REGULATORY IMPACT ANALYSIS FOR THE PROPOSED
LONG TERM 1 ENHANCED SURFACE WATER TREATMENT AND
               FILTER BACKWASH RULE

             EPA DOCUMENT NO. 815-R-00-005


                   FEBRUARY 15,2000


                    PREPARED FOR:

          U.S. ENVIRONMENTAL PROTECTION AGENCY
       OFFICE OF GROUND WATER AND DRINKING WATER
      MR. STEVE POTTS, TASK ORDER PROJECT MANAGER


                     PREPARED BY:

                THE CADMUS GROUP, INC.
              1901 NORTH FORT MYER DRIVE
                    SUITE No. 1016
                 ARLINGTON, VA 22209

     SCIENCE APPLICATIONS INTERNATIONAL CORPORATION
                 1710 GOODRIDGE DRIVE
                MCLEAN, VA 22102-3701

       EPA CONTRACT No. 68-C-99-206, TASK ORDER 021

-------
                                TABLE OF CONTENTS

1.      Executive Summary	1-1
       1.1     Need for the Proposal	1-1
       1.2     Consideration of Regulatory Alternatives	1-3
       1.3     Baseline Analysis	1-4
       1.4     Benefits of the LT1FBR	1-5
       1.5     Costs of the LT1FBR	1-6
       1.6     Economic Impact Analysis	1-8
       1.7     Weighing of Benefits and Costs	1-9
2.      Need for the Proposal	2-1
       2.1     Description of the Issue 	2-1
       2.2     Public Health Concerns 	2-5
              2.2.1  Contaminants and Their Associated Health Effects  	2-5
              2.2.2  Sensitive Subpopulations 	2-6
              2.2.3  Sources of Contaminants 	2-6
              2.2.4  Waterborne Disease Outbreaks	2-12
              2.2.5  Filter Backwash and Other Process Streams: Occurrence and
                    Impact Studies	2-14
              2.2.6  Current Control and Potential for Improvement	2-18
       2.3     Regulatory History and Current Controls	2-19
              2.3.1  1979 Total Trihalomethane Rule 	2-19
              2.3.2  Total Coliform Rule	2-19
              2.3.3  Surface Water Treatment Rule	2-20
              2.3.4  Information Collection Rule	2-20
              2.3.5  Interim Enhanced Surface Water Treatment Rule	2-20
              2.3.6  Stage 1 Disinfection Byproduct Rule	2-21
              2.3.7  Stakeholder Involvement 	2-21
       2.4     Economic Rationale	2-22
              2.4.1  Introduction 	2-22
              2.4.2  Statutory Authority for Promulgating the Rule	2-22
              2.4.3  The Economic Rationale for Regulation	2-23
       2.5     Summary of the Proposed Rule  	2-24
              2.5.1  Enhanced Filtration Provisions	2-24
              2.5.2  Disinfection Benchmarking Provision  	2-25
              2.5.3  Other LT1 Provisions	2-27
              2.5.4  Recycle Provisions	2-27
3.      Consideration of Regulatory Alternatives	3-1
       3.1     Individual Filter Turbidity Monitoring  	3-1
       3.2     Applicability Monitoring  	3-4
       3.3     Disinfection Profiling and Benchmarking	3-6
       3.4     Recycle Provisions	3-8
4.      Baseline Analysis	4-1
       4.1     Baseline Assumptions	4-1
       4.2     Industry  Profile	4-3
       4.3     Number of Systems Under the Turbidity Provisions  	4-4

RIA for the Proposed LT1FBR                       -ii-                              February 15, 2000

-------
              4.3.1   Estimate of the Number of Systems Subject to 2 log
                     Cryptosporidium Removal Requirement 	4-4
              4.3.2   Systems Subject to Strengthened CFE Turbidity Standards	4-5
              4.3.3   Estimate of the Number of Systems Subject to
                     Individual Filter Monitoring Requirements	4-6
       4.4    Systems Affected by Disinfection Benchmarking Provision 	4-7
       4.5    Systems Affected by the Recycle Provisions  	4-7
              4.5.1   Recycle Practice	4-8
       4.6    Contaminant Exposure	4-20
5. Benefits Analysis	5-1
       5.1    Introduction 	5-1
              5.1.1   Expected Benefits from Turbidity Provisions	5-1
              5.1.2   Expected Benefits from Recycle Provisions	5-2
              5.1.3   Expected Benefits from Other Provisions	5-3
       5.2    Health Benefits from Turbidity Provisions	5-3
              5.2.1   Contaminants and Their Health Effects  	5-3
              5.2.2   Risk Assessment: Methods  and Assumptions	5-6
              5.2.3   Baseline and Reduced Health Risk of the Turbidity Provisions	5-19
              5.2.4   Monetization of Reduced Risks 	5-29
              5.2.5   Health Effects to Sensitive Subpopulations  	5-34
       5.3    Other Benefits  of Turbidity Provisions	5-35
              5.3.1   Reduction in Outbreak Risk	5-35
              5.3.2   Enhanced Aesthetic Water Quality	5-35
              5.3.3   Avoided Costs of Averting  Behavior	5-35
       5.4    Benefits from Other Rule Provisions  	5-37
              5.4.1   Benefits of Recycle Provision	5-37
              5.4.2   Benefits of Disinfection Benchmark Provision	5-41
              5.4.3   Benefits of Covered Finished Water Reservoirs	5-41
              5.4.4   Benefits from Including Cryptosporidium in the GWUDI Definition  . 5-42
              5.4.5   Benefits from Including Cryptosporidium in Watershed
                     Requirements for Unfiltered Systems	5-42
              5.4.6   Risk Reduction from Emerging Pathogens	5-42
       5.5    Summary  	5-43
              5.5.1   Summary of Quantified and Monetized Benefits  	5-43
              5.5.2   Summary of Non-Quantified Benefits  	5-44
              5.5.3   Summary of Uncertainties  	5-45
6.     Cost Analysis	6-1
       6.1    Introduction 	6-1
              6.1.1   Cost Assumptions	6-1
       6.2    Turbidity Provisions Cost Analysis  	6-3
              6.2.1   Combined Filter Effluent Provision:  Turbidity Treatment Costs	6-4
              6.2.2   Individual Filter Monitoring Provision: Monitoring Costs	6-9
              6.2.3   Individual Filter Monitoring Provision: Exceptions Reporting
                     Costs	6-13
       6.3    Disinfection Benchmarking Provision Cost Analysis	6-15
              6.3.1   System Costs  	6-15

February 15, 2000                              -in-                      RIA for the Proposed LT1FBR

-------
              6.3.2   State Costs  	6-23
       6.4    Covered Finished Water Reservoir Provision Cost Analysis 	6-24
              6.4.1   Unit Cost  	6-24
              6.4.2   Compliance Forecast  	6-24
              6.4.3   System Costs 	6-24
       6.5    Recycle Provisions Cost Analysis  	6-25
              6.5.1   Recycle to New Return Location	6-27
              6.5.2   Direct Recycle Provision Costs  	6-30
              6.5.3   Direct Filtration Provision Costs	6-35
              6.5.4   Summary of Costs by Regulatory Alternative  	6-38
       6.6    Summary of Costs  	6-39
              6.6.1   Omissions, Biases, and Uncertainty	6-41
       6.7    Household Costs	6-43
              6.7.1   Household Cost Estimation Method	6-44
              6.7.2   Results of Household Cost Analysis	6-45
       6.8    Cost Effectiveness  	6-46
7.     Economic Impact Analysis	7-1
       7.1    Impacts on Governments and Business Units	7-1
              7.1.1   Unfunded Mandates Reform Act 	7-1
              7.1.2   Indian Tribal Governments	7-10
              7.1.3   Regulatory Flexibility Act and Small Business Regulatory
                     Enforcement Fairness Act	7-10
              7.1.4   Effect of Compliance With the LT1FBR on the Technical,
                     Financial,  and Managerial Capacity of Public Water Systems  	7-18
              7.1.5   Paperwork Reduction Act	7-20
       7.2    Impacts on Subpopulations	7-21
       7.3    Environmental Justice	7-23
8.     Weighing the Benefits and the Costs  	8-1
       8.1    Incremental and Marginal Analysis  	8-1
       8.2    Benefit-Cost Comparisons  	8-1
       8.3    Breakeven Analysis for the Recycle Provisions	8-5
       8.4    Summary of Uncertainty/Sensitivity Analysis  	8-8
       8.5    Combined Regulatory Effects with Other Rules	8-9
9.     References	9-1
RIA for the Proposed LT1FBR                      -iv-                              February 15, 2000

-------
                                  LIST OF EXHIBITS

Exhibit 1-1.   Key Differences Between the Preferred Alternatives for the Proposed
             LT1FBR and the IESWTR Requirements	1-4
Exhibit 1-2.   Number of Systems Under LT1FBR Provisions 	1-5
Exhibit 1-3.   Summary of Annual Benefits Associated with Avoided Illnesses and
             Mortalities for the Turbidity Provisions (January 1999 Dollars)	1-6
Exhibit 1-4   Total Annual Costs for Two Combinations of Alternatives
             (January 1999 Dollars)	1-7
Exhibit 1-5.   Comparison of Annual Benefit, Cost, and Net Benefit Ranges (January 1999
             Dollars, Millions)  	1-10
Exhibit 2-1.   Flow Schematics for Conventional Water Treatment Systems	2-2
Exhibit 2-2.   Flow Schematics for Direct Filtration Systems	2-3
Exhibit 2-3.   Flow Schematics for Systems that Recycle	2-4
Exhibit 2-4.   Surface Water Survey and Monitoring Data for Cryptosporidium Oocysts  .... 2-7
Exhibit 2-5.   U.S. GWUDI Monitoring Data for Cryptosporidium Oocysts	2-10
Exhibit 2-6.   Comparison of Outbreaks and Outbreak-related Illnesses from Ground
             Water and Surface Water for the Period 1971-1996  	2-12
Exhibit 2-7.   Cryptosporidiosis Outbreaks in U.S. Drinking Water Systems	2-13
Exhibit 2-8.   Cryptosporidium Occurrence in Filter Backwash and Other Recycle Streams . 2-15
Exhibit 2-9.   Summary of How the Proposed LT1FBR is Organized	2-29
Exhibit 2-10. Illustration of How the Provisions Apply to Different Types of Surface  	2-30
Exhibit 3-1.   Filter Turbidity Monitoring Alternatives	3-3
Exhibit 3-2.   Applicability Monitoring Alternatives	3-6
Exhibit 3-3.   Disinfection Profiling and Benchmarking Alternatives	3-7
Exhibit 3-4.   Filter Backwash Alternatives  	3-11
Exhibit 4-1.   System Population Size Categories and Total Population	4-2
Exhibit 4-2.   Systems Utilizing Surface Water or GWUDI Serving Less Than Under
             10,000 People	4-3
Exhibit 4-3.   Systems Utilizing Surface Water or GWUDI Serving More Than 10,000
             People	4-3
Exhibit 4-4.   Estimate of Systems Subject to 2 log Cryptosporidium Removal Requirement . 4-5
Exhibit 4-5.   Estimate of Systems Subject to Strengthened CFE Turbidity Standards for
             Conventional and Direct Filtration Systems	4-6
Exhibit 4-6.   Estimate of Systems Subject to Disinfection Benchmarking Provision	4-7
Exhibit 4-7.   Recycle Practice at ICR Plants	4-8
Exhibit 4-8.   ICR Recycle Plants by Population Served	4-9
Exhibit 4-9.   ICR Plants by Treatment Train Type	4-9
Exhibit 4-10. Source Water Use by ICR Recycle Plants  	4-10
Exhibit 4-11. Treatment of Recycle at ICR Plants  	4-10
Exhibit 4-12. Recycle Return Point  	4-11
Exhibit 4-13. Distribution of FAX Survey Plants by Plant Capacity  	4-12
Exhibit 4-14. Number of Plants per State Included in AWWA Recycle Survey 	4-13
Exhibit 4-15. Source Water Used by FAX Survey Plants  	4-14
Exhibit 4-16. Treatment Trains of FAX Survey Plants	4-14
Exhibit 4-17. Recycle Return Point of FAX  Survey Plants  	4-15
February 15, 2000
RIAfor the Proposed LT1FBR

-------
Exhibit 4-18. Recycle Treatment at FAX Survey Plants	4-15
Exhibit 4-19. Recycle Practice Based on Population Served  	4-16
Exhibit 4-20. Options to Recycle as Reported by FAX Survey Plants	4-17
Exhibit 4-21. Summary of the Number of Systems Affected by Recycle Provisions	4-19
Exhibit 5-1.  Overview of LT1FBR Benefit Categories and Associated Components	5-2
Exhibit 5-2.  Symptoms of 205 Patients with Confirmed Cases of Cryptosporidiosis during
             the Milwaukee Outbreak  	5-5
Exhibit 5-3.  Steps in the Health Risk Assessment for Cryptosporidium	5-7
Exhibit 5-4.  Cryptosporidium Exposure Probabilities and Characteristics of Public
             Drinking Water Systems	5-9
Exhibit 5-5.  Summary of Hazard Identification Assumptions  	5-10
Exhibit 5-6.  Baseline Expected National Source Water Cryptosporidium Distributions
             (oocysts/lOOL)  	5-12
Exhibit 5-7.  Baseline Expected National Finished Water Cryptosporidium Distributions,
             Based on Current Treatment (oocysts/lOOL)  	5-14
Exhibit 5-8.  Summary of Systems and Population Potentially Modifying Treatment under
             the LT1FBR Turbidity Provisions  	5-14
Exhibit 5-9.  Cumulative Probability Distribution of Aggregate Pilot Plant Data for C.
             parvum Removal 	5-16
Exhibit 5-10. Improved Cryptosporidium Removal Assumptions  (Additional
             Cryptosporidium Log Removal with the Proposed Rule)	5-16
Exhibit 5-11. Expected National Source Water and Finished Water Cryptosporidium
             Distributions with Improved Removal  	5-17
Exhibit 5-12. Summary of Exposure  Assessment Assumptions	5-18
Exhibit 5-13  Distribution of Annual Individual Risks of Illness Due to Cryptosporidium
             for the Baseline and Improved Removal Scenarios  	5-22
Exhibit 5-14. Number of Illnesses and Illnesses Avoided	5-23
Exhibit 5-15. Distribution of Annual  Individual Risks of Mortality Due to
             Cryptosporidium for the Baseline and Improved Removal Scenarios 	5-27
Exhibit 5-16. Number of Mortalities  and Mortalities Avoided among Exposed Population . . 5-28
Exhibit 5-17. Losses per Case of Giardiasis by Category	5-30
Exhibit 5-18. Number and Value of Illnesses Avoided Annually from Turbidity
             Provisions ($Millions)	5-32
Exhibit 5-19. Number and Cost of Mortalities Avoided Annually from Turbidity
             Provisions ($Millions)	5-34
Exhibit 5-20. Potential Benefit for a System Serving 1,900 People 	5-41
Exhibit 5-21. Potential Benefit Range for System  Serving 25,108 People	5-41
Exhibit 5-22. Summary of Annual Benefits Associated with Avoided Illnesses and
             Mortalities for the Turbidity Provisions ($Millions)  	5-44
Exhibit 5-23. Summary of Non-Quantified Benefits	5-45
Exhibit 5-24. Damages/Benefits Summary of Bias and Uncertainty	5-46
Exhibit 6-1.  Summary of the Estimated Number of Small Systems Affected by Turbidity
             Provision	6-4
Exhibit 6-2a. Treatment Process Improvement Capital Costs per System
             (January 1999 dollars)  	6-6
Exhibit 6-2b. Treatment Process Improvement Operation & Maintenance
RIAfor the Proposed LT1FBR
February 15, 2000

-------
              Costs per System (January 1999 dollars)  	6-7
Exhibit 6-3.   Number of Systems Modifying Treatment Practices to Meet New Turbidity
              Standards 	6-9
Exhibit 6-4.   Annual Cost Estimates for Turbidity Treatment Requirements
              (January 1999 dollars, millions) 	6-9
Exhibit 6-5.   Summary of Labor Requirements for Turbidity Monitoring Alternatives  .... 6-11
Exhibit 6-6.   Annual System Turbidity Start-Up and Monitoring Cost by  Alternative
              (January 1999 dollars)  	6-12
Exhibit 6-7.   State Turbidity Start-Up and Monitoring Annual Cost
              (January 1999 dollars)  	6-13
Exhibit 6-8.   Exceptions Reporting, Individual Filter Assessment, and  Comprehensive
              Performance Evaluation Requirements for the Individual Filter Turbidity
              Monitoring Alternatives 	6-13
Exhibit 6-9.   System and State Costs for Exception Reports, Individual Filter Turbidity
              Assessments, and Comprehensive Performance Evaluations (January 1999
              dollars)  	6-15
Exhibit 6-10.  Disinfection Benchmark Development Start-up Costs by System  Size
              (January 1999 dollars)  	6-16
Exhibit 6-11.  Summary of Proposed Applicability Monitoring Sampling Alternatives	6-17
Exhibit 6-12.  Disinfection Benchmark Applicability Monitoring Costs by Alternative and
              System Size (January 1999 dollars)  	6-18
Exhibit 6-13.  Summary of Alternative System-Level Data Collection Requirements
              for the Disinfection Profile	6-20
Exhibit 6-14.  Summary of System Disinfection Profile and Benchmarking Costs
              by Alternative and System Size (January 1999 dollars)  	6-22
Exhibit 6-15.  State Disinfection Benchmarking Costs (January 1999 dollars)	6-23
Exhibit 6-16.  Unit Cost Assumptions to Cover New Finished Water Reservoirs	6-24
Exhibit 6-17.  Total Cost Estimates to Cover New Finished Water Reservoirs
              (January 1999 dollars)  	6-25
Exhibit 6-18.  Systems Potentially Affected by the Proposed Recycling  Provisions 	6-26
Exhibit 6-19.  Total Annualized System Costs for Reporting Proposed New Return
              Location by  System Size (January 1999 dollars) 	6-29
Exhibit 6-20.  State Cost Estimate to Review and Approve Plans to Move Recycle
              Return Location (January 1999 dollars)	6-29
Exhibit 6-21.  Total Annualized Costs for Recycling to Return Location by System Size (January
              1999 dollars)	6-30
Exhibit 6-22.  Total System Start-up and Self Assessment Costs by System  Size for the
              Direct Recycle Provision (January 1999 dollars) 	6-32
Exhibit 6-23.  Total State Start-up and Review Costs for the Direct Recycle Provision
              (January 1999 dollars)  	6-33
Exhibit 6-24.  Total Annualized Costs to Modify Recycling Practices for the Direct
              Recycle Provision by System Size and Regulatory Alternative
              (January 1999 dollars)  	6-34
Exhibit 6-25.  Total System Start-up and Reporting Costs for the Direct Filtration
              Provision by System Size (January 1999 dollars)	6-36
Exhibit 6-26.  Total State Start-up and Review Costs for the Direct Filtration Provision
February 15, 2000
RIAfor the Proposed LT1FBR

-------
              (January 1999 dollars)  	6-37
Exhibit 6-27.  Annualized Costs for Altering Recycle Practices for the Direct Filtration Provision
              by System Size (January 1999 dollars)	6-38
Exhibit 6-28.  Total Annual System and State Costs by Recycling Provision and
              Regulatory Alternative (January 1999  dollars)	6-38
Exhibit 6-29.  Total Annual Costs for Two Combinations of Alternatives
              (January 1999 dollars)  	6-39
Exhibit 6-30.  Summary of Total Annual Costs for the Preferred Alternatives by System
              Size (January 1999 dollars)	6-41
Exhibit 6-31.  Summary of Cost Analysis Uncertainty	6-42
Exhibit 6-32.  Distributions of Annual Household Costs for the Turbidity, Benchmarking,
              and Covered Finished Water Provisions and Recycle Provisions	6-46
Exhibit 7-1.    Geographic Distribution of Annual LT1FBR Costs to States	7-6
Exhibit 7-2.    Small Surface and GWUDI System Distribution	7-6
Exhibit 7-3.    LT1FBR Costs as a Percentage of State Drinking Water  Expenditures	7-7
Exhibit 7-4.    Per Capita Expenditures for LT1FBR and Current Drinking Water Programs .... 7-8
Exhibit 7-5.    Number and Percent of Public and Private System, by Size of System	7-12
Exhibit 7-6.    Small Entities Affected by the Turbidity Monitoring and Turbidity Treatment
              Provisions of LT1FBR	7-12
Exhibit 7-7.    Small Entities Affected by the Benchmarking Provisions of LT1FBR	7-12
Exhibit 7-8.    Small Entities Affected by the Filter Backwash Recycle Provisions of LT1FBR  . . 7-13
Exhibit 7-9.    Results of Comparison of Mean Sales, Revenues, and Operating
              Expenditures to Costs 	7-14
Exhibit 7-10.  Summary of Respondents, Responses, Burden, and Costs for PWSs and
              States for the ICR Approval Period	7-21
Exhibit 8-1.    Summary of Annual Mean Benefits Associated with Avoided Illnesses and Mortalities
              for the Turbidity Provisions of the Proposed Rule 	8-2
Exhibit 8-2.    Total Annual Costs for Two Combinations of Alternatives	8-2
Exhibit 8-3.    Comparison of Annual Benefit, Cost, and Net Benefit Ranges	8-3
Exhibit 8-4.    Summary of Benefit and Cost Analysis Completeness	8-4
Exhibit 8-5.    Breakeven Analysis Summary	8-6
Exhibit 8-6.    Combined Annual Number of Illnesses Remaining After LT1FBR and
              IESWTR	8-7
Exhibit 8-7.    Percent Reduction in Illnesses Needed to Break Even by Alternative	8-8
Exhibit 8-8.    Annualized Costs for Recent and Forthcoming Rules that Affect Surface Water
              and GWUDI Drinking Water Systems   	8-9
RIAfor the Proposed LT1FBR
February 15, 2000

-------
                               LIST OF APPENDICES







Appendix A:  Benefits Analysis Assuming 2.0 Baseline Log Removal




Appendix B:  Benefits Analysis Assuming 2.5 Baseline Log Removal




Appendix C:  Compliance Forecasts and Total Capital Costs of Treatment




Appendix D:  Annualized Capital and O&M Costs at 3 Percent Cost of Capital




Appendix E:  Annualized Capital and O&M Costs at 7 Percent Cost of Capital




Appendix F:  Annualized Capital and O&M Costs in Cents per Kilo-gallon




Appendix G:  State and Utility Monitoring Activities and Costs




Appendix H:  Household Costs

-------
                             1.  Executive Summary

1.1    Need for the Proposal

Drinking water contamination is one of the most important environmental risks, and disease-causing
microbial contaminants (i.e., bacteria, protozoa, and viruses) are probably the greatest remaining health
risk management challenge for drinking water suppliers according to EPA's Science Advisory Board
(SAB), an independent panel of experts established by Congress (U.S. EPA/SAB, 1990).  The
proposed Long Term 1 Enhanced Surface Water Treatment and Filter Backwash Rule (LT1FBR)
undertakes the challenge to improve the control of microbial pathogens such as Cryptosporidium in
public drinking water systems.

Cryptosporidium., which is common in the environment, is transported in watersheds from sources of
oocysts (e.g., agricultural runoff and untreated wastewater) to water bodies that serve as drinking water
sources.  If system treatment operates inefficiently, oocysts may enter finished water at levels that pose
health risks.  Cryptosporidium is of particular concern to EPA as it develops the LT1FBR
because—unlike pathogens such as viruses and bacteria—it is difficult to inactivate Cryptosporidium
oocysts using standard disinfection practices. Therefore, the control of Cryptosporidium is dependent
on physical removal processes.  Other emerging disinfection-resistant pathogens such as
Microsporidia, Cyclospora, and Toxoplasma are also a concern for similar reasons.

Cryptosporidiosis is the disease caused by ingesting Cryptosporidium oocysts. Dupont, et al. (1995)
found that a dose of even a few C. parvum oocysts is sufficient to cause infection in healthy adults.
Cryptosporidiosis is a common protozoa! infection that usually causes 7 to 14 days of diarrhea with
possibly a low-grade fever, nausea, and abdominal cramps in individuals with healthy immune systems
(Juranek, 1998).  There is currently no therapeutic cure for Cryptosporidiosis, but the disease is
self-limiting in healthy individuals.  It does, however, pose serious health and mortality risks for sensitive
subpopulations including children, the elderly, pregnant women, and the immunocompromised1 (Gerba
et al., 1996;  Payer and Ungar, 1986; U.S. EPA  1998a), which represents almost 20 percent of the
population in the United States (Gerba et al., 1996).

Cryptosporidium oocysts in drinking water treated by small systems pose both an endemic and an
epidemic health risk.  The nature of endemic risks prevents the resulting illnesses from appearing in
databases that track waterborne diseases. Consequently, there are no data to determine the extent of
the endemic health impact in the United States. Evidence on epidemic risk, however, suggests that
improving small  system performance will generate health benefits. Between 1984 and  1994, six of the
ten documented epidemics associated with drinking water systems occurred in systems serving fewer
        For instance, a follow-up study of the 1993 Milwaukee waterborne disease outbreak reported that at least
50 Cryptosporidium-associated deaths occurred among the severely immunocompromised (Hoxie et al., 1997).

      RIA for the Proposed LT1FBR                   1-1                          February 15,2000

-------
than 10,000 people (Moore et al., 1993; Kramer et al., 1996; Levy et al., 1998; Craun et al., 1996;
Craun, 1998).2

The two primary methods for treating drinking water for microbial contaminants are chemical
disinfection (inactivation) and physical removal. The main goal of LT1FBR, which is discussed in
greater detail in Chapter 2, is to improve the physical removal of microbial contaminants through the
enhancement of turbidity treatment processes and management of recycle practices.  Chemical
disinfection is also addressed through a disinfection benchmark provision that maintains microbial
protection levels while systems alter disinfection practices to reduce health risks associated with
disinfection byproducts.

The strategy of the proposed rule is to enhance turbidity treatment practices at small systems that use
filtration and to address risks posed by recycle practices. EPA believes that improved finished water
turbidity levels are indicative of improved physical removal and, therefore, reduced risk of
Cryptosporidium-related illnesses.  Recycle practices are of concern because recycle of flows such as
filter backwash and thickener supernatant within the treatment process can potentially return a
significant number of oocysts to the treatment plant in a short amount of time, particularly if the recycle
is returned to the treatment process without prior treatment, equalization, or some other type of
hydraulic detention. Should recycle  disrupt normal treatment operations or should treatment not
function efficiently due to other deficiencies, high concentrations of oocysts may pass  through the plant
into finished drinking water.

As  a result, the proposed rule addresses two requirements of the Safe Drinking Water Act (SDWA) as
amended in  1996. Those amendments established a number of regulatory deadlines, including
schedules for a Stage 1 and a Stage 2 Disinfection Byproduct Rule, two stages of the Enhanced
Surface Water Treatment Rule 1412(b)(2)(C), and a requirement that EPA promulgate regulations to
"govern" filter backwash recycling within the treatment process of public utilities (Section 1412(b)(14)).
The proposed LT1FBR is the second part of the first stage of the Enhanced Surface Water Treatment
Rule. The other part, the Interim Enhanced Surface Water Treatment Rule (IESWTR) promulgated in
December 1998, established requirements to improve control of microbial pathogens in public water
systems serving 10,000 or more people that use surface water or GWUDI. The LT1FBR extends
those requirements to small systems  and also addresses the filter backwash recycling requirement.

The proposed LT1FBR applies to public drinking water systems using surface water or ground water
under the direct influence of surface  water (GWUDI) as a source and serving fewer than 10,000
people, with the exception of a recycle control provision that also applies to large systems (i.e., systems
serving 10,000 or more people).  LT1FBR builds on the 1989  Surface Water Treatment Rule (SWTR)
(54 FR 27486, June 19, 1989). It will also protect the public against increases in microbial risk as
systems alter their disinfection  practices to meet new disinfection byproduct (DBF) standards
        The documented outbreaks do not account for all of the epidemic illnesses. The number of identified and
reported outbreaks in the Centers for Disease Control database represents a small percentage of actual waterborne
disease outbreaks because there are numerous opportunities for the reporting framework to fail to register an
outbreak (National Research Council, 1997; Bennett et al., 1987).

      RIA for the Proposed LT1FBR                  1-2                           February 15,2000

-------
promulgated under the Stage 1 Disinfectants/Disinfection Byproducts Rule (Stage 1 DBPR) (63 FR
69477, December 16, 1998). Chapters 2 and 3 describe the provisions of the proposed rule in detail;
briefly, they are:

•      2-log Cryptosporidium removal requirements for systems that are required to filter under the
       SWTR

•      Strengthened combined filter effluent (CFE) turbidity performance standards of 1.0 NTU as a
       maximum and 0.3 NTU as the 95th percentile monthly, based on continuous monitoring for
       systems using conventional or direct filtration, and related requirements for systems using other
       filtration technologies

•      Requirements for individual filter turbidity monitoring for plants using conventional or direct
       filtration

•      A disinfection benchmark provision with applicability monitoring, profiling, and benchmarking
       components to insure that microbial protection is not undercut as facilities take the necessary
       steps to comply with new disinfection byproduct standards

•      Inclusion of Cryptosporidium in the definition of GWUDI systems and in the watershed control
       requirements for unfiltered public water systems

•      A requirement that new finished water storage facilities are covered.

       Reporting requirements and recycle practice modifications for systems that practice recycle

1.2    Consideration of Regulatory Alternatives

The public health decision-making process is particularly demanding for this rule because it primarily
affects small drinking water systems, which have limited capabilities for investing in new treatment
technologies or maintaining complex monitoring regimes compared to large systems. Nevertheless,
their customers are entitled to health protection comparable to the protection afforded to large system
customers by the IESWTR (63 FR 69389, December 16, 1998). Thus, EPA has carefully weighed
the feasibility  of small system implementation against the public health risks posed by microbial
contaminants to customers of small systems throughout the rule development process. EPA
incorporated stakeholder inputs from small systems operators, consumers, and States with the
substantive elements of the IESWTR to design a proposed rule that provides small system customers
with a comparable level of protection while minimizing the costs to small systems.

To reduce the potential burden of the proposed LT1FBR on small systems, EPA developed and
evaluated the cost implications of several regulatory alternatives for the following provisions: individual
filter monitoring, disinfection benchmark applicability monitoring, disinfection benchmark profiling, and
recycle practice.  EPA started with the regulatory framework for the IESWTR and worked with

      WA for the Proposed LT1FBR                  1-3                          February 15,2000

-------
stakeholder groups to determine how to make the requirements less burdensome for small systems
without compromising the health benefits of the proposed rule.

Chapter 3 discusses the alternatives in detail and describes why EPA selected the preferred alternatives
for the proposed rule. Exhibit 1-1 compares the preferred LT1FBR alternatives with their IESWTR
counterparts. With the exception of the recycle provision, which is new under the proposed rule, the
comparison shows that the preferred LT1FBR alternatives will place less data collection and data
analysis burden on small systems.

     Exhibit 1-1.  Key Differences Between the Preferred Alternatives for the Proposed
                         LT1FBR and the IESWTR Requirements
Provision
Individual filter monitoringC
events requiring an exceptions
report1
Disinfection benchmarkC
applicability monitoring
Disinfection benchmarkC
profile development
Preferred LT1FBR Alternative
Individual filter turbidity exceeds
1 NTU in two consecutive
measurements
Optional TTHM and HAAS sample at
maximum residence point during the
month of warmest water temperature
Total samples: 1 (optional)
Collect data weekly for 1 year
Total profile data points: 52
IESWTR Requirements
Individual filter turbidity exceeds
1 NTU in two consecutive
measurements
Individual filter turbidity exceeds
0.5 NTU in two consecutive
measurements after first 4 hours
of filter operation
TTHM and HAAS samples at four
locations in each of 4 quarters
Total samples: 16
Collect data daily for 1 year
Total profile data points: 365
 1. The proposed LT1FBR and the IESWTR have comparable filter self assessment and comprehensive performance evaluation
 requirements.

1.3    Baseline Analysis

Each provision of the proposed LT1FBR affects a different subset of surface water or GWUDI
systems, and Chapter 4 discusses how EPA estimated the number of affected systems by
provision. For example, the proposed LT1FBR establishes new combined filter effluent
requirements for surface water and GWUDI systems that serve fewer than 10,000 people and filter
drinking water, and an individual filter monitoring requirement that applies to the subset of systems
using conventional and direct filtration.  The LT1FBR also establishes management of recycle
flow requirements, which apply to all surface water and GWUDI systems using rapid granular
filtration regardless of size.

The methods and sources used to identify the number of systems that are included under each
provision are detailed in Chapter 4.  Exhibit 1-2 provides the system size categories and the
number of systems that may be affected by the provisions of the LT1FBR.  The number of systems
using filtration that may be affected by one or more of the turbidity provisions was developed from
     RIAfor the Proposed LT1FBR
1-4
February 15,2000

-------
an analysis of current finished water turbidity, and an analysis of the types of filtration employed
(e.g., conventional, direct, slow sand, and membranes).  The number of systems that will have to
perform applicability monitoring and/or develop disinfection profiles and establish a benchmark is
based on the type of drinking water system (i.e., community, nontransient noncommunity) and
chronic exposure. Estimates of systems that may be regulated under the recycle provision were
developed from an analysis of primary water treatment and recycling practices.

                Exhibit 1-2.  Number of Systems Under LT1FBR Provisions
System
Size
• 100
101B500
50181,000
1,00183,300
3,30189,999
10,000850,000
50,0018100,000
100,00181,000,000'
Total
Turbidity
Provisions
2,201
2,031
1,109
2,150
1,643
N/A
N/A
N/A
9,133
Applicability Monitoring
and Disinfection
Benchmarking
1,404
2,333
1,301
2,553
1,859
N/A
N/A
N/A
9,450
Recycle
Provisions
502
670
486
993
887
830
141
127
4,636
 1. This system estimate does not include seven individual plants that belong to systems serving more than one million people,
 which were included in the cost analysis because they may be affected by the recycle provisions.


1.4    Benefits of the LT1FBR

Chapter 5 provides EPA's analysis of potential benefits of the proposed rule.  According to the
risk assessment performed for this RIA, the turbidity provisions in the proposed LT1FBR are
estimated to reduce the mean annual number of Cryptosporidium illnesses by improving filtration
in drinking water treatment plants serving fewer than  10,000 people. The risk assessment predicts
that improved filtration will result in mean reductions of 22,800 to 83,600 annual illnesses
depending on which of the scenarios describing baseline removal (2.0 or 2.5 log) and improved
Cryptosporidium removal (low-, mid-, or high-removal) is assumed. Based on these
cryptosporidiosis reductions, the mean annual estimated benefits from reducing illnesses are
between  $53.9 million and $199.5 million per year. This calculation assumes a mean cost of
illness of approximately $2,400 per illness.
     RIA for the Proposed LT1FBR
1-5
February 15,2000

-------
The LT1FBR risk assessment also predicts a mean annual reduction of 3 to 10 mortalities by
improving filtration practices, depending on the baseline removal and improved removal
assumptions.  These annual mortality reductions produce benefits in the range of $16.2 million to
$59.8 million, based on a mean value of $5.7 million per statistical life saved. Exhibit 1-3
summarizes annual benefits accruing from the LT1FBR turbidity provisions.

      Exhibit 1-3. Summary of Annual Benefits  Associated with Avoided Illnesses and
              Mortalities for the Turbidity Provisions (January 1999 dollars)
Improved Log-Removal Assumption
Low Removal
Avoided Illnesses
Mortalities
Total
Mid Removal
Avoided Illnesses
Mortalities
Total
High Removal
Avoided Illnesses
Mortalities
Total
Daily Drinking Water Ingestion and
Baseline Cryptosporidium Log-Removal Assumptions
(S millions)
Mean = 1.2 Liters per person
2.0 log
$150.3
$45. 0
$195.3
$185.3
$55.5
$240.8
$199.5
$59.8
$259.4
2.5 log
$53.9
$16.2
$70.1
$66.2
$19.9
$86.1
$71.1
$21.3
$92.4
 Totals may not equal detail due to rounding.


The calculated turbidity benefits are the lower bound of total benefits accruing from the proposed
LT1FBR. Additional nonquantified benefits come from the recycle provisions for small and large
drinking water systems, the disinfection benchmark provision, the requirement that all new finished
water reservoirs be covered, and the inclusion of Cryptosporidium in the definition of GWUDI
and the watershed control requirements for small unfiltered systems. These components combine
to reduce health effects to sensitive subpopulations, reduce the risk of outbreaks, enhance aesthetic
water quality, and minimize expenditures associated with averting behavior, as well as reduce risk
from other pathogens (e.g., Giardia lamblid). Data were not available to quantify benefits for
these categories and provisions; however, qualitative analysis suggests that these benefits will be
positive and significant. In particular, the recycle provisions will prevent the accumulation of
Cryptosporidium within the treatment plant and minimize the risk of oocysts entering finished
water by improving filter performance and reducing hydraulic disruptions.


1.5    Costs of the LT1FBR


Chapter 6 summarizes the methods EPA used to analyze costs for the proposed rule. EPA
estimates that the annualized cost of the preferred alternatives for the proposed rule will be $87.6
or $97.5 million depending on the discount rate.  These estimates include capital costs for turbidity
     RIAfor the Proposed LT1FBR
1-6
February 15,2000

-------
treatment and recycle practice changes and start-up labor costs for monitoring and reporting
activities that have been annualized assuming either a 3 percent or a 7 percent discount rate over a
20-year period.  They also include annual operating and maintenance costs for turbidity treatment
and recycle practice changes and annual labor for turbidity monitoring activities. The labor costs
incorporate both system and State burden estimates.

Exhibit 1-4 summarizes costs by provision for the set of preferred alternatives.  Costs for the
turbidity provisions, which include treatment changes to meet the revised CFE requirements and
individual filter monitoring  activities, account for 70 or 72 percent of total costs depending on the
discount rate assumption.  The recycle provisions, which include costs for some systems moving
their recycle return location, direct recycle systems conducting a self assessment, and direct
filtration reporting their recycle practices account for 23 to 25 percent of total costs.3  System
expenditures for all provisions are approximately 93 percent of total costs; State expenditures make
up the remainder.
           Exhibit 1-4. Total Annual Costs for Two Combinations of Alternatives
                                   (January 1999 dollars)
Compliance
Activity
Turbidity Provisions
Disinfection Benchmarking
Covered Finished Storage
Recycle Provisions
Total Costs
Preferred Alternatives
(S millions)
3%
$63.4
$1.3
$2.5
$20.4
$87.6
7%
$68.6
$1.8
$2.6
$24.5
$97.5
IESWTR Alternatives
($ millions)
3%
$116.7
$5.9
$2.5
$20.4
$145.5
7%
$121.9
$8.3
$2.6
$24.5
$157.3
 Detail may not add to total due to independent rounding.


To reduce the potential cost to small systems, EPA developed and evaluated the cost implications
of several regulatory alternatives for the following provisions: turbidity monitoring, disinfection
benchmark applicability monitoring, disinfection benchmark profiling, and recycle practice. Many
of the alternatives reduce the labor burden on small systems relative to what it would be if the
proposed rule incorporated the same requirements as the IESWTR. Exhibit 1-4 also reports what
the costs would be under alternatives similar to IESWTR requirements. Comparing these costs
with the costs of the preferred alternatives shows that the preferred alternatives reduce total costs
by approximately 38 to 40 percent, primarily by reducing the labor burdens associated with
individual filter monitoring activities.
        The recycle cost estimate includes indirect capital and operating and maintenance costs based on EPA's
estimates of how many direct recycle and direct filtration systems may be required to alter their recycle practices.
This represents the high range of EPA cost estimate for the preferred recycle alternative.
     RIAfor the Proposed LT1FBR
1-7
February 15,2000

-------
Increased drinking water production costs may be passed on to consumers, including households,
in the form of higher fees. EPA estimated the potential impact on households by developing a
distribution of costs across all affected systems and converting that distribution to a per-household
basis. Approximately 6.6 million households could be affected by the turbidity, disinfection
benchmarking, and covered finished water storage provisions. EPA estimates that the mean
annual incremental cost per household would be $8.66, and the incremental annual cost would be
less than $10 for approximately 86 percent of households and less than $120 (i.e., $10 per month)
for 99 percent of households. The recycling provisions could affect approximately 12.9 million
households based  on EPA's assumptions, and the mean cost per household is $1.79.  Annual
incremental cost would be less than $10 for about 99 percent of households and less than $120 for
99.9 percent of households.


1.6    Economic  Impact Analysis

As part of the rule promulgation process, EPA is required to perform a series of distributional
analyses that address the potential regulatory burden placed on entities that are directly or indirectly
effected by the rule.  The distributional impacts considered were the cost of compliance for State,
local, and Tribal governments, and small businesses; the effect of rule implementation on sensitive
subpopulations; and  the potential for disproportionate impacts on low-income and minority
populations.  Chapter 7 discusses EPA's economic impact analyses and findings.

A distributional impact analysis was performed as part of the requirements under the Unfunded
Mandate Reform Act.  The analysis looked at budgetary impacts across  system sizes as well as
geographically over the United States. Across system sizes, the greatest impact in terms of
cost-to-revenue ratios would be on those systems serving 500 or fewer people, although over 70
percent of system  costs will accrue to systems serving more than  1,000 people.

For States, budgetary impacts were considered using three different perspectives: annual
compliance costs per State; the percentage increase in drinking water program costs per State; and
State per capita drinking water program expenditures. The evidence from the three perspectives of
budgetary impacts does not suggest that there would be a disproportionate budgetary effect
resulting from the rule.  There is no evidence of a geographic concentration of higher impact. Nor
does any one State consistently show relatively high impacts across all three analyses.

The Regulatory Flexibility Act, as amended by the Small Business Regulatory Enforcement
Fairness Act (SBREFA), requires the EPA to consider the financial impact of LT1FBR on small
business entities. EPA conducted an analysis of the budgetary impact of the rule on three different
categories of small system ownership: public, private, and non-profit. Using assumptions
regarding the costs that systems of various sizes will incur to comply with the LT1FBR rule, EPA
was able to generate a hypothetical distribution of per-system costs.  EPA compared this cost
distribution to financial data to determine how many systems might incur costs in excess of 3
percent or 1 percent  of revenue. The results of the analysis suggest that a significant number of
small systems will be substantially impacted by the rule. Consequently, EPA prepared an Initial
Regulatory Flexibility Analysis, which is included in Chapter 7.

     WA for the Proposed LT1FBR                 1-8                          February 15,2000

-------
A primary purpose of the proposed LT1FBR is to improve control of microbial pathogens,
specifically the protozoan Cryptosporidium.  Under Executive Order 13045, EPA must evaluate
the environmental health or safety effects of the planned rule on children, and explain why the
planned regulation is preferable to other potentially effective and reasonably feasible alternatives
considered by EPA. Because of the severity of illness and high costs for treatment experienced by
sensitive subpopulations—including young children—as a result of Cryptosporidium infection,
LT1FBR is expected to have a disproportionately positive impact on children.

As required under Executive Order 12898, EPA must identify and address disproportionately high
and adverse human health or environmental effects the adoption of LT1FBR may have on
minority and low-income populations.  The Agency has considered environmental justice-related
issues concerning the potential impacts of this action and has consulted with minority and
low-income stakeholders. Furthermore, the proposed LT1FBR extends the types of health risk
reductions achieved by the 1998 IESWTR to consumers served by small public water systems.
Therefore, the minority and impoverished populations served by small systems will realize the
health protection benefits currently provided to populations served by larger systems.

1.7    Weighing  of Benefits and Costs

For the comparison of benefits and costs in Chapter 8, EPA subtracted the incremental costs of the
proposed rule from the incremental benefits to obtain net benefits.  Assuming incremental
annualized costs can be represented by the costs for the preferred alternatives shown in Exhibit
1-4 ($87.6 million to $97.5 million across discount rates) and incremental benefits can be
represented by the range in Exhibit 1-3 (mean values of $70.1 million to $259.4 million across the
removal assumptions), net benefits potentially range from a negative value of $27.4 million to
positive value of $171.8 million. The low net benefit estimate equals the low benefit minus the
high cost estimate.  Conversely, the high net benefit estimate is the difference between the high
benefit estimate and the low cost estimate. The chart in Exhibit 1-5  compares these ranges, and
shows that the range of potential net benefits lies primarily in the positive quadrant.  Thus, benefits
will most likely exceed costs.
     RIA for the Proposed LT1FBR                  1-9                         February 15,2000

-------
          Exhibit 1-5. Comparison of Annual Benefit, Cost, and Net Benefit Ranges
                               (January 1999 dollars, millions)
         Net Benefits
                Costs
             Benefits
                      -$50    $0    $50   $100  $150   $200   $250  $300
Although this analysis suggests that monetized net benefits may be as low as negative $27.4 million for
the preferred alternatives, total social net benefits are likely to be positive taking the qualitative benefits
into consideration.  Costs were estimated for most of the provisions of the proposed LT1FBR, but
benefits were only estimated for the turbidity provisions; EPA did not quantify benefits associated with
three provisions that accounted for approximately one third of total costs: recycle practices, disinfection
benchmarking, and covered finished storage. Furthermore, the quantified benefits for the turbidity
provisions do not include categories of benefits such as reducing exposure to other pathogens (e.g.,
Giardia lamblid) and avoiding the cost of averting behavior.  The nonquantified benefits could
represent substantial additional economic value.  Overall, EPA expects the rule to provide benefits for
more than 18 million households.  If the aggregate benefit per household for these nonquantified benefits
is at least $1.52 per year, then even the low range of net benefits will be positive.

This RIA provides background on the proposed rule, summarizes the key components, discusses
alternatives to the proposed rule, and estimates costs and benefits to the public and State governments.
     RIA for the Proposed LT1FBR
1-10
February 15,2000

-------
                             2.  Need for the Proposal

This document analyzes the impact of the proposed Long Term 1 Enhanced Surface Water
Treatment and Filter Backwash Rule (LT1FBR). Executive Order 12866, Regulatory Planning
and Review, requires EPA to estimate the costs and benefits of regulations in a regulatory impact
analysis (RIA) and to submit the analysis in conjunction with publishing the proposed rule.

The proposed Long  Term 1  Enhanced Surface Water Treatment and Filter Backwash Rule applies
to public drinking water systems using surface water or ground water under the direct influence of
surface water (GWUDI) as a source and serving fewer than 10,000 people, with the exception of
the recycle provisions, which also apply to large systems (systems serving 10,000 or more people).
LT1FBR builds on the 1989 Surface Water Treatment Rule (SWTR) (54 FR 27486, June 29,
1989). LT1FBR will improve control of microbial pathogens such as Cryptosporidium as well as
assure there will be no significant increase in microbial risk for those systems that may need to
change their disinfection practices in order to meet new disinfection byproduct (DBF) standards
under Stage  1 Disinfectants/Disinfection Byproducts Rule (Stage 1 DBPR) (63 FR 69389,
December 16, 1998).

This RIA provides background on the proposed rule, summarizes the key components, discusses
alternatives to the proposed rule, and estimates costs and benefits to the public and private sectors.
This chapter summarizes the technical and regulatory issues associated with the need for the
proposed rule. It explains the nature of surface water treatment for microbial pathogens, identifies
the public health concerns addressed by the proposed rule, and summarizes the key components of
the proposed rule.

Chapters 3 through 8 are intended to meet the requirements of Executive Order 12866 by
responding to specific analytical questions.  Chapter 3 reviews alternative approaches considered
as the proposed rule  was being developed.  Chapter 4 presents public water system (PWS) data to
establish a baseline of information for use in the following four chapters. Chapter 5 examines the
proposed rule's potential benefits, reviewing occurrence data, treatment  efficiencies and dose
response relationships.  Chapter 6 presents an estimate of the costs to implement the proposed rule.
Chapter 7 reviews the  distribution of the costs and benefits of the proposed rule on various entities
and subpopulations.  Chapter 8 weighs the overall benefits and costs of the various alternatives
considered for the proposed rule.

2.1    Description  of the Issue

The primary issue of concern being addressed by LT1FBR is improved  public health protection
through the treatment of drinking water for microbial contaminants. The two primary methods for
treating drinking water for microbial contaminants are chemical  disinfection (inactivation) and
physical removal. Chemical disinfection has been addressed by several  other regulations and is
not the primary issue being addressed by LT1FBR. The main goal of LT1FBR is to improve the
physical removal of  microbial contaminants by enhancing filtration and other physical removal

February 15, 2000                              2-1                       RIA for the Proposed LT1FBR

-------
processes and management of recycle practices to protect these physical treatment processes.

The types of water treatment plants or unit operations that may be used to physically remove
particles include: conventional treatment, direct filtration, package plants, softening, solids contact
clarification, slow sand filtration, and diatomaceous earth filtration. LT1FBR does not apply to the
latter two types of treatment.

Conventional treatment, the most widely used plant type, consists of chemical coagulation, rapid
mix, flocculation, and sedimentation followed by filtration.  A general flow schematic for a
conventional water treatment plant is presented in Exhibit 2-1.  Source water is treated with
chemical coagulant(s), such as aluminum sulfate (alum), ferric or ferrous sulfate, ferric chloride,
and/or a coagulant aid to destabilize suspended particles and improve sedimentation as the water
enters the treatment system. Coagulant aids promote the attachment of suspended particles to the
polymers, and coagulate to form a heavy floe that is easily removed in the settling process. The
flow is then subjected to rapid mixing that blends the coagulant into the raw water. In the
flocculation step, coagulated water is gently stirred to allow particles to collide and combine to
form larger particles.  This produces a dense and readily settleable floe. The flocculated water
flows into a sedimentation basin where the dense floe settles over time leaving clarified water
above it.  Sedimentation should provide a high level of particulate removal and significantly reduce
the turbidity level. This clarified water is filtered to remove particles or turbidity that remains after
sedimentation.
         Exhibit 2-1. Flow Schematics for Conventional Water Treatment Systems
  Raw Water
                  Coagulants
Rapid M ix


Flocculation






Filtration
In the direct filtration process, suspended solids are removed solely through filtration process
(AWWA/ASCE, 1998).  As depicted in Exhibit 2-2, direct filtration consists of coagulation
followed by rapid mixing, flocculation, and filtration.  Unlike conventional treatment, the
chemically conditioned and flocculated water is applied directly to the filters. No separate
sedimentation process is used in direct filtration.  A variation of direct filtration, in-line filtration,
excludes the flocculation process and instead relies on flocculation to occur in the piping between
the rapid mix and the filters. In both direct filtration and in-line filtration, the filters are the only
means of suspended solid, particle, and pathogen removal.
RIAfor the Proposed LT1FBR
2-2
February 15, 2000

-------
                 Exhibit 2-2.  Flow Schematics for Direct Filtration Systems
          Raw Water
                           Coagulants
Rapid Mix


Flocculation


Filtration
                                          (a) Direct Filtration
                           Coagulants
          Raw Water
                                          (b) In-Line Filtration
Small surface water systems may use several other technologies.  Some small surface water
systems are package plants.  Package plants can be defined as a complete modular treatment plant,
designed as a factory assembled, skid mounted unit.  A complete modular treatment plant typically
consists of chemical coagulation, flocculation, settling and filtration.  Most modular systems utilize
high rate treatment processes, with shorter detention time than that of custom-engineered
conventional treatment plants. Another type treatment used by some systems is the softening plant.
Softening plants utilize the same basic treatment process as conventional treatment plants, except
that they also remove hardness (calcium and magnesium ions) through precipitation, followed by
solids removal. In the contact clarification treatment process, the flocculation and sedimentation
(and often the rapid mix) processes are combined in one unit, that being an upflow solids contactor
or contact clarifier.  In addition, small systems may also utilize microfiltration and bag and
cartridge filters to remove turbidity.

In the treatment processes just discussed, pathogenic microorganisms are removed during the
sedimentation and/or filtration processes in a water treatment plant. As Exhibit 2-3  shows, recycle
streams generated during treatment, such as spent filter backwash water, liquids from dewatering,
or thickener supernatant are often concentrated and returned to the treatment train. These recycle
streams, therefore, may contain high concentrations of pathogens, including disinfection-resistant
Cryptosporidium oocysts, in addition to  chemicals added during the stages of the treatment process
February 15, 2000
2-3
RIAfor the Proposed LT1FBR

-------
(e.g., oxidants, softeners, coagulants, polymers).  If the recycle enters the treatment train after the
point of primary coagulation, the recycle can degrade the treatment process, by causing an
inappropriate chemical dose, hydraulic surge and potentially overwhelming the plant's multi-
barriers with a large concentration of pathogens.

                   Exhibit 2-3.  Flow Schematics for Systems that Recycle
    Raw Water
    Raw Water
                                           Direct Recycle
                    Coagulants
/ >
f
Rapid Mix
k



h

Sedimentation
k

Filtration
                  .                     -^	  Recycle Flows             I                  I
                                            Treated Recycle
                     Coagulants
                               Recycle Flows
                  I	
  Equalization
     or
 Sedimentation
RIAfor the Proposed LT1FBR
2-4
February 15, 2000

-------
2.2    Public Health Concerns

In 1990, EPA's Science Advisory Board (SAB), an independent panel of experts established by
Congress, cited drinking water contamination as one of the most important environmental risks and
indicated that disease-causing microbial contaminants (i.e., bacteria, protozoa and viruses) are
probably the greatest remaining health risk management challenge for drinking water suppliers
(U.S. EPA/SAB, 1990). Information on the number of waterborne disease outbreaks from the
U.S. Centers for Disease Control and Prevention (CDC) underscores this concern.  CDC indicates
that 401 waterborne disease outbreaks were reported between 1980 and 1996, with over 750,000
associated cases of disease (CDC, 1996).

2.2.1  Contaminants and Their Associated Health Effects

Waterborne disease caused by Cryptosporidium is of particular concern to the LT1FBR, as
Cryptosporidium oocysts are not inactivated with standard disinfection practices (unlike pathogens
such as viruses and bacteria), and there is currently no therapeutic cure for cryptosporidiosis
(unlike giardiasis).  Cryptosporidium is not generally inactivated in systems using standard
disinfection practices, therefore, the control of Cryptosporidium is entirely dependent on physical
removal processes. Other emerging disinfection resistant pathogens, such as microsporidia,
Cyclospora, and Toxoplasma are also a concern of LT1FBR for similar reasons.

Waterborne disease is usually acute (i.e., sudden onset and typically lasting a short time in healthy
people).  Some pathogens (e.g., Giardia and Cryptosporidium) may cause extended illness, lasting
weeks  or longer in otherwise healthy individuals, and the infection can prove fatal for sensitive
populations such as the immunocompromised. Most waterborne pathogens cause gastrointestinal
illness, with diarrhea, abdominal discomfort, nausea, vomiting, and/or other symptoms. Other
waterborne pathogens cause, or at least are associated with, more serious disorders such as
hepatitis, gastric cancer, peptic ulcers, myocarditis, swollen lymph glands, meningitis, encephalitis,
and many other diseases.

Cryptosporidiosis is caused by ingestion of Cryptosporidium oocysts, which are readily carried by
the waterborne route.  The most common source of oocysts in water is the feces of infected hosts
(Walker et al., 1998).  Dupont, et al. (1995) found through a human feeding study that a low dose
of C. parvum is sufficient to cause infection in healthy adults. Infected humans and other animals
may excrete Cryptosporidium oocysts, which can then be transmitted to others. Transmission of
cryptosporidiosis often occurs through ingestion of the infective oocysts from contaminated food or
water,  but may also result from direct or indirect contact with infected persons or  animals
(Casemore, 1990; Cordell and Addiss, 1994).

Cryptosporidiosis is a common protozoal infection that usually causes 7-14 days of diarrhea with
possibly a low-grade fever, nausea, and abdominal cramps in individuals with healthy immune
systems (Juranek, 1998).  There appears to be an immune response to Cryptosporidium, but it is
not known if this results in complete protection (Payer and Ungar, 1986). When  prior exposure or
chronic contamination of the water by low levels of oocysts confers short-term immunity to
February 15, 2000                               2-5                       RIA for the Proposed LT1FBR

-------
immunocompetent residents in a community (Okhuysen et al., 1998), most cases of symptomatic
illness in that community will then occur in newly exposed individuals, such as young children,
visitors, and new residents (Frost et al., 1997).

2.2.2  Sensitive Subpopulations

There are a number of sensitive populations that are at greater risk of serious illness (morbidity) or
mortality from either epidemic or endemic infection by Cryptosporidium pathogens than is the
general population (Frost et al., 1997). In sensitive populations, gastrointestinal illness caused by
cryptosporidiosis may be chronic. These sensitive populations include children, especially the very
young; the elderly; pregnant women; and the immunocompromised.  This sensitive segment
represents almost 20  percent of the population in the United States (Gerba et al.,  1996; Payer and
Ungar, 1986; U.S. EPA 1998b).

EPA has a particular concern regarding drinking water exposure to Cryptosporidium, especially in
severely immunocompromised persons, because there is no effective therapeutic drug to cure the
disease (Framm and Soave, 1997). Therefore, prevention of infection is critical (Petersen, 1992).
The severity and duration of illness is often greater in immunocompromised persons than in
healthy individuals and may be fatal among this population. For instance, a follow-up study of the
1993 Milwaukee waterborne disease outbreak reported that at least 50  Cryptosporidium-
associated deaths occurred among the severely immunocompromised (Hoxie et al., 1997).

2.2.3  Sources of Contaminants

Cryptosporidium is common in the environment (Rose, 1997, Soave, 1995, LeChevallier et al.,
199 la).  Runoff from unprotected watersheds allows the transport of these microorganisms from
sources of oocysts (e.g., feces of wildlife, untreated wastewater, and  agricultural runoff) to water
bodies used as intake sites for drinking water treatment plants. If treatment operates inefficiently,
oocysts may enter the finished water at levels of public health concern. Increasing disinfection
dosages (i.e., chlorine or chloramines) is not an effective strategy for controlling Cryptosporidium.,
because the Cryptosporidium  oocyst is especially resistant to disinfection practices.

Cryptosporidium oocysts have been detected in wastewater, pristine  surface water, surface water
receiving agricultural runoff or contaminated by sewage, ground water under the direct influence
of surface water (GWUDI), water for recreational use, and drinking  water (Rose, 1997; Soave,
1995). Over thirty environmental surveys have reported Cryptosporidium source water
occurrence data from surface water and GWUDI  (presented in Exhibits 2-4 and 2-5), which
typically involved the collection of a few water samples from a number of sampling locations
having different characteristics (e.g., polluted vs. pristine; lakes or reservoirs vs. rivers).

Each of the studies cited in Exhibits 2-4 and 2-5  presents Cryptosporidium source water
occurrence information, including: 1)  the number of samples collected, 2) the number of samples
positive,  and 3) both  the means and ranges for the concentrations of Cryptosporidium detected

RIA for the Proposed LT1FBR                       2-6                              February 15, 2000

-------
(where available).  However, the recovery and detection for Cryptosporidium in water samples
using the immunoflourescence assay method is limited.  Additionally, the method does not indicate
with certainty whether the oocysts detected are viable or infective to humans (Frey et al., 1997).
Despite these limitations, the occurrence information generated from these studies is valuable as a
measure of the incidence of Cryptosporidium in source waters. EPA compiled information on the
following source waters: rivers, reservoirs, lakes, streams, raw water intakes, springs, wells under
the influence of surface water, and infiltration galleries.

