EPA815-R-00-013
Proposed Arsenic in Drinking Water Rule
Regulatory Impact Analysis
Developed for:
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
John B. Bennett
Work Assignment Manager
Developed by:
Abt Associates Inc
4800 Montgomery Lane
Bethesda, MD20814
Gerald D. Stedge, Ph.D.
Principal Investigator
June 2000
-------
EPA815-R-00-013
Proposed Arsenic in Drinking Water Rule
Regulatory Impact Analysis
Developed for:
Office of Ground Water and Drinking Water
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
John B. Bennett
Work Assignment Manager
Developed by:
Abt Associates Inc.
4800 Montgomery Lane
Bethesda, MD20814
Gerald D. Stedge, Ph.D.
Principal Investigator
June 2000
-------
-------
Contents
1 Executive Summary 1-1
1.1 Regulatory Background 1-1
1.2 Health Effects of Arsenic 1-1
1.3 Regulatory Alternatives Considered 1-2
1.4 Benefits and Costs of the Proposed Rule 1-3
2 Need for the Proposal 2-1
2.1 Introduction 2-1
2.2 Public Health Concerns to be Addressed 2-2
2.2.1 Health Effects of Arsenic 2-2
2.2.2 Sources and Mechanisms of Exposure 2-4
2.3 Regulatory History 2-4
2.4 Rationale for the Regulation 2-7
2.4.1 Statutory Authority 2-7
2.4.2 Economic Rationale for Regulation 2-8
3. Consideration of Regulatory Alternatives 3-1
3.1 Regulatory Approaches 3-1
3.1.1 Determining the Standard 3-1
3.1.2 Determining the MCLG 3-1
3.1.3 Determining an MCL 3-2
3.1.4 Variances and Exemptions 3-3
3.1.5 Analytic Methods 3-3
3.2 Regulatory Alternatives Considered and Proposed Rule 3-4
3.2.1 Applicability 3-4
3.2.2 MCL 3-4
3.2.3 Monitoring 3-5
3.2.4 Compliance Technologies and Variances 3-5
3.2.5 Monitoring Waivers 3-5
3.2.6 Implementation 3-6
Table of Contents iii Proposed Arsenic in Drinking Water Rule RIA
-------
4. Baseline Analysis 4-1
4.1 Introduction 4-1
4.2 Industry Profile 4-1
4.2.1 Definitions 4-1
4.2.2 Sources of Industry Profile Data 4-2
4.2.3 Number and Size of Public Water Systems 4-2
4.2.4 System Size and Population Served 4-4
4.2.5 Number of Entry Points 4-6
4.2.6 Number of Households 4-6
4.2.7 Production Profile 4-7
4.2.8 Treatment Profile 4-10
4.2.9 Financial Profile 4-10
4.3 Occurrences of Arsenic 4-11
5. Benefits Analysis 5-1
5.1 Nature of Regulatory Benefits 5-1
5.2 Health Effects 5-1
5.2.1 Overview 5-1
5.2.2 Carcinogenic Effects 5-2
5.2.3 Noncarcinogenic Effects 5-3
5.2.4 Susceptible Subgroups 5-5
5.2.4.1 Definition 5-5
5.2.4.2 Children 5-6
5.2.4.3 Genetic Predispositions and Dietary Insufficiency 5-6
5.2.4.4 Individuals with Pre-existing Organ Susceptibilities 5-7
5.2.4.5 Individuals Exposed via Non-water Sources 5-7
5.3 Quantitative Benefits of Avoiding Bladder Cancer 5-7
5.3.1 Risk Assessment for Bladder Cancer Resulting from Arsenic Exposure . 5-8
5.3.1.1 Risk Assessment Methodology 5-8
5.3.1.2 Risk Assessment Results and Benefit Estimates 5-16
5.4 Other Benefits of Reductions in Arsenic Exposure 5-24
5.4.1 Ecological Effects 5-24
5.4.2 Drinking Water Quality and Public Perception 5-24
Table of Contents iv Proposed Arsenic in Drinking Water Rule RIA
-------
6. Cost Analysis 6-1
6.1 Introduction 6-1
6.2 Methodology 6-1
6.2.1 Description of Available Technologies 6-1
6.2.2 Unit Costs and Compliance Assumptions 6-4
6.2.3 Monitoring and Administrative Costs 6-11
6.2.4 Predicting Compliance Decisions (Compliance Decision Tree) 6-14
6.2.5 Calculating Costs 6-15
6.3 Results 6-24
6.3.1 National Costs 6-24
6.3.2 Costs by System Size and Type 6-27
6.3.3 Costs per Household 6-29
7. Comparison of Costs and Benefits 7-1
7.1 Introduction 7-1
7.2 Summary of National Costs and Benefits 7-1
7.2.1 National Cost Estimates 7-1
7.2.2 National Benefits Estimates 7-1
7.3 Comparison of Benefits and Costs 7-2
7.3.1 National Net Benefits and National Benefit-Cost Comparison 7-2
7.3.2 Cost-Effectiveness 7-5
7.4 Other Benefits 7-7
8. Economic Impact Analyses 8-1
8.1 Introduction 8-1
8.2 Regulatory Flexibility Act and Small Business Regulatory Enforcement
Fairness Act 8-2
8.2.1 Summary of EPA's Small Business Consultations 8-2
8.2.2 Definition of Small Entity for the Arsenic Rule 8-5
8.2.3 Requirements for the Initial Regulatory Flexibility Analysis 8-6
8.2.4 Small Entity Impacts 8-7
8.2.5 Small System Affordability 8-21
Table of Contents V Proposed Arsenic in Drinking Water Rule RIA
-------
8.3 Coordination With Other Federal Rules 8-24
8.4 Minimization of Economic Burden 8-25
8.5 Unfunded Mandates Reform Act 8-26
8.5.1 Social Costs and Benefits 8-27
8.5.2 State and Local Administrative Costs 8-28
8.5.3 Future Compliance Costs and Disproportionate Budgetary Effects .... 8-28
8.5.4 Macroeconomic Effects 8-34
8.5.5 Consultation with State, Local, and Tribal Government 8-34
8.5.6 State, Local, and Tribal Government Concerns 8-36
8.5.7 Regulatory Alternatives Considered 8-36
8.5.8 Impacts on Small Governments 8-36
8.6 Effect of Compliance With the Arsenic Rule on the Technical,
Financial, and Managerial Capacity of Public Water Systems 8-37
8.7 Paperwork Reduction Act 8-38
8.8 Protecting Children From Environmental Health Risks and Safety Risks .... 8-39
8.9 Environmental Justice 8-39
8.10 Health Risk Reduction and Cost Analysis 8-40
8.10.1 Quantifiable and Non-Quantifiable Health Risk Reduction Benefits . . 8-41
8.10.2 Quantifiable and Non-Quantifiable Costs 8-43
References R-l
Appendix A: Decision Tree and Decision Matrix A-l
Appendix B: Bladder Cancer Risk Analysis B-l
Table of Contents vi Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibits
Exhibit 1-1 Estimated Monetized Total Cancer Health Benefits and
Non-Quanitifiable Health Benefits from Reducing Arsenic in CWSs 1-4
Exhibit 1-2 Total National Cost of Compliance 1-5
Exhibit 1-3 Net Benefits and Benefit-Cost Ratios of Each Regulatory Option 1-6
Exhibit 4-1 Total Number of Systems by Size, Type and Ownership 4-3
Exhibit 4-2 Total Population Served of Water Systems by Source Water,
System Type and Service Population Category 4-4
Exhibit 4-3 Characteristics of NCWSs Affected by the Arsenic Rule 4-5
Exhibit 4-4 Average Number of Entry Points per Ground Water System 4-6
Exhibit 4-5 Water Consumption per Residential Connection and Number
of Residential Connections per System 4-7
Exhibit 4-6 Design Capacity of CWS Plants by Source, Ownership and System Size 4-8
Exhibit 4-7 Daily Production of CWS Plants by Source, Ownership, and System Size 4-9
Exhibit 4-8 Percentage of CWSs with Various Treatments In-Place 4-10
Exhibit 4-9 Baseline Revenues and Expenses for CWSs 4-11
Exhibit 4-10 Arsenic Occurrence in CWSs at Various Concentration Levels 4-13
Exhibit 4-11 Number of CWSs Exceeding Various Arsenic MCL Concentrations 4-13
Exhibit 5-1 Adverse Noncarcinogenic Health Effects Reported in Humans in NRC (1999)
as Potentially Associated with Arsenic, by Organ System Affected 5-4
Exhibit 5-2 Components of the Bladder Cancer Risk Assessment 5-9
Exhibit 5-3 Source of Water Consumed 5-9
Exhibit 5-4 EPA Assumed Life-Time Bladder Risk Estimates for Bladder
Cancer Among Males 5-11
Table of Exhibits Proposed Arsenic in Drinking Water
vii Rule RIA
-------
Exhibit 5-5 Life-Long Relative Exposure Factors 5-11
Exhibit 5-6 Exposure Factors Used in the NTNC Risk Assessment 5-13
Exhibit 5-7 Composition of NTNCs 5-14
Exhibit 5-8 School Children Risk Associated with Current Arsenic Exposure in NTNCs ... 5-15
Exhibit 5-9 Mean Bladder Cancer Risks for U.S. Populations Exposed At
or Above MCL Options, after Treatment 5-17
Exhibit 5-10 Exposed Population at 10"4 Risk or Fligher for Bladder Cancer After
Treatment (Community Water Consumption Data) 5-17
Exhibit 5-11 Exposed Population at 10"4 Risk or Fligher for Bladder Cancer After
Treatment (Total Water Consumption Data) 5-17
Exhibit 5-12 Annual Bladder Cancer Cases Avoided from Reducing Arsenic in CWSs 5-18
Exhibit 5-13 Potential Annual Lung Cancer Cases Avoided
from Reducing Arsenic in CWSs 5-19
Exhibit 5-14 Lifetime Avoided Medical Costs for Survivors (preliminary estimates) 5-21
Exhibit 5-15 Estimated Monetized Bladder Cancer Health Benefits and
Non-Quantifiable Health Benefits from Reducing Arsenic in CWSs 5-22
Exhibit 5-16 Mean Annual Bladder Cancer Risks, Exposed Population, and
Annual Cancer Cases Avoided in NTNCs 5-23
Exhibit 5-17 Sensitive Group Evaluation of Lifetime Risks 5-23
Exhibit 6-1 Arsenic Rule Treatment Trains by Compliance Technologies
Component, with Associated Removal Efficiencies 6-3
Exhibit 6-2 Average Compliance Technology Costs 6-5
Exhibit 6-3 (a) Estimated One-Time State Resources Required for Initiation
of the Arsenic Rule 6-13
Exhibit 6-3(b) Estimated One-Time System Resources Required for Initiation
of the Arsenic Rule 6-13
Table of Exhibits Proposed Arsenic in Drinking Water
viii Rule RIA
-------
Exhibit 6-4 Unit Resources Required for Monitoring, Implementation and Administration . . 6-14
Exhibit 6-5 Flow Regression Parameters by Water Source and System Ownership 6-18
Exhibit 6-6 Arsenic Occurrence by Water Source 6-18
Exhibit 6-7 Summary of Recommended Cost of Capital Estimates 6-20
Exhibit 6-8 Annual Treatment Costs for Three Large CWSs Expected to Undertake
or Modify Treatment Practice to Comply with the Arsenic Rule 6-23
Exhibit 6-9 Annual National System and State Compliance Costs
(CWSs and NTNCs Comply With MCL) 6-25
Exhibit 6-10 Annual National System and State Compliance Costs
(CWSs Comply With MCL /NTNCs Monitor) 6-26
Exhibit 6-11 Total Annual CWS Treatment Costs Across MCL Options by
System Size and Type 6-27
Exhibit 6-12 Total Annual CWS Monitoring and Administrative Costs Across
MCL Options by System Size and Type 6-28
Exhibit 6-13 Number of Households in CWSs Expected to Treat by Size
Category and MCL Option 6-29
Exhibit 6-14 Mean Annual Household Costs Across MCL Options by System Size 6-30
Exhibit 6-15 Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving <10,000 People (MCL 3|ig/L) 6-31
Exhibit 6-16 Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving <10,000 People (MCL 5|ig/L) 6-31
Exhibit 6-17 Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving <10,000 People (MCL 10|ig/L) 6-32
Exhibit 6-18 Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving <10,000 People (MCL 20|ig/L) 6-32
Exhibit 7-1 Summary of Annual National Net Benefits and Benefit-Cost Ratios,
CWSs Comply With MCL /NTNCs Only Monitor 7-3
Table of Exhibits Proposed Arsenic in Drinking Water
ix Rule RIA
-------
Exhibit 7-2 Comparison of Costs and Benefits of Bladder Cancer Cases Avoided
(CWSs Comply with MCL / NTNCs Monitor, 7% Discount Rate) 7-4
Exhibit 7-3 Cost per Bladder Cancer Case Avoided for the Proposed Arsenic Rule 7-5
Exhibit 7-4 Comparison of Annual Costs to Cases of Bladder Cancer per Year
(7% Discount Rate) 7-6
Exhibit 7-5 Incremental Cost per Incremental Bladder Cancer Case Avoided
(CWSs, 7% Discount Rate) 7-7
Exhibit 8-1 Profile of the Universe of Small Water Systems Regulated Under
the Arsenic Rule 8-6
Exhibit 8-2 Number of CWSs Expected to Undertake or Modify Treatment Practice 8-7
Exhibit 8-3 Number of CWSs Expected to Undertake or Modify Treatment
Practice (MCL 3|ig/L) 8-8
Exhibit 8-4 Number of CWSs Expected to Undertake or Modify Treatment
Practice (MCL 5|ig/L) 8-8
Exhibit 8-5 Number of CWSs Expected to Undertake or Modify Treatment
Practice (MCL 10|ig/L) 8-9
Exhibit 8-6 Number of CWSs Expected to Undertake or Modify Treatment
Practice (MCL 20|ig/L) 8-9
Exhibit 8-7 Average Annual System Compliance Costs for CWSs 8-10
Exhibit 8-8 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving <100 People (MCL 3|ig/L) 8-11
Exhibit 8-9 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving <100 People (MCL 5|ig/L) 8-11
Exhibit 8-10 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving <100 People (MCL 10|ig/L) 8-12
Exhibit 8-11 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving <100 People (MCL 20|ig/L) 8-12
Table of Exhibits Proposed Arsenic in Drinking Water
X Rule RIA
-------
Exhibit 8-12 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 101-500 People (MCL 3|ig/L) 8-13
Exhibit 8-13 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 101-500 People (MCL 5|ig/L) 8-13
Exhibit 8-14 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 101-500 People (MCL 10|ig/L) 8-14
Exhibit 8-15 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 101-500 (MCL 20|ig/L) 8-14
Exhibit 8-16 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 501-1,000 People (MCL 3|ig/L) 8-15
Exhibit 8-17 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 501-1,000 (MCL 5|ig/L) 8-15
Exhibit 8-18 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 501-1,000 People (MCL 10|ig/L) 8-16
Exhibit 8-19 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 501-1,000 People (MCL 20|ig/L) 8-16
Exhibit 8-20 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 1,001-3,300 People (MCL 3|ig/L) 8-17
Exhibit 8-21 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 1,001-3,300 People (MCL 5|ig/L) 8-17
Exhibit 8-22 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 1,001-3,300 People (MCL 10|ig/L) 8-18
Exhibit 8-23 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 1,001-3,300 People (MCL 20|ig/L) 8-18
Exhibit 8-24 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 3,301-10,000 People (MCL 3|ig/L) 8-19
Exhibit 8-25 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 3,301-10,000 People (MCL 5|ig/L) 8-19
Table of Exhibits Proposed Arsenic in Drinking Water
xi Rule RIA
-------
Exhibit 8-26 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 3,301-10,000 People (MCL 10|ig/L) 8-20
Exhibit 8-27 Comparison of CWS Baseline and Post-Compliance Total
Expenses for Systems Serving 3,301-10,000 People (MCL 20|ig/L) 8-20
Exhibit 8-28 Mean Annual Costs to Households Served by CWSs, by Size Category 8-22
Exhibit 8-29 Average Annual Cost per CWS by Ownership 8-30
Exhibit 8-30 Annual Compliance Costs per Household for CWSs Exceeding MCLs 8-32
Exhibit 8-31 Annual Compliance Costs per Household for CWSs Exceeding
MCLs, as a Percentage of Median Household Income 8-33
Exhibit 8-32 Mean Bladder Cancer Risks, Exposed Population, and Annual
Cancer Cases Avoided in CWSs 8-41
Exhibit 8-33 Annual Bladder Cancer Cases Avoided from Reducing Arsenic in CWSs 8-42
Exhibit 8-34 Potential Annual Lung Cancer Cases Avoided from Reducing
Arsenic in CWSs 8-42
Exhibit 8-35 Estimated Monetized Total Cancer Health Benefits and
Non-Quantifiable Health Benefits from Reducing Arsenic in CWSs 8-43
Exhibit 8-36 Summary of the Total Annual National Costs of Compliance 8-44
Exhibit 8-37 Mean Annual Costs per Household in CWSs 8-45
Exhibit 8-38 Cost per Bladder Cancer Case Avoided for the Proposed Arsenic Rule 8-46
Table of Exhibits Proposed Arsenic in Drinking Water
xii Rule RIA
-------
Chapter 1: Executive Summary
1.1 Regulatory Background
An enforceable standard of 50 |ig/L currently exists for arsenic in community water systems under
the National Interim Primary Drinking Water Regulations (NPDWR) (40 CFR 59566). In
§1412(b)(12)(A) of the SDWA, as amended in 1996, Congress specifically directed EPA to issue
a final regulation by January 1, 2001. At the same time, Congress directed EPA to develop a
research plan to reduce the uncertainty in assessing health risks from low levels of arsenic by
February 2, 1997, and conduct the research in consultation with the National Academy of
Sciences, other Federal agencies, and interested public and private entities.
This document analyzes the impacts of the proposed rule which revises the current standard as
follows:
1) reduces the current MCL for arsenic in community water systems from 50 |ig/L to 5
2) requires nontransient non-community water systems (NTNC) to perform compliance
monitoring; and
3) revises the current monitoring requirements to make them consistent with the Standard
Monitoring Framework (40 CFR 141.23(c)).
1.2 Health Effects of Arsenic
Arsenic's carcinogenic role was noted over 100 years ago (NCI, 1999) and has been studied ever
since. The Agency has classified arsenic as a Class A human carcinogen, "based on sufficient
evidence from human data. An increased lung cancer mortality was observed in multiple human
populations exposed primarily through inhalation. Also, increased mortality from multiple internal
organ cancers (liver, kidney, lung, and bladder) and an increased incidence of skin cancer were
observed in populations consuming drinking water high in inorganic arsenic."
A 1999 NRC report on arsenic states that "epidemiological studies ... clearly show associations of
arsenic with several internal cancers at exposure concentrations of several hundred micrograms
per liter of drinking water." Ten epidemiological studies covering eight organ systems have
quantitative data for risk assessment (NRC, 1999, Table 4-1). The organ systems where cancers
in humans have been identified include skin, bladder, lung, kidney, nasal, liver, and prostate.
Table 10-6 of the same NRC report provides risk parameters for three cancers: bladder, lung, and
liver cancer. Considering all cancers in aggregate, the NRC states that "considering the data on
bladder and lung cancer in both sexes noted in the studies ... a similar approach for all cancers
could easily result in a combined cancer risk on the order of 1 in 100" (at the current MCL of 50
Chapter 1, Executive Summary 1-1 Proposed Arsenic in Drinking Water Rule RIA
-------
New data provide additional health effects information on both carcinogenic and noncarcinogenic
effects of arsenic. A recent study by Tsai et al. (1999) of a population that has been studied over
many years in Taiwan has provided standardized mortality ratios (SMRs) for 23 cancerous and
non-cancerous causes of death in women and 27 causes of death in men at statistically significant
levels in an area of Taiwan with elevated arsenic exposures (Tsai et al., 1999). SMRs are an
expression of the ratio between deaths that were observed in an area with elevated arsenic levels
and those that were expected to occur, based on the mortality experience of the populations in
nearby areas without elevated arsenic levels. Drinking water (250-1,140 |ig/L) and soil (5.3-11.2
mg/kg) in the Tsai et al. (1999) population study had very high arsenic content.
