 1996). The severity and duration of
illness is often greater in sensitive subpopulations than in healthy individuals, and may
occasionally result in death.
When humans are exposed to and infected by an enteric pathogen, the pathogen becomes
capable of reproducing in the gastrointestinal tract. As a result, healthy humans shed pathogens
in their feces for a period ranging from days to weeks. This shedding of pathogens often occurs
in the absence of any signs of clinical illness. Regardless of whether a pathogen causes clinical
illness in the person who sheds it in his or her feces, the pathogen being shed may infect other
people directly by person-to-person spread, contact with contaminated surfaces, and other means
which are referred to as secondary spread. As a result, waterborne pathogens that are initially
waterborne may subsequently infect other people through a variety of routes (WHO 2004).
Waterborne pathogens include pathogens of both fecal and non-fecal origin. Non-fecal
pathogens are found in the soil and soil water interface and include Legionella, Naegleria fowleri
and probably Helicobacter pylori, or are components of the distribution system environment and
can include Mycobacterium avium complex (MAC). Examples of common fecal bacterial
pathogens include pathogenic E. coli, Salmonella, Shigella, and Campylobacter jejuni. Some
1 Lifetime costs associated with a new case of diabetes, for example, assuming an average illness duration of 30
years, are estimated at $255,833 using a three percent discount rate and $161,967 using a seven percent discount rate
(year 2007 dollars). For dilated cardiomyopathy, the lifetime (21 year average) cost is $68,870 (seven percent
discount rate, year 2007 dollars).
Economic Analysis for the Proposed RTCR 2-3
June 2010
-------
waterborne bacterial pathogens cause disease by rapid growth and dissemination (e.g.,
Salmonella) while others primarily cause disease via toxin production (e.g., Shigella, E. coli
0157, Campylobacter jejuni). Campylobacter jejuni, E. coli, and Salmonella have a host range
that includes both animals and humans; Shigella is associated only with humans (Geldreich,
1996). Unlike viruses, bacteria are able to reproduce outside of the host.
Because a positive TC assay is an indicator of ambient water entry into or growth within
the distribution system, it represents an indication of potential for both fecal and non-fecal
contamination. Because E. coli originates in the gut of warm-blooded animals, it is an indication
of fecal contamination. Together, the TC IE. coli indicator system can be used to identify the
potential for fecal contamination by monitoring. Since fecal contamination can contain
waterborne pathogens including bacteria, viruses, and parasitic protozoa, in general a reduction
in fecal contamination should also reduce the potential risk from these other contaminants.
The group of bacteria known as E. coli contains both pathogenic and non-pathogenic
isolates. The methods approved by EPA to assay for E. coli do not identify most of the
pathogenic E. coli strains. The most dangerous E. coli bacteria contain the gene for producing
shiga-toxins. E. coli 0157:H7 is the most widespread shiga-toxin producing E. coli but at least
81 serotypes have been identified (Prager et al. 2005). Release of toxins in the body can result in
kidney failure, shock and death in otherwise healthy individuals, especially small children.
Typically, kidney failure occurs in 2-7% of illnesses. Death or end-stage renal disease occurs in
about 12% of patients four years after diarrhea-associated kidney failure (Garg et al. 2003).
Twenty five percent of kidney failure survivors demonstrate long-term renal consequences (Garg
et al. 2003). For patients with moderate and severe gastroenteritis caused by E. coli, long-term
study shows that they have an increased risk of irritable bowel syndrome, hypertension and
reduced kidney function (Garg et al. 2005). The Centers for Disease Control and Prevention
(CDC) indicates that drinking water is responsible for a portion of the 73,000 illnesses each year
from I'], coli 0157:H7 (Mead et al. 1999).
Ground water outbreaks due to E. coli 0157:H7 are prominent because of the fatal
outcomes associated with those outbreaks. In Walkerton, Ontario, 6 individuals died and 27
developed kidney failure from ground water contaminated withE. coli 0157:H7 (and
Campylobacter jejuni) (Health Canada, 2000). The outbreak coincided with a reduction in
chlorination treatment coincident with a large fecal contamination event. In Washington County,
NY, two individuals (including an otherwise healthy two year old child) died from E. coli
0157:H7 contamination of a county fair water supply system. (The fair ground's water supply
system was not recognized as a PWS because it was used for less than 60 days each year.) Four
individuals died in Cabool, MO due to E. coli 0157:H7 (Swerdlow etal. 1992) in an outbreak
that was likely due to source water contamination. Another E. coli 0157:H7 outbreak occurred in
ground water in Minnesota but no kidney failures resulted.
Although manure is often considered to be the source of shiga-toxin producing E. coli,
they have also been isolated from municipal sewage (Holler etal. 1999). E. coli 0157:H7 was
found to survive on a pasture surface for almost 4 months. About 4-15% of cases are acquired
via secondary transmission (Parry and Salmon, 1998). In addition to the shiga-toxin producing E.
coli, there are a substantial number of other E. coli bacteria that are pathogenic, mostly through
production of other toxins. Little data are available on the hazard associated with waterborne
transmission for most of the pathogenic E. coli other than E. coli 0157:H7.
Economic Analysis for the Proposed RTCR 2-4
June 2010
-------
As the result of a seven year long study of individuals infected or ill during the
Walkerton, Ontario waterborne disease outbreak due to E. coli 0157 and Campylobacter jejuni,
researchers noted an increased incidence in chronic disease sequellae. Within a few years after
the outbreak, irritable bowel syndrome, hypertension, reactive arthritis and reduced kidney
function increased 33-38% (Clark etal. 2008).
Shigella bacteria are distinct because they are often associated with bloody diarrhea
(bacillary dysentery). The enterohemorraghic E. coli bacteria acquired the capability to produce
toxins by exchanging plasmids with Shigella (Murray et al., 2007). Thus, Shigella often also
causes kidney failure and chronic kidney disease. Shigella contamination only results from
human fecal contamination and thus it is probably less common than E. coli contamination,
which has both human and animal sources.
Campylobacter (like Salmonella) are very common contaminants of food and water.
Campylobacter is commonly associated with animal manure, especially cow and chicken
manure. Campylobacter and Salmonella are associated with many common source (food or
water) exposure deaths, probably in large part because there are a large number of illnesses.
More deaths may be associated with Campylobacter and Salmonella than with viruses. Uniquely,
Campylobacter is often associated with Guillain-Barre paralysis that can last for weeks or
months. About 1 paralysis case occurs for every 1000 cases of campylobacteriosis (Altekruse et
al. 1999). About 20% of paralysis patients are left with some disability and approximately 5%
die. Campylobacterosis is also associated with Reiter syndrome (reactive arthritis).
Approximately 1% of patients with camplybacterosis have arthritis onset in one or more joints
(especially the knee) in the 7 to 10 days after diarrheal onset (Altekruse etal. 1999). Arcobacter
(now a separate genus from Campylobacter) was responsible for a ground water outbreak at a
camp in Coeur d'Alene, Idaho (McMillan, 1996).
Salmonella causes typhoid fever, once a common and dangerous waterborne disease.
Typhoid is no longer a problem in the United States, and in recent years, Salmonella has become
increasingly less common as a common source outbreak agent while campylobacterosis
outbreaks have correspondingly increased. The reasons for this change are unclear. Seven deaths
that occurred due to Salmonella contamination in a ground water PWS in Gideon, Missouri were
due to bird entry into a storage tank (Angulo et al. 1997). Salmonella resulted in a very large
outbreak in a ground water utility in Riverside, California during the 1960s (16,000 illnesses, 70
hospitalizations and 3 deaths) prior to the advent of the Total Coliform Rule
(Boring et al. 1971). A 2008 waterborne outbreak in Alamosa, Colorado was due to Salmonella.
Salmonella-positive samples were collected from the distribution system but no E. coli was
identified in other distribution samples collected on the same day.
Legionella are opportunistic bacterial pathogens that colonize water distributions
systems. An estimated 8,000-10,000 cases of Legionaires disease and Pontiac fever occur in the
U.S. each year due to Legionella. Twenty-one of 48 known species are able to infect humans.
Helicobacter pylori is often associated with ground water (Hegarty et al. 1999; Rolle-
Kampzczyk et al. 2004) and is known to cause gastric ulcers. However, Helicobacter is not
culturable and so occurrence data are fairly uncertain. Improved hygiene and water treatment
have together reduced the number of ulcers caused by this organism over the last few decades
but it is impossible to quantify that decrease.
Economic Analysis for the Proposed RTCR 2-5
June 2010
-------
Most of the waterborne bacterial pathogens cause gastrointestinal illness, but some can
cause other severe illnesses as well. For example, Legionella causes Legionnaires Disease, a
form of pneumonia that has a fatality rate of about 15 percent. It can also cause Pontiac Fever,
which is a milder respiratory infection form of Legionnaires Disease. Several strains of E. coli
can cause severe disease, including kidney failure. Some bacterial pathogens are opportunistic
(i.e., they are only infectious in the presence of another, preexisting condition or weakness).
Opportunistic pathogens usually cause illness only in immunocompromised persons or in other
sensitive subpopulations, such as the very young or the elderly. Other pathogens, such as
Salmonella, Shigella, and Campylobacter jejuni, are not entirely opportunistic but result in
certain diseases with greater frequency and severity in immunocompromised persons (Framm
and Soave, 1997).
Examples of common viral pathogens include norovirus and the enteroviruses
(coxsackievirus, echovirus, poliovirus, enterovirus 70 and 71). The virus can cause a spectrum of
mild to severe clinical illness, including paralytic disease and death. There is considerable
information that Type 1 diabetes may be associated with enterovirus infection, including
infection with coxsackievirus and echoviruses (Maria et al. 2005; Vreugdenhil et al. 2000).
Dilated cardiomyopathy can follow myocarditis caused by echovirus and other enterovirus infection.
Mild enteroviral illness includes nonspecific febrile illness, respiratory illness,
photophobia or sensitivity, stiff neck, and gastrointestinal illness. Aseptic meningitis may or may
not require a doctor's visit, but more severe illnesses such as viral encephalitis, myocarditis, and
non-polio flaccid paralysis are likely to require hospitalization. Most likely to be hospitalized are
infants less than 3 months old with non-specific febrile illnesses that require treatment to rule out
and expectantly treat serious bacterial illness. Dilated cardiomyopathy can follow myocarditis
caused by echovirus and other enterovirus infection.
The health effects of norovirus illness include acute onset of nausea, vomiting, abdominal
cramps and diarrhea. Vomiting is more prevalent among children. Many adults experience
vomiting as well as diarrhea. Constitutional symptoms (e.g., headache, fever, chills, and
myalgia) are frequently reported. Although rare, severe dehydration caused by norovirus
gastroenteritis can be fatal, with this outcome occurring among susceptible persons (e.g., older
persons with debilitating health conditions). No long-term sequelae of norovirus infection have
been reported (CDC, 2001a). Duration of illness is typically 12-60 hours.
Hepatitis A (HAV) virus is the only waterborne virus that is reportable to CDC. About
28,000 HAV cases are reported to CDC each year, although that number is expected to decline
with time because children are currently vaccinated for HAV in high risk states and newer
recommendations are for increased vaccination coverage. More generally, however, because
HAV is more severe as an adult disease, an aging U.S. population may have greater disease
burden. Mead et al. (1999) estimate about 83,000 HAV cases each year, with a hospitalization
rate of 13% and a mortality rate of 0.3%.
Hepatitis E (HEV) virus is another fecal oral virus that is potentially waterborne. Using
serology and case histories of individual patients, HEV is established as endemic within the
United States (Tsang et al. 2000). However, the data suggest that only one or a few percent of
the population has been infected. Unlike HAV, no waterborne outbreaks of HEV have occurred
in the United Sates, although they have occurred in China and Somalia. The disease is severe,
with up to 20% mortality among pregnant women in developing countries.
Economic Analysis for the Proposed RTCR 2-6
June 2010
-------
The adenoviruses are a large group of viruses that produce diverse symptoms. Two
adenovirus serotypes, adenovirus 40 and 41, produce primarily enteric symptoms, but several
other adenoviruses are also capable of producing such symptoms. Some cause conjunctivitis.
Most significantly, adenoviruses caused a fatal outcome in otherwise healthy young males in
military settings (CDC, 2001b). All adenoviruses, no matter the infection site and characteristic
illness, are shed copiously through the gut. Thus, the adenoviruses are fecal\oral viruses (Carter,
2005) potentially transmissible via water to infection sites in the gut or, for the respiratory
adenoviruses, in the lung, via waterborne aerosol drops.
Astrovirus is commonly acquired in child care settings and causes mild disease in
children. Most children are exposed at an early age. However, a small percentage of that large
population may suffer more significant health effects and may require in-patient care. The
disease burden in older children and adult populations is underestimated because the disease is
mild (Carter, 2005). Astroviruses are shed in stool at large numbers. In France, prospective
epidemiology studies have implicated water as a route of infection (Gofti-Laroche, eial. 2003).
The Centers for Disease Control and Prevention (CDC) has determined that the incidence
of rotavirus diarrhea can reach 0.30 episodes/child/year by age two, with a cumulative incidence
approaching 0.80 episodes/child by age five (Glass et al. 1996). Hospitalizations for rotavirus
diarrhea are most common in children 6 months to 3 years of age (Parashar et al. 1998), while
self-limiting norovirus infections are prevalent in school-age children (LeBaron et al. 1990).
Although deaths from infectious diarrhea have generally declined among U.S. children since
1965 because of re-hydration therapy, newborn children, especially infants born prematurely,
remain at risk of death from severe diarrheal illness (Kilgore et al. 1995). Common strain
rotaviruses typically affect young children, particularly those less than 3 years old, but other
strains (Gl, G2 and G9) have been found to be common in adults and the elderly in nursing
homes, and found to be responsible for more severe illness in children (Griffin et al. 2000). In
2006, the U.S. Food and Drug Administration (FDA) approved a new pentavalent vaccine for
rotavirus (Vesikari et al. 2006) in children. The vaccine protects against the most common
rotaviruses now found in the U.S. (serotypes Gl, G2, G3, G4, G9) but does not protect again all
rotaviruses found in the U.S. or all rotaviruses.
Reovirus is recovered in environmental samples using the BGM cell line and is
commonly found co-occurring with the enteroviruses in PWS wells. Carducci et al. (2002) found
that, in some cases, enterovirus detection was limited because reovirus reproduction was so
highly favored. Reovirus is more closely related to rotavirus and thus has some similar
characteristics. Reovirus is now recognized as a human pathogen in children (Tyler et al. 2004).
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 oocysts are 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. Emerging disinfectant-resistant pathogens, such as
Microsporidia, Cyclospora, and Toxoplasma, are also a concern for similar reasons.
Cryptosporidiosis is a protozoal infection that usually causes 7 tol4 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
Cryptosporidium oocysts, which are readily carried in water. The most common source of
Economic Analysis for the Proposed RTCR 2-7
June 2010
-------
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. 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 AIDS patients, the elderly with other underlying illnesses, and other immuno-
compromised individuals. In a Cryptosporidium outbreak in Milwaukee in 1994, 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. 1996).
Giardiasis is a protozoal infection that usually causes a diarrhea, possibly accompanied
by fever, nausea and abdominal cramps in individuals with healthy immune systems. It is caused
by ingestion of Giardia cysts.
In 2003, two five-year old boys living in the same water service area near Phoenix AZ
died in the same week from Primary Amoebic Meningioencephalitis (PAM) (Marciano-Cabral et
al. 2003). Both boys lived in homes supplied by untreated PWS wells. Atypically, the wells in
that area provide water at elevated temperatures, representing the elevated geothermal gradient in
the subsurface. Seventeen samples taken from the boys homes were positive for Naegleria
fowleri and N. fowleri was also responsible for the boy's deaths from PAM. It is likely that the
elevated ground water temperature provided a suitable habitat for TV. fowleri colonization and
growth either in the aquifer, the well, the distribution system or the household plumbing.
2.2.2 Total Coliforms as Indicators of Treatment Effectiveness and Integrity of the
Distribution System
TC is a group of closely related bacteria that, with a few exceptions, are not harmful to
humans. Many bacteria in the TC group are soil bacteria but some, such as E. coli, originate in
the gut of warm-blooded animals. Coliform bacteria may be transported to surface water by run-
off or to ground water by infiltration. Total coliforms are common in ambient water and may be
injured by environmental stresses such as lack of nutrients, and water treatment chlorine
disinfection, in a manner similar to most bacterial pathogens and many viral enteric pathogens
(including fecal pathogens). EPA considers TC to be a useful indicator of a potential pathway
through which fecal contamination can enter the distribution system. The absence (versus the
presence) of total coliforms in the distribution system indicates a reduced likelihood that fecal
contamination and/or waterborne pathogens are occurring in the distribution system.
Under the current TCR, each TC-positive sample is assayed for either fecal coliform oris.
coli. Fecal coliform bacteria are a subgroup of TC that traditionally has been associated with
fecal contamination. Since the promulgation of the TCR, more information and understanding of
the suitability of fecal coliform and E. coli as indicators have become available. Study has shown
that the fecal coliform assay is imprecise and too often captures bacteria that do not originate in
Economic Analysis for the Proposed RTCR 2-8
June 2010
-------
the human or mammal gut (Edberg et al. 2000). On the other hand, E. coli is a more restricted
group of coliform bacteria that almost always originate in the human or animal gut (Edberg et al.
2000). Thus, E. Coli is a better indicator of fecal contamination than fecal coliforms.
2.2.3 Sanitary Defects
As part of the proposed RTCR, TC will be used as an indicator to start an evaluation
process that, where necessary, will require the PWS to correct sanitary defects, which are defined
as "a defect that could provide a pathway of entry for microbial contamination into the
distribution system or that is indicative of a failure or imminent failure in a barrier that is already
in place." "Sanitary defect" is a term specific to the TCR assessment and correction provisions.
Sanitary defects are not intended to be linked directly to the "significant deficiencies" under the
SWTR and GWR, although some problems could meet either definition. The proposed RTCR is
not intended to limit the existing authorities of primacy agencies under other regulations.
2.2.4 Occurrence of Fecal Contamination and/or Waterborne Pathogens
2.2.4.1 Presence of Fecal Contamination
Fecal contamination is a very general term that includes all of the organisms found in
feces, both pathogenic and nonpathogenic. Fecal contamination can occur in drinking water both
through use of contaminated source water as well as direct intrusion into the drinking water
distribution system. Biofilms in distribution systems may harbor waterborne pathogens and
accumulate enteric viruses and parasitic protozoa (Skraber et al. 2005; Helmi eial. 2008).
Waterborne pathogens in biofilms may have entered the distribution system as fecal
contamination from humans or animals.
Co-occurrence of indicators and waterborne pathogens is difficult to measure. The analytical
methods approved by EPA to assay for E. coli do not specifically identify most of the pathogenic
E. coli strains. There are at least 700 recognized E. coli strains. About 10% of the recognized E.
coli strains are pathogenic to humans (Feng 1995, Hussein 2007, Kaper et al. 2004). Pathogenic
E. coli include E. coli 0157:H7, which is the primary cause of hemolytic uremic syndrome
(HUS) in the United States (Rangel et al. 2005). The CDC estimates that there are 73,000 cases
of illness each year in the U.S. due to E. coli 0157:H7 (Mead et al. 1999). The CDC estimates
that about 15% of all reported E. coli 0157:H7 cases are due to water contamination (Rangel et
al. 2005). Active surveillance by CDC shows that 6.3 percent of E. coli 0157:H7 cases progress
to HUS (Griffin and Tauxe 1991; Gould et al. 2009) and about 12% of HUS cases result in death
within four years (Garg et al. 2003). About 4 to 15% of STEC 0157 cases are transmitted within
households by secondary transmission (Parry and Salmon 1998).
EPA believes that coli is a meaningful indicator for fecal contamination and the
potential presence of associated pathogen occurrence. Because EPA-approved standard methods
for E. coli do not typically identify the presence of the pathogenic E. coli strains, an E. coli-
positive monitoring result is an indicator of fecal contamination but is not necessarily a definitive
measure of waterborne pathogen occurrence. Specialized assays and methods must be used to
identify waterborne pathogens, including pathogenic E. coli and all others. These specialized
Economic Analysis for the Proposed RTCR 2-9
June 2010
-------
assays and methods are generally considered too costly and time consuming to be effective for
routine monitoring of drinking water distribution systems.
One notable exception are the data reported by Cooley et al. (2007), which showed high
concentrations of pathogenic E. coli strains in samples containing high concentrations of fecal
indicator E. coli. These data are from streams and other poor quality surface waters surrounding
California spinach fields associated with the recent E. coli 0157:H7 foodborne outbreak. Data
equivalent to these are not available from drinking water samples collected under the TCR.
2.2.4.2 Waterborne Disease Outbreaks
The CDC defines a waterborne disease outbreak as occurring when at least two persons
experience a similar illness, or one person is ill with amoebic meningioencephalitis, after the
ingestion of drinking water or following exposure to water used for recreational purposes, in
cases where the epidemiologic evidence implicates water as the probable source of the illness
(Kramer et al. 1996). The CDC maintains a database on waterborne disease outbreaks in the
United States. The database is based upon responses to a voluntary and confidential survey form
that is completed by state and local public health officials.
The National Research Council strongly suggests that the number of identified and
reported outbreaks in the CDC database for surface and ground waters represents a small
percentage of the actual number of waterborne disease outbreaks (Safe Water From Every Tap,
National Research Council 1997; Bennett et al. 1987, Hopkins et al. 1985 for Colorado data).
Under-reporting occurs because most waterborne outbreaks in community water systems are not
recognized until a sizable proportion of the population is ill (Perz et al. 1998; Craun 1996),
perhaps 1% to 2% of the population (Craun 1996).
EPA drinking water regulations are designed to protect against endemic waterborne
disease and to minimize waterborne outbreaks. Endemic waterborne disease may be defined as
any waterborne disease not associated with an outbreak.
2.3 Statutory Authority for Promulgating the Rule
As mentioned previously, SDWA requires EPA to review and revise, as appropriate, each
existing NPDWR at least once every six years (SDWA section 1412(b)(9), 42 U.S.C. § 300g-
1(b)(9)). Section 1412(b)(9) of SDWA states:
The Administrator shall, not less often than every 6 years, review and revise, as
appropriate, each national primary drinking water regulation promulgated under this title.
Any revision of a national primary drinking water regulation shall be promulgated in
accordance with this section, except that each revision shall maintain, or provide for
greater, protection of the health of persons.
In 2003, EPA completed its review of the TCR and 68 NPDWRs for chemicals that were
established prior to 1997 (USEPA 2002b, 67 FR 19030, April 17, 2002). In the Six-Year Review
determination published in July 2003, EPA stated its intent to revise the 1989 TCR (USEPA
2003b, 68 FR 42907, July 18, 2003).
Economic Analysis for the Proposed RTCR 2-10
June 2010
-------
2.4 Economic Rationale
As a revision to an existing regulation, the primary goal of the proposed RTCR is to
achieve the objectives of the 1989 TCR more effectively and efficiently, taking into account the
changes in regulatory framework for implementing the SDWA over the past 20 years and
experience with the TCR since it was promulgated in 1989. In this context, the overall economic
rationale behind the promulgation of the proposed RTCR is the same as the rationale for the
promulgation of the original TCR. This section addresses the economic rationale for choosing a
regulatory approach under the proposed RTCR (and TCR) rather than non regulatory options.
An economic rationale for the rule is required by Executive Order 12866, Regulatory
Planning and Review (58 FR 51735, October 1993), which states:
"[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 where 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
individual 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.
Many PWSs 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 (e.g., two supplier networks in a
single market) usually makes this situation unprofitable. The result is a market suitable for a
single supplier and hostile to alternative suppliers. In such natural monopolies, suppliers have
fewer incentives for providing quality services or maintaining competitive prices. In these
situations, governments often intervene to help protect the public interest.
For example, 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 structure may not facilitate the transmission of 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) most often represent a higher water cost
Economic Analysis for the Proposed RTCR 2-11
June 2010
-------
to consumers than the purchase from PWSs. Therefore, the water supplier may have little
incentive to improve water quality.
The public may also not understand the health and safety issues associated with poor
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. EPA has implemented a Consumer
Confidence Report (CCR) Rule (63 FR 44512, August 1998) that makes water quality
information more easily 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. Furthermore, even if informed consumers are able to
engage PWSs in a dialogue about health issues, the transaction costs of such interaction
(measured in personal time and monetary outlays) present another significant impediment to
consumer expression of risk reduction preferences.
SDWA regulations are intended to provide a level of protection from exposure to
drinking water contaminants. The regulations set minimum performance requirements to protect
consumers from exposure to contaminants. SDWA regulations are not intended to restructure
market mechanisms or 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 Proposed RTCR 2-12
June 2010
-------
3. Consideration of Regulatory Options
3.1 Introduction
This chapter describes the regulatory options considered during the development of the
proposed Revised Total Coliform Rule (RTCR) and evaluated as part of this Economic Analysis
(EA).
3.2 Total Coliform Rule/Distribution System Advisory Committee
In the July 2003 NPDWR review decision, EPA provided public notice of its intent to
revise the TCR. In 2007, EPA decided to establish a committee under the Federal Advisory
Committee Act (FACA) called the Total Coliform Rule/Distribution System Advisory
Committee (TCRDSAC) (USEPA, 2007a). The advisory committee was charged with
developing an Agreement in Principle (AIP) with recommendations for the TCR revision. The
advisory committee was to also provide EPA with recommendations needed to understand and
address possible public health impacts from potential degradation of drinking water quality in
distribution systems. Specifically, the major objectives of the advisory committee were to
provide advice and recommendations on:
• Revisions to the TCR that would improve implementation while maintaining or
improving public health protection and distribution system water quality. Examples
of the issues the TCRDSAC considered include: the TCR monitoring framework,
sanitary survey provisions, definition of Maximum Contaminant Level (MCL)
violations and potential follow-up corrective actions, and public notification of
violations.
• What data should be collected, research conducted, and/or risk management strategies
evaluated to better project distribution system contaminant occurrence and associated
public health risks in the distribution systems. This was intended to "initiate a process
for addressing cross connection control and backflow prevention requirements and
consider additional distribution system requirements related to significant health
risks" recommended by the M/DBP Federal Advisory Committee.
EPA held a series of 13 meetings of the advisory committee between July 2007 and
September 2008. The 15 committee members, along with other stakeholders, discussed options
for revising the TCR and began the process of developing the AIP by discussing the purpose,
efficacy and applicability of the current TCR as well as data collection and research needed to
better understand potential distribution system risks. The advisory committee formed a Technical
Work Group (TWG) to provide data analysis and information to inform the advisory
committee's discussions. The advisory committee also discussed the relevant provisions of the
Ground Water Rule (GWR) and discussed the extent to which other rules contribute to the
objectives of the current TCR. All advisory committee members signed the AIP in September
2008. All of the recommendations of the AIP, on which all the members of the committee
agreed, are found in the signed AIP (USEPA 2009, 74 FR 1683, January 13, 2009). The AIP and
Economic Analysis for the Proposed RTCR 3-1
June 2010
-------
details about the advisory committee can be found at EPA's website at:
http://www.epa.gov/safewater/disinfection/tcr/regulation revisions.html.
3.3 Regulatory Options Considered
EPA evaluated three regulatory options as part of the development of the RTCR proposal
- the current TCR option, the AIP option, and the Alternative option. The following discussion
provides an overview of the three regulatory options considered followed by a detailed table
comparing specific regulatory components. A detailed evaluation and comparison of the benefits
and costs of each of the three regulatory options for the proposed RTCR is found in later
chapters of this EA.
Baseline Option: Current Total Coliform Rale
The baseline option for the proposed RTCR EA represents the current TCR as written,
with adjustments made to reflect the effects of the GWR which PWSs need to comply with by
December 1, 2009. Analysis of this option is primarily based on data collected under the TCR, as
described in Chapter 4 (Baseline Conditions) of this EA.
The current TCR requires Public Water Systems (PWSs) to monitor routinely for Total
Coliform (TC) and fecal coliforms oris. coli. Monitoring requirements are based on system type,
population served by the PWS and source water type. Each PWS must sample according to a
written sample siting plan. Plans are subject to state review and revision.
The current TCR specifies a Maximum Contaminant Level Goal (MCLG) of zero for TC
(including fecal coliforms and E. coli) as well as an MCL where compliance is based on the
presence or absence of TC and/or fecal coliforms oris, coli in the samples. A PWS is required to
take repeat samples following a Total Coliform-positive (TC+) sample and to test the TC+
sample for E. coli. A monthly MCL violation is triggered if: (1) a PWS collecting fewer than 40
samples per month has > 1 routine/repeat sample per month which is TC+; or (2) a PWS
collecting at least 40 samples per month has > 5.0% of the routine/repeat samples in a month are
TC+. An acute MCL violation is triggered if any PWS has any Fecal Coliform- oris, coli -
positive (FC+ / EC+) repeat sample or has a FC+ or EC+ routine sample followed by a TC+
repeat sample. A PWS must demonstrate compliance with the MCL for total coliforms each
month it serves water to the public (or each calendar month that sampling occurs for PWSs on
reduced monitoring). MCL violations must be reported to the state no later than the end of the
next business day after the PWS learns of the violation. The public must also be notified
depending on the severity of the MCL violation (within 30 days for monthly MCL violations and
within 24 hours for acute MCL violations).
Agreement in Principle Option (as developed by the advisory committee, reflecting the proposed
RTCR language)
The AIP option is designed to trigger PWSs that exceed specified levels of TC is. coli in
their finished water to do an assessment, to identify whether sanitary defect(s) is (are) present,
and to correct such defects accordingly. The advisory committee believed that the AIP option is
an improvement over the current TCR framework because it takes a more preventive approach to
identifying and fixing problems with public health implications. Under this option, EPA
Economic Analysis for the Proposed RTCR 3-2
June 2010
-------
establishes an MCLG of zero for is. coli, an MCL for is. coli based on TC and E. coli monitoring
results, and a coliform treatment technique for protection against potential fecal contamination.
The AIP option maintains the current TCR requirement for all PWSs to monitor for TC
and E. coli according to a sample siting plan and schedule specific to the PWS. PWSs are
required to take repeat samples and to test for E. coli following a TC+ sample.
Under the AIP option, the PWS monitoring frequencies take into account the unique
characteristics of various PWS types and sizes. Small, well-operated PWSs using ground water
may be able to reduce monitoring frequencies by meeting specific criteria, thus reducing their
monitoring and reporting burden. Criteria for reduced monitoring include monitoring results that
reflect a clean compliance history as well as preventive practices that are designed to continue to
maintain the integrity of the distribution system.
The AIP option requires PWSs to complete either Level 1 or Level 2 assessments
following certain triggers (see Exhibit 3.1). The purpose of Level 1 and 2 assessments is to
identify the presence of sanitary defects and defects in distribution system coliform monitoring
practices. The minimum elements of both Level 1 and 2 assessments include review and
identification of:
• Inadequacies in sample sites, sampling protocol, and sample processing;
• Atypical events that may affect distributed water quality or indicate that distributed
water quality was impaired;
• Changes in distribution system maintenance and operation that may affect distributed
water quality including water storage;
• Source and treatment considerations that bear on distributed water quality, where
appropriate (e.g. small ground water PWSs); and
• Existing water quality monitoring data.
A Level 1 assessment (completed by the PWS) consists of a simple examination of the
system and relevant operational practices. The PWS must complete the assessment form and
submit it to the state for review within 30 days after determination that the PWS has exceeded
the Level 1 trigger. The completed assessment form must includeassessments conducted, all
sanitary defects identified (or indicate if no sanitary defect was found), corrective actions
completed, and a timetable for any corrective actions not already completed. Upon completion
and submission of the assessment form by the PWS, the state will determine if the PWS has
identified a likely cause for the Level 1 trigger and determine whether the PWS has corrected the
problem, or has included a schedule acceptable to the state for correcting the problem. If the state
determines the assessment insufficient, it will consult with the PWS. If necessary after
consultation, the PWS must submit a revised assessment form to the state on an agreed-upon
schedule not to exceed 30 days from the date of consultation. Upon completion and submission
of the assessment form by the PWS, the state must determine if the PWS has identified a likely
cause for the Level 1 trigger and, if so, establish that the PWS has corrected the problem, or has
included a schedule acceptable to the state for correcting the problem.
The Level 2 assessment must be conducted either by the state, a third party approved by
the state, or the PWS where the PWS has the staff or management with the required certification
Economic Analysis for the Proposed RTCR 3-3
June 2010
-------
or qualifications specified by the state. A Level 2 assessment is a more detailed examination of
the system's monitoring and operational practices than a Level 1 assessment. As with the Level 1
assessment, the completed assessment form for a Level 2 assessment must also be submitted to
the state for review within 30 days after determination that the PWS has exceeded the Level 2
trigger. The PWS must also indicate in the completed assessment form the sanitary defects
detected (or indicate if no sanitary defect was found), corrective actions completed, and a
timetable for any corrective actions not already completed. Upon completion and submission of
the assessment form by the PWS, the state will determine if the PWS has identified a likely cause
for the Level 2 trigger and determine whether the PWS has corrected the problem, or has
included a schedule acceptable to the state for correcting the problem. If the state determines the
assessment insufficient, it will consult with the PWS. If necessary after consultation, the PWS
must submit a revised assessment form to the state on an agreed-upon schedule not to exceed 30
days from the date of consultation. Upon completion and submission of the assessment form by
the PWS, the state must determine if the PWS has identified a likely cause for the Level 2 trigger
and, if so, establish that the PWS has corrected the problem, or has included a schedule
acceptable to the state for correcting the problem.
The AIP option requires PWSs to correct sanitary defects found through either a Level 1
or Level 2 assessment. The current TCR does not require PWSs that have TC MCL violations to
perform corrective actions. For corrections that are not completed by the time the PWS submits
the completed assessment form to the state, the PWS must complete the corrective action(s) on a
schedule determined by the state in consultation with the PWS. The PWS must notify the state
when each scheduled corrective action is completed.
The AIP option specifies violations corresponding to different degrees of potential public
health concern and public notification. A violation of an E. coli MCL occurs when:
• A routine sample is TC+ and one of its associated repeat samples is EC+; or
• A routine sample is EC+ and one of its associated repeat samples is TC+ (or as a
corollary to this condition, a routine sample is EC+ and one of its associated repeat
samples is EC+); or
• A system fails to test for E. coli when any repeat sample is TC+; or
• A system fails to take all required repeat samples following a routine sample that is
EC+.
An E. coli MCL violation requires a Tier 1 public notification. A coliform treatment technique
violation occurs when the PWS exceeds a treatment technique trigger (see Exhibit 3.1) and then
fails to conduct the required assessment or corrective action within the required timeframes. A
coliform treatment technique violation requires a Tier 2 public notification. If the system does
conduct the assessment and satisfies the requirements of the coliform treatment technique
(including corrective action when a sanitary defect is identified), no public notification is
required. A monitoring violation occurs when a PWS fails (1) to take every required routine or
additional routine sample in a compliance period, or (2) to analyze for E. coli following a TC+
sample. A reporting violation occurs when (1) a PWS properly conducts monitoring or an
Economic Analysis for the Proposed RTCR 3-4
June 2010
-------
assessment but fails to submit a monitoring report or a completed assessment form, or (2) a PWS
fails to notify the state following an EC+ sample. Both monitoring violation and reporting
violation require a Tier 3 public notification.
The provisions of the AIP consider the implication and linkages to other rules
promulgated by EPA under SDWA (e.g., GWR, Surface Water Treatment Rules (SWTRs), and
the Disinfectant/ Disinfection Byproduct (DBP) Rules). Compliance activities for other rules,
such as the sanitary survey, are used under the proposed RTCR as criteria to assess the integrity
of the PWS and to implement the reduced monitoring provisions in a cost-effective manner.
Ground water PWSs will have completed the sanitary survey component of the GWR at least
once no later than January 2015 so sanitary survey findings can be used by states in the
determination of the monitoring frequency.
Alternative Option: All PWSs Initially Sample for TC/EC on Monthly Basis
The third option is the Alternative option which parallels the AIP in most ways but
includes variations of some of the provisions that were discussed by the advisory committee
before consensus was reached on the AIP. Under the Alternative option, at the compliance date,
all PWSs would be required to sample monthly for an initial period until they meet the eligibility
criteria for reduced monitoring. This more stringent approach differs from the AIP option which
would allow PWSs to continue to monitor at their current frequencies (with an additional annual
site visit / voluntary Level 2 assessment requirement for PWSs wishing to remain on annual
monitoring) until they are triggered into an increased sampling frequency. Under the Alternative
option, no PWSs would be allowed to reduce monitoring to an annual basis. EPA defined the
Alternative option this way and included it in the economic analysis to assess the relative
impacts of a more stringent rule and to better understand the balance between costs and public
health protection.
3.3.1 Comparative Summary of Regulatory Options
Exhibit 3.1 below summarizes the components of each regulatory option considered for
the proposed RTCR.
Economic Analysis for the Proposed RTCR 3-5
June 2010
-------
Exhibit 3.1 Comparison of Proposed RTCR Regulatory Options
Element
Current TCR option
AIP option
Alternative option
Rule construct
• MCLG of zero for TC (including FC
and EC).
• TC monthly MCL based on the
number/percent of TC+ samples in
a month.
• FC / EC acute MCL based on
FC+/EC+ samples.
• Tier 1 Public Notification (PN)
required for MCL violations.
• No MCLG/MCL for TC, FC no longer used.
• EC MCLG of zero, and an EC MCL and a coliform treatment
technique (TT) based on TC and/or EC results.
• Assessment (and corrective action if necessary) required if PWS
has a coliform treatment technique trigger.
• Tier 1 PN not required for only TC occurrence.
Same as AIP.
Transition to the
New Rule
N/A
• PWSs continue on their existing TCR monitoring schedule
provided they meet criteria.
• Noncommunity Water Systems (NCWSs) on quarterly/annual
monitoring remain on that schedule unless/until they have an
event that triggers increased monitoring.
• Community Water Systems (CWSs) on reduced monitoring
remain on that schedule unless/until they have an event that
triggers return to routine monitoring.
• Monitoring schedules will be evaluated during the "special
monitoring evaluation" conducted by the state as part of the
periodic sanitary survey to determine if the monitoring frequency
is appropriate.
No transition. PWSs
start new requirements
as soon as the rule is
effective.
Routine
Monitoring
• 1 sample per quarter for NCWS
<1,000 Ground Water (GW).
• 1 sample per month for NCWS
<1,000 Surface Water (SW) and
all CWS <1,000.
• For all PWS >1,000, the number of
samples per month is based on
population served.
• Same as current TCR, with more explicit criteria to qualify for
reduced monitoring (see "Reduced Monitoring").
• 1 sample per month for seasonal systems2 with criteria to qualify
for reduced monitoring. Seasonal systems must also demonstrate
completion of a state -approved start-up procedure.
A minimum frequency
of monthly monitoring
for all PWSs.
Economic Analysis for the Proposed RTCR
3-6
June 2010
-------
Element
Current TCR option
AIP option
Alternative option
Reduced
Monitoring
• NCWS <1,000 (GW) can reduce to
1 sample per year if system is free
of sanitary defects.
• CWS <1,000 (GW) can reduce to 1
sample per quarter if no history of
TC contamination, no sanitary
defects, and protected GW source.
• No other systems are eligible for
reduced monitoring.
• NCWS < 1,000 (GW) - same as in current TCR, but more criteria
to qualify and remain on reduced annual monitoring.
o Most recent sanitary survey shows that system is free of
sanitary defects1, has a protected water source, and meets
approved construction standards;
o Clean compliance history for a minimum of 12 months;
o Level 2 assessment by party approved by state within the last
12 months and correction of all identified sanitary defects.
System must also have an annual site visit every year
thereafter to remain on annual monitoring.
• Seasonal systems can be eligible for reduced monitoring by
having an approved sample site plan that designates the time
period for monitoring based on demand and vulnerability
o For quarterly monitoring the system must also have a sanitary
survey or site visit or Level 2 assessment within last 12 months;
a protected water source; a clean compliance history for a
minimum of 12 months, and be free of sanitary defects.
o To reduce to 1 sample per year, the system must meet the
criteria specified above for quarterly monitoring and have in
place or adopt one or more additional enhancements to barriers
to contamination (cross connection control, certified operator,
meet disinfection criteria, maintenance of at least 4-log removal
or inactivation of viruses, other equivalent enhancements).
• CWS < 1,000 (GW) - same as in current TCR, but more criteria to
qualify and remain on reduced quarterly monitoring.
o State certified operator;
o Most recent sanitary survey shows that system is free of
sanitary defects (or has an approved plan and schedule to
correct them), has a protected water source, and meets
approved construction standards;
o Clean compliance history for a minimum of 12 months;
• NCWS < 1,000 (GW)
(including seasonal
systems) - reduced
quarterly monitoring
(criteria same as
AIP). Reduced
annual monitoring is
not allowed.
• CWS < 1,000 (GW) -
same as AIP.
• No other systems
are eligible for
reduced monitoring.
Economic Analysis for the Proposed RTCR
3-7
June 2010
-------
Element
Current TCR option
AIP option
Alternative option
o Meets at least one of the following criteria: annual site visit by
the state or a voluntary Level 2 assessment by a party
approved by the state or meeting criteria established by the
state; cross connection control; meet disinfection criteria;
maintenance of at least 4-log removal or inactivation of viruses;
other equivalent enhancements to water systems as approved
by the state.
• No other systems are eligible for reduced monitoring.
Increased
Monitoring
(NCWS) and
Return to Routine
Monitoring (CWS)
• N/A (none specified) - NCWS.
• N/A (none specified) - CWS
• NCWS < 1,000 (GW only) increases from quarterly or annual
monitoring to monthly monitoring if one of the following occurs:
o Triggered Level 2 assessment or a 2nd Level 1 assessment in
12 months;
o EC MCL violation;
o Coliform TT violation; or
o Two monitoring violations within 12 months if on quarterly
monitoring or one monitoring violation if on annual.
• CWS < 1,000 (GW only) on quarterly monitoring return to monthly
monitoring based on same criteria above.
Same as AIP with the
exception that reduced
annual monitoring is
not allowed under this
option.
Economic Analysis for the Proposed RTCR
3-8
June 2010
-------
Element
Current TCR option
AIP option
Alternative option
Return to
Reduced
Monitoring (After
Being Triggered
to Increased
Monitoring)
N/A
• NCWS < 1,000 (GW) must meet the following criteria to return to
routine quarterly monitoring after being triggered to increased
monitoring:
o Wthin the last 12 months, system must have completed a
sanitary survey or a site visit by the state or a voluntary Level 2
assessment, must be free of sanitary defects, and must have a
protected water source; and
o Clean compliance history for a minimum of 12 months.
• NCWS < 1,000 (GW) must meet the following criteria to return to
reduced annual monitoring in addition to meeting the criteria for
returning to routine quarterly monitoring:
o An annual site visit by the state or a voluntary Level 2
assessment and correction of all identified sanitary defects; and
o Adoption of one or more additional enhancements to the water
system barriers to contamination (cross connection control,
certified operator, meet disinfection criteria, maintenance of at
least 4-log removal or inactivation of viruses, other equivalent
enhancements).
• CWS < 1,000 (GW) must meet the same criteria for qualifying for
reduced quarterly monitoring.
Same as AIP with the
exception that reduced
annual monitoring is
not allowed under this
option.
Repeat
Monitoring
• PWS serving > 1,000 people must
take 3 repeat samples after a TC+
sample.
• PWS serving <1,000 must take 4
repeat samples after a TC+
sample.
• For GWS PWS, 1 sample can be a
source water sample to comply
with the GWR.
• All PWSs must take 3 repeat samples after a TC+ sample.
• PWS < 1,000, the number of repeat samples is reduced from 4
samples to 3.
• GW PWS must still take an additional source sample to comply
with the GWR.
• For GW PWS < 1,000, a single sample can meet both the
triggered source water requirements of the GWR and the repeat
sample requirements of the AIP option, but only if the state
approves the use of the single sample to meet both rule
requirements and the use of EC as the fecal indicator.
Same as AIP.
Economic Analysis for the Proposed RTCR
3-9
June 2010
-------
Element
Current TCR option
AIP option
Alternative option
Additional
Routine
Monitoring
PWS taking < 5 routine samples per
month (PWS serving < 4,100) must
take at least 5 routine samples the
month after a TC+ sample.
• For PWS taking samples less than monthly, the number of
samples required after a TC+ is reduced from 5 to 3.
• For PWS taking at least 1 sample per month, the additional
routine sample requirement is eliminated (they take their usual
number of samples the following month).
Same as AIP.
Sample Siting
Plan
• Sampling must occur at sites
representative of water quality in
the distribution system.
• Sample siting plans are subject to
state review/ revision.
• Special purpose samples are not
used for determining compliance.
• Take at least one repeat sample
within 5 connections up- and
downstream of the TC+ site.
Same as current except:
• Specifically allows for dedicated sampling stations.
• Provides more flexibility for systems in determining the locations
for taking repeat samples.
• States may allow entry point sampling at ground water systems if
the overall plan remains representative of water quality in the
distribution system.
Same as AIP.
Assessment
N/A - none required in the current
TCR.
• The PWS must conduct a Level 1 (self-assessment) if it exceeds
any of the following triggers:
o For systems taking > 40 samples per month, the PWS exceeds
5.0% TC+ samples for the month; or
o For systems taking < 40 samples per month, the PWS has > 2
TC+ samples for the month; or
o The PWS fails to take every required repeat sample after any
single routine TC+ sample.
• The PWS must ensure that a Level 2 assessment is conducted
either by the state or a state -approved 3rd party (including
qualified PWS employee) if it exceeds any of the following
triggers:
o The PWS has an E. coli MCL violation.
o The PWS has a second Level 1 trigger within a rolling 12-month
period, or in 2 consecutive years for systems on annual
Same as AIP.
Economic Analysis for the Proposed RTCR
3-10
June 2010
-------
Element
Current TCR option
AIP option
Alternative option
monitoring.
• Assessment results and description of corrective action taken will
be submitted to the state within 30 days.
Corrective Action
N/A - none required in the current
TCR.
• System must correct all sanitary defects found in the assessment.
• If none found, state must be satisfied with the assessment.
• For corrections not completed by the time the assessment form is
submitted, the systems must be in compliance with a state
determined schedule and must notify the state when completed.
Same as AIP.
Violations and
Public
Notification (PN)
• EC/FC MCL violation - acute
violation, Tier 1 PN.
• Monthly TC MCL violation - Tier 2
PN.
• Monitoring or reporting violation -
Tier 3 PN.
• PWS must notify state regarding
single EC+/FC+ result.
• EC MCL violation - Tier 1 PN. Failure to take repeat samples
following an EC+ is also an EC MCL violation.
• Monthly TC MCL violation is dropped - conditions that trigger a
monthly TC MCL violation under the TCR trigger assessment and
corrective action instead.
• Coliform TT violation occurs when a PWS fails to conduct required
assessment and corrective action - Tier 2 PN.
• Monitoring and reporting violations will be tracked separately -
Tier 3 PN.
Same as AIP.
The advisory committee recommended the following definition of "sanitary defect"-"a defect that could provide a pathway of entry for microbial
contamination into the distribution system or that is indicative of a failure or imminent failure in a barrier that is already in place"
2 As described in the preamble of the proposed RTCR, the AIP defines a seasonal system as "one which operates less than four calendar quarters per
year." (AIP, p 10) EPA has interpreted this to mean that a seasonal system is one that is shut down for at least one full calendar quarter (i.e., operates
"in" three or fewer calendar quarters).
Economic Analysis for the Proposed RTCR
3-11
June 2010
-------
3.4 Proposed Rule Requirements
Through coordination and consultation with the advisory committee and stakeholders as
described in Section 3.2, EPA selected the AIP option as the preferred option for proposal. The
AIP option was developed by the advisory committee with the primary goal of achieving the
objectives of the 1989 TCR more effectively and efficiently, taking into account the changes in
regulatory framework for implementing the SDWA over the past 20 years and experience with
the TCR since it was promulgated in 1989. Additional discussion supporting the selection of the
AIP option as the preferred regulatory option in terms of benefits and costs is provided in
Chapters 6 (benefits), 7 (costs), and 9 (benefits and costs).
Economic Analysis for the Proposed RTCR 3-12
June 2010
-------
4. Baseline Conditions
4.1 Introduction
A primary step in the development of the Economic Analysis (EA) for the proposed
Revised Total Coliform Rule (RTCR) was to estimate the baseline conditions. The U.S.
Environmental Protection Agency (EPA) used available data to develop an occurrence and
predictive model for Public Water Systems (PWSs) serving 4,100 people or fewer based
primarily on the 2005 Six-Year Review 2 data to predict total coliform and E. coli occurrence,
Level 1 and Level 2 assessment triggers, corrective actions, and violations, both at the baseline
and over time. EPA developed another, simpler, predictive model for PWSs serving more than
4,100 people based primarily on the 2005 Safe Drinking Water Information System-Federal
Version (SDWIS/FED) violation data (USEPA, 2005a) that predicts triggers, assessments,
corrective actions, and violations both at the baseline and over time, but not total coliform and E.
coli occurrence. Because five years of Ground Water Rule (GWR) implementation prior to the
effective date of the proposed RTCR are expected to cause changes to Ground Water Systems
(GWSs), the baseline conditions that EPA developed for the EA account for the effects of the
GWR, as described in this chapter and Chapter 5. The remainder of this chapter describes the
data sources used to develop the baselines and how the data were used. The resulting estimate of
baseline conditions serves as a reference point for understanding net impacts of the proposed rule
revisions, as discussed in Chapters 6, 7, and 9 of this EA.
4.1.1 Background and Purpose
The baseline analysis is a characterization of the water industry and its current operations
in effect before systems make changes to meet requirements of the proposed RTCR. This chapter
presents estimates of the number of systems, the population affected, water quality measures, and
subpopulations affected. The data collected for this profile will serve as the baseline used in the
proposed RTCR EA. The development of the baseline analysis consists of the following steps:
• Compilation of a profile of water systems - identifying and collecting information on all
PWSs.
• Characterization of current monitoring requirements, including testing requirements,
sample siting plans, routine and repeat sampling requirements, and sanitary surveys.
• Characterization of current PWS compliance information, including acute and monthly
Maximum Contaminant Level (MCL) violations, triggered monitoring, and public
notification reporting.
4.1.2 Chapter Organization
The remainder of this document is organized into three general sections:
Section 4.2 Describes general information on data sources.
Economic Analysis for the Proposed RTCR
4-1
June 2010
-------
Section 4.3
Describes the baseline conditions.
Section 4.4 Estimates baseline impact on sensitive populations
Section 4.5 Presents a summary of assumptions used to develop baseline estimates,
associated uncertainty and variability, and potential impacts on the results
of the RTCR EA.
4.2 Data Sources
Several primary data sources were used for the EA to characterize the proposed RTCR
baseline to create system and population baselines, including the SDWIS/FED (USEPA, 2007b),
the 2005 Six-Year Review of National Primary Drinking Water Regulations (Six-Year Review
2), the Ground Water Rule EA, and the U.S. Census. Each of these data sources is further
explained in Sections 4.2.1 - 4.2.3 below.
4.2.1 Background on SDWIS/FED Data
SDWIS/FED1 is EPA's national regulatory compliance database for the drinking water
program and is the main source of PWS inventory and violation data for the proposed RTCR
baseline. SDWIS/FED contains information on each of the approximately 155,000 active PWSs
as reported by primacy agencies2, EPA Regions, and EPA headquarters personnel. SDWIS/FED
does not include sample results, but does include identification of Maximum Contaminant Level
(MCL) violations and monitoring and reporting violations (both routine and repeat and minor
and major)3. It also contains information to characterize the U.S. inventory of PWSs, namely:
1 Further information on SDWIS/FED can be found on EPA's website:
http://www.epa.gov/safewater/databases/sdwis/index.htinl.
2 States and Indian Tribes are given primary enforcement responsibility (primacy) for regulations pertaining to
public water systems in their State/jurisdiction if they meet certain requirements specified under the primacy
regulations at 40CFR142, Subpart B (revised most recently in 1998).
3 Under the current TCR, a non-acute (monthly) MCL violation occurs when >1 routine and/or repeat sample are
Total Colifonn-positive (TC+) for systems that take fewer than 40 samples monthly, or greater than 5.0% of
monthly samples are TC+ for systems that take at least 40 samples. An acute MCL violation is triggered if a PWS
has a positive fecal colifonn (FC) or E. coli (EC) REPEAT sample (following a TC+) or has a positive FC/EC
ROUTINE sample followed by a TC+ REPEAT sample. Monitoring violations occur when a system 1) does not
satisfy the sample siting plan requirement; 2) does not sample in accordance with its required schedule; 3) does not
analyze a positive TC ROUTINE sample for FC or EC; 4) does not collect REPEAT samples within 24 hours and
analyze for TC following a ROUTINE positive TC; and 5) does not test REPEAT samples that are positive for TC
for FC or EC. If a system fails to conduct some of the required routine samples in a compliance period, the system
incurs a Minor Routine Monitoring Violation. Equivalently, a system that fails to conduct some or all of its required
repeat samples following a positive TC routine sample incurs a Minor Repeat MR Violation or a Major Repeat MR
Violation, respectively. Reporting violations occur when a PWS does not report to the state or to the public
according to the schedule for each violation type. For Acute MCL violations, a PWS must report to the state by end
of day and to the public within 24 hours (by posting, hand delivery, or mass media subject to primacy agency
approval). For Monthly MCL violations, a PWS must report to the state by end of next day, and to the public within
30 days by mail, hand delivery, or other methods approved by the primacy agency. For Monitoring violations, a
PWS must notify the public with 1 year in accordance with general public notification requirements approved by the
primacy agency.
Economic Analysis for the Proposed RTCR
4-2
June 2010
-------
system name and location; retail population served; source water type (i.e., Ground Water (GW),
Surface Water (SW), or Ground Water Under the Direct Influence of Surface Water (GWUDI));
whether or not systems disinfect the water; and PWS type, as described below.
EPA defines a PWS as a system that provides water for human consumption through
pipes or other constructed conveyances if such a system has at least 15 service connections or
regularly serves an average of at least 25 individuals per day for at least 60 days per year. PWSs
are categorized as follows:
• Community Water Systems (CWS) are PWSs that supply water to the same population
year-round.
• Noncommunity Water Systems (NCWS) are PWSs that supply water to a varying
population or one that is served less than year-round. They are sub-categorized as
follows:
- Nontransient Noncommunity Water Systems (NTNCWS) are public water
systems that are not community water systems and that regularly supply water to
at least 25 of the same people at least six months per year, for example, schools.
- Transient Noncommunity Water Systems (TNCWS) are NCWSs that provide
water in places such as gas stations or campgrounds where people do not remain
for long periods of time.
Additionally, PWS are analyzed in this EA according to categories defined by the number
of people they serve. The following nine categories of populations served by PWSs are used
throughout this document:
• <100.
• 101-500.
• 501-1,000.
• 1,001-4,100.
• 4,101-33,000.
• 33,001-96,000.
• 96,001-500,000.
• 501,001-1 Million.
• Over 1 Million.
These population categories are mostly consistent with those analyzed for other rules and
consider the distinctions in cost and system operation that are meaningful when considering the
economic effects of rule requirements. In particular, under the current Total Coliform Rule
(TCR), PWSs serving more than 1,000 people all test for Total Coliform (TC) monthly; PWSs
serving fewer than 1,000 people may monitor quarterly or even annually and (assuming that they
collect fewer than 5 samples per month) must have a sanitary survey every 5 years. The 33,000
threshold is significant as it corresponds with sampling requirements (at least 40 samples per
month are taken) that affect how compliance with the rule is calculated. (PWSs with >33,000
persons served must take 40 or more samples in a month, and may trigger an MCL violation
based on the percentage of TC+ samples, in this case, >5.0 percent.) Finally, the size category
Economic Analysis for the Proposed RTCR
4-3
June 2010
-------
thresholds of 1,000; 4,100; 33,000; and 96,000 are consistent with the size categories used in the
TCR to determine the monitoring regimen.
4.2.1.1 Data Used from SDWIS/FED
To create the system and population baseline, EPA obtained the most current PWS
inventory data by downloading data from the 4th quarter of 2007 from SDWIS/FED (USEPA,
2007b). These data represent all current, active PWSs and the population served by these
systems. This information composes the PWS inventory baseline for this EA, and is summarized
in Section 4.3.
EPA also used MCL violations data from SDWIS/FED to validate model predictions for
systems serving <4,100 people as described in Section 5.3.3.2, and to predict E. coli (acute)
MCL violations (current TCR, the Agreement in Principle or AIP, and Alternative option), total
coliform (non-acute or monthly) MCL violations (current TCR) and Level 1 and Level 2
assessment triggers (AIP option and Alternative option) for systems serving more than 4,100
people, as described in Section 4.3.4.2.
4.2.1.2 Data Cleaning
Data obtained from SDWIS/FED comprised PWS inventory data for the 4th quarter of
2007 and violations data from 3rd quarter 2005. The 4th quarter inventory data included the
number of each type of PWS and the retail population served by each system for all active
PWSs, and was used from SDWIS/FED without manipulation for the purpose of this analysis.
For the SDWIS/FED violations dataset, EPA did exclude some data representing OH, tribes, and
territories as described in Section 4.3.4.2. The data set taken from the 2005 Six-Year Review 2
database was cleaned as described in Section 4.2.2 of this EA.
4.2.1.3 Representativeness and Quality of SDWIS/FED Data
As noted above, SDWIS/FED is the source of PWS inventory data, and includes
information on all of the approximately 155,000 PWSs to the extent that such data was entered
by primacy agencies, EPA Regions, and EPA headquarters personnel. In 1998, EPA began a
major effort to assess the quality of the drinking water data in SDWIS, including performance of
a data quality assessment that has been published triennially since 2002. The most recent report
from this periodic assessment, the 2006 Drinking Water Data Reliability Analysis and Action
Plan (2006 report) (USEPA, 2008a), evaluated data for 2002 - 2004 and found the PWS
inventory data quality to be high (87%) and the TCR violations data quality to be moderately
high (81%). For the PWS inventory data used in this analysis, EPA found specifically that:
"... The overall data quality of the eight inventory (water system identification) parameters
assessed was 87%. In other words, 87% of PWSs from Data Verification4 (DV) states
4 EPA routinely evaluates state programs by conducting Data Verification (DV) audits, which evaluate the accuracy
of a state's decisions regarding PWS compliance with SDWA regulations and the accuracy with which inventory
and compliance data is reported (entered or uploaded into) SDWIS/FED.
Economic Analysis for the Proposed RTCR
4-4
June 2010
-------
between 2002 and 2004 had consistent data for all eight inventory data elements between
their state files and SDWIS/FED database, or 13% of PWSs had at least one data element
reported with a discrepancy. The highest discrepancy rate was for the administrative contact
address element."
The report's assessment indicates a high degree of completeness and accuracy in
inventory data submitted by states to SDWIS/FED, and suggests that the information is largely
representative of the PWSs in the United States. Although the proposed RTCR EA uses
inventory data from 2007, the assessment above applies to the most recent period analyzed in
this report (2002-2004); thus EPA believes that the inventory data from SDWIS/FED in 2007 is
also likely to be representative of the approximately 155,000 PWSs and that uncertainty in the
data with respect to numbers of systems, water source type5, and size classification is low.
For the TCR MCL violations data, the overall data quality estimate was 81%, based on a
rating of 84% for completeness of data and 97% accuracy of the data in SDWIS/FED as compared to
data observed in states' databases during the verification process6. As described in the report, the
84% completeness assessment indicates that for approximately 16% of those systems found by EPA
to have TCR MCL violations, states did not report the data in SDWIS/FED. Additionally, data
completeness for monitoring and reporting violations as a whole for all SDWA rules in SDWIS/FED
is low at 29%, although this metric specific to the TCR is not provided in the report. Low compliance
with monitoring and reporting may occur if systems would rather incur an MZR violation rather than
risk an MCL violation by sampling. These factors together suggest that violation rates would likely
be higher if systems were fully in compliance with monitoring and reporting requirements, and if
states were fully reporting those results to SDWIS/FED. Despite this potential downward bias in the
MCL violations data, the TCRDSAC was still able to note systematic trends in the data (across
time and among PWS types and sizes) and informed its decisions based on key observations:
• Most systems with violations are NCWSs.
• Lower violation rates observed among larger systems.
• Higher violation rates observed among GWSs than SWSs.
• Lower violation rates among disinfecting GWSs than other GWSs.
An additional concern with the representativeness of the violations data in SDWIS/FED that
would not be accounted for in the data quality estimate from the 2006 report is the potential bias
introduced by oversampling. In a recent paper by (Bennear et al., 2009), the authors attempt to
5 In particular, in the 2006 report the source water type (disinfecting, nondisinfecting, or status unknown) inventory
item was assessed at 99.3%, indicating that less than 1% of systems from DV states analyzed between 2002 and
2004 had a discrepancy in this information between State files and SDWIS/FED. For the purpose of the RTCR EA,
systems with status listed as "unknown" were included in the nondisinfecting group.
6The Overall Data Quality Estimate in SDWIS/FED measures how many noncompliance determinations are
correctly reported in SDWIS/FED among all noncompliance determinations (that are either violations or false-
positive violations). This quantity is estimated based on the violations confirmed by EPA and correctly reported to
SDWIS/FED out of all violations found by EPA or in the state files and SDWIS/FED. When the false positive rate is
0%, this measure is the product of Completeness and Accuracy. The False Positive rate of the violation data in
SDWIS/FED describes how much of the reported violation data in SDWIS are, in fact, false violations, expressed as
a percentage. Further description can be found in the 2006 Drinking Water Data Reliability Analysis and Action
Plan (2006 report) (USEPA, 2008a).
Economic Analysis for the Proposed RTCR
4-5
June 2010
-------
determine how often oversampling is an effort to "sample out" of a potential violation versus just
overcompliance - taking additional samples to increase the PWS's diagnostic power is assessing
the quality of the water. The authors did find evidence that strategic oversampling was occurring
in the one state included in their analysis7; some systems were reducing their probability of
incurring monthly violations (on the basis of TC positives as a percent of total monthly samples
taken) by increasing the total number of samples taken. Although the study was not broad
enough in scope to warrant adjusting the data used from SDWIS/FED, the issue is included in
the discussion of net benefits in Chapter 9 of this EA. Moving forward, removal of the public
notification requirement for monthly violations should help to reduce some of the incentive for
strategic sampling of this kind.
4.2.2 Background on 2005 Six-Year Review 2 Data
Through an Information Collection Request (USEPA, 2006b), EPA requested that states
voluntarily submit monitoring data (sample results) that were collected between January 1998
and December 2005 that were available electronically for specified chemical, radiological, and
microbiological contaminants. This request included data collected in compliance with the TCR
regarding the presence/absence of TC, E. coli, and/or fecal coliforms (FC) and any disinfection
data collected at TCR monitoring sites. (Surface water systems are required to monitor for the
presence of a disinfectant residual when collecting coliform samples in the distribution system.)
These data are an important component in supporting EPA's second Six-Year Review of
National Primary Drinking Water Regulations (NPDWRs). EPA encouraged each primacy
agency to submit its contaminant occurrence information because these data directly contribute
to EPA's understanding of national contaminant occurrence, populations exposed to regulated
contaminants, and exposure reductions associated with the current regulations.
EPA requested the TCR monitoring results with the intent of conducting analyses and
developing models to assess the potential impacts of changes to the rule. Models of the
occurrence of TC and E. coli were developed using the TCR data to assess revisions to
monitoring requirements and compliance decisions for different types of PWSs as suggested by
the stakeholders during the TCR Distribution System Advisory Committee (TCRDSAC) process,
described further in Chapter 3 of this EA.
For the RTCR EA, EPA used the 2005 Six Year Review data to develop the key parameters
used in the RTCR predictive model to forecast water quality (in terms of TC and E. coli
occurrence) under the current TCR, the AIP option, and the Alternative option. Additionally,
2005 Six Year Review data was used to inform EPA's assumptions regarding the proportion of
GWSs serving 1,000 or fewer people that sample monthly, quarterly, or annually (presented in
Exhibit 4.4).
7 Bennear et al. found that violations in the state of Massachusetts may have been approximately 30% greater had
the oversampling not taken place.
Economic Analysis for the Proposed RTCR
4-6
June 2010
-------
4.2.2.1 2005 Six-Year Review 2 Data Used
As described in the Total Coliform Rule Compliance Monitoring Data Quality and
Completion Report (USEPA, 2010e) (Data Quality Report) prepared during the TCRDSAC
process prior to this EA, the 2005 Six-Year Review 2 data is based on coliform monitoring data
voluntarily provided to EPA by 37 primacy agencies (35 states and 2 tribes). The data consist of
over nine million TCR records collected between 1998 and 2005. The data elements include the
following:
• PWS information - system type, population, source water type.
• Sampling types- routine, repeat, special, unknown.
• Sampling locations - entry point to the distribution system, distribution system, and,
for repeat samples, original location, downstream, upstream, and other.
• Methods - for testing for TC and E. coli.
• Monitoring results - presence/absence for total coliforms, fecal coliforms, and E. coli.
EPA prepared a tool that states using the SDWIS/state database system could use to
easily extract data from their databases and submit it to EPA. The data extraction tool was a
query that collected the data elements requested in the letter to the states. The states that used this
extraction tool are categorized as Tier 1 states. States submitting data in formats not compatible
with the EPA extraction tool are categorized as Tier 2 states. States that submitted incomplete or
problematic data are categorized as Tier 3, and states that did not provide data are categorized as
Tier 4 (e.g., California and Pennsylvania). The complete lists of states included in each of the
Tiers 1 - 4 are provided in the Data Quality Report.
EPA used 2005 data exclusively in developing this EA. This decision was made for
several reasons:
• The 2005 data, being the most recent in the Six-Year Review, were judged to be more
representative of present conditions than the less recent data, especially in terms of
the percent of TC records that were positive.
• There were fewer data in the years before 2005, especially 2001 and earlier, and
therefore data in earlier years was judged to be too sparse to be comparable with the
2005 data. The difference in the base of PWSs studied year-to-year might have
introduced an unknown bias.
• As noted earlier, the 2005 data had more records than other years, and enough to
represent the full spectrum of PWSs within states that provided the data.
• Using only a single year of data was beneficial in that is simplified the analysis to
include a fairly static set of PWSs, and did not require interpretation of the meaning
of differences in occurrence data observed across multiple years. Understanding these
types of differences would involve analysis of changing environmental factors,
Economic Analysis for the Proposed RTCR
4-7
June 2010
-------
program participation and administration factors, and possibly other factors before the
data could be used as a baseline in the model. Additionally, a full year (and not less)
of observed data was required for consistency with the basic unit of time in
occurrence estimates, which is one year; and a period of 12 consecutive months is
assumed to capture seasonal variation normally experienced in water quality.
• A single year of data would allow for use of a single database to store all the data for
easier analysis.
4.2.2.2 Data Cleaning of 2005 Six-Year Review 2 Data
The Data Quality Report (USEPA, 2010e) describes how TCR monitoring data were
obtained, evaluated, and modified where necessary to make the database internally consistent
and usable for analysis. Exhibit 2.1 in the Data Quality Report provides a complete list of states
or territories that submitted data and a description of the use of these data.
To determine whether a PWS's sampling data were complete in the 2005 Six-Year Review
2 dataset, a representative month for a system was defined as having at least 50 percent of the
number of samples expected based on the size of system and current TCR requirements. Using
this 50% criterion, months that were not representative of a PWS were excluded; this cleaning
resulted in a decrease of only 2% of submitted data, and is unlikely to have skewed occurrence
results, as shown by Exhibits 3.1 and 3.2, respectively, of the Data Quality Report. EPA
determined that a 50% criterion systematically eliminated systems that were only reporting
positive results or otherwise not reporting many of its samples, while retaining data for systems
that had increased from one size category to the next (and were therefore expected to have, for
example, 15 samples per month instead of 10). The Technical Work Group of the Federal
Advisory Committee agreed with this approach, which is explained further in Section 3.2 of the
Data Quality Report.
4.2.2.3 Representativeness of 2005 Six-Year Review 2 Data
In the course of cleaning the 2005 Six-Year Review 2 data, EPA examined data from the
10 largest PWSs in each state and found, by matching with SDWIS inventory information, that
many larger systems were not included in the data submitted by the states. For the largest
systems that are represented in the 2005 Six-Year Review 2 data, EPA also compared the
expected number of samples based on population served (i.e., the TCR requires a system serving
a million people to sample 300 times per month) to the number of samples in the actual data to
evaluate completeness. In some cases it was found that only data from the state laboratory were
included. Since many large PWSs have their own laboratory and do not use the state laboratories,
data for these large systems were not available. In many cases, states handled large PWSs with
negative samples in a special way, and those data were not captured by the data extraction tool.
The end result was that large PWSs had a lower percentage of samples included in the data than
did smaller systems; therefore, data for systems serving more than 4,100 were not used from this
database. Instead, as described in Section 4.3.4.2, EPA used SDWIS/FED violations data to
model changes in the number of corrective actions to be implemented and incremental costs
under the proposed RTCR for PWSs serving greater than 4,100 people.
Economic Analysis for the Proposed RTCR
4-8
June 2010
-------
Based on the number of smaller PWSs (serving fewer than 4,101 people) in states whose
data were excluded for data quality/quantity issues or for non-submittal of data, approximately
39 percent of systems were excluded from the analysis. Specifically, the following states or
territories submitted data but were excluded because of issues with the quality or quantity of data
submitted: AL, American Somoa, DC, ME, MD, NJ, SC, SD, TN, and the Navajo Nation.
Additionally, the following states/territories did not submit data: CA, Guam, HI, LA, MA, MS,
Northern Mariana Islands, PA, Puerto Rico, TX, Virgin Islands, and WA. The remaining 70
percent of PWSs' data were used to simulate occurrence of TC and E. coli under the current TCR
and AIP option regulatory alternatives and to develop parameters for use in the RTCR predictive
model, described in Chapter 5 of this EA.
Although approximately 61 percent of the systems are represented in the analysis, only an
estimated 32% of the population served by PWSs is represented due to the omission of some
large population states (i.e., CA, PA, and TX)8. However, since EPA is not predicting changes in
ocurrence for the largest systems (i.e., the majority of the population is served by very large
CWSs) the under-representation of population served in the data is not expected to have a
significant impact on the analyses performed.
4.2.3 Background on Other Data and Information Used
SDWIS/FED data and 2005 Six-Year Review 2 data represent the largest portion of
information used for the proposed RTCR baseline analysis. In addition, EPA incorporated
information from the GWR, the U.S. Census, and the current TCR to develop assumptions for
the model. EPA also relied upon the knowledge and experience of stakeholders representing
industry, states, small systems, and the public to inform the process throughout development of
the proposed RTCR.
4.2.3.1 Current TCR
Under the current TCR, PWSs experienced reduced occurrence of fecal indicator
contamination after the effective date in 1989 as a result of increased sampling, reporting, and to
some degree corrective actions. In recent years under the TCR, the occurrence of fecal indicators
has reached a steady state, as reflected in SDWIS/FED violation data (see Section 4.3.4.2 for
discussion). Additional reductions in occurrence are expected prior to proposed RTCR
implementation that reflect water quality improvements due to implementation of the GWR.
Because the water quality data used in this analysis is the Six-Year Review 2 data for 2005,
adjustments are made to the 2005 data in the modeling to reflect this GWR impact. Further
reductions in occurrence of fecal indicators under the proposed RTCR can then be properly
modeled based on more focused assessments and corrective actions in response to TC+ or EC+
samples.
8 As described further in the Data Quality Report, submission of the data was voluntary; CA, PA, and TX did not
submit data.
Economic Analysis for the Proposed RTCR
4-9
June 2010
-------
4.2.3.2 GWR data
The GWR was promulgated to address microbial contamination of ground water sources
used to supply drinking water and, specifically, to address the concern about potential adverse
health risks that may be associated with fecal contamination. Before the GWR was promulgated,
there were no federal regulations to require filtration or disinfection of ground water sources to
remove microbial contaminants. The GWR requirements include sanitary surveys, triggered
monitoring in response to TCR samples testing positive for contaminant indicators, source water
monitoring for indicators of fecal contamination such as E. coli,9 and corrective actions which
may include disinfection.
Both the GWR and the proposed RTCR seek to decrease the level of fecal contamination
in drinking water. Since the proposed RTCR applies to all public water systems including
systems that use ground water sources, the implementation of the GWR (requiring compliance
by December 1, 2009) in advance of the proposed RTCR is expected to modify the TCR baseline
(see Sections 4.3.1 - 4.3.2 below) taken from the 2007 SDWIS/FED (4th quarter) and 2005 Six
Year Review data (see Sections 4.2.1 and 4.2.2 for discussion of SDWIS/FED and 2005 Six
Year Review data, respectively). The number of disinfecting and non-disinfecting public
groundwater systems prior to the start of GWR implementation in October 2009 (based on 4th
quarter 2007 SDWIS/FED data) is 54,469 and 85,999, respectively. By October 2012 when the
GWR is fully implemented, EPA predicts that the number of systems disinfecting will have
increased from the 2007 4th quarter baseline. The proposed RTCR occurrence model estimates
and incorporates the estimated change in the number of systems with disinfection into this EA
baseline (Section 5.3.1).
The RTCR occurrence model (described in further detail in Section 5.3 of this EA)
accounts for GWR implemention in the following ways:
• The baseline number of disinfecting systems in the 4th quarter 2007 data of
SDWIS/FED will be increased to account for the additional systems that are
estimated to begin disinfecting under the GWR.
• Sanitary surveys, as required by th GWR, will be phased in for applicable PWSs in
the model (over five years for systems on a five year sanitary survey cycle; and over
three years for systems on a three year cycle), resulting in reduced occurrence for
these systems indefinitely to represent that systems are correcting deficiencies
identified in the survey. Further detail on this component of the model is provided in
Chapter 5 of this EA.
• GWR compliance monitoring will be incorporated with an additional decrease in
probability of occurrence to represent anticpated efforts by systems to identify and
correct problems.
9 Standridge, J. (2008) discusses the use of E. coli as an indicator of drinking water quality.
Economic Analysis for the Proposed RTCR
4-10
June 2010
-------
• The baseline estimates will account for the decreased levels of TC occurrence and
fecal contamination achieved under the GWR. This decrease will be reflected in
model output for the following proposed RTCR rule components:
- Repeat monitoring
- Additional routine monitoring
- Triggers for level 1 and 2 assessments
- Corrective actions
- Public notification
• Under the GWR, if a PWS experiences a TC+ while conducting TCR monitoring, the
system must sample the source water for the presence of a fecal indicator (E. coli,
enterococci, or coliphage, as determined by the state). Ground water PWSs that have
a positive initial fecal indicator sample from the source are required to take five
repeat fecal indicator samples if they are not required by the state to take corrective
action. Under the proposed RTCR, PWSs having less than 4-log treatment and
serving 1,000 people or fewer would only be required to take 3 repeat samples. To
meet GWR requirements, these systems would need to take one additional sample at
the source.
4.2.3.3 U.S. Census Data
The baseline analysis incorporates data from the U.S. Census that are publically available
and include surveys of the U.S. population, economics, industry, and geography. Census 2000
data were used in the baseline analyses to estimate the sensitive subpopulations in the U.S. who
may be more susceptible to illness as a result of poor drinking water quality, as explained in
Section 4.4 of this EA. Additionally, the Census 2008 Annual Social and Economic Supplement
was used in the estimates of household size as described in Section 4.3.3.
4.3 Baseline Profile
The proposed RTCR applies to all PWSs, regardless of their size, water source (ground
water or surface water), or type (community, noncommunity non-transient, noncommunity
transient). This section estimates the baseline number of PWSs and the size of the population
subject to the proposed RTCR in each of these subcategories. EPA used 4th quarter 2007 data
from SDWIS to develop this inventory baseline because. At the time when the TCRDSAC began
discussing the proposed rule and its analysis, 2007 data were the most recent inventory data
available. As described in Section 4.2, the best available occurrence data for smaller PWSs was
collected in 2005. For larger systems (serving >4,100 people) as Section 4.3.4.2 describes, EPA
used SDWIS-FED violations data for modeling benefits and costs for PWSs serving more than
4,100 people because the 2005 Six-Year Review 2 dataset (containing sampling data) for larger
systems was not as robust. Third quarter 2005 SDWIS/FED data is used for the violation
baseline since EPA compiles violation data in the 3rd quarter every year (as opposed to the 4th
quarter freeze used for annual compilation inventory data) and was the most recently available
data at the time that this analysis was begun. Use of the most recent year of data was deemed to
Economic Analysis for the Proposed RTCR
4-11
June 2010
-------
be equivalent to an average of multiple years due to the steady numbers of violations seen in the
data over the past several years.