Finally, sedimentation and filtration processes in drinking water treatment plants remove
pathogenic microorganisms from the source water.  Recycle streams generated during water
treatment (e.g., spent filter backwash water, sedimentation basin sludge, or thickener supernatant)
are often concentrated and returned to the plant and combined with raw source water entering the
plant. Recycle combined with raw water can elevate the influent pathogen concentrations because
recycle streams may contain high concentrations of pathogens, including disinfection-resistant
Cryptosporidium oocysts. Fligh oocyst concentrations challenge the treatment process and
increase the risk of pathogen breakthrough to the finished water.
   Exhibit 2-4. Surface Water Survey and Monitoring Data for Cryptosporidium Oocysts
Sample
Source
Rivers
River
Reservoirs/Rivers
(polluted)
Reservoir (pristine)
Impacted river
Lake
Stream
Filtered water
Raw water
River (pristine)
River (polluted)
Lake/reservoir
(pristine)
Lake/reservoir
(polluted)
River
(all samples)
Protected drinking
water supply
Number of
Samples
(n)
25
6
6
6
11
20
19
82
85
59
38
34
24
36
6
Samples
Positive for
Cryptosporidiu
m (percent)1
100
100
100
83
100
71
74
27
87
32
74
53
58
97
81
Range of Oocyst
cone.
(oocysts/L)
2-112
2-5,800
0.19-3.0
0.01-0.13
2-1 122
0-22
0-240
0.001-0.48
0.07^184
NR
<0.001^142
NR
<0.001-3.82
0.15-0.45 (pristine)
10-63.5 (agricultural)
0.15-0.42
Mean
(oocysts/L)
25.1
l,920(a)
0.99(a)
0.02(a)
25(g)
0.58(g)
1.09(g)
0.015
2.7 (g)
detectable
0.29(g)
0.66(g)
0.093(g)
1.03(g)
0.2 (pristine)
18.3 (agricultural)
0.24(g)
Reference
Ongerth and Stibbs 1987
Madoreetal. 1987
Rose 1988
Rose 1988
Roseetal. 1988a2
Rose et al. 1988b2
Rose et al. 1988b2
LeChevallier et al. 1991b
LeChevallier et al. 1991c
Roseetal. 1991
Roseetal. 1991
Roseetal. 1991
Roseetal. 1991
Hansen and Ongerth 1991
Hansen and Ongerth 1991
February 15, 2000
2-7
RIAfor the Proposed LT1FBR

-------
   Exhibit 2-4. Surface Water Survey and Monitoring Data for Cryptosporidium Oocysts
Sample
Source
Pristine river, forestry
area
River below rural
community in forested
area
River below dairy
farming agricultural
activities
Reservoirs
Streams
Rivers
Finished water
(clearwell)
Finished water (filter
effluents)
Site 1 — River source
(high turbidity)
Site 1 — Filter effluent
Site 2 — River source
(moderate turbidity)
Site 2 — Filter
effluent
Site 3 — Reservoir
source(low turbidity)
Site 3 — Filter
effluent
Lakes
Streams
Finished water
River/lake
River/lake
River 1
River 2
Dairy farm stream
Number of
Samples
(n)
6
6
6
56
33
37
14
118
10
10
10
10
10
10
179
210
262
262
147
15
15
13
Samples
Positive for
Cryptosporidiu
m (percent)1
100
100
100
45
48
51
14
26
100
70
70
10
70
10
6
6
13
52
20
73
80
77
Range of Oocyst
cone.
(oocysts/L)
0.46-6.97
0.54-3.6
3.3-63.5
NR
NR
NR
NR
NR
0.82-71.9
0.01-0.04
0.42-5.1
0.005
0.77-8.7
0.02
0-22.4
0-20.0
0.0029-0.57
0.065-65.1
0.3-9.8
0-22.3
0-14.7
0-11.1
Mean
(oocysts/L)
1.62(g)
1.07(g)
10.72(g)
NR
NR
NR
NR
NR
4.8
NR
2.5
NA
2.5
NA
0.033 (median)
0.07 (median)
0.33 (detectable)
2.4 (detectable)
2.0
1.88 (a) all samples
0.43 (g) detected
1 .47 (a) all samples
0.61 (g) detected
1 .26 (a) all samples
0.55 (g) detected
Reference
Hansen and Ongerth 1991
Hansen and Ongerth 1991
Hansen and Ongerth 1991
Consonery et al. 1992
Consonery et al. 1992
Consonery et al. 1992
Consonery et al. 1992
Consonery et al. 1992
LeChevallier and Norton
1992
LeChevallier and Norton
1992
LeChevallier and Norton
1992
LeChevallier and Norton
1992
LeChevallier and Norton
1992
LeChevallier and Norton
1992
Archer et al. 1995
Archer et al. 1995
LeChevallier and Norton
1995
LeChevallier and Norton
1995
LeChevallier et al. 1995
States etal. 1995
States etal. 1995
States etal. 1995
RIAfor the Proposed LT1FBR
2-8
February 15, 2000

-------
    Exhibit 2-4.  Surface Water Survey and Monitoring Data for Cryptosporidium Oocysts
Sample
Source
Finished water
Reservoir inlets
Reservoir outlets
River (polluted)
Source water
First flush (storm
event)
Filtered (non
•storm event)
Grab (non
•storm event)
River 1
Stream by
dairy farm
River 2 (at
slant intake)
Settled water
(prefiltration)
Finished
Reservoirs (unfiltered
system)
Raw water intakes
Finished water
Raw water intakes
(rural)
Number of
Samples
(n)
1,237
60
60
72
NR
20
87
21
24
22
24
24
24
NR
148
155
NR
Samples
Positive for
Cryptosporidiu
m (percent)1
7
5
12
40
24
35
10
19
63
82
63
29
8 (confirmed)
13 (presumed)
37-52"
25
2.5
NR
Range of Oocyst
cone.
(oocysts/L)
NR
0.007-0.24
0.012-1.07
0.2-2.8
0.01-53. 93
0^117
0^1.2
0-6.5
0-14.7
0-23
0-22
0-0.35
0-0.006
0.1 5-0. 43 (maxima)4
0.0004-0.18
0.0002-0.008
0.4^1
Mean
(oocysts/L)
NR
0.019 (g)
0.016 (median)
0.061 (g)
0.6 (median)
0.24 (g)
7.4 (a)3
0.71 (g)3
NR
NR
NR
0.58 (g)
0.42 (g)
0.31 (g)
0-12(g)
0.005 (g)
0.008-0.0144
0.003
0.002
NR
Reference
Rosen etal. 19965
LeChevallier et al. 1997b
LeChevallier et al. 1997b
LeChevallier et al. 1997a
Swertfeger et al. 1997
Stewart etal. 1997
Stewart etal. 1997
Stewart etal. 1997
States etal. 1997
States etal. 1997
States etal. 1997
States etal. 1997
States etal. 1997
Okun et al. 1997
Consonery et al. 1997
Consonery etal. 1997
Swiger et al. 1998
 1. Rounded to nearest percent.
 2. As cited in Lisle and Rose (1995).
 3. Based on presumptive oocyst count
 4. Combined monitoring results for multiple sites in large urban water supply.
 5. As cited in States et al. 1997
 (a) = arithmetic average
 (g) = geometric average
 NR = not reported, NA = not applicable
February 15, 2000
2-9
RIAfor the Proposed LT1FBR

-------
          Exhibit 2-5. U.S. GWUDI Monitoring Data for Cryptosporidium Oocysts
Sample Source
Ground water well
jround water sources
(all categories)
Vertical wells
Springs
infiltration galleries
Horizontal wells
jround water sources
Spring-fed cistern (Potter
:o.,PA,Nov. 1991)
Spring (Spring Twp., Center
:o.,PA, May 1995)
Vertical well Lemont Well #4
Center Co., PA, Aug. 1992)
Vertical well (Boggs Twp.
Well #1, Center Co., PA,
\pr. 1997)
Vertical well (Boggs Twp.
Well #1, Center Co., PA,
\pr. 1997)
Vertical well (Douglas Co.,
3R, Feb. 1996, Feb. 1997)
nfiltration gallery (Salem,
3R, Jan. 1994 through
Sept. 1996)
ianney collector (St.
Helens, OR, May 1993
hroughMar. 1997)
Well (Marshfield, MA, June
1997)
Shallow, vertical well
BraymerWell#4,MO,
May 1995)
Number
of Samples
(n)
17
199 sites2
149 sites2
35 sites2
4 sites2
11 sites2
18
1
1
6
1
2
3
31
31
1
3
Samples Positive
for Cryptosporidium
Oocysts (percent)
(1 sample)
II2
52
202
502
452
5.61
100
100
66.7
100
50
33.3
35.5
3.2
(one sample)
33.3
Range of
Positive
Values
(oocysts/L)
1/1,175 L
0.002-0.45d
NR
NR
NR
NR
0.3
Unknown
0.3
NR
0.52
0.79
NR
0.3-12.7
NR
NR
NR
Mean
(oocysts/L)1
NA
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
NR
12
NR
4.5
NA
0.3
Reference
\rcheretal. 1995
Hancock et al. 1998
Hancock et al. 1998
Hancock etal. 1998
Hancock etal. 1998
Hancock etal. 1998
•lose etal. 1991
:onrad!997;PADEP
1997
:onrad!997;PADEP
1997
Conrad 1 997; PADEP
1997
:onrad!997;PADEP
1997
Conrad 1 997; PADEP
1997
3ebaldl997
Salis 1997
Salis 1997
Smith 1997
^edbetter (undated)
1.  Geometric mean reported unless otherwise indicated
2.  Data are presented as the percentage of positive sites
NA = not applicable
NR = not reported

The LeChevallier and Norton (1995) study collected the most samples and repeat samples from the
largest number of surface water plants nationally. LeChevallier and Norton conducted the study to
determine the level of Cryptosporidium in surface water supplies and plant effluent water. In total,
surface water sources for 72 treatment plants in 15 States and two Canadian provinces were
RIAfor the Proposed LT1FBR
2-10
February 15, 2000

-------
sampled.  The generated data set covered a two-year monitoring period (March, 1991 to January,
1993) that was combined with a previous set of data (October, 1988 to June, 1990) collected from
most of the same set of systems to create a database containing at least five immunofluorescence
assay analyses for 94 percent of the 67 systems sampled. Cryptosporidium oocysts were detected
in 135  (51.5 percent) of the 262 raw water samples collected between March 1991 and January
1993, while 87 percent of samples were positive during the survey period from October, 1988 to
June, 1990.  The geometric mean of detectable Cryptosporidium was 240 oocysts per 100 liters
(L), with a range from 6.5-6510 oocysts/lOOL.

LeChevallier and Norton (1995) also detected Cryptosporidium oocysts in 35 of 262 plant effluent
samples (13.4 percent) analyzed between 1991 and 1993. When detected, the oocyst levels
averaged 3.3 oocysts/lOOL (ranging from 0.29 to 57 oocysts/lOOL). A summary of occurrence
data for all samples in filtered effluents for the years 1988 to 1993 showed that 32 of the water
treatment plants  (45 percent) were consistently negative for Cryptosporidium. Forty-four of the
plants (62 percent) were positive for Giardia, Cryptosporidium, or both at one time or another
(LeChevallier and Norton, 1995).

The oocyst recoveries and densities reported by LeChevallier and Norton (1995) are comparable
to the results of another survey of treated, untreated, protected (pristine) and feces-contaminated
(polluted) water  supplies (Rose et al., 1991).  Six of thirty six samples (17 percent) taken from
potable drinking water were positive for Cryptosporidium, and concentrations in these waters
ranged from .5-1.7 oocysts/lOOL.  In addition, a total of 188 surface water samples were analyzed
from rivers,  lakes, or springs in 17  States. The average oocyst concentrations ranged from  less
than 1 to 4,400 oocysts/lOOL, depending  on the type of water analyzed. Cryptosporidium oocysts
were found in 55 percent of the surface water samples at an average concentration of 43
oocysts/lOOL.

It should be noted that the Information Collection Rule (ICR) will provide 18 months of
Cryptosporidium monitoring data for the development of a national source water Cryptosporidium
occurrence distribution.  Although the data collection efforts have been completed (January 2000), the
last 6 months of data are still undergoing quality assurance review.  EPA's supplementary survey is also
providing Cryptosporidium and other microbial source water occurrence data; the full set of
supplementary survey data will not be available for analysis until July 2000.  The Technical Working
Group supporting the Federal Advisory Committee involved with LT2ESWTR negotiation has been
deliberating over the appropriate data analysis methods to create the national source water distribution
for Cryptosporidium occurrence.  This issue will continue to be discussed during the remainder of the
LT2ESWTR Regulatory Negotiation process, scheduled to end in July 2000.  It is likely that the data
will undergo peer review only after the closure of the Regulatory Negotiation process. Due to the ICR
data evaluation and peer review time frame, EPA does not envision being able to utilize these data in
the LT1FBR regulatory impact analyses and instead intends to incorporate the data into the impact
analysis for the LT2ESWTR.

Despite analytical method limitations, Cryptosporidium has been detected in source waters. In
general, oocysts  are detected more frequently and in higher concentrations in rivers and streams

February 15, 2000                              2-11                       RIA for the Proposed LT1FBR

-------
than in lakes and reservoirs (LeChevallier et al., 1991a; Rose et al., 1988a,b). Madore et al. (1987)
found high concentrations of oocysts in a river affected by agricultural runoff (5,800 oocysts/L).
Such concentrations are especially significant if the contaminant removal process (sedimentation,
filtration) of the treatment plant is not operating effectively.  Oocysts may pass through to the
finished water, as LeChevallier and Norton (1995) also found, and infect drinking water
consumers, evident through waterborne disease outbreaks.

2.2.4  Waterborne Disease Outbreaks

The CDC, EPA, and the Council of State and Territorial Epidemiologists have maintained a
collaborative surveillance program for collection and periodic reporting of data on waterborne
disease outbreaks since 1971. The CDC database and biennial CDC-EPA surveillance
summaries include data reported voluntarily by the States on the incidence and prevalence of
waterborne illnesses. According to the CDC-EPA database, between 1971 and 1996 a total of
652 outbreaks and 572,829 cases of illnesses were reported (see Exhibit 2-6).  The total number of
outbreaks reported includes outbreaks resulting from protozoan contamination, virus
contamination, bacterial contamination, chemical contamination, and unknown factors.
    Exhibit 2-6. Comparison of Outbreaks and Outbreak-related Illnesses from Ground
                   Water and Surface Water for the Period 1971-19961
Water
Source
Ground
Surface
Other
All Systems3
Total
Outbreaks2
371 (57%)
223 (34%)
58 (9%)
652 (100%)
Cases of
Illnesses2
90,815(16%)
471,375(82%)
10,639 (2%)
572,829 (100%)
Outbreaks
in CWSs
113
148
30
291
Outbreaks
in NCWSs
258
43
19
320
 1. Modified from Craun and Calderon (1994) and additional 1995-1996 data.
 2. Includes outbreaks in CWSs + NCWSs + Private wells.
From 1984 to 1994, there were 19 reported outbreaks of cryptosporidiosis in the United States
(Craun, 1998).  As mentioned previously, C. parvum was not identified as a human pathogen until
1976. Furthermore, Cryptosporidiosis outbreaks were not reported in the United States prior to
1984. Ten of these cryptosporidiosis outbreaks were documented in CWSs, NCWSs, and a
private water system (Moore et al., 1993; Kramer et al., 1996; Levy et al., 1998; Craun, 1996;
Craun et al., 1998). The remaining nine outbreaks were associated with water-based recreational
activities (Craun et al., 1998). The ten cryptosporidiosis outbreaks in drinking water systems are
summarized in Exhibit 2-7.
RIAfor the Proposed LT1FBR
2-12
February 15, 2000

-------
         Exhibit 2-7. Cryptosporidiosis Outbreaks in U.S. Drinking Water Systems
Year
1984
1987
1991
1992
1992
1993
1993
1993
1994
1994
Location
CWS, NCWS,
Private
3raun Station,
rx, cws
Carrollton, GA,
:ws
Berks County,
DA, NCWS
VIedford (Jackson
bounty), OR,
:ws
Talent, OR,
CWS
Vlilwaukee, WI,
:ws
fakima, WA,
private
Cook County,
VIN, NCWS
Clark County, NV,
cws
Walla WaUa,WA,
cws
Cases of Illness
(Estimated)
117(2,000)
(13,000)
(551)
(3,000; combined
total
for Jackson County
and Talent, below)
see Medford, OR
(403,000)
7
27
103; many
confirmed for
Cryptosporidiosis
were HIV positve
134
Source
Water
Well
River
Well
Spring/River
Spring/River
Lake
Well
Lake
River/Lake
Well
Treatment
Chlorination
Conventional filtration/
dilorination; inadequate
jackwashing of some
liters
Chlorination
Chlorination/package
filtration plant
Chlorination/package
filtration plant
Conventional filtration

"iltered, chlorinated
r'rechlorination,
filtration and post-
filtration chlorination
STone reported
Suspected
Cause
Sewage-contaminated
well
Treatment
deficiencies
Ground water
under the influence of
surface water.
Source not identified
Treatment
deficiencies
High source water
contamination and
treatment deficiencies
Ground water under
the influence of
surface water
Possible sewage
backflow from
toilet/septic tank
Source not identified
Sewage
contamination
Adapted from Craun, et al. (1998)

Six of the ten Cryptosporidiosis outbreaks reported in Exhibit 2-7 originated from surface water or
possibly GWUDI water supplied by public drinking water systems serving fewer than 10,000
persons. The first outbreak involved 117 known cases and 2,000 estimated cases of illness in
Braun Station, Texas in 1984.  It was caused by sewage leakage into a ground water well
suspected to be under the influence of surface water.  A second outbreak in Pennsylvania in 1991
(551 cases of illness) occurred at a well also under the influence of surface water.  The third and
fourth (multi-episodic) outbreaks took place in Jackson County, Oregon in 1992 (3,000 cases of
illness) and were linked to treatment deficiencies in the Talent surface water system. A fifth
outbreak (27 cases of illness) in Minnesota, in 1993, occurred at a resort supplied by lake water.
February 15, 2000
2-13
RIAfor the Proposed LT1FBR

-------
Finally, a sixth outbreak (134 cases of illness) in Washington in 1994, occurred due to sewage-
contaminated wells at a CWS.

Three of the ten outbreaks (Carollton, Georgia (1987); Talent, OR (1992); Milwaukee, WI (1993)
were caused by water supplied by water treatment plants where process stream recycle was
implicated as a possible cause of the outbreak. In total, the nine outbreaks occurring in PWSs
caused approximately 419,939 cases of illness. These outbreaks illustrate that when treatment is
not operating optimally or when source water is highly contaminated, Cryptosporidium may enter
the finished drinking water and infect drinking water consumers, ultimately resulting in waterborne
disease outbreaks.

The occurrence of outbreaks of waterborne gastrointestinal infections including cryptosporidiosis
may be much greater than suggested by reported surveillance data (Craun and Calderon 1996).
The CDC database is based on responses to a voluntary and confidential survey that is completed
by State and local public health officials.

The U.S. National Research Council strongly suggests that the number of identified and reported
outbreaks in the CDC database (both for surface and ground waters) represents a small percentage
of actual waterborne disease outbreaks  (National Research Council, 1997; Bennett et al., 1987).
In practice, 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 one to two
percent of the population (Craun,  1996).

In addition, healthy adults with cryptosporidiosis may not suffer severe symptoms from the disease;
therefore, infected individuals may not seek medical assistance, and their cases go unreported.
Even if infected individuals consult a physician, Cryptosporidium is not  analyzed by routine
diagnostic tests for gastroenteritis and, therefore, tends to be under-reported in the general
population (Juranek, 1995; Craun, 1996).  Such obstacles to outbreak reporting indicate that the
incidence of disease and outbreaks of cryptosporidiosis may be much higher than officially
reported by the CDC.

Endemic waterborne disease is a factor that should also be considered. Endemic waterborne
disease is defined as any waterborne disease not associated with an outbreak.  EPA,  however, is
not aware of any currently available data for the United States that documents the incidence of
waterborne endemic cryptosporidiosis.  For example, 14^10 percent of the normal gastrointestinal
illness in a community in Quebec was associated with treated drinking water from a surface water
source (Payment et al., 1997). Given the lack of endemic waterborne disease occurrence data,
combined with the strong possibility that outbreaks are under-reported, it is likely that there are
greater instances of cryptosporidiosis and other waterborne diseases than are currently recorded.
RIA for the Proposed LT1FBR                       2-14                               February 15, 2000

-------
2.2.5  Filter Backwash and Other Process Streams:  Occurrence and Impact  Studies

EPA, in conjunction with the American Water Works Association (AWWA), the American Water
Works Service Company (AWWSCo), and Cincinnati Water Works, compiled issue papers on
each of the following recycle streams:  spent filter backwash water, sedimentation basin solids,
combined thickener supernatant, ion-exchange regenerate, membrane concentrate, lagoon decant,
mechanical dewatering device concentrate, monofill leachate, sludge drying bed leachate, and
small-volume streams (e.g., floor, roof, lab drains) (EE&T, 1999). In addition, EPA compiled the
existing Cryptosporidium occurrence data and occurrence data on other constituents of the recycle
streams with the data presented in AWWA's white papers. Through these efforts,
Cryptosporidium occurrence data have been found for five types of recycle streams: untreated
spent filter backwash water, gravity settled spent filter backwash water, combined gravity
thickener supernatant (a combination of spent filter backwash and clarification process solids),
gravity thickener supernatant from clarification process solids, and mechanical dewatering device
liquids.  Nine studies have  reported the occurrence of Cryptosporidium for these process streams.
Each study's scope and results are presented in Exhibit 2-8, and brief narratives on  each major
study follow the table. Note that the results of the studies, if not presented in the published report
as oocysts/lOOL, have been converted into oocysts/lOOL.
  Exhibit 2-8.  Cryptosporidium Occurrence in Filter Backwash and Other Recycle Streams

Name/

Location
of study
Drinking water
treatment
facilities


Farmoor water
treatment plant,
England



Potable water
supplies in
17 States

STame/Location
not reported




Number

of
samples
2




not reported





not reported



not reported







Type
of sample
backflush
waters from
rapid sand
filters

backwash
water from
rapid sand filter



filter backwash
from rapid sand
filters (10 to 40-
L sample vol.)
raw water

initial
backwash
water



Cyst/oocyst
concentration
sample 1 : 26,000 oocysts/gal
(calc. as 686,900 oocysts/lOOL);

sample 2: 92,000 oocysts/gal (calc
as 2,430,600 oocysts/lOOL)
Over 1,000,000 oocysts/lOOL in
backwash water on 2/1 9/89

100,000 oocysts/lOOL in
supernatant from settlement
tanks during the next few days
217 oocysts/lOOL
(geometric mean)


7 to 108 oocysts/lOOL

detected at levels 57 to 61 times
higher than in the raw water

Number of

treatment
plants
sampled
2




1





not
reported


not reported








Reference
Roseetal. 1986




Colbourne 1989





Roseetal. 1991



LeChevallier
etal. 1991a



February 15, 2000
2-15
RIAfor the Proposed LT1FBR

-------
  Exhibit 2-8. Cryptosporidium Occurrence in Filter Backwash and Other Recycle Streams

Name/

Location
of study
Bangor Water
Treatment Plant
(PA)





Mo shannon
Valley Water
Treatment Plant





Plant "C"











Pittsburgh
Drinking Water
Treatment Plant


'Plant
Slumber 3"



Number

of
samples
Round 1:
1 (8-hour
composite)

Round 2:
1 (8-hour
composite)

Round 1:
1 (8-hour
composite)


Round 2:
1 (8-hour
composite)
11 samples
using
continuous
flow
centrifu-
gation;
39 samples
jsing

cartridge
filters


24 (two
years of
monthly
samples)

not reported






Type
of sample
raw water
filter backwash
supernatant
recycle
raw water
filter backwash
supernatant
recycle
spent
backwash
supernatant
recycle raw
water sludge
raw water
supernatant
recycle
backwash
water from
rapid sand
filters; samples
collected from
sedimentation
basins during
sedimentation
phase of

backwash

water at depths
of 1,2, 3, and
3.3m.
filter backwash




raw water

spent
backwash



Cyst/oocyst
concentration
6 oocysts/lOOL
902 oocysts/lOOL
141 oocysts/lOOL

140 oocysts/lOOL
850 oocysts/lOOL
750 oocysts/lOOL

16,613oocysts/100L
82 oocysts/lOOL
13oocysts/100L

2,642 oocysts/lOOL
20 oocysts/lOOL
420 oocysts/lOOL

continuous flow: range 1 to 69
oocysts/lOOL; 8 of 1 1 samples
positive

cartridge filters: ranges 0.8
to252/100L; 33 of 39 samples
positive







328 oocysts/lOOL (mean);
(38 percent occurrence rate)


non-detect-13,158 oocysts/lOOL
140 oocysts/lOOL

850 oocysts/lOOL

Number of

treatment
plants
sampled
1







1







1











1




not reported







Reference
Comwell and Lee
1993a,b






Comwell and Lee
1993a,b






Karanisetal. 1996











States etal. 1997




Cornwell 1997



RIAfor the Proposed LT1FBR
2-16
February 15, 2000

-------
  Exhibit 2-8. Cryptosporidium Occurrence in Filter Backwash and Other Recycle Streams
Name/
Location
of study
'Plant C"
(see Karanis, et
al, 1996)
'Plant A"




Number
of
samples
12
50

1


Type
of sample
raw water
backwash
water from
rapid sand
filters
rapid sand filter
(sample taken
10 min. after
start of
backwashing)
Cyst/oocyst
concentration
avg. 23.2 oocysts/lOOL (max. 109
oocysts/lOOL) in 8 of 12 samples
avg. 22.1 oocysts/lOOL (max. 257
oocysts/lOOL) in 41 of 50 samples

150oocysts/100L


Number of
treatment
plants
sampled
1





Reference
Karanis etal. 1998
(TableS, p. 14)





The occurrence data available and reported are primarily for raw and recycle stream water. Filtered
effluent water was not sampled in most cases, and effluent occurrence would not provide a good proxy
for the implications of process stream recycle on the efficacy of the treatment process. There is
generally a plant-specific latency period for filter backwash to re-enter the treatment train.  If filter
backwash does enter the treatment train as a slug load and disrupts the treatment process, its effects
would possibly not register in the finished water until several hours after the start of backwashing. In
addition, the recovery efficiencies of the IF A detection method complicate measurements in dilute
effluent waters. However, the generally large concentrations of oocysts flushed from the filters,
sedimentation basins, and other areas of the plant and present in recycle streams can potentially enter
the finished water and cause cryptosporidiosis outbreaks, should the treatment plant not operate
efficiently or become disrupted due to recycle practice.

As shown in Exhibit 2-8, the concentrations of oocysts in backwash water and other recycle streams
are greater than the concentrations generally found in raw water.  For example, four studies (Cornwell
and Lee, 1993b; States et al., 1997; Rose et al., 1986; and Colbourne, 1989) have reported
Cryptosporidium oocyst concentrations in filter backwash water exceeding 10,000 oocysts/lOOL, in
some instances by several orders of magnitude.  Such concentrations illustrate that the treatment plant
has been removing oocysts from the influent water during the sedimentation and/or filtration processes.
As expected, the oocysts have concentrated on the filters and/or in the sedimentation basin sludge.
Therefore, the recycling of such process streams (e.g., filter backwash, thickener supernatant,
sedimentation basin sludge) re-introduces high concentrations of oocysts to the drinking water treatment
train.  Recycle can potentially return a significant number of oocysts  to the treatment plant in a short
amount of time, particularly if the recycle is returned to the treatment process without prior treatment,
equalization, or some other type of hydraulic detention. Should recycle disrupt normal  treatment
operations or should treatment not function efficiently due to other deficiencies, high
February 15, 2000
2-17
RIAfor the Proposed LT1FBR

-------
concentrations of oocysts may pass through the plant into finished drinking water. The major
recycle stream studies presented in Exhibit 2-8 will be described in further detail below.

Rose, etal.

Rose, et al. (1991) reported the geometric mean of the backwash samples at 217 Cryptosporidium
oocysts/lOOL. This was the highest reported average Cryptosporidium concentration of any of the
water types tested, which included polluted and pristine surface and ground water sources,
drinking water sources in addition to backwash water.

LeChevallier, etal.


In the analysis of pathogen concentrations in the  raw water and filter backwash water of the water
treatment process, LeChevallier et al. (1991c) found very high oocyst levels in backwash water of
systems that had low raw water parasite concentrations. Cryptosporidium levels in the initial
backwash water were 57 to 61 times higher than in the raw water supplies. Raw water samples were
found to contain from 7 to 108 oocysts/lOOL.  LeChevallier et al. (1991c) also noted that when
Cryptosporidium (12 of 13 times) were detected in plant effluent samples, the organisms were also
observed in the backwash samples.  They conclude that the consistency of these results shows that
accumulation of parasites in the treatment filters (and subsequent release in the backwash water)
could be related to subsequent penetration of the treatment barriers.

Cornwell and Lee

Cornwell and Lee (1993b) detected Cryptosporidium concentrations of over 15,000
Cryptosporidium oocysts/lOOL in the spent filter backwash at an adsorption clarifier plant
(Moshannon Valley) and over 800 Cryptosporidium oocysts/lOOL in backwash water from a
direct filtration plant (Bangor).  The parasite levels in the backwash samples were significantly
higher than concentrations found in raw source water, which contained Cryptosporidium oocyst
concentrations of 6-140 oocysts/lOOL at the Bangor plant and 13-20 oocysts/lOOL at Moshannon
Valley.

In addition, Cornwell and Lee determined oocyst concentrations for two other recycle streams,
combined thickener supernatant and sedimentation basin solids. The supernatant pathogen
concentrations was reported at  141 Cryptosporidium oocysts/lOOL at the  Bangor plant, and levels
were reported at 82  to 420 cysts/1 OOL for the Moshannon plant in Rounds 1 and 2 of sampling,
respectively.  The sedimentation basin sludge was reported at 2,642 Cryptosporidium
oocysts/lOOL in the clarifier sludge from the Moshannon Valley plant.
RIA for the Proposed LT1FBR                      2-18                              February 15, 2000

-------
States, et al.

Cryptosporidium occurred in the raw Allegheny river water supplying the plant with a geometric
mean of 31 oocysts/lOOL in 63 percent of samples collected, and ranged from non-detectto 2,333
oocysts/lOOL (States et al., 1997).  Of the filter backwash samples, a geometric mean of 328
oocysts/lOOL was found at an occurrence rate of 38 percent of samples, with a range from non-
detectto 13,158 oocysts/lOOL  The fact that the mean concentration of Cryptosporidium oocysts in
backwash water can be substantially higher than the oocyst concentration in untreated river water
suggests that recycling untreated filter backwash water can be a significant source of this parasite
within the treatment process.

2.2.6  Current Control and Potential for Improvement

Existing turbidity limits were created to remove large parasite cysts such as  Giardia, and,
therefore, must be strengthened to control for the smaller Cryptosporidium oocysts passing through
the treatment plant removal processes. In addition, degradation in treatment performance caused
by improper plant process stream recycle or other treatment deficiencies may harm efforts to
control Giardia lamblia and emerging pathogens,  in addition to Cryptosporidium, particularly
during periods of heavy precipitation or high runoff.

 In spite of filtration and disinfection, Cryptosporidium oocysts have been found in filtered
drinking water (LeChevallier, et al., 1991a; U.S. EPA, 1993), and many of the individuals affected
by waterborne disease outbreaks caused by Cryptosporidium were served by filtered surface water
supplies (Solo-Gabriele and Neumeister, 1996). It appears that surface water systems that filter
and disinfect may still be vulnerable to Cryptosporidium, depending on the  source water quality
and treatment effectiveness. However, today's proposal will ensure that treatment is operating
efficiently to control Cryptosporidium (see Section IV. A and IV.D).  Treatment practices that
control Cryptosporidium will control other microbiological contaminants of concern (e.g.,
Giardia).

One of the key regulations EPA has developed and implemented to counter pathogens in drinking
water is the  Surface Water Treatment Rule (SWTR) (54 FR 27486, June 29, 1989).  Among the
provisions of the rule, the SWTR requires that a surface water system have sufficient treatment to
reduce the source water concentration of Giardia and viruses by at least 99.9 percent (3 logs) and
99.99 percent (4  logs), respectively. A shortcoming of the SWTR is that the rule does not
specifically control for the protozoan Cryptosporidium.  The first report of a recognized outbreak
caused by Cryptosporidium was published during the development of the SWTR (D'Antonio et
al., 1985).

In 1998, the Agency finalized the IESWTR, designed to enhance the SWTR protections from
microbial pathogens, specifically Cryptosporidium, for systems serving 10,000 or more persons.
The IESWTR provisions  included a Maximum Contaminant Level Goal (MCLG) of zero for
Cryptosporidium. In addition, the  IESWTR requires a minimum 2 log (99 percent) removal of
Cryptosporidium, linked to enhanced combined filter effluent and individual filter turbidity
February 15, 2000                              2-19                      RIA for the Proposed LT1FBR

-------
monitoring provisions, although this requirement currently applies only to surface and GWUDI
systems serving 10,000 people or more that must filter under the SWTR.

Several provisions of today's proposed rule, the LT1FBR, are designed  to address the concerns
covered by the IESWTR, improving control of Cryptosporidium and other microbial
contaminants, for the portion of the public served by smaller PWSs (i.e., serving fewer than 10,000
persons). The LT1FBR also addresses the concern that for all PWSs that practice process stream
recycling, Cryptosporidium (and other emerging pathogens resistant to standard disinfection
practice) are reintroduced to the treatment process of PWSs by the recycle of spent filter backwash
water, solids treatment residuals, and other process streams.

Insufficient treatment practices have been cited as the cause of several reported waterborne disease
outbreaks (Rose, 1997).  Rose (1997) also found that reducing turbidity is indicative of a more
efficient filtration process.  Therefore, the turbidity and filter monitoring requirements of today's
proposed LT1FBR would ensure that the removal process necessary to protect the public from
cryptosporidiosis is operating properly, and the recycle stream provisions would ensure that the
treatment process is not disrupted or operating inefficiently. The regulatory history that led up to
development of the LT1FBR is summarized in the following section.

2.3    Regulatory History and Current Controls


2.3.1   1979 Total Trihalomethane Rule

In November 1979 (44 FR 68624), EPA set an interim MCL for total trihalomethanes  (TTHM -
the sum of chloroform, bromoform, bromodichloromethane, chlorodibromomethane) of 0.10 mg/1
as an annual average.

The interim TTHM standard applies to community water systems using surface water and/or
ground water serving at least 10,000 people that add a disinfectant to the drinking water during any
part of the treatment process. At their  discretion, States may extend coverage to smaller water
systems; however, most States have not exercised this option.

2.3.2   Total Coliform Rule

The Total Coliform Rule (TCR) (54 FR 27544, June 19, 1989) applies to all public water systems.
The TCR sets compliance with the Maximum Contaminant Level (MCL) for total coliforms (TC)
as follows.  If a system exceeds the MCL, it must notify the public using mandatory language
developed by the EPA. All systems must have a written plan identifying where samples are to be
collected.
RIA for the Proposed LT1FBR                      2-20                             February 15, 2000

-------
If a system has a TC-positive sample, it must test that sample for the presence of fecal coliforms or
E. coli.  The system must also collect a set of repeat samples, and analyze for TC (and fecal
coliform or E. coli within 24 hours of the first TC-positive sample).

The TCR also requires an on-site inspection (referred to as a sanitary survey) every 5 years for
each system that collects fewer than five samples per month. This requirement is extended to
every 10 years for noncommunity systems using only protected and disinfected ground water.

2.3.3   Surface Water Treatment Rule

Under the Surface Water Treatment Rule (SWTR) (54 FR 27486, June 29, 1989), EPA set
maximum contaminant level goals of zero for Giardia lamblia, viruses, and Legionella; and
promulgated regulatory requirements for all PWSs using surface water sources or ground water
sources under the direct influence of surface water.  The SWTR includes treatment technique
requirements for filtered and unfiltered systems that are intended to protect against the adverse
health effects of exposure to Giardia lamblia, viruses,  and Legionella, as well as  many other
pathogenic organisms. Briefly, those requirements include 1) requirements for maintenance of a
disinfectant residual in the distribution  system; 2) removal and/or inactivation of 3 log (99.9
percent) for Giardia and 4 log (99.99 percent) for viruses;  3)  combined filter effluent turbidity
performance standard of 5 nephelometric turbidity units (NTU) as a maximum and 0.5 NTU at the
95th percentile monthly, based on four-hour monitoring for treatment plants using conventional
treatment or direct filtration (with separate standards for other filtration technologies); and 4)
watershed protection and other requirements for unfiltered systems.

2.3.4   Information Collection Rule

The Information Collection Rule (ICR), which was promulgated on May 14, 1996 (61 FR 24354)
applied to large public water systems serving populations over 100,000; a more limited set of ICR
requirements pertain to ground water systems serving between 50,000 and 100,000 people.

The purpose of the ICR was to collect occurrence and treatment information to help evaluate the
need for possible changes to the current microbial requirements and existing microbial treatment
practices, and to help evaluate the need for future regulation for disinfectants and disinfection
byproducts (DBFs).  The ICR will provide EPA with additional information on the national
occurrence in drinking water of (1) chemical byproducts that form when disinfectants used for
microbial control react with naturally occurring compounds already present in source water; and
(2) disease-causing microorganisms, including Cryptosporidium, Giardia, and viruses. The ICR
also provided engineering data on how PWSs currently control for such contaminants. The ICR
monthly sampling data will  also provide information on the quality of the recycle waters via
monthly monitoring (for 18  months) of pH, alkalinity, turbidity, temperature, calcium and total
hardness, TOC, UV254, bromide,  ammonia, and disinfectant residual (if disinfectant is used).  This
data will provide some indication of the treatability of the water, the extent to which contaminant
concentration effects may occur, and the potential for contribution to DBF formation.

February 15, 2000                              2-21                      RIA for the Proposed LT1FBR

-------
2.3.5  Interim Enhanced Surface Water Treatment Rule

Public water systems serving 10,000 or more people that use surface water or ground water under
the direct influence of surface water are required to comply with the IESWTR (63 FR 69477,
December 16, 1998) by December 2001.  The purposes of the IESWTR are to improve control of
microbial pathogens,  specifically the protozoan Cryptosporidium, and address risk trade-offs
between pathogens and disinfection byproducts. Key provisions established by the rule include: a
MCLG of zero for Cryptosporidium; 2 log Cryptosporidium removal requirements for systems that
filter; strengthened combined filter effluent turbidity performance standards of 1.0 NTU as a
maximum and 0.3 NTU at the 95th percentile monthly, based on four-hour monitoring for
treatment plants using conventional treatment or direct filtration; requirements for individual filter
turbidity monitoring;  disinfection benchmark provisions to assess the level of microbial protection
provided as facilities take the necessary steps to comply with new disinfection byproduct
standards; inclusion of Cryptosporidium in the definition of ground water under the direct
influence of surface water and in the watershed control requirements for unfiltered public water
systems; requirements for covers on new finished water reservoirs; and sanitary surveys for all
surface water systems regardless of size.

2.3.6  Stage 1 Disinfection Byproduct Rule

The  Stage 1 DBPR (63 FR 69389, December 16, 1998) applies to all PWSs that  are either a
community water system (CWS) and nontransient noncommunity water system that treat their
water with a chemical disinfectant for either primary or residual treatment. In addition, certain
requirements for chlorine dioxide apply to transient noncommunity water systems.  The Stage 1
DBPR was published at the same time as the IESWTR (63  FR 69477, December 16, 1998).

The  Stage 1 DBPR finalizes maximum residual disinfectant level goals (MRDLGs) for chlorine,
chloramines, and chlorine dioxide; MCLGs for four trihalomethanes (chloroform,
bromodichloromethane, dibromochloromethane, and bromoform), two haloacetic acids
(dichloroacetic acid and trichloroacetic acid), bromate,  and chlorite; and NPDWRs for three
disinfectants (chlorine, chloramines, and chlorine dioxide), two groups of organic disinfection
byproducts TTHMs and HAAS and two inorganic disinfection byproducts, chlorite and bromate.
The NPDWRs consist of maximum residual disinfectant levels (MRDLs) or maximum
contaminant levels (MCLs) or treatment techniques for these disinfectants and their byproducts.
The NPDWRs also include monitoring, reporting, and public notification requirements for these
compounds.  The Stage 1 DBPR rule includes the best available technologies (BATs) upon which
the MRDLs and MCLs are based.  EPA believes the implementation of the Stage 1 DBPR will
reduce the levels of disinfectants and disinfection byproducts in drinking water supplies.  The
Agency believes the rule will provide public health protection for an additional 20 million
households that were not previously covered by drinking water rules for disinfection byproducts.

2.3.7  Stakeholder Involvement
RIA for the Proposed LT1FBR                      2-22                              February 15, 2000

-------
EPA has conducted two stakeholder meetings to solicit feedback and information from the
regulated community and other concerned stakeholders on issues relating to today's proposed rule.
The first meeting was held July 22 and 23, 1998 in Lakewood, Colorado. EPA presented
potential regulatory components for the LT1FBR. Breakout sessions with stakeholders were held
to generate feedback on the regulatory provisions being considered and to solicit feedback on next
steps for rule development and stakeholder involvement.  Additionally, information was presented
summarizing ongoing research and data gathering activities regrding the recycle of filter backwash.
The presentations generated useful discussion and provided substantial feedback to EPA regarding
technical issues, stakeholder concerns, and possible regulatory options.

The second stakeholder meeting was held in Dallas, Texas on March 3 and 4, 1999.  EPA
presented new analysis, summaries of current research, and revised regulatory options and data
collected since the July stakeholder meeting. Regional perspectives on turbidity and disinfection
benchmarking components were also  discussed with presentations from EPA Region VI and the
Texas Natural Resources Conservation Commission. Four break-out sessions were extremely
useful and generated a wide range of information, issues,  and technical input from a diverse group
of stakeholders.


2.4    Economic Rationale

2.4.1   Introduction

This section of the RIA discusses the statutory authority and the economic rationale for choosing a
regulatory approach to protect public health from drinking water contamination.  The economic
rationale is provided in response to Executive Order 12866, Regulatory Planning and Review,
which states,

       [E]ach agency shall  identify the problem that it intends to address (including, where
       applicable, the failures of the private market or public institutions that warrant new
       agency action) as well as assess the significance of that problem (Sect.  lb(l)).

In addition, 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 (p. 3)."  Therefore, the economic rationale laid out in this section should not be
interpreted as the Agency's approach to implementing the Safe Drinking Water Act (SDWA).
Rather, it is the Agency's economic analysis, as required by the Executive Order, to support a
regulatory approach to the public health issue at hand.

2.4.2   Statutory Authority for Promulgating the Rule

Section 1412(b)(l)(A) of SDWA requires EPA to establish National Primary Drinking Water
Regulations for contaminants that may have an adverse public health effect, that are known  to
occur, or present a substantial likelihood of occurrence in public water  systems at a frequency and
February 15, 2000                               2-23                       RIA for the Proposed LT1FBR

-------
level of public health concern and that present a meaningful opportunity for health risk reduction
for persons served by PWSs.

Section 1412(b)(2)(C) of SOW A states that, "The Administrator shall promulgate an Interim
Enhanced Surface Water Treatment Rule, a Final Enhanced Surface Water Treatment Rule, a
Stage I Disinfection Byproducts Rule, and a Stage II Disinfection Byproducts Rule..." The above
section of the statute gives specific authority to promulgate the LT1FBR.  Section 1412(b)(2)(c) is
supplemented with an additional provision regarding the recycle of process streams, which states,
"The Administrator shall promulgate a regulation to govern the recycling of filter backwash water
within the treatment process of a public water system."(1412(b)(14))

EPA is authorized to promulgate a National Primary Drinking Water Regulation "that requires the
use of a treatment technique in lieu of establishing a MCL," if the Agency finds that "it is not
economically or technologically feasible to ascertain the level of the contaminant." A treatment
technique has been selected to control  Cryptosporidium because it is currently not feasible to
measure the concentration of Cryptosporidium for regulatory purposes.

2.4.3   The Economic Rationale for Regulation

In addition to the statutory directive to regulate surface water treatment and recycling there is also a
strong economic rationale for government regulation.  The need for regulation is a direct result of
the structure of the market for publically provided drinking water. Economic theory suggests that
society's well being is maximized when goods are produced and sold in well functioning
competitive markets. A perfectly competitive market is said to exist when there are many
producers of a product selling to many buyers,  and both producers and consumers have complete
knowledge regarding the products of each firm. There must also be no barriers to entry in the
industry, and firms in the industry must not have any advantage over potential new producers.
Two major factors in the public water supply industry do not satisfy the requirements for a
competitive market and lead to market failures that require regulation.

First, the public water market has monopolistic tendencies.  These monopolies tend to exist
because it is not economically efficient to have  multiple suppliers competing to build multiple
systems of pipelines, reservoirs, wells, and other facilities in the same locality. Instead, a single
firm or government entity performs these functions under public control. Under monopolistic
conditions, consumers are provided only one level of service with respect to the quality attribute of
the product, in this case  drinking water quality.  Since water purveyors often operate in such a
monopolistic environment they may not respond to the usual market incentive to satisfy their
consumers' desire for safety and high drinking  water quality.

Second, there are high information and transaction costs that impede public understanding of the
health and safety issues concerning drinking water quality.  The type of health risk potentially
posed by trace quantities of drinking water contaminants involve analysis and distillation of
complex  toxicological data and health sciences. EPA has finalized the development of the
Consumer Confidence Report Rule that makes  water quality information more easily available to
RIA for the Proposed LT1FBR                       2-24                              February 15, 2000

-------
consumers.  This Rule requires community water systems to post or mail their customers an annual
report on local drinking water quality. However, consumers would still have to analyze this
information for its health risk implications. Even if informed consumers are able to engage
systems regarding these health issues, the costs of such engagement-transaction costs (measured in
personal time and commitment) present significant impediment to consumer expression of risk
preference.

SDWA regulations are intended to provide a level  of protection from exposure to drinking water
contaminants that would not otherwise occur in the existing market environment for public water
supply.  The regulations set minimum performance requirements for all public water supplies in
order to protect all consumers from exposures to contaminants.  SDWA regulations are not
intended to restructure flawed market mechanisms  or to establish competition in supply, but rather,
to regulate the "product" produced within these markets.  In other words, SDWA standards
establish the level of service to be provided in  order to better reflect public preferences for safety.
Also, the Federal regulations remove the high  information and transaction costs that would be
required for consumers to make informed purchasing decisions by acting on behalf of all
consumers in balancing the risk reduction and the social costs of achieving this reduction.


2.5    Summary of the Proposed Rule

EPA is proposing the following requirements to meet the public health protection goals of the LT1FBR,
which will provide a level of protection for small systems that is comparable to the IESWTR, and to
fulfill the statutory requirements of the SDWA. Exhibit 2-9 shows that the proposed rule includes two
sets of provisions—the set of small  system provisions that are parallel to the IESWTR requirements
(enhanced filtration requirements, disinfection benchmark requirements, and additional requirements),
and the set of provisions that address recycle practices.  The flow chart in Exhibit 2-10 illustrates how
a system using surface water or GWUDI as a source  determines which provisions apply to it.


2.5.1   Enhanced Filtration Provisions

The proposed rule established a requirement for 2 log (i.e., 99 percent) removal of
Cryptosporidium oocysts for surface water and GWUDI systems serving fewer than 10,000
people and filtering their water under the SWTR. This requirement applies between a point where
the raw water is not subject to recontamination by  surface water runoff and a point downstream
either before or at the first customer. Compliance with the combined filter effluent turbidity
requirements, described below, insures compliance with the 2 log removal requirement.

There are two turbidity provisions in the proposed LT1FRB. One  provision establishes revised
combined filter effluent requirements for small systems that use filtration.  These requirements differ
across filtration technologies. A second provision establishes individual filter monitoring requirements
for the subset of systems that use conventional or direct filtration.
February 15, 2000                               2-25                       RIA for the Proposed LT1FBR

-------
For conventional and direct filtration systems, the proposed rule revises the existing combined filter
effluent requirement such that the turbidity level of representative samples of a system's combined filter
effluent water must be less than or equal to 0.3 nephelometric turbidity units (NTU) in at least 95
percent of the measurements taken each month.  The turbidity level of representative samples of a
system's filtered water must not exceed 1 NTU at any time.

For systems using membrane filtration (i.e., microfiltration, ultrafiltration, nanofiltration, and reverse
osmosis) the proposed rule requires that the turbidity level of representative samples of a system's
combined filter effluent water must be less than or equal to 0.3 NTU in at least 95 percent of the
measurements taken each month.  The turbidity level of representative samples of a system's filtered
water must not exceed 1 NTU at any time.  EPA included turbidity limits for membrane systems to
allow such  systems the ability to opt out of a demonstration of their ability to remove Cryptosporidium.
In lieu of these turbidity limits, a public water system that utilizes membrane filtration may demonstrate
(using pilot plant studies or other means) to the State that membrane filtration—in combination with
disinfection treatment—consistently achieves 3 log removal and/or inactivation of Giardia lamblia
cysts, 4 log removal and/or inactivation  of viruses, and 2 log removal of Cryptosporidium oocysts.
For each approval, the State will set turbidity performance requirements that the system must meet at
least 95 percent of the time and that the system may not exceed at any time that are consistent with
these removal and/or inactivation requirements.

Systems utilizing slow sand or diatomaceous earth filtration must continue to meet the combined filter
effluent limits established for these technologies under the SWTR. Namely, the turbidity level of
representative samples of a system's filtered water must be less than or equal to 1 NTU in at least 95
percent of the measurements taken each month and the turbidity level  of representative samples of a
system's filtered water must at no time  exceed 5 NTU.

For all alternative filtration technologies (i.e., other than conventional,  direct, slow sand,  diatomaceous
earth, or membrane), public water systems must demonstrate to the State for purposes of approval
(using pilot plant studies or other means) that the alternative filtration technology—in combination with
disinfection treatment—consistently achieves 3 log removal and/or inactivation of Giardia lamblia
cysts, 4 log removal and/or inactivation  of viruses, and 2 log removal of Cryptosporidium oocysts.
For each approval, the State will set turbidity performance requirements that the system must meet at
least 95 percent of the time and that the system may not exceed at any time at a level that consistently
achieves these removal  and/or inactivation requirements.

The proposed individual filter monitoring requirement applies to all surface water and GWUDI systems
that serve populations fewer than 10,000 and utilize direct or conventional filtration. These systems are
required to  conduct continuous monitoring (i.e., one turbidity measurement recorded every 15 minutes)
for each individual filter. A system must provide an exceptions report to the State as part of the existing
combined effluent reporting process if any individual filter turbidity measurement exceeds 1 NTU,
unless the system can show that the next reading is less than 1 NTU. Furthermore, if a system is
required to  submit an exceptions report  for the same filter in three consecutive months, the system must

RIA for the Proposed LT1FBR                       2-26                               February 15, 2000

-------
conduct a self-assessment of the filter.  Finally, if a system is required to submit an exceptions report
that contains an accedence of 2 NTU for the same filter in two consecutive months, the system must
arrange for a comprehensive performance evaluation to be conducted by the State or a third party
approved by the State.

2.5.2  Disinfection Benchmarking Provision


Small systems are already required to comply with the Stage 1 DBPR. The proposed LT1FBR follows
the principles set forth in earlier negotiations, i.e., if systems consider making changes to their
disinfection practices to comply with Stage 1 DBPR, they cannot undercut their current level of
microbial protection.  The disinfection benchmarking requirements are designed to ensure that risk from
one contaminant is not increased while risk from another contaminant is decreased.  The requirements,
which apply to all small systems (i.e., serving fewer than 10,000 people) that use surface water or
GWUDI as a source and are not  a transient noncommunity system, have three components:


       •   Applicability monitoring to determine which systems have annual average TTHM and
           HAAS levels close enough to their respective MCL (i.e., equal to or greater than 80
           percent of the MCL) such that they may need to consider altering their disinfection
           practices to comply with Stage 1 DBPR

       •   Disinfection profiling to develop a baseline of current microbial inactivation over one year


       •   Disinfection benchmarking to determine the month with the lowest average level of
           microbial inactivation from the profile.


All systems subject to the rule must develop a disinfection profile unless they choose to monitor
TTHM values and can demonstrate levels less than 80 percent of their respective MCLs.  The
disinfection profile is developed by measuring four parameters:  disinfectant residual, contact time,
water temperature, and pH. These values are used to derive the level of microbial inactivation and
must be measured on a weekly  basis for one year starting in the first week of January 2003.

If a system that is required to develop a disinfection profile decides to make a significant change in
disinfection practices it must calculate its disinfection benchmark, which is the lowest level of
inactivation achieved over the course of the year, and consult with the State before implementing
such a change. Significant changes in disinfection practice are defined as: moving the point of
disinfection (other than routine  seasonal changes already approved by the State); changing the type
of disinfectant;  changing the disinfection process; or making other modifications designated as
significant by the State. Supporting materials for the consultation with the State must include a
description of the proposed change, the disinfection profile and benchmark for Giardia lamblia
(and, if necessary, viruses for systems using ozone or chloramines), and an analysis of how the
proposed change might affect the current level of Giardia lamblia inactivation.
February 15, 2000                                2-27                       RIA for the Proposed LT1FBR

-------
Systems serving fewer than 500 persons have the option to request assistance from the State in
calculating disinfection benchmark. This option is contingent on the system providing the State
with the necessary operational data, and State agreement to perform the profile and benchmark
calculations.

In the proposed rule, the applicability monitoring component is optional, which differs from the
IESWTR requirements. If a system has TTHM and HAAS data taken during the month of
warmest water temperature (one sample for each to be taken anytime from 1998-2002) and taken
at the point of maximum residence time in the distribution system, they may submit this data to the
State beginning two years after the publication date. If the data shows that TTHM and HAAS
levels are less than 80 percent of their respective MCLs, the system does not have to develop a
disinfection profile.  If the data shows that either TTHM or HAAS levels or both are equal to or
greater than 80 percent of the MCLs, the system would be required to develop a disinfection
profile in 2003. If a system does not have, or does not gather such TTHM and HAAS data during
the month of warmest water temperature and at the point of maximum residence time in the
distribution system as described, then the system would automatically be required to develop a
disinfection profile starting the first week of January 2003.

2.5.3  Other LT1 Provisions

The proposed rule also modifies the definition of GWUDI to include Cryptosporidium for systems
serving fewer than 10,000 persons. Under the SWTR, States were required to determine whether
systems using ground water were using ground water under the direct influence of surface water.
State determinations were required to be completed by June 29, 1994, for community water
systems and by June 29, 1999, for  noncommunity water systems.  EPA does not believe that it is
necessary to make a new determination of GWUDI for this rule based on the addition of
Cryptosporidium to the definition of GWUDI because the current screening methods appear to
adequately address the possibility of Cryptosporidium in the ground water.

The proposed rule extends the existing  watershed control regulatory requirements for unfiltered
systems serving fewer than 10,000 people to include the control of Cryptosporidium, which will
be included in the watershed control provisions for these systems wherever Giardia lamblia is
mentioned. Affected public water  systems must maintain their watershed control programs to
minimize the  potential for contamination by Cryptosporidium oocysts as well as Giardia lamblia
and viruses in the water. The  State must determine whether the watershed  control program is
adequate to meet this goal.  The adequacy of a program to limit potential contamination by Giardia
lamblia cysts, Cryptosporidium oocysts, and viruses must be based on the comprehensiveness of
the watershed review, the effectiveness of the system's program to monitor and control detrimental
activities occurring in the watershed, and the extent to which the water system has maximized land
ownership and/or controlled land use within the watershed.

The proposed rule also requires surface water and GWUDI systems that serve fewer than 10,000
people to cover all new reservoirs,  holding tanks, or other storage facilities for finished water for
RIA for the Proposed LT1FBR                      2-28                              February 15, 2000

-------
which construction begins after 60 days after publication of the final rule.  This requirement does
not apply to existing uncovered finished water reservoirs.

2.5.4  Recycle Provisions

As Exhibit 2-9 shows, there are three recycle provisions in the proposed rule.  These provisions
apply to large and small systems.  The first provision requires all surface water and GWUDI
systems that employ rapid granular filtration to return spent filter backwash, thickener supernatant,
or liquids from solid/liquid separation processes prior to the point of primary coagulant addition;
systems that must move their recycle are required to submit a schematic diagram of proposed
changes in the location of returned recycled water to the State.  Plants that require an alternative
recycle location to maintain optimal finished water quality, plants that are designed to employ
recycle flow as an intrinsic component of the treatment process, or plants with unique treatment
requirements or processes may apply to the State for a waiver to return recycle flows to an
alternative location.

The second provision requires all  surface water and GWUDI systems employing rapid granular
filtration that practice direct recycle, typically  employ 20 or fewer filters to meet production
requirements during the highest production month in the 12-month period prior to the LT1FBR
compliance date and recycle spent filter backwash or thickener supernatant to the primary
treatment process to conduct a self assessment and report the results to the State.  The State will
determine whether recycle practices need to be changed. Prior to conducting the self assessment, a
system must submit a monitoring plan to the State describing how it will monitor recycle flows and
source water influent during the month of highest production and determine whether and how
frequently it exceeds State-approved capacity.  States are required to review the self assessments
and report to EPA their determinations regarding whether modifications to recycle practices are
necessary and provide a brief summary of the reasons for making those determinations.

The third provision requires that surface water and GWUDI systems with direct filtration and
recycling to the main treatment process report their recycling practices to the State. The State is
required to determine whether recycle practices must be changed.  States are also required to report
these determinations to EPA along with brief explanations.
February 15, 2000                               2-29                      RIA for the Proposed LT1FBR

-------
            Exhibit 2-9. Summary of How the Proposed LT1FBR is Organized
RIAfor the Proposed LT1FBR
2-30
February 15, 2000

-------
 Exhibit 2-10. Illustration of How the Provisions Apply to Different Types of Surface Water
                                   or GWUDI Systems
Is system a PWS?
1
Yes
T
February 15, 2000
2-31
RIAfor the Proposed LT1FBR

-------
                  3. Consideration of Regulatory Alternatives

In addition to the proposed rule described in the previous section, EPA considered several
alternative regulatory options. EPA started with the regulatory framework for the IESWTR,
which applies to systems serving 10,000 or more people, in developing this proposed rule.  Then
several less burdensome alternatives tailored to small water systems were considered. For the
purposes of developing the RIA, EPA considered three different individual filter turbidity
alternatives, four applicability monitoring alternatives,  and three disinfection profiling alternatives.
In addition, four alternatives for the recycle provisions were considered. The remainder of the
chapter provides a detailed description of these regulatory alternatives and EPA's rationale for
considering them.


3.1    Individual  Filter Turbidity Monitoring

The proposed LT1FBR establishes an individual filter turbidity monitoring requirement which
applies to all surface  water and GWUDI systems using conventional and direct filtration and
serving fewer than 10,000 people. In developing this requirement, the Agency evaluated several
alternatives in an attempt to reduce the burden faced by small systems while still providing a level
of public health protection comparable to the IESWTR.  In addition, the Agency wanted the
individual filter turbidity monitoring requirement to serve as an early warning tool systems can use
to detect and correct problems with filters.

Alternative Tl

The first alternative considered by the Agency would require direct and conventional filtration
systems serving populations fewer than  10,000 to meet the same requirements as established for
systems serving 10,000 or more people in the IESWTR.  This alternative would require that
systems conduct continuous monitoring  of turbidity (one turbidity measurement every 15 minutes)
for each individual filter. Based on this  monitoring, systems would need to provide an exceptions
report to the State as  part of the existing  combined filter effluent reporting process for any of the
following circumstances:

       Any individual filter has a turbidity level greater than 1.0 NTU based on two consecutive
       measurements taken 15 minutes  apart

•          Any individual filter has a turbidity level greater than 0.5 NTU at the end of the first 4
           hours of filter operation based on two consecutive measurements taken 15 minutes
           apart

•          Any individual filter has turbidity levels greater than 1.0 NTU based on two
           consecutive measurements taken 15 minutes apart at any time in each of 3 consecutive
February 15, 2000                                3-1                      RIA for the Proposed LT1FBR

-------
           months (the system must, in addition to filing an exceptions report, conduct a self
           assessment of the filter)

       Any individual filter has turbidity levels greater than 2.0 NTU based on two consecutive
       measurements taken 15 minutes apart at any time in each of 2 consecutive months (the
       system must file an exceptions report and must arrange for the conduct of a comprehensive
       performance evaluation (CPE) by the State or a third party approved by the State).

Under the first two circumstances identified above, a system must produce a filter profile if no
obvious reason for the abnormal filter performance can be identified.

Alternative T2

The second alternative considered by the Agency represents a slight modification from the
individual filter monitoring requirements for large systems.  The 0.5 NTU exceptions report trigger
was omitted in an effort to reduce the burden associated with daily data evaluation.  Additionally,
the filter profile requirement was removed. This alternative still requires that all conventional and
direct filtration systems conduct continuous monitoring (one turbidity measurement every 15
minutes) for each individual filter, and the additional requirements are stated as follows.

•      A system must provide an exceptions report to the State as part of the existing combined
       effluent reporting process if any individual filter turbidity measurement exceeds 1.0 NTU,
       unless the system can show that the next reading is less than 1.0 NTU.

       If a system is required to submit an exceptions report for the same filter in 3 consecutive
       months, the system must conduct a self assessment of the filter

       If a system is required to submit an exceptions report for the same filter in 2 consecutive
       months that contains an exceedance of 2.0 NTU by the same filter, the system must arrange
       for the conduct of a CPE by the State or a third party approved by the State.

Alternative T3

The third alternative considered by the Agency would include new triggers for reporting and
follow-up action in an effort to reduce the daily burden associated with data review.  This
alternative still requires that all conventional and direct filtration systems conduct continuous
monitoring (one turbidity measurement every 15 minutes) for each individual filter, and it has three
other requirements.

•      A system must provide an exceptions report to the State as part of the existing combined
       effluent reporting process if filter samples exceed 0.5 NTU in at least 5 percent of the
RIAfor the ProposedLT1FBR                3-2                          February 15, 2000

-------
       measurements taken each month and/or any individual filter measurement exceeded 2.0
       NTU (unless the system can show that the following reading was less than 2.0 NTU).

       If a system is required to submit an exceptions report for the same filter in 3 consecutive
       months the system must conduct a self assessment of the filter.

•      If a system is required to submit an exceptions report for the same filter in 2 consecutive
       months that contains an exceedance of 2.0 NTU by the same filter, the system must arrange
       for the conduct of a CPE by the State or a third party approved by the State.

For all three alternatives, the requirements for conducting filter self assessments and CPEs are the
same.  If a CPE is required, the system must arrange for the State or a third party approved by the
State to conduct the CPE no later than 30 days following the exceedance. The CPE must be
completed and submitted to the State no later than 90 days following the exceedance. If a self
assessment is required, it must take place within 14 days of the exceedance and the system must
report to the State that the self assessment was conducted.  The self assessment must consist of at
least the following components:

       •   Assessment of filter performance
       •   Development of a filter profile
       •   Identification and prioritization of factors limiting filter performance
       •   Assessment of the applicability of corrections
           Preparation of a filter self assessment report.

Exhibit 3-1 summarizes the key differences between the alternatives. It includes EPA's estimate
of how frequently a system would  need to review turbidity data to comply with the reporting
requirements.
February 15, 2000                                3-3                     RIA for the Proposed LT1FBR

-------
                   Exhibit 3-1. Filter Turbidity Monitoring Alternatives
Key Differences
Minimum frequency
with which turbidity data
are analyzed
Exceptions report trigger
based on



Alternative Tl
Daily
> 1 NTU in
two consecutive
measurements
or
> 0.5 NTU in
two consecutive
measurements
after first 4 hours
of filter operation
Alternative T2
Weekly
> 1 NTU in
two consecutive
measurements


Alternative T3
Monthly
> 0.5 NTU in
at least 5%
of measurements
in a month
or
> 2 NTU in
two consecutive
measurements
In considering the above alternatives, the Agency attempted to reduce the burden faced by small
systems. Each of the three alternatives was judged to provide comparable levels of public health
protection.  All three were also considered useful diagnostic tools for small systems to evaluate the
performance of filters and correct problems before follow-up action was necessary. The first
alternative was viewed as significantly more challenging to implement and burdensome for smaller
systems because of the amount of data review. This evaluation was also expressed by small entity
representatives during the Agency's SBREFA process as well as stakeholders at each of the public
meetings held to discuss issues related to the proposed rule.  Although Alternative T3 reduced the
system burden associated with data review, it would institute a very different trigger for small
systems than that established for large systems by the IESWTR. This was viewed as problematic
by several stakeholders who stressed the importance of maintaining similar requirements in order
to limit transactional costs and additional State burden.  Therefore, the Agency is proposing
Alternative T2 as described above, which allows operators to expend less time evaluating turbidity
data.  Alternative T2 maintains a comparable level of public health protection as those afforded
large systems, reduces much of the burden associated with daily data collection and review, yet
still serves as a self-diagnostic tool for operators and provides the mechanism for State follow-up
when significant performance problems exist.