Tsai et al. (1999) identified "bronchitis, liver cirrhosis, nephropathy, intestinal cancer, rectal
cancer, laryngeal cancer, and cerebrovascular disease" as possibly "related to chronic arsenic
exposure via drinking water," which had not been reported before. In addition, the study area had
upper respiratory tract cancers previously only related to occupational inhalation. High male
mortality rate (SMR > 3) existed for bladder, kidney, skin, lung, and nasal cavity cancers and for
vascular disease. However, the authors noted that the mortality range was marginal for leukemia,
cerebrovascular disease, liver cirrhosis, nephropathy (kidney), and diabetes. Females also had
high mortalities for laryngeal cancer. There are, of course, possible differences between the
population and health care in Taiwan and the United States. For example, arsenic levels in the
U.S. are not as high as they were in the study area of Taiwan. However, the study gives an
indication of the types of health effects that may be associated with arsenic exposure via drinking
water.
Arsenic interferes with a number of essential physiological activities, including the actions of
enzymes, essential cations, and transcriptional events in cells (NRC, 1999). A wide variety of
adverse health effects have been associated with chronic ingestion of arsenic in drinking water,
occurring at various exposure levels. Exhibit 5-1 lists the effects on specific organ systems
reported in humans exposed to arsenic and provides descriptive information on the specific
diseases and/or symptoms associated with categories of diseases.
1.3. Regulatory Alternatives Considered
In regulating a contaminant, EPA first sets a maximum contaminant level goal (MCLG), which
establishes the contaminant level at which no known or anticipated adverse health effects occur.
MCLGs are non-enforceable health goals. For this rulemaking, EPA is proposing an MCLG of
zero. EPA then sets an enforceable maximum contaminant level (MCL) as close as technologically
possible to the MCLG. In addition, EPA may use its discretion in setting the MCL by choosing
an MCL that is protective of public health while also insuring that the quantified and non-
quantified costs are justified by the quantified and non-quantified benefits of the rule. For this
rulemaking, EPA is proposing an MCL of 5 |ig/L. Chapter 3 describes the process by which EPA
determined both the MCLG and the MCL.
Chapter 1, Executive Summary 1-2 Proposed Arsenic in Drinking Water Rule RIA
-------
EPA considered a range of MCLs in developing the proposed Arsenic Rule, including MCLs of 3,
5, 10, and 20 //g/L. EPA evaluated the following five factors to determine the proposed
Maximum Contaminant Level (MCL):
the analytical capability and laboratory capacity,
likelihood of water systems choosing various compliance technologies for several
sizes of systems based on source water properties,
the national occurrence of arsenic in water supplies.
quantified and non-quantified costs and health risk reduction benefits likely to
occur at the MCLs considered, and
the effects on sensitive subpopulations.
After evaluating the above factors, EPA considered an MCL of 3 jig/L since this is the level that
has been determined to be as close to the MCLG as is feasible. However, the Agency is using its
discretionary authority in Section 1412(b)(6)(A) to consider setting MCL at a less stringent level.
The statute requires that the alternative less stringent level be one which maximizes health risk
reduction at a level where costs and benefits are balanced. As a result, EPA considered the
alternative MCL options of 5, 10, and 20 |ig/L. These alternative MCL options were considered
because they also provide assurance that the residual risk for both bladder and lung cancer
endpoints will be in the 10"4 range, but at lower anticipated national costs.
The Agency also considered two regulatory options related to the applicability of the proposed
MCL. Specifically, EPA investigated applying both the monitoring and treatment requirements of
the proposed rule to both community water systems (CWS) and non-transient non-community
water systems (NTNC). A CWS is defined as a system that provides piped water to at least 25
people or with at least 15 service connections year-round. A NTNC is a public water systems
that is not defined as a CWS and that regularly serves at least 25 of the same people for at least
six months of the year. After considering the costs and benefits of the proposed rule with regard
to both CWSs and NTNCs, EPA proposes to require CWS to comply with all facets of the
proposed rule, while only requiring NTNCs to comply with the monitoring components of the
rule. The benefit-cost analysis upon which this decision is based is provided in Chapters Five, Six,
and Seven of this RIA. Transient non-community systems, which provide potable water to
continuously changing populations, will not be subject to the proposed rule.
The proposed rule also includes modifications to the current monitoring requirements, including
the availability of monitoring waivers. A detailed discussion of these changes can be found in
Chapter 3.
1.4 Benefits and Costs of the Proposed Rule
Quantitative risk metrics (e.g., slope factors or reference doses) are necessary to evaluate cancer
or non-cancer risks. Although arsenic causes numerous health effects, bladder cancer is the only
endpoint for which an Agency-approved metric for evaluating arsenic related risk currently exists.
Chapter 1, Executive Summary 1-3 Proposed Arsenic in Drinking Water Rule RIA
-------
This cancer slope factor (SF) for bladder cancer is used to calculate cases potentially avoided due
to EPA's proposed drinking water standards. Benefits estimates for avoided cases of bladder
cancer were calculated using mean population risk estimates at various MCL levels. Lifetime risk
estimates were converted to annual risk factors, and applied to the exposed population to
determine the number of cases avoided. These cases were divided into fatalities and non-fatal
cases avoided, based on survival information. The avoided premature fatalities were valued based
on the VSL estimates discussed earlier, as recommended by EPA current guidance for cost/benefit
analysis. The avoided non-fatal cases were valued based on the willingness to pay estimates for
the avoidance of chronic bronchitis. The upper bound estimates include the possibility of the
incidence rate being understated, depending on the survival rate for bladder cancer in the study
area of Taiwan during the Chen study.
The "What if?" scenario for lung cancer benefits was used to estimate benefits for avoided cases
of lung cancer. This scenario is based on the statement in the NRC report "Arsenic in Drinking
Water," which states that "some studies have shown that excess lung cancer deaths attributed to
arsenic are 2-5 fold greater than the excess bladder cancer deaths (NRC, 1999, pg. 8)." Two-to-
five fold greater would be 3.5 fold greater on average. Also in the U.S. the mortality rate from
bladder cancer is 26% and the mortality rate of lung cancer is 88%. This suggests that if the risk
of contracting lung cancer were identical to the risk of contracting bladder cancer, one would
expect 3.4 times the number of deaths from lung cancer as from bladder cancer. Since these
numbers are essentially the same, it seems reasonable to assume that the risk of contracting lung
cancer is essentially the same as the rate of contracting bladder cancer,l in the context of this
"what-if' scenario. If the risk of contracting lung cancer from arsenic in drinking water is
approximately equal to the risk of contracting bladder cancer, then the combined risk estimates of
contracting either bladder or lung cancer would be approximately double the risk estimates of
bladder cancer alone.
Numerous other health effects that are likely to be avoided as a result of this rule may generate
significant benefits, and should not be discounted based on the fact that they can't be quantified at
this time. The estimated total national monetized benefits of the proposed rule and the other rule
options considered are provided in Exhibit 1-1.
'If "X" is the probability of contracting bladder cancer, then 0.26X is the probability of mortality from
bladder cancer. If lung cancer deaths are 2 to 5 times as high as bladder cancer, then they are, on average, 3.5
times as high and the average probability of mortality from lung cancer would be 3.5 times 0.26X, or 0.9IX. Since
we also know that there is a 88% mortality rate from lung cancer, then if the probability of contracting lung cancer
is "Y," the probability of mortality from lung cancer can also be represented as 0.88Y. Setting the two ways of
deriving the probability of mortality from lung cancer equal, or 0.91X = 0.88Y, one can solve for Y
(Y= (0.91/0.88) X). Thus Y is approximately equal to X, and the rate of contracting lung cancer is approximately
the same as the rate of contracting bladder cancer.
Chapter 1, Executive Summary 1-4 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 1-1
Estimated Monetized Total Cancer Health Benefits and
Non-Quantifiable Health Benefits from Reducing Arsenic in CWSs
Arsenic
Level
(ug/L)
3
5
10
20
Annual
Bladder Cancer
Health Benefits
(Smillions)12
$43.6 -$104.2
$31 .7 -$89.9
$17. 9 -$52.1
$7.9 - $29.8
"What-if ' Scenario and Potential Non-Quantifiable
Health Benefits
"What-if" Scenario
Annual
Lung Cancer
Health Benefits
(Smillions)13
$47.2 - $448.0
$35.0 - $384.0
$19.6 -$224.0
$8.8 -$128.0
Potential Non-Quantifiable
Health Benefits
Skin Cancer
Kidney Cancer
Cancer of the Nasal Passages
Liver Cancer
Prostate Cancer
Cardiovascular Effects
Pulmonary Effects
Immunological Effects
Neurological Effects
Endocrine Effects
Reproductive and Developmental
Effects
1. May 1999 dollars.
2. The lower-end estimate is calculated using the lower-end number of bladder cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was
100 percent. The upper-end estimate is calculated using the upper-end number of cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was 80
percent.
3. These estimates are based on the "what if" scenario for lung cancer, where the risks of a fatal lung cancer
case associated with arsenic are assumed to be 2-5 times that of a fatal bladder cancer case.
For the proposed MCL of 5 ug/L, the estimated monetized bladder cancer health benefits range
from $31.7 million to $89.9 million. Potential lung cancer health benefits, based on the "What-if
scenario, range from $35.0 million to $384.0 million. More detail about these benefit estimates
are found in Chapter 5. Exhibit 1-2 shows the estimated national cost of compliance of the
proposed rule and the other rule options that were considered. At the proposed MCL of 5 ug/L,
the estimated national cost of compliance is $379 million at a discount rate of 3 percent, and $442
million at a discount rate of 7 percent.
Chapter 1, Executive Summary
1-5
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 1-2
Total National Cost of Compliance ($ millions)
Discount Rate
cws
3% 7%
NTNC*
3% 7%
TOTAL
3% 7%
MCL = 3 ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$639.2 $746.4
$2.2 $3.0
$1.6 $1.9
$643.1 $751.4
$0.9 $1.1
$0.6 $0.7
$1.5 $1.8
$639.2 $746.4
$3.1 $4.1
$2.2 $2.6
$644.6 $753.2
MCL =5 nQ/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$374.0 $436.0
$2.0 $2.8
$1.3 $1.6
$377.3 $440.4
$0.9 $1.1
$0.6 $0.7
$1.6 $1.8
$374.0 $436.0
$2.9 $3.9
$2.0 $2.3
$378.9 $442.2
MCL = 10 ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$160.4 $186.7
$1.8 $2.5
$1.1 $1.3
$163.3 $190.5
$1.0 $1.1
$0.6 $0.7
$1.6 $1.9
$160.4 $186.7
$2.8 $3.7
$1.7 $2.1
$164.9 $192.4
MCL =20 ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$58.9 $68.3
$1.7 $2.4
$1.0 $1.2
$61.6 $71.8
$1.0 $1.1
$0.7 $0.7
$1.6 $1.9
$58.9 $68.3
$2.7 $3.5
$1.6 $1.9
$63.2 $73.7
*Costs include treatment, O&M, monitoring, and administrative costs to CWSs, monitoring and administrative costs to
NTNCWSs, and State costs for administration of water programs.
-------
The net benefits and benefit-cost ratios of each regulatory option are provided in Exhibit 1-3,
when only health benefits for bladder cancer cases avoided are quantified. At the proposed MCL
of 5 ng/L, the net benefits range from a high of-$287.4 million to a low of-$345.6 million, at a
discount rate of 3 percent. These net benefits correspond to benefit-cost ratio of 0.24 and 0.08
(also at a 3 percent rate of discount). At a 7 percent discount rate the net benefits range from a
high of -$350.5 million to a low of -$408.7 million. These net benefits correspond to benefit-cost
ratio ranging from 0.20 to 0.07 (also at a seven percent rate of discount).
Exhibit 1-3
Net Benefits and Benefit-Cost Ratios of Each Regulatory Option
(Bladder Cancer Cases Only, in $ millions)
MCL dog/L)
3
5
10
20
3% Discount Rate
lower bound
upper bound
Net Benefits
Benefit/Cost Ratio
Net Benefits
Benefit/Cost Ratio
$ (599.5)
0.07
$ (538.9)
0.16
$ (345.6)
0.08
$ (287.4)
0.24
$ (145.4)
0.11
$ (111 .2)
0.32
$ (53.7)
0.13
$ (31.8)
0.48
7% Discount Rate
a
^
o
£2
a>
_o
upper bound
Net Benefits
Benefit/Cost Ratio
Net Benefits
Benefit/Cost Ratio
$ (707.8)
0.06
$ (647.2)
0.14
$ (408.7)
0.07
$ (350.5)
0.20
$ (172.6)
0.09
$ (138.4)
0.27
$ (63.9)
0.11
$ (42.0)
0.42
"Costs include treatment, O&M, monitoring, and administrative costs to CWSs, monitoring and
administrative costs to NTNCWSs, and State costs for administration of water programs.
The lower-end estimate of bladder cancer cases avoided is calculated using the lower-end risk estimate
(see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group
was 100 percent. The upper-end estimate of bladder cancer cases is calculated using the upper-end risk
estimate (see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese
study group was 80 percent.
Chapter 1, Executive Summary
1-7
Proposed Arsenic in Drinking Water Rule RIA
-------
As mentioned above, there are a number of important non-monetized benefits of reducing arsenic
exposure that are not include in the net benefit and benefit-cost calculations. Chief among these
are certain health impacts known to be caused by arsenic (such as skin cancer). In 1988 EPA
published a risk estimate which used the skin cancer data from Taiwan. EPA calculated a
Maximum Likelihood Estimate (MLE) for skin cancer of 3 x 10"5 for females and 7 x 10"5 for
males drinking 2 liters a day contaminated with 1 |ig/L of arsenic. At the current MCL of 50
|ig/L and two liters per day, the risk would be 5xlO"3. A number of epidemiologic studies
conducted in several countries (e.g., Taiwan, Japan, England, Hungary, Mexico, Chile, and
Argentina) report an association between arsenic in drinking water and skin cancer in exposed
populations. Studies conducted in the U.S. have not demonstrated an association between
inorganic arsenic in drinking water and skin cancer. However, these studies may not have
included enough people in their design to detect these types of effects.
The potential monetized benefits associated with skin cancer reduction would not change the total
benefits of the rule to an appreciable degree, even if the assumption were made that the risk of
skin cancer were equivalent to that of bladder cancer, using EPA's 1988 risk assessment. Skin
cancer is highly treatable (at a cost of illness of less than $3,500 for basal and squamous cell
carcinomas vs. a cost-of-illness of $178,000 for non-fatal bronchitis) in the U.S., with few
fatalities (less than one percent).
In addition to potentially reducing the risk of skin and lung cancer, there are also a large number
of other health effects associated with arsenic, as presented in Exhibit 1-1, which are not
monetized in this analysis, due to lack of appropriate data.
Other benefits not monetized in this analysis include customer peace of mind from knowing
drinking water has been treated for arsenic and reduced treatment costs for currently unregulated
contaminants that may be co-treated with arsenic. To the extent that reverse osmosis is used for
arsenic removal, these benefits could be substantial. Reverse osmosis is the primary point of use
treatment, and it is expected that very small systems will use this treatment to a significant extent.
(These benefits of avoided treatment cannot currently be monetized; however, they can be readily
monetized in the future, as decisions are made about which currently unregulated contaminants to
regulate.)
Chapter 1, Executive Summary 1-8 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 2: Need for the Proposal
2.1 Introduction
The Safe Drinking Water Act (SDWA), as amended in 1996, requires the EPA to identify and
regulate substances in drinking water that may have an adverse effect on public health and that are
known or anticipated to occur in public water supplies. National Primary Drinking Water
Regulations (NPDWRs) address risks to public health, and secondary regulations address
aesthetic qualities (such as taste, odor, or color) that relate to public acceptance of drinking
water. For NPDWRs, EPA must either establish a Maximum Contaminant Level (MCL) or, if it is
not economically or technically feasible to monitor the contaminant in drinking water, specify a
treatment technique to remove the contaminant or reduce its concentration in the water supply.
An enforceable standard of 50 |ig/L currently exists for arsenic in community water systems under
the National Interim Primary Drinking Water Regulations (40 CFR 59566). In §1412(b)(12)(A)
of the SDWA, as amended in 1996, Congress specifically directed EPA to propose a NPDWR for
arsenic by January 1, 2000 and issue the final regulation by January 1, 2001. At the same time,
Congress directed EPA to develop a research plan to reduce the uncertainty in assessing health
risks from low levels of arsenic by February 2, 1997 and conduct the research in consultation with
the National Academy of Sciences, other Federal agencies, and interested public and private
entities.
This document analyzes the impacts of the proposed rule which revises the current standard as
follows:
1) reduces the current MCL for arsenic in community water systems from 50 |ig/L to 5 |ig/L;
2) requires nontransient non-community water systems (NTNC) to perform compliance
monitoring; and
3) revises the current monitoring requirements to make them consistent with the Standard
Monitoring Framework (40 CFR 141.23(c)).
Executive Order 12866, Regulatory Planning And Review, requires EPA to estimate the costs
and benefits of the Arsenic Rule in a regulatory impact analysis (RIA). This chapter of the RIA
discusses the public health concerns being addressed by the rule, describes the history of
regulatory efforts concerning arsenic, and discusses the economic rationale for the rule.
Subsequent chapters will accomplish the following:
discuss the regulatory options considered by EPA (Chapter 3),
present the results of the baseline analysis (Chapter 4),
examine the benefits of the proposed rule (Chapter 5),
present the results of the cost analysis (Chapter 6),
compare the costs and benefits of the proposed rule and the regulatory options
considered by EPA (Chapter 7), and
discuss the potential economic impacts of the rule (Chapter 8).
Chapter 2, Need for the Proposal 2-1 Proposed Arsenic in Drinking Water Rule RIA
-------
2.2 Public Health Concerns to be Addressed
This section describes the public health concerns addressed by the proposed Arsenic Rule. A
description of potential health effects associated with arsenic, including effects in sensitive
subpopulations, along with the sources of human exposure to arsenic is presented. In addition,
the section describes current controls that address exposure to arsenic.
2.2.1 Health Effects of Arsenic
Arsenic is a naturally occurring element present in the environment in both organic and inorganic
forms. Inorganic arsenic, the more toxic form, is found in ground water, surface water and many
foods. Chronic exposure to arsenic has been found to result in a variety of adverse health effects,
including skin and internal cancers and cardiovascular and neurological effects. The available
evidence on the health effects of arsenic has recently been reviewed in a report by the National
Research Council (NRC) of the National Academy of Sciences (NAS). The health effects of
inorganic arsenic are summarized here and are described in more detail in Chapter 5.
Exposures to organic forms of arsenic also occur through ingestion of food and metabolism of
ingested inorganic arsenic. Experimental data on the effects of organic forms of arsenic are not as
well characterized as those for inorganic arsenic, and thus are the subject for future research.
Limited data in animals suggest that some organic forms of arsenic also produce cancer and
non-cancer health effects.
Cancer
EPA has identified arsenic as a group A "known" human carcinogen, based on increased risks of
lung cancer in workers exposed to airborne arsenic and dose-dependent increases in skin cancer
risk in Taiwan. Using data from Taiwan, EPA calculated a Maximum Likelihood Estimate (MLE)
of 3 x 10"5 for females and 7 x 10"5 for males drinking 2 liters a day contaminated with 1 |ig/L of
arsenic. The values were combined to give an overall risk of 5 x 10"5 /(|ig/L) or 2 |ig/L = 1 x 10"4
risk level. The International Agency for Research on Cancer (IARC) has also classified arsenic as
a human carcinogen. Epidemiological studies have shown evidence of carcinogenic risk by both
inhalation and ingestion.