The baseline described in the following sections is used as a reference point for informing
net costs and benefits as described in Chapters 6 and 7 of this EA.
4.3.1 Pre-GWR Baseline
Estimates of the number of ground water PWSs subject to the proposed RTCR are
presented in Exhibit 4.1 below. These numbers reflect SDWIS/FED inventory from the fourth
quarter of 2007, prior to implementation of the GWR. Ground water system inventories indicate
disinfection status as disinfecting or unreported; since reporting is voluntary, the status of a
system that did not report disinfection is actually unknown, although EPA assumes in this EA
that the system is not disinfecting. The number of disinfecting PWSs includes those achieving
less than 4-log disinfection, which matches the categorization of systems used to estimate the
underlying distributions of occurrence of TC and E. coli modeled, as described in Chapter 5.
Exhibit 4.2 presents the number of surface water PWSs subject to the proposed RTCR,
also derived from the SDWIS/FED database. GWR implementation does not affect the
disinfection status of surface water systems.
Exhibit 4.3 presents the population associated with each PWS category of system size
and type prior to incorporating the effects of the GWR into the model; this is the pre-GWR
baseline affected population. After incorporating GWR effects, some number of systems shift
into the disinfecting category, and some of the population shifts accordingly.
Lastly, Exhibit 4.4 presents the distribution of PWSs by frequency of sampling -
monthly, quarterly, or annually - under the current TCR for those systems that qualify for
reduced monitoring (all other systems are required to sample monthly). This is the baseline
distribution of sampling frequencies used for analyses in this EA. This distribution was derived
as part of the work done by the Technical Workgroup (TWG) supporting the TCRDSAC, and
validated by the Association of state Drinking Water Administrators (ASDWA). Workgroup
members analyzed 2005 Six Year Review and other readily available state data to derive the
estimates shown in Exhibit 4.4.
Economic Analysis for the Proposed RTCR
4-12
June 2010
-------
Exhibit 4.1 Pre-GWR Baseline Number of GW Systems
PWS Size
Number of GW PWSs (Pre-GWR)
CWS
NT NCWS
TNCWS
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
<100
6,132
5,806
2,907
5,919
13,558
46,642
101-500
9,260
4,632
2,753
3,860
5,341
13,934
501-1,000
3,502
965
868
850
673
1,290
1,001-4,100
5,405
1,038
542
270
269
348
4,101-33,000
2,798
358
56
14
27
40
33,001-96,000
307
28
2
-
-
2
96,001-500,000
62
1
-
-
-
1
500,001-1 Million
4
-
-
-
-
1
> 1 Million
3
-
-
-
-
-
Total
27,473
12,828
7,128
10,913
19,868
62,258
Combined Total
40,301
18,041
82,126
Source: Data extracted from SDWIS/FED PWS Inventory, 2007 4th Quarter Data based on listed disinfection status.
PWSs listed as "unknown" disinfection status in SDWIS/FED are counted as non-disinfecting.
Exhibit 4.2 Baseline Number of SW Systems
PWS Size
Number of SW PWSs
CWS
NTNCWS
TNCWS
<100
1,170
250
1,339
101-500
2,150
253
497
501-1,000
1,173
88
88
1,001-4,100
2,938
72
67
4,101-33,000
3,164
22
18
33,001-96,000
720
2
-
96,001-500,000
308
1
-
500,001-1 Million
31
-
-
> 1 Million
17
-
1
Total
11,671
688
2,010
Source: Data extracted from SDWIS/FED PWS Inventory, 2007 4 Quarter Data.
SW count includes GWUDI systems.
Economic Analysis for the Proposed RTCR
4-13
June 2010
-------
Exhibit 4.3 Pre-GWR Baseline Population Served by GW Systems
PWS Size
Population Served by GW PWSs (Pre-GWR)
CWS
NTNCWS
TNCWS
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
<100
387,558
342,243
168,482
326,342
645,949
2,139,529
101-500
2,375,507
1,085,555
708,424
948,500
1,231,077
3,005,703
501-1,000
2,570,662
703,562
630,071
600,643
516,707
956,781
1,001-4,100
11,307,740
2,074,139
991,971
456,985
449,757
599,171
4,101-33,000
29,346,057
3,627,365
418,368
122,865
194,136
353,290
33,001-96,000
15,587,186
1,527,861
89,405
-
-
119,700
96,001-500,000
9,935,500
107,323
-
-
-
100,000
500,001-1 Million
2,670,841
-
-
-
-
725,000
> 1 Million
4,389,948
-
-
-
-
-
Total
78,570,999
9,468,048
3,006,721
2,455,335
3,037,626
7,999,174
Combined Total
88,039,047
5,462,056
11,036,800
Source: Data extracted from SDWIS/FED PWS Inventory, 2007 4th Quarter Data based on listed disinfection status.
Populations for PWSs listed as "unknown" disinfection status in SDWIS/FED are counted as non-disinfecting.
Exhibit 4.4 Percent Distribution of Ground Water PWS Monitoring Frequencies by
PWS Size and Type for Current TCR
GW or
PWS
No.
Disinf?
SW?
Type
Size
PWSs
Monthly
Quarterly
Annual
Y
GW
TNCWS
<100
13,558
4.8%
62.9%
32.3%
Y
GW
TNCWS
101-1000
6,014
7.9%
66.9%
25.2%
Y
GW
TNCWS
1001-4100
269
100.0%
0.0%
0.0%
Y
GW
NTNCWS
<100
2,904
19.3%
64.6%
16.0%
Y
GW
NTNCWS
101-1000
3,621
18.5%
66.7%
14.7%
Y
GW
NTNCWS
1001-4100
542
100.0%
0.0%
0.0%
Y
GW
CWS
<100
6,132
86.6%
13.4%
0.0%
Y
GW
CWS
101-1000
12,762
88.5%
11.5%
0.0%
Y
GW
CWS
1001-4100
5,405
100.0%
0.0%
0.0%
N
GW
TNCWS
<100
46,642
4.8%
62.9%
32.3%
N
GW
TNCWS
101-1000
15,224
7.8%
66.8%
25.3%
N
GW
TNCWS
1001-4100
348
100.0%
0.0%
0.0%
N
GW
NTNCWS
<100
5,913
19.3%
64.6%
16.0%
N
GW
NTNCWS
101-1000
4,710
18.5%
66.7%
14.8%
N
GW
NTNCWS
1001-4100
270
100.0%
0.0%
0.0%
N
GW
CWS
<100
5,806
86.6%
13.4%
0.0%
N
GW
CWS
101-1000
5,597
88.6%
11.4%
0.0%
N
GW
CWS
1001-4100
1,038
100.0%
0.0%
0.0%
Note: All other system sizes and types are required to sample monthly. Source: Based on EPA and TWG
analysis of 2005 Six-Year data and individual state statutes during the TCRDSAC and review by ASDWA.
Economic Analysis for the Proposed RTCR
4-14
June 2010
-------
4.3.2 Post-GWR Baseline
PWSs must comply with requirements of the GWR by December 1, 2009, which will be
approximately five years prior to the effective date of the proposed RTCR. To account for the
impact of the GWR on baseline occurrence for those public water systems using a ground water
source, EPA performed a number of adjustments to the 2005 data prior to its use in this EA.
Using 2007 SDWIS inventory data (as explained in Section 4.3.1) as a pre-GWR
inventory baseline, EPA applied probabilities of Psampie and PWeii10 to determine the probability of
a positive sample for a fecal indicator in non-disinfecting PWSs. A fraction of those systems
with a positive sample is expected to move to a disinfecting status in the five years prior to
proposed RTCR implementation. This modification ensured that systems in the analysis that
disinfect drew from the probability distribution for TC+ and EC+ for disinfecting systems, which
is different from the distribution for those that do not disinfect. The model continues to apply
PWeii and PSampie throughout the proposed RTCR period of analysis (the 25 years following
promulgation). The application of this model parameter, as it is applied in the context of other
aspects of the predictive occurrence model, is explained in further detail in Chapter 5 of this EA.
Exhibit 4.5 shows the baseline inventory of ground water systems by disinfection status in model
year 5, which is the year immediately prior to the effective date of the proposed RTCR. The
estimates in Exhibit 4.5 of non-disinfecting PWSs include a number of systems that selected a
corrective action other than converting to a disinfecting system in response to fecal
contamination at the source.
As explained in Chapter 5 of this EA, a number of PWSs will undertake sanitary surveys
(for all GW systems) and compliance monitoring (for disinfecting GW systems) prior to
implementation of the proposed RTCR; although the costs are not considered in this EA (they
were considered in the GWR EA), the systems are assigned a reduced occurrence, which is
reflected in the proposed RTCR baseline. The predictive model incorporates the contribution of
sanitary survey results to the qualification of systems for reduced monitoring by assuming a 10%
reduction in the number of systems finding a TC+, which is applied to a portion of systems each
year and retained throughout the period of analysis. For noncommunity water systems, the
reduction is applied to one-fifth of systems per year for the five years of GWR implementation
prior to the RTCR effective date; for community water systems, it is applied to one-third of
systems in each of the first three years of the GWR implementation period. Similarly, the system
incorporates the effect of compliance monitoring by disinfecting GW systems by applying an
additional 10% reduction for systems that are disinfecting as of the RTCR effective date, and
additionally to those determined in the model to select disinfection as a corrective action; the
compliance monitoring reduction in occurrence is assumed to apply in all remaining years of the
analysis once a system qualifies for the reduction.
As described in Section 5.3.2.2 of this EA, model output was generated for a 30-year
period to encompass 5 years of the effects of GWR implementation and 25 years of the effects of
RTCR implementation. Chapter 5 presents this output for the 30-year period. For the purpose of
10 The term "Pweu" refers to the probability that a randomly selected well across the United States will ever test
positive in its source water for a virus or indicator species, such as TC; "Psampie" is the probability that given a
contaminated well, a random sample at the well will test positive for a virus or indicator. The GWR EA (November
2006) (USEPA, 2006a), and in particular the Baseline Chapter of that document, provides a detailed explanation of
the analysis used to generate these Pwen and Psampie estimates.
Economic Analysis for the Proposed RTCR
4-15
June 2010
-------
the benefit-cost analysis, years 3 through 27 of this period are included in the calculations and
presented in the results shown in Chapter 6 (Benefits Analysis), Chapter 7 (Cost Analysis), and
Chapter 9 (Comparison of Benefits and Costs). Initially in the modeled period, all systems under
the current TCR option and the AIP option are assumed to sample based on the sampling
regimen applicable to their system type, size, type of water source, and whether or not the system
disinfects(Exhibit 4.4).11 Under the Alternative option, all systems are assumed to sample
monthly for the first five years after the RTCR effective date. Following an assessment period as
described in Section 5.3.2.2, monitoring frequencies are adjusted from baseline for GW systems
serving for fewer than 1,000 people. All other systems remain on monthly monitoring.
The estimates for distribution of PWSs across the M/Q/A monitoring schedules as
adjusted from the baseline distribution are presented in Chapter 5 for the current TCR, AIP
option, and Alternative option (Exhibits 5.10a- 5.10c).
Exhibit 4.5 Post-GWR Baseline Number of GW Systems
PWS Size
Number of GW PWSs (Post-GWR)
CWS
NTNCWS
TNCWS
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
<100
6,308
5,630
2,937
5,889
13,781
46,419
101 -500
9,326
4,566
2,777
3,836
5,459
13,816
501-1,000
3,516
951
873
845
685
1,278
1,001-4,100
5,423
1,020
547
265
274
343
4,101-33,000
2,798
358
56
14
27
40
33,001-96,000
307
28
2
-
-
2
96,001-500,000
62
1
-
-
-
1
500,001-1 Million
4
-
-
-
-
1
> 1 Million
3
-
-
-
-
-
Total
27,746
12,555
7,192
10,849
20,226
61,900
Combined Total
40,301
18,041
82,126
Source: proposed RTCR Predictive Occurrence Model Output.
4.3.3 Baseline Population Served
PWS population characteristics are important to this analysis for determining the number
of people, both prior to and following implementation of the GWR, for which risk changes under
each component of the proposed RTCR, as discussed in Chapter 6 of this EA. These population
estimates are based on SDWIS/FED 2007 4th quarter data as described in Section 4.2.1, and
stratified by PWS size, type, whether the system disinfects or not and whether the system uses
GW or SW sources. The estimates represent the number of systems which disinfect water and
their populations served; both estimates are adjusted upward based on predicted impacts of the
GWR in the five years prior to the effective date of proposed RTCR requirements, as shown in
Exhibits 4.6 and 4.7 below. (See Exhibits 4.1 and 4.3 for PWS counts and population served by
disinfection status prior to affects of GWR.)
11 Based on EPA and TWG analysis of Six-Year data and individual state statutes during the TCRDSAC and review
byASDWA.
Economic Analysis for the Proposed RTCR
4-16
June 2010
-------
Exhibit 4.6 Post-GWR Baseline Population Served by GW Systems
PWS Size
Population Served by Post-GWR PWSs
CWS
NTNCWS
TNCWS
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
Disinfecting
Non-Disinfecting
<100
397,933
331,868
170,149
324,675
656,169
2,129,309
101 -500
2,390,906
1,070,156
714,232
942,692
1,256,597
2,980,183
501-1,000
2,580,602
693,622
633,813
596,901
525,325
948,163
1,001-4,100
11,343,429
2,038,450
999,929
449,027
458,707
590,221
4,101-33,000
29,346,057
3,627,365
418,368
122,865
194,136
353,290
33,001-96,000
15,587,186
1,527,861
89,405
-
-
119,700
96,001-500,000
9,935,500
107,323
-
-
-
100,000
500,001-1 Million
2,670,841
-
-
-
-
725,000
> 1 Million
4,389,948
-
-
-
-
-
Total
78,642,402
9,396,645
3,025,896
2,436,160
3,090,935
7,945,865
Combined Total
88,039,047
5,462,056
11,036,800
Source: Estimates calculated based on the proportion of PWSs changing disinfection status, which is an output from
the predictive occurrence model, as described in Chapter 5 of this EA.
Exhibit 4.7 Baseline Population Served by SW Systems
PWS Size
Population Served bySW PWSs
CWS
NTNCWS
TNCWS
<100
56,740
13,297
57,454
101-500
608,084
68,083
116,529
501-1,000
885,400
67,958
70,147
1,001-4,100
6,628,597
140,168
143,347
4,101-33,000
38,700,554
174,408
147,423
33,001-96,000
39,034,554
121,446
-
96,001-500,000
58,489,936
203,000
-
500,001-1 Million
22,327,506
-
-
> 1 Million
37,363,275
-
2,000,000
Total
204,094,646
788,360
2,534,900
Source: Data extracted from SDWIS/FED PWS Inventory, 2007 4th Quarter Data.
Population figures for SW systems include GWUDI systems.
Number of Households Served
Because PWS costs are often passed onto customers in the form of water rate increases,
the proposed RTCR EA also includes analyses to assess the impact of the rule provisions at a
household level. The number of households served by CWSs expected to be subject to the
proposed RTCR is estimated by dividing the population for each PWS size category (Exhibits
4.6 and 4.7 above) by the average number of people per household, which was estimated as 2.56
for the year 2007 (U.S. Census Bureau Annual Social and Economic Supplement, 2008). Exhibit
4.8 below shows the number of households served in groundwater and surface water systems,
respectively.
Economic Analysis for the Proposed RTCR
4-17
June 2010
-------
Exhibit 4.8 Household Baseline
PWS Size
Number of Households
GW CWSs
SW CWSs
Disinfecting
Non-Disinfecting
Disinfecting
<100
151,390
133,689
22,164
101-500
927,932
424,045
237,533
501-1,000
1,004,165
274,829
345,859
1,001-4,100
4,417,086
810,211
2,589,296
4,101-33,000
11,463,304
1,416,939
15,117,404
33,001-96,000
6,088,745
596,821
15,247,873
96,001-500,000
3,881,055
41,923
22,847,631
500,001-1 Million
1,043,297
-
8,721,682
> 1 Million
1,714,823
-
14,595,029
Total
30,691,796
3,698,456
79,724,471
Sources: U.S. Census Bureau, Current Population Survey, 2008 Annual Social and Economic
Supplement; SDWIS/FED PWS Inventory, 2007 4th Quarter Data
4.3.4 Baseline Water Quality
The following sections provide an overview of summary statistics relating to baseline
water quality. The source data from which these summary statistics are derived form the basis of
further analysis in the proposed RTCR occurrence and risk assessment models as described in
later chapters of the EA.
4.3.4.1 Percent of TC and EC Positive Samples Based on 2005 Six-Year Review 2
Data
Exhibit 4.9 below shows the percent of TC+ and EC+, samples based on PWS type and
size. As described in Section 4.2.2.2, the 2005 Six-Year Review 2 data was cleaned using a
criterion that a given system-month of data should include a minimum of 50% of the expected
number of samples based on the system's population served and water system category.
The "TC+ samples" column was calculated by taking the total number of routine TC+
samples and dividing by the number of routine TC samples. For small PWSs, additional routine
TC samples in the month following a TC+ were included in the denominator. To calculate the
EC+ rate, the total number of EC+ samples was divided by the total number of TC- samples12
plus the number of TC+ samples that were tested for E. coli. This EC+ computation did not
include additional routine samples taken in the month following a TC+ because a significant
number of PWSs did not provide data on EC+ or EC- samples, and some systems tested for FC
and not E. coli.
12 EPA assumes that a PWS will not test for EC if the TC assay is negative.
Economic Analysis for the Proposed RTCR
4-18
June 2010
-------
Exhibit 4.9 Total Coliform and E. coli Percent Positive by System Size and Type
PWS
Source
Population
TC
TC
TC
EC
EC
EC
Type
Water
Served
(# Samples)
(+ Samples)
(% Positive)
(# Samples)1
(+ Samples)
(% Positive)2
<100
93,105
2,479
2.66%
1,172
72
0.08%
101 -500
125,490
2,500
1.99%
1,639
61
0.05%
501-1,000
48,265
736
1.52%
483
20
0.04%
GW
1,001-4,100
110,391
1,176
1.07%
732
21
0.02%
4,101-33,000
183,721
877
0.48%
458
22
0.01 %
33,001-100,000
96,361
214
0.22%
44
2
0.00%
>100,000
64,965
289
0.44%
34
1
0.00%
Total GW
722,298
8,271
1.15%
4,562
199
0.03%
cws
<100
6,735
95
1.41%
64
6
0.09%
101 -500
19,716
227
1.15%
159
10
0.05%
501-1,000
12,828
90
0.70%
70
7
0.05%
SW
1,001-4,100
55,310
314
0.57%
233
17
0.03%
4,101-33,000
175,758
525
0.30%
399
41
0.02%
33,001-100,000
112,894
157
0.14%
106
5
0.00%
>100,000
112,143
235
0.21%
99
2
0.00%
Total SW
495,384
1,643
0.33%
1,130
88
0.02%
GW&SW
Total CWS
1,217,682
9,914
0.81 %
5,692
287
0.02%
<100
163,730
7,820
4.78%
5,820
316
0.20%
101 -500
52,891
2,418
4.57%
1,869
99
0.19%
GW
501-1,000
6,952
299
4.30%
217
4
0.06%
>1,000
7,062
143
2.02%
85
2
0.03%
Total GW
230,635
10,680
4.63%
7,991
421
0.18%
TNCWS
<100
6,723
150
2.23%
141
17
0.25%
101 -500
2,854
75
2.63%
69
13
0.46%
SW
501-1,000
523
19
3.63%
19
-
0.00%
>1,000
988
6
0.61 %
37
-
0.00%
Total SW
11,088
250
2.25%
266
30
0.27%
GW&SW
Total TNCWS
241,723
10,930
4.52%
8,257
451
0.19%
<100
46,505
1,476
3.17%
1,061
34
0.07%
101 -500
33,084
893
2.70%
628
19
0.06%
GW
501-1,000
9,531
166
1.74%
103
2
0.02%
>1,000
13,138
177
1.35%
103
5
0.04%
Total GW
102,258
2,712
2.65%
1,895
60
0.06%
NTNCWS
<100
1,668
32
1.92%
30
4
0.24%
101 -500
2,304
9
0.39%
9
2
0.09%
SW
501-1,000
932
6
0.64%
5
-
0.00%
>1,000
1,316
1
0.08%
1
-
0.00%
Total SW
6,220
48
0.77%
45
6
0.10%
GW&SW
Total NTNCWS
108,478
2,760
2.54%
1,940
66
0.06%
Source: Derived using 2005 Six-Year Review 2 Data, which was filtered by including a month of data for a given
system (system-month) only if it represented at least 50% of the expected number of samples for a month based on
the system's size. The Total Coliform Rule Compliance Monitoring Data Quality and Completion Report (USEPA,
201 Oe) includes a detailed description of this data cleaning process.
Notes: 1 "#EC samples" is the denominator of the EC% positive calculation, and includes # TC- samples plus #TC+
samples that were tested for EC.2 Percent EC+ was calculated as (# EC+ samples)/("# EC samples")
Economic Analysis for the Proposed RTCR
4-19
June 2010
-------
4.3.4.2 Violation Levels Based on SDWIS/FED Data
Non-acute violations are triggered under the current TCR by water quality violations.
They are defined as >1 routine or repeat monitoring sample testing positive for TC in a given
compliance period (month, quarter, or year) for PWSs serving fewer than 33,001 people, or by
>5.0 percent of samples being positive for systems taking >40 samples per month (typically
those systems serving greater than 33,000 people). Acute violations are triggered when any PWS
has a repeat sample with an EC+ or FC+ or has an FC+ or EC+ routine sample followed by a
TC+ repeat sample in a given month.
Exhibit 4.10 below presents an assessment of water quality prior to the proposed RTCR
promulgation in terms of the number of acute and non-acute violations incurred by PWSs. The
number of violations from this data are directly input into the cost model for PWSs serving more
than 4,100 people to derive estimates of the number of assessments and corrective actions that
will be undertaken under the proposed RTCR. As noted in Exhibit 4.10, the data used is from
2005 3rd quarter. Exhibit 4.11 presents the seven years of PWS violation data13 evaluated by the
TWG. Appendix G presents further detail on this assessment, including queries used in
SDWIS/FED and the resulting datasets downloaded.
In addition to the acute and non-acute system violation data, Exhibit 4.11 also presents
data on the numbers of PWSs with monitoring and reporting violations (minor and major for
routine and repeat monitoring and reporting). Although PWS monitoring and reporting violation
data is not used directly in the quantitative anlyses performed for the EA, the TCRDSAC did
consider the data in its deliberations on the impact of the regulation. In particular, the TCRDSAC
determined that the proposed revisions to the RTCR may significantly reduce the high numbers
of monitoring and reporting violations by reducing the numbers of additional routine and repeat
samples required and providing more flexibility in development of sampling site plans.
The data verification process was used to review individual states' data in comparison to
SDWIS/FED data and to compare violations rates across states. This process revealed many
differences between states in their implementation of the TCR, and allowed EPA to identify and
review outliers in the national database. Specifically, EPA did not include violation data from
Ohio, U.S. territories, or tribal PWSs in the summaries presented in Exhibits 4.10 and 4.11.
Review of the data verification information revealed that Ohio's broad interpretation of what
constitutes a violation led to abnormally high violation counts compared (on average) to other
states. For U.S. territories and tribal systems, it was established that unique environmental
factors and operating conditions contribute to abnormally high TC+ results and associated
violations. Thus, inclusion of these results would skew the national averages used for analysis,
and were therefore excluded.
13 Exhibit 4.11 presents PWS counts (PWSs with at least one violation during the year) to be consistent with the
metric evaluated by the TCRDSAC TWG. Exhibit 4.10 presents total violation counts, which are used as inputs to
the predictive modeling for the EA.
Economic Analysis for the Proposed RTCR
4-20
June 2010
-------
Exhibit 4.10 Baseline Number of TCR Violations by System Size and Type (2005)
Num ber of Violations in GW
PWSs
Num ber of Violations in SW
PWSs
Total Num ber
of GW PWSs
(without OH &
PR)
Total Number
of SW PWSs
(without OH &
PR)
Non-Acute
Acute
Non-Acute
Acute
A
B
C
D
E
F
CWSs
<100
905
52
16
3
11,709
1,133
101-500
809
34
50
7
13,508
2,067
501-1,000
203
13
16
3
4,299
1,112
1,001-3,300
272
8
55
7
5,490
2,406
3,301-10,000
171
8
75
3
2,628
2,015
10,001-50,000
125
8
78
4
1,247
1,752
50,001-100,000
11
2
5
4
147
362
100,001-250,000
1
1
3
1
61
306
250,001-500,000
-
-
-
-
500,001-1 Million
-
-
1
-
> 1 Million
-
-
-
-
3
15
Total
2,497
126
299
32
39,092
11,168
NT NCWSs
<100
514
34
7
2
8,392
249
101-500
346
20
4
-
6,294
246
501-1,000
57
6
2
-
1,622
87
1,001-3,300
58
4
-
-
726
66
3,301-10,000
9
2
1
-
102
23
10,001-50,000
1
-
-
-
11
4
50,001-100,000
-
-
-
-
-
1
100,001-250,000
-
-
-
-
-
1
250,001-500,000
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
> 1 Million
-
-
-
-
-
-
Total
985
66
14
2
17,147
677
TNCWSs
<100
2,665
278
19
5
58,396
1,334
101-500
833
76
11
1
18,184
492
501-1,000
133
11
4
-
1,868
85
1,001-3,300
58
2
1
-
578
57
3,301-10,000
5
-
1
-
77
22
10,001-50,000
-
-
-
-
11
3
50,001-100,000
-
-
-
-
3
-
100,001-250,000
-
-
-
-
1
250,001-500,000
-
-
-
-
500,001-1 Million
-
-
-
-
> 1 Million
-
-
-
-
-
1
Total
3,694
367
36
6
79,118
1,994
Source: SDWIS/FED annual data for period ending 3 quarter 2001 - 2007 . OH, U.S. territories
excluded. Data was not available for 4th quarter (it is posted in the 3rd quarter annually).
Economic Analysis for the Proposed RTCR
4-21
June 2010
-------
Exhibit 4.11 Number of PWSs with Violations by System Type (2001 - 2007)
PWS Type
Year
2001
2002
2003
2004
2005
2006
2007
Acute MCL Violations
cws
143
144
185
171
151
171
171
NTNCWS
51
53
70
58
65
68
45
TNCWS
261
278
322
351
349
361
295
All
455
475
577
580
565
600
511
Non-Acute MCL Violations
CWS
2,074
2,110
2,204
2,314
2,196
2,095
1,996
NTNCWS
601
679
725
750
753
735
655
TNCWS
2,707
2,934
3,036
3,132
3,039
3,244
3,209
All
5,382
5,723
5,965
6,196
5,988
6,074
5,860
Major Monitoring and Reporting Violations (Routine and Repeat)
CWS
3,312
3,327
3,900
3,924
3,760
3,659
3,415
NTNCWS
1,503
1,561
1,676
1,679
1,588
1,468
1,377
TNCWS
10,360
10,531
11,230
11,043
10,426
10,630
10,389
All
15,175
15,419
16,806
16,646
15,774
15,757
15,181
Minor Monitoring and Reporting Violations (Routine and Repeat)
CWS
1,218
1,257
1,389
1,445
1,302
1,285
1,244
NTNCWS
187
182
200
233
194
214
183
TNCWS
701
842
883
903
875
843
839
All
2,106
2,281
2,472
2,581
2,371
2,342
2,266
Note: PWSs counts are of systems that had at least one violation during the year.
Source: SDWIS/FED annual data for period ending 3rd quarter 2001 - 2007. OH, U.S. territories, tribal PWS data
excluded. Data was not available for 4th quarter (it is posted in the 3rd quarter annually).
4.4 Sensitive Sub-populations
Under the Safe Drinking Water Act (SDWA) Amendments of 1996 (PL 104-182), EPA
must analyze health impacts of rulemaking on sensitive subpopulations. Sensitive populations
include "infants, children, pregnant women, the elderly, individuals with a history of serious
illness, and other subpopulations that are identified as likely to be at a greater risk of adverse
health effects due to exposure to contaminants in drinking water than the general population."
(USEPA, 1996a)
Exhibit 4.12 presents the EPA's estimates of the number of U.S. individuals who are at
increased risk of developing more severe symptoms from illnesses caused by waterborne
pathogens. Persons suffering from certain diseases and/or conditions are believed to be sensitive
to microbes and chemicals in drinking water. These subgroups of the population include
pregnant women, the very young, the elderly, and the immunocompromised. In total, these
subgroups represent approximately 20 percent of the current population of the United States.
Economic Analysis for the Proposed RTCR
4-22
June 2010
-------
Exhibit 4.12 Estimates of Sensitive Subpopulations in the United States
Sensitive
Population
Individuals
Approximate
Percent of
U.S.
Population1
Citation/Notes
Pregnant women and neonates
Pregnant Women
6,240,000
2.2
Vital and Health Statistics, CDC (Ventura et
al., 2000)
Neonates (under one
month)2
317,137
0.1
U.S. 2000 Census (US Census Bureau,
2001a)
Age-based sensitive populations
Children (< 5 years
old)
19,175,798
6.8
U.S. 2000 Census (US Census Bureau,
2001a)
Elderly (> 65 years
old)
34,991,753
12.4
U.S. 2000 Census (US Census Bureau,
2001a)
Compromised immune status
Bone marrow
transplant Recipients
20,000
0.01
National Marrow Donor Program
http://. ma rrow.org/MEDIA/facts_figures.pdf
AIDS Patients
816,149
0.3
HIV/AIDS Surveillance Report, cases
through 2001 (CDC, 2002)
Organ transplant
recipients
23,143
0.01
U.S. Census Bureau, Statistical Abstract of
the U.S., based on 1998 data (2001b)
Total
61,583,980
21.8
Based on U.S. Census estimate (July 2000).
21/12 of the 2000 census population for age <1 year.
4.5 Summary of Baseline Assumptions
In Section 4.3 (Data Sources), EPA discusses the representativeness and quality of the data
used from the 2005 Six-Year Review 2 data and SDWIS/FED. An additional source of
uncertainty is introduced into the baseline in the monitoring frequencies estimated for systems,
shown in Exhibit 4.4. As described in that section, to arrive at this distribution EPA and the
Technical Workgroup conducted an analysis of 2005 Six-Year Data and individual state statutes,
and had the results validated by ASDWA during the TCRDSAC deliberations.
Exhibit 4.13 below presents a summary of the assumptions made in developing the RTCR
EA baseline. These assumptions may introduce uncertainty into the predictive model and its
outcomes in terms of benefits and costs. However, all of the assumptions influencing the EA
baseline (current TCR) do so in the same way as for the proposed RTCR and Alternative option.
Therefore, EPA does not expect the net results of the analyses presented in this EA to be
significantly influenced by the uncertainty in the assumptions applied in developing the RTCR
EA baseline.
Economic Analysis for the Proposed RTCR
4-23
June 2010
-------
Exhibit 4.13 Summary of Baseline Assumptions Influencing RTCR Estimates
Uncertainty Factor
Current Assumption
Section with
Full
Discussion
of
Uncertainty
or Factor
Contributes
Variability,
Uncertainty,
or Constant?
Most Likely Effect on
Benefits (Note 1)
Most Likely Effect on Costs
(Note 1)
Under-
estimate
Over-
estimate
Unknown
Impact
Under-
estimate
Oxer-
estimate
Unknown
Impact
Pre-GWR baseline distribution of
systems across sampling
frequency categories (M/Q/A)
Varies per system category as
shown in Ex. 4.4
4.3.1
Variability,
Uncertainty
X
X
Selection of 2005 data from Six
Year Review for use as baseline
Data are adequately
representative of U.S. PWSs and
of sufficient quality. (Note 2)
4.2.2
Variability,
Uncertainty
X
X
Use of SDWIS/FED for inventory
data (2007 Q3) and violations
data (2001 - 2007 Q3)
Data are adequately
representative of U.S. PWSs and
of sufficient quality.
4.2.1
Variability,
Uncertainty
X
X
Distribution of systems across
sampling frequency categories
(M/Q/A) following
implementation of GWR sanitary
surveys prior to RTCR
implemlentation
Varies per system category as
shown in Ex. 5.14a - 5.14c
5.3.2.2
Variability,
Uncertainty
X
X
Pwell - the portion of GWSs
having viral pathogens in their
source waters (adopted from
GWR EA)
21.58%
5.3.1
Constant,
Uncertainty
X
X
Psample - the probability that a
random sample will test positive
for viral pathogens gi\en a
contaminated source water
(adopted from GWR EA)
Variable drawn from a beta
distribution with a range of alpha
and beta estimates having a
median value of 5.8 percent and
an expected \alue of 12.4
percent.
5.3.1
Variability,
Uncertainty
X
X
Reduced occurrence for sanitary
surveys performed under the
GWR
90% of baseline occurrence;
applied to equal number of
systems annually over the
sanitary survey cycle (20% each
year of 5 years for CWSs; 33.3%
each year for 3 years for
NCWSs)
5.3.1
Constant,
Uncertainty
X
X
Reduced occurrence for GWR
compliance monitoring (applies
to subset of GWSs that
disinfect).
90% of baseline occurrence
5.3.1
Constant,
Uncertainty
X
X
% of non-disinfecting GWSs
choosing disinfection CA \s non-
disinfection CA in response to
source water quality issue
(adopted from GWR)
Range: High end is the
percentage of CWS entry points
employing disinfection by
system size, low end assumed
to be 10% based on discussions
with State representatives.
5.3.1
Constant,
Uncertainty
X
X
Population and sensitive sub
population estimates
Varies by category
4.2.3.3 and
4.4
Variability,
Uncertainty
X
X
Notes: 1) All baseline variables or factors were incorporated into the predictive model in the same way for the Current
TCR (baseline option) and the other two regulatory options considered in this EA (AIP option and Alternative option);
therefore, EPA expects that any under- or over-estimation would affect the baseline and other options similarly,
resulting in no signficant net effect on the results of the analysis. 2) A complete discussion of the data cleaning and
assumption used in preparing 6 Year Review data for use in the RTCR baseline is provided in the Total Coliform Rule
Compliance Monitoring Data Quality and Completion Report (USEPA, 201 Oe).
Economic Analysis for the Proposed RTCR
4-24
June 2010
-------
5. Occurrence and Predictive Model
5.1 Introduction
Based on evaluation of the Safe Drinking Water Information System-Federal Version
(SDWIS/FED) violation data, the U.S. Environmental Protection Agency (EPA) assumes that the
national occurrence of total coliform (TC) and E. coli (EC) has reached a steady state in recent
years under the current Total Coliform Rule (TCR). Cycles of normal deterioration and
repair/replacement appear to be occurring at the individual system level, while the numbers of
violations at the national level remain relatively unchanged. Chapter 4 (Exhibit 4.11) presents the
number of Public Water Systems (PWSs) with TCR violations over the last several years, which
shows that national violation rates have remained relatively steady. Revisions to the TCR would
affect this steady state, likely resulting in a reduction of the underlying occurrence and associated
violations. However, prior to the proposed Revised Total Coliform Rule (RTCR)
implementation, the Ground Water Rule (GWR), which became effective December 2009, would
also have an effect on the steady state. This chapter explains the development of a model that
both reproduces the steady-state occurrence conditions under the current TCR and predicts the
effects of both the proposed RTCR and the GWR in further reducing occurrence from current
estimated levels.
The occurrence and predictive model was used for PWSs serving 4,100 people or fewer
and it has two components. Occurrence and predictive modeling focuses on small PWSs (those
serving <4,100 people) because (a) there are adequate data to support development of the
occurrence distributions that are needed for the predictive components of the model, (b) these
systems are known to have higher occurrence levels than larger systems, and (c) these systems,
in particular, are subject to changes in the monitoring requirements under the proposed RTCR.
The first component of the model characterizes how the presence or positive rates of TC and EC
detections vary across the population of small (serving 4,100 people or fewer) public water
systems in the U.S. These rates vary by the type of sample (routine or repeat), by analyte (TC or
EC), and by system type (Community Water System (CWS), Noncommunity Water System
NCWS, or Transient Noncommunity Water Systems (TNCWS)) and size. EPA determined from
the Six-Year Review data1 that systems differ considerably with respect to the observed
occurrence of TC and EC in both routine and repeat sampling. Some of this variability can be
explained by the type, size, water source, and disinfection practices of the systems. However,
even among systems having the same characteristics, some systems were found to rarely
experience positive assays, while others often did. The second component of the model uses the
TC and EC occurrence distributions, together with the various sampling and response
requirements, to simulate a set of nationally-representative systems within the context of the
three regulatory options (current TCR, Agreement in Principle or AIP, and Alternative) to
predict changes in TC and EC occurrence, triggers, assessments, corrective actions over time,
and violations. The model generates estimates of reduced TC and EC occurrence based on
requirements of the proposed RTCR to perform assessments and corrective actions not explicitly
required in the current TCR. (In addition, the models take into account reductions attributable to
the implementation of the GWR, the effective date of which is December 2009.)
1 A discussion of the use of data from the Six-Year Review of SDWA regulations is presented in Chapter 4 of this
EA.
Economic Analysis for the Proposed RTCR 5-1
June 2010
-------
The occurrence model produces outputs over a 30-year time horizon2 that include
national estimates of TC and EC occurrence in PWSs across the U.S., and the resulting Level 1
and Level 2 assessments3 and corrective actions performed. These outputs are used to
characterize reduced exposure to the potential contamination under each regulatory scenario
considered, as presented in Chapter 6 of this EA, and to estimate incremental costs for PWSs
across the U.S., as described in Chapter 7 of this EA.
The two major components of the occurrence and predictive model used for systems
serving <4,100 people are described in Sections 5.2 and 5.3 of this chapter.
For systems serving more than 4,100 people, EPA estimated violation and trigger rates
using SDWIS/FED (USEPA, 2005a) because the Six-Year Review data for PWSs serving more
than 4,100 people were not as robust as the Six-Year Review data for systems serving 4,100 or
fewer people. EPA did not quantify changes in violation or trigger rates for systems serving more
than 4,100 people among the current TCR, AIP option, and Alternative option because of: (1)
limited Six-Year Review data to characterize these systems, (2) the essentially unchanged
monitoring requirements across options for these systems, and (3) the level of effort already
occurring to implement the current TCR.
A complete description of the methods for developing the occurrence estimates as
described above is found in the remaining sections of this chapter:
Section 5.2 Describes development of parameter estimates a, /? based on empirical
occurrence data for use in the model developed for the proposed
RTCR.
Section 5.3 Describes the occurrence model developed for systems serving <4,100
people (Sections 5.3.1-5.3.2), predictions from the model (5.3.3) and
uncertainty in the model (5.3.3.1), including a comprehensive list of
assumptions contributing to uncertainty, a sensitivity analysis, and a
discussion of model validation.
Section 5.4 Describes how outcomes were predicted for systems serving more than
4,100 people.
2 The model includes an additional 5 years prior to the commencement of AIP option monitoring to account for
effects of the GWR after it is fully implemented, plus 25 years of RTCR analysis, for a total of 30 modeled years.
3
The AIP and Alternative option require PWSs to complete either a Level 1 or Level 2 assessment following
triggers specified in Exhibit 3.1 of this EA. The purpose of Level 1 and 2 assessments is to identify the presence of
sanitary defects and deficiencies in distribution system coliform monitoring practices, similar to an annual
inspection but more focused on determining the cause of a contamination event. In order to complete a Level 2
assessment the PWS must perform a more detailed examination of the system than for a Level 1 assessment,
including its monitoring and operational practices.
Economic Analysis for the Proposed RTCR 5-2
June 2010
-------
5.2 Modeling of Current Total Coliform and E. coli Occurrence for Systems
Serving up to 4,100 People
This section describes the development of parameters used to define variability across
and within groups of similar PWSs with regard to occurrence of TC and EC. Details related to
this section are provided in Appendix F to this EA.
5.2.1 Distributional Model and Notation
The Six-Year Review data demonstrate that, even among systems of similar size and
type, there are some PWSs that rarely find routine samples to be TC-positive and others that
often find TC-positive samples. For example, among the very small (serving <101) TNCWSs
using nondisinfected ground water, 47 systems reported ten or more routine TC positives while
assaying 40 or fewer samples during 2005. Among the same set of small TNCWSs with
nondisinfected ground water are 60 that assayed 20 or more routine samples, finding all of them
TC-negative. Clearly, routine TC-positive rates vary considerably within this set of systems.
Significantly different fractions positive can be found among any basic subsets of systems
(defined by system type, water type, and population served, as described in Section 5.2.2 of this
EA) and for any type of assay (both routine and repeat TC and both routine and repeat E. coli).
Within a group of similar systems, pRTTC;, which is defined as the probability that a
routine sample taken from system i will test positive, is assumed to vary as a beta random
variable. The beta distribution is commonly used to model varying probabilities because it is
limited to values between 0 and 1 and can assume a wide variety of shapes. EPA used a beta
distribution to describe how probabilities of virus-positive samples vary among virus-
contaminated wells in the Economic Analysis for the Ground Water Rule (USEPA 2006a).
Samples that test negative for TC are generally not tested further for the presence of EC,
as are TC-positive samples. The probability that a TC-positive routine (RT) sample taken from
system i will also test EC-positive, pRTEQ, is assumed to vary among similar systems as
another beta random variable, but with parameters that can be quite different from those of the
routine TC samples. Although systems with frequent routine TC positives necessarily generated
more EC assays, the EC positive rates for these systems did not appear to be significantly
different than the EC positive rates of systems with rare TC positives. The variables pRTTC and
pRTEC are therefore modeled as independent beta-distributions.
Similar assumptions are extended to the repeat positive probabilities, pRPTC, and
pRPEC;. Like pRTTC, and pRTEQ, these are assumed to be independently beta-distributed
among sets of similar systems. A process for determining "sets of similar systems" is described
in Sections 5.2.3 and 5.2.4. The process is informed by system-specific data that are reduced
from the sample-specific data described in Chapter 4 of this EA, as discussed in the following
section.
Economic Analysis for the Proposed RTCR 5-3
June 2010
-------
5.2.2 Data Reduction
The occurrence model assumes that positive routine TC measurements for system i result
from some unobserved probability (pRTTC,). This probability may vary from month-to-month
and season-to-season. However, in this model each result, whether positive or negative, is
assumed to be an independent trial with positive probability pRTTC;. Each of these is called a
Bernoulli trial. The total number positive observed for system i during a year (KRTTC,) is
therefore a binomial random variable, whose parameters are the system's positive probability
(pRTTC;) and the total number of samples assayed (NRTTC,). What is observed for the system
are the number of positive samples (KRTTC;) and the number of samples assayed (NRTTC;).
The order of the positives and negatives is unimportant, so the routine TC data needed for any
system can be reduced to these two values: NRTTC; and KRTTC;. The leading character "N"
designates the number of samples tested and "K" designates the number positive.
Similarly, the routine EC data for system i can be reduced to the number of routine TC-
positive samples that were assayed for EC (NRTEC;) and, of those, the number testing positive
(KRTEC;). For many systems, both NRTEC and KRTEC will be zero because they encountered
no routine TC positives during 2005 (KRTTC = 0), and EPA assumes that in the case of a TC-
negative, no EC assay is performed.
Finally, the data for repeat samples for system i can be reduced to four numbers: NRPTC,
KRPTC;, NRPEC;, and KRPEC;. Again, the leading characters, "N" and "K", designate numbers
of samples assayed and numbers positive, respectively. "RP" designates that these are repeat
samples, while "TC" and "EC" designate assays for TC and EC, respectively. For many systems,
all four of these numbers are zero because no routine samples were TC positive.
It is important to note that only TC-positive samples are assayed for EC. The positive
probabilities for EC are therefore conditional, that is, they apply only to samples that have tested
TC positive. All TC-negative samples are assumed to be EC negative. The overall
(unconditional) probability that a routine sample from system i will be EC positive is the product
pRTTC;*pRTEC; and the overall (unconditional) probability that a repeat sample from the
system will be EC-positive is the product pRPTC;*pRPEC;.
The reduced dataset includes eight integer values (for NRTTC, KRTTC, NRTEC,
KRTEC, NRPTC, KRPTC, NRPEC, KRPEC) for each of approximately 93,000 systems.
Although none of the NRTTC are zero (all systems assayed at least one routine TC sample
during 2005), the great majority of the other values are zeros.
5.2.3 Basic Subsets of Systems
The 92,747 systems serving <4,100 people in the Six-Year Review dataset (which
represent the approximately 147,000 total systems that exist in this size range) can be classified
by system type, source water type and disinfection status, and population served. The three types
of systems are CWS, TNCWS and nontransient noncommunity (NTNCWS). The three source
water types are surface water (SW), which is always disinfected, nondisinfected ground water
(NondisGW) and disinfected ground water (DisGW). Three important size categories are small
systems (serving <100 people), systems serving 101 to 1,000 people, and systems serving 1,001
to 4,100 people. Systems serving more than 4,100 people are not included here, for reasons
Economic Analysis for the Proposed RTCR 5-4
June 2010
-------
discussed in Chapter 4 of this EA and earlier in this chapter. In total, there are 27 possible
categories that each PWS may belong to (3 system types x 3 source water types x 3 population
served sizes). Each PWS may only be in one of the 27 categories. Exhibit 5.1 lists the 27 basic
subsets and corresponding numbers of systems in the dataset.
Exhibit 5.1 Basic Classifications of PWSs Used for Occurrence Modeling
Subset
System Type
Water Type
Population
Served Group
Number of
Systems in 6-
Year Review
Data
1
TNCWS
SW
<100
838
2
TNCWS
sw
101-1000
358
3
TNCWS
SW
1001-4100
30
4
NTNCWS
sw
<100
114
5
NTNCWS
sw
101-1000
172
6
NTNCWS
sw
1001-4100
33
7
CWS
sw
<100
484
8
CWS
sw
101-1000
2,034
9
CWS
sw
1001-4100
1,699
10
TNCWS
DisGW
<100
8,112
11
TNCWS
DisGW
101-1000
3,007
12
TNCWS
DisGW
1001-4100
119
13
NTNCWS
DisGW
<100
1,577
14
NTNCWS
DisGW
101-1000
1,780
15
NTNCWS
DisGW
1001-4100
252
16
CWS
DisGW
<100
3,662
17
CWS
DisGW
101-1000
7,678
18
CWS
DisGW
1001-4100
2,707
19
TNCWS
NondisGW
<100
32,028
20
TNCWS
NondisGW
101-1000
10,217
21
TNCWS
NondisGW
1001-4100
217
22
NTNCWS
NondisGW
<100
4,026
23
NTNCWS
NondisGW
101-1000
3,264
24
NTNCWS
NondisGW
1001-4100
152
25
CWS
NondisGW
<100
3,788
26
CWS
NondisGW
101-1000
3,755
27
CWS
NondisGW
1001-4100
644
TOTAL
92,747
Note: Water Type includes disinfection status, as informed by information from SDWIS-FED. Systems with no
indication of disinfection status in SDWIS-FED were assumed to not disinfect (Nondis).
Economic Analysis for the Proposed RTCR 5-5
June 2010
-------
5.2.4 Estimation Methodology
The methodology described in this section applies to all four kinds of measurements,
routine TC (RTTC), routine E. coli (RTEC), repeat TC (RPTC) and repeat coli (RPEC).
However, to simplify the presentation, the description will be expressed only in terms of RTTC.
Results presented in Section 5.2.5 show that the final system classifications used to model TC
(both routine and repeat) are quite different from the final subsets used to model EC occurrence.
For any individual subset of systems from Exhibit 5.1 or grouping of these subsets, two
methods were used to estimate parameters of the beta-distributed RTTC: maximum likelihood
estimation and Bayesian Markov Chain Monte Carlo (MCMC) sampling. The MCMC samples
are used to test whether two basic subsets have equal average positive rates and to check the
maximum likelihood estimates (MLEs). This check was conducted to ensure that the algorithm
used to find the MLE had not stopped too soon.
Both the MLE and MCMC methodologies require computation of the likelihood function,
L(data | oc,(3), where a and (3 are parameters of beta-distributed pRTTC. For a specific system in
a set of similar systems, the number of positive RTTC samples is a binomial random variable
with parameters NRTTC and pRTTC, which are the numbers of RTTC samples assayed by the
system and the system's unobserved probability of a positive, respectively. The probability of a
positive, pRTTC, is a beta random variable, whose parameters are a and [i. Exhibit 5.2 is a
directed graph showing the model structure. At the top are the two high-level parameters (a and
P) and arrows from them showing their influence on the distribution of the system-specific
parameters, the positive probabilities or pRTTCs. Finally, the exhibit shows how the pRTTCs,
together with the numbers of samples assayed (NRTTCs), influenced the numbers estimated to
be positive (KRTTCs).
Economic Analysis for the Proposed RTCR 5-6
June 2010
-------
Exhibit 5.2 Directed Graph of Model Used for RTTC Occurrence
^pRTTC^
<
^"pRTTC^
• * • (
%RTTC^
*
1
1
^
KRTTC,
KRTTC,
• • •
KRTTCn
,
"
s,
NRTTC,
NRTTC,
• • •
NRTTCn
This same model could have been equally well represented with all of the arrows pointing up to
show the direction of inference, starting with the data (NRTTCs and KRTTCs), which inform
estimates of the unobserved positive probabilities (pRTTCs), and then those informing estimates
of the high level parameters (a and P).
5.2.4.1 Maximum Likelihood Estimation
NOTE: From this point through the end of this chapter, the approach for estimating
parameters for routine total coliform occurrence (RTTC) is the same approach used for repeat
total coliform (RPTC) and both routine and repeat E. coli occurrence (RTEC and RPEC). To
simplify the notation, the designation of measurement type (e.g., RTTC) will be dropped. For
example, KRTTC is reduced to K, NRTTC is reduced to N, and pRTTC is reduced to p.
Subscript i is used to denote a particular system in a basic subset of systems.
For a single system in a subset, if p; were known, the likelihood of the system's RTTC data
(K; positives observed among N, routine samples) would be given by the binomial probability
mass function:
dbinon^Kj, N;, p;) =
N: !
-K;)
;Pi
'(1-pJ
.Ni-K;
But pi is not known, nor can it be observed explicitly. It is, therefore, treated as a beta
random variable. The expected value of the likelihood function is obtained by integrating the
Economic Analysis for the Proposed RTCR
5-7
June 2010
-------
product of the beta probability density function and the binomial probability mass function. The
beta probability density function is:
r("p> .(pf-'ii-pf-1
r(a)-r(P) v M v M
And the expected likelihood is:
-1
Ni! l"(a + P) a-l+K; S-l+Ni-K;
•(P) (1 " P) dp
K.^N.-K.)! r(a)T(P)
'0
Because both N; and K, are constant (observed data), the function only depends on the
factors involving p, a, and (3, so the factorial terms in the above may be ignored when seeking to
maximize the likelihood function. This simplifies the computation considerably, and the
resulting integral has a closed-form solution, which is:
l~(a + P>r(a + K.j-r(p + N. - K.j Beta^a + K., p + N. - K.j
r(a)-r(p)-r(a + P + N.) Beta(a,P)
The total likelihood for a set of systems is the product of the system-specific likelihoods:
Beta|a + K.,(3 + N. - K.j
Likelihood(a,(3) = | j -
Beta(a + (3)
In order to find maximum likelihood parameter estimates of parameters a, and (3, some
additional steps were taken:
1. Log likelihood, rather than likelihood, was computed to avoid the problems of
numerical overflow and underflow. Because the log transformation is positive and
monotonic, parameters found to maximize the log likelihood function will also
maximize the untransformed likelihood function:
f Beta|a + K.,(3 + N. - K.j ^
LogLikelihood(a,B) = ^In ,
^ ^ Beta(a + (3) J
2. Parameters of the Beta distribution were expressed in terms of the mean (a=a/(a+|3))
and dispersion (b=l/sqrt(a+|3)). In terms of these new parameters, the log likelihood
function is:
LogLikelihood(a,b) =
In
1
( Betaf— + K.,-—- + N. - K.^
b2 ' b2 ' 'j
V
Beta
f
a
1 ~a^
V
b2 + b2 j j
3. Bayesian MCMC estimation, as described in Appendix F was used to check the
MLEs and to support decisions about combining (pooling) different categories or
subsets of systems.
4. Where subsets of systems were found to have insufficient data to support maximum
likelihood estimation, engineering judgment was used to decide if combining with
Economic Analysis for the Proposed RTCR 5-8
June 2010
-------
other subsets made sense, for the purpose of parameter estimation. Rationale and
strategies for this are described in Appendix F.
5.2.5 Results
Appendix F of this EA describes how systems and their data were combined, or pooled,
for the purpose of RTTC, RTEC, RPTC, and RPEC modeling. Section 5.2.5.1 and 5.2.5.2 report
the resulting MLEs in terms of a, b and a, /?.
RTTC data were addressed first because a) so much more data were available for RTTC
than for the other three measurement types (RTEC, RPTC, and RPEC); and b) summary statistics
suggested that systematic differences due to system size and other factors were larger for TC
than for EC measurements.
When evaluating data for any one type of measurement, system size was evaluated first
because it was expected to be important for a number of reasons, primarily: a) Larger systems
have more resources and are more likely to employ full-time professional operators than smaller
systems; and b) Having more resources, the larger systems may provide better treatment and
distribution systems. System type and water type were evaluated after system size, but the order
in which these were evaluated does not appear to have been important.
5.2.5.1 Maximum Likelihood Estimates
Exhibit 5.3 provides MLEs for routine and repeat TC, while Exhibit 5.4 provides MLEs
for routine and repeat EC. Note that both of these exhibits present the a and b parameters as
defined in Section 5.2.4.
Economic Analysis for the Proposed RTCR 5-9
June 2010
-------
Exhibit 5.3 Maximum Likelihood a and b Parameter Estimates for RTTC and
RPTC
Disinf?
GW or
SW?
Sys Type
Size
No. Systems
a RTTC
bRTTC
a RPTC
bRPTC
Y
sw
TNCWS
<100
838
0.02312
0.2992
0.1439
1.108
Y
SW
TNCWS
101-1000
358
0.02312
0.2992
0.1439
1.108
Y
sw
TNCWS
1001-4100
30
0.02312
0.2992
0.1439
1.108
Y
sw
NTNCWS
<100
114
0.01584
0.3576
0.1064
1.078
Y
sw
NTNCWS
101-1000
172
0.01127
0.2458
0.1064
1.078
Y
sw
NTNCWS
1001-4100
33
0.01127
0.2458
0.1064
1.078
Y
sw
CWS
<100
484
0.01502
0.2090
0.06464
0.6107
Y
sw
CWS
101-1000
2034
0.009458
0.1490
0.06464
0.6107
Y
sw
CWS
1001-4100
1699
0.005864
0.1275
0.03430
0.2721
Y
GW
TNCWS
<100
8112
0.02312
0.2992
0.1439
1.108
Y
GW
TNCWS
101-1000
3007
0.02312
0.2992
0.1439
1.108
Y
GW
TNCWS
1001-4100
119
0.02312
0.2992
0.1439
1.108
Y
GW
NTNCWS
<100
1577
0.01584
0.3576
0.1064
1.078
Y
GW
NTNCWS
101-1000
1780
0.01127
0.2458
0.1064
1.078
Y
GW
NTNCWS
1001-4100
252
0.01127
0.2458
0.1064
1.078
Y
GW
CWS
<100
3662
0.01592
0.2243
0.09415
0.8493
Y
GW
CWS
101-1000
7678
0.01243
0.2142
0.09415
0.8493
Y
GW
CWS
1001-4100
2707
0.007767
0.1841
0.05221
0.4916
N
GW
TNCWS
<100
32028
0.04758
0.3990
0.2788
1.068
N
GW
TNCWS
101-1000
10217
0.04758
0.3990
0.2487
0.9359
N
GW
TNCWS
1001-4100
217
0.02467
0.3121
0.1694
0.4052
N
GW
NTNCWS
<100
4026
0.03658
0.4044
0.2575
0.9319
N
GW
NTNCWS
101-1000
3264
0.02690
0.2840
0.2575
0.9319
N
GW
NTNCWS
1001-4100
152
0.02690
0.2840
0.2575
0.9319
N
GW
CWS
<100
3788
0.03147
0.2680
0.1898
0.8407
N
GW
CWS
101-1000
3755
0.02690
0.2840
0.1898
0.8407
N
GW
CWS
1001-4100
644
0.02690
0.2840
0.1312
0.6938
Note: Disinfection status ("Disinf?") is from information downloaded from SDWIS/FED. Systems with no indication
of disinfection status in SDWIS/FED are assumed to not disinfect.
Economic Analysis for the Proposed RTCR 5-10
June 2010
-------
Exhibit 5.4 Maximum Likelihood a and b Parameter Estimates for RTEC and
RPEC
Disinf?
GW or
SW?
Sys Type
Size
No. Systems
a RTEC
bRTEC
aRPEC
bRPEC
Y
sw
TNCWS
<100
838
0.1291
0.6287
0.05175
1.153
Y
SW
TNCWS
101-1000
358
0.1291
0.6287
0.05175
1.153
Y
sw
TNCWS
1001-4100
30
0.1291
0.6287
0.05175
1.153
Y
sw
N TNCWS
<100
114
0.1291
0.6287
0.05175
1.153
Y
sw
N TNCWS
101-1000
172
0.1291
0.6287
0.05175
1.153
Y
sw
N TNCWS
1001-4100
33
0.1291
0.6287
0.05175
1.153
Y
sw
CWS
<100
484
0.1291
0.6287
0.05175
1.153
Y
sw
CWS
101-1000
2034
0.1291
0.6287
0.05175
1.153
Y
sw
CWS
1001-4100
1699
0.07514
0.8767
0.05175
1.153
Y
GW
TNCWS
<100
8112
0.08599
0.8796
0.07063
1.196
Y
GW
TNCWS
101-1000
3007
0.08599
0.8796
0.07063
1.196
Y
GW
TNCWS
1001-4100
119
0.08599
0.8796
0.07063
1.196
Y
GW
N TNCWS
<100
1577
0.08599
0.8796
0.07063
1.196
Y
GW
NTNCWS
101-1000
1780
0.08599
0.8796
0.07063
1.196
Y
GW
NTNCWS
1001-4100
252
0.08599
0.8796
0.07063
1.196
Y
GW
CWS
<100
3662
0.05449
0.6387
0.03950
1.168
Y
GW
CWS
101-1000
7678
0.05449
0.6387
0.03950
1.168
Y
GW
CWS
1001-4100
2707
0.05449
0.6387
0.03950
1.168
N
GW
TNCWS
<100
32028
0.04223
0.8129
0.0357
1.059
N
GW
TNCWS
101-1000
10217
0.04223
0.8129
0.03569
1.059
N
GW
TNCWS
1001-4100
217
0.04223
0.8129
0.03569
1.059
N
GW
NTNCWS
<100
4026
0.04223
0.8129
0.03569
1.059
N
GW
NTNCWS
101-1000
3264
0.04223
0.8129
0.03569
1.059
N
GW
NTNCWS
1001-4100
152
0.04223
0.8129
0.03569
1.059
N
GW
CWS
<100
3788
0.04223
0.8129
0.03569
1.059
N
GW
CWS
101-1000
3755
0.04223
0.8129
0.03569
1.059
N
GW
CWS
1001-4100
644
0.04223
0.8129
0.03569
1.059
Note: Disinfection status ("Disinf?") is from information downloaded from SDWIS/FED. Systems with no indication
of disinfection status in SDWIS/FED are assumed to not disinfect.
5.2.5.2 Derivation of a, ft from Maximum Likelihood Estimates
Using the estimates of a and b presented in Exhibits 5.3 and 5.4, EPA derived a, (5
estimates of the variation in PWSs within groups of systems, and incorporated these into the
occurrence model as described in Section 5.2.4.
These parameters are derived from a and b as follows:
a = a / b2
|3 = (1- a) / b2
Exhibits 5.5 and 5.6 present the estimates of a, (5 corresponding to estimates of a and b in
exhibits 5.3 and 5.4.
Economic Analysis for the Proposed RTCR 5-11
June 2010
-------
Exhibit 5.5 a and /? Parameter Estimates for RTTC and RPTC
Disinf?
GW or
SW?
Sys Type
Size
No. Systems
a RTTC
pRTTC
a RPTC
pRPTC
Y
sw
TNCWS
<100
838
0.2584
10.92
0.117
0.6972
Y
SW
TNCWS
101-1000
358
0.2584
10.92
0.117
0.6972
Y
sw
TNCWS
1001-4100
30
0.2584
10.92
0.117
0.6972
Y
sw
NTNCWS
<100
114
0.1239
7.697
0.092
0.7697
Y
sw
NTNCWS
101-1000
172
0.1866
16.37
0.092
0.7697
Y
sw
NTNCWS
1001-4100
33
0.1866
16.37
0.092
0.7697
Y
sw
CWS
<100
484
0.3438
22.54
0.173
2.508
Y
sw
CWS
101-1000
2034
0.4262
44.63
0.173
2.508
Y
sw
CWS
1001-4100
1699
0.3608
61.17
0.463
13.04
Y
GW
TNCWS
<100
8112
0.2584
10.92
0.1172
0.6972
Y
GW
TNCWS
101-1000
3007
0.2584
10.92
0.1172
0.6972
Y
GW
TNCWS
1001-4100
119
0.2584
10.92
0.1172
0.6972
Y
GW
NTNCWS
<100
1577
0.1239
7.697
0.09164
0.7697
Y
GW
NTNCWS
101-1000
1780
0.1866
16.37
0.09164
0.7697
Y
GW
NTNCWS
1001-4100
252
0.1866
16.37
0.09164
0.7697
Y
GW
CWS
<100
3662
0.3166
19.57
0.1305
1.256
Y
GW
CWS
101-1000
7678
0.2709
21.53
0.1305
1.256
Y
GW
CWS
1001-4100
2707
0.2291
29.26
0.2160
3.922
N
GW
TNCWS
<100
32028
0.2989
5.983
0.2444
0.6323
N
GW
TNCWS
101-1000
10217
0.2989
5.983
0.2839
0.8577
N
GW
TNCWS
1001-4100
217
0.2532
10.01
1.03175
5.059
N
GW
NTNCWS
<100
4026
0.2236
5.890
0.2965
0.8550
N
GW
NTNCWS
101-1000
3264
0.3336
12.07
0.2965
0.8550
N
GW
NTNCWS
1001-4100
152
0.3336
12.07
0.2965
0.8550
N
GW
CWS
<100
3788
0.4381
13.49
0.2685
1.146
N
GW
CWS
101-1000
3755
0.3336
12.07
0.2685
1.146
N
GW
CWS
1001-4100
644
0.3336
12.07
0.2726
1.805
Note: Disinfection status ("Disinf?") is from information downloaded from SDWIS/FED. Systems with no indication
of disinfection status in SDWIS/FED are assumed to not disinfect.
Economic Analysis for the Proposed RTCR 5-12
June 2010
-------
Exhibit 5.6 a and /? Parameter Estimates for RTEC and RPEC
Disinf?
GW or
SW?
Sys Type
Size
No. Systems
aRTEC
PRTEC
aRPEC
PRPEC
Y
sw
TNCWS
<100
838
0.3266
2.203
0.03891
0.7129
Y
SW
TNCWS
101-1000
358
0.3266
2.203
0.03891
0.7129
Y
sw
TNCWS
1001-4100
30
0.3266
2.203
0.03891
0.7129
Y
sw
N TNCWS
<100
114
0.3266
2.203
0.03891
0.7129
Y
sw
N TNCWS
101-1000
172
0.3266
2.203
0.03891
0.7129
Y
sw
N TNCWS
1001-4100
33
0.3266
2.203
0.03891
0.7129
Y
sw
CWS
<100
484
0.3266
2.203
0.03891
0.7129
Y
sw
CWS
101-1000
2034
0.3266
2.203
0.03891
0.7129
Y
sw
CWS
1001-4100
1699
0.0978
1.203
0.03891
0.7129
Y
GW
TNCWS
<100
8112
0.1112
1.181
0.04936
0.6496
Y
GW
TNCWS
101-1000
3007
0.1112
1.181
0.04936
0.6496
Y
GW
TNCWS
1001-4100
119
0.1112
1.181
0.04936
0.6496
Y
GW
N TNCWS
<100
1577
0.1112
1.181
0.04936
0.6496
Y
GW
NTNCWS
101-1000
1780
0.1112
1.181
0.04936
0.6496
Y
GW
NTNCWS
1001-4100
252
0.1112
1.181
0.04936
0.6496
Y
GW
CWS
<100
3662
0.1336
2.318
0.02896
0.7041
Y
GW
CWS
101-1000
7678
0.1336
2.318
0.02896
0.7041
Y
GW
CWS
1001-4100
2707
0.1336
2.318
0.02896
0.7041
N
GW
TNCWS
<100
32028
0.0639
1.449
0.03184
0.8602
N
GW
TNCWS
101-1000
10217
0.0639
1.449
0.03184
0.8602
N
GW
TNCWS
1001-4100
217
0.0639
1.449
0.03184
0.8602
N
GW
NTNCWS
<100
4026
0.0639
1.449
0.03184
0.8602
N
GW
NTNCWS
101-1000
3264
0.0639
1.449
0.03184
0.8602
N
GW
NTNCWS
1001-4100
152
0.0639
1.449
0.03184
0.8602
N
GW
CWS
<100
3788
0.0639
1.449
0.03184
0.8602
N
GW
CWS
101-1000
3755
0.0639
1.449
0.03184
0.8602
N
GW
CWS
1001-4100
644
0.0639
1.449
0.03184
0.8602
Note: Disinfection status ("Disinf?") is from information downloaded from SDWIS/FED. Systems with no indication
of disinfection status in SDWIS/FED are assumed to not disinfect.
5.3 Predictive Modeling of Occurrence for Systems Serving Up to 4,100 People
For the purpose of understanding both the potential to reduce exposure to potential
contamination and for estimating incremental costs that may be incurred under the proposed
RTCR, EPA developed a model to predict the changes in TC and EC occurrence over the
modeled period with respect to requirements of the proposed RTCR regulatory scenarios. This
predictive occurrence model takes into account the requirements of existing regulations (TCR
and GWR) and develops estimates of changes in occurrence (as well as the frequency of the new
Level 1 and Level 2 assessments and corrective actions) based on requirements of the regulatory
alternatives considered under the proposed RTCR. The resulting estimates are used to generate
cost estimates that can be compared across the three regulatory scenarios4 - the current TCR
4 The three regulatory scenarios considered in the RTCR EA (current TCR option, AIP option, and Alternative
option) are described in detail in Chapter 3 of this EA. Briefly, the AIP and Alternative options both differ from the
current TCR primarily in the requirements for Level 1 and Level 2 assessments and corrective actions and for
reduced monitoring eligibility. The AIP allows a continuation of current monitoring frequency while the Alternative
Economic Analysis for the Proposed RTCR 5-13
June 2010
-------
option, the AIP option, and the Alternative option, as described in Chapter 7 of this EA. Potential
benefits are compared in terms of changes in the level of occurrence associated with each
regulatory scenario, as presented in Chapter 6 of this EA.
The occurrence model focuses on a 30-year period beginning with full implementation of
the GWR and covering a 25-year period of analysis from the anticipated proposed RTCR
effective date of 2015. The predictive model is implemented as a Monte Carlo simulation and
currently uses 10,000 iterations (simulated systems) within each of the 27 type, source and size
categories to predict the effects of changes to the current rule over the modeling period.
The PWSs included in this model are those serving <4,100 people primarily because
these are the systems that will experience the major changes in monitoring frequency and other
requirements that will affect their TC and EC occurrence. For systems serving more than 4,100
people, which will not have significant monitoring frequency changes, a separate analysis was
performed based on historical violations rates instead of occurrence rates, as explained in Section
5.4 of this EA. The characteristics of all systems are described in Section 4.3 of this EA.
5.3.1 Summary of GWR factors and timing affecting the TCR and Proposed RTCR
The triggered and compliance monitoring requirements of the GWR must be
implemented by December 2009, which will be approximately five years prior to the anticipated
compliance date of the proposed RTCR. Sanitary survey (SS) requirements must be implemented
by states by December 2012 for most CWSs, and not until December 2014 for higher performing
CWSs and NCWSs. To account for the impact of the GWR requirements on baseline occurrence
for those public water systems using a ground water source (and serving <4,100 people), the
model performs a number of adjustments to the existing data prior to its application in the model
as a baseline.
Using 2005 SDWIS data (as explained in Section 4.3.1 of this EA) as a pre-GWR
inventory baseline, EPA applies probabilities of Psampie and Pweii5 to determine the number of
nondisinfecting systems that are expected to move to a disinfecting status in the five years prior
to proposed RTCR implementation. EPA derived Psampie and Pweii6 in the GWR EA (November
2006); PWeii is a fixed value of 21.58 percent, and Psampie is drawn from a beta distribution with a
range of alpha and beta estimates having a median value for each system of 5.8 percent and an
expected value of 12.4 percent. Incorporation of these factors into the model ensures that the
corrective actions anticipated to be performed under the GWR are reflected in reduced
option requires systems to sample monthly for an initial period. Both allow reduced monitoring for qualified
systems, but only the AIP allows annual monitoring. Systems on annual would require an annual site visit or an
annual voluntary Level 2 assessment.
5 The term "Pwen" refers to the probability that a randomly selected well across the United States will test positive for
a virus or fecal indicator species, such as EC, in its source water; "Psampie" is the probability that given a
contaminated well, a random sample at the well will eventually test positive for a virus or fecal indicator. The GWR
EA (November 2006), and in particular the Baseline Chapter of that document, provide a detailed explanation of the
analysis used to generate these Pwen and Psampie estimates.
6 Under the GWR, GWSs with less than 4 log of treatment for viruses must sample their source for a fecal indicator
(e.g., E. coli) when they incur a TC+ under the TCR (or AIP option). If sample water is fecal indicator positive, the
system may be required to implement remedial actions at the source, one of which achieves disinfection to 4 log for
viruses.
Economic Analysis for the Proposed RTCR 5-14
June 2010
-------
occurrence for systems implementing them. In the predictive model, simulated nondisinfecting
ground water systems having a Total Coliform-positive (TC+) take one or more source water
samples per the GWR triggered monitoring requirement and test the samples for fecal indicator
presence (EC is assumed to be the fecal indicator used)7. The probability of observing an E. coli-
positive (EC+) is based on the product of Pweii and the Psampie value drawn for that system. If
positive, a second assessment is done to determine if that system implements disinfection or a
nondisinfection corrective action based on the estimated proportion choosing those options as
defined in the GWR EA for the various types and sizes of systems.8
The simulation model keeps track of systems that begin as nondisinfecting ground water
systems at the start of the analysis period and elect disinfection at some point during the 30-year
period from the effective date of the GWR (December 2009) through the end of the analysis
period for the proposed RTCR. The inventory for disinfecting and nondisinfecting ground water
systems are then adjusted accordingly for subsequent stages of the analysis.
Exhibit 4.4 in Chapter 4 of this EA presents the baseline inventory of ground water
systems by disinfection status at the beginning of proposed RTCR implementation. Those
Ground Water Systems (GWSs) achieving < 4 log disinfection must sample at their source.
Sampling costs were accounted for in the GWR EA (USEPA 2006a).
Throughout the modeling period, Ground Water (GW) systems that do not currently
disinfect or add disinfection may have EC+ source water samples and implement a
nondisinfection corrective action. Modeling in the GWR EA did not incorporate estimates of the
effectiveness of implementing GWR requirements because the GWR was only effective as of
December 2009. Absent this data, EPA applied best professional judgment in assuming that
these corrective actions would have approximately the same effectiveness as a Level 2 corrective
action since they are similar in nature to an acute violation under TCR. As shown in Exhibit 5.8,
if a nondisinfecting system performs a nondisinfecting corrective action, the model assumes that
occurrence is 0% for the remainder of that year and 2 full years after that. As a result of this
assumption, baseline occurrence will be significantly reduced as a function of movement from
non-disinfecting to disinfecting status. The modeling then assumes the P values are only 25
percent of baseline values for an additional five years. The model assumes the system returns to
baseline P values after that (approximately) seven year period.
Additionally, the model incorporates the effect of SSs performed under the GWR as of
January 2010, which is estimated by EPA to reduce TC and EC occurrence by 10 percent for the
remainder of the period of analysis. For transient PWSs, which are on a 5 year SS cycle under
the GWR, this reduction is initially applied to 20 percent of qualifying systems for each of the 5
years from 2010 to 2015. For CWSs, which are on a 3 year cycle, this reduction is initially
applied to 1/3 of systems annually from 2010 to 2012. EPA considers the estimate of 10 percent
in reduction of occurrence to be a conservative estimate of the effectiveness of the sanitary
survey provision of the GWR based on best professional judgment of the TCRDSAC. However,
absent empirical information to assume otherwise and because the factor is similarly
incorporated by the model into the estimates of occurrence for all three regulatory scenarios,
EPA assumes that this conservatism does not significantly affect the net results for the AIP
option or Alternative option, which are calculated by comparison to the current TCR (baseline).
7 Edberg, R. (2000) discusses the use of E. coli as an indicator of drinking water quality.
8 Costs for compliance with the GWR were considered in the GWR EA (November 2006) and are not included in
the RTCR EA.
Economic Analysis for the Proposed RTCR 5-15
June 2010
-------
A third GWR requirement that the model incorporates is compliance monitoring by the
subset of GWSs that disinfect, for which the model assigns a 10 percent reduction in occurrence
in effect throughout the period of analysis. Again, this GWR effectiveness estimate may be
conservative, but EPA assumes that because it is incorporated by the model into occurrence
estimates for all three regulatory scenarios, the net effect on results of the AIP option and
Alternative options (in comparison to the current TCR) is unlikely to be significant. Taking into
account the effect of both SS and compliance reductions, the disinfecting GWSs are expected to
ultimately have TC and EC occurrence rates that are 81 percent of the values derived from the
occurrence distributions (from 90 percent for sanitary surveys x 90 percent for compliance
monitoring).
A fourth and last GWR requirement that the model incorporates concerns the repeat
number of samples following a TC positive. The GWR allows PWSs to use of one of the
required four repeat samples for source water testing (for GWSs serving <1,000 people for EC
following a case of a TC+ assay) under the current TCR. By comparison, under the AIP option,
Alternative option, and the current TCR scenarios for systems serving >1,000 people, only 3
repeat samples are required and in these cases an additional, separate sample is used for the
GWR source water EC assay.
As mentioned above with regard to SS and disinfecting system GWR compliance, the
actual reductions in occurrence that will result from implementation of GWR requirements may
differ from those predicted based on assumptions used in this model. EPA believes this would
affect neither the net analysis nor how the AIP or Alternative options compare to each other,
since both would be affected similarly. Section 5.3.3.1 describes the various sources of
uncertainty and variability considered in this EA and presents analysis of key variables.
5.3.2 Summary of Predictive Model
As the following Sections 5.3.2.1 - 5.3.2.2 describe, EPA modeled the expected trend in
occurrence of TC and EC positive assays over a pre-RTCR monitoring period of 5 years (to
capture applicable GWR effects), and a subsequent 25-year period of analysis for the three
regulatory scenarios: continuation of the current TCR, AIP option, and Alternative option9. EPA
sought to capture the changes in occurrence from the implementation of the GWR requirements,
and in turn, the implementation of requirements of the AIP option and the Alternative option.
Along with changes in TC and EC occurrence, the model predicts behavioral changes: the
number of Level 1 and Level 2 assessments (and associated number of Level 1 or Level 2
corrective actions) to be performed and further resulting adjustments to occurrence, and changes
in sampling regimens as systems qualify for reduced monitoring requirements.
5.3.2.1 Approach to Estimating Parameters
As described in Section 5.2 of this chapter, EPA derived Beta distributions, characterized
by their a, /? parameters, based on analyses of the variation in TC and EC positive assays that
occurred both between and within 27 groupings of systems (as defined in Section 5.2.); the
9 Footnote 4 briefly describes the three regulatory scenarios.
Economic Analysis for the Proposed RTCR 5-16
June 2010
-------
resulting a, ft parameter estimates are presented in Exhibits 5.5 and 5.6. Based on an initial
analysis, the 27 groups were condensed into the groups presented in Appendix F of this EA. The
final Beta distributions were based on these groupings to make the best use of the available data,
combining groups of systems when their occurrence levels were similar to avoid having small
sample size in any of the final groups.
5.3.2.2 Description of Predictive Model
Model output in this chapter (Exhibits 5.10 - 5.21) was generated for a 30-year period to
encompass 5 years of the effects of GWR implementation and 25 years of the effects of RTCR
implementation. For the purpose of the benefit-cost analysis, years 3 through 27 of this period
are included in the calculations and presented in the results shown in Chapter 6 (Benefits
Analysis), Chapter 7 (Cost Analysis), and Chapter 9 (Comparison of Benefits and Costs).