3.2    Applicability Monitoring

EPA considered four alternatives for disinfection byproduct (DBF) applicability monitoring
and, thereby, which systems would be required to develop a disinfection profile.  Although the
applicability monitoring alternatives and the profile alternatives (see Section 3.3) are discussed and
analyzed separately, EPA considered them in conjunction with one another to select the preferred
alternatives because they work together to protect public health. In exploring alternatives for both
activities, the Agency focused on reducing the burden on small systems while providing a level  of
health protection with respect to the risk-risk tradeoff between microbial and DBF contaminants
that is comparable to IESWTR.
RIAfor the ProposedLT1FBR
3-4
February 15, 2000

-------
Alternative A1

The IESWTR required that large systems monitor each quarter for total trihalomethanes (TTHMs)
and five haloacetic acids (HAA5)4 at four points in the distribution system.  At least one of those
samples must be taken at a point that represents the maximum residence time of the water in the
system. The remaining three must be taken at representative locations in the distribution system,
taking into account the number of persons served, the different sources of water, and the different
treatment methods employed.  The results of all analyses per quarter are averaged and reported to
the State.

EPA considered applying this  alternative to systems serving fewer than 10,000 people and
requested input from small system operators and other interested parties, including the public.
Based on the feedback EPA received, two other alternatives were  developed for consideration.

Alternative A2

EPA considered requiring systems serving fewer than 10,000  people to monitor for TTHM and
HAAS according to the following schedule, which has requirements that differ by system  size.
Systems serving between 500  and 10,000 would need to collect a  sample for TTHM and HAAS
analyses no less than once per quarter per treatment plant operated. Systems serving 500 and
fewer would need to collect a  sample for TTHM and HAAS analyses no less than once per year
per treatment plant during the month of warmest water temperature. If systems wish to take
additional samples, however, they would be permitted to do so.  This alternative provides  an
applicability monitoring frequency that is identical to the DBF monitoring frequency under the
Stage 1 DBPR that systems will have to comply with in 2004.

Alternative A3

EPA considered requiring all systems serving fewer than  10,000 people to monitor once per year
per system during the month of warmest water temperature.

Under Alternatives Al, A2, and A3, systems may consult with States and elect not to perform
TTHM and HAAS monitoring and proceed directly with the development of a disinfection profile.

During the SBREFA process and during stakeholder meetings, EPA received some positive
comments regarding Alternative A3 as the least burdensome approach. Other stakeholders,
however, pointed out that Alternative A3 does not allow systems to measure seasonal variation as
in Alternative A2 for systems  serving populations between 500 and 10,000.  Several stakeholders
agreed that despite the costs, the information obtained from applicability monitoring would be
       4This is the sum of the concentrations of mono-, di-, and trichloroacetic acids and mono- and
dibromoacetic acids.

February 15, 2000                               3-5                     RIA for the Proposed LT1FBR

-------
useful. EPA agrees that it is valuable to systems to monitor and understand the seasonal variation
in TTHM and HAAS values. Alternative A2 is most consistent with EPA's goal of reducing the
burden on small systems, while maintaining comparable levels of public health protection.
Consequently, the Agency considered proposing Alternative A2 as part of today's proposed rule.

During subsequent discussions, which included feedback from States, the Agency reconsidered
Alternative A2. Due to the statutory provisions in SDWA that require States have 2 years
to develop their own regulations as part of their primacy requirements, EPA recognized
that requiring applicability monitoring prior to the completion of the 2-year period after
promulgation could pose a significant burden on States. In response to these concerns, the Agency
developed a new preferred alternative, Alternative A4.

Alternative A4

Applicability monitoring is optional and not a requirement under this alternative; systems do not
need to obtain State approval to forego monitoring. If a system has TTHM and HAAS data taken
during the month of warmest water temperature (from 1998-2002) and taken at the point of
maximum residence time, it may submit these data to the State 2 years after the rule is published
and prior to January 1, 2003. If the data show TTHM and HAAS levels less than 80 percent of
their respective MCLs, the system does not have to develop a disinfection profile. If the data show
TTHM or HAAS levels or both at or greater than 80  percent of their respective MCLs, the  system
would be required to  develop a disinfection profile in 2003.  If a system does not have or does not
gather TTHM and HAAS data during the month of warmest water temperature, and at the point  of
maximum residence time in the distribution system as described above, then the system would
automatically be required to develop a disinfection profile starting the first week of January 2003.

Exhibit 3-2 summarizes the sampling requirements across the four alternatives. The combination
of fewer sampling periods and fewer samples per period imply a lower system burden  for A4
compared to the other alternatives. Furthermore, systems can opt to forego monitoring, which
further reduces the potential burden of the proposed rule.
RIAfor the ProposedLT1FBR               36                          February 15, 2000

-------
                    Exhibit 3-2.  Applicability Monitoring Alternatives
Key
Differences
Sample
collection
frequency

Sample
collection
location

Total
samples
Alternative
Al
Once per
quarter for
1 year
At four locations
including one at
the point of
maximum
residence time
161
Alternative
A2
501-9,999: no less than once
per quarter for 1 year; • 500:
no less than once during the
month of warmest water
temperature

At the point of maximum
residence time

501-9,999: 41
• 500: I1
Alternative
A3
Once during
the month of
warmest water
temperature

At the point of
maximum
residence time

I1
Alternative
A4
Optional: one sample
during the month of
warmest water temperature
prior to January 1, 2003

At the point of maximum
residence time

I2
 1. Systems may obtain State approval to forego applicability monitoring and begin profiling.
 2. Applicability monitoring is optional for Alternative A4; State approval is not required.


3.3    Disinfection Profiling and Benchmarking


EPA considered three alternatives for requiring systems to develop the disinfection profile.  These
alternatives consider a range of sampling requirements that EPA developed to evaluate the trade-
off between less frequent sampling and the robustness of the resulting disinfection profile and
benchmark.

Alternative Bl

The IESWTR requires systems  serving 10,000 or more persons to measure four parameters
(disinfectant residual, contact time, water temperature, and pH) and develop a profile of microbial
inactivation on a daily basis if they have TTHM or HAAS levels that equal or exceed  80 percent
of their respective MCLs.  EPA considered extending this requirement to systems serving fewer
than 10,000 persons, and requested input from small system operators and other interested
stakeholders. EPA received feedback that this requirement would place a significant burden on
the small system operator for at least two reasons:

       •   Small system operators are not present at the plant every day

           Small systems often have only one operator at a plant who is responsible for all aspects
           of maintenance, monitoring, and operation.

Recognizing the potential burdens that the profiling procedures placed on small systems, EPA
considered two additional alternatives that are evaluated in the RIA.
February 15, 2000
3-7
RIA for the Proposed LT1FBR

-------
Alternative B2

EPA considered requiring all systems serving fewer than 10,000 persons to develop a disinfection
profile based on weekly parameter measurements for one year. A system with TTHM and HAAS
levels less than 80 percent of their respective MCLs (based on either required or optional
monitoring as described above) would not be required to conduct disinfection profiling. This
alternative would save the operator time and still provide information on seasonal variation over
the period of 1 year.

Alternative B3

EPA considered a daily monitoring requirement only during a 1-month critical monitoring period
to be determined by the State.  In general, colder temperatures reduce disinfection efficiency.  For
systems in warmer climates, or climates that do not change very much during the course of the
year, the State would identify other critical periods or conditions.  This alternative reduces the
number of times the operator has to calculate the microbial inactivation.

Initially, EPA considered a fourth alternative of not requiring the disinfection profile at all.  After
consideration of the feedback of small system operators and other interested stakeholders, EPA
believes that there is a strong benefit in the plant operator knowing the level of microbial
inactivation, and that this information is essential to ensuring that systems continue to provide
adequate microbial protection while they comply with the requirements of the Stage 1 DBPR.
Consequently, this alternative was excluded from further analysis.  Exhibit 3-3 summarizes the
three alternatives that are evaluated in the RIA.
            Exhibit 3-3. Disinfection Profiling and Benchmarking Alternatives
Key Differences
Profiling data is collected
Inactivation observations
Alternative Bl
Once/day for a year
365
Alternative B2
Once/week for a year
52
Alternative B3
Once/day for a month
30
EPA considered all of the above alternatives and EPA selected Alternative B2 as the preferred
alternative for the proposed rule for the following reasons.  First, given early implementation
concerns, the timing of this alternative appears to be the most appropriate in balancing early
implementation issues with the need for systems to prepare for implementation of the Stage 1
DBPR and ensuring adequate and effective microbial protection.  Second, it allows systems and
States that have been proactive in conducting applicability monitoring to reduce costs for those
systems that can demonstrate low TTHM and HAAS levels. Third, this alternative allows systems
and States the opportunity to understand seasonal variability in microbial disinfection. Finally, this
alternative takes into account the flexibility needed by the smallest systems while maintaining
comparable levels of public health protection with the larger systems.
RIA for the ProposedLT1FBR               3-8                         February 15, 2000

-------
3.4    Recycle Provisions

EPA considered four alternatives for the recycle provisions.  All of the alternatives require select
recycle flows to be returned prior to the point of primary coagulant addition. Alternatives R2, R3,
and R4 place additional requirements on systems that practice direct recycle or direct filtration, as
well as other conventional systems that recycle. Each of these alternatives are discussed in detail
in the paragraphs that follow.

Alternative Rl

The  first alternative considered by the Agency requires that rapid granular filtration plants using
surface water or GWUDI as a source return filter backwash, thickener supernatent, and liquids
from dewatering processes prior to the point of primary coagulant addition.5  Plants that require an
alternative recycle return location to maintain optimal finished water quality (as indicated by
finished water or intra-plant turbidity levels), plants that are designed to employ recycle flow as an
intrinsic component of the treatment  process, or plants with unique treatment requirements or
processes may apply to the State for  a waiver to return recycle flows to an alternative location.
Softening systems may recycle process solids, but not spent filter backwash, thickener supernatant,
or liquids from dewatering processes, at the point of lime addition immediately preceding the
softening process to improve treatment efficiency.  Contact clarification systems may recycle
process solids, but not spent filter backwash, thickener supernatant, or liquids from dewatering
processes, directly into the contactor  to improve treatment efficiency.

Alternative R2

In addition to requiring plants to return select recycle flows prior to the point of primary coagulant
addition, this alternative also requires some direct recycle systems to perform a self assessment of
their recycle practice and report the results to the State. The public water systems that would be
required to conduct a self assessment are those that meet all of the following criteria:

•      Use surface water or GWUDI as a source and employ conventional rapid granular
       filtration treatment

•      Employ 20 or fewer filters to meet production requirements during the highest production
       month in the 12-month period prior to LTlFBR's compliance date

•      Recycle spent filter backwash or thickener supernatant directly to the treatment process
       (i.e., recycle flow is returned  within the treatment process of a PWS without first passing
       5The recycle provisions apply to individual plants because some large systems have two or
more plants treating water, some of which may not recycle flow to the treatment process and, therefore,
are not subject to the recycle provisions.
February 15, 2000                                 3-9                      RIA for the Proposed LT1FBR

-------
       the recycle flow through a treatment process designed to remove solids, a raw water
       storage reservoir, or some other structure with a volume equal to or greater than the volume
       of spent filter backwash water produced by one filter backwash event).

The systems that meet all the above criteria are required to develop and submit a recycle self
assessment monitoring plan to the State no later than 3 months after the rule's effective date.  At a
minimum, the monitoring plan must identify the month during which monitoring will be
conducted, contain a schematic identifying the location of raw and recycle flow monitoring
devices, describe the type of flow monitoring devices to be used, and describe how data from the
raw and recycle flow monitoring devices will be simultaneously retrieved and recorded.

Systems are required to submit a self assessment report to the State within 1 month of completing
the self assessment monitoring.  At a minimum, the report must provide the following information:

•      All source and recycle flow measurements taken and the dates they were taken. For all events
       monitored, report the times the filter backwash recycle event was initiated, the flow
       measurements taken at three minute intervals, and the time the filter backwash recycle event
       ended. Report the number of filters in use when the backwash recycle event is monitored.

•      All data used and calculations performed to determine whether the system exceeded operating
       capacity during monitored recycle events and the number of event flow values that exceeded
       State approved operating capacity.
+
•      A plant schematic showing the origin  of all recycle flows, the hydraulic conveyance used
       to transport them, and their final destination in the plant

•      A list of all the recycle flows and the frequency at which they are returned to the plant

•      Average and maximum backwash flow through the filters and the average and maximum
       duration of backwash events in minutes, for each monitoring event

       Typical filter run length, number of filter typically employed, and a written summary of
       how filter run length is determined (e.g., preset run time, headloss, or turbidity level).

Systems are required to submit the self assessment to the  State within 3  months of completing the
last day of source and recycle flow monitoring.

EPA is proposing that the State review all self assessments  submitted by PWSs and report to the
Agency the one of the following for each individual plant:
RIAfor the ProposedLT1FBR              3  10                         February 15, 2000

-------
       A finding that modifications to recycle practice are necessary, followed by a brief
       description of the required change and a summary of the reason(s) the change is required
       A finding that changes to recycle practice are not necessary and a brief description of the
       reason(s) this determination was made.

Alternative R2 also requires direct filtration plants using surface water and GWUDI that recycle to
the treatment process to report certain data that characterize their recycle practice to the State:

       Whether recycle flow treatment or equalization is in place

       The type of treatment provided for the recycle flow

       If equalization, sedimentation, or some type of clarification process is used, the following
       information should be provided:  the physical dimensions of the unit sufficient to allow
       calculation of its volume, and the type, typical dose, and frequency at which treatment
       chemicals are used

•      The minimum and maximum hydraulic loading the treatment unit experiences

       The maximum backwash rate, duration, typical filter run length, and the number of filters at
       the plant.

The purpose of this requirement is to allow States to assess whether the existing recycle practice of
direct filtration plants addresses the potential risks posed by recycle. The Agency believes that
direct filtration plants need to remove oocysts from recycle flow prior to reintroducing it to the
treatment process. States are required to report their determination for each system to EPA and
provide a brief explanation of the reason(s) for the decision.

Alternative R3

The Agency also considered requiring all recycle plants without existing recycle flow equalization
or treatment to install recycle flow equalization.  This  option would not require a self assessment.
Under Alternative R3, systems would also still be required to return select recycle  flows prior to
the point of primary coagulant addition.  Direct filtration plants would have to report data on
recycle treatment to the State.

Alternative R4

Finally, the Agency considered requiring conventional filtration plants that recycle within the
treatment process to provide sedimentation or more advanced recycle treatment. This option
would not require direct recycle systems to perform a  self assessment, nor would it require direct

February 15, 2000                                3-11                      RIA for the Proposed LT1FBR

-------
filtration plants to report on their recycle practices.  Similarly, direct filtration plants would also
need to provide sedimentation or more advanced recycle treatment. Under Alternative R4,
systems would also still be required to return select recycle flows prior to the point of primary
coagulant addition.
                         Exhibit 3-4.  Filter Backwash Alternatives

Recycle return
location
Direct recycle
systems
Direct filtration
systems
Alternative Rl
Prior to primary
coagulant addition
No Provision
No Provision
Alternative R2
Prior to primary
coagulant addition
Report self
assessment to State
Report recycle
practices to State
Alternative R3
Prior to primary
coagulant addition
Equalization for
recycle flows
Report recycle
practices to State
Alternative R4
Prior to primary
coagulant addition
Sedimentation or better
for recycle flows*
Sedimentation or better
for recycle flows
* Note:  This requirement would apply to all conventional filtration systems that do not provide sedimentation or
       more advanced treatment for their recycle flows.
EPA considered all of the above alternatives and is proposing Alternative R2.  EPA concluded
that a national treatment requirement is inappropriate at this time due data deficiencies.  However,
the Agency believes that the available information supports requiring recycle to be returned prior
to the point of primary coagulant.  In addition, providing the States with information from the
direct recycle self assessments and the direct filtration recycle practices will aid them in targeting
recycle treatment in higher risk recycle practices.
RIAfor the ProposedLT1FBR
3-12
February 15, 2000

-------
                                4.  Baseline Analysis

This chapter discusses the methods used to identify the number of systems and populations
affected by each of the rule provisions.  Additionally, the sources, the resulting statistics, and any
assumptions developed for the analysis are identified.  Much of the data used to develop the RIA
are provided in the Occurrence Assessment for the Long Term 1 Enhanced Surface Water
Treatment and Filter Backwash Recycle Rule (U.S. EPA 1999a) and the Cost and Technology
Document for the Long Term 1 Enhanced Surface Water Treatment and Filter Backwash Recycle
Rule (U.S. EPA 1999b). However, additional data and assumptions were needed to complete the
economic analysis for this document and are discussed in this chapter.

4.1    Baseline Assumptions

The Agency developed estimates of the number of systems that would be affected by components
of the proposed  rule by utilizing three primary sources: Safe Drinking Water Information System
(SDWIS), Community Water Supply Survey, and WaterAStats.  A brief overview of each of the
data sources is provided in the following paragraphs.

       Safe Drinking Water Information System (SDWIS^
       SDWIS  contains information about public water systems including violations of EPA' s
       regulations for safe drinking water. Pertinent information in this database includes system
       name and identification number, population served, geographic location, type of source
       water, and type of treatment (if provided).  EPA utilized the 1997 State-verified version of
       SDWIS to develop the total universe of systems that utilize surface water or ground water
       under direct influence (GWUDI) as sources.

       Community Water System Survey
       EPA conducted the 1995 Community Water System Survey to obtain data to support its
       development and evaluation of drinking water regulations. The survey consisted of a
       stratified random sample of 3,700 water systems nationwide (surface water and ground
       water). The survey asked 24 operational and 13 financial questions.

       WaterA Stats (WaterStats^
       WaterStats is an in-depth database of water system information compiled by the American
       Water Works Association.  The database consists of 898 systems and provides a variety of
       data including treatment information.

System population characteristics are important to this analysis in several ways.  First, all systems
are categorized by the size of the population served. For this RIA, only small systems (i.e., serving
fewer than  10,000 people) were included for LT1 provisions.  Both small and large systems are
analyzed for the recycle provisions along with individually analyzing these provisions for systems
serving over 1,000,000. Systems are divided into the seven size categories used throughout the

February 15, 2000                              4-1                       RIA for the Proposed LT1FBR

-------
analysis; these categories are consistent with industry definitions of system size categories.  The
size categories are shown in Exhibit 4-1.
        Exhibit 4-1. System Population Size Categories and Total Population
System Size
25-100
101-500
501-1,000
1,001-3,300
3,301-9,999
10,000-50,000
50,001-100,000
100,001-1,000,000
over 1,000,000
Turbidity Provisions
•
•
•
•
•




Recycle Control
Provisions
•
•
•
•
•
•
•
•
•
Average and system design flow rates are integrated into the national compliance cost model in
estimating unit costs, determining treatment developed for compliance forecast or decision trees,
and sizing equipment.  Average and system design flows, expressed in millions of gallons per day
(mgd), were developed separately from the cost model but are key components in generating unit
costs (U.S. EPA, 1999b).  Flows are used to estimate equipment size, basin dimensions, filter bed
and media requirements, and energy costs.

Purchased water systems are included in this analysis even though many of the provisions of the
LT1FBR apply only to systems that treat their water. The Agency chose to include these
purchased water systems to estimate the total population affected by LT1FBR. Purchased water
systems must be included  because population served information in SDWIS represents only the
direct retail population served by each system. It does not include the total number of people using
water treated by a system  if the system sells water to other systems. Furthermore, the Agency
chose to include purchased water systems in the analysis as  a proxy to determine the amount of
treatment costs systems selling water to purchased water systems will face.  Costs are estimated
from water production flow equations based on population served information. Systems that only
sell water to other systems may have little or no population  served in SDWIS, therefore including
the purchased water systems accounted for the production of water sold to other systems.

Noncommunity water systems were assumed to have similar treatment characteristics to
community water systems in a given size category.  This assumption was made because specific
information on treatment technologies utilized by nontransient noncommunity (NTNC) and
RIAfor the Proposed LT1FBR
4-2
February 15, 2000

-------
transient noncommunity (TNC) water systems were not available. In some instances,
noncommunity water systems may in fact have less treatment in place than community water
systems.

4.2    Industry Profile
Data on systems and their capacity to achieve treatment levels were analyzed to develop the
national compliance cost estimate. The scope of this rule is confined to water systems that utilize
surface water or GWUDI as sources.  This universe consists of 11,593 systems serving fewer than
10,000 persons, and 2,096 systems serving 10,000 or more persons. Exhibit 4-2 provides the
number of systems using surface water or GWUDI in each size category for systems serving fewer
than 10,000 people. Exhibit 4-3 provides the same information for systems serving more than
10,000 people.

        Exhibit 4-2.  Systems Utilizing Surface Water or GWUDI Serving Fewer Than
                                      10,000 People
System Type
Community
Noncommunity
NTNC
Total
Population Served
<100
1,131
1,400
273
2,804
101-500
2,046
527
287
2,860
501-1,000
1,198
98
103
1,399
1,001-3,30
0
2,475
78
78
2,631
3,301-9,99
9
1,839
40
20
1,899
Total
Number of
Systems
8,689
2,143
761
11,593
             Exhibit 4—3.  Systems Utilizing Surface Water or GWUDI Serving
                                 10,000 or More People
System Type
Community
Noncommunity
NTNC
Total
Population Served
10,000-
50,000
1,539
18
4
1,561
50,001-
100,000
269
3
1
273
100,001-
1,000,000
245
0
1
246
> 1,000,000
16
0
0
16
Total
Number of
Systems
2,069
21
6
2,096
February 15, 2000
4-3
RIAfor the Proposed LT1FBR

-------
The rule has three main components—turbidity provisions, disinfection benchmarking provisions,
and recycle provisions. The LT1 provisions pertain to surface water and GWUDI systems serving
populations fewer than 10,000 that use filtration.  The recycle provisions of the rule include all
surface water and GWUDI systems that practice rapid granular filtration and recycle filtered water.

Primarily, the number of systems is derived from data collected from SDWIS, the CWSS (U.S.
EPA, 1997a), and WaterStats. These primary sources were supplemented with data from the 1996
Information Collection Rule (ICR) and the Water Industry Baseline Handbook (U.S. EPA
1999c).6

The 1996 ICR data provided information on the number of systems recycling water and the
locations of recycle return. In addition, WaterStats data was augmented by data from a survey
performed by the American Water Works Association (AWWA, 1998) and was used to
characterize the specific treatment processes used by systems that recycle.

Steps taken to identify the number of systems included for each rule provision are discussed below
along with the data used to develop that universe.  These data were used to quantify benefits and
develop national costs of the rule.


4.3    Number of Systems Under the Turbidity  Provisions

To determine the number of systems affected by the turbidity provision, the total number of
systems serving less than 10,000 people utilizing surface water and GWUDI was tabulated from
SDWIS and reported in the WIBH (U.S. EPA, 1999c). These systems were aggregated by system
size and by system type.  The WIBH does not provide detailed treatment information for systems.
To develop the number of systems that filter and those that use rapid granular filtration
(conventional or direct filtration) the WIBH values were multiplied by the percent of systems
practicing those filtration techniques as reported in CWSS (U.S. EPA,  1997a).

4.3.1   Estimate of the Number of Systems Subject to 2 log
       Cryptosporidium Removal Requirement

Using the baseline described in Exhibits 4-2 and 4-3, the Agency applied the percentages of
surface water and GWUDI systems that filter (as noted in the CWSS) to develop an estimate of the
number of systems that filter and serve fewer than 10,000 persons.  These percentages range from
78.5 percent for the smallest systems to 86.5 percent and are shown in  Appendix I.  This resulted
in an estimated 9,133 surface water and GWUDI systems that filter and would be subject to the
       6The Water Industry Baseline Handbook (WIBH) was developed by EPA to support the analyses
required under the 1996 SDWA amendments. To complete the analyses required under SDWA, EPA
developed the WIBH to serve as a single integrated set of data that defines baseline characteristics or
conditions of the regulated community, the customers, and governmental entities.


RIA for the Proposed LT1FBR                      4-4                              February 15, 2000

-------
proposed removal requirement.  Exhibit 4-4 provides this estimate broken down by system size
and type.
     Exhibit 4-4. Estimate of Systems Subject to 2 log Cryptosporidium Removal
     Requirement
System Type
Percent of
systems that filter1
Community
Noncommunity
NTNC
Total
Population Served
<100
78.5%
888
1,099
214
2,201
101-50
0
71.0%
1,453
374
204
2,031
501-1,00
0
79.3%
950
78
82
1,109
1,001-3,30
0
81.7%
2,022
64
64
2,150
3,301-9,99
9
86.5%
1,591
35
17
1,643
Total
Number of
Systems

6,903
1,649
581
9,133
Note: Columns and row might not add to total due to rounding
1. Source: CWSS
4.3.2  Systems Subject to Strengthened CFE Turbidity Standards

Using the estimate of 9,133 systems that filter and serve fewer than 10,000 persons, the Agency
used information in the CWSS database to estimate the number of systems that utilized specific
types of filtration.  The data were segregated based on the type of filtration and the population size
of the system. Percentages were derived for each of the following types of filtration:

       •    Conventional and direct filtration
           Slow sand filtration
           Diatomaceous earth filtration
           Alternative filtration technologies.

The percentages were applied to the estimate of the number of systems that filter for each of the
respective system size categories.  The percent of filtered systems that are conventional or direct
filtration range from 38 percent for the smallest size category to 90 percent for the largest size
category.  Based on this analysis, the Agency estimates 5,897 conventional and direct filtration
systems will be subject to the strengthened combined filter effluent turbidity standards. Exhibit
4-5 provides the number of conventional and direct filtration systems by system size category.
February 15, 2000
4-5
RIAfor the Proposed LT1FBR

-------
      Exhibit 4-5. Estimate of Systems Subject to Strengthened CFE Turbidity Standards
                        for Conventional and Direct Filtration Systems




System Type
Percent of filtration systems
that are Conventional or
Direct1
Community
Noncommunity
NTNC
Total
Population Served



<100

38.0%

337
418
81
836


101-50
0

55.0%

799
206
112
1,117


501-
1,000

73.0%

694
57
60
810


1,001-
3,300

77.0%

1,557
49
49
1,655


3,301-
9,999

90.0%

1,432
31
16
1,478
Total
Number
of System
s



4,819
760
318
5,897
Note: Columns and rows might not add to total due to rounding.
1. Source: CWSS.

EPA estimates 1,756 systems utilize slow sand or diatomaceous earth filtration, and must continue
to meet turbidity standards set forth in the SWTR.  The remaining 1,482 systems are estimated to
use alternative filtration technologies and will be required to meet turbidity standards as set forth by
the State upon analysis of a 2 log Cryptosporidium demonstration conducted by the system.

4.3.3  Estimate of the Number of Systems Subject to
       Individual Filter Monitoring Requirements

EPA believes that the support and underlying principles regarding the IESWTR individual filter
monitoring requirements are also applicable for the LT1FBR. The Agency has estimated that
5,897 conventional and direct filtration systems will be subject to the proposed individual filter
turbidity requirements. The Agency has analyzed information regarding turbidity spikes and filter
masking and concluded potential improvements in finished water quality justify individual filter
monitoring in addition to CFE monitoring.

Monitoring the performance of individual filters  within a treatment plant is important to ensuring
low turbidity in finished water.  Poor performance of one filter—accompanied by potential
pathogen breakthrough—can be masked by optimal performance in other filters, such that there is
no discernable increase in combined filter effluent turbidity. Individual  filters are also susceptible
to short turbidity spikes that are be captured by existing four-hour combined filter effluent
measurements. To address these shortcomings in large systems, EPA established individual filter
monitoring requirements in the IESWTR.  The Agency believes it's appropriate and necessary to
extend individual filter monitoring requirements to systems serving populations fewer than 10,000.
RIAfor the Proposed LT1FBR
4-6
February 15, 2000

-------
This provision applies to small surface water and GWUDI systems that use conventional or direct
filtration, which are reported above in Exhibit 4-5.

4.4    Systems Affected by Disinfection Benchmarking Provision


The disinfection benchmarking requirement applies to all small water systems utilizing surface
water or GWUDI that are community water systems or nontransient noncommunity water systems.
Exhibit 4-6 provides the number of systems in each system size category that will be required to
comply with the disinfection benchmarking requirement.
 Exhibit 4-6.  Estimate of Systems Subject to Disinfection Benchmarking Provision
System Type
Community
Noncommunity
NTNC
Total
Population Served
<100
1,131
0
273
1,404
101-50
0
2,046
0
287
2,333
501-1,00
0
1,198
0
103
1,301
1,001-3,30
0
2,475
0
78
2,553
3,301-9,99
9
1,839
0
20
1,859
Total
Number of
Systems
8,689
0
761
9,450
Columns and rows may not add to totals due to rounding
4.5    Systems Affected by the Recycle Provisions

To determine the systems affected by the recycle provisions, data from the Information Collection
Rule and the AWWA Survey (1998) were analyzed. The following section describes those data
sources and summarizes their results. Using the results from the analysis of the two data sources,
the baseline number of systems that will be affected by the provisions are estimated and presented
at the end of this section.

Information Collection Rule

Public water systems subject to the ICR were required to report whether recycle is practiced for
sample washwater (i.e., recycle flow) between the washwater treatment plant (if one existed) and
the point at which recycle is added to the process train. Sampling of plant recycle flow was
required prior to blending with the process train. Systems were also required to measure recycle
flow at the time of sampling, the twenty four hour average flow prior to sampling, and report
whether treatment of the recycle  was provided and, if so, the type of treatment.  Reportable
treatment types were plain sedimentation, coagulation and sedimentation, filtration, disinfection, or
a description of an alternative treatment type. Plants were also required to submit a plant schematic
February 15, 2000
4-7
RIAfor the Proposed LT1FBR

-------
to identify sampling locations.  EPA used the sampling schematics and other reported information
to compile a database of national recycle practices.  The results are summarized below.

4.5.1  Recycle Practice

The Agency developed a database from the ICR sampling schematics and other reported
information. Exhibit 4-7 summarizes the plants in the database.  Of the 502 plants in the database
at the time the analysis was performed, 362 used rapid granular filtration.


                     Exhibit 4-7.  Recycle Practice at ICR Plants
Plant Classification
All ICR Plants
Filtration Plants1
Filtration Plants Recycling2
Recycle Plants Serving • 100,000
Recycle Plants Serving < 100,000
Filtration Plants Treating Recycle
Number
502
362
226
168
58
148
These plants are classified as conventional, lime softening, other softening, and direct filtration.
The remaining 140 plants in the database do not employ rapid granular filtration capability and
generally provide disinfection for ground water. Of the 362 filtration plants in the database, 226
(62.4 percent) reported recycling to the treatment process.  Seventy-four percent of the plants that
recycle serve populations greater than 100,000 and 26 percent serve populations below 100,000.
Exhibit 4-8 shows the distribution of plants by treatment type and Exhibit 4-9 shows the
distribution of plants by population served. Exhibit 4-10 shows that 88 percent of ICR recycle
plants use surface water.  An additional 1  percent use GWUDI and another 1 percent use a
combination of ground water and surface water. Therefore, 90 percent of ICR recycle plants use a
source water that could contain Cryptosporidium.
RIA for the Proposed LT1FBR                       4-8                              February 15, 2000

-------
              Exhibit 4^8. ICR Recycle Plants by Treatment Train Type
                Exhibit 4-9. ICR Recycle Plants by Population Served
February 15, 2000
4-9
RIAfor the Proposed LT1FBR

-------
                 Exhibit 4-10. Source Water Use by ICR Recycle Plants
Source Water Type
Total Number of Recycle Plants
Surface Water
Ground Water Under the Direct
Influence
Ground Water and Surface Water
Ground Water Only
Number of
Plants
226
199
3
2
22
Percent of Recycle Plants
100%
88%
1%
1%
10%
 Exhibit 4-11 shows that 65 percent of ICR recycle plants report providing treatment for the
 recycle flow.  The percentage of plants providing treatment is the same for the subsets of plants
 serving greater than and less than 100,000 people. Sedimentation is the most widely reported
 treatment method, as 77 percent of plants providing treatment employ it. The database does not
 provide information on the solids removal efficiency of the sedimentation units. All direct filtration
 plants practicing recycle reported providing treatment for the recycle flow.
                  Exhibit 4-11.  Treatment of Recycle at ICR Plants*
ICR Recycling Plants
Number of Recycle Plants
Practice Recycle Treatment
Use Sedimentation
Use Sedimentation/Coagulation
Use Two or More Treatments
Other Treatment
Number of Plants
226
147
114
14
14
5
Percentage of Recycle
Plants
100%
65%
77%
10%
10%
3%
* Disinfection not counted as treatment because it does not inactivate Cryptosporidium..
 Exhibit 4-12 indicates that 75 percent of ICR recycle plants return recycle prior to the point of
 primary coagulant addition. Fifteen percent return it prior to sedimentation, and ten percent of
 plants return it prior to filtration.  These percentages hold for the subsets of plants serving greater
 than and less than 100,000 people.  The data indicate that introducing recycle prior to rapid mix
 may be a common practice. EPA believes that introducing recycle flow prior to the point of
 primary coagulant addition is the best recycle return location because it limits the possibility that
 residual treatment chemicals in the recycle flow will disrupt treatment chemistry.
 RIAfor the Proposed LT1FBR
4-10
February 15, 2000

-------
                          Exhibit 4-12. Recycle Return Point
Point of Recycle Return
Number of Recycle Plants
Prior to Point of Primary Coagulant
Addition
Prior to Sedimentation
Prior to Filtration
Number of Plants
224*
169
34
21
Percent of Plants
100%
75%
15%
10%
*Recycle return point could not be determined for two plants.

The data provide the following conclusions regarding the recycle practice of ICR plants:

•      The recycle of spent filter backwash and other process streams is a common practice

•      The great majority of recycle plants in the database use filtration and surface water sources

•      A majority of plants in the database that recycle provide treatment for recycle flow

       A large maj ority of plants in the database that recycle (approximately three out of four)
       return recycle flows prior to the point of primary coagulant addition.

Recycle FAX Survey

The AWWA sent a FAX survey (AWWA, 1998) to its membership in June 1998 to gather
information on recycle practices. Plants were not selected based on source water type, the type of
treatment process employed, or any other factor. The survey was sent to the broad membership to
increase the number of responses. Responses indicating a plant recycled spent filter backwash or
other flows were compiled to create a database.  The resulting database included 335 plants.  The
database does not contain information from respondents who reported recycle was not practiced.
Data from some of the FAX survey respondents is also included in the ICR database. Plants in the
database are well distributed geographically and represent a broad range of plant sizes as measured
by capacity.  Exhibit 4-13 shows plant distribution by capacity and Exhibit 4-14 by geographic
location. The following discussion of FAX survey data is divided into two sections. The first
discusses national recycle practice and the  second discusses options for recycle disposal in lieu of
returning recycle to the treatment process.
February 15, 2000
4-11
RIAfor the Proposed LT1FBR

-------
                 Exhibit 4-13.  Distribution of FAX Survey Plants by Plant Capacity
   Number of plants
   90
   80
   70
   60
   50
   40
   30
   20
   10
    0
                             
-------
 4
 I
                                                                                                                CD
                                                                                                                CD
                                                                                                                CD
                                                                                                                O
                                                                                                                C
                                                                                                                O)
                                                                                                                ^
                                                                                                                O
                                                                                                                c
                                                                                                                .c
                                                                                                                O
                                                                                                                CD
                                                         T3
                                                          C
                                                          CD

                                                          O)
                                                         ^C
                                                         'l_
                                                          CD
                                                          CD
                                                          C
                                                         'cn
                                                          c
                                                         LJJ
                                                                                                                c
                                                                                                                CD
                                                                                                                E
                                                                                                                C
                                                                                                                LJJ


                                                                                                                CD
                                                                                                                O

                                                                                                                D
                                                                                                                O
                                                                                                                C/)
February 15, 2000
4-13
RIAfor the Proposed LT1FBR

-------
Recycle Practices of FAX Survey Plants

Data summarized in Exhibit 4-15 show that 78 percent of plants in the database rely on a surface
water as their source. The percentage of plants using source water influenced by a surface water
(which may contain Cryptosporidium) could be higher because the data do not report whether
wells were pure ground water or GWUDI.
                Exhibit 4-15.  Source Water Used by FAX Survey Plants
Source Water Type
Surface Water
River
Reservoir
Lake
Other
Well*
Percent of Plants
78%
27%
28%
16%
7%
22%
* Wells sources can be either ground water or ground water under the direct influence of surface water.
Exhibit 4-16 shows that a wide variety of treatment process types are included in the data, with
conventional filtration (rapid mix, coagulation, sedimentation, filtration) representing over half of
the plants submitting data.  Upflow clarification is the second most common treatment process
reported.  Ten percent of plants in the database use direct filtration.  Only 4 percent of plants do
not use rapid granular filtration.
                 Exhibit 4—16. Treatment Trains of FAX Survey Plants
Treatment Process Type
Conventional filtration
Upflow Clarifier
Softening
Direct Filtration
Other
Percent of Plants*
51%
21%
14%
10%
4%
 : 96 percent of plant in the database provide filtration.
Exhibit 4-17 indicates that a vast majority of plants recycle prior to the point of primary coagulant
addition. Only 6 percent of plants returned recycle in the sedimentation basin or just prior to
filtration.
RIAfor the Proposed LT1FBR
4-14
February 15, 2000

-------
              Exhibit 4-17. Recycle Return Point of FAX Survey Plants
Return Point
Prior to Point of Primary Coagulant Addition
Pre-sedimentation
Sedimentation basin
Before filtration
Percent of Plants
83%
11%
4%
2%
Exhibit 4-18 shows that the majority of plants in the database provide some type of treatment for
the recycle flow prior to its reintroduction to the treatment process. Approximately 70 percent of
plants reported providing treatment, with sedimentation being employed by over half of these
plants. Equalization, defined as a treatment technology by the survey, is practiced by 20 percent of
plants in the database. Fourteen percent of plants reported using both sedimentation and
equalization.
                Exhibit 4-18. Recycle Treatment at FAX Survey Plants
Treatment Type
No Treatment
Treatment
Sedimentation
Equalization
Sedimentation and Equalization
Lagoon
Others
Percent of Plants
30%
70%
54%
20%
14%
5%
7%
Exhibit 4-19 summarizes recycle treatment practice and frequency of direct recycle based on
population served. The table illustrates that, for plants supplying data, treatment of recycle with
sedimentation is provided more frequently as plant service population deceases. Plants serving
populations of less than 10,000 direct recycle (23 percent) less frequently than plants serving
populations greater than 100,000 (42 percent).  The data indicate that a majority of small plants in
the database may have installed equalization or sedimentation treatment to protect treatment
process integrity from recycle induced hydraulic disruption.  All direct filtration plants in the FAX
survey provide recycle treatment or equalization.
February 15, 2000
4-15
RIAfor the Proposed LT1FBR

-------
             Exhibit 4-19. Recycle Practice Based on Population Served1
Population Served
< 10,000
10,000-50,000
50,001-100,000
> 100,000
Recycle Practice
Number
of
Plants
43
79
35
65
Equalization2
4 (9%)
8 (10%)
6 (17%)
23 (35%)
Sedimentation2
29 (67%)
45 (57%)
19 (54%)
15 (23%)
Direct Recycle
10 (23%)
26 (33%)
10 (29%)
27 (42%)
 1 Based on 222 surface water plants suppling all necessary data to make determination.
 2 Number of plants (percent of plants) in category.
FAX survey data support the following conclusions regarding the recycle practice of plants
supplying data:  1) the recycle of spent filter backwash and other process streams is a common
practice, 2) the majority of recycle plants use surface water as their source and are thereby at risk
from Cryptosporidium, 3) a large majority of plants providing data recycle prior to the point of
primary coagulant addition, and 4) a majority of plants supplying data provide treatment for
recycle waters prior to reintroducing them to the treatment plant. The FAX survey provides an
informative snapshot of national recycle practices due to the number of recycle plants it includes,
the geographic distribution of respondents, and the good representation of plants serving
populations of less than 10,000 people.

Options to Recycle

The  FAX survey collected information about: 1) whether feasible alternatives to recycle are
available (i.e., National Pollutant Discharge Elimination System (NPDES) surface water discharge
permit or pretreatment permit for discharge to a Publicly Owned Treatment Work (POTW)), 2) the
importance of recycle to optimizing treatment performance and meeting production requirements,
and 3) whether recycle flow monitoring had been performed. The related responses are
summarized in Exhibit 4-20.

Exhibit 4-20 shows that approximately 20 percent of respondents could not obtain either an
NPDES surface water discharge permit or a pretreatment permit for discharge to a POTW.
Approximately 85 percent of respondents stated that recycle flow is not important to meet typical
demand. Twenty-four percent of all respondents stated that returning recycle to the treatment
process is important for optimal operation. "Optimal operation" was not defined by the survey and
respondents may have considered not changing current plant operation (e.g., not changing current
recycle practice) an aspect of optimal treatment, rather than addressing whether recycle practice is
important for the plant to produce the highest quality finished water.
RIAfor the Proposed LT1FBR
4-16
February 15, 2000

-------
        Exhibit 4-20. Options to Recycle as Reported by FAX Survey Plants*
Question
Able to obtain NPDES surface
discharge permit?
Able to obtain pretreatment
permit for POTW discharge?
Can obtain either an NPDES or
a POTW discharge permit?
Is recycle important to meet
peak demand?
Is recycle important to meet
typical demand?
Is recycle important to optimal
operation? (All plants in survey)
Is recycle important to optimal
operation?** (softening plants
only)
Yes (Percent)
131 (41%)
137 (43%)
192 (60%)
44 (14%)
28 (9%)
75 (24%)
3 (13%)
No (Percent)
120 (37%)
136 (42%)
63 (19.5%)
257 (80%)
272 (85%)
225 (70%)
19 (83%)
Unknown
(Percent)
70 (22%)
48 (15%)
66 (20.5%)
20 (6%)
21 (6%)
21 (6%)
1 (4%)
* Number of plants varies from question to question due to different response rates.
"Optimal operation not defined by survey. May include overall plant operation rather than importance of recycle to
producing highest possible quality finished water.

Summary of Analysis

The ICR and FAX survey data are complimentary, as the ICR data supplies a wealth of data
regarding recycle practices at large capacity plants, while the FAX survey provides data on recycle
practices over a range of plant capacities.  Taken together, the two data sets provide a good picture
of current recycle practice. The data indicate that recycle is a common practice for the plants
sampled. Approximately half of the respondents providing data return recycle flow to the
treatment process and 70 percent provide some type of recycle treatment. Sedimentation and
equalization are the two most commonly employed treatment technologies for plants supplying
data.  Approximately 80 percent of plants sampled return recycle prior to the point of primary
coagulant addition. Examining the recycle practices of plants in the ICR and FAX survey data
shows that small plants (i.e., fewer than 10,000 people served) are more than twice as likely as
large plants (i.e., greater than 10,000 people served) to provide sedimentation for recycle treatment
(58 versus  26 percent).

The FAX survey responses show that approximately half of plants providing data have an option
to recycle return, whether it be an NPDES surface water discharge permit or discharge to  a
POTW. Eighty percent of respondents stated that recycle flow is not important to meet peak
February 15, 2000
4-17
RIAfor the Proposed LT1FBR

-------
demand. Less than a quarter of respondents have monitored pathogen concentrations in recycle
return flows and fewer than half have any monitoring data to characterize the quality of the recycle
return flows.

The proposed filter backwash recycle provisions apply to both large and small surface water
systems. The number of systems is a subset of the rapid granular filtration systems discussed under
the turbidity provisions of the LT1FBR and systems regulated under the IESWTR promulgated in
December 1998. The total number of conventional and direct filtration systems was multiplied by
the percent of systems practicing filtration techniques cited in the CWSS. EPA estimates that 60
percent of these systems recycle filter backwash (Cornwell and Lee, 1994).  Exhibit 4-21 provides
the total number of systems, those practicing filtration and those recycling filter backwash water.

Data from the ICR and FAX survey  indicate that 75 and 83 percent of plants, respectively, return
recycle prior to the point of primary  coagulant addition.  The "point of primary coagulant addition"
was defined in both analyses as the return of recycle prior to the rapid mix unit.  The FAX survey
data indicate that 77 percent of plants serving under 10,000 people recycle prior to the point of
primary coagulant addition. It also showed that 83 percent of all plants in the database return
recycle there, which suggests that plants serving smaller populations may return recycle prior to the
point of primary coagulant addition as frequently as plants serving larger populations.

The number of systems treating recycle flows is derived from the percentages estimated in Exhibit
4-19.  The percentages in the FAX survey were used to estimate the number of systems present in
Exhibit 4-21 that are direct recycle systems and the number of systems that treat recycle with
equalization or sedimentation/clarification.

Twenty-three direct filtration plants that used surface water responded to the FAX survey. In the
FAX survey, plants could report whether they provide recycle flow equalization, sedimentation, or
some other type of treatment.  Of the respondents, 21 reported providing treatment for the recycle
flow and two plants reported providing only equalization. In the ICR database, there were 23
direct filtration plants  and 14 of them recycled to the treatment process. All  fourteen plants provide
recycle treatment.  It is not possible to determine the level of oocyst removal that FAX survey and
ICR plants achieve with available data.

Similar to the methodology used for  the turbidity provision, the number of systems identified in the
WIBH were multiplied by the percent of systems practicing specific treatment and recycling
practice to provide a distribution of systems by size, type, and treatment practice.

Exhibit 4-21 provides the number of systems for each of the treatment and recycle practice and the
multiplier used to derive the value. It is important to note specific treatments are subsets
of previous developed subsets. For example, systems that practice direct recycle return,
equalization of recycle water, or sedimentation of recycled water are a subset of the conventional
filtration systems that recycle.
RIA for the Proposed LT1FBR                       4-18                              February 15, 2000

-------






=
o
Q_
^•^
—i*
>>
U
*
J2
•o
u
w
i
4*
K
O
•
o>
s
*
.=
^^
o
^
•_
SS
s
s
s
C^
^H
fs|
^
,£
'.c
W











o °r,
o o
^
08
^ as"
fO
^ 0
Ofo
Cv
o
o
o
^H
^
o
l/^
o
o
*?
o
^H

o
o
^H
0
a
H
i






o
o
-o
o
o
-o


i
o
-o


o
o
-o

i
o
-o
i
o
-o


\0
o
o
-o


\p
o
o
-o
Percent of Direct or Conventional
filtration systems that recycle
>o
m
vq_
^


(N
5




m
oo




oo
oo


f~>
ON
ON


OO
^^




o
(^^
-o



(N
O


| Number of Systems I












































o
m
(N
O
m
(N


i
m
(N


O
m
(N

i
m
(N
o
m'
(N


\o
O
m'
(N


\p
O
m
(N
b
o^
o^
15
^
t«
§
•£
5?
<+H
0 X
S '§
B^
SI
VO
VO

^H


ON
(N
m




ON




O
(N


oil
(N
(N


S




•^T
^



i/-j
,


I
GO
5












































O
(N
O
ON
(N


i
m
m


O
(N

i
m
(N
o
m'
(N


\o
O
m'
(N


\p
o
m
(N
Percent of conventional systems with
direct recycle
ON
ON
O
^H


O
06
m




(N




O
ON


(N


O




m
2



[^^
o


1 Number of Systems 1












































o
m
o
o'
^H


i
O



o
ON


i
ON

i
ON



\o
O
ON


\p
o
ON

Percent of conventional systems that
equalize recycle flows
t^
m
^^



^^
(N




[^^




f^«


m
oo


^




*o
i^



(N



1 Number of Systems 1












































o
m'
(N
O
•^-'
i/^


i
o'



o
06
VO

i
06
VO
i
06
VO


\0
o
06
VO


\p
o
06
VO
Percent of conventional systems that
treat recycle flows with sedimentation
i/~}
f^
t~~
(N


(N
r^




^^
^^




VO


oil
(N
VO


O
m




i^-
(N



f^
m


1 Number of Systems 1












































o
o'
o
o'



o
[^^



o
o'


o
£^

o
f^^



x°
O
^


\p
o
o'

1 Direct Filtration Systems that Recycle I
v^
(N
m



ON
O




oo




(N
VO


^


m




^
^^



i/~j
m


1 Number of Systems 1

February 15, 2000
4-19
RIAfor the Proposed LT1FBR

-------
4.6    Contaminant Exposure

The proposed LT1FBR rule will increase the level of protection to public health through
reductions in turbidity and waterborne pathogens, particularly Cryptosporidium.  The LT1FBR
turbidity provisions are the small drinking water system companion piece to IESWTR. The
proposed provisions are intended to reduce turbidity, which is indicative of a more efficient
filtration process (Rose, 1997). The LT1FBR recycling provisions apply to all drinking water
systems, large and small, that recycle flows including filter backwash as described in Chapter 3.
EPA's data indicates that current spent filter backwash and other recycle streams may introduce
Cryptosporidium oocysts, in excess of oocysts present in the  source water, to the treatment process
(U.S. EPA, 1999a).  Oocysts added to the treatment process through recycle water may increase
the risk of oocysts occurring in finished water supplies, and thereby threaten public health.

Several sources were used to assess the health effects and hazards posed by Cryptosporidium in
drinking water.  Data from the Centers for Disease Control and Prevention (CDC) provided the
number of reported outbreaks and resulting cases of cryptosporidiosis (Centers for Disease
Control, 1996).  Other publications provided information on symptoms and the incidence of
hospitalization and fatalities for the Milwaukee outbreak (Mackenzie, et al., 1994). Information on
the toxicity, dose-response relationship, and ingestion assumptions were derived from recent peer-
reviewed articles (see Chapter 5).  These sources described recent studies on the infection and
illness in human volunteers subjected  to controlled exposure to oocysts of Cryptosporidium to
arrive at an estimate of the risk and toxicity of Cryptosporidium.

The analysis described in Chapter 6 of the Occurrence Assessment for the LT1FBR (U.S. EPA,
1999a), which includes a characterization of national finished water Cryptosporidium distribution,
was used to assess the population exposure to Cryptosporidium in finished water supplies.

Estimating the benefits of reducing exposure to Cryptosporidium requires performing a risk
assessment to determine the number of illnesses reduced by the rule and then assigning a value to
those reductions.  Risk assessments require information on health effects, toxicity, and exposure.
Benefits analysis requires information on the value of reducing health and other potential damages.
Data to estimate the benefits associated with reducing health damages (cost-of-illnesses avoided)
were derived from previous survey research on the costs for a giardiasis outbreak (Harrington, et
al., 1985 and 1989). The data and any assumptions used to complete the benefits analysis are
detailed in Chapter 5.

For each rule component the same methodology used to develop the number of systems was
applied to population statistics. Appendix I provides the potential populations used in the risk
assessment and benefits analysis in Chapter 5.  Each exhibit shows the progression used to develop
the exposed populations.  Population data were taken from the WIBH (U.S. EPA, 1999d) and
further developed using percentages from CWSS (U.S. EPA, 1997a).
RIA for the Proposed LT1FBR                      4-20                              February 15, 2000

-------
                                 5.  Benefits Analysis


5.1    Introduction


The health benefits of a drinking water standard come from reducing the probability that
consumers will suffer health damages and other losses. The value of the benefits is captured in the
consumer's willingness-to-pay (WTP) for the change in drinking water quality (Freeman, 1979).
Often, this value is estimated to be the health damages (medical cost and lost productivity) that will
be avoided as a result of enforcing the drinking water standard—referred to as 'cost-of-illness'
(COI). COI measures, however, are thought to understate total benefits because they do not
capture the full value that consumers place on reducing risk and avoiding illness. This chapter
describes how avoided health injuries are estimated using a risk assessment approach, and how
those injuries are valued using WTP or COI estimates from the economic valuation literature. The
proposed Long Term 1  Enhanced Surface Water Treatment and Filter Backwash Rule (LT1FBR)
contains turbidity, recycle, and other provisions intended to increase the level of protection to
public health through reductions in waterborne pathogens, particularly Cryptosporidium.  This
section of the document discusses the proposed rule's turbidity and recycle provisions and the
benefit that may be realized from each provision.  Section 5.2 discusses and monetizes the health
benefits associated with reducing human exposure to Cryptosporidium in regulated drinking water
systems through implementation of the proposed LT1FBR turbidity provisions. Sections 5.3 and
5.4 describe qualitative benefits associated with the proposed rule, including benefits from the
turbidity provisions, recycle provisions,  disinfection benchmark provisions, requirements for
covers on new finished water reservoirs, and inclusion of Cryptosporidium in the definition of
ground water under the direct influence (GWUDI) and in the watershed control requirements for
unfiltered public water systems. Section 5.5 concludes the benefits chapter with a summary of
calculated annual benefits from the proposed rule and a discussion of how omissions, biases, and
uncertainties may affect the results of the benefits analysis.

5.1.1   Expected Benefits from Turbidity Provisions

The proposed LT1FBR turbidity provisions  are the small drinking water system companion pieces
to IESWTR.  These provisions will reduce finished water turbidity, which is indicative of a more
efficient filtration process (Rose, 1997). Improved removal of Cryptosporidium and other
waterborne pathogens is likely to occur as the filtration process improves (U.S.  EPA, 1999a),
resulting in reduced endemic illnesses and associated health benefits, as well as  other non-health
related benefits (Exhibit 5-1).  The benefits of improved filtration that have been quantified and
monetized in this analysis are due to the decreased probability of cryptosporidiosis, the infection
caused by Cryptosporidium. Reduced exposure to other pathogenic protozoa, such as Giardia, or
other waterborne bacterial or viral pathogens, are additional benefits of the proposed LT1FBR
turbidity provisions that have  not been quantified. Furthermore, additional benefits of reduced
averting costs (e.g., purchasing bottled water or boiling tap water) associated with improvements in
drinking water quality were not quantified because of the difficulties in making  such assessments.
February 15,2000                                5-1                     RIA for the Proposed LT1FBR

-------
     Exhibit 5-1. Overview of LT1FBR Benefit Categories and Associated Components
Health Benefits
Reduced illness (morbidity and
mortality)
• Reduced risk of Cryptosporidium and other disinfection resistant
pathogens occurring in finished water (endemic and outbreak-related)
Non-Health Benefits
Avoided costs of averting behavior
Enhanced aesthetic water quality
Avoided outbreak responses
• Bottled water and point-of-use (POU) devices
• Improved perception of drinking water quality
• Avoided costs to affected water systems and local governments
(provision of alternative water, issuing warnings and alerts, and costs
associated with negative publicity)
• Time spent on averting behavior during outbreaks, e.g.,
hauling/boiling water
5.1.2  Expected Benefits from Recycle Provisions

The proposed LT1FBR recycle provisions apply to large and small surface water and GWUDI
drinking water systems that recycle treatment process flows within the primary treatment process.
Benefits associated with the recycle provisions are similar to the benefits outlined in Exhibit 5-1.
EPA's research indicates that spent filter backwash and other recycle streams may introduce additional
Cryptosporidium oocysts to the treatment process.  Since Cryptosporidium is not inactivated by
standard disinfection practice, any oocysts returned to the treatment process in recycle flow are a threat
to enter the finished water and cause disease.  Further, hydraulic and chemical treatment disruption
caused by recycle flow may lower log removal performance, increasing the public health risk from
oocysts in finished water supplies.

The proposed rule contains three recycle provisions.  First, all plants must return spent filter backwash,
thickener supernatant, and liquids from dewatering prior to the point of primary coagulation addition.
This ensures that recycle flows pass through as many physical removal processes as possible to provide
maximum opportunity for oocyst removal and maintain the integrity of chemical dosing.  Second, plants
meeting specific criteria must perform a self assessment to determine the impact of recycle flows on
plant operations.  Results from the self assessment must be reported to the State. The self assessment
and reporting process will identify plants that may challenge oocyst removal performance by
exceeding design capacity during recycle events and  allow States to require changes to recycle
practices to protect public health.  Third, direct filtration plants must report their recycle practices to
the State, including whether flow equalization or treatment is provided for recycle flow prior to its
return to the treatment process.  The purpose of this requirement is to ensure that the recycle
practice of direct filtration plants is assessed to determine whether existing plant practice addresses
the potential risk posed by  recycle. The improved recycle practices under the LT1FBR will
reduce the public's exposure to Cryptosporidium and other waterborne pathogens in drinking
water, thus resulting in public health benefits.
RIAfor the Proposed LT1FBR
5-2
February 15, 2000

-------
5.1.3  Expected Benefits from Other Provisions

The proposed LT1FBR contains three additional provisions that provide positive health benefits to
customers of small drinking water systems:

       •   Disinfection benchmark provisions
       •   Requirements for covers on new finished water reservoirs
       •   Inclusion of Cryptosporidium in the definition of GWUDI and in the watershed
           control requirements for unfiltered public water systems.

Disinfection benchmarking provisions ensure continued microbial protection of drinking water
while facilities take the necessary steps to comply with new disinfection byproduct standards. The
disinfection benchmarking requirements are designed to ensure that there will be no unintended
reduction in microbial protection as a result of significant modifications to disinfection practices
that may be made to reduce DBFs.  The proposed rule requires that all new potable finished water
reservoirs serving small drinking water systems be covered to prevent contamination from various
sources including:  animals, microbes, algae, swimmers, and storm water run-off  The proposed
rule also includes Cryptosporidium in the definition of GWUDI and in watershed protection
regulatory requirements.


5.2    Health Benefits from Turbidity Provisions

Section 5.2 describes the risk assessment approach, the dose-response equation used for hazard
identification, and the exposure assumptions used to estimate Cryptosporidium risk to populations
served by regulated drinking water  systems. This section also presents the health and economic
benefits that accrue from the proposed LT1FBR turbidity provisions.  Benefits are estimated  on an
annual basis because benefits and costs occur concurrently in the proposed rule (i.e., changing
treatment and monitoring practices immediately reduces health risks) and remain relatively constant
after the start up period.  Thus, the annual approach generates a benefit/cost ratio that is
comparable to using the net present value (NPV) approach.  Furthermore, annual net benefits can
be used to derive an estimate of NPV of net benefits over a complete policy horizon.
Consequently, the use of annual benefits provides comparable results and streamlines the analysis.

5.2.1  Contaminants and Their Health Effects

Drinking water supplies can be contaminated by a number of pathogens that have been identified
as the cause of waterborne disease outbreaks (Centers for Disease Control, 1996).  In particular,
drinking water supplies contaminated with the parasite Cryptosporidium pose a health risk to the
public because the parasite is highly infectious, resistant to inactivation by chlorine, widespread
among many animal species, and small in size and consequently difficult to filter (Guerrant, 1997).
This benefits analysis of the proposed rule estimates the potential benefits of reducing human
exposure to Cryptosporidium in drinking water supplies through improved operation and
performance of the drinking water filtration process. In addition, nonquantified public health
benefits from the proposed rule include reduced  exposure to Giardia lamblia and other emerging
pathogens in drinking water.

February 15,2000                                5-3                     RIA for the Proposed LT1FBR

-------
The presence of Cryptosporidium in surface water sources is relatively common. Exhibits 2-4 and
2-5 provide a summary of the current research and information available on the occurrence of
Cryptosporidium. The ranges of concentrations cited in the 45 studies in these exhibits describe
source and finished water. Cryptosporidium concentrations in rivers, creeks, and streams range
between 0 and 417 oocysts per liter.  Results from lake and reservoir studies show
Cryptosporidium concentrations to range between 0 and 22 oocysts per liter. Researchers have
identified Cryptosporidium concentrations in finished water of up to 0.57 oocyst per liter.
Additional information on the level and occurrence of Cryptosporidium in surface and finished
water can be found in Chapter 2 and the Occurrence Document (U.S. EPA, 1999a).

Because Cryptosporidium is exceptionally resistant to inactivation by chlorine, physical removal
by clarification and filtration is extremely important to control this organism. Because of the
turbidity provisions in the proposed rule, many water systems would be expected to place an
increased emphasis on improving overall filtration performance.  The result of improving overall
and individual filter performance will be a reduction in the number of Cryptosporidium oocysts
that make it through the treatment process to finished water supplies with the ability to infect
humans and cause illness. In addition to improving overall filter performance, monitoring
requirements for individual filters in the proposed rule will ensure that water treatment plant
operators can identify problems with the filters and subsequently improve the performance of
individual filters.

Ingesting Cryptosporidium oocysts can cause cryptosporidiosis, which is an acute, self-limiting
illness lasting 7 to 14 days with symptoms that include diarrhea, abdominal cramping, nausea,
vomiting,  and fever (Juranek, 1995). There is no effective treatment for cryptosporidiosis
(Guerrant,  1997).

Several subpopulations are more sensitive to cryptosporidiosis, including the young, elderly,
malnourished, disease impaired (especially those with diabetes), and a broad category of those
with compromised immune systems (Rose,  1997). Subpopulations with compromised immune
systems include AIDS patients, those with Lupus or cystic fibrosis, transplant recipients, and those
on chemotherapy (Rose, 1997). Symptoms in the immunocompromised subpopulations are much
more severe, including debilitating voluminous diarrhea that may be accompanied by severe
abdominal cramps, weight loss, malaise, and low grade fever (Juranek, 1995).  Mortality is a
substantial threat to the immunocompromised infected with Cryptosporidium:

       "The duration and severity of the disease are significant: whereas 1 percent of
       the immunocompetent population may be hospitalized with very little risk of
       mortality (< 0.001), Cryptosporidium infections are associated with a high rate
       of mortality in the immunocompromised (50 percent)" (Rose, 1997).