Unlike most environmental contaminants, there is a large human database available for inorganic
arsenic. However, there is substantial debate among the scientific community over the
interpretation of these data and their application in risk assessment. A number of epidemiologic
studies conducted in several countries (e.g., Taiwan, Japan, England, Hungary, Mexico, Chile,
and Argentina) report an association between arsenic in drinking water and skin cancer in exposed
populations. Increased mortality from internal cancers of liver, bladder, kidney, and lung have
also been reported.
In 1996, EPA requested that the National Research Council (NRC) of NAS conduct an
independent review of the arsenic toxicity data. NRC was asked to review EPA's current criteria
(50 |ig/L and 0.018 |ig/L), evaluate use of recent Taiwan data and other studies to assess the
carcinogenic and non-carcinogenic health effects of arsenic, and recommend changes to EPA's
risk characterization for arsenic. NRC issued its report on March 23, 1999. A summary of the
Chapter 2, Need for the Proposal 2-2 Proposed Arsenic in Drinking Water Rule PJA
-------
NRC's results are provided below.
Non-Cancer Health Effects
In addition to cancer, arsenic exposures have been reported to result in other adverse health
effects. These include thickening of the skin, effects on the nervous system such as tingling and
loss of feeling in limbs, hearing impairment, effects on the heart and circulatory system, diabetes,
developmental effects, and effects on the gastrointestinal system and liver. Many of these effects
are observed at concentrations where cancer effects were observed in the epidemiology studies.
Sensitive Subpopulations
Certain sensitive individuals may be at a greater risk of serious illness from exposure to arsenic
than the general population. The NRC report noted that human sensitivity to the toxic effects of
inorganic arsenic exposure is likely to vary based on genetics, metabolism, diet, health status, sex,
and other possible factors. For example, reduced ability to methylate arsenic (converting
inorganic arsenic into less acutely toxic and more readily excreted forms) may result in retention
of more arsenic in the body and increased risk of toxic effects.
The following groups are cited in various studies as being particularly susceptible to arsenic:
Children are identified as especially susceptible to health effects from arsenic
because their dose of arsenic will be, on average, higher than that of adults
exposed to similar concentrations due to their higher fluid and food intake relative
to body weight. The NRC report cited one study that suggests that children may
have a lower arsenic-methylation efficiency than adults.
Pregnant and lactating women are especially vulnerable because of the adverse
reproductive and developmental effects of arsenic.
People with poor nutritional status may have a reduced ability to methylate
arsenic.
Individuals with pre-existing diseases that affect specific organs - in
particular, kidney and liver problems - may be more susceptible to the effects of
arsenic, because these organs act to detoxify arsenic in the body. In addition,
arsenic can directly damage these and other organ systems, as described above.
Individuals with pre-existing damage or congenital defects in these systems are
more susceptible to health effects from exposure to arsenic. The elderly are more
likely as a group to have pre-existing conditions in the susceptible organ systems.
Section 5.2.4 discusses the susceptibility of these subgroups in more detail. Due to a lack of
available data, no quantitative analysis of the specific risks to sensitive populations was performed
as part of this RIA.
Chapter 2, Need for the Proposal 2-3 Proposed Arsenic in Drinking Water Rule RIA
-------
2.2.2 Sources and Mechanisms of Exposure
Arsenic is an element that occurs in the earth's crust. Accordingly, there are natural sources of
exposure. Erosion and weathering of rocks deposit arsenic in water bodies and lead to the uptake
of arsenic by animals and plants. Consumption of food and water are the major sources of arsenic
exposure for the majority of U.S. citizens. People may also be exposed from industrial sources, as
arsenic is used in semiconductor manufacturing, petroleum refining, wood preservatives, animal
feed additives and herbicides.
Arsenic can combine with other elements to form inorganic and organic arsenicals. In general,
inorganic derivatives are regarded as more toxic than the organic forms. While food contains both
inorganic and organic arsenicals, primarily inorganic forms are present in water.
Recently, EPA developed estimates of human exposure to arsenic in drinking water, food, and air
using data from numerous Federal sampling surveys analyzing the occurrence of arsenic in public
water supplies, dietary foods, and ambient air. EPA's national air sampling data bases indicate
very low concentrations of arsenic in both urban and non-urban locations, at levels typically
ranging from about 0.003-0.03 |ig/m3. Air is therefore an insignificant source of arsenic intake,
representing typically less than one percent of overall exposure.
EPA reviewed several local and regional studies for comparison purposes. Using the Food and
Drug Administration's (FDA) Total Diet Study, recent dietary analyses indicate that the average
adult's total arsenic intake is about 53 ng/day. However, the FDA analytical methodology does
not differentiate between the organic and inorganic forms of arsenic. For most people living in the
U.S., inorganic arsenic exposure is primarily from food and water sources. Since the inorganic
forms are considered to be more toxic, it is important to estimate the amount of inorganic arsenic
in the diet. To accomplish this estimation, EPA used the FDA data along with a separate study
that characterized arsenic species in foods. This separate characterization indicated that about 20
percent of daily intake of dietary arsenic is in the inorganic form. Conversely, most arsenic
present in drinking water is in the form of inorganic arsenic species.
Accounting for the organic forms of arsenic in food, the dietary intake of inorganic arsenic was
estimated to be approximately 14 ng/day. An adult drinking 2 L/day of water containing 10 |ig/L
of arsenic, would obtain 20 |ig/day from drinking water, so that drinking water would contribute
about 60 percent of total intake of inorganic arsenic. On the other hand, an adult drinking water
containing 2 |ig/L of arsenic would obtain almost 80 percent of the daily inorganic arsenic from
food.
2.3 Regulatory History
This section provides a chronology and overview of regulatory actions affecting arsenic in
drinking water and recent efforts that have led to this proposed rule-making. The major studies
and data collection efforts that have highlighted the need for new regulation are also summarized.
Current MCL: In 1975, EPA set the National Interim Primary Drinking Water Regulation at 50
Chapter 2, Need for the Proposal 2-4 Proposed Arsenic in Drinking Water Rule RIA
-------
(40 FR 59566 December 24, 1975.) This standard was equal to the standard set in 1943 by
the U.S. Public Health Service (US PHS) for interstate water carriers which was not based on a
risk assessment. The MCL was based on EPA's assertion that daily consumption of two liters of
water provides approximately 10 percent of total ingested arsenic of 900 ng/day. Commenters
recommended an MCL of 100 jig/L based on no observed adverse health effects. EPA noted
long-term chronic effects at 300 to 2,750 |ig/L, but no chronic effects at 120 |ig/L (NIPDWRs,
EPA-570/9-76-003).
Water Quality Criteria: In 1980, EPA announced the availability of Water Quality Criteria
Documents to protect surface water bodies of water from pollutants under the Clean Water Act
(45 FR 79318, November 28, 1980). These criteria are used as guidance to the States in
establishing surface water quality standards and discharge limits for effluents. The criteria for
protection of human health from ingestion of arsenic in contaminated water and aquatic organisms
was 2.2 mg/L, or 0.0022 |ig/L. In 1992, the Clean Water Act criteria were recalculated with
updated cancer slope factor data to yield 0.018 jig/L for arsenic (57 FR 60848, December 22,
1992.).
1985 Proposed MCL: In an Advanced Notice of Proposed Rulemaking (ANPR) published
October 5, 1983 (48 FR 45502), EPA requested comment on whether the arsenic MCL should
consider carcinogenicity, other health effects, and nutritional requirements; and whether MCLs
are necessary for separate valence states. In 1985, EPA then proposed a non-enforceable
Maximum Contaminant Level Goal (MCLG) of 50 |ig/L based on a NAS conclusion that 50 |ig/L
balanced toxicity and possible essentiality. EPA also requested comment on alternate MCLGs of
100 |ig/L based on non-carcinogenic effects and 0 |ig/L based on carcinogenicity (50 FR 46936,
November 13, 1985).
1986 SWDA Amendments: The 1986 SDWA Amendments converted the 1975 interim arsenic
standard to a NPDWR, subject to revision by 1989.
1988 Risk Assessment Forum Report: EPA's Risk Assessment Forum wrote the Special Report
on Ingested Inorganic Arsenic: Skin Cancer: Nutritional Essentiality (EPA/625/3-87/013), in part,
to evaluate the validity of applying the Taiwan 1968/1977 data to dose-response assessments in
the U.S. At the 50 jig/L standard, the calculated U.S. lifetime risk ranged from
1 x 10'3to3x 10'3.
1989: After reviewing EPA's arsenic health effects studies in June 1988, the Science Advisory
Board (SAB) stated in its August 14, 1989, report the following:
The essentiality of arsenic is suggestive but not definitive;
Hyperkeratosis may not be a precursor of skin cancer;
The Taiwan data are adequate to conclude that high doses of ingested arsenic can
cause skin cancer;
The Taiwan study is inconclusive to determine cancer risk at levels ingested in the
U.S.; and
As (III) levels below 200 to 250 jig per day may be detoxified.
SAB concluded that the dose-response is nonlinear and reported that the 1988 Forum Report did
not apply nonlinearity in its risk assessment.
Chapter 2, Need for the Proposal 2-5 Proposed Arsenic in Drinking Water Rule RIA
-------
1989: Uncertainty about arsenic risk assessment issues caused the Agency to miss the 1989
deadline for proposing a revised NPDWR, and a citizen suit was filed against EPA. A consent
decree was entered by the court in June, 1990, and was amended several times thereafter before
being dismissed after passage of the 1996 SDWA Amendments.
1994: EPA thoroughly reviewed the available information and determined that:
There is evidence of an association between internal cancer and arsenic;
The risk of internal cancer cannot be quantified using the available epidemiological
data; and
The risk assessment will be based on the existing quantified skin cancer risk with a
hazard identification for internal cancer.
The Safe Drinking Water Act Amendments of 1996, in Section 1412(b)(12)(A) pertaining to
arsenic, directed EPA to take the following actions:
Develop an arsenic research strategy within 180 days of enactment;
Propose a revised MCL by January 1, 2000;
Issue a final regulation by January 1, 2001;
Assess health effects for sensitive populations;
List both compliance and variance treatment technologies for small systems;
Evaluate the incremental costs and benefits of different regulatory options,
accounting for the changes that may result from implementation of other rules;
Issue an MCL that maximizes health benefits at a cost that is justified by the
benefits;
Review MCLs every six years or sooner.
The 1996 amendments also made the following changes:
The effective date of MCLs is three-to-five years after promulgation of the final
rule.
Compliance can be achieved by use of point-of-use (POU) or point-of-entry (POE)
devices that are maintained by the public water system;
Congress authorized $2.5 million per year from 1997-2000 for the studies. In 1996 and 1997,
Congress appropriated $1 million each year for arsenic research.
EPA is proposing this rule to meet the deadlines specified in the 1996 Amendments. At the same
time, EPA is proceeding with its Arsenic Research Plan, which will address a variety of issues
related to exposure, treatment, and health effects.1
NRC Report: As mentioned above, In 1996, EPA requested that the National Research Council
(NRC) of NAS conduct an independent review of the arsenic toxicity data and evaluate the
scientific validity of EPA's 1988 risk assessment for arsenic in drinking water. In addition, NRC
was asked to review EPA's current criteria (50 jig/L and 0.018 |ig/L), evaluate use of recent
Taiwan data and other studies to assess the carcinogenic and non-carcinogenic health effects of
arsenic, and recommend changes to EPA's risk characterization for arsenic. NRC issued its
lrThe Arsenic Research Plan is published at http://www.epa.gov/ORDAVebPubs/final/arsenic.pdf.
Chapter 2, Need for the Proposal 2-6 Proposed Arsenic in Drinking Water Rule RIA
-------
report on March 23, 1999. The report had four conclusions:
The Taiwan studies provide the best available evidence on the human health effects
of arsenic, and are supported by studies in Chile and Argentina which report
similar results. These studies show that chronic ingestion of inorganic arsenic at
high doses causes bladder and lung cancer, as well as skin cancer.
Noncancer effects from chronic ingestion of arsenic have been detected at doses of
0.01 mg/kg per day and higher.
There is a need for more research to characterize the dose-response relationship
for both cancer and non-cancer endpoints, especially at low doses.
The current 50 |ig/L MCL is not adequately protective of human health, and
therefore requires downward revision as promptly as possible.2
2.4 Rationale for the Regulation
This section discusses the economic rationale for choosing a regulatory approach to address the
public health consequences of drinking water contamination. EPA proves the economic rationale
in response to Executive Order Number 12866, Regulatory Planning and Review, which states:
[EJach agency shall identify the problem that it intends to address (including, where
applicable, the failures of the private markets or public institutions that warrant new
agency action) as well as assess the significance of that problem (Section 1, b(l)).
In addition, 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 presented in this section should not be
interpreted as the agency's approach to implementing the Safe Drinking Water Act (SDWA).
Instead, it is the agency's justification, as required by the Executive Order, for a regulatory
approach to this public health issue.
2.4.1 Statutory Authority
Section 1412(b)(l)(A) of the Safe Drinking Water Act 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 occurring once in public water systems
(PWSs), at a frequency and level of public health concern and that present a meaningful
opportunity for health risk reduction for persons served by PWSs. This general provision is
supplemented with an additional requirement under Section 1412(b)(12)(A) that EPA propose a
revised MCL for arsenic by January 1, 2000, and issue a final regulation by January 1, 2001.
The NRC report is available at http://www.nap.edu/readingroom/enter2.cgi70309063337.html
Chapter 2, Need for the Proposal 2-7 Proposed Arsenic in Drinking Water Rule RIA
-------
2.4.2 Economic Rationale for Regulation
In addition to the statutory directive to regulate arsenic, there is also economic rationale for
government regulation. In a perfectly competitive market, market forces guide buyers and sellers
to attain the best possible social outcome. A perfectly competitive market occurs when there are
many producers of a product selling to many buyers, and both producers and buyers have
complete knowledge regarding the products of each firm. There must also be no barriers to entry
in the industry, and producers in the industry must not have any advantage over potential new
producers. Several factors in the public water supply industry do not satisfy the requirements for
a perfect market and lead to market failures that may require regulation.
First, water utilities are natural monopolies. A natural monopoly exists because it is not
economically efficient to have multiple suppliers competing to build multiple systems of pipelines,
reservoirs, wells, and other facilities.3 Instead, a single firm or government entity performs these
functions generally under public control. Under monopoly conditions, consumers are provided
only one level of service with respect to the quality of the product, in this case drinking water
quality. If consumers do not believe the margin of safety in public health protection is adequate,
they cannot simply switch to another water utility or perceived higher quality source of supply
(e.g., bottled water) without incurring additional cost.
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 risks potentially
posed by trace quantities of drinking water contaminants involve analysis and distillation of
complex toxicological data and health sciences. EPA recently developed the Consumer
Confidence Report rule to make water quality information more easily available to consumers.
The Consumer Confidence Report rule requires community water systems to mail their customers
an annual report on local drinking water quality. However, consumers will still have to analyze
this information for its health risk implications. Even if informed consumers are able to engage
utilities regarding these health issues, the costs of such engagement, known as "transaction costs,"
(in this case measured in personal time and commitment) present another 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 of public water
supply. The regulations set minimum performance requirements for all public water supplies in
order to reduce the risk confronted by all consumers from exposure to drinking water
contaminants. SDWA regulations are not intended to restructure market mechanisms or to
establish competition in supply. Rather, SDWA standards establish the level of service to be
3Mansfield (1975) states that natural monopolies exists because the average cost of producing the product
reaches a minimum at an output rate that is enough to satisfy the entire market at a price that is profitable.
Multiple producers competing would produce the product at higher than minimum long-run average cost.
competition to achieve lower average costs would drive prices down until a single supplier is victorious.
Chapter 2, Need for the Proposal 2-8 Proposed Arsenic in Drinking Water Rule RIA
-------
provided in order to better reflect public preference for safety. The Federal regulations remove
the high information and transaction costs by acting on behalf of all consumers in balancing the
risk reduction and the social costs of achieving this reduction.
Chapter 2, Need for the Proposal 2-9 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 2, Need for the Proposal 2-10 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 3: Consideration of Regulatory Alternatives
3.1 Regulatory Approaches
The Safe Drinking Water Act (SDWA) establishes EPA's responsibility for ensuring the quality of
drinking water, and defines the mechanisms available to the Agency to protect public health.
Specifically, the SDWA requires EPA to set enforceable MCLs when technically or economically
feasible or otherwise establish treatment technique requirements for specific contaminants in
drinking water. In meeting this mandate, EPA sets water quality standards by identifying which
contaminants should be regulated and establishing the levels of the contaminant water systems
must attain. This section discusses the approach EPA used in determining the regulatory
alternatives that have been considered.
3.1.1 Determining the Standard
In regulating a contaminant, EPA first sets a maximum contaminant level goal (MCLG), which
establishes the contaminant level at which no known or anticipated adverse health effects occur.
MCLGs are non-enforceable health goals. For this rulemaking, EPA is proposing an MCLG of
zero. EPA then sets an enforceable maximum contaminant level (MCL) as close as technologically
possible to the MCLG. In addition, EPA may use its discretion in setting the MCL by choosing
an MCL that is protective of public health while also insuring that the quantified and non-
quantified costs are justified by the quantified and non-quantified benefits of the rule. For this
rulemaking, EPA is proposing an MCL of 5 |ig/L. The following sections describe the process by
which EPA determined both the MCLG and the MCL.
3.1.2 Determining the MCLG
Carcinogens: For many years, Congress supported a goal of zero tolerance for carcinogens in
food and water, and that goal was incorporated into the SDWA of 1974. Under this policy,
contaminants that are classified as probable human carcinogens have had MCLGs set at zero.
EPA's Office of Science and Technology (OST) (in the Office of Water) develops a cancer risk
range that quantifies the probability that a person will develop cancer during a lifetime of ingesting
water containing the regulated contaminant. Mathematical models have been used to calculate the
drinking water concentration that would lead to excess cancer risks of 10"5 to 10"6 from exposure
to the carcinogen.
Data used in risk estimates usually come from lifetime exposure studies in animals. To predict the
risk for humans, the oral doses used in animal studies are corrected for differences in animal and
human size and surface area.
In 1986, EPA published Guidelines for Carcinogen Risk Assessment in the Federal Register (51
FR 33992). At that time EPA's default assumptions included low-dose linearity to extrapolate the
cancer risk range, which assumes that carcinogenic effects do not exhibit a threshold and that
carcinogens pose risks to humans at any concentration. EPA proposed revised Guidelines for
Carcinogen Risk Assessment in 1996 (61 FR 17960).
Chapter 3, Consideration of Regulatory Alternatives 3-1 Proposed Arsenic in Drinking Water Rule RIA
-------
Noncarcinogens: MCLGs for noncarcinogens are based on Reference Doses (RfDs) and their
Drinking Water Equivalent Levels (DWELs).
The Reference Dose (RfD, formerly the Acceptable Daily Intake, or ADI), estimates the daily
amount of chemical a person, including sensitive humans, can ingest over a lifetime with little risk
of causing adverse health effects. RfDs are usually expressed in milligrams of chemical per
kilogram of body weight per day (mg/kg/day). Data from chronic (usually two years) or sub-
chronic (usually 90 days) studies of humans or animals provide estimates of the No- or- Lowest-
Observed-Adverse-Effect Level (NOAEL or LOAEL). The NOAEL (or LOAEL) is divided by a
total uncertainty factor (UF) of 1 to 10,000 to obtain the RfD. In the final National Primary
Drinking Water Regulations published on January 30, 1991 (56 FR 3532), EPA applies a UF of 1,
3, or 10 when a NOAEL from a human study is used to account for intraspecies variation and an
uncertainty factor of 100 to a human LOAEL to account for lack of a NOAEL and for species
variation. The UFs provide a margin for variations in species responses, data gaps, and less than
lifetime exposures. Scientific judgement is used to select the total UF factor for specific risk
assessments.