Initially in the modeled period, all systems under the current TCR option and the AIP option are
assumed to sample based on the sampling regimen applicable to their system type, size, type of
water source, and whether or not the system disinfects.10 Under the Alternative option, all
systems are assumed to sample monthly for the first five years after the RTCR effective date.
For the AIP option, the model adjusts these monitoring frequencies for ground water
systems serving fewer than 1,000 people in years 9 and 11 of the 30-year period, or 3 and 5 years
after the RTCR effective date, respectively, for CWSs and NCWSs. The adjustments are made
based on the acute and non-acute violations incurred during these 3 and 5 year periods,11 and the
period of assessment coincides with the sanitary survey for CWSs and NCWSs. The adjusted
monitoring frequencies then apply from years 9-30 and years 11 - 30 of the analysis for CWSs
and NCWSs, respectively.
For the Alternative option, the model adjusts the monitoring frequencies from the all-
monthly regimen based on the acute and non-acute violations12 predicted during the assessment
period, which is the first five years after the RTCR effective date, during which all systems
sample monthly. The model applies the schedule thus adjusted in years 11 -30 of the analysis.
The timing for the adjustments made to the monitoring frequency schedules - effective in
years 9 and 11 for CWSs and NCWSs, respectively, following GWR implementation under the
AIP and Alternative options - are consistent with the length of time of the respective sanitary
survey cycles. Although EPA uses the sanitary cycle to define the period of assessment for
updating the schedule, the results of the surveys are not predicted in this model and are not
directly incorporated into the calculations in any way. Assumptions for the percentage of systems
10 Based on EPA and TWG analysis of Six-Year data and individual state statutes during the TCRDSAC and review
by ASDWA.
11 Systems incurring 1 or more acute or non-acute violations in this period are assumed to be on monthly
monitoring. For the AIP option, the remaining systems (incurring no acute or non-acute violations) are distributed
between quarterly and annual monitoring according to the percentage distribution from the initial M/Q/A estimates
prior to GWR implementation (Exhibit 4.4). Constraints to these adjustments include: 1) The final estimates of the
proportion of community disinfecting systems on monthly monitoring should not decrease from the current baseline
estimate. 2) For the noncommunity systems, there are caps on the % that move to annual monitoring under the AIP:
TNCWS <500: 13.8%; TNCWS >500: 8.6%; NTNCWS < 500: 7.3%; NTNCWS>500: 3.0%.
12 Systems incurring 1 or more acute or non-acute violations in this period are assumed to be on monthly
monitoring. For the Alternative option, all of the remaining systems (incurring no acute or non-acute violations)
move to quarterly monitoring.
Economic Analysis for the Proposed RTCR 5-17
June 2010
-------
on monthly, quarterly, or annual monitoring upon proposed RTCR implementation are shown in
Exhibit 5.9a - 5.9c.
Using the monitoring frequencies described above, the model applies the baseline
occurrence associated with each category of PWS at the start of each modeled year (i.e., the level
of occurrence in 2005 under the current TCR updated to reflect GWR implementation, as
described in Section 5.3.1), as described in Exhibits 5.7 and 5.8 for surface and ground water
systems, respectively. The TC and EC occurrence (both from routine and repeat samples) are
modeled by combining the Beta distributions described previously with the binomial distribution
to simulate TC and EC positives (i.e., successes) given the number of samples (i.e., trials) each
month.
For each simulated system in the model, a probability (P) is drawn randomly for each of
the RTTC, RPTC, RTEC, and RPEC types of samples from their corresponding Beta
distributions. These P values are used in the binomial distribution, together with the number of
samples taken each month, to predict the number of TC and EC positives. In the absence of any
corrective actions being performed under the proposed RTCR, those P values remain the same
for that system for the entire modeling period. If TC or EC occurrence is found that results in a
specific corrective action under the proposed RTCR, temporary reductions in these P values are
made as described further below. In addition, when a system has a TC+ in a particular month that
does not result in a specific corrective action, it is assumed, (for both the current TCR and the
proposed RTCR), that the RTTC rate in the following month will change to reflect either some
worsening of the problem or an improvement based on some other actions taken that are not part
of the specific corrective actions being considered. This was carried out by selecting at random a
different RTTC P value in those months following a TC+ for which explicit corrective actions
were not initiated. In the second month following a TC+ that does not result in a corrective
action, the RTTC rate reverts to the rate applicable in the month in which the TC+ occurred.
EPA determined that this approach provided overall results for the current TCR that
approximated what was observed in the 6-Year Review data and SDWIS data as discussed in the
Section 5.3.3.2 (model validation) below.
Based on the frequency of sampling and the occurrence probabilities for each system, the
model determines what percentage of these systems will be required to take an action (i.e.,
perform a Level 1 or Level 2 assessment in response to exceeding the threshold of multiple TC
positive assays or one EC positive assay, and, in some cases, implement a corrective action).
For the percentage of systems predicted to require a Level 1 assessment following the
requisite number of TC positives, the model assumes that 10 percent will find and address the
exact source of the problem under the AIP option and Alternative option. This represents an
incremental increase over the number expected to implement such a corrective action under the
current TCR. The model assumes that systems that successfully identify and correct the problem
will not have a positive assay for the remainder of that year nor 1 additional full year after the
assessment, after which the probability of a TC or EC positive will be 50 percent that of the
baseline P value for 3 additional years.13 After this period of reduced occurrence, the model
13Using best professional judgment informed by deliberations within the TCRDSAC, EPA assumes that some
systems will have 0% occurrence for some number of years longer than one year, and some will have occurrence
levels that are >0% within the first year. The assumption used in the analysis (systems will have 1 year with 0%
occurrence, plus 2 to 3 years of reduced occurrence from their baseline occurrence) is meant to represent an average
system that experiences occurrence that is between these two scenarios.
Economic Analysis for the Proposed RTCR 5-18
June 2010
-------
assumes that occurrence probabilities for this subset of systems return to those that were initially
assigned. Because the Level 1 and Level 2 assessments are a new requirement, empirical data on
their effectiveness in determining an appropriate corrective action is not available. Absent this
evidence, EPA has applied best professional judgment in choosing what it believes to be a
significant but conservative success rate - 10% of assessments will result in determining the root
of the problem and effectively implementing an appropriate corrective action.
Similarly, for the systems predicted to require a Level 2 assessment based on observing
the requisite number of TC and/or EC positives, under the AIP option and Alternative options an
incremental 10 percent more of those systems than under the current TCR are assumed to then
implement a corrective action that addresses the cause of the problem. For those systems, the
model assumes that occurrence is 0% for the remainder of that year and then for 2 full years after
that. The modeling then assumes the P values are only 25 percent of baseline values for an
additional five years. The model assumes the system returns to baseline P values after that
(approximately) seven year period.
For nondisinfecting GWSs, a percentage of positive samples is estimated (from the Pweii
and Psample parameters) to be related to a source water problem. All nondisinfecting GWSs
having sourcewater EC positives14 will implement disinfection or nondisinfecting corrective
action. Those adding disinfection essentially get removed from the nondisinfecting category and
are reassigned to the disinfecting category for the remainder of the modeling period. For the
portion of this subset that are assumed to implement a nondisinfecting corrective action15, the
systems are assumed to experience the same reduced occurrence levels as those addressing Level
2 Assessments (i.e., 0% occurrence for 2 years, then occurrence at 25 percent of baseline for an
additional 5 years). As Chapter 7 of this EA explains, no costs are attributed to the proposed
RTCR for any corrective action related to source water EC occurrence since such costs have
already been considered in the GWR EA (November 2006).
Outputs generated from this predictive simulation model (for PWSs serving 4,100 or
fewer people) include the following for each modeled year: occurrence of TC and EC positive
assays, the number of Level 1 and Level 2 assessments conducted, and the number of Level 1
and Level 2 corrective actions implemented. Based on these model outputs and the criteria for
reduced monitoring, the model also provides estimates of the number of systems on monthly,
quarterly, and annual sampling regimes (and by implication the number of routine, repeat, and
additional routine samples taken annually) under the AIP option and Alternative option. These
requirements are detailed in Chapter 3 of this EA.
The results of the proposed RTCR predictive occurrence model are presented in the
following Section, 5.3.3.
14 The model determines source water occurrence separately based on the subset of nondisinfecting GW systems, as
described in Section 5.3.1 (2nd paragraph) of this chapter.
15 Estimates of the number of systems implementing a nondisinfection corrective action are taken from Exhibit
6.21b ("Estimated Distribution of Source Water Contamination Corrective Actions") of the GWR EA (USEPA
2006x). The GWR EA estimated the number of systems choosing a nondisinfection corrective action based on a
range; the high end was the percentage of CWS entry points employing disinfection at that time by system size , and
the low end was assumed to be 10% based on discussions with state representatives.
Economic Analysis for the Proposed RTCR 5-19
June 2010
-------
Exhibit 5.7 Simulated Impact of the AIP and Alternative Options on Systems
Serving <4,100 People - Surface Water Systems
Note 1: For Level 1 and Level 2 assessments, systems are
estimated to find the exact source of the problem 10% of the
time under the AIP and Alternative Analysis.
Economic Analysis for the Proposed RTCR 5-20
June 2010
-------
Exhibit 5.8 Simulated Impact of the AIP and Alternative Options on Systems
Serving <4,100 People - Ground Water Systems
Note 2: For Level 1 and Level 2 assessments, systems are estimated to find
the exact source of the problem 10% of the time under the AIP and Alternative
Analysis.
Note 3: For modeling purposes, adjustments to a new steady state distribution
between monthly, quarterly, and annual monitoring regimens are made once in
the period of analysis - after 5 years for noncommunity water systems and 3
years for community water systems. This time frame is meant to correspond
with a full sanitary survey cycle for these systems based on the Agency's
assumption that States may want to coordinate their evaluation of PWS
eligibility for reduced monitoring with the sanitary survey cycle.
Economic Analysis for the Proposed RTCR 5-21
June 2010
-------
Exhibit 5.9a Percent of GW Systems Assumed to be on Monthly, Quarterly, and
Annual (M/Q/A) Monitoring by System Category under Current TCR (Baseline) -
Initial Estimates (Post GWR Implementation)
Disinf?
GW or
SW?
Sys Type
Size
No.
Systems
Monthly
Quarterly
Annual
Y
GW
TNCWS
<100
13,558
4.8%
62.9%
32.3%
Y
GW
TNCWS
101-1000
6,014
7.9%
66.9%
25.2%
Y
GW
TNCWS
1001-4100
269
100.0%
0.0%
0.0%
Y
GW
NTNCWS
<100
2,904
19.3%
64.6%
16.0%
Y
GW
NTNCWS
101-1000
3,621
18.5%
66.7%
14.7%
Y
GW
NTNCWS
1001-4100
542
100.0%
0.0%
0.0%
Y
GW
CWS
<100
6,132
86.6%
13.4%
0.0%
Y
GW
CWS
101-1000
12,762
88.5%
11.5%
0.0%
Y
GW
CWS
1001-4100
5,405
100.0%
0.0%
0.0%
N
GW
TNCWS
<100
46,642
4.8%
62.9%
32.3%
N
GW
TNCWS
101-1000
15,224
7.8%
66.8%
25.3%
N
GW
TNCWS
1001-4100
348
100.0%
0.0%
0.0%
N
GW
NTNCWS
<100
5,913
19.3%
64.6%
16.0%
N
GW
NTNCWS
101-1000
4,710
18.5%
66.7%
14.8%
N
GW
NTNCWS
1001-4100
270
100.0%
0.0%
0.0%
N
GW
CWS
<100
5,806
86.6%
13.4%
0.0%
N
GW
CWS
101-1000
5,597
88.6%
11.4%
0.0%
N
GW
CWS
1001-4100
1,038
100.0%
0.0%
0.0%
Notes:
1) EPA assumes that the frequencies noted in this exhibit for the current TCR would also apply under the AIP
option in the first 10 years of the analysis period. Under the Alternative option, all systems would monitor monthly
for the first 5 years after the RTCR effective date.
2) Estimates for the size categories presented in this exhibit are produced from weighted averages from the
following size categories (and corresponding estimates of proportion sampling monthly, quarterly, or annually):
<100, 101-500, 501-1000, 1001-2500, 2501-3300, and 3301-4100. Water Type includes disinfection status, as
informed by information from SDWIS-FED.
3) Systems with no indication of disinfection ("Disinf?") status in SDWIS-FED were assumed to not disinfect.
4) Some figures may not add due to rounding in the model.
Economic Analysis for the Proposed RTCR 5-22
June 2010
-------
Exhibit 5.9b Percent of GW Systems Predicted to be on M/Q/A Monitoring by
System Category under AIP Option - Adjusted Estimates (Post RTCR
Implementation)
Disinf?
GW or
SW?
Sys Type
Size
No.
Systems
Monthly
Quarterly
Annual
Y
GW
TNCWS
<100
13,558
6.4%
79.8%
13.8%
Y
GW
TNCWS
101-1000
6,014
6.5%
80.2%
13.2%
Y
GW
TNCWS
1001-4100
269
100.0%
0.0%
0.0%
Y
GW
NTNCWS
<100
2,907
3.0%
89.7%
7.3%
Y
GW
NTNCWS
101-1000
3,621
2.5%
91.2%
6.3%
Y
GW
NTNCWS
1001-4100
542
100.0%
0.0%
0.0%
Y
GW
CWS
<100
6,132
86.6%
13.4%
0.0%
Y
GW
CWS
101-1000
12,762
88.5%
11.5%
0.0%
Y
GW
CWS
1001-4100
5,405
100.0%
0.0%
0.0%
N
GW
TNCWS
<100
46,643
14.4%
71.8%
13.8%
N
GW
TNCWS
101-1000
15,225
16.1%
70.5%
13.4%
N
GW
TNCWS
1001-4100
348
100.0%
0.0%
0.0%
N
GW
NTNCWS
<100
5,919
15.1%
77.6%
7.3%
N
GW
NTNCWS
101-1000
4,711
11.4%
82.0%
6.5%
N
GW
NTNCWS
1001-4100
270
100.0%
0.0%
0.0%
N
GW
CWS
<100
5,806
86.6%
13.4%
0.0%
N
GW
CWS
101-1000
5,597
88.6%
11.4%
0.0%
N
GW
CWS
1001-4100
1,038
100.0%
0.0%
0.0%
Notes:
1) M/Q/A percentage estimates are based on model predictions of acute and non-acute violations from years 6-10
(NCWSs) and 6-8 (CWSs) of the 30-year modeled period. Systems incurring 1 or more acutes or non-acutes in
this period are assumed to be on monthly monitoring for years 11 - 25 of the analysis; the remaining systems are
distributed between quarterly and annual monitoring according to the percentage distribution from the initial M/Q/A
estimates (Exhibit 5.9a). Other constraints apply to the determination of the M/Q/A distribution of systems as
described in Section 5.3.2 above.
2) Estimates for the size categories presented in this exhibit are produced from weighted averages from the
following size categories (and corresponding estimates of proportion sampling monthly, quarterly, or annually):
<100, 101-500, 501-1000, 1001-2500, 2501-3300, and 3301-4100.
3) Systems with no indication of disinfection ("Disinf?") status in SDWIS-FED were assumed to not disinfect.
4) Some figures may not add due to rounding in the model.
Economic Analysis for the Proposed RTCR 5-23
June 2010
-------
Exhibit 5.9c Percent of GW Systems on M/Q/A Monitoring by System Category
under Alternative option - Adjusted Estimates (Post RTCR Implementation)
Disinf?
GW or
SW?
Sys Type
Size
No.
Systems
Monthly
Quarterly
Annual
Y
GW
TNCWS
<100
13,558
13.0%
87.0%
0.0%
Y
GW
TNCWS
101-1000
6,014
12.4%
85.7%
0.0%
Y
GW
TNCWS
1001-4100
269
100.0%
0.0%
0.0%
Y
GW
NTNCWS
<100
2,907
4.9%
95.1%
0.0%
Y
GW
NTNCWS
101-1000
3,621
4.5%
95.5%
0.0%
Y
GW
NTNCWS
1001-4100
542
100.0%
0.0%
0.0%
Y
GW
CWS
<100
6,132
86.6%
13.4%
0.0%
Y
GW
CWS
101-1000
12,762
88.5%
11.5%
0.0%
Y
GW
CWS
1001-4100
5,405
100.0%
0.0%
0.0%
N
GW
TNCWS
<100
46,642
25.6%
74.4%
0.0%
N
GW
TNCWS
101-1000
15,224
25.9%
74.1%
0.0%
N
GW
TNCWS
1001-4100
348
100.0%
0.0%
0.0%
N
GW
NTNCWS
<100
5,919
22.9%
77.1%
0.0%
N
GW
NTNCWS
101-1000
4,710
17.2%
82.8%
0.0%
N
GW
NTNCWS
1001-4100
270
100.0%
0.0%
0.0%
N
GW
CWS
<100
5,806
86.6%
13.4%
0.0%
N
GW
CWS
101-1000
5,597
88.6%
11.4%
0.0%
N
GW
CWS
1001-4100
1,038
100.0%
0.0%
0.0%
Notes:
1) M/Q/A percentage estimates are based on model predictions from years 6-10 (NCWSs) and 6-8 (CWSs) of
the period of analysis. Systems incurring 1 or more acutes or non-acutes in this period are assumed to be on
monthly monitoring for years 11 - 25 of the analysis; the remaining systems are assumed to be on quarterly
monitoring. Other constraints apply to the determination of the M/Q/A distribution of systems as described in Section
5.3.2 above.
2) Estimates for the size categories presented in this exhibit are produced from weighted averages from the
following size categories (and corresponding estimates of proportion sampling monthly, quarterly, or annually):
<100, 101-500, 501-1000, 1001-2500, 2501-3300, and 3301-4100.
3) Systems with no indication of disinfection ("Disinf?") status in SDWIS-FED were assumed to not disinfect.
4) Some figures may not add due to rounding in the model.
5.3.3 Predictive Model Results
As noted above, the predictive model produces output that includes national estimates of
sampling and occurrence in PWSs across the United States, the resulting Level 1 and Level 2
assessments, corrective actions resulting from Level 1 and Level 2 assessments, and reductions
in sampling and occurrence for these systems in accordance with requirements under the current
TCR option, the AIP option, and the Alternative option. In Exhibits 5.10 through 5.15, output
includes the model years 1-30 described in Section 5.3.2.2. Years 1 - 5 are summed for the
current TCR only and represent the period prior to the start of monitoring under the proposed
RTCR, including 5 years after the GWR effective date and 3 overlapping years of RTCR
implementation prior to the RTCR effective date. Year 5 in the analysis (also for just the current
Economic Analysis for the Proposed RTCR 5-24
June 2010
-------
TCR) is also shown separately because it represents the baseline year, reflecting conditions just
prior to the start of monitoring under the proposed RTCR. Model years 6 through 30 are summed
for the current TCR, the AIP option, and the Alternative option because they represent the 25-
year period of analysis following promulgation of the proposed RTCR.
For all types of PWSs, the model output is presented for the current TCR option (i.e., the
baseline regulatory scenario for this EA), the AIP option, and the Alternative option. The TCR
with the GWR effects incorporated serves as the appropriate baseline for comparisons with the
AIP option.
EPA's general expectations for TC occurrence for GW systems shown in Exhibits 5.16 -
5.18 are as follows:
1. Years 1-5 show a decline in the TC+ rates due to GWR implementation effects. These
include expected reductions in RTTC rates for all GW systems related to sanitary
surveys, for disinfecting GW systems related to the compliance monitoring, and for non-
disinfecting GW systems due to corrective actions taken following a source water EC+
sample.
2. Years 6-30 reflect additional reductions (beyond GWR effects) in TC+ rates due to the
AIP option or Alternative option. The AIP option and Alternative option occurrence rates
are generally lower than those for TCR with the GWR primarily because of the additional
corrective actions (Level 1 and Level 2). In some subgroups, some counts are higher
under the Alternative option than under the current TCR due to the increased sampling
e.g., routine TC and routine EC positives for GW NTNCWSs (Ex. 5.11). The balance of
the two effects of reduced sampling (reduced additional and repeat samples) and the
additional assessments and corrective actions, which move the counts in opposite
directions, is explored in the sensitivity analysis in Section 5.3.3.1 of this chapter.
3. In Years 6 through 10 and years 11-30 differences in TC+ rates between the AIP option
and Alternative option are due to different proportions of systems that are performing
monthly, quarterly, or annual monitoring. In general, more monitoring (e.g., more doing
monthly than quarterly, or more doing quarterly than annual) - in conjunction with the
corrective actions - will push the TC occurrence levels down (more samples are expected
to be taken and fewer are likely to be positive).
a. In years 6-10, the Alternative option has a lower TC+ rate than the AIP option
because all systems under the Alternative option begin the five year period on
monthly sampling.
b. By year 11, the systems are sampling according to their "steady state" regimens,
when all of the systems that qualified for reduced monitoring during their
respective periods of assessment (as described in Section 5.3.2.2) are assigned to
their reduced regimens in the model. For the AIP option, the proportions of
systems on M/Q/A sampling is identical to those under the current TCR; under
the Alternative option, however, systems that would qualify under the current
TCR for annual monitoring can only reduce to quarterly monitoring.
Economic Analysis for the Proposed RTCR 5-25
June 2010
-------
In Exhibit 5.16 (Community GW systems), the graph shows the expected drop in TC
rates for years 1-5 as the GWR is implemented prior to AIP option. There are only very small
differences between AIP option and Alternative options, and these are not observable on this
graph. A comparison of the estimates in Exhibit 5.10 clarifies these small differences. Most of
these systems are already on monthly sampling, so there is no significant difference seen
between the AIP and Alternative options in years 6-8 for those few not on monthly and it is
assumed that the monthly/quarterly distribution after this period returns to that under the current
TCR. Therefore, the TC+ rates over the 25 year period for these systems show no difference
between the AIP and Alternative options.
In Exhibit 5.17 (Non-transient Noncommunity (NTNC) GW systems), the graph also
shows the expected drop in TC+ rates for years 1-5 as the GWR is implemented prior to the
proposed RTCR. For the years 6-10 the occurrence rates decrease due to the implementation of
corrective actions, and the rates for the Alternative option are lower than for the AIP option since
all systems are on monthly sampling under the Alternative option for this period (having the
effect of reducing the occurrence level as noted above). However, from year 11 to 30, the M/Q/A
distributions for these two are essentially the same and therefore the TC occurrence levels are the
same over this period.
In Exhibit 5.18 (TNCWS GW), the graph also shows the expected drop in TC+ rates for
years 1-5 as the GWR is implemented prior to the proposed RTCR. For the years 6-10, the TC
rates drop further as the corrective actions implemented in response to Level 1 and Level 2
assessments are begun for both the AIP and Alternative options. TC occurrence is markedly
lower during this period for the Alternative option where all are on monthly, whereas the
majority of systems under the AIP option are on quarterly. Beginning in Year 11, there is a shift
under the AIP option toward more systems on monthly than under the TCR, and more systems
on quarterly than under the TCR, and hence the overall TC+ rate drops. For the AIP option,
beginning in Year 11, a substantial portion move from the all monthly sampling during years 6-
10 to quarterly sampling (but none on annual). Because there are slightly more systems under
Alternative option that end up on monthly and quarterly than under the AIP option for years 11-
30, the TC+ rate for the AIP option is slightly higher than for the Alternative option.
EPA's general expectations for TC+ rates for SW systems shown in Exhibits 5.19-5.21
are as follows:
1. The current TCR is provided as the baseline for comparison with the proposed RTCR
(the GWR does not apply to surface water systems and therefore its effect is not included
in these graphs).
2. Because all surface water systems of all sizes do monthly monitoring (no systems are on
quarterly or annual monitoring), the applicable provisions of the proposed RTCR are
identical for the AIP and Alternative options.
3. The reductions in TC+ rates shown by the proposed RTCR relative to the current TCR
therefore reflect solely the effects of the corrective actions implemented in response to
Level 1 and Level 2 assessments.
Economic Analysis for the Proposed RTCR 5-26
June 2010
-------
In Exhibit 5.19 (Community SW systems), the graph shows little difference in TC+ rates
between the current TCR and proposed RTCR. This is primarily because the current TC+ rate is
very low for these systems, and therefore there are few corrective actions implemented over the
25 year period.
In Exhibit 5.20 (NTNC SW), the graph shows that there is a slightly greater difference
between the current TCR and the proposed RTCR for the NTNC SW systems than for the
community systems shown in Exhibit 5.19. This is because the baseline TC+ rate is slightly
higher and, therefore, more corrective actions are performed.
In Exhibit 5.21 (TNCWS SW), the graph indicates that this group exhibits the most
significant change between the current TCR and proposed RTCR since the baseline TC+ rate for
these systems is approximately twice that of the NTNC and community systems.
A comparison of results for SW versus GW systems reveals a pattern of lower occurrence
for SW systems; this is explained by the uniform requirement that SW systems disinfect. Many
GW systems are not required to disinfect, and the result is a level of occurrence that is
periodically elevated in some systems, raising the overall average in GW system categories. This
difference is larger in the noncommunity GW systems than for the community water systems
based on the relatively large proportion of noncommunity water systems that do not disinfect.
The difference is largest among the transient noncommunity water systems, a category in which
the largest proportion of GW systems do not disinfect.
Complete results from the model (by year and system size) and additional graphs are
shown in Appendices A and B, respectively. The implications of these results in terms of
changes in the level of risk associated with contamination of PWS water supplies are discussed
in Chapter 6 of this EA. Model results are used to calculate national estimates of incremental
costs for PWSs in Chapter 7 of this EA.
Economic Analysis for the Proposed RTCR 5-27
June 2010
-------
Exhibit 5.10 Ground Water Community Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1 CA
L2CA
GWR
ND
GWR
D
sum years
6-30
TCR No GWR
12,712,859
1,052,957
816,079
0
215,787
111,640
10,267
4,344
0
55,639
5,773
34,556
30,657
0
0
0
0
TCR
12,761,616
864,605
588,252
93,331
176,379
66,797
7,236
2,106
872
39,769
3,335
26,947
18,545
0
0
714
158
AIP
12,984,469
19,027
488,523
83,140
162,841
48,053
6,438
1,347
816
30,271
2,362
22,112
11,667
2,220
1,155
665
151
Alt
13,073,297
16,686
490,683
83,299
163,561
47,829
6,562
1,595
781
29,955
2,511
21,917
11,660
2,242
1,175
622
160
sum years
1-5
TCR No GWR
2,542,182
254,662
197,228
0
52,150
27,053
2,471
1,033
0
13,486
1,391
8,410
7,393
0
0
0
0
TCR
2,550,295
180,676
122,391
19,799
36,791
14,208
1,532
460
312
8,358
711
5,542
4,103
0
0
203
109
Year 5
TCR No GWR
508,401
42,520
32,969
0
8,719
4,613
402
176
0
2,259
240
1,415
1,241
0
0
0
0
TCR
510,266
35,345
24,060
3,864
7,231
2,655
297
70
48
1,605
130
1,107
731
0
0
31
17
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Exhibit 5.11 Ground Water Nontransient Noncommunity Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1 CA
L2CA
GWR
ND
GWR
D
sum years
6-30
TCR No GWR
2,661,114
328,889
264,574
0
68,232
59,455
3,484
2,338
0
24,614
2,723
14,759
14,021
0
0
0
0
TCR
2,668,723
271,756
178,204
42,339
55,965
34,716
2,534
1,269
446
17,922
1,778
11,658
8,991
0
0
351
95
AIP
2,564,440
90,884
145,401
37,384
48,467
25,164
2,115
841
435
13,642
1,213
10,157
5,088
1,015
499
350
85
Alt
3,274,430
67,297
181,342
47,000
60,447
30,963
2,566
1,058
484
16,967
1,506
11,673
7,422
1,161
745
378
105
sum years
1-5
TCR No GWR
532,239
66,134
53,191
0
13,714
11,904
693
446
0
4,925
531
2,956
2,794
0
0
0
0
TCR
533,205
58,599
38,489
9,192
12,094
7,867
574
288
150
3,924
403
2,477
2,084
0
0
99
52
Year 5
TCR No GWR
106,453
13,330
10,711
0
2,760
2,425
137
89
0
1,001
102
603
555
0
0
0
0
TCR
106,725
10,980
7,219
1,717
2,267
1,416
95
48
23
729
68
473
362
0
0
16
6
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Economic Analysis for the Proposed RTCR
5-28
June 2010
-------
Exhibit 5.12 Ground Water Transient Noncommunity Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1 CA
L2CA
GWR
ND
GWR
D
sum years
6-30
TCR No GWR
7,508,835
1,766,220
1,489,330
0
374,523
373,849
17,881
14,195
0
143,238
15,757
87,899
79,745
0
0
0
0
TCR
7,524,239
1,458,180
966,919
253,032
305,814
213,958
12,859
7,416
2,651
105,646
10,326
70,031
51,552
0
0
2,134
517
AIP
9,780,374
674,328
981,521
283,807
327,174
191,401
13,121
5,792
2,993
99,961
8,421
71,817
39,668
7,188
4,039
2,455
538
Alt
13,983,988
459,373
1,309,762
378,723
436,587
245,373
17,405
7,559
3,417
128,639
11,120
82,133
63,375
8,039
6,253
2,837
580
sum years
1-5
TCR No GWR
1,501,831
350,417
295,149
0
74,227
73,576
3,577
2,709
0
28,251
3,150
17,357
15,692
0
0
0
0
TCR
1,503,848
311,414
206,844
54,227
65,441
47,969
2,873
1,791
942
22,651
2,360
14,769
11,528
0
0
610
332
Year 5
TCR No GWR
300,376
70,103
59,023
0
14,844
14,766
686
612
0
5,627
610
3,545
3,008
0
0
0
0
TCR
300,923
59,747
39,565
10,447
12,536
8,813
506
331
129
4,269
429
2,821
2,101
0
0
87
42
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Exhibit 5.13 Surface Water Community Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1
CA
L2
CA
GWR
ND
GWR
D
sum years
6-30
TCR
3,516,224
137,063
98,139
0
27,729
5,244
2,909
293
0
3,322
666
2,829
1,237
0
0
0
0
AIP
3,565,200
0
81,187
0
27,062
3,860
2,866
180
0
2,755
502
2,415
901
243
93
0
0
Alt
3,565,200
0
81,187
0
27,062
3,860
2,866
180
0
2,755
502
2,415
901
243
93
0
0
sum years
1-5
TCR
703,236
27,458
19,651
0
5,550
1,066
584
57
0
675
137
573
255
0
0
0
0
Year 5
TCR
140,622
5,559
3,956
0
1,117
224
119
13
0
138
29
115
55
0
0
0
0
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Economic Analysis for the Proposed RTCR
5-29
June 2010
-------
Exhibit 5.14 Surface Water Nontransient Noncommunity Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1
CA
L2
CA
GWR
ND
GWR
D
sum years
6-30
TCR
222,494
13,728
10,670
0
2,796
1,127
360
51
0
458
105
292
294
0
0
0
0
AIP
225,900
0
8,183
0
2,728
675
350
37
0
330
76
232
187
24
19
0
0
Alt
225,900
0
8,183
0
2,728
675
350
37
0
330
76
232
187
24
19
0
0
sum years
1-5
TCR
44,500
2,740
2,131
0
558
221
72
10
0
90
21
57
59
0
0
0
0
Year 5
TCR
8,898
554
429
0
113
43
15
2
0
17
4
11
12
0
0
0
0
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Exhibit 5.15 Surface Water Transient Noncommunity Water System Model Output Cumulative Endpoints
Period
Regulatory
Scenario
Reg Rout
Samples
Reg Rout
Samples
if Addis
Taken
Repeat
Samples
GWR
Samples
RTTC+
RPTC+
RTEC+
RPEC+
GWR
EC +
Non-
acutes
Acutes
L1
Assmt
L2
Assmt
L1
CA
L2
CA
GWR
ND
GWR
D
sum years
6-30
TCR
609,514
71,280
57,424
0
14,612
8,365
1,855
398
0
3,148
765
1,893
2,214
0
0
0
0
AIP
625,200
0
39,905
0
13,302
4,311
1,679
218
0
2,037
437
1,357
1,200
130
118
0
0
Alt
625,200
0
39,905
0
13,302
4,311
1,679
218
0
2,037
437
1,357
1,200
130
118
0
0
sum years
1-5
TCR
121,900
14,265
11,476
0
2,920
1,671
382
81
0
628
152
370
449
0
0
0
0
Year 5
TCR
24,373
2,891
2,314
0
589
331
77
17
0
124
30
74
88
0
0
0
0
Source: RTCR predictive model.
Note: Years 1 - 5 in the model reflect incorporation of GWR effects (explained in Section 5.3.3) into pre-GWR baseline data; Year 5 in the model reflects baseline
conditions immediately prior to the effective date of the RTCR; and Years 6-30 represent a period of 25 years after monitoring begins under the RTCR.
Economic Analysis for the Proposed RTCR
5-30
June 2010
-------
Exhibit 5.16 Ground Water Community Water System TC Occurrence
0.018
0.016
0.014
2 0.012
ns
fo"
0.01
~ 0.008
O
Q-
U
(C
3
C
C
<
0.006
0.004
¦TCR No GWR
•TCR
0.002
AIP
Alt
—I—
10
—I 1—
15 20
Years
—i—
25
—i
35
30
Source: RTCR Predictive Model
Notes:
TCR No GWR: Represents the predicted TC occurrence, absent the effects of GWR implementation, under the
current TCR requirements. The current TCR is the baseline regulatory scenario for the RTCR EA.
TCR: Represents the projected TC occurrence under the current TCR requirements, including the effect of GWR
implementation. The TCR is the baseline regulatory scenario for the EA.
AIP: Represents the projected TC occurrence under AIP requirements, including the effects of GWR
implementation.
Alt: Represents the projected TC occurrence under the Alternative option requirements, including the effects of
GWR implementation.
Steady state: By Year 11, all systems that qualify for reduced monitoring during their assessment period (Section
5.3.2.2) have moved to their reduced sampling schedule under the AIP and Alternative options. This period in the
model (from years 11 - 30) reflects a steady state with regard to proportions of systems sampling on monthly,
quarterly, or annual regimens.
Economic Analysis for the Proposed RTCR 5-31
June 2010
-------
Exhibit 5.17 Ground Water Nontransient Noncommunity Water System TC
Occurrence
0.025
0 5 10 15 20 25 30 35
Years
Source: RTCR Predictive Model
Notes:
TCR No GWR: Represents the predicted TC occurrence, absent the effects of GWR implementation, under the
current TCR requirements. The current TCR is the baseline regulatory scenario for the RTCR EA.
TCR: Represents the projected TC occurrence under the current TCR requirements, including the effect of
GWR implementation. The TCR is the baseline regulatory scenario for the EA.
AIP: Represents the projected TC occurrence under AIP requirements, including the effects of GWR
implementation.
Alt: Represents the projected TC occurrence under the Alternative option requirements, including the effects of
GWR implementation.
Steady state: By Year 11, all systems that qualify for reduced monitoring during their assessment period (Section
5.3.2.2) have moved to their reduced sampling schedule under the AIP and Alternative options. This period in the
model (from years 11 - 30) reflects a steady state with regard to proportions of systems sampling on monthly,
quarterly, or annual regimens.
Economic Analysis for the Proposed RTCR 5-32
June 2010
-------
Exhibit 5.18 Ground Water Transient Noncommunity Water System TC
Occurrence
0.045
>
3 0.025
Qi
;1 0.02
O
a.
u
0.015
A3
3
C
< 0.01
0.005
—i—
10
—I 1—
15 20
Years
—i—
25
¦TCR No GWR
¦TCR
AIP
Alt
30
—i
35
Source: RTCR Predictive Model
Notes: TCR No GWR: Represents the predicted TC occurrence, absent the effects of GWR implementation,
under the current TCR requirements. The current TCR is the baseline regulatory scenario for the RTCR EA.
TCR: Represents the projected TC occurrence under the current TCR requirements, including the effect of
GWR implementation. The TCR is the baseline regulatory scenario for the EA.
AIP: Represents the projected TC occurrence under AIP requirements, including the effects of GWR
implementation.
Alt: Represents the projected TC occurrence under the Alternative option requirements, including the effects of
GWR implementation.
Steady state: By Year 11, all systems that qualify for reduced monitoring during their assessment period (Section
5.3.2.2) have moved to their reduced sampling schedule under the AIP and Alternative options. This period in the
model (from years 11 - 30) reflects a steady state with regard to proportions of systems sampling on monthly,
quarterly, or annual regimens.
Economic Analysis for the Proposed RTCR 5-33
June 2010
-------
Exhibit 5.19 Surface Water Community Water System TC Occurrence
Source: RTCR Predictive Model
Notes:
TCR: Represents the projected TC occurrence under the current TCR requirements. The TCR is the baseline
regulatory scenario for the EA.
AIP and Alt: Represents the projected TC occurrence under the AIP and Alternative option, which have the
same requirements for SW systems, resulting in the same TC occurrence under the two options.
Economic Analysis for the Proposed RTCR 5-34
June 2010
-------
Exhibit 5.20 Surface Water Nontransient Noncommunity Water System TC
Occurrence
Source: RTCR Predictive Model
Notes:
TCR: Represents the projected TC occurrence under the current TCR requirements, which compose the
baseline regulatory scenario for the RTCR EA.
AIP and Alt: Represents the projected TC occurrence under the AIP and Alternative option, which have the
same requirements for SW systems, resulting in the same TC occurrence under the two options.
Economic Analysis for the Proposed RTCR 5-35
June 2010
-------
Exhibit 5.21 Surface Water Transient Noncommunity Water System TC
Occurrence
Source: RTCR Predictive Model
Notes:
TCR: Represents the projected TC occurrence under the current TCR requirements. The TCR is the baseline
regulatory scenario for the EA.
AIP and Alt: Represents the projected TC occurrence under the AIP and Alternative option, which have the
same requirements for SW systems, resulting in the same TC occurrence under the two options.
5.3.3.1 Uncertainty and Sensitivity Analysis in Occurrence Modeling
There are two primary sources of uncertainty in the proposed RTCR occurrence
modeling: uncertainty related to the baseline occurrence model due to the limited data on which
the model is based, and uncertainty regarding the frequency and effectiveness of corrective
actions and their effects on subsequent TC and EC occurrence.
Data limitations in part derive from a lack of observed data on the effects of the
requirements of the GWR, for which implementation began as of December 2009. In the absence
of this data, EPA made assumptions regarding the effectiveness of each of the relevant
requirements, as described in Section 5.3.1 of this chapter. Each of these assumptions affects the
modeled current TCR, which is the baseline for this EA. The predictive model uses this same
baseline (current TCR) in determining the net influence of both the AIP option and Alternative
option requirements on occurrence. Therefore, although the actual reductions in occurrence
resulting from GWR implementation may differ from the assumptions used in this model, this
Economic Analysis for the Proposed RTCR 5-36
June 2010
-------
difference would likely not affect the net results of the AIP option and Alternative option relative
to the current TCR, or their performance relative to each other. Other data limitations are
discussed in Section 4.2 of this EA.
Many other assumptions incorporated into the predictive model also have no net effect on
results. More generally, assumptions that contribute to uncertainty in the predictive model results
can be separated into two categories: those that have no net effect on results and those that may
bias results downward or upward. Exhibit 5.22a below presents assumptions that influence the
baseline (current TCR) and the RTCR options considered (AIP option and Alternative option) in
the same way, and therefore are expected to have no significant effect on net results. Exhibit 22b
presents those assumptions that may have a significant effect on net results because they
influence only the AIP and Alternative options, but not the current TCR option. Exhibits 5.22a
and 5.27b also include EPA's best estimate of whether the assumption would tend to over-
estimate, under-estimate, or have an unknown impact.
Based on a lack of empirical data, EPA did not include in the model an estimate for the
number of systems that might be required to return to more frequent monitoring after having
been operating on a reduced schedule. A result of more frequent monitoring when combined
with assessments and corrective actions as required under the AIP option (or Alternative option),
as described elsewhere in this chapter, is expected to be an increase in TC and EC hits, and a
corresponding increase in assessments and Corrective Actions (CAs), resulting in improved
water quality. EPA expects that the result of not accounting for movement of systems back to a
more rigorous monitoring schedule will tend to underestimate benefits and costs for the AIP and
Alternative options relative to the current TCR option. It may have more of an impact on net
results for the AIP option than for the Alternative option because the AIP option is expected to
have a larger proportion of systems on more reduced monitoring.
Economic Analysis for the Proposed RTCR 5-3 7
June 2010
-------
Exhibit 5.22a Summary of Model Parameters Influencing RTCR Baseline and
Options
Model Parameter
Cu rre nt Assu m ption
Section with
Full
Discussion of
Uncertainty or
Factor
Contributes
Variability,
Uncertainty, or
Constant?
Most Likely Effect on Benefits
Most Likely Effect on Costs
Under-
estimate
Over-
estimate
Unknown
Impact
Under-
estimate
Over-
estimate
Unknown
Impact
Applies to Current TCR, AIP Option, and Alternative Option
Distribution of systems across
sampling frequency categories.
Varies per system
category as shown in Ex.
5.14a-5.14c
5.3.2.2
Variability,
Uncertainty
X
(Note 1)
X
(Note 1)
Alpha, beta parameters developed as
model inputs describing distribution of
occurrence based on Six Year Data
(2005)
Varies per system
category as shown in Ex.
5.10-5.11
Appendix F
Variability,
Uncertainty
X
(Note 1)
X
(Note 1)
Pwell - the portion of GWSs having
viral pathogens in their source waters
(adopted from GWR EA)
21.58%
5.3.1
Constant,
Uncertainty
X
(Note 1)
X
(Note 1)
Psample - the probability that a
random sample will test positive for
viral pathogens given a contaminated
source water (adopted from GWR EA)
Variable drawn from a beta
distribution with a range of
alpha and beta estimates
having a median value of
5.8 percent and an
expected value of 12.4
percent.
5.3.1
Variability,
Uncertainty
X
(Note 1)
X
(Note 1)
Reduced occurrence for sanitary
surveys performed under the GWR.
90% of baseline
occurrence; applied to
equal number of systems
annually over the sanitary
survey cycle (20% each
year of 5 years for CWSs;
33.3% each year for 3
years for NCWSs)
5.3.1
Constant,
Uncertainty
X
(Note 1)
X
(Note 1)
Reduced occurrence for GWR
compliance monitoring (applies to
subset of GWSs that disinfect).
90% of baseline
occurrence
5.3.1
Constant,
Uncertainty
X
(Note 1)
X
(Note 1)
% of non-disinfecting GWSs choosing
disinfection CA vs non-disinfection CA
in response to source water quality
issue (adopted from GWR)
Range: High end is the
percentage of CWS entry
points employing
disinfection by system
size, low end assumed to
be 10% based on
discussions with State
representatives.
5.3.1
Constant,
Uncertainty
X
(Note 1)
X
(Note 1)
Seasonal variation in TC hit rates.
TC hit rates are assumed
to be constant over the
course of the year.
5.3.1
Variability,
Uncertainty
X
(Note 1)
X
(Note 1)
The RTTC rate in the month following
TC+ not resulting in CA assumed to
change to reflect either some
worsening of the problem or an
improvement based on some other
actions taken that are not part of the
specific corrective actions being
considered.
This was carried out by
selecting at random a
different pRTTC value in
those months following a
TC+ without explicit
corrective actions being
initiated.
5.3.2.2
Variability,
Uncertainty
X
(Note 1)
X
(Note 1)
Note: For variables or factors that were incorporated into the predictive model in the same way for the Current TCR
(baseline) and the other two regulatory options considered in this EA (AIP and Alternative options), EPA expects that
any under- or over-estimation would affect the baseline and other options in the same way, resulting in no significant
effect on the net results for the AIP or Alternative option.
Economic Analysis for the Proposed RTCR 5-38
June 2010
-------
Exhibit 5.22b Summary of Model Parameters Influencing Proposed RTCR
Options Only
Model Parameter
Cu rre nt Assu m ption
Section with
Full
Discussion of
Uncertainty or
Factor
Contributes
Variability,
Uncertainty, or
Constant?
Most Likely Effect on Benefits
Most Likely Effect on Costs
Under-
estimate
Over-
estimate
Unknown
Impact
Under-
estimate
Over-
estimate
Unknown
Impact
Applies Only to AIP Option and Alternative Option
% Implementing an Effective
Corrective Action after a Level 1 or
Level 2 Assessment
10%
5.3.3.1
Constant,
Uncertainty
X
X
Hit rate immediately following CA
implementation for period of 3 or 5
years (Level 1 and 2, respectively) (or
non-disinfecting GWS choosing a non-
disinfection CA)
0%
5.3.3.1
Constant,
Uncertainty
X
X
Duration of initial phase of reduced hit
rate (0%) following CA implementation
Remainder of year
+ 1 or 2 years
(Level 1 and 2,
respectively)
5.3.3.1
Constant,
Uncertainty
X
X
Hit rate in second phase following CA
implementation (after initial phase of
0% hit rate)
50% or 25% of baseline
occurrence
(Level 1 and 2,
respectively)
5.3.3.1
Constant,
Uncertainty
X
X
Duration of reduced hit rate in second
phase following CA implementation
3 or 5 years
(Level 1 and 2,
respectively)
5.3.3.1
Constant,
Uncertainty
X
X
Duration ot reduced nit rate (U%) in
initial phase following CA
implementation (applies to non-
disinfecting GWSs choosing non-
disinfecting CA)
Remainder of year
+ 2 years
reference
discussion in
5.3.3.1
Constant,
Uncertainty
X
X
Hit rate in second phase following CA
implementation (after initial phase of
0% hit rate) (applies to non-
disinfecting GWSs choosing non-
disinfecting CA)
25% of baseline
occurrence
reference
discussion in
5.3.3.2
Constant,
Uncertainty
X
X
Duration of reduced hit rate in second
phase following CA implementation
(applies to non-disinfecting GWSs
choosing non-disinfecting CA)
5 years
reference
discussion in
5.3.3.3
Constant,
Uncertainty
X
X
For systems serving more than 4,100
people: Application of a rate that
varies by type of system and water
source but is held constant through
time (does not change over the 25
years of analysis)
Assumption that these
groups would incur the
same number of violations
annually throughout the 25
years of the analysis (i.e.,
no reduction in violations in
the years immediately
following CA
implementation.) This
number of violations is
used to calculate the
number of Level 1 and Level
2 assessments and
corrective actions by PWS
category that systems will
implement under the
proposed RTCR
alternatives.
5.4.1
Variability,
Uncertainty
X
X
Note: For variables and factors that affect the options considered (AIP and Alternative options) but do not have a
parallel influence on the current TCR (baseline), EPA expects that any under- or over-estimation will have an effect
as noted in the above exhibit on the net results for the AIP or Alternative option.
Economic Analysis for the Proposed RTCR 5-39
June 2010
-------
Key Factors Driving Analysis Uncertainty
The remainder of this section focuses on those assumptions identified in Exhibit 5.22b as
contributing the most uncertainty to the net results of the analysis: the frequency and
effectiveness of corrective actions (i.e., % reduction in occurrence and duration of reduction
period resulting from correction actions) applied in the predictive model. These are the key
drivers in determining net results, that is, the difference between the AIP option or Alternative
options and the current TCR option.
As described previously in Section 5.3, EPA incorporated the following assumptions
(also shown in Exhibits 5.12 and 5.13) for the frequency and effectiveness of the corrective
actions that systems will implement following a Level 1 or a Level 2 assessment:
Level 1: Following a Level 1 Assessment in a particular year, 10 percent of the systems
performing the assessment will implement corrective actions that will reduce the TC and EC
occurrence levels to 0% for the remainder of that year and for one full year after that. Then, for
an additional 3 full years, the TC and EC occurrence levels for each of these systems will be
reduced to 50 percent of their initial values. After that, the TC and EC occurrence levels return to
their initial values.
Level 2: Following a Level 2 Assessment in a particular year, 10 percent of the systems
performing the assessment will implement corrective actions that will reduce the TC and EC
occurrence levels to 0% for the remainder of that year and for 2 full years after that. Then, for an
additional 5 full years, the TC and EC occurrence levels for each of these systems will be
reduced to 25 percent of their initial values. After that, the TC and EC occurrence levels return to
their initial values.
Also, for the nondisinfecting ground water systems that are found in the simulation to
have an EC+ in their source water but do not move to disinfection, the assumptions for the
effectiveness of the nondisinfecting corrective actions are identical to those above for Level 2
corrective actions. (However, note that all nondisinfecting systems that discover EC+ source
water that do not go to disinfection will implement these corrective actions (under GWR), not
just 10 percent as assumed for systems that already disinfect.)
To assess the influence of these assumptions on the model results, EPA ran the model
with alternative assumptions reflecting corrective actions that are less effective (50% of the
assumption of efficacy in the primary analysis in this EA) and corrective actions that are more
effective (2x the assumption of efficacy in the primary analysis). The assumptions made in the
primary analysis, and the alternative assumptions made in this sensitivity analysis, do not
represent a sum of the effectiveness of all individual CAs, but indicate the average frequency
across all systems with which a given system will correctly diagnose and effectively implement a
corrective action, mitigating the source of the problem. The alternative assumptions assume this
frequency will be half or twice that used in the primary analysis. Using these alternative
assumptions, key model outputs that can serve as proxy indicators of the costs and benefits of the
proposed RTCR were compared with those same outputs from the main model assumptions for
the proposed RTCR.
Economic Analysis for the Proposed RTCR 5-40
June 2010
-------
For the less effective corrective actions, the alternative assumptions for Level 1 and Level
2 are:
Level 1: Following a Level 1 Assessment in a particular year, 5 percent of the systems
performing the assessment will implement corrective actions that reduce the TC and EC
occurrence levels to 0% for the remainder of that year only. Then, for an additional 2 full years,
the TC and EC occurrence levels for each of these systems will be reduced to 50 percent of their
initial values. After that, the TC and EC occurrence levels return to their initial values.
Level 2: Following a Level 2 Assessment in a particular year, 5 percent of the systems
performing the assessment will implement a corrective action that will reduce the TC and EC
occurrence levels to 0% for the remainder of that year and for 1 full year after that. Then, for an
additional 3 full years, the TC and EC occurrence levels for that system will be reduced to 25
percent of their initial values. After that, the TC and EC occurrence levels return to their initial
values. (These alternative Level 2 durations were also applied to the nondisinfecting GWR
corrective actions.)
For the more effective corrective actions, the alternative assumptions for Level 1 and
Level 2 are:
Level 1: Following a Level 1 Assessment in a particular year, 20 percent of the systems
performing the assessment will implement corrective actions that will reduce the TC and EC
occurrence levels to 0% for the remainder of that year and for two full years after that. Then, for
an additional 6 full years, the TC and EC occurrence levels for that system will be reduced to 50
percent of their initial values. After that, the TC and EC occurrence levels return to their initial
values.
Level 2: Following a Level 2 Assessment in a particular year, 20 percent of the systems
performing the assessment will implement corrective actions that will reduce the TC and EC
occurrence levels to 0% for the remainder of that year and for 4 full years after that. Then, for an
additional 10 full years, the TC and EC occurrence levels for that system will be reduced to 25
percent of their initial values. After that, the TC and EC occurrence levels return to their initial
values. (Again, these alternative Level 2 durations were also applied to the nondisinfecting GWR
corrective actions.)
A summary of the assumptions used in these alternative analyses is provided in Exhibit
5.22a - 5.22b. The analyses were run in the predictive model for the AIP option for the
approximately 60,000 nondisinfecting transient noncommunity groundwater systems serving
<500 people. This subgroup contains the largest number of systems and has on average the
largest number of violations of all subgroups; as a group they stand to experience the largest
change in activity under the proposed rule because, unlike the larger systems that have more
resources at their disposal, these systems are not generally going beyond requirements of the
current TCR, and generally have less frequent requirements for sampling under the current TCR,
than the larger systems. To conduct the sensitivity analysis described in this section, EPA applied
to the predictive model the alternative assumptions that are presented in Exhibit 5.23 along with
assumptions applied in the primary analysis for comparison.
Economic Analysis for the Proposed RTCR 5-41
June 2010
-------
Exhibit 5.23 Sensitivity Analysis Assumptions for Frequency and Effectiveness
of Corrective Actions following Level 1 or 2 Assessments
Type of
Assess
ment
% performing
Corrective Action
after a Level 1 or
Level 21
Assessment
Reduced hit rate
immediately after
Corrective Action
Period of time for
reduced hit rate2
Additional
period of
reduced hit rate
(yrs)
Reduction from
initial hit rate in
additional reduced
period
Predictive Model
Level 1
10%
0%
Remainder + 1
3
50%
Level 2
10%
0%
Remainder + 2
5
25%
Sensitivity 1 -
Less Effective
Level 1
5%
0%
Remainder only
2
50%
Level 2
5%
0%
Remainder + 1
3
25%
Sensitivity 2 -
More effective
Level 1
20%
0%
Remainder + 2
6
50%
Level 2
20%
0%
Remainder + 4
10
25%
Notes:
1 Level 2 assumptions for the predictive model and sensitivity analyses are also applied to 100% of nondisinfecting
ground water systems that incur an EC+ in the source water and implement a nondisinfecting corrective action.
2"Remainder" refers to the balance of the year following implementation of a corrective action.
Exhibit 5.24a Cumulative Effect of Alternative Assumptions for Corrective Action
Effectiveness and Duration on AIP Option Model Results
(Nondisinfecting TNCWS Serving <500 People over 25 Years)
Corrective
Actions
Assumption
Model Output
RTTC+
RTTC+
as rate
RPTC+
RTEC+
RPEC+
GWR EC +
Nonacutes
Acutes
LI Assmt
L2 Assmt
LI CA
L2CA
Low CAs1
300,209
3.81%
204,124
11,585
6,613
3,364
104,161
9,162
68,941
48,658
3,539
2,395
AIP2
277,776
3.55%
174,200
9,937
5,038
2,906
90,389
7,170
64,275
36,090
6,448
3,671
High CAs3
240,248
3.11%
124,822
8,271
3,942
2,240
67,662
5,352
53,823
20,710
10,664
3,918
% Change from AIP
Low CAs1
8.1%
7.42%
17.2%
16.6%
31.2%
15.8%
15.2%
27.8%
7.3%
34.8%
-45.1%
-34.7%
High CAs3
-13.5%
-12.43%
-28.3%
-16.8%
-21.8%
-22.9%
-25.1%
-25.4%
-16.3%
-42.6%
65.4%
6.7%
Relative Change
Low CAs1
1.08
1.08
1.17
1.17
1.31
1.16
1.15
1.28
1.07
1.35
0.55
0.65
High CAs3
0.86
0.83
0.72
0.83
0.78
0.77
0.75
0.75
0.84
0.57
1.65
1.07
Notes:
1 "Low CAs" uses half of the primary analysis estimate, or 5% incremental increase in CAs beyond the current TCR (50% decrease in the assumption).
2,1 AIP" refers to the AIP option under assumptions used in the primary analysis, which includes an assumption that 10% more CAs adddressing the root cause of a
TC/EC event will be implemented than under the current TCR (0% change in the assumption).
3 "High CAs" doubles the primary analysis CAs estimate from 10% to 20% (100% increase in the assumption).
Note: Cumulative effects shown in this exhibit are based on application in the predictive occurrence model of the
alternative assumptions for frequency and effectiveness of corrective actions following Level 1 or 2 Assessments, as
presented in Ex. 5.23. The effectiveness of corrective actions is based both on the effectiveness of the Level 1 or Level
2 assessments performed in diagnosing the correct system deficiency(ies) and upon the effectiveness of implementation
of the intended corrective action.
The results of these runs indicate that, as would be expected, the less frequent and less
effective corrective action assumptions result in finding more TC+ and EC+ samples and having
more nonacute and acute events than the main assumptions. In addition, more LI and L2
assessments are conducted in response to those additional events, but fewer corrective actions are
performed.
Economic Analysis for the Proposed RTCR 5-42
June 2010
-------
Conversely, the more frequent and more effective corrective action assumptions lead to
fewer TC+, EC+, nonacute, and acute events occurring and more LI and L2 CAs being
performed.
The fewer number of TC+, EC+, nonacute and acute events occurring with the more
effective corrective action assumptions relative to the main assumptions can be viewed as events
that are prevented from occurring as a result of those more effective actions relative to the main
assumptions. Similarly, the increased number of those events occurring with the less effective
corrective actions can be viewed as those that would not be prevented relative to the main
assumptions.
It is important to note that this sensitivity analysis is not intended as a rigorous,
quantitative comparison of the alternatives but rather as a general indicator of the magnitude and
direction of change in these outputs relative to the magnitude and direction of the changes in
these inputs. Broadly speaking, since the input assumptions reflect changes that are
approximately (but not precisely) factors of one half (0.5) and twice (2.0) the main assumptions,
outputs that have ratios that are substantially less than 0.5 or significantly greater than 2.0 could
be considered to be very sensitive to the changes in these input assumptions. Outputs with ratios
that fall between 0.5 and 2.0 can be considered to be much less sensitive to the input assumptions
(ratios equal to or approaching 1.0 indicate that the output is essentially not sensitive to changes
in the input assumptions.
As indicated in Exhibit 5.24, all of the ratios for outputs fall within the 0.5 to 2.0 range,
indicating that these outputs do not appear to be highly sensitive to these alternative assumptions
for corrective action effectiveness.
The outputs that could serve as indicators of the sensitivity of the benefits on these
assumptions are the numbers of TC and EC positives predicted as well as the number of
nonacute and acute events predicted. These outputs have factors that generally fall well within
the 0.5 to 2.0 range, indicating that these changes in the inputs assumptions appear to result in
relatively small changes in these outputs. That is, the less effective corrective action assumptions
do not appear to result in "missing" a disproportionately greater number of these events, and the
more effective corrective actions appear to result in "finding" a disproportionately greater
number of these events.
The outputs that could serve as indicators of the sensitivity of the costs on these
assumptions are the Level 1 and Level 2 assessments and the Level 1 and Level 2 corrective
actions. The assessments have ratios that, like those for the outputs like TC and EC positives, are
generally well within the 0.5 to 2.0 range. The ratios for the corrective actions are also within
this range but tend to be close to the 0.5 and 2.0 values. The input assumptions used for the
number of corrective actions done is directly proportional to the number of assessments
performed, so these changes would be expected to be more directly linear (closer to 0.5 or 2.0).
For example, using the more effective corrective action assumptions, the number of TC+
samples predicted is reduced, as are the number of Level 1 assessments that are related to TC+
events. The ratio for Level 1 assessments under the more effective corrective action assumptions
is 0.84, indicating that under these assumptions systems will only need to perform about 85% of
the Level 1 assessments over the 25 year period (that is, the others are prevented by the more
effective actions). However, this set of alternative assumptions includes a factor that 20% of
Economic Analysis for the Proposed RTCR 5-43
June 2010
-------
those doing Level 1 assessments go on to do Level 1 corrective actions (which is twice the 10%
assumption in the main model). Therefore, we see that almost twice as many of these Level 1
corrective actions are performed as under the main model. More specifically, the ratio for Level
1 corrective actions is 1.65 which is approximately 2x the 0.84 ratio noted for Level 1
assessment predicted.
An additional analysis was run to test the model's sensitivity to the primary analysis
assumption that 10% more CAs addressing the root cause of a TC/EC event will be implemented
than under the current TCR. The additional sensitivity analysis investigates the implications of a
much greater rate of CA implementation - 50% rather than 10%, or a 400% increase from the
primary analysis assumption. All other parameters in this sensitivity test, including the
effectiveness and duration of the CAs, are identical to those of the AIP primary analysis.
Exhibit 5.24b Cumulative Effect of 50% Assumption for Corrective Action
Implementation Rate on AIP Option Model Results
(Nondisinfecting TNCWS Serving <500 People over 25 Years)
Corrective
Actions
Assumption
Model Output
RTTC+
RTTC+
as rate
RPTC+
RTEC+
RPEC+
GWR EC +
Nonacutes
Acutes
LI Assmt
L2 Assmt
LI CA
L2 CA
AIP1
271,716
3.55%
174,200
9,937
5,038
2,906
90,389
7,170
64,275
36,090
6,448
3,671
50% Option2
229,210
2.97%
110,839
7,887
3,225
2,641
61,424
4,534
53,896
12,506
27,227
6,288
% Change from AIP
50% Option2
-17.5%
-16.22%
-36.4%
-20.6%
-36.0%
-9.1%
-32.0%
-36.8%
-16.1%
-65.3%
322.3%
71.3%
Relative Change
50% Option2
0.83
0.77
0.64
0.79
0.64
0.91
0.68
0.63
0.84
0.35
4.22
1.71
Notes:
1"AIP" refers to the AIP option under assumptions used in the primary analysis, which includes an assumption that 10% more CAs addressing the root cause of a
TC/EC event will be implemented than under the current TCR (0% change in the assumption).
2"50% option" uses a CAs estimate of 50% instead of the primary analysis assumption of 10% (400% increase in the assumption).
As might be expected, this modification causes rates of Level 1 and Level 2 CAs to
increase. This in turn decreases the RTTC+ rate and other levels of occurrence, including the
number of Level 1 and Level 2 assessments. It is interesting to note that the rate of Level 1 CAs
increases much more dramatically than the rate of Level 2 CAs. One possible explanation for this
is that with the higher CA implementation rate, more systems move to a reduced TC occurrence
level during the first years of RTCR. Thereafter these systems are more likely to trigger Level 1
assessments than Level 2 assessments, causing the ratio of Level 1 CAs to Level 2 CAs to shift
substantially from that predicted in the primary analysis.
Since this analysis involves changing a parameter by a factor of five, changes in model
output greater than a factor of five would indicate a high degree of sensitivity to the parameter.
In the table it is evident that no changes on that scale occur. Nonetheless, the analysis does
indicate that an increase in the rate of implementation of CAs would reduce the long-run
incidence of TC/EC events and that it would increase the number of Level 1 CAs more
dramatically than it would increase the number of Level 2 CAs.
Economic Analysis for the Proposed RTCR 5-44
June 2010
-------
Additional Factors Assessed
As described in the introduction to this section, there were assumptions in addition to the
frequency and effectiveness of corrective actions that EPA considered might influence the
analysis. These assumptions were categorized as those that were unlikely to have an impact on
net results (5.22a), and those that affected the RTCR options differently than the current TCR
(5.22b), thus having the potential to impact net results.
Because the effect of seasonality on occurrence is commonly a concern when using
annual average occurrence rates in an analysis, and because RTTC occurrence rates were found
to be twice as high in the summer and fall as in the winter and spring, EPA ran a sensitivity
analysis on the influence of seasonal trends. This analysis was performed on the subset including
transient non-disinfecting groundwater systems serving fewer than 101 people because of its
relatively large number of systems and high occurrence rate. For simplicity, the sensitivity
analysis was limited to only the AIP option.
EPA performed the sensitivity analysis through two runs of the RTCR AIP model. The
first run established a baseline, and the second multiplied the assigned baseline pRTTC values by
a factor of 0.67 during December-May and by a factor of 1.33 during June-November. (No
seasonality factors were applied for systems on annual sampling.)
Summary results in Table 5.25 show expected yearly averages for selected occurrence
metrics during the first 25 years under full implementation of the RTCR. Overall the differences
between the model results are small. Also, they are not consistently higher for one model (e.g.
when accounting for seasonality, total TC occurrence is slightly higher but the number of L2
Corrective Actions is slightly lower). These variations are likely within the range of Monte Carlo
error. Therefore, EPA determined that not incorporating seasonality into the occurrence model is
an acceptable simplification.
Exhibit 5.25 Effect of Seasonality on Occurrence Analysis Endpoints
RTCR AIP Metric
No
Seasonality
Seasonality
Total TC Positives
8524
8556
Total Level 1 Triggers
2742
2784
Total Acutes1
240
240
LI Corrective Actions
200
194
L2 Corrective Actions
107
104
Note: Estimates represent a yearly average during the first 25 years under full
implementation of the RTCR.
includes those Level 2 Triggers resulting from an EC+ (does not include those
resulting from a second Level 1 trigger occurring within a 12-month rolling period).
5.3.3.2 Model Validation
There are a limited number of ways in which outputs from the predictive model for the
systems serving <4,100 people can be compared to observed data. EPA has identified two types
of analyses where comparisons can be made. One of these is to compare the model's predictions
Economic Analysis for the Proposed RTCR 5-45
June 2010
-------
of average annual TC and EC occurrence levels for various types and sizes of systems under the
current TCR against the observations from the 2005 Six-Year Review data used to parameterize
the model. The other is to compare the model's predictions for the current TCR against those
reported in SDWIS/FED.
It is necessary to keep in mind that the primary purpose of the modeling effort for the
systems serving <4,100 people was not to exactly match observed results but to provide a
framework for comparing relative changes across regulatory options. Nevertheless, it is
important that, to the extent possible, the model outputs for conditions that can be checked
against observed data compare reasonably well. This provides assurance that the model is
operating in a manner that is a reasonable simulation of how the TCR currently operates and
provides some measure of confidence that the relative changes seen for the rule options are
meaningful.
Exhibit 5.25 provides a comparison of the model results for average annual TC+ assays
under the current TCR with those observed in the 2005 Six-Year Review data.
Economic Analysis for the Proposed RTCR 5-46
June 2010
-------
Exhibit 5.25 Comparison of TC+ Occurrence Predicted as a 25-Year Annual
Average under the Current TCR with 2005 Six-Year Review Data
System
Type
Source
Water
Population
Served
Six Year review
Predictive Model
TC
{% Positive)
EC
{% Positive)
TC
{% Positive)
EC
{% Positive)
cws
GW
<=100
2.66%
0.08%
2.24%
0.10%
101-500
1.99%
0.05%
1.64%
0.08%
501-1,000
1.52%
0.04%
1.52%
0.07%
1,001-4,100
1.07%
0.02%
0.92%
0.05%
SW
<=100
1.41%
0.09%
1.46%
0.18%
101-500
1.15%
0.05%
0.91%
0.12%
501-1,000
0.70%
0.05%
0.90%
0.12%
1,001-4,100
0.57%
0.03%
0.57%
0.04%
TNCWS
GW
<=100
4.78%
0.20%
2.18%
0.19%
101-500
4.57%
0.19%
2.18%
0.19%
501-1,000
4.30%
0.06%
3.92%
0.20%
1,001-4,100
2.02%
0.03%
2.14%
0.13%
SW
<=100
2.23%
0.25%
2.19%
0.28%
101-500
2.63%
0.46%
2.18%
0.28%
501-1,000
3.63%
0.00%
2.19%
0.28%
1,001-4,100
0.61%
0.00%
1.98%
0.26%
NTNCWS
GW
<=100
3.17%
0.07%
2.89%
0.14%
101-500
2.70%
0.06%
1.99%
0.10%
501-1,000
1.74%
0.02%
1.90%
0.10%
1,001-4,100
1.35%
0.04%
1.46%
0.09%
SW
<=100
1.92%
0.24%
1.42%
0.18%
101-500
0.39%
0.09%
1.05%
0.13%
501-1,000
0.64%
0.00%
1.08%
0.14%
1,001-4,100
0.08%
0.00%
1.03%
0.13%
Source: Six Year Review data from Exhibit 4.9. Current TCR data from model output.
The comparison of the predictive model output with the Six-Year data for TC and EC
occurrence levels (based on number of TC routine samples taken) shows a reasonable
concordance for all types and sizes of systems. With respect to the differences, the model does
not appear to be either systematically overestimating or underestimating the occurrence levels
compared with the Six-Year Review data (except where the Six-Year Review data show 0%
observed, the model does show a low level of occurrence).
Exhibit 5.26 provides a comparison of the model results of the average annual nonacute
and acute violations for the current TCR with the same metric from SDWIS data for 2005 3Q.
(Violations data are downloaded from SDWIS on an annual basis in 3rd quarter only, and are not
disaggregated by month.)
Economic Analysis for the Proposed RTCR 5-4 7
June 2010
-------
Exhibit 5.26 Comparison of SDWIS Data for Nonacute and Acute Violations with
Predictive Model Annual Results for the Current TCR
Non-Acute Violations
Acute Violations
GW
SW
GW
SW
SDWIS
No GWR
(Avg)
SDWIS
No GWR
(Avg)
SDWIS
No GWR
(Avg)
SDWIS
No GWR
(Avg)
CWSs
<100
905
955
16
31
52
97
3
7
101-500
809
733
50
34
34
81
7
8
501-1,000
203
191
16
19
13
19
3
5
1,001-4,100
338
347
83
49
12
35
6
7
Totals
2,255
2,226
165
133
111
231
19
27
NTNCWSs
<100
514
560
7
7
34
60
2
2
101-500
346
265
4
6
20
30
-
1
501-1,000
57
61
2
2
6
7
-
0
1,001-4,100
62
98
1
3
6
12
-
1
Totals
979
985
14
18
66
109
2
4
TNCWSs
<100
2,665
4,109
19
81
278
445
5
19
101-500
833
1,371
11
31
76
156
1
8
501-1,000
133
144
4
5
11
16
-
1
1,001-4,100
58
105
2
9
2
13
-
2
Totals
3,689
5,730
36
126
367
630
6
31
Grand Total
6,923
8,940
215
277
544
970
27
61
Source: SDWIS 2005 Q3 download and occurrence and predictive model results.
The comparison of the predictive model output with the SDWIS violations data also
shows a reasonable concordance for all types and sizes of systems. EPA believes there is some
under-counting of actual violations in SDWIS due to monitoring and reporting violations, which
the model does not incorporate. Therefore, EPA generally expects that the model would generate
results that are higher than the observed SDWIS data. While a comparison reveals that some of
the model estimates are lower than the SDWIS statistics, the grand totals for the model are
consistently higher than those observed in SDWIS.
5.4 Occurrence Analysis for Systems Serving More Than 4,100 People
Systems serving populations greater than 4,100 are similar in many ways. In particular,
these systems generally have the resources and are managed in ways that lead to generally high
expectations for the integrity of the distribution systems and the ability of the systems to identify
and correct problems. Some of these differences include:
• Most of these systems operate with certified operators, are staffed at all operating times,
and usually operate continuously.
• Most of these systems have operators for whom producing water is their primary activity,
and most are owned by communities rather than run as ancillary businesses.
Economic Analysis for the Proposed RTCR 5-48
June 2010
-------
• As a result of the professional operations, these systems have a much lower incidence of
monitoring and reporting violations.
• Many of these systems have their own laboratory for analyzing samples.
• Many more of these systems disinfect their water.
• These systems take at least 5 samples per month. The largest of these systems take
hundreds of samples per month. There is no requirement for "additional next month
samples" as there is for smaller systems because these larger systems already take at least
5 samples each month.
• The proposed monitoring changes to the TCR mostly do not affect systems serving more
than 4,100 people; all take at least 5 samples per month.
For systems serving more than 4,100 people, EPA assumes that occurrence may change
based on the extra distribution system awareness created by: a) applying the Level 1 and Level 2
assessments in lieu of prior assessments that may, in some cases, have been less structured, and
b) reporting the assessment results. Therefore, EPA developed a simple model to predict the
effects of the AIP and Alternative options on PWSs and to compare those data to the baseline
data predicted for the current TCR. EPA did not quantify changes in violation or trigger rates for
systems serving more than 4,100 people among the current TCR, AIP option, and Alternative
options because of: (1) limited Six-Year Review data to characterize these systems, (2) the
essentially unchanged monitoring requirements across options for these systems, and (3) the
level of effort already occurring to implement the current TCR.
5.4.1 Model
As input for this model, EPA first considered 2005 TCR data on sampling and positive
assays that were compiled under the Six-Year Review, as was incorporated into the model for
smaller systems. However, as discussed in Section 4.2.2.3, detailed sampling results from the Six
Year Review data were not representative of the universe of larger systems. EPA instead found
that violations data were adequately representative of the universe of large systems. This input to
the model, based on 2005 SDWIS data (see Exhibit 4.10), was used to predict violations that
systems serving more than 4,100 people will incur in the 25-year period of AIP option analysis,
and the behavioral responses (Level 1 and Level 2 assessments and Corrective Actions) that
would result.
The model uses the same stratification as smaller systems - by system type (CWS,
TNCWS, or NTNCWS) and by water source (SW or GW). For each system in a given category
the model applies the 2005 SDWIS violations rate for that category. This rate is applied in each
year of the analysis, that is, the violations rate is a constant in the model. For estimating the
effects on violations under the AIP and Alternative option, the model estimates that 10 percent of
the assessments resulting from violations will result in corrective actions, the same value applied
in the analysis of the smaller systems. This 10% represents the incremental increase in efficacy
of addressing the root cause of system contamination under either regulatory alternative for the
proposed RTCR as compared to the current TCR.
Economic Analysis for the Proposed RTCR 5-49
June 2010
-------
Repeat samples are reduced in the model under both regulatory alternatives from the
current TCR. They are calculated by applying the ratios of repeat samples to routine samples
from the smaller system category (serving 1,001 - 4,100 people) to the systems serving more
than 4,100 people.
The model output, shown in Section 5.4.2, includes predictions of the annual number of
violations (non-acute = Level 1, acute = Level 2), and the number of Level 1 and Level 2
assessments and corrective actions to be implemented.
5.4.2 Model Results
Exhibits 5.27 through 5.29 present model results for systems serving more than 4,100
people by size category.
5.4.3 Model Uncertainty
As explained in Section 5.4.1 of this chapter, EPA does not expect systems serving
>4,100 people to experience changes in routine monitoring or repeat sample regimes under the
regulatory alternatives; furthermore, all systems in this size range are on monthly sampling.
Therefore, the larger systems model was relatively simple compared to that for the smaller
systems, and EPA did not develop an uncertainty analysis for this group of systems.
Economic Analysis for the Proposed RTCR 5-50
June 2010
-------
Exhibit 5.27 Results for Systems Serving more than 4,100 People - Current TCR
PWS Size
(Population
Served)
Num ber of
Routine
Monitoring
Sam pies
Num ber of
Additional
Routine
Monitoring
Sam pies
Num ber of
Repeat
Monitoring
Sam pies
PWSs
Performing
Additional
Annual Site
Inspections
Num ber of
No n-Acute
Violation
Assessments
(Single
Violations)
Nu m be r of
Corrective
Actions
(based on
Single Non-
Acute
Violation
Assessment)
Num ber of
Acute
Violation
Assessments
Nu m be r of
Non-Acute
Violation
Assessments
(Multiple
Violations)
Nu m be r of
Corrective
Actions
(based on
Acute and
Multiple Non-
Acute Violation
Assessm ents)
A
B
C
D
E
F
G
H
I
Comm unity Water Systems (CWSs) - SW
4,101-33,000
10,636,296
186,781
2,152
197
33,001-96,000
11,058,960
194,204
534
56
96,001-500,000
10,190,400
178,951
233
24
500,001-1 Million
2,019,600
35,466
22
> 1 Million
1,686,960
29,624
Comm unity Water Systems (CWSs) - GW
4,101-33,000
9,145,224
230,201
4,545
263
33,001-96,000
4,884,000
122,938
656
53
96,001-500,000
1,945,680
48,976
129
10
500,001-1 Million
253,440
6,380
> 1 Million
269,280
6,778
Nontransient Noncom m unity Water Systems (NTNCWSs) - SW
4,101-33,000
50,424
1,632
5
33,001-96,000
34,320
1,111
96,001-500,000
31,680
1,025
500,001-1 Million
> 1 Million
Nontransient Noncom m unity Water Systems (NTNCWSs) - GW
4,101-33,000
153,648
5,936
123
9
33,001-96,000
23,760
918
4
96,001-500,000
500,001-1 Million
> 1 Million
Transient Noncom m unity Water Systems (TNCWSs) - SW
4,101-33,000
40,656
8
33,001-96,000
96,001-500,000
500,001-1 Million
> 1 Million
102,960
Transient Noncom m unity Water Systems (TNCWSs) - GW
4,101-33,000
156,288
8,909
116
4
33,001-96,000
34,320
1,956
96,001-500,000
26,400
1,505
500,001-1 Million
63,360
3,612
> 1 Million
Note: Corrective actions output from the model are incremental additions to those already performed under the
current TCR (baseline conditions); therefore, columns F and I (corrective actions) are blank (zero) for the current
TCR.