Waterborne disease outbreak data from the Centers for Disease Control (CDC) for the period
1993-1994 estimates that Cryptosporidium was responsible for over 400,000 cases of
gastrointestinal infection (Craun et al.,  1998). The vast majority of these  cases occurred in one
outbreak in Milwaukee, Wisconsin, the largest recorded outbreak of waterborne disease in the
United States. Using standard epidemiological methods for estimating cases of illness, CDC
estimated that of the approximately 800,000 persons served by the water  system, over 400,000 (50
RIA for the Proposed LT1FBR                      5-4                               February 15, 2000

-------
percent) became ill (Exhibit 5-2).  Of those, 4,000 required hospitalization (approximately
1 percent of those becoming ill), and there were at least 50 cryptosporidiosis-associated deaths
among immunocompromised individuals (as reported on death certificates) (Mackenzie et al.,
1994; Hoxie et al., 1997). Exhibit 5-2 contains detailed information on some of the symptoms of
patients with cryptosporidiosis observed during the Milwaukee outbreak.

 Exhibit 5-2.  Symptoms of 205 Patients with Confirmed Cases of Cryptosporidiosis during
                                 the Milwaukee Outbreak
Symptom
Diarrhea
Abdominal Cramps
Weight Loss
Fever
Vomiting
Percent of Patients
93
84
75
57
48
Mean
Duration: 12 days
N/A
10 pounds
100.9'F
N/A
Range
1-55 days
N/A
1^10 pounds
99.0«-104.9«F
N/A
 Source: Mackenzie et al., 1994


Although the Milwaukee outbreak represents the largest number of cases in a single
cryptosporidiosis outbreak in the United States, most cryptosporidiosis outbreaks have occurred in
small systems serving fewer than 10,000 persons (Exhibit 2-7).  Between 1991 and 1996 there
were  16 small water system outbreaks caused by either Cryptosporidium or Giardia lamblia
resulting in 1,036 reported cases of cryptosporidiosis and 518 reported cases of giardiasis (U.S.
EPA  1999a). Two of the 16 outbreaks were associated with Cryptosporidium in small surface-
water systems, and four Cryptosporidium outbreaks occurred in ground water assumed to be under
the direct influence of surface water (see Chapter 2) (U.S. EPA, 1999a).  During small system
outbreaks, the rate of morbidity (i.e., the percent of the exposed population becoming ill) varies
from  8 to 80 percent of the exposed population.

Outbreak data represent only a portion of the incidence of cryptosporidiosis. Only large outbreaks
of cryptosporidiosis cases concentrated in a specific location have a chance of being detected and
reported. Isolated cases (endemic) are much less likely to be reported. Many, perhaps most,
infected individuals may not seek medical treatment for their symptoms.  If the infected individuals
do seek medical treatment, primary care physicians may not be able to isolate Cryptosporidium as
the cause of the illness. If diagnosed, physicians may not report the information to the CDC.
These compounded impacts could lead to gross under-reporting and under-estimating of
cryptosporidiosis cases (Okun et al., 1997).

5.2.2   Risk Assessment: Methods and Assumptions

Risk assessment is an analytical tool that can be used to characterize and estimate the potentially
adverse health effects associated with exposure to an environmental hazard, in this case
Cryptosporidium (Rose, 1997). The risk assessment developed by Rose was used to estimate
potential benefits and follows a standard methodology employed by EPA and the Federal
government (National Research Council, 1983).  The standard methodology requires the use of
scientific data or, if data are not available, reasonable assumptions to produce estimates when there
is considerable uncertainty about the exact nature, extent, and degree of the risk. This particular
February 15, 2000
5-5
RIA for the Proposed LT1FBR

-------
health risk assessment makes use of ranges and probability distributions to take into account
scientific uncertainty.

Risk assessment generally involves three basic steps (National Research Council, 1983).

•      Hazard identification identifies the potential health effects associated with exposure to the
       hazard and the exposure threshold (e.g., dose) above which the health effects may occur.

       Exposure assessment estimates the number of people exposed to the hazard and the level
       of exposure.

•      Risk characterization combines the hazard identification and exposure assessment to
       characterize overall risk to the exposed population.

The three possible health endpoints to risk characterization are infection, illness (morbidity), and
death (mortality). For the purpose of deriving benefits estimates, this analysis calculates the
number of illnesses and the associated number of premature deaths attributable to infection from
Cryptosporidium. Exhibit 5-3 displays the steps in the risk assessment process for characterizing
the endemic risk of morbidity and mortality from Cryptosporidium in drinking water.

To quantify the health effects due to Cryptosporidium  in drinking water, the following input
variables are necessary:

       Ingested dose (concentration of oocysts in the daily ingestion of finished water)
•      Percent of ingested oocysts that are viable
•      Dose-response function, which relates ingestion to infection
•      Morbidity rate resulting from the infection
•      Size of population exposed to Cryptosporidium in drinking water.

The following sections describe the assumptions and derivation of these variables used in the risk
assessment.
RIA for the Proposed LT1FBR                       5-6                               February 15, 2000

-------
             Exhibit 5-3.  Steps in the Health Risk Assessment for Cryptosporidium
                       Hazard
                     Identification
  Exposure
Assessment
      Risk
Characterization
                          ;. it  Intss
                           efmvtali*
                                      X
                          Heahh Bids (*jf
                                                     1 .*i ri. -
                                     Ctnc tnt-jl jn jl
                                 r=nnec er n%tmtc1  ir tig
                                   Irestnait ("icvsts/llier)
           Cincmtrstinn ef
        Crj^asAU*»»i? £f£^
       i lii ehH li 
-------
Hazard Identification

Hazard identification characterizes the incidence of the health effect in relationship to the dose
administered (dose-response relationship).  Dose-response information for Cryptosporidium is
represented by the following general model defining the probability of infection in an individual
given a single exposure to a dose of Cryptosporidium (Haas et al., 1996):

                                   •   = 1 • exp (•£>/£).

Where:              •  = probability of infection
              D = average dose
              k = slope parameter (relation of ingestion to infection)

The benefit analysis evaluates the effect of daily exposures to drinking water because an individual
could be repeatedly exposed to and infected by Cryptosporidium over the course of a year.
Calculating the overall probability of being infected only once, or twice, or three times, or some
other multiple over the course of a year is more difficult than calculating the probability of not
being infected at all.  The probability of not being infected in a single exposure, • n is:

                     . n = l- •  = l •  [l •  exp (• D / k )] = exp (• D / k).

The probability of not being infected at all during 365 exposures over the course of a year is:

                                 (• n)365 = [exp(-D/£)]365.

Finally, the probability of being exposed at least once in a year is

                              1 * (* n )365 = 1 *  [exp (• D/£)]365.
The probability of being exposed to Cryptosporidium at least once in a year depends upon the
number of exposures (i.e., not all individuals will be exposed 365 times over the course of the
year). EPA's model accounts for this by calculating exposure probabilities for three types of
public drinking water systems: community water systems (CWSs); nontransient noncommunity
water systems (NTNCs);  and transient noncommunity water systems (TNCs). Exhibit 5-4
describes the three types of public drinking water systems and the annual number of exposures
modeled for individuals served by these water systems.
RIA for the Proposed LT1FBR                       5-8                              February 15, 2000

-------
        Exhibit 5-4.  Cryptosporidium Exposure Probabilities and Characteristics of
                              Public Drinking Water Systems
Type of
Water System
Community Water System
Nontransient
Noncommunity Water
System
Transient Noncommunity
Water System
Annual Number of Exposures
to Crytosporidium
350
250
10
Defining
Characteristics
CWSs supply water to the same population
year-round.
NTNCs regularly supply water to at least 25 of
the same people for at least six months per
year, but not year-round.
TNCs supply drinking water in places where
people do not remain for long periods of time.
Using lognormally distributed data from human ingestion trials of Cryptosporidium parvum
(C. parvum) Iowa strain, the estimated best fit value for & is 238.6, with a 90 percent confidence
interval of 132.0 to 465.4 (Haas et al., 1996). Infection was defined as excretion of oocysts in the
stool 36 hours or longer following the challenge dose.  These trials were conducted with 29
healthy, medically screened individuals. Consequently, the slope parameter k may be different for
sensitive subpopulations (i.e., it is possible that a lower dose may induce a response in sensitive
individuals equivalent to what a higher dose induces in healthy individuals).

Infectivity varies among isolates of C. parvum., the Cryptosporidium species that is infectious to
humans.  In a comparison of the "Iowa" isolate of C. parvum used in the original human challenge
studies by DuPont et al. (1995) with two other C. parvum isolates, Chappell et al. (1997) observed
similar incubation periods and duration of illness among all isolates, but a 1 log lower ID50 (the average
dose required to infect 50 percent of exposed persons) for the TAMU isolate of C. parvum.  A third
isolate tested, the UCP isolate, had a 1 log higher ID50 than the Iowa isolate (Chappell et al., 1997).
Some researchers are also questioning the taxonomic designations of the various species of
Cryptosporidium, and further research is needed to clearly identify the genetic similarities of isolates
infective to humans (Tzipori and Griffiths, 1998).  Until more complete experimental data are available,
the dose-response relationship for the Iowa isolate of C. parvum (DuPont et al., 1995; Haas et al.,
1996) will be used as a proxy for all species and strains of Cryptosporidium.

The analysis uses a log-normal distribution for the dose-response relationship that runs from a low value
of 78 to a high value of 782 (mean of 238.6), a one order of magnitude spread.  This distribution should
adequately characterize the potential variability of the dose-response relationship across different strains
and different population sensitivities.

Not all infections will result in illness and observable symptoms. The probability of becoming ill given
infection is called the morbidity rate.  A change in dose has not been found to affect the morbidity rate
based on preliminary human ingestion trials. Therefore, the morbidity rate has been incorporated into
the risk assessment independent of dose. Haas et al. (1996) provided information suggesting a
morbidity rate value of 0.39, with 90 percent confidence bounds of 0.19 and 0.62. These data were
used to develop a triangular distribution of the morbidity rate for use in the Monte Carlo simulation as
described further below.
February 15, 2000
5-9
RIA for the Proposed LT1FBR

-------
Combining the underlying risk of infection with the morbidity rate (M), the annual probability of at least
one illness per year is:

                                   M x (1  •  [exp(« D/£)]n).

Where n represents the average annual number of exposures to Cryptosporidium.  The average annual
number of exposures depends on the type of system providing water (Exhibit 5-4).

The preliminary human ingestion trials were conducted on healthy individuals with no evidence of
previous C. Parvum infection (DuPont et al, 1995). Recently, however, it was found that after
repeated exposure to C. parvum (Iowa strain) the morbidity rate was the same as for the initial
exposure, but the symptoms were less severe and fewer oocysts were shed by re-infected subjects
(Okhuysen et al., 1998).  In addition, Chappell et al. (1997) observed that the diarrheal attack rate was
significantly higher for the TAMU or UCP isolates of C. parvum in comparison with the attack rate for
the Iowa isolate first studied by DuPont et al. (1995). Given these results and the variability of attack
rates of C. parvum during reported outbreaks (Exhibit 2-7), it may be expected that the actual
morbidity ratio may vary with the type of isolate to which a population is exposed as well as with the
immune status of the exposed population. In the absence of scientific evidence on the direction and
magnitude of such differences, the analysis will assume that a triangular morbidity rate distribution with a
mode of 0.39 and endpoints of 0.62 and 0.19 characterizes the range of uncertainty.

Exhibit 5-5 summarizes the parameters used to characterize the infection and illness hazards associated
with ingesting Cryptosporidium  oocysts.

                 Exhibit 5-5.  Summary of Hazard Identification Assumptions
 Annual dose/response relationship reflecting the probability of being exposed at lease once in a year:
 1 •  (•  „)"=!•  [exp(« D/£)]n

    k value:     mean = 238.6, 5th percentile = 132.0, 95th percentile = 465.4 (data fit to log normal
                distribution)

 Morbidity:     mode = 0.39, minimum = 0.19, maximum = 0.62 (assumed
                triangular distribution)

         n:      number of days per year of exposure to drinking water:  CWSs = 350;
                NTNCs; 250; TNCs = 10.

 Source: Haasetal., 1996.
Exposure Assessment
RIA for the Proposed LT1FBR                       5-10                               February 15, 2000

-------
In general, the exposure assessment focuses on characterizing an individual's daily dosage, which is
denoted D in the equations previously discussed. Estimating the daily exposure to Cryptosporidium
requires five basic pieces of information:

•      The concentration of Cryptosporidium in source water

•      The concentration of Cryptosporidium removed or inactivated during treatment

•      The concentration of Cryptosporidium remaining in finished water supplies

•      The percent viability of Cryptosporidium oocysts in finished water supplies (i.e., the number
       that are potentially infectious) the amount of drinking water consumed on a daily basis.

The benefit analysis estimates exposure under two sets of conditions to evaluate the potential human
health impacts of the proposed rule:

•      Baseline conditions, which characterize how many infectious Cryptosporidium oocysts an
       individual may ingest under current conditions

•      Rule conditions, which characterize how many infectious Cryptosporidium oocysts an
       individual may ingest under improved removal conditions proposed by the rule.

The baseline and rule conditions use the same assumptions about source water quality,
Cryptosporidium oocyst viability, and daily drinking water intake. The two differ with respect to
assumptions about the concentrations of Cryptosporidium oocysts removed during filtration and the
concentrations in finished water. The following describes each set of assumptions.

Source Water Quality

The source water quality distribution (i.e., the distribution of Cryptosporidium in source water) is
based on a survey of Cryptosporidium oocyst occurrence in source water (LeChevallier and Norton,
1995) that was analyzed by EPA in  1996. These data were also the basis for the source water quality
distribution used to estimate benefits for IESWTR and are shown in Exhibit 5-6.
February 15,2000                                 5-11                     RIA for the Proposed LT1FBR

-------
          Exhibit 5-6.  Baseline Expected National Source Water Cryptosporidium
                                Distributions (oocysts/lOOL)
Percentile
25
50
75
90
95
Mean
Standard Deviation
Source Water Concentration
103
231
516
1,064
1,641
470
841
The mean concentrations at the 69 sites from the eastern and central United States appears to be
represented by a lognormal distribution. Although limited by the small number of samples per site (i.e.,
1 to 16 samples with most sites sampled 5 times), variation within each site appears to be described by
the lognormal distribution. The distribution of Cryptosporidium oocysts used in this analysis is
lognormal with a mean concentration of 470 and a standard deviation of 841  oocysts per 100 liters.
Exhibit 5-6 reports concentrations for selected points in the cumulative density function of the
lognormal distribution.  EPA continues to evaluate the potential biases caused by limited geographical
data and analytical methods for Cryptosporidium recovery. EPA assumed that geographic and
analytic uncertainties introduced off-setting biases in the IESWTR benefit analysis, and this approach is
carried through in this analysis.

Cryptosporidium Oocyst Viability

The concentration of Cryptosporidium oocysts in finished water refers to a count of the total number
of oocysts in the water and does not take into account whether the oocysts are viable and potentially
infectious. The viability of oocysts after treatment is an area of scientific uncertainty.  One study
(LeChevallier et al., 1991a) found that one tenth to one third of oocysts in untreated water are viable
and potentially infectious. Oocyst viability is defined by the presence of one or more internal
morphological structures (nuclei, axonemes, or median bodies). Empty oocysts are assumed to be
non-viable (LeChevallier et al., 1997a).

This analysis uses the same viability assumptions employed for the IESWTR benefit analysis. In that
analysis, EPA chose to use a viability range about 50 percent lower than the range suggested by
LeChevallier et al. (1991a). The lower range was chosen to account for uncertainly regarding the lack
of specificity for species detection (many of which may not be infectious) and inability of research
methods to distinguish between a live and dead oocyst. The percentage of potentially viable and
infectious oocysts  in finished water was assumed to be a uniform distribution ranging from 5 percent to
15 percent with a mean value of 10 percent
RIA for the Proposed LT1FBR                       5-12                                February 15, 2000

-------
Daily Drinking Water Consumption

       In the Interim Enhanced Surface Water Treatment Rule, EPA assumed the daily water ingestion
of healthy adults to be lognormally distributed with a mean of 1.948 liters per person. This value was
used in developing the benefits of the IESWTR.  EPA's Office of Water has subsequently evaluated
drinking water consumption data from USDA's 1994-1996 Continuing Survey of Food Intakes by
Individuals (CSFII) study. EPA's analysis of the CSFn study resulted in a daily water ingestion
lognormally distributed with a mean of 1.2 liters per person. The risk and benefit analysis contained
within the LT1FBR RIA reflect this distribution.

       EPA has conducted additional risk and benefit analyses using water consumption distributions
with means of 0.9 liters and 1.9 liters per day for comparative purposes. These analyses are found in
the Appendices to the RIA.  The 0.9 liters per day distribution is another CSFH-based distribution that
reflects an alternative approach to characterizing water consumption from public water supplies.

Removal and Finished Water Concentrations: Baseline Conditions

Recognizing the uncertainty in knowing the current removal rates of Cryptosporidium being achieved
by water supplies subject to the proposed LT1FBR, EPA has adopted two alternative assumptions in
this analysis for characterizing the baseline:

•      Median 2.0 log removal
•      Median 2.5 log removal

These removal assumptions are similar, but not identical, to the  assumptions used in the IESWTR RIA
(2.5 and 3.0 logs). EPA based the  removal assumptions for IESWTR on historical studies of
Cryptosporidium and Giardia lamblia removal efficiencies by rapid granular filtration as discussed in
the IESWTR Notice of Data Availability (62 FR 59485, November 3, 1997), which noted an
observed range across different source water concentrations and treatment plant efficiencies of 2 to 6
log removal of Cryptosporidium oocysts.

In the IESWTR RIA, EPA stated that the SWTR and the Partnership for Safe Water have influenced
the removal range of typical plant performance upward from 2.0 or 2.5 log removal to 2.5 or 3.0 log
removal.  The Partnership for Safe  Water is a voluntary program that works closely with systems to
help optimize their performance; however, few systems serving under 10,000 individuals participate in
the Partnership for Safe Water.  In  addition, EPA's turbidity performance data shows higher finished
water NTU levels for plants serving fewer than 10,000 customers than for systems serving more than
10,000 customers. Thus, EPA assumes that systems serving under 10,000  are likely to achieve slightly
lower removal on average than systems serving 10,000 or more. To further characterize the variability
in Cryptosporidium removal currently being achieved by water systems subject to the proposed
LT1FBR,  EPA incorporated the assumed alternative log removal rates of 2.0 and 2.5 as distributions in
the Monte Carlo analysis.  Specifically, EPA has characterized the variability in the current log removal

February 15,2000                                 5-13                    RIA for the Proposed LT1FBR

-------
being achieved nationally as normal distributions with a mean of 2.0 or 2.5 log, and a standard deviation
of 0.63 log for both distributions.

Exhibit 5-7 presents expected baseline national finished water Cryptosporidium distributions derived
from the source water occurrence distributions and the baseline log removed distributions.
         Exhibit 5-7. Baseline Expected National Finished Water Cryptosporidium
                 Distributions, Based on Current Treatment (oocysts/lOOL)
Percentile
25
50
75
90
95
Mean
Standard Deviation
2.0 log
1.16
3.45
10.21
27.14
48.71
12.60
44.30
2.5 log
0.20
0.73
2.59
8.10
16.04
4.26
24.53
Removal and Finished Water Concentrations: Rule Conditions

EPA assumes that the turbidity provisions in the proposed rule will result in lower exposure to
Cryptosporidium, reflecting improvements in overall and individual filter performance.

Exhibit 5-8 gives the total number of small surface water systems currently using filtration, population
served, and the number of systems expected to need additional removal due to the new treatment
standard. The source for the number of systems and the number expected to need additional treatment
is described in Chapter 4. The remainder of this section discusses the treatment and removal
assumptions used in the exposure assessment.
RIAfor the Proposed LT1FBR
5-14
February 15, 2000

-------
 Exhibit 5-8.  Summary of Systems and Population Potentially Modifying Treatment under
                             the LT1FBR Turbidity Provisions
System Size
(population
served)
25-100
101-500
501-1,000
1,001-3,300
3,301-9,999
Total
Total Small Surface Water Systems
Number of
Systems
836
1,117
810
1,655
1,478
5,896
Total Population
Served
41,463
305,346
609,188
3,259,323
8,792,326
13,007,647
Systems Potentially Modifying Treatment
Number of
Systems3
341
456
331
675
603
2,406
Total Population
Served"
16,912
124,653
248,940
1,329,331
3,587,126
5,306,963
 a.  Estimates of the share of systems potentially affected by size category are based on the share of systems using filtration in
    Community Water System Survey, Volume II (62 FR 59485).
 b.  Population estimates by system size category are based on Water Industry Baseline Handbook. U.S. EPA, 1999c.
The assumed finished water Cryptosporidium distributions that would result from additional
log removal under the proposed rule were based on the removal distributions in the IESWTR analysis.
Those distributions were derived assuming that additional log removal was dependent on current
removal, i.e., that plants currently achieving the worst filtered water turbidity performance levels would
show the largest improvements or high improved removal assumption (for example, plants now failing to
meet a 0.4 NTU limit would show greater removal improvements than plants now meeting a 0.3 NTU
limit). The analysis also assumed dependence between the distribution of Cryptosporidium and
turbidity level.

It should be noted that the ICR will provide 18 months of Cryptosporidium monitoring data for the
development of a national source water Cryptosporidium occurrence distribution.  Although the data
collection efforts have been completed (January  2000), the last 6 months of data are still undergoing
quality assurance review. EPA's supplementary survey is also providing Cryptosporidium and other
microbial source water occurrence data; the full set of supplementary survey data will not be available
for analysis until July 2000. The Technical Working Group supporting the Federal Advisory
Committee involved with LT2ESWTR negotiation has been deliberating over the appropriate data
analysis methods to create the national source water distribution for Cryptosporidium occurrence.
This issue will continue to be discussed during the remainder of the LT2ESWTR Regulatory
Negotiation process, scheduled to end in July 2000. It is likely that the data will undergo peer review
only after the closure of the Regulatory Negotiation process. Due to the ICR data evaluation and peer
review time frame, EPA does not envision  being able to utilize these data in the LT1FBR regulatory
impact analyses and instead intends to incorporate the data into the impact analysis for the
LT2ESWTR.

Exhibit 5-9, based on a study by Patania et al. (1995), shows the relationship between C. parvum and
removal efficiencies by rapid granular filtration as discussed in the IESWTR Notice of Data Availability
(62 FR 59485, November 3, 1997).  This study  showed that a filter effluent turbidity of 0.1 NTU or
February 15, 2000
5-15
RIA for the Proposed LT1FBR

-------
less resulted in the most effective oocyst removal. The improved removal shown under the high
removal assumptions for IESWTR are based on this observed level of oocyst removal. An incremental
decrease in filter effluent turbidity  from 0.3 to 0.1 NTU increased oocyst removal by up to one log.
This oocyst removal range is the basis for the mid- and low- removal assumptions. Exhibit 5-10
contains the assumptions used in IESWTR to generate the new treatment distribution for a low-, mid-,
and high-log removal assumptions.
RIA for the Proposed LT1FBR                       5-16                               February 15, 2000

-------
  13
   g
   E
   Kt
       Exhibit 5-9. Cumulative Probability Distribution of Aggregate Pilot Plant Data
                                  for C. parvum Removal
l U
•
•
5,0-
4.0'.
3.0-
*
2-0~
I
i.o-J
I




j***

^C^"
^jj?
r ooiT
'"' R Turbidity <
• Turbidity 5
u (C, /wrfwrn
" TuAidity >
" Turbiditj' >
(C. f^arvum
1 I 1 1 t 1 I II
C. parv^m
o
pO** " n
= " "



0.1 KTL'
0. 1 NTU
removal > indicated value)
0.1 NTU
0.1 NTU
removal > indicated value)
i a l 	 1













-01 .1 I 5 10 20 30 50 70 80 90 95 99 99.9 99.99
                      Percent of Samples Less Than or Equal to a Given Value
Pataniaetal. 1995.
              Exhibit 5-10. Improved Cryptosporidium Removal Assumptions
             (Additional Cryptosporidium Log Removal with the Proposed Rule)

Plants now meeting 0.2 NTU Standard
Plants now meeting 0.3 NTU Standard
Plants now meeting 0.4 NTU Standard
Plants now failing to meet 0.4 NTU Standard
Log Removal Assumption Scenarios
Low
None
0.15
0.35
0.50
Mid
None
0.25
0.50
0.75
High
None
0.3
0.6
0.9
The effect on finished water quality using these removal assumptions, based on current log removal of
2.0 and 2.5, is displayed in Exhibit 5-11.
February 15, 2000
5-17
RIA for the Proposed LT1FBR

-------
    Exhibit 5-11. Expected National Source Water and Finished Water Cryptosporidium
                            Distributions with Improved Removal
Assuming Current Log Removal of 2.0
Percentile
25
50
75
90
95
Source Water
Concentrations
(oocysts/lOOL)
103
231
516
1064
1641
Mean
Standard Deviation
Finished Water Concentration (oocysts/lOOL)
Current
Treatment
1.16
3.45
10.21
27.14
48.71
12.60
44.30
Improved Removal
Low
1.14
1.77
3.94
8.585
15.40
4.52
14.96
Mid
0.97
1.40
2.51
4.83
8.66
2.80
8.37
High
0.85
1.24
1.90
3.42
6.13
2.13
5.90
Assuming Current Log Removal of 2.5
25
50
75
90
95
103
231
516
1064
1641
Mean
Standard Deviation
0.20
0.73
2.59
8.10
16.04
4.26
24.53
0.21
0.37
1.01
2.56
5.07
1.45
6.53
0.18
0.29
0.64
1.44
2.85
0.87
3.66
0.16
0.25
0.49
1.02
2.02
0.65
2.59
Using the assumption of a 2.0 current log removal and mid-case improvement in removal, the turbidity
provisions are estimated to reduce the mean concentration of oocysts from 12.60 oocysts per 100 liters
to 2.80 oocysts per 100 liters, a reduction of 78 percent (from Exhibit 5-11).  Using the assumption of
a 2.5 current log removal and mid-case improvement in removal, the turbidity provisions are estimated
to reduce the mean concentration of oocysts from 4.26 oocysts per 100 liters to 0.87 oocysts per 100
liters, a reduction of 80 percent.

The final element required for the exposure assessment is an estimate of the number of people
potentially exposed to Cryptosporidium by  consuming drinking water from small systems.  As
presented earlier in Exhibit 5-8, EPA estimated the population is served by small surface water systems
and ground water systems under the influence of surface water.  Exhibit 5-8 also provides estimates of
the number of systems and associated population that are expected to be affected by the proposed rule.

Exhibit 5-12 summarizes the assumptions used to characterize an individual's exposure to viable
Cryptosporidium oocysts.
RIAfor the Proposed LT1FBR
5-18
February 15, 2000

-------
                 Exhibit 5-12.  Summary of Exposure Assessment Assumptions
  Source Water Quality
  The source water concentration of Cryptosporidium oocysts is lognormally distributed (U.S. EPA 1998a):
         •   mean value of 470 oocysts/lOOL with a standard deviation of 841 oocysts/lOOL.

  Cryptosporidium oocyst viability
  The viability or infectivity of oocysts in finished water is uniformly distributed (LeChevallier and Norton,
  1992):
             low = 5%
         •   average = 10%
             high = 15%

  Daily Water Intake
  The drinking water consumption distribution used in this version of the benefits analysis (averaging
  approximately 1.2 liters per day) reflects Department of Agriculture CSFII data.  This distribution is
  currently being considered as being reflective of water consumption among the population that consumes
  drinking water from either community or noncommunity water supplies.

  Removal and Finished Water Concentrations:
  Baseline Conditions
  The median national Cryptosporidium removal efficiency for current conditions is estimated to be 2.0 log
  or 2.5 log, reflecting uncertainty in that value. Variability in these alternative removal rates is characterized
  using normal distributions with:
         •   Mean 2.0; standard deviation 0.63
         •   Mean 2.5; standard deviation 0.63
  The two resulting baseline distributions of finished water Cryptosporidium concentrations (oocysts/lOOL)
  are right skewed distributions with the characteristics of:
         •   Mean 12.60, standard deviation 44.30; median 3.45 (for 2.0 log removal)
         •   Mean 4.26, standard deviation 25.43; median 0.73 (for 2.5 log removal)

  Removal and Finished Water Concentrations:
  Rule Conditions
  Lognormal finished water Cryptosporidium oocyst concentrations (oocysts/1 OOL) for:
  2.0 log baseline
         •   low improved removal: mean 5.59 with standard deviation of 13.24
         •   mid improved removal: mean 4.07 with standard deviation of 9.88
         •   high improved removal: mean 3.45 with standard  deviation of 6.38
  2.5 log baseline
         •   low improved removal: mean 1.94 with standard deviation of 6.99
         •   mid improved removal: mean 1.33 with standard deviation of 4.01
         •   high improved removal: mean 1.12 with standard deviation of 3.09
Risk Characterization


The above assumptions are inputs to a model that estimates the annual number of Cryptosporidium
infections and illnesses.  The model uses the exposure assessment information provided above to

February 15,2000                                   5-19                      RIAfor the Proposed LT1FBR

-------
calculate the ingested dose parameter (D) in the dose-response equation. Ingested dose is the number
of potentially infectious oocysts an individual ingests daily, defined as:
                                        D = CF x V x Q

where CF is the concentration of oocysts in finished water (oocysts/liter)
       V is the percent of oocysts that are viable and potentially infectious
       Q is the quantity of water ingested daily (liters/day).

The ingested dose parameter (D) combined with the slope parameter (k) forms the hazard quotient
(D/k) portion of the dose-response relationship. The dose-response relationship describes an
individuals daily Cryptosporidium infection risk, or an annual risk if taken to the nth exponent, where n
is one of the annual exposures in Exhibit 5^1. The probability that an individual will experience at least
one Cryptosporidium illness per year is predicted by multiplying annual risk of infection by the
morbidity rate. Population risk is individual risk multiplied by the total exposed population.
                              I = P x M x (l • [exp (-
 Where:                      I  = total number of illnesses
                             P  = population exposed
                             M = morbidity rate
                             D = ingested dose (concentration of oocysts in finished water x daily
                                  ingestion of water)
                             k  = slope parameter (relation of ingestion to infection)
                             n = number of days of exposure

5.2.3  Baseline and Reduced Health Risk of the Turbidity Provisions

Based on the assumptions and methodology described previously, risk assessment allows the estimation
of existing risk under the current conditions (called baseline risk) and reduced risk when the turbidity
provisions of the LT1FBR have been implemented. The endemic risk (isolated cryptosporidiosis cases
that are not reported) is estimated by this risk assessment and is expected to be reduced due to
improved overall filter performance resulting in the greater removal of oocysts on a regular basis
(Exhibit  5-11). Outbreak-related risk is also expected to be reduced as the enhanced monitoring and
tighter control over individual filter operations allow operators to detect and prevent breaches in
treatment. Outbreak-related risk is more difficult to quantify using a standard risk assessment and is not
included  in the baseline or rule condition risk estimated in the following section.
RIA for the Proposed LT1FBR                        5-20                                February 15, 2000

-------
Risk of Infection and Illness

EPA has developed the following types of risk characterizations for the LT1FBR impact
assessment:

          •   Individual risk experienced by an average exposed person, and individual risk
              experienced by a highly exposed person
•      General population risk
•      Sensitive subpopulation risk.

The results of the risk characterization analyses performed are described in the following sections.

Individual Risk

The annual risk of infection and illness has been estimated for both an average individual having
central tendency or typical exposed conditions,  and for an individual who is highly exposed due to
the combination of both high raw water Cryptosporidium levels and a high daily drinking water
consumption rate.

Two LT1FBR treatment scenarios are modeled, the first in which systems achieve a median
baseline oocyst removal efficiency of 2.0 logs, and the second in which systems achieve a median
baseline removal efficiency of 2.5.

The average exposed individual is assumed to be served by a system with an average finished
water oocyst concentration, equivalent to the mean of the finished water concentrations of 12.6
oocysts per 100 liters and 4.26 oocysts per 100 liters calculated for the 2.0 and 2.5 log treatment
scenarios, respectively, as shown in Exhibit 5-7. It is also assumed that the average exposed
individual has a daily drinking water consumption of 1.24 liters per day.

It is assumed that a highly exposed individual would be served by a system with a high finished
water oocyst concentration, equivalent to the 90th percentile of the finished water distributions
shown in Exhibit 5-7 for the 2.0 and 2.5 log treatment scenarios (27.14 oocysts per 100 liters and
8.10 oocysts per 100 liters, respectively).  The highly exposed individual is also assumed to have a
daily drinking water consumption of 2.35 liters per day, the 90th percentile of the custom consumption
distribution described previously.

Point estimates for other parameters used in the model include: viability of treated oocysts = 0.1;
for the dose-response equation, a lvalue = 238.6, the mean of the distribution, where k is the
average number of oocysts required to initiate an infection; and a morbidity rate of 0.39, the mean
of the morbidity distribution (Haas et al.,1996).
February 15,2000                                5-21                     RIA for the Proposed LT1FBR

-------
Based on these assumptions, the following individual risk estimates were obtained:

Average exposed individual (assuming 2.0 log removal)

       Annual risk of infection:  2.27 x W2
       Annual risk of illness:     8.84 x lO'3

Average exposed individual (assuming 2.5 log removal)

       Annual risk of infection:  7.72 x 10"3
       Annual risk of illness:     3.01 x icr3

Highly exposed individual (assuming 2.0 log removal)

       Annual risk of infection:  8.91 x icr2
       Annual risk of illness:     3.48 x 10'2

Highly exposed individual (assuming 2.5 log removal)

       Annual ri sk of infecti on:  2.7 5 x 1Q'2
       Annual risk of illness:     1.07 x 10'2

Note that the individual annual risks shown above are for persons served by CWSs (approximately
95 percent of the affected population). The individual annual risks for consumers using NTNCs
(approximately 2 percent of the affected population) are very similar to those for the CWSs. The
individual annual risks for consumers using TNCs (approximately 3 percent of the affected
population) are substantially lower (approximately 2 orders of magnitude) than the estimated
individual risks for consumers using  CWSs.  For the most part, this lower risk value reflects the
assumption that consumers using TNCs have fewer days of exposure, as compared with users of
CWSs (10 days per individual versus 350 days at community supplies).

General Population Risk


EPA used  a Monte Carlo simulation to develop estimates of the  range of risks of infection and illness
experienced in the general population, and of the number of annual infections and illnesses resulting from
those risks. The algorithms used for calculating individual risk and the resulting number of infections and
illnesses in the overall population at risk, as well as the details on the forms of the distributions used in
the Monte Carlo simulation employing those algorithms, have been described previously in this chapter

Using the  Monte Carlo simulation analysis, estimates of the distributions of annual risk of illness
for the baseline and three improved removal  assumptions were obtained. These are summarized in
Exhibit  5-13.  Note that these population estimates include exposures from CWSs, NTNCs, and
RIA for the Proposed LT1FBR                       5-22                              February 15, 2000

-------
TNCs (summing to approximately 13 million). Exhibit 5-14 summarizes the calculated infections
and illnesses reduced (difference between the baseline and improved removal scenarios as modeled in
the Monte Carlo simulation) for each of the two current log removal assumptions under low-, mid-, and
high-case improved removal scenarios. The mean value presented in the tables represents the statistical
expected value of the distribution. The 10th and 90th percentiles implies that there is a 10 percent
chance that the estimated value could be  as low as the 10th percentile and that there is a 10 percent
chance that the estimated value could be  as high as the 90th percentile.

Based on this risk assessment at an ingestion rate of 1.2 liters per day, LT1FBR is estimated to result in
77,500 fewer illnesses at the 2.0 log removal baseline and the mid-removal assumption and 27,900
fewer illnesses at the 2.5 log removal baseline.
 Exhibit 5-13  Distribution of Annual Individual Risks of Illness Due to Cryptosporidium for
                       the Baseline and Improved Removal Scenarios
Annual Risk of
Illness range
>io-2
10-3tolQ-2
10-4to 10'3
lO^tolO'4
io-6toicr5
io-2
10-3tolQ-2
10-4to 10'3
icr'toicr4
10-6tolQ-5

-------
                    Exhibit 5-14.  Number of Illnesses and Illnesses Avoided
Improved Log-Removal Assumption
Daily Drinking Water Ingestion and
Baseline Cryptosporidium Log-Removal Assumptions
Mean = 1.2 Liters per person
2.0 log
2.5 log
Current Treatment/Baseline
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
103,400
1,700
231,000
36,000
280
66,000
Low Improved CryptosporidiumRemoval
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
40,600
1,500
81,900
13,200
250
22,600
Illnesses Avoided with Low Improved CryptosporidiumRemoval Assumption
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
62,800
0
152,000
22,800
0
43,900
Mid Improved CryptosporidiumRemoval
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
25,900
1,400
51,100
8,100
230
13,700
Illnesses Avoided with Mid Improved CryptosporidiumRemoval Assumption
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
77,500
0
184,000
27,900
0
52,900
High Improved CryptosporidiumRemoval
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
19,900
1,300
39,100
6,000
220
10,300
Illnesses Avoided with High Improved CryptosporidiumRemoval Assumption
Annual Illnesses — Mean
Annual Illnesses — 10th Percentile
Annual Illnesses — 90th Percentile
83,600
0
196,000
30,000
0
56,500
 Note: Illnesses avoided derived from Monte Carlo simulation may not precisely match values derived arithmetically.
RIAfor the Proposed LT1FBR
5-24
February 15, 2000

-------
Sensitive Subpopulation Risk

In addition to estimating the risks of illness for the general population, EPA has developed separate
estimates of the risk of illness for specific sensitive subpopulation groups.  The sensitive subgroups
considered include both age-based and health-based characteristics.

Data from Gerba et al. (1996) Sensitive Populations: Who Is at the Greatest Risk? and Bureau of
the Census data are used to estimate the fraction of the exposed population whose sensitivity to
Cryptosporidium is based on health status.

It is assumed that all infants, toddlers, children up to 5 years old, and elderly persons are more
sensitive to Cryptosporidium infection than the general population.  The toddler/young child
subgroup is called out separately to allow flexibility for possible future calculations of person-to-
person secondary spread common among children in this age group.

The seriously ill subgroups addressed in these calculations include:

       Non-hospitalized cancer patients
       Organ transplant patients
       AIDS patients
       Nursing home and related care facility residents.

The first three  of the seriously ill subgroups comprise approximately 1 percent of the general
population, calculated from population numbers presented by Gerba et al.  (1996), and it is assumed
that these persons are divided equally among all age groups. For the fourth category, the nursing
home or related care facility residents, it is assumed that these persons are all older than 65 years;
they constitute another 0.6 percent of the general population (calculated from population numbers
presented by Gerba et al., 1996). The age-adjusted factor for nursing home patients is:  (0.006 -f-
0.126) = 4.8 percent of the population over 65 years old.

Assumptions regarding the number of pregnant women in the population are based on data cited in
Gerba et al. (1996). There were 5,657,900 reported pregnancies in 1989 (Gerba et al.,  1996). The
1990 Census (www.census.gov/statab/fireq/98s0014.txt) reports that the total US population was
248,765,000 persons in that time frame.  Assuming that each pregnant woman is pregnant only
once per year,  it can be estimated that pregnant women represent about 2.27 percent (5,657,900 +
248,765,000) of the general population.

This factor is age-adjusted to the age group of most childbearing women, ages >16 to 50 years old.
In 1990, there  were 63,316,800 women ages >16 to 50 years.  This age group makes up about
51.3 percent of the general population. The age adjusted factor for pregnant women is: (0.023 +
0.513) = 4.5 percent of the general population from >16 to 50 years old. This factor is equivalent
to about 8.9 percent of women ages >16 to 50 years old.
February 15,2000                                5-25                    RIA for the Proposed LT1FBR

-------
In all, approximately 23 percent of the population is assumed to be in the increased-sensitivity
subpopulation.  This breaks down as:

                           Infants:  2.8%
                         Toddlers:  4.4%
                          Elderly: 12.6%
                      Seriously ill:  1.0%
                 Pregnant women:  2.3%

Although it is reasonable to expect that the subpopulations identified here are likely to be at an
increased risk of both infection and illness from exposure to Cryptosporidium relative the general
population, there is no specific data available addressing those increased risks quantitatively.  To
develop estimates of the number of infections and illnesses in these subgroups, the following
assumptions were made.

Infectivity: A higher rate of infectivity is assumed, and is  modeled by setting k in the dose-
response function to 132, the lower bound of the 95 percent confidence limits on this factor (Haas
etal., 1996).

Morbidity: A higher rate of illness given an infection is assumed by setting the morbidity rate to
0.62, a point estimate equal to the upper bound of the triangular distribution of morbidity rate for
the general population (Haas et al., 1996).

All other assumptions used in calculating the incidence of cryptosporidiosis infections and illnesses
in sensitive subgroups are the same as for the general population under average exposure
conditions.

The estimated annual baseline illnesses for the sensitive subpopulations in the aggregate are
approximately 60,000 assuming systems are currently achieving 2.0 log removal, and 21,100
assuming systems are currently achieving 2.5 log removal.

The reader is cautioned that these numbers cannot be directly added to the general population
estimates of annual illness presented earlier. Doing so would result in some double counting since
these sensitive subgroups were included in the general population estimates without accounting
directly for their increased risks there.  An estimate of the number of annual illnesses resulting if
these higher risk factors were included for these subpopulations can be made by subtracting out
approximately 23 percent of the illnesses estimated in the baseline (currently attributable to these
groups) and adding back the specific estimates noted above.  Doing so would result in an increase
in the baseline number of annual illnesses of approximately 35- 40 percent. However, the reader is
reminded that the key quantitative assumptions used in performing this analysis on  sensitive
populations (specifically, increasing the infectivity and morbidity rates) are based on best
professional judgement in view of the limited relevant data available to describe the actual risks to
these groups.
RIA for the Proposed LT1FBR                       5-26                               February 15, 2000

-------
Risk of Mortality

Cryptosporidiosis poses a serious risk of death in sensitive subpopulations, such as those with
compromised immune systems.  Based on data from the Milwaukee outbreak, the mortality rate can
be estimated at approximately 0.0125 percent (0.0125 percent of all illnesses would result in a
mortality—50 mortalities/400,000 cases) in a mixed population of exposed persons. This figure was
derived based on death certificate reporting (50 additional deaths associated with cryptosporidiosis as
reported on the death certificate, of which 46 had AIDS as the underlying cause of death) and should
be regarded as a minimum estimate (Hoxie et al., 1996).

The mortality rate from the Milwaukee outbreak may not be reflective of overall mortality rates from
low-level endemic exposure.  The estimated levels of Cryptosporidium in the finished water supplies
during the Milwaukee outbreak were much higher than the levels expected in systems complying with
the existing SWTR.  Thus, the higher level of Cryptosporidium in the water supply could have resulted
in a higher mortality rate if more significant symptomatic response were associated with infection
influenced by higher ingested dosages.  No data are yet available, however, to support this hypothesis;
data are available to indicate only a higher probability of infection resulting from higher ingested dose
levels.  There is some evidence that the mortality rate among susceptible subpopulations may not be
linked to community-wide exposure levels (Rose, 1997).  The majority of mortalities identified from the
Milwaukee outbreak (46 of 50) were among individuals with AIDS (Hoxie et al., 1997).  In another
outbreak in Las Vegas, similar mortality rates were observed in ADDS patients (52.6 percent among
ADDS patients in Las Vegas compared with 68 percent among ADDS patients in Milwaukee), although
it was hypothesized that the drinking water had been contaminated over an extended period of time
with intermittent low levels of oocysts, unlike Milwaukee's massive contamination (Rose, 1997).

The Milwaukee mortality rate might also not be representative of the national mortality rate if there are
larger or smaller sensitive subpopulations in Milwaukee than nationally. According  to Hoxie et al.
(1996), "in 1992, just prior to the outbreak, the annual reported AIDS case rate in  the Milwaukee
metropolitan area ranked 78th among 98 metropolitan areas in the United States with populations
500,000 or more." Thus, the greater presence of sensitive subpopulations in some areas might indicate
a greater susceptibility to cryptosporidiosis.  At this time, there is no basis for adjusting the Milwaukee
outbreak mortality rate to the general population.

Exhibit 5-15 provides a distribution of the estimated individual annual risks of mortality derived
from the Monte Carlo simulation.  Assuming the Milwaukee mortality rate of 0.0125  percent, Exhibit
5-16 displays the estimated range of mortalities and mortalities prevented as modeled in the Monte
Carlo simulation.
February 15,2000                                  5-27                     RIAfor the Proposed LT1FBR

-------
 Exhibit 5-15. Distribution of Annual Individual Risks of Mortality Due to Cryptosporidium
                    for the Baseline and Improved Removal Scenarios
Annual Risk of
Mortality Range
>10'2
10-3tolQ-2
icr4to icr3
icr'toicr4
io-6toio-5
10'2
10-3tolQ-2
10-4to 10'3
lO-'tolO'4
lO^tolO'5
<10'6
2.5 Log Removal
Baseline
0
0
0
52,000
825,000
12,131,000
Low Improved
Removal
0
0
0
10,000
249,000
12,748,000
Mid Improved
Removal
0
0
0
5,000
125,00
12,877,000
High Improved
Removal
0
0
0
3,000
79,000
12,925,000
RIAfor the Proposed LT1FBR
5-28
February 15, 2000

-------
        Exhibit 5-16.  Number of Mortalities and Mortalities Avoided among Exposed
                                            Population
Improved Log-Removal Assumption
Current Treatment/Baseline
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
Daily Drinking Water Ingestion and
Baseline Cryptosporidium Log-Removal Assumptions
Mean = 1.2 Liters per person
2.0 log 2.5 log
13
0
29
5
0
8
Low Improved CryptosporidiumRemoval
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
5
0
10
2
0
3
Mortalities Avoided with Low Improved CryptosporidiumRemoval Assumption
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
9
0
19
3
0
5
Mid Improved CryptosporidiumRemoval
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
3
0
6
1
0
2
Mortalities Avoided with Mid Improved CryptosporidiumRemoval Assumption
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
10
0
23
3
0
7
High Improved CryptosporidiumRemoval
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
2
0
5
1
0
1
Mortalities Avoided with High Improved CryptosporidiumRemoval Assumption
Annual Mortalities — Mean
Annual Mortalities — 10th Percentile
Annual Mortalities — 90th Percentile
10
0
25
4
0
7
Note: Mortalities avoided derived from Monte Carlo simulation may not precisely match values derived
arithmetically.
February 15, 2000
5-29
RIA for the Proposed LT1FBR

-------
5.2.4  Monetization of Reduced Risks

The health benefits of the LT1FBR can be evaluated in terms of two valuation measures; 1) COI
avoided and 2) WTP to reduce the probability of suffering an adverse health effect (Freeman, 1979).
COI avoided due to adverse health effects includes medical costs, lost income, reduced productivity,
and averting expenditures. These goods have observable market values and are, therefore, easier to
quantify than WTP values.

The WTP concept goes beyond the expected value of avoided COI to include the total value of health
benefits.  In principle, WTP is a comprehensive measure of the welfare effect of a change in risk and is
generally expected to exceed the out-of-pocket financial effect of the change (Chestnut and Alberini,
1997). WTP includes the intuitive notion that illness is disagreeable and that one would be willing to
pay to avoid the pain and suffering associated with an adverse health effect beyond the cost of the
illness. Since there are no markets for avoided pain and suffering, there are no observable market
transactions by which their value can be measured.

Another reason that WTP for reduced health risk is likely to exceed the expected value of avoided
COI is the general acceptance of additional costs to avoid risk. WTP values for avoidance of
premature death include the value of reductions in the risk of out-of-pocket costs (i.e., COI) plus the
value of reduced risk of the lost enjoyment of life (Chestnut and Alberini, 1997).  The use of expected
COI, instead of WTP, tends to understate the economic value of risk reduction because COI does not
incorporate nonpecuniary benefits such as avoided pain and suffering.

Expenditures on averting behavior also comprise a part of WTP. In the context of reducing endemic
Cryptosporidium risk, averting behaviors involve the day-to-day, routine activities that consumers
undertake with respect to drinking water, including consumption of bottled water or use of individual
filtration devices. The reasons for undertaking these behaviors are numerous (i.e., taste, odor, reduced
exposure to chemical contaminants) with the motivation of reducing specifically the risk from
Cryptosporidium a minor factor. Expenditures on averting behaviors during outbreaks are discussed
in Section 5.3.

Monetization of Illness

Information is not available on direct measurements of either COI or WTP to reduce risk specifically
for Cryptosporidium. For the purposes of this analysis, an adjusted giardiasis COI is used as a proxy
for the COI of cryptosporidiosis.  The costs incurred during an outbreak of waterborne giardiasis in
1983 in Pennsylvania were based on a survey of 370 people who had "confirmed" cases of giardiasis,
i.e., a positive stool sample. The study estimated direct medical costs paid for either by the victim or
insurance company, including the costs of doctor visits, emergency room visits, hospital visits,
laboratory fees, and medication. The study also estimated other costs, including time costs for medical
care, value of work loss days, loss of productivity, and loss of leisure time (Harrington et al,  1989).
However, this  COI study did not include averting expenditures or value the "pain, suffering, stress, and
RIA for the Proposed LT1FBR                       5-30                               February 15, 2000

-------
anxiety, or any other psychological or resulting physiological consequences of the outbreak."
(Harrington et al., 1985).
Exhibit 5-17 contains a summary of the average losses for confirmed cases of giardiasis in 1984 dollars
and updated using the Consumer Price Index to a January  1999 price level.


                    Exhibit 5-17.  Losses per Case of Giardiasis by Category
Loss Category
Direct Medical Costs:
Doctor visits
Hospital visits
Emergency room visits
Laboratory tests
Medication
Subtotal
Indirect Medical Costs:
Time costs for medical care
Value of work loss days
Loss of work productivity
Loss of leisure time
Subtotal
Mean total expected losses per case —
giardiases
Mean total expected losses per case -
cryptosporidiosis
Average Losses (1984 S)
(Harrington et al., 1985)

$36
100
27
63
28
$254

$18
359C
371C
876C
$1, 624
$1,878

CPI Update
Factor

2.31a
2.31
2.31
2.31
2.31


1.58b
1.58
1.58
1.58



Average Losses
(1999 S)
$83
231
62
146
65
$587
$28
567
586
1,384
$2,565
$3,152
$2,403
 a.   Consumer Price Index, Medical Care: 246.6 (January 1999)7106.9 (1984 average)
 b.   Consumer Price Index, All Items: 164.3 (January 1999)7103.9 (1984 average)
 c.   Based on the assumption that the wage rate for the unemployed, homemakers, and retirees equals the wage rate for employed
     persons in the sample. Use of an alternative assumption or labor rate will result in different indirect costs.
The average losses per case of giardiasis reported in the survey are approximately $3,150 at the
current price level (1999 $). The average losses per case of cryptosporidiosis could be less than those
of giardiasis because cryptosporidiosis is self-limiting in immunocompetent subjects, with infections
lasting a shorter duration (7 to  14 days) than giardiasis infections (30 days median length-of-illness in
sample).  To take into account the shorter duration of cryptosporidiosis, the estimates for non-direct
medical costs of giardiasis are adjusted by the ratio of the duration of cryptosporidiosis over the
duration of giardiasis. The ratio and adjusted costs are estimated using a Monte Carlo simulation to
model the distribution of potential duration for each illness. Data from the Milwaukee outbreak indicate
that the duration of cryptosporidiosis is lognormally distributed, with a range of 1  to 55 days, a mean of
12 days, and a median of 9 days (Mackenzie et al., 1994). Data from the Pennsylvania outbreak
indicate that the duration of giardiasis is lognormally distributed, with a mean of 41.6 days  and a
standard deviation of 45 days  (Harrington et al., 1985).  The resulting adjusted COI distribution
February 15, 2000
5-31
RIA for the Proposed LT1FBR

-------
derived for cryptosporidiosis has a mean of approximately $2,400 and a median of approximately
$1,400. This mean value is the value presented in Exhibit 5-17.

It is important to note that the values in the above distribution reflect the potential COI avoided, not the
full WTP to reduce the probability of suffering a cryptosporidiosis infection.  The estimates do not take
into account the value of avoiding pain and suffering, the economic premium associated with risk
aversion, or the costs of averting behaviors.  Therefore the full value of the economic benefit to reduce
cryptosporidiosis may be higher than the $2,400 COI avoided per case mean estimate.  Exhibit 5-18
contains the values of annual illnesses avoided by the LT1FBR turbidity provisions, using the
distribution of adjusted COI estimates.

To compare these results against previous studies, Mauskopf and French (1991) estimated WTP to
avoid food borne illnesses based on the nature and length of the illness, integrated with the value of a
statistical life and indices of self-reported health status to value the losses in quality and length of life.
The WTP estimates (1999 $) for illnesses similar to cryptosporidiosis range from $166 to $7,424 for
mild to moderate cases of botulism (5 to 21 days of weakness, vomiting,  and nausea) and $284 to
$1,139 for salmonellosis (3 to 7 days of similar symptoms). Using these estimates, the value for
cryptosporidiosis (7 to 14 day duration) could range from $233  ($33.25/day for 7 days) to $4,942
($353/day for 14 days).  The cost of illness estimates (with a mean of $2,403) fall within this range and
are a reasonable approximation of the value to avoid health damages associated with  cryptosporidiosis,
recognizing that some costs (such as averting expenditures, and pain and suffering) have not been
monetized.
Exhibit 5-18 that follows displays the potential benefits from preventing illnesses using the COI
estimates as described above.
RIA for the Proposed LT1FBR                        5-32                                February 15, 2000

-------
             Exhibit 5-18.  Number and Cost of Illnesses Avoided Annually from
                               Turbidity Provisions* (SMillions)
Improved Log-Removal Assumption
Illnesses Avoided with Low Improved Cryptospor
Mean
lOthPercentile
90th Percentile
Daily Drinking Water Ingestion and
Baseline Cryptosporidium Log-Removal Assumptions
Mean = 1.2 Liters per person
2.0 log 2.5 log
uUiiin Removal Assumption
62,800
0
152,000
22,800
0
43,900
COI Avoided with Low Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$150.3
$0.0
$288.2
$53.9
$0.0
$81.4
Illnesses Avoided with Mid Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
77,500
0
184,000
27,900
0
52,900
COI Avoided with Mid Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$185.3
$0.0
$350.9
$66.2
$0.0
$98.8
Illnesses Avoided with High Improved Cryptosporidium Removal Assumption
Mean
lOthPercentile
90th Percentile
83,600
0
196,000
30,000
0
56,500
COI Avoided with High Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$199.5
$0.0
$376.7
$71.1
$0.0
$105.8
 ! All values presented are in January 1999 dollars.
February 15, 2000
5-33
RIA for the Proposed LT1FBR

-------
Monetization of Mortality

Studies that assess the value per statistical life (VSL) saved (i.e., reduced risk of premature death)
generally have central point estimates between $5 million and $8 million dollars with a range from $2
million to $14 million (Chestnut and Alberini, 1997). A recent EPA study characterized the VSL saved
as a lognormal distribution with a mean of $4.8 million with a standard deviation of $3.24 million,
capped at $13.5 million (in 1990 price level), based on 26 individual study estimates (62 FR 59485,
November 3, 1997).  Updating the VSL for current price levels results in a distribution with a mean of
$5.7 million and a standard deviation of $3.16 million, truncated at $16.87 million.

Because cryptosporidiosis mortalities are expected to occur primarily in sensitive subpopulations, there
may be some arguments for adjusting the VSL.  The typical valuation methodology used to derive the
VSL generally measure the individuals' WTP to reduce the risk of a premature death by a small
amount.  The small reduction in risk is then spread across a broad population. The mortality risk
associated with cryptosporidiosis is different in that a smaller sensitive subpopulation faces a higher
baseline risk. The valuation literature is unclear on whether this type of a risk would have a higher or
lower WTP although one study found that respondents favored programs that affect smaller populations
facing higher baseline risks, assuming the same number of lives are saved (Van Houtven, 1997). A
review of existing empirical literature with respect to adjusting  the VSL saved by drinking water
programs does not, however, provide a strong basis for specific adjustments (up or down) to the VSL
(Van Houtven et al, 1997).

For the purposes of this RIA, Exhibit 5-19 displays the potential benefits from preventing mortalities
using the updated VSL distribution, recognizing the uncertainties inherent in this or any available
valuation methodology.
RIA for the Proposed LT1FBR                        5-34                                February 15, 2000

-------
            Exhibit 5-19. Number and Cost of Mortalities Avoided Annually from
                               Turbidity Provisions* (SMillions)
Improved Log-Removal Assumption
Mortalities Avoided with Low Improved Crypto
Mean
lOthPercentile
90th Percentile
Daily Drinking Water Ingestion and
Baseline Cryptosporidium Log-Removal Assumptions
Mean = 1.2 Liters per person
2.0 log
sporidium Removal Assumption
9
0
19
2.5 log
3
0
5
Cost of Mortalities Avoided with Low Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$45.0
$0.0
$101.7
$16.2
$0.0
$28.8
Mortalities Avoided with Mid Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
10
0
23
3
0
7
Cost of Mortalities Avoided with Mid Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$55.5
$0.0
$123.3
$19.9
$0.0
$34.8
Mortalities Avoided with High Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
10
0
25
4
0
7
Cost of Mortalities Avoided with High Improved CryptosporidiumRemoval Assumption
Mean
lOthPercentile
90th Percentile
$59.8
$0.0
$132.0
$21.3
$0.0
$37.3
 * All values presented are in January 1999 dollars.


5.2.5  Health Effects to Sensitive Subpopulations

The health effects of Cryptosporidium on sensitive Subpopulations is much more severe and debilitating
than the health effects on the general public.  The estimated COI avoided calculated earlier likely does
not capture the full value of costs to sensitive Subpopulations, health trials were only conducted with
healthy individuals and symptomatic responses are more severe in sensitive populations. For example,
the duration of cryptosporidiosis in those with compromised immune systems is considerably longer
than in those with competent immune systems, with more severe symptoms often  requiring lengthy
hospital stays. COI from cryptosporidiosis is expected to be much larger than $2,400 per case for
February 15, 2000
5-35
RIA for the Proposed LT1FBR

-------
sensitive subpopulations. During the Milwaukee outbreak, 33 AIDS patients with Cryptosporidium
accounted for 400 hospital days at an additional cost of nearly $760,000 (Rose, 1997). COI due to
these hospital days alone was estimated at $23,000 per case ($760,000/33 patients). Although the
COI for sensitive populations is expected to be greater than the general population, no attempt was
made to quantify these effects for the purposes of this regulatory impact analysis. Also, the cost of
averting expenditures could be higher in sensitive subpopulations.  Sensitive subpopulations are more
susceptible to Cryptosporidium infections, thus these individuals may purchase bottled water, boil
water, or take other health precautions on a daily basis.

5.3    Other Benefits of Turbidity Provisions

Section 5.3 describes qualitative benefits of the turbidity provisions from the reduction in outbreak risk,
enhanced aesthetic water quality, and avoided costs of averting behavior.

5.3.1  Reduction in Outbreak Risk

Besides reducing the endemic risk of cryptosporidiosis, the LT1FBR will reduce the likelihood of major
outbreaks, such as the Milwaukee outbreak, from occurring.  The economic value of reducing the risk
of outbreaks could be quite high when the magnitude of potential costs is considered. For example, if
the $2,400 per cryptosporidiosis infection estimate is applied to the estimated 2,000 cryptosporidiosis
cases attributed to a sewage contaminated well in Braun Station, Texas (Craun et al., 1998), health
damages could reach $4.8 million. 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
provision of alternative water supplies and issuing customer water use warnings and health alerts.
Commercial establishments (e.g., restaurants) and their customers may incur costs due to interrupted
and lost service (e.g., lost producer and consumer surplus). Local businesses, institutions, and
households may incur costs associated with undertaking averting and defensive actions.  To the extent
that LT1FBR reduces the likelihood of waterborne disease outbreaks, avoided response costs are
potentially numerous and significant.

5.3.2  Enhanced Aesthetic Water Quality

Economic theory suggests that improving the aesthetic quality of drinking water produces benefits
separate from improvements in health. Consumers,  presumably, would be willing to pay to protect the
aesthetic quality of drinking water from high turbidity levels. Aesthetic improvements from the
proposed rule may not be noticeable to the general public and, therefore, these benefits are not
quantified for this analysis.
RIA for the Proposed LT1FBR                       5-36                                February 15, 2000

-------
5.3.3  Avoided Costs of Averting Behavior

During outbreaks or periods of high turbidity, consumers and businesses may use alternative water
sources or practice behaviors to reduce risk, such as boiling water. If the rule reduces the need for
these averting behaviors, an economic benefit will accrue. During an outbreak of giardiasis,
expenditures on averting behaviors, such as hauling in safe water, boiling water, and purchasing bottled
water, were estimated at between $1.74 to $5.53 per person per day during the outbreak (Harrington
et al., 1989).  If these figures are applied to a small drinking water system serving 10,000 customers,
total expenditures on averting behavior during a Cryptosporidium outbreak could range between
$17,400 and $55,300 per day. Determining the precise reduction in outbreak risk and resulting
benefits due to reduced or avoided averting behavior is not possible given current information, but
potential benefits are expected to be substantial.

Five additional studies were identified that used the averting cost approach to estimate household and
other costs attributable to short-term contamination of drinking water supplies (Abdalla, 1990; Abdalla
et al., 1992; Harrington et al., 1985; Sun et al.,  1992; Van Houtven, et al., 1997). The most relevant
of these for the LT1FBR analysis is a study by Harrington et al., (1985), that  analyzes the costs
associated with drinking water contamination by Giardia in Luzerne County,  Pennsylvania.  The
December  1983 outbreak resulted in 366 confirmed giardiasis cases resulting from sewage leaking into
the unfiltered source water. The total affected population was 75,000 individuals across Pittston
Burough and 17 other municipalities. The Harrington study also developed a theoretical and empirical
example of how outbreak costs  are incurred, based on the Luzerne County example.