The Drinking Water Equivalent Level (DWEL) is calculated by multiplying the RfD by an
assumed adult body weight of 70 kg (approx. 154 pounds) and dividing by an average adult water
consumption of 2 liters per day (L/day). The DWEL assumes that 100 percent of the exposure
comes from drinking water. The MCLG is then determined by multiplying the DWEL by the
percentage of the total daily exposure contributed by drinking water (relative source
contribution), set at 20% by default when adequate data are not available, but set between 20 and
80 percent when adequate data are available to estimate exposure. Based on the 1993 RfD (1993
Draft Criteria) for arsenic (0.3 jig/kg/day), the calculated DWEL would be 0.3 jig/kg/day times
70 kg divided by 2 L/day, or 10 |ig/L. Due to the three-fold uncertainties noted in the IRIS file
on arsenic, the DWEL could be 3 to 30 jig/L. It should be noted that the toxicological studies
used to determine the effect level and the derivation of the RfD are different from the analysis
conducted in 1975. Additionally, the current policy on relative source contribution, including the
default policy are also different from those used in 1975.
3.1.3 Determining an MCL
Once an MCLG is established, EPA sets an enforceable standard - in most cases, a Maximum
Contaminant Level (MCL). The MCL is the maximum permissible level of a contaminant in water
that is delivered to any user of a public water system. EPA must set the MCL as close to the
MCLG as feasible. The SWDA defines feasible as the level that may be achieved with the use of
the best available technology, treatment techniques, and other means which EPA finds are
available (after examination for efficacy under field conditions), taking cost to large systems into
consideration.
After determining an MCL based on affordable technology for large systems, EPA must complete
an economic analysis to determine whether the benefits of the standard justify the costs. If not,
EPA may adjust the MCL for a particular class or group of systems to a level that "maximizes
health risk reduction benefits at a cost that is justified by the benefits" (§1412(b)(6)).
3.1.4 Variances and Exemptions
Chapter3, Consideration of Regulatory Alternatives 3-2 Proposed Arsenic in Drinking Water Rule RIA
-------
The 1996 SDWA identifies two classes of technologies for small systems: compliance and
variance technologies. A compliance technology is one that achieves compliance with the MCL
or treatment technique requirement. The 1996 Amendments require EPA to list affordable
compliance technologies for three categories of small systems: those serving 25 to 500 people,
those serving 501 to 3,300 people, and those serving 3,301 to 10,000 people. If EPA cannot
identify an affordable compliance technology for a particular system category, it must then
identify a variance technology instead. The variance technology must achieve the maximum
reduction that is affordable, considering the size of the system and the quality of the source water,
and must be protective of public health. If EPA lists such a variance technology, small systems
will be eligible to apply to the States for a small system variance. States are authorized to grant
variances from standards for systems serving up to 3,300 people if the system cannot afford to
comply with a rule and the system installs the EPA-approved variance technology. States can
grant exemptions to systems serving 3,301 - 10,000 people with EPA approval.
The Agency published draft national-level affordability criteria in the August 6, 1998, Federal
Register (63 FR 42302). The draft criteria discuss affordable treatment technology
determinations for contaminants regulated before 1996. An average expenditure level of up to
$500 per year was considered affordable for those contaminants. Since EPA identified treatment
technologies for all pre-1996 contaminants with average per household costs below $500 per
year, the Agency did not list any small system variance technologies.
EPA expects the national-level affordability criteria to be lower than $500 per household per year
for the arsenic rule because water rates are currently increasing faster than median household
income, and because the baseline for annual water bills will rise as treatment is installed to comply
with regulations promulgated after 1996. As part of this RIA, a household level cost analysis was
done to determine if EPA needs to list variance technologies (Chapters 6 and 8).
3.1.5 Analytic Methods
Determination of an MCL depends on the ability of laboratories to reliably measure the
contaminant at the MCL. The SDWA directs EPA to set an MCL "if in the judgement of the
Administrator, it is economically and technologically feasible to ascertain the level of such
contaminant in water in public water systems (Section 1401 (l)(c)(ii))." EPA must therefore
evaluate the available analytical methods to determine a Practical Quantification Level (PQL),
which is the minimum reliable quantification level that most laboratories can be expected to meet
during day-to-day operations.
EPA has approved several analytical methods to support compliance monitoring of arsenic at the
current MCL (40 CFR 141.23). In 1994, EPA evaluated available data and determined the PQL
for arsenic to be 2.0 jug/L at an acceptable limit of ± 40%. In its July 1995 report, EPA's Science
Advisory Board recommended that EPA set the PQL for arsenic using acceptance limits similar to
those applied for other inorganics (± 20% or ± 30 %). Based on more recent information and
these acceptance limits, EPA has established a 1999 PQL for arsenic of 3 //g/L (EPA, MOM
1999) with an acceptance limit of ± 30 percent. While the PQL represents a stringent target for
laboratory performance, the Agency believes most laboratories, using appropriate quality
assurance and quality control procedures, have the capacity to achieve this level on a routine
Chapters, Consideration of Regulatory Alternatives 3-3 Proposed Arsenic in Drinking Water Rule RIA
-------
basis. Available data suggest that 75 percent of EPA Regional and State laboratories and 62
percent of non-EPA laboratories were capable of achieving acceptable results at 3 jug/L.
3.2 Regulatory Alternatives Considered and Proposed Rule
This section describes the components of the proposed rule and the alternatives that were
considered by the Agency.
3.2.1 Applicability
The Agency investigated applying both the monitoring and treatment requirements of the
proposed rule to both community water systems (CWS) and non-transient non-community water
systems (NTNC). A CWS is defined as a system that provides piped water to at least 25 people or
with at least 15 service connections year-round. A NTNC is a public water systems that is not
defined as a CWS and that regularly serves at least 25 of the same people for at least six months
of the year. After considering the costs and benefits of the proposed rule with regard to both
CWSs and NTNCs, EPA proposes to require CWS to comply with all facets of the proposed rule,
while only requiring NTNCs to comply with the monitoring components of the rule. The benefit-
cost analysis upon which this decision is based is provided in Chapters Five, Six, and Seven of this
RIA. Transient non-community systems, which provide potable water to continuously changing
populations, will not be subject to the proposed rule. The rule applies to CWSs and NTNCs that
produce water from either primarily ground or surface water sources.
3.2.2 MCL
EPA considered a range of MCLs in developing the proposed Arsenic Rule, including MCLs of 3,
5, 10, and 20 //g/L. EPA evaluated the following five factors to determine the proposed
Maximum Contaminant Level (MCL):
the analytical capability and laboratory capacity,
likelihood of water systems choosing various compliance technologies for several
sizes of systems based on source water properties,
the national occurrence of arsenic in water supplies.
quantified and non-quantified costs and health risk reduction benefits likely to
occur at the MCLs considered, and
the effects on sensitive subpopulations.
An MCL of 3 //g/L was considered since this is the level that has been determined to be as close
to the MCLG as is feasible. However, the Agency is using its discretionary authority in Section
1412(b)(6)(A) to consider setting MCL at a less stringent level. The statute requires that the
alternative less stringent level be one which maximizes health risk reduction at a level where costs
and benefits are balanced. As a result, EPA considered the alternative MCL options of 5, 10, and
20 jug/L. These alternative MCL options were considered because they also provide
assurance that the residual risk for both bladder and lung cancer endpoints will be in the 10"4
range, but at lower anticipated national costs.
3.2.3 Monitoring
Chapters, Consideration of Regulatory Alternatives 3-4 Proposed Arsenic in Drinking Water Rule RIA
-------
The current monitoring requirements for arsenic (40 CFR 141.23(1)) apply to community water
systems only. EPA is proposing to change the current monitoring requirements and require
systems to monitor for arsenic in accordance with the provisions of 40 CFR 141.23(c), the
Standard Monitoring Framework (SMF). This change will make the arsenic requirements
consistent with the requirements for inorganic contaminants (lOCs) regulated under the Phase
II/V regulations. The proposal would make the following changes to the monitoring requirements
for arsenic:
NTNC systems will be required to monitor for arsenic for the first time.
MCL exceedances will trigger quarterly monitoring, as opposed to the current
requirements for three additional samples within one month when exceedances occur.
The State will determine when the system is "reliably and consistently" below the
MCL, after a minimum number of samples following an exceedance (two sample for
ground water systems and four for surface water systems), and can return to the
default sampling frequency. (Currently, the system automatically returns to the default
monitoring frequency when a minimum of two consecutive samples are below the
MCL).
States may grant a nine year monitoring waiver to a system, if it finds that arsenic
detections are the result of natural occurrence and not from human activity.
(Currently, no monitoring waivers are permitted).
3.2.4 Compliance Technologies and Variances
EPA reviewed several technologies as BAT candidates for arsenic removal. Those technologies
capable of removing arsenic from source water that fulfill the SDWA requirements for BAT
determinations for arsenic are as follows:
ion exchange;
activated alumina;
reverse osmosis ;
coagulation assisted microfiltration;
modified coagulation/filtration;
modified lime softening;
point-of-use RO/AA; point-of-entry AA ; and
oxidation/filtration (including greensand filtration).
EPA has further determined that these technologies are affordable for all system size categories
and has therefore not identified a variance technology for any system size or source water
combination at the proposed MCL.
3.2.5 Monitoring Waivers
Under the proposed Arsenic Rule (§141.23(c)(3)), States may grant a nine year monitoring
waiver from sampling requirements to water systems based on the analytical results from previous
sampling and a vulnerability assessment or the assessment from an approved source water
assessment program (provided that the assessments were designed to collect all of the necessary
information needed to complete a vulnerability assessment for a waiver). States issuing waivers
Chapter3, Consideration of Regulatory Alternatives 3-5 Proposed Arsenic in Drinking Water Rule RIA
-------
must consider the requirements in 40 CFR 141.23(c)(2)-(6). In order to qualify for a waiver,
there must be three previous samples from a sampling point (annual for surface water and three
rounds for ground water) with analytical results reported below the proposed MCL (i.e., the
reporting limit must be < 0.005 mg/L). The use of grandfathered data collected after January 1,
1990, that is consistent with the analytical methodology and detection limits of the proposed
regulation may be used for issuing sampling point waivers.
The current arsenic regulations §141.23(l)-(q) do not permit the use of monitoring waivers.
However, a State could now use the analytical results from the three previous compliance periods
(1993 to 1995, 1996 to!998, and 1999 to 2001) to issue ground water sampling point waivers.
Surface water systems must collect annual samples so a State could use the previous three years
sampling data (1999, 2000, and 2001) to issue sampling point waivers. One sample must be
collected during the nine-year compliance cycle that the waiver is effective, and the waiver must
be renewed every nine years. Vulnerability assessments must be based on a determination that the
water system is not susceptible to contamination and arsenic is not a result of human activity (i.e.,
it is naturally occurring).
Although the approved analytical methods can measure to 0.005 mg/L, not all States have
required systems to report arsenic results below 50 |ig/L. In this case, the States would not have
adequate data to grant waivers until enough data is available to make the determinations.
EPA believes that some States may have been regulating arsenic under the proposed standardized
inorganic framework. If so, those States will have to ensure that existing monitoring waivers
have been granted using data reported below the new proposed MCL. Otherwise States will have
to notify the systems of the new lower reporting requirements that need to be met to qualify for a
waiver for the proposed MCL.
3.2.6 Implementation
The following schedule is proposed for implementation of the proposed rule:
States must submit applications for primacy revisions within two years after
promulgation, unless the State requests and is granted a two year extension.
The rule will be effective three years after promulgation (January 1, 2004).
All systems must complete initial sampling by December 31, 2004.
If capital improvements are needed to achieve compliance, systems may apply to the
State for a two year extension.
If EPA makes a national finding that capital improvements are necessary at the majority of
systems, then the rule will become effective five years after promulgation (January 1, 2006). In
this case, systems will have to complete the initial round of monitoring by December 31, 2007 (for
ground water systems), or December 31, 2006 (for surface water systems.)
Chapter 3, Consideration of Regulatory Alternatives 3-6 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 4: Baseline Analysis
4.1 Introduction
This chapter presents baseline information to describe the operational and financial characteristics
of water systems in the absence of the proposed Arsenic Rule. The baseline information provides
a basis for EPA's analysis of the costs, benefits and economic impacts of the regulatory options
considered. This chapter includes data on the number of water systems regulated, the population
affected, current treatment practices, raw and treated water quality, and socio-economic impacts.
In addition, this chapter provides information to support the Agency's national-level affordability
determination.
The baseline is assumed to be current conditions, as reflected in the most recent available data. In
some cases, changes in the industry have occurred or will occur that are not reflected in the
available data; for example, changes in operations induced by a regulation that will take effect
prior to the Arsenic Rule.
4.2 Industry Profile
4.2.1 Definitions
According to EPA's definition, public water systems (PWSs) include community water systems
(CWSs) and non-community water systems (NCWSs). NCWSs are further classified as either
transient or non-transient. The proposed rule will affect all public water systems except for
transient non-community water systems. The following definitions will help the reader follow the
discussion in this chapter:
Public water systems (PWS) serve 25 or more people or has 15 or more service
connections and operates at least 60 days per year. A PWS can be publicly-owned or
privately-owned.
Community water systems (CWS) serve at least 15 service connections used by year-
round residents, or regularly serve at least 25 year-round residents.
Non-community water systems (NCWS) do not have year-round residents, but serve at
least 15 service connections used by travelers or intermittent users for at least 60 days
each year, or serve an average of 25 individuals for at least 60 days a year.
Non-transient non-community water systems (NTNC) serve at least 25 of the same
persons over six months per year (e.g., factories, schools, office buildings, and hospitals).
Transient non-community water systems (TNC) serve fewer than 25 of the same
persons over six months per year (e.g., many restaurants, rest stops, parks).
Public water systems are also classified by their water source as being surface water (e.g., drawn
from lakes, streams, rivers, etc.) or ground water (e.g., drawn from wells or springs).
Chapter 4, Baseline Analysis 4-1 Proposed Arsenic in Drinking Water Rule RIA
-------
4.2.2 Sources of Industry Profile Data
EPA uses two primary sources of data to characterize the universe of ground water systems: the
Safe Drinking Water Information System (SDWIS) and the Community Water System Survey
(CWSS).
EPA's SDWIS contains data on all PWSs as reported by States and EPA regions. This source
reflects both mandatory and optional reporting components. States must report the system
location, system type (CWS, NTNC, or TNC), primary raw water source (ground water or
surface water), and violations. Optional reporting fields include type of treatment and ownership
type. Because providing some data is discretionary, EPA does not have complete data on every
system for these parameters. This is particularly common for non-community systems.
The second source of information, the CWSS, is a detailed survey of surface and ground water
CWSs conducted by EPA in 1995 and published in 1997 (EPA, 1997). The CWSS is stratified to
represent the complete population of CWSs across the U.S. The CWSS includes information
such as revenues, expenses, treatment practices, source water protection measures, and plant
capacity. There is no equivalent survey such as the CWSS to define treatment practices in non-
community water systems.
4.2.3 Number and Size of Public Water Systems
Exhibit 4-1 shows the number of systems in the U.S. by source water (ground or surface) and
system size (measured by the number of people served), based on the December 1998 SDWIS
data. In the U.S. there are a total of 63,984 ground water systems and 11,843 surface water
systems, including CWSs and NTNCs. All are potentially affected by the proposed Arsenic Rule.
Some ground water sources (e.g., riverbank infiltration/galleries) are directly impacted by adjacent
source water bodies and are separately identified in SDWIS as ground water under the direct
influence of surface water (GWUDI). Since these systems would have similar occurrence as
surface water systems, GWUDI systems are considered surface water systems in this analysis.
SDWIS also provides system data by ownership. As previously described, PWSs include both
publicly-owned and privately-owned systems. This detail is also provided in Exhibit 4-1, where
any systems referred to as "other" in the SDWIS database has been presented as a privately-
owned system.
The majority (95%) of PWSs are small systems that serve fewer than 10,000 people. 89 percent
of PWSs serve 3,300 people or fewer; 77 percent serve fewer than 1,000 people; 67 percent serve
fewer than 500 people; and 34 percent serve fewer than 100 people.
Chapter 4, Baseline Analysis 4-2 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 4-1
Total Number of Systems by Size, Type, and Ownership
SOURCE
<100
101-
500
501-
1,000
1,001-
3,300
3,301-
10,000
10,001-
50,000
50,001-
100,000
100,001-
1,000,000
TOTAL
CWS
Ground Water
Public
Private
Total
Surface Water
Public
Private
Total
Total
1,335
12,942
14,277
394
698
1,092
15,369
4,678
10,380
15,058
1,117
886
2,003
17,061
2,868
1,821
4,689
917
303
1,220
5,909
4,167
1,547
5,714
2,012
408
2,420
8,134
1,993
466
2,459
1,656
188
1,844
4,303
1,011
205
1,216
1,436
171
1,607
2,823
105
26
131
260
40
300
431
50
11
61
217
44
261
322
16,207
28,303
44,510
8,009
3,053
11,062
54,352
NTNCWS
Ground Water
Public
Private
Total
Surface Water
Public
Private
Total
Total
1,725
7,965
9,690
58
213
271
9,961
3,108
3,930
7,038
63
232
295
7,333
1,163
815
1,978
19
87
106
2,084
337
355
692
24
56
80
772
23
39
62
6
17
23
85
9
5
14
3
1
4
18
0
0
0
1
0
1
1
0
0
0
1
0
1
1
6,365
13,109
19,474
175
606
781
20,255
Source: Safe Drinking Water Information System (SDWIS), December 1998 freeze.
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA
-------
4.2.4 System Size and Population Served
All PWSs are potentially subject to the proposed requirements of the Arsenic Rule, with the
exception of TNCs. The majority of systems to be regulated are community water systems, which
also serve, on average, more people than NTNCs. Exhibit 4-2 provides information on the
average populations served by CWSs for each system size category.
Exhibit 4-2
Total Population Served of Water Systems by
Source Water, System Type, and Service Population Category
Service
Population
Category
< 100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
> 1,000,000
Total
Community
Ground water
859,777
3,741,017
3,457,163
10,631,422
14,095,015
25,004,779
8,609,455
14,575,556
2,855,494
83,829,678
Surface Water
61,450
570,448
921,449
4,797,855
10,995,980
36,819,575
20,500,370
65,375,183
28,658,586
168,700,896
Non-Transient
Non-Community
-
-
-
-
-
-
-
-
-
31,968,181
Source: Safe Drinking Water Information System (SDWIS), December 1998 freeze.
Those NTNCs determined to be affected by the Arsenic Rule are presented in Exhibit 4-3 by type
of system. The NCWSs are much smaller than CWSs on average and vary substantially in their
characteristics. Schools account for more than half of the affected NCWSs (8,414 of 16,410),
followed by office parks (950), daycare centers (809) and food manufacturing facilities (768).
Prisons serve the largest number of people on average (1,820). All other system types serve an
average of 500 people or fewer.