Economic Analysis for the Proposed RTCR 5-51
June 2010
-------
Exhibit 5.28 Results for Systems Serving more than 4,100 People - AIP option
PWS Size
(Population
Served)
Num ber of
Routine
Monitoring
Sam pies
Num ber of
Additional
Routine
Monitoring
Sam pies
Num ber of
Repeat
Monitoring
Sam pies
PWSs
Perform ing
Additional
Annual Site
Inspections
Num ber of
Level 1
Assessm ents
Num ber of
Corrective
Actions (based
on Level 1
Assessments)
Num ber of
Level 2
Assessments
(based on
Acute
Violations)
Num ber of
Level 2
Assessments
(based on Non
Acute
Violations)
Num ber of
Corrective
Actions (based
on Level 2
Assessm ents)
A
B
C
D
E
F
G
H
I
Com m unity Water Systems (CWSs) - SW
4,101-33,000
10,636,296
-
175,689
-
2,152
215
197
-
20
33,001-96,000
11,058,960
-
182,671
-
534
53
56
-
6
96,001-500,000
10,190,400
-
168,324
-
233
23
24
-
2
500,001-1 Million
2,019,600
-
33,360
-
22
2
-
-
-
> 1 Million
1,686,960
-
27,865
-
-
-
-
-
-
Com m unity Water Systems (CWSs) - GW
4,101-33,000
9,145,224
-
217,321
-
4,545
454
263
-
26
33,001-96,000
4,884,000
-
116,060
-
656
66
53
-
5
96,001-500,000
1,945,680
-
46,236
-
129
13
10
-
1
500,001-1 Million
253,440
-
6,023
-
-
-
-
-
-
> 1 Million
269,280
-
6,399
-
-
-
-
-
-
Nontransient Noncom m unity Water Systems (NTNCWSs) - SW
4,101-33,000
50,424
-
1,395
-
5
0
-
-
-
33,001-96,000
34,320
-
950
-
-
-
-
-
-
96,001-500,000
31,680
-
877
-
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Million
-
-
-
-
-
-
-
-
-
Nontransient Noncom m unity Water Systems (NTNCWSs) - GW
4,101-33,000
153,648
-
5,157
-
123
12
9
-
1
33,001-96,000
23,760
-
797
-
4
0
-
-
-
96,001-500,000
-
-
-
-
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Million
-
-
-
-
-
-
-
-
-
Transient Noncom m unity Water Systems (TNCWSs) - SW
4,101-33,000
40,656
-
2,040
-
8
1
-
-
-
33,001-96,000
-
-
-
-
-
-
-
-
-
96,001-500,000
-
-
-
-
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Million
102,960
-
5,167
-
-
-
-
-
-
Transient Noncom m unity Water Systems (TNCWSs) - GW
4,101-33,000
156,288
-
7,188
-
116
12
4
-
0
33,001-96,000
34,320
-
1,578
-
-
-
-
-
-
96,001-500,000
26,400
-
1,214
-
-
-
-
-
-
500,001-1 Million
63,360
-
2,914
-
-
-
-
-
-
> 1 Million
-
-
-
-
-
-
-
-
-
Note: Estimates of the number of assessments and corrective actions are incremental increases in activity predicted
to occur under the AIP and Alternative option relative to the current TCR; estimates of "zero" reflect that no additional
such activity occurs as compared to baseline (the current TCR).
Economic Analysis for the Proposed RTCR 5-52
June 2010
-------
Exhibit 5.29 Results for Systems Serving more than 4,100 People - Alternative
option
PWS Size
(Population
Served)
Num ber of
Routine
Monitoring
Sam pies
Num ber of
Additional
Routine
Monitoring
Sam pies
Num ber of
Repeat
Monitoring
Sam pies
PWSs
Perform ing
Additional
Annual Site
Inspections
Num ber of
Level 1
Assessments
Num ber of
Corrective
Actions (based
on Level 1
Assessments)
Num ber of
Level 2
Assessm ents
(based on
Acute
Violations)
Num ber of
Level 2
Assessments
(based on Non
Acute
Violations)
Num ber of
Corrective
Actions (based
on Level 2
Assessments)
A
B
C
D
E
F
G
H
I
Com m unity Water System s (CWSs) - SW
4,101-33,000
10,636,296
175,689
2,152
215
197
20
33,001-96,000
11,058,960
182,671
534
53
56
6
96,001-500,000
10,190,400
168,324
233
23
24
2
500,001-1 Million
2,019,600
33,360
22
2
> 1 Million
1,686,960
27,865
Com m unity Water System s (CWSs) - GW
4,101-33,000
9,145,224
217,321
4,545
454
263
26
33,001-96,000
4,884,000
116,060
656
66
53
5
96,001-500,000
1,945,680
46,236
129
13
10
1
500,001-1 Million
253,440
6,023
> 1 Mllion
269,280
6,399
Nontransient Noncom m unity Water Systems (NTNCWSs) - SW
4,101-33,000
50,424
1,395
5
0
33,001-96,000
34,320
950
96,001-500,000
31,680
877
500,001-1 Million
> 1 Million
Nontransient Noncom m unity Water Systems (NTNCWSs) - GW
4,101-33,000
153,648
5,157
123
12
9
1
33,001-96,000
23,760
797
4
0
96,001-500,000
500,001-1 Million
> 1 Million
Transient Noncom m unity Water Systems (TNCWSs) - SW
4,101-33,000
40,656
2,040
8
1
33,001-96,000
96,001-500,000
500,001-1 Million
> 1 Mllion
102,960
5,167
Transient Noncom m unity Water System s (TNCWSs) - GW
4,101-33,000
156,288
7,188
116
12
4
0
33,001-96,000
34,320
1,578
96,001-500,000
26,400
1,214
500,001-1 Million
63,360
2,914
> 1 Million
Note: Estimates of the number of assessments and corrective actions are incremental increases in activity predicted
to occur under the AIP and Alternative option relative to the current TCR; estimates of "zero" reflect that no additional
such activity occurs as compared to baseline (the current TCR).
Economic Analysis for the Proposed RTCR 5-53
June 2010
-------
5.5 Summary of Key Drivers for Benefit and Cost Analyses Output from the
Predictive Model
As presented in Exhibits 5.10- 5.15, the model for small systems (serving <4,100 people)
produces estimates of the number of samples, the number of Level 1 and Level 2 assessments,
and the number of corrective actions implemented as a result of Level 1 and Level 2 assessments
under each regulatory scenario. For the AIP and Alternative options, the model calculates the
CAs performed as a net increase from those performed under the current TCR. Those being
performed under the current TCR are described in Chapter 7 in the discussion of current TCR
costs. The net change in each of the model outputs, in particular sampling regimens and
implementation of assessments and corrective actions, relative to the baseline (current TCR)
under both alternative regulations considered provides the basis for estimating the benefits in the
rule, described in Chapter 6 of this EA. Similarly, the cost model applies unit costs to the
activities described by these outputs to produce net costs for each regulatory scenario, as
described in Chapter 7.
As described in the uncertainty discussion and sensitivity analysis presented in Section
5.3.3.1, these key drivers represent the most significant source of uncertainty in the analysis.
However, the sensitivity analysis showed that a 100% increase in the implementation and
effectiveness assumptions for CAs (the proportional increase in the number of effective
corrective actions implemented and the duration and extent of reduced occurrence) implemented
under the AIP option would result in only approximately a 30% decrease in acute events.
Conversely, a 50% reduction in CA implementation and effectiveness was estimated to induce
less than an 18% increase in acute events. These results are expected to apply similarly to all
categories of systems. Based on these results, EPA concludes that changing the assumptions
within the likely range (a doubling or halving of current assumptions) would not have a
significant effect on the conclusions drawn from this EA. Similarly, varying the assumptions by
size and type of system within a likely range would not be expected to have a significant effect
on the results.
For larger systems (serving >4,100 people), incremental costs are estimated for the
additional reporting activity required under the rule, as well as a slight increase in the number of
corrective actions implemented for systems serving more than 4,100 people. As described in
Section 5.4, however, the sampling regimens are not expected to change for these systems, and
they'll experience minimal changes in risk and cost, as described in Chapter 6 and Chapter 7 of
this EA, respectively.
Economic Analysis for the Proposed RTCR 5-54
June 2010
-------
6. Benefits Analysis
6.1 Introduction
This chapter considers the overall change in risk of contamination to Public Water
Systems (PWSs), as indicated by the presence of total coliforms (TC) and E. coli, associated with
compliance with the current Total Coliform Rule (current TCR option) and with the proposed
Revised Total Coliform Rule (RTCR) (Agreement in Principle or AIP) and the Alternative
option. Since E. coli is an indicator of fecal contamination, EPA assumed that a decrease in E.
coli occurrence in the distribution system would be associated with a decrease in fecal
contamination in the distribution system. In general, this decrease in fecal contamination should
reduce the potential risk to human health for PWS customers. Thus, any reduction in E. coli
occurrence is considered a benefit of the proposed Revised Total Coliform Rule (RTCR). Also,
since fecal contamination may contain waterborne pathogens including bacteria, viruses, and
parasitic protozoa, in general, a reduction in fecal contamination should also reduce the risk from
these other contaminants.
Based on limitations in available data as described further in Section 6.3 of this chapter,
EPA determined that benefits could not be calculated in terms of avoided costs or other
quantified benefits related to avoided morbidity or mortality. Therefore, this Economic Analysis
(EA) focuses on a qualitative analysis of risk, which is supported by quantitative analysis of
predicted outcomes for each regulatory option. The quantitative discussion focuses on net
changes in occurrence of contaminant indicators under both regulatory options considered as
compared to the current TCR option. The qualitative analysis considers the direction of
anticipated changes in risk related to changes in sampling and corrective action regimens under
each regulatory scenario (Section 6.2). EPA considered the results of the qualitative and
quantified assessments to determine how the current TCR option compares to the AIP and
Alternative option in terms of overall change in risk to the population served by PWSs across the
United States.
The remainder of this chapter is organized as follows:
6.2 Qualitative Benefits Analyses
6.3 Assessment of Predictive Analysis Results
6.4 Uncertainty Sensitivity Analyses
6.5 Other Potential Benefits
Economic Analysis for the Proposed RTCR
6-1
June 2010
-------
6.2 Qualitative Benefits Analyses
When revising an existing drinking water regulation, one of the main concerns is to
ensure that backsliding on water quality and public health protection does not occur. Risk
reduction for the AIP is characterized by the activities performed that are presumed to reduce
risk of exposing the public to contaminated water. These activities are considered under each
rule component discussed in Sections 6.2.1 - 6.2.8 and summarized in Figure 6.1 below. The
discussion in Section 6.2.9 considers how expected overall decreases in this risk for some rule
components are offset by expected overall increases resulting from other components, and
provides a net assessment of the direction in change of risk for the regulatory options considered.
The qualitative analysis uses the best professional judgment of EPA as informed by Total
Coliform Rule Distribution System Advisory Committee (TCRDSAC) deliberations1, as well as
certain quantitative estimates, to estimate the directional change in risk for each regulatory
component of the AIP and the Alternative option. Quantitative estimates considered include the
changes in total coliform occurrence and counts of systems conducting assessments and
implementing corrective actions shown in Exhibits 6.2 - 6.6 below. Exhibit 6.1 below presents a
summary of this evaluation for the AIP and Alternative option as compared to the current TCR
option. The qualitative analysis discusses the influence of the individual regulatory components
under each regulatory option considered on the occurrence of Total Coliform-positive (TC+) and
E. 6'o//-positive (EC+) samples. Since a dose-response relationship between exposure to the fecal
indicator E. coli and adverse health effects from waterborne pathogens that can be present with
fecal contamination is not adequately quantified, the resulting risk to human health is discussed
only in terms of the anticipated change in direction of risk.
6.2.1 Implementation Activities
Rule implementation activities are expected to be similar under the AIP and Alternative
option. These activities are primarily administrative in nature and include items such as reading
and understanding the rule, training, and development of reporting and recordkeeping protocols.
Because of the similarities in expected implementation activities under the regulatory options,
they are not expected to have an observably different affect on overall risk relative to current
regulations. Both PWSs and states would incur additional burden and costs for transitioning to
operations under the AIP or Alternative option requirements. Because the activities undertaken
to make the transition are primarily administrative, the additional activities are not expected to
have any direct impact on risk.
6.2.2 Routine Monitoring
EPA expects that more frequent monitoring would decrease the risk of contamination in
PWSs based on an enhanced ability to diagnose and mitigate system issues in a more timely
fashion. Conversely, EPA assumes that a less frequent monitoring schedule would result in
increased risk. Real-time continuous sampling would mitigate the most risk possible based on
1 TCRDSAC deliberations are described in Chapter 3 of this EA.
Economic Analysis for the Proposed RTCR
6-2
June 2010
-------
sampling schedule; however, it would cost prohibitively more than the periodic sampling
practiced under the current TCR and included in the AIP and the Alternative option. EPA's
objective in proposing the sampling schedules included in the AIP and Alternative option was to
find an appropriate balance between the factors of risk mitigation and cost management.
Current TCR Option
The current TCR option requires that a large portion (approximately 40%) of water
systems (including all ground water and surface water systems) to perform monthly monitoring
for total coliforms. Some systems are eligible for reduced monitoring, as follows.
• Ground water Noncommunity Water Systems (NCWSs) serving <1,000 people are
required to monitor quarterly. States may allow annual monitoring if a sanitary
survey conducted in the past five years shows the system is supplied solely by
protected Ground Water (GW) sources and is free of sanitary defects.
• Ground water Community Water Systems (CWSs) serving <1,000 people may
monitor quarterly if they have no history of total coliform contamination in their
current configurations and a sanitary survey conducted in the past five years
shows that the system is supplied solely by a protected GW source and has no
sanitary defects.
AIP
The eligibility requirements for ground water systems to qualify for reduced monitoring
under the AIP are more stringent than under the current TCR option, leading to a higher total
number of routine samples being taken over the 26-year period of analysis. The primacy agency
has the discretion to reduce monitoring frequency for well operated ground water NCWSs and
ground water CWSs serving <1,000 people. Under the AIP, eligibility for reduced monitoring for
NCWSs and CWSs, respectively, is discussed as follows:
NCWSs
To be eligible to qualify for and remain on annual monitoring after the compliance
effective date, noncommunity ground water systems serving <1,000 people must meet each of
the following criteria:
• The most recent sanitary survey shows the system is free of sanitary defects and has a
protected water source and meets approved construction standards;
• The system must have a clean (TCR) compliance history (no MCL violations, Level 1
triggers, Level 2 triggers, treatment technique violations or monitoring violations) for
a minimum of 12 months;
• An annual site visit (recurring) by the primacy agency within the last 12 months and
correction of all identified sanitary defects. A voluntary Level 2 assessment by a party
approved by the primacy agency may be substituted for the primacy agency annual
site visit; and
Economic Analysis for the Proposed RTCR
6-3
June 2010
-------
• The primacy agency should encourage additional enhancements to the barriers
protecting the distribution system from contamination. These measures could include
but are not limited to the following:
Cross connection control, as approved by the primacy agency;
An operator certified by an appropriate primacy agency certification program,
which may include regular visits by a circuit rider;
Continuous disinfection entering the distribution system and a residual in the
distribution system in accordance with criteria specified by the primacy
agency; and
Maintenance of at least a 4-log inactivation of viruses each day of the month
based on daily monitoring as specified in the Ground Water Rule (GWR)
(with allowance for a 4-hour exception).
Other equivalent enhancements to water system barriers as approved by the
primacy agency.
CWSs
To be eligible for quarterly reduced monitoring, community ground water systems
serving <1,000 people on monthly monitoring after the compliance effective date must be in
compliance with primacy agency certified operator provisions and meet each of the following
criteria:
• The most recent sanitary survey shows the system is free of sanitary defects (or has
an approved plan and schedule to correct them), has a protected water source, and
meets approved construction standards;
• The system must have a clean (TCR) compliance history (no MCL violations, Level 1
or Level 2 triggers, treatment technique violations or monitoring violations) for a
minimum of 12 months; and
• Meet at least one of the following criteria:
An annual site visit by the primacy agency or a voluntary Level 2 assessment
by a party approved by the primacy agency and correction of all identified
sanitary defects (or an approved plan and schedule to correct them), or
A cross connection control program, as approved by the primacy agency, or
The system must maintain continuous disinfection entering the distribution
system and a residual in the distribution system in accordance with criteria
specified by the primacy agency, or
The system must maintain at least a 4-log inactivation of viruses each day of
the month based on daily monitoring as specified in the GWR (with allowance
for a 4-hour exception); or
Economic Analysis for the Proposed RTCR
6-4
June 2010
-------
Other equivalent enhancements to water systems as approved by the primacy
agency.
Based on the additional protection provided by the more stringent criteria to qualify for
reduced monitoring, the AIP is expected to reduce risk for Ground Water Systems (GWSs) that
qualify for reduced monitoring compared to the eligibility requirements under the current TCR
option. In addition, those systems that no longer qualify for reduced monitoring are expected to
have a reduced risk as a function of the increased numbers of samples taken.
Seasonal systems (noncommunity water systems that are operating three or fewer
quarters per calendar year) that sample on a reduced schedule under the current TCR option may
retain their current schedules unless a state determines that more frequent sampling is
appropriate. For seasonal systems required to move to an increased monitoring schedule, risk
would decrease as a function of the increase in monitoring. In addition, more explicit procedural
requirements related to monitoring under the AIP option (i.e., samples must be taken during the
period of highest vulnerability or peak usage) would also be expected to reduce risk for seasonal
systems.
Alternative option
Under the Alternative option, all systems must initially collect total coliform samples
monthly regardless of size or type (i.e., water source), which is more frequent for those systems
on reduced monitoring schedules under the current TCR option or AIP option. Over time, in the
second year of rule implementation and beyond, the Alternative option would allow some ground
water PWSs to reduce to quarterly monitoring if the systems meet the same qualifications
required for quarterly monitoring under the AIP option. Reduced monitoring on an annual
schedule is not allowed under the Alternative option, creating a further increase in samples taken
over the current TCR option and AIP option. Overall, the more frequent monitoring requirements
(i.e., all PWSs monitor monthly in the first few years after promulgation and no annual reduced
monitoring is allowed) would reduce risk as compared to the current TCR option to a greater
extent than under the AIP option.
6.2.3 Repeat Monitoring
Under the current TCR option, PWS serving <1,000 people take 4 repeat samples at and
within five service connections upstream and downstream of the initial total coliform positive
(TC+) occurrence event over the course of 24 hours following the event. Three repeat samples
are required for PWSs serving >1,000 people, including one sample at the site of the initial TC+
and two additional samples within five service connections up or downstream of that site.
Under the AIP option and Alternative option, PWSs serving <1,000 people are only
required to take three repeat samples. The reduction in the number of repeat samples required
could increase risk for any individual sampling event. However, the effect of the reduced
sampling is expected to be minor. Analysis of the data (see Appendix H for the complete
analysis) shows that for all PWSs serving 1,000 people or fewer, two or more of the repeat
Economic Analysis for the Proposed RTCR
6-5
June 2010
-------
samples are positive in 75 percent of those instances in which there are any positive repeat
samples. For those 75 percent of instances, reducing the number of repeat samples from four to
three would have no effect on the number of systems that would be triggered to conduct an
assessment of the system under the AIP option. In these cases, at least one of the remaining
repeat samples would still be total coliform-positive, and only one positive repeat sample is
required to trigger an assessment.
The data show that one repeat sample is positive in 25 percent of the instances in which
any of the four repeat samples is positive. For these instances, EPA estimates that if only three
repeat samples had been taken instead of four, three out of four (or 75 percent) of these positive
samples would still have been encountered. The analysis indicates that if the number of required
repeats were reduced from four to three, there would still be almost as many (approximately 94
percent) situations leading to an assessment being triggered for the system (0.75 + 0.75*0.25 =
0.94).
Although dropping the required number of repeat samples from four to three means that
some fraction of triggers may be missed representing an increase in risk, EPA believes that the
other provisions of the AIP option as described in this chapter compensate for that change and
that, taken as a whole, the provisions of the AIP option provide for greater protection of public
health.
6.2.4 Additional Routine Monitoring
Under the current TCR option, PWSs serving <4,100 people must conduct additional
routine monitoring in the month following a TC+ sample. All systems must collect and test a
minimum of five samples in the month following the TC+ sample, unless the primacy agency
finds that additional sampling is unnecessary or the primacy agency determines the cause of the
TC+ sample and establishes that the system has corrected or would correct the problem.
EPA is proposing to retain the requirement of taking additional routine samples the
month following a TC+ sample for systems on quarterly or annual monitoring. The advisory
committee recognized both the benefits and the limitations of additional routine monitoring.
Under the AIP option, a system that has a Level 1 trigger must conduct an assessment, and if a
problem is found, the system must take corrective action. Under such circumstances, the
advisory committee believed that additional samples collected the following month are
appropriate to help to recognize the problem if it still persists. Without the provision of
additional monitoring, systems on annual or quarterly monitoring would not take any samples the
following month. Systems having a Level 2 assessment are triggered into a monthly monitoring
and therefore have less need for additional routine monitoring to indicate if a problem persists.
For systems required to take the additional routine samples the following month (i.e.,
systems on quarterly or annual monitoring), the AIP option changes the requirement from taking
a total of five routine samples to a requirement of just three routine samples. The advisory
committee recognized that it is appropriate to drop from five to three samples the following
month to reduce monitoring costs while still maintaining a substantial likelihood of identifying a
Economic Analysis for the Proposed RTCR
6-6
June 2010
-------
problem if a problem persists. EPA recognizes that a reduction in the number of samples taken
could also mean a reduction in the number of positive samples found. However, the reduction in
the number of additional routine samples in conjunction with the new assessment and corrective
action provisions of the proposed RTCR (discussed in section III.A.5 of the proposed RTCR
preamble (USEPA, 2010c) leads to a rule that is ultimately more protective of public health (i.e.,
more E. coli MCL violations being prevented) and improvement in water quality (i.e., decrease
in the total coliform and E. co/z'-positive hit rates observed as shown by EA occurrence modeling
results).
For systems taking at least one sample monthly, the advisory committee recommended
no additional routine samples for these systems for the following reason. Taking no additional
routine samples the following month substantially reduces monitoring costs. The assessment and
corrective action provisions will give systems the ability to identify and prevent the occurrence
of problems. EA modeling results show that although there is a decrease in the number of E. coli
MCL violations found with the decrease in the number of additional routine samples taken (i.e.,
going from five samples to one during the month following a TC+ sample), the assessment and
corrective action provisions leads to more E. coli MCL violations being prevented compared to
the current TCR (see Exhibit 6.7 for more details).
In addition, whenever a TC+ sample occurs during routine sampling, there is also a
requirement to conduct repeat sampling to clarify if potential pathways to contamination persist.
For small systems serving 1,000 people or fewer, if a repeat sample is TC+, at least a Level 1
assessment will be triggered. If a sanitary defect(s) is (are) found, the system is required to
correct the sanitary defect(s). The absence of any repeat positive sample provides some
indication that the problem is not persisting. For systems on monthly monitoring, these two
conditions mitigate the need for additional routine sampling for the following month.
Although the changes to the additional routine monitoring provisions mean that some
fraction of triggers may be missed representing an increase in risk, EPA believes that the other
provisions of the proposed RTCR as described in this chapter compensate for that change and
that, taken as a whole, the provisions of the RTCR provide for greater protection of public
health.
6.2.5 Annual Site Inspections
The current TCR option does not include any requirements for annual site inspections.
However, based on discussions with stakeholders, some states do perform annual site visits for
any systems on an annual sampling schedule.
Under the AIP, ground water NCWSs serving <1,000 people must, within one year of the
compliance effective date, have an initial (and annually thereafter) visit by the state or an annual
voluntary Level 2 assessment by a party approved by the state to remain on an annual monitoring
schedule. Because of the cost differential between conducting annual site inspections and the
alternative (quarterly monitoring), EPA has estimated that only those states that already
voluntarily conduct annual site inspections under the current TCR option would also do so under
Economic Analysis for the Proposed RTCR
6-7
June 2010
-------
the AIP. Therefore, no risk reduction is expected for these systems (and thus overall) from this
regulatory component under the AIP.
The Alternative option does not allow systems to reduce to a frequency of annual
monitoring and therefore does not include an annual site inspection requirement. However, based
on discussions with stakeholders, those states that currently conduct annual site assessments
under the current TCR option may no longer have the resources to continue the inspections and
conduct quarterly monitoring under the Alternative option. For NCWSs on annual monitoring,
the TCRDSAC believed that requiring a system to have an annual site visit or a Level 2
assessment provides at least an equivalent level of diagnosis of problems and vulnerabilities that
might exist as compared to quarterly monitoring without an annual site visit. This trade-off
between annual monitoring with site inspections and three additional routine samples (i.e.,
quarterly monitoring with no site inspections) would potentially result in an increased risk for the
Alternative option compared to the AIP.
6.2.6 Assessments
Under the current TCR option, although there is no explicit "assessment" required, some
level of investigation is necessary to evaluate the cause of a problem. In particular, a system must
investigate a problem and is required to notify the public and state under different scenarios.
Specifically, a PWS must:
• Report any acute MCL violation to the state no later than the end of the business after
the system learns of a violation;
• Notify the public within 24 hours of an acute MCL violation.
• Notify the public within 30 days of a monthly/non-acute MCL violation;
Under the AIP and Alternative option2, PWSs are required to perform and submit a
Level 1 assessment if:
• Systems taking 40 or more samples per month have more than 5.0% total coliform
positive samples;
• Systems taking less than 40 samples per month have two or more total coliform
positive samples in one month; or
• A system fails to take all required repeat samples after a single total coliform positive
sample.
2 Requirements for the current TCR, proposed RTCR, and Alternative option are described in detail in Chapter 3 of
this EA.
Economic Analysis for the Proposed RTCR
6-8
June 2010
-------
A more detailed examination of the system, including its monitoring and operational
practices (a Level 2 assessment), is required if a system has:
• An E. coli MCL violation;
• An E. coli monitoring violation; or
• A second Level 1 trigger within a rolling 12 month period.
Mandatory assessments are a new requirement under the AIP and Alternative option, and
also represent an increased focus on problem solving from the less defined investigations
conducted under the current TCR option. Because of the more explicit requirements of the
assessments, it is expected that more problems would be identified and resolved. As a result, the
risk relative to the current TCR option is assumed to decrease.
6.2.7 Corrective Actions
Corrective actions are not explicitly required under the current TCR option. However,
systems do perform corrective actions based on investigations they perform (see Section 6.2.6) in
response to positive samples.
Under the AIP and Alternative option, PWSs are responsible for correcting any sanitary
defects found during either a Level 1 or Level 2 assessment. For corrections not completed by
the time of submission of the assessment form, systems must complete the corrective action(s)
on a schedule determined by the state in consultation with the system. Systems would be
required to notify the state when they have completed each corrective action. Failure to
implement a corrective action is considered a treatment technique violation, subject to public
notification.
EPA does not have data on the existing rates at which corrective actions are taken under
the current TCR option, so an assumption of the incremental percentage of assessments resulting
in corrective action (10 percent)3 is made as part of this EA and subsequent evaluation of
changes in risk. Overall, increased protection provided by this net increase in corrective action as
a result of requiring systems to implement a correction action through an enforceable mechanism
would reduce risk under both the AIP and Alternative option.
3 The 10 percent assumption is based on EPA discussions with stakeholders regarding experiences with
implementing the current TCR and the expected impact of proposed RTCR revisions. A sensitivity analysis
evaluating alternative assumptions was conducted. Results of these analyses are discussed in Chapter 5 (section
5.3.3.1).
Economic Analysis for the Proposed RTCR
6-9
June 2010
-------
6.2.8 Public Notification
Monthly/Non-acute MCL Violations
The current TCR option requires public notification within 30 days of a monthly/non-
acute MCL violation or 24 hours of an acute MCL violation.
Both the AIP and Alternative option would require:
• Tier 1 public notification within 24 hours of an E. coli MCL violation;
• Tier 2 public notification within 30 days of a treatment technique (TT) violation; and;
• Tier 3 public notification within a year following either a routine monitoring violation
or a reporting violation.
However, under either option, the public notification requirements for monthly/non-acute MCL
violations would no longer be required (the AIP and Alternative options do not include
monthly/non-acute MCL violations). Since monthly/non-acute violations account for a large
number of violations under the current TCR option, there is expected to be a large decrease in the
number of notices presented to the public. If it is assumed that such notices provide information
that aid in risk avoidance by water customers, risk may increase as a result of reducing this
public notification requirement to the extent that a risk exists. Because PWSs are no longer
required to send out notices for monthly/non-acute MCL violations or the monitoring results4
that triggered them under the current TCR, there is also the potential that some PWSs may
become less responsive in addressing TC hits or preventing them from occurring in the first
place.
The TCRDSAC evaluation of the current TCR option public notification also concluded
that the numbers of monthly/non-acute MCL violation notices that go out to the public are
confusing. Unlike acute MCL violations associated with positive E. coli samples, monthly/non-
acute MCL violations may have no connection to a direct health risk because TC+ samples do
not indicate a direct threat to public health. To the extent that a high number of notices issued for
monthly/non-acute MCL violations result in false alarms, consumers may not appreciate a notice
of an acute MCL violation. Therefore, the EPA concluded, using best professional judgment
informed by TCRDSAC's evaluation, that risk may decrease through the elimination of the
Public Notification (PN) requirement for monthly/non-acute MCL violations under the AIP and
Alternative option. Additionally, resources used to issue high numbers of monthly/non-acute
MCL violations and (as well as the time spent responding to customer inquiries about the
violations) may be better employed on other PWS issues which could result in further reduced
risk.
As discussed above, the influences on risk of eliminating the PN requirements for
monthly/non-acute MCL violations may move risk in both directions. A decrease in the overall
4 Under the current TCR option, for systems taking fewer than 40 samples per month, an acute violation occurs if 2
or more samples are TC+; for systems taking 40 or more samples per month, an acute violation is triggered by >5%
of TC+ samples.
Economic Analysis for the Proposed RTCR
6-10
June 2010
-------
information received by consumers may result in reduced ability to undertake averting behaviors
necessary to avoid potentially contaminated drinking water when appropriate, increasing
potential risk. Conversely, overly abundant notices may cause confusion or indifference that may
result in consumers not taking averting actions when appropriate (i.e., ignoring an acute violation
notice). Thus focusing on fewer, yet more "serious" notices may result in a decrease in potential
risk. Risk may also decrease as a result of PWSs being able to better employ resources currently
used on notice issuance and follow-up.
EPA also considered the effect of Tier 2 PN requirements for TT violations, which allow
for up to a 30-day time delay between incurrence of the violation and notification to the public.
This time delay could equate to some increased risk for the public relative to a scenario where an
immediate notification was required. However, the change in risk relative to the current TCR
option, which does not require assessments or corrective actions, is a decrease in risk. In
allowing PWSs the flexibility to take up to a month before issuing PN, EPA is attempting to
balance the potential public health benefits associated with having prompt notification for all TT
violations with the potential inefficiency and extra costs related to cases where the same level of
expedience may not be necessary.
In summary, EPA assumes that there would not be an overall change in risk by changing
PN requirements for monthly/non-acute MCL violations under the proposed RTCR, but given
the contradictions in potential risk reductions the potential change in risk is best characterized as
"unknown."
Monitoring and Reporting Violations
The TCRDSAC also evaluated the effects of revised PN requirements for monitoring and
reporting violations under the AIP and Alternative option and concluded that significant
reductions in monitoring and reporting violations and associated costs may be realized through
the revised regulatory framework of the AIP. For the current TCR option, monitoring and
reporting violations result in Tier 3 PN. Under the AIP and Alternative option, monitoring
violations are separate from the reporting violations, and the monitoring PN requirements are
more stringent as compared to the monitoring and reporting violations under the current TCR.
For repeat sampling, a PWS that fails to take every required repeat sample following a
routine EC+ sample must:
• Initiate Tier 1 PN within 24 hours;
• Initiate consultation with the state no later than 24 hours after learning of the
violation, to determine additional PN requirements, if any;
• Perform a level 2 assessment/corrective action: and
• Increase to a minimum of monthly monitoring.
For a PWS that fails to take every required repeat sample after any single TC+ sample, the PWS
must perform a level 1 assessment/corrective action. Failure to perform a required assessment
Economic Analysis for the Proposed RTCR
6-11
June 2010
-------
and/or corrective action results in a treatment technique violation (Tier 2) and a minimum of
monthly monitoring.
For routine and additional routine monitoring, a PWS that does not take every required
routine sample, or every required additional routine sample, in a compliance period is still
subject to Tier 3 PN; noncompliance with sampling requirements does not necessarily increase
the likelihood that a PWS will be contaminated with fecal contamination and/or a waterborne
pathogen and so does not present a direct or immediate public health risk. However, if the PWS
has monitoring violations in 2 of 4 quarters (for systems on quarterly monitoring) or misses its
required annual sample (for systems on annual monitoring), the PWS must revert to monitoring
no less than monthly.
Overall, the added PN stringency for monitoring violations is expected to decrease
potential risk under the AIP and Alternative option as PWSs opt to perform required sampling to
avoid transition to increased monitoring requirements or other additional actions.
Summary Exhibit
The component discussions in Sections 6.2.1 through 6.2.8 above address the individual
effects under each rule component on the various system types and sizes. The terms "increase,"
"decrease," and "no change" in Exhibit 6.1 indicate the direction of change in risk under the AIP
and Alternative option relative to the current TCR option. Risk may change for some system
sizes or types and not for others under a given rule component. In such cases, Exhibit 6.1 reflects
the overall change in risk direction and does not necessarily apply to all types and sizes of
systems.
Economic Analysis for the Proposed RTCR
6-12
June 2010
-------
Exhibit 6.1 Potential Changes in Risk under the AIP and Alternative Option Relative to the Current TCR
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
AIP option
Alternative
option
AIP option
Alternative
option
AIP option
Alternative
option
Implementation
Activities
None
None
None
None
No change
No change
Routine
Monitoring
(Including
Reduced
Monitoring)
None
None
Increased stringency
in requirements to
qualify for reduced
monitoring along with
requirement to return
to baseline
monitoring upon loss
of these criteria is
expected to result in
decreased risk (i.e.,
fewer PWSs will
qualify and therefore
more will monitor
more frequently).
PWSs all monitor
monthly in the first few
years of implementation
of the RTCR, which is an
increase in sampling
frequency for systems
that monitor quarterly or
annually under the
current TCR. After the
first few years, systems
may reduce to quarterly,
but none may reduce to
annual monitoring,
creating a decrease in
risk for systems on
annual monitoring under
the current TCR.
Decrease
Decrease
Economic Analysis for the Proposed RTCR
6-13
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
Repeat
Monitoring
Required repeat
samples reduced
from 4 to 3 for
systems serving
<1,000 people
Required repeat
samples
reduced from 4
to 3 for systems
serving <1,000
people
None
None
Increase
Increase
Additional
Routine
Monitoring
Additional routine
samples are no
longer required for
PWSs monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional routine
samples from 5 to
3.
Additional
routine samples
are no longer
required for
PWSs
monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional
routine samples
from 5 to 3.
None
None
Increase
Increase
Annual
Inspections
None (only states
currently
performing annual
inspections are
expected to
continue)
Based on
discussions with
stakeholders,
states that
currently
conduct annual
site
assessments
under the
current TCR
may no longer
have the
None (only states
currently performing
annual inspections
are expected to
continue)
None
No change
Increase
Economic Analysis for the Proposed RTCR
6-14
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
resources to
continue the
inspections and
conduct
quarterly
monitoring
under the
Alternative
option.
Assessments
None
None
Mandatory
assessments are a
new requirement.
Mandatory assessments
are a new requirement.
Decrease
Decrease
Corrective
Actions
None
None
Mandatory corrective
actions are a new
requirement.
Mandatory corrective
actions are a new
requirement.
Decrease
Decrease
Public
Notification -
Monthly/Non-
Acute MCL
Violations
Reduction in
available public
information
Possible PWS
complacency
Reduction in
available public
information
Possible PWS
complacency
Less confusion (PN
more in line with
potential health risks)
PWS resources used
more efficiently
Less confusion (PN
more in line with
potential health risks)
PWS resources used
more efficiently
Unknown
Unknown
Economic Analysis for the Proposed RTCR
6-15
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
Public
Notification -
Monitoring and
Reporting
Violations
None
None
Increased stringency
of PNs motivates
PWSs to conduct
required sampling.
Increased stringency of
PNs motivates PWSs to
conduct required
sampling.
Decrease
Decrease
Overall
Decrease
Decrease
Note: Detailed discussion of the rationale for determinations of potential risk for each rule component is presented in Ch. 6 (Section 6.2) of this EA. Implementation
activities consist of administrative activities by PWSs and states to implement the rule.
Assessment of potential changes in risk for monitoring components is an overall assessment. Potential changes (or static state) of risk for particular system sizes
and types differ according to individual regulatory requirements and are discussed in Section 6.2. Chapter 3 provides a detailed description of the regulatory
components for all three regulatory scenarios, and the Preamble to the RTCR provides additional discussion of the TCRDSAC process and the rationale underlying
the structure of the regulatory options considered.
Economic Analysis for the Proposed RTCR
6-16
June 2010
-------
6.2.9 Summary of Qualitative Benefits Analyses
The discussions above describe relative risks in terms of individual components of the
regulatory options. Based on the discussions presented above, EPA has used best professional
judgment informed by the TCRDSAC to qualitatively estimate the relative risk of each
regulatory option. These assessments were made with contributions from a wide range of
drinking water experts, including public health scientists, engineers, administrators, and
regulatory experts. The overall change in risk relative to the current TCR baseline is a result of
the complex interactions of all regulatory components. As the discussion above shows, under
repeat and additional routine monitoring provisions for the AIP and Alternative option, there is a
potential for increased risk for PWS customers because total coliform monitoring frequency may
be reduced for some PWSs. However, this increase is expected to be more than offset by
potential decreases in risk from increased routine monitoring and the addition of the assessments
and corrective action provisions that will find and fix problems identified by monitoring.
The consensus opinion resulting from the TCRDSAC deliberations was that the proposed
RTCR, as described in the AIP, would achieve a net risk reduction compared to the current TCR
option. The committee applied best professional judgment in determining that the increased
protection provided by the new requirements for implementing focused assessments and
implementing appropriate corrective actions would more than offset any potential increase in risk
introduced by the reduction in samples and other changes resulting from the AIP. The committee
considered that the Alternative option presents potential for further reduction in risk due to the
increased numbers of samples taken, especially in the first several years of implementation.
However, the additional burden5 of requiring all PWSs to initially monitor on a monthly basis
(regardless of PWS size or type) and limiting reduced monitoring to quarterly (disallowing
annual monitoring) would fall disproportionately on small systems based on their proportionately
large increase in activity under the AIP.
6.3 Assessment of Predictive Analysis Results
Based on discussions in and information developed for the TCRDSAC meetings
(described in Chapter 3 of this EA), EPA anticipated prior to beginning this EA that the proposed
RTCR would not be a significant rule in terms of costs (i.e., less than $100 million annually).
However, EPA considered the feasibility of performing a traditional risk assessment that would
produce quantified estimates of costs, benefits, and net costs and benefits, as outlined in section
1412 (b)(3)(C) of the Safe Drinking Water Act (SDWA). For this type of analysis, the minimum
information requirements comprise data on contaminant occurrence; exposure rates in the
population defined for the various pathways (i.e., water consumption, inhalation, and dermal
contact); potential health effects associated with exposure to contaminated water; and a dose-
response relationship. A quantified benefits analysis would use this information to estimate the
number of avoided cases of morbidity or mortality associated with the rule, which would then be
valued in terms of saved lives and preserved quality of life and work capacity.
5 Additional costs under the Alternative option are discussed Chapter 7 of this EA.
Economic Analysis for the Proposed RTCR
6-17
June 2010
-------
For the proposed RTCR EA, the available information includes data on the incidence of
positive TC+ samples collected under the current TCR option over the 6 year period from 1998 -
2005, from which EPA selected one complete year (2005) to use for analyses in this EA, as
explained in Section 4.2 of this EA. For the year 2005, EPA also has data on TC+ samples that
subsequently test positive for E. coli (using EPA-approved standard methods available at that
time). No other national occurrence data are available.
As discussed in Chapter 2 of this EA, the presence of E. coli is an indication that the
water contains fecal contamination. SomeE. coli strains (or serotypes) such as E. coli 0157:H7,
for example, are pathogens. However, EPA recognizes that the EPA-approved standard methods
available for E. coli do not typically identify the presence of the pathogenic E. coli strains, such
as E. coli 0157:H7. Thus, E. coli occurrence, as used in this EA, serves as an indication of fecal
contamination but not necessarily pathogenic contamination. E. coli occurrence does not confer
any significant quantitative information about the likelihood of health effects (e.g. acute
gastrointestinal illness or chronic illnesses as described in Chapter 2 of this EA) from consuming
drinking water contaminated with fecal indicator organisms.
There are few data reporting the co-occurrence in a single sample of fecal indicator E.
coli (assayed using EPA-approved standard methods) and pathogenic E. coli strains. One notable
exception are the data reported by Cooley et al. (2007), which showed high concentrations of
pathogenic E. coli strains in samples containing high concentrations of fecal indictor E. coli.
These data are from streams and other poor quality surface waters surrounding California
spinach fields associated with the recent E. coli 0157:H7 foodborne outbreak. Data equivalent to
these are not available from drinking water samples collected under the current TCR option.
Absent any definitive data on co-occurrence of fecal indicator (E. coli) and pathogenic E.
coli, EPA did not estimate the cases of morbidity or mortality avoided. Instead, EPA estimated
changes in occurrence and the resulting changes in assessments and corrective actions performed
for systems serving <4,100 people. For systems serving >4,100 people, EPA applied the 2007
violations data to estimate the increase in effective corrective actions implemented. Discussion
of reductions in risk, then, considers the change in occurrence and corrective actions
implemented for systems serving <4,100 and the changes in corrective actions implemented for
systems serving >4,100.6
For all systems, EPA also estimated the behavioral response of the regulated community
based on projected occurrence rates or violations under the AIP and Alternative option, including
the frequency of Level 1 or Level 2 assessments and the type and number of corrective actions
implemented by PWSs to address the problems identified. EPA expects that the effects of these
changes on risk will be varied, as described in Section 6.3.1.
6 The rationale for using different metrics as proxies for risk reduction in systems serving <4,100 people and those
serving > 4,100 is explained in Chapters 4 and 5 of this EA.
Economic Analysis for the Proposed RTCR
6-18
June 2010
-------
6.3.1 Assessment of Predictive Analysis Results for Small (Serving <4,100 People) and
Large (Serving >4,100 People) Systems
Because the PWSs serving 4,100 people or fewer have a higher initial E. coli occurrence
and will be triggered into more assessments and corrective actions than larger PWSs, the
increase in benefits for these small systems will be more evident as compared to the larger
systems. In particular, model results suggest that customers of small ground water TNCWSs
serving 100 or fewer people, which constitute approximately 40 percent of PWSs, experience the
most improvement in water quality under the proposed RTCR. That is, the occurrence of E. Coli
is predicted to decrease more for these systems that for other systems types.
6.3.1.1 Small Systems (Serving <4,100 People)
For small systems7 (serving <4,100 people), EPA developed a model to simulate
regulatory responses for a 30-year period of analysis, including years 1 - 5 in which GWR8 is in
effect, years 3 - 5 in which RTCR is being implemented, and years 6-30 during which RTCR
requirements are in effect. The 30-year modeled time period includes a total of 28 years of
RTCR impacts post promulgation. Although this EA considers benefits and costs for only 25
years post promulgation, the 30-year period was presented in Chapter 5 for broader consideration
of the appropriateness of the model in terms of its results. The 25 years post promulgation that
are considered in this EA encompass years 3 - 27 of the 30-year modeled period. A complete
description of the model is provided in Chapter 5 of this EA, and data sources used are described
in Chapter 4.
Output from the smaller system simulation model includes the following estimates for
each year of analysis:
- Samples taken (routine, additional routine, and repeat);
- Number of positive results (TC+ and/or EC+)
- Level 1 and Level 2 assessments conducted based on non-acute violations;
- Level 2 assessments conducted based on acute violations; and
- Corrective actions based on Level 1 assessments and Level 2 assessments.
The results of analyses in terms of expected changes in hit rates (positive samples/totals
samples taken) over time are presented in chapter 5 (Section 5.3.3) and Appendix B. Predicted
TC+ hit rate results reflect the overall increase in water quality expected over time under the AIP
and Alternative option. Exhibits 6.2 - 6.4 present a summary of the additional endpoints listed
above for the 25-year period of analysis following rule promulgation.
7 The threshold for defining PWSs as "small" varies depending on the context. For the purpose of this EA, small
systems are defined as those serving <4,100 people.
8 Reductions in EC+ occurrence or violations are only attributable to the proposed RTCR if they result from
requirements of the proposed RTCR alternatives; reductions resulting from the GWR are not attributable to the
proposed RTCR and are not considered further in this chapter. As described in sections 4.2, 4.3, and 5.2 of this EA,
GWR effects are incorporated to adjust source data to the appropriate baseline for this EA.
Economic Analysis for the Proposed RTCR
6-19
June 2010
-------
As discussed earlier in this section, because a dose-response relationship between EC+
occurrence (i.e., acute violations) and illness is not available, EPA is focusing on changes in EC+
occurrence. EPA assumes that reduced occurrence in PWSs should generally correlate with a
reduction in risk of contamination to PWS drinking water. Therefore, the acute violation rates
predicted under the AIP and Alternative option (Exhibits 6.3 - 6.4) represent a level of risk of
contamination that is reduced from that predicted under the current TCR option (Exhibit 6.2).
The numbers of predicted acute violations have two major drivers. First, improvements in water
quality are predicted to result in fewer acute violations. Second, the monitoring frequency
impacts the number of acute violations found, regardless of water quality. Both decreases in
acute violations attributable to water quality improvements and increases due to additional
diagnostic ability of more samples taken result in reduced risk. The combination of these two
influences in the predictive model means that the risk is expected to be lower even though the
number of predicted acute violations may actually be higher or lower than under the current TCR
option as reflected in exhibits 6.2 - 6.4. This concept is discussed in the uncertainty analysis in
Section 6.4 of this chapter.
The changes in the steady state estimates of annual acute violations from the current TCR
option to the AIP and Alternative option are shown in Exhibit 6.5 and 6.6 in absolute numbers
and as a percentage change from the TCR, respectively. The steady state in the model refers to
the period beginning in years 7 (CWSs) and 9 (NCWSs) following promulgation, after the
proportions of systems sampling on monthly, quarterly, or annual regimens are adjusted
following a period of assessment.9 Systems that qualify for reduced monitoring will begin their
new regimens in years 7 and 9 after promulgation, respectively, for CWSs and NCWSs. The
estimates shown in Exhibit 6.5 for small systems are from the predictive model and reflect the
average annual estimates for the 25-year period of analysis, which includes 3 years of
implementation followed by 22 years in which new rule requirements are in effect. To accurately
reflect the results under this schedule, the model output for the 22 years of post-implementation
activity is divided by 22, rather than the entire 25 years, which would distort the results
downward. The steady state reductions in the number of annual acute violations found under the
AIP and Alternative option primarily reflect the benefits of corrective actions under these two
options in preventing many of the acute violations that would otherwise occur over this period.
These results show that under the AIP, no subgroups are predicted to experience an
increase in annual acute violations. While most categories of systems/sizes would experience a
decrease in predicted numbers of violations under the Alternative option, six categories would
actually experience an increase (a positive percent value in this exhibit): NTNC ground water
9 The effective date of the proposed RTCR occurs after 3 years of implementation, at the start of year 4 post
promulgation. For CWSs, years 4-6 post promulgation are the period of assessment for considering potential to
move to reduced monitoring for systems; for NCWSs, years 4 - 8 are the period of assessment. In Chapter 5 of this
EA, these transitions are presented on the 30-year scale of the modeling period rather than the 25-year scale of the
RTCR EA period of analysis. In Chapter 5, Ex. 5.16-5.21 show that the steady state begins in model year 9
(CWSs) and model year 11 (NCWSs). Under the Alternative option, the period of assessment occurs in years 4-8
post promulgation; this represents model years 6 - 10. PWSs under the Alternative option begin their steady state
regarding reduced monitoring schedules in year 9 post promulgation (model year 11). Chapter 5 provides a
description of the criteria used to determine the percent of systems qualifying for reduced monitoring under the
proposed RTCR and Alternative option. This assessment is not made for simulating systems operating under the
current TCR; they are assumed to remain on the initial schedule presented in Chapter 4 (baseline).
Economic Analysis for the Proposed RTCR
6-20
June 2010
-------
systems serving 501-1,000 people and Transient Noncommunity (TNC) ground water systems
serving <100, 101-500, and 501-1,000 people. It is important to note that in all cases except two,
the absolute increases in annual acute violations predicted under either option is very small (<1
annual acute violation as shown in Exhibit 6.5), but translate into observable percentage changes
in Exhibit 6.6. The two categories with an increase of greater than one annual acute violation are
the TNC GW systems serving <100 people and those serving 101-500 people, for which
increases of approximately 52 and 16 annual acute violations, respectively, are predicted. For
these small TNCWSs under the Alternative option, increased monitoring is expected to lead to
an overall increase in annual acute violations (and is also the driver of the greater total number of
annual acute violations predicted).
As discussed earlier, a decrease in acute violations may be caused by an improvement in
water quality, which in this model would result from an increased number of effective corrective
actions being implemented (i.e., occurrence events are "prevented"). An increase in acute
violations can be attributable to a decrease in water quality or a PWS improving its ability to
diagnose more issues (e.g., through more sampling). Alternatively, a decrease in acute violations
could be caused by a decreased ability in PWSs to diagnose the occurrence of TC/EC because of
a reduced sampling schedule, resulting in undetected or "missed" occurrence events. Section 6.4
of this chapter presents a stepwise analysis to discern the relative significance of the effects of
reduced additional routine samples and increased corrective action efficacy as shown by
predictions of "prevented", "found", and "missed" acute violations under the AIP.
6.3.1.2 Large Systems (Serving >4,100 People)
The number of acute and non-acute MCL violations for a given group (based on system
type and water source) of PWSs serving >4,100 people was estimated using 2005 Safe Drinking
Water Information System-Federal Version (SDWIS/FED) data (USEPA 2005a). These
estimates were used to determine the number of Level 1 and Level 2 assessments and associated
corrective actions triggered for large systems in the period of analysis. EPA made a simplifying
assumption that for PWSs serving >4,100 people the number of annual assessment triggers
would remain constant throughout the 22 years following proposed RTCR implementation. As
does the occurrence model for smaller systems, this analysis assumes that systems responding to
a Level 2 assessment trigger (i.e., an acute MCL violation under the current TCR option) would
identify and specifically address the cause of the contamination at a rate increased by 10 percent
under the AIP and Alternative option compared to the current TCR option10.
Exhibits 6.2 - 6.4, described in Section 6.3.1.1 for small systems, also include the number
of activities (assessments and corrective actions) EPA expects PWSs to implement under each of
the regulatory options considered for large systems. Any reduction in risk is estimated to derive
from the additional corrective actions predicted per the explanation above. Exhibit 6.5 also
reflects that EPA is not quantifying any potential change in the number of annual acute
violations for PWSs serving >4,100 people during the period. This assumption simplifies the
analysis and is consistent with EPA's understanding based on TCRDSAC deliberations and best
professional judgment that larger systems are at a relatively steady state with regard to
10 See footnote 3.
Economic Analysis for the Proposed RTCR
6-21
June 2010
-------
operations. Larger systems have been in a position relative to smaller systems to diagnose more
and address the cause of acute violations more, and they have likely had more resources to apply
in maintaining and updating their systems on a regular basis. Therefore, EPA does not believe
that larger systems will make many changes based on RTCR implementation, or that they will
experience large changes in their occurrence rates for TC and EC after RTCR promulgation.
Economic Analysis for the Proposed RTCR
6-22
June 2010
-------
Exhibit 6.2 Predicted Outcomes (25-Year Period of Analysis) under Current
TCR
PWS Size
(Population
Served)
PWSs Subject
to Current
TCR
Number of
Routine
Monitoring
Samples
Num ber of
Additional
Routine
Monitoring
Samples
Num ber of
Repeat
Monitoring
Sam pies
Num ber of
No n-Acute
Violation
assessments
(Single
Violations)
Number of
Corrective
(based on
Single Non-
Acute Violation
Assessm ent)
Num ber of
Acute
Violation
Number of
No n -A cute
Violation
Vssessm ents
(Multiple
Violations)
Num ber of
Corrective
Actions
(based on
Acute and
Multiple Non-
Acute Violation
essm ents)
Com m unity V\feter Systei
33,001-96,000
96,001-500,000
500,001-1 Million
inity VVkter Systei
96,001-500,000
500,001-1 Million
Nontransient Nonce
inity Water Syste m s (NTNCWSs) - SW
33,001-96,000
96,001-500,000
500,001-1 Million
Nontransient No n com m unity Water Systems (NTNCWSs) -GW
33,001-96,000
96,001-500,000
500,001-1 Million
Transient Noncomm unity Water Systems (TNCWSs) - SW
96,001-500,000
500,001-1 Million
Transient Noncomm unity Water Systems (TNCWSs) - GW
33,001-96,000
96,001-500,000
500,001-1 Million
Grand Total
Notes: 1) Predicted outcomes for systems serving <4,100 are output from the occurrence model
detailed in Ch. 5 of this EA; those for systems serving >4,100 people are from the larger systems
model based on 2007 SDWIS violations data. 2) Zero corrective actions are indicated in the two
Number of Corrective Actions columns (F and I) because these columns represent incremental
changes from the level of corrective actions that EPA estimates would be implemented under the
current TCR option. 3)Results differ slightly from those presented in Ex. 5.10 - 5.15 because
they're capturing slightly different time periods of the 30 modeled years. For completeness in
discussing the simulation model, Chapter 5 exhibits show 30 years of results, beginning with 5
years of GWR in effect, the last 3 of which also include RTCR implementation, followed by 25
years of RTCR in effect. Alternatively, Ex. 6.2 - 6.4 include the modeled period that encompasses
the 3 years of proposed RTCR implementation and the 22 years that follow of RTCR in effect
(Years 3-28 of the 30-year period). During the 3 years of RTCR implementation, under either the
AIP or Alternative option, systems are still working under the requirements of the GWR while
incurring costs for implementing the new rule requirements.
Economic Analysis for the Proposed RTCR
6-23
June 2010
-------
Exhibit 6.3 Predicted Outcomes (25-Year Period of Analysis) for AIP
Number of
Number of
Number of
Number of
Level 2
Level 2
Number of
Level 1 and
Number of
Additional
Number of
Corrective
Assessments
Assessments
Corrective
Level 2 (based
Routine
Routine
Repeat
Number of
Actions (based
(based on
(based on non>
Actions (based
on non-acute
PWS Size
PWSs Subject
Monitoring
Monitoring
Monitoring
Level 1
on Level 1
Acute
acute
on Level 2
triggers)
(Population
to RTCR
Samples
Samples
Samples
Assessments
Assessments)
Violations)
triggers)
Assessments)
Assessments
Served)
A
B
C
D
E
F
G
H
I
J=E+H
Community Water Systems (CWSs) - SW
<100
1,170
308,849
12,875
363
39
112
73
17
437
101-500
2,150
567,486
15,333
465
43
135
55
20
520
501-1,000
1,173
309,610
8,366
254
23
74
30
11
284
1,001-4,100
2,938
1,950,717
32,222
846
83
161
96
23
942
4,101-33,000
3,164
10,636,296
175,689
2,152
215
197
20
2,152
33,001-96,000
720
11,058,960
182,671
534
53
56
6
534
96,001-500,000
308
10,190,400
168,324
233
23
24
2
233
500,001-1 Mllion
31
2,019,600
33,360
22
2
22
> 1 Mllion
17
1,686,960
27,865
Totals
11,671
38,728,879
656,704
4,869
482
758
253
100
5,122
Community Water Systems (CWSs) - GW
<100
11,938
2,870,075
8,760
156,897
8,004
791
853
3,523
398
11,527
101-500
13,892
3,391,200
6,127
136,906
6,502
669
696
2,399
335
8,901
501-1,000
4,467
1,085,730
1,844
39,659
1,780
168
188
626
85
2,406
1,001-4,100
6,443
4,079,328
96,939
3,208
318
342
1,705
206
4,913
4,101-33,000
3,156
9,145,224
217,321
4,545
454
263
26
4,545
33,001-96,000
335
4,884,000
116,060
656
66
53
5
656
96,001-500,000
63
1,945,680
46,236
129
13
10
1
129
500,001-1 Mllion
4
253,440
6,023
> 1 Mllion
3
269,280
6,399
Totals
40,301
27,923,956
16,731
822,439
24,824
2,480
2,405
8,253
1,056
33,077
Nontransient Non community Water Systems (NTNCWSs) - SW
<100
250
65,986
2,840
69
7
27
33
6
102
101-500
253
66,766
2,073
66
7
24
21
4
87
501-1,000
88
23,223
721
23
2
8
7
2
30
1,001-4,100
72
42,751
1,183
38
4
12
18
3
56
4,101-33,000
22
50,424
1,395
5
0
5
33,001-96,000
2
34,320
950
96,001-500,000
1
31,680
877
500,001-1 Mllion
> 1 Mllion
Totals
688
315,151
10,038
201
20
72
79
15
280
Nontransient Noncommunity Water Systems (NTNCWSs) - GW
<100
8,826
932,025
48,142
68,123
4,797
493
559
2,010
254
6,807
101-500
6,613
678,688
25,630
35,860
2,794
271
315
757
107
3,552
501-1,000
1,718
180,145
6,166
8,601
675
66
79
168
24
843
1,001-4,100
812
473,352
15,887
690
68
114
530
65
1,221
4,101-33,000
70
153,648
5,157
123
12
9
1
123
33,001-96,000
2
23,760
797
4
0
4
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Totals
18,041
2,441,617
79,938
134,426
9,084
912
1,077
3,466
450
12,550
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
1,339
353,461
20,787
1,004
99
336
426
80
1,430
101-500
497
131,149
7,816
372
34
125
165
28
537
501-1,000
88
23,222
1,384
66
6
22
29
5
95
1,001-4,100
67
42,209
2,118
90
9
32
62
9
152
4,101-33,000
18
40,656
2,040
8
1
8
33,001-96,000
96,001-500,000
500,001-1 Mllion
> 1 Mllion
1
102,960
5,167
Totals
2,010
693,657
39,312
1,541
149
515
682
122
2,222
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
60,200
6,076,163
446,166
631,105
47,190
4,755
5,477
20,628
2,593
67,818
101-500
19,275
1,940,946
135,822
194,697
13,780
1,363
1,608
5,694
799
19,474
501-1,000
1,963
206,130
14,078
20,078
1,396
143
177
585
76
1,982
1,001-4,100
617
348,480
16,027
773
77
117
638
76
1,412
4,101-33,000
67
156,288
7,188
116
12
4
0
116
33,001-96,000
2
34,320
1,578
96,001-500,000
1
26,400
1,214
500,001-1 Mllion
1
63,360
2,914
> 1 Mllion
Totals
82,126
8,852,088
596,065
874,801
63,256
6,349
7,383
27,546
3,544
90,801
Grand Total
154,837
78,955,347
692,734
2,537,720
103,775
10,391
12,210
40,278
5,287
144,053
Notes: 1) Predicted outcomes for systems serving <4,100 are output from the occurrence model detailed
in Ch. 5 of this EA; those for systems serving >4,100 people are from the larger systems model based on
2007 SDWIS violations data. 2) The two Number of Corrective Actions columns (F and I) represent the
incremental changes in the level of corrective actions that EPA estimates would be implemented under
the regulatory option in comparison to the current TCR option. 3) Results differ slightly from those
presented in Ex. 5.10 - 5.15 because they're capturing slightly different time periods of the 30 modeled
years. For completeness in discussing the simulation model, Chapter 5 exhibits show 30 years of results,
beginning with 5 years of GWR in effect, the last 3 of which also include RTCR implementation, followed
by 25 years of RTCR in effect. Alternatively, Ex. 6.2 - 6.4 include the modeled period that encompasses
the 3 years of proposed RTCR implementation and the 22 years that follow of RTCR in effect (Years 3 -
28 of the 30-year period). During the 3 years of RTCR implementation, under either the AIP or
Alternative option, systems are still working under the requirements of the GWR while incurring costs for
implementing the new rule requirements.
Economic Analysis for the Proposed RTCR
6-24
June 2010
-------
Exhibit 6.4 Predicted Outcomes (25-Year Analysis Period) for Alternative Option
Number of
Number of
Number of
Number of
Level 2
Level 2
Number of
Level 1 and
Number of
Additional
Number of
Corrective
Assessments
Assessments
Corrective
Level 2 (based
Routine
Routine
Repeat
Number of
Actions (based
(based on
(based on non
Actions (based
on non-acute
PWS Size
PWSs Subject
Monitoring
Monitoring
Monitoring
Level 1
on Level 1
Acute
acute
on Level 2
triggers)
(Population
to RTCR
Samples
Samples
Samples
Assessments
Assessments)
Violations)
triggers)
Assessments)
Assessments
Served)
A
B
C
D
E
F
G
H
I
J=E+H
Community Water Systems (CWSs) - SW
<100
1,170
308,880
13,764
400
36
100
102
21
501
101 - 500
2,150
567,600
15,660
539
56
119
75
20
615
501-1,000
1,173
309,672
8,708
277
27
75
40
12
317
1,001-4,100
2,938
1,951,224
33,326
920
95
146
132
29
1,052
4,101-33,000
3,164
10,636,296
181,661
2,152
215
197
20
2,152
33,001-96,000
720
11,058,960
188,880
534
53
56
6
534
96,001-500,000
308
10,190,400
174,046
233
23
24
2
233
500,001-1 Mllion
31
2,019,600
34,493
22
2
22
>1 Mllion
17
1,686,960
28,812
Totals
11,671
38,729,592
679,350
5,076
507
717
349
110
5,425
Community Water Systems (CWSs) - GW
<100
11,938
2,908,469
7,545
158,439
7,871
812
926
3,272
432
11,143
101-500
13,892
3,428,876
5,264
137,959
6,495
667
747
2,543
322
9,038
501-1,000
4,467
1,098,488
1,616
39,580
1,772
174
203
607
83
2,379
1,001-4,100
6,443
4,079,328
96,939
3,208
318
342
1,705
206
4,913
4,101-33,000
3,156
9,145,224
217,321
4,545
454
263
26
4,545
33,001-96,000
335
4,884,000
116,060
656
66
53
5
656
96,001-500,000
63
1,945,680
46,236
129
13
10
1
129
500,001-1 Mllion
4
253,440
6,023
> 1 Mllion
3
269,280
6,399
Totals
40,301
28,012,784
14,425
824,956
24,675
2,504
2,544
8,127
1,077
32,802
Nontransient Noncommunity Water Systems (NTNCWSs) - SW
<100
250
66,000
3,040
75
8
28
41
7
116
101-500
253
66,792
2,169
69
7
19
24
4
93
501-1,000
88
23,232
756
24
2
6
9
2
33
1,001-4,100
72
42,768
1,228
37
4
13
23
4
59
4,101-33,000
22
50,424
1,448
5
0
5
33,001-96,000
2
34,320
985
96,001-500,000
1
31,680
910
500,001-1 Mllion
> 1 Mllion
Totals
688
315,216
10,536
209
22
67
98
17
306
Nontransient Noncommunity Water Systems (NTNCWSs) - GW
<100
8,826
1,314,175
36,965
91,416
5,673
562
723
3,390
399
9,063
101-500
6,613
976,627
19,382
48,269
3,551
356
446
1,333
183
4,884
501-1,000
1,718
249,760
4,802
11,817
814
81
99
298
41
1,112
1,001-4,100
812
473,352
15,887
690
68
114
530
65
1,221
4,101-33,000
70
153,648
5,157
123
12
9
1
123
33,001-96,000
2
23,760
797
4
0
4
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Totals
18,041
3,191,322
61,149
173,343
10,855
1,080
1,393
5,551
689
16,406
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
1,339
353,496
23,122
796
76
250
425
66
1,221
101-500
497
131,208
8,192
278
27
90
154
25
432
501-1,000
88
23,232
1,533
50
5
17
25
4
75
1,001-4,100
67
42,240
2,312
73
7
29
69
10
142
4,101-33,000
18
40,656
2,225
8
1
8
33,001-96,000
96,001-500,000
500,001-1 Mllion
> 1 Mllion
1
102,960
5,636
Totals
2,010
693,792
43,020
1,204
116
386
674
105
1,878
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
60,200
9,524,123
333,524
912,589
57,597
5,737
7,796
37,532
4,450
95,129
101-500
19,275
3,021,771
104,732
282,740
17,358
1,616
2,441
10,924
1,337
28,282
501-1,000
1,963
304,534
10,412
27,932
1,661
163
230
1,015
123
2,676
1,001-4,100
617
348,480
16,027
773
77
117
638
76
1,412
4,101-33,000
67
156,288
7,188
116
12
4
0
116
33,001-96,000
2
34,320
1,578
96,001-500,000
1
26,400
1,214
500,001-1 Mllion
1
63,360
2,914
> 1 Mllion
Totals
82,126
13,479,275
448,667
1,252,181
77,506
7,605
10,589
50,109
5,986
127,615
Grand Total
154,837
84,421,981
524,241
2,983,387
119,526
11,834
15,695
64,908
7,983
184,433
Notes: 1) Predicted outcomes for systems serving <4,100 are output from the occurrence model detailed
in Ch. 5 of this EA; those for systems serving >4,100 people are from the larger systems model based on
2007 SDWIS violations data. 2) The two Number of Corrective Actions columns (F and I) represent the
incremental changes in the level of corrective actions that EPA estimates would be implemented under
the regulatory option in comparison to the current TCR option. 3) Results differ slightly from those
presented in Ex. 5.10 - 5.15 because they're capturing slightly different time periods of the 30 modeled
years. For completeness in discussing the simulation model, Chapter 5 exhibits show 30 years of results,
beginning with 5 years of GWR in effect, the last 3 of which also include RTCR implementation, followed
by 25 years of RTCR in effect. Alternatively, Ex. 6.2 - 6.4 include the modeled period that encompasses
the 3 years of proposed RTCR implementation and the 22 years that follow of RTCR in effect (Years 3 -
28 of the 30-year period). During the 3 years of RTCR implementation, under either the AIP or
Alternative option, systems are still working under the requirements of the GWR while incurring costs for
implementing the new rule requirements.
Economic Analysis for the Proposed RTCR
6-25
June 2010
-------
Exhibit 6.5 Predicted Average Annual Acute Violations by Reg.
Option and System Type1
PWS Size
(Population
Alte rnative
PWS Type
Served)
Number of Systems
TCR
AIP Option
Option
<100
1,170
7
5
5
101-500
2,150
8
5
5
501-1,000
1,173
5
3
3
1,001-4,100
2,938.0
7.4
6.6
6.6
cws sw
4,101-33,000
3,164
9
9
9
33,001-96,000
720
3
3
3
96,001-500,000
308
1
1
1
500,001-1 Million
31
0
0
0
> 1 Million
17
0
0
0
Totals
11,671
39
33
33
<100
11,938
52
39
42
101-500
13,892
47
32
34
501-1,000
4,467
13
9
9
1,001-4,100
6,443
22
16
16
CWS GW
4,101-33,000
3,156
12
12
12
33,001-96,000
335
2
2
2
96,001-500,000
63
0
0
0
500,001-1 Million
4
0
0
0
> 1 Million
3
0
0
0
Totals
40,301
148
109
116
<100
250
2
1
1
101-500
253
1
1
1
501-1,000
88
0
0
0
1,001-4,100
72
1
1
1
NTNC SW
4,101-33,000
22
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
1
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
0
0
0
0
Totals
688
4
3
3
<100
8,826
39
25
33
101-500
6,613
20.31
14.31
20.29
501-1,000
1,718
4.31
3.61
4.51
1,001-4,100
812
8
5
5
NTNC GW
4,101-33,000
70
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
0
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
0
0
0
0
Totals
18,041
72
49
63
<100
1,339
20
11
11
101-500
497
8
4
4
501-1,000
88.00
1.25
0.79
0.79
1,001-4,100
67.0
2.3
1.3
1.3
TNC SW
4,101-33,000
18
0
0
0
33,001-96,000
0
0
0
0
96,001-500,000
0
0
0
0
500,001-1 Million
0
0
0
0
> 1 Million
1
0
0
0
Totals
2,010
31
18
18
<100
60,200
302
249
354
101-500
19,275
94.93
73.11
110.96
501-1,000
1,963
10.07
8.04
10.47
1,001-4,100
617
8
5
5
TNC GW
4,101-33,000
67
0
0
0
33,001-96,000
2
0
0
0
96,001-500,000
1
0
0
0
500,001-1 Million
1
0
0
0
> 1 Million
0
0
0
0
Totals
82,126
416
336
481
All System s
Grand Total
154,837
710
547
713
Source: Output from Proposed RTCR models as described in Sections
5.3 and 5.4 of this EA for smaller systems and larger systems,
respectively.
Notes: 1 Average violation estimates are for the 22-year period following
the effective date of the RTCR and 3 years of implementation activities.
Economic Analysis for the Proposed RTCR
6-26
June 2010
-------
Exhibit 6.6 Predicted Change in Average Annual Acute Violations1 by Regulatory
Option and System Type
PWS Size
(Population
AIP
Alternative
PWS Type
Se rved)
Option
Option
cwssw
<100
-36%
-36%
101-500
-29%
-29%
501-1,000
-27%
-27%
1,001-4,100
-10%
-10%
Totals
-17%
-17%
CWSGW
<100
-25%
-19%
101-500
-32%
-27%
501-1,000
-34%
-29%
1,001-4,100
-28%
-28%
Totals
-26%
-22%
NTNC SW
<100
-20%
-20%
101-500
-35%
-35%
501-1,000
-32%
-32%
1,001-4,100
-30%
-30%
Totals
-28%
-28%
NTNC GW
<100
-35%
-15%
101-500
-30%
0%
501-1,000
-16%
5%
1,001-4,100
-32%
-32%
Totals
-32%
-12%
TNC SW
<100
-42%
-42%
101-500
-47%
-47%
501-1,000
-37%
-37%
1,001-4,100
-43%
-43%
Totals
-43%
-43%
TNC GW
<100
-18%
17%
101-500
-23%
17%
501-1,000
-20%
4%
1,001-4,100
-37%
-37%
Totals
-19%
16%
All Systems
Grand Total
-23%
1%
Source: Exhibit 6.5. Percentages may not match results
calculated directly from exhibit 6.5 due to rounding in exhibit
6.5.
Notes: 1 Average violation estimates are for the 22 year
period (following 3 years of implementation) after systems
have fully implemented their new sampling regimes.
Negative changes indicate reductions in the number of acute
violations under either option in comparison to the current
TCR option; positive changes indicate increases in acute
violations. As described further in Section 6.4, a net increase
in acute violations is caused by increased diagnostic power
from increased sampling.
Systems serving >4,100 were omitted from this table because
there were no changes predicted in the number of violations
to be incurred.
Economic Analysis for the Proposed RTCR
6-27
June 2010
-------
6.3.1.3 Overall Assessment of Predictive Analysis Results
For PWSs serving <4,100 people, EPA expects risk to decrease under the AIP and
Alternative option as compared to the current TCR option. As shown by the year by year detail
of the output provided in Appendix A of this EA and summarized in Exhibits 5.21 - 5.26, risk is
further reduced for PWSs under the Alternative option relative to baseline (the current TCR
option) for the first 5 years after the effective date as compared to the AIP relative to baseline.
This additional reduction in risk under the Alternative option is driven by the increase in
additional sampling that would occur while all systems were still on monthly sampling. Over
time, some of these systems could qualify to move from monthly to quarterly under either
option, or from quarterly to annual sampling under the AIP. Unlike the AIP, no systems under
the Alternative option would be able to sample annually. However, additional costs are incurred
under the Alternative option for this increased monitoring, especially for the systems most
impacted by the AIP (TNCWSs serving <500 people). Chapter 7 presents the full discussion of
costs associated with implementation of the regulatory options considered.
When considering the period beginning in Year 9 after RTCR promulgation through the
end of the period of analysis (model year 11 in Exhibits 5.21 - 5.26), the AIP and Alternative
option generally have similar estimates of occurrence. Year 9 after RTCR promulgation
represents the first year of the steady state for monitoring regimens, when all PWSs that qualified
for reduced monitoring are following their new regimens. However, for some categories and
sizes of PWSs the AIP actually has a lower rate of occurrence than the Alternative option. This
may occur because PWSs under the AIP may have more resources available to perform the
annual inspection more regularly based on states needing fewer resources for monitoring than
under the Alternative option. Appendix B includes graphs of predicted occurrence for each of the
size and PWS categories considered in this analysis.
EPA does not expect changes in risk for the PWSs serving >4,100 people under the AIP
to be as obvious as for the smaller systems. The larger systems are starting from a smaller
baseline level of occurrence than smaller systems and have a correspondingly lower level of
triggered assessment and correction action activity. This suggests that percentage increases in
these activities will result in less obvious changes. Additionally, monitoring requirements for
PWSs serving >4,100 people would remain essentially unchanged under either the AIP or
Alternative option as compared to the current TCR option. Thus the observed overall net
increase in benefits (and costs) for PWSs serving >4,100 people is driven by the requirements to
conduct assessments and to correct any sanitary defects that are found. The increase of 10% from
baseline in effective CAs implemented applies to these larger systems just as it does to systems
serving <4,100 people; however, as shown in Chapter 4 (Exhibit 4.8), the larger systems have a
much smaller level of violations than smaller systems, based on reasons discussed in Chapter 5
(Section 5.4). Therefore, increases in CAs will be less evident in absolute (not percentage) terms.
6.4 Uncertainty and Sensitivity Analyses
Key sources of uncertainty in the estimates provided in this EA include the data used to
develop baseline estimates and the assumptions made regarding model input variables, as
Economic Analysis for the Proposed RTCR
6-28
June 2010
-------
described in Chapters 4 and 5 of this EA, respectively. The quality and representativeness of the
data used in this EA are discussed in Section 4.2, while uncertainty in model input variables is
discussed in Section 5.3.3.1. A summary of assumptions made in developing the baseline and
input parameters to the predictive model that contribute uncertainty to the analysis are included
in Chapters 4 and 5 (Exhibits 4.13 and 5.22a-b). By definition, those uncertainty factors that are
incorporated into the baseline will have a similar effect on the current TCR option, the AIP and
Alternative option; therefore, EPA believes that they will not significantly affect the net results
of the EA. These assumptions are shown in Ex. 5.22a, and include the estimates representing
GWR effects. Although EPA believes its GWR efficacy estimates are conservative, the estimates
are applied in similar fashion to the baseline (current TCR option) and AIP and Alternative
option, therefore any bias that could be introduced is essentially canceled out in the net analysis.
By contrast, the assumptions listed in Ex. 5.22b include those assumptions that will affect only
the AIP and Alternative option, such as the assumed increase in the number of effective CAs
implemented as compared to the current TCR option. These types of assumptions are not
canceled out in the net analysis, and are expected to have some effect on net results. Therefore,
EPA has identified the key drivers of the analysis among this type of assumption and performed
a sensitivity analysis on their values (Chapter 5, Section 5.3.3.1).
This discussion focuses on the provisions of the AIP and Alternative option for
conducting corrective actions based on the results of Level 1 and Level 2 assessments performed,
and the reductions in repeat samples following TC positives and in the additional routine samples
required in the month following a TC positive. The analyses performed and presented here are
intended to provide insight into the overall impact of these two changes in rule provisions under
the proposed RTCR, each of which moves risk (as defined in this EA) in opposite directions.
Unlike the uncertainty analysis presented in Chapter 5 (Occurrence and Predictive Model), this
analysis does not consider variations on the assumptions related to corrective actions, but instead
tests the relative impact of changes in the sampling regimen given the assumptions for corrective
actions applied in the primary analysis of this EA (and summarized in Section 5.3).
The primary benefit of the proposed RTCR is a potential reduction in exposure to
microbial contaminants from drinking water provided by public water systems. Two features of
the proposed RTCR are expected to influence the exposure reduction.
The first feature is the requirement to perform corrective actions based on the results of
the Level 1 and Level 2 assessments performed in response to their individual triggers.
Implementation of additional corrective actions beyond the level currently implemented under
the TCR will reduce exposure to microbial contamination both by addressing the immediate
problem identified by the Level 1 or Level 2 assessment and by preventing some additional
future exposures to fecal contamination.
The second feature, which primarily affects the smaller systems, is the reduction in the
numbers of additional routine and repeat samples that systems are required to take whenever
routine samples are found to be TC positive. This reduction in sampling may contribute to
increased exposure and risk because fewer samples provide fewer opportunities to identify and
address TC and EC exposures.
Economic Analysis for the Proposed RTCR
6-29
June 2010
-------
The analyses that were performed used the predictive model with the occurrence input
parameters for the nondisinfecting ground water TNCWS serving <101 people. This set of
systems was chosen because 1) it contains the largest number of systems of the 27 occurrence
sets (46,642 systems in this set); 2) it is subject to all of the changes in sampling requirements;
and 3) it has the highest routine TC hit rate so that effects on this set of systems are likely to be
more pronounced (and therefore more clearly observable) than other sets. The analysis was
carried out by running simulations of 10,000 systems each for monthly, quarterly, or annual
monitoring schedules.