The four stages associated with  a waterborne outbreak that may impose costs on society  are discovery,
survey and testing, reaction and aftermath. (Harrington et al.,  1985).  These are described below.

           Discovery.  Health care providers or State, local, or hospital laboratory technicians send
           reports to State authorities notifying them of the need for further investigation when the rate
           of new cases suddenly increases above the normal rate.

       •   Survey and testing. A host of epidemiological  surveys may be conducted,  along with
           tests of the water supply, once a few cases are confirmed.

       •   Reaction. Local authorities and the water system may issue boil-water advisories, or
           other warnings to reduce exposure  once  a link is made between the drinking  water supply
           and the disease outbreak.  Businesses as well as households may be affected by such
           action, requiring government agencies to begin surveillance and enforcement activities and
           in some cases, provide alternative water sources.

           Aftermath.  This final stage involves discussions of any long-term solutions to the
           problem, and how the costs of the outbreak and prevention of future ones may be shared.
           These discussions can only take place once the outbreak is contained by actions taken
           during the previous phase.
February 15,2000                                  5-37                      RIA for the ProposedLT1FBR

-------
The Luzerne County outbreak resulted in losses, due to actions taken by individuals to avoid the
contaminated water, estimated to be between $20.8 million and $61.8 million.  The predominant cost
was time lost to boiling water.  Losses due to averting actions for restaurants, bars, schools and other
businesses during the outbreak averaged $1.0 million.  The burden for government agencies was
$230,000  and the outbreak cost the water supply utility $1.8 million.  These costs do not include legal
fees, outbreak effects on businesses that were not investigated, leisure activities, or net losses due to
substituting more expensive beverages for tap water.

5.4   Benefits from Other Rule Provisions

Section 5.4 describes qualitative benefits associated with the proposed rule including benefits from the
recycle provisions, disinfection benchmarking, covered finished water reservoirs, including
Cryptosporidium in the GWUDI definition, and provisions modifying watershed regulatory
requirements for unfiltered systems.

5.4.1  Benefits  of Recycle Provision

EPA has identified four primary public health concerns arising from the recycle of spent filter backwash
and other  recycle  streams within the treatment process of public water systems.

1.      Data establishes recycle flows can contain Cryptosporidium oocysts, often at higher
       concentrations than plant source waters, and recycling these flows may increase the number of
       oocysts entering the plant, reaching the filters, and entering the finished water. Since
       Cryptosporidium is not inactivated by standard disinfection practice, it is critical that all
       available physical removal processes (coagulation, flocculation, clarification, filtration) be
       protected  from the hydraulic and chemical treatment disruptions recycle events may cause.
       Note that recycle returns oocysts to the plant at precisely the time treatment efficiency may be
       challenged by hydraulic and chemical disruption induced by recycle events. This may cause
       more oocysts to enter the finished water.

2.      Returning spent filter backwash, thickener supernatant, and liquids from dewatering into, or
       downstream of, the point of primary  coagulant addition may disrupt treatment chemistry by
       introducing residual coagulant or other treatment chemicals to the process stream. Recycle
       flow returned to the sedimentation basin may not reside in the basin long enough for recycled
       oocysts to settle, or it may create hydraulic currents within the basin that lower the unit's overall
       oocyst removal efficiency. Additionally, recycle can cause large variations in influent flow,
       which may result in harming of treatment efficiency by chemical under or over dosing (Patania et
       al., 1995; Edzwald and Kelley, 1998; Bellamy et al., 1993; Conley, 1965; Dugan et al., 1999;
       Robeck et al., 1964).

3.      The direct recycle of spent filter backwash without first providing treatment, equalization, or
       some form of hydraulic detention for the flow, may cause plants to exceed State-approved
       operating capacity during recycle events. Exceeding operating capacity can cause
RIA for the Proposed LT1FBR                       5-38                                February 15, 2000

-------
        sedimentation/clarification and filter loading rates to be exceeded, which may lower overall
        oocyst removal provided by the plant and increase finished water oocyst concentrations.

4.      Direct filtration plants do not employ a sedimentation basin in their primary treatment process to
        remove solids and oocysts; all oocyst removal is achieved by the filters.  If treatment for the
        recycle flow is not provided prior to its return to the plant, all of the oocysts captured by a filter
        during a filter run will be returned to the plant and again loaded to the filters. This may lead to
        ever increasing levels of oocysts being applied to the filters and could increase the concentration
        of oocysts in finished water.

The LT1FBR recycle provisions are based on the assumption that improving the aforementioned
recycle practices will prevent the accumulation of Cryptosporidium within the treatment plant and
minimize the risk of oocysts entering into the finished water. EPA expects these provisions to reduce
the incidence of cryptosporidiosis in two ways. First, endemic risk is likely to be reduced because
improved recycle processes will consistently reduce Cryptosporidium occurrence in finished water
relative to the recycle baseline.  Second, endemic and outbreak risk is likely to be reduced by returning
certain recycle flows (spent filter backwash, thickener supernatant, and liquids from dewatering) to the
plant prior to the point of primary coagulant addition because all available physical removal processes
will be employed to remove oocysts before they reach the filters. Returning these recycle flows prior to
the point of primary coagulant addition will also protect the integrity of chemical  dosing, which
determines the treatment efficacy of sedimentation/clarification and filtration, by minimizing the potential
for large fluctuations in plant influent flow volume and chemistry that can render  chemical doses less
than optimal.

EPA has not developed a national benefit estimate because the overall impact on finished water quality
of different treatment changes brought about by the provisions depends on a wide variety of system
operational parameters that cannot be easily modeled. In order to model the affect of recycle practice,
data regarding the ability of a wide range of unit processes (sedimentation, DAF, contact clarification,
filtration) to remove oocysts from a wide variety of source water types, under a range of treatment
conditions, is needed to calibrate the model. This data is currently not extensive enough to model the
impact of recycle on a wide variety of treatment configurations. Due to limited calibration data, EPA
did not quantify  benefits for these provisions.  However, data show that oocysts occur in recycle
streams and in the finished water of normally operated plants (unchallenged, well performing plants).
Recycle adds additional oocysts to the plant and risks lowering plant treatment efficiency during recycle
events by means of hydraulic and chemical disruption. The following discussion provides a qualitative
description of how the filter backwash provisions are expected to reduce health risks.

Returning spent  filter backwash, thickener supernatant, and liquids from dewatering prior to the point of
primary coagulant addition, will pass recycled oocysts through available physical removal processes
and protect the integrity of treatment chemistry, thereby improving log removal of Cryptosporidium
oocysts during recycle. Returning these flows prior to the point of primary chemical addition, which is
included in each of the recycling alternatives of the proposed rule, generates positive health benefits by
controlling pathogens and improving treatment chemistry during recycle events.  For example,
eliminating return of recycle flow to the flocculation or clarification basin mitigates the possibility the

February 15,2000                                  5-39                      RIA for the Proposed LT1FBR

-------
recycle flow will generate disruptive currents that can harm the oocyst removal efficiency of these
processes.  Furthermore, returning these flow prior to the point of primary coagulant addition, rather
than at this location or downstream of it, improves the accuracy of chemical treatment and protects the
integrity of the process, as the dose can be targeted for the mixture of recycle and source water rather
than just source water.

Plants that recycle directly (i.e., return recycle to treatment process without providing equalization,
treatment, or some other form of hydraulic detention) may exceed State approved operating capacity
during recycle events. Even if a system reduces or eliminates its raw water influent flow for the duration
of the recycle event, the filter loading rate of some plants may still exceed State-approved operating
capacity during recycling.  Also, overly abrupt changes in filtration rate may occur - such changes have
been shown to cause particles lodged in filter media to pass into the  filter effluent (Cleasby et al,  1963;
Glasgow and Wheatley, 1998; McTigue et al., 1998). The potential benefits of the proposed rule
would differ among the proposed alternatives to the extent that modification to recycle practice differ
across the alternatives.  No benefits are realized under alternative Rl for hydraulic surge reduction,
because the option does not contain a provision to address hydraulic surge. Under R2, the States
determine whether systems are required to modify recycle practice to address public health risk;
therefore, the number of affected systems is uncertain.  The option allows States to determine whether
recycle practice needs to be modified to allow the consideration of site-specific factors. Under
alternatives R3 and R4  all systems either install a flow equalization basin or a sedimentation basin,
respectively.  Installing equalization basins will hold the recycle in tanks and gradually release it back
into the treatment stream, thereby reducing the risk of hydraulic disruption and the associated health
risk. Installing sedimentation basins will cause a majority of oocysts in recycle flows to settle before
being returned to the primary treatment process, thereby eliminating the possibility they will pass
through the filter and risk public health. EPA believes the greatest benefit will be realized by systems
with the fewest number of filters because they are the most vulnerable to hydraulic and treatment
chemistry upset induced by recycle events. Since the volume of recycle flow is a larger percent of plant
influent at plants with fewer filters, they are more vulnerable to disruptions, in terms of both hydraulics
and water quality.

Similarly, changes in recycle practices among direct filtration systems will differ across the proposed
alternatives, and the resulting health benefits will differ. Under Rl, direct filtration plants are required to
return spent filter backwash prior to the point of primary coagulant addition. However, there are no
expected benefits under Rl because the analysis assumes direct filtration plants return recycle to the
required location. Under R2 and R3, States will determine whether  modifications to recycle practice is
required to reduce health risks, but R4 requires that all direct filtration plants install a sedimentation
basin if they do not already provide recycle flow treatment equivalent to or superior to sedimentation
for  recycle flow.  The reductions in health risks will depend on the number of systems that ultimately
modify recycle practice and the extent to which the modification increases oocyst removal from the
recycle flow.

Under any of the proposed alternatives, facilities may choose to alter their recycle practices by directly
discharging recycle flows to surface waters or publicly owned treatment works (POTW). In terms of
finished water quality, direct discharge generates the largest possible health benefits because

RIA for the Proposed LT1FBR                        5-40                                 February 15, 2000

-------
recycle flows containing oocysts are completely removed from the treatment plant, thereby
eliminating the risk of introducing oocysts from recycle flows into finished water.

Finally, in addition to the benefits from reduced occurrence of Cryptosporidium in the finished water,
the proposed recycle provisions are likely to reduce the occurrence of other contaminants in the
finished water. Giardia lamblia will likely be more effectively removed by the primary treatment
process under the proposed recycle provision. Furthermore, the changes in recycle practices that result
from LT1FBR may reduce the risk from other emerging disinfection resistant pathogens that may exist
in source water such as Toxoplasma, microsporidia, and Cyclospora.

Sensitivity Analysis for the Recycle Provisions

Available research literature demonstrates that increased hydraulic loading or disruptive hydraulic
currents, such as may be experienced when plants exceed State-approved operating capacity or when
recycle is returned directly into the sedimentation basin, can disrupt filter performance (Cleasby, 1963;
Glasgow and Wheatley, 1998; and McTigue et al., 1998) and sedimentation performance (Fulton,
1987; Logsdon, 1987; and Cleasby, 1990).  However, the literature does not quantify the extent to
which performance can be lowered and, more specifically, does not quantify the log reduction in
Cryptosporidium removal that may be experienced during direct recycle events.

In the absence of quantified log reduction data, EPA performed a sensitivity analysis at the system-level
for small and large systems to estimate a range of potential benefits for the recycle provisions. For the
analysis, EPA assumed both system sizes would meet the proposed 2.0 log removal for
Cryptosporidium provision except for problems caused by recycle practices that disrupt filter or
sedimentation performance.  The analysis incorporates the effect of these recycle practices by reducing
the average baseline log removal7 by a range of values (0.05 logs to 0.50 logs) to account for the
reduction in removal performance plants may experience if they exceed State-approved operating
capacity or return recycle to the sedimentation basin.  EPA assumed that installing equalization to
eliminate exceedences of State-approved operating capacity or moving the recycle return location from
the sedimentation basin to prior to the point of primary coagulant addition will result in health benefit by
returning the system to a 2.0 log removal of Cryptosporidium and thereby improving finished water
quality. The benefit estimate is conservative, because it does not account for the fact that recycle also
returns additional oocysts to the plant.

The difference between the number of illnesses that result from the 2.0 log removal assumption and the
reduced performance assumptions (i.e., 1.95 or 1.50 log removal) is used to calculate the annual
benefit using the $2,400 COI value.  EPA compared the benefit to cost estimates for returning recycle
prior to the point of primary coagulant additional  and installing equalization for two service populations:
       7The reduction in baseline log removal is an average over periods when recycle is and is not
occurring.  The actual reduction will be greater during recycle periods than other production periods.
For this sensitivity analysis, EPA assumed that the potential health impacts of recycle practices could be
captured by an average overall reduction in log removal.
February 15,2000                                  5-41                     RIA for the Proposed LT1FBR

-------
a service population of 1,900 persons, which represents a plant serving fewer than 10,000 people, and
a service population of 25,108, which represents a plant serving greater than 10,000 people.  Annual
benefits and annualized costs are summarized in Exhibits 5-20 and 5-21.
              Exhibit 5-20.  Potential Benefit for a System Serving 1,900 People
Log Removal
Reduction
0.05
0.50
Benefit for Population of
1,900
$1,400
$30,700
Cost of Moving Recycle
Return1
$5,200
$5,200
Cost of Installing Equalization1
$25,200
$25,200
 'Costs are annualized assuming a 7 percent discount rate over 20 years.
           Exhibit 5-21. Potential Benefit Range for System Serving 25,108 People
Log Removal
Reduction
0.05
0.50
Benefit for Population
of25,108
$18,700
$405,800
Cost of Moving
Recycle Return1
$18,700
$18,700
Cost of Installing
Equalization1
$57,200
$57,200
 'Costs are annualized assuming a 7 percent discount rate over 20 years.


Although research literature does not quantify the log reduction caused by specific recycle practices, the
results of the sensitivity analysis show that the benefit a plant serving 25,108 people would realize by
improving its baseline performance to 2.0 logs would range from $18,700 to $405,800. Benefits
would range from $1,400 to $30,700 for a plant serving 1,900. This benefit range supports EPA's
determination that unquantified benefits will justify costs.

5.4.2  Benefits of Disinfection Benchmark Provision

Disinfection benchmarking helps ensure that existing microbial protection is not significantly reduced or
undercut as a result of steps taken to comply with maximum contaminant levels (MCLs) for total
trihalomethanes (TTHMs) and 5 haloacetic acids (HAAS) set forth in Stage 1 DBF. The disinfection
benchmark provision will prevent future incremental illnesses associated with pathogens that are
controlled by current disinfection practices. However, it is not possible to quantify the health benefits
from disinfection benchmarking  on a national basis. The level of benefits will depend on how individual
systems alter their disinfection practices and how those alterations might have increased pathogen risks
without the disinfection benchmark provision in the proposed LT1FBR.
RIAfor the Proposed LT1FBR
5-42
February 15, 2000

-------
5.4.3  Benefits of Covered Finished Water Reservoirs

The quality of water in finished water reservoirs is subject to similar environmental influences as surface
water, including deposition of airborne chemicals, surface water runoff, animal carcasses, animal or bird
droppings, and growth of algae and other aquatic organisms.  In one study, sea gulls contaminated a 10
million gallon reservoir and increased bacteriological growth, and in another study waterfowl were
found to elevate coliform levels in small recreational lakes by 20 times their normal levels (Morra,
1979). Algal growth increases the biomass in the reservoir, which reduces dissolved oxygen and
thereby increases the release of iron, manganese, and nutrients from the sediments.  This, in turn,
supports more algal growth (Cooke and Carlson, 1989). Algae can cause drinking water taste and
odor problems.  Further, uncovered finished water reservoirs may be subject to contamination by illegal
swimming and dumping. Documented water quality problems in open finished water reservoirs include
increased algal cells; heterotrophic plate count (HPC) bacteria; turbidity; color; particle counts;
biomass; and decreased chlorine residuals (Pluntze, 1974; AWWA, 1983; Silverman et al.,  1983;
LeChevallier et al, 1997b).

Finished water may not be treated again prior to consumption, so any contamination in the uncovered
reservoir may be passed directly to the customer. Therefore, requirements to cover new finished water
reservoirs will result reduce the risk of contamination and result in positive health benefits.  Data are not
available, however, to quantify the benefits associated with covering all new finished water reservoirs.

5.4.4  Benefits from Including Cryptosporidium in the GWUDI Definition

Including Cryptosporidium in the definition of GWUDI will change drinking water treatment
requirements for a nonquantified number of small drinking water systems. Although EPA does not
currently have data on the number of systems that will be required to change treatment practices, the
Agency anticipates that the health benefits from increased oocyst log removal will be positive when
these small systems are reclassified as GWUDI.

5.4.5  Benefits from Including Cryptosporidium in Watershed Requirements for
       Unfiltered Systems

The proposed rule requires small unfiltered surface water and GWUDI systems to control
Cryptosporidium contamination within the watershed.  EPA expects that control of Cryptosporidium
will reduce the incidence of cryptosporidiosis in populations served by these small drinking water
systems.  EPA does not currently have data on the number of unfiltered or GWUDI systems that will be
required to control for Cryptosporidium; however, EPA anticipates that the health benefits will be
positive as these systems take proactive steps to minimize the potential for oocyst contamination within
watersheds.
February 15,2000                                 5-43                      RIA for the Proposed LT1FBR

-------
5.4.6 Risk Reduction from Emerging Pathogens

While the benefits analysis for the LT1FBR only includes reductions in illness and mortality attributable
to Cryptosporidium, the LT1FBR is expected to increase the level of protection from exposure to
other pathogens (i.e. Giardia or other waterborne bacterial or viral pathogens such as Cyclospora and
Microsporidium).  Strengthened filtration requirements will translate to increased removal of additional
pathogens and a resulting reduction in risk. This may prove essential, as the susceptibility of emerging
pathogens to inactivation by chlorination is not well established.  Unfortunately, EPA is unable to
quantify the resultant benefit associated with a reduction in risk from emerging pathogens due to current
data limitations.


5.5   Summary

EPA estimated the potential health benefits of the proposed rule using a health risk assessment
approach to characterize baseline infections, illnesses, and mortalities caused by exposure to
Cryptosporidium oocysts in treated drinking water from small surface water systems and small
GWUDI systems for the turbidity provisions.  Baseline estimates were compared with risk assessment
results incorporating the improved Cryptosporidium removal rates that are assumed to  occur because
of the requirements in the proposed rule that will alter treatment practices.

5.5.1  Summary of Quantified and Monetized Benefits

Exhibit 5-22 presents the mean value of avoided illnesses from the LT1FBR turbidity provisions under
the 1.2 liter per day daily water consumption assumption (see Exhibit 5-12).  The mean  value of
avoided illnesses with this consumption rate under a 2.5 log baseline removal  ranges from $53.9 million
under the low removal assumption to $92.4 million under the high removal assumption.  For the 2.0 log
removal baseline, mean benefits range from $150.3 million under low removal to $199.5  million under
the high removal assumption. Mortality results suggest that the mean value of avoided deaths under the
2.0 log removal baseline ranges from $45.0 million under low removal  to $59.8 million under the high
removal assumption.  The value of avoided mortalities under the 2.5 log removal baseline ranges from
$16.2 million to $21.3 million across the low and high removal assumptions.

EPA's Office of Water is continuing to evaluate drinking water  consumption data from USDA's 1994-
1996 CSFn study.  The drinking water consumption distribution used in this version of the benefits
analysis (averaging approximately 1.2 liters per day) reflects CSFn data. This distribution is currently
being considered as being reflective of water consumption among the population that consumes drinking
water from either community or noncommunity water supplies.
RIA for the Proposed LT1FBR                       5-44                               February 15, 2000

-------
      Exhibit 5-22. Summary of Annual Benefits Associated with Avoided Illnesses and
                     Mortalities for the Turbidity Provisions* (SMillions)
Log Removal Assumption
Low Removal
Avoided Illnesses
Mortalities
Total
Mid Removal
Avoided Illnesses
Mortalities
Total
High Removal
Avoided Illnesses
Mortalities
Total
Daily Drinking Water Ingestion and
Baseline Cryptosporidium\jog Removal Assumptions
Mean = 1.2 Liters per person
2.0 log
$150.3
$45.0
$195.3
$185.3
$55.5
$240.8
$199.5
$59.8
$259.4
2.5 log
$53.9
$16.2
$70.1
$66.2
$19.9
$86.1
$71.1
$21.3
$92.4
 * All values are in January, 1999 dollars. Totals may not equal detail due to rounding.


5.5.2 Summary of Non-Quantified Benefits

As noted in Sections 5.3 and 5.4, several types of potential benefits were not included in the
quantitative analysis. Exhibit 5-23 shows how the rule provisions that have not been quantified would
be expected to affect the overall benefits derived from LT1FBR.
February 15, 2000
5-45
RIA for the Proposed LT1FBR

-------
                           Exhibit 5-23.  Summary of Non-Quantified Benefits
                    Item
Potential Effect
  on Benefits
                    Comments
 Reducing mortality and morbidity rates by
 changes to recycle practices in small and
 large surface water and GWUDI treatment
 facilities.
                   Changing recycle practices is expected to
                   generate positive benefits by lowering the risk of
                   contracting cryptosporidiosis from drinking
                   water. See Section 5.4.1.
 Reducing risks to sensitive
 subpopulations
                   The study probably does not capture the full
                   value of benefits to sensitive subpopulations
                   because of a lack of scientific data.  See Section
                   5.2.5.
 Reducing outbreak risks and response
 costs
                   Determining the precise reduction in outbreak risk
                   and resulting benefits is not possible given
                   current information; however, the positive
                   benefits associated with a reduction in outbreak
                   risk are expected to be significant. See Sections
                   5.3.1, and 5.3.3.
 Improving aesthetic water quality
  +/no change
It is not clear that this rule will improve aesthetic
water quality and generate any associated
positive benefits; however, the rule is not
expected to reduce aesthetic water quality and
generate negative benefits. See Section 5.3.2.
 Reducing averting behavior (e.g., boiling
 tap water or purchasing bottled water).
                   Averting behavior is associated with both out-of-
                   pocket costs (e.g., purchase of bottled water) and
                   opportunity costs (e.g., time required to boil
                   water) to the consumer.  Reductions in averting
                   behavior are expected to have a positive impact
                   on benefits from the proposed rule. See Section
                   5.3.3.
 Covering new finished water reservoirs
                   Although insufficient data were available to
                   qualify benefits, the reduction of contaminants
                   introduced to finished water reservoirs would
                   produce positive public health benefits. See
                   Section 5.4.3
 Including Cryptosporidium in the
 definition of GWUDI
                   The change in GWUDI definition will change the
                   treatment practices at an undetermined number of
                   system. Although data to quantify the number of
                   systems are not available, EPA anticipates
                   increased health benefits.  See Section 5.4.4
 Including Cryptosporidium in watershed
 requirements for unfiltered systems
                   Similar to the inclusion of Cryptosporidium in
                   GWUDI, EPA anticipates that the proposed rule
                   increases health benefits.  See Section 5.4.5
 Reducing exposure to other pathogenic
 protozoa, waterborne bacteria, or viral
 pathogens
                   Exposure to other pathogenic protozoa, such as
                   Giardia, or other waterborne bacterial or viral
                   pathogens, are almost certainly reduced by the
                   recommended turbidity provisions but are not
                   quantified. See Section 5.4.6
 + = resolving the omission, bias, or uncertainty will tend to increase benefits.
 • = resolving the omission, bias, or uncertainty will tend to reduce benefits.
 +/• = the effect of the omission, bias, or uncertainty on benefits is undetermined.
RIAfor the Proposed LT1FBR
         5-46
                                  February 15, 2000

-------
5.5.3   Summary of Uncertainties

The benefits analysis has several biases and uncertainties. There is one source of identifiable bias in the
analysis.  The values for avoided illnesses were COI-based and not WTP-based; using WTP values
might increase benefits. WTP values tend to be greater than COI values because they include
nonpecuniary benefits of avoiding illness (i.e., benefits aside from avoiding out-of-pocket costs).

Exhibit 5-24 describes how uncertainties may affect the benefit analysis.  Although several  of the
sources of uncertainty were incorporated in the Monte  Carlo analysis, other sources could not be
incorporated in a quantitative manner.  These are described in the table.
               Exhibit 5-24.  Damages/Benefits Summary of Bias and Uncertainty
                   Item
Potential Effect
  on Benefits
Comments
  Biases
  COI values were used to monetize
  morbidity risk reductions
                 WTP values are generally higher than the
                 expected value of COI. No WTP values were
                 identified in the literature that were usable for
                 this analysis.
 Uncertainties
  The slope parameter k may be different
  for sensitive subpopulations.
                 If lower doses of Cryptosporidium to members
                 of sensitive subpopulations produce equivalent
                 responses as higher doses in healthy
                 individuals, a reduction in finished water oocyst
                 concentrations may only have positive benefits
                 to healthy populations and no change to
                 sensitive subpopulations.
 Different strains of Cryptosporidium may
 produce different dose-response
 relationships.
                 The dose-response relationship for C. Parvum
                 is used as a proxy for all Cryptosporidium
                 species until more complete experimental data
                 becomes available. Some strains are more
                 infectious, but less common while others are
                 equivalent to C. Parvum.
  Source water quality for the proposed
  rule is assumed to be identical to source
  water quality used to estimate IESWTR
  benefits.
                 Data are not available to show that source water
                 quality for small systems is different than, or the
                 same as, source water quality for large systems.
                 (Note: ICR data that could clarify this issue are
                 not yet available - see P. 5-15 for a discussion of
                 ICR data availability)
February 15, 2000
          5-47
   RIA for the Proposed LT1FBR

-------
                Exhibit 5-24.  Damages/Benefits Summary of Bias and Uncertainty
                    Item
Potential Effect
  on Benefits
Comments
  IESWTR source water measurements
  were from the eastern and central United
  States and may not be representative of
  the United States as a whole.
                   LeChevallier and Norton may have sampled
                   poorer quality source water (than the United
                   States as a whole) resulting in a higher
                   measured distribution and overstatement of
                   benefits by the model used in this RIA. The
                   outbreak information in Exhibit 2-7, however,
                   shows several outbreaks in the western and
                   southern United States.  Other appropriate
                   datasets were not identified for the current RIA
                   model.
  The existing analytic method provides
  poor Cryptosporidium recovery.
                   Poor recovery acts to produce a lower than
                   expected source water distribution and possible
                   understatement of benefits.
  Removal efficiencies for small systems
  may or may not be comparable to large
  systems.
                   EPA is currently evaluating whether small
                   system removal efficiencies are comparable to
                   large system removal efficiencies. Small systems
                   may have comparable removal efficiencies,
                   which could lead to overstatement of benefits
                   by the current model.
  The mortality rate of 0.0125 percent used
  in the model is based on data from the
  Milwaukee outbreak.
                   The mortality rate from Milwaukee may reflect
                   overall mortality rates from low level exposure
                   by immunocompromised individuals. The
                   population of immunocompromised individuals
                   in other towns and cities may be different than
                   Milwaukee and, thus, result in different mortality
                   rates.
  The VSL was not adjusted to reflect
  mortalities in sensitive subpopulations.
                   The valuation literature is unclear as to whether
                   risk to sensitive subpopulations is associated
                   with a higher or lower WTP.  Review of existing
                   literature does not provide a strong basis for
                   adjusting the VSL up or down.
  Transient Systems may not be accurately
  characterized.
                   There is uncertainty as to the precise number of
                   transient systems.  Also, it is assumed that
                   transient systems use approximately the same
                   technologies as community water systems. This
                   may overestimate the number of systems
                   affected by the rule. Also, the number of people
                   served by these systems might be over or
                   underestimated, to a lesser degree.
  + = resolving the omission, bias, or uncertainty will tend to increase benefits.
  • = resolving the omission, bias, or uncertainty will tend to reduce benefits.
  +/• = the effect of the omission, bias, or uncertainty on benefits is undetermined.
RIA for the Proposed LT1FBR
        5-48
              February 15, 2000

-------
                                     6.  Cost Analysis

6.1    Introduction

This chapter reports national cost estimates for the proposed Long Term 1 Enhanced Surface
Water Treatment and Filter Backwash Rule (LT1FBR) and discusses the methods EPA used to
estimate implementation costs incurred by drinking water systems and States.1  EPA anticipates
that water system compliance with the proposed LT1FBR provisions will increase monitoring and
reporting burdens and entail adjustments to existing treatment processes and plant operations. EPA
also expects the proposed rule to increase the labor requirements for additional compliance
tracking activities among States. Consequently, the cost analysis includes labor costs associated
with additional monitoring, reporting, and compliance requirements, as well as capital and
operating and maintenance (O&M) expenditures associated with changes in water treatment
processes.2

Sections 6.2 through 6.5 provide detailed cost information for each component included in the
quantitative analysis. Each section includes the cost assumptions and data elements used in
the analysis, describes how the costs were estimated, and reports the results.  Additional
documentation for the cost estimates in this  chapter can be found in Appendices C through H.
Total national costs are summarized in Section 6.6, which also includes a discussion of the impact
of potential biases, omissions, and uncertainties on the national cost estimate.  Section 6.7 discusses
how system-level costs were translated into  annual cost increases per household. Section 6.8
summarizes a cost effectiveness analysis that estimates costs per illness avoided.

6.1.1   Cost Assumptions

EPA estimated costs at the water system and State level, then multiplied these costs by the number
of affected entities to obtain total costs.  EPA used existing data sources and stakeholder inputs to
determine system and State responses to the proposed LT1FBR.  This included identifying
treatment process improvements that systems may implement and estimating labor burdens for
monitoring and reporting activities.  EPA estimated costs for these various responses using industry
cost models, equipment prices, and wage rates from standard engineering sources, stakeholder
inputs, as well as assumptions in the IESWTR RIA.  System costs were estimated for several
        1 Throughout the cost analysis, the term State refers to the 56 States, Commonwealths, Territories, and the
District of Columbia that are eligible for primary enforcement authority or primacy. This definition is consistent
with the assumption used for the cost analysis in the final IESWTR RIA. Currently, however, Wyoming and the
District of Columbia do not have primacy; EPA regional offices administer their drinking water programs. Indian
Tribes are also eligible for primacy, although none have yet obtained it and EPA regional offices also administer
drinking water programs for Tribal lands.

        2 This analysis of social costs is limited to compliance cost estimates. Consequently, costs may be
overstated because consumer and producer responses that minimize cost impacts are not incorporated. The
analysis assumes that drinking water systems pass incremental costs on to consumers in the form of higher water
prices and there are no impacts such as system closure.


February 15, 2000                                  6-1                       RIA for the Proposed LT1FBR

-------
different service population size categories because systems that serve larger populations will often
incur higher costs.
To be consistent with the annual basis used in the benefit analysis, EPA also estimated costs on an
annual basis. All one-time costs such as investments in capital equipment or training were
annualized before they were added to annual O&M expenditures or recurring annual labor costs.
Capital costs for most process improvements3 were annualized over a 20-year period to reflect a
typical capital investment lifetime. Two results are reported because EPA used two different
discount rates that have been recommended for policy analysis: a 7 percent rate, which is
recommended by OMB guidance (OMB, 1993 and 1996), and a 3 percent rate,  which is
recommended in Guidelines for Preparing Economic Analyses (U.S. EPA, 1999c). Start-up labor
costs were also  annualized over 20 years to obtain equivalent annual values.

The process improvements that EPA used to estimate costs for the turbidity provisions were
revised from a set of technology enhancements used in the IESWTR RIA cost analysis to
reflect conditions at smaller facilities. For the recycled provisions, the set of feasible process
improvements was limited. EPA relied on information provided in  stakeholder  and SBREFA
meetings,  best professional judgment, the schematic of ICR systems, and the AWWA survey
(1998) of recycling practices, which indicated the current range of recycling techniques to develop
a list of potential treatment changes.

System-level cost estimates for all turbidity or recycle modifications are described in detail in The
Cost and Technology Document for the Long Term 1 Enhanced Surface Water Treatment and
Filter Backwash Rule  (U.S. EPA,  1999b). Most capital and O&M costs are functions of system
flow rates, which were obtained from the Community Water System Survey database.
Consequently, these costs are more representative of costs for community water systems, but EPA
used the same costs for noncommunity systems as well  because of a lack of data concerning the
flow capacities  and technologies employed by  these systems. Thus, the cost analysis may over
estimate costs for noncommunity systems.

Overall costs for the provisions affecting only small systems will also be overestimated because
EPA included "purchased water" systems in the baseline. These systems purchase water from
other systems and,  therefore, are not likely to incur treatment or monitoring costs.  EPA included
these systems in the baseline to obtain a better  estimate  of benefits.  The SDWIS database
associates the population served with these systems and not the wholesale systems that treat and
supply the water.  Consequently, if EPA included only the wholesale systems in the RIA, the
affected population would be under estimated;  including the purchased systems  adds the missing
population, but  also adds systems to the cost analysis.

EPA's labor cost estimates incorporate assumptions about incremental system and State labor
hours and  hourly labor rates for managerial and technical labor categories. For systems, EPA used
labor rates based on a range of rates recommended by the Technology Design Panel (TDP). To
verify that these labor rates were consistent with the January 1999 basis for benefits and costs,
       3 The exception is individual filter turbidimeter installation, which was annualized over a 7-year period to
reflect a shorter equipment lifetime.


RIA for the Proposed LT1FBR                       6-2                               February 15, 2000

-------
EPA compared the rates to the rates for water and wastewater treatment system operators in the
1998 1999 Occupational Outlook Handbook (BLS, 1999b).  The TDP-recommended ranges,
particularly the low to medium rates, were consistent with the reported labor rates in BLS (1999b),
after escalating the latter to December 1998 dollars using the Employment Cost Index (BLS,
1999a).4

The loaded technical labor rate for systems with design flows under 1.0 million gallons per day
(i.e., systems serving 3,300 or fewer) is $28 per hour, and the rate for systems with design flows
above 1.0 mgd (i.e., systems serving 3,301 or more) is $40 per hour.  EPA assumed that systems
serving 1,000 or more also have a management labor category, which has a loaded labor rate of
$56 per hour.  All of the labor rates in the cost analysis incorporate a 1.4 load factor on top of a
base hourly rate to account for fringe benefits and other nonwage costs.5 Thus, the base labor rates
are $20 per hour for technical labor at systems serving 3,300 or fewer; approximately $28.60 for
technical labor for larger systems, and $40 per hour for managerial labor.

State labor rate assumptions are the same as those used in the IESWTR RIA, escalated from June
to December 1998 dollars using the Employment Cost Index for State and local employees (BLS,
1999a). The unloaded hourly rate for technical  staff is $15.21 and the unloaded rate for
managerial staff is $22.31.  Loaded rates, assuming the same 1.4 load factor used in the system
estimates  above, are $21.29 and $31.23, respectively.

To reduce the potential burden on small systems, EPA developed and evaluated the cost
implications of several regulatory alternatives for the following provisions: individual filter
turbidity monitoring, disinfection benchmark applicability monitoring, disinfection benchmark
profiling,  and recycle provisions.  Chapter 3 describes the alternatives in detail, they are briefly
summarized below in the cost  analysis sections for these provisions.

6.2   Turbidity Provisions Cost Analysis

The national annual cost estimate includes costs for two provisions that address finished water
turbidity levels. The combined filter effluent (CFE) provision has requirements that differ across
filtration methods. Systems using conventional or direct filtration must meet a 95th percentile
turbidity value of 0.3 nephelometric treatment units (NTUs) and a maximum turbidity value of 1
NTU.  Systems using membrane filtration will be required to meet these standards or standards
determined by the State not to exceed 1 NTU in 95 percent of monthly measurements and a
maximum of 5 NTU.  Systems using alternative filtration methods will need to meet turbidity
standards  determined by the State, not to exceed 1 NTU in 95 percent of monthly measurements
       4 The quarterly index does not allow an exact match with January 1999 units for all other cost items. The
December 1998 index value is a closer approximation than the March 1999 value; the growth rate between these
two index values was 0.6 percent and any partial adjustment is not expected to affect reported costs, which are
generally reported in tenths of millions of dollars.

       5 The load factor for fringe benefits and supplemental pay for professional and technical occupations is
approximately 1.3 (BLS, 1999a), and the IESWTR cost analysis used a slightly higher load factor of 1.4 to include
other costs. The same rate is used to develop labor costs for the proposed LT1FBR.


February 15, 2000                                 6-3                     RIA for the Proposed LT1FBR

-------
and a maximum of 5 NTU.  The second provision requires all systems that use conventional or
direct filtration to monitor individual filter turbidity levels and submit exceptions reports when
turbidity levels exceed reporting thresholds.

To comply with the CFE provision, some conventional and direct filtration systems will need
to implement process improvements. The applicable annual cost estimate for those systems include
the annualized capital cost of the new equipment and the annual cost of incremental labor and
supplies (e.g., electrical power and polymer stock) needed to operate and maintain the equipment.
Systems that use other filtration methods will need to demonstrate that their systems meet the
microbial removal or inactivation goals noted in the proposed rule, which will form the basis for
State determinations regarding their turbidity standards.  Annual costs will comprise annualized
demonstration and determination costs.

Exhibit 6-1 summarizes EPA's estimate of the number of systems potentially affected by each
element of the turbidity provisions.  The costs associated with each system and the methods for
estimating the numbers of affected systems are described in more detail in Sections 6.2.1 through
6.2.3.
             Exhibit 6-1. Summary of the Estimated Number of Small Systems
                             Affected by Turbidity Provision
Provision
System Size Category
•100
101-500
501-1,000
1,001-3,300
3,301-9,999
Total
Combined Filter Effluent Provision
Turbidity Treatment
Modifications
(Section 6.2.1)
341
456
331
675
603
2,406
Individual Filter Monitoring Provision
Monitoring
(Section 6.2.2)
Exceptions Reporting
(Section 6.2.3)
Individual Filter
Assessment (Section 6.2.3)
Comprehensive
Performance Evaluation
(Section 6.2.3)
836
150-167
33
17
1,117
201-223
45
22
810
146-162
32
16
1,655
298-331
66
33
1,478
266-296
59
30
5,896
1,061-1,179
236
118
 Detail may not add to total due to rounding.

6.2.1  Combined Filter Effluent Provision: Turbidity Treatment Costs

Unit Costs

EPA identified 24 treatment process improvements that small conventional and direct filtration
systems might implement to improve finished water quality to meet the proposed CFE turbidity
RIAfor the ProposedLT1FBR
6-4
February 15, 2000

-------
standards.  The unit costs developed for these process improvements are based on cost models,6
best engineering judgment, and existing cost and technology documents. In most cases, EPA
derived costs from estimates of system design and average flow rates. The Cost and Technology
Document for the Long Term 1 Enhanced Surface Water Treatment and Filter Backwash Rule
(U.S. EPA, 1999b) describes the methods and assumptions used to develop unit costs.  Exhibit
6-2a reports capital costs per system for each treatment process improvement by system size
category.  These are actual costs; they do not reflect annualized values. Exhibit 6-2b reports the
annual O&M costs.

Compliance Forecast

EPA based the compliance forecast on its understanding of current levels of finished water
turbidity and the requirements in the proposed rule. Systems generally measure turbidity in two
ways: as the output from an individual filter and as a combined stream of all  individual filter
outputs (i.e., combined filter effluent). During the development of the rule, EPA analyzed CFE
maximum  and 95th percentile turbidity results from finished water turbidity data provided by States
to determine how many systems currently fail to meet the proposed turbidity standards. Predicted
compliance was measured as meeting a limit such as 0.3 NTU at least 95 percent of the time and
not exceeding a CFE maximum such as 1 NTU.  In general, plants that expect to meet a 0.3 NTU
limit 95 percent of the time will target operations to achieve 0.2 NTU to ensure that they would
consistently meet the 0.3 NTU level.  EPA took this into consideration when it developed
estimates of the numbers of systems expected to modify treatment.

Using the baseline information on filtration practices and finished water turbidity results, which
was discussed in Chapter 4, EPA developed a compliance forecast for the proposed 95th percentile
turbidity standard of 0.3 NTU.  EPA also developed forecasts for two other standards, 0.2 NTU
and 0.1 NTU, to evaluate the effect of more stringent standards on marginal treatment costs (see
Appendix tables C-l through C-3  for detailed compliance forecasts). A compliance forecast first
estimates the number of systems expected to modify their treatment practices to meet the turbidity
requirements, and then it identifies the process modifications they would likely  select.  The number
of systems requiring each process modification were multiplied by the unit costs in Section 6.2.1 to
obtain total costs per treatment alternative. Costs were then summed across the treatment
alternatives to obtain total costs and then annualized using the 3 percent and 7 percent discount
rates.

The compliance forecast estimates in Appendix tables  C-l through C-3 vary across the five
system size categories because compliance needs will vary by system size. Furthermore, the
treatment modifications are generally not mutually exclusive (i.e., some systems are expected to
       6 EPA used three cost models: the Very Small System (VSS) Model, which is a spreadsheet model based on
cost equations in Very Small Systems. Best Available Technology Document (U.S. EPA, 1993); the WATER Model,
which is a spreadsheet model based on cost equation in Estimation of Small System Water Treatment Costs (U.S.
EPA, 1984); and the WAV COSTS Model, which contains cost estimating routines based on cost equations in
Estimating Water Treatment Costs.  Volume 2.  Cost Curves Applicable to 1 to 200 mgd Treatment Plants
(U.S. EPA, 1979).


February 15, 2000                                 6-5                      RIA for the Proposed LT1FBR

-------
           Exhibit 6-2a.  Treatment Process Improvement Capital Costs per System
                                      (January 1999 dollars)
System Population Size Categories
•100
101-500
501-1,000
1,001-3,30
0
3,301-9,999
Chemical Addition
Install coagulant aid polymer feed capability
Install backwash water polymer feed capability
Install pH adjustment for enhancing alkalinity
surposes
$9,016
$9,016
$8,137
$9,016
$9,016
$8,137
$9,016
$9,016
$8,137
$9,016
$9,016
$8,137
$9,016
$9,016
$8,137
Coagulant Improvements
Primary coagulant feed points, control, measurement
$9,016
$9,016
$9,016
$9,016
$9,016
Rapid Mixing
Rapid mix improvements — mechanical
Rapid mix improvements — structural
$2,444
$2,852
$2,444
$2,852
$3,157
$3,768
$3,768
$4,583
$6,212
$9,064
Flocculant Improvements
Flocculation improvements — mechanical
Flocculation improvements — structural
$8,351
$15,888
$9,064
$22,406
$12,730
$39,210
$16,397
$52,755
$32,081
$100,418
Settling Improvements
Equipment modification — weirs, inf/effl, etc.
Add tube settlers
$519
$2,953
$1,348
$10,694
$3,157
$28,882
$6,824
$70,476
$14,665
$207,150
Filtration Improvements
Filter media additions
Filter media overhaul
Backwashing — increase flow/velocity
Backwashing — install surface wash
Post backwash filter-to -waste
Filter control systems
Individual filter turbidimeter installation '
Membrane (microfiltration)
$415
$7,333
$9,879
$10,388
$3,055
$2,139
$3,941
$56,523
$933
$28,516
$15,277
$17,212
$5,194
$4,176
$3,941
$162,441
$1,660
$76,281
$22,609
$30,451
$10,490
$7,027
$3,941
$341,482
$3,259
$186,170
$65,282
$117,324
$46,543
$29,229
$7,862
$741,014
$12,221
$423,466
$144,516
$162,339
$55,505
$36,358
$10,816
$1,635,914
Administrative Culture Improvements
Plant staffing — increase (1 or 2 persons) 2
Staff qualifications 2
$0
$0
$0
$0
$0
$0
$0
$0
$0
$0
Laboratory Modifications
Bench top turbidimeter purchase — replace obsolete
units
Jar test apparatus purchase
Alternative process control testing equipment
$1,293
$2,342
$8,534
$1,293
$2,342
$8,534
$1,293
$2,342
$8,534
$1,293
$2,342
$8,534
$1,293
$2,342
$8,534
Process Control Testing Modifications
Staff training (advanced)
$4,888
$4,888
$4,888
(T/l OQQ
J^OOO
$4,888
 'Turbidimeter installation was included with other capital costs for modeling purposes although this treatment
 change will be undertaken because of the individual turbidity monitoring provision instead of the CFE provision.
 2 There are no capital costs for this treatment process improvement.
RIAfor the ProposedLT1FBR
6-6
February 15, 2000

-------
  Exhibit 6-2b.  Treatment Process Improvement Operation & Maintenance Costs per System
                                      (January 1999 dollars)
System Population Size Categories
•100
101-500
501-1,000
1,001-3,30
0
3,301-9,999
Chemical Addition
Install coagulant aid polymer feed capability
Install backwash water polymer feed capability
Install pH adjustment for enhancing alkalinity purposes
$2,908
$2,908
$5,581
$2,908
$2,908
$5,707
$2,908
$2,908
$5,991
$2,908
$2,908
$6,814
$4,081
$4,074
$12,246
Coagulant Improvements
Primary coagulant feed points, control, measurement
$2,908
$2,908
$2,908
$2,908
$4,081
Rapid Mixing
Rapid mix improvements — mechanical
Rapid mix improvements — structural
$2,803
$2,803
$2,803 $2,953
$2,803
$2,953
$3,157
$3,157
$5,296
$5,296
Flocculant Improvements
Flocculation improvements — mechanical
Flocculation improvements — structural
$2,750
$2,852
$2,852
$2,852
$2,852
$2,852
$2,953
$2,953
$4,787
$4,787
Settling Improvements
Equipment modification — weirs, inf/effl, etc.
Add tube settlers
$741
$741
$741
$741
$2,224
$2,224
$2,224
$2,224
$6,355
$6,355
Filtration Improvements
Filter media additions2
Filter media overhaul2
Backwashing — increase flow/velocity
(10-20% increase)
Backwashing — install surface wash
Post backwash filter-to -waste
Filter control systems
Individual filter turbidimeter installation1
Membrane (microfiltration)
$0
$0
$3,870
$1,731
$3,768
$2,750
$825
$13,647
$0
$0
$4,277
$2,037
$4,074
$3,157
$825
$23,730
$0
$0
$4,583
$2,241
$4,379
$3,972
$825
$46,543
$0
$0
$6,212
$4,583
$5,907
$8,555
$825
$90,132
$0
$0
$17,619
$5,805
$8,555
$18,332
$825
$251,146
Administrative Culture Improvements
Plant staffing — increase (1 or 2 persons)
Staff qualifications
$14,828
$672
$14,828
$672
$29,657
$672
$29,657
$713
$42,367
$1,080
Laboratory Modifications
Bench top turbidimeter purchase — replace
obsolete units
Jar test apparatus purchase2
Alternative process control testing equipment
$76
$0
$1,483
$76
$0
$1,483
$76
$0
$1,483
$76
$0
$1,483
$76
$0
$2,118
Process Control Testing Modifications
Staff training (advanced)2
$0
$0
$0
$0
$0
 'Turbidimeter installation was included with other O&M costs for modeling purposes although this treatment change
 will be undertaken because of the individual turbidity monitoring provision instead of the CFE provision.
2There are no O&M costs for this treatment process improvement.
February 15, 2000
6-7
RIAfor the Proposed LT1FBR

-------
adopt more than one of the treatment process improvements). Consequently, the sum of
percentages across the treatment process improvements exceeds 100 for each system size category.

The percentages provided in the compliance forecast tables in Appendix C generally indicate the
percent of the systems expected to modify treatment that will adopt each treatment process change.
For individual filter turbidimeter installation, however, the percentage applies to all systems
affected by the proposed individual filter turbidity monitoring provision, not just to those systems
that are expected to modify their treatment to meet the turbidity levels in the rule.  All systems that
practice conventional or direct filtration will be required to install individual filter turbidimeters
under the rule, regardless of current performance. For the IESWTR RIA, EPA assumed that
approximately 20 percent of systems already have turbidimeters in place and 80 percent will need
to install a turbidimeter for each filter. EPA has no data to suggest that small systems are any more
or less likely to have turbidimeters installed.  Thus, EPA will use the same assumption until better
data for small systems is available.

EPA estimated that 2,406 or 41 percent of the 5,896 conventional and direct filtration systems
incur treatment modification costs to meet a revised turbidity standard of 0.3 NTU.7  Exhibit 6-3
summarizes the number of systems needing to modify treatment by size category. Treatment
modifications for the proposed 0.3 NTU standard and 0.2 NTU sensitivity analysis included a
wide variety of technologies. For the 0.1 NTU sensitivity analysis, however, EPA assumed that
increased protection would be achieved primarily through adoption of membrane technology
rather than altering other treatment practices to reduce turbidity.

System Costs

Exhibit 6-4 summarizes annual cost estimates including annualized capital costs and annual O&M
expenditures by system size and turbidity level. Total annualized costs for the proposed 0.3 NTU
standard are $47.4 million to $52.2 million, depending on the discount rate assumption. The
sensitivity analysis shows that costs increase rapidly for more stringent turbidity standards.  Total
costs for the 0.2 NTU case are approximately 157 percent higher than costs for the 0.3 NTU
standard (at the 7 percent discount rate), and costs for the 0.1 NTU case are approximately 675
percent higher than the 0.3 NTU costs. As noted, the cost estimates for the 0.2 NTU and 0.1 NTU
cases are likely to be under estimated because the number of systems modifying treatment was
assumed to be the same; if more systems would need to modify treatment to meet the stronger
standards, costs would be higher than those reported in Exhibit 6-A. EPA identified several cost
drivers for the 0.3 NTU and 7 percent discount rate assumptions. O&M expenditures account for
59 percent of annual costs; the remaining 41  percent is annualized
       7 EPA assumed that the number of systems expected to modify treatment remains the same for the 0.2 NTU
and 0.1 NTU sensitivity analyses, but altered the mix of treatment changes. This potentially under estimates the
marginal costs for the 0.2 NTU and 0.1 NTU standards because it excludes costs that might accrue to systems that
currently meet the 0.3 NTU standard (i.e., that currently achieve 0.2 NTU turbidity levels at least 95 percent of the
time), but do not meet a 0.2 NTU or a 0.1 NTU standard.


RIA for the Proposed LT1FBR                        6-8                                February 15, 2000

-------
          Exhibit 6-3. Number of Systems Modifying Treatment Practices to Meet
                                 New Turbidity Standards
System Size
• 100
101-500
501-1,000
1,001-3,300
3,301-9,999
Total
Number of Conventional
and Direct Systems
836
1,117
810
1,655
1,478
5,896
Number of Systems Expected
to Modify Treatment
341
456
331
675
603
2,406
        Exhibit 6-4. Annual Cost Estimates for Turbidity Treatment Requirements
                               (January 1999 dollars, millions)
System Size
• 100
101-500
501-1,000
1,001-3,300
3,301-9,999
Total
0.3 NTU
3%
$4.3
$5.6
$5.5
$14.4
$17.5
$47.4
7%
$4.5
$6.0
$5.9
$16.0
$19.9
$52.2
0.2 NTU
3%
$5.4
$8.7
$10.1
$33.7
$62.2
$120.0
7%
$5.7
$9.4
$11.1
$38.0
$69.9
$134.1
0.1 NTU
3%
$8.0
$17.8
$24.8
$96.7
$212.6
$360.0
7%
$8.6
$19.8
$27.9
$109.9
$238.5
$404.6
Results for the 3 percent discount rate are in Appendix D and results for the 7 percent discount rate are in Appendix E.
Detail may not add to total due to rounding.
capital costs. Approximately 36 percent of total O&M expenditures are for plant staffing
increases. Plant staffing is one of four process improvements that together account for almost 50
percent of total turbidity treatment costs:

           Filter control systems (8 percent of total costs)
       •   Filter media overhaul (8 percent of total costs)
       •   Backwashing-install surface wash (10 percent of total costs)
           Plant staffing increase (21 percent of total costs).

6.2.2  Individual Filter Monitoring Provision:  Monitoring Costs

The proposed rule requires that all small surface water or GWUDI systems using conventional or
direct filtration continuously monitor turbidity for each filter in their system.  This section discusses
EPA's estimate of monitoring costs for systems and States. Turbidity monitoring costs include
both start-up and annual costs for systems and States.  In each case, the underlying estimation
approach is the same. Costs for monitoring activities reported below, however, do not include the
February 15, 2000
6-9
RIAfor the Proposed LT1FBR

-------
capital and O&M costs of the turbidimeters, which were included in the previous discussion on
turbidity treatment.  EPA estimated that total annualized costs for turbidimeters will be
approximately $9.0 million assuming a 3 percent discount rate or $9.8 million assuming a 7 percent
discount rate.  Annual O&M expenses for calibration materials are $3.9 million and annualized
capital expenses account for the remainder.

The following analysis includes cost estimates for three alternatives, which are described below.
These alternatives will affect both the frequency and duration of system monitoring activities as
well as the number of exceptions reported, which is discussed in Section 6.2.3.

       •   Alternative Tl: Individual filter monitoring and exceptions reporting requirements are
           identical to the final IESWTR. They include a requirement to submit an exceptions
           report for any individual filter that exceeds 1 NTU in two consecutive measurements
           taken 15 minutes apart at any time or for any individual  filter that exceeds 0.5 NTU in
           two consecutive measurements taken 15 minutes apart at the end of 4 hours of filter
           operation, which necessitates daily analysis and review of turbidity data gathered by
           the turbidimeters. Finally, a filter profile is required when any of the above
           exceedances cannot be explained.

       •   Alternative T2: Individual filter monitoring and exceptions reporting requirements are
           slightly revised from the final IESWTR provisions of Tl to exclude  the exceptions
           report for an individual filter that exceeds 0.5 NTU in two consecutive measurements
           taken 15 minutes apart at the end of 4 hours of filter operation, which allows systems
           to shift from daily to weekly analysis and review of the monitoring data if they so
           chose. For the cost analysis, EPA assumed that all systems review data weekly.
           Finally, the filter profile requirement does not apply.

       •   Alternative T3: Individual filter monitoring and exceptions reporting requirements are
           revised to a distributional standard.  Only systems that exceed 0.5 NTU in more than 5
           percent of monthly measurements or exceed 2 NTU in two consecutive measurements
           are required to submit an exceptions report. Thus, monthly analysis review of data
           should be sufficient to detect instances that require an exceptions report. The filter
           profile requirement does not apply.

System Costs

System start-up activities were based on the list of activities included in the IESWTR cost analysis,
which were discussed with small entity representatives and stakeholders during the SBREFA and
stakeholder meetings. System start-up activities include reading and understanding the rule,
mobilization and planning, and employee training.  The cost analysis assumes durations for each
activity that adequately reflect staffing and expertise typically found  in small systems. First, it
assumes that system managers (or system operators for systems serving 1,000 or fewer, which  are
assumed to have no managerial staff) would spend 6 hours reviewing the rule to understand the
monitoring provision and how it affects their operations.  Second, it  assumes that 30 hours are
required for mobilization and planning activities, (e.g., assessing current plant operations and
employee schedules to develop a strategy for monitoring the turbidity data.)  Finally, it assumes 16

RIA for the ProposedLT1FBR                       6-10                               February 15, 2000

-------
hours for additional staff training among systems serving 1,001 to 3,300, and 40 hours for training
among systems serving 3,301 to 9,999.  There are no training costs for the three smallest system
size categories because the first two activities are assumed to be sufficient given their small staff
size.

Annual system monitoring activities at the plant level include data analysis, data review, and
recordkeeping.  Exhibit 6-5 summarizes the activities for each of the three monitoring alternatives
EPA developed, and shows that labor hour assumptions differ by system size.  The larger size
categories will require more time for data analysis and review because they have more filters and,
therefore, more turbidimeter readings to review.

   Exhibit 6-5. Summary of Labor Requirements for Turbidity Monitoring Alternatives
Compliance
Activity
Data Analysis
Data Review
Recordkeeping
System Size
• 1,000
1,001-9,999
• 1,000
1,001-9,999
all systems
Alternative Tl
(minutes/day)
15
30
15
30
2 hours/month
Alternative T2
(preferred option)
(minutes/week)
10
15
10
15
2 hours/month
Alternative T3
All systems will require at
most 2 hours/month
for data analysis, review,
and recordkeeping.
The burden estimate per system for Alternative Tl is about five times larger than the estimate for
Alternative T2 because it requires more data analysis. The burden estimate for Alternative T3 is
the smallest because the streamlined exceptions reporting requirement substantially reduces the
amount of time operators need to review data. EPA assumed that operators can complete the
required Alternative T3 analysis in approximately 24 hours per year compared to slightly higher
burdens for Alternative T2 (i.e., 41 to 50 hours across size categories), and substantially higher
burdens for Alternative Tl (i.e., 207 to 389 hours).

Estimated annual costs to systems for turbidity monitoring range from $5.6 million for Alternative
T3 to $63.3 million for Alternative Tl (Exhibit 6-6).  The labor burden for annual monitoring and
reporting requirements ranges from 140,000 (T3) to 1.8 million (Tl) hours per year. The
annualized national system start-up and implementation costs are $1.2 million assuming a 7 percent
discount rate. The total labor burden associated with system start-up activities is almost 300,000
hours.
February 15, 2000
6-11
RIAfor the Proposed LT1FBR

-------
   Exhibit 6-6.  Annual System Turbidity Start-Up and Monitoring Cost by Alternative
                                   (January 1999 dollars)
Compliance
Activities
Annualized Monitoring Start-Up Cost (7%)'
Annual Monitoring Cost
Total Annual Cost2
Annual Cost by Alternative ($ millions)
Alternative
Tl
$1.2
$62.1
$63.3
Alternative T2
(preferred)
$1.2
$8.8
$10.1
Alternative
T3
$1.2
$4.4
$5.6
 Detail may not add to total due to independent rounding. See Appendices G-la through G-lf for detail.
 'Total start-up cost of $13.0 million is annualized over 20 years assuming a 7 percent discount rate. Using the 3 percent discount
 rate, annualized start-up cost is $0.9 million.
 2Total annual cost assuming a 3 percent discount rate is $63.0 million (Tl), $9.7 million (T2), and $5.3 million (T3).
State Costs

One-time State start-up activities include 12 hours to review the final rule, 120 hours for
mobilization and planning activities, and 120 hours for State staff training.  The cost analysis
assumes that managerial staff account for about 80 percent of these hours and technical staff
account for the remaining 20 percent of hours.  These assumptions are similar to the cost analysis
in the final IESWTR RIA and may overstate costs if similar activities undertaken to implement the
IESWTR reduce the subsequent start-up burden for the proposed LT1FBR.

The State's annual responsibility under the rule includes ensuring that all systems are in
compliance by reviewing monthly reports from each system.  These reports indicate whether
individual filter monitoring occurred.  State activities also include reviewing exceptions reports,
record keeping, and determining compliance.  State activities for the proposed rule are based on
assumptions made for the cost analysis in the final IESWTR RIA, which were based on interviews
with State officials, a review of similar regulatory requirements, and confirmation by the M-DBP
Committee.  The burdens have been adjusted from the IESWTR RIA assumptions to reflect effort
levels more appropriate for tracking compliance for small systems.

Exhibit 6-7 summarizes the estimated State cost of implementing the individual filter turbidity
monitoring provision. The rule would collectively cost States an estimated $413,000 in start-up
costs.  Amortizing this cost at 7 percent results in an annual cost of almost $40,000. The national
labor burden for the State program start-up is estimated to exceed 14,000 hours.  Annual
monitoring costs are $838 per system  and the total cost for 5,896 systems is approximately $4.9
million.  The annual  labor burden is approximately 212,000 hours.  These costs are more
applicable for alternatives Tl and T2; costs for T3 might be higher because States may need to
establish two tracking systems—one for small systems and one for large systems—because the
exceptions reporting requirements for small systems differ from the requirements in IESWTR.
Maintaining two reporting systems might impose annual costs on States that offset the potential
cost-savings of T3. Consequently, EPA believes that T2 may actually be more cost effective in
the long run even though the estimated costs of T3 are lower.
RIA for the Proposed LT1FBR
6-12
February 15, 2000

-------
            Exhibit 6-7. State Turbidity Start-Up and Monitoring Annual Cost
                                    (January 1999 dollars)
Respondents Unit Cost Annual Cost
Compliance Activities Affected ($) (S millions)
Annualized Monitoring Start-Up Cost (7%)'
Annual Monitoring Cost
56 Entities
5,896 Systems
$7,373
$838
Total Annual Cost2
$0.04
$4.94
$4.98
 See Appendices G-la, G-lb, and G-lg for detail.
 'Total start-up cost of $0.4 million is annualized assuming a 20-year time period and a 7 percent discount rate. Assuming a 3
 percent discount rate, annualized start-up cost is $0.03 million.
 Total annual cost assuming a 3 percent discount rate is $4.97 million.

6.2.3  Individual Filter Monitoring Provision:  Exceptions Reporting Costs

If monitoring activities indicate that individual filter turbidity levels exceed certain thresholds, the
proposed rule requires that systems submit an exceptions report to the State. If exceedances are
persistent,  systems may be required to conduct an Individual Filter Assessment (IFA).  States will
need to review the  exceptions reports, and may need to complete Comprehensive Performance
Evaluations (CPE).

The  regulatory alternatives differ primarily with respect to what turbidity levels trigger an
exceptions report.  These differences generated a wide range of burden estimates for data
collection and analysis activities, which was discussed in the previous section.  EPA expects,
however, that the overall effect on the number of exceptions reported will be minimal,  reflecting
comparable levels of filter problem detection and health protection across the alternatives. Exhibit
6-8  summarizes the exceptions reporting requirements for the alternatives.  It also describes the
conditions under which  an IFA and CPE are required.