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 4-3
Characteristics of NCWSs Affected by the Arsenic Rule
Service Area Type
Daycare Centers
Highway Rest Areas
Hotels/Models
Interstate Carriers
Medical Facilities
Mobile Home Parks
Restaurants
Schools
Service Stations
Summer Camps
Water Wholesalers
Agricultural
Products/Services
Airparks
Construction
Churches
Campgrounds/RV Parks
Fire Departments
Federal Parks
Forest Service
Golf and Country Clubs
Landfills
Mining
Amusement Parks
Military Bases
Migrant Labor Camps
Misc. Recreation Services
Nursing Homes
Office Parks
Prisons
Retailers (Non-food related)
Retailers (Food related)
State Parks
Non-Water Utilities
Manufacturing: Food
Manufacturing: Non-Food
Sum
Weighted Average
# of Systems
809
15
351
287
367
104
418
8,414
53
46
266
368
101
99
230
123
41
20
107
116
78
119
159
95
33
259
130
950
67
695
142
83
497
768
0
16,410
Avg. Pop.
76
407
133
123
393
185
370
358
230
146
173
76
60
53
50
160
98
39
42
101
44
113
418
395
63
87
107
136
1,820
174
322
165
170
372
0
981
Design Flow/System (mpd)
0.005114
0.008895
0.01892
0.002853
0.116623
0.026164
0.003907
0.033269
0.00511
0.021758
0.163659
0.019882
0.002608
0.000914
0.005287
0.021428
0.018623
0.006528
0.001351
0.011796
0.005253
0.012322
0.017084
0.069538
0.010212
0.002468
0.04107
0.00769
0.532196
0.003787
0.005789
0.004828
0.013288
0.045384
0.015723
0.028
Avg. Flow/System (mpd)
0.001068
0.001970
0.004540
0.00056
0.033945
0.006498
0.000793
0.008476
0.001067
0.005299
0.049381
0.004796
0.000507
0.000159
0.001108
0.00521
0.004461
0.001399
0.000245
0.002692
0.0011
0.002825
0.004055
0.019159
0.002295
0.000477
0.0107
0.001677
0.182
0.000766
0.001225
0.001002
0.003071
0.01195
0.003785
0.007
Source: EPA, 1999a. Geometries and Characteristics of Public Water Systems, Table 7.5 and Table 7.6.
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA
-------
4.2.5 Number of Entry Points
If water systems employ more than one water supply source, they may have more than one
treatment facility. For estimation purposes this analysis assumes a treatment facility at every entry
point to the distribution system. As a result, the total number of entry points is an important
determinant of compliance costs. Exhibit 4-4 presents the distribution of entry points per ground
water CWS by system service population category.
Exhibit 4-4
Average Number of Entry Points per Ground Water System
Upper Bound
95%
Confidence
Service Population Category
<100
101-
500
501-
1,000
1,001 -
3,300
3,301 -
10,000
10,001 -
50,000
50,001 -
100,000
> 100,000
Percentile
Mean
5th
50th (median)
95th
1
1
1
2
1
1
1
3
2
1
1
3
2
1
1
5
2
1
2
5
4
1
3
12
6
1
4
22
9
1
5
28
Source: EPA, 1999a. Geometries and Characteristics of Public Water Systems, Table 5.2.
In this respect, surface water systems are unlike ground water systems in that little variation in the
number of entry points was reported among surface water systems. Even for large population
categories, the majority of surface water systems reported only one or two entry points. (EPA,
1999a). This finding was supported by data recently collected from the Information Collection
Request for large surface water systems.
4.2.6 Number of Households
Another method for estimating the effect of regulations on customers is to determine the cost per
household. This measure is often used instead of per capita cost because it is a more accurate
representation of how customers are billed, per household, not per person. Exhibit 4-5 shows the
average number of connections for CWSs by size, water source, and ownership. The number of
connections, or households, ranges from an average of 30 residential service connections for
CWSs serving fewer than 100 people to an average of more than 56,000 residential connections
for CWSs serving more than 100,000 people (EPA, 1999b. Drinking Water Baseline Handbook,
B4.2.2(b)).
Household consumption does not vary substantially across size category or ownership type. The
mean water consumption ranges from 81,000 gallons per year to 127,000 gallons per year per
household.
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 4-5
Water Consumption per Residential Connection and
Number of Residential Connections per System
Population
< 100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
_ . Mean Water Connections per System
System
' Consumption* by Source**
ype (kgal/yr) Ground Surface
Public
Private
Public
Private
Public
Private
Public
Private
Public
Private
Public
Private
Public
Private
Public
Private
81
92
93
110
97
88
82
102
87
124
108
110
122
96
127
114
46
29
128
99
314
321
663
601
2,130
1,572
6,075
7,432
20,278
27,423
59,969
55,047
24
37
192
105
315
321
774
565
2,304
2,120
6,737
6,862
19,427
24,432
69,985
76,833
Source: *EPA, 1997. CWSS, Vol. II: Detailed Summary Result Tables and
Methodology Report, Table 1-14;
** EPA, 1995. CWSS: mean residential connections exclude purchased
systems.
4.2.7 Production Profile
Exhibit 4-6 shows the average design capacity (in thousands of gallons) of CWS plants by source,
ownership, and system size categories. Design capacity is the maximum amount of water a plant
can deliver. Exhibit 4-7 provides the daily production of CWSs (in thousands of gallons) for the
same categories. Daily production is the average amount of water a plant delivers in a day.
Chapter 4, Baseline Analysis
4-7
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 4-6
Design Capacity of CWS Plants
by Source, Ownership, and System Size
(Thousands of Gallons)
Primary Source/
Ownership Type
Ground water
Public
Private
Purchased-Public
Purchased-Private
Surface Water
Public
Private
Purchased-Public
Purchased-Private
GW under
influence
Public
Private
Purchased-Public
Purchased-Private
Service Population Category
<25
6.27
4.84
6.50
-
0.89
1.30
1.14
3.19
0.04
1.12
-
-
-
-
-
25-100
21.86
29.46
21.34
5.71
4.99
20.32
25.79
18.13
5.71
4.99
22.16
33.29
21.53
-
2.54
101-500
86.86
123.67
77.30
27.37
24.78
92.60
130.90
75.69
29.01
24.65
87.20
111.32
81.77
30.21
29.83
501-1.000
251.0
305.0
232.1
81.4
79.5
239.3
318.2
214.2
81.8
73.6
247.5
291.2
227.4
97.1
94.3
1,001-
3.300
619.5
740.3
560.6
223.0
200.6
617.9
807.8
527.3
241.1
213.8
631.6
760.0
618.5
209.3
-
3,301-
10.000
1864
2152
1683
801
824
1818
2218
1582
854
719
1779
2077
1802
461
905
10,001-
50.000
6673
7365
6347
3380
2748
6682
7887
6165
3698
2933
7499
8992
-
2319
-
50,001-
100.000
20785
22614
18234
19796
8690
19707
22337
15869
13206
12788
18482
20195
-
-
-
100,001-
1.000.000
67379
67994
75629
26765
-
69224
77298
61381
43650
29270
-
-
-
-
-
>1. 000.000
392939
401175
-
-
-
554759
584889
296609
-
-
-
-
-
-
-
Source: Drinking Water Baseline Handbook, Table B1.5.3.
-------
Exhibit 4-7
Daily Production of CWS Plants
by Source, Ownership, and System Size
(Thousands of Gallons)
Primary Source/
Ownership Type
Ground water
Public
Private
Purchased-Public
Purchased-Private
Surface Water
Public
Private
Purchased-Public
Purchased-Private
GW under
influence
Public
Private
Purchased-Public
Purchased-Private
Service Population Category
<25
1.35
0.96
1.39
-
0.85
0.39
0.28
0.95
0.04
1.06
-
-
-
-
-
25-100
5.33
6.72
4.80
5.11
4.54
6.91
7.51
6.15
5.11
4.54
5.41
7.70
4.85
-
2.36
101-500
24.40
33.20
20.30
23.50
21.60
33.70
41.60
28.60
24.90
21.50
24.50
29.50
21.30
25.90
25.9
501-1.000
78.50
90.50
69.30
68.20
67.30
90.70
106.20
87.30
68.50
62.40
77.30
86.00
67.70
80.90
79.4
1,001-
3.300
212
243
18
182
166
244
284
230
197
176
217
250
207
171
-
3,301-
10.000
715
796
635
634
656
753
823
748
675
575
679
765
686
370
719
10,001-
50.000
2,914
3,129
2,802
2,585
2,119
2,932
3,133
3,225
2,821
2,258
3,313
3,907
-
1,789
-
50,001-
100.000
10,187
10,900
9,121
14,496
6,502
9,069
9,387
8,907
9,766
9,472
8,951
9,611
-
-
-
100,001-
1.000.000
37,224
37,095
44,760
19,455
-
33,667
34,749
38,094
31,351
21,215
-
-
-
-
-
>1. 000.000
259,751
267,256
-
-
-
295,680
293,439
206,950
-
-
-
-
-
-
-
Source: EPA, 1999b. Drinking Water Baseline Handbook, Table B1.5.1.
-------
4.2.8 Treatment Profile
Exhibit 4-8 below presents information regarding in-place treatment technologies that affect
arsenic concentrations in delivered water. The current treatment in place will determine the likely
remedy that systems will select in order to come into compliance with the new MCL.
Exhibit 4-8
Percentage of CWSs with Various Treatments In-Place
Primary
Source/ Type
of Treatments
Service Population Category
<100
101-500
501-
1,000
1,001-
3,300
3,301-
10,000
10,001-
50,000
50,001-
100,000
100,001-
1,000,000
> 1,000,000
Ground Water Systems
Ion Exchange
Reverse
Osmosis
Coagulation/
Flocc.
Lime/Soda Ash
Softening
0.7%
0.0%
1 .5%
2.1%
1 .6%
1 .2%
5.4%
3.7%
3.8%
0.0%
4.2%
4.1%
1 .9%
0.9%
3.4%
5.2%
4.6%
1 .2%
8.1%
7.0%
3.3%
0.7%
15.1%
12.2%
1 .2%
1 .2%
24.2%
17.4%
0.0%
0.0%
25.2%
32.4%
-
-
-
-
Surface Water Systems
Ion Exchange
Reverse
Osmosis
Coagulation/
Flocc.
Lime/Soda Ash
Softening
0.0%
0.0%
27.5%
3.9%
0.0%
0.0%
52.6%
8.1%
0.0%
0.0%
70.2%
20.5%
0.0%
0.0%
78.5%
17.5%
0.0%
0.0%
95.4%
10.8%
0.0%
0.0%
94.5%
6.9%
0.0%
0.0%
93.7%
5.7%
0.0%
0.0%
99.5%
5.1%
-
-
-
-
Source: EPA, 1999a. Geometries and Characteristics of Public Water Systems, Tables 6-1 and 6-2.
4.2.9 Financial Profile
EPA developed a baseline financial profile of all CWSs based on CWSS data. Exhibit 4-9 shows
revenues, expenses, and net revenues by size and ownership. Revenues and expenses are reported
in thousands of dollars and are based on 1995 data. The smallest systems, i.e., those serving 100
people or fewer, have an average revenue of less than $10,000 per year, while the largest systems,
i.e., those serving more than 100,000 people, have an average annual revenue of over $35 million.
Exhibit 4-9 presents the mean total expenses for all community water systems by service
population category. Annual expenses range from a mean of approximately $7,000 per year for
the smallest systems to over $28 million for the largest systems. In general, revenues exceed
expenses; however, for systems that serve fewer than 500 people, expenses frequently exceed
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA.
-------
revenues.
Exhibit 4-9
Baseline Revenues and Expenses for CWSs
Publicly-Owned
Revenues
Expenses
Net Revenues
Privately-Owned
Revenues
Expenses
Net Revenues
All Systems
Revenues
Expenses
Net Revenues
System Size Category
<100
$9,677
$14,747
($5,070)
$7,009
$7,366
($356)
$7,486
$8,683
($1,198)
101-500
$39,023
$38,781
$242
$36,206
$37,681
($1,476)
$37,644
$38,243
($599)
501-1,000
$103,981
$96,015
$7,966
$110,375
$101,612
$8,764
$105,999
$97,782
$8,218
1,001-
3,300
$238,163
$217,900
$20,263
$234,610
$226,307
$8,304
$237,326
$219,879
$17,447
3,301-
10,000
$714,414
$626,242
$88,172
$702,382
$621,296
$81,086
$711,788
$625,162
$86,626
10,001-
50,000
$2,701,280
$2,188,092
$513,188
$3,728,501
$2,973,931
$754,570
$2,829,324
$2,286,048
$543,277
50,001-
100,000
$8,933,319
$8,010,010
$923,309
$10,188,865
$8,188,989
$1,999,877
$9,075,183
$8,030,233
$1,044,950
100,001-
1,000,000
$35,439,984
$31,492,090
$3,947,895
$38,755,625
$33,333,252
$5,422,372
$35,750,697
$31,664,628
$4,086,070
Source: EPA, 1995. CWSS.
Expense data include operating costs, interest payments, and "other" expenses from systems reporting non-zero values.
Revenue data include total revenues from systems reporting zero values.
Unfortunately, there is no available information characterizing non-transient non-community
water system revenues. This is primarily due to the fact that these systems generally do not
operate water systems to generate revenue, but to support their primary business. It should also
be noted that there are potential limitations with using the revenue and expense data from the
CWSS since the data are based on only one year of information during which time all systems may
not have accounted for all water-related revenues.
4.3 Occurrences of Arsenic
EPA has relied on a variety of data sources to evaluate the occurrence of arsenic in community
water systems and non-transient non-community systems. This information supports EPA's
assessment of baseline conditions, including (1) the number of systems expected to exceed various
MCL options, and (2) the population exposed to different levels of arsenic.
In 1992, EPA conducted an analysis of the number of systems that would be impacted by various
arsenic MCL options, ranging 0.5 //|ig/L to > 50 //|ig/L. These projections were based on the
following national surveys:
1984-1986 National Inorganic and Radionuclide Survey (MRS) for ground water
systems;
1976-1977 National Organic Monitoring Survey for surface water systems;
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA.
-------
1978-1980 Rural Water Survey for surface water systems; and
1978 Community Water Supply Survey for surface water systems.
EPA estimated that approximately 150 CWSs and NTNCs with ground water sources and five
CWSs and NTNCs with surface water sources would exceed 50 //g/L, and that approximately
4,500 CWSs and NTNCs with ground water sources and 350 CWSs and NTNCs with surface
water sources would exceed 5 //g/L.
These data sources have several limitations. First, the surveys used for surface water systems
were conducted primarily before 1980. It is likely that arsenic occurrence has changed in the past
two decades due to changes in raw water sources or the addition of filtration treatment to comply
with the Surface Water Treatment Rule (SWTR). In addition, many of the survey responses had
relatively high minimum reporting limits (5 //g/L). Therefore, it is statistically difficult to
extrapolate low-level arsenic occurrence.
EPA has subsequently received new data from EPA offices, States, public water utilities, and
associations supporting a new evaluation of baseline occurrences. These data, based on recent
samples, benefit from improved analytical techniques with lower detection limits and lower
reporting limits. The new data include the following:
Arsenic MCL compliance monitoring data for ground and surface water
community water systems in 25 states;
A 1992-1993 national survey of 140 large ground and surface water systems
(greater than 10,000 people), which was performed by the Metropolitan Water
District of Southern California (MWDSC);
A 1993 survey examining low levels of arsenic occurrence in surface water and
ground water in the State of California (Association of California Water Agencies;
ACWA); and
A 1996 national survey of approximately 500 ground and surface water systems
serving more than 1,000 people performed by American Water Works Association
(AWWA).
In addition, the US Geological Survey (USGS) worked extensively with EPA to share arsenic
ambient ground water occurrence data through an Interagency Agreement.
EPA (1999c) used the MCL compliance monitoring data from 25 states to develop an improved
estimate of national baseline arsenic occurrence. The estimates based on this data are comparable
to those based on the other sources listed above.
Chapter 4, Baseline Analysis 4-12 Proposed Arsenic in Drinking Water Rule RIA.
-------
Exhibit 4-10
Arsenic Occurrence in CWSs at Various Concentration Levels (ug/L)
Source
GW
SW
% of systems
2
27.20
9.93
3
19.90
6.01
5
12.10
2.90
10
5.43
0.75
15
3.13
0.40
greater
than
20
2.
0.
06
26
(H9/L)
25
1.45
0.19
30
1.08
0.14
40
0.66
0.09
50
0.44
0.07
Source: EPA, 1999c. Arsenic Occurrence in Public Drinking Water Supplies.
EPA used statistical techniques to assess
(1) the national distribution of mean arsenic concentrations in water systems,
(2) the distribution of source means within systems, and
(3) the number of systems with at least one source above various MCLs.
Exhibit 4-10 shows the percentage of systems with an arsenic occurrence in excess often different
concentration levels, ranging from 2 |ig/L to 50 |ig/L. Less than one percent of ground water and
surface water systems have a concentration level of arsenic greater than 50 jig/L. In contrast, 27
percent of ground water systems and 10 percent of surface water systems have an arsenic
concentration greater than 2 |ig/L.
Exhibit 4-11 provides a summary of the number of systems expected to exceed various MCLs,
based on the results of EPA's revised occurrence estimates.
Exhibit 4-11
Number of CWSs Exceeding Various Arsenic MCL Concentrations (ug/L)
MCLS
GW
SW
MCLS
GW
SW
MCL 10
GW
SW
MCL 20
GW
SW
<100
3,024
62
1,898
32
874
10
343
3
101-500
3,256
109
2,048
57
934
18
377
5
501-1,000
1,058
67
671
34
312
11
126
3
1,001-
3,300
1,406
137
893
70
424
23
177
5
3,301-
10,000
696
94
444
50
218
16
91
5
10,001-
50,000
434
76
286
41
144
13
61
3
50,001-
100,000
53
16
35
8
19
3
8
1
100,001-
1,000,000
33
16
21
8
11
4
5
1
Chapter 4, Baseline Analysis
Proposed Arsenic in Drinking Water Rule RIA.
-------
Chapter 4, Baseline Analysis 4-14 Proposed Arsenic in Drinking Water Rule RIA.
-------
Chapter 5: Benefits Analysis
5.1 Nature of Regulatory Benefits
The benefits associated with reductions of arsenic in drinking water arise from a reduction in
adverse human health effects. To a lesser degree benefits may also accrue from positive
ecological effects and an avoidance of costly consumers behaviors aimed at avoiding exposure,
such as the cost of bottled water.
The value to consumers of a reduction in the risk of adverse health effects includes the following
components:
The avoidance of medical costs and productivity losses associated with illness;
The avoidance of the pain and suffering associated with illness;
The losses associated with risk and uncertainty, also called the "risk premium;" and
The reduction in risk of premature mortality.
This conceptual valuation framework goes beyond valuing out-of-pocket medical costs and lost
time to include the value consumers place on avoiding pain and suffering and the risk premium.
The risk premium represents the damages associated with risk and uncertainty, captured in the
expression of consumer's "willingness-to-pay" for the reduction in risk of illness (Freeman, 1979).
This chapter first presents information on the multiple adverse health effects associated with
arsenic, followed by a quantitative risk analysis of a single arsenic related endpoint, bladder
cancer. Due to the large number of potential health effects can't be quantified, it is likely that the
estimated benefits associated with avoidance of bladder cancer considerably underestimate the
total benefits of a reduction of arsenic in drinking water.
5.2 Health Effects
5.2.1 Overview
Exposure to arsenic has many potential health effects which have been described in two recent
publications: Arsenic in Drinking Water by the National Research Council (NRC, 1999), and the
Agency for Toxic Substances and Disease Registry's Draft Toxicological Profile for Arsenic
(ATSDR, 1998). These two sources provide descriptions of health effects which are summarized
in this section, along with additional information provided from the recent literature.
Ingestion of inorganic arsenic can result in both cancer and non-cancer health effects (NRC,
1999). Exposure may also occur via other routes of exposure including inhalation and dermal
exposure. There is a large human effects database available for inorganic arsenic. However, the
effects of organic forms of arsenic are not as well characterized as those for inorganic arsenic.
The proposed rule addresses both organic and inorganic forms of arsenic.