There were 5 sampling and corrective action "regimens" considered in the model. As
shown in Exhibit 6.7, Regimens 1 through 3 assume a routine sampling regimen equivalent to
the current TCR option (i.e., 1 regular routine sample and a minimum of 5 routine samples in the
month following a TC positive). Regimen 1 is the only one of the 5 regimens that assumes there
are no corrective actions (incremental beyond those already conducted under the current TCR
option); regimens 2 through 5 all assume corrective actions are conducted in accordance with
requirements under the AIP. Regimens 3 through 5 decrease the repeat samples from 4 to 3.
Regimen 4 also decreases the minimum number of routine samples in the month following a TC
positive from 5 to 3; Regimen 5 uses the sampling requirements of the RTCR where systems
doing monthly sampling need only take their 1 regular routine sample in a month following a TC
positive, while for those that monitor quarterly and annually, the requirement remains for a
minimum of 3 additional routine samples in the month following a TC positive.
In this analysis, Regimen 1 describes the current TCR option and Regimen 5 describes
the AIP. Going from Regimen 1 to Regimen 2, where corrective actions are brought in without
any change in sampling, provides key insights to the benefits (reduction in exposure) derived
from the corrective action aspect of the AIP. Going from Regimen 2 through Regimen 5 provides
some insight into how much of those corrective action benefits might be foregone because of the
small reductions in the number of TC samples required.
The metric used for comparing the relative impacts of these 5 regimens was the number
of acute violations based on EC-positive assays (referred to as "acutes" throughout this section),
because this was considered to be the most relevant measure of the potential microbial health
risk-based benefits of the proposed RTCR. The numbers presented are the average annual
numbers of acute violations.
It is important to note that in order to isolate the effects of the proposed RTCR corrective
actions and sampling changes, this analysis excluded any effect of the GWR. To further
understand the effects of sampling regimen, this analysis also presents the results separately for
monthly, quarterly, and annual sampling since one of the changes in the proposed RTCR applies
only to those on monthly sampling. Exhibit 6.7 provides a summary of the results of this
analysis.
Economic Analysis for the Proposed RTCR
6-30
June 2010
-------
Exhibit 6.7 Relative Impacts Analysis for TNCWS Serving <101 People
for the Range of Corrective Action and Sampling Regimens Predicted
Regimen
Minimum
Samples in
Month
Following a
TC+
Repeat
Samples
L1&L2
CAs?
Average Total
TC Samples per Year
Average Acutes Per Year
Found
Average Acutes
PerYear
Prevented
Average Acutes
PerYear
Missed
M
Q
A
M
Q
A
M
Q
A
M
Q
A
M
Q
A
1
5
5
5
4
No
168,252
60,980
15,317
279
106
25
2
5
5
5
4
Yes
160,394
59,693
15,103
198
92
26
80
14
-1
0
0
0
3
5
5
5
3
Yes
156,779
57,851
14,894
172
79
25
70
12
-1
37
14
2
4
3
3
3
3
Yes
151,500
55,411
14,031
189
82
21
77
13
-1
13
11
5
5
1
3
3
3
Yes
146,129
175
71
32
Ratios of Prevented to Found:
0.41
0.15
-0.05
Key: (CA) = correction action; L1 and L2 = Level 1 and Level 2; M = monthly; Q = quarterly; and A = annual.
Definitions: Acutes "found" are acute violations the model predicts will be identified; acutes "prevented" are those
avoided by implementation of CAs; acutes "missed" are those found under the sampling regimen of the current TCR
option (Regimen 1) that were not found under the reduced sampling of Regimens 3, 4, and 5 (calculated by
subtracting acutes found under Regimens 3, 4, and 5 from those found under Regimen 1).
Notes:
Results shown are averages based on simulations of 10,000 systems each for monthly, quarterly, and
annual monitoring using occurrence inputs for nondisinfecting GW TNCWS serving <101 people.
Assumptions for L1 and L2 corrective action efficacy and resulting duration of reduced occurrence are those
used in the primary analysis.
For the purpose of isolating the relative impacts of changes in monitoring regimens and implementation of
L1 and L2 corrective actions, no GWR effects (including GWR corrective actions) are modeled in these runs.
Ratios of acutes "prevented" to those "found" are based on Regimen 2 results, which reflect the current TCR
option sampling regimen but includes implementation of L1 and L2 correction actions.
Regimen 1, representing the current TCR option, provides a "baseline" against which the
other regimens can be compared. Although arguably more acute violations could be found if
more routine and repeat samples were taken, the numbers of acute violations shown here as
"found" (monthly = 279, quarterly = 106, and annual = 25) represent the maximum number that
can be found given the amount of sampling done. This finding is based on analysis using 10,000
simulated systems each for monthly, quarterly, and annual; the sampling scheme under the
current TCR option; and no additional corrective actions being performed.
Regimen 2 uses the exact same sampling scheme as Regimen 1, but includes the
performance of Level 1 and Level 2 corrective actions consistent with the AIP. Here there is a
reduction observed in the number of acutes found for monthly and quarterly sampling although
not for annual sampling. Since there is no change in the sampling requirements, differences (and
similarities) between Regimen 1 and Regimen 2 numbers are due to: a) the implementation of
corrective actions under Regimen 2; and b) and random variation that is endemic to the Monte
Carlo simulation.
For monthly sampling, which requires a large number of TC samples per year, Monte
Carlo variation is relatively small with respect to the effect of the parameter adjustments. Thus,
80 acute violations occur under monthly sampling for Regimen 1 (279 acutes) but not under
Regimen 3 (198 acutes) as the result of corrective actions. Quarterly and annual sampling
Economic Analysis for the Proposed RTCR
6-31
June 2010
-------
regimens require far fewer TC samples per year, and therefore the Monte Carlo variation may
obscure the salient data trends in those model results. For instance, the increase in the number of
acutes under annual sampling under Regimen 1 from 25 to 26 under Regimen 2, which is not
consistent with the trend seen in the larger sample size in the simulation of monthly systems, is
likely to be reflective of stochastic noise from the Monte Carlo simulation peformed rather than
of the impact of corrective actions. For this reason most of the discussion that follows on the
trends across these regimens focuses on the monthly sampling results.
Regimen 3 reflects the reduction in repeat sampling requirements of 4 to 3 following a
TC+. As would be expected, this also results in a small reduction in the number of acutes found.
For example, in the monthly sampling group this falls from 198 to 172. Using an assumption that
the ratio of acutes prevented to acutes found observed in Regimen 2 apply to the other regimens
as well, then results for the monthly sampling group (where 80/198 = 0.41) indicate that 70
acutes are still prevented. The sum of those found (179) and those prevented (70) for the monthly
group totals 249. This implies, then, that as a result of the reduction in repeat sampling from 4 to
3, 37 acutes are missed (279 "found" under the current TCR option sampling regimen - 172 - 70
= 37). That is, they are neither found nor prevented, but are assumed to occur unobserved.
Regimen 4 reduces the number of routines in a month following a TC positive from 5 to
3, but retains the 3 repeats and the corrective action requirements. The number of acutes now
"missed" changes to 13, 11, and 5 respectively for those on monthly, quarterly, and annual
sampling.
Regimen 5, which represents the proposed RTCR as described by the AIP, excludes any
additional routine samples in the month following a TC+ for those on monthly sampling. Here
the number of "missed" acutes for those on monthly increases from 13 to 32. There are no
changes for those on quarterly or annual sampling relative to Regimen 4 since systems sampling
quarterly and annually are treated identically in both scenarios.
In addition to displaying the changes in the number of annual average acutes for these 5
regimens, Exhibit 6.6 also shows the average annual number of TC samples (regular routine,
additional routines and repeat samples) taken. The number of samples declines considerably
across the regimens due both to the reduction in actual sampling requirements and to reduced
number of additional routines and repeats that need to be taken because of the prevention of TC
occurrence by the corrective actions taken.
This analysis points first and foremost to the highly positive public health benefits of
including the corrective actions as part of the proposed RTCR (based on the modest assumptions
regarding their effectiveness used in the occurrence analysis detail in Chapter 5 of this EA). This
is seen not only in the number of acutes that are found for which corrective actions may be
performed, but also - and perhaps more importantly - in the large number of additional acutes
that are prevented from occurring again in those systems as a result of the corrective actions.
While some of these potential benefits are "missed" as a result of the reductions in
additional routine and repeat sampling requirements, these numbers (shown in Exhibit 6.7) seem
small when compared to the numbers found and prevented. This is particularly important when
Economic Analysis for the Proposed RTCR
6-32
June 2010
-------
considering the potential cost savings from reducing the number of TC samples that are taken, as
discussed in Chapter 9 (Net Benefits) of this EA.
The group of 46,642 TNCWSs serving <100 currently has, and will continue to have, a
vast majority of systems on either quarterly or annual sampling in the steady state of the analysis
period, beginning in approximately Year 9 after RTCR promulgation. Since quarterly and annual
estimates of acutes are less stable, predictions for this specific subset of systems are less reliable.
Nonetheless, since the vast majority of systems affected by RTCR are on monthly sampling, the
trends in the monthly numbers cited above should be indicative of trends for other types and
sizes of systems included in this model (those serving fewer than 4,100 people). That is, relative
to the current TCR option, the inclusion of the Level 1 and Level 2 corrective actions will always
result in more benefit accruing from corrective actions being performed and acute violations
being avoided than the relatively small reductions in sample numbers will result in "missing"
acute violations.
Because all of the other system types have lower overall TC and EC hit rates than the
nondisinfecting ground water TNCWS serving <101 people, on a "per system basis" their missed
acute violations would be fewer than those estimated for the set of systems addressed in this
analysis. Furthermore, for the systems serving 1,001 to 4,100 people where the proposed RTCR
change in the minimum number of samples to be taken in the month following a TC+ is smaller,
the number of "missed" acute violations would be less than that seen in this analysis for the
smaller systems. For example, for systems serving 3,301 to 4,100 (which are all on monthly
sampling), the minimum number of next month samples is reduced from 5 to 4 compared with a
reduction from 5 to 1 for the systems serving fewer than 1,001 people that perform monthly
sampling. Thus this analysis presents an estimate that is near the higher end of the range for
various system types in expected "missed" acute violations due to reduced sampling
requirements under the proposed RTCR. In spite of this conservative estimate, the effect of the
proposed RTCR requirements for corrective actions more than balances this effect with a larger
change in risk in the opposite (reducing) direction, as shown by the relatively large number of
"prevented" to "missed" acute violations.
6.5 Other Potential Benefits
A number of benefits may accrue to PWS and their populations served that are not
included in the qualitative relative risk comparison or reductions in occurrence discussed in
Sections 6.1 - 6.4 (above), they are described in the following sections.
6.5.1 Increased System Knowledge
By requiring additional assessments focused on isolating and identifying system
problems in response to TC+ or EC+ samples, the proposed RTCR will increase the likelihood
that PWS operators, in particular those of systems triggered to conduct assessments, will develop
further general understanding of system operations and potential issues. This heightened
familiarity with the system may increase preventive maintenance, or may increase the efficiency
Economic Analysis for the Proposed RTCR
6-33
June 2010
-------
with which future problems are identified, decreasing risk to the PWS population served in both
cases. Delaying system component replacement costs or avoiding an increase in treatment costs
may also result in cost savings for some communities.
6.5.2 Accelerated Infrastructure Repair/Replacement
As described in 6.5.1 (above), the increased familiarity of operators with their systems
may encourage an increase in preventive maintenance, preempting some potential contamination
issues and decreasing risk for the PWS population served. Some systems may see additional non-
quantified benefits associated with the acceleration of their capital replacement fund investments
in response to early identification of impending problems with large capital components.
Although such capital investment would have occurred anyway, earlier investment may ensure
that problems are addressed in a preventive manner and may preclude some decrease in
protection that may have occurred otherwise. At the very least, the increased operator awareness
should reduce the occurrence of unplanned capital expenditures in any given year. However,
because of the difficulty of projecting when capital replacements would occur, EPA has not
costed this acceleration of capital replacement, so there would also be a non-quantified cost of
making such investments sooner.
6.5.3 Reduction in Averting Behavior
As the risk of contamination is reduced in PWSs over time following RTCR
promulgation, EPA expects that the public will eventually become aware of increased quality,
and consistency of quality, of their water. This may lead to PWS customers becoming
increasingly comfortable with drinking tap water and cause them to exercise less averting
behavior (e.g., drinking bottled water or using Point of Use (POU) filtration devices). Such
averting behaviors are costly relative to consuming tap water.
6.5.4 Reduction of Co-Occurring and Other Contaminants
In addition to the specific E. coli endpoint targeted by the proposed RTCR, there are
many potentially co-occurring and emerging pathogens (such as viruses, parasitic protozoa
and/or other bacteria) that may be avoided as part of any avoided contamination event. To the
extent that E. coli co-occurs with pathogens sufficiently in abundance to result in health affects,
the proposed RTCR offers the potential for additional morbidity and mortality prevention.
Potential benefits from the proposed RTCR include avoidance of a full range of health
effects, including acute and chronic illness, endemic and epidemic disease, associated outbreaks
and death that may occur from the consumption of fecally contaminated drinking water. Also
since fecal contamination may contain waterborne pathogens including bacteria, viruses, and
parasitic protozoa, in general, a reduction in fecal contamination should also reduce the risk from
these other contaminants.
Economic Analysis for the Proposed RTCR
6-34
June 2010
-------
Systems may choose corrective actions that also address other drinking water
contaminants. For example, correcting for a pathway of potential contamination into the
distribution system can mitigate a variety of potential contaminants. Due to a lack of
contamination co-occurrence data that quantify the effect that treatment corrective action may
have on contamination entering through distribution system pathways, EPA has not quantified
such potential benefits.
6.5.5 Reduction in Outbreak Risk and Response Costs
Besides reducing the endemic risk of illnesses from waterborne pathogens, the proposed
RTCR would reduce the likelihood of major outbreaks from occurring. These avoided illnesses
and other costs are not estimated or included in the proposed RTCR analyses and would be
difficult to quantify. The economic value of reducing the risk of outbreaks could be quite high
when the magnitude of potential costs is considered. The Agency was unable to quantify or
monetize the cost associated with acute and chronic illnesses or death acquired from consuming
water contaminated with waterborne pathogens because cases avoided could not be calculated, as
described in Section 6.2. Examples of potential illnesses associated with ingestion of waterborne
pathogens and their potential costs are described in Chapter 2 of this EA.
Other types of 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 provisions 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. Thus, to the extent that the proposed
RTCR reduces the likelihood of waterborne disease outbreaks, avoided response costs are
potentially numerous and significant. For example, an analysis of the economic impacts of a
waterborne disease outbreak in Walkerton, Ontario (population 5,000) estimated the economic
impact excluding medically related costs to be over $45.9 million11 in Canadian dollars
(approximately 42.8 million U.S. dollars)12 (Livernois, 2002). The author believed that this was a
conservative estimate.
11 Households and businesses in the town of Walkerton were unable to use municipal water for eight months
following the contamination event. The response involved a massive effort at all levels of government in terms of
public health response and investigation of the cause. The paper asserts that the impact of the crisis extended beyond
Walkerton to nearby towns and the countryside, resulting in economic costs that the author expects were at a
minimum equal to the costs presented in the paper (of which only the non-medical costs are included here).
12 Updated from $43 million in Canadian dollars, $32 million U.S. dollars in 2000. Costs updated from 2000
dollars to 2007 dollars using the Canadian core CPI (= 95.71667 109.8167 = 1.14731) and the 2007 exchange rate
(1.0734 Canadian dollars /1.0 U.S. dollars).
Economic Analysis for the Proposed RTCR
6-35
June 2010
-------
7. Cost Analysis
7.1 Introduction
This chapter presents estimates of the total national and household costs of the proposed
Revised Total Coliform Rule (RTCR). To estimate the national costs of the proposed RTCR, the
United States Environmental Protection Agency (EPA or Agency) calculated the net change in
costs of rule components associated with state1 practices and activities required under the
proposed revised rule compared to the current Total Coliform Rule (TCR). The remainder of this
chapter provides detailed discussion of the methodology used and results from the cost analyses
and is organized as follows:
• Section 7.2 describes the cost model and general costing and compliance assumptions
used to estimate national costs of the proposed RTCR.
• Section 7.3 describes the methodology of projecting costs over a 25-year period
(discounted at 3 and 7 percent, respectively) according to the proposed RTCR
compliance schedule, estimating the present value of each cost, and annualizing each
over a 25-year period.
• Section 7.4 describes the methodology for developing costs for all rule activities.
• Section 7.5 presents household cost estimates.
• Section 7.6 presents a discussion of nonquantified costs.
• Section 7.7 presents a discussion of uncertainties in cost estimates.
• Section 7.8 presents a comparison of cost estimates for all regulatory alternatives.
7.2 General Cost Assumptions and Methodology
The proposed RTCR cost model builds on the baseline data, occurrence analysis, and
benefits model results described in Chapters 4-6. Based on these analyses, the annual and
cumulative numbers of Public Water Systems (PWSs) that would be required to comply with
each rule component of the proposed RTCR over the 25-year compliance period are provided in
Appendix A (Exhibits 6.2 - 6.4 provide a summary of this information). In general, the numbers
of PWSs presented in Appendix A are multiplied by the unit cost assumptions described in this
chapter to calculate total annual costs.
There are also several general costing assumptions that are unique to the costing process
and are used as inputs to the cost model. The derivation of these inputs is discussed in detail
below.
1 The term "state" in the context of this chapter refers to any state or other primacy agency that has oversight
authority for drinking water programs.
Economic Analysis for the Proposed RTCR 7-1
June 2010
-------
7.2.1 Labor Rates
For costing purposes, EPA estimates the labor needs and hourly labor rates of PWSs and
states. EPA recognizes that there may be significant variation in labor rates across all PWSs.
However, for purposes of this Economic Analysis (EA), and to implement national policy, EPA
uses national-level estimates from Labor Costs for National Drinking Water Rules (USEPA,
2003a) (as used in the Ground Water Rule Economic Analysis (USEPA, 2006a)). For the
proposed RTCR cost analyses, these labor rates were inflated to 2007$ using the Employee Cost
Index (ECI), and weighted based on the PWS size categories used in the proposed RTCR EA. To
account for the general composition of staff at PWSs of smaller sizes (e.g., PWSs serving 3,300
or fewer), EPA uses only the technical rate. For PWSs serving more than 3,300 people, EPA uses
a ratio of 80 percent technical labor to 20 percent managerial labor to arrive at a labor cost, or
weighted labor rate. The actual ratio between technical and managerial rates employed may vary
by PWS and among the different compliance activities under the proposed RTCR. However, for
simplicity, the 80/20 ratio is used as a general assumption for costing purposes in this EA. A full
description of the derivation of the labor rates used is provided in the Technology and Cost
Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d). The weighted labor
rates ($2007) are shown in Exhibit 7.1.
Exhibit 7.1 Labor Rates by PWS Size (2007$)
PWS Size (population served)
Weighted Labor Rate ($/hour)
< 100
$
25.10
101 -500
$
27.03
501 - 1,000
$
28.96
1,001 -4,100
$
29.73
4,101 -33,000
$
36.00
33,001 -96,000
$
36.39
96,001 -500,000
$
41.01
500,001-1 Million
$
41.01
> 1 Million
$
41.01
Note: Labor rates for each size category are assumed to be the same regardless
of system type (CWS, NTNCWS, and TNCWS).
Source: Proposed RTCR T&C Document
For states, the administrative and field engineer labor rates from the 2001 State Drinking
Water Needs Analysis (ASDWA, 2001) are used in the proposed RTCR EA (as used in the
Ground Water Rule (GWR) EA (USEPA, 2006a)). These rates include a 60 percent overhead
rate and were inflated to 2007$ using the ECI. EPA recognizes that there may be significant
variation in labor rates across all states. The state labor rates in 2007$ are $39.22 for an
administrative state employee and $43.58 for a state field engineer. EPA assumes that the state
field engineer would conduct annual site visits, and the administrative state employee would
work with PWSs on all remaining aspects of the proposed RTCR. Because this separation
between field engineer and administrative employee is used, the 80/20 weighting ratio between
technical and managerial rates is not used to develop state costs.
Economic Analysis for the Proposed RTCR 7-2
June 2010
-------
7.2.2 TCR Monitoring Costs per Sample
A cost per sample is associated with distribution system monitoring. For the purpose of
this cost analysis, PWSs would perform total coliform (TC) monitoring, supplemented by E.coli
analyses as required. EPA estimated the sample analysis cost for both in-house and commercial
laboratory analysis. The weighted unit costs for monitoring provided in Exhibit 7.2 are based on
the percentage of PWSs conducting in-house and commercial laboratory analysis based on
conversations of the Total Coliform Rule Distribution System Advisory Committee (TCRDSAC)
Technical workgroup (TWG). For in-house sample analysis, the estimated burden includes
sample collection and analysis and also accounts for Operations and Maintenance (O&M) costs
such as equipment and maintenance. For commercial laboratory analysis, the estimated burden
includes sample collection, shipping and delivery, and the laboratory analysis fee. These
estimates reflect a national average; however, individual PWSs may realize collection burden
that is either less than or greater than this average depending on the locations of sampling points
in a particular PWS.
Rates may vary due to regional variations in laboratory fees, the number of samples
processed (quantity discounts), and laboratory capacity. As shown in Exhibit 7.2, the cost per
sample decreases as more samples are taken, and as PWSs take advantage of savings from bulk
shipping. A full description of the derivation of the monitoring costs per sample is provided in
the Technology and Cost Document for the Proposed Revised Total Coliform Rule (USEPA,
2010d).
Exhibit 7.2 Monitoring Costs per Sample (2007$)
PWS Size
Numbers of Samples Taken and Delivered at the Same Time
(population served)
1
2
3
4
>5
< 100
$
50.54
$
42.64
$
40.01
$
38.69
$
37.90
101 -500
$
51.55
$
43.63
$
40.99
$
39.67
$
38.87
501 - 1,000
$
59.81
$
51.86
$
49.21
$
47.88
$
47.09
1,001 -4,100
$
60.40
$
52.45
$
49.79
$
48.47
$
47.67
4,101 -33,000
$
65.26
$
57.23
$
54.55
$
53.21
$
52.40
33,001 - 96,000
$
60.57
$
56.55
$
55.21
$
54.54
$
54.14
96,001 - 500,000
$
72.38
$
71.57
$
71.30
$
71.17
$
71.09
500,001-1 Million
$
72.38
$
71.57
$
71.30
$
71.17
$
71.09
> 1 Million
$
72.38
$
71.57
$
71.30
$
71.17
$
71.09
Note: Per sample monitoring costs for each size category are assumed to be the same regardless of PWS type (CWS,
NTNCWS, and TNCWS).
Source: Proposed RTCR T&C Document
7.2.3 Technology Unit Costs and Compliance Forecasts
EPA has assumed that PWSs may use a variety of existing Best Management Practices
(BMPs) and technologies to address distribution system deficiencies discovered during Level 1
and Level 2 assessments. These BMPs and technologies include a combination of flushing
programs, training of personnel to collect samples, replacing valves/pipes/hydrants/meters,
installing new water mains, modifying operation of storage facilities, booster disinfection,
physical security devices, etc. For a full list of technologies and BMPs that are anticipated to be
used to meet rule requirements, see Appendix D. EPA estimated unit costs for these various
Economic Analysis for the Proposed RTCR 7-3
June 2010
-------
components using equipment price lists and quotes, costs associated with BMPs from PWSs,
engineering cost data sources (e.g., R.S. Means), consultations with the technical workgroup
supporting the TCRDSAC Federal Advisory Committee (FAC), and other relevant assumptions
used in economic analyses performed for existing drinking water rules (e.g., GWR). Detailed
explanations of the unit cost derivations for these BMPs and technologies are presented in the
Technology and Cost Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d).
Compliance forecasts (or technology selection forecasts) are estimates of which
technologies PWSs undergoing corrective action would use. Section 7.4.6 provides details on
compliance forecasts for PWSs performing corrective actions based on Level 1 and Level 2
assessments.
7.2.4 Cost Model
National costs are estimated using a cost model specifically developed for the proposed
RTCR. The model builds on the occurrence model described in Chapter 5. Within the modeling
structure, costs for PWSs serving >4,100 retail customers are analyzed differently from smaller
PWSs to capture differing baseline structures and to account for differences in available
occurrence data as described in Chapter 4.
PWS costs are estimated for different PWS types and size categories (nine size categories
are used based on population served, consistent with the Technology and Cost Document for the
Proposed Revised Total Coliform Rule (USEPA, 2010d)). PWS cost analyses include estimates
to implement the rule; to revise sampling plans; to conduct routine monitoring, additional routine
monitoring, and repeat monitoring; to perform Level 1 and Level 2 assessments and implement
corrective actions; and to provide public notification. State cost analyses include estimates of the
labor burdens that states would face, including staff training on proposed RTCR requirements
and conducting annual administration, reviewing monitoring reports, reviewing and approving
corrective action plans, and recordkeeping. Section 7.4 provides detailed discussion on the
underlying cost-buildup for each rule component analyzed within the cost model.
7.2.5 Modeled Variability and Uncertainty in National Costs
As noted throughout this EA, there is variability among many of the input parameters to
the proposed RTCR cost model and several rule compliance assumptions based on PWS size and
type (e.g., population served, labor rates, TC hit rates, and occurrence distributions are different
for different sizes and types of PWSs). However, there is insufficient information to fully
characterize the distribution of variability (i.e., calculating confidence bounds) within each of
these PWS size and type classifications on a national scale; therefore, EPA uses mean values for
these input parameters.
EPA also recognizes that there is uncertainty in the national cost estimates, much of
which has the same impact on the modeled results of both the TCR and proposed RTCR options
(e.g., baseline assumptions and effects of GWR implementation). Because the EA analyses focus
on net changes between the TCR and the proposed RTCR options, these common sources of
uncertainty cancel each other out in the net change analyses. For assumptions that are major
drivers of the analysis and differ between the TCR and proposed RTCR options (e.g., corrective
Economic Analysis for the Proposed RTCR 7-4
June 2010
-------
action compliance forecast), EPA has evaluated uncertainty and performed sensitivity analyses to
qualitatively and quantitatively characterize the potential impacts of alternative input parameters.
Chapter 5 discusses uncertainty and presents sensitivity analyses pertaining to the predictive
occurrence model results, which also impact the cost calculations. Section 7.7 discusses
uncertainty and provides sensitivity analysis results as they specifically pertain to the cost
analyses.
7.3 Projecting and Discounting National Costs
Costs must be expressed in common units so they can be added together to calculate total
annual costs and compared to benefits to compute net benefits. For the proposed RTCR, the
performance of activities varies over time in response to regulatory requirements and monitoring
results. To compare the values of performing these activities the year or years in which all costs
are expended must be determined and the costs must be calculated as a net present value. For the
purposes of this EA, one-time and yearly costs were projected over a 25-year time period to
allow comparison with other drinking water regulations using the same analysis period. The net
present values of costs are calculated using discount rates of 3 and 7 percent based on EPA
policy and guidance from the Office of Information and Regulatory Affairs of the Office of
Management and Budget (OMB).2 A summary of the steps used in making adjustments to the
national-level costs presented in this EA is as follows:
• Estimate all costs (noncorrective action, corrective action, and state) over a 25-
year time horizon based on the rule implementation schedule.
• Calculate total net present value costs using 3 and 7 percent discount rates.
• Annualize the costs over 25 years using the same discount rates.
Appendix C presents step by step results for the projection and discounting of costs to
show how yearly costs for each rule component are accounted for by the cost model for
Community Water Systems (CWSs), Nontransient Noncommunity Water Systems (NTNCWSs),
Transient Noncommunity Water Systems (TNCWSs), and states. Exhibits C.l through C.9 show
the nominal costs projected over the rule schedule and the present value of each cost calculated
to the expected year of rule implementation for the TCR. Exhibits C.10 through C.45 show the
results for the Agreement in Principle (AIP) and the Alternative option.
7.4 Derivation of Costs for PWSs and States
This section presents the methodology and unit costs used to derive national costs for
PWSs and states to perform TCR- and proposed RTCR -related activities. Chapter 3 contains
detailed summaries of the activities under TCR and the proposed RTCR options (AIP and
Alternative option) considered as part of this EA. The following subsections provide a brief
2 The choice of an appropriate discount rate is a complex and controversial issue among economists and policy
makers. Therefore, the Agency compares streams of future national level costs and benefits using two alternative
discount rates, 3 and 7 percent. The underlying logic for each discount rate can be found in Guidelines for Preparing
Economic Analyses (USEPA, 2000c).
Economic Analysis for the Proposed RTCR 7-5
June 2010
-------
summary of each activity and the assumptions used to estimate the burden and costs attributable
to both PWSs and states for:
7.4.1 Rule Implementation and Annual Administration
7.4.2 Revising Sampling Plans
7.4.3 Monitoring
7.4.4 Annual Site Visits
7.4.5 Assessments
7.4.6 Corrective Actions
7.4.7 Public Notification
This chapter uses information from the baseline analysis in Chapter 4 as a starting point
for analysis of PWSs subject to each rule requirement. Additional baseline information and
detailed intermediate model outputs are provided in Appendix A.
There are also 57 states and primacy agencies that would incur costs as a result of the
proposed rule.
7.4.1 Rule Implementation and Annual Administration
PWSs
Under the AIP and Alternative option, all PWSs subject to the proposed RTCR would
incur one-time costs that include time for staff to read the rule and become familiar with its
provisions and to train employees on rule requirements. No additional implementation burden or
costs are incurred by PWSs under the TCR to implement the proposed RTCR, as these PWSs
have already performed implementation and are continuing to perform annual administration
activities under the TCR. Under the AIP and Alternative option, all PWSs subject to the
proposed RTCR would perform additional or transitional implementation activities. The labor
rates presented in Section 7.2.1 are used along with estimates of labor hours to generate
estimated implementation costs for all PWSs. Based on previous experience with rule
implementation and consistent with estimates used in the GWR EA, EPA estimates that PWSs
would require a total of 4 hours to read and understand the rule, and a total of 8 hours to plan and
mobilize (i.e., assign appropriate personnel and resources to carry out rule activities).
Economic Analysis for the Proposed RTCR 7-6
June 2010
-------
Exhibit 7.3 Net Change in PWS Unit Burden and Cost Estimates for Rule
Implementation (2007$)
FWS Size
(Population
Se rved)
Labor Cost
(per hour)
Read and
Understand Rule
(hours/system)
Planning and
Mobilization
(hours/system)
Unit Cost
A
B
C
[>A*(B+C)
Community Water Systems (CWSs)
<100
$ 25.10
4.0
8.0
$ 301.20
101-500
$ 27.03
4.0
8.0
$ 324.36
501-1,000
$ 28.96
4.0
8.0
$ 347.52
1,001-4,100
$ 29.73
4.0
8.0
$ 356.76
4,101-33,000
$ 36.00
4.0
8.0
$ 432.00
33,001-96,000
$ 36.39
4.0
8.0
$ 436.68
96,001-500,000
$ 41.01
4.0
8.0
$ 492.12
500,001-1 Million
$ 41.01
4.0
8.0
$ 492.12
> 1 Million
$ 41.01
4.0
8.0
$ 492.12
Nontransient Noncommunity Water Systems (NTNCWSs)
<100
$ 25.10
4.0
8.0
$ 301.20
101-500
$ 27.03
4.0
8.0
$ 324.36
501-1,000
$ 28.96
4.0
8.0
$ 347.52
1,001-4,100
$ 29.73
4.0
8.0
$ 356.76
4,101-33,000
$ 36.00
4.0
8.0
$ 432.00
33,001-96,000
$ 36.39
4.0
8.0
$ 436.68
96,001-500,000
$ 41.01
4.0
8.0
$ 492.12
500,001-1 Million
$ 41.01
4.0
8.0
$ 492.12
> 1 Million
$ 41.01
4.0
8.0
$ 492.12
Transient Noncommunity Water Systems (TNCWSs)
<100
$ 25.10
4.0
8.0
$ 301.20
101-500
$ 27.03
4.0
8.0
$ 324.36
501-1,000
$ 28.96
4.0
8.0
$ 347.52
1,001-4,100
$ 29.73
4.0
8.0
$ 356.76
4,101-33,000
$ 36.00
4.0
8.0
$ 432.00
33,001-96,000
$ 36.39
4.0
8.0
$ 436.68
96,001-500,000
$ 41.01
4.0
8.0
$ 492.12
500,001-1 Million
$ 41.01
4.0
8.0
$ 492.12
> 1 Million
$ 41.01
4.0
8.0
$ 492.12
Notes:
FWS burden and cost estimates for implementation activities are assumed to be identical
under the AlPand Alternative Option.
Sources:
(A) Labor rates for FWSs from Exhibit 7.1.
(B), (C) EPA estimates based on best professional judgement.
Economic Analysis for the Proposed RTCR 7-7
June 2010
-------
States
States would incur administrative costs to implement the proposed RTCR. These
implementation costs are not directly required by specific provisions of the proposed RTCR
alternatives but are necessary for states to ensure that the provisions of the proposed RTCR are
properly carried out. States would need to allocate time for their staff to establish and maintain
the programs necessary to comply with the proposed RTCR, including developing and adopting
state regulations and modifying data management systems to track new required PWS reports to
the states. As a one-time burden covered under the 520 hours allocated to modifying data
management systems, each state would modify their data management system to be able to track
the changes in monitoring regimes. Note, on average no on-going annual cost is assumed for
modifying data management because the tracking system would already be in place and
oversight would be accounted under state review of sampling results, assessments, etc. Time
requirements for a variety of state agency activities and responses are estimated in this EA.
Exhibit 7.4 lists the activities required to revise the program following promulgation of the
proposed RTCR along with their respective costs and burden. Because time requirements for
implementation and annual administration activities vary among state agencies, EPA recognizes
that the burden and cost estimates presented in Exhibit 7.4 may be an over- or under-estimate for
some states.
Exhibit 7.4 Net Change in State Unit Burden and Cost Estimates for Rule
Implementation (2007$)
Labor Cost
(per hour)
Hours
FTEs
Cost
Compliance Activity
A
B
C=B/2,080
D=A*B
Read and Understand Rule
$ 39.22
15
0.01
$ 588
Regulation Adoption and Program Development
$ 39.22
260
0.13
$ 10,197
Initial Laboratory Certification
$ 39.22
-
-
$
Modify Data Management Systems
$ 39.22
520
0.25
$ 20,393
FWS Training and Technical Assistance
$ 39.22
520
0.25
$ 20,393
Staff Training
$ 39.22
130
0.06
$ 5,098
Per State Total
National Totals (57 States/Primacy Agencies)
1,445
82,365
$ 56,670
$ 3,230,201
Notes: Detail may not add due to independent rounding.
State burden and cost estimates for Implementation activities are assumed to be identical under the AlPand
Alternative Option.
Sources: (A) Labor rate for state employee from Section 7.2.1.
(B) Labor hours for start-up activities are based on GWR estimates. Because the proposed RTCR is a
revision of the existing TCR, one fourth of the State unit start up burden from GWR is used in the proposed
RTCR.
(C) Full-time equivalent (FTE) assumes individual w orking 40 hours per w eek, 52 w eeks per year.
In addition to these one-time costs, states would use resources to continue
administrative activities. On an annual basis, states must coordinate with their particular EPA
regional office to be certain that the state's program is consistent with federal requirements.
States would also continue to train state and PWS staffs, maintain laboratories' certifications,
and report PWS compliance information to the Safe Drinking Water Information System
(SDWIS). However, based on discussions with stakeholders, once the initial implementation
activities are complete, the annual burden (on average) for general administrative tasks for the
Economic Analysis for the Proposed RTCR 7-8
June 2010
-------
proposed RTCR would not be any higher than the burden incurred under the current TCR
requirements. In some cases, the general administrative burden for TCR may actually decrease as
PWSs experience better performance, and thus require less state interaction, under the proposed
RTCR. Therefore, no continuing annual administrative costs are estimated for the EA.
States would also spend time responding to specific requirements under the proposed
RTCR (i.e., review assessment reports, consult with PWSs, etc.). In these cases, the state costs
are estimated under the costing for that particular rule requirement.
Implementation Net Cost Summary
Because EPA does not anticipate early implementation of the proposed RTCR, EPA
expects that implementation activities would take place in years 1 through 3 of the 25-year
compliance period before PWSs begin monitoring activities. Annualized costs estimates for
PWSs and states to perform implementation activities are estimated by multiplying the number
of PWSs or states required to comply with the proposed RTCR (i.e., all PWSs) by the unit costs
presented in Exhibits 7.3 and 7.4. Annualized and net change cost estimates for PWSs and states
to perform implementation activities under the TCR, AIP, and Alternative option are presented
in Exhibit 7.5.
Exhibit 7.5 Annualized Cost Estimates for Rule Implementation ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR-Total
$
$
$
$
$
$
AIP - Total
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
AIP - Net Change
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Alternative Option - Total
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Alternative Option - Net Change
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Notes:
1) Detail may not add due to independent rounding.
2) FWS and state burden and cost estimates for irrplementation activities are assumed to be identical under the AIP and Alternative Option.
Source: Roposed RTCR Cost Model.
7.4.2 Revising Sampling Plans
PWSs
Under the AIP and Alternative option, all PWSs subject to the proposed RTCR would
incur one-time costs to revise existing sampling plans to identify sampling locations and
collection schedules that are representative of water throughout the distribution system. System
sample siting plans must include routine and repeat sample sites and any sampling points
necessary to meet GWR requirements. Under the TCR, no additional burden or costs are
expected to be incurred by PWSs to revise sampling plans, as these PWSs are already collecting
total coliform samples in accordance with a written sampling plan. The labor rates presented in
Section 7.2.1 are used along with estimates of labor hours to generate sampling plan costs for all
PWSs. Based on previous experience, EPA estimates that PWSs would require 2-8 hours for
revising their sampling plan, depending on PWS size. Estimates of PWS unit costs to revise
sampling plans are presented in Exhibit 7.6.
Economic Analysis for the Proposed RTCR 7-9
June 2010
-------
States
Under the AIP and Alternative option, states are expected to incur one-time costs to
review sampling plans and recommend any revisions to PWSs. Under the TCR, no additional
burden or costs are incurred by states to review sampling plans, as these PWSs' sampling plans
have already been reviewed and approved. State costs are based on the number of PWSs
submitting revised sampling plans to PWSs each year. The state labor rate presented in Section
7.2.1, the number of PWSs in each PWS size category required to revise sampling plans, and
estimates of labor hours are used to generate sampling plan costs incurred by states. Based on
previous experience, EPA estimates that states would require 1-4 hours to review revised
sampling plans and provide any necessary revisions to PWSs, depending on PWS size. Estimates
of state unit costs to revise sampling plans are presented in Exhibit 7.6.
Exhibit 7.6 Net Change in PWS and State Burden and Cost Estimates for
Revising Sampling Plans (2007$)
PWS Size
(Population
Served)
PWSs
States
PWS Labor Cost
(per hour)
Revise
Sampling Plan
(hours/system)
Unit Cost
State Labor Cost
(per hour)
Review and
Revise
Sampling Plan
(hours/system)
Unit Cost
A
B
C=A*B
D
E
^D*E
Community Water Systems (CWSs)
<100
$ 25.10
2.0
$ 50.20
$ 39.22
1.0
$ 39.22
101-500
$ 27.03
2.0
$ 54.06
$ 39.22
1.0
$ 39.22
501-1,000
$ 28.96
4.0
$ 115.84
$ 39.22
2.0
$ 78.44
1,001-4,100
$ 29.73
4.0
$ 118.92
$ 39.22
2.0
$ 78.44
4,101-33,000
$ 36.00
6.0
$ 216.00
$ 39.22
3.0
$ 117.65
33,001-96,000
$ 36.39
8.0
$ 291.12
$ 39.22
4.0
$ 156.87
96,001-500,000
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
500,001-1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
> 1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
Nontransient Noncommunity Water Systems (NTNCWSs)
<100
$ 25.10
2.0
$ 50.20
$ 39.22
1.0
$ 39.22
101-500
$ 27.03
2.0
$ 54.06
$ 39.22
1.0
$ 39.22
501-1,000
$ 28.96
4.0
$ 115.84
$ 39.22
2.0
$ 78.44
1,001-4,100
$ 29.73
4.0
$ 118.92
$ 39.22
2.0
$ 78.44
4,101-33,000
$ 36.00
6.0
$ 216.00
$ 39.22
3.0
$ 117.65
33,001-96,000
$ 36.39
8.0
$ 291.12
$ 39.22
4.0
$ 156.87
96,001-500,000
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
500,001-1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
> 1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
Transient Noncommunity Water Systems (TNCWSs)
<100
$ 25.10
2.0
$ 50.20
$ 39.22
1.0
$ 39.22
101-500
$ 27.03
2.0
$ 54.06
$ 39.22
1.0
$ 39.22
501-1,000
$ 28.96
4.0
$ 115.84
$ 39.22
2.0
$ 78.44
1,001-4,100
$ 29.73
4.0
$ 118.92
$ 39.22
2.0
$ 78.44
4,101-33,000
$ 36.00
6.0
$ 216.00
$ 39.22
3.0
$ 117.65
33,001-96,000
$ 36.39
8.0
$ 291.12
$ 39.22
4.0
$ 156.87
96,001-500,000
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
500,001-1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
> 1 Million
$ 41.01
8.0
$ 328.08
$ 39.22
4.0
$ 156.87
Notes:
FWS and state burden and cost estimates for review ing and revising sampling plans are assumed to be identical under the AIP and
Alternative Option.
Sources:
(A) Labor rates for systems from Exhibit 7.1.
(B) FWS labor hours for reviewing and revising sampling plan reflect EPA estimate.
(D) Labor rates for state employee from Section 7.2.1.
(E3 State labor hours for reviewing and revising sampling plan reflect EPA estimate.
Economic Analysis for the Proposed RTCR 7-10
June 2010
-------
Sampling Plan Net Cost Summary
PWSs are expected to revise sampling plans before monitoring begins. For modeling
purposes costs are split between years 2 and 3 of the 25-year compliance period (monitoring is
required starting in year 4). Annualized cost estimates for PWSs to revise sampling plans and
states to review the revised sampling plans (and consult with PWSs if necessary) under the TCR,
AIP, and Alternative Option are presented in Exhibit 7.7.
Exhibit 7.7 Annualized Cost Estimates to Revise (PWSs) and Review (States)
Sampling Plans ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR-Total
$
$
$
$
$
$
AIP - Total
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
AIP - Net Change
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Alternative Option - Total
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Alternative Option - Net Change
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Note: Detail may not add due to independent rounding.
Source: Proposed RTCR Cost Model.
7.4.3 Monitoring
PWSs
Monitoring costs for PWSs are calculated by multiplying the total numbers of routine,
additional routine, and repeat samples required under the TCR, AIP, and Alternative Option
(Exhibit 7.9) by the monitoring costs per sample presented in Exhibit 7.2. Differences in
monitoring requirements between the TCR and the AIP and Alternative Option as summarized in
Exhibit 7.8 below drive the differences in monitoring costs between the options. Chapters 4
(Exhibit 4.4) and 5 (Exhibit 5.14) show the distribution of monitoring frequencies prior to and
after rule implementation, respectively, that are used to further inform the analysis. The effects
of the differences in monitoring regimes on net changes in costs are described in further detail
following the exhibit.
Economic Analysis for the Proposed RTCR 7-11
June 2010
-------
Exhibit 7.8 Summary of Monitoring Requirements Under the TCR, AIP, and Alternative Options
Monitoring Requirement
Current TCR
AIP
Alternative Option
Default Routine
Monitoring Frequency
The default TC monitoring frequency for
ground water Noncommunity Water
Systems (NCWSs) serving <1,000 people is
quarterly
The default TC monitoring frequency for all
other PWSs is monthly forTC
The default TC monitoring frequency for non-
seasonal ground water NCWSs serving <1,000
people is quarterly
The default TC monitoring frequency for all
other PWSs is monthly
PWSs would be permitted to transition to the
proposed RTCR at their current TC monitoring
frequencies3
The default TC monitoring frequency for all
PWSs is monthly
All PWSs would start on monthly TC
monitoring
Red uced/l ncreased
Routine Monitoring
Frequency
Ground water NCWSs serving <1,000
people can reduce to annual monitoring if
no sanitary defects and served only by
protected GW sources
Ground water CWSs serving <1,000 can
reduce to quarterly monitoring if no history
of TC+ in current configuration, no sanitary
defects, and served only by protected GW
sources
All other PWSs are ineligible for reduced
monitoring
PWSs not meeting criteria for reduced
monitoring return to default monitoring (no
increased monitoring provision under
current TCR)
Ground water NCWSs serving <1,000 people
can reduce to annual monitoring if no sanitary
defects, clean TCR compliance history, and
annual site visit or Level 2 assessment, and
correction of all identified sanitary defects
Ground water CWSs serving <1,000 people
can reduce to quarterly monitoring if no
sanitary defects, clean TCR compliance
history, and at least one of the following:
1) Annual site visit or Level 2 assessment and
correction of all identified sanitary defects;
2) An approved cross connection control
program;
3) Continuous disinfection & a residual;
4) 4-log inactivation of viruses daily as per
GWR (4 hr exception allowed); or
5) Other equivalent measures as approved by
the primacy agency.
All other PWSs are ineligible for reduced
monitoring
Ground water NCWSs serving <1,000 people
can reduce to quarterly monitoring if no
sanitary defects, clean TCR compliance
history, and annual site visit or Level 2
assessment, and correction of all identified
sanitary defects.
Ground water CWSs serving <1,000 people
can reduce to quarterly monitoring if no
sanitary defects, clean TCR compliance
history, and at least one of the following:
1) Annual site visit or Level 2 assessment and
correction of all identified sanitary defects;
2) An approved cross connection control
program;
3) Continuous disinfection & a residual;
4) 4-log inactivation of viruses daily as per
GWR (4 hr exception allowed); or
5) Other equivalent measures as approved by
the primacy agency.
All other PWSs are ineligible for reduced
3 In order for PWSs to be able to transition to the proposed RTCR at their current TC monitoring frequencies, an annual site visit or voluntary Level 2 assessment would
be needed in the first year for PWSs on annual monitoring, including those transitioning.
Economic Analysis for the Proposed RTCR
7-12
June 2010
-------
Monitoring Requirement
Current TCR
AIP
Alternative Option
Ground water PWSs serving <1,000 people on
quarterly or annual monitoring that experience
any of the following events will be required to
begin monthly monitoring:
1) System triggers a Level 2 assessment (or a
2nd Level 1 assessment in a rolling 12 month
period);
2) System has an E. coli Maximum
Contaminant Level (MCL) violation;
3) System has an RTCR treatment technique
violation (either Level 1 or 2); or
4) System has two routine monitoring
violations in a rolling 12-month period.
monitoring
Ground water PWSs serving <1,000 people on
quarterly monitoring that experience any of the
following events will be required to begin
monthly monitoring (i.e., return to default
monitoring frequency):
1) System triggers a Level 2 assessment (or a
2nd Level 1 assessment in a rolling 12 month
period);
2) System has an E. coli MCL violation;
3) System has an RTCR treatment technique
violation (either Level 1 or 2); or
4) System has two routine monitoring
violations in a rolling 12-month period.
Additional Routine
Monitoring Frequency
All PWSs serving <4,100 people must take
at least 5 samples in the month following a
TC+ unless state performs site visit and
deems additional sampling unnecessary OR
determines the reason for the TC+ and
PWS has or will correct problem.
Not required for PWSs serving >4,100
people
Ground water NCWSs serving <1,000 people
and monitoring quarterly or annually must take
at least 3 samples in the month following TC+
Ground water CWSs serving <1,000 people
and monitoring quarterly must take at least 3
samples in the month following TC+
Not required for PWSs monitoring monthly
Ground water PWSs (NCWS and CWS)
serving <1,000 people and monitoring
quarterly must take at least 3 samples in the
month following TC+
Not required for PWSs monitoring monthly
Repeat Monitoring
Frequency
All PWSs serving <1,000 people must take
4 repeat TC samples
All PWSs serving >1,000 people must take
3 repeat TC samples
All PWSs must take 3 repeat TC samples
All PWSs must take 3 repeat TC samples
Economic Analysis for the Proposed RTCR
7-13
June 2010
-------
Routine Monitoring
Under the AIP, the increased stringency to qualify for reduced monitoring results in more
routine samples being taken over time (fewer PWSs are on reduced monitoring) for most PWS
sizes and types. The only PWSs predicted to have a decrease in routine monitoring under the AIP
are ground water NTNCWSs serving <1,000 people due to a more PWSs monitoring quarterly
than under the current TCR. For the Alternative option, this effect would be combined with the
requirement that all PWSs start the implementation period on monthly monitoring, resulting in
more routine samples being taken for all PWS sizes and types. The Alternative option also
prohibits annual monitoring for an even greater increase in the number of routine samples. The
resulting increases in total national costs due to increased monitoring on the national level are
reflected in the routine monitoring costs shown in Exhibit 7.10.
Additional Routine Monitoring4
The overall reductions in additional routine samples required under the AIP and
Alternative option result in reduced costs (Exhibit 7.10). Under the AIP and Alternative options,
additional routine monitoring is no longer required for systems that monitor at least monthly, and
when additional routine monitoring is required, the number of samples required is reduced from
five to three. Cost reductions are greater under the Alternative option than under the AIP because
all PWSs start on monthly monitoring and are not required to take additional routine samples
during that period.
Repeat Monitoring
Under the AIP and Alternative option, all PWSs are only required to take three repeat
samples. However, EPA assumes that ground water PWSs treating to less than 4-log would still
take an additional source water sample to comply with the GWR (no change in cost).
Additionally, the number of repeat samples taken is a function of the number of regular and
additional routine samples taken, which in turn affects the number of TC+ samples found (i.e.,
the more samples taken, the greater chance of finding a TC+). Thus, the overall increase in
routine sampling under the AIP and Alternative option would result in more repeat samples
while the decreases in additional routine samples under both of these options would lead to fewer
repeat samples.
In most PWS size and type categories, the large reductions in additional routine
monitoring samples is the major driver, leading to decreases in repeat samples. Under the AIP,
only groundwater TNCWSs serving <100 people are predicted to see an increase repeat sampling
due to the large increase in routine sampling within this category. Under the Alternative option,
ground water NTNCWSs and TNCWSs serving <1,000 people are predicted to have increased
repeat monitoring as a function of large increases in routine monitoring. The overall effect on the
national level is a reduction in the number (and cost) of required repeat samples under the AIP
and an increase under the Alternative option compared to the current TCR (Exhibit 7.10).
4 For modeling purposes and discussion throughout the cost chapter, regular routine monitoring samples taken in the
month following a total coliform-positive are included in the additional routine monitoring sample counts.
Economic Analysis for the Proposed RTCR 7-14
June 2010
-------
Cumulative Monitoring Summary
Exhibit 7.9 summarizes the total number of samples taken by PWS size and category for
routine, additional, and repeat monitoring under the TCR, AIP, and Alternative option over the
entire 25 year period of analysis. Appendix A presents additional information on samples taken
for each individual year during the analysis period.
States
Under the current TCR, states are estimated to incur a monthly 15-minute burden to
review each PWS's sample results. This estimate reflects the methodology used to calculate
reporting and recordkeeping burden under the current TCR in the Draft Information Collection
Request for the Microbial Rules (USEPA, 2008b). Because the existing methodology is
calculated on a per PWS basis and the total number of PWSs is the same for cost modeling under
the TCR and both proposed RTCR options, the net change in costs for reviewing monitoring
results is estimated to be zero for the AIP and Alternative option. Specific actions by states
related to positive samples are accounted for under the actions required in response to those
samples. Maintenance of sample results in SDWIS is accounted for under general
implementation and administrative activities, which are discussed in Section 7.4.1.
Monitoring Net Cost Summary
The annualized net present value total and net change cost estimates for PWSs and states
to perform monitoring under the TCR, AIP, and Alternative option are presented in Exhibit 7.10.
All monitoring is modeled to begin in year 4 of the 25-year analysis period.
The overall estimated increase in monitoring costs seen under the AIP is driven by
increases in routine monitoring due to stricter requirements to qualify for reduced monitoring.
However, this is mostly offset by reductions in additional routine and repeat monitoring required
under the revised regulations. For the Alternative option, the requirement for all PWSs to sample
on a monthly basis at the beginning of rule implementation results in a much large cost
differential that is only partially offset by reduced costs due to reductions in additional routine
monitoring requirements. Although not shown in Exhibit 7.10, costs for individual PWS
categories (size and type) are expected to move in the same direction relative to the numbers of
samples taken as reflected in Exhibit 7.9. Exceptions to the general trends on the national level
are discussed in the subsections describing routine, additional routine, and repeat sampling
above.
Economic Analysis for the Proposed RTCR 7-15
June 2010
-------
Exhibit 7.9 Cumulative Number of Samples over 25-Year Period of Analysis
TCR
AIP
Alternative Option
Additional
Additional
Additional
Routine
Routine
Repeat
Routine
Routine
Repeat
Routine
Routine
Repeat
PWS Size
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
Monitoring
(Population
Samples
Samples
Samples
Sam pies
Samples
Sam pies
Samples
Sam pies
Sam pies
Served)
A
B
C
D
E
F
G
H
I
Comm unity Water Systems (CWSs) - SW
<100
304,098
23,747
19,160
308,849
-
12,875
308,849
-
12,875
101-500
562,214
26,578
21,338
567,486
-
15,333
567,486
-
15,333
501-1,000
306,733
14,501
11,641
309,610
-
8,366
309,610
-
8,366
1,001-4,100
1,920,789
55,202
33,730
1,950,717
-
32,222
1,950,717
-
32,222
4,101-33,000
10,636,296
-
186,781
10,636,296
-
175,689
10,636,296
-
175,689
33,001-96,000
11,058,960
-
194,204
11,058,960
-
182,671
11,058,960
-
182,671
96,001-500,000
10,190,400
-
178,951
10,190,400
-
168,324
10,190,400
-
168,324
500,001-1 Million
2,019,600
-
35,466
2,019,600
-
33,360
2,019,600
-
33,360
> 1 Mllion
1,686,960
-
29,624
1,686,960
-
27,865
1,686,960
-
27,865
Total
38,686,051
120,028
710,896
38,728,879
-
656,704
38,728,879
-
656,704
Comm unity Water Systems (CWSs) - GW
<100
2,815,951
286,073
194,462
2,870,075
8,760
156,897
2,908,469
7,545
158,439
101-500
3,344,578
243,895
171,252
3,391,200
6,127
136,906
3,428,876
5,264
137,959
501-1,000
1,072,202
70,803
51,673
1,085,730
1,844
39,659
1,098,488
1,616
39,580
1,001-4,100
3,997,293
160,710
100,618
4,079,328
-
96,939
4,079,328
-
96,939
4,101-33,000
9,145,224
-
230,201
9,145,224
-
217,321
9,145,224
-
217,321
33,001-96,000
4,884,000
-
122,938
4,884,000
-
116,060
4,884,000
-
116,060
96,001-500,000
1,945,680
-
48,976
1,945,680
-
46,236
1,945,680
-
46,236
500,001-1 Million
253,440
-
6,380
253,440
-
6,023
253,440
-
6,023
> 1 Mllion
269,280
-
6,778
269,280
-
6,399
269,280
-
6,399
Total
27,727,648
761,481
933,279
27,923,956
16,731
822,439
28,012,784
14,425
824,956
Nontransient Noncommunity Water Systems (NTNCWSs]
-SW
<100
65,009
4,918
4,005
65,986
-
2,840
65,986
-
2,840
101-500
66,038
3,734
3,008
66,766
-
2,073
66,766
-
2,073
501-1,000
22,970
1,299
1,046
23,223
-
721
23,223
-
721
1,001-4,100
41,740
2,147
1,351
42,751
-
1,183
42,751
-
1,183
4,101-33,000
50,424
-
1,632
50,424
-
1,395
50,424
-
1,395
33,001-96,000
34,320
-
1,111
34,320
-
950
34,320
-
950
96,001-500,000
31,680
-
1,025
31,680
-
877
31,680
-
877
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Mllion
-
-
-
-
-
-
-
-
-
Total
312,182
12,097
13,179
315,151
-
10,038
315,151
-
10,038
Nontransient Noncomm unity Water Systems (NTNCWSs]
-GW
<100
971,538
128,775
84,992
932,025
48,142
68,123
1,281,321
34,581
89,002
101-500
725,785
66,525
43,597
678,688
25,630
35,860
952,008
18,114
46,996
501-1,000
190,649
16,037
10,680
180,145
6,166
8,601
247,132
4,674
11,689
1,001-4,100
460,470
28,214
17,790
473,352
-
15,887
473,352
-
15,887
4,101-33,000
153,648
-
5,936
153,648
-
5,157
153,648
-
5,157
33,001-96,000
23,760
-
918
23,760
-
797
23,760
-
797
96,001-500,000
-
-
-
-
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Mllion
-
-
-
-
-
-
-
-
-
Total
2,525,850
239,551
163,913
2,441,617
79,938
134,426
3,131,221
57,369
169,528
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
345,443
39,654
32,349
353,461
-
20,787
353,461
-
20,787
101-500
128,109
15,365
12,541
131,149
-
7,816
131,149
-
7,816
501-1,000
22,683
2,720
2,220
23,222
-
1,384
23,222
-
1,384
1,001-4,100
39,816
3,990
2,590
42,209
-
2,118
42,209
-
2,118
4,101-33,000
40,656
-
-
40,656
-
2,040
40,656
-
2,040
33,001-96,000
-
-
-
-
-
-
-
-
-
96,001-500,000
-
-
-
-
-
-
-
-
-
500,001-1 Million
-
-
-
-
-
-
-
-
-
> 1 Mllion
102,960
-
-
102,960
-
5,167
102,960
-
5,167
Total
679,667
61,730
49,700
693,657
-
39,312
693,657
-
39,312
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
4,493,808
905,554
600,315
6,076,163
446,166
631,105
9,100,613
289,563
867,968
101-500
1,614,924
316,238
210,714
1,940,946
135,822
194,697
2,886,164
91,085
268,893
501-1,000
177,264
32,730
22,064
206,130
14,078
20,078
295,928
9,578
27,084
1,001-4,100
335,283
29,957
19,113
348,480
-
16,027
348,480
-
16,027
4,101-33,000
156,288
-
8,909
156,288
-
7,188
156,288
-
7,188
33,001-96,000
34,320
-
1,956
34,320
-
1,578
34,320
-
1,578
96,001-500,000
26,400
-
1,505
26,400
-
1,214
26,400
-
1,214
500,001-1 Million
63,360
-
3,612
63,360
-
2,914
63,360
-
2,914
> 1 Mllion
-
-
-
-
-
-
-
-
-
Total
6,901,647
1,284,478
868,188
8,852,088
596,065
874,801
12,911,552
390,226
1,192,865
Grand Total
76,833,044
2,479,366
2,739,154
78,955,347
692,734
2,537,720
83,793,244
462,020
2,893,404
Note: (B), (E), (H) For modeling purposes, additional routine sample counts include regular routine samples taken in the same month.
Source: Appendix A - Total FW3 Counts (A.1z, A.2z, A.3z)
Economic Analysis for the Proposed RTCR 7-16
June 2010
-------
Exhibit 7.10 Annualized PWS and State Cost Estimates for Monitoring Costs
($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
Routine Monitoring
TCR-Total
$ 170.59
$
$ 170.59
$ 163.94
$
$ 163.94
AIP-Total
$ 174.71
$
$ 174.71
$ 167.74
$
$ 167.74
AIP - Net Change
$ 4.12
$
$ 4.12
$ 3.80
$
$ 3.80
AIP - Percent Change
2.42%
0.00%
2.42%
2.32%
0.00%
2.32%
Alternative Option - Total
$ 186.34
$
$ 186.34
$ 181.49
$
$ 181.49
Alternative Option - Net Change
$ 15.75
$
$ 15.75
$ 17.56
$
$ 17.56
Alternative Option - Percent Change
9.23%
0.00%
9.23%
10.71%
0.00%
10.71%
Additional Monitoring
TCR - Total
$ 3.87
$
$ 3.87
$ 3.72
$
$ 3.72
AIP - Total
$ 1.12
$
$ 1.12
$ 1.09
$
$ 1.09
AIP - Net Change
$ (2.75)
$
$ (2.75)
$ (2.63)
$
$ (2.63)
AIP - Percent Change
-71.09%
0.00%
-71.09%
-70.74%
0.00%
-70.74%
Alternative Option - Total
$ 0.68
$
$ 0.68
$ 0.58
$
$ 0.58
Alternative Option - Net Change
$ (3.18)
$
$ (3.18)
$ (3.14)
$
$ (3.14)
Alternative Option - Percent Change
-82.31%
0.00%
-82.31%
-84.35%
0.00%
-84.35%
Repeat Monitoring
TCR - Total
$ 5.11
$
$ 5.11
$ 4.91
$
$ 4.91
AIP - Total
$ 4.82
$
$ 4.82
$ 4.64
$
$ 4.64
AIP - Net Change
$ (0.29)
$
$ (0.29)
$ (0.27)
$
$ (0.27)
AIP - Percent Change
-5.66%
0.00%
-5.66%
-5.57%
0.00%
-5.57%
Alternative Option - Total
$ 5.50
$
$ 5.50
$ 5.45
$
$ 5.45
Alternative Option - Net Change
$ 0.39
$
$ 0.39
$ 0.54
$
$ 0.54
Alternative Option - Percent Change
7.68%
0.00%
7.68%
10.99%
0.00%
10.99%
Total
TCR - Total
$ 179.57
$
$ 179.57
$ 172.57
$
$ 172.57
AIP - Total
$ 180.65
$
$ 180.65
$ 173.46
$
$ 173.46
AIP - Net Change
$ 1.08
$
$ 1.08
$ 0.90
$
$ 0.90
AIP - Percent Change
0.60%
0.00%
0.60%
0.52%
0.00%
0.52%
Alternative Option - Total
$ 192.53
$
$ 192.53
$ 187.53
$
$ 187.53
Alternative Option - Net Change
$ 12.96
$
$ 12.96
$ 14.96
$
$ 14.96
Alternative Option - Percent Change
7.22%
0.00%
7.22%
8.67%
0.00%
8.67%
Notes:
1) Detail may not add due to independent rounding.
2) For modeling purposes, additional routine sample counts include regular routine samples taking in the same month.
3) State costs are premised on a per system basis. State costs for monitoring are expected to be identical under the TCR, AIP, and
Alternative Option, and are therefore not included in the total costs.
Source: Proposed RTCRCost Model.
7.4.4 Annual Site Visits
Under the AIP, any PWS on an annual monitoring schedule would be required to also
have an annual site visit conducted by the state or state-designated third party. A voluntary Level
2 assessment can also satisfy the annual site visit requirement. In many cases a sanitary survey
performed during the same year can also be used to satisfy this requirement.5 Although similar
5 In some instances, the performance of an assessment (especially a Level 2 assessment) may overlap with a
scheduled sanitary survey. To the extent that the requirements for performing a sanitary survey may be satisfied as
part of the assessment process, PWSs and states may realize a cost savings compared to performing a separate
sanitary survey. This potential for indirect cost savings is not captured in the cost model, resulting in a potential
overestimate of costs stemming from proposed RTCR implementation.
Economic Analysis for the Proposed RTCR 7-17
June 2010
-------
annual site visits are not currently required under the existing TCR, discussions with states
during the TCRDSAC proceedings revealed that some do, in fact, conduct such inspections for
PWSs on annual monitoring schedules. Because of the high cost for an annual site visit by a
state, for this analysis, EPA assumes that no states would choose to conduct annual site visits
unless they already do so under the current TCR. Therefore, for overall costing purposes, no net
change in state or PWS costs are assumed for annual monitoring site visits under the AIP or
Alternative option.
7.4.5 Assessments
PWSs
Level 1 Assessments
Under the AIP and Alternative option, all PWSs experiencing a Level 1 trigger must
complete a Level 1 assessment of the PWS. A Level 1 trigger under the AIP and Alternative
Option is defined as:
• For PWSs taking >40 samples per month, TC+ exceeds 5.0% for a given month;
• For PWSs taking <40 samples per month, two or more TC+ in a month; or
• Failure to take all required repeat samples after a single TC+ sample.
The current TCR does not require a specific assessment to be performed in response to
events comparable to the Level 1 triggers described above (i.e., non-acute violations). However,
PWSs do perform some level of activity similar to a Level 1 assessment in response to
violations. This effort is taken into account in the cost model to accurately assess the net cost of
changes attributable to the proposed RTCR.
A Level 1 assessment (or comparable assessment under the current TCR) includes a
simple examination of the PWS and relevant operational practices. The Level 1 assessment
would be intended as a self-assessment (EPA anticipates that these would be completed by the
PWS and reviewed by the state). If the state determines the completed assessment form
insufficient, it would consult with the PWS. Additionally, as part of the Level 1 assessment,
PWSs would be required to submit to the state a form identifying sanitary defects detected,
corrective actions completed, and a timetable for any corrective actions not already completed.
Additional detail on the requirements of a Level 1 assessment and the derivation of associated
unit burden (labor hours) for performance of the assessment is provided in the Technology and
Cost Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d).
Economic Analysis for the Proposed RTCR 7-18
June 2010
-------
Level 2 Assessments
Under the AIP and Alternative option, all PWSs experiencing a Level 2 trigger must
complete a Level 2 assessment of the PWS. A Level 2 trigger under the AIP and Alternative
Option is defined as:
• An E. coli MCL violation;
• An E. coli monitoring violation; or
• A second Level 1 trigger, within a rolling 12 month period, unless the primary agency
has determined a likely reason that the initial Level 1 samples were TC+ and
establishes that the PWS has corrected the problem.
• For PWSs with approved reduced annual monitoring, a Level 1 trigger in two
consecutive years.
As with Level 1 assessments, the current TCR does not require a specific assessment to
be performed in response to events comparable to the Level 2 triggers described above (i.e.,
acute violations), but PWSs do currently perform some level of activity similar to a Level 2
assessment in response to acute violations. These actions are taken into account in the cost model
to properly assess the net cost of changes attributable to the proposed RTCR regulatory
alternatives.
A Level 2 assessment (or comparable assessment under the current TCR) would be more
involved than a Level 1 assessment. A Level 2 assessment would be a more detailed examination
of the PWS, its monitoring program and results, and its operational practices. It would comprise
essentially the same elements as a Level 1 assessment, but each element would be investigated in
greater detail. The level of effort and resources required to implement the Level 2 assessments
would be commensurate with a more comprehensive investigation, a higher level review of
available information, and may involve the engagement of additional parties and expertise.
Additionally, as part of the Level 2 assessment, PWSs must submit to the state a form identifying
sanitary defects detected, corrective actions completed, and a timetable for completion of any
corrective actions that not already completed. Additional detail on the derivation of associated
unit burden (labor hours) for performance of the assessment is provided in the Technology and
Cost Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d).
The labor rates presented in Section 7.2.1 are used along with estimates of labor hours as
presented in the Technology and Cost Document for the Proposed Revised Total Coliform Rule
(USEPA, 2010d) to generate Level 1 and Level 2 assessment unit costs by PWS size and type.
Labor hours provided are assumed to include time for reporting and recordkeeping activities.
Estimates of PWS unit costs for Level 1 and Level 2 assessments are presented in Exhibits 7.11
and 7.12. Additionally, the numbers of Level 1 and level 2 assessments over the 25-year
compliance period (used to calculate total costs) are presented in Exhibit 7.13.
Economic Analysis for the Proposed RTCR 7-19
June 2010
-------
Exhibit 7.11 PWS Unit Costs Estimates for Assessment Activities under Current
TCR (2007$)
PWS Size
(Population
Served)
Labor Cost
(per hour)
Non-Acute
Violations
(single
violation)
Unit Cost
Acute
Violations
(hours)
Unit Cost
Non-Acute
Violations
(m ultiple
violations) (hours)
Unit Cost
A
B
C=A*B
D
E=A*D
F
G=A*F
Com m unity Water Systems (CWSs) - SW
<100
$ 25.10
11.0
$ 276.10
14.0
$ 351.40
14.0
$ 351.40
101-500
$ 27.03
11.0
$ 297.33
14.0
$ 378.42
14.0
$ 378.42
501-1,000
$ 28.96
13.0
$ 376.48
15.0
$ 434.40
15.0
$ 434.40
1,001-4,100
$ 29.73
22.0
$ 654.06
29.0
$ 862.17
29.0
$ 862.17
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Com m unity Water Systems (CWSs) - GW
<100
$ 25.10
11.0
$ 276.10
14.0
$ 351.40
14.0
$ 351.40
101-500
$ 27.03
11.0
$ 297.33
14.0
$ 378.42
14.0
$ 378.42
501-1,000
$ 28.96
13.0
$ 376.48
15.0
$ 434.40
15.0
$ 434.40
1,001-4,100
$ 29.73
22.0
$ 654.06
29.0
$ 862.17
29.0
$ 862.17
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Nontransient Noncomm unity Water Systems (NTNCWSs) - SW
<100
$ 25.10
4.0
$ 100.40
6.0
$ 150.60
6.0
$ 150.60
101-500
$ 27.03
4.0
$ 108.12
6.0
$ 162.18
6.0
$ 162.18
501-1,000
$ 28.96
4.0
$ 115.84
6.0
$ 173.76
6.0
$ 173.76
1,001-4,100
$ 29.73
4.0
$ 118.92
6.0
$ 178.38
6.0
$ 178.38
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Nontransient Noncomm unity Water Systems (NTNCWSs) - GW
<100
$ 25.10
4.0
$ 100.40
6.0
$ 150.60
6.0
$ 150.60
101-500
$ 27.03
4.0
$ 108.12
6.0
$ 162.18
6.0
$ 162.18
501-1,000
$ 28.96
4.0
$ 115.84
6.0
$ 173.76
6.0
$ 173.76
1,001-4,100
$ 29.73
4.0
$ 118.92
6.0
$ 178.38
6.0
$ 178.38
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Transient Noncom m unity Water Systems (TNCWSs) - SW
<100
$ 25.10
4.0
$ 100.40
6.0
$ 150.60
6.0
$ 150.60
101-500
$ 27.03
4.0
$ 108.12
6.0
$ 162.18
6.0
$ 162.18
501-1,000
$ 28.96
4.0
$ 115.84
6.0
$ 173.76
6.0
$ 173.76
1,001-4,100
$ 29.73
4.0
$ 118.92
6.0
$ 178.38
6.0
$ 178.38
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Transient Noncom m unity Water Systems (TNCWSs) - GW
<100
$ 25.10
4.0
$ 100.40
6.0
$ 150.60
6.0
$ 150.60
101-500
$ 27.03
4.0
$ 108.12
6.0
$ 162.18
6.0
$ 162.18
501-1,000
$ 28.96
4.0
$ 115.84
6.0
$ 173.76
6.0
$ 173.76
1,001-4,100
$ 29.73
4.0
$ 118.92
6.0
$ 178.38
6.0
$ 178.38
4,101-33,000
$ 36.00
30.0
$ 1,080.00
36.0
$ 1,296.00
36.0
$ 1,296.00
33,001-96,000
$ 36.39
59.0
$ 2,147.01
75.0
$ 2,729.25
75.0
$ 2,729.25
96,001-500,000
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
500,001-1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
> 1 Million
$ 41.01
108.0
$ 4,429.08
117.0
$ 4,798.17
117.0
$ 4,798.17
Note:
(F) EPA assumes that the burden incurred by operators to assess their FWSs follow ing a second non-acute violation is equal to the burden incurred
by an assessment follow ing an acute violation.
Sources:
(A) Labor rates for FWSs from Exhibit 7.1.
(B), (D), (F) Ftoposed RTCRT&CDocument.
Economic Analysis for the Proposed RTCR 7-20
June 2010
-------
Exhibit 7.12 PWS Unit Costs Estimates for Level 1 and Level 2 Assessments (AIP
and Alternative Option) (2007$)
PWS Size
(Population
Served)
Labor Cost
(per hour)
Level 1 Assessments
Level 2 Assessments
Non-Acute
Trigger
(single trigger)
(hours)
Unit Cost
Acute
Violations
(hours)
Unit Cost
Level 2 Triggers
(triggered by
multiple Level 1s)
(hours)
Unit Cost
A
B
C=A*B
D
E=A*D
F
G=A*F
Community Water Systems (CWSs) - SW
<100
$ 25.10
19.0
$ 476.90
23.0
$ 577.30
22.0
$ 552.20
101-500
$ 27.03
19.0
$ 513.57
23.0
$ 621.69
22.0
$ 594.66
501-1,000
$ 28.96
20.0
$ 579.20
24.0
$ 695.04
23.0
$ 666.08
1,001-4,100
$ 29.73
31.0
$ 921.63
48.0
$ 1,427.04
46.0
$ 1,367.58
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Community Water Systems (CWSs) - GW
<100
$ 25.10
19.0
$ 476.90
23.0
$ 577.30
22.0
$ 552.20
101-500
$ 27.03
19.0
$ 513.57
23.0
$ 621.69
22.0
$ 594.66
501-1,000
$ 28.96
20.0
$ 579.20
24.0
$ 695.04
23.0
$ 666.08
1,001-4,100
$ 29.73
31.0
$ 921.63
48.0
$ 1,427.04
46.0
$ 1,367.58
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Nontransient Noncom m unity Water Systems (NTNCWSs) - SW
<100
$ 25.10
7.0
$ 175.70
21.0
$ 527.10
9.0
$ 225.90
101-500
$ 27.03
7.0
$ 189.21
21.0
$ 567.63
9.0
$ 243.27
501-1,000
$ 28.96
7.0
$ 202.72
21.0
$ 608.16
9.0
$ 260.64
1,001-4,100
$ 29.73
8.0
$ 237.84
29.0
$ 862.17
10.0
$ 297.30
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Nontransient Noncom m unity Water Systems (NTNCWSs) - GW
<100
$ 25.10
7.0
$ 175.70
21.0
$ 527.10
9.0
$ 225.90
101-500
$ 27.03
7.0
$ 189.21
21.0
$ 567.63
9.0
$ 243.27
501-1,000
$ 28.96
7.0
$ 202.72
21.0
$ 608.16
9.0
$ 260.64
1,001-4,100
$ 29.73
8.0
$ 237.84
29.0
$ 862.17
10.0
$ 297.30
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Transient Noncom m unity Water Systems (TNCWSs) - SW
<100
$ 25.10
7.0
$ 175.70
21.0
$ 527.10
9.0
$ 225.90
101-500
$ 27.03
7.0
$ 189.21
21.0
$ 567.63
9.0
$ 243.27
501-1,000
$ 28.96
7.0
$ 202.72
21.0
$ 608.16
9.0
$ 260.64
1,001-4,100
$ 29.73
8.0
$ 237.84
29.0
$ 862.17
10.0
$ 297.30
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Transient Noncom m unity Water Systems (TNCWSs) - GW
<100
$ 25.10
7.0
$ 175.70
21.0
$ 527.10
9.0
$ 225.90
101-500
$ 27.03
7.0
$ 189.21
21.0
$ 567.63
9.0
$ 243.27
501-1,000
$ 28.96
7.0
$ 202.72
21.0
$ 608.16
9.0
$ 260.64
1,001-4,100
$ 29.73
8.0
$ 237.84
29.0
$ 862.17
10.0
$ 297.30
4,101-33,000
$ 36.00
41.0
$ 1,476.00
71.0
$ 2,556.00
69.0
$ 2,484.00
33,001-96,000
$ 36.39
68.0
$ 2,474.52
121.0
$ 4,403.19
116.0
$ 4,221.24
96,001-500,000
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
500,001-1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
> 1 Million
$ 41.01
159.0
$ 6,520.59
252.0
$ 10,334.52
238.0
$ 9,760.38
Sources:
(A) Labor rates for FWSs from Exhibit 7.1.
(B), (D), (F) Proposed RTCRT&C Document.