            Exhibit 6-8. Exceptions Reporting, Individual Filter Assessment, and
                Comprehensive Performance Evaluation Requirements for the
                     Individual Filter Turbidity Monitoring Alternatives
Activity
Exceptions Reporting
Individual Filter
Assessment
Comprehensive
Performance Evaluation
Alternative
Tl1
>1NTU
>0.5NTU2
> 1 NTU for 3
consecutive months
> 2 NTU for 2
consecutive months
Alternative T21
(preferred)
>1NTU
> 1 NTU for 3
consecutive months
> 2 NTU for 2
consecutive months
Alternative
T31
>0.5NTU(«5%)3
>2NTU
If an exceptions report is required
for 3 consecutive months
> 2 NTU for 2 consecutive
months
 'All standards are based on any two consecutive measurements taken 15 minutes apart, except those noted below.
 2Based on two consecutive measurements taken 15 minutes apart at the end of the first 4 hours of filter operation.
  Based on turbidity levels exceeding 0.5 NTU in at least 5 percent of measurements taken in a month.


Individual filter turbidity measurements are assumed to trigger the equivalent of one monthly
exceptions report to the State at 20 percent of all systems each year for Alternatives Tl and T3,
February 15, 2000
6-13
RIAfor the Proposed LT1FBR

-------
generating about 1,179 exceptions reports per year.8 Based on existing turbidity data, EPA
estimated that systems will exceed 0.5 NTU in 5 percent of their measurements each month with
the same frequency that they exceed 1 NTU in two consecutive measurements. For Alternative
T2, EPA assumed that the equivalent of 18 percent of systems will generate one monthly
exceptions report (i.e., 1,061 reports per year); omitting the 0.5 NTU exceptions report trigger (see
Exhibit 6-8) leads to this slight reduction. Preparation,  submission, and review time is estimated to
take 1 hour per exceptions report for all system size categories and all regulatory alternatives.
Under Tl, EPA assumes that all systems will also require an additional 30 minutes to develop a
filter profile.  This over estimates costs because some systems will not need to develop a profile.
EPA does not have the necessary data, however, to estimate the fraction of systems that will not
require a filter profile.

For all alternatives, EPA assumed that 4 percent of all systems each year will conduct an
IFA.  Consequently, the levels of health protection provided by the monitoring alternatives are
expected to be comparable among one another. At this percentage, approximately 236 IF As will
be conducted each year. The cost per IFA assumes that it takes 10 hours to complete at a labor
rate of $28/hour for systems with 1,000 or fewer served (costing $280 per IFA). IF As for systems
in the two larger size categories (1,001-3,300 and 3,301-9,999) will cost more ($336 and $432,
respectively) because of the more costly mix of technical and managerial labor.

For all alternatives, EPA assumes that each year 2 percent of all systems will require a CPE.
Approximately  118 CPEs are assumed conducted each year.  The cost estimate assumes that it will
require States 60 hours to complete a CPE at systems serving 1,000 or fewer and 120 hours to
complete a CPE at larger systems. The assumed labor rate of $100/hour is the  same rate used in
the IESWTR RIA to approximate a third-party cost including travel expenses.

Exhibit 6-9 summarizes estimated annual costs for water systems and States. System costs for the
preferred alternative, which include filing exceptions reports and conducting IF As, total
approximately $0.12 million. States are expected to incur annual costs of $0.09 million to review
the exceptions reports  and $1.08 million to perform CPEs. Cumulative annual  costs for exceptions
reports, IF As, and CPEs total $1.29 million under alternative T2.

Costs for monitoring alternatives Tl and  T3 are approximately the same; the difference is that T3
does not include the filter profile requirement for systems.  Costs for alternative T2 are lower
because of the reduced number of exceptions reports and because it does not include the filter
profile requirement for water systems.
       8  This does not mean that 1,179 individual systems will submit a report because systems requiring an IFA
or CPE will submit multiple reports.


RIA for the Proposed LT1FBR                      6-14                              February 15, 2000

-------
       Exhibit 6-9. System and State Costs for Exceptions Reports, Individual Filter
                Assessments, and Comprehensive Performance Evaluations
                                  (January 1999 dollars)
Annual Cost ($ millions)
Compliance Annual Alternative
Activities Occurrence Tl
System Costs
Annual Exceptions Reports
Annual IFAs
Total System Cost
State Costs
Annual Exceptions Reports
Annual CPEs
Total State Cost

1,061-1, 179 Reports
236 IFAs


1,061-1, 179 Reports
118 CPEs

Total Annual Cost

$0.07
$0.08
$0.15

$0.10
$1.08
$1.18
$1.33
Alternative T2 Alternative
(preferred) T3

$0.05
$0.08
$0.12

$0.09
$1.08
$1.17
$1.29

$0.05
$0.08
$0.13

$0.10
$1.08
$1.18
$1.31
 Detail may not add to total due to independent rounding. See Appendices G-2a and G-2b as well as G-lc through G-lg for
 detail.

6.3    Disinfection Benchmarking Provision Cost Analysis

To comply with the Stage 1 Disinfectants and Disinfection Byproducts Rule (Stage 1 DBPR) (63
FR 69389, December 16, 1998), some small systems that use disinfection may need to alter
practices to reduce the presence of disinfection byproducts in finished water. Systems disinfect to
reduce the risk of microbial contamination in drinking water. Chemical reactions between the
disinfection products and organic material in source water, however, produce disinfection
byproducts. Chronic exposure to these byproducts over a long period of time has been associated
with health risks such as cancer. Consequently, this provision does not apply to drinking water
systems that are classified as transient noncommunity water systems, because the population using
these systems changes over time. The benchmarking provision of the proposed LT1FBR will
provide information on the current level of Giardia inactivation to ensure that altered disinfection
practices do not increase risks of microbial infection. It is assumed that the 9,450 small surface
water and GWUDI systems that are classified as community or nontransient noncommunity
systems are subject to the disinfection benchmarking provision of the proposed rule.

Section 6.3.1 describes the activities systems will undertake to implement this provision and
estimates the associated costs.  Section 6.3.2 provides the cost analysis for  State activities.

6.3.1  System Costs

Systems will incur startup costs and they will implement this provision in two distinct phases: an
applicability monitoring phase, which determines whether a system needs to develop a benchmark,
and a profile and benchmark development phase. To minimize the potential burden of this
provision on systems, EPA developed a set of alternative regulatory requirements for each phase.
Startup Costs
February 15, 2000
6-15
RIAfor the Proposed LT1FBR

-------
Initial start-up activities—reading and understanding the rule, and mobilizing and planning—and
record keeping are assumed to require 20 hours of managerial or technical labor time per affected
system. Costs will range from $560 to $864 across the system size categories because wage rates
and hour allocations differ.  Exhibit 6-10 summarizes total start-up costs.

            Exhibit 6-10. Disinfection Benchmark Development Start-up Costs
                           by System Size (January 1999 dollars)
System Size Category
(# systems)
• 500(3,737)
501-1,000(1,301)
1,001-3,300(2,553)
3,301-9,999(1,859)
Unit Cost
($)
$560
$560
$672
$864
Total
Annualized Cost (3%)
Annualized Cost (7%)
Total Cost
($ millions)
$2.1
$0.7
$1.7
$1.6
$6.1
$0.4
$0.6
 Detail may not add due to independent rounding. See Appendices G-3a through G-3f for detail.

Applicability Monitoring

Each small system may also be required to obtain water samples to test for total trihalomethanes
(TTHM) and five haloacetic acids (HAAS) concentrations, which will determine whether it must
develop a disinfection profile  and benchmark. EPA evaluated four monitoring alternatives that
differ with respect to data collection and analysis burdens on systems.

       •   Alternative Al: The TTHM and HAAS monitoring requirements for small systems
           are the same as the final IESWTR provisions for large systems, which required all
           systems to obtain  four samples in each of 4 quarters.  A system may request that the
           State waive its applicability monitoring requirement.

       •   Alternative A2: Small systems serving more than 500 are required to obtain one
           sample in each of 4 quarters. Small systems serving 500 and fewer are required to
           obtain one sample during the critical period, which is determined by the State, and
           systems may choose to obtain an optional second sample. A system may request that
           the State waive its applicability monitoring requirement.

       •   Alternative A3: All small systems must sample once during a critical period, which is
           determined by the State. A system may request that the  State waive its applicability
           monitoring requirement.

       •   Alternative A4: Applicability monitoring is optional and systems do not need to
           request a State waiver.  All systems are required to develop a disinfection profile
           beginning January 1, 2003, based on weekly calculation of log inactivation ofGiardia
RIAfor the ProposedLT1FBR
6-16
February 15, 2000

-------
           lamblia, unless a system opts to perform the applicability monitoring during the month
           of warmest water temperature in 2002 (i.e., collect one sample), and demonstrates that
           TTHM and HAAS levels are less than 80 percent of their respective MCLs. This
           alternative can only be paired with the second profile development alternative, which
           is described in the next section. Systems that choose to forego applicability monitoring
           under the LT1FBR may still need to gather TTHM and HAAS samples after January
           1, 2003, under the Stage 1 DBPR.

Exhibit 6-11  summarizes the disinfection byproduct sampling requirements across the alternatives.
EPA assumed that each sample can be used for both a TTHM analysis and an HAAS analysis.
The exhibit summarizes the total number of chemical analyses for each individual alternative (e.g.,
the total number of analyses for Alternative Al is 32: 4 sample locations x 2 analyses x 4
quarters).  EPA assumes that each sample requires 2 hours of operator time to collect and process,
and that total lab fee will be $360 (assuming that it costs $180 to analyze each individual
contaminant). Consequently, the labor burden and laboratory fees both decline as the number of
required samples declines.

   Exhibit 6-11. Summary of Proposed Applicability Monitoring Sampling Alternatives1
System Size
Category
• 500
501-9,999
Alternative
Al
Sample 4 times per
quarter for 4 quarters
(32 analyses)
Sample 4 times per
quarter for 4 quarters
(32 analyses)
Alternative Alternative Alternative A4
A2 A3 (preferred)
Sample once during
critical monitoring
period2'3
(2 analyses)
Sample once per
quarter for 4 quarters
(8 analyses)
Sample once during
critical monitoring
period2
(2 analyses)
Sample once during
critical monitoring
period2
(2 analyses)
Optional sample
during warmest water
temperature month
(2 analyses)
Optional sample
during warmest water
temperature month
(2 analyses)
 'Each sample will be used for a TTHM analysis and an HAAS analysis.
 2The State will determine the critical monitoring period, usually the month of warmest water temperature.
 3Systems may obtain an additional sample.

Exhibit 6-12 shows total costs by system size category and monitoring alternative. System-level
costs for the two smallest size categories are identical, so the exhibit combines those categories.
These costs differ across the categories because the labor rate assumptions differ, as noted in
Section 6.2. Total annualized cost for the preferred alternative, A4, is either $0.03 million or $0.04
million depending on the discount rate assumption. By comparison, annualized cost for Al is $4.3
million or $6.0 million.
February 15, 2000
6-17
RIAfor the Proposed LT1FBR

-------
          Exhibit 6-12. Disinfection Benchmark Applicability Monitoring Costs by
                      Alternative and System Size (January 1999 dollars)

System Size
Category
(# systems)1
• 500
(3,737 / 463)
501-1,000
(1,301 / 128)
1,001-3,300
(2,553/253)
3,301-9,999
(1,859/116)
Total
Annualized
Cost (3%)
Annualized
Cost (7%)
Alternative
Al
Unit
Cost
($)
$6,68
4
$6,68
4
$6,71
2
$7,09
6
Total
Cost
($ million)
$25.0
$8.7
$17.1
$13.2
$64.0

$4.3

$6.0
Alternative
A2
Unit
Cost
($)
$444
$1,69
2
$1,72
0
$1,81
6
Total
Cost2
($ million)
$1.9
$2.2
$4.4
$3.4
$11.9

$0.8

$1.1
Alternative
A3
Unit
Cost
($)
$444
$444
$472
$496
Total
Cost
($ million)
$1.7
$0.6
$1.2
$0.9
$4.4

$0.3

$0.4
Alternative A4
(preferred)
Unit
Cost
($)
$444
$444
$472
$496
Total
Cost
($ millions)
$0.21
$0.06
$0.12
$0.06
$0.44

$0.03

$0.04
 Detail may not add due to independent rounding. See Appendices G-3a through G-3f for detail.
 'The first number of systems are those assumed to conduct applicability monitoring under Al, A2, and A3; the second number is
 those assumed to conduct applicability monitoring under A4.
  The total cost for Alternative A2 assumes that 15 percent of systems in the • 500 size category will collect an optional second
 quarter sample to get an average DBF level.  This assumption is based on the estimate that 29 percent of all  systems will need to
 develop a profile and benchmark and half this amount will collect an optional second sample. Consequently, total cost is slightly
 higher than the total cost for A3 although unit costs are the same.

The preferred alternative, A4, allows systems to forego applicability monitoring and begin
disinfection profile development in 2003. The cost estimates for alternative A4 reported in Exhibit
6-12 assume that almost all systems choose to forego applicability monitoring. This flexibility
combined with the lower cost per system reduces applicability monitoring costs by about 99
percent compared to the Alternative Al, which is the most similar to the IESWTR provision for
large systems.  EPA assumed that the only  systems that would choose conduct applicability
monitoring would be the  960 systems in States where total organic carbon levels tend to be 2 ppm
or lower (Washington,  Oregon, Nevada, Idaho, Alaska, Wisconsin,  and West Virginia). EPA
assumed that these systems could reasonably expect to have TTHM and HAAS levels below the
profiling thresholds and, thereby, would only incur applicability monitoring costs. EPA assumed
that all other systems would choose to forego applicability monitoring rather than potentially incur
both applicability monitoring and profile development costs.  Thus, systems implement either
applicability monitoring or profiling; no system performs both. This alternative is more cost
effective than A3, which  requires the same  number of samples per system, but makes monitoring
mandatory.
RIAfor the ProposedLT1FBR
6-18
February 15, 2000

-------
Profile and Benchmark Development Activities

Systems must develop a disinfection profile and may need to develop a benchmark if they choose
to forego monitoring under alternative A4 or if applicability monitoring data under any
applicability monitoring alternative show that either TTHM or HAAS are equal to or exceed 80
percent of their respective maximum contaminant levels (MCLs):

           TTHM levels are at least 80 percent of the MCL (i.e., 0.064 mg/L)
           HAAS levels are at least 80 percent of the MCL (i.e., 0.048 mg/L).

EPA based its assumption regarding the number of small systems on its earlier assumptions for the
IESWTR and the Stage  1 DBPR RIAs. The final IESWTR RIA determined that 29 percent of all
large systems would need to develop disinfection benchmarks based on data in the 1996 Water
Industry Database (WIDE). This percentage reflects the number of systems with levels that are
equal to or greater than either 0.064 mg/L for TTHM or 0.048 mg/L for HAAS.9 The final Stage
1 DBPR RIA reported that approximately 24 percent of large systems would exceed at least one of
the thresholds, based on another analysis of 1996 WIDE data. Furthermore, the Stage 1 DBPR
RIA compliance forecast assumed that the rate would be applicable to small systems. This
analysis makes a similar assumption, but applies the higher rate,  29 percent, to small systems.
Preliminary data reviewed by EPA suggest this assumption is within an acceptable range of
uncertainty for small systems.  Additional data, however, are still under review.

A disinfection benchmark is based on a disinfection profile, which is a compilation of Giardia
lamblia log inactivation levels (as well as viral inactivation levels for systems using either
chloramines or ozone for primary disinfection). System operators will compute log inactivation
levels based on measurements of operational data (i.e., disinfectant residual concentration at the
first customer and just prior to each additional point of disinfectant addition, contact time during
peak flow conditions, temperature, and pH). The disinfection benchmark is lowest level of
inactivation in the disinfection profile. It will be used by the system in consultation with the State
to evaluate potential changes to disinfection practices that systems may make to comply with the
Stage 1 DBPR.
       9 The 1996 Water Industry Data Base (WIDE) includes annual average TTHM and HAAS figures from 574
plants (comprising 399 systems). Analysis of the 78 systems in the 1996 WIDE for which TTHM and HAAS data
exist shows that 29 percent had TTHM levels greater than 0.064 mg/L and/or HAA levels greater than 0.048 mg/L.


February 15, 2000                                 6-19                     RIA for the Proposed LT1FBR

-------
EPA considered three alternative data collection regimes for profile development, which are
summarized in Exhibit 6-13.10 The alternatives differ with respect to the total number of profile
observations, which depend on the frequency of data collection and the length of the collection
period:
       •    Alternative B1: Daily data collection for 1  year

            Alternative B2: Weekly data collection for 1 year

       •    Alternative B3: Daily data collection for 1 month.

     Exhibit 6-13.  Summary of Alternative System-Level Data Collection Requirements
                                  for the Disinfection Profile
Collection Requirement
Frequency
Duration
Total profile observations
Profile burden (hours)
Benchmark burden (hours)
Alternative Bl
Daily
1 year
365
91
48
Alternative B2
(preferred with A4)
Weekly
1 year
52
13
48
Alternative B3
Daily
1 month
30
8
48
The burden estimate assumed that system operators will use a spreadsheet to calculate inactivation
levels and that they will require 15 minutes per observation to collect, enter, and review data.  If a
benchmark is required because a system plans to alter its disinfection practices, then additional
effort is required to develop the benchmark and prepare a report for the State (40 hours) that
describes the profile and benchmark calculations and how the proposed change may affect
inactivation levels in comparison with the benchmark. Furthermore, the system will require a total
of 8 hours on average to meet with the State. Exhibit 6-13 reports the total burden per  system for
each alternative in hours. Regardless of the differences in system burden, all three alternatives are
expected to achieve comparable levels of health protection.

EPA  assumes that only 29 percent of affected systems calculate benchmarks, regardless of how
many gather profile data because benchmarks need only be calculated when disinfection practices
must  change.  Thus, Exhibit 6-14 reports profile and benchmark costs separately. Exhibit 6-14
also reports supplementary costs for developing viral inactivation profiles and benchmarks, which
will be required of a subset of systems changing disinfection practices because they use
chloromines or ozone. According to CWSS data, between 0.3 and 2.9 percent of small systems
utilize chloromines or  ozone and would require the additional profile and benchmark effort.
       10 The Preamble contains an additional "no action" option that is not included in the cost analysis
because feedback from small system operators indicated that it is beneficial for them to know their system's level
of microbial inactivation. Furthermore, the public health protection principles developed during the regulation
negotiation and Federal Advisory Committee should be applied to small system as well as large systems.
RIAfor the ProposedLT1FBR
6-20
February 15, 2000

-------
Total costs range from $5.4 million (B3) to $13.6 million (Bl).  These cost estimates are
conservative because they do not include any of potential cost savings from waivers.  Systems
serving 500 or fewer people can apply for waivers to allow them to submit observed data to the
State rather than calculate inactivation levels for the profile. Shifting part of the system's labor
burden to the State would decrease the cost estimates because the hourly labor rate assumptions
used for State staff are somewhat lower than the system rates.

Overall, B2 costs about 60 percent more than B3  (the lowest cost alternative) because EPA
assumed that substantially more systems would forego monitoring and develop profiles. Yet, EPA
selected B2 as the preferred alternative because it provides a profile over an entire year rather than
one month for a small incremental increase in system-level cost.
February 15, 2000                                 6-21                     RIA for the Proposed LT1FBR

-------
      Exhibit 6-14.  Summary of System Disinfection Profile and Benchmarking Costs
                     by Alternative and System Size (January 1999 dollars)
System Size Category
(# systems)
Alternative
Bl
Unit Cost
(S)
Total Cost
($ million)
Alternative B2
(preferred with A4)
Unit Cost
($)
Total Cost
($ million)
Alternative
B3
Unit Cost
($)
Total Cost
($ million)
Profile Development
• 500
(1,084/3,274)'
501-1,000
(377/1,173)'
1,001-3,300
(740/2,300)'
3,301-9,999
(539/1,743)'
Subtotal
$2,583
$2,583
$3,463
$4,193
$2.8
$1.0
$2.6
$2.3
$8.6
$392
$392
$541
$645
$1.3
$0.5
$1.2
$1.1
$4.1
$238
$238
$336
$396
$0.3
$0.1
$0.2
$0.2
$0.8
Benchmark Development
• 500(1,084)
501-1,000(377)
1,001-3,300(740)
3,301-9,999 (539)
Subtotal
$1,344
$1,344
$1,680
$2,112
$1.5
$0.5
$1.2
$1.1
$4.3
$1,344
$1,344
$1,680
$2,112
$1.5
$0.5
$1.2
$1.1
$4.3
$1,344
$1,344
$1,680
$2,112
$1.5
$0.5
$1.2
$1.1
$4.3
Supplemental Viral Profile/Benchmark
• 500(11)
501-1,000 (4)
1,001-3,300(74)
3,301-9,999 (39)
Subtotal
Total
Annualized Cost (3%)
Annualized Cost (7%)
$3,675
$3,675
$4,751
$5,865
$0.04
$0.01
$0.35
$0.23
$0.6
$13.6
$0.9
$1.3
$1,484
$1,484
$1,829
$2,317
$0.01
$0.01
$0.14
$0.09
$0.2
$8.7
$0.6
$0.8
$1,330
$1,330
$1,624
$2,068
$0.01
$0.01
$0.12
$0.08
$0.2
$5.4
$0.4
$0.5
 Detail may not add to total due to independent rounding.  See Appendices G-3a through G-3f for detail.
 'in the profile development section, the first number is the number of systems affected by Bl and B3; the second is the number of
 systems affected by B2, assuming it is paired with A4.
RIAfor the ProposedLT1FBR
6-22
February 15, 2000

-------
6.3.2  State Costs

Similar to IESWTR, each State will review disinfection benchmarks and meet with systems to
approve any significant changes in disinfection practice (e.g., move point of disinfection, change
the type of disinfectant, change the disinfection process, or make other changes designated as
significant by the State).  Supporting materials for these consultations must include a description of
the proposed change, the disinfection benchmark, and an analysis of how the proposed change will
alter the effectiveness of disinfection.

State activities considered applicable to the disinfection benchmark process included reading and
understanding the rule changes, mobilization and planning, training of State staff, and providing
training in protocols for systems and consultants.  The cost analysis assumes these start-up
activities require about 144 hours and cost $3,472 per State.  For each of the 9,450 systems
affected by the rule, the State must track compliance (e.g., track whether an applicability
monitoring or profiling notification was submitted) and keep records.  EPA assumes an average
burden of 8 hours or $190 per system for all such activities.  The burden associated with Giardia
lamblia benchmark activities such as reviewing data, approving significant changes in disinfection
practices, and meeting with systems  are assumed to require 32 hours per system and cost $761 per
system. Supplementary viral benchmark data review and determination activities will require 16
additional hours and cost $380 per system.  The supplementary viral benchmark time estimate is
lower because the meeting cost can address both benchmarks.  Although meetings require 4 hours
for systems, EPA assumed a higher burden for States to account for travel time, which is assumed
to average 4 hours per meeting.

Exhibit 6-15 summarizes State costs for reviewing system disinfection benchmarks including start-
up and benchmark review costs per State, and total costs for all  States. As noted above, the State
burden would increase if systems serving fewer than 500 people apply for a waiver that would
shift part of the system's profiling and benchmark burden to the State.

        Exhibit 6-15. State Disinfection Benchmarking Costs (January 1999 dollars)
Compliance Activities
Start-up Cost1
Compliance Tracking/Record Keeping Cost
Giardia Profile and Benchmark Cost
Supplemental Viral Benchmark Cost
Respondents
Affected
56 Entities
9,450 Systems
2,741 Systems
128 Systems
Unit Cost (S)
$3,472
$190
$761
$380
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Total Cost
(S millions)
$0.19
$1.80
$2.09
$0.05
$4.13
$0.28
$0.39
 Detail may not add to total due to independent rounding. See Appendices G-3a, G-3b, and G-3g for detail.
 'The $3,472 unit cost is applicable for Bl and B2. The unit cost for B3 of $3,662 is higher because States need to determine the
 critical monitoring period.  The total startup cost is $0.19 million.
February 15, 2000
6-23
RIAfor the Proposed LT1FBR

-------
6.4    Covered Finished Water Reservoir Provision Cost Analysis

The proposed LT1FBR requires that small public water systems using surface water or GWUDI
cover all new finished water reservoirs, holding tanks, or other storage facilities for finished water.
Finished water reservoirs open to the atmosphere are subject to the same environmental factors as
surface waters, depending on site-specific characteristics and the degree of protection provided.
These include contamination by persons swimming, disposal of garbage into the reservoir,
microbial organisms, small mammals, birds, fish, and the growth of algae. This contamination is
marked by increases in algal cells, bacteria, turbidity, total and fecal coliforms (e.g., E. coli\ and
pathogens.

6.4.1   Unit Cost

The calculations for this rule element use a model finished water reservoir, assuming a 10-foot
depth for systems serving 3,300 or fewer people, and a 20-foot depth for  systems serving 3,301 to
9,999 people. It assumes a reservoir storage volume equal to 1 day of average water flow capacity
for each system size category. Cover costs are approximately $2.00 per square foot for a floating
cover.  O&M costs include visual inspections, cleaning, and repair expenses.  Exhibit 6-16
summarizes the capital and O&M costs by system size.

       Exhibit 6-16.  Unit Cost Assumptions to Cover New Finished Water Reservoirs
System Size
• 100
101-500
501-1,000
1,001-3,300
3,301-9,999
Reservoir Volume (ft3)
936
3,743
10,027
28,476
100,000
Cover Area (ft2)
94
375
1,003
1,424
5,000
Capital Cost (S)
$188
$750
$2,006
$2,848
$10,000
O&M Cost (S)
$3,370
$2,860
$4,423
$4,067
$7,501
6.4.2  Compliance Forecast

The analysis of costs for covering new finished water reservoirs is complicated by the lack of data
regarding the construction of reservoirs over the next 20 years.  The precise number of systems
constructing finished water reservoirs is unknown.  Because the proposed rule requires all systems
constructing finished water reservoirs to cover them, its cost impact is only on those that were not
originally planning to construct covers.  EPA assumes that future construction rates can be
approximated by historical rates and assumes that no small systems would include storage covers
without the proposed rule. Historical construction rates suggest that new reservoirs over the next
20 years will roughly equal 5 percent of the existing number of systems. Exhibit 6-17 summarizes
the number of new storage reservoirs affected by the proposed rule.

6.4.3  System Costs

Exhibit 6-17 summarizes the total cost by system size.  Total annual costs, including annualized
capital costs and 1 year of O&M costs, are expected to be $2.55 million or $2.59 million
RIAfor the ProposedLT1FBR
6-24
February 15, 2000

-------
depending on the discount rate.  This estimate over states costs because it assumes that the new
storage facilities are built in the first year rather than over a 20-year period.

       Exhibit 6-17.  Total Cost Estimates to Cover New Finished Water Reservoirs
                                  (January 1999 dollars)
System Size
• 100
101-500
501-1,000
1001-3,300
3,301-9,999
Total
Number of New
Reservoirs
140
143
70
132
95
580
Total Annualized Cost (3%)
(S millions)
$0.474
$0.416
$0.319
$0.560
$0.776
$2.545
Total Annualized Cost (7%)
(S millions)
$0.475
$0.419
$0.323
$0.570
$0.802
$2.589
 Detail may not add to total due to independent rounding. See Appendix C-9 for detail.

6.5    Recycle Provisions Cost Analysis

Unlike the provisions discussed above, the proposed recycle provisions apply to both large and
small surface water or GWUDI systems. The recycle provisions primarily affect three types of
systems. Exhibit 6-18  summarizes the number of systems potentially affected by system type and
size category.

       •   Systems that do not return recycle prior to the point of primary coagulant
           addition, defined as systems employing rapid granular filtration that currently return
           spent filter backwash, thickener supernatant, or liquids from dewatering processes
           concurrent with or downstream from the point of primary coagulant addition, will need
           to move the reintroduction point unless the State grants a waiver for an alternative
           location.

           Direct recycle systems that employ conventional rapid granular filtration treatment,
           use 20 or fewer filters to meet production requirements during the highest production
           month in the  12-month period prior to the proposed rule's compliance date, and
           recycle spent filter backwash or thickener supernatent to the primary treatment process
           will be required to conduct a recycle self assessment to determine whether it exceeds
           its State approved operating capacity during recycle events and consult with the State
           to determine whether changes to recycle practices are necessary. Alternatively, EPA
           also considered requirements that all direct recycle systems construct a flow
           equalization basin or a sedimentation basin.

           Direct filtration systems that recycle to the primary treatment process will be required
           to report their recycling practices to the State, which will decide if changes to recycle
           practices are necessary.  EPA also evaluated an alternative requirement that these
           systems install a sedimentation basin.
February 15, 2000
6-25
RIAfor the Proposed LT1FBR

-------
      Exhibit 6-18.  Systems Potentially Affected by the Proposed Recycling Provisions
Type of System Affected by a Provision1
Total rapid granular filtration systems that recycle
Systems moving recycle return to point prior to
primary coagulant addition
Direct recycle systems
Direct filtration systems
System Size
Small Systems
<10,000
3,538
569
757
248
Large Systems
• 10,000
1,098
221
342
77
Total
4,636
791
1,099
325
 Detail may not add to total due to independent rounding.
 Note: To be consistent with the baseline numbers in Chapter 4, these system estimates exclude seven individual
 plants that belong to systems serving more than one million people, which have been included in the cost
 analysis as indicated in Appendices C-13, D-l 1, E-l 1, and G^l through G-6. See footnote 1 1 for discussion.
   EPA is considering an alternative under which conventional filtration systems that currently have flow
 equalization basins would be required to provide sedimentation or more advanced treatment. This alternative
 would affect an additional 296 small conventional systems and an additional 141 large conventional systems for
 a total of 437 conventional systems.
EPA considered four regulatory alternatives for the recycle provisions.  The alternatives are
discussed in detail in Chapter 3 (see Exhibit 3-4 for a summary). The alternatives differ with
respect to their affect on the three types of systems described above.

        •    Alternative Rl requires all systems to return spent filter backwash, thickener
            supernatant, and liquids from dewatering to a point prior to primary coagulant
            addition, unless the State grants a waiver. Systems that must change their recycle
            practices are required to submit a plant schematic, which shows the current return
            location(s) and the proposed new return location, to the State. This provision is the
            same for all  of the alternatives.

            Alternative R2 requires direct recycle systems to conduct a self assessment and report
            the results to the State, and it requires direct filtration systems to report their recycle
            practices to the State. In both instances, the State will make determinations regarding
            changes in recycle practices.  The requirement for the return flow location is the same
            asRl.

            Alternative R3 has the same requirements as R2 for direct filtration systems and the
            recycle return location.  It differs from R2 in that all direct recycle systems are required
            to install a flow equalization basin; no self assessment is required, although States must
            still review basin installation plans.

        •    Alternative R4 requires that all systems provide treatment for recycle flows that is
            equivalent to or more advanced than sedimentation. This affects all systems that
RIAfor the ProposedLT1FBR
6-26
February 15, 2000

-------
            practice direct recycle, direct filtration systems that do not have recycle treatment, and
            conventional filtration systems that only provide flow equalization for recycle streams.
            The requirement for recycling prior to the point of primary coagulant addition is the
            same as Rl.

Sections 6.5.1 through 6.5.3 describe the system expenditures to modify practices and labor
burdens for systems and States across the various alternatives for the new return location, direct
recycle, and direct filtration provisions, respectively. Section 6.5.4 summarizes costs for the four
regulatory alternatives.

Cost estimates for all three provisions include capital and O&M costs. As described in The Cost
and Technology Document for the Long Term 1 Enhanced Surface Water  Treatment and Filter
Backwash Rule (U.S. EPA, 1999b), unit capital and O&M costs were estimated using engineering
models and system-level flow rates. Using these unit costs to develop cost estimates for large
systems introduced some uncertainty because total system flows at large systems—especially
systems serving more than 100,000—may be treated by two or more plants, some of which may
not recycle flow to the treatment process.  Consequently, EPA potentially overestimated
compliance costs for large systems that do not need to change recycle practices at all of their
plants.  Conversely,  EPA may have underestimated compliance costs for large systems that need to
change recycle practices for all of their plants because installing new equipment at two or more
plants with smaller flow rates may cost more than estimated unit cost of installing equipment at a
single large plant that handles the same flow rate. Although these biases will tend to offset one
another, EPA cannot determine whether total costs are more likely to be over or under estimated
because it does not have details about the plant  configurations of all large plants that recycle.11

6.5.1    Recycle to New Return Location

Under Alternative Rl, systems that do not return select recycle flows prior to the point of primary
coagulant addition will be  required to move  the return point to this location. As noted in Exhibit
6-18, an estimated 791 systems will need to move their recycling return point. EPA based this
estimate on information provided by a sample of large and small systems that responded to a 1998
AWWA survey on recycle practices (AWWA,  1998) and plant schematics gathered under the
Information Collection Rule (61 FR 24354,  May 14, 1996).  The data were not sufficient to allow
EPA to distinguish practices for system size categories and system types.  EPA was able to obtain,
however, a single percentage (25 percent) for all direct recycle system size categories, and two
percentage estimates for the systems with recycle treatment (15 percent for small systems and  20
        11 The exception to this approach is EPA's analysis of systems serving more than 1 million. EPA used the
schematic of ICR systems and SDWIS to determine whether these systems would be affected by the recycle
provisions. First, EPA identified 17 systems in SDWIS that serve populations greater than 1 million.  Then EPA
identified schematics of the individual plants within these systems. Of the 24 plants identified, only seven would
be affected by the rule. Two plants (both serving 10,000 to 50,000) would have to move their recycle return
location, one plant (serving 10,000 to 50,000) would be required to perform a self assessment, and four direct
filtration plants (two serving 50,000 to 100,000 and two serving more than 100,000) would be required to submit
data on their recycle practice to the State. EPA included these individual plants in the cost analysis, with the
exception of one direct filtration plant that alone served more than 1 million.


February 15, 2000                                  6-27                      RIA for the ProposedLT1FBR

-------
percent for large systems). Further analysis of the data for direct recycle systems showed that large
and small systems recycle prior to the point of primary coagulant addition at roughly the same
frequency.

EPA excluded direct filtration plants from this cost analysis because available data on the return
location of these plants suggest that almost all of them already return recycle prior to the point
of primary coagulant addition.  Only one plant out of 37 direct filtration plants in EPA's database
returned recycle after primary coagulant addition. Many direct filtration plants may be configured
in a manner that makes returning recycle prior to primary coagulant addition the logical location.

The proposed rule requires each of these systems to prepare documentation that describes its plans
to move recycle return to a new location.  This documentation will be submitted to the State for
review. EPA estimated system and State burdens for these activities. Subsequent to State
approval of plans, systems will need to install additional pipe and may need to install additional
pump capacity to recycle to the new location. Additional energy will be required to pump water
the extra distance, which raises annual operating costs. Thus, EPA also estimated capital and
O&M expenditures associated with these changes in recycling practices. The aggregate burden,
capital, and O&M costs, all of which are described below, do not differ across the alternatives
because this provision remains the same.

Some systems may require a recycle return location other than the one specified in the proposed
rule to maintain optimal performance.  The proposed rule allows these plants to apply for a waiver
to recycle to an alternative location. States will review the waivers and decide whether to approve
them. EPA did not have information sufficient to estimate how many of the approximately 791
systems that return recycle concurrent with or below the point of primary coagulant addition will
apply for and qualify for a waiver. Consequently, EPA's cost analysis assumed that all of these
systems will develop and implement plans to move their point of recycle return.  This assumption
overstates costs for these systems because the cost of applying for a waiver will be less than the
combined cost of planning and moving the recycle influent.

System Start-up and Reporting Costs

EPA assumed that systems will incur start-up and reporting costs for the following activities: read
and understand the rule, mobilize and plan, prepare and submit plan to State, meet with State, and
maintain records. The system-level burden across these activities is 50  hours. Exhibit 6-19
summarizes total costs by system size category.  Total cost is $1.5 million, and annualized cost is
approximately $0.10 million to $0.14 million depending on the discount rate assumption.
RIA for the ProposedLT1FBR                       6-28                               February 15, 2000

-------
     Exhibit 6-19.  Total Annualized System Costs for Reporting Proposed New Return
                       Location by System Size (January 1999 dollars)
System Size (# systems)
• 1,000(267)
1,001-3,300(160)
3,301-9,999 (143)
• 10,000(221)'
Alternatives Rl, R2, R3, and R4
Unit Cost (S)
$1,400
$1,764
$2,208
$2,224
Total
Annualized Cost (3%)
Annualized Cost (7%)
Total Cost (S millions)
$0.4
$0.3
$0.3
$0.5
$1.5
$0.10
$0.14
 Detail may not add to total due to independent rounding. See Appendices G^a through G^lc for detail.
 'Total cost includes two plants for systems serving more than 1 million in addition to the 221 systems serving 10,000 to 1
 million.

State Start-up and Review Costs

State start-up activities include reading and understanding the rule, mobilizing and planning, and
training State staff. EPA estimated that these activities will require an average of 51 hours per
State.  States will also need to review the plans submitted by systems, and meet with systems to
discuss proposed recycling changes. These activities will require about 12 hours  per system. Total
costs for all State activities, which are summarized in Exhibit 6-20, are $0.29 million; annualized
costs are $0.02 million to $0.03 million depending on the discount rate assumption.
 Exhibit 6-20. State Cost Estimate to Review and Approve Plans to Move Recycle Return
                               Location (January 1999 dollars)
Compliance Activities
State Start-up Cost
State Plan Review Cost'
Respondents
Affected
56 Entities
791 Systems
Unit Cost ($)
$1,187
$278
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Total Cost
($ millions)
$0.07
$0.22
$0.29
$0.02
$0.03
 Detail may differ from total due to independent rounding.  See Appendices G^a, G^lb, and G^d for detail.
 'Total cost includes 791 systems serving 10,000 to 1 million and two plants that belong to systems serving more than 1 million.

Recycle to New Return Location Capital Costs

Appendices C-l 1 and C-12 summarize the capital and O&M costs per system for conventional
filtration systems that need to redirect their recycle flows prior to the point of primary coagulant
addition.  EPA (1999b) discusses how these costs were derived.
To obtain total annualized costs for this provision, unit costs were first multiplied by the 791
systems EPA assumed would move their recycle return location. Capital costs were  annualized
February 15, 2000
6-29
RIAfor the Proposed LT1FBR

-------
over a 20-year period assuming either a 3 percent or a 7 percent discount rate.  EPA added 1 year
of O&M expenditures to annualized capital costs to obtain total annualized costs.  Appendices D
and E provide detail cost estimates by system size category and Exhibit 6-21 summarizes costs.
The total cost of this provision is $13.8 million or $16.7 million, depending on the discount rate
assumption.  The cost to move recycle prior to primary coagulant addition is the same for all  four
alternatives.

   Exhibit 6-21.  Total Annualized Costs for Recycling to Return Location by System Size
                                    (January 1999 dollars)
System Size (# systems)
• 1,000(267)
1,001-3,300 (160)
3,301-9,999 (143)
• 10,000(221)'
Total
3% Discount Rate
(S millions)
$0.9
$0.7
$1.0
$11.2
$13.8
7% Discount Rate
(S millions)
$1.0
$0.8
$1.1
$13.8
$16.7
 Detail may differ from total due to independent rounding. See Appendices D-12 and E-12 for detail.
 'Total cost includes expected modification costs (i.e., probability of modification multiplied by unit cost) for the two plants that
 belong to systems serving 1 million or more.

6.5.2  Direct Recycle Provision Costs

EPA considered three alternative approaches to address the risks posed by direct recycle practices,
and costs for each provision include a burden component as well as capital and O&M
expenditures.  Alternative R2 requires systems to conduct a single one-month hydraulic self
assessment and report the findings to the State. Subsequent requirements to modify recycle
practices are at the discretion of the State. EPA estimate a cost range for this alternative to
incorporate uncertainty regarding the indirect costs of State determinations.

Alternative R3  does not require a self assessment, although system and State start-up costs are still
applicable because it requires that all direct recycle systems provide flow equalization for their
recycle flows.  Consequently, the cost analysis for that alternative includes capital costs and O&M
costs associated with this change in recycle practice.  Alternative R4 also has capital and O&M
cost components for sedimentation basins as well as system and State start-up costs, although no
self assessment is required.  This section describes the start-up costs, self assessment costs, and
treatment costs for all three alternatives.
System Start-up Costs

Under all three alternatives, systems will incur start-up costs.  System start-up costs include reading
and understanding the rule, mobilization and planning, and record keeping. EPA estimated these
activities will require approximately 44 hours for systems serving 1,000 and fewer and 46 hours
for larger systems.  The cost per system ranges from approximately $1,200 to $2,000 depending on
system size.
RIAfor the ProposedLT1FBR
6-30
February 15, 2000

-------
System Reporting and Consultation Costs

Under R2, a system will also need to perform a recycle self assessment for Alternative R2 if it
satisfies all of the four following criteria.

       •   The system uses surface water or GWUDI as a source and employ conventional rapid
           granular filtration treatment

       •   The system employs 20 or fewer filters to meet production requirements during the
           highest production month in the 12-month period prior to LTlFBR's compliance date

       •   The system recycles spent filter backwash or thickener supernatant directly to the
           treatment process (i.e., recycle flow is returned within the treatment process of a PWS
           without first passing the recycle flow through a treatment process designed to remove
           solids, a raw water storage reservoir, or some other structure with a volume equal to or
           greater than the volume of spent filter backwash water produced by one filter
           backwash event).

The proposed rule  requires that each affected system identify the month with the highest water
production in the calendar year preceding the proposed rule's effective date.  During the 12-month
period after the effective date, a system must monitor one recycle event per day for that month and
estimate the combined raw water influent and recycle flow rate. It must prepare and submit a self
assessment report to the State that provides all of the flow rate monitoring data along with other
descriptions of filter operation and recycling practices.  Prior to conducting the self assessment,
each system must submit a monitoring plan to the State that describes how it will conduct the self
assessment.

EPA estimated that the monitoring plan and self assessment activities will require 46 to 54 hours,
which includes 45  minutes per day for monitoring and flow rate calculation and 16 hours to
prepare the self assessment report for the State. Any systems needing to modify recycle practices
will also need to consult with the State to review the modifications.  EPA assumed this will require
a total of 8 hours.  Costs per system range from $1,500 to $2,800 and vary with system size
because the labor rate assumptions and labor mix between operator and manager as well as the
burden assumptions differ by  size.

Exhibit 6-22 summarizes total direct recycle  provision cost estimates by system size category,
combining the three smallest categories in one entry and all of the large systems in another. For
Alternatives R3 and R4, EPA assumed that all direct recycle systems incur consultation costs in
addition to start-up costs because these alternatives require significant changes in recycle practices.
Consultation costs  are for meetings with the State to review plans to install either a flow
equalization basin  or a sedimentation basin.

Exhibit 6-22 also reports total cost by system size category.  Total cost ranges from $2.2 million
(R3) to $3.9 million (R2).  Annualized cost for the preferred alternative is $0.26 to $0.36 million
depending on the discount rate assumption.
February 15, 2000                                 6-31                     RIA for the Proposed LT1FBR

-------
   Exhibit 6-22. Total System Start-up and Self Assessment Costs by System Size for the
                       Direct Recycle Provision (January 1999 dollars)
System Size (# systems) 1
• 1,000(355/494)
1,001-3,300 (212/296)
3,301-9,999 (190/264)
• 10,000 (342/4S3)2
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Alternative R2
(preferred)
(S millions)
$0.9
$0.7
$0.8
$1.5
$3.9
$0.26
$0.36
Alternative
R3
($ millions)
$0.5
$0.4
$0.5
$0.8
$2.2
$0.15
$0.21
Alternative
R4
($ millions)
$0.7
$0.6
$0.6
$1.2
$3.1
$0.21
$0.30
 Detail may differ from total due to independent rounding. See Appendices G-5a through G-5c for detail.
 1 Total cost for R2 and R3 is based on the first system estimate, which is the number of direct recycle systems. Total cost for
 R4 is based on the second system estimate, which includes all direct recycle systems and all other conventional filtration
 systems that do not already have a sedimentation basin for their recycle stream.
 2 The cost estimates include one plant belonging to a system that serves more than 1 million in addition to the 342 or 483
 systems serving 10,000 to 1 million.

State Start-up and Review Costs

State activities under this provision include start-up activities such as reading and understanding
the rule,  mobilizing and planning and staff training,  and activities that depend on the number
of systems such as reviewing the monitoring plans and self assessments,  determining which
systems  need to change their recycling practices, reviewing changes to recycle practices, and
record keeping. EPA assumed that start-up activities will require 106 hours per State and that
reviewing the monitoring plans and self assessments and making determinations under R2 will
require 31  hours per system on average. Additional activities such as meeting with systems to
discuss changes to recycle practices and keeping records will require  12 hours per system, and
follow-up inspections will require 8 hours per system.  Under alternatives R3 and R4,  all these
burdens  will accrue to States except the 31 hours to  review the monitoring plan  and the self
assessment and make determinations.

Exhibit 6-23  summarizes State costs across the  alternatives.  Total cost ranges from $0.50 million
to $0.77  million.  The annualized value for the preferred alternative is $0.05 million or $0.07
million depending on the discount rate assumption.
RIAfor the ProposedLT1FBR
6-32
February 15, 2000

-------
   Exhibit 6-23. Total State Start-up and Review Costs for the Direct Recycle Provision
                                  (January 1999 dollars)
Compliance Activities
Start-up Costs
Review and Follow-up Costs
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Alternative R2
(preferred)
(S millions)
$0.14
$0.63
$0.77
$0.05
$0.07
Alternative
R3
($ millions)
$0.14
$0.36
$0.50
$0.03
$0.05
Alternative
R4
($ millions)
$0.14
$0.50
$0.64
$0.04
$0.06
 Detail may differ from total due to independent rounding. See Appendices G-5a, G-5b, and G-5d for detail.

Direct Recycle Capital and O&M Costs

Under Alterative R2, States will determine which systems need to change recycle practices under
this provision. For this alternative, EPA estimated the potential indirect costs of the proposed rule
with respect to these follow-on investments.  To develop a compliance forecast, EPA first
estimated how many systems are likely to exceed capacity during recycle events based on the
results of the AWWA Survey (AWWA, 1998). Then, EPA determined the types of process
changes systems might implement to ensure they remain below State approved operating capacity
during recycle events, and estimated how many of the systems exceeding capacity would
implement each one.  These compliance forecast results  were multiplied by annualized per system
capital costs and O&M costs to obtain total annual costs.

This method may over estimate costs for R2 because State determinations may lead to fewer
changes in recycle practices than EPA estimated. Consequently, EPA developed a cost range
to account for uncertainty regarding State determinations.  The high cost for the range is based on
the method described above. EPA has no additional information, however, to develop a plausible
estimate for minimum indirect costs. Consequently, the  lower cost estimate is a bounding estimate
based the assumption that State determinations do not require any systems to alter their recycle
practices.

For Alternative R2, EPA identified eight modifications that systems could implement. Appendix
C lists these modifications and shows the unit capital and O&M costs by system size. The Cost
and Technology Document for the Long Term 1 Enhanced Surface Water Treatment and Filter
Backwash Rule (U.S. EPA,  1999b) describes how EPA derived the unit costs using engineering
models, existing cost and technology documents, and best engineering judgment.

Appendix C also reports the number of systems that EPA assumes will implement each
modification under this regulatory alternative. Using responses to the AWWA (1998) survey of
recycling practices, EPA determined that 359 systems potentially exceed their design capacity
during recycle events. To make this determination, EPA calculated the sum of an instantaneous
backwash flow rate and an instantaneous peak system flow rate and compared this sum to the
design flow rates reported in the AWWA survey.
February 15, 2000
6-33
RIAfor the Proposed LT1FBR

-------
Under Alternative R3, all direct recycle systems would be required to provide flow equalization
treatment for recycle flows. This alternative has higher costs than R2 because it requires all of the
1,099 direct recycle systems to make modifications.  EPA assumed that all systems would install a
flow equalization basin, which may over estimate costs because some systems may choose a lower
cost option such as discharging recycle flows.

Alternative R4 requires that all systems that recycle provide treatment for their recycle stream
equivalent to or more advanced than sedimentation.  EPA assumed that all of the 1,099 direct
recycle systems would install sedimentation basins under this alternative.  Furthermore, the
provision would affect approximately 437 conventional filtration plants that currently have flow
equalization basins. The analysis assumes that these conventional  systems would also install
sedimentation basins.

Exhibit 6-24 summarizes total annualized costs by system size for each alternative.  Annualized
costs include annualized capital costs (assuming a 20-year period and either a 3 or 7 percent
discount rate) and one year of O&M expenditures.  The treatment costs for Alternative R4 are
substantially higher than costs for R2 or R3.

  Exhibit 6-24.  Total Annualized Costs  to Modify Recycling Practices for the Direct Recycle
         Provision by System Size and Regulatory Alternative (January 1999 dollars)
System Size
(Ssystems)1
• 1,000
(142/355/494)
1,001-3,300
(85/212/296)
3,301-9,999
(76/190/264)
• 10,0003
(57/342/483)
Total
Alternative R2
(preferred)
(S millions)2
3%
$0-$0.6
SO-S0.9
$0-$1.4
$0-$1.7
SO-S4.6
7%
$0-$0.6
$0-$1.0
SO-S1.5
$0-$2.3
SO-S5.4
Alternative R3
(S millions)
3%
$3.0
$4.2
$4.6
$17.5
$29.3
7%
$3.3
$5.3
$5.9
$23.3
$37.8
Alternative R4
($ millions)
3%
$10.4
$7.3
$9.1
$75.6
$102.4
7%
$13.6
$9.5
$11.7
$98.2
$132.9
 Detail may differ from totals due to independent rounding. See Appendices D-12 and E-12 for detail.
 'Costs for Alternative R2 are based on the first system number, which is direct recycle systems currently exceeding capacity; costs
 for R3 are based on the second system number, which is all direct recycle systems; and costs for R4 are based on the third system
 number, which is all direct recycle systems plus other conventional filtration systems without sedimentation treatment for recycle
 streams.
 The cost range for R2 assumes a lower bound of no costs for recycle practice modifications and an upper bound based on the
 modification costs of EPA's assessment of the number of systems that exceed State approved capacity.
 The costs include expected costs for one  plant that belongs to a system serving more than 1 million in addition to the estimates of
 affected systems serving 10,000 to 1 million.
RIAfor the ProposedLT1FBR
6-34
February 15, 2000

-------
6.5.3  Direct Filtration Provision Costs

EPA estimated that direct filtration plants account for approximately 7 percent of all plants that use
conventional or direct filtration. Because these plants do not have sedimentation basins in their
main treatment train, recycling can lead to higher concentrations of Cryptosporidium oocysts in the
system compared to conventional filtration systems, unless recycle streams are treated to remove
oocysts.  Based on the AWWA Survey (1998) and the schematics from ICR systems, EPA
estimated that 7 percent of all direct filtration systems do not currently provide treatment for their
recycle streams.  This equals approximately 23 systems across all size categories.

Alternatives R2 and R3 require that all direct filtration systems report their recycling practices to
the State, which will determine whether changes in those practices are necessary.  EPA also
evaluated a provision, Alternative R4, that requires these systems to install sedimentation basins or
more advanced treatment if they do not already provide treatment for recycle streams.

The cost analysis includes system and State reporting costs and capital and O&M costs to modify
recycle practices. System modification costs for R2 and R3 will ultimately depend on the States'
determinations, but EPA developed a cost range using a method similar to the one described above
for the direct recycle provision.

System Start-up and Reporting Costs

EPA estimates that it will require about 23 hours per system to complete start-up activities such as
reading and understanding the rule, and mobilization and planning. Small and large systems will
require another 6 or 12 hours, respectively, to compile the information for the report, which
includes:

           Whether recycle flow treatment or equalization is in place

       •    The type of treatment provided for the recycle flow

           If equalization, sedimentation, or some type of clarification process is used, the
           physical dimensions of the unit (i.e., sufficient for calculating its volume) and the type,
           typical dose, and frequency at which treatment chemicals are used.

       •    The minimum and maximum hydraulic loading the unit experiences

           The maximum backwash rate, duration, typical filter run  length, and the number of
           filters at the plant.

Record keeping will require an additional 4 hours per system.  Finally, systems that are  required to
modify recycling practices will spend approximately 8 hours meeting with the State to discuss the
modifications.

For R4, EPA assumes that all direct filtration systems will spend 8 hours reading and
understanding the rule. Only those systems that do not currently provide treatment for recycle

February 15, 2000                                 6-35                     RIA for the Proposed LT1FBR

-------
streams will also incur costs for mobilizing and planning (15 hours), meeting with the State (8
hours) to review plans to install sedimentation basins or provide more advanced treatment, and
record keeping (4 hours).

Exhibit 6-25 reports total costs and annualized costs for system start-up and reporting activities by
system size category.  Total costs range from $0.12 million to $0.42 million across the alternatives;
costs for Alternatives R2 and R3 are identical because the reporting requirements are the same.
Annualized costs for the preferred alternative are $0.03 million to $0.04 million depending on the
discount rate assumption.

      Exhibit 6-25. Total  System  Start-up and Reporting Costs for the Direct Filtration
                       Provision by System Size (January 1999 dollars)
System Size (# systems)1
• 1,000(116/8)
1,001-3,300 (70/5)
3,301-9,999 (62/4)
• 10,0002 (77/5)
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Alternative R2
(preferred)
($ millions)
$0.11
$0.08
$0.09
$0.14
$0.42
$0.03
$0.04
Alternative R3
($ millions)
$0.11
$0.08
$0.09
$0.14
$0.42
$0.03
$0.04
Alternative R4
($ millions)
$0.03
$0.02
$0.03
$0.03
$0.12
$0.01
$0.01
 Detail may differ from totals due to independent rounding. See Appendices G-6a through G-6c for detail.
 'The first system numbers indicate systems that incur all start-up and reporting costs under R2 and R3, and start-up costs under
 R4. The second number indicates systems that incur State consultation costs to discuss changes to recycling practices under all
 three alternatives.
 2Costs include three plants that belong to systems serving more than 1 million in addition to the estimated number of systems
 serving  10,000 to 1 million.

State Start-up and Review Cost

Under Alternatives R2 and R3, States should use the reports to determine which plants need to
change  their recycle practice to provide additional public health protection.  State start-up activities
include 51 hours for reading and understanding the rule, mobilization and planning, and training.
For these two alternatives, EPA assumed that States will spend 6 hours reviewing a system's
report, 20 hours making a determination for each system, and 4 hours for record keeping. The
State may waive the reporting requirement at the plant operator's request if the State already has
sufficient data to determine whether a plant has recycle treatment in place, and has information to
make an assessment of treatment provided.  EPA did not estimate how many systems this waiver
provision might affect so the cost estimate may overstate costs. EPA's analysis assumes that State
follow-up activities such  as consultations and inspections will require 16 hours for each system
making modifications to its recycle practices. Total costs for all States range from $0.10 million to
$0.16 million (Exhibit 6-26) and the annualized costs for the preferred alternative are $0.011
million or $0.015 million depending  on the discount rate assumption.
RIAfor the ProposedLT1FBR
6-36
February 15, 2000

-------
   Exhibit 6-26. Total State Start-up and Review Costs for the Direct Filtration Provision
                                   (January 1999 dollars)
Compliance Activities
State Start-up Costs
State Review and Follow-up Costs
Total Cost
Annualized Cost (3%)
Annualized Cost (7%)
Alternative R2
(preferred)
(S millions)
$0.07
$0.09
$0.16
$0.011
$0.015
Alternative R3
($ millions)
$0.07
$0.09
$0.16
$0.011
$0.015
Alternative R4
($ millions)
$0.07
$0.04
$0.10
$0.007
$0.010
 Detail may differ from totals due to independent rounding. See Appendices G-6a through G-6c for detail.
Direct Filtration Capital and O&M Costs

For Alternatives R2 and R3, States will determine whether modifications to recycle practices are
necessary.  Consequently, the potential capital and O&M costs for these alternatives are uncertain.
EPA estimated a cost range to reflect the degree of uncertainty. For the high cost, EPA assumed
that States would require modifications for the 23 direct filtration systems that EPA estimated do
not provide recycle treatment.  EPA identified four modifications that direct filtration systems
might use to treat or discharge recycle flows.  Appendix C summarizes the unit costs for each
option by system size category. The low cost assumes no modifications are required by States.
For Alternative R4, EPA assumed that 23 systems would install a sedimentation basin.

For each system size category, unit costs were multiplied by the number of affected systems to
obtain total costs for that category. Amortized capital costs (assuming a 20-year period and either
3 percent or 7 percent discount rate) and annual O&M costs were summed across the system size
categories to obtain total annual costs. Exhibit 6-27 summarizes costs by system size and discount
rate and includes low and high cost ranges for Alternatives R2 and R3.  Annualized high costs
range from $1.56 million to $1.65 million across the alternatives, assuming a  7 percent discount
rate.
February 15, 2000
6-37
RIAfor the Proposed LT1FBR

-------
   Exhibit 6-27. Annualized Costs for Altering Recycle Practices for the Direct Filtration
                       Provision by System Size (January 1999 dollars)
System Size
(# systems)
• 1,000(8)
1,001-3,300(5)
3,301-9,999 (4)
• 10,000(5)'
Total
Alternative R2 (preferred)
($ millions)
3%
$0-$0.04
S0-$0.05
S0-$0.10
$0-$1.32
SO-S1.51
7%
$0-$0.05
$0-$0.06
$0-$0.11
$0-$1.44
SO-S1.65
Alterative R3
(S millions)
3%
$0-$0.04
$0-$0.05
$0-$0.10
$0-$1.32
SO-S1.51
7%
$0-$0.05
SO-S0.06
so-so. 11
SO-S1.44
SO-S1.65
Alternative R4
($ millions)
3%
$0.17
$0.12
$0.15
$0.76
$1.20
7%
$0.22
$0.16
$0.19
$0.99
$1.56
 Detail may not add to total due to independent rounding.
 'Costs include expected costs (i.e., unit costs multiplied by the probability of each system incurring costs) for three plants that
 belong to systems serving more than 1 million in addition to the estimated number of systems serving 10,000 to 1 million.

6.5.4   Summary of Costs by Regulatory Alternative

Exhibit 6-28 summarizes the results of the cost analyses described in the previous sections by
provision and regulatory  alternative. Annualized costs for the preferred alternative, R2, are $14.3
million or $24.5 million depending on the discount rate assumption.  The alternative with the
lowest cost is Rl, which  does not address direct recycle or direct filtrationsystems. Costs for
Alternative R2 are the lowest of the three alternatives that address direct recycle and direct
filtration systems because it gives systems the greatest flexibility regarding recycle practice
changes.

       Exhibit 6-28.  Total Annual System  and State Costs by Recycling Provision and
                         Regulatory Alternative (January 1999 dollars)
Provision
Alternative Rl
($ millions)
Alternative R2
(preferred)
($ millions)
Alternative R3 Alternative R4
($ millions) ($ millions)
3% Discount Rate
Recycle Location
Direct Recycle
Direct Filtration
Total
$13.9
—
—
$13.9
$13.9
$0.3-$4.9
$0.04-$1.5
$14.3-$20.4
$13.9
$29.5
$0.04-$1.5
$43.5-$45.0
$13.9
$102.6
$1.2
$117.8
7% Discount Rate
Recycle Location
Direct Recycle
Direct Filtration
Total
$16.9
—
—
$16.9
$16.9
$0.4-$5.9
$0.1-$1.7
$17.4-$24.5
$16.9
$38.1
$0.1-$1.7
$55.0-$56.7
$16.9
$133.3
$1.6
$151.8
 Detail may not add to total due to independent rounding.
RIAfor the ProposedLT1FBR
6-38
February 15, 2000

-------
6.6    Summary of Costs

National costs for the proposed LT1FBR are the sum of costs across the compliance forecasts.
The turbidity treatment and recycle modifications selected for the forecasts omit potential double-
counting of costs across the provisions. Exhibit 6-29 summarizes the estimate of total annual
national costs for the preferred LT1FBR alternatives:

       •   Turbidity Monitoring: Alternative T2
       •   Disinfection Benchmarking Applicability Monitoring: Alternative A4
       •   Disinfection Profiling and Benchmarking: Alternative B2
       •   Recycle: Alternative R2.
          Exhibit 6-29. Total Annual Costs for Two Combinations of Alternatives
                                   (January 1999 dollars)
Compliance Activity
Preferred Alternatives ($ millions)
3%
7%
IESWTR Alternatives ($ millions)
3%
7%
System Costs
Turbidity Treatment
Turbidity Monitoring
Turbidity Exceptions
Disinfection
Benchmarking
Covered Finished Storage
Recycle Return Location
Direct Recycle1
Direct Filtration1
Total System Costs
$47.42
$9.71
$0.12
$1.03
$2.55
$13.91
$4.86
$1.54
$81.14
$52.23
$10.06
$0.12
$1.44
$2.59
$16.88
$5.80
$1.69
$90.82
$47.42
$62.98
$0.15
$5.63
$2.55
$13.91
$4.86
$1.54
$139.04
$52.23
$63.34
$0.15
$7.90
$2.59
$16.88
$5.80
$1.69
$150.59
State Costs
Turbidity Monitoring
Turbidity Exceptions
Disinfection
Benchmarking
Recycle Return Location
Direct Recycle
Direct Filtration
Total State Costs
Total Costs
$4.97
$1.17
$0.28
$0.02
$0.05
$0.01
$6.50
$87.64
$4.98
$1.17
$0.39
$0.03
$0.07
$0.01
$6.65
$97.48
$4.97
$1.18
$0.28
$0.02
$0.05
$0.01
$6.51
$145.55
$4.98
$1.18
$0.39
$0.03
$0.07
$0.01
$6.66
$157.25
 Detail may not add to total due to rounding.
        high cost estimate from Alternative R2 is reported.
February 15, 2000
6-39
RIAfor the Proposed LT1FBR

-------
The exhibit compares these costs with the costs that EPA estimates small systems would have
incurred if the turbidity, disinfection benchmarking, and covered finished water provisions were to
reflect the provisions promulgated for the IESWTR:

       •   Turbidity Monitoring: Alternative Tl
       •   Disinfection Benchmarking Applicability Monitoring: Alternative Al
       •   Disinfection Profiling and Benchmarking: Alternative B1
       •   Recycle (not a component of IESWTR): Alternative R2.