The nature of the health effects avoided by reducing arsenic levels in drinking water is a function
Chapter 5, Benefits Analysis 5-1 Proposed Arsenic in Drinking Water Rule RIA
-------
of characteristics unique to each individual and the level and timing of exposure. Therefore, the
relationship between exposure and response is quite complex. This section describes potential
health effects, but does not reach conclusions about specific effects that might occur due to the
current levels of arsenic in our country's drinking water.
5.2.2 Carcinogenic Effects
Arsenic's carcinogenic role was noted over 100 years ago (NCI, 1999) and has been studied since
that time. The Agency has classified arsenic as a Class A human carcinogen, "based on sufficient
evidence from human data. An increased lung cancer mortality was observed in multiple human
populations exposed primarily through inhalation. Also, increased mortality from multiple internal
organ cancers (liver, kidney, lung, and bladder) and an increased incidence of skin cancer were
observed in populations consuming drinking water high in inorganic arsenic." (EPA, IRIS web
site extracted 8/99).
The International Agency for Research on Cancer (IARC) concluded that inhalation of inorganic
arsenic caused skin and lung cancer in humans. The 1999 NRC report on arsenic states that
"epidemiological studies ... clearly show associations of arsenic with several internal cancers at
exposure concentrations of several hundred micrograms per liter of drinking water" (NRC, 1999).
Ten epidemiological studies, covering eight organ systems, present quantitative data useful for
risk assessment (NRC, 1999, Table 4-1). The organ systems where cancers in humans have been
identified include skin, bladder, lung, kidney, nasal, liver, and prostate.
Table 10-6 of the NRC report provides risk parameters for three cancers: bladder, lung, and liver
cancer. Considering all cancers in aggregate, the NRC states in their Risk Characterization
section that "considering the data on bladder and lung cancer in both sexes noted in the studies in
chapter 4, a similar approach for all cancers could easily result in a combined cancer risk on the
order of 1 in 100" (at the current MCL of 50 ng/L).
New data provide additional health effects information on both carcinogenic and noncarcinogenic
effects of arsenic. A recently study by Tsai et al. (1999) of a population that has been studied
over many years in Taiwan has provided statistically significant standardized mortality ratios
(SMRs) for 23 cancerous and non-cancerous causes of death in women and 27 causes of death in
men. SMRs are an expression of the ratio between deaths that were observed in an area with
elevated arsenic levels and those that were expected to occur, based on the mortality experience
of the populations in nearby areas without elevated arsenic levels. Drinking water (250-1,140
|ig/L) and soil (5.3-11.2 mg/kg) in the Tsai et al. (1999) population study had very high arsenic
content.
Tsai et al. (1999) identified "bronchitis, liver cirrhosis, nephropathy, intestinal cancer, rectal
cancer, laryngeal cancer, and cerebrovascular disease" as possibly "related to chronic arsenic
exposure via drinking water." In addition, the study area had upper respiratory tract cancers
previously only related to occupational inhalation. High male mortality rate (SMR > 3) existed
for bladder, kidney, skin, lung, and nasal cavity cancers and for vascular disease. However, the
authors noted that the mortality range was marginal for leukemia, cerebrovascular disease, liver
cirrhosis, nephropathy (kidney), and diabetes. Females also had high mortalities for laryngeal
Chapter 5, Benefits Analysis 5-2 Proposed Arsenic in Drinking Water Rule RIA
-------
cancer. The SMRs calculated by Tsai et al. (1999) used the one cause of death noted on the
death certificates. Many chronic diseases, including some cancers, do not result in mortality.
Consequently, the impact indicated by the SMR will underestimate the total impact of these
diseases.
There are, of course, possible differences between the population and health care in Taiwan and
the United States. For example, arsenic levels in the U.S. are not nearly as high as they were in
the study area of Taiwan. However, the study gives an indication of the types of health effects
that may be associated with arsenic exposure via drinking water.
5.2.3 Noncarcinogenic Effects
Arsenic interferes with a number of essential physiological activities, including the actions of
enzymes, essential cations, and transcriptional events in cells (NRC, 1999). A wide variety of
adverse health effects have been associated with chronic ingestion of arsenic in drinking water,
occurring at various exposure levels.
Effects on specific organ systems reported in humans exposed to arsenic are listed below in
Exhibit 5-1 (NRC, 1999). Exhibit 5-1 provides descriptive information on the specific diseases
and/or symptoms associated with categories of diseases.
Chapter 5, Benefits Analysis 5-3 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 5-1
Adverse Noncarcinogenic Health Effects Reported in Humans in NRC (1999) as
Potentially Associated with Arsenic, by Organ System Affected
cutaneous effects
hyperpigmentation
hyperkeratoses
melanosis
gastrointestinal and hepatic
effects
noncirrhotic portal hypertension
gastrointestinal hemorrhage secondary to esophageal varices
hepatic enlargement
splenic enlargement
periportal fibrosis of the liver
obliterative intimal hypertrophy of intrahepatic venules resulting in obstruction of portal
venous flow, increased splenic pressures, and hypersplenism, and cirrhosis of the liver
diarrhea
cramping
cardiovascular and peripheral
vascular effects
peripheral vascular disease (blackfoot disease)
gangrene of the feet
coldness and numbness in the extremeties
intermittent claudication
ulce ration
spontaneous amputation
Raynaud's syndrome
acrocyanosis
ischemic heart disease
cardiovascular and peripheral
vascular effects (in children)
arterial spasms in fingers and toesm
esenteric artery thrombosis
cerebrovascular disease
extensive coronary occlusions
cerebrovascular occlusions
ischemia of the tongue
Raynaud's syndrome
gangrene in extremities
hematological effects
anemia - normocytic, megoblastic
leukopenia - neutropenia, lymphopenia, eosinophilia
thrombocytopenia
reticulocytosis
erythroid hyperplasia
pulmonary effects
chronic cough
restrictive and obstructive lung disease
emphysema
immunological effects
impaired immune response (more specific effects observed in human cell studies and
animal studies- see source)
neurological effects
peripheral neuropathy
endocrine effects
diabetes mellitus
reproductive and developmental
effects
spontaneous abortion
perinatal death
stillbirth
low birth weight
birth defects including coarctation of the aorta and others
reproductive and developmental
effects*
neural tube defects
ophthalmic abnormalities
numerous skeletal abnormalities
urogenital abnormalities
growth retardation
*Notes in parenthesis indicate where health effects were observed in animal studies rather than human studies.
NRC reports results of numerous animal reproductive and developmental studies and notes that there are "very
few" human studies.
Chapter 5, Benefits Analysis
5-4
Proposed Arsenic in Drinking Water Rule RIA
-------
5.2.4 Susceptible Subgroups
This section discusses the nature of special susceptibilities and identifies population subgroups
that may be at higher risk than the general population when exposed to arsenic.
5.2.4.1 Definition
A susceptible subgroup exhibits a response that is different or enhanced when compared to the
responses of most people exposed to the same level of arsenic (ATSDR, 1998). Many diseases
affect certain subgroups of the population disproportionately. The subgroups may be defined by
age, gender, race, ethnicity, socioeconomic status, pre-existing medical conditions, behavioral or
physiological differences, or other characteristics. For example, there are pre-existing medical
conditions that will increase susceptibility to most toxins, such as a pre-existing disease in the
toxin's target organ. Very few diseases affect all population groups (ages, sexes, races) equally.
For purposes of evaluating potential benefits to different segments of the population, it is useful to
evaluate whether there are susceptible subpopulations that require consideration. The benefit of
reducing their exposure may be considerably higher than the benefit associated with reducing
exposure among the general population (on a per capita basis).
Special susceptibilities may be indicated by known differences in biological processes that are
essential to detoxification of a toxin. In addition to identifying susceptible subgroups based on
biological processes, susceptible subgroups are often identified by observing higher-than-average
rates of the disease of interest. Increases in the rates of reported diseases may be due to a variety
of factors. Some of these indicate an increased susceptibility; others are matters of personal
choice and may not be considered relevant in a benefits analysis. One way to approach this issue is
to evaluate increased susceptibility when it is based on an increased risk of disease due to factors
reasonably beyond the control of the subpopulation. Factors that are usually beyond the control
of the individual that may cause increased susceptibility include:
Constitutional limitations (e.g., illnesses, genetic abnormalities, birth defects such as enzyme
deficiencies);
Concurrent synergistic exposures that cannot reasonably be controlled (e.g., at home or in
the workplace); and
Normal constitutional differences (i.e., differences based on sex, age, race, ethnicity, etc).
Other factors that are not usually considered beyond the individual's control include personal
choices, such as smoking, drinking, and drug use. Choice of place of residence or work may or
may not be treated as a relevant factor. Ultimately, which types of factors should be included in
identifying susceptible subgroups is a matter of public policy.
No studies were located by ATSDR (1998) that focused exclusively on evaluating unusual
susceptibility to arsenic. However, some members of the population are likely to be especially
susceptible due to a variety of factors. These factors include increased dose (intake per unit of
body weight) in children, genetic predispositions, and dietary insufficiency (ATSDR, 1998), as
Chapter 5, Benefits Analysis 5-5 Proposed Arsenic in Drinking Water Rule RIA
-------
well as pre-existing health conditions.
5.2.4.2 Children
One often-identified susceptible subgroup is children. Due to their increased fluid and food intake
in relation to their body weight (NAS, 1995), their dose (milligrams per kilogram of body weight
per day - mg/kg/day) of arsenic will be, on average, greater than that of adults. For example, an
intake of 1.2 liters per day in a 70 kg adult yields an overall water intake of 0.017 liters per kg of
body weight. An infant who consumes 1 liter per day and weighs 10 kg is consuming 0.1 liter per
kg of body weight, which is more than 5 times the water intake per kg of an adult. Any
contaminant which is present in the water will be delivered at a correspondingly higher level, on a
daily basis. Foy et al., noted that in studies of chronic exposure, children appear to be more
severely affected, probably due to a higher exposure per body weight (1992 citation, reported in
ATSDR, 1998).
The increased daily dose in children can be effectively considered for noncarcinogenic effects
because toxicity is evaluated in terms of exposures that can range from relatively short-term to
long-term exposure. However, carcinogenic effects (i.e., bladder cancer) are evaluated based on a
lifetime of exposure, which takes into consideration the elevated dose that occurs in children.
Because the only health effect measured in this benefits assessment is bladder cancer, a sensitivity
analysis to consider higher doses of arsenic during childhood was not necessary. However, the
numerous noncarcinogenic effects listed in Exhibit 5-1 may be of greater concern for children than
adults. As the table indicates, many severe cardiovascular effects have been observed in children.
Avoidance of these effects constitutes an unqualified benefit of the rule.
The adverse reproductive effects listed in Exhibit 5-1 give evidence of the susceptibility of a child.
The toxic mechanism of action of arsenic involves the inhibition of proliferation of cells, as well as
impairment of the embryonal cell division and mitosis (cell reproduction) (Dong and Luo, 1993,
Jha et al., 1992, Petres et al., 1977, Leonard and Lauwerys, 1980, Li and Chou, 1992, Mottet and
Ferm, 1983 reported in ATSDR, 1998). Arsenic also causes chromosomal aberrations (Jha et al.,
1992, Leonard and Lauwerys, 1980, reported in ATSDR, 1998). These all have serious adverse
implications for critical developmental processes.
In addition, arsenic crosses the placenta and preferentially accumulates in the embryonic
neuroepithelium, as well as occurring in breast milk (Somogyi and Beck, 1994, reported in
ATSDR, 1998). The neuroepithelium is particularly susceptible during development because the
process of neurulation is being carried out, which involves the development of the neural tube and
other essential structures (Dallaire and Beliveau, 1992, Edelman, 1992, Gunn et al., 1992, Li and
Chou, 1992, Morris-Kay et al., 1994, Scheonwolf and Smith, 1990, Taubeneck et al., 1994,
reported in ATSDR, 1998).
5.2.4.3 Genetic Predispositions and Dietary Insufficiency
Methylation of arsenic in the liver is a pathway for the detoxification of inorganic arsenic.
Individuals who are deficient in essential enzymes for this process, or who have a dietary
deficiency of methyl donors (choline or methionine), will be at greater risk following inorganic
Chapter 5, Benefits Analysis 5-6 Proposed Arsenic in Drinking Water Rule RIA
-------
arsenic exposure (Buchet and Lauwerys, 1987; Vahter and Marafante, 1987; Brouwer et al., 1992
cited in ATSDR, 1998). However, liver disease does not appear to increase risk at low levels of
arsenic exposure (Buchet et al., 1982; Geubel et al., 1988 cited in ATSDR, 1998). Therefore,
these factors are not expected to increase risk levels for a significant portion of the U.S.
population.
5.2.4.4 Individuals with Pre-existing Organ Susceptibilities
Individuals may have increased susceptibilities based on specific organ-related factors. Those
with pre-existing diseases (e.g., kidney disease), as well as those with congenital defects (a single
kidney) will be at greater risk from a toxin that either causes additional damage to that organ, or
that relies on that organ for detoxification. In the case of arsenic, both the kidneys and liver are
used to detoxify and remove the contaminant. Both single high doses and long-term low doses
may cause an accumulation of arsenic in the liver and kidneys, which can impair function. In
addition, these organs may be directly damaged by arsenic exposure. A review of Tables 5-1 and
5-2 indicates other organ systems that are targets of arsenic toxicity, including the cardiovascular
system (heart, veins, arteries), hematopoietic system, endocrine system, cutaneous system,
pulmonary system, gastrointestinal system, immune system, and peripheral nervous system. In
individuals with pre-existing damage to these systems or congenital defects in the systems, the
likelihood of risk is greater. Due to the higher incidence of most types of disease among the
elderly, they are more likely to have pre-existing conditions in these organ systems.
5.2.4.5 Individuals Exposed via Non-water Sources
Although arsenic is ubiquitous at low levels, it is not generally found at levels of concern in food
or air, in the absence of elevated local sources. Where background levels are high, however,
(e.g., elevated levels in water) it is reasonable to consider the contribution to total exposure that
may occur from soil, food, and other local sources. When anthropogenic sources are known to
generate elevated arsenic levels in water (e.g., a local smelter), it is more likely that other media
may be contaminated as well. The total exposure from all sources is a critical component of
evaluating potential health risks and the benefits of avoiding contaminated drinking water in these
cases. A reduction in arsenic in drinking water will reduce the overall exposure to individuals in
living in contaminated areas (e.g., around certain Superfund sites) or workers exposed to arsenic
on the job. Total exposure from all sources is of particular concern for noncancer risks, because
background levels from non-drinking water sources will determine whether the total exposure
leads to an exceedence of a threshold for effects.
5.3 Quantitative Benefits of Avoiding Bladder Cancer
Quantitative risk metrics (e.g., slope factors or reference doses) are necessary to evaluate cancer
or non-cancer risks. Although arsenic causes numerous health effects, bladder cancer is the only
endpoint for which an Agency-approved metric for evaluating arsenic related risk currently exists.
This cancer slope factor (SF) for bladder cancer is adequate data to perform a risk assessment and
was used to calculate cases potentially avoided due to EPA's proposed drinking water standards.
Benefits estimates for avoided cases of bladder cancer were calculated using mean population risk
estimates at various MCL levels. Lifetime risk estimates were converted to annual risk factors,
Chapter 5, Benefits Analysis 5-7 Proposed Arsenic in Drinking Water Rule RIA
-------
and applied to the exposed population to determine the number of cases avoided. Due to a lack
of risk data, health benefit estimates for lung cancer were quantified based on the assumption that
the risks of a fatal lung cancer case associated with arsenic are 2-5 times that of a fatal bladder
cancer case.
5.3.1 Risk Assessment for Bladder Cancer Resulting from Arsenic Exposure
5.3.1.1 Risk Assessment Methodology
Risk assessment is based on the analysis of scientific data to determine the likelihood, nature, and
magnitude of harm to public health associated with particular agents, and involves three main
analytical components: hazard identification (dose-response assessment), exposure assessment,
and risk characterization. Exhibit 5-2 illustrates the steps in a traditional risk assessment process
for characterizing the potential human cancer associated with contaminants in drinking water.
Community Water Systems
The following sections summerize how risk reductions were calculated for populations exposed to
arsenic levels at or above 3 |ig/L. The approach for this analysis included five components, which
are described in more detail below. First, EPA used data from the recent EPA water consumption
study. Second, Monte-Carlo simulations were used to develop relative exposure factors. Third,
arsenic occurrence estimates identified the population exposed to levels above 3 jig/L. Fourth,
risk distributions were chosen for the analysis from the 1999 NRC report. Finally, EPA
developed estimates of the actual risks faced by exposed populations using Monte-Carlo
simulations, using the relative exposure factors, occurrence, and risk distributions mentioned
above. A more detailed description of the risk methodology is provided in Appendix B.
Water Consumption.
EPA recently updated its estimates of per capita daily average estimates of water consumption
(EPA, 1999). The estimates used data from the combined 1994, 1995, and 1996 Continuing
Survey of Food Intakes by Individuals (CSFII), conducted by the U.S. Department of Agriculture
(USD A). The CSFII is a complex, multistage area probability sample of the entire U.S. and is
conducted to survey the food and beverage intake of the U.S. Estimates of water consumed
include direct water, indirect water and total water. "Direct" water is tap water consumed
directly as a beverage. "Indirect" water is defined as water added to foods and beverages during
final preparation at home or by food service establishments such as school cafeterias and
restaurants. For the purpose of the report, indirect water did not include "intrinsic" water which
consists of water found naturally in foods (biological water) and water added by commercial food
and beverage manufactures (commercial water). "Total" water refers to combined direct and
indirect water consumption.
Chapter 5, Benefits Analysis 5-8 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 5-2
Components of the Bladder Cancer Risk Assessment
HAZARD
IDENTIFICATION
Toxicity
(dose-response)
EXPOSURE
ASSESSMENT
Exposure
t
Population Size and Distribution
t
Ingestion/Dose Human Intake Factors
t
Concentration of Contaminant in
Finished Drinking Water Supply and
Available for Human Consumption
t
Concentration of Contaminant in Source
Water
RISK
CHARACTERIZATION
Health Effects
Exhibit 5-3
Source of Water Consumed
Source
Community Tap
Other Tap Sources
Total
Direct Tap Water
(drinking)
X
X
X
Indirect Tap Water
(from food and beverages)
X
X
X
Bottled water
X
Per capita water consumption estimates are reported by source. Sources include community tap
water, bottled water, and water from other sources, including water from household wells and
rain cisterns, and household and public springs. For each source, the mean and percentiles of the
distribution of average daily per capita consumption are reported. The estimates are based on an
average of 2 days of reported consumption by survey respondents.
The estimated mean daily average per capita consumption of community tap water by individuals
Chapter 5, Benefits Analysis
5-9
Proposed Arsenic in Drinking Water Rule RIA
-------
in the U.S. population is 1 liter/person/day. For total water, which includes bottled water, the
estimated mean daily average per capita consumption is 1.2 liters per/person/day. These estimates
of water consumption are based on a sample of 15,303 individuals in the 50 States and the District
of Columbia. The sample was selected to represent the entire population of the U.S. based on
1990 census data.
The estimated 90th percentile of the empirical distribution of daily average per capita
consumption of community tap water for the U.S. population is 2.1 liters/person/day; the
corresponding number for daily average per capita consumption of total water is 2.3
liters/person/day. In other words, current consumption data indicate that 90 percent of the U.S.
population consumes approximately 2 liters/person/day, or less.
Water consumption estimates for selected subpopulations in the U.S. are described in the CSFII,
including per capita water consumption by source for gender, region, age categories, economic
status, race, and residential status and separately for pregnant women, lactating women, and
women in childbearing years. The water consumption estimates by age were used in the
computation of the relative exposure factors discussed below.
Monte-Carlo analysis.