Economic Analysis for the Proposed RTCR 7-21
June 2010
-------
Exhibit 7.13 Number of Level 1 and Level 2 Assessments over the 25-Year
Compliance Period
PWS Size
(Population
Served)
TCR
AIP
Alternative Option
Level 1
Assessments
Level 2
Assessments
Level 1
Assessments
Level 2
Assessments
Level 1
Assessments
Level 2
Assessments
Non-Acute
Violations (single
violation)
Acute
Violations
Non-Acute
Violations
(multiple
violations)
Non-Acute
Trigger
(single trigger)
Acute
Violations
Level 2
Triggers
(triggered by
multiple
Level 1s)
Non-Acute
Trigger
(single trigger)
Acute
Violations
Level 2
Triggers
(triggered by
multiple
Level 1s)
A
B
C
D
E
F
G
H
I
Community Water Systems (CWSs) - SW
<100
541
156
168
363
112
73
363
112
73
101-500
621
181
112
465
135
55
465
135
55
501-1,000
339
99
61
254
74
30
254
74
30
1,001-4,100
931
149
123
846
161
96
846
161
96
4,101-33,000
2,152
197
2,152
197
2,152
197
33,001-96,000
534
56
534
56
534
56
96,001-500,000
233
24
233
24
233
24
500,001-1 Million
22
22
22
> 1 Million
Total
5,373
861
463
4,869
758
253
4,869
758
253
Community Water Systems (CWSs) - GW
<100
9,772
1,141
5,383
8,004
853
3,523
7,871
926
3,272
101-500
8,169
1,025
4,214
6,502
696
2,399
6,495
747
2,543
501-1,000
2,250
284
1,050
1,780
188
626
1,772
203
607
1,001-4,100
3,545
477
2,808
3,208
342
1,705
3,208
342
1,705
4,101-33,000
4,545
263
4,545
263
4,545
263
33,001-96,000
656
53
656
53
656
53
96,001-500,000
129
10
129
10
129
10
500,001-1 Million
> 1 Million
Total
29,066
3,253
13,455
24,824
2,405
8,253
24,675
2,544
8,127
Nontransient Noncommunity Water Systems (NTNCWSs) - SW
<100
93
38
67
69
27
33
69
27
33
101-500
80
29
39
66
24
21
66
24
21
501-1,000
28
10
14
23
8
7
23
8
7
1,001-4,100
43
16
33
38
12
18
38
12
18
4,101-33,000
5
5
5
33,001-96,000
96,001-500,000
500,001-1 Million
> 1 Million
Total
248
92
153
201
72
79
201
72
79
Nontransient Noncommunity Water Systems (NTNCWSs) - GW
<100
5,581
856
3,829
4,797
559
2,010
5,506
703
3,314
101-500
3,130
447
1,273
2,794
315
757
3,450
435
1,307
501-1,000
744
95
298
675
79
168
804
98
295
1,001-4,100
818
169
974
690
114
530
690
114
530
4,101-33,000
123
9
123
9
123
9
33,001-96,000
4
4
4
96,001-500,000
500,001-1 Million
> 1 Million
Total
10,400
1,577
6,373
9,084
1,077
3,466
10,578
1,359
5,447
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
1,277
508
943
1,004
336
426
1,004
336
426
101-500
487
202
372
372
125
165
372
125
165
501-1,000
86
36
66
66
22
29
66
22
29
1,001-4,100
102
52
122
90
32
62
90
32
62
4,101-33,000
8
8
8
33,001-96,000
96,001-500,000
500,001-1 Million
> 1 Million
Total
1,960
797
1,504
1,541
515
682
1,541
515
682
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
44,730
6,649
25,425
47,190
5,477
20,628
54,441
7,386
35,932
101-500
14,530
2,089
8,864
13,780
1,608
5,694
16,390
2,313
10,467
501-1,000
1,477
221
896
1,396
177
585
1,602
224
989
1,001-4,100
927
186
1,138
773
117
638
773
117
638
4,101-33,000
116
4
116
4
116
4
33,001-96,000
96,001-500,000
500,001-1 Million
> 1 Million
Total
61,780
9,149
36,324
63,256
7,383
27,546
73,323
10,043
48,026
Grand Total
108,828
15,728
58,272
103,775
12,210
40,278
115,187
15,291
62,614
Note: Detail may not add due to rounding.
Sources:
(A), (B), (C) - A.1.Z
(D), (E), (F) - A.2.z
(G),(H), (l)-A.3.z
Economic Analysis for the Proposed RTCR 7-22
June 2010
-------
States
Level 1 and Level 2 Assessments
Under the TCR, AIP and Alternative option, states would incur burden to review
completed assessment forms required to be filed by PWSs (or similar reports required by states
under the current TCR). Although some states may choose to conduct assessments for their
PWSs, EPA does not quantify these costs. State costs are based on the number of PWSs
submitting assessment reports. The state labor rate presented in Section 7.2.1 and estimates of
labor hours are used to generate state Level 1 and Level 2 assessment unit costs. EPA estimates
that state burden to review PWS assessment forms would range from one to eight hours
depending on PWS size and type, as well as the level of the assessment. This burden includes
any time required to consult with the PWS about the assessment report. Estimates of state unit
costs for Level 1 and Level 2 assessments are presented in Exhibit 7.14.
Assessment Net Cost Summary
Annualized cost estimates for Level 1 and Level 2 assessments under the TCR, AIP, and
Alternative Option are calculated by multiplying the number of assessments estimated by the
predictive modeling (summarized in Exhibit 7.13) by the unit costs presented in Exhibits 7.11,
7.12, and 7.14. Exhibit 7.13 presents the estimated totals of non-acute and acute MCL violations
(current TCR) and Level 1 and Level 2 assessments (AIP and Alternative options). The model
predicts a total of approximately 109,000 single non-acute MCL violations, 58,000 cases of a
second non-acute MCL violation, and 16,000 acute MCL violations for the current TCR under
which some PWSs do currently engage in assessment activity which may or may not meet the
proposed RTCR criteria (see section 7.4.5). For the AIP option, the model predicts
approximately 104,000 Level 1 assessments and 52,000 Level 2 assessments. For the Alternative
option, the model predicts approximately 115,000 Level 1 assessments and 78,000 Level 2
assessments. (Appendix A provides a detailed breakout of the number of Level 1 and Level 2
assessments estimated by the occurrence model.) Annualized cost estimates are presented in
Exhibit 7.15 below.
Under the proposed RTCR, all PWSs are required to conduct assessments of their
systems when they exceed Level 1 or Level 2 treatment technique triggers. While PWSs are not
required to conduct assessments under the current TCR, some PWSs do currently engage in
assessment activity (which may or may not meet the proposed RTCR criteria) following
nonacute and acute MCL violations. EPA estimates both the costs to PWSs to conduct
assessments under the proposed RTCR as well as the level of effort that PWSs already put
towards assessment activities under the current TCR; these estimates are based on the work of
the stakeholders in the TWG during the proceedings of the TCRDSAC. These estimates allowed
EPA to determine the average net costs to conduct assessments under the proposed RTCR. EPA
assumes that the numbers of nonacute and acute MCL violations would remain steady under a
continuation of the current TCR (based on review of SDWIS/FED violation data). Under the
proposed RTCR, EPA assumes that the numbers of assessments decreases from the steady state
level seen under the current TCR over time to a new steady state level as a function of reduced
fecal indicator occurrence associated with the effects of requiring assessments and corrective
action.
Economic Analysis for the Proposed RTCR 7-23
June 2010
-------
The overall number of assessments increases for under the Alternative option. This is a
result of the initial monthly monitoring requirements for all PWSs under this analysis. The
modeling results indicate that the much higher sampling early in the implementation period,
results in more positive samples and associated assessments despite the predicted long term
reductions in occurrence. This increase in total assessments performed, combined with the higher
unit cost of performing assessments compared to existing practices under the TCR, results in a
higher net cost increase than under the AIP. For states, the increase in the number of assessments
is estimated to translate directly to a cost increase. The total net change in cost for the Alternative
option is estimated to be positive, and greater than under the AIP.
Economic Analysis for the Proposed RTCR 7-24
June 2010
-------
Exhibit 7.14 State Unit Cost Estimates for Review of Level 1 and Level 2
Assessments under TCR, AIP, and Alternative Options (2007$)
PWS Size
(Population
Served)
Labor Cost
(per hour)
Level 1 Assessments
Level 2 Assessments
Non-Acute
Trigger
(single trigger)
(hours)
Unit Cost
Acute
Violations
(hours)
Unit Cost
Level 2 Triggers
(triggered by
m ultiple Level 1s)
(hours)
Unit Cost
A
B
C=A*B
D
E=A*D
F
G=A*F
Com m unity Water Systems (CWSs) - SW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Com m unity Water Systems (CWSs) - GW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Nontransient Noncom m unity Water Systems (NTNCWSs) - SW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Nontransient Noncom m unity Water Systems (NTNCWSs) - GW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
101-500
$ 39.22
1.0
$ 39.22
2.0
$ 78.44
2.0
$ 78.44
501-1,000
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
1,001-4,100
$ 39.22
2.0
$ 78.44
4.0
$ 156.87
4.0
$ 156.87
4,101-33,000
$ 39.22
3.0
$ 117.65
6.0
$ 235.31
6.0
$ 235.31
33,001-96,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
96,001-500,000
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
500,001-1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
> 1 Million
$ 39.22
4.0
$ 156.87
8.0
$ 313.74
8.0
$ 313.74
Sources:
(A) Labor rates for state employee from Section 7.2.1.
(B), (D), (E) Labor hour assumptions based on best professional judgement.
Economic Analysis for the Proposed RTCR 7-25
June 2010
-------
Exhibit 7.15 Annualized PWS and State Cost Estimates for Level 1 and Level 2
Assessments ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
Level 1 Assessment
TCR-Total
$
1.13
$
0.21
$
1.34
$
1.08
$
0.21
$
1.29
AIP - Total
$
1.63
$
0.20
$
1.84
$
1.57
$
0.20
$
1.77
AIP - Net Change
$
0.50
$
(0.01)
$
0.49
$
0.49
$
(0.01)
$
0.48
Alternative Option - Total
$
1.73
$
0.23
$
1.95
$
1.69
$
0.22
$
1.91
Alternative Option - Net Change
$
0.60
$
0.01
$
0.61
$
0.60
$
0.02
$
0.62
Level 2 Assessment
TCR - Total
$
0.70
$
0.26
$
0.96
$
0.68
$
0.25
$
0.93
AIP - Total
$
0.90
$
0.19
$
1.09
$
0.88
$
0.18
$
1.06
AIP - Net Change
$
0.20
$
(0.07)
$
0.12
$
0.20
$
(0.07)
$
0.13
Alternative Option - Total
$
1.23
$
0.28
$
1.51
$
1.27
$
0.30
$
1.57
Alternative Option - Net Change
$
0.52
$
0.02
$
0.55
$
0.60
$
0.05
$
0.65
Note: Detail may not add due to independent rounding.
Source: Proposed RTCR Cost Model.
7.4.6 Corrective Actions
PWSs
Under the AIP and Alternative option, all PWSs would be required to correct sanitary
defects found through the performance of Level 1 or Level 2 assessments. For modeling
purposes, EPA estimated only the net change in the number of corrective actions performed
under the AIP and Alternative option6 compared to the current TCR. Based on discussions with
state representatives, EPA estimates that additional corrective actions (on top of those already
being performed under the existing TCR) would be performed for only 10% of the assessments
undertaken as a result of the proposed RTCR. Because only the net change in costs is estimated,
no additional costs for corrective actions are modeled for the existing TCR (it is assumed that
PWSs are already performing corrective actions under the existing TCR).
To estimate the costs incurred for the correction of sanitary defects, EPA estimated the
percent distribution of PWSs that would perform different types of corrective actions as
presented in the compliance forecast below (Exhibit 7.16a). The compliance forecast presented
below focuses on the higher level categorization of corrective actions anticipated. The categories
of anticipated corrective actions were informed by both EPA judgment and discussions of the
TCRDSAC TWG and are essentially the same as those presented to the advisory committee. For
each of the categories listed, a PWS is assumed to take a specific action that falls under that
general category. Exhibit 7.16b lists the specific corrective actions that fall under the higher level
categorization, along with the percent of PWSs estimated to perform that corrective action based
on level 1 and level 2 assessments. Detailed compliance forecasts showing the specific corrective
actions used in the cost analysis are provided in Appendix D, along with summary tables of the
unit costs used in the analysis. Each corrective action in the detailed compliance forecast is also
assigned a representative unit cost. Detailed descriptions of the derivation of unit costs are
6 Any corrective actions based on a positive source water sample are assumed to be accounted for under the GWR
and not under the proposed RTCR.
Economic Analysis for the Proposed RTCR 7-26
June 2010
-------
provided in Exhibits 5-1 through 5-47 of the Technology and Cost Document for the Proposed
Revised Total Coliform Rule (USEPA, 2010d).
As shown in the compliance forecast in Exhibit 7.16a, EPA estimates that corrective
actions found through Level 1 assessments would result in corrective actions that focus more on
actions by PWS staff such as flushing or training (columns A and B) than on permanent fixes to
the PWS. This reflects the assumption that Level 1 assessments would generally be less involved
than Level 2 assessments and may not result in finding more complex problems. Corrective
actions taken as a result of Level 2 assessments are expected to find a higher proportion of
structural/technical issues (columns C-K) resulting in material fixes to the PWSs and distribution
system. Consistent with the discussions of the TCRDSAC regarding major structural fixes or
replacements, EPA did not include these major costs in the analysis. Distribution system
appurtenances such as storage tanks generally have a useful life that is accounted for in water
system capital planning and the assessments conducted in response to RTCR triggers could
identify when that useful life has ended but are not solely responsible for the need to correct the
defect.7
It is not anticipated that the overall effectiveness of corrective actions will differ
significantly by the type of corrective action. Some activities, such as spot flushing, may have a
shorter, transient impact on water quality. However, most PWSs are expected to institute
ongoing flushing programs that will have a continuing impact on water quality. Other corrective
actions in the compliance forecast are expected to have ongoing benefits, though the exact
duration may be highly variable depending on the exact conditions at the PWS. The overall
duration of reduced occurrence modeled as a result of corrective actions initiated in response to
either a Level 1 or Level 2 assessment is intended to reflect an average effectiveness of
corrective actions. Uncertainties and associated sensitivity analyses that inform the effect of the
reduced occurrence predicted from corrective actions are discussed further in Section 7.7.
PWSs would also incur reporting and recordkeeping burden to notify the state upon
completion of each corrective action. PWSs may also consult with the state or with outside
parties to determine the appropriate corrective action to be implemented. PWS reporting and
recordkeeping costs (including consultations) are derived by multiplying PWSs labor rates
(Section 7.2.1) by an EPA-estimated labor burden. Exhibit 7.17 presents the estimated unit costs
for this reporting and recordkeeping burden.
States
For each corrective action performed under AIP and Alternative option, states would
incur recordkeeping and reporting burden to review and coordinate with PWSs. This includes
burden incurred from any optional consultations states may conduct with PWSs or outside
parties to determine the appropriate corrective action to be implemented. The state labor rate
presented in Section 7.2.1 and estimates of labor hours are used to generate state unit corrective
7 Additionally, EPA ran two sensitivity analyses to assess the potential impacts of different distributions within the
compliance forecast. Results of the sensitivity analyses are presented in Section 7.7, and indicate that the low bound
estimates of annualized net change in costs at three percent discount rate are approximately $3M for the AIP option
and $15M for the Alternative option, and the high bound estimates are approximately $25M for the AIP option and
$40M for the Alternative option. Varying the assumptions about the percentage of corrective actions identified and
the effectiveness of those actions had less than a linear effect on outcomes, and the AIP option continues to be less
costly than the Alternative option under all scenarios modeled.
Economic Analysis for the Proposed RTCR 7-27
June 2010
-------
action costs. Exhibit 7.17 presents the estimated PWS and state reporting and recordkeeping unit
costs (including consultations) by PWS size.
Exhibit 7.16a Compliance Forecast for Corrective Actions based on Level 1 and
Level 2 Assessments
Maintenance
Cross-
Addition or
Development
Replace/Repair
of appropriate
connection
Upgrade of
and
of Distribution
Maintenance
Hydraulic
Storage
Control and
On-line
Addition
Implementation
PWS Size
PWS
Sam pier
System
of Adequate
Residence
Facility
Booster
Back flow
Monitoring
of Security
of an
(Population
Flushing
Training
Com ponents
Pressure
Time
Maintenance
Disinfection
Prevention
and Control
Measures
Operations Plan
Served)
A
B
C
D
E
F
G
H
I
J
K
Level 1 Compliance Forecast
<100
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
101-500
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
501-1,000
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
1,001-4,100
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
4,101-33,000
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
33,001-96,000
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
96,001-500,000
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
500,001-1 Million
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
> 1 Million
39%
15%
12%
9%
8%
6%
4%
1%
3%
1%
2%
Level 2 Compliance Forecast
<100
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
101-500
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
501-1,000
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
1,001-4,100
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
4,101-33,000
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
33,001-96,000
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
96,001-500,000
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
500,001-1 Million
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
> 1 Million
15%
4%
18%
15%
15%
11%
8%
2%
6%
2%
4%
Source: (A) - (K) Ftercent of FWSs performing corrective actions based on Level 1 and Level 2 assessments reflect EPA estimate.
Economic Analysis for the Proposed RTCR 7-28
June 2010
-------
Exhibit 7.16b Detailed PWS Compliance Forecast for Corrective Actions based
on Level 1 and Level 2 Assessments
General Corrective Action
Category
Specific Corrective Action
Level 1
Compliance
Forecast
Level 2
Compliance
Forecast
PWS Flushing
Routine Flushing
30%
10%
Spot Flushing
9%
5%
Sampler Training
Operator Training/Certification
15%
4%
Replace/Repair of Distribution
System Components
Replace Valve
2%
3%
Installing a Dedicated Sampling Tap
2%
3%
Replace Iron Pipe
2%
3%
Replace Fittings
2%
3%
Replace Hydrant
2%
3%
Replace Meter
2%
3%
Maintenance of Adequate
Pressure
New Booster Pump Station
2%
3%
Modify/Replace Existing Pumps
2%
3%
Install Variable Frequency Drives
2%
3%
Maintenance of Adequate
Pressure
New Elevated Storage Tank
0%
0%
Install Surge Relief valve
2%
3%
Install Surge Control Tank
1%
3%
Maintenance of appropriate
Hydraulic Residence Time
Install Auto Flushing Devices
2%
3%
Modify Inlet/Outlet Piping
1%
2%
Install Mixing Devices
1%
2%
Loop Dead Ends
1%
2%
Install Appropriate Main Sizes
1%
2%
Modify Storage Operation
1%
2%
Decommission Storage
1%
2%
Storage Facility Maintenance
Inspect & Clean Storage Tanks
2%
3%
Line Storage Tanks
1%
2%
Vent Repair/Replace Vent Screen
1%
2%
Storage Facility Maintenance
Repair/Replace Tank Hatch
1%
2%
Repair Storage Tank
1%
2%
Booster Disinfection
Install Permanant CI Booster Station
1%
2%
Install Temp CI Booster Station
1%
2%
Install Permanant NH2CI Booster Station
1%
2%
Install Temp NH2CI Booster Station
1%
2%
Cross-connection Control and
Backflow Prevention Assembly
1%
2%
Addition or Upgrade of On-line
Monitoring and Control
Online CI Monitoring & Programming
1%
2%
Online NH2CI Monitoring & Programming
1%
2%
Online Pressure Monitoring & Programming
1%
2%
Addition of Security Measures
Installation of Additional Security Measures
1%
2%
Development and
Implementation of an
Operations Plan
Develop/1 mplement/Maintain Operations
Plan
1%
2%
Operator Training/Certification
1%
2%
Economic Analysis for the Proposed RTCR 7-29
June 2010
-------
Exhibit 7.17 Net Change in PWS and State Unit Costs Estimates for Reporting
and Recordkeeping for Corrective Actions (2007$)
Systems
States
PWS Size
(Population
Labor Cost
(per hour)
Corrective Action Burden
(hours/corrective action)
Unit Cost
Labor Cost(per
hour)
Corrective Action Burden
(hours/corrective action)
Unit Cost
Served)
A
B
C=A*B
D
E
F=D*E
<100
$
25.10
0.5
$
12.55
$
39.22
0.5
$
19.61
101-500
$
27.03
0.5
$
13.52
$
39.22
0.5
$
19.61
501-1,000
$
28.96
0.5
$
14.48
$
39.22
0.5
$
19.61
1,001-4,100
$
29.73
0.5
$
14.87
$
39.22
0.5
$
19.61
4,101-33,000
$
36.00
0.5
$
18.00
$
39.22
0.5
$
19.61
33,001-96,000
$
36.39
0.5
$
18.20
$
39.22
0.5
$
19.61
96,001-500,000
$
41.01
0.5
$
20.51
$
39.22
0.5
$
19.61
500,001-1 Million
$
41.01
0.5
$
20.51
$
39.22
0.5
$
19.61
> 1 Million
$
41.01
0.5
$
20.51
$
39.22
0.5
$
19.61
Notes:
FWS and state burden estimates identical for all FWS types under TCR, AIP, and Alternative Option.
Sources:
(A) Labor rates for FWSs from Exhibit 7.1.
(B) FWS labor hours for corrective action recordkeeping/reporting reflect EPA estimate.
(D) Labor rates for state employee from Section 7.2.1.
(E) State labor hours for corrective action recordkeeping/reporting reflect EPA estimate.
Corrective Action Net Cost Summary
Annualized net cost estimates for PWSs and states to perform corrective actions are
estimated by multiplying the number of Level 1 and Level 2 corrective actions estimated by the
predictive model (i.e., 10 percent of Level 1 and Level 2 assessments representing the net
increase in corrective actions found), by the percentages in the compliance forecast and unit
costs of corrective actions and associated reporting and recordkeeping. The incremental and net
change costs of corrective actions are shown in Exhibit 7.18.
Because only the net change in corrective actions taken is modeled, no costs are
estimated under the TCR. The differences in the net change in corrective action costs between
the AIP and Alternative Option are a function different number of assessments estimated to be
performed in the predictive model as discussed in Section 7.4.5 above.
Economic Analysis for the Proposed RTCR 7-30
June 2010
-------
Exhibit 7.18 Annualized PWS and State Cost Estimates for Corrective Actions
based on Level 1 and Level 2 Assessments ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
Corrective Actions based on Level 1 Assessments
TCR-Incremental
$
$
$
$
$
$
AIP - Incremental
$ 9.17
$ 0.01
$ 9.18
$ 7.77
$ 0.01
$ 7.77
AIP - Net Change
$ 9.17
$ 0.01
$ 9.18
$ 7.77
$ 0.01
$ 7.77
Alternative Option - Incremental
$ 9.39
$ 0.01
$ 9.40
$ 8.01
$ 0.01
$ 8.02
Alternative Option - Net Change
$ 9.39
$ 0.01
$ 9.40
$ 8.01
$ 0.01
$ 8.02
Corrective Actions based on Level 2 Assessments
TCR-Incremental
$
$
$
$
$
$
AIP - Incremental
$ 2.72
$ 0.00
$ 2.72
$ 2.41
$ 0.00
$ 2.41
AIP - Net Change
$ 2.72
$ 0.00
$ 2.72
$ 2.41
$ 0.00
$ 2.41
Alternative Option - Incremental
$ 3.53
$ 0.01
$ 3.53
$ 3.36
$ 0.01
$ 3.37
Alternative Option - Net Change
$ 3.53
$ 0.01
$ 3.53
$ 3.36
$ 0.01
$ 3.37
Note: Detail may not add due to independent rounding.
Source: Roposed RTCR Cost Model.
7.4.7 Public Notification
PWSs
Tier 1 Public Notification
Acute violations (E.coli MCL violations) would require Tier 1 public notification under
all regulatory scenarios (TCR, AIP, and Alternative option). PWSs with acute violations must
report the violation to the state by the end of the business day after the PWS learns of the
violation and must notify the public within 24 hours.
Tier 2 Public Notification
PWSs with non-acute violations under the current TCR must report the violation to the
state by the end of the business day after the PWS learns of the violation, and must provide Tier
2 public notification within 30 days. PWSs with similar violations under the AIP and Alternative
option are not required to perform Public Notification (PN). These PWSs are required to perform
Tier 2 notification for a treatment technique violation (failure to perform Level 1 or 2 assessment
if triggered; failure to correct all sanitary defects), but because the cost model assumes full
compliance with proposed RTCR requirements, no cost is estimated for these violations. Overall,
costs decrease significantly for Tier 2 public notification under the proposed RTCR for both the
AIP and Alternative option.
Tier 3 Public Notification
Under the TCR, AIP, and Alternative option, Tier 3 public notification for monitoring
and reporting violations are assumed to be reported once per year as part of the Consumer
Confidence Report (CCR). Because of the use of the CCR to communicate Tier 3 public
notification on a yearly basis, no cost differential between the current TCR and the AIP and
Alternative option is estimated in the cost model. However, although they are not quantitatively
evaluated as part of the EA, the TCRDSAC concluded that significant reductions in monitoring
Economic Analysis for the Proposed RTCR 7-31
June 2010
-------
and reporting violations may be realized through the revised regulatory framework of the
proposed RTCR. Resources used to process monitoring violation notices for the Consumer
Confidence Report (CCR) (as well as the time spent responding to customer inquiries about the
notices) may be reduced if significant reductions are realized. Exclusion of this potential cost
savings may lead to an underestimate of the PN cost savings under both the AIP and Alternative
option.
Estimates of PWS unit costs for PN are derived by multiplying PWS labor rates from
Section 7.2.1 and burden hour estimates derived from the Draft Information Collection Request
for the Public Water System Supervision Program (USEPA, 2008c). PWS PN unit cost estimates
are presented in Exhibit 7.19.
Economic Analysis for the Proposed RTCR 7-32
June 2010
-------
Exhibit 7.19 PWS Unit Cost Estimates for Public Notification (2007$)
Tier 1 (acute)
Tier 2 (non-acute)
PWS Size
(Population
Served)
Labor Cost
(per hour)
Average Number
of Service
Connections per
System
Preparation
(labor) (hours/
violation)
Distribution
(labor)
(hours/
violation)
Distribution
(O&M Cost/
notice)
($/service
connection)
Unit Cost
Preparation
(labor)
(hours/
violation)
Distribution
(labor) (hours/
violation)
Distribution
(O&M Cost/notice)
($/service
connection)
Unit Cost
A
B
C
D
E
F=A*(C+D)+(B*E)
G
H
I
J=A*(G+H)+(B*I)
Community Water Systems (CWSs) - SW
<100
$ 25.10
440
8.5
12.0
$ 0.05
$ 536.55
3.5
9.0
$ 0.05
$ 335.75
101-500
$ 27.03
387
8.5
12.0
$ 0.05
$ 573.44
3.5
9.0
$ 0.05
$ 357.20
501-1,000
$ 28.96
303
8.5
12.0
$
$ 593.68
3.5
30.0
$ 0.26
$ 1,049.05
1,001-4,100
$ 29.73
850
8.5
12.0
$
$ 609.47
3.5
30.0
$ 0.26
$ 1,216.99
4,101-33,000
$ 36.00
4,288
9.2
12.0
$
$ 763.66
3.5
30.0
$ 0.25
$ 2,259.65
33,001-96,000
$ 36.39
17,273
10.0
12.0
$
$ 800.58
3.5
30.0
$ 0.23
$ 5,191.88
96,001-500,000
$ 41.01
56,465
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 14,360.67
500,001-1 Million
$ 41.01
205,609
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 48,663.97
> 1 Million
$ 41.01
448,564
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 104,543.57
Community Water Systems (CWSs) - GW
<100
$ 25.10
41
8.5
12.0
$ 0.05
$ 516.62
3.5
9.0
$ 0.05
$ 315.82
101-500
$ 27.03
99
8.5
12.0
$ 0.05
$ 559.05
3.5
9.0
$ 0.05
$ 342.81
501-1,000
$ 28.96
315
8.5
12.0
$
$ 593.68
3.5
30.0
$ 0.26
$ 1,052.07
1,001-4,100
$ 29.73
756
8.5
12.0
$
$ 609.47
3.5
30.0
$ 0.26
$ 1,192.61
4,101-33,000
$ 36.00
3,495
9.0
12.0
$
$ 757.48
3.5
30.0
$ 0.25
$ 2,076.93
33,001-96,000
$ 36.39
16,366
10.0
12.0
$
$ 800.58
3.5
30.0
$ 0.23
$ 4,983.17
96,001-500,000
$ 41.01
50,564
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 13,003.50
500,001-1 Million
$ 41.01
209,220
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 49,494.49
> 1 Million
$ 41.01
473,641
10.0
12.0
$
$ 902.22
3.5
30.0
$ 0.23
$ 110,311.34
Nontransient Noncommunil
y Water System s (NTNCWSs) - SW
<100
$ 25.10
128
8.5
12.0
$ 0.05
$ 520.94
3.5
9.0
$ 0.05
$ 320.14
101-500
$ 27.03
21
8.5
12.0
$ 0.05
$ 555.16
3.5
9.0
$ 0.05
$ 338.92
501-1,000
$ 28.96
46
8.5
12.0
$
$ 593.68
3.5
9.0
$ 0.05
$ 364.29
1,001-4,100
$ 29.73
47
8.5
12.0
$
$ 609.47
3.5
9.0
$ 0.05
$ 373.97
4,101-33,000
$ 36.00
176
8.7
12.0
$
$ 745.36
3.5
9.0
$ 0.05
$ 458.08
33,001-96,000
$ 36.39
94
10.0
12.0
$
$ 800.58
3.5
9.0
$ 0.02
$ 456.75
96,001-500,000
$ 41.01
2,181
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 556.25
500,001-1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
> 1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
Nontransient Noncommunil
y Water System s (NTNCWSs) - GW
<100
$ 25.10
4
8.5
12.0
$ 0.05
$ 514.76
3.5
9.0
$ 0.05
$ 313.96
101-500
$ 27.03
8
8.5
12.0
$ 0.05
$ 554.52
3.5
9.0
$ 0.05
$ 338.28
501-1,000
$ 28.96
11
8.5
12.0
$
$ 593.68
3.5
9.0
$ 0.05
$ 362.56
1,001-4,100
$ 29.73
42
8.5
12.0
$
$ 609.47
3.5
9.0
$ 0.05
$ 373.71
4,101-33,000
$ 36.00
130
8.7
12.0
$
$ 744.94
3.5
9.0
$ 0.05
$ 456.02
33,001-96,000
$ 36.39
75
10.0
12.0
$
$ 800.58
3.5
9.0
$ 0.02
$ 456.38
96,001-500,000
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
500,001-1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
> 1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
Transient Noncommunity Water Systems (TNCWSs) -SW
<100
$ 25.10
9
8.5
12.0
$ 0.05
$ 514.99
3.5
9.0
$ 0.05
$ 314.19
101-500
$ 27.03
30
8.5
12.0
$ 0.05
$ 555.62
3.5
9.0
$ 0.05
$ 339.38
501-1,000
$ 28.96
49
8.5
12.0
$
$ 593.68
3.5
9.0
$ 0.05
$ 364.45
1,001-4,100
$ 29.73
58
8.5
12.0
$
$ 609.47
3.5
9.0
$ 0.05
$ 374.50
4,101-33,000
$ 36.00
57
8.8
12.0
$
$ 747.00
3.5
9.0
$ 0.05
$ 452.55
33,001-96,000
$ 36.39
-
10.0
12.0
$
$ 800.58
3.5
9.0
$ 0.02
$ 454.88
96,001-500,000
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
500,001-1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
> 1 Million
$ 41.01
2
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.67
Transient Noncommunity Water Systems (TNCWSs) -GW
<100
$ 25.10
5
8.5
12.0
$ 0.05
$ 514.82
3.5
9.0
$ 0.05
$ 314.02
101-500
$ 27.03
15
8.5
12.0
$ 0.05
$ 554.87
3.5
9.0
$ 0.05
$ 338.63
501-1,000
$ 28.96
30
8.5
12.0
$
$ 593.68
3.5
9.0
$ 0.05
$ 363.51
1,001-4,100
$ 29.73
43
8.5
12.0
$
$ 609.47
3.5
9.0
$ 0.05
$ 373.77
4,101-33,000
$ 36.00
39
8.7
12.0
$
$ 746.87
3.5
9.0
$ 0.05
$ 451.77
33,001-96,000
$ 36.39
14
10.0
12.0
$
$ 800.58
3.5
9.0
$ 0.02
$ 455.15
96,001-500,000
$ 41.01
9
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.81
500,001-1 Million
$ 41.01
1
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.65
> 1 Million
$ 41.01
-
10.0
12.0
$
$ 902.22
3.5
9.0
$ 0.02
$ 512.63
Notes:
(B) Service connections per system is consistent w ith SCWIS 2007 4th Quarter Freeze data. Data for certain size categories (e.g., <100, 101-500, 501-1,000) may seem counterintuitive. EPA
is investigating the SCWIS database for any data discrepancies.
(F), (J) used to derived TCR PN costs; (F) used to derive AIP and Alternative Option PN costs.
Sources:
(A) Labor rates for FW3s from Exhibit 7.1.
(B) SCWIS 2007 4th Quarter Freeze.
(C), (D), (G), (H) Labor hour assumptions based on best professional judgement as carried forw ard from the Draft Information Collection Request for the Public Water System Supervision
Program (USEPA, 2008x).
(E), (I) Distribution cost assumptions based on best professional judgement as carried forw ard from the original PN ICR.
Economic Analysis for the Proposed RTCR 7-33
June 2010
-------
States
Under the TCR, AIP, and Alternative option, states would incur recordkeeping and
reporting burden to provide consultation, review the public notification certification, and file the
report of the violation. State unit PN costs are based on the state labor rate presented in Section
7.2.1 and burden hour estimates derived from the Draft Information Collection Request for the
Public Water System Supervision Program (USEPA, 2008c). Estimates of state unit costs for PN
are presented in Exhibit 7.20.
Exhibit 7.20 State Unit Costs Estimates for Public Notification (TCR, AIP,
Alternative Options)
Labor Cost
(per hour)
Tier 1
Consultation
(labor)
(hours/violation)
Tier 2
Consultation
(labor)
(hours/violation)
Receive/Review PN
Certification
(labor)
(hours/violation)
File Reports
(labor)
(hours/violation)
Tier 1
Unit Cost
Tier 2
Unit Cost
A
B
C
D
E
F=A*(B+I>-E)
G=A*(C+OE)
$ 39.22
3.0
1.1
$ 0.20
$ 0.10
$ 129.42
$ 54.91
Sources:
(A) Labor rate for state errployee from Section 7.2.1.
(B), (C), (D), (E) Labor hour assumptions based on best professional judgement as carried forward fromthe Draft Information Collection
Request for the Rjblic Water System Supervision Rogram (USEPA, 2008x).
Public Notification Net Cost Summary
Total and net change in annualized net present value costs for PN are estimated by
multiplying the model estimates of PWSs with acute (Tier 1 public notification) and non-acute
(Tier 2 public notification) violations by the PWS and state unit costs for performing PN
activities. The proposed RTCR Cost Model assumes that all violations are addressed following
initial PN, and no burden would be incurred by PWSs or states for repeat notification. Exhibit
7.21 summarizes the total number of Tier 1 and Tier 2 public notifications that would be
prepared by PWS size and category for under the TCR, AIP, and Alternative Option over the
entire 25 year period of analysis. Annualized total and net cost estimates for PWSs and states to
perform public notification under the TCR, AIP, and Alternative Option are presented in Exhibit
7.22.
A significant reduction in costs is estimated due to the elimination of Tier 2 public
notification under the AIP and Alternative option. Because state costs are calculated on a per-
violation basis, state costs decline. Under the Alternative option, some of this cost decrease is
offset by additional Tier 1 public notification.
Economic Analysis for the Proposed RTCR 7-34
June 2010
-------
Exhibit 7.21 Number of Tier 1 and Tier 2 Public Notifications over the 25-Year
Compliance Period
PWS Size
(Population
Served)
TCR
AIP
Alternative
Option
Tier 1 PN
Tier 2 PN
Tier 1 PN
Tier 1 PN
A
B
C
D
Community Water Systems (CWSs) - SW
<100
156
708
112
112
101-500
181
733
135
135
501-1,000
99
400
74
74
1,001-4,100
149
1,055
161
161
4,101-33,000
197
2,152
197
197
33,001-96,000
56
534
56
56
96,001-500,000
24
233
24
24
500,001-1 Mllion
22
> 1 Mllion
Total
861
5,836
758
758
Community Water Systems (CWSs) - GW
<100
1,141
15,155
853
926
101-500
1,025
12,383
696
747
501-1,000
284
3,300
188
203
1,001-4,100
477
6,354
342
342
4,101-33,000
263
4,545
263
263
33,001-96,000
53
656
53
53
96,001-500,000
10
129
10
10
500,001-1 Mllion
> 1 Mllion
Total
3,253
42,521
2,405
2,544
Nontransient Noncommunity Water Systems (NTNCWSs) - SW
<100
38
160
27
27
101-500
29
120
24
24
501-1,000
10
42
8
8
1,001-4,100
16
75
12
12
4,101-33,000
5
33,001-96,000
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Total
92
401
72
72
Nontransient Noncommunity Water Systems (NTNCWSs) - GW
<100
856
9,411
559
703
101-500
447
4,402
315
435
501-1,000
95
1,042
79
98
1,001-4,100
169
1,792
114
114
4,101-33,000
9
123
9
9
33,001-96,000
4
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Total
1,577
16,774
1,077
1,359
Transient Noncommunity Water Systems (TNCWSs) - SW
<100
508
2,220
336
336
101-500
202
859
125
125
501-1,000
36
152
22
22
1,001-4,100
52
225
32
32
4,101-33,000
8
33,001-96,000
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Total
797
3,464
515
515
Transient Noncommunity Water Systems (TNCWSs) - GW
<100
6,649
70,156
5,477
7,386
101-500
2,089
23,394
1,608
2,313
501-1,000
221
2,373
177
224
1,001-4,100
186
2,065
117
117
4,101-33,000
4
116
4
4
33,001-96,000
96,001-500,000
500,001-1 Mllion
> 1 Mllion
Total
9,149
98,104
7,383
10,043
Grand Total
15,728
167,100
12,210
15,291
Note: Cetail may not add due to rounding.
Sources:
(A), (B)-A.1.z
(C) - A.2.Z
(D) - A.3.Z
Economic Analysis for the Proposed RTCR 7-35
June 2010
-------
Exhibit 7.22 Annualized PWS and State Cost Estimates for Public Notification
($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR-Total
$ 3.75
$ 0.44
$ 4.19
$ 3.60
$ 0.42
$ 4.03
AIP-Total
$ 0.26
$ 0.06
$ 0.33
$ 0.26
$ 0.06
$ 0.32
AIP - Net Change
$ (3.49)
$ (0.38)
$ (3.87)
$ (3.35)
$ (0.36)
$ (3.71)
AIP - Percent Change
-93%
-86%
-92%
-93%
-86%
-92%
Alternative Option - Total
$ 0.34
$ 0.08
$ 0.42
$ 0.35
$ 0.08
$ 0.43
Alternative Option - Net Change
$ (3.41)
$ (0.36)
$ (3.77)
$ (3.26)
$ (0.34)
$ (3.60)
Alternative Option - Percent Change
-91%
-82%
-90%
-90%
-80%
-89%
Note: Detail may not add due to independent rounding.
Source: Proposed RTCRCost Model.
7.4.8 Uncertainty in Unit Costs
As stated in Section 7.2.5, EPA recognizes that there are both variability and uncertainty
in unit cost estimates used to develop national costs for the proposed RTCR. Variability is
expected in the actual costs that would be experienced by different PWSs of similar size
conducting the same corrective action. Otherwise similar PWSs may experience different capital
and/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 distribution of this variability would be on a national scale; therefore,
EPA uses mean values for these input parameters.
The unit costs used in this EA are developed as average or representative estimates of
what these unit costs would be nationally, and are not the unit costs of any particular PWS. Some
components of monitoring costs, such as the purchase and wear-and-tear of vehicles, are not
quantified because of either limited data or inability to attribute these costs directly to the
proposed RTCR. Additionally, the PWS and state labor hours for each rule component
(specifically for Level 1 and Level 2 assessments) are meant to capture national averages for the
purposes of developing national cost estimates and making comparisons between regulatory
alternatives. Thus, the unit costs presented in this document may over- or under-estimate the unit
costs of any particular PWS. Detailed information on the derivation of unit costs for each rule
component is provided in the Technology and Cost Document for the Proposed Revised Total
ColiformRule (USEPA, 2010d).
7.5 Household Costs
The household cost analysis considers the impact that the costs incurred by CWSs have
on the households they serve. This analysis considers the potential increase in a household's
annual water bill if a CWS passed the entire cost increase resulting from the proposed rule on to
their customers. This analysis is a tool to gauge potential impacts and should not be construed as
a precise estimate of potential changes to household water bills. State costs and costs to
TNCWSs and NTNCWSs are not included in this analysis since their costs are not typically
passed through directly to households.
Economic Analysis for the Proposed RTCR 7-36
June 2010
-------
To calculate household costs (which are in the units of $ per household per year), the
CWS population subject to the proposed RTCR by PWS size is divided by the average number
of people per household, estimated as 2.56 for the year 2007 (United States Census Bureau,
2008), to calculate the number of households subject to the proposed RTCR by PWS size. The
cost of the rule, by size category, is then divided by that number of households to determine a
per-household cost.
The first section of Exhibit 7.23 presents net costs per household under the AIP and
Alternative option for all rule components spread across all CWSs. In this scenario, comparison
to the current TCR shows a cost savings for households in the largest size category. For those
households that are expected to see a cost increase, the average annual water bill would be
expected to increase by less than ten cents on average. Although this average cost per household
is very low, customers served by PWSs that incur greater costs to comply with the proposed
RTCR would be higher.
While the average increase in annual household water bills to implement the AIP option
is less than a dollar, customers served by a small CWS that have to take corrective actions as a
result of the proposed rule would incur slightly larger increases in their water bills. The
subsequent sections of the exhibit present net costs per household for different subsections of
CWSs (e.g., CWSs that perform assessments but no corrective actions, CWSs that do perform
corrective actions, and CWSs that do not perform assessments or corrective actions). As shown
in the second section of Exhibit 7.23, approximately 67% of households belong to CWSs that
would perform assessments but would not perform corrective actions (because no sanitary
defects are found). These households would experience a slight cost savings on an annual basis.
The 9% of households belonging to CWSs that would perform corrective actions would
experience an increase in annual net household costs of less than $1 to approximately $23 on an
annual basis. The final section of the exhibit presents the 24% of households belonging to CWSs
that would not perform assessments or corrective actions. Households of this category would
experience an increase in cost savings comparable to those performing corrective actions, and a
decrease in cost savings compared to those performing assessments but no corrective actions.
This decrease in costs savings is because no PN costs are associated with systems not performing
assessments. Overall, the main driver of additional household costs under the proposed RTCR is
whether or not additional corrective actions are performed.
Economic Analysis for the Proposed RTCR 7-3 7
June 2010
-------
Exhibit 7.23 Summary of Net Annual Per-Household Costs for the Proposed
RTCR (2007$)
PWS Size
(Population
Served)
Number of
Households
(AIP)
Num ber of
Households
(Alternative
Option)
3% Discount Rate
7% Discount Rate
AIP Net
AIP Net Cost per
Houshold
Alternative
Option Net
Alternative
Option Net
Cost per
Household
AIP Net
AIP Net
Cost per
Houshold
Alternative
Option Net
Alternative
Option Net
Cost per
Household
A
B
C
D=C/A
E
F=BB
G
H=G/A
I
J=l/B
All Com m unity Water Systems (CWSs)
<100
307,243
307,243
$ 68,433
$ 0.223
$ 180,905
$ 0.589
$ 170,417
$ 0.555
$ 310,914
$ 1.012
101-500
1,589,510
1,589,510
$ 323,373
$ 0.203
$ 417,352
$ 0.263
$ 430,801
$ 0.271
$ 557,863
$ 0.351
501-1,000
1,624,853
1,624,853
$ 35,455
$ 0.022
$ 75,901
$ 0.047
$ 86,936
$ 0.054
$ 138,624
$ 0.085
1,001-4,100
7,816,592
7,816,592
$ 335,846
$ 0.043
$ 335,846
$ 0.043
$ 390,122
$ 0.050
$ 390,122
$ 0.050
4,101-33,000
27,997,647
27,997,647
$ 2,295,169
$ 0.082
$ 2,295,169
$ 0.082
$ 2,007,071
$ 0.072
$ 2,007,071
$ 0.072
33,001-96,000
21,933,438
21,933,438
$ 1,547,791
$ 0.071
$ 1,547,791
$ 0.071
$ 1,284,124
$ 0.059
$ 1,284,124
$ 0.059
96,001-500,000
26,770,609
26,770,609
$ 1,463,800
$ 0.055
$ 1,463,800
$ 0.055
$ 1,199,384
$ 0.045
$ 1,199,384
$ 0.045
500,001-1 Million
9,764,979
9,764,979
$ 153,975
$ 0.016
$ 153,975
$ 0.016
$ 122,635
$ 0.013
$ 122,635
$ 0.013
> 1 Million
16,309,853
16,309,853
$ (4,876)
$ (0.000)
$ (4,876)
$ (0.000)
$ (4,023)
$ (0.000)
$ (4,023)
$ (0.000)
Total
114,114,724
114,114,724
$ 6,218,966
$ 0.054
$ 6,465,863
$ 0.057
$ 5,687,467
$ 0.050
$ 6,006,714
$ 0.053
Com m unity Water Systems (CWSs) performing Level 1/Level 2 Assessments (and no Corrective Actions)
<100
125,340
124,920
$ (187,747)
$ (1.498)
$ (151,107)
$ (1.210)
$ (126,913)
$ (1.013)
$ (77,760)
$ (0.622)
101-500
460,577
464,568
$ (138,181)
$ (0.300)
$ (104,157)
$ (0.224)
$ (83,851)
$ (0.182)
$ (41,251)
$ (0.089)
501-1,000
394,643
401,009
$ (132,431)
$ (0.336)
$ (122,868)
$ (0.306)
$ (110,140)
$ (0.279)
$ (97,592)
$ (0.243)
1,001-4,100
2,341,578
2,341,578
$ (261,760)
$ (0.112)
$ (261,760)
$ (0.112)
$ (212,565)
$ (0.091)
$ (212,565)
$ (0.091)
4,101-33,000
24,827,588
24,827,588
$ (207,674)
$ (0.008)
$ (207,674)
$ (0.008)
$ (88,472)
$ (0.004)
$ (88,472)
$ (0.004)
33,001-96,000
19,232,570
19,232,570
$ (187,918)
$ (0.010)
$ (187,918)
$ (0.010)
$ (156,652)
$ (0.008)
$ (156,652)
$ (0.008)
96,001-500,000
23,912,325
23,912,325
$ (162,082)
$ (0.007)
$ (162,082)
$ (0.007)
$ (146,737)
$ (0.006)
$ (146,737)
$ (0.006)
500,001-1 Million
5,524,188
5,524,188
$ (45,790)
$ (0.008)
$ (45,790)
$ (0.008)
$ (42,392)
$ (0.008)
$ (42,392)
$ (0.008)
> 1 Million
-
-
$
$
$
$
$
$
$
$
Total
76,818,809
76,828,746
$ (1,323,584)
$ (0.017)
$ (1,243,356)
$ (0.016)
$ (967,721)
$ (0.013)
$ (863,420)
$ (0.011)
Community Water Systems (CWSs) performing Corrective Actions
<100
13,927
13,880
$ 325,679
$ 23.385
$ 347,280
$ 25.020
$ 296,928
$ 21.321
$ 318,549
$ 22.950
101-500
51,175
51,619
$ 469,264
$ 9.170
$ 462,084
$ 8.952
$ 412,574
$ 8.062
$ 410,784
$ 7.958
501-1,000
43,849
44,557
$ 164,754
$ 3.757
$ 167,578
$ 3.761
$ 143,922
$ 3.282
$ 146,752
$ 3.294
1,001-4,100
260,175
260,175
$ 556,364
$ 2.138
$ 556,364
$ 2.138
$ 484,508
$ 1.862
$ 484,508
$ 1.862
4,101-33,000
3,170,059
3,170,059
$ 2,526,599
$ 0.797
$ 2,526,599
$ 0.797
$ 2,132,792
$ 0.673
$ 2,132,792
$ 0.673
33,001-96,000
2,700,868
2,700,868
$ 1,736,797
$ 0.643
$ 1,736,797
$ 0.643
$ 1,446,819
$ 0.536
$ 1,446,819
$ 0.536
96,001-500,000
2,858,284
2,858,284
$ 1,624,577
$ 0.568
$ 1,624,577
$ 0.568
$ 1,345,503
$ 0.471
$ 1,345,503
$ 0.471
500,001-1 Million
613,799
613,799
$ 194,889
$ 0.318
$ 194,889
$ 0.318
$ 161,003
$ 0.262
$ 161,003
$ 0.262
> 1 Million
-
-
$
$
$
$
$
$
$
$
Total
9,712,136
9,713,240
$ 7,598,924
$ 0.782
$ 7,616,167
$ 0.784
$ 6,424,050
$ 0.661
$ 6,446,711
$ 0.664
Com m unity Water Systems (CWSs) not performing Level 1/Level 2 Assessments, or Corrective Actions
<100
167,976
168,442
(69,499)
$ (0.414)
(15,268)
$ (0.091)
402
$ 0.002
70,124
$ 0.416
101-500
1,077,758
1,073,324
(7,710)
$ (0.007)
59,425
$ 0.055
102,079
$ 0.095
188,331
$ 0.175
501-1,000
1,186,361
1,179,288
3,132
$ 0.003
31,191
$ 0.026
53,154
$ 0.045
89,464
$ 0.076
1,001-4,100
5,214,839
5,214,839
41,242
$ 0.008
41,242
$ 0.008
118,178
$ 0.023
118,178
$ 0.023
4,101-33,000
-
-
-
$
-
$
-
$
-
$
33,001-96,000
-
-
-
$
-
$
-
$
-
$
96,001-500,000
-
-
-
$
-
$
-
$
-
$
500,001-1 Million
3,626,992
3,626,992
-
$
-
$
-
$
-
$
> 1 Million
16,309,853
16,309,853
-
$
-
$
-
$
-
$
Total
27,583,779
27,572,738
$ (32,836)
$ (0.001)
$ 116,591
$ 0.004
$ 273,813
$ 0.010
$ 466,097
$ 0.017
Source:
(C), (E), (G), (I) Exhibit 7.28.
7.6 Nonqualified Costs
EPA believes that all of the rule elements that are the major drivers of the net changes in
costs of the current TCR have been quantified to the greatest degree possible. However, cost
reductions related to fewer monitoring and reporting violations are not specifically accounted for
in the cost analysis, and their exclusion from consideration may result in an overestimate of net
change in cost between the TCR and the AIP or Alternative option.
Economic Analysis for the Proposed RTCR 7-38
June 2010
-------
In addition, under the TCR, AIP, and Alternative option, Tier 3 public notification for
monitoring and reporting violations are assumed to be reported once per year as part of the CCR.
Because of the use of the CCR to communicate Tier 3 public notification on a yearly basis, no
cost differential between the current TCR and the AIP and Alternative option is estimated in the
cost model. However, the advisory committee concluded that significant reductions in
monitoring and reporting violations may be realized through the revised regulatory framework of
the proposed RTCR. These possible reductions have not been quantified. System resources used
to process monitoring violation notices for the CCR and respond to customer inquiries about the
notices as well as state resources to remind systems to take samples, may be reduced if
significant reductions are realized. Exclusion of this potential cost savings may lead to an
underestimate of the PN cost savings under both the AIP and Alternative option. Such costs
savings to states may be significant given the high occurrence of monitoring and reporting
violations under the current TCR.
Additionally, as an underlying assumption to the costing methodology, EPA has assumed
that all PWSs subject to the proposed RTCR requirements are already complying with the
current TCR. There may be some PWSs that are not in full compliance with the current TCR,
and if so, additional costs would be incurred.
7.7 Uncertainty Analysis
There are two primary sources of uncertainty in the proposed RTCR cost modeling. The
first is related to the underlying estimates of events resulting from the rule revisions as generated
from the occurrence and predictive modeling. The occurrence and predictive modeling
(discussed in detail in Chapter 5) does not explicitly consider uncertainty, and therefore does not
generate "confidence intervals" on the predicted outcomes. However, EPA evaluated the model
inputs to determine which of the inputs were likely to have a significant effect on the results and
subjected those to further review through sensitivity analyses. In particular, EPA evaluated the
impacts of alternative estimates of the net change percentage of PWSs predicted to take
corrective actions in response to an assessment (10 percent in the model) and the associated
effectiveness of those corrective actions (Section 5.3.3.1 discusses these analyses in detail). The
results of these analyses suggest that changes in the major assumptions about the net change
percentage of corrective actions identified and the effectiveness of those actions have a less than
linear effect on predicted outcomes. When applied to costs estimates they would have even less
impact.
In the case where PWSs increase the percentage of corrective actions taken in response to
an assessment, costs are expected to increase corresponding to the greater number of corrective
actions taken. Using the increases predicted for the small subset of PWSs analyzed in the
analysis, it is estimated that corrective actions (and related costs) would increase by 65% for
Level 1 corrective actions and 7% for Level 2 corrective actions. This is in response to a
doubling of the net change percentage of corrective actions taken in response to an assessment.
However, the increased effectiveness of more corrective actions has a dampening effect on other
costs that would offset the increased corrective action costs. For instance, although the numbers
of corrective actions increase due to the higher percentages found during assessments, the actual
number of assessments decreases (-16% Level 1 and -43% Level 2) because fewer positive
samples (ranging from -13% routine TC+ to -28% repeat TC+) are predicted to be found. The
Economic Analysis for the Proposed RTCR 7-39
June 2010
-------
overall effect of the cost increases and decreases in response to the major predictive occurrence
model uncertainties is expected to be minimal.
The other major area of uncertainty that may affect the resulting cost calculations is the
distribution of corrective actions taken by PWSs (the compliance forecast) in response to finding
a problem during a Level 1 or Level 2 assessment. The compliance forecast presented in Exhibit
7.16 represents EPA's best estimate of a distribution of corrective actions that may be taken by
PWSs to respond to problems identified under the proposed RTCR. Because there is a wide
variation in the unit costs of the actual corrective actions underlying the compliance forecast (see
Appendix D for unit cost detail), EPA ran two sensitivity analyses to assess the potential impacts
of different distributions within the compliance forecast. These two sensitivity runs attempt to
bound the analysis of corrective action costs by generating a low and high cost bound to the
estimates.
Low Bound Estimate
During discussions in the TCRDSAC meetings, several stakeholders suggested that
almost all additional corrective actions taken in response to the proposed RTCR would be in
response to transient contamination events or poor sampling techniques. These are also the least
expensive corrective actions. To examine the effects of greater emphasis on these types of
corrective actions, EPA reran the cost model to reflect 90% selection of either spot flushing or
sampler training as corrective actions. The remaining 10% of costs were distributed across the
least costly corrective actions under each of the other compliance forecast categories. This results
in an approximate 80 percent decrease in total net costs for the AIP and an approximate 44
percent decrease in total net costs for the Alternative option. Exhibit 7.24 shows the change in
the overall costs with this change.
High Bound Estimate
PWSs may also take actions that result in higher corrective action costs than those
predicted by the current compliance forecast, although this would be a less likely scenario based
on stakeholder discussions. Purely economic considerations also suggest that, given the option,
PWSs would opt for the least costly option to address any issues identified. However, to test a
potential high end of corrective action costs, EPA ran the cost model with the compliance
forecast set to take the highest cost corrective action in each compliance forecast category. In this
scenario, only 5% of corrective actions were predicted for flushing and sampler training, and
10%) were estimated for each of the other compliance categories, and unit costs were assigned for
the highest unit cost corrective action in each. This results in an increase in total net costs for the
AIP by approximately a factor of 1.8, and an increase in total net costs by approximately a factor
of 1.5 for the Alternative option. Exhibit 7.24 shows the annualized total and net change cost
estimates for PWSs and states to comply with proposed RTCR under the TCR, AIP, and
Alternative options based on low bound and high bound estimates in the compliance forecast.
Economic Analysis for the Proposed RTCR 7-40
June 2010
-------
Exhibit 7.24 Sensitivity Analysis - Annualized Net Change in Costs based on
Changes in Compliance Forecast ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
AIP Net Change
$ 13.54
$ 0.15
$ 13.69
$ 13.24
$ 0.42
$ 13.66
AIP Low Bound Net Change
$ 2.54
$ 0.15
$ 2.68
$ 3.84
$ 0.42
$ 4.26
AIP High Bound Net Change
$ 24.82
$ 0.15
$ 24.96
$ 23.38
$ 0.42
$ 23.80
Alternative Option Net Change
$ 26.95
$ 0.29
$ 27.24
$ 29.10
$ 0.59
$ 29.69
Alternative Option Low Bound Net Change
$ 15.03
$ 0.29
$ 15.31
$ 18.64
$ 0.59
$ 19.23
Alternative Option High Bound Net Change
$ 40.11
$ 0.29
$ 40.40
$ 41.46
$ 0.59
$ 42.05
Note: Detail may not add due to independent rounding.
Source: Proposed RTCRCost Model.
Overall, EPA recognizes that there is uncertainty in various parts of its estimates that
could result in either an over- or underestimate of the costs as presented in this chapter. Exhibit
7.25 presents a summary of these issues, references the location in the EA where the information
is introduced, and estimates the effects that each may have on national costs. All of the
assumptions influencing the EA baseline (current TCR) do so in the same way for the AIP and
Alternative option. Therefore, EPA does not expect the net results of the analyses presented in
this EA to be significantly influenced by the uncertainty in the assumptions applied in
developing the RTCR cost analysis. EPA has been careful to use the best available data, to
account for uncertainty quantitatively when possible, and to avoid any consistent biases in
assumptions and the use of data.
Economic Analysis for the Proposed RTCR 7-41
June 2010
-------
Exhibit 7.25 Cost Uncertainty Summary
Uncertainty
Section with Full
Discussion of
Uncertainty
Most Likely Effect
of National Costs
Underestimate
of Current Assumpt
Overestimate
ons on Estimate
Unknown Impact
Labor Rate
7.2.1
X
Labor Burden
Estimates
7.4.1 -7.4.8
X
Unit Costs
7.4.1 -7.4.8
X [note 1]
Percentage of
systems on monthly,
quarterly, and annual
monitoring
frequencies
7.4.3
X
Number of acute
violations
7.4.5
X
Number of systems
performing level 1
and level 2
assessments
7.4.5
X
Compliance
forecasts for
corrective actions
based on level 1 and
level 2 assessments
7.4.6 and 7.7
X [note 2]
Corrective Actions
based on level 1 and
level 2 Assessments
7.4.6
X
Notes:
1) All unit costs (with the exception of those for corrective actions, which are not applied to the current TCR)
were used in the proposed RTCR cost code in the same way for the current TCR, AIP, and Alternative
option. Therefore, EPA expects that any under- or over-estimation would affect the baseline and other
options similarly, resulting in no significant net effect on the results of the analysis.
2) The compliance forecast represents EPA's best estimate of a distribution of corrective actions that may be
taken by PWSs to respond to problems identified under the proposed RTCR. EPA ran two sensitivity
analyses to assess the potential impacts of different distributions within the compliance forecast (see
Section 7.7). These two sensitivity runs showed that under the low bound and high bound estimates, the net
change costs for the AIP were smaller than those for the Alternative Option.
7.8 Comparison of Total and Net Annualized Costs for All Regulatory
Alternatives
Based on information presented previously in this chapter, EPA developed national cost
estimates for the TCR, AIP, and Alternative option. Exhibit 7.26 presents the total8 and net
change in annualized costs to PWSs and states at 3 and 7 percent discount rates. Exhibit 7.27
presents the total and net change in annualized costs for the TCR, AIP, and Alternative option by
rule component at 3 and 7 percent discount rates. Exhibit 7.28 presents the total and net change
in annualized costs for the TCR, AIP, and Alternative Option by PWS size and type at 3 and 7
8 Because only the incremental costs of some rule components are considered as part of the cost analysis, references
to "total" costs in this section do not refer to the complete costs for regulatory implementation, but only to the
specific costs considered to calculate net changes in costs.
Economic Analysis for the Proposed RTCR 7-42
June 2010
-------
percent discount rates. Exhibit 7.29 presents the costs shown in Exhibit 7.28 on a per-PWS basis.
Further discussion of the results follows these exhibits.
Exhibit 7.26 Comparison of Total and Net Change from Current TCR in
Annualized Present Value Costs ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR - Total
$ 185
$ 0.9
$ 186
$ 178
$ 0.9
$ 179
AIP-Total
$ 199
$ 1.1
$ 200
$ 191
$ 1.3
$ 192
AIP - Net Change
$ 14
$ 0.1
$ 14
$ 13
$ 0.4
$ 14
AIP - Percent Change
7%
16%
7%
7%
48%
8%
Alternative Option - Total
$ 212
$ 1.2
$ 213
$ 207
$ 1.5
$ 209
Alternative Option - Net Change
$ 27
$ 0.3
$ 27
$ 29
$ 0.6
$ 30
Alternative Option - Percent Change
15%
32%
15%
16%
67%
17%
Note: Detail may not add due to independent rounding. Because only the incremental costs of some rule components
are considered as part of the cost analysis, references to "total" costs in this exhibit do not refer to the complete costs
for regulatory implementation, but only to the specific costs considered to calculate net changes in costs.
Source: Proposed RTCR Cost Model.
Economic Analysis for the Proposed RTCR 7-43
June 2010
-------
Exhibit 7.27 Comparison of Total and Net Change in Annualized Present Value
Costs by Rule Component ($Millions, 2007$)
PWSs | State | Total
PWSs | State | Total
3%Discount Rate
7%Discount Rate
Rule Implementation
TCR-Total
$
$
$
$
$
$
AIP- Total
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
AIP - Net Change
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Alternative Option - Total
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Alternative Option - Net Change
$ 2.77
$ 0.18
$ 2.95
$ 4.00
$ 0.26
$ 4.26
Revising Sampling Plans
TCR-Total
$
$
$
$
$
$
AIP-Total
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
AIP-Net Change
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Alternative Option - Total
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Alternative Option - Net Change
$ 0.59
$ 0.42
$ 1.01
$ 0.84
$ 0.59
$ 1.42
Routine Monitoring
TCR-Total
$ 170.59
$
$ 170.59
$ 163.94
$
$ 163.94
AIP-Total
$ 174.71
$
$ 174.71
$ 167.74
$
$ 167.74
AIP-Net Change
$ 4.12
$
$ 4.12
$ 3.80
$
$ 3.80
Alternative Option - Total
$ 186.34
$
$ 186.34
$ 181.49
$
$ 181.49
Alternative Option - Net Change
$ 15.75
$
$ 15.75
$ 17.56
$
$ 17.56
Additional Routine Monitoring
TCR-Total
$ 3.87
$
$ 3.87
$ 3.72
$
$ 3.72
AIP-Total
$ 1.12
$
$ 1.12
$ 1.09
$
$ 1.09
AIP-Net Change
$ (2.75)
$
$ (2.75)
$ (2.63)
$
$ (2.63)
Alternative Option - Total
$ 0.68
$
$ 0.68
$ 0.58
$
$ 0.58
Alternative Option - Net Change
$ (3.19)
$
$ (3.19)
$ (3.14)
$
$ (3.14)
Repeat Monitoring
TCR-Total
$ 5.11
$
$ 5.11
$ 4.92
$
$ 4.92
AIP-Total
$ 4.82
$
$ 4.82
$ 4.64
$
$ 4.64
AIP-Net Change
$ (0.29)
$
$ (0.29)
$ (0.28)
$
$ (0.28)
Alternative Option - Total
$ 5.50
$
$ 5.50
$ 5.45
$
$ 5.45
Alternative Option - Net Change
$ 0.39
$
$ 0.39
$ 0.54
$
$ 0.54
Annual Site Visits
TCR-Total
$
$
$
$
$
$
AIP-Total
$
$
$
$
$
$
AIP - Net Change
$
$
$
$
$
$
Alternative Option - Total
$
$
$
$
$
$
Alternative Option - Net Change
$
$
$
$
$
$
Level 1 Assessment
TCR-Total
$ 1.13
$ 0.21
$ 1.34
$ 1.08
$ 0.20
$ 1.29
AIP-Total
$ 1.63
$ 0.20
$ 1.84
$ 1.57
$ 0.20
$ 1.77
AIP-Net Change
$ 0.50
$ (0.01)
$ 0.49
$ 0.49
$ (0.01)
$ 0.48
Alternative Option - Total
$ 1.73
$ 0.23
$ 1.95
$ 1.69
$ 0.22
$ 1.91
Alternative Option - Net Change
$ 0.60
$ 0.01
$ 0.61
$ 0.60
$ 0.02
$ 0.62
Level 2 Assessment
TCR-Total
$ 0.70
$ 0.26
$ 0.96
$ 0.68
$ 0.25
$ 0.92
AIP-Total
$ 0.90
$ 0.19
$ 1.09
$ 0.88
$ 0.18
$ 1.06
AIP-Net Change
$ 0.20
$ (0.07)
$ 0.12
$ 0.20
$ (0.07)
$ 0.13
Alternative Option - Total
$ 1.23
$ 0.28
$ 1.51
$ 1.27
$ 0.30
$ 1.57
Alternative Option - Net Change
$ 0.52
$ 0.02
$ 0.55
$ 0.60
$ 0.05
$ 0.65
Corrective Actions based on Level 1 Assessments
TCR-Total
$
$
$
$
$
$
AIP-Total
$ 9.17
$ 0.01
$ 9.18
$ 7.77
$ 0.01
$ 7.77
AIP - Net Change
$ 9.17
$ 0.01
$ 9.18
$ 7.77
$ 0.01
$ 7.77
Alternative Option - Total
$ 9.39
$ 0.01
$ 9.40
$ 8.01
$ 0.01
$ 8.02
Alternative Option - Net Change
$ 9.39
$ 0.01
$ 9.40
$ 8.01
$ 0.01
$ 8.02
Corrective Actions based on Level 2 Assessments
TCR-Total
$
$
$
$
$
$
AIP-Total
$ 2.72
$ 0.00
$ 2.72
$ 2.41
$ 0.00
$ 2.41
AIP-Net Change
$ 2.72
$ 0.00
$ 2.72
$ 2.41
$ 0.00
$ 2.41
Alternative Option - Total
$ 3.53
$ 0.01
$ 3.53
$ 3.36
$ 0.01
$ 3.37
Alternative Option - Net Change
$ 3.53
$ 0.01
$ 3.53
$ 3.36
$ 0.01
$ 3.37
Public Notification
TCR-Total
$ 3.75
$ 0.44
$ 4.19
$ 3.60
$ 0.42
$ 4.02
AIP-Total
$ 0.26
$ 0.06
$ 0.33
$ 0.26
$ 0.06
$ 0.32
AIP - Net Change
$ (3.48)
$ (0.38)
$ (3.86)
$ (3.34)
$ (0.36)
$ (3.70)
Alternative Option - Total
$ 0.34
$ 0.08
$ 0.42
$ 0.35
$ 0.08
$ 0.43
Alternative Option - Net Change
$ (3.41)
$ (0.36)
$ (3.77)
$ (3.25)
$ (0.34)
$ (3.59)
Notes:
1) Detail may not add due to independent rounding. Because only the incremental costs of some rule components are considered as
part of the cost analysis, references to "total" costs in this exhibit do not refer to the complete costs for regulatory implementation,
but only to the specific costs considered to calculate net changes in costs.
2) For modeling purposes, additional routine sample counts include regular routine samples taken in the same month.
Source: Proposed RTCRcost model.
Economic Analysis for the Proposed RTCR 7-44
June 2010
-------
Exhibit 7.28 Total and Net Change in Annualized Costs to PWSs by PWS Size and
Type ($Millions, 2007$)
PWS Size
(Population
Served)
3%Discount Rate
7% Discount Rate
TCR-Total
AIP - Total
AIP - Net
Alternative
Option - Total
Alternative
Option - Net
TCR-Total
AIP - Total
AIP - Net
Alternative
Option -
Total
Alternative
Option - Net
A
B
C=B-A
D
E=D-A
F
G
H=G-F
I
J=l-F
Community Water Systems (CWSs)
<100
$7.4
$7.5
$0.1
$7.6
$0.2
$7.1
$7.3
$0.2
$7.4
$0.3
101-500
$9.0
$9.3
$0.3
$9.4
$0.4
$8.6
$9.1
$0.4
$9.2
$0.6
501-1,000
$3.7
$3.8
$0.0
$3.8
$0.1
$3.6
$3.7
$0.1
$3.7
$0.1
1,001-4,100
$13.2
$13.6
$0.3
$13.6
$0.3
$12.7
$13.1
$0.4
$13.1
$0.4
4,101-33,000
$42.4
$44.7
$2.3
$44.7
$2.3
$40.7
$42.7
$2.0
$42.7
$2.0
33,001-96,000
$34.9
$36.4
$1.5
$36.4
$1.5
$33.5
$34.8
$1.3
$34.8
$1.3
96,001-500,000
$34.7
$36.2
$1.5
$36.2
$1.5
$33.4
$34.6
$1.2
$34.6
$1.2
500,001-1 Million
$6.5
$6.7
$0.2
$6.7
$0.2
$6.2
$6.4
$0.1
$6.4
$0.1
> 1 Million
$5.6
$5.5
($0.0)
$5.5
($0.0)
$5.3
$5.3
($0.0)
$5.3
($0.0)
Total
$157.4
$163.6
$6.2
$163.9
$6.5
$151.3
$157.0
$5.7
$157.3
$6.0
Nontransient Noncommunity Water Systems (NTNCWSs)
<100
$2.6
$2.6
$0.1
$3.6
$1.0
$2.5
$2.6
$0.2
$3.8
$1.3
101-500
$1.9
$2.0
$0.1
$2.7
$0.9
$1.8
$2.0
$0.2
$2.9
$1.1
501-1,000
$0.6
$0.6
$0.0
$0.8
$0.3
$0.6
$0.6
$0.1
$0.9
$0.3
1,001-4,100
$1.2
$1.3
$0.1
$1.3
$0.1
$1.1
$1.2
$0.1
$1.2
$0.1
4,101-33,000
$0.4
$0.5
$0.1
$0.5
$0.1
$0.4
$0.5
$0.0
$0.5
$0.0
33,001-96,000
$0.1
$0.1
$0.0
$0.1
$0.0
$0.1
$0.1
$0.0
$0.1
$0.0
96,001-500,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
500,001-1 Million
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
> 1 Million
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
Total
$6.9
$7.2
$0.4
$9.1
$2.3
$6.6
$7.2
$0.6
$9.4
$2.8
Transient Noncommunity Water Systems (TNCWSs)
<100
$13.4
$18.5
$5.1
$26.7
$13.4
$12.8
$18.0
$5.1
$27.8
$14.9
101-500
$4.9
$6.4
$1.5
$9.1
$4.2
$4.7
$6.2
$1.5
$9.4
$4.7
501-1,000
$0.6
$0.8
$0.2
$1.1
$0.5
$0.6
$0.8
$0.2
$1.2
$0.5
1,001-4,100
$0.9
$1.0
$0.1
$1.0
$0.1
$0.9
$1.0
$0.1
$1.0
$0.1
4,101-33,000
$0.4
$0.5
$0.1
$0.5
$0.1
$0.4
$0.5
$0.0
$0.5
$0.0
33,001-96,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
96,001-500,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
500,001-1 Million
$0.2
$0.2
($0.0)
$0.2
($0.0)
$0.2
$0.2
($0.0)
$0.2
($0.0)
> 1 Million
$0.3
$0.3
$0.0
$0.3
$0.0
$0.3
$0.3
$0.0
$0.3
$0.0
Total
$20.9
$27.8
$6.9
$39.1
$18.2
$20.1
$27.1
$7.0
$40.4
$20.3
Grand Total
$185.2
$198.7
$13.5
$212.1
$26.9
$177.9
$191.2
$13.2
$207.0
$29.1
Note: Detail may not add due to independent rounding. Because only the incremental costs of some rule components are considered as part of the cost analysis, references to "total" costs
in this exhibit do not refer to the complete costs for regulatory implementation, but only to the specific costs considered to calculate net changes in costs.
Source: Proposed RTCR cost model.
Economic Analysis for the Proposed RTCR 7-45
June 2010
-------
Exhibit 7.29 Total and Net Change in Annualized Per PWS Costs by PWS Size
and Type (2007$)
3% Discount Rate
7% Discount Rate
Alternative
Alternative
PWS Size
Option -
Alternative
Option -
Alternative
(Population
TCR-Total
AIP-Total
AIP - Net
Total
Option - Net
TCR-Total
AIP-Total
AIP - Net
Total
Option - Net
Served)
A
B
C=B-A
D
E=D-A
F
G
H=G-F
1
J=l-F
Community Water Systems (CWSs)
<100
$564
$570
$5
$578
$14
$542
$555
$13
$566
$24
101-500
$561
$581
$20
$587
$26
$539
$566
$27
$574
$35
501-1,000
$662
$669
$6
$676
$13
$637
$652
$15
$661
$25
1,001-4,100
$1,411
$1,447
$36
$1,447
$36
$1,356
$1,397
$42
$1,397
$42
4,101-33,000
$6,707
$7,070
$363
$7,070
$363
$6,445
$6,763
$318
$6,763
$318
33,001-96,000
$33,053
$34,520
$1,467
$34,520
$1,467
$31,764
$32,981
$1,217
$32,981
$1,217
96,001-500,000
$93,619
$97,565
$3,946
$97,565
$3,946
$89,968
$93,201
$3,233
$93,201
$3,233
500,001-1 Million
$185,678
$190,077
$4,399
$190,077
$4,399
$178,435
$181,939
$3,504
$181,939
$3,504
> 1 Million
$277,732
$277,488
($244)
$277,488
($244)
$266,900
$266,698
($201)
$266,698
($201)
Total
$3,029
$3,148
$120
$3,153
$124
$2,911
$3,020
$109
$3,026
$116
Nontransient Noncommunity Water Systems (NTNCWSs)
<100
$284
$292
$8
$394
$111
$273
$291
$18
$413
$141
101-500
$273
$289
$16
$397
$124
$263
$289
$26
$416
$153
501-1,000
$321
$347
$26
$465
$144
$308
$346
$38
$486
$178
1,001-4,100
$1,325
$1,436
$111
$1,436
$111
$1,273
$1,380
$106
$1,380
$106
4,101-33,000
$4,819
$5,365
$546
$5,365
$546
$4,631
$5,110
$478
$5,110
$478
33,001-96,000
$32,004
$33,310
$1,305
$33,310
$1,305
$30,757
$31,853
$1,096
$31,853
$1,096
96,001-500,000
$91,171
$90,807
($364)
$90,807
($364)
$87,617
$87,298
($319)
$87,298
($319)
500,001-1 Million
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
> 1 Million
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Total
$366
$387
$20
$488
$121
$352
$382
$30
$501
$149
Transient Noncommunity Water Systems (TNCWSs)
<100
$217
$300
$83
$434
$217
$209
$292
$83
$451
$243
101-500
$247
$323
$76
$459
$212
$237
$314
$77
$474
$237
501-1,000
$311
$405
$94
$548
$237
$299
$393
$94
$564
$265
1,001-4,100
$1,363
$1,524
$161
$1,524
$161
$1,310
$1,457
$147
$1,457
$147
4,101-33,000
$5,068
$5,658
$589
$5,658
$589
$4,871
$5,390
$519
$5,390
$519
33,001-96,000
$38,507
$38,155
($352)
$38,155
($352)
$37,008
$36,701
($307)
$36,701
($307)
96,001-500,000
$77,791
$77,043
($748)
$77,043
($748)
$74,762
$74,090
($672)
$74,090
($672)
500,001-1 Million
$186,699
$184,841
($1,858)
$184,841
($1,858)
$179,429
$177,725
($1,704)
$177,725
($1,704)
> 1 Million
$287,010
$301,525
$14,515
$301,525
$14,515
$275,817
$289,883
$14,066
$289,883
$14,066
Total
$248
$331
$83
$465
$216
$238
$322
$83
$480
$241
Grand Total
$1,196
$1,283
$87
$1,370
$174
$1,149
$1,235
$39
$1,337
$141
Note: Detail may not add due to independent rounding. Because only the incremental costs of some rule components are considered as part of the cost analysis,
references to "total" costs in this exhibit do not refer to the complete costs for regulatory implementation, but only to the specific costs considered to calculate net
changes in costs.
Source: Proposed RTCR cost model.
Evaluation of Regulatory Options
Exhibit 7.26 summarizes the comparison of total and annualized present value of the net
change difference in cost from the baseline for each of the three regulatory scenarios (existing
TCR, AIP Option, and Alternative option). A continuation of the existing TCR would result in
no net change in costs. The net change in mean annualized present value national costs of the
AIP Option is estimated to be approximately $14M. The net change in mean annualized present
value national costs for the Alternative option are estimated to be approximately $27M using a
3% discount rate and $30M using a 7% discount rate.