For both the preferred alternatives and the IESWTR alternatives, annual costs include annualized
capital and start-up costs as well as annual O&M and labor costs.

On an annual basis, the cost of the preferred alternatives is $87.64 million to $97.48 million,
depending on the discount rate assumption.  This combination of alternatives, designed to
minimize the impact of the proposed rule on small systems, represents a cost savings of about 38 to
40 percent, depending on the discount rate, compared to the estimated cost for alternatives that are
similar to the IESWTR provisions. Excluding the recycle costs, the cost savings for the provisions
included in both rules is approximately 45 percent.

The turbidity treatment costs account for approximately 54 percent of total costs under the
preferred alternatives, and aggregate system and State turbidity monitoring costs of $16.0 million
to $16.3 million represent approximately 17 to 18 percent of total costs.  In contrast, under the
alternatives that closely match the IESWTR provisions, aggregate turbidity monitoring costs are
$69.3 million to $69.6 million, which represents approximately 44 to 48 percent of total costs.
This comparison illustrates the relative importance of the turbidity monitoring cost savings under
the preferred alternative in terms of the reducing the overall cost of the proposed rule.

Community water systems account for approximately 88 percent of total system costs.  Transient
noncommunity systems, which represent approximately 17 percent of affected systems, incur only
8 percent of costs, largely because they are not affected by the disinfection benchmark provision.
Nontransient, noncommunity systems account for 4 percent of system costs.

Exhibit 6-30 shows costs for the preferred combination of alternatives broken down by system
size.  The  five small system size categories are shown separately; large systems are aggregated into
a single column. The three smallest size categories incur approximately 27 percent of the total
costs of the proposed rule even though they account for more than 60 percent of the affected
systems. In comparison, large systems, which are only affected by the recycling provisions and
account for only 5 percent of total affected systems, incur approximately 18 percent of total costs.
RIA for the ProposedLT1FBR                       6-40                               February 15, 2000

-------
      Exhibit 6-30.  Summary of Total Annual Costs for the Preferred Alternatives
                             System Size (January 1999 dollars)
                                     by
Compliance Activity
Total Annual Costs by System Size Category ($ millions)
•100
101-500
501-1,000
1,001-3,30
0
3,301-9,999
• 10,000
System Costs
Turbidity Treatment
Turbidity Monitoring
Turbidity Exceptions
Disinfection
Benchmarking
Covered Finished Storage
Recycle Return Location
Direct Recycle
Direct Filtration
Total System Costs
S4.50
$1.05
S0.01
$0.24
$0.47
$0.29
$0.17
$0.02
$6.74
$5.97
$1.40
$0.02
$0.24
$0.42
$0.45
$0.27
$0.02
$8.80
$5.90
$1.01
$0.01
$0.17
$0.32
$0.34
$0.28
$0.02
$8.05
$15.98
$2.96
$0.04
$0.42
$0.57
$0.83
$1.11
$0.06
$21.97
$19.89
$3.64
$0.04
$0.38
$0.80
$1.11
$1.55
$0.12
$27.52
—
—
—
—
—
$13.87
$2.43
$1.45
$17.76
State Costs
Turbidity Monitoring
Turbidity Exceptions
Disinfection
Benchmarking
Recycle Return Location
Direct Recycle
Direct Filtration
Total State Costs
Total Costs
Share of Total Costs
$0.71
$0.17
$0.06
$0.003
$0.01
$0.002
$0.94
$7.68
7.9%
$0.94
$0.22
$0.10
$0.004
$0.01
$0.002
$1.28
$10.08
10.3%
$0.68
$0.16
$0.05
$0.003
$0.01
$0.002
$0.91
$8.96
9.2%
$1.40
$0.33
$0.11
$0.01
$0.02
$0.003
$1.86
$23.83
24.4%
$1.25
$0.29
$0.08
$0.01
$0.01
$0.003
$1.64
$29.16
29.9%
—
—
—
$0.01
$0.02
$0.004
$0.03
$17.79
18.2%
 Detail may not add to total due to rounding.

6.6.1  Omissions, Biases, and Uncertainty

There are several omissions, biases, and uncertainties that affect EPA's estimate of total costs,
which are summarized in Exhibit 6-31. The cost analysis does not include the costs of two
provisions: including Cryptosporidium in the definition of GWUDI systems and including
Cryptosporidium in watershed requirements for unfiltered systems. EPA does not have
data sufficient to estimate either the number of systems that will be affected by including
Cryptosporidium in the definition of GWUDI systems or the potential effects on these systems
of including them under rules that apply to GWUDI systems.
February 15, 2000
6-41
RIAfor the Proposed LT1FBR

-------
                    Exhibit 6-31.  Summary of Cost Analysis Uncertainty
Item
Potential Effect
on Costs
Comments
Omissions
Excluded analysis of provision that
includes Cryptosporidium in the
definition of GWUDI
Excluded analysis of provision that
includes Cryptosporidium in the
watershed requirements for unfiltered
systems
Excluded analysis of demonstrations
+
+
+
Any affected systems will incur incremental costs
of changes in compliance requirements.
Any affected systems will incur incremental costs
of changes in compliance requirements.
Systems and States will incur costs associated
with preparing and reviewing demonstrations.
EPA is gathering data to develop cost estimates.
Biases
Assumed no market responses to system
cost increases
Assumed system-level costs for
community systems were applicable to
noncommunity systems
Included purchased water systems
Assumed all affected systems would move
their recycle location
-

-
—
Demand responses to price changes may mitigate
total costs.
Noncommunity systems may have lower flow
rates than community systems, which would
generate lower system-level costs using the
engineering cost models.
Excluding these systems from the baseline and
compliance forecast would reduce costs.
Some of these systems may only incur the cost of
preparing a waiver to allow another return
location.
Uncertainties
Cost estimates based on model drinking
water systems and aggregate costs based
on compliance forecasts constructed from
SDWIS, AWWA, and ICR data
+/•
The engineering models and burden analyses are
based on model systems or expected burdens.
Actual costs and burdens will differ across
systems. The compliance forecasts are based on
sample data; the actual number of systems
implementing treatment changes will most likely
differ from EPA's projections.
  = resolving the omission, bias, or uncertainty will tend to increase costs.
  = resolving the omission, bias, or uncertainty will tend to reduce costs.
 -/• = the effect of the omission, bias, or uncertainty on costs is undetermined.
Under the SWTR, unfiltered systems are required to meet watershed control requirements that
include developing a watershed control program to minimize the potential for source water
contamination by Giardia lamblia and viruses.  Because the sources of contamination for both
Giardia lamblia and Cryptosporidium are the same (e.g., wild animal populations, wastewater
treatment plants, grazing animals, feedlots, and recreational activities), EPA believes existing
watershed programs will not require significant modification to comply with the LT1FBR.
Therefore, the Agency has not developed costs for this component of the rule. In addition, EPA
RIAfor the ProposedLT1FBR
6-42
February 15, 2000

-------
does not have sufficient data to estimate costs associated with the preparation and review of
demonstrations.

Some of EPA's assumptions introduced bias into cost estimates. As noted in Section 6.1, the cost
analysis estimates only implementation costs, which potentially overstate social costs because it
excludes market responses to changes in drinking water production costs. Demand-side responses
to price increases may reduce social costs in the long run by reducing demand for water or by
shifting demand from systems that may incur large cost increases to systems that operate at a lower
cost.

The system estimates in the compliance forecasts include purchased water systems. The majority
of these systems will not actually incur the costs discussed in this chapter because they purchase
treated water from a wholesale  system.  This approach will over state costs for the provisions that
only affect small systems because EPA is including costs for small wholesale systems and for the
small systems that purchase treated water from them, although only the wholesale systems treat
water.  There are some wholesale systems, however, that report a small number of retail customers
in SDWIS. These systems are included among the smallest system size categories in the baseline,
thereby inflating these numbers while decreasing the larger system estimates. Although these
systems are classified as serving fewer customers than their production rates suggest, costs for
these systems will not be underestimated because EPA has included costs for the wholesale system
and the systems that purchase the treated water, so all of the production is captured in the cost
analysis. Finally, EPA may be  including costs and benefits for small systems that purchase water
from large systems, both of which properly accrue to the IESWTR, but these will offset one
another in the net benefit analysis. EPA cannot determine the extent of this effect on the cost
analysis and chose to retain all small systems to develop consistent cost and benefit analyses.

Finally, the methods used to estimate costs introduced uncertainty into the analysis because  actual
system and State-level costs will vary from the modeled treatment costs or estimated burden costs.
Furthermore, the compliance forecasts are EPA's estimates of the numbers of systems potentially
affected by various provisions.  These forecasts are based on a variety of sources including  sample
data from  the AWWA recycle survey and information gathered under the ICR.  They may over or
under estimate the actual number of systems affected by various proposed provisions (e.g., the
number of direct recycle systems) and/or the number of systems altering treatment practices. EPA
cannot determine whether the methods and data tend to over or under estimate total costs.

6.7   Household Costs

Water system cost increases are often passed on to customers, including households, in the form of
higher monthly water bills.  This section approximates potential household impacts of the proposed
LTlFBRby estimating two distributions of household costs based on the system costs discussed
above:

       •   Costs for the turbidity, disinfection benchmarking, and covered finished water
           reservoir provisions

       •   Costs for the recycle provisions.

February 15, 2000                                6-43                    RIA for the Proposed LT1FBR

-------
EPA estimated two distributions because it cannot determine how many small systems incurring
turbidity treatment costs will also incur recycle treatment costs. Of the 5,896 small conventional
and direct filtration systems, EPA expects approximately 2,406 systems to alter turbidity treatment
practices and 889 to alter recycling practices; the extent to which these two subsets overlap is
uncertain.

6.7.1   Household Cost Estimation Method

Most annual system costs were the same for all systems and these costs could be readily converted
to household costs by dividing per-system costs by the average number of households reported in
SDWIS. Turbidity treatment costs, however, needed to be allocated across systems because only a
subset of systems incur these costs. To obtain estimates of maximum potential household impacts,
EPA developed an allocation method that distributed costs across systems in a way that maximizes
costs for a subset of systems. First, EPA identified a subset of process improvements that are
mutually exclusive (i.e., systems would not implement them together), then distributed the costs for
these improvements across the systems expected to modify treatment.  For example, systems may
implement any or all three of the chemical addition activities, so the analysis assumes that some
systems (i.e., based on the minimum compliance estimate across the activities) undertake all three.
These changes, however, are substitutes for adding or overhauling filter media, which are applied,
therefore, to a different subset of systems. Appendix H illustrates these distributions.  Then the
remaining process improvement costs, which any system might incur, were allocated  across the
systems from highest to lowest cost.  Thus, some systems incur the highest cost combination from
the first set as well as costs for all of the other improvements, totaling as many as 12 treatment
changes. Because the compliance forecast differs across the five  small system size categories,
Appendix H shows distributions by size category.

EPA added expected costs for the turbidity monitoring, benchmark, and covered finished water
provisions to the treatment cost distributions to calculate the aggregate cost per system.  The
percentiles from the allocation process determine how many systems incur aggregate  costs for each
part of the distribution.  This distribution was then converted to a household basis.  System costs
were converted to household costs by dividing total annual  costs  by the mean number of
households per community water system, and numbers of community water systems were
multiplied by mean households per system. EPA limited the analysis to community water systems
because only those systems serve residential customers. Aggregating these results across system
size categories EPA obtained a cumulative distribution of cost per household for the turbidity,
benchmarking, and covered finished water provisions.  Household cost estimate details are shown
in Appendix H.

This method tends to overestimate the highest costs.  To be on the upper bound of the curve,  a
system would have to implement a large number of the treatment process improvements.  Such
system-level costs are unlikely to occur because there are less costly alternatives such as
purchasing a package plant or connecting to a larger regional water system. The degree of
overestimation, however, is less  severe than the  method used for the household cost analysis of the
IESWTR. That analysis identified only four mutually exclusive treatment activities, so some
RIA for the ProposedLT1FBR                      6-44                              February 15, 2000

-------
systems incurred costs for as many as 28 treatment changes. As a sensitivity analysis, EPA
replicated this approach for the proposed LT1FBR and reports the results below.

The distribution of costs for the recycle provisions is less complex because the only overlap in
recycle process changes is among direct recycle systems, some of which may need to move
recycle to the point prior to primary coagulant addition as well as install a flow equalization basin.
EPA estimated this overlap by size category based on data in the AWWA Survey (1998). EPA
calculated a system cost distribution across all of the recycle provisions and converted it to a
household cost distribution using the approach described above (see Appendix H for details).

6.7.2   Results of Household Cost Analysis

Exhibit 6-32 illustrates the cumulative distribution of household costs for all small systems for the
LT1 provisions (e.g., turbidity, benchmarking, and covered finished water).  The mean cost per
household is $8.66.  The chart shows per-household costs of $10 per year or less for 86 percent of
the 6.6 million households affected by those provisions, and costs of $120 per year (i.e., $10 per
month) or less for approximately 99 percent of households.  Per-household cost exceeds $240 per
year (i.e., $20 month) for  approximately 12,000 households.  It exceeds $500 per year for fewer
than 600 households. Costs exceeding $500 per household occur only for the smallest size
category, and the number of affected households represent about 34 of the smallest systems. The
highest per-household cost estimate is $2,177.  This extreme estimate, however, is an artifact of the
way the system cost distribution was generated. As noted above, it is unlikely that any small
system will incur annual costs of this magnitude because less costly options are available. In
comparison, the maximum cost per household would be $3,147 using the cost allocation method
developed for the IESWTR RIA because that approach would tend to shift treatment costs toward
a smaller set of systems.  Consequently, 90 percent of households would have costs below $10 per
year.

Exhibit 6-32 also illustrates the distribution of household costs for the recycle provisions. The
mean cost per household is $1.79 and the cost per household is less than $10 for 99 percent of 12.9
million households potentially affected by the proposed rule. The cost per household exceeds
$120 for approximately 1,800 households  and it exceeds $500 for approximately 100 households.
The maximum cost of $1,238 would only be incurred if a direct filtration system in the smallest
size category installed a sedimentation basin.
February 15, 2000                                6-45                     RIA for the Proposed LT1FBR

-------
         Exhibit 6-32. Distributions of Annual Household Costs for the Turbidity,
       o
       0)
      Q_
         100% -r
          90% -
          80% -
          70% -
          60% -
          50% -
          40% -
30% -
          20% -
          10% -
           0%
            $0.00
             $0.01
$0.10      $1.00      $10.00     $100.00

   Annual Cost per Household (log scale)
$1,000.00   $10,000.00
                                    •Recycle Provisions
                                             • LT1 Provisions
      Benchmarking, and Covered Finished Water Provisions and Recycle Provisions
There are approximately 1.5 million households served by small drinking water systems that may
be affected by the recycling provisions in addition to the turbidity, benchmarking, and covered
finished water provisions. The expected aggregate annual cost to these households can be
approximated by the sum of the expected cost for each distribution, which is $10.45.

6.8     Cost Effectiveness

The cost effectiveness of the proposed rule can be measured as the cost per case of avoided illness.
The quantified benefit of mean avoided cases of illness per year ranges from 22,800 to 83,600
avoided cases for the turbidity provisions alone. Dividing the associated costs of $68.6 million
(assuming a 7 percent discount rate) by this range, the resulting cost per case of avoided illness
ranges from $800 to $3,000.

The overall cost of the preferred combination of regulatory alternatives for LT1FBR is $97.5
million (assuming a 7 percent discount rate); this includes the $68.6 million in costs attributed to
the turbidity provisions.  Dividing this preferred combination total by the quantified benefit of total
avoided cases (i.e., 22,800 to 83,600) would overstate the cost per case estimate.  If the other
provisions are equally effective in reducing illness, the per case cost of avoided illness would be
the same (i.e., $800 to $3,000 per case). If the other provisions are less effective than the turbidity
provisions, then the costs would be higher, not to exceed $4,300 as a highest estimate (i.e., the
RIAfor the ProposedLT1FBR
                                   6-46
                                              February 15, 2000

-------
$97.5 million divided by 22,800 cases—all presumably attributable, in this worst case estimate, to
the turbidity provisions). If, however, the other provisions are more effective than the turbidity
provisions in avoiding illness, then the cost per case would be lower than the $800 to $3,000 per
case estimate range.
February 15, 2000                                 6-47                      RIA for the ProposedLT1FBR

-------
                            7.  Economic Impact Analysis

The rule promulgation process requires EPA to perform a series of distributional analyses that address
the potential regulatory burden placed on entities directly or indirectly affected by the rule. This chapter
contains all or part of EPA's analyses and statements with regard to six Federal mandates:

        (1)    The Unfunded Mandates Reform Act (UMRA) of 1995;
        (2)    Executive Order 12886 (Regulatory Planning and Review);
        (3)    the Regulatory Flexibility Act (RFA) of 1980, as amended by the Small Business
              Regulatory Enforcement Fairness Act (SBREFA) of 1996;
        (4)    Technical, Financial, and Managerial Capacity Assessment required by Section
              1420(d)(3) of the 1996 amendments to the Safe Drinking Water Act (SOWA);
        (5)    Executive Order 13045 (Protection of Children From Environmental Health Risks and
              Safety Risks); and
        (6)    Executive Order 12989 (Federal Actions to Address Environmental Justice in Minority
              Populations and Low-Income Populations).

The preparation of this regulatory impact analysis for the LT1FBR is a response to the requirements of
Executive Order 12886. This chapter is a response to the remaining five mandates.  In addition, this
chapter contains a summary of the analysis conducted to fulfill requirements set forth by the Paper
Work Reduction Act.  A separate Information Collection Request (ICR) document, entitled the
LT1FBR Information Collection Request,  contains the complete analysis.

This chapter is organized into three sections. The first section addresses how the proposed rule
pertains to those mandates concerning potential impacts to government and business entities.  The next
section considers the impact of the proposed rule on possible sensitive subpopulations, such as children.
The final section addresses the potential impact to minority and low-income populations.

7.1    Impacts on Governments and Business Units

The following sections contain the analyses necessary to fulfill Executive Orders pertaining to
governments  and businesses.  Section 7.1.1  provides the UMRA analysis. Section 7.1.2 discusses
possible impacts to Indian Tribal Governments.  Section 7.1.3 is the required RFA and SBREFA
analysis. Section 7.1.4 is the Capacity analysis, and Section 7.1.5 gives a summary of the ICR.

7.1.1   Unfunded Mandates Reform Act

Title II of the UMRA, P.L. 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 UMRA
Section 202, EPA generally must prepare a written statement, including a  cost-benefit analysis, for
proposed and final rules with "Federal mandates" that may result in expenditures to State, local, and
Tribal governments, in the aggregate, or to the private sector, of $100 million or more in any  1 year.
EPA estimated annual aggregate State, local, and Tribal government expenditures for the 7 percent

February 15, 2000                                 7-1                       RIA for the Proposed LT1FBR

-------
discount rate assumption scenario by adding State program costs of $6.7 million to the share of system
costs potentially incurred by publicly owned systems. These systems account for approximately 73.9
percent of the $90.8 million in total annual system costs, which is $67.1 million per year.  Thus, State
program costs and publicly owned system costs total $73.8 million per year.

Although this falls below $100 million, the cost figure is close enough to the threshold that the Agency
expects it to surpass the threshold within the 20-year analysis period due to inflation at some point in
the future.  Therefore, EPA has determined that this rule contains a Federal mandate that may
eventually result in expenditures of $100 million or more for State, local, and Tribal governments, in the
aggregate and the private sector in any 1 year.  Accordingly, under Section 202 of the UMRA, EPA is
obligated to prepare a written statement addressing:

       •    The authorizing legislation

       •    Cost-benefit analysis including an analysis of the extent to which the Federal government
            will pay for the costs of State, local and Tribal governments

            Estimates of future compliance costs and disproportionate budgetary effects

       •    Macroeconomic  effects

            A summary of EPA's consultation with State, local, and Tribal governments and their
            concerns, including a summary  of the Agency's evaluation of those comments and
            concerns

            Identification and consideration of regulatory alternatives and the selection of the least
            costly, most cost-effective or least burdensome alternative that achieves the objectives of
            the rule.

Chapter 2 details the authorizing legislation. Cost-benefit analyses, disproportional budgetary effects,
macroeconomic effects, and consultations are addressed in the rest of this chapter. Future compliance
costs are discussed in Chapter 6. And both Chapters 3 and 6 address the potential regulatory
alternatives, with Chapter 6 showing that the preferred alternatives are the most cost effective ones that
achieve the public health objectives.

Before promulgating a rule that requires a written statement, Section 205 of the UMRA generally
requires EPA to identify and consider a reasonable number of regulatory alternatives and then adopt the
least costly, most cost effective or least burdensome alternative that achieves the objectives of the  rule.
However, the provisions of Section 205 do not apply when they are inconsistent with applicable law.

Under Section 1412(b), the SDWA requires that MCLs be set as close to MCLGs "as is feasible,"
except when EPA determines that the cost of a standard at that level are not justified by the benefits, or
when certain "risk-risk" considerations apply. Whereas, MCLGs are nonenforceable health goals
based only on health effects  and exposure information, MCLs are enforceable standards that

RIA for the ProposedLT1FBR                        7-2                                February 15, 2000

-------
SDWA directs EPA to set with the use of the best technology, treatment techniques, and other
means that the Administrator finds available.  Also, SDWA requires the Agency to identify the
best available technology (BAT) that is feasible for meeting the MCL for each contaminant.
Under Section 1412 (b)(7)(A), if it is not economically or technically feasible to ascertain the level of a
contaminant in drinking water, EPA may require the use of a prescribed treatment technique instead of
an MCL.

As a result of this mandate set forth by the SDWA, EPA can choose an alternative that is not the most
cost effective if it determines that this is necessary to attain health goals as close to MCLGs as feasible.
Moreover, Section 205 allows EPA to adopt an alternative other than the least costly, most cost
effective, or least burdensome alternative if the Administrator publishes an explanation why that
alternative was not adopted within the final rule.

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 the 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.

Social Costs and Benefits

The social benefits are those that accrue to the public through an increased level of protection from
exposure to Cryptosporidium and other pathogens in drinking water.  Chapter 5 presents the benefit
analysis, which includes both qualitative and monetized benefits of improvements to health and safety.
Because of scientific uncertainty regarding the exposure assessment and the risk assessment for
LT1FBR, the Agency has used statistical methods to assess the benefits of LT1FBR. The  methods
quantified and valued the cryptosporidiosis illnesses and mortalities avoided due to the revised
combined filter effluent standards. This rule, however, may also decrease illness from Giardia and
other emerging disinfection resistant pathogens further increasing the benefits of the rule.  Additional
benefits of the rule include reduced risks of outbreaks and enhanced aesthetic water quality.

Measuring the social costs of the rule requires identifying affected entities by ownership (public or
private), considering regulatory alternatives, calculating regulatory compliance costs,  and estimating any
disproportionate impacts. Chapter 6 of this document details the cost analysis  performed for the
LT1FBR. EPA expects the proposed rule to have a total annualized cost of approximately $87.6 or
$97.5 million depending on the discount rate.

The Federal government may defray a portion of the cost of the rule by providing financial  assistance to
State, local, and Tribal governments in complying with this rule.  The Federal government provides
funding to States that have primary enforcement responsibility for their drinking water programs through
February 15, 2000                                 7-3                      RIA for the Proposed LT1FBR

-------
the Public Water Systems Supervision Grants program.12 Additional funding is available from other
programs administered either by EPA or other Federal agencies. These include the Drinking Water
State Revolving Fund (DWSRF) and Housing and Urban Development's Community Development
Block Grant Program. For example, SDWA authorizes the Administrator of the EPA to award
capitalization grants to States, which in turn can provide low-cost loans and other types of assistance to
water systems for eligible projects.  The DWSRF assists  public water systems with financing the costs
of the infrastructure needed to achieve or maintain compliance with SDWA requirements. Each State
has considerable flexibility to determine the design of its program and to direct funding toward its most
pressing compliance and public health protection needs.  The Drinking Water State Revolving Fund
Program Guidelines detail a variety of ways that States can use funds to assist small systems (U.S.
EPA, 1997b). The State must use a minimum of 15 percent of the DWSRF grant to provide
infrastructure loans to small systems. Furthermore, the State may use  2 percent of the grant to
provide technical assistance to small systems. For disadvantaged small systems, the State can use
up to 30 percent of its DWSRF money to increase loan subsidies. States may also, on a matching
basis, use up to 10 percent of their DWSRF allotments for each fiscal year to assist in running the State
drinking water program.

Disproportionate Impacts

This section examines disproportionate impacts on geographic or social segments of the nation. In
general, the costs that a PWS, whether publicly or privately owned, would incur to comply with this
rule would depend on many  factors that are independent of location. However, the data needed to
confirm this assessment and to analyze other impacts of this problem is not available; therefore, EPA
looked at four other factors:

        •   The impacts of small versus large systems and the impacts within the  five small system size
           categories

           The costs to public versus private water systems

        •   The costs to households (See Chapter 6)

           The distribution of costs across States.

First, small systems will experience a greater impact than large systems under LT1FBR because large
systems are subject only to the recycle provisions; the Interim Enhanced Surface Water Treatment Rule
(IESWTR) promulgated turbidity, benchmarking, and covered finished storage provisions for large
systems in December, 1998.  However,  small systems have realized cost  savings  over time due to their
exclusion from the IESWTR.
       12The Federal government also defrays State costs by providing for the administrative cost and burden for States and
Territories that do not have primacy. In addition, the Federal government administers many of the treatment plants in national
forests and military installations.

RIA for the ProposedLT1FBR                        7-4                                February 15, 2000

-------
The second measure of impact is the relative total cost to privately owned water systems compared to
that incurred by publicly owned water systems. A majority of the systems are publicly owned (60
percent of the total). As a result, publicly owned systems will incur a larger share of the total costs of
the rule. However, EPA has no basis for expecting the cost per system to differ systematically with
ownership.

The third measure of impact is at the household level. Chapter 6 includes this analysis, as part of the
overall cost analysis.

The fourth measure of budgetary impacts is geographically across States. There is nothing to suggest
that costs to individual systems would vary significantly from State to State.  Yet this does not preclude
a specific State or region from being significantly impacted more by the rule. Therefore, EPA
conducted an analysis of the potential geographic impact of LT1FBR on the various States.13 For State
budgetary impacts, the costs for starting and annually administering the LT1FBR rule are combined with
information on the distribution of PWSs across States.

Exhibit 7-1 shows the distribution of annual costs to States for the proposed rule. From the map it is
apparent that Texas, New York, California, Oklahoma, and Illinois are the States with the highest
annual costs. As mentioned in Chapter 6, the turbidity monitoring provision of the LT1FBR would be
the source of the greatest financial burden to States.  Since the turbidity provision only applies to small
surface and GWUDI systems, those States with the greatest number of small systems would incur the
highest costs. Exhibit 7-2 shows the distribution of surface and GWUDI systems serving fewer than
10,000 over the 50 States.  A comparison of the two maps suggests that the geographic distribution of
costs is closely  correlated with the distribution of small systems.

Estimates of the cost increase are only one measure of potential budgetary impacts, and other
comparisons can provide additional perspective.  For instance, the five States with the highest potential
costs also had the highest number of small systems.  Of these five States, four have very large
populations.14  States with larger populations may already have larger budgets for program
expenditures, and the proportional cost increase due to LT1FBR may be smaller than or comparable to
other States. Therefore, EPA compared the estimated percentage increase to overall State drinking
water program  costs that would result from the rule.
        13The information on systems per State was derived from the Water Industry Baseline Handbook (U.S. EPA,1999c).
This information is from 1997 and previous years, whereas the cost analysis in Chapter 6 is based upon the most recent
information contained in the SDWIS, therefore there are some discrepancies in the total number and distribution of systems
between the two analyses.

        14California, Texas, New York, and Illinois will be the first, second, third, and sixth most populated States, respectively,
in 2000 according to population projections of the United States Census Bureau.
www.census.gov/population/projections/state/stpjpop.txt

February 15, 2000                                   7-5                       RIA for the Proposed LT1FBR

-------
           Exhibit 7-1. Geographic Distribution of Annual LT1FBR Costs to States
                                                  Annual LT1FBR Costs for
                                               States (in thousands of Dollars)
                                                  D > $500,000       (3)
                                                  D $300,009 $500,000 (3)
                                                  D $100,009 $300,000(17)
                                                  Q< $100,000      (27)
                 Exhibit 7-2.  Small Surface and GWUDI System Distribution
RIAfor the ProposedLT1FBR
7-6
February 15, 2000

-------
Exhibit 7-3 shows the cost increase to the States as a percentage of overall State drinking water
expenditures that would result from implementation the LT1FBR.15  Over the entire country, the
average increase in program expenditures would be 4.1  percent.  The States of Colorado, Alabama,
Oklahoma, Texas, West Virginia, Kansas, and Kentucky would have the greatest increase in
expenditures as a result of the rule. However, if the State's per capita drinking water expenditures are
relatively low, in relation to other States then a large percentage increase still may not result in a
significant burden to the State.16 Conversely, if the State already has higher than average expenditures
per person, then the percentage increase may underestimate the actual burden of the rule.

      Exhibit 7-3:  LT1FBR Costs as a Percentage of State Drinking Water Expenditures
                                                      LT1FBR Costs as a % of State
                                                      Drinking Water Expenditures
                                                           No Expenditure Data  (4)
                                                         D>=15%         (1)
                                                                        (6)
                                                            -10%        (6)
                                                            5%          (33)
Exhibit 7-4 shows the geographic distribution of per capita drinking water expenditures. From this
exhibit it is apparent that, if the rule were implemented, none of the seven States mentioned in the
previous paragraph would likely be required to spend more than $1.50 per person annually.  The
        15State expenditure data was not available for the States of Alaska, Connecticut, Louisiana, Montana, Nevada, New
Hampshire, New Mexico, North Dakota, and Wyoming. Annual revenue data was available for Alaska, Louisiana, Nevada, New
Mexico, and North Dakota (U.S. EPA, 1999d).  This data was used as an approximate measure of likely expenditures for the
respective States.  These expenditures and revenues do not incorporate the compliance costs for the IESWTR and the Stage 1
DBPR, so the percentage increases are overstated and the overall costs are underestimated.

        16Per capita expenditure is based upon total State Drinking water expenditures for the 1997 divided by the number of
people served by Public Drinking Water Systems in 1997  (U.S. EPA, 1999d).
February 15, 2000
7-7
RIAfor the Proposed LT1FBR

-------
average annual State cost per person for just LT1FBR and then for current drinking water expenditures
plus LT1FBR costs would be $0.04 and $1.05, respectively.  Exhibit 1-4 does show that combined
drinking water costs would require Alaska and Arkansas to spend more than $3 per person.
   Exhibit 7-4. Per Capita Expenditures for LT1FBR and Current Drinking Water Programs
                                                       Per Capita Drinking Water
                                                     Expenditures (Including LT1FBR)
                                                          Expenditure
                                                      • >$3
                                                        $2 - $3
                                                          -$2
                                                      n<$i
Yet examining potential budgetary impacts alone, provides an incomplete description of the likely
financial burden to the States.  For Alaska the cost of LT1FBR would be $0.18 per person, bringing its
drinking water expenditures to $4.56 per person. However, in 1997 Alaska acquired three-fourths of
its drinking water budget in the form of Federal PWSS grant money.17 By contrast, Arkansas
expenditures would be $3.47 after promulgation of LT1FBR, but it received only 17 percent of its
1997 drinking water budget from Federal funds.  From this perspective Arkansas would bear the
greater financial burden.

In conclusion, the evidence exhibited on the four maps (Exhibits 7-1 to 7-4) does not suggest that there
would be a disproportionate budgetary effect resulting from the rule. Exhibits 7-1, 7-3 and 7^1 do not
show evidence of a geographic concentration of higher impact attributed to the rule.  Nor does any one
State consistently fall into the top two categories of impact in all four  exhibits. Furthermore, it is
possible that the financial impact of the rule to States could be offset partially by additional PWSS grant
money from the Federal Government.
        17In 1997, 75 percent of Alaska's Drinking water program budget came from Federal PWSS grant money, 22 percent
from State General Funds, and 3 percent from other revenue sources (U.S. EPA, 1999d).
RIAfor the ProposedLT1FBR
7-8
February 15, 2000

-------
Macroeconomic Effects

EPA is required, under UMRA Section 202, to estimate the potential macroeconomic effects of the
regulation. Macroeconomic effects tend to be measurable in nationwide econometric models only if the
economic impact of the regulation reaches 0.25 to 0.5 percent of Gross Domestic Product (GDP).  In
1998, the GDP was $8,511 billion so a rule would have to cost at least $21 billion annually to have a
measurable effect.  A regulation with a smaller aggregate effect is unlikely to have any measurable
impact unless it is highly focused on a particular geographic region or economic sector.  The
macroeconomic effects on the national economy from the preferred alternatives of FBR would be
negligible based on the estimated total annual cost range of $87.6 to $97.5 million.  In addition, from
the analysis in the previous section EPA does not expect that costs would be highly concentrated on
any particular geographic region.

Summary of Consultation Efforts with State,  Tribal, and Local Governments

Under UMRA Section 202, EPA is  to provide a summary of its consultation with elected
representatives (or their designated authorized employees) of affected State, local and Tribal
governments in this rulemaking. EPA initiated consultations with governmental entities and the
private sector affected by this rule through various means. This included participation on a
Regulatory Negotiation Committee, chartered under the Federal Advisory Committee Act
(FACA), in  1992-93 that included stakeholders representing State and local governments, public
health organizations, public water systems, elected officials, consumer groups, and environmental
groups.

In accordance with the Regulatory Flexibility Act (RFA), as amended by the SBREFA, EPA
convened a Small Business Advocacy Review Panel. The Review Panel allows small regulated
entities to provide advice and perspective to EPA early in the regulatory development process.
EPA also provided an informal draft of the preamble to SBREFA representatives and individuals
who attended either of the two stakeholder meetings.  Because this was an informal review
process, EPA did not prepare formal responses to the comments, however, the Agency reviewed
the comments carefully and considered their merit when developing the regulatory  provisions in
the proposed rule.

To inform and involve Tribal governments in the rulemaking process, EPA presented the LT1FBR
at the 16th Annual Consumer Conference of the National Indian Health Board, the Annual
Conference of the National  Tribal Environmental Council, and the OGWDW/Inter Tribal Council
of Arizona, Inc. Tribal consultation  meeting. More than 900 attendees representing Tribes from
across the country attended the National Indian Health Board's Consumer Conference and over
100 Tribes were represented at the annual  conference of the National Tribal Environmental
Council. At both conferences, an OGWDW representative conducted two workshops on EPA's
drinking water program and upcoming regulations, including the LT1FBR. At the OGWDW/Inter
Tribal Council of Arizona meeting,  representatives from 15 Tribes participated. In addition, EPA
sent the presentation materials and meeting summary to more than 500 Tribes and Tribal
organizations.
February 15, 2000                                7-9                     RIA for the Proposed LT1FBR

-------
The primary concern expressed by the governments was the ability of the smallest systems to staff
drinking water treatment facilities adequately to perform the monitoring and reporting associated
with the new requirements. The proposed rule attempts to minimize the monitoring and reporting
burden to the greatest extent feasible and still accomplish its objective of protecting public health.
The Agency believes the monitoring and reporting requirements are necessary to ensure
consumers served by small systems  receive a comparable level of public health protection as
consumers served by large systems.

7.1.2  Indian Tribal Governments

Under Executive Order 13084, EPA may not issue a regulation, which is not required by statute, that
significantly or uniquely affects the communities of Indian Tribal governments, and that imposes
substantial direct compliance costs on those communities, unless the Federal government provides the
funds necessary to pay the direct compliance costs incurred by the Tribal governments or EPA consults
with those governments. If EPA complies by consulting, Executive Order 13084 requires EPA to
provide to the Office of Management and Budget, in a separately identified section of the preamble to
the rule, a description of the extent of EPA's prior consultation with representatives of affected Tribal
governments, a summary of the nature of their concerns, and a statement supporting the need to issue
the regulation. In addition, Executive Order 13084 requires EPA to develop an effective process
permitting elected officials and other representatives of Indian Tribal governments "to provide
meaningful and timely input in the development of regulatory policies on matters that significantly or
uniquely affect their communities."

EPA has concluded that this rule will significantly affect communities of Indian Tribal governments.18 It
will also impose substantial direct compliance costs on such communities, and the Federal government
will not provide the funds necessary to pay the direct costs incurred by the Tribal governments in
complying with the rule. In developing this rule, EPA consulted with representatives of Tribal
governments pursuant to UMRA and both Executive Order 12875 and Executive Order 13084. As
described in the UMRA discussion in the previous section, EPA held extensive meetings that
provided the opportunity for meaningful and timely input in the development of the proposed rule.
The public docket for this  rulemaking includes summaries of the meetings.

7.1.3  Regulatory Flexibility Act and Small Business
       Regulatory Enforcement Fairness Act

The provisions of the Regulatory Flexibility Act, 5 U.S.C. 601 et seq., as amended by the Small
Business Regulatory Enforcement Fairness Act of 1996, require EPA to prepare a regulatory flexibility
analysis unless the Agency certifies that the rule will not have "a significant economic impact on a
substantial number of small  entities." A regulatory flexibility analysis describes the impact of the
regulatory action on small entities as part of the rule promulgation process. The 1996 amendments to
the SDWA define a small public water system as a system serving fewer than 10,000 persons. This
       18There are approximately 60 small PWSs that use surface water or GWUDI classified as Tribal systems in
the SDWIS.
RIA for the ProposedLT1FBR                       7-10                               February 15, 2000

-------
definition reflects the fact that the original 1979 standard for total trihalomethanes applied only to
systems serving at least 10,000 people. The definition thus recognizes that baseline conditions from
which systems serving fewer than 10,000 people will approach disinfection byproduct control and
simultaneous control of microbial pathogens is different from that for systems serving 10,000  or more
persons.

Background and Quantitative Analysis

When a proposed or final rule may potentially have an adverse effect on one or more small entities, the
RFA and SBREFA require EPA to determine the extent of the impact and the number of small entities
affected. If it is determined that the rule would not have a "significant impact on a substantial  number of
small entities," then the Agency can certify the rule. If the Agency determines that the rule would have
an impact then the RFA/SBREFA requires that EPA prepare an Initial Regulatory Flexibility  Analysis
(IRFA) for a proposed rule, or Final Regulatory Flexibility Analysis for a final rule.  Chapter 4 of this
document provides data on the small entities potentially affected by LT1FBR, and Chapter 6 discusses
the changes systems would have to make and the likely costs.  Using information found in these two
chapters, along with additional information from SDWIS and CWSS, EPA conducted a quantitative
analysis to assist in determining whether to  certify the rule or prepare an IRFA.

The Agency recognizes that economic characteristics will vary among entities affected by a given rule.
Therefore, EPA evaluated the  potential economic impact by comparing compliance costs as a
percentage of sales, revenues,  and operating expenses for small businesses, governments, and non
profit organizations respectively. Data on water systems changes frequently, which makes it difficult to
describe the universe of surface water systems with specificity. Similarly, ownership data is difficult to
ascertain as most data sets, such as SDWIS (Safe Drinking Water Information System), do not
maintain such information.  For this analysis, the number of publicly and privately owned water systems
was derived using the ratio of public to private water systems as reported in the 1995 Community
Water System Survey (CWSS).  Using SDWIS and  CWSS, EPA estimates that the changes to the
Surface Water Treatment Rule will potentially affect  11,593 surface water systems  and GWUDI
systems.  Of these systems, EPA estimates that small businesses own 37.2 percent, 56.7 percent are
small governments, and 6.1 percent are nonprofit organizations.19  While it was not possible to use
existing data to establish the exact profile of water system ownership, EPA used information in the
Water Industry Baseline Handbook (U.S. EPA, 1999c) to approximate an ownership profile. As
shown in Exhibit 7-5, the data suggest that  a majority of small systems are publicly  owned.20
        19The Water Industry Baseline Handbook separates system ownership data into public, private or other (U.S. EPA,
1999c).  For this analysis, EPA assumed for the small systems that public represents small government, private represents small
business, and other represented small non-profit.

        20A public water system provides piped water for human consumption. The term "public water system" applies not only
to water utilities, but also to wide range of privately or publicly owned businesses and entities that provide drinking water (e.g.,
campgrounds, factories, restaurants, and schools). Public water systems are classified as community, nontransient noncommunity,
or transient noncommunity systems.

February 15, 2000                                   7-11                      RIA for the Proposed LT1FBR

-------
      Exhibit 7-5.  Number and Percent of Public and Private System, by Size of System
System
Type
Public
Private
Other
Total
System Size
<100
691 (24.6%)
1,755(62.6%)
358 (12.8%)
2,804 (24%)
101-500
1,308(45.7%)
1,350(47.2%)
202(7.1%)
2,860 (25%)
501-1,000
913 (65.3%)
418(29.8%)
68 (4.9%)
1,399 (12%)
1,001-3,300
2,020 (76.8%)
555(21.1%)
56(2.1%)
2,631 (23%)
3,301-9,999
1,638(86.3%)
236 (12.4%)
25(1.3%)
1,899 (16%)
Total
6,570 (56.7%)
4,314(37.2%)
709(6.1%)
11,593 (100%)
 Note'. Number (percent) within system size category.

The LT1FBR proposed rule contains provisions for turbidity monitoring and treatment, disinfection
benchmarking, and filter backwash recycling.  Chapter 6 discusses these provisions and Exhibits 7-6
through 7-8 summarize EPA's estimate of the number of small entities that LT1FBR provisions will
affect.21
  Exhibit 7-6. Small Entities Affected by the Turbidity Monitoring and Turbidity Treatment
                                   Provisions of LT1FBR
System Size
(population served)
<100
101-500
501-1,000
1,001-3,300
3,301-9,999
Totals
Total Number
of Systems
836
1,117
810
1,655
1,478
5,896
Systems to Modify
Treatment and Monitor
341
456
331
675
603
2,406
Systems to
Monitor Only
495
661
479
980
875
3,490
      Exhibit 7-7. Small Entities Affected by the Benchmarking Provisions of LT1FBR
System Size
(population served)
<100
101-500
501-1,000
1,001-3,300
3,301-9,999
Totals
Total Number
of Systems
1,404
2,333
1,301
2,553
1,859
9,450
Systems to
Do Applicability
Monitoring
162
301
128
253
116
960
Systems
Disinfection
Profiling
1,242
2,032
1,173
2,300
1,743
8,490
Systems to Develop
Benchmarks
407
677
377
740
539
2,741
        'The numbers of systems potentially affected by each provision is based upon the preferred alternative set.
RIAfor the ProposedLT1FBR
7-12
February 15, 2000

-------
                  Exhibit 7-8.  Small Entities Affected by the Filter Backwash
                                Recycle Provisions of LT1FBR
System Size
(population served)
<100
101-500
501-1,000
1,001-3,300
3,301-9,999
Totals
Systems Moving
Recycle Return
Location
81
108
78
160
143
569
Systems
Performing
Self Assessments
107
143
104
212
190
757
Systems with
Direct Filtration
35
47
34
70
62
248
The major impact of the rule is the requirement to install and operate water filtration equipment to meet
turbidity standards of quality in delivered water.  These requirements pertain to systems that use either
conventional or direct filtration to treat their water.  Systems that purchase treated water from another
source may see an increase in their wholesale costs, but a database sufficient to track all the wholesale
treated water transactions in the country does not exist. Impacts are therefore evaluated as though all
small systems treat water. The data with which to characterize the capacities and flows of these
facilities that treat water does exist and provides an  adequate basis for assessing total capital and
operating costs. In Chapter 6 of this document,  as part of a household cost analysis, EPA developed
assumptions regarding the steps that systems of various sizes will need to take to comply with the
LT1FBR rule.  EPA was then able to generate a hypothetical distribution of per-system costs.

For the quantitative analysis, EPA used data from the CWSS to estimate mean sales, revenues, and
expenditures for each system size and ownership category.22  EPA then used the cost distributions for
the turbidity, benchmark, and covered finish storage provisions to estimate cost-to-revenue ratios with
each system size and ownership category to determine the potential extent of the financial impact on
small entities. Some of these systems may also incur costs under the recycle provisions, but EPA
cannot determine the number of systems for which this cost overlap would occur, and the majority of
any system's recycle costs would be for changes in recycle practices, which will be determined by the
State. However, not including the potential costs of the recycle provisions as part of the analysis does
not change EPA's conclusion regarding the potential impact of the proposed rule on small entities.

Exhibit 7-9 presents the results from the analysis of the potential financial burden on PWSs from the
proposed rule.  The exhibit shows a comparison between the  distribution of system costs developed in
Chapter 6 and the financial data from CWSS.  The complete distribution of costs for each system size
category is located in Appendix H. Exhibit 7-9 summarizes the financial data and the number of
        22Due to insufficient data, EPA could not determine the sales, revenues, or expenditures distributions by system within
each size category. Instead, EPA estimated the mean sales, revenues, or expenditures for a system within each size category. This
mean was then used to develop the cost-to-revenue ratios. Ideally, costs would be compared to financial data for every entity, or a
distribution of that data.
February 15, 2000
7-13
RIAfor the Proposed LT1FBR

-------
systems incurring costs in excess of 1 percent and 3 percent of sales, revenues, or expenditures. As
shown in the exhibit, EPA estimates that the cost of the LT1FBR rule may exceed 3 percent for 2,575
small systems. As a result of this analysis, EPA determined that there would be a significant impact on
a substantial number of small systems and thus prepared an IRFA for this rule.

 Exhibit 7-9. Results of Comparison of Mean Sales, Revenues, and Operating Expenditures
                                          to Costs
Population Category
100 and
below
101-
500
501-
1,000
1,001-
3,300
3,301-
10,000
Total #
of Systems
Total %
of Systems
Annualized Cost Distribution Summary1
High Range of
Compliance Cost
Low Range of
Compliance Cost
$47,009
$24
$51,784
$21
$74,204
$16
$122,905
$29
$186,510
$40

Small Business: Includes Private and Ancillary Systems as recorded in CWSS
Mean Total Sales
Number of Systems >1%
Number of Systems >3%
$11,481
523
523
$49,682
527
422
$83,937
241
193
$343,153
279
57
$776,189
52
22

1,624
1,218
37.6%
28.2%
Small Governments: Includes Public Systems as recorded in CWSS
Mean Total Revenues
Number of Systems >1%
Number of Systems >3%
$13,220
206
206
$86,459
510
408
$127,171
529
151
$317,751
1,017
207
$806,802
364
156

2,626
1,129
40.0%
17.2%
Small Nonprofit Organizations: Includes Homeowners' Associations as recorded in CWSS
Mean Operating
Expenditures
Number of Systems >1%
Number of Systems >3%
Total:
Number of Systems >1%
Number of Systems >3%
$5,815
107
107
836
836
$43,348
79
79
1,117
909
$121,538
40
11
810
356
$79,949
35
28
1,331
292
$528,204
15
3
432
181

276
228
4,526
2,575
38.9%
32.1%
39.0%
22.2%
 'Compliance costs are based on the preferred alternative set. See Chapter 6 for a discussion of the regulatory
 alternatives.

Requirements for the Initial Regulatory Flexibility Analysis

Because EPA is not certifying the proposed rule under SBREFA, the Regulatory Flexibility Act
requires EPA to complete an IRFA addressing the following:

       •   The need for the rule

           The obj ectives of and legal basis for the proposed rule

       •   A description of, and where feasible, an estimate of the number of small entities to which
           the rule will apply
RIAfor the ProposedLT1FBR
7-14
February 15, 2000

-------
            A description of the proposed reporting, record keeping, and other compliance
            requirements of the rule, including an estimate of the types of small entities, which will be
            subject to the requirements and the type of professional skills necessary for preparation of
            reports or records

            An identification, to the extent practicable, of all relevant Federal rules that may duplicate,
            overlap, or conflict with the proposed rule

            A description of "any significant regulatory alternatives" to the proposed rule that
            accomplish the stated objectives of the applicable statutes, and that minimize any significant
            economic impact of the proposed rule on small entities; the analysis is to discuss significant
            regulatory alternatives such as:

            •    Establishing different compliance or reporting requirements or timetables that take into
                account the resources of small entities

            •    Clarifying, consolidating, or simplifying compliance and reporting requirements under
                the rule for small entities

            •    Using performance rather than design standards

            •    Exempting small entities from coverage of the rule or any part of the rule.

EPA has considered and addressed the above  requirements as part of this RIA for the proposed
LT1FBR rule.  The following is a summary of how this and the preceding chapters met the various
requirements.  Chapter 2 explains the need, objectives of, and legal basis for the rule.  The previous
Section 7.1.3 and Chapter 6 provide a description and estimate of the small entities affected. Section
7.1.3 also discusses the coordination with other Federal rules.  Chapter 3, Chapter 6, and the current
chapter provide a discussion of regulatory alternatives. The compliance requirements are discussed in
Chapter 6 as well as section 7.1.5  of this chapter concerning the Paperwork Reduction Act and in the
Information Collection Request prepared for the LT1FBR.

Coordination With Other Federal Rules

The proposed rule does not directly overlap with any other existing or proposed rules, yet the
development of LT1FBR has occurred in coordination with several other rules, all of which are a
direct result from amendments to the SWDA. To better understand how the proposed rule relates
to other proposed and existing rules, it is necessary to briefly summarize several earlier Drinking
water regulations. For a more comprehensive history of these regulations  refer to Section 2.3 of
Chapter 2.

Three initial rules that addressed both the control of specific pathogens and disinfection byproducts
preceded the amendments to the SWDA in 1996. These were the Total Trihalomethane Rule,
February 15, 2000                                 7-15                     RIA for the Proposed LT1FBR

-------
passed in November 1979 (44 FR 68624); the Total Coliform Rule (TCR) (54 FR 27544, June 29,
1989); and the SWTR (54 FR 27486), passed on June 29, 1989.
Under the Total Trihalomethane Rule, EPA set an interim MCL for total trihalomethanes
(TTFJJVI—the sum of chloroform, bromoform, bromodichloromethane, chlorodibromomethane) of 0.10
mg/1 as an annual average. The interim TTFJJVI standard applied to community water systems using
surface water and/or ground water serving at least 10,000 people that add a disinfectant to the drinking
water during any part of the treatment process. At their discretion, States could extend coverage to
smaller PWSs; however, to date few States have chosen to exercise this option.

The TCR, which applies to all public water  systems, sets compliance with the MCL for total
coliforms (TC). All systems must have a written plan identifying where samples are to be
collected.  If a system has a TC-positive sample, it must test that sample for the presence of fecal
coliforms or E. coli.  The system must also collect a set of repeat samples, and analyze for TC  (and
fecal coliform or E. coli within 24 hours of the first TC-positive sample).  The TCR also requires
an on-site inspection (referred to as a sanitary survey) every 5 years for each system that collects
fewer than five samples per month.

Under the SWTR, EPA set maximum contaminant level goals of zero for Giardia lamblia, viruses,
and Legionella; and promulgated regulatory requirements for all PWSs using surface water sources
or groundwater sources under the direct influence of surface water. The SWTR includes treatment
technique requirements for filtered and unfiltered systems intended to protect against the adverse
health effects of exposure to  Giardia lamblia, viruses, and Legionella, as well as many other
pathogenic organisms.

In 1992 EPA instituted a formal regulatory negotiation (RegNeg) process with potentially affected
parties (57 FR 53866; November 13, 1992), to consider potential amendments to the SDWA.
Through an extensive consensus-building effort, the RegNeg Committee agreed that EPA should
propose: an Information Collection Rule (ICR) (final in 1996); a staged Enhanced Surface Water
Treatment Rule, and a staged Disinfectants/Disinfection Byproducts Rule. These rules formed the basis
for the provisions of the 1996  amendments. Those amendments established a number of regulatory
deadlines, including schedules for a Stage 1 and a Stage 2 Disinfection Byproduct Rule  (DBPR), and
for two stages of the Enhanced Surface Water Treatment Rule 1412(b)(2)(C).  The SDWA as
amended also requires EPA to promulgate regulations to "govern" filter backwash recycling within the
treatment process of public systems (Section 1412(b)(14)) and to promulgate regulations specifying
criteria for requiring disinfection "as necessary" for ground water systems.  The LT1FBR if approved
will be part of the first stage of the Enhanced Surface Water Treatment Rule, and will address recycling
requirements.

The Stage 1 DBPR (63  FR 69389, December 16, 1998) applies to all PWSs that are CWSs and
NTNCWs that treat their water with a chemical disinfectant for either primary or residual treatment. In
addition, certain requirements  for chlorine dioxide  apply to TNCWSs. The Stage 1 DBPR finalizes
maximum residual disinfectant level goals (MRDLGs) for chlorine, chloramines, and chlorine dioxide;
MCLGs for four trihalomethanes (chloroform, bromodichloromethane, dibromochloromethane, and
bromoform), two haloacetic acids (dichloroacetic acid and trichloroacetic acid), bromate,  and chlorite;
RIA for the ProposedLT1FBR                       7-16                               February 15, 2000

-------
and NPDWRs for three disinfectants (chlorine, chloramines, and chlorine dioxide), two groups of
organic disinfection byproducts TTHMs and HAAS and two inorganic disinfection byproducts, chlorite
and bromate.

As part of the first stage of the Enhanced Surface Water Treatment Rule, the Agency promulgated the
Interim Enhanced Surface Water Treatment Rule (IESWTR) in December 1998, in conjunction with
the Stage 1 DBPR The purposes of the IESWTR are to improve control of microbial pathogens,
specifically the protozoan Cryptosporidium, and address risk takeoffs between pathogens and
disinfection byproducts. The provisions of IESWTR only pertain to public water systems serving
10,000 or more people that use surface water or GWUDI. Key provisions of the rule include: a
Maximum  Contaminant Level Goal (MCLG) of zero for Cryptosporidium; 2 log Cryptosporidium
removal requirements for systems that filter; strengthened combined filter effluent turbidity
performance standards of 1.0 NTU as a maximum and 0.3 NTU at the 95th percentile monthly,
based on 4-hour monitoring for treatment plants using conventional treatment or direct filtration;
requirements for individual filter turbidity monitoring;  disinfection benchmark provisions to assess
the level of microbial protection provided as facilities take the necessary steps to comply with new
disinfection byproduct standards; inclusion of Cryptosporidium in the definition of ground water
under the direct influence of surface water and in the watershed control requirements for unfiltered
public water systems; requirements for covers on new finished water reservoirs; and sanitary
surveys for all surface water systems regardless of size.

The proposed turbidity monitoring, disinfection benchmarking, and covered finished reservoir
provisions  of the LT1FBR, parallel several IESWTR provisions, extending them to surface and
GWUDI systems serving fewer than 10,000 people. In addition, the LT1FBR recycling
provisions, which apply to all surface and GWUDI systems that recycle, are meant to control
pathogens in filter backwash that may increase as a result of changes in disinfection practices.

Minimization of Economic Burden

On an annual basis, the cost of the preferred alternatives is $87.6 to $97.5 million, depending on the
discount rate assumption. This combination of alternatives, designed to minimize the impact of the
proposed rule on small systems,  represents a cost savings of about 38 to 40 percent over the estimated
cost for alternatives that are similar to the IESWTR provisions. For the preferred alternative, EPA
streamlined monitoring requirements, which reduced annual monitoring costs.  The turbidity monitoring
costs are 17 to 19 percent of total costs under the preferred alternatives in contrast to 45 to 48 percent
under the alternatives that closely match the IESWTR provisions. Also by staggering the
implementation of the changes to the SWTR, smaller systems will gain from the experience of larger
systems on  how to most cost effectively comply with the LT1FBR Larger systems will generate a
significant amount of treatment and cost data from the IESWTR ICR and in their efforts to achieve
compliance with the IESWTR requirements. EPA intends to summarize this information and make it
available through guidance manuals. EPA believes this information will assist smaller systems in
achieving compliance with the LT1FBR.
February 15, 2000                                 7-17                     RIA for the Proposed LT1FBR

-------
7.1.4  Effect of Compliance With the LT1FBR on the Technical, Financial, and Managerial
       Capacity of Public Water Systems

Section 1420(d)(3) of the SDWA as amended requires that, in promulgating an NPDWR, the
Administrator shall include an analysis of the likely effect of compliance with the regulation on the
technical, financial, and managerial capacity of public water systems.  The following analysis fulfills this
statutory obligation. In EPA guidance (EPA 816-R-98-006) (U.S. EPA, 1998) the Agency defines
water system capacity as the ability to plan for, achieve, and maintain compliance with applicable
drinking water standards.  Capacity has three components: technical, managerial, and financial.

Technical capacity is the physical and operational ability of a water system to meet SDWA
requirements. Technical capacity refers to the physical infrastructure of the water system, including the
adequacy of source water and the adequacy of treatment,  storage, and distribution infrastructure. It
also refers to the ability of system personnel to adequately operate and maintain the system and to
otherwise implement requisite technical knowledge. Examining key issues and questions can determine
a water system's technical capacity, including:

            Source water adequacy. Does the system have a reliable source of drinking water? Is the
            source of generally good quality and adequately protected?

            Infrastructure adequacy.  Can the system provide water that meets SDWA standards?
            What is the condition of its infrastructure, including well(s) or source water intakes,
            treatment, storage, and distribution? What is the infrastructure's life expectancy? Does the
            system have a capital improvement plan?

            Technical knowledge and implementation. Is the system's operator certified? Does the
            operator have sufficient technical knowledge of applicable standards? Can the operator
            effectively implement this technical knowledge? Does the operator understand the system's
            technical and operational characteristics? Does the system have an effective operation and
            maintenance program?

Managerial capacity is the ability of a water system to conduct its affairs in a manner enabling the
system to achieve and maintain compliance with SDWA requirements. Managerial capacity can be
assessed through key issues and questions, including:

       •    Ownership accountability. Are the system owner(s) clearly identified?  Can they be held
            accountable for the system?

       •    Staffing and organization.  Are the system operators) and manager(s) clearly identified? Is
            the system properly organized and staffed? Do personnel understand the management
            aspects of regulatory requirements and system operations? Do they have adequate
            expertise to manage water system operations? Do personnel have the necessary licenses
            and certifications?
RIA for the ProposedLT1FBR                       7-18                               February 15, 2000

-------
           Effective external linkages. Does the system interact well with customers, regulators, and
           other entities? Is the system aware of available external resources, such as technical and
           financial assistance?

Financial capacity is a water system's ability to acquire and manage sufficient financial resources to
allow the system to achieve and maintain compliance with SDWA requirements. Financial capacity can
be assessed through key issues and questions, including:

           Revenue sufficiency.  Do revenues cover costs? Are water rates and charges adequate to
           cover the cost of water?

           Credit worthiness.  Is the system financially healthy? Does it have access to capital through
           public or private sources?

           Fiscal management and controls. Are adequate books and records maintained? Are
           appropriate budgeting, accounting, and financial planning methods used? Does the system
           manage its revenues effectively?

A total of 13,689 large and small systems are potentially subject to the proposed LT1FBR. Of these,
EPA estimates that 10,850 systems would need to take some action to come into compliance with the
rule. Much of the activities undertaken by these systems would be one time start-up activities
associated with reporting requirements. For example, there are 9,450 small systems that will need to
comply with the disinfection benchmark provision. Approximately 5,136 of these systems are among
the 5,896 small systems that would undertake turbidity monitoring are the only systems that would incur
annual monitoring costs, and of these 2,406 of them may need to modify their turbidity treatment
process. EPA estimates that the maximum number of systems that would be required to change their
recycling treatment practices is 1,172.  The expected number of small systems needing to comply with
the covered finished water provision is 580. Some large or small systems may require significantly
increased technical, financial, or managerial capacity to comply with these new requirements.

Systems needing to modify treatment will do so to meet turbidity or recycling provisions.  EPA
identified numerous process improvements that small systems might implement to improve finished
water quality to meet the proposed turbidity standards. The unit costs developed for these process
improvements are based on cost models, best engineering judgement, and existing cost and technology
documents. In most cases, cost estimates were derived by system size based on estimates of system
design and average flow rates. Exhibits 6-2a and 6-2b detail the capital costs and O&M costs for the
turbidity provisions. The Agency estimates turbidity capital and O&M costs for small systems to be
$62.4 million using a 7 percent discount rate.  The Cost and Technology Document for the Long
Term 1 Enhanced Surface  Water Treatment Filter Backwash Rule  (U.S. EPA, 1999b) describes in
detail the methods and assumptions used to develop unit costs.

Systems modifying treatment to meet recycling provisions may need to move recycle return to the plant
headworks or alter treatment processes. As noted in Exhibit 6-18,  the compliance forecast estimated
that 791 plants would need to move their recycling return flow to the head of the plant. Exhibits in

February 15, 2000                                 7-19                     RIA for the Proposed LT1FBR

-------
Appendix C detail the unit capital and O&M costs for direct and conventional filtration plants that need
to redirect theirs recycle flows.  Direct recycle plants would need to conduct self assessments and
report the findings to the State; subsequent requirements to alter treatment practices are at the
discretion of the State.  Appendix C reports the number of systems that would need to implement
treatment changes and the unit capital costs.  EPA estimates that approximately 359 direct recycle
systems potentially exceed their design capacity during recycling events and would need to alter
treatment. Direct filtration plants will report their recycling practices to the State which would
determine if treatment changes are necessary. EPA assumes that 23 direct filtration systems that
recycle would be required to install treatment. Appendix A summarizes the treatment process options,
the forecast of how many systems would need to alter treatment, and the unit cost for each option.
Annualized capital costs associated with the recycling provisions are approximately $13.6 million
(assuming a 7 percent discount rate) and the  O&M costs are approximately $10.2 million.

7.1.5  Paperwork Reduction Act

The information collected as a result of this rule will  allow the States and EPA to evaluate PWS
compliance with the rule. For the first 3 years after promulgation of this rule, the major information
requirements pertain to monitoring, and compliance reporting. Responses to the request for information
are mandatory (Part 141). The information collected is not confidential.