Monte-Carlo analysis is a technique for analyzing problems where there are a large number of
combinations of input values which makes it impossible to calculate every possible result. A
random number generator is used to select input values from pre-defined distributions. For each
set of random numbers a single scenario's result is calculated. As the simulation runs, the model
is recalculated for each new scenario that continues until a stopping criteria is reached. For the
risk distributions calculated in this report, the simulations were carried out 2,000 times. For each
simulation, a relative exposure factor, occurrence estimate, and individual risk estimate were
calculated. These calculations resulted in estimates of the actual risks faced by populations
exposed to arsenic concentrations in their drinking water. The underlying risk distribution are
described below.
NRC risk distributions.
In its 1999 report, "Arsenic in Drinking Water," the NRC analyzed bladder cancer risks using data
from Taiwan. In addition, NRC examined evidence from human epidemiological studies in Chile
and Argentina, and concluded that risks of bladder and lung cancer had comparable risks to those
"in Taiwan at comparable levels of exposure (NRC, 1999, page 7)." The NRC also examined the
implications of applying different statistical analyses to the newly available Taiwanese data for the
purpose of characterizing bladder cancer risk. For Taiwanese male bladder cancer, using a
Poisson regression model and no data on unexposed populations, yielded a risk at the current
MCL of 1 to 1.347 per 1,000 (NRC, 1999, Table 10-11). In Table 10-12 of the report, excess
lifetime risk estimates for bladder cancer in males, calculated using EPA's 1996 cancer guidelines,
is presented (EPA 1996). EPA selected two of these distributions to be representative of the risks
and uncertainty involved (selecting relatively high and relatively low estimates). These
Chapter 5, Benefits Analysis 5-10 Proposed Arsenic in Drinking Water Rule RIA
-------
distributions are shown in Exhibit 5-4. l
Exhibit 5-4
EPA Assumed Life-Time Bladder Risk Estimates for Bladder Cancer Among Males
(Risk per 1,000 Men)
Mean
0.731
1.237
95% Upper Confidence Limit
0.807
1.548
Relative Exposure Factors.
A Monte-Carlo analysis generated male and female relative exposure factors (REFs) for each of
the broad age categories used in the water consumption study. Lifetime male and female REFs
were then calculated, where the life-long REFs indicate the sensitivity of exposure to an individual
relative to the sensitivity of exposure of an "average" person weighing 70 kilograms and
consuming 2 liters of water per day. These life-long REFs can be directly multiplied by the
average drinking water consumption to provide estimates of individual lifetime consumption
practices. In this analysis, EPA combined the water consumption data with data on population
weight from the U.S. Census. Distributions for both community tap water and total water
consumption were used because the community tap water estimates may underestimate actual tap
water consumption. The weight data included a mean and a distribution of weight for male and
females on a year-to-year basis. The means and standard deviations of the life-long REFs derived
from this analysis are shown in Exhibit 5-5.
Exhibit 5-5
Life-Long Relative Exposure Factors
Male
Female
Community Water Consumption Data
Mean = 0.60
s.d. = 0.61
Mean =0.64
s.d. = 0.6
Total Water Consumption Data
Mean =0.73
s.d. = 0.62
Mean =0.79
s.d. = 0.61
Non-Transient Non-Community Water Systems
Determination of system and individual exposure factors.
In the past, the Agency has directly used SDWIS population estimates for assessing the risks
posed to users of NTNC water systems. In other words, it was assumed that the same person
received the exposure on a year round basis. Under this approach it was generally assumed that all
1 All of these risk distributions are linear in the mean, and thus may be conservative assumptions, as the
NRC report suggested the true relationship may be sublinear. If the true relationship is sublinear, i.e., lower than
the straight line from 50 ug/L to zero, the true risks at levels below 50 ug/L are being overestimated.
Chapter 5, Benefits Analysis
5-11
Proposed Arsenic in Drinking Water Rule RIA
-------
NTNC users were exposed for 270 days out of the year and obtained fifty percent of their daily
consumption from these systems. As a comparison, TNC users are assumed to use the system for
only ten days per year.
With the recent completion of Geometries and Characteristics of Public Water Systems (EPA,
1999), however, the Agency has developed a more comprehensive understanding of NTNC water
systems. These systems provide water in due course as part of operating another line of business.
Many systems are classified as NTNC, rather than TNC water systems, solely because they
employ sufficient workers to trigger the "25 persons served for over six months out of the year"
requirement. Client utilization of these systems is actually much less and more similar to exposure
in TNC water systems. For instance, it is fairly implausible that highway rest areas along
interstate highways serve the same population on a consistent basis (with the exception of long
distance truckers). Nevertheless, there are highway rest areas in both NTNC and TNC system
inventories. The Geometries and Characteristics of Public Water Systems report suggests that
population figures reported in SDWIS which have been used for past risk assessments generally
appear to reflect the number of workers in the establishment coupled with peak day customer
utilization.
Under these conditions use of the SDWIS figures for population greatly overestimates the actual
individual exposure risk for most of the exposed population and also severely underestimates the
number of people exposed to NTNC water2. Adequately characterizing individual and population
risks necessitates some adjustment of the SDWIS population figures. For chronic contaminants,
such as arsenic, health data reflect the consequences of a lifetime of exposure. Consequently, risk
assessment requires the estimation of the portion of total lifetime drinking water consumption that
any one individual would receive from a particular type of water system. In turn, one needs to
estimate the appropriate portions for daily, days per year, and year per lifetime consumption.
These estimates need to be prepared for both the workers at the facility and the "customers" of
the facility.
This adjustment was accomplished through a comprehensive review of government and trade
association statistics on entity utilization by SIC code. These figures, coupled with SDWIS
information relating to the portion of a particular industry served by non-community water
systems, made possible the development of two estimates needed for the risk assessment:
customer cycles per year and worker per population served per day. These numbers are required
to distinguish the more frequent and longer duration exposure of workers from that of system
customers3. A more detailed characterization of the derivation of these numbers is contained in
2For example, airports constitute only about a hundred of the NTNC water systems. Washington's Reagan
National and Dulles, Dallas/Fort Worth, Seattle/Tacoma, and Pittsburgh airports are the five largest of the airports.
SDWIS reports that these five airports serve about 300,000 people. In actuality, Bureau of Transportation Statistics
suggest that they serve about eleven million passengers per year. Examination of this information and other BTS
statistics suggests that these airports serve closer to seven million unique individuals over the course of a year and
that exposure occurs on an average of ten times per year per individual customer, not 270 times.
3For example, travel industry statistics provide information on total numbers of hotel stays, vacancy rates,
traveler age ranges, and average duration of stay. These figures can be combined with the SDWIS peak day
Chapter 5, Benefits Analysis 5-12 Proposed Arsenic in Drinking Water Rule RIA
-------
the docket. Exhibit 5-6 provides the factors used in the NTNC risk assessment to account for the
intermittent nature of exposure.
Exhibit 5-6
Exposure Factors Used in the NTNC Risk Assessment
NTNCWS
Water
wholesalers
Nursing homes
Churches
Golf/country clubs
Food retailers
Non-food retailers
Restaurants
Hotels/motels
Prisons/jails
Service stations
Agricultural
products/services
Daycare centers
Schools
State parks
Medical facilities
Campgrounds/RV
Federal parks
Highway rest areas
Misc. recreation
service
Forest Service
Interstate carriers
Amusement parks
# cycles
peryr
1.00
1.00
1.00
4.50
2.00
4.50
2.00
86.00
1.33
7.00
7.00
1.00
1.00
26.00
16.40
22.50
26.00
50.70
26.00
26.00
93.00
90.00
worker/
pop/day
0.000
0.230
0.010
0.110
0.070
0.090
0.070
0.270
0.100
0.060
0.125
0.145
0.073
0.016
0.022
0.041
0.016
0.010
0.016
0.016
0.304
0.180
worker
fraction
daily
-
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.50
1.00
0.50
0.50
worker
days/yr
-
250
250
250
250
250
250
250
250
250
250
250
200
250
250
180
250
250
250
250
250
250
worker
exposure
years
-
40
40
40
40
40
40
40
40
40
40
10
40
40
40
40
40
40
40
40
40
10
customer
fraction
daily
0.25
1.00
0.50
0.50
0.25
0.25
0.25
1.00
1.00
0.25
0.25
0.50
0.50
0.50
1.00
1.00
0.50
0.50
1.00
1.00
0.50
0.50
days of
use/yr
270
365
52
52
185
52
185
3.4
270
52
52
250
200
14
6.7
5
14
7.2
14
14
2
1
customer
exposure
years
70
10
70
70
70
70
70
40
3
54
50
5
12
70
10.3
50
70
70
70
50
70
70
Exhibit 5-6
Exposure Factors Used in the NTNC Risk Assessment (continued)
population estimates to allocate daily population among workers, customers and vacancies. The combination of
these factors provides an estimate of the number of independent customer cycles experienced in a year.
Chapter 5, Benefits Analysis
5-13
Proposed Arsenic in Drinking Water Rule RIA
-------
NTNCWS
Summer camps
Airports
Military bases
Non-water utilities
Office parks
Manufacturing: Food
Manufacturing:
Non-food
Landfills
Fire departments
Construction
Mining
Migrant labor camps
#
cycles
peryr
8.50
36.50
worker/
pop/day
0.100
0.308
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
worker
fraction
daily
1.00
0.50
0.50
0.50
0.50
0.50
0.50
1.00
1.00
1.00
1.00
1.00
worker
days/yr
180
250
250
250
250
250
250
250
250
250
250
250
worker
exposure
years
10
40
40
40
40
40
40
40
40
40
40
40
customer
fraction
daily
1.00
0.25
days of
use/yr
7
10
customer
exposure
years
10
70
Once the population adjustment factors were derived, it was possible to determine the actual
population served by NTNC water systems. Exhibit 5-7 provides a breakout of these figures by
type of establishment. Although not included in Exhibit 5-7, there are other equally
important characteristics to note about these systems. With notable exceptions (such as the
airports in Washington, DC and Seattle), the systems generally serve a fairly small population on
any given day. In fact, 99 percent of the systems serve less than 3,300 users on a daily basis. This
means that water production costs will be relatively high on a per gallon basis.
Exhibit 5-7
Composition of NTNCs
(Percentage of Total NTNCWS Population Served by Sector)
Schools
Manufacturing
Airports
Office Parks
9.7
2.7
26.1
0.6
Medical
Facilities
Restaurants
Non-food Retail
Hotels/Motels
8
0.9
1.6
9.2
Interstate Carriers
State Parks
Amusement Parks
Highway Rest Area
7.1
8.6
17.7
1.0
Campgrounds
Misc.
Recreation
Other
1.3
1.8
3.5
Risk calculation.
Calculations of individual risk were prepared for each industrial sector. Even within a given
sector, however, risk varies as a function of an individual's relative water consumption, body
weight, vulnerability to arsenic exposure, and the water arsenic concentration. Computationally,
risks were estimated by performing Monte-Carlo modeling. The approach used was similar to the
modeling technique applied in estimating the community water system risk estimation, but with
Chapter 5, Benefits Analysis
5-14
Proposed Arsenic in Drinking Water Rule RIA
-------
two notable exceptions. First, each realization in a given sector was multiplied by the portion of
lifetime exposure factor presented in Exhibit 5-4 to reflect the decreased consumption associated
with the NTNC system. Secondly, relative exposure factors were limited to age specific ratings
where appropriate4. For example, in the case of school children, water consumption rates and
weights for six to eighteen year-olds were used.
To illustrate the process, it was assumed that a child would attend only NTNCWS-served schools
for all twelve years, a somewhat improbable likelihood. Further, it was assumed that a child
would get half of their daily water consumption at school (for an average first grader this would
correspond to roughly nine ounces of water per school day). Finally, it was assumed that the
child would have perfect attendance and attend school for 200 days per year. Exhibit 5-8 below
provides a sample output for the upper bound individual risk distribution to school children
resulting from exposure to the range of untreated arsenic observed in community groundwater
systems5, as well as an estimate based on more moderate assumptions of four ounces per day and
150 days attendance for four years. Upper and lower bound risk distributions were prepared for
both workers and "customers" at all types of NTNC water systems and are contained in Appendix
B.
Exhibit 5-8
School Children Risk Associated with Current Arsenic Exposure in NTNCs
Exposure Risks
Mean Lifetime Risk
90th Percentile Lifetime Risk
Lifetime Bladder Cancers [out of 575,000 students]
Moderate Exposure
Scenario
0.0087 x10'4
0.019x10-4
0.5 x10'4
Upper Bound
Scenario
0.079 x10'4
0.17 x10'4
4.5 x10'4
The distribution of overall population risks was determined as part of the same simulation by
developing sector weightings to reflect the total portion of the NTNCWS population served by
each sector. Population weighted proportional sampling of the individual sectors provided an
overall distribution of risk among those exposed at NTNC systems.
4For example, water consumption among school children was weighted to reflect consumption between
ages 6 and 18, while factory worker consumption was weighted over ages 20 to 64.
'Community groundwater occurrence information was used since NTNC systems are almost exclusively
supplied by groundwater sources. Further, as there was no depth dependence of arsenic levels observed in the
community information, it is believed that the data are an adequate approximation.
Chapter 5, Benefits Analysis
5-15
Proposed Arsenic in Drinking Water Rule RIA
-------
5.3.1.2 Risk Assessment Results and Benefit Estimates
Community Water Systems
Estimated Risk Reductions.
Estimated risk reductions for bladder cancer at various MCL levels were developed using Monte-
Carlo simulations. These simulations combined the distributions of relative risk factors associated
with arsenic levels at or above 3|ig/L, and the distribution of general bladder cancer risk taken
from the National Research Council report. Since the relative risk and occurrence distributions
represent primarily population and occurrence variability, and the cancer risk distributions
represent primarily uncertainty about the true risk, the combined distributions contain both
variability and uncertainly. These combined distributions provide more accurate estimates of the
actual risks faced by the exposed population, including the portion of the population facing
various levels of risk.
Estimated risk levels for bladder cancer at various MCL levels are shown in Exhibit 5-9. Results
based on both the community water consumption data and the total water consumption data are
shown. Populations at or above 10"4 risk levels are also summed in Exhibit 5-9. Since there is
uncertainty about these numbers, it is assumed that the range 1 - 1.5 x 10"4 represents a risk level
of essentially 10"4. It is then assumed that risks above 1.5 x 10"4 represent risks greater than 10"4.
The after treatment occurrence distributions were assumed to reflect treatment to 80 percent of
the MCL level. The latter assumption is made since water systems tend to treat below the MCL
level in order to provide a margin of safety.
Exhibit 5-10 provides an estimate about percentages of the exposed populations and the number
of people exposed at 10"4 risk levels and above, and, using the stated definition for an over 10"4
risk level, above 10"4. The numbers in this table are based on community water consumption data
and show that at an MCL of 3 |ig/L, only a small number (not quantified) face a level of risk
greater than 10"4. At an MCL of 5 |ig/L, about 0.3 to 0.8 million face such risk levels. At an
MCL of 10 |ig/L, 0.8 to 4 million are at risk, and at an MCL of 20 |ig/L, about 2.4 to 6.4 million
would be exposed at such levels. Exhibit 5-11 gives similar information using total water
consumption data.
Chapter 5, Benefits Analysis 5-16 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 5-9
Mean Bladder Cancer Risks for U.S. Populations
Exposed At or Above MCL Options, after Treatment1
MCL
(ug/L)
3
5
10
20
Mean Exposed
Population Risk
(Community Water
Consumption data)
2.1 -3.6 x10'5
3.6-6.1 x10'5
5.5-9.2x10-5
6.9- 11.6x10-5
Mean Exposed
Population Risk
(Total Water
Consumption data)
2.6-4.5x10-5
4.4-7.5x10-5
6.7-11.4x10-5
8.4- 13.9x10'5
Mean Exposed
Population Risk
(composite of available
consumption data)
2.1 -4.5x10-5
3.6-7.5x10-5
5.5-11.4x10-5
6.9- 13.9x10'5
1The bladder cancer risks presented in this table provide our "best" estimates at this time. Actual risks could be
lower, given the various uncertainties discussed, or higher, as these estimates assume that the probability of illness
from arsenic exposure in the U.S. is equal to the probability of death from arsenic exposure among the Taiwanese
study group.
Exhibit 5-10
Exposed Population at 10"4 Risk or Higher for Bladder Cancer After Treatment1
Community Water Consumption data)
MCL
(ug/L)
3
5
10
20
% at 10" Risk
or higher
<1 - 2.6%
1.5-12%
1 1 - 34%
19.5-41%
Population at 10"* risk or
higher (millions)
< 0.3 -0.7
0.4-3.2
2.9-9.1
5.2-11
% over 10"*
< 1%
< 1 - 3%
3- 15%
9 - 24%
Population over
10" (millions)
t
< 0.3 -0.8
0.8-4
2.4-6.4
1The bladder cancer risks presented in this table provide our "best" estimates at this time. Actual risks could be
lower, given the various uncertainties discussed, or higher, as these estimates assume that the probability of illness
from arsenic exposure in the U.S. is equal to the probability of death from arsenic exposure among the Taiwanese
study group.
*where over 10~4 means 1.5 x 10~4 or above; *too low to calculate
Exhibit 5-11
Exposed Population at 10"4 Risk or Higher for Bladder Cancer After Treatment1
(Total Water Consumption data)
MCL
(ug/L)
3
5
10
20
% at 10" Risk
or higher
< 1 - 3%
3- 18%
16-50%
26 - 53%
Population at 10" risk or
higher (millions)
< 0.3 -0.8
0.8-4.8
4.3-13.4
7- 14.2
% over 10"*
<1%
< 1 - 4%
4 - 23%
13-33%
Population over
10" (millions)
t
0.3- 1.1
1.1 -6.2
3.5-8.9
1The bladder cancer risks presented in this table provide our "best" estimates at this time. Actual risks could be
lower, given the various uncertainties discussed, or higher, as these estimates assume that the probability of illness
from arsenic exposure in the U.S. is equal to the probability of death from arsenic exposure among the Taiwanese
study group.
*where over 10~4 means 1.5 x 10~4 or above; *too low to calculate
Chapter 5, Benefits Analysis
5-17
Proposed Arsenic in Drinking Water Rule RIA
-------
One adjustment was necessary in order to use the risk distributions contained in the NRC report
to calculate arsenic induced bladder cancer cases at each MCL. As mentioned above, the NRC
risk distributions were based on the Taiwanese studies. In these studies, information on arsenic
related bladder cancer deaths was reported. In order to use these data to determine the
probability of contracting bladder cancer as a result of exposure to arsenic, we must assume a
probability of mortality given the onset of arsenic induced bladder cancer among the Taiwanese
study population. We have decided to bracket the uncertainty about this parameter. For the
lower-end estimate of bladder cancer cases, we assume that this conditional probability is 100
percent. In other words, we are assuming that everyone in the Taiwanese study group that
contracted bladder cancer died of it. Therefore, for the lower-end estimate of bladder cancer
cases, no adjustment to the number of cases estimated using the NRC risk distribution is needed.
For the upper-end estimate of bladder cancer cases, we assume that this conditional probability is
80 percent. In other words, we are assuming that for every arsenic induced bladder cancer death
recorded among the Taiwanese study population, 1.25 people actually had arsenic induced
bladder cancer. Therefore, we multiplied the number of bladder cases derived using the upper-end
NRC risk distribution by 1.25 to determine our estimate of the upper-end number of bladder
cancer cases at each MCL.
The number of bladder cancer cases avoided at each MCL are shown in Exhibit 5-12, and range
from 22 to 52 at an MCL of 3 ug/L, 16 to 45 at an MCL of 5 ug/L, 9 to 26 at an MCL of 10
ug/L, and 4 to 15 at an MCL of 20 ug/L.