The total net change in national annualized present value costs for PWSs serving >4,100
people (approximately $5M) are the same under the AIP and Alternative option. This is expected
because the provisions for PWSs serving >4,100 are the same under both the AIP and Alternative
option. Monitoring requirements for PWSs serving >4,100 people would remain essentially
Economic Analysis for the Proposed RTCR 7-46
June 2010
-------
unchanged under either the AIP or Alternative option. The observed overall net increase in costs
for PWSs serving >4,100 people is driven primarily by the requirements to conduct assessments
and to correct any sanitary defects that are found.
Under the AIP, PWSs are estimated to incur approximately 97% to 99% of the net
annualized present value costs. States are expected to incur the remaining costs. EPA recognizes
that state labor may be more expensive than PWS labor, and that for some states, state costs may
be passed onto PWSs. However, state costs are only a very small percentage of the total net
change costs for the RTCR.
Exhibit 7.27 presents the comparison of total and net change in annualized present value
costs by rule component. The exhibit shows that, for the AIP and Alternative option, corrective
action costs are the most significant contributor to the net increase in costs for PWSs. Under the
AIP, state costs to review revised sampling plans contribute most to the cost increase. For both
PWSs and states, a net decrease in costs associated with PN requirements helps to offset the total
net cost increase.
The large difference in net cost increases between the AIP and the Alternative option is
primarily driven by the increased number of routine samples taken under the Alternative option
in comparison to the AIP. A larger number of samples are also estimated to result more in Level
1 and Level 2 triggers and to subsequently require corrective actions based on Level 1 and Level
2 assessments. Overall, the total net costs of the AIP are estimated at 50% of the net costs of the
Alternative option. This cost difference is an important consideration in the selection of the AIP
option over the Alternative option as the preferred proposed RTCR option.
Exhibit 7.28 presents the total and net change in annualized costs to PWSs by size and
type for the three regulatory options. No net change in costs would result from a continuation of
the current TCR. Among PWSs serving <4,100 people, the largest increase in net costs would be
incurred by the TNCWSs serving <100 people under either the AIP ($5.1M) or Alternative
option ($13.4M). As shown in Exhibit 7.29, on a per system basis, this translates to a net
annualized present value increase of approximately $83 per system under the AIP and $217 per
system under the Alternative option for the TNCWSs serving <100 people. Significant impacts
are also estimated for CWSs serving greater than 4,101 people, driven by the relatively high
number of CWSs in this category, combined with relatively high unit costs for implementing
corrective actions for PWSs of this size. However, PWSs of this size (and CWSs in general) are
expected to be better able to absorb cost increases by passing costs through to a larger base of
customers. Overall, the most important drivers of total net costs are the numbers of PWSs and
the underlying baseline occurrence estimates in a given size category. Taken together, these
influence the numbers of samples and ultimately the numbers of corrective actions taken, which
drive the cost estimates.
Exhibit 7.29 shows the costs from Exhibit 7.28 on a per-PWS basis. On this basis, the
annual impact of the proposed RTCR generally increases with PWS size and the magnitude of
the annual per-PWS net costs do not appear to be prohibitive, even for the most heavily impacted
PWS categories. However, the range of per-PWS costs is expected to be fairly wide and some
individual PWSs may be more heavily impacted.
Economic Analysis for the Proposed RTCR 7-4 7
June 2010
-------
8. Economic Impact Analysis
8.1. Introduction
As part of the rulemaking process, the United States Environmental Protection Agency
(EPA) is required to address the direct and indirect burdens that the Revised Total Coliform Rule
(RTCR) may place on certain types of governments, businesses, and populations. This chapter
presents analyses performed by the EPA in accordance with the following 12 federal mandates
and statutory reviews:
1. Executive Order 12866: Regulatory Planning and Review.
2. Paperwork Reduction Act (the Information Collection Request (ICR) document
for the proposed RTCR contains the complete analysis (USEPA, 2010b)).
3. The Regulatory Flexibility Act (RFA) of 1980, as amended by the Small Business
Regulatory Enforcement Fairness Act (SBREFA) of 1996.
4. Unfunded Mandates Reform Act (UMRA) of 1995.
5. Executive Order 13132: Federalism.
6. Executive Order 13175: Consultation and Coordination with Indian Tribal
Governments.
7. Executive Order 13045: Protection of Children from Environmental Health Risks
and Safety Risks.
8. Executive Order 13211: Action Concerning Regulations That Significantly Affect
Energy Supply, Distribution, or Use.
9. National Technology Transfer and Advancement Act (NTTAA).
10. Executive Order 12898: Federal Action to Address Environmental Justice in
Minority Populations and Low-Income Populations.
11. Consultations with the Science Advisory Board (SAB), National Drinking Water
Advisory Council (NDWAC), and the Secretary of Health and Human Services as
Required by Section 1412 (d) and (e) of the Safe Drinking Water Act (SDWA).
12. Impacts on Sensitive Subpopulations as Required by Section 1412(b)(3)(c)(i) of
the 1996 Amendments to the SDWA.
13. Plain Language as Required by Executive Order 12866.
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). Others are designed to assess the financial and health effects of the rule on sensitive,
low-income, and tribal populations as well as on small Public Water Systems (PWSs).
Economic Analysis for the Proposed RTCR 8-1
June 2010
-------
8.2. Executive Order 12866: Regulatory Planning and Review
Under Executive Order 12866 (58 Federal Register (FR) 51735, October 4, 1993), the
Agency must determine whether the regulatory action is "significant" and therefore subject to
Office of Management and Budget (OMB) review and the requirements of the Executive Order.
The Order defines "significant regulatory action" as one that is likely to result in a rule that may:
• Have an annual effect on the economy of $100 million or more or adversely affect
in a material way the economy, a sector of the economy, productivity,
competition, jobs, the environment, public health or safety, or state, local, or tribal
governments or communities;
• Create a serious inconsistency or otherwise interfere with an action taken or
planned by another agency;
• Materially alter the budgetary impact of entitlement, grants, user fees, or loan
programs or the rights and obligations of recipients thereof; or
• Raise novel legal or policy issues arising out of legal mandates, the President's
priorities, or the principles set forth in the Executive Order.
Under Executive Order 12866, this action is a "significant regulatory action."
Accordingly, EPA submitted this action to the OMB for review under Executive Order 12866
and any changes made in response to OMB recommendations have been documented in the
docket for this action.
In addition, EPA prepared an analysis of the potential costs and benefits associated with
this action. EPA estimates that the proposed RTCR will have an overall impact on PWSs of $14
M and that the impact on small entities (PWSs serving 10,000 people or fewer) will be $9.4 M -
$9.8 M annualized at 3 and 7 percent discount rates, respectively. These impacts are described in
Section 8.4 as well as sections VI and VII.C of the proposed RTCR preamble (USEPA 2010c),
respectively.
8.3. Paperwork Reduction Act
The information collection requirements for the proposed RTCR have been submitted for
approval to the OMB under the Paperwork Reduction Act (PRA), 44 U.S.C. 3501 et seq. The
ICR document prepared by EPA has been assigned EPA ICR number 1895.06.
The PRA requires EPA to estimate the burden on PWSs and state/primacy agencies of
complying with the rule. The information collected as a result of EPA's efforts toward proposing
the proposed RTCR should allow states/primacy agencies and EPA to determine appropriate
requirements for specific systems and evaluate compliance with the proposed RTCR. Burden is
defined at 5 CFR 1320.3(b) and means the total time, effort, and financial resources required to
generate, maintain, retain, disclose, or provide information to or for a federal agency. The burden
includes the time needed to conduct the following state and PWS activities:
Economic Analysis for the Proposed RTCR 8-2
June 2010
-------
State activities:
• Read and understand the rule;
• Mobilize (including primacy application), plan, and implement;
• Train PWS and consultant staff;
• Track compliance;
• Analyze and review PWS data;
• Review sampling plans and recommend any revisions to PWSs;
• Make determinations concerning PWS monitoring requirements;
• Respond to PWSs with positive samples;
• Recordkeeping;
• Review completed assessment forms and consult with the PWS about the
assessment report;
• Review and coordinate with PWSs to determine optimal corrective actions to be
implemented; and
• Provide consultation, review public notification certifications, and file reports of
violations.
PWS activities:
• Read and understand the rule;
• Planning and mobilization activities;
• Revise existing sampling plans to identify sampling locations and collection
schedules that are representative of water throughout the distribution system;
• Conduct routine, additional routine, and repeat monitoring;
• Complete a Level 1 Assessment if the PWS experiences a Level 1 trigger, and
submit a timetable to the state to identify sanitary defects detected, corrective
actions completed, and a timetable for any corrective actions not already
completed;
• Complete a Level 2 Assessment if the PWS experiences a Level 2 trigger, and
submit a timetable for any corrective actions not already completed;
Economic Analysis for the Proposed RTCR 8-3
June 2010
-------
• Correct sanitary defects found through the performance of Level 1 or Level 2
assessments;
• Develop and distribute Tier 1 public notices when E. coli MCL violations occur;
• Develop and distribute Tier 2 public notices when the PWSs failed to take
corrective action; and
• Develop and distribute Tier 3 public notices when the PWSs failed to comply
with the monitoring requirements or with mandatory reporting of required
information within the specified timeframe.
For the first three years after publication of the final rule in the Federal Register, the
major information requirements apply to 154,894 respondents. The total incremental burden
associated with the change in moving from the information requirements of the current TCR to
those in the proposed RTCR over the three years covered by the ICR is 2,518,878 hours, for an
average of 839,526 hours per year. The total incremental cost over the three year clearance
period is $71.3 million, for an average of $23.8 million per year (simple average over three
years). (The ICR estimate is higher than what is presented in the Economic Analysis (EA) for the
proposed rule because in the EA, the up-front costs that occur in the first three years, as well as
future costs, are annualized over a 25-year time horizon.) The average burden per response (i.e.,
the amount of time needed for each activity that requires a collection of information) is 5.4
hours; the average cost per response is $153.4. The collection requirements are mandatory under
SDWA (42 U.S.C. 300h etseq.). Detail on the calculation of the proposed rule information
collection burden and costs can be found in the ICR for the Proposed RTCR and Chapter 7 of
this EA. A summary of the burdens and costs of the proposed collection is presented in Exhibit
8.1 below.
Exhibit 8.1 Average Annual Net Change Burden and Costs for the Proposed RTCR
Information Collection Request
Respondent Type
Annual Burden
Hours
Cost
Annual
Responses
Annual Labor
Cost
Annual O&M
Cost
Annual Capital
Cost
Total Annual
Cost
PWSs
747,848
$ 20,171,639
$
$
$ 20,171,639
103,225
States and
Territories
91,678
$ 3,595,421
$
$
$ 3,595,421
51,669
TOTAL
839,526
$ 23,767,060
$
$
$ 23,767,060
154,894
Notes: Detail may not add exactly to total due to independent rounding.
"Annual Burden Hours" reflects an annual average for all system sizes over the 3-year ICR period.
Source: ICR for the proposed RTCR (USEPA 2010b).
An agency may not conduct or sponsor, and a person is not required to respond to, a
collection of information unless it displays a currently valid OMB control number. The OMB
control numbers for EPA's regulations in 40 CFR are listed in 40 CFR part 9.
Economic Analysis for the Proposed RTCR 8-4
June 2010
-------
To comment on EPA's need for this information, the accuracy of the provided burden
estimates, and any suggested methods for minimizing respondent burden, EPA has established a
public docket for this proposed rule, which includes this ICR, under Docket ID number EPA-
HQ-OW-2008-0878. Submit any comments related to the ICR to EPA and OMB. See
ADDRESSES section at the beginning of this notice for where to submit comments to EPA.
Send comments to OMB at the Office of Information and Regulatory Affairs, Office of
Management and Budget, 725 17th Street, NW, Washington, DC 20503, Attention: Desk Officer
for EPA. Since OMB is required to make a decision concerning the ICR between 30 and 60 days
after [INSERT DATE OF PUBLICATION IN THE FEDERAL REGISTER], a comment to
OMB is best assured of having its full effect if OMB receives it by [INSERT DATE 30 DAYS
AFTER PUBLICATION IN THE FEDERAL REGISTER] The final rule will respond to
any OMB or public comments on the information collection requirements contained in this
proposal.
8.4. The Regulatory Flexibility Act
The RFA generally requires an agency to prepare a regulatory flexibility analysis of any
rule subject to notice and comment rulemaking requirements under the Administrative Procedure
Act or any other statute unless the agency certifies that the rule will not have a significant
economic impact on a substantial number of small entities. Small entities include small
businesses, small organizations, and small governmental jurisdictions.
The RFA provides default definitions for each type of small entity. Small entities are
defined as: (1) a small business as defined by the Small Business Administration's (SBA)
regulations at 13 CFR 121.201; (2) a small governmental jurisdiction that is a government of a
city, county, town, school district, or special district with a population of less than 50,000; and
(3) a small organization that is any "not-for-profit enterprise which is independently owned and
operated and is not dominant in its field." However, the RFA also authorizes an agency to use
alternative definitions for each category of small entity, "which are appropriate to the activities
of the agency" after proposing the alternative definition(s) in the Federal Register and taking
comment. 5 USC 601(3) - (5). In addition, to establish an alternative small business definition,
agencies must consult with SBA's Chief Counsel for Advocacy.
For purposes of assessing the impacts of the proposed RTCR on small entities, EPA
considered small entities to be PWSs serving fewer than 10,000 people. This is the cut-off level
specified by Congress in the 1996 Amendments to the SDWA for small system flexibility
provisions. As required by the RFA, EPA proposed using this alternative definition in the
Federal Register (63 FR 7620, February 13, 1998), requested public comment, consulted with
the SB A, and finalized the alternative definition in the Consumer Confidence Reports regulation
(63 FR 44524, August 19, 1998). As stated in that Final Rule, the alternative definition would be
applied for all future drinking water regulations.
After considering the economic impacts of today's proposed rule on small entities, I
certify that this action will not have a significant economic impact on a substantial number of
small entities. The small entities directly regulated by this proposed rule are small PWSs serving
fewer than 10,000 people. These include small community water systems (CWSs), non-transient
noncommunity water systems (NTNCWSs), and transient noncommunity water systems
(TNCWSs), entities such as municipal water systems (publicly and privately owned), and
Economic Analysis for the Proposed RTCR 8-5
June 2010
-------
privately-owned PWSs and for profit businesses where provision of water may be ancillary, such
as mobile home parks, day care centers, churches, schools and homeowner associations. We
have determined that only 61 of 150,672 small systems (0.04 percent) will experience an impact
of more than one percent of revenues, and that none of the small systems will experience an
impact of three percent or greater of revenue. Exhibit 8.2 provides a summary of the numbers
and percentages of small systems for which costs exceed one percent of revenues, by size
category and system type using a three percent discount rate. Exhibit 8.3 provides a summary of
the average costs to average revenues on a per system basis.
Exhibit 8.2 Average Annualized Revenue by System Size and Percent of Systems
with Costs Exceeding One Percent and Three Percent of Revenue (Three Percent
Discount Rate)
System
Type
System
Size
Total
Systems
Average
Revenue/
System
l%of
Revenue
Systems
Exceeding
l%of
Revenue
Percent of
Systems
Exceeding
l%of
Revenue
3% of
Revenue
Systems
Exceeding
3% of
Revenue
Percent of
Systems
Exceeding
3% of
Revenue
A
B
C=B*0.01
D
E=D/A
F=B*0.03
G
H=G/A
CWS
<500
29,150
$ 199,000
$ 2,000
60
0.21%
$ 6,000
0
0.00%
501-4,100
15,021
$ 1,114,000
$ 11,000
0
0.00%
$ 33,000
0
0.00%
4,101-10,000
3,672
$ 4,432,000
$ 44,000
1
0.03%
$ 133,000
0
0.00%
NTNCWS
<500
15,942
$ 3,075,000
$ 31,000
0
0.00%
$ 92,000
0
0.00%
501-4,100
2,690
$ 11,793,000
$ 118,000
0
0.00%
$ 354,000
0
0.00%
4,101-10,000
80
$ 61,663,000
$ 617,000
0
0.00%
$ 1,850,000
0
0.00%
TNCWS
<500
81,311
$ 1,811,000
$ 18,000
0
0.00%
$ 54,000
0
0.00%
501-4,100
2,735
$ 3,192,000
$ 32,000
0
0.00%
$ 96,000
0
0.00%
4,101-10,000
71
$ 11,280,000
$ 113,000
0
0.00%
$ 338,000
0
0.00%
All
<10,000
150,672
N/A
N/A
61
0.04%
N/A
0
0.00%
Sources: (A) SDWIS 2007 (B) Average CWS revenues from 2006 CWSS survey. Average NTNCWS and TNCWS
revenues calculated based on representative revenues for specific business categorizations (Appendix I provides
additional detail.)
(D, G) Number of systems with costs exceeding one percent and three percent of revenue calculated based on
distributions of costs for each rule component by system size compared to the average revenues. As a function of
calculations in the cost model, fractional system counts may be generated. Analyses in the exhibit are based only on
whole system counts. (Appendix I provides additional detail.) Consistent with the rest of the EA, the population break
is at 4,100 people rather than 3,300. This population break was based on TCRDSAC deliberations and consideration
that concluded that breaking out the analysis at the 4,100 point would be most informative because of changes in the
rule provisions at the 4,100 people point. (Chapter 4 provides further detail on data selection and size break outs.)
Economic Analysis for the Proposed RTCR 8-6
June 2010
-------
Exhibit 8.3 Average Costs per System and as Percentage of Revenue (2007$)
System
Size
Number
of
Systems
Average
Annual
Net Cost/
System
Average
Revenue/
System1
Average Annual
Net Costs as a
Percentage of
Revenue
A
B
C
D=(B/C)*100
<500
126,403
$ 57
$ 1,599,000
0.004%
501-4,100
20,446
$ 43
$ 2,797,000
0.002%
4,101 -10,000
3,823
$ 262
$ 5,757,000
0.005%
All PWSs
<10,000
150,672
$ 60
$ 1,867,000
0.003%
1
Includes water revenues and non-water revenues (e.g., revenues related to the primary
entities that operate a water system to support their business or municipal general revenue
for publicly owned and operated systems).
Although this proposed rule will not have a significant economic impact on a substantial
number of small entities, EPA nonetheless has tried to reduce the impact of this rule on small
PWSs. Provisions in the proposed RTCR that result in reduced costs for many small entities
include:
• Reduced routine monitoring for qualifying PWS serving 1,000 or fewer people.
• Reduced number of repeat samples required.
• Reduced additional routine monitoring for PWS serving 4,100 or fewer people.
• Reduced public notification requirements for all systems, including small
systems.
For some PWSs, cost savings may be offset in whole or in part by increased costs of
more stringent assessment requirements, stricter rules for qualifying for reduced monitoring, or
performing specific corrective actions. Additionally, seasonal PWSs are subject to increased
monitoring.
A description of activities that small systems perform under the proposed RTCR for each
rule component is provided in Section 7.4 of this EA along with the associated change in cost.
Exhibit 8.4 below provides the distribution of total costs to small entities by rule component.
Underlying these estimates are EPA's assumptions regarding the types of corrective actions that
will be implemented and the number of small PWSs predicted to implement each type. These
assumptions are detailed in Appendix D of this EA.
Economic Analysis for the Proposed RTCR 8-7
June 2010
-------
Exhibit 8.4 Annualized Net Rule Costs Predicted for Small Entities (PWSs serving
<10,000) by Rule Component Using Three Percent and Seven Percent Discount
Rates (2007$)
Rule Component
Net Costs at 3%
Net Costs at 7%
Rule Implementation
$ 2,673,000
$ 3,849,000
Revising Sampling Plans
$ 535,000
$ 755,000
Routine Monitoring
$ 4,120,000
$ 3,799,000
Additional Routine Monitoring
$ (2,750,000)
$ (2,631,000)
Repeat Monitoring
$ (193,000)
$ (186,000)
Site Inspections
$
$
Level 1 Assessment
$ 414,000
$ 401,000
Level 2 Assessment
$ 172,000
$ 179,000
Correction Actions based on Level 1
Assessments
$ 4,712,000
$ 4,055,000
Correction Actions based on Level 2
Assessments
$ 2,542,000
$ 2,255,000
Public Notification
$ (2,789,000)
$ (2,675,000)
Total
$ 9,435,000
$ 9,801,000
Source: Derived from cost model outputs (Appendix C).
EPA also conducted outreach to small entities and convened a Small Business Advocacy
Review (SBAR) Panel to obtain advice and recommendations of representatives of the small
entities that potentially would be subject to the proposed rule's requirements. EPA consulted with
small entity representatives before and during the review by the Panel. These small entity
representatives included representatives from small water systems of various types and sizes,
representatives from associations that assist and /or advocate for small systems, and federal
agencies that operate small systems. Panel members included representatives from OMB, SBA,
and the EPA Office of Ground Water and Drinking Water. The consultation led to the
development of a report providing recommendations to EPA on how to revise the TCR to
address small system concerns, which EPA considered in drafting this proposed RTCR (SBAR
Panel 2008). EPA also made presentations to the advisory committee on the recommendations of
the Panel so the advisory committee could consider their recommendations in developing the
Agreement in Principle (AIP).
Consistent with the RFA/SBREFA requirements, the Panel evaluated the assembled
materials and small-entity comments on issues and prepared a final report to the EPA
Administrator. A copy of the Panel report is included in the docket for this proposed rule. The
proposed rule is consistent with the Panel recommendations to use total coliforms as a trigger for
investigation and/or corrective action, to balance monitoring requirements and costs with risk, to
further differentiate requirements based on differences in water systems, to coordinate
requirements with other related rules, and to consider reporting and recordkeeping costs in
estimating burden. Consistent with the Panel recommendation to evaluate which parameters are
most appropriate for routine monitoring and as potential triggers for investigative and corrective
actions, EPA is conducting a review of existing methods for total coliform and E. coli analysis
and is evaluating its Alternative Test Procedure protocol for approving new methods as
described in section III.9 of the proposed RTCR preamble. EPA is also one of the founding
Economic Analysis for the Proposed RTCR 8-8
June 2010
-------
members of a Research and Information Collection Partnership, described in section V of the
proposed RTCR preamble, which is considering research and information needs to evaluate the
magnitude of risks and potential risk mitigation options related to potential distribution system
contamination.
EPA further reviewed the potential impacts of the AIP option on small entities in terms
of the assumption incorporated into SDWA major rulemakings that households or residential
water users are most vulnerable to cost increases1 (SBAR Panel 2008). In an Assessment of the
Vulnerability of Noncommunity Water Systems to SDWA Cost Increases (USEPA, 2008d), EPA
considered the burden of SDWA rule costs in comparison to the average revenues of various
categories of noncommunity water systems. All of the NCWS categories reviewed were found
to be less vulnerable to SDWA-related increases than a typical household. The report notes that
in some categories of businesses, costs are more easily passed on to the customer base than in
others. However, in each NCWS category, expenditures on water were found to be a relatively
small percentage of total revenues. In nearly all cases, water expenditures (including
expenditures for sewer service and miscellaneous other utilities) totaled less than one percent of
total revenues, and were not more than 1.3 percent of total revenues for any category. Several
caveats were put forth in this report, including one that considered the potential for
underestimating the impact to golf courses, which were grouped in with other recreational
entities whose use of water was less significant to the core business than the golf courses.
Despite the significant caveats listed, the report strongly suggests that TNCWS and NTNCWS
should not be considered particularly vulnerable to operating cost increases resulting from
SDWA rulemakings.
The consistency of many of the aspects of the AIP option with the recommendations of
the SBAR Panel further support that the formulation of the proposed RTCR EA is conducive to
minimizing net impacts on small entities that will be affected by the rule. Overall, the economic
analysis summarized in Exhibit 8.2 shows that net cost increases from the AIP option over time
are low relative to revenue. Based on this result, EPA certifies that there will not be a significant
impact on a substantial number of small entities under the AIP option.
8.5. Unfunded Mandates Reform Act
The UMRA seeks to protect state, local, and tribal governments from the imposition of
unfunded federal mandates. In addition, the Act seeks to strengthen the partnership between the
federal government and state, local, and tribal governments and ensure that the federal
government covers the costs incurred during compliance with federal mandates.
Title II of the UMRA of 1995, Public Law 104-4, establishes requirements for federal
agencies to assess the effects of their regulatory actions on state, local, and tribal governments
and the private sector. Under section 202 of UMRA, EPA generally must prepare a written
statement, including a cost-benefit analysis, for proposed and final rules with "federal mandates"
that may result in expenditures to state, local, and tribal governments, in the aggregate, or to the
private sector, of $100 million or more in any one year.
1 Major SDWA rulemakings generally include a section in the Economic Analysis that presents an analysis of costs
of the rule per household.
Economic Analysis for the Proposed RTCR 8-9
June 2010
-------
Before promulgating an EPA rule for which a written statement is needed, section 205 of
UMRA generally requires EPA to identify and consider a reasonable number of regulatory
options and adopt the least costly, most cost-effective or least burdensome option that achieves
the objectives of the rule. The provisions of section 205 do not apply when they are inconsistent
with applicable law. Moreover, section 205 allows EPA to adopt an option other that the least
costly, most cost-effective or least burdensome option if the Administrator publishes with the
final rule and explanation why that option was not adopted.
Before EPA establishes any regulatory requirements that may significantly or uniquely
affect small governments, including tribal governments, it must have developed under section
203 of UMRA a small government agency plan. The plan must provide for notifying potentially
affected small governments, enabling officials of affected small governments to have meaningful
and timely input in the development of EPA regulatory proposals with significant federal
intergovernmental mandates, and informing, educating, and advising small governments on
compliance with the regulatory requirements.
This proposed rule does not contain a federal mandate that may result in expenditures to
state, local, and tribal governments, in the aggregate, or to the private sector, of $100 million or
more in any one year. Expenditures associated with compliance, defined as the incremental costs
beyond the current TCR, will not surpass $100 million in the aggregate in any year. Thus, this
proposed rule is not subject to the requirements of sections 202 and 205 of UMRA.
This proposed rule is also not subject to the requirements of section 203 of UMRA
because it contains no regulatory requirements that might significantly or uniquely affect small
governments. Costs to small entities are generally not significant, as described previously in
Section 8.3 and in section VII.C of the proposed RTCR preamble. The regulatory requirements
of the proposed RTCR are not unique to small governments, as they apply to all PWSs regardless
of size.
8.6. Executive Order 13132: Federalism
Executive Order 13132, entitled "Federalism" (64 FR 43255, August 10, 1999), requires
EPA to develop an accountable process to ensure "meaningful and timely input by state and local
officials in the development of regulatory policies that have federalism implications." "Policies
that have federalism implications" is 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."
This action does not have federalism implications. It will not 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 as
specified in Executive Order 13132. The net change in cost for state, local, and tribal
governments in the aggregate is estimated to be approximately $0.1M and $0.4M at three percent
and seven percent discount rates, respectively. Thus, Executive Order 13132 does not apply to
this proposed rule.
Economic Analysis for the Proposed RTCR 8-10
June 2010
-------
Although section 6 of Executive Order 13132 does not apply to the proposed RTCR,
EPA conducted a federalism consultation, consistent with Executive Order 13132, in July 2008.
The consultation included a stakeholder meeting where EPA requested comments on the impacts
of the potential revisions to the TCR with respect to state, county and local governments. EPA
did not receive any comments in response to this consultation. In addition, the advisory
committee included representatives of state, local and tribal governments, and through this
process EPA consulted with state, local, and tribal government representatives to ensure that
their views were considered when the AIP recommendations for the RTCR were developed.
In the spirit of Executive Order 13132, and consistent with EPA policy to promote
communications between EPA and state and local governments, EPA specifically solicits
comment on this proposed action from state and local officials.
8.7. Executive Order 13175: Consultation and Coordination with 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.
This action does not have tribal implications, as specified in Executive Order 13175.
Because the requirements of the proposed RTCR as defined by the AIP option are estimated to
result in low net cost increases (or in many cases, net cost savings) compared to the existing TCR
requirements, the AIP option is not anticipated to have a negative impact on tribal PWSs. Thus,
Executive Order 13175 does not apply to this action.
Although Executive Order 13175 does not apply to this action, EPA consulted with tribal
officials in developing this action. EPA has consulted with tribal governments through the EPA
American Indian Environmental Office, included a representative of the Native American Water
Association on the advisory committee which developed recommendations regarding the
proposed rule and signed the AIP, and has addressed tribal concerns throughout the regulatory
development process, as appropriate. The consultation included participation in three tribal
conference calls (EPA regional tribal call (February 2008), National Indian Workgroup call
(March 2008), and National Tribal Water Conference (March 2008)). EPA requested comments
on the current TCR, requested suggestions for current TCR revisions (March 2008), and
presented possible revisions to the current TCR to the National Tribal Council (April 2008). In
addition, the advisory committee included entities representing tribal governments, and through
Economic Analysis for the Proposed RTCR 8-11
June 2010
-------
this process EPA ensured that their views were considered when the AIP recommendations for
the RTCR were developed. None of these consultations identified issues that were particular to
tribal entities. As a result of the tribal consultations and other tribal outreach, EPA has
determined that the proposed RTCR is not anticipated to have a negative impact on tribal
systems. Thus, Executive Order 13175 does not apply to this action.
EPA specifically solicits additional comment on this proposed action from tribal officials.
8.8. Executive Order 13045: 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 options considered by EPA.
The proposed RTCR is not subject to Executive Order 13045 because it is not
economically significant as defined in Executive Order 12866. This action's health and risk
assessments are contained in section VI.K. 1 of the proposed RTCR preamble and Chapter 6 of
the EA. This EA uses a qualitative approach in assessing the changes in risk anticipated for each
regulatory option relative the baseline (current TCR). EPA expects that the proposed RTCR
would provide additional protection, through the additional assessments and corrective action
required, to both children and adults who consume drinking water supplied from PWSs. EPA
also assumes that the benefits of the proposed rule, including reduced health risk, will
disproportionally accrue more to children because young children are more susceptible than
adults to some waterborne illnesses and are more likely to experience more serious effects from
infection. For example, the risk of mortality resulting from diarrhea is often greatest in the very
young and elderly (Rose 1997; Gerba et al. 1996), and viral and bacterial illnesses often
disproportionately affect children. Any overall benefits of the rule would reduce this mortality
risk for children.
The public is invited to submit comments or identify peer-reviewed studies and data that
assess effects of early life exposure to drinking water that contains fecal contaminants.
8.9. Executive Order 13211: Action 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
Economic Analysis for the Proposed RTCR 8-12
June 2010
-------
notices of proposed rulemaking, and notices of proposed rulemaking: (l)(i) that is a significant
regulatory action under Executive Order 12866 or any successor order, and (ii) is likely to have a
significant adverse effect on the supply, distribution, or use of energy; or (2) that is designated by
the Administrator of the Office of Information and Regulatory Affairs as a significant energy
action."
The proposed RTCR is not a "significant energy action" as defined in Executive Order
13211, because it is not likely to have a significant adverse effect on the supply, distribution, or
use of energy. This proposed rule is not a significant regulatory action under Executive Order
12866, and has not been designated by the Administrator of the Office of Information and
Regulatory Affairs as a significant energy action, for the reasons described as follows.
Energy Supply
The proposed RTCR does not regulate power generation, either directly or indirectly, and
public and private PWSs that the proposed RTCR applies to do not, as a rule, generate power.
Further, the energy cost increases borne by customers of PWSs as a result of the AIP option are a
low percentage of the total cost of water. Therefore, power generation utilities that purchase
water as part of their operations are unlikely to face any significant effects as a result of the AIP
option.
Energy Distribution
The proposed RTCR does not regulate any aspect of energy distribution and PWSs that
are regulated by the proposed RTCR already have electrical service. The rule is not expected to
increase peak electricity demand at PWSs. Therefore, EPA assumes that the existing
connections are adequate and that the AIP option has no discernable adverse effect on energy
distribution.
Energy Use
Because the proposed RTCR modifies existing regulations, very few PWSs are expected
to make modifications or changes that will alter energy use patterns. Therefore, EPA does not
expect any noticeable effect on the national levels of power generation in terms of average and
peak loads.
8.10. National Technology Transfer and Advancement Act
Section 12(d) of the NTTAA of 1995, Public Law No. 104-113, 12(d) (15 U.S.C. 272
note) directs EPA to use voluntary consensus standards in its regulatory activities unless to do so
would be inconsistent with applicable law or otherwise impractical. Voluntary consensus
standards are technical standards (e.g., materials specifications, test methods, sampling
procedures, and business practices) that are developed or adopted by voluntary consensus
standards bodies. NTTAA directs EPA to provide Congress, through OMB, explanations when
EPA decides not to use available and applicable voluntary consensus standards.
The proposed RTCR involves technical voluntary consensus standards. EPA proposes to
use several analytical methods to monitor for total coliforms and/or E. coli as they are described
Economic Analysis for the Proposed RTCR 8-13
June 2010
-------
in Standard Methods for the Examination of Water and Wastewater, 20th and 21st editions
(Clesceri et al. 1998; Eaton et al. 2005). Methods included in Standard Methods are voluntary
consensus standards. The proposed rule includes 11 methods that can be used to test for total
coliform, four of which are described in Standard Methods.
EPA welcomes comments on this aspect of the proposed rulemaking and, specifically,
invites the public to identify potentially-applicable voluntary consensus standards and to explain
why such standards should be used in this regulation.
8.11. Executive Order 12898: Federal Actions to Address Environmental Justice in
Minority Populations and Low-Income Populations
Executive Order 12898 (59 FR 7629, February 16, 1994) establishes federal executive
policy on environmental justice. Its main provision directs federal agencies, to the greatest extent
practicable and permitted by law, to make environmental justice part of their mission. Agencies
must do this by identifying and addressing as appropriate any disproportionately high and
adverse human health or environmental effects of their programs, policies, and activities on
minority populations and low-income populations in the United States.
EPA has determined that this proposed rule will not have disproportionately high and
adverse human health or environmental effects on minority or low-income populations because it
increases the level of environmental protection for all affected populations without having any
disproportionately high and adverse human health or environmental effects on any population,
including any minority or low-income population. The proposed RTCR applies uniformly to all
PWSs. Consequently, the proposed RTCR provides health protection equally to all income and
minority groups served by PWSs. The proposed RTCR and other drinking water regulations are
expected to have a positive effect on human health regardless of the social or economic status of
a specific population. To the extent that contaminants in drinking water might be
disproportionately high among minority or low-income populations (which is unknown), the
proposed RTCR contributes toward removing those differences by assuring that all PWSs meet
drinking water standards and take appropriate corrective action whenever appropriate. Thus, the
proposed RTCR meets the intent of the federal policy requiring incorporation of environmental
justice into federal agency missions.
8.12. Consultations with the Science Advisory Board, National Drinking Water
Advisory Council, and the Secretary of Health and Human Services as
Required by Section 1412 (d) and (e) of the SDWA
In accordance with section 1412 (d) and (e) of the SDWA, EPA consulted with the
Science Advisory Board (SAB), National Drinking Water Advisory Council (NDWAC), and the
Secretary of the US Department of Health and Human Services on the proposed RTCR.
EPA met with the SAB Drinking Water Committee (DWC) and considered their
recommendations in developing data requirements to better understand the impacts of the RTCR.
In response to the SAB DWC recommendations, EPA also conducted sensitivity analyses, to
explore a wider range of assumptions regarding the percentage of assessments leading to
corrective actions and to demonstrate that using an annual average for occurrence provided
Economic Analysis for the Proposed RTCR 8-14
June 2010
-------
results comparable to varying the occurrence based on the season. In addition, EPA added an
exhibit in the EA that summarizes all significant model parameters and assumptions, their
influence on variability and uncertainty, and their most likely effect on benefits or costs. A copy
of the SAB report (SAB 2010) is available in the docket for the proposed RTCR.
EPA also consulted with NDWAC and considered their recommendations in developing
the public notification requirements for the proposed rule.
EPA completed its consultation with the US Department of Health and Human Services,
as required by SDWA section 1412(d). In addition, EPA provided an informational briefing to
the Food and Safety Group of the Food and Drug Administration.
Details about EPA's consultations with SAB, NDWAC, and the US Department of
Health and Human Services can be found at Section VII.K of the proposed RTCR preamble.
8.13. Impacts on Sensitive Subpopulations as Required by Section 1412(b)(3)(c)(i)
of the 1996 Amendments to the Safe Drinking Water Act (SDWA)
EPA is required to seek public comment regarding the effects of contamination
associated with the proposed RTCR on the general population and sensitive subpopulations.
Sensitive subpopulations include "infants, children, pregnant women, the elderly, individuals
with a history of serious illness, or other subpopulations that are identified as likely to be at
greater risk of adverse health effects due to exposure to contaminants in drinking water than the
general population" (SDWA section 1412(b)(3)(C)(i)(V), 42 U.S.C 300g-l(b)(3)(C)(i)(V)).
Pregnant and lactating women may be at an increased risk from pathogens as well as act
as a source of infection for newborns. Infection during pregnancy may also result in the
transmission of infection from the mother to the child in utero, during birth, or shortly thereafter.
Since very young children do not have fully developed immune systems, they are at increased
risk and are particularly difficult to treat.
Infectious diseases are also a major problem for the elderly because immune function
declines with age. As a result, outbreaks of waterborne diseases can be devastating on the elderly
community (e.g., nursing homes) and may increase the possibility of significantly higher
mortality rates in the elderly than in the general population.
Immunocompromised individuals are a growing proportion of the population with the
continued increase in HIV/AIDS, the aging population, and the escalation in organ and tissue
transplantations. Immunocompromised individuals are more susceptible to severe and invasive
infection. These infections are particularly difficult to treat and can result in a significantly
higher mortality than in immunocompetent persons.
It is anticipated that the requirements of the proposed RTCR will help reduce pathways of
entry for fecal contamination and/or waterborne pathogens into the distribution system, thereby
reducing exposure and risk from these contaminants in drinking water to the entire general
population. The proposed RTCR seeks to provide a similar level of drinking water protection to
all groups including sensitive subpopulations, thus meeting the intent of this federal policy.
Economic Analysis for the Proposed RTCR 8-15
June 2010
-------
8.14. Plain Language as Required by Executive Order 12866
Executive Order 12866 requires each agency to write its rules in plain language.
Readable regulations help the public find requirements quickly and understand them easily.
Readable regulations may also increase compliance, strengthen enforcement, and decrease
mistakes, frustration, phone calls, appeals, and distrust of government. EPA has made every
effort to write this preamble to the proposed rule in as clear, concise, and unambiguous manner
as possible. EPA requests comments on how to improve rule language to enhance readability and
make it easier to understand.
Economic Analysis for the Proposed RTCR 8-16
June 2010
-------
9. Comparison of Benefits and Costs
9.1 National Benefits and Costs of the AIP Option Considered in Comparison to
the Current Total Coliform Rule and Alternative Option
The Environmental Protection Agency (EPA) is proposing a regulatory option consistent
with the Total Coliform Rule/Distribution System Advisory Committee (TCRDSAC) Agreement
in Principle (AIP) for the Revised Total Coliform Rule (RTCR) (the AIP option). The primary
discussion of costs and benefits in this chapter focuses on the requirements of the AIP option in
comparison to the current Total Coliform Rule (TCR) and the Alternative option considered.
The primary benefit of the AIP option is to further reduce the risk of fecal contamination
of public drinking water from the current baseline risk under the existing TCR. The indicator of
this benefit that is most feasible to predict is the reduction in E. coli occurrence in Public Water
Systems (PWSs). As a fecal contamination indicator, E. coli can also co-occur with other
pathogenic organisms shed in feces, such as viruses, bacteria, and pathogenic protozoa at a rate
that is not quantified.1 EPA believes that a reduction in E. coli occurrence, together with other
AIP option actions, will result in reduced PWS fecal contamination and may facilitate a
consequent reduction in endemic and epidemic waterborne disease in the United States.
The current TCR mandates sampling for total coliforms (TC) and E. coli in PWSs in the
U.S. However, under the current TCR, PWSs that encounter samples testing positive for E. coli
have the option to implement an effective long-term corrective action, a short-term corrective
action, or they can choose not to implement any corrective action. This wide variability in
effectiveness of corrective actions increases the potential for exposure to fecally-contaminated
drinking water when the response to E. coli occurrence is inadequate or ineffective.
EPA proposes to augment the current TCR with, among other components, the
requirement for systems to implement mandatory assessments and corrective actions. PWSs
would be required to conduct either a Level 1 or Level 2 assessment, depending on the severity
of a trigger (based on TC vs. E. coli and the number of triggers). As described in Chapter 3, the
Level 1 and 2 assessments require a PWS to more formally pursue the cause of a non-acute or
acute violation, and to address the problem with an appropriate corrective action. It is the
combination of these two requirements (assessments and corrective actions) from which most of
the net benefits and costs of the AIP option would derive.
A significant amount of the costs of the additional activities required under the AIP
option would be offset by the reduced costs from some decreases in monitoring (additional
routine and repeat) and Public Notification (PN). The decreases in monitoring may limit the
ability of PWSs to identify E. coli positive samples at the levels found under the current TCR
when using the same monitoring techniques. However, predictive modeling results show that
significantly more contamination events will be prevented by assessments that lead to additional
corrective actions under the AIP option than would be missed by reduced additional routine and
1 Chapter 2 of this EA provides examples of waterborne pathogens that can occur in PWSs. Also, Edberg, R. (2000)
discusses the use of E. coli as an indicator of drinking water quality.
Economic Analysis for the Proposed RTCR
9-1
June 2010
-------
repeat monitoring for systems serving 4,100 or fewer people.2 Some of this reduction in both
monitoring and costs would be offset by the increase in routine monitoring predicted for systems
serving 1,000 or fewer people because fewer systems are likely to qualify for reduced monitoring
based on the more stringent requirements under the AIP option3.
Further cost reductions may be achieved by the requirement for a mandatory sampling plan
under the AIP option, which would increase monitoring efficiency by having the operator submit
a planned approach, including regular sampling sites that are representative of the water
throughout the distribution system. The plan may also specify repeat sample locations that may,
if approved by the State, satisfy source water sampling requirements as well; this benefit would
likely apply to some of the PWSs with limited or no distribution systems.
These increases in routine monitoring,along with more efficient monitoring based on
approved monitoring plans, are expected to increase the ability of PWSs to detect EC a larger
percentage of the time when it is present. This increase in detection will be partially offset by
reductions in additional routine and repeat monitoring. For the reasons discussed this chapter and
in analyses and discussion presented throughout this Economic Analysis (EA), EPA concludes
that the AIP option will improve public health protection compared to the current TCR.
The Alternative option considered in this EA would provide benefits equal to or greater
than the AIP option in terms of decreased risk of PWSs delivering contaminated water to the
public and subsequent decreases in potential health risks. However, monitoring costs under the
Alternative option would increase significantly, particularly for small systems, causing a
significant challenge to the effective implementation of the Alternative option.
The remainder of this chapter provides further detail on the costs and benefits and how
they compare under the AIP and Alternative options:
• National benefits (Section 9.1.1)
• National costs (Section 9.1.2)
• Uncertainty and nonquantified benefits (Section 9.2)
• Comparison of regulatory alternatives in cost/benefit terms (Section 9.3)
2 Chapter 6 of this EA explores the trade-off in a stepwise analysis of the risk benefit resulting from implementation
of assessments resulting in corrective actions vs. the potential increase in risk due to decreased sampling frequency.
That analysis concludes that for each type of system and size category, the AIP would avoid a larger number of
acute events than it would fail to diagnose.
3 Chapter 5 (Section 5.3) includes a description of the determination of PWS sampling frequencies that are
applicable beginning in year 6 following RTCR promulgation.
Economic Analysis for the Proposed RTCR 9-2 June 2010
-------
9.1.1. National Benefits of the Regulatory Options Considered
The benefit analyses performed in this EA and summarized in this section consider the
net changes in TC and E. coli occurrence under the AIP option as compared to the current TCR4.
In promulgating the RTCR as defined by the AIP option, EPA expects to further reduce the risk
of contamination of public drinking water from the current baseline risk under the existing TCR.
The options considered during development of this proposed rule and analyzed as part of this EA
are designed to achieve this reduction while maintaining public health protection in a cost-
effective manner.
This section examines the benefits in terms of trade-offs between compliance with the
current TCR and the AIP and Alternative options. Based on limitations in available data
(described further in Chapter 6, Section 6.3), EPA determined that benefits could not be
calculated in terms of avoided cases of (or costs related to) morbidity or mortality. EPA used
several methods to qualitatively evaluate the benefits of the AIP option. The qualitative
evaluation uses both the judgment of EPA as informed by the TCRDSAC deliberations as well
as quantitative estimates of changes in TC occurrence and counts of systems implementing
corrective actions. The evaluation characterizes, in relative terms, the reduction in risk for each
regulatory scenario as compared to baseline conditions.
Since E. coli is an indicator of fecal contamination, EPA assumed that a decrease in E.
coli occurrence in the distribution system would be associated with a decrease in fecal
contamination in the distribution system. In general, this decrease in fecal contamination should
reduce the potential risk to human health for PWS customers. Thus, any reduction in E. coli
occurrence is considered a benefit of the proposed RTCR. Also, since fecal contamination may
contain waterborne pathogens including bacteria, viruses, and parasitic protozoa, in general, a
reduction in fecal contamination should also reduce the risk from these other contaminants.
As presented in Exhibit 4.9, the percentages of samples that are positive for TC and E.
coli are generally higher for PWSs serving 4,100 or fewer people than those serving more than
4,100 people. PWSs with higher TC and E. coli occurrence are more likely to be triggered into
assessments and corrective action. As discussed previously, EPA believes that the assessments
and corrective action will lead to a decrease in TC and E. coli occurrence. Because the PWSs
serving 4,100 people or fewer have a higher initial E. coli occurrence and will be triggered into
more assessments and corrective actions than larger PWSs, the increase in benefits for these
small systems will be more evident as compared to the larger systems. In particular, model
results suggest that customers of small ground water Transient Noncommunity Water Systems
(TNCWSs), which constitute approximately 30% of PWSs, would experience the most obvious
benefit under the AIP option. That is, the occurrence of E. coli is predicted to decrease more for
TNCWSs than for other systems types.
As discussed above, because there was insufficient data on the co-occurrence of E. coli
and waterborne pathogens, it was not possible to quantify the adverse health effects in the human
population served by PWSs. Thus, EPA employed a qualitative approach as the primary method
for analyzing the expected change in exposure to fecal contamination based on each individual
4 The "current TCR" is the baseline used in the RTCR EA, and consists of the rule components and resulting
activities of the 1989 TCR with effects of the Ground Water Rule (GWR) implementation in 2010 incorporated.
Chapter 4 of this EA presents information used to develop this baseline, and Chapter 5 (Section 5.3.1) describes how
the occurrence and predictive model incorporates the GWR effects.
Economic Analysis for the Proposed RTCR
9-3
June 2010
-------
rule component. The qualitative analysis considers the anticipated changes in sampling and
corrective action regimens and estimates the corresponding anticipated reduction or increase in
exposure to fecal contamination. Because the reduced exposure to fecal contamination conveys
inadequate information about pathogen exposure, the analysis in this section is a hazard analysis
rather than a risk analysis. Nevertheless, because the intent is to reduce risk by reducing hazard
(potential exposure to fecal contamination and/or waterborne pathogens), risk is used as the
discussion endpoint. Section 9.1.1.1 below summarizes the results of the qualitative analysis, and
Chapter 6 (Section 6.2) presents this information in more detail.
Although a qualitative analysis was the primary method employed for analyzing
potential changes in risk, quantitative measures were used to support the qualitative conclusions,
providing additional perspective on potential changes in risk. In particular, EPA considered
modeled predictions of the reductions in the numbers of violations/assessment triggers and of the
number of corrective actions to be performed as they may relate to risk. Section 9.1.1.2 below
discusses the predicted changes in these outcomes. More detailed discussions of the predictive
model and analyses results are presented in Chapters 5 and 6.
9.1.1.1. Qualitative Comparison - Relative Risks of the AIP and Alternative Options
Compared to Baseline
When revising an existing drinking water regulation, one of the main concerns is to
ensure that backsliding on water quality and public health protection does not occur. Risk
reduction for the AIP option is characterized by the activities performed that are presumed to
reduce risk of exposing the public to contaminated water. These activities are considered under
each rule component presented in Exhibit 9. lb. Under repeat and additional routine monitoring
provisions for both rule options, there is a potential to contribute to increased risk for PWS
customers because TC monitoring frequency may be reduced for some PWSs. However, this
increase in risk is expected to be more than offset by potential decreases in risk from increased
routine monitoring and the addition of the assessments and corrective action provisions that will
find and fix problems identified by monitoring.
Exhibit 9.1a illustrates the predicted reduced frequency at which TC-positives occur
subsequent to the implementation of the AIP and Alternative options. TC occurrence is used as a
surrogate for indicating the existence of potential pathways through which waterborne pathogens
may enter a PWS. Exhibit 9.1a illustrates the combined effects on TC occurrence resulting from
changes in monitoring and the effects of assessments and corrective actions for the different rule
options illustrated. The relative trends evident in Exhibit 9.1a for TNCWSs also pertain to other
PWS categories as illustrated in Chapter 5 of proposed RTCR EA. EPA chose to include the
characterization for TNCWSs because they represent the system category of largest influence on
the national impacts.
Economic Analysis for the Proposed RTCR
9-4
June 2010
-------
Exhibit 9.1a Ground Water (GW) Transient Noncommunity Water System TC
Occurrence
0.045
0.04
0.035
8.
0.03
S 0.025
;i o.o2
i/i
o
CL
H 0.015
ns
3
< 0.01
0.005
¦TCR No GWR
•TCR
AIP
Alt
10
15 20
Years
25
30
35
Note: 1) Six Year 2005 TC-positive occurrence is representative of all GW TNCWS. The rate presented may
underestimate the occurrence for systems serving 25-4,100 individuals. 2) Graph shows the 30-year modeled period
discussed in Ch. 5. Model years 3-27 represent the 25-year period of analysis for this EA. Model year 11 begins
the steady state, during which systems that qualified for reduced monitoring are now sampling on their reduced
schedules. The criteria and timing of this monitoring adjustment is discussed in Section 5.3.2.2 of this EA. 3) The
results represented by the curves for TCR, AIP, and Alt all incorporate the effects of the GWR.
The effect that the proposed changes to public notification requirements for monthly/non-
acute MCL violations have on risk is difficult to predict. Some factors, such as reduction in
available public information and possible PWS complacency, lead to a potential increase in risk
and other factors, such as less confusion (PN more in line with potential health risks) and PWSs
resources used more efficiently, lead to a potential decrease, as discussed in Exhibit 9.1b. This
change to PN is addressing a key concern expressed by various stakeholders in the advisory
committee and during the Six-Year Review 1 comment solicitation process. By eliminating the
requirement and replacing it with assessment and corrective action requirements, the Agency
expects less public confusion, more effective use of resources, and increased transparency. Other
proposed rule components are expected to have a negligible effect on risk. However, the overall
effect of the proposed RTCR is expected to be a further reduction in risk from the current
baseline risk under the current TCR. Chapter 6 presents a detailed discussion of the potential
influence on health risk for each proposed rule component.
Economic Analysis for the Proposed RTCR
9-5
June 2010
-------
Exhibit 9.1b Potential Changes in Risk under the AIP and Alternative Options Relative to the Current TCR
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
AIP option
Alternative
option
AIP option
Alternative
option
AIP option
Alternative
option
Implementation
Activities
None
None
None
None
No change
No change
Routine
Monitoring
(Including
Reduced
Monitoring)
None
None
Increased stringency
in requirements to
qualify for reduced
monitoring along with
requirement to return
to baseline
monitoring upon loss
of these criteria is
expected to result in
decreased risk (i.e.,
fewer PWSs will
qualify and therefore
more will monitor
more frequently).
PWSs all monitor
monthly in the first few
years of implementation
of the RTCR, which is an
increase in sampling
frequency for systems
that monitor quarterly or
annually under the
current TCR. After the
first few years, systems
may reduce to quarterly,
but none may reduce to
annual monitoring,
creating a decrease in
risk for systems on
annual monitoring under
the current TCR.
Decrease
Decrease
Economic Analysis for the Proposed RTCR
9-6
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
Repeat
Monitoring
Required repeat
samples reduced
from 4 to 3 for
systems serving
<1,000 people
Required repeat
samples
reduced from 4
to 3 for systems
serving <1,000
people
None
None
Increase
Increase
Additional
Routine
Monitoring
Additional routine
samples are no
longer required for
PWSs monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional routine
samples from 5 to
3.
Additional
routine samples
are no longer
required for
PWSs
monitoring
monthly.
Ground water
PWSs serving
<1,000 people
would reduce
additional
routine samples
from 5 to 3.
None
None
Increase
Increase
Annual
Inspections
None (only states
currently
performing annual
inspections are
expected to
continue)
Based on
discussions with
stakeholders,
state s that
currently
conduct annual
site
assessments
under the
current TCR
may no longer
have the
None (only state s
currently performing
annual inspections
are expected to
continue)
None
No change
Increase
Economic Analysis for the Proposed RTCR
9-7
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
resources to
continue the
inspections and
conduct
quarterly
monitoring
under the
Alternative
option.
Assessments
None
None
Mandatory
assessments are a
new requirement.
Mandatory assessments
are a new requirement.
Decrease
Decrease
Corrective
Actions
None
None
Mandatory corrective
actions are a new
requirement.
Mandatory corrective
actions are a new
requirement.
Decrease
Decrease
Public
Notification -
Monthly/Non-
Acute MCL
Violations
Reduction in
available public
information
Possible PWS
complacency
Reduction in
available public
information
Possible PWS
complacency
Less confusion (PN
more in line with
potential health risks)
PWS resources used
more efficiently
Less confusion (PN
more in line with
potential health risks)
PWS resources used
more efficiently
Unknown
Unknown
Economic Analysis for the Proposed RTCR
9-8
June 2010
-------
Proposed Rule
Component
Factors Leading to a Potential
Increase in Risk
Factors Leading to a Potential Decrease in
Risk
Overall Predicted Change in
Risk
Public
Notification -
Monitoring and
Reporting
Violations
None
None
Increased stringency
of PNs motivates
PWSs to conduct
required sampling.
Increased stringency of
PNs motivates PWSs to
conduct required
sampling.
Decrease
Decrease
Overall
Decrease
Decrease
Note: Detailed discussion of the rationale for determinations of potential risk for each rule component is presented in Ch. 6 (Section 6.2) of this EA. Implementation
activities consist of administrative activities by PWSs and states to implement the rule.
Assessment of potential changes in risk for monitoring components is an overall assessment. Potential changes (or static state) of risk for particular system sizes
and types differ according to individual regulatory requirements and are discussed in Section 6.2. Chapter 3 provides a detailed description of the regulatory
components for all three regulatory scenarios, and the preamble to the RTCR provides additional discussion of the TCRDSAC process and the rationale underlying
the structure of the regulatory options considered.
Economic Analysis for the Proposed RTCR
9-9
June 2010
-------
9.1.1.2. Comparison of Quantified Benefits between Baseline and AIP and
Alternative Options
The quantified portion of the benefits analysis focuses on several measures that contribute to
the changes in risk expected under the AIP option. Specifically, EPA modeled the predicted
outcomes based on each regulatory option considered - baseline (current TCR), the AIP option,
and the Alternative option - in the form of estimates of non-acute violations for the current TCR
and assessment triggers for the AIP and Alternative options; E. coli violations; and the number of
corrective actions implemented under each option.
This section presents a summary of the estimated impacts of both regulatory options
considered in comparison to the current TCR. Evaluation of each of these endpoints informed
EPA's understanding of potential changes to the underlying quality of drinking water. In
particular, the number of corrective actions performed has a strong relationship to potential
improvements in water quality and public health. For a given level of TC and E. coli occurrence,
an increase in the number of corrective actions implemented would lead to improved water
quality. However, a reduction in sampling leads to a reduction in TC and E. coli positives being
found, which in turn leads to a reduction in assessments and corrective actions being
implemented. The number of TC and E. coli positives that are prevented, missed, or found under
each regulatory option considered in comparison to those predicted under the current TCR
results in estimates of annual non-acute and acute violations (current TCR) and assessment
triggers (AIP and Alternative options). Section 6.4 of this EA presents a step-wise uncertainty
analysis of the competing effects of additional protective activity (assessments and corrective
actions) and decreased additional routine and repeat sampling of the regulatory alternatives
compared to the current TCR. The results of this uncertainty analysis showed that for all
categories of systems, more TC and E. coli positives would be prevented than missed.
For each of the graphs presented in Exhibit 9.2 through Exhibit 9.7 there are two main
model drivers that impact the endpoints depicted; the total number of samples taken over time
(including routine, additional routine, and repeat) and corrective actions taken. When looking at
the comparisons between the TCR with the AIP option across all PWSs, the overall impact of the
total numbers of samples taken is negligible because the total number of samples predicted to be
taken throughout the period of analysis is almost the same (approximately 82 million samples)
under both the TCR and AIP option, as shown in Exhibits 6.2 - 6.4 of this EA. For the
Alternative option, the analysis predicts that almost 87 million total samples will be taken over
the period of analysis. Based on the relationships of total samples taken between the TCR, AIP
option, and Alternative option, the best way to interpret the graphs presented in this section is in
a step-wise manner.
The first comparison is between the TCR and proposed RTCR as defined by the AIP option.
Because similar total numbers of samples are taken under each, the major effect seen in the
graphs can be isolated to the effects that implementation of corrective actions will have on
underlying occurrence and how that occurrence influences the endpoint in question. In each
graph, this is depicted by a marked reduction in the endpoint under the AIP option compared to
the TCR and is a reflection of overall better water quality. The second comparison can then be
made of the Alternative option against the AIP option. In each graph, the endpoints for the
Economic Analysis for the Proposed RTCR
9-10
June 2010
-------
Alternative option are above those for the AIP option and represent an additional benefit over the
AIP option. This additional benefit is a primarily a function of the additional diagnostic abilities
gained through increased monitoring under the Alternative option, and is especially prominent in
the early years of the analysis when all systems are required to monitor at least monthly.
More detailed descriptions of each endpoint considered in terms of the evaluation process
described previously are provided in this section as they apply to the individual graphs in
Exhibits 9.2 through 9.7. The graphs shown in this section are presented first in nondiscounted
terms and then based on a discount rate of 3% to reflect the reduced valuation of potential
benefits over time, consistent with the presentation of costs in the section that follows. Graphs of
benefits discounted using 7% discount rates are presented in Appendix B to this EA.
Exhibit 9.2 shows the effect (on average across all PWSs) of the AIP and the Alternative
options on the annual number of non-acute violations (assessment triggers under the AIP and
Alternative options) over time. The estimated reduction of approximately 1,000 trigger events in
moving from the current TCR option to the AIP option is a net result from the following effects
described previously in this section: improved water quality (events prevented); retained
diagnostic power (events found), which in turn leads to prevention of some future events; and
reduced diagnostic power (missed events). The amount of this reduction would be some amount
lower if the analysis could account for events where contamination was present but not
diagnosed ("missed" events). This is why a similar, but smaller, reduction from the current TCR
is seen under the Alternative option. During the first 9 years of sampling under the AIP option
and Alternative option sampling regimens, diagnostic power under the Alternative option is at a
maximum based on the "All-monthly" sampling requirement. In year 9, following a 5-year
period of assessment5, all systems that qualified for reduced monitoring begin sampling
according to their new regimens. This is the start of the relatively steady state of the period of
analysis, driven by the static nature of the percentages of systems following monthly, quarterly,
or annual sampling regimens throughout the remainder of the analysis (years 9 through 25).
Trigger events under the Alternative option remain at a higher steady state than under the AIP
option based on the increased diagnostic ability provided by more frequent sampling under the
Alternative option. The additional number of triggers identified by increased sampling under the
Alternative option translate into greater potential benefits than those under the AIP option.
Exhibit 9.3 shows the effect (on average across all PWSs) of the AIP option and the
Alternative option with respect to acute (E. coli) violations found over the 25-year period of
analysis in comparison to the current TCR. The overall reduction in annual acute violations
under the AIP option of more than 100 events is a measure that should correlate more closely
with expected benefits (i.e., reductions in adverse health outcomes) than non-acute violations
because acute violations are a direct result of measurement of E. coli in water. Again a similar,
but smaller, reduction is seen under the Alternative option after steady state is achieved. This is
the result of two off-setting effects. The "true" number of steady state violations under the
Alternative option is lower, because there is a greater likelihood that violations will be found and
fixed. However, the additional monitoring leads to a higher percentage of violations being
detected. This second effect outweighs the first, so that the total number of detected violations in
5 The assessment period is 3 years for CWSs, and CWSs begin sampling on their reduced regimens in year 7 after
promulgation.
Economic Analysis for the Proposed RTCR
9-11
June 2010
-------
the steady state is higher than for the AIP, even though the underlying "true" number of
violations is lower. This lower number of "true" violations means that the Alternate option is
more protective of public health, even though more violations are detected.
Exhibit 9.4 presents estimates over the 25-year period of analysis of the increase in
corrective actions (on average across all PWSs) attributable to the regulatory options considered.
The performance of these additional corrective actions is expected to result in the most direct
benefits under the AIP option. Because only the numbers of corrective actions estimated under
the AIP option and Alternative option net of those estimated under the current TCR were
modeled, the reference point for comparison to the current TCR is the base (zero) line in the
graph. The RTCR EA assumes that corrective actions are being performed under the current
TCR. These are taken into account by assuming only a modest net increase in effective
corrective actions implemented, calculated as 10% of assessments performed under each
regulatory alternative that are net of those performed under the current TCR.
Exhibit 9.4 indicates that more corrective actions would be implemented under the
Alternative option than under the AIP option. This is driven, again, by the increased diagnostic
power of more sampling and reflects additional potential benefits beyond those gained under the
AIP option. However, not quantified are the numbers of corrective actions that would be
identified by the annual inspections required to qualify for annual monitoring under the AIP
option but not under the Alternative option. Inclusion of the additional corrective actions taken
under the AIP option due to the annual inspection requirement could narrow the difference
between the potential benefits of the Alternative option and AIP option.
Taken together, Exhibits 9.2 - 9.4 indicate that the modeled endpoints for the AIP option
and Alternative option predict positive benefits in comparison to the current TCR; in particular,
the Alternative option would capture more benefits than the AIP option. These outcomes are
consistent with the qualitative assessment of the benefits summarized in Section 9.1.1.1 above.
The assessment based on Exhibits 9.5 - 9.7 is also consistent with the qualitative analysis
in Section 9.1.1.1: For each of the discounted endpoints presented over time in Exhibits 9.5 -
9.7, the graphs show that (on average across all PWSs) the Alternative option would provide
more benefits than the AIP option, and both provide more benefits than the current TCR. The
major difference between the AIP option and Alternative option is the increased monitoring that
would be required under the Alternative option. The increased diagnostic ability of the extra
samples taken under the Alternative option is seen in the large difference in the endpoint counts
through the first several years in the graphs below. Absent this effect, the Alternative option
would mirror the AIP option in the graphs. Thus, even though the predicted endpoints are greater
than the current TCR at first, it is due to initially finding more problems through monitoring.
This would reflect a frontloading of benefits under the Alternative option at the beginning of the
implementation period. The benefits, however, would tend to even out over time between the
AIP option and Alternative option as eligible systems qualify for less frequent (quarterly)
monitoring under the Alternative option.
Economic Analysis for the Proposed RTCR
9-12
June 2010
-------
Exhibit 9.2 Estimates of Non-Acute Violations (TCR) and Level 1 Assessment
Triggers (AIP and Alternative Options)
Q_
<
£
CD
O)
,D)
'Z
1-
-4-»
t/>
c
a)
t/>
T3
a)
a)
tn
-4-»
c
<
3
o
o
Q)
w
>
-Q
Q.
t/>
O
c
+-»
O
'•*->
<
ra
"D
O
C
(0
>
Q)
3
O
c
o
2
14,000
13,000
12,000
11,000
10,000
9,000
8,000
7,000
6,000
¦TCR - Non-Acute Violations
- AIP Option - Level 1 Assessment Triggers
-AltOption- Level 1 Assessment Triggers
5,000 i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1—
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) The X-axis begins at Year 4 after rule promulgation, which is the first year under full implementation of the AIP or
Alternative options.
2) The annual rates of non-acute violations (TCR) and Level 1 assessment triggers (AIP and Alternative options) as
predicted by the model reach a steady state beginning in approximately Year 9, by which time PWSs that would be
expected to meet the criteria for reduced monitoring would have begun it.
3) Non-acute violations/Level 1 assessment triggers are predicted to be lower under the AIP option because,
although both options are preventing contamination events as discussed in Section 6.4 of this EA, the Alternative
option is finding more because of its increased sampling and correspondingly higher diagnostic power.
Economic Analysis for the Proposed RTCR
9-13
June 2010
-------
Exhibit 9.3 Estimates of Acute Violations (TCR) and E. coli MCL Violations (AIP
and Alternative Options)
<
"O
c
(0
Q.
<
(/>
c
O
tf)
+-»
a)
CO
D
o
(0
>
>
-1
o
¦a
$
c
3
o
o
o
o
w
ui
-a
c
"O
o
c
2
(0
a:
t/>
o
H
c
O
'•*->
t/>
c
Q_
O
O
-4—'
(0
o
i>
a)
+-»
3
O
<
1,300
1,200
1,100
1,000
900
800
700
600
500
"TCR - Acute Violations
AIP Option - E. coli MCL Violations
Alt Option - E. coli MCL Violations
' \ \
\ /
x /
* - - ' /A
400 i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 r~
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) X-axis begins at Year 4 after rule promulgation, which is the first year under full implementation of the AIP or
Alternative options.
2) The annual rates of acute violations (TCR) and EC violations (AIP and Alternative options) as predicted by the
model reach steady state in approximately Year 9, by which time PWSs that would be expected to meet the criteria
for a reduced monitoring schedule would have begun it. Estimates represent the annual number of acute violations
found by each option and the TCR.
3) Acute violations are predicted to be lower under the AIP option because, although both options are preventing
contamination events as discussed in Section 6.4 of this EA, the Alternative option is finding more because of its
increased sampling and correspondingly higher diagnostic power.
Economic Analysis for the Proposed RTCR
9-14
June 2010
-------
Exhibit 9.4 Estimates of Corrective Actions
• AIP Option - Corrective Actions
Alt Option - Corrective Actions
1,600
1,500
1,400
1,300
1,200
1,100
1,000
900
800
700
600
500
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) X-axis begins at Year 4 after rule promulgation, which is the first year under full implementation of the AIP option or
Alternative option. The annual rates of corrective actions as predicted by the model reach a steady state beginning
approximately in Year 9, by which time PWSs that would be expected to meet the criteria for reduced monitoring
would have begun it.
2) Includes L1 and L2 corrective actions. All corrective actions performed are in addition to activity under the current
TCR, which does not require corrective actions. Therefore the current TCR is not included in this graph.
3) Corrective actions are predicted to be higher under the Alternative option because, although both options are
preventing contamination events as discussed in Section 6.4 of this EA, the Alternative option is finding more
because of its increased sampling and correspondingly higher diagnostic power.
in
0)
3
£
¦o
c
o
o
<
'•P
o
£
o
o
Economic Analysis for the Proposed RTCR
9-15
June 2010
-------
Exhibit 9.5 Discounted Estimates of Non-Acute Violations (TCR) and Level 1
Assessment Triggers (AIP and Alternative Options) (3% Discount Rate)
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) X-axis begins at Year 4 after rule promulgation, which is the first year under full implementation of the AIP option or
Alternative option.
2)The annual rates of non-acute violations (TCR) and Level 1 assessment triggers (AIP and Alternative options) as
predicted by the model reach a steady state beginning in approximately Year 9, by which time PWSs that would be
expected to meet the criteria for reduced monitoring would have begun it.
3) Non-acute violations/Level 1 assessment triggers are predicted to be lower under the AIP option because,
although both options are preventing contamination events as discussed in Section 6.4 of this EA, the Alternative
option is finding more because of its increased sampling and correspondingly higher diagnostic power.
Economic Analysis for the Proposed RTCR
9-16
June 2010
-------
Exhibit 9.6 Discounted Estimates of Acute Violations (TCR) and E. coli MCL
Violations (AIP and Alternative Options) (3% Discount Rate)
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) X-axis begins at Year 4 after rule promulgation, which is the first year under full implementation of the AIP option or
Alternative option.
2) The annual rates of acute violations (TCR) and EC violations (AIP and Alternative options) as predicted by the
model reach steady state in approximately Year 9, by which time PWSs that would be expected to meet the criteria
for reduced monitoring would have begun it.
3) Acute violations are predicted to be lower under the AIP option because, although both options are preventing
contamination events as discussed in Section 6.4 of this EA, the Alternative option is finding more because of its
increased sampling and correspondingly higher diagnostic power.
Economic Analysis for the Proposed RTCR
9-17
June 2010
-------
Exhibit 9.7 Discounted Estimates of Corrective Actions
(3% Discount Rate)
CC
+->
c
3
O
o
w
C
c
o
o
<
'•P
o
a;
o
o
1,400
1,300
1,200
1,100
1,000
900
800
700
600
500
400
300
200
100
¦ AIP Option - Corrective Actions
Alt Option - Corrective Actions
—i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1—
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Source: Proposed RTCR occurrence model output.
Notes:
1) X-axis begins at Year 4, which is the first year under full implementation of the AIP or Alternative options.
2) Includes L1 and L2 corrective actions. The annual rates of corrective actions as predicted by the model reach a
steady state beginning in approximately Year 9, by which time PWSs that would be expected to meet the criteria for
reduced monitoring would have begun it. All corrective actions performed are in addition to activity under the current
TCR, which does not require corrective actions. Therefore the current TCR is not included in this graph.
3) Corrective actions are predicted to be higher under the Alternative option because, although both options are
preventing contamination events as discussed in Section 6.4 of this EA, the Alternative option is finding more
because of its increased sampling and correspondingly higher diagnostic power.
Economic Analysis for the Proposed RTCR
9-18
June 2010
-------
9.1.2 National Cost Summary
To understand the net impacts of the AIP option on public water systems and states, EPA
first used available data, information, and best professional judgment to characterize how PWSs
and states are currently implementing the current TCR. Then, EPA considered the net change in
costs that would result from implementing the AIP option or Alternative option as compared to
the costs of continuing with the current TCR. The objective was to provide the net change in
costs resulting from revisions to the current TCR rather than absolute totals. The exhibits below
present this net change in costs several different ways to help analyze the impacts of the AIP
option.
Costs are estimated for different PWS types and size categories (eight size categories are
used based on population served) using unit costs developed in the Technology and Cost
Document for the Proposed Revised Total Coliform Rule (USEPA, 2010d). Cost analyses for
PWSs include estimates to implement the rule; to revise sampling plans; to conduct routine
monitoring, additional routine monitoring, and repeat monitoring; to perform Level 1 and Level
2 assessments and implement corrective actions; and to provide public notification in the case of
violations. State cost analyses include estimates of the labor burdens that states would incur,
including staff training on RTCR requirements and conducting annual administration, reviewing
monitoring reports, reviewing and approving corrective action plans, and for recordkeeping.
Chapter 7 of this EA provides detailed discussion on the underlying cost-buildup for each rule
component analyzed within the cost model.
Considering costs over time, Exhibits 9.8 through 9.10 show that most of the cost
difference between the AIP option and Alternative option will be experienced in the first 8 years
after promulgation. Nevertheless, the costs of the AIP option remain lower than those of the
Alternative option throughout the period of analysis. The lower costs of the AIP option are
driven by the monitoring requirements of the respective options (e.g., AIP option allows annual
monitoring for qualifying PWSs while the Alternative option requires at least quarterly
monitoring for all PWSs).
Exhibit 9.11 presents the total and net change in costs for the TCR and the regulatory
options considered for both systems and states. The estimated net change in costs is relatively
small for PWSs and states on a national level; for the AIP option, total costs would increase from
the current TCR by approximately 7%, or $14 million (using a 3 percent discount rate). The
Alternative option, based on the increased sampling regimen, increases total costs from baseline
by approximately 15%, or about $27 million (using a 3 percent discount rate). Given these
relatively low net costs, both options are below the threshold of a "significant regulatory action"
under Executive Order 12866. The approximate $13 million difference between the two options,
and the net cost of each option, are not particularly large in absolute terms compared to a
threshold for economic significance of $100 million annually.
As shown in Exhibit 9.11, the largest portion of the net cost increase is borne by PWSs,
which incur approximately 99% of the proposed revised rule's net annualized present value costs
(using a 3 percent discount rate). States are expected to incur the remaining costs. PWS costs
represent only a 7% and 15% increase, respectively, for the AIP option and Alternative option, in
Economic Analysis for the Proposed RTCR
9-19
June 2010
-------
costs over those estimated for the current TCR. By comparison, state costs are expected to
increase by 16% and 32% of the TCR costs under the AIP and Alternative options, respectively;
however, state costs remain relatively low in absolute terms. The higher net change in PWS costs
reflects the potential impact on the subsets of PWSs (primarily small TNCWSs) that would be
most affected by the RTCR under either regulatory option. Exhibit 9.13 shows that the greatest
cost impact is on the smallest subset of TNCWSs, for whom the net change is more than double
under the Alternative option as compared to the AIP option. The front loading of costs (shown in
Exhibits 9.8 through 9.10) would further amplify the impacts on small TNCWSs during this
period.
Comparison of Costs of the Regulatory Options Considered
EPA estimated the net change in costs of the AIP and Alternative options as compared to
the baseline (current TCR).
Exhibit 9.12 presents the total and net change in costs to PWSs and states by rule
component for each of the regulatory alternatives considered. For PWSs, the main driver of
additional costs under the alternatives would be additional corrective actions and increases in
routine monitoring. Offsetting the cost increase are reductions in costs for additional routine and
repeat monitoring and PN.
For states, the main drivers of net cost increases are reviews of revised sampling plans,
states are expected to see a reduction in costs for assessment activities and PN, although they
would experience an overall estimated increase in net costs of $0.1 million and $0.3 million,
respectively, for the AIP and Alternative options. The additional costs associated with the
Alternative option would result from the increased sampling (and reporting) frequency under that
option, and the resulting increase in assessments and corrective actions.
Exhibit 9.13 presents the total and net change in costs to PWSs by system size and type.
For PWSs, the most significant total net cost impact is for the smallest category of TNCWSs
(serving <100 people). Approximately 38% of the net cost increase under the AIP option is
estimated to fall on these systems. This is an important factor in evaluation of potential impacts
when comparing regulatory options.
Economic Analysis for the Proposed RTCR
9-20
June 2010
-------
Exhibit 9.8 Net Costs to PWSs (2007$)
n.
o
o
CNJ
o
o
¦O
-------
Exhibit 9.9 Discounted Net Costs to PWSs Over Time (3% Discount Rate) (2007$)
+-»
C
o
o
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Source: Proposed RTCR cost model.
Note: X-axis begins at Year 4, which is the first year of implementation of the AIP or Alternative options. Steady state
begins in approximately Year 9, by which time PWSs that would be expected to meet the criteria would have qualified
for and would begin reduced monitoring.
Economic Analysis for the Proposed RTCR
9-22
June 2010
-------
Exhibit 9.10 Discounted Net Costs to PWSs (7% Discount Rate) (2007$)
Source: Proposed RTCR cost model.
TCR
AIP
Alternative Option
$230.00
$205.00
$180.00
$155.00
$130.00
$105.00
$80.00
$55.00
$30.00
-1 1 1 1 1 1 1 1 1 1 1 i i i i i
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Time (Years)
Note: X-axis begins at Year 4, which is the first year of implementation of the AIP or Alternative options. Steady state
begins in approximately Year 9, by which time PWSs that would be expected to meet the criteria would have qualified
for and would begin reduced monitoring.
Economic Analysis for the Proposed RTCR
9-23
June 2010
-------
Exhibit 9.11 Comparison of Total and Net Change in Annualized Present Value
Costs ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
TCR - Total
$ 185
$ 0.9
$ 186
$ 178
$ 0.9
$ 179
AIP-Total
$ 199
$ 1.1
$ 200
$ 191
$ 1.3
$ 192
AIP - Net Change
$ 14
$ 0.1
$ 14
$ 13
$ 0.4
$ 14
AIP - Percent Change
7%
16%
7%
7%
48%
8%
Alternative Option - Total
$ 212
$ 1.2
$ 213
$ 207
$ 1.5
$ 209
Alternative Option - Net Change
$ 27
$ 0.3
$ 27
$ 29
$ 0.6
$ 30
Alternative Option - Percent Change
15%
32%
15%
16%
67%
17%
Source: Proposed RTCR cost model.
Note:
1) Detail may not add due to independent rounding.