The Paper Work Reduction Act requires EPA to estimate the burden on PWS, States, and territories
for complying with the final rule.  Burden refers to the total time, effort, or financial resources expended
by persons to generate, maintain, retain, or disclose or provide information to or for a Federal agency.
This includes the time needed to review instructions;  develop, acquire, install, and utilize technology and
systems for the purposes of collecting, validating, and verifying information, processing and maintaining
information, and disclosing and providing information; adjust the existing ways to comply with any
previously applicable instructions and requirements;  train personnel to be able to respond to a collection
of information; search data  sources; complete and review the collection of information; and transmit or
otherwise disclose the information.

Information collection activities of PWSs required under this rule during the first 3 years following
promulgation will result in an average annual burden of 248,978 person-hours collectively for the
PWSs. For States and territories, the analysis suggests the annual average burden to be 62,508 hours.
The national, average annual labor cost of PWSs associated with information collection activities
averages $9.3 million annually.  The average annual costs incurred by PWSs for turbidimeter installation
would be $2.7 million in capital costs and $1.8 million in average O&M costs.  The annual cost for
laboratory analyses would be $115,000.  Nationally, States and territories will incur annual average
costs of $1.5 million.  Exhibit 7-10 presents a summary of the burden hours and costs associated with
the LT1FBR rule for the 3 years covered by the ICR.

The costs and burdens during ICR approval period do not reflect all the costs and burdens that would
result from the rule, as several provisions are  not scheduled for implementation until after the first  3
years.  The recycling provisions would be implemented in the fourth and fifth years following
promulgation. The total burden and costs to systems for the recycle provisions would be 157,768

RIA for the ProposedLT1FBR                        7-20                                 February 15, 2000

-------
hours and $5.7 million, respectively.  For States the recycling provisions would result in a 52,411 hour
burden and a $1.2 million cost.

          Exhibit 7-10.  Summary of Respondents, Responses, Burden, and Costs
                     for PWSs and States for the ICR Approval Period

PWSs
States
Total
Number
Respondents
Average
5,980
39
6,019
Number
Responses
Average
9,222
7,275
16,497
Average
Annual
Burden
248,978
62,508
311,486
Average
Annual Labor
Cost
$9,313,919
$1,513,000
$10,826,919
Average
Annual
Laboratory
Costs
$115,200
$0
$115,200
Average Annual
Turbidimeter Costs
Capital
$2,713,815
$0
$2,713,815
O&M
$1,783,395
$0
$1,783,395
Turbidity monitoring, the only annual activity that the proposed rule requires, would not begin until the
fourth year after promulgation. By the sixth year after promulgation all start-up activities would be
completed and the remaining compliance activity associated with the rule would be annual turbidity
monitoring.  In year 6 the annual burden and cost for all PWSs would be 274,274 hours and $9.0
million, respectively. Similarly, for States in the sixth year the burden and cost would be 219,286 hours
and $6.1 million. The derivation of all LT1FBR burdens and costs for start-up and annual information
collection activities can be found in LT1FBR Information Collection Request.

7.2    Impacts on Subpopulations

A primary purpose of the proposed LT1FBR is to improve control of microbial pathogens,
specifically the protozoan Cryptosporidium.  The health effect of cryptosporidiosis on sensitive
subpopulations is much more severe and debilitating than on the general population. Several
Subpopulations are more sensitive to cryptosporidiosis, including the young, elderly, malnourished,
disease impaired (especially those with diabetes), and a broad category of those with compromised
immune systems, such as ADDS patients, those with Lupus or cystic fibrosis, transplant recipients,
and those on chemotherapy (Rose, 1997).

Mortality as a result of cryptosporidiosis infection is a much greater risk for sensitive
subpopulations than it is for the general population, particularly for the immunocompromised.

       The duration and severity of the disease are significant: whereas the disease may
       hospitalize  1 percent of the immunocompetent population with very little risk of mortality
       (< 0.001), Cryptosporidium infections are associated with  a high rate of mortality in the
       immunocompromised (50 percent) (Rose, 1997).

The duration of cryptosporidiosis in those with compromised immune systems is considerably
longer than in those with competent immune systems, with more severe symptoms often requiring
lengthy hospital stays.  In those subpopulations, the cost of illness (COI) from cryptosporidiosis
would be much greater than for the general populace. During a 1993 outbreak in Milwaukee,  33
AIDS patients with Cryptosporidium accounted for 400 hospital days at an additional cost of
February 15, 2000
7-21
RIAfor the Proposed LT1FBR

-------
nearly $760,000 (Rose, 1997).  COI due to these hospital days alone is estimated at $23,000 per
case ($760,000/33 patients).

Because of the severity of illness and high costs for treatment experienced by sensitive subpopulations,
as a result of Cryptosporidium infection, the Agency expects LT1FBR to have a disproportionately
positive impact on the subpopulations mentioned earlier.

Protecting 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, 1997, or proposed after April 21, 1998,  that 1) is determined to be "economically significant"
as defined under Executive Order 12866 and 2) concerns an environmental health or safety risk that
EPA has reason to believe may have a disproportionate effect on children. If the regulatory action
meets both criteria, EPA must evaluate the environmental health or safety effects of the planned rule on
children,  and explain why the planned regulation is preferable to other potentially effective and
reasonably feasible alternatives considered by EPA.

In promulgating the LT1FBR EPA recognizes that the health risks associated with exposure to the
protozoan Cryptosporidium are of particular concern for certain sensitive subpopulations, including
children and immunocompromised individuals. Cryptosporidiosis acquired by the drinking water
exposure pathway can spread quickly among children in group settings, especially diapered children in
day care  centers (Juranek, 1998).  This type of transmission is called secondary transmission, and can
result in spread of the disease to both children and adults. Evidence of such secondary transmission
of Cryptosporidiosis from children to household and other close contacts has been found in many
outbreak investigations (Casemore, 1990;  Cordell, et al., 1997; Frost, et al., 1997).  During the
1993 Milwaukee outbreak, 74 percent of day care centers interviewed reported Cryptosporidiosis
among children or staff members, although only 3.4 percent of facilities closed because of the epidemic
(Cordell, et al., 1997). Having a child under 5 years of age in a household was a risk factor (i.e., with a
matched  odds ratio of 17) among laboratory-confirmed Cryptosporidiosis cases during the post-
outbreak period in Milwaukee (Osewe, et al., 1996).

Malnourished and immunocompromised children are at greater risk of developing chronic  diarrhea
when infected with Cryptosporidium due to impaired immune response (Griffiths, 1998, Casemore,
1990).  The progression from acute Cryptosporidiosis to chronic diarrhea and death among
malnourished children is not well understood (Griffiths, 1998). However, information on mortality
from diarrhea  shows the greatest risk of mortality occurring among the very young and elderly
(Gerba, et al.,  1996).  Specifically, young children are a vulnerable population  subject to infectious
diarrhea  caused by Cryptosporidium (CDC, 1994). Cryptosporidiosis is prevalent worldwide, and
its occurrence is higher in children than in adults (Payer and Ungar, 1986).  Moreover,
Cryptosporidiosis appears to be more prevalent in populations that may not have established
immunity against the disease and may be in greater contact with environmentally contaminated
surfaces, such as infants (DuPont, et al., 1995).  Once a child is infected, it may spread the disease
to other children or family members.
RIA for the ProposedLT1FBR                       7-22                               February 15, 2000

-------
These concerns were considered as part of the regulatory development process, particularly in the
establishment of the MCLG at zero for Cryptosporidium in drinking water established under
IESWTR.  The proposed LT1FBR continues to take into account the need to protect sensitive
populations (e.g., children) and provide for an adequate margin of safety. For public water systems
that use surface water, filter, and serve less than 10,000 people, EPA is establishing physical removal
treatment requirements to control for Cryptosporidium.  For systems that use conventional, direct, or
membrane filtration, EPA is strengthening the existing turbidity standards for finished water and is also
requiring individual filter monitoring for conventional and direct filtration systems to assist in controlling
pathogen breakthrough during the treatment process.

7.3    Environmental Justice

Executive Order 12898 (59 FR 7629) establishes a Federal policy for incorporating environmental
justice into Federal agency missions by directing agencies to identify and address disproportionately
high and adverse human health or environmental effects of its programs, policies, and activities on
minority and low-income populations.  For example, to assist in identifying the need for ensuring
protection of populations who principally rely on fish or wildlife for subsistence, the Executive Order
directs agencies, whenever practicable and appropriate, to collect, maintain, and analyze information on
the consumption patterns of those populations, and to communicate to the public the risks of those
consumption patterns.

The Agency has considered environmental justice related issues concerning the potential impacts of this
proposed action and has consulted with minority and low-income stakeholders. Three aspects of the
rule comply with the Environmental Justice Executive Order and they  can be classified as follows:  1)
The overall nature of the rule;  2) the inclusion of sensitive subpopulations in the regulatory development
process; and 3) the convening of a stakeholder meeting specifically to address environmental justice
issues. The overall nature of the LT1FBR rule mimics the 1998 IESWTR by regulating small public
water systems to improve control of microbial pathogens. Therefore, the minority and impoverished
populations served by small systems will realize the health protection benefits currently provided to
populations served by larger systems.  The Water Industry Baseline Handbook provides limited
information concerning the distribution of impoverished households served by small systems.

In addition, the proposed rule  includes concerns of the sensitive sub-populations through the Reg. Neg.
and M-DBP Advisory Committee process undertaken to  craft the regulation. Both Committees were
chartered under the FACA authorization, and included a  broad cross-section of regulators, regulated
communities, industry, public  interest groups, and State and local public health officials.
Representatives of sensitive subpopulations, in particular people with ADDS, participated in the
regulatory development process.  Extensive discussion on setting treatment requirements that provide
the maximum feasible protection took place, and the final consensus that resulted in the rule considered
issues of affordability, equity,  and safety.

Finally, as part of EPA's responsibilities to comply with Executive Order 12898, the Agency held a
stakeholders' meeting to address various components of pending drinking water regulations; and how
they may impact sensitive subpopulations, minority populations, and low-income populations. Topics
February 15, 2000                                 7-23                      RIA for the Proposed LT1FBR

-------
discussed included treatment techniques, costs and benefits, data quality, health effects, and the
regulatory process. Participants included national,  State, Tribal, municipal, and individual stakeholders.
The major objectives for the meeting were:

        •    Solicit ideas from environmental justice stakeholders on known issues concerning current
            drinking water regulatory efforts

            •Identify key issues of concern to stakeholders

            •Receive suggestions from stakeholders concerning ways to increase representation of
            communities in OGWDW regulatory efforts.

Furthermore, EPA developed a plain-English guide specifically for this meeting to assist stakeholders in
understanding the multiple and sometimes complex issues surrounding drinking water regulation.
RIA for the ProposedLT1FBR                        7-24                                 February 15, 2000

-------
                     8. Weighing the Benefits and the Costs

This chapter compares the benefit estimates discussed in Chapter 5 with the cost estimates
discussed in Chapter 6. The comparison method used is an estimate of net benefits, which is the
difference between benefits and costs. Most of the potential range of quantifiable benefits, which
exclude many types of benefits discussed in Chapter 5, exceeds the cost estimates associated with
the preferred alternatives. Consequently, even though there are costs associated with the proposed
rule, EPA expects the benefits to exceed costs such that the net benefits will be positive.

8.1    Incremental and Marginal Analysis

EPA estimated the incremental benefits and costs of the proposed rule, which are the benefits and
costs expected to accrue compared to a baseline in which the provisions of the rule are not
implemented. Both benefits and costs are expressed as annual values.  Annual benefits are
monetized estimates of avoided morbidity or mortality cases.  Annual costs include annualized
capital investments to alter treatment processes and annual O&M expenditures. They also include
annualized start-up cost estimates for all of the provisions of the proposed rule.

EPA compared the costs of the proposed CFE requirement (i.e., 0.3 NTU 95th percentile) with the
costs that would be associated with more stringent requirements of 0.2 NTU and 0.1  NTU.  For
these two sensitivity tests, EPA developed assumptions about which process changes systems
might implement to meet the requirement and how many systems would adopt each change.  The
decision trees in Appendix C summarize these assumptions, which were inputs to the cost model
that calculated total annual costs.  The comparison of total compliance cost estimates  in Chapter 6
showed that marginal costs of more stringent requirements rise rapidly.  Compared to the proposed
0.3 NTU requirement, the turbidity treatment costs of the 0.2 NTU sensitivity  test were 157
percent higher and the costs of the 0.1 NTU case were 675  percent higher.

8.2    Benefit-Cost Comparisons

EPA compared annual benefits and costs by calculating net benefits, which is the difference
between the two.  The assessment of benefits was based on a health risk assessment that
characterized the scientific uncertainty regarding the exposure assessment and  health hazards
associated with exposure to Cryptosporidium through drinking water.  The benefits analysis used
Monte Carlo simulations to derive a distribution of estimates, rather than a single point estimate.

Exhibit 8-1 summarizes the mean expected value of potential annual benefits for the turbidity
provisions under both baselines, the three improved removal assumptions, and two daily  consumption
distributions. Assuming a 2.0 log removal baseline for small systems, benefits range from $195.3
million to $259.4 million.  Alternatively, if a 2.5 log removal baseline is more representative of current
small system performance, the benefits are less, ranging from $70.1 million to $92.4 million. The range
across these assumptions, is therefore, $70.1 million to $259.4 million.  The corresponding mean
estimates of avoided illnesses, which account for almost 80 percent of benefits, range from
approximately 22,800  cases to 83,600 cases per year.
February 15, 2000                                8-1                     RIA for the Proposed LT1FBR

-------
  Exhibit 8-1.  Summary of Annual Mean Benefits Associated with Avoided Illnesses and
    Mortalities for the Turbidity Provisions of the Proposed Rule (January 1999 dollars)
Improved Log Removal Assumption
Low Removal
Mid Removal
High Removal
Baseline Cryptosporidium Log-Removal Assumptions
($ millions)
Mean = 1.2 Liters per person
2.0 log Baseline
$195. 3
$240.8
$259.4
2.5 log Baseline
$70.1
$86.1
$92.4
Exhibit 8-2 summarizes the annual cost of the rule at the 3 percent and 7 percent discount rates for
two combinations of alternatives. Annual costs for the combination of preferred options are $87.6
or $97.5 million depending on the discount rate, and they are $145.5 or $157.3 million for the
combination of alternatives that closely resembles provisions in the IESWTR.  This comparison
shows that the proposed rule reduces total cost by 38 to 40 percent compared to what costs would
be without tailoring the turbidity and disinfection benchmark provisions to meet the needs and
circumstances of small systems while fulfilling public health protection goals that are comparable
to the IESWTR.

          Exhibit 8-2. Total Annual Costs for Two Combinations  of Alternatives
                                      (January  1999)
Compliance
Activity
Turbidity Provisions
Disinfection Benchmarking
Covered Finished Storage
Recycle Provisions
Total Costs
Preferred Alternatives
(T2, A4, B2 with R2)
(S millions)
3%
$63.4
$1.3
$2.5
$20.4
$87.6
7%
$68.6
$1.8
$2.6
$24.5
$97.5
IESWTR Alternatives
(Tl, Al, Bl with R2)
($ millions)
3%
$116.7
$5.9
$2.5
$20.4
$145.5
7%
$121.9
$8.3
$2.6
$24.5
$157.3
 Detail may not add to total due to independent rounding.

The bottom bar in Exhibit 8-3 shows the range of expected benefits from $70.1 million to $259.4
million.  The middle bar shows the range between the two total cost estimates for the preferred
alternatives, $87.6 million and $97.5 million. Finally, it plots a range of potential net benefits,
which compares the quantified benefits for the turbidity provisions with the total  costs for all
provisions. The lowest potential net benefit is the difference between the lowest  benefit estimate
and the highest cost estimate (i.e., $70.1 million - $97.5 million = -$27.4 million). The highest
potential net benefit is the difference between the highest benefit estimate and the lowest cost
estimate (i.e., $259.4 million - $87.6 million = $171.8 million).  As the chart shows, the range of
potential net benefits from negative $27.4 million to $171.8  million lies primarily to the right of
zero, showing that benefits are likely to exceed costs.
RIAfor the ProposedLT1FBR
February 15, 2000

-------
         Exhibit 8-3.  Comparison of Annual Benefit, Cost, and Net Benefit Ranges
                              (January 1999 dollars, millions)
        Net Benefits
               Costs
             Benefits
                     -$50    $0     $50   $100   $150  $200  $250  $300
Although monetized net benefits may be as low as negative $27.4 million, total social benefits are
likely to be positive if the qualitative benefits are taken into consideration. Exhibit 8-4 shows that
although costs have been estimated for most of the provisions, benefits have not. Benefits were
only estimated for the turbidity provisions. Furthermore, the benefits that were quantified for the
turbidity provisions exclude some benefits of those provisions such as reducing exposure to other
pathogens (e.g., Giardia lamblia) and avoiding the cost of averting behavior.  The nonquantified
benefits could represent substantial additional economic value. Overall, EPA expects the proposed
rule to improve health protection for more than 18 million households, which are the households
served by systems EPA estimates will alter treatment or practices under the rule.  If the aggregate
benefit per household for these nonquantified benefits is at least $1.52 per year, then even the low
range of net benefits will be positive.
February 15, 2000
S-3
RIAfor the Proposed LT1FBR

-------
             Exhibit 8-4.  Summary of Benefit and Cost Analysis Completeness
                                  (January 1999, millions)

Compliance
Activity
Small System Provisions Comparable to IESWTR:
• Turbidity Provisions
• Disinfection Benchmarking
• Covered Finished Storage
• Cryptosporidium in GWUDI Definition
• Cryptosporidium in Watershed Requirements
Subtotal forLTl Provisions
Recycle Provisions
Total Costs
Range of Costs for
Preferred Alternatives
(T3, A4, B2, R2)

$63.4 -$68.6
$1.3-$1.8
$2.5 -$2.6
not estimated
not estimated
$67.2 -$73.0
$20.4 -$24. 5
$87.6 - $97.5

Range of
Benefits

> $70.1 -$259.4
not estimated
not estimated
not estimated
not estimated
> $70.1 -$259.4
not estimated
> $70.1 -$259. 4
Under the 1996 Amendments to the Safe Drinking Water Act, EPA is required to make a
determination of whether benefits justify costs for the rulemaking. EPA has determined that the
benefits of the LT1FBR justify the costs. EPA made this determination for both the LT1 (i.e., the
five provisions in the first row of Exhibit 8-4) and the FBR (i.e., the recycle provisions) portions
of the rule separately as described below.

EPA has determined that the benefits of the LT1 provisions justify their costs on a quantitative
basis. The LT1 provisions include enhanced filtration, disinfection benchmarking, and other
nonrecycle-related provisions.  The quantified benefits of $70.1  million to $259.4 million annually
exceed the annual cost range of $67.2 million to $73.0 million over a substantial portion of the
range of benefits. In addition, the nonquantified benefits include avoided outbreak response costs,
avoided costs of averting behavior, and reduced uncertainty about drinking water quality.

For the recycle provisions, the Agency has determined that the benefits  of the provisions justify
their cost on a qualitative basis although the analysis in the following section provides a
quantitative perspective on the health benefits  needed to break even given EPA's cost estimates.
The recycle provisions will reduce the potential for certain recycle practices to lower or upset
treatment plant performance during recycle events.  Therefore, the provisions will help prevent
Cryptosporidium oocysts from entering finished drinking water supplies and will increase the level
of public health protection.

Returning Cryptosporidium to the treatment process in recycle flows, if performed in a manner that
is inconsistent with fundamental engineering and water treatment principles, can increase public
health risks.  EPA believes that there are three instances, in particular, that increase health risks.
 First, returning recycle flow directly to the plant—without equalization or treatment—can cause
large variations in the influent flow magnitude and the influent water quality. If chemical dosing is
not adjusted to reflect these variations, then less than optimal chemical dosing can occur, which
may lower sedimentation and filtration performance. Returning recycle flows prior to the point of
RIAfor the ProposedLT1FBR
8-4
February 15, 2000

-------
primary coagulant addition will help diminish the likelihood of this occurring.  Second, exceeding
State-approved operating capacity, which is more likely to occur if recycle equalization or
treatment is not in place, can hydraulically overload plants and diminish the ability of individual
unit processes to remove Cryptosporidium. Exceeding approved operating capacity violates
fundamental engineering principles and water treatment objectives. States set limits on plant
operating capacity and loading rates for individual unit processes to ensure that treatment plants
and individual treatment processes operate within their capabilities and, thereby, provide the
necessary levels of public health protection.  Third, returning recycle flows directly into
flocculation or sedimentation basins can generate disruptive hydraulic currents, which lower the
performance of these units and increase the risk Cryptosporidium will be present in finished water
supplies.

The objective of the recycle provisions is to eliminate practices that are counter to common sense,
sound engineering judgement, and that create additional and preventable risk to public health.
EPA's proposed rule addresses these practices while providing States and affected systems with
the flexibility necessary to implement the most cost-effective solutions. Consequently, EPA
believes the public health protection benefit provided by the recycle provisions justifies their cost.

8.3   Breakeven Analysis for the Recycle Provisions

The LT1FBR recycle provisions are expected to improve recycle practices, thereby preventing the
accumulation of Cryptosporidium within the treatment plant and minimizing the risk of oocysts entering
into the finished water. Estimating the magnitude of the risk posed by existing recycle practices and the
risk reduction resulting from the recycle provisions, however, is complicated by the lack of scientific
data to support assumptions and analyses. Given the large population served by systems that are
potentially subject to the proposed rule (66.8 million—the total population served by all rapid granular
filtration systems that practice recycle), even a small risk reduction could have a substantial impact. To
assess the costs of the recycle provisions against the possible range of risks, EPA used a breakeven
analysis to explore net benefits of the alternatives.  Breakeven analysis represents an approach to
assessing the benefits of the  recycle provisions given the scientific uncertainties surrounding the risk
posed by recycling practices.

Breakeven is a standard benchmark of cost effectiveness and  economic efficiency and is essentially the
point where the benefits of the recycle provisions would be equal to the costs. Normally, the benefits
and costs of an  option are calculated separately and then compared to assess whether and by what
amount benefits exceed costs. In the case of the recycle provisions, independently estimating benefits is
difficult, if not impossible, because of the uncertainty surrounding the risk and resulting risk reduction.
Instead, the breakeven analysis works backwards from those variables that are less uncertain.  In this
case, implementation costs for the rule and the monetary value associated with the health endpoints are
used to calculate what risk reduction estimates are needed for the rule to just pay for itself in avoided
health damages associated with cryptosporidiosis.

The first step in the breakeven analysis is to calculate the number of cryptosporidiosis cases that would
need to be avoided for the benefits of avoiding those cases to be equal to the cost of the rule. The
simple calculation is to divide the annual costs of the rule by the value per cryptosporidiosis case to

February 15, 2000                                  8-5                       RIA for the Proposed LT1FBR

-------
derive the number of cryptosporidiosis cases needed to cover the costs of the rule.  The value of a
cryptosporidiosis case differs based on whether the case results in illness or fatality.  Fatal cases are
valued at the VSL with a mean of $5.7 million as mentioned earlier. It is assumed that .0125
percent of all cryptosporidiosis cases are fatal and are, therefore, assigned the VSL.

For the cryptosporidiosis illnesses (nonfatal), comprising 99.9875 percent of the total, the cost of illness
estimate of $2,400 was used as a valuation. It is important to note that this value  reflects the potential
COI avoided, not the full WTP to reduce the probability  of suffering a cryptosporidiosis infection. The
estimates do not take into account the value of avoiding pain and suffering, the economic premium
associated with risk aversion, or the costs  of averting behaviors.  Therefore the full value of the
economic benefit to reduce cryptosporidiosis may be higher than the $2,400 COI avoided per case
mean estimate.

The average cost per cryptosporidiosis case avoided is approximately $3,100 [($5.7 million x
0.000125) + ($2,400 x 0.999875). This value is divided into the implementation costs for the
various rule alternatives to estimate the  number of cryptosporidiosis cases reduced needed to break
even (Exhibit 8-5). For the preferred alternative at a 7 percent rate, the recycle provisions would
have to prevent 5,600 to 7,900 illnesses annually to break even.

                        Exhibit 8-5. Breakeven Analysis Summary

Implementation cost
($ millions)
Total number of
cases prevented to
break even
Fatal cases
Nonfatal cases
3% Cost of Capital
Alternative R2
Low
$14.3
4,600
1
4,599
High
$20.4
6,600
1
6,599
7% Cost of Capital
Alternative
Rl
$16.9
5,500
1
5,499
Alterative R2
Low
$17.4
5,600
1
5,599
High
$24.5
7,900
1
7,899
Alterative R3
Low
$55.0
17,700
3
17,697
High
$56.7
18,300
3
18,297
Alternative
R4
$151.8
49,000
7
48,993
The breakeven number of cases provides only part of the information needed to assess under what
conditions the recycle provisions will break even. Two other factors, the baseline number of
cryptosporidiosis illnesses and the percent reduction in risk due to the provisions, combine to give
the number of illnesses avoided by the rule. In general, these two factors have an inverse
relationship with respect to the breakeven point:  the higher the baseline number of illnesses, the
lower the reduction needs to be to break even. Conversely, the lower the baseline number of
cases, the higher the reduction in risk needs to be.

The baseline number of illnesses can be estimated using the risk characterizations performed for
the IESWTR (for systems serving 10,000 and over) and the turbidity provisions of the proposed
LT1FBR (for systems serving less than 10,000). The baseline number of illnesses remaining after
the implementation of the LT1FBR and IESWTR range from a low estimate of 62,000 to a high
estimate 344,600 (Exhibit 8-6).
RIAfor the ProposedLT1FBR
8-6
February 15, 2000

-------
                    Exhibit 8-6.  Combined Annual Number of Illnesses
                         Remaining After LT1FBR and IESWTR
Improved Log
Removal Assumption
Low Improved Removal
Mid Improved Removal
High Improved Removal
Baseline Assumption (LT1FBR/IESWTR)
2.0/2.5
344,600
235,900
199,900
2.0/3.0
139,600
92,900
75,900
2.5/2.5
317,200
218,100
186,000
2.5/3.0
1 12,200
75,100
62,000
Exhibit 8-7 contains a graph that addresses the question: "Given the range of baseline risk, what
risk reduction due to the recycle provisions would you need to reach the breakeven illnesses in
Exhibit 8-5?"

Exhibit 8-7 displays three sets of breakeven lines assuming a 7 percent discount rate, one for
Alternative R4, and two lines each for Alternatives R2 and R3; the line for Alternative Rl lies on
top of the low line for Alternative R2 and is therefore not shown in the graph. Each point on the
line represents the combination of baseline illnesses and percent reduction needed to produce the
breakeven cases. For Alternative 4, the combination of baseline number of illnesses and percent
reduction at each point on the line produces 56,200 cases avoided.  Graphically, at any
combination of baseline risk and percent reduction in the area to the right of the lines, the recycle
provisions would exceed a breakeven point (i.e., benefits would exceed costs) and at any
combination in the area to the left of the lines, the costs of compliance would exceed the benefits.

Alternative R4 is expected to result in the greatest reduction in risk. If the estimated baseline risk is
low (below about 115,000 cases), Alternative R4 would need to reduce baseline cases by more
than 50 percent to break even. At the highest baseline risk, Alternative R4 would need to result in
about a 15 percent reduction in risk to break even.

For Alternative R3, if the baseline risk is low (115,000 cases), the rule would need to reduce 20
percent of cases to breakeven. At the highest baseline risk, Alternative R3 would need to result in
about a 5 percent reduction in risk to break even.

For Alternative R2,  the preferred alternative, if the baseline risk is low (115,000 cases) the rule
would need to reduce only about 6 percent to 9 percent of the risk to break even.  At the high end
of baseline risk, the rule would break even at a minimal reduction in risk (about 2 percent). At the
mid range of baseline risk, the required reduction is under 5 percent. EPA selected Alternative R2
because it was felt to have a greater likelihood to be able to achieve reductions in risk due to
Cryptosporidium that would break even and justify the costs.
February 15, 2000
8-7
RIAfor the Proposed LT1FBR

-------
     Exhibit 8-7. Percent Reduction in Illnesses Needed to Break Even by Alternative

100%
90% -
80% -
1 70%"
O /-no/
OU/0
g 50% -
1 40% -
""O
& 30% -
20% -
10% -
\
>
	 v 	
\
	 \ 	
k
\
>
«
>

R2 Low: 6,400 cases
R2 High: 9,100 cases
	 R3 Low: 20,400 cases
	 R3 High: 21,000 cases •
— - -R4: 56,200 cases

w ^
^ ^-^
^ ^ 	
^^^1^ w"*- 	 	 , 	 " • 	
~~ " 	

0"o H 	 1 	 1 	 1 	 1 	 1 	 1 	
50,000 100,000 150,000 200,000 250,000 300,000 350,000
Baseline Cases
8.4    Summary of Uncertainty/Sensitivity Analysis

The results of the benefit and costs analyses were based on several assumptions that introduce
uncertainties. Areas where important sources of uncertainty enter the benefits assessment include
the following:

       •   Occurrence of Cryptosporidium oocysts in source waters

       •   Baseline occurrence of Cryptosporidium oocysts in finished waters

       •   Reduction of Cryptosporidium oocysts due to improved treatment, including filtration
           and disinfection

       •   Viability of Cryptosporidium oocysts after treatment

       •   Infectivity of Cryptosporidium

           Incidence of infections (including impact of under reporting)

           Characterization of the health risk

       •   Willingness-to-pay to reduce risk and avoid costs.

The benefit analysis incorporates all of these uncertainties in either the Monte Carlo simulations or
the assumption of two baselines—2.0 log removal and 2.5 log removal—as discussed in Chapter
5. The results in Exhibit 8-1  show that benefits are more sensitive to the baseline log removal and
RIAfor the ProposedLT1FBR
8-8
February 15, 2000

-------
the daily water consumption assumptions than the range of low to high improved removal
assumptions.  Other unquantified benefits stemming from illnesses avoided from other rule
provisions were also discussed in Chapter 5.

Cost analysis uncertainties are primarily caused by baseline assumptions made about how many
systems will be affected by various provisions and how they will likely respond.  Capital and
O&M expenditures account for a majority of total costs. EPA derived these costs for a "model"
system in each size category using engineering models, best professional judgement, and existing
cost and technology documents.  Costs for systems affected by the proposed rule could be higher
or lower, which would affect total costs.

8.5    Combined Regulatory Effects with Other Rules

The proposed LT1FBR is one of several rules that potentially affect surface water and GWUDI
drinking water systems. Exhibit 8-8 summarizes recent and forthcoming rules.  Thus far, EPA's
cost estimates suggest that costs associated with small systems will total $59 million under the
existing IESWTR and Stage 1 DBPR, and could increase by another $73 million under the
proposed rule assuming a 7 percent discount rate. Costs associated with larger systems are
substantially larger, totaling $569 for the existing rules, and may reach $587 million including costs
for the proposed LT1FBR. Adding costs across rules may over state the social costs of regulatory
activities because actions to satisfy one rule may reduce the costs of subsequent rules that have
related public health goals.
   Exhibit 8-8. Annualized Costs for Recent and Forthcoming Rules that Affect Surface
                      Water and GWUDI Drinking Water Systems
Rule
IESWTR (12/98)
Stage 1 DBPR (12/98)
LT1FBR (11/00)
Stage 2 DBPR (11/02)
Stage 2 LTESWTR (05/02)
Total Costs
Small Systems
(serving < 10,000 customers)
$2 million
$57 million
$73 million
yet to be determined
yet to be determined
$132 million
Large Systems
(serving • 10,000 customers)
$291 million
$278 million
$18 million
yet to be determined
yet to be determined
$587 million
February 15, 2000
8-9
RIAfor the Proposed LT1FBR

-------
                                   9.  References

Abdalla, C.W. 1990. "Measuring Economic Losses from Ground Water Contamination: An
       Investigation of Household Avoidance Costs."  Water Resources Bulletin 26: 451-63.

Abdalla, C.W., B.A. Roach, and DJ. Epp.  1992.  "Valuing Environmental Quality Changes
       Using Averting Expenditures: An Application to Groundwater Contamination." Land
       Economics 68: 163-69.

American Water Works Association (AWWA).  1998.  Survey ofAWWA Systems and Recycle
       Practices.

American Water Works Association (AWWA) Committee Report.  1983. Deterioration of Water
       Quality in Large Distribution Reservoirs (Open Reservoirs). AWWA Committee on
       Control of Water Quality in Transmission and Distribution Systems. Journal of American
       Water Works Association:  75(4):  191-195.

AWWA/ASCE (American Water Works  Association; American Society of Civil Engineers).
       1998. Water Treatment Plant Design, 3rd edition. New York: McGraw Hill.

Archer, J.R., J.R. Ball, J.H. Standridge, S.R. Greb, P.W. Rasmussen, J.P. Masterson, and L.
       Boushon. 1995. Cryptosporidium spp. oocysts and Giardia spp. cyst occurrence,
       concentrations and distribution in  Wisconsin waters. Wisconsin Department of Natural
       Resources. August.  PUBL-WR420-95.

Bellamy W., J. Cleasby, G. Logson, and M. Allen. 1993. Assessing Treatment Plant Performance.
       J. AWWA (85:12:34-38).

Craun, G.F., S. Hubbs, F.J. Frost, R. Calderon, and S. Via. 1998. Waterborne outbreaks of
       cryptosporidiosis. J. AWWA (90:9:81-91).

Bennett J.V., S.D. Holmberg, M.F. Rogers, and S.L. Solomon. 1987.  Infectious and parasitic
       diseases. AM J Prev Med #: 102-113. In: R.W. Ambler and H.B. Dull (eds). Closing the
       gap: the burden of unnecessary illness. Oxford: Oxford University Press.

Bureau of Labor Statistics.  1999a.  Employer Costs for Employee Compensation Summary-
       March 1999.  accessed July 13, 1999.

Bureau of Labor Statistics.  1999b.  1998 1999  Occupational Outlook Handbook. Washington,
       DC:  U.S. Government Printing Office.  Also available at
       

Casemore, D.P. 1990. "Epidemiological aspects of human cryptosporidiosis." Epidemiol. Infect.
       104:  1-28.
February 15, 2000                              9-1                    RIA for the Proposed LT1FBR

-------
Centers for Disease Control (CDC). 1996. Surveillance for Waterborne-Disease
       Outbreaks-United States, 1993 1994. 45(SS-1)

CDC. 1994. Addressing Emerging Infectious Disease Threats: A Prevention Strategy for the
       United States. Executive Summary.

Chappell, C.L., H.L. DuPont, P.C. Okhuysen.  1997.  Virulence Factors in Cryptosporidium and
       Infective Dose in Humans. Progress Report for Grant R8 2-4 759.  Progress Report for
       Grant R82-4759, U.S. EPA to University of Texas-Houston School of Public Health,
       Center for Infectious Diseases.

Chestnut, L.G., and A. Alberini. 1997. Monetary Valuation of Human Mortality Risks in Cost-
       Benefit Analyses of Environmental Programs:  Background Paper and Bibliography.

Cleasby,  J.  1990. Filtration, Chapter 8, IN: (F. Pontius, ed) Water Quality and Treatment.
       AWWA, Denver, 14pp.

Cleasby,  J.L., M.M. Williamson, and E.R. Baumann. 1963. "Effect of Filtration Rate Changes on
       Quality." J. AWWA (55:7:869-880).

Colbourne, J. 1989. Thames Water Utilities Experience with Cryptosporidium. AWWA Water
       Quality Technology Conference, pp. 275-278.

Conley, R.  1965. "Integration of the Clarification Process." Journal of the American Water
       Works Association: 57(10):  133

Conrad, L.  1997. Personal communication from Larry Conrad of the Pennsylvania Department of
       Environmental Protection, North Central Region, Willamsport, PA to Henry Willis of
       Science Applications International Corporation, Middleborough, PA. September 11, 1997.

Consonery, P.J., D.N. Greenfield, and JJ. Lee.  1997. Evaluating and optimizing surface water
       treatment plants: How good is good enough? Pennsylvania Department of Environmental
       Protection, Harrisburg, PA.

Consonery  P.J., M.G. McGowan, and R.L. Diehl. 1992. Preliminary Results of Cryptosporidium
       Analysis at Filtered and Unfiltered Public Water Supplies, Study Period May 1990 through
       May 1992. Pennsylvania Department of Environmental Resources.

Cooke, G.D. and R.E. Carlson. 1989. Reservoir Management for Water Quality and THM
       Precursor Control. AWWARF. Denver, CO.

Cordell, R.L., and D.G. Addiss. 1994.  Cryptosporidiosis in child care settings: a review of the
       literature and recommendations for prevention and control. Pediatr. Infect. Dis. Jour.
       13(4): 310-317.
RIA for the ProposedLT1FBR                       9-2                              February 15, 2000

-------
Cordell, R.L., P.M. Thor, D.G. Addiss, J. Theurer, R. Lichterman, S.R. Ziliak, D.D. Juranek, J.P.
       Davis. 1997. "Impact of a massive waterborne cryptosporidiosis outbreak on child care
       facilities in metropolitan Milwaukee, Wisconsin." Pediatr. Infect. Dis. J.  16 (7): 639- 644.

Cornwell, D. 1997. Treatment of Recycle and Backwash Streams. Water Residuals and Biosolids
       Management: WEF/AWWA.

Cornwell, D. and R. Lee.  1994.  Waste Stream Recycling: Its Effects on Water Quality, Journal
       of the American Water Works Association. 86: 50-63.

Cornwell, D., and R. Lee. 1993a. Impacts of Recycle Streams on Finished Water Quality. 1993
       Annual Conference Proceedings, American Water Works Association, Water Quality. San
       Antonio, TX, June 6-10, 1993. AWWA Paper No. 35-4, pp. 611-630.

Cornwell, D., and R. Lee. 1993b. Recycle Stream Effects on Water Treatment. Sponsored by
       AWWA Research Foundation, Denver, CO.  Published by the AWWA Research
       Foundation and American Water Work Association.  153 pp.

Craun, G.F.  1998. Waterborne Outbreaks 1995-96. Memorandum to Valerie Blank, U.S. EPA
       OGWDW, June 20.

Craun, G.F. 1996. Task 1. Cases of illness required to initiate a waterborne outbreak investigation.
       September 18. Task 2. Epidemiological and water supply information from waterborne
       outbreaks of giardiasis and levels of cysts detected in water during outbreaks. September
       23. Task 3. Epidemiological and water supply information from waterborne outbreaks of
       cryptosporidiosis and levels of oocysts detected in water during outbreak investigations.
       September 20. Task 4. Waterborne disease in the USA: reported waterborne outbreaks,
       1993-94. September 27. 68-C3-0368. Wade Miller Associates, Inc.

Craun, G.F., and R. Calderon.  1996. "Microbial Risk in Groundwater Systems: Epidemiology of
       Waterborne Outbreaks. Under the Microscope: Examining Microbes in Groundwater.
       Proceedings of the Groundwater Foundation's 12th Annual Fall Symposium, Boston,
       September.

Craun, G.F., R.  Calderon, and F.J. Frost. 1996. An introduction to epidemiology. J. AWWA
       88(9): 54-65.

Craun, G.F., S. Hubbs, F.J. Frost, R. Calderon, and S. Via. 1998. Waterborne outbreaks of
       cryptosporidiosis. J. AWWA (90:9:81-91).

D'Antonio R.G., R.E. Winn, J.P. Taylor, T.L. Gustafson, W.L. Current, M.M. Rhodes, G.W.
       Gary, and R.A. Zajac. 1985. A waterborne outbreak of cryptosporidiosis in normal hosts.
       Ann. Intern. Med. 103: 886-888.

Dugan, N., K. Fox, R. Miltner, D. Lytle, D. Williams, C. Parrett, C. Feld, and J. Owens. 1999.
       "Control of Cryptosporidium Oocysts by Steady-State Conventional Treatment".

February 15, 2000                               9-3                    RIA for the Proposed LT1FBR

-------
       Proceedings of the U.S. Environmental Protection Agency 6th National Drinking Water
       and Wastewater Treatment Technology Transfer Workshop, Kansas City, MO (August 2-
       4, 1999), 19pp.

DuPont, H.L., C.L. Chappell, C.R. Sterling, P.C. Okhuysen, J.B. Rose and W. Jakubowski.
       1995. "Infectivity of Cryptosporidium parvum in Healthy Volunteers." New England
       Journal of Medicine: 332:855-859.

Edzwald, J.K., M.B. Kelley. 1998. "Control of Cryptosporidium: From Reservoirs to Clarifiers to
       Filters." Water Science Technology:  37:1*8.

Environmental Engineering &  Technology, Inc (EE&T). 1999. State-of-the-art papers for various
       potential recycle streams. February 9.

EE&T. 1998. (chapter 4 exhibits) Environmental Engineering and Technology. 1998. Presentation
       to Technical Working Group.

Payer, R. andB.L.P.Ungar. 1986. Cryptosporidium spp. and cryptosporidiosis. Microbiol. Rev.
       50(4): 458-483.

Framm, S.R., andR. Soave. 1997. Agents of diarrhea. Med. Clin. N. Amer. 81(2): 427-447.

Freeman, A.M.  1979. "The Benefits of Environmental Improvement: Theory and Practice."
       Baltimore: Johns Hopkins University Press.

Frey, M.M., C. Hancock, G.S. Logsdon. 1997. Cryptosporidium: Answers to Questions
       Commonly Asked by Drinking Water Professionals. AWWARF. Denver.

Frost, F., G.F. Craun, R. Calderon, and S.A. Hubbs. 1997. So many oocysts, so few outbreaks. J.
       AWWA 89(12): 8-10.

Fulton, P.  1987. Upgrading Filtration to Meet Pending Standards. Public Works (August: 68-72).

Gerba C.P., J.B. Rose, and C.N. Haas. 1996.  Sensitive populations: who is at the greatest risk?
       International Journal of Food Microbiology 30: 113-123.

Glasgow, G.D.E., and A.D. Wheatley.  1998. "The Effects of Surges on the Performance of
       Rapid Gravity Filtration." Wat. Sci. Tech. (37:2:75-81).

Griffiths, J.K. 1998. "Human cryptosporidiosis: epidemiology, transmission, clinical disease,
       treatment, and diagnosis." Adv. Parisitol. 40: 37-85.

Guerrant, R.L. 1997. "Cryptosporidiosis: An Emerging, Highly Infectious Threat." Emerging
       Infectious Diseases: 3(1):  1-9.
RIA for the ProposedLT1FBR                       9-4                              February 15, 2000

-------
Haas, C.N., and J.B. Rose. 1995. "Developing an Action Level for Cryptosporidium'' Journal
       of the American Water Works Association: 89(9).

Haas, C.N., C.S. Crockett, J.B. Rose, C.P. Gerba, and A.M. Fazil.  1996.  "Assessing the Risk
       Posed by Oocysts in Drinking Water." Journal of the American Water Works Association.

Hancock, C.M., J.B. Rose, G.J. Vasconcelos, S.I. Harris, P.T. Klonicki, and G.D. Sturbaum.
       1998. Correlation of Cryptosporidium and Giardia Occurrence in Groundwater with
       Surface Water Indicators. American Water Works Association, Annual Conference and
       Exposition.

Hansen, J.S., and J.E. Ongerth. 1991. Effects of time and watershed characteristics on the
       concentration of Cryptosporidium oocysts in river water. Appl. Environ. Microbiol.
       57(10): 2790-2795.

Harrington, W., A.J. Krupnick, and W.O. Spofford, Jr.  1985. "The Benefits of Preventing an
       Outbreak of Giardiasis Due to Drinking Water Contamination." EPA/Resources for the
       Future Report.

Harrington, W., A.J. Krupnick, and W.O. Spofford, Jr. 1989.  "The Economic Loss of a
       Waterborne Disease Outbreak." Journal of Urban Economics: 25.

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, WI. Amer. J. Publ. Health 87(12): 2032-2035.

Juranek, D.D. 1998. Cryptosporidiosis: Sources of infection and guidelines for prevention.
       Division of Parasitic Diseases, National Center for Infectious Diseases. Article originally
       appeared in Clin. Infect. Dis. 21(Suppl. 1): 57-61 [www.cdc.gov/ncidod/dpd/sources.htm]

Juranek, D. 1995. Cryptosporidiosis:  Sources of infection and guidelines for prevention. Clin.
       Infect. Dis. 21(Suppl. 1): S57-S61.

Karanis, P., D. Schoenen, and H.M. Seitz. 1998. Distribution and removal of Giardia and
       Cryptosporidium in water supplies in Germany. Wat. Sci. Tech 37(2): 9-18.

Karanis, P., D. Schoenen, and H.M. Seitz. 1996. Giardia and Cryptosporidium in backwash water
       from rapid sand filter used for drinking water production. Zbl. Bakt. 284: 107-114.

Kramer, M.H., B.L. Herwaldt, G. F. Craun, R.L. Calderon and D.D. Juranek. 1996. Waterborne
       Disease:  1993 and 1994 (Fig 4). J. AWWA 88(3): 66-80.

LeChevallier, M.W., and W.D. Norton.  1995.  "Giardia and Cryptosporidium in Raw and
       Finished Water." Journal of the American Water Works Association 87: 54-68.
February 15, 2000                               9-5                     RIA for the Proposed LT1FBR

-------
LeChevallier, M.W., and W.D. Norton. 1992. Examining relationships between particle counts
       and Giardia, Cryptosporidium and turbidity.  Journal AWWA (Dec 1992): 52-60.

LeChevallier, M.W., W.D. Norton, M. Abbaszadegan, T. Atherholt, and J. Rosen. 1997a.
       Variations in Giardia and Cryptosporidium in Source Water: Statistical Approaches to
       AnalyzingICR Data. Presented at the American Water Works Association's 1997 Water
       Quality Treatment Conference.

LeChevallier, M.W., W.D. Norton, and T.B. Atherholt. 1997b.  Protozoa in open reservoirs.
       Journal AWWA (Sept 1997), 89(9): 84-96.

LeChevallier, M.W., W.D. Norton, J.E. Siegel, and M. Abbaszadegan. 1995. Evaluation of the
       Immunoflourescence Procedure for Detection of Giardia cysts and Cryptosporidium
       oocysts in Water. Appl. Environ. Microbiol. 61: 690-697.

LeChevallier, M.W., W.D. Norton, and R.G. Lee. 199la. "Giardia and Cryptosporidium spp. in
       Filtered Drinking Water Supplies." Applied Environmental Microbiology: 57(9):
       2617-2621.

LeChevallier, M.W., W.D. Norton, and R.G. Lee. 1991b. "Occurrence of Giardia and
       Cryptosporidium spp. in Surface Water Supplies."  Applied Environmental Microbiology:
       57(9): 2610-2616.

LeChevallier, M.W., W.D. Norton, and R.G. Lee. 1991c. "Giardia and Cryptosporidium in
       Water Supplies. AWWARF.

Ledbetter, B.K. No date. Microscopic Paniculate Analysis Study of Missouri Flood Impacted Well
       Supplies of 1993 and 1995. Missouri Department of Natural Resources, Division of
       Environmental Quality, Public Drinking Water Program.

Levy, D.A., M.S. Bens, G.F. Craun, R.L.  Calderon,  and B.L. Herwaldt.  1998. Surveillance for
       Waterborne Disease Outbreaks—United States, 1995-1996. MMWR 47(55-5): 1-34.

Lisle, J.T., and J.B. Rose. 1995. Cryptosporidium contamination of water in the USA and UK: a
       mini-review. Jour. Water SRT-Aqua. 44(3):  103-117.

Logdson, G. 1987. Evaluating Treatment Plants for  Particulate Contaminant Removal. J. AWWA
       (79:9:82-92).

MacKenzie, W., N. Hoxie, M. Proctor, M. Gradus, K. Blari, D. Peterson, J. Kazmierczak,
       D. Addiss, K. Fox, J. Rose and J. Davis.  1994.  "A Massive Outbreak in Milwaukee of
       Cryptosporidium Infection Transmitted Through the Public Water Supply." New England
       Journal of Medicine: 33(1):161-167.
RIA for the ProposedLT1FBR                      9-6                             February 15, 2000

-------
Madore, M.S., J.B. Rose, C.P. Gerba, MJ. Arrowood, and C.R. Sterling. 1987. Occurrence of
       Cryptosporidium oocysts in sewage effluents and selected surface waters. J. Parasit. 73(4):
       702-705.

Mauskopf, J., and M. T. French. 1991. "Estimating the Value of Avoiding Morbidity and
       Mortality from Food borne Illnesses." Risk Analysis 11: 619-31.

McTigue, N., LeChevallier, M., Arora, H., and Clancy, J. 1998. National Assessment of Particle
       Removal by Filtration. AWWARF. Denver, 256pp

Moore, A.C., B.L. Herwaldt, G.F. Craun, R.L. Calderon, A.K. Highsmith, and D.D. Juranek.
       1993. Surveillance for waterborne disease outbreaks—United States, 1991-1992. MMWR
       42(SS-5): 1-22.

Morra, J. J.  1979. A Review of Water Quality Problems Caused by Various Open Distribution
       Storage Reservoirs,  pp. 316-21.

National Research Council. 1997. Safe Water from Every Tap.

National Research Council. 1983.  Risk Assessment in the Federal Government: Managing the
       Process.  Washington, DC: National  Academy Press.

Norton, W. D., and M.W. LeChevallier. 1997. Survey of Open Finished Water Reservoirs for
       Giardia Cysts and Cryptosporidium Oocysts. Prepared for the New Jersey Department of
       Environmental Protection. September.

Office of Management and Budget (OMB).  1993. Guidelines and Discount Rates for Benefit-
       Cost Analysis of Federal Programs. Circular No. A-94 (Revised, Transmittal Memo No.
       64). Washington, DC. Available electronically at
       

Office of Management and Budget (OMB).  1996. Economic Analysis of Federal Regulations
       Under Executive Order 12866. Washington, DC.  Available electronically at
       

Okhuysen, P.C., C.L. Chappell, C.R. Sterling, W. Jakubowski, and H.L.  DuPont.  1998.
       "Susceptibility and  Serologic Response of Health Adults to Reinfection with
       Cryptosporidium parvum."  Infection  and Immunity:  66(2): 441-443.

Okun, D., G. Craun, J. Edzwald, J. Gilbert, and J. Rose. 1997. New York City: To filter or not to
       filter? J. AWWA 89(3): 62-74.

Ongerth,  I.E., and H.H. Stibbs. 1987. Identification  of Cryptosporidium oocysts in river water.
       Appl. Environ. Microbiol. 53(4): 672-676.
February 15, 2000                               9-7                     RIA for the Proposed LT1FBR

-------
Osewe, P., D.G. Addiss, K.A. Blair, A. Hightower, M.L. Lamb, and J.P. Davis. 1996.
       "Cryptosporidiosis in Wisconsin: a case-control study of post-outbreak transmission."
       Epidemiol. Infect. 117 (2): 297-304.

PADEP (Pennsylvania Department of Environmental Protection). 1997. Data obtained from file
       records at North Central Region, Williamsport, PA. September 16, 1997.

Patania, N.L., J.G. Jacangelo, L. Cummings, A. Wilczak, K. Riley, and J. Oppenheimer. 1995.
       Optimization of Filtration for Cyst Removal. AWWARF.

Payment, P., J. Siemiatycki, L. Richardson, R. Gilles, E. Franco, and M. Prevost. 1997. "A
       Prospective epidemiological Study of Gastrointestinal Health Effects Due to the
       Consumption of Drinking Water." International J. Environmental Health Research 7, 5-31.

Perz J., F. Ennever, and S. Le Blanq.  1998. "Cryptosporidium in tap water; Comparison of
       predicted risks with observed levels of disease." Amer. J. Epidem. (147:289-301).

Petersen, C. 1992. Cryptosporidiosis in patients infected with the human immunodeficiency virus.
       Clin. Infect. Dis.  15: 903-909.

Pluntze, J.D.  1974.  "Health Aspects of Uncovered Reservoirs."  Journal of the American
       Waterworks Association. 66(8): 432-37.

Robeck, G., K. Dostal, and R. Woodward. 1964. Studies of Modification in Water Filtration. J.
       AWWA (56:2:198-213).

Rose, J.B. 1997. "Environmental Ecology of Cryptosporidium and Public Health Implications,"
       Annual Review of Public Health: 18:135-161.

Rose, J.B. 1988. Occurrence and significance of Cryptosporidium in water. Removing
       Cryptosporidium using multimedia filters. J. AWWA 87(12): 83-89.

Rose, J.B., C.P. Gerba, and W. Jakubowski. 1991.  Survey of Potable Water Supplies for
       Cryptosporidium and Giardia."  Environ. Sci. Technol. 25(8): 1392-1400.

Rose, J.B., H. Darbin, and C. P. Gerba. 1988a. "Correlations of the Protozoa Cryptosporidium
       and Giardia with Water Quality Variables in a Watershed."  Proc. Int. Conf. Water
       Wastewater Microbiol. Newport Beach,  CA.

Rose, J.B., D. Kayed, M. S. Madore, C. P. Gerba, MJ. Arrowood,  and C. R. Sterling. 1988b.
       "Methods for the Recovery of Giardia and Cryptosporidium from Environmental Waters
       and Their Comparative Occurrence." In: P. Wallis and B. Hammond, eds. Advances in
       Giardia Research. Calgary, Alberta, Canada:  University of Calgary Press.
RIA for the ProposedLT1FBR                       9-8                              February 15, 2000

-------
Rose, J.B., A. Cifrino, M.S. Madore, C.P. Gerba, C.R. Sterling, andMJ. Arrowood. 1986.
       Detection of Cryptosporidium from wastewater and freshwater environments. Water Sci.
       Technol. 18: 233-239.

Rosen, J.S., M. LeChevallier, and A. Roberson. 1996. Development and Analysis of a National
       Protozoa Database.

Salis, K. 1997. Faxed data sheets sent by Kari Salis, Department of Human Resources, Health
       Division, Oregon to Diane Loy, SAIC, McLean, VA. September 4, 1997.

Sebald, H. 1997.  Fax sent from Hans Sebald, Hydro Resources, Portland, Oregon to Diane Loy
       at SAIC, McLean, VA.  September 4, 1997.

Silverman, G.S., L.A. Nagy, and B.H. Olson.  1983.  "Variations in Particulate Matter, Algae, and
       Bacteria in Uncovered Finished Drinking Water Reservoirs." Journal of the American
       Waterworks Association.  75(4): 191-195.

Smith, J. 1997. Fax sent by Julie Smith, Department of Environmental Protection, Boston,
       Massachusetts to Diane Loy, SAIC, McLean, VA.  September, 1997.

Soave, R. 1995. Editorial Response: Waterborne Cryptosporidiosis—setting the stage for control
       of an emerging pathogen. Comparison of conventional protozoal detection methods with
       the polymerase chain reaction (PCR). Proc. of Water Qual. Technol. Conf. New Orleans,
       1663-1671.

Solo-Gabriele H, and S Neumeister. 1996. U.S. Outbreaks of Cryptosporidiosis. Journal AWWA
       (Sept 1996), 88:  76-86.

States, S.K. Stadterman, L. Ammon, P. Vogel, J. Baldizar, D. Wright, L. Conley, and J. Sykora.
       1997. Protozoa in river water: Sources, occurrence, and treatment. 15:  903-909.

States, S.K., J. Sykora, K. Stadterman, D. Wright, J. Baldizar, andL. Conley. 1995. Sources,
       occurrence and drinking water treatment removal of Cryptosporidium and Giardia in the
       Allegheny River. Proc. of Water Qual. Technol. Conf. New Orleans,  1587-1601.
Stewart, M.H., D.M. Ferguson, R. DeLeon, and W.D. Taylor. 1997. "Monitoring Program to
       Determine Pathogen Occurrence in Relationship to Storm Events and Watershed
       Conditions." Proceeding AWWA Water Quality Technology Conference, Denver, CO.

Sun, H., J.C. Bergstrom, and J.H. Dorfman.  1992.  "Estimating the Benefits of Groundwater
       Contamination Control."  Southern Journal of Agricultural Economics: pp.  63-71.

Swertfeger, J., F. Cossins, J. DeMarco, D.H. Metz, and D.J. Hartman.  1997. Cincinnati: Six
       Years of Parasite Monitoring at a Surface Water Treatment Plant. AWWA International
       Waterborne Cryptosporidium Workshop, Newport Beach, CA.

February 15, 2000                               9-9                     RIA for the Proposed LT1FBR

-------
Swiger, S.J., P.R. Scheuerman, and P.R. Musich. 1998. Determination of potential risk associated
       with Cryptosporidium and Giardia in a rural water source. Abstracts of the  1998 General
       Meeting of the American Society for Microbiology, p.459.

Tzipori, S., and J.K. Griffiths.  1998.  "Natural History and Biology of Cryptosporidium Parvan,"
       Adv. Parasitol. 40: 5-36.

U.S. Environmental Protection Agency (U.S. EPA). 2000. Estimated Per Capita Water Ingestion
       in the United States. Prepared for the Office of Water, Office of Science and Technology,
       U.S. EPA, Washington, D.C., by  Science Applications International Corporation, Mclean,
       VA.

U.S. EPA.  1999a. Occurrence Assessment of the Long Term 1 Enhanced Surface Water
       Treatment and Filter Backwash Recycle Rule. Office of Groud Water and Drinking Water.

U. S. EPA.  1999b. The Cost and Technology Document for the Long Term 1 Enhanced Surface
       Water Treatment and Filter Backwash Rule. Office of Groud Water and Drinking Water
       Washington, DC.

U.S. EPA.  1999c.  Water Industry Baseline Handbook. First Edition, EPA Contract No. 68-C6-
       0039. Prepared by International Consultants, Inc.

U.S. EPA.  1999d. Draft Report: Estimated Per Capita Water Ingestion In The United States.
       Office of Science and  Technology, Washington, DC.

U.S. EPA.  1998a.  Occurrence Assessment of the Interim Enhanced Surface Water Treatment
   Rule. Office of Ground Water and Drinking Water.

U.S. EPA. 1998b. Demographic Distribution of Sensitive Population Groups. Final Report.
   Prepared by SRA Technologies, Inc., Falls Church, VA. Work Assignment No. B-l 1/22
   (SRA 557-05/14). February 24.

U.S. EPA. 1998c. Guidance on Implementing the Capacity Development Provisions of the Safe
   Drinking Water Act Amendments of 1996. EPA Document Number: 816-R-98-006.

U.S. EPA. 1998d. National Primary Drinking Water Regulations: Interim Enhanced Surface
   Water Treatment Rule; Final Rule. 63 FR 69477, December 16, 1998.

U.S. EPA. 1998e. National Primary Drinking Water Regulations: Disinfectants and Disinfection
   Byproducts; Final Rule. 63 FR 69389, December 16, 1998.

U.S. EPA. 1997a. Community Water System Survey.  Volume I and II.  Office of Water. EPA
   815-R-97-001a and 815-R-97-001b
RIAfor the ProposedLT1FBR                      9-10                              February 15, 2000

-------
U.S. EPA.  1997b.  Drinking Water State Revolving Fund Program Guidelines. Office of Water,
   Washington, D.C.

U.S. EPA.  1997c.  IESWTR Notice of Data Availability. 62 FR 58485, November 3,  1997.

U.S. EPA.  1996. National Primary Drinking Water Regulations: Monitoring Requirements for
   Public Drinking Water Systems; Final Rule. 61 FR 24354, May 14, 1996.

U.S. EPA.  1993.  Very Small Systems. Best Available Technology Document.  Office of Ground
   Water and Drinking Water, Washington, DC.

U.S. EPA.  1992. Notice of regulatory negotiation (RegNeg) process. 63 FR 53866, November
    13, 1992.

U.S. EPA.  1989b. National Primary Drinking Water Regulations: Disinfection; Turbidity;
   Giardia Lamblia; Viruses; Legionella; and Heterotropic Bacteria; Final Rule. 54 FR 27486,
   June 29, 1989.

U.S. EPA.  1989c. National Primary Drinking Water Regulations: Total Coliforms (including
   Fecal Coliforms and E. Coli); Final Rule. 54 FR 27544, June 29, 1989.

U.S. EPA.  1984.  Estimation of Small System Water Treatment Costs. EPA Contract No.
   68-03-3093.

U.S. EPA.  1979a.  Estimating Water Treatment Costs.  Volume 2.  Cost Curves Applicable to 1
   to 200 mgd Treatment Plants.  EPA 600/2-79-162b.

U.S. EPA. 1979b. National Interim Primary Drinking Water Regulations: Control of
   Trihalomethanes in Drinking Water. 44 FR 68624, November 29, 1979.

U.S. EPA.  1979c. National Primary Drinking Water Regulations: Control of Trihalomethanes in
   Drinking Water; Final Rule. 44 FR 68624, November 29, 1979.

U.S. Environmental Protection Agency/Science Advisory Board (U.S. EPA/SAB). 1990.
   Reducing Risk: Setting Priorities and Strategies for Environmental Protection. September.

Van Houtven, G.L., J.C. Whitehead, T.H. Bingham, and B. Depro.  1997. Valuing Drinking
    Water Benefits: Theory, Methods,  and Research Needs.  Draft Report Prepared for U.S. EPA
   Office of Groundwater and Drinking Water. Research Triangle Park, NC: Research Triangle
   Institute.

Van Houtven, G.L. 1997. "Altruistic Preferences for Life-Saving Public Programs: Do Baseline
   Risks Matter." Risk Analysis. 17:85-92.
February 15, 2000                              9-11                    MA for the Proposed LT1FBR

-------
Walker, M.J., C.D. Montemagno, and M.B. Jenkins. 1998. Source water assessment and nonpoint
    sources of acutely toxic contaminants: A review of research related to survival and transport of
    Cryptosporidiumparvum.  Wat. Resour. Res. 34(12): 3383-3392.
RIAfor the ProposedLT1FBR                       9-12                              February 15, 2000

-------