2) Annualized costs are greater using the 7% than 3% discount rate for the Alternative option. This occurs because
the costs under the Alternative option are front-loaded in the 25-year time period due to the all-monthly sampling
requirement (which doesn't occur under the AIP option). Discounting a given stream using 3% and 7% will always
result in a higher present value using 3%; annualization of an identical value (any value) using 3% and 7% would
result in a higher value under 7%. Depending on how costs accrue over the period, and how long the period is, one
effect will be stronger than the other. Generally, the discounting effect is stronger, resulting in the pattern commonly
seen where the 3% annualized amount is greater than the 7% one. However, in cases like the Alternative option
where the costs accrue faster early in the time period, the annualization effect can more than compensate for the
discounting effect, resulting in a higher annualized value under 7% compared to 3%.
Economic Analysis for the Proposed RTCR
9-24
June 2010
-------
Exhibit 9.12 Comparison of Total and Net Change in Annualized Present
Value Costs by Rule Component ($Millions, 2007$)
PWSs
State
Total
PWSs
State
Total
3% Discount Rate
7% Discount Rate
Rule Implementation
TCR- Total
$
$
$
$
$
$
AIP ¦ Total
$
2.77
$
0.18
$
2.95
$
4.00
$
0.26
$
4.26
AIP - Net Change
$
2.77
$
0.18
$
2.95
$
4.00
$
0.26
$
4.26
Alternative Option - Total
$
2.77
$
0.18
$
2.95
$
4.00
$
0.26
$
4.26
Alternative Option - Net Change
$
2.77
$
0.18
$
2.95
$
4.00
$
0.26
$
4.26
Revising Sampling Plans
TCR-Total
$
$
$
$
$
$
AIP ¦ Total
$
0.59
$
0.42
$
1.01
$
0.84
$
0.59
$
1.42
AIP - Net Change
$
0.59
$
0.42
$
1.01
$
0.84
$
0.59
$
1.42
Alternative Option - Total
$
0.59
$
0.42
$
1.01
$
0.84
$
0.59
$
1.42
Alternative Option - Net Change
$
0.59
$
0.42
$
1.01
$
0.84
$
0.59
$
1.42
Routine Monitoring
TCR-Total
$
170.59
$
$
170.59
$
163.94
$
$
163.94
AIP - Total
$
174.71
$
$
174.71
$
167.74
$
$
167.74
AIP - Net Change
$
4.12
$
$
4.12
$
3.80
$
$
3.80
Alternative Option - Total
$
186.34
$
$
186.34
$
181.49
$
$
181.49
Alternative Option - Net Change
$
15.75
$
$
15.75
$
17.56
$
$
17.56
Additional Routine Monitoring
TCR-Total
$
3.87
$
$
3.87
$
3.72
$
$
3.72
AIP - Total
$
1.12
$
$
1.12
$
1.09
$
$
1.09
AIP - Net Change
$
(2.75)
$
$
(2.75)
$
(2.63)
$
$
(2.63)
Alternative Option - Total
$
0.68
$
$
0.68
$
0.58
$
$
0.58
Alternative Option - Net Change
$
(3.18)
$
$
(3.18)
$
(3.14)
$
$
(3.14)
Repeat Monitoring
TCR-Total
$
5.11
$
$
5.11
$
4.91
$
$
4.91
AIP - Total
$
4.82
$
$
4.82
$
4.64
$
$
4.64
AIP - Net Change
$
(0.29)
$
$
(0.29)
$
(0.27)
$
$
(0.27)
Alternative Option - Total
$
5.50
$
$
5.50
$
5.45
$
$
5.45
Alternative Option - Net Change
$
0.39
$
$
0.39
$
0.54
$
$
0.54
Site Ins
pections
TCR-Total
$
$
$
$
$
$
AIP - Total
$
$
$
$
$
$
AIP - Net Change
$
$
$
$
$
$
Alternative Option - Total
$
$
$
$
$
$
Alternative Option - Net Change
$
$
$
$
$
$
Level 1 Assessment
TCR-Total
$
1.13
$
0.21
$
1.34
$
1.08
$
0.21
$
1.29
AIP - Total
$
1.63
$
0.20
$
1.84
$
1.57
$
0.20
$
1.77
AIP - Net Change
$
0.50
$
(0.01)
$
0.49
$
0.49
$
(0.01)
$
0.48
Alternative Option - Total
$
1.73
$
0.23
$
1.95
$ 1.69
$
0.22
$
1.91
Alternative Option - Net Change
$
0.60
$
0.01
$
0.61
$
0.60
$
0.02
$
0.62
Level 2 Assessment
TCR-Total
$
0.70
$
0.26
$
0.96
$
0.68
$
0.25
$
0.93
AIP - Total
$
0.90
$
0.19
$
1.09
$
0.88
$
0.18
$
1.06
AIP - Net Change
$
0.20
$
(0.07)
$
0.12
$
0.20
$
(0.07)
$
0.13
Alternative Option - Total
$
1.23
$
0.28
$
1.51
$
1.27
$
0.30
$
1.57
Alternative Option - Net Change
$
0.52
$
0.02
$
0.55
$
0.60
$
0.05
$
0.65
Corrective Actions based on Level 1 Assessments
TCR-Total
$
$
$
$
$
$
AIP - Total
$
9.17
$
0.01
$
9.18
$
7.77
$
0.01
$
7.77
AIP - Net Change
$
9.17
$
0.01
$
9.18
$
7.77
$
0.01
$
7.77
Alternative Option - Total
$
9.39
$
0.01
$
9.40
$
8.01
$
0.01
$
8.02
Alternative Option - Net Change
$
9.39
$
0.01
$
9.40
$
8.01
$
0.01
$
8.02
Corrective Actions based on Level 2 Assessments
TCR-Total
$
$
$
$
$
$
AIP - Total
$
2.72
$
0.00
$
2.72
$
2.41
$
0.00
$
2.41
AIP - Net Change
$
2.72
$
0.00
$
2.72
$
2.41
$
0.00
$
2.41
Alternative Option - Total
$
3.53
$
0.01
$
3.53
$
3.36
$
0.01
$
3.37
Alternative Option - Net Change
$
3.53
$
0.01
$
3.53
$
3.36
$
0.01
$
3.37
Public Notification
TCR-Total
$
3.75
$
0.44
$
4.19
$
3.60
$
0.42
$
4.03
AIP - Total
$
0.26
$
0.06
$
0.33
$
0.26
$
0.06
$
0.32
AIP - Net Change
$
(3.49)
$
(0.38)
$
(3.87)
$
(3.35)
$
(0.36)
$
(3.71)
Alternative Option - Total
$
0.34
$
0.08
$
0.42
$
0.35
$
0.08
$
0.43
Alternative Option - Net Change
$
(3.41)
$
(0.36)
$
(3.77)
$
(3.26)
$
(0.34)
$
(3.60)
Notes:
1) Detail may not add due to independent rounding.
2) For modeling purposes, additional routine sample counts include
regular routine samples taken in the same month.
3) The "Site Inspections" component has $0 cost because it's not
required by the rule. A PWS may request a site visit by the state as part
of the process of moving towards qualifying for reduced monitoring.
Economic Analysis for the Proposed RTCR
9-25
June 2010
-------
4) Not all TCR components are quantified. For components not
quantified for the TCR, the AIP and Alternative option totals include only
an estimate of the net increase for those same rule components (e.g.,
corrective action costs).
Source: Proposed RTCR cost model.
Exhibit 9.13 Total and Net Change in Annualized Costs to PWSs by PWS Size and
Type ($Millions, 2007$)
PWS Size
(Population
Served)
3%Discount Rate
7% Discount Rate
TCR-Total
AIP - Total
AIP - Net
Alternative
Option - Total
Alternative
Option - Net
TCR-Total
AIP - Total
AIP - Net
Alternative
Option -
Total
Alternative
Option - Net
A
B
C=B-A
D
E=D-A
F
G
H=G-F
I
J=l-F
Community Water Systems (CWSs)
<100
$7.4
$7.5
$0.1
$7.6
$0.2
$7.1
$7.3
$0.2
$7.4
$0.3
101-500
$9.0
$9.3
$0.3
$9.4
$0.4
$8.6
$9.1
$0.4
$9.2
$0.6
501-1,000
$3.7
$3.8
$0.0
$3.8
$0.1
$3.6
$3.7
$0.1
$3.7
$0.1
1,001-4,100
$13.2
$13.6
$0.3
$13.6
$0.3
$12.7
$13.1
$0.4
$13.1
$0.4
4,101-33,000
$42.4
$44.7
$2.3
$44.7
$2.3
$40.7
$42.7
$2.0
$42.7
$2.0
33,001-96,000
$34.9
$36.4
$1.5
$36.4
$1.5
$33.5
$34.8
$1.3
$34.8
$1.3
96,001-500,000
$34.7
$36.2
$1.5
$36.2
$1.5
$33.4
$34.6
$1.2
$34.6
$1.2
500,001-1 Million
$6.5
$6.7
$0.2
$6.7
$0.2
$6.2
$6.4
$0.1
$6.4
$0.1
> 1 Million
$5.6
$5.5
($0.0)
$5.5
($0.0)
$5.3
$5.3
($0.0)
$5.3
($0.0)
Total
$157.4
$163.6
$6.2
$163.9
$6.5
$151.3
$157.0
$5.7
$157.3
$6.0
Nontransient Noncommunity Water Systems (NTNCWSs)
<100
$2.6
$2.6
$0.1
$3.6
$1.0
$2.5
$2.6
$0.2
$3.8
$1.3
101-500
$1.9
$2.0
$0.1
$2.7
$0.9
$1.8
$2.0
$0.2
$2.9
$1.1
501-1,000
$0.6
$0.6
$0.0
$0.8
$0.3
$0.6
$0.6
$0.1
$0.9
$0.3
1,001-4,100
$1.2
$1.3
$0.1
$1.3
$0.1
$1.1
$1.2
$0.1
$1.2
$0.1
4,101-33,000
$0.4
$0.5
$0.1
$0.5
$0.1
$0.4
$0.5
$0.0
$0.5
$0.0
33,001-96,000
$0.1
$0.1
$0.0
$0.1
$0.0
$0.1
$0.1
$0.0
$0.1
$0.0
96,001-500,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
500,001-1 Million
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
> 1 Million
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
$0.0
Total
$6.9
$7.2
$0.4
$9.1
$2.3
$6.6
$7.2
$0.6
$9.4
$2.8
Transient Noncommunity Water Systems (TNCWSs)
<100
$13.4
$18.5
$5.1
$26.7
$13.4
$12.8
$18.0
$5.1
$27.8
$14.9
101-500
$4.9
$6.4
$1.5
$9.1
$4.2
$4.7
$6.2
$1.5
$9.4
$4.7
501-1,000
$0.6
$0.8
$0.2
$1.1
$0.5
$0.6
$0.8
$0.2
$1.2
$0.5
1,001-4,100
$0.9
$1.0
$0.1
$1.0
$0.1
$0.9
$1.0
$0.1
$1.0
$0.1
4,101-33,000
$0.4
$0.5
$0.1
$0.5
$0.1
$0.4
$0.5
$0.0
$0.5
$0.0
33,001-96,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
96,001-500,000
$0.1
$0.1
($0.0)
$0.1
($0.0)
$0.1
$0.1
($0.0)
$0.1
($0.0)
500,001-1 Million
$0.2
$0.2
($0.0)
$0.2
($0.0)
$0.2
$0.2
($0.0)
$0.2
($0.0)
> 1 Million
$0.3
$0.3
$0.0
$0.3
$0.0
$0.3
$0.3
$0.0
$0.3
$0.0
Total
$20.9
$27.8
$6.9
$39.1
$18.2
$20.1
$27.1
$7.0
$40.4
$20.3
Grand Total
$185.2
$198.7
$13.5
$212.1
$26.9
$177.9
$191.2
$13.2
$207.0
$29.1
Note: Detail may not add due to independent rounding. Because only the incremental costs of some rule components are considered as part of the cost analysis,
references to "total" costs in this exhibit do not refer to the complete costs for regulatory implementation, but only to the specific costs considered to calculate net
changes in costs.
Source: Proposed RTCR cost model.
Economic Analysis for the Proposed RTCR
9-26
June 2010
-------
9.1.3 Comparison of National Benefits and Costs of the Regulatory Options Considered
Because the potential benefits of the AIP and Alternative options have not been
monetized, direct questions of how benefits compare to costs cannot be quantitatively answered
(i.e., whether monetary benefits exceed costs and by how much). However, EPA's collective
evaluation of the results presented in this EA lead to the conclusion that the benefits of the AIP
option will, in fact, exceed costs. Further evaluation of the cost effectiveness of the AIP option in
comparison to the Alternative option considered is provided in Section 9.3 below. The break-
even analysis described in Section 9.3.2 entails determining how many cases of morbidity and/or
mortality would need to be avoided in order for the benefit of those avoided cases to equal the
cost of the rule option. It does not predict cases of illness avoided as a result of the rule. Instead,
it is included to help provide some context for considering the magnitude of cases of waterborne
illness that would need to be avoided to breakeven with rule costs.
9.2 Effect of Uncertainties and Nonqualified Benefit/Cost Estimates on the
Estimation of National Benefits and Costs
Throughout the development of the analyses in this EA, EPA closely evaluated
uncertainties in data used. EPA presents a discussion and comprehensive list of assumptions that
introduce uncertainty into the analysis in Sections 4.5 and 5.3.3.1 (Exhibits 4.13 and 5.22a-b,
respectively). Certain assumptions are incorporated into the predictive model for all three
regulatory scenarios in a similar fashion (current TCR as well as the AIP and Alternative
options). Because the analysis is based on net changes from the current TCR to the AIP and
Alternative options, incorporating these uncertainties into the current TCR and the options
results in no significant net effect on the analysis results. As discussed in Chapters 4 and 5,
although GWR assumptions are considered by EPA to be conservative estimates of the
effectiveness of the GWR, they are incorporated into the model in similar fashion for all three
regulatory options considered, and therefore are not expected to significantly affect net results of
the EA. However, assumptions for certain variables identified as key drivers in producing the net
results of this EA (the frequency and effectiveness of CA's) affect the AIP and Alternative
options but not the current TCR. These assumptions are further evaluated in a sensitivity analysis
in Ch. 5, which is summarized in Section 9.2.1 below.
EPA also assessed nonquantified costs and benefits to estimate how they would affect the
overall conclusions of the analyses performed. Potential impacts of nonquantified costs and
benefits are discussed in Chapter 7 and Chapter 6, respectively, of the EA. Section 9.2.2
summarizes the major nonquantified costs and benefits evaluated.
9.2.1 Summary of Major Uncertainties in EA Analyses
The primary sources of uncertainty in the AIP option are introduced into the analysis in
the occurrence model and cost models. In the occurrence model, uncertainty is introduced in two
types of information. First, data on TC/ E. coli occurrence and acute and non-acute violations
data are adapted from the Six-Year Review and Safe Drinking Water Information System-
Federal Version (SDWIS/FED). Second, EPA used best professional judgment in choosing the
Economic Analysis for the Proposed RTCR
9-27
June 2010
-------
values of model inputs to represent the net increase in effective corrective actions to be
implemented under the AIP option, and the resulting period of reduced occurrence (both in
length of time and percent reduction). The data limitations are discussed in Chapter 4 of this EA.
The influence of selected inputs on the model output was tested in a sensitivity analysis, using
substitute values that were factors of 0.5 and 2.0 of the original values. Model runs based on the
0.5 decrease in original inputs yielded results that decreased by a factor less than 0.5 (i.e., the
results were between 0.5 and 1.0 of the original value). Similarly, model runs based on inputs
that were 2X the value of the original inputs yielded results that were less than a multiple of 2X
the original output. These results, showing that input variable selection has less than a one-to-one
influence on model output, indicate that the model output is not highly sensitive to changes in
these input variables.
Additionally, uncertainty in the analysis results stemmed from the complexity of
interactions between the effects of the various rule components. The occurrence model produced
outputs that showed a decrease in acute violations under the AIP option and an increase under
the Alternative option. Given that the Alternative option incorporated the same risk reducing
features (assessments and corrective actions) as the AIP option, EPA considered that the
additional sampling under the Alternative option was having a larger upwards effect (increase)
on occurrence compared to the downward effect of the additional corrective actions. EPA
developed a stepwise sensitivity analysis to test the relative importance of the following
competing effects under the AIP option relative to the current TCR:
• New assessments and corresponding corrective actions
• Reduced repeat and additional routine monitoring.
The results (Section 6.4) revealed that acute violations under the AIP option as compared
to the current TCR differed as follows:
• Many acute violations were prevented from occurring by implementation of
assessments and corresponding corrective actions.
• Some acute violations were missed by the reduction in repeat and additional routine
sampling.
From these results (described in Section 6.4), EPA concluded that under the AIP option and the
Alternative option, more acute violations are prevented than are missed. Additionally, the
increase in routine monitoring under the AIP option finds additional acute violations as
compared to the current TCR, which explains the net increase in the acute violations found as
compared to the current TCR. Exhibits 9.2 - 9.7 show the larger net difference (between the
alternative and baseline) seen in levels of acute violations, non-acute violations, and corrective
actions implemented under the AIP option as compared to the Alternative option.
The sources of uncertainty embedded within the results of the occurrence model are also
incorporated into the cost model. As described in reference to the occurrence model, EPA used
best professional judgment in choosing 10% as the factor of net increase over corrective actions
Economic Analysis for the Proposed RTCR
9-28
June 2010
-------
implemented under the current TCR. A sensitivity analysis of the effect of an increase in this
factor on costs revealed that as the number of corrective actions increase, a portion of the
corresponding cost is offset by the reduced number of assessments required as water quality
improves (Section 7.7 of this EA).
A potentially significant additional source of uncertainty in costs is introduced in the
distribution of correction actions selected (i.e., the number of each type implemented) for the
10% additional corrective actions. Because there is a wide variation in the unit costs of the actual
corrective actions underlying the compliance forecast (see Appendix D for unit cost detail), EPA
ran two sensitivity analyses to assess the potential impacts of a high and low cost distribution in
the compliance forecast. The results provide a low bound and a high bound on the original
analysis results: approximately $lmillion/$19 million for the AIP option and $16 million/$39
million for the Alternative option. This indicates that although the estimates of the number of
systems that would choose each type of corrective action has a significant impact on costs of the
rule, the rule remains a relatively low cost rule under the most conservative (costly) assumptions.
9.2.2 Summary of Nonqualified Costs and Benefits
EPA believes that all of the rule elements that are the major drivers of the net change in
costs from the current TCR have been quantified to the greatest degree possible. However, cost
reductions related to fewer monitoring and reporting violations are not specifically accounted for
in the cost analysis, and their exclusion from consideration may result in an overestimate of net
change in cost between the TCR and the AIP option or Alternative option. Using E. coli as the
endpoint, benefits have been estimated and predictions made about the reduced hazard associated
with reduced exposure to fecal contamination and the potential benefits from corrective actions
that, if effective, will reduce exposure. No attempt was made to monetize these benefits by
predicting infections, illnesses or deaths and to compare these values directly with the monetized
costs because of lack of data.
Because the cost analyses in this EA focus on net changes in costs, consideration of
nonquantified costs can result in either an under- or over-estimate of the net costs for the AIP
option, as described in Section 7.7 of this EA. Overall, the nonquantified costs identified would
potentially lower the net cost of the AIP option, and include the following:
• Reduced costs due to overlap between sanitary surveys and required assessments; and
• Reductions in costs related to less frequent PN for reporting and monitoring
violations.
Additionally, EPA identified a potential cost and benefit increase for the AIP option and
Alternative option related to the baseline assumption that all PWSs are currently meeting
requirements of the TCR. To the degree that PWSs are not currently meeting TCR requirements,
the current TCR costs and benefits are overstated in the RTCR EA, resulting in an underestimate
of costs and benefits under the AIP option.
Economic Analysis for the Proposed RTCR
9-29
June 2010
-------
EPA identified five additional potential areas of nonquantified benefits:
• Reductions in primary and secondary endemic disease burden, both acute and
chronic, associated with consuming PWS drinking water;
• Reductions in risk of outbreak and associated illness and costs;
• Benefits from increased systems knowledge (including more consistent water
quality);
• Accelerated infrastructure repair/replacement; and
• Reductions in averting behavior (due to higher consumer confidence in water quality).
Taken together, these items represent a potential for significant additional health benefits. For the
first four items listed above, benefits may be realized directly through the elimination, reduction,
and/or prevention of pathogen occurrence. Reductions in averting behavior may lead to benefits
related to a more efficient use of resources (i.e., using tap water instead of more expensive, yet
not necessarily safer, alternatives).
9.3 Comparison of the Regulatory Options Considered
The following subsections present further analysis of the AIP option in comparison to the
Alternative option. Comparisons based on incremental costs and benefits and several
effectiveness measures are presented to inform the selection of the AIP option over the
Alternative option,
9.3.1 Incremental Costs and Benefits
Incremental costs and benefits are those that are incurred or realized to reduce pathogen
illnesses and deaths from one alternative to the next more stringent alternative. Estimates of
incremental costs and benefits are useful when considering the economic efficiency of different
regulatory alternatives considered by EPA. One goal of an incremental analysis is to identify the
regulatory alternatives where net social benefits are maximized. Incremental net benefits analysis
is not possible when benefits are not monetized as in the case with the RTCR. However,
incremental analysis can still provide information on relative cost-effectiveness of different
regulatory options.
For the RTCR only costs were monetized. An indirect proxy for benefits was quantified.
To compare the additional net cost increases and associated incremental benefits of the AIP and
Alternative options, benefits are presented in terms of corrective actions performed since
performance of corrective actions is expected to have an impact that is most directly translatable
into potential health benefits.
Economic Analysis for the Proposed RTCR
9-30
June 2010
-------
To compare the additional net cost increases and associated benefits of the AIP and
Alternative options (shown as numbers of corrective actions performed), analyses of both
incremental costs and benefits are presented in Exhibit 9.14. The results represent that portion of
costs for each rule alternative that are an addition to the cost of the next least stringent rule
option. Benefits are also incrementally presented in terms of corrective actions performed, since
performance of corrective actions is expected to have an impact that is most directly translatable
into potential health benefits.
Exhibit 9.14 shows that the Alternative option is incrementally approximately the same
annualized cost as the AIP option using a 3 discount rate ($13.5 million vs. $13.7 million,
respectively). Consistent with the lower incremental costs, the incremental benefits in terms of
corrective actions performed are fewer under the Alternative option than under the AIP. The new
requirement under both the AIP and Alternative options for PWSs to implement assessments and
corrective actions will capture the bulk of improvements to be made in comparison to the current
TCR option. The additional sampling required under the Alternative option relative to the AIP
option is providing an approximately equal incremental increase in costs but a lower incremental
effectiveness relative to the portion of costs that occurs under both the AIP and Alternative
options. This suggests that beyond a certain level of corrective action implementation, PWSs
would experience diminishing net marginal returns for their investment in correction actions.
The relationship between the incremental costs and benefits is examined further with respect to
cost effectiveness in Section 9.3.2 below.
Exhibit 9.14 Incremental Net Change in Annualized Present Value Costs
($Millions, 2007$) and Benefits (Number of Corrective Actions)
Regulatory Option
Costs
Benefits
(L2 Corrective Actions)
3%
7%
3%
7%
Current TCR
$ 186.1
$ 178.8
No change3
No change3
AIP
$ 199.8
$ 192.5
202
189
Incremental AIP1
$ 13.7
$ 13.7
202
189
Alternative Option
$ 213.3
$ 208.5
326
331
Incremental Alternative Option2
$ 13.5
$ 16.0
124
143
1 Represents the incremental net change of the AIP option over the current TCR option.
2 Represents the incremental net change of the Alternative option over the AIP option. Add incremental
net change for Alternative option to incremental net change for AIP option to calculate the total net
change of the Alternative option over the current TCR option. Note: The RTCR occurrence model yields
the number of corrective actions that are expected to be implemented in addition to (net of) those already
implemented under the current TCR. The model does not incorporate an estimate of the number of
corrective actions implemented per year under the current TCR and does not yield a total for the AIP and
Alternative option that includes the current TCR corrective actions. Benefits shown include corrective
actions based on L2 assessments. Detailed benefits and cost information is provided in Appendices A
and C, respectively, of the Proposed RTCR EA (USEPA 2009a).
3 As explained in section VI.F.2.f of the preamble, for modeling purposes, EPA estimates the net change
only in the number of corrective actions performed under the AIP and Alternative options compared to the
current TCR and thus did not quantify the (non-zero) baseline number of corrective actions performed
under the Current TCR.
Economic Analysis for the Proposed RTCR
9-31
June 2010
-------
9.3.2 Cost Effectiveness Measures
To further differentiate between the AIP option and Alternative option results, EPA
performed analyses to assess the cost effectiveness of each regulatory scenario. In this section,
EPA provides an estimate of the cost per corrective action performed under the regulatory
alternatives, which is a measure of the cost effectiveness of each proposed rule option. As noted
elsewhere in this chapter, corrective actions are considered a clear enhancement to the efforts to
reduce E. coli (and co-occurring pathogen) occurrence in PWSs, and the most direct benefit in
this EA. Corrective actions are also a main driver of the net increases in costs for both rule
options considered. In addition, the modeling of corrective actions is done only on a net basis
relative to the level performed under the current TCR, thus providing a clear dividing line
between impacts under the current TCR and each rule option. The results (Exhibit 9.15) of this
EA show that, on average, the annual net cost of the rule per corrective action performed beyond
baseline is approximately $23,000 and $35,000, respectively, for the AIP and Alternative options
(using a 3 percent discount rate).
Exhibit 9.15 Total Net Annual Cost Per Corrective Action Implemented under AIP
Option and Alternative Options, Annualized Using 3% and 7% Discount Rates
($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Net Cost
$ 13.7
$ 13.7
AIP Net Corrective Actions (L1 & L2)
598
555
AIP Cost Effectiveness Analysis (CEA )(net rule cost/CA)
$ 22,902
$ 24,613
Alternative Option Net Cost
$ 27.2
$ 29.7
Alternative Option Net Corrective Actions (L1 & L2)
785
765
Alternative Option CEA (net rule cost/CA)
$ 34,721
$ 38,815
Note: Corrective actions include those conducted as a result either Level 1 or Level 2 assessments. Total rule costs
are shown in Exhibit 9.14. Detailed benefits and cost information is provided in Appendices A and C, respectively, of
this EA.
EPA also considered the incremental cost effectiveness of the AIP as compared to the
Alternative option to determine the additional benefit associated with that portion of Alternative
option cost beyond the cost of the AIP option. Exhibit 9.16 shows that the AIP option has a far
lower marginal cost per corrective action than the Alternative option ($23,000 for the AIP option
vs. $73,000 for the Alternative option). This indicates that the additional corrective actions
performed under the Alternative option are much more costly on a unit basis than those
performed under the AIP option.
EPA further considered the issue of additional costs under the Alternative option as
compared to the AIP option in terms of the effect on small TNCWSs using GW (serving <100).
The additional net cost increases under the Alternative option could cause substantial additional
burden on small TNCWSs, which incur the largest portion of costs of the rule among all
categories of systems. However, these same TNCWSs would also receive the most benefits from
the RTCR (increased assessments and corrective actions following a trigger). EPA considered
the group of 60,200 TNCWSs using GW (serving <100), which are the largest subset of systems
by size/type, and which are expected to bear the highest burden of all system categories under
Economic Analysis for the Proposed RTCR
9-32
June 2010
-------
the AIP option. Exhibit 9.17 shows that in incremental terms, the cost effectiveness of the AIP
option is significantly greater than of the Alternative option (a cost per corrective action of
$18,000 vs. $65,000, respectively) and both options are more cost effective for this subset
(TNCWSs using GW serving <100) than for the entire population of PWSs taken as a whole.
The two incremental analyses (Exhibits 9.16 and 9.17) together indicate that, using a 3 percent
discount rate to compare incremental benefits and costs, the AIP option is significantly more cost
effective (by a factor of approximately 3.5) than the Alternative option for the most burdened
subset of systems, and is significantly more cost effective (by a factor of approximately 3.2) than
the Alternative option when considering all PWSs together.
Exhibit 9.16 Incremental Rule Cost Per Corrective Action (CA) Implemented for
All PWSs under AIP and Alternative Options, Annualized Using 3% and 7%
Discount Rates ($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Incremental Net Costs ($ millions)1
$ 13.7
$ 13.7
AIP Incremental Net Corrective Actions (L1 & L2)1
598
555
AIP Incremental Cost per CA($)
$ 22,902
$ 24,613
Alternative Option Incremental Net Costs ($ millions)2
$ 13.5
$ 16.0
Alternative Option Incremental Net Corrective Actions (L1 & L2)2
187
210
Alternative Option Incremental Cost per CA ($)
$ 72,582
$ 76,299
Note: Exhibit includes only the number of corrective actions predicted by the RTCR occurrence model to be
implemented in addition to those implemented under the current TCR. Includes both L1 & L2 corrective actions. Total
net costs for each option and total CAs (not incremental) are shown in Exhibit 9.15. Detailed benefits and cost
information is provided in Appendices A and C, respectively, of this EA.
1 Represents the incremental increase of the AIP option over the current TCR.
2 Represents the incremental increase of the Alternative option over AIP option. Add incremental net values for
Alternative option to incremental net values for AIP option to calculate total net values of Alternative option over
current TCR.
Exhibit 9.17 Incremental Rule Cost Per CA Implemented for GW TNCWSs under
AIP and Alternative Options, Annualized Using 3% and 7% Discount Rates ($2007)
Regulatory Scenario
3% Discount Rate
7% Discount Rate
AIP Incremental Net Costs ($ millions)1
$ 5.1
$ 5.1
AIP Incremental Corrective Actions (L1 & L2) (TNCWS <101 only)1
279
257
AIP Incremental Cost per CA($)
$ 18,219
$ 19,965
Alternative Option Incremental Net Costs ($ millions)2
CO
CO
$ 9.8
Alternative Option Incremental Corrective Actions (L1 & L2) (TNCWS <101 only)2
128
145
Alternative Option Incremental Cost per CA ($)
$ 64,731
$ 67,762
Note: Incremental Net Costs are based on TNCWS serving <101 people. Detailed benefits and cost information is
provided in Appendices A and C, respectively, of this EA.
1 Represents the incremental increase of the AIP option over the current TCR.
2 Represents the incremental increase of the Alternative option over AIP option. Add incremental net values for
Alternative option to incremental net values for AIP option to calculate total net values of Alternative option over
current TCR.
Economic Analysis for the Proposed RTCR
9-33
June 2010
-------
9.3.3 Break-Even Analysis
As described in Chapter 6 of this EA, this analysis does not include a fully quantified risk
assessment because there are insufficient data on the co-occurrence in PWSs of fecal indicator E.
coli and pathogenic strains of E. coli or other bacteria, viruses, and parasitic protozoa that can
cause waterborne disease. Given the absence of E. coli and waterborne pathogen co-occurrence
data to predict health effects in PWSs, EPA developed a break-even analysis to inform the
discussion of whether the benefits justify the costs of the regulation. A break-even analysis
typically compares the rule cost with the estimated value of the medical cases (and related costs
such as time off from work) that would be avoided annually by the rule to determine if the rule
benefits "break even" with avoided costs. In the case of the EA, where cases avoided cannot be
estimated, the break-even analysis informs the benefit/cost discussion by providing estimates of
the number of fatal and non-fatal cases that the rule would need to avoid in order to break even
with rule costs.
The underlying question this analysis seeks to inform is whether corrective actions
performed under the AIP option will sufficiently reduce potential exposure to fecal
contamination and/or waterborne pathogens to compensate for the net costs of the rule. One
result of the implementation of the RTCR predicted by the model and presented in Exhibit 6.5 is
that some number of acute events will be prevented.6 As discussed in Chapter 6, sufficient data
are not available to quantify the numbers of illnesses and deaths potentially avoided by
preventing such acute events. However, the presence of E. coli in a water system is an indicator
of the presence of fecal contamination;7 it suggests that a pathway for the introduction of
waterborne pathogens may exist. The total numbers of acute events and potentially affected
people are small (most events are predicted in small NCWSs), but nonetheless there will be a
positive impact on public health by avoiding these events. The magnitude of this impact will
depend heavily on the specific elements underlying any acute event avoided: who and how
many in the population are affected, and the etiologic agent to which exposure is avoided. An
event can be endemic to a single person or small group of persons, or it may reach the level of an
outbreak. The cause of an event may be one or more of a range of bacterial, viral, or parasitic
organisms, as described in Chapter 2 of this EA. Although there may be a range of waterborne
pathogens mitigated by implementation of the proposed RTCR, two potential waterborne
pathogens are considered in this analysis to provide input for developing the break even analysis:
shiga toxin-producingE. coli 0157:H7 (STEC 0157)8 and Salmonella9
6Results presented in Exhibit 6.5 of this EA indicate that fewer EC positive events would occur under the AIP option
than under the current TCR. These results are consistent with the net improvement in PWS water quality (as
indicated by a decrease in fecal indicator EC) suggested by the step-wise analysis presented in Section 6.4 of this
EA. Chapter 6 of the EA presents this information in further detail.
7 Standridge, J. (2008) discusses the use of E.coli as an indicator of drinking water quality.
8 According to the website of the American Academy of Family Physicians
(http://www.aafp.org/afp/20000401/tips/ll.html). "Shiga toxin-producing Escherichia coli is a group of bacteria
strains capable of causing significant human disease. The pathogen is transmitted primarily by food and has become
an important pathogen in industrialized North America. The subgroup enterohemorrhagic E. coli includes the
relatively important serotype 0157:H7, and more than 100 other non-0157 strains."
9 Salmonella, according to the CDC website, is actually a group of bacteria that can cause diarrheal illness in
humans. They are microscopic living creatures that pass from the feces of people or animals to other people or other
animals. Salmonella serotype Typhimurium. which caused the 2008 Alamosa, CO outbreak, and Salmonella
serotype Enteritidis are the most common in the United States. Every year, approximately 40,000 cases of
Economic Analysis for the Proposed RTCR
9-34
June 2010
-------
The Centers for Disease Control and Prevention (CDC) estimates that STEC 0157
pathogens from all sources are responsible for approximately 73,000 illnesses in the United
States annually. The CDC has found that 15% of outbreak cases of STEC 0157 are waterborne
(Rangel et al., 2005); if that rate applies as well to endemic cases, then approximately 11,000
cases of STEC 0157 infection would be due to waterborne exposure. For Salmonella infections,
the Economic Research Service (ERS) uses an estimate of 1,397,187 cases annually based on
1999 estimates from the CDC that included an estimate of non-reported illnesses. The 1999
estimate stated that the share of cases due to consumption of contaminated food was 95 percent,
indicating that approximately 70,000 cases were due to other sources. While the portion of these
cases due to waterborne exposure is unknown, some public systems have documented cases of
waterborne Salmonellosis. In March of 2008, Salmonella in the water system of Alamosa,
Colorado caused an outbreak in which hundreds of the 10,000 residents were sickened and one
person died. The deep mountain aquifer from which the water was drawn was confirmed not to
be the source of the bacteria10, but two storage tanks had several inches of sediment and
breaches, and one tank had breaches large enough for birds and animals to enter (CDPHE, 2009).
For the break-even analysis, medical cases are classified as non-fatal or fatal, and
analyses are presented for these two classifications.11 The valuations for medical cases used in
the break-even analysis are developed using the ERS Foodborne Illness Calculator12 (ERS
calculator), as described in Appendix E of the RTCR, for the two example pathogens considered:
STEC 0157 and Salmonella. For most ERS calculator inputs, default or suggested alternative
values are used to estimate costs. However, for the valuation of fatal cases within the model,
EPA used a Value of a Statistical Life (VSL)13 estimate of $8.8 million14 to be consistent with
EPA methodology for calculating the value of this measure.15
salmonellosis are reported in the United States; most of these are foodborne, but some, as in the case of Alamosa,
are waterborne. (http://www.cdc.gov/nczved/dfbmd/disease listing/salmonellosis gi.html#2. downloaded 10-6-09)
10 Berg, Rebecca. 2008. The Alamosa Salmonella Outbreak: A Gumshoe Investigation. Journal of Environmental
Health. 71(2):54-55.
11 Evaluation of fatal cases in this analysis includes consideration of the underlying non-fatal component of the
illness prior to death.
12 A full detailed description of the ERS Foodborne Illness Calculator, including detailed discussion of the items
discussed above and derivation of the underlying values used in calculations can be found online at:
http://www.ers.usda.gov/Data/FoodborneIllness/ecoli_Intro.asp.
13 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 by 1/1,000,000 for one 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 risk reduction anticipated from the regulation, the VSL would be $5 million ($5 X
1,000,000). Appendix E of this EA provides further detail of the derivation of the updated VSL estimate.
14 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 2000c) and endorsed by the Science Advisory Board (SAB) Arsenic review
panel (USEPA 2001b). For use in the break-even analysis, adjustments to the base value from the EPA study were
made to account for income and price level changes over time. The mean VSL, after all adjustments were made, was
$8.8 million in year 2007 dollars. These adjustments are explained further in Appendix E of this EA.
15 Use of different VSL estimates corresponding to different government agency policy and methodologies is
specifically suggested by ERS, and EPA's methodology is given as a specific example of an alternate value that may
be used.
Economic Analysis for the Proposed RTCR
9-35
June 2010
-------
To represent a range of estimates, EPA used the ERS calculator to calculate break-even
thresholds for a low- and high-end valuation of cases avoided corresponding to "traditional" and
"enhanced" methodologies for estimating Cost of Illness (COI) values. For the traditional COI
valuation, EPA used the ERS calculator default values for STEC 0157, with the exception of the
VSL applied to fatalities, which EPA updated as shown in Appendix E of the RTCR EA. To
estimate an enhanced COI valuation, EPA applied the basic settings used for the traditional COI
valuation (including the EPA-updated VSL value) but with alternative values for estimated
productivity and disutility losses. The traditional COI, as developed in this analysis based on the
default setting in the ERS calculator, places no value on lost nonmarket work time, while the
enhanced COI method values nonmarket work time based on opportunity costs (equivalent to
market work time lost). The alternative valuation used (a suggested option within the calculator)
is to apply the productivity losses to all lost time for market and non-market work (e.g., for
individuals who stay home to raise children or care for the elderly). The ERS calculator default
value is zero for other forms of disutility caused by the illness16, including pain and suffering,
inconvenience, and lost non-work time; this setting is used for the traditional approach in this
analysis. The alternative valuation, used for the enhanced approach in this analysis (also a
suggested option within the calculator), is based on values (greater than zero) used by the Food
and Drug Administration.
Exhibits 9.18 and 9.19 show the values calculated for avoided cases using the ERS
Foodborne Illness Calculator for STEC 0157 and Salmonella, as summarized above and
described in Appendix E of this EA. In addition to separate calculations based on traditional and
enhanced COI valuation methodologies, estimates are presented separately for non-fatal and fatal
cases.
Exhibit 9.18 Average Estimated Value Per STEC 0157 Case Avoided (2007$)
Case Description
Value Per Case
Traditional COI
Enhanced COI
Non-fatal cases only
$ 1,756
$ 5,269
Fatal cases only1
$ 8,820,401
$ 8,821,911
illness prior to death.
Notes: Estimates for a weighted average estimate of the cost of "all cases"
that includes both fatal and non-fatal cases is shown in Appendix E. COI
estimates derived using USDA ERS Foodborne Illness Cost Calculator:
STEC 0157. Traditional COI estimates are based on ERS default values for
all inputs except for the VSL estimate, for which EPA substituted its own
estimate of approximately $8.8 million. Enhanced COI estimates are based
on the same values except for the following changes: productivity losses are
applied to all cases rather than just those employed in the formal market, and
FDA disutility values are applied rather than the ERS default value of 0.
16 Disutility is a specific category defined in the ERS calculator, as mentioned above, to include: pain and suffering,
inconvenience, and lost non-work time.
Economic Analysis for the Proposed RTCR
9-36
June 2010
-------
Exhibit 9.19 Average Estimated Value Per Salmonella Case Avoided (2007$)
Case Description
Value Per Case
Traditional COI
Enhanced COI
Non-fatal cases only
$ 210
$ 1,397
Fatal cases only1
$ 8,802,998
$ 8,803,551
underlying illness prior to death.
Notes: Estimates for a weighted average estimate of the cost of "all
cases" that includes both fatal and non-fatal cases is shown in
Appendix E. COI estimates derived using USDA ERS Foodborne
Illness Cost Calculator: Salmonella. Traditional COI estimates are
based on ERS default values for all inputs except for the VSL
estimate, for which EPA substituted its own estimate of approximately
$8.8 million. Enhanced COI estimates are based on the same values
except for the following changes: productivity losses are applied to all
cases rather than just those employed in the formal market, and FDA
disutility values are applied rather than the ERS default value of 0.
Using the valuations per case described above, EPA estimated the numbers of cases
needed to break even by dividing the total net cost of the AIP option by the values from Exhibits
9.18 and 9.19. The resulting estimates of break-even thresholds are presented in Exhibit 9.20 and
9.21 separately for non-fatal and fatal cases for STEC 0157 and Salmonella, respectively.
Estimates are presented using both three and seven percent discount rates.
Exhibit 9.20 Estimated Annual Break-Even Threshold for Avoided Cases of
STEC 0157
COI
Discount
AIP Option
Alternative Option
Methodology
Rate
Non-fatal
Fatal cases
Non-fatal
Fatal cases
cases only
o
3
<
cases only
o
3
<
i—4
Traditional
3%
8,000
1.6
16,000
3.1
COI
7%
8,000
1.5
17,000
3.4
Enhanced COI
3%
3,000
1.6
5,000
3.1
7%
3,000
1.5
6,000
3.4
Calculations for fatal cases include the non-fatal COI component for the underlying
illness prior to death.
Notes: The number of cases needed to reach break-even threshold calculated by dividing
the net change in costs for the RTCR (Exhibit 9.5) by the average estimated value of
avoided cases (Exhibit 9.18). Threshold estimates based on a weighted average of the
cost of "all cases" that includes both fatal and non-fatal cases is shown in Appendix E.
STEC 0157 is only an example of a pathogenic endpoint that could have been used for
this analysis. Use of additional pathogenic contaminants in addition to this single endpoint
would result in lower threshold values. Detail may not add due to independent rounding.
The break-even threshold is higher using a 7% discount rate than a 3% discount rate
under the Alternative option. This result is consistent with the annualized costs of the
Alternative option being higher using the 7% discount rate, which is caused by the
frontloading of costs in the period of analysis, as explained further in Chapter 7 of the EA.
Economic Analysis for the Proposed RTCR
9-37
June 2010
-------
Exhibit 9.21 Estimated Annual Break-Even Threshold for Avoided Cases of
Salmonella
COI
Discount
AIP Option
Alternative Option
Methodology
Rate
Non-fatal
Fatal cases
Non-fatal
Fatal cases
cases only
o
3
<
cases only
o
<
i—4
Traditional
3%
65,000
1.6
130,000
3.1
COI
7%
65,000
1.6
141,000
3.4
Enhanced COI
3%
10,000
1.6
20,000
3.1
7%
10,000
1.6
21,000
3.4
Note: Calculations for fatal cases include the non-fatal COI component for the
underlying illness prior to death.
Note: The number of cases needed to reach break-even threshold calculated by
dividing the net change in costs for the RTCR (Exhibit 9.5) by the average estimated
value of avoided cases (Exhibit 9.19). Threshold estimates based on a weighted
average of the cost of "all cases" that includes both fatal and non-fatal cases is shown
in Appendix E. Salmonella is only an example of a pathogenic endpointthat could have
been used for this analysis. Use of additional pathogenic contaminants in addition to
this single endpoint would result in lower threshold values. Detail may not add due to
independent rounding.
The break-even threshold is higher using a 7% discount rate than a 3% discount rate
under the Alternative option. This result is consistent with the annualized costs of the
Alternative option being higher using the 7% discount rate, which is caused by the
frontloading of costs in the period of analysis, as explained further in Chapter 7 of the
EA.
The estimates shown in Exhibits 9.20 and 9.21 for STEC 0157 and Salmonella,
respectively, represent the number of fatal cases that would need to be avoided annually by the
rule for its benefits and costs to society to be equal, or break even, with rule costs. Based on
either example pathogen considered in this break-even analysis, the results show that both the
AIP and Alternative options would need to avoid a relatively small number of fatal cases
annually.
Under the AIP option, just two deaths would need to be avoided annually using a 3%
discount rate based on consideration of the bacterial pathogen STEC 0157. Alternatively,
approximately 3,000 or 8,000 non-fatal cases, using the enhanced or traditional benefits
valuations approaches17,, respectively, would need to be avoided to break even with rule costs.
As expected based on its costs, the Alternative option would require that a higher number of
17 Both traditional and enhanced COI approaches count the value of the direct medical costs and of time lost that
would been spent working for a wage, but differ in their assessment of the value of time lost that would be spent in
nomnarket work (e.g. housework, yardwork, and raising children) and leisure (e.g. recreation, family time, and
sleep). They also differ in their valuation of (other) disutility, which encompasses a range of factors of well being,
including both inconvenience and any pain and suffering. A complete discussion of the traditional and enhanced
COI approaches can be found in Appendix E.
Economic Analysis for the Proposed RTCR
9-38
June 2010
-------
cases be avoided annually for that option to break even (approximately 60% to 100% more than
the AIP option under the traditional and enhanced approaches, respectively).
As Exhibit 9.21 shows, approximately 2 deaths would need to be avoided from a
Salmonella infection for the rule to break even. The estimated number of non-fatal Salmonella
cases that would need to be avoided to break even is approximately 10,000 or 65,000 cases under
the enhanced and traditional benefits valuations approaches, respectively.
As the discussion presented in Chapter 2 of this EA describes, disease and deaths are
attributable to drinking water contamination across the country. The CDC has found that 15% of
outbreak cases of STEC 0157:H7 are waterborne (Rangel et al., 2005); if that rate applies as well
to endemic cases, then approximately 11,000 cases would be due to waterborne exposure to
STEC 0157:H7. The revised rule could mitigate pathways of contamination into PWSs and
prevent some of these endemic cases. For just this one contaminant, if just 3,000 non-fatal cases
under the enhanced COI approach, 8,000 cases under the traditional COI approach, or
approximately 2 fatal cases of STEC 0157:H7 are avoided by the proposed RTCR, the rule
would break even. If more than one contaminant was reduced or prevented from occurring in
PWSs, the rule would be that much more likely to break even. Avoided illness and death from
secondary transmission of infection could also be significant, and would increase the likelihood
that total avoided cases reach a break-even threshold. An additional consideration is that a larger
number of avoided cases would be predicted if all the benefits of the rule as described in Chapter
6 could be quantified in the analysis. If increased assessments and corrective actions result in a
level of system knowledge that enables earlier mitigation of potential pathways of
contamination, then additional illnesses or deaths may be avoided beyond those suggested by the
decrease in acute events that the model predicts. In the case of the outbreak of Salmonella
infections in Alamosa, Colorado, as described previously in this section of the EA, some of the
key factors that contributed to the outbreak include significant levels of sediment and the
presence of breaches to the integrity of the storage tank.
9.3.4 Summary of Conclusions
The preferred option for this rulemaking is the AIP option. The analyses performed as
part of this EA support the collective judgment and consensus of the advisory committee that the
proposed RTCR requirements, as defined by the AIP, provide for effective and efficient
revisions to current TCR regulatory requirements. The following is a summary of points that
must be considered when weighing the benefits and costs of the rule options:
• The estimated net annual cost of the AIP option is small (approximately $14 million)
compared to the Alternative option ($27 million) (exhibit 9.11). This cost differential is
especially important considering the potential concentration of impacts on the smallest
TNCWSs (exhibit 9.13) and the potential front loading of costs under the Alternative
option (exhibits 9.8 - 9.10).
• The benefits are expected to be positive under the AIP option and no backsliding in
overall risk is predicted (Section 9.1.1.1) from the current TCR. The total benefits under
Economic Analysis for the Proposed RTCR
9-39
June 2010
-------
the Alternative option would be greater than under the AIP option (Section 9.1.1.2), but
the AIP option is more cost effective both on an average basis and an incremental basis
(Exhibits 9.15-9.17).
Of the various outputs of the occurrence and cost models developed in this EA, the
estimate of corrective actions to be implemented has the clearest connection to
potential benefits. Based on this metric, the AIP option is more cost effective on
average (i.e., costs less per corrective action achieved) than the Alternative option
(exhibit 9.15) by a factor of approximately 1.5.
- EPA conducted an incremental cost effectiveness analysis based on the increasing
stringency of the options from the current TCR to the AIP option to the Alternative
option. In comparing the ratio of additional costs and benefits associated with the
AIP option compared to the current TCR and of the Alternative option compared
to the AIP option, EPA found the following results:
¦ Exhibit 9.16 shows that the incremental cost effectiveness of the AIP
option far exceeds that of the Alternative option, by a factor of
approximately 3.2.
¦ Exhibit 9.17 shows that for TNCWSs the cost effectiveness of the AIP
option far exceeds that of the Alternative option, by a factor of
approximately 3.5.
• Although data limitations precluded estimating the number of fatal or non-fatal illnesses
that would be avoided by implementation of the regulatory alternatives considered, the
break-even analysis (Section 9.3.3) shows that just over 2 deaths would need to be
avoided annually for the rule to break even based on estimated net costs (exhibit 9.19)
using a three percent discount rate. Considering the average costs of non-fatal illnesses
resulting from STEC 0157 infection, the AIP would break even if it avoided 3,000 or
8,000 such cases annually based on the enhanced or traditional cost of illness approach,
respectively. These thresholds under the Alternative option are approximately 60% and
100% more than under the AIP option using the enhanced and traditional approaches,
respectively. Under the AIP, the estimated break-even threshold for non-fatal Salmonella
cases is 10,000 or 65,000 for the enhanced and traditional approaches, respectively, and
approximately 50% more (using both enhance and traditional approaches) for the
Alternative option. These break-even thresholds would be reduced if more than one type
of pathogen is avoided through implementation of the RTCR, or if cases of secondary
transmission are avoided. Given the large variety of pathogens that may occur in PWS
source water and the relatively low net costs of the AIP option, EPA believes the AIP
option is likely to break even.
• There are several items that are not quantified in the EA that represent a potential for
significant additional benefits (Section 9.2.2). These additional benefits may be realized
due to reductions in co-occurring contaminants, increased systems knowledge (including
more consistent water quality), reductions in outbreak risks, accelerated infrastructure
Economic Analysis for the Proposed RTCR
9-40
June 2010
-------
repair/replacement, and reductions in averting behavior (due to higher consumer
confidence in water quality).
As a result of these considerations, EPA believes that the estimated benefits of the
proposed RTCR as defined by the AIP option will likely exceed the modest increases in
annualized national costs and will be effective in reducing risk to consumers from exposure to
fecal contamination in drinking water.
Economic Analysis for the Proposed RTCR
9-41
June 2010
-------
10. References
Altekurse, S. F., N.J. Stern, P.I. Fields, andD.L. Swerdlow. 1999. Camplybacter jejuni - An
emerging foodborne pathogen. Emerging Infectious Diseases. 5(10): 1-10.
Angulo, F.J., S. Tippen, D. Sharp, B.J. Payne, C. Collier, J. Hill, T.J. Barrett, R.M. Clark, E.
Geldreich, H.D. Donnell, and D.L. Swerdlow. 1997. A community waterborne outbreak of
salmonellosis and the effectiveness of a boil water order. American Journal of Public Health.
87(4):580-584.
Association of State Drinking Water Administrators (ASDWA). 2001. Drinking water program
resource needs assessment. Version 9. November 27, 2001.
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 andMortality Weekly
Report. 49(SS04):l-35.
Bennear, L., K. Jessoe, and S. Olmstead. 2009. Sampling Out: Regulatory Avoidance and the
Total Coliform Rule. Environmental Science and Technology. 43:5176-5182.
Bennett, J.V. et al. 1987. Infectious and parasitic diseases. In: Amler, Closing the Gap: The
Burden of Unnecessary Illness, R.W. and H.B. Dull, (eds.). p. 102-14. New York, NY: Oxford
Press.
Berg, Rebecca. 2008. The Alamosa Salmonella Outbreak: A Gumshoe Investigation. Journal of
Environmental Health. 71 (2): 54-5 5.
Boring, J.R., W.T. Martin, and L.M. Elliott. 1971. Isolation of salmonella typhimurium from
municipal water, Riverside, California, 1965. American Journal of Epidemiology. 93:49-54.
Carducci A., L. Caniania, R. Moscatelli, B. Casini, E. Rovini, F. Mazzoni, A. Giuninti, and M.
Verani.2002. Interference between enterovirus and reovirus as a limiting factor in environmental
virus detection. Letters in Applied Microbiology. 34:110-113.
Carter, M.J. 2005. A review: Enterically infecting viruses: pathogenicity, transmission and
significance for food and waterborne infection. Journal of Applied Microbiology. 98:1354-1380.
Casemore, DP 1990. Epidemiologic aspects of human cryptosporidiosis. Epidemiology and
Infection 104:1-28.
CDC. 2001a. Norwalk-like viruses, public health consequences and outbreak management.
Morbidity and Mortality Weekly Report. 50(RR-9):18.
CDC. 2001b. Two fatal cases of adenovirus-related illness in previously healthy young adults -
Illinois, 2000. Morbidity and Mortality Weekly Report. 50(26): 553-555.
Economic Analysis for the Proposed RTCR
10-1
June 2010
-------
Centers for Disease Control (CDC). 2002. HIV AIDS Surveillance Report: U.S. HIV and AIDS
cases reported through December 2001. Vol. 13. No. 2.
CDC and Prevention Division of Foodborne Bacterial and Mycotic Diseases (DFBMD).
Salmonellosis. 2008.
http://www.cdc.gOv/nczved/dfbmd/disease_listing/salmonellosis_gi.html#2; downloaded 10-6-
09
Clark, W. F., J.J Macnab, and J.M. Santrop. 2008. The Walkerton Health Study, 2002-2008:
Final Report to the Ontario Ministry of Health and Long-Term Care. London Health Sciences
Center. University of Western Ontario. Available at:
www.lhrionhealth.ca/feature/Walkerton.pdf.
Colorado Department of Public Health and Environment (CDPHE). 2009. Waterborne
Salmonella Outbreak in Alamosa, Colorado March and April 2008: Outbreak identification,
response, and investigation. Available at:
http://www.cdphe.state.co.us/wq/DrinkingWater/pdf/AlamosaInvestRpt.pdf
Cooley, M., D. Carychao, L. Crawford-Miksza, M.T. Jay, C. Myers, C. Rose, C. Keys, J. Farrar,
and R.E. Mandrell. 2007. Incidence and tracking of Escherichia coli 0157:H7 in a major produce
production region in California. PLoS ONE. November 2007. 14;2(1 l):el 159.
Craun, G.F. 1996. "Cases of illness required to initiate a waterborne outbreak investigation."
Unpublished report to SAIC.
Craun, G.F., R.L. Calderon and T.J. Wade. 2006. Assessing waterborne risks: an introduction.
Journal of Water and Health 4(2):3-18.
Edberg, R. 2000. E. coli, the best biological drinking water indicator for public health protection.
Journal of Applied Microbiology. 88: 106S-116S.
ERS Foodborne Illness Cost Calculator: STEC 0157. Available online:
http://www.ers.usda.gov/Data/FoodborneIllness/ecoli_Intro.asp.
Feng, P. 1995. Escherichia coli serotype 0157:H7: Novel vehicles of infection and emergence of
phenotypic variants. Emerging Infectious Diseases l(2):47-52.
Framm, S.R. and R. Soave. 1997. Agents of diarrhea. The Medical clinics of North America.
81(2):427-447.
Garg, A.X., R.S. Suri, N. Barrowman, F. Rehman, D. Matsell, M. Patricia Rosas-Arellano, M.
Salvadori,R.B. Haynes and W.F. Clark. 2003. Long-term renal prognosis of diarrhea-associated
hemolyticuremic syndrome: A systematic review, meta-analysis and meta-regression. Journal of
American Medical Association. 290(10): 1360-1370.
Economic Analysis for the Proposed RTCR
10-2
June 2010
-------
Garg, A. X., L. Moist, D. Matsell, H.R. Thiessen-Philbrook, R.B. Haynes, R.S.Suri, M.
Salvadori, J. Ray, and W.F. Clark. 2005. Risk of hypertension and reduced kidney function after
acutegastroenteritis from bacteria-contaminated drinking water. Canadian Medical Association
Journal. 173(3): 1-8.
Geldreich, E.E. 1996. Microbial quality of water supply in distribution systems. Boca Raton, FL:
Lewis Publishers.
Gelman, A., J.B. Carlin, H.S. Stern, and D.R. Rubin. 1995. Bayesian Data Analysis. London:
Chapman & Hall.
Gerba, C.P., J.B. Rose, and C.N. Haas. 1996a. Sensitive subpopulations: who is at the greatest
risk? International Journal of Food and Microbiology. 30:113-123.
Glass, R.I., P.E. Kilgore, R.C. Holman, S. Jin, J.C. Smith, P.A. Woods, M.J. Clarke, M.S. Ho,
and J.R. Gentsch. 1996. The epidemiology of rotavirus diarrhea in the United States: surveillance
and estimates of disease burden. Journal of Infectious Diseases. 174(Suppl 1): S5—S11.
Gofti-Laroche, L., B. Gratacap-Cavallier, D. Demanse, O. Genoulaz, J.M. Seigncurin, andD.
Zmirou. 2003. Are waterborne astrovirus implicated in acute digestive morbidity (EMIRA
Study)? Journal of Clinical Virology. 27:74-82.
Gould, L. H., L. Demma, T. F. Jones, S. Hurd, D.J. Vugia, K. Smith, B. Shiferaw, S.Segler, A.
Palmer, S. Zansky, P. M. Griffin, and the Emerging Infections Program FoodNet Working
Group. 2009. Hemolytic Uremic Syndrome and Death in Persons with Escherichia coli 0157:H7
Infection, Foodborne Diseases Active Surveillance Network Sites, 2000-2006. Clinical
Infectious Diseases. 49:1480-5.
Griffin, D.D., C. D. Kirkwood, U. D. Parashar, P. A. Woods, J. S. Bresee, R. I. Glass, and J. R.
Gentsch. 2000. National Rotavirus Strain Surveillance System Collaborating Laboratories.
Surveillance of Rotavirus Strains in the United States: Identification of Unusual Strains. Journal
of Clinical Microbiology. 38(7):2784-2787.
Griffin, P. M. and R. V. Tauxe. 1991. The Epidemiology of Infections Caused by Escherichia
coli 0157: H7, Other Enterohemorrhagic E. coli, and the Associated Hemolytic Uremic
Syndrome. Epidemiologic Reviews. 13(l):60-98.
Guerrant, R. L. 1997. Cryptosporidiosis: An Emerging, Highly Infectious Threat. Emerging
Infectious Diseases. 3(1).
Health Canada. 2000. Waterborne outbreak of gastroenteritis associated with a contaminated
municipal water supply, Walkerton, Ontario, May-June 2000. Communicable Disease Report.
26-20.
Hegarty, J.P., M.T. Dowd and K.H. Baker. 1999. Occurrence of helicobacter pylori in surface
water in the United States. Journal of Applied Microbiology. 87:697-701.
Economic Analysis for the Proposed RTCR
10-3
June 2010
-------
Helmi, K., S. Skraber, C. Gantzer, R. Willame, L. Hoffmann, and H.M. Cauchie. 2008.
Interactions of Cryptosporidium parvum, Giardia lamblia, vaccinal poliovirus type 1, and
bacteriophages phiX174 and MS2 with a drinking water biofilm and a wastewater biofilm.
Applied and Environmental Microbiology. 74:2079-2088.
Holler, C., S. Koschinsky and D. Witthuhn. 1999. Isolation of enterohaemorrhagic Escherichia
coli from municipal sewage. Lancet. 353:2039.
Hopkins, R. S., P. Shillam, B. Gaspard, L. Eisnach, and R. J. Karlin. 1985. Waterborne disease in
Colorado: three years' surveillance and 18 outbreaks. American Journal of Public Health.
75(3):254-257.
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. American Journal of Public Health. 87(12):2032-2035.
Hunter P.R., S. Hughes, S. Woodhouse, N. Raj, Q. Syed, R. M. Chalmers, N. Q. Verlander, and
J. Goodacre. 2004. Health Sequelae of Human Cryptosporidiosis in Immunocompetent Patients.
Clinical Infectious Diseases. 39:504-510.
Hussein, H.S. 2007. Prevalence and pathogenicity of Shiga toxin-producing Escherichia coli in
beef cattle and their products. Journal of Animal Science 85(E. Suppl.): E63-E72
Juranek, D. D. 1998. Cryptosporidium and Water: A Public Health Handbook. Atlanta, Georgia:
Working Group on Waterborne Cryptosporidiosis.
Kaper, J.B, J.P. Nataro, H.L.T. Mobley. 2004. Pathogenic Escherichia coli. Nature Reviews
Microbiology 2(2): 123-140
Kilgore, P.E., R.C. Holman, M.J. Clarke, and R.I. Glass. 1995. Trends of diarrheal disease—
associated mortality in U.S. children, 1968 through 1991. Journal of the American Medical
Association.!!4(14): 1143-1148.
Kramer, M.H., B.L. Herwaldt, R.L. Calderon, and D.D. Juranek. 1996. Surveillance for
waterborne-disease outbreaks—United States, 1993-1994. Morbidity and Mortality Weekly
Report. 45(SS-l):l-33.
LeBaron, C.W., M.D. Furutan, J.F. Lew, J R. Allen, D.V. Gouvea, D C. Moe, D.S. Monroe.
1990. Viral agents of gastroenteritis; public health importance and outbreak management.
Morbidity and Mortality Weekly Report. 39(RR-5):l-24.
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. 51(SS08):l-28.
Economic Analysis for the Proposed RTCR
10-4
June 2010
-------
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. Morbidity andMortality Weekly
Report. 47(55-5): 1-34.
Marciano-Cabral, F., R. MacLean, A. Mensah, L. LaPat-Polasko. 2003. Identification of
Naegleria fowleri in domestic water wources by nested PCR. Applied and Environmental
Microbiology. 69(10):5864-5869.
Maria, H., A. Elshebani, O. Anders, T. Torsten, and F. Gun. 2005. Simultaneous type 1 diabetes
onset in mother and son coincident with an enteroviral infection. Journal of Clinical Virology
33:158-167.
McMillan, S. 1996. Camp Four Echoes outbreak investigation. Unpublished report by the
Panhandle Health District, Environmental Health Division, Coeur d'Alene, Idaho.
Mead, P.S., L. Slutsker, V. Dietz, L.F. McCaig, J.S. Bresee, C. Shapiro, P.M. Griffin, and R.V.
Tauxe. 1999. Food-related illness and death in the United States. Emerging Infectious Diseases
5(5):607-625.
Means. R.S. 1998. Mechanical Cost Data. Kingston: R.S. Means Company, Inc.
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. Morbidity and
Mortality Weekly Report. Surveillance Summary SS-5, US Centers for Disease Control and
Prevention.
Murray, P.R., E.J. Baron, J.H. Jorgensen, M.L. Landry, and M.A. Pfaller, eds. 2007. Manual of
Clinical Microbiology. Volume 1. Washington, D.C.: ASM Press. 9th Edition. 2256pp.
National Research Council. 1997. Safe water from every tap, improving water service to small
communities. Washington, DC: National Academy Press.
Ntzoufras, I. 2009. Bayesian Modeling Using WinBUGS. Hoboken: John Wiley & Sons, Inc.
Parashar, U.D., R.C. Holman, M.J. Clarke, J.S. Bresee, and R.I. Glass. 1998. Hospitalizations
associated with rotavirus diarrhea in the United States, 1993 through 1995: surveillance based on
the new ICD-9-CM rotavirus-specific diagnostic code. Journal of Infectious Diseases. 177:13-
17.
Parry, S.M. and R. L. Salmon. 1998. Sporadic STEC 0157 infection: secondary household
transmission in Wales. Emerging Infectious Diseases. 4(4): 1-6.
Perz, J. F., F. K. Ennever, and S. M. Le Blancq. 1998. Cryptosporidium in Tap Water:
Comparison of Predicted Risks with Observed Levels of Disease. American Journal of
Epidemiology. 147(3): 289-301.
Economic Analysis for the Proposed RTCR
10-5
June 2010
-------
Prager, R., S. Annemuller and H. Tschape. 2005. Diversity of virulence patterns among shiga
toxin-producing Escherichia coli from human clinical cases - need for more detailed diagnostics.
International Journal of Medical Microbiology. 295:29-38.
Quignon, F., L. Kiene, Y. Levi, M. Sardin, L. Schwartzbrod. 1997. Virus behavior within a
distribution system. HEALTH-RELATED WATER MICROBIOLOGY 1996. 311-318.
Rangel, J.M., P.H. Sparling, C.Crowe, P.M. Griffin andD. L. Swerdelow. 2005. Epidemiology
of the Escherichia coli 0157:H7 outbreaks, United States, 1982-2002. Emerging Infectious
Diseases. 11(4): 1-11.
Rolle-Kampczyk, U.E., G.J. Fritz, U. Diez, I. Lehman, M. Richter, and O. Herbarth. 2004. Well
water - one source of Helicobacter pylori colonization. International Journal of Hygiene and
Environmental Health. 207:363-368.
Rose, J.B. 1997. Environmental ecology of Cryptosporidium and public health implications.
Annual Review of Public Health. 18:135-161.
SAB (Science Advisory Board). 2010. Science Advisory Board Drinking Report on the
Agency's Supporting Analyses for Revisions of the Total Coliform Rule. Final Report. March
19, 2010.
SBAR Panel (Small Business Advocacy Review Panel). 2008. Small Business Advocacy Review
Panel on EPA Planned Revisions to Public Water System Requirements: Revisions to the Total
Coliform Monitoring and Analytical Requirements and Consideration of Distribution System
Issues. Final Report. March 31, 2008.
Skraber, S., J. Schijven, C. Gantzer, and A.M. de Roda Husman. 2005. Pathogenic viruses in
drinking-water biofilms: a public health risk? Biofilms. 2:105-117.
Standridge, J. 2008. E. coli as a public health indicator of drinking water quality. Journal AWWA
100(2):65-75.
Swerdlow, D.L., B.A. Woodruff, R.C. Brady, P.M. Griffin, S. Tippen, H. Donnel, Jr., E.
Geldreich, B.J. Payne, A. Meyer, Jr., J.G. Wells, K.D. Greene, M. Bright, N.H. Bean, and P.A.
Blake. 1992. A waterborne outbreak in Missouri of Escherichia coli 0157:H7 associated with
bloody diarrhea and death. Annals ofInternal Medicine. 117(10):812-819.
Tsang, T.H.F., E.K. Denison, H.V. Williams, L.V. Venczel, M.M. Ginsberg and D. J. Vugia.
2000. Acute Hepatitis E infection acquired in California. Clinical Infectious Diseases. 30:618-
619.
Tyler, K.L., E.S. Barton, M.L. Ibach, C. Robinson, J.A. Campbell, S.M. O'Donnell, T. Valy-
Nagy, P. Clarke, J. D. Wetzel, and T. Dermody. 2004. Isolation and molecular characterization
of a novel type 3 reovirus from a child with meningitis. Journal of Infectious Diseases.
189:1664-1675.
Economic Analysis for the Proposed RTCR
10-6
June 2010
-------
United States Census Bureau. 2001a. Households and Families: 2000. Census 2000 Brief.
C2KBR/01-8.
United States Census Bureau. 2001b. Statistical Abstract of the United States: 2001.
Washington, DC. Economics and Statistics Administration.
United States Census Bureau. 2008. Current Population Survey. Annual Social and Economic
Supplement. Population Division, Education & Social Stratification Branch.
United States Environmental Protection Agency (USEPA). 1979. National Interim Primary
Drinking Water Regulations: Control of Total Trihalomethanes in Drinking Water: Final Rule.
Federal Register 44(231):68624. November 29, 1979.
United States Environmental Protection Agency (USEPA). 1989a. National Primary Drinking
Water Regulations: Filtration, Disinfection; Turbidity, Giardia lamblia, Viruses, Legionella, and
Heterotrophic Bacteria; Final Rule (SWTR); Final Rule. Federal Register 54(124): 27486. June
29, 1989.
United States Environmental Protection Agency (USEPA). 1989b. National Primary Drinking
Water Regulations: Total Coliforms (Including Fecal Coliform and E. colij; Final Rule. Federal
Register 54(124):27544. June 29, 1989.
United States Environmental Protection Agency (USEPA). 1993. Executive Order Regulatory
Planning and Review. Federal Register 58(190):51735. October 4, 1993.
United States Environmental Protection Agency (USEPA). 1994. Executive Order12898 Federal
Actions To Address Environmental Justice in Minority Populations and Low-Income
Populations. Federal Register 59(32):7629.
United States Environmental Protection Agency (USEPA). 1996a. 1996 Amendments to the Safe
Drinking Water Act. Public Law 104-182 104th Congress. PL 104-182.
United States Environmental Protection Agency (USEPA). 1996b. Economic analysis of Federal
regulations under Executive Order 12866. Office of Management and Budget. January 11, 1996.
United States Environmental Protection Agency (USEPA). 1996c. National Primary Drinking
Water Regulations: Monitoring Requirements for Public Drinking Water Supplies:
Cryptosporidium, Giardia, Viruses, Disinfection Byproducts, Water Treatment Plant Data and
Other Information Requirements; Final Rule. Office of Ground Water and Drinking Water.
Federal Register 61(94): 24354. May 14, 1996.
USEPA. 1996d. Ground Water Disinfection and Protective Practices in the United States.
Prepared by EPA and Science Applications International Corporation. Office of Ground Water
and Drinking Water, Washington, D.C.
Economic Analysis for the Proposed RTCR
10-7
June 2010
-------
United States Environmental Protection Agency (USEPA). 1997. Executive Order 13045
Protection of Children From Environmental Health Risks and Safety Risks. Federal Register
62(48): 19885. April 23, 1997.
United States Environmental Protection Agency (USEPA). 1998a. National Primary Drinking
Water Regulations: Consumer Confidence Reports; Final Rule. Federal Register 63(160): 44512.
August 19, 1998.
United States Environmental Protection Agency (USEPA). 1998b. National Primary Drinking
Water Regulations: Disinfectants and Disinfection Byproducts; Final Rule. Office of Ground
Water and Drinking Water. Federal Register 63(241):69389. December 16, 1998.
United States Environmental Protection Agency (USEPA). 1998c. National Primary Drinking
Water Regulations: Interim Enhanced Surface Water Treatment. Office of Ground Water and
Drinking Water. Federal Register 63(241): 69477-69521. December 16, 1998.
United States Environmental Protection Agency (USEPA). 1998d. National Primary Drinking
Water Regulations: Regulatory Flexibility Act. Federal Register 63(30):7620. February 13, 2008.
United States Environmental Protection Agency (USEPA). 1999. Executive Order 13132
Federalism. Federal Register 64(153): 43255. August 10, 1999.
United States Environmental Protection Agency (USEPA). 2000a. Development of Cost of
Capital Estimates for Public Water Systems, Final Report.
United States Environmental Protection Agency (USEPA). 2000b. Executive Order 13175
Consultation and Coordination With Indian Tribal Governments. Federal Register 65(218):
67249.
United States Environmental Protection Agency (USEPA). 2000c. Guidelines for Preparing
Economic Analyses. U.S. EPA Office of the Administrator. EPA/240-R-00-003. September
2000.
United States Environmental Protection Agency (USEPA). 2001a. Executive Order 13211
Actions Concerning Regulations That Significantly Affect Energy Supply, Distribution, or Use.
Federal Register 66(99):28355.
United States Environmental Protection Agency (USEPA). 2001b. National Primary Drinking
Water Regulations; Arsenic and Clarifications to Compliance and New Source Contaminants
Monitoring; Final Rule. Federal Register 66(14):6976. January 22, 2001.
United States Environmental Protection Agency (USEPA). 2001c. Third Edition of the Water
Industry Baseline Handbook (Baseline Handbook). May 2001.
Economic Analysis for the Proposed RTCR
10-8
June 2010
-------
United States Environmental Protection Agency (USEPA). 2002a. National Primary Drinking
Water Regulations: Long Term 1 Enhanced Surface Water Treatment Rule : Final Rule. Federal
Register 67(9):1812. January 14, 2002.
United States Environmental Protection Agency (USEPA). 2002b. National Primary Drinking
Water Regulations; Announcement of the Results of EPA 's Review of Existing Drinking Water
Standards and Request for Public Comment. Federal Register 67(74): 19030. April 17, 2002.
United States Environmental Protection Agency (USEPA). 2003a. Labor Costs for National
Drinking Water Rules.
United States Environmental Protection Agency (USEPA). 2003b. National Primary Drinking
Water Regulations; Announcement of Completion of EPA 's Review of Existing Drinking Water
Standards: Notice. Federal Register 68(138):42907. July 18 2003.
United States Environmental Protection Agency (USEPA). 2005a. The Safe Drinking Water
Information System - Federal Version (SDWIS/FED) Violation data. Available online:
http://epa.gov/enviro/html/sdwis/ [14 Apr 2009],
United States Environmental Protection Agency (USEPA). 2006a. Ground Water Rule Economic
Analysis. Office of Water. EPA 815-R-06-014. October 2006.
United States Environmental Protection Agency (USEPA). 2007a. Establishment of the Total
Coliform Rule Distribution System Advisory Committee: Federal Register 72(125): 35869. June
29, 2007.
United States Environmental Protection Agency (USEPA). 2007b. The Safe Drinking Water
Information System - Federal Version (SDWIS/FED) data. Available online:
http://epa.gov/enviro/html/sdwis/ [29 Apr 2009],
United States Environmental Protection Agency (USEPA). 2008a. 2006 Drinking Water Data
Reliability Analysis and Action Plan, EPA 816-R-07-010. March 2008. Available online:
www. epa. gov/ safewater.
United States Environmental Protection Agency (USEPA). 2008b. Draft Information Collection
Request for the Microbial Rules. OMB Control Number: 2040-0205. EPA Tracking Number:
1895.04.
United States Environmental Protection Agency (USEPA). 2008c. Draft Information Collection
Request for the Public Water System Supervision Program. OMB Control Number: 2040-0090.
EPA Tracking Number: 0270.43.
United States Environmental Protection Agency (USEPA). 2008d. Assessment of the
Vulnerability of Non-Community Water Systems to SDWA Cost Increases.
Economic Analysis for the Proposed RTCR
10-9
June 2010
-------
United States Environmental Protection Agency (USEPA). 2009. Total Coliform Rule
Distribution Systems Advisory Committee Agreement in Principle: Notice of agreement in
principle. Federal Register 74(8): 1683. January 13, 2009.
United States Environmental Protection Agency (USEPA). 2010a. Information Collection
Request for Contaminant Occurrence Data in Support of EPA's Second Six-Year Review of
National Primary Drinking Water Regulations. EPA ICR Number 2231.01., OMB Control No.
2040
United States Environmental Protection Agency (USEPA). 2010b. Information Collection
Request for the Proposed Revised Total Coliform Ride.
United States Environmental Protection Agency (USEPA). 2010c. Preamble for the Proposed
Revised Total Coliform Ride. June 2010.
United States Environmental Protection Agency (USEPA). 2010d. Technology and Cost
Document for the Proposed Revised Total Coliform Rule. June 2010.
United States Environmental Protection Agency (USEPA). 2010e. Total Coliform Rule
Compliance Monitoring Data Quality and Completion Report. June 2010.
US Transplant. 2008. Scientific Registry of Transplant Recipients. Accessed 1 December 2008.
Data updated June, 2000. Available online: www.ustransplant.org
Ventura, S. J., Mosher, W.D., Curtin, S.C., Abma J.C., and S. Henshaw. 2000. Trends in
pregnancies and pregnancy rates by outcome: estimates for the United States, 1976-96. Vital
and health statistics. Series 21, Data on natality, marriage, and divorce 2000;(56):l-47.
Vesikari T. 2006. Safety and Efficacy of a Pentavalent Human-Bovine (WC3) Reassortant
Rotavirus Vaccine. The New England Journal of Medicine. 354:23-33.
Vreugdenhil G.R., N.C. Schloot, A. Hoorens, C. Rongen, D.G. Pipeleers, W.J.G. Melchers, B.O
Roep, and J.M.D. Galama. 2000. Acute onset of type 1 diabetes mellitus after severe echovirus 9
infection: putative pathogenic pathways. Clinical Infectious Diseases 31: 1025-1031.
WHO. 2004. Guidelines for Drinking-water Quality. Third Edition, Volume 1,
Recommendations. World Health Organization. Geneva, Switzerland.
Economic Analysis for the Proposed RTCR
10-10
June 2010
------- |