Exhibit 5-12
Annual Bladder Cancer Cases Avoided from Reducing Arsenic in CWSs
Arsenic Level
(ug/L)
3
5
10
20
Reduced Mortality
Cases**
6-14
4- 12
2-7
1 -4
Reduced Morbidity
Cases**
16-39
12-33
7-19
3- 11
Total Bladder Cancer
Cases Avoided*
22-52
16-45
9-26
4- 15
* The lower-end estimate of bladder cancer cases avoided is calculated using the lower-end risk estimate (see
Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group was 100
percent. The upper-end estimate of bladder cancer cases avoided is calculated using the upper-end risk estimate
(see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group was
80 percent.
**Assuming 20-year mortality rate in the U.S. of 26 percent.
Using the "What if?" Scenario for Lung Cancer Benefits
The "What if?" scenario for lung cancer benefits was used to estimate benefits for avoided cases
of lung cancer. This scenario is based on the statement in the NRC report "Arsenic in Drinking
Water," which states that "some studies have shown that excess lung cancer deaths attributed to
arsenic are 2-5 fold greater than the excess bladder cancer deaths (NRC, 1999, pg. 8)." Two-to-
five fold greater would be 3.5 fold greater on average. Also in the U.S. the mortality rate from
bladder cancer is 26% and the mortality rate of lung cancer is 88%. This suggests that if the risk
Chapter 5, Benefits Analysis
5-18
Proposed Arsenic in Drinking Water Rule RIA
-------
of contracting lung cancer were identical to the risk of contracting bladder cancer, one would
expect 3.4 times the number of deaths from lung cancer as from bladder cancer. Since these
numbers are essentially the same, it seems reasonable to assume that the risk of contracting lung
cancer is essentially the same as the rate of contracting bladder cancer,6 in the context of this
"what-if' scenario. If the risk of contracting lung cancer from arsenic in drinking water is
approximately equal to the risk of contracting bladder cancer, then the combined risk estimates of
contracting either bladder or lung cancer would be approximately double the risk estimates of
bladder cancer alone.
The number of lung cancer cases avoided for reducing arsenic in CWSs is presented in Exhibit 5-
13, and range from 8 to 80 at an MCL of 3 ug/L, 6 to 68 at an MCL of 5 |ig/L, 3 to 40 at an
MCL of 10 |ig/L, and 2 to 23 at an MCL of 20 |ig/L.
Exhibit 5-13
Potential Annual Lung Cancer Cases Avoided from Reducing Arsenic in CWSs
Arsenic Level
(ug/L)
3
5
10
20
Reduced Mortality
Cases
7-70
5-60
3-35
1 -20
Reduced Morbidity
Cases
1 - 10
1 -8
0-5
0-3
Total Lung Cancer
Cases Avoided
8-80
6-68
3-40
2-23
Economic Measurements of the Value of Risk Reduction
The evaluation stage in the analysis of risk reductions involves estimating the value of reducing
the risks. The following sections describe the use of the benefits valuation techniques to estimate
the value of the risk reductions attributable to the regulatory options for arsenic in drinking water.
First, the approach for valuing the reductions in fatal risks is described, followed by a description
of the approach for valuing the reductions in nonfatal risks.
The benefits described in this RIA are assumed to begin to accrue on the effective date of the rule
and are based on a calculation referred to as the "value of a statistical life" (VSL). Of the many
VSL studies, the Agency recommends using estimates from 26 specific studies that have been
6If "X" is the probability of contracting bladder cancer, then 0.26X is the probability of mortality from
bladder cancer. If lung cancer deaths are 2 to 5 times as high as bladder cancer, then they are, on average, 3.5
times as high and the average probability of mortality from lung cancer would be 3.5 times 0.26X, or 0.9IX. Since
we also know that there is a 88% mortality rate from lung cancer, then if the probability of contracting lung cancer
is "Y," the probability of mortality from lung cancer can also be represented as 0.88Y. Setting the two ways of
deriving the probability of mortality from lung cancer equal, or 0.91X = 0.88Y, one can solve for Y
(Y= (0.91/0.88) X). Thus Y is approximately equal to X, and the rate of contracting lung cancer is approximately
the same as the rate of contracting bladder cancer.
Chapter 5, Benefits Analysis
5-19
Proposed Arsenic in Drinking Water Rule RIA
-------
peer reviewed and extensively reviewed within the Agency.7 These estimates, which are derived
from wage-risk and contingent valuation studies, range from $0.7 million to $16.3 million and
approximate a Wiebull distribution with a mean of $4.8 million (in 1990 dollars). Most of these
26 studies examine willingness to pay in the context of voluntary acceptance of higher risks of
immediate accidental death in the workplace in exchange for higher wages. This value is sensitive
to differences in population characteristics and perception of risks being valued.
EPA updated the VSL estimate from the The Benefits and Costs of the Clean Air Act, 1970 to
1990 report to a value of $5.8 million in 1997 dollars, according to internal guidance on economic
analyses (personal communication, John Bennett 5/15/00). In order to directly compare the
estimated national costs of compliance, the VSL used in this analysis was updated from the
January 1997 value to $6.06 million in May 1999 dollars, using the Consumer Price Index (CPI-
U) for all items.
Several factors may influence the estimate of economic benefits associated with avoided cancer
fatalities, including:
1. a possible "cancer premium" (i.e., the additional value or sum that people may be willing to
pay to avoid the experiences of dread, pain and suffering, and diminished quality of life
associated with cancer-related illness and ultimate fatality);
2. the willingness of people to pay more over time to avoid mortality risk as their income rises;
3. a possible premium for accepting involuntary risks as opposed to voluntary assumed risks;
4. the greater risk aversion of the general population compared to the workers in the wage-risk
valuation studies;
5. "altruism" or the willingness of people to pay more to reduce risk in other sectors of the
population; and
6. a consideration of health status and life years remaining at the time of premature mortality.
Use of certain of these factors may significantly increase the present value estimate. EPA
therefore believes that adjustments should be considered simultaneously. The Agency also
believes that there is currently neither a clear consensus among economists about how to
simultaneously analyze each of these adjustments nor is there adequate empirical data to support
definitive quantitative estimates for all potentially significant adjustment factors. As a result, the
primary estimates of economic benefits presented in the analysis of this proposed rule rely on the
unadjusted estimate.
To estimate the monetary value of reduced fatal risks (i.e., risks of premature death from cancer)
predicted under different regulatory options, value of a statistical life (VSL) estimates are
multiplied by the number of premature fatalities avoided. VSL does not refer to the value of an
identifiable life, but instead to the value of small reductions in mortality risks in a population. A
"statistical" life is thus the sum of small individual risk reductions across an entire exposed
population. For example, if 100,000 people would each experience a reduction of 1/100,000 in
7 U.S. Environmental Protection Agency, The Benefits and Costs of the Clean Air Act, 1970 to 1990,
October 1997, Appendix I; and U.S. Environmental Protection Agency, Guidelines for Preparing Economic
Analysis (Review Draft), June 1999, Chapter 7.
Chapter 5, Benefits Analysis 5-20 Proposed Arsenic in Drinking Water Rule RIA
-------
their risk of premature death as the result of a regulation, the regulation can be said to "save" one
statistical life (i.e., 100,000 x 1/100,000). If each member of the population of 100,000 were
willing to pay $20 for the stated risk reduction, the corresponding value of a statistical life would
be $2 million (i.e., $20 x 100,000). VSL estimates are appropriate only for valuing small changes
in risk; they are not values for saving a particular individual's life.
Estimates of the willingness to pay to avoid treatable, nonfatal cancers are the ideal economic
measures used to value reductions in nonfatal risks. Unfortunately, this information is not
available for bladder cancer. However, willingness to pay (WTP) data to avoid chronic bronchitis
is available, and has previously been employed by OGWDW (the microbial/disinfection by-
product (MDBP) rulemaking) as a surrogate to estimate the WTP to avoid non-fatal bladder
cancer. A WTP central tendency estimate of $607,162 (May 1999$) is used to monetize the
benefits of avoiding non-fatal cancers (this value was updated from the $536,000 value EPA
updated to 1997$ from the Viscusi et al. 1991 study).
To ground-truth the use of the chronic bronchitis WTP value as a proxy for bladder cancer WTP,
EPA has also developed cost of illness estimates for bladder cancer, as reported in Exhibit 5-14.
These estimates of direct medical costs are derived from a study conducted by Baker et al.
(1989), which uses data from a sample of Medicare records for 1974 - 1981. These data include
the total charges for inpatient hospital stays, skilled nursing facility stays, home health agency
charges, physician services, and other outpatient and medical services. EPA combined these data
with estimates of survival rates and treatment time periods to determine the average costs of
initial treatment and maintenance care for patients who do not die of the disease. This value of
$178,405 at a 3 percent discount rate, serves as a low-end estimate of the WTP to avoid bladder
cancer and does not include the value of avoided pain and suffering, lost productivity, or risk
premium.
Exhibit 5-14
Lifetime Avoided Medical Costs for Survivors
(preliminary estimates1)
Type of
Cancer
Bladder
Date Data
Collected
1974-1981
Number of Cases Studied
5% of 1974
Medicare patients
(sample from national statistics)
Estimated
Mortality Rate
26 percent
(after 20 years)
Mean Value per
Nonfatal Case
(Discount Rate)1
$178,405(3%)
$147,775(7%)
(for typical individual
diagnosed at age 70)
1. May 1999 dollars
Source: U.S. Environmental Protection Agency, Cosf of Illness Handbook (draft), September 1998.
Estimates of Cancer Health Benefits of Arsenic Reduction
Benefits estimates were calculated using mean population risk estimates at various MCL levels
(the composite values of the mean population risks presented in Exhibit 5-9). Lifetime risk
estimates were converted to annual risk factors, and applied to the exposed population to
determine the number of bladder cases avoided per year (Exhibit 5-12). Exhibit 5-13 describes
Chapter 5, Benefits Analysis
5-21
Proposed Arsenic in Drinking Water Rule RIA
-------
the number of lung cancer cases avoided per year. These cases were divided into fatal and non-
fatal cases avoided, based on survival information. The avoided premature fatalities were valued
based on the VSL estimates discussed earlier, as recommended by current EPA guidance for
cost/benefit analysis (EPA, 1997). The avoided non-fatal cases were valued based on the
willingness to pay estimates for the avoidance of chronic bronchitis.
The results of the benefits valuation are presented in Exhibit 5-15. Total annual health benefits
resulting from bladder cancer cases avoided range from $43.6 to $104.2 at an MCL of 3 ug/L,
$31.7 to $89.9 at an MCL of 5 ug/L, $17.9 to $52.1 at an MCL of 10 ug/L, and $7.9 to $29.8 at
an MCL of 20 ug/L. Potential annual health benefits from avoided cases of lung cancer, when
estimated based on the "what if scenario in which the risks of a fatal lung cancer case associated
with arsenic is assumed to be two to five times that of a fatal bladder cancer case, range from
$47.2 to $448.0 at an MCL of 3 ug/L, $35.0 to $384.0 at an MCL of 5 ug/L, $19.6 to $224.0 at
an MCL of 10 ug/L, and $8.8 to $128.0 at an MCL of 20 ug/L. In addition, other potential non-
quantifiable health benefits are summarized in Exhibit 5-15.
Exhibit 5-15
Estimated Monetized Total Cancer Health Benefits and
Non-Quantifiable Health Benefits from Reducing Arsenic in CWSs
Arsenic
Level
(ug/L)
3
5
10
20
Annual
Bladder Cancer
Health Benefits
(Smillions)12
$43.6 -$104.2
$31 .7 -$89.9
$17. 9 -$52.1
$7.9 - $29.8
"What-if ' Scenario and Potential Non-Quantifiable
Health Benefits
"What-if" Scenario
Annual
Lung Cancer
Health Benefits
(Smillions)13
$47.2 - $448.0
$35.0 - $384.0
$19.6 -$224.0
$8.8 -$128.0
Potential Non-Quantifiable
Health Benefits
Skin Cancer
Kidney Cancer
Cancer of the Nasal Passages
Liver Cancer
Prostate Cancer
Cardiovascular Effects
Pulmonary Effects
Immunological Effects
Neurological Effects
Endocrine Effects
Reproductive and Developmental
Effects
1. May 1999 dollars.
2. The lower-end estimate is calculated using the lower-end number of bladder cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was
100 percent. The upper-end estimate is calculated using the upper-end number of cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was 80
percent.
3. These estimates are based on the "what if" scenario for lung cancer, where the risks of a fatal lung cancer
case associated with arsenic are assumed to be 2-5 times that of a fatal bladder cancer case.
Chapter 5, Benefits Analysis
5-22
Proposed Arsenic in Drinking Water Rule RIA
-------
Non-Transient Non-Community Water Systems
Exhibit 5-16 presents a summary of the risk and benefit analyses for regulation of arsenic in
NTNC water systems. Exhibit 5-17 presents risk figures for three particular sets of individuals:
children in daycare centers and schools, and construction workers. Construction and other
strenuous activity workers comprise an extremely small portion of the population served by
NTNC systems (less than 0.1 percent), but face the highest relative risks of all NTNCWS users
(90th percentile risks of 0.7 to 1.6 x 10"4 lifetime risk).
Exhibit 5-16
Mean Bladder Cancer Risks, Exposed Population,
and Annual Cancer Benefits in NTNCs1
Arsenic
Level
(M9/L)
3
5
10
20
baseline
Mean Exposed Population
Risk (10^)
lower
bound
0.0046
0.0077
0.012
0.015
0.019
upper
bound
0.01
0.017
0.026
0.033
0.042
Total Bladder Cancer
Cases Avoided
per Year
lower
bound
0.132
0.104
0.064
0.039
-
upper
bound
0.294
0.229
0.147
0.088
-
Annual Benefits
($millions)2
lower
bound
$0.610
$0.481
$0.296
$0.180
-
upper
bound
$2.717
$2.116
$1.359
$0.814
-
1. Note that this table does not include lung cancer benefits.
2. May 1999 dollars
Exhibit 5-17
Sensitive Group Evaluation of Lifetime Risks
Group
Forest Service, Construction and Mining Workers
School Children
Day Care Children
Mean Risk
3.2-7x10-5
3.8-7.9x10-6
3.4-6.8x10-6
90th Percentile Risk
7.2-16x10-5
0.84- 1.7 x10'5
0.74- 1.5 x10'5
However, there is considerable uncertainty about these exposure numbers, as it is quite likely that
they overestimate consumption. The risks for children are much lower with an upper bound, 90th
percentile estimate of 1.7 x 10"5 lifetime risk. What is not possible to determine from the analysis
of NTNC systems is the extent to which there is overlap of individual exposure between the
various sectors. NTNC establishments generally constitute a small portion of their SIC sectors.
In conjunction with the observation that NTNC populations would only serve about eleven
percent of the total population if all sectors were mutually exclusive, it would seem reasonable to
treat the SIC groups independently. However, it is equally plausible that there are communities
where one individual might go from an NTNC day care center to a series of NTNC schools and
then work in an NTNC factory. Unfortunately, the Agency presently has no basis for
Chapter 5, Benefits Analysis
5-23
Proposed Arsenic in Drinking Water Rule RIA
-------
quantitatively estimating the extent to which this would occur.
The Agency is quite concerned about the potential for local issues to arise with respect to
combined arsenic exposures. In the rare community where all ground water is contaminated with
the highest levels of arsenic, risks could be outside of the Agency's traditionally allowable range.
5.4 Other Benefits of Reductions in Arsenic Exposure
Although health effects are the primary focus of this analysis, arsenic also has negative impacts on
the ecology and on public perceptions and acceptance of drinking water. These are briefly
discussed below.
5.4.1 Ecological Effects
Ecological effects are not quantified in this benefits assessment. However, they are anticipated to
occur at levels of severity that are in proportion to the levels of contamination that occur. Since
drinking water is ultimately reintroduced into the environment, arsenic contamination of drinking
water is of concern for its potential adverse impacts on the ecology. The avoidance of ecological
effects resulting from the proposed rule constitutes an non-quantified benefit of the rule.
Arsenic has numerous ecological effects on multiple biological systems, as indicated in the health
effects listed in Exhibit 5-1. While there are differences in the function and disease induction
between humans and animals, there are also striking similarities. The anatomy and physiology of
most animal systems share many common elements, including most of the basic organ systems and
biochemical processes. Effects observed in humans are generally assumed to be similar to those
observed in animals, with the exception of higher cognitive functions. This is the basis for the
extensive animal laboratory testing programs and requirements that EPA uses to evaluate the
toxicity of chemicals on humans.
It is likely that most if not all of the effects observed in humans (listed in Exhibit 5-1) will also be
observed in animals. As noted above, there are also numerous animal studies that have
demonstrated effects in various species. Arsenic, a heavy metal, bioaccumulates in many
biological materials and can move through the food chain through a variety of pathways. In
addition to damage caused directly on ecological systems, ecological effects may indirectly cause
adverse effect in humans, through ingestion of contaminated plants or animals.
5.4.2 Drinking Water Quality and Public Perception
It is well established that the public often avoids the use of tap water that is suspected to be
contaminated. In this context, contamination may suggest biological, chemical, or other water
quality issues. When public perception of water quality declines, consumers purchase bottled
water if they have the means to do so. In addition or as an alternative, they may avoid the use of
tap water, ingesting and cooking with other liquids, substituting pre-mixed baby formula, and
using other strategies to limit ingestion. Consumer avoidance of tap water sources usually results
in costs to the consumers, either in the cost of obtaining substitute fluids or potential health
impacts of reduced fluid intake. In addition, there are numerous cases where government
Chapter 5, Benefits Analysis 5-24 Proposed Arsenic in Drinking Water Rule RIA
-------
agencies have provided bottled water due to biological or chemical contamination. The levels of
contamination at which the government activities occur vary depending on a variety of factors.
The relationship between arsenic in tap water and changes in consumer behavior or government
interventions is a complex one. Factors that impact the choice to avoid tap water depend on
public information that is provided on levels of contamination, potential health effects, individual
aversions to risk taking, and other considerations. A quantitative evaluation of these responses
and the potential benefits of avoiding associated costs to the consumer or governments is not
included in this benefits assessment. However, it is clear that many consumers purchase bottled
water (a multimillion dollar industry) or invested in other methods of improving drinking water
quality, such as point-of-use (POU) devices, specifically to avoid ingestion of contaminants such
as arsenic. Thus, it is reasonable to conclude that a reduction in arsenic contamination will have
the long-term effect of restoring some level of consumer confidence in the water supply.
Chapter 5, Benefits Analysis 5-25 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 5, Benefits Analysis 5-26 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 6: Cost Analysis
6.1 Introduction
This chapter presents the national cost estimates for the proposed Arsenic Rule. The costs
associated with the proposed rule include: 1) costs borne by water systems to comply with the
new MCL standard and modified monitoring requirements, and 2) costs to the States to
implement and enforce the rule. Section 6.2 describes the inputs and methodologies used to
estimate costs, including the following:
a description of the technologies that may be used by systems to achieve the MCL
(Section 6.2.1),
the unit costs of different technologies for complying with the MCLs (Section 6.2.2),
system and State unit costs for monitoring and administration functions (Section 6.2.3),
the methods used to predict systems' compliance methods and the methods used to
calculate costs (Section 6.2.4).
Section 6.3 presents the results of the cost analysis, including the following:
a summary of national costs for the different regulatory options (Section 6.3.1),
costs by system size and type for the proposed option (Section 6.3.2),
and household costs (Section 6.3.3).
6.2 Methodology
6.2.1 Description of Available Technologies
In 1993, EPA developed a document entitled Treatment and Occurrence-Arsenic in Potable
Water Supplies (EPA, 1993) which summarized the results of pilot-scale studies examining
low-level arsenic removal, from 50 parts per billion (ppb or |ig/L) down to 1 ppb or less. EPA
convened a panel of outside experts in January 1994 to review this document and comment on the
ability of the technologies to achieve various MCLs. The Agency has since sought stakeholder
input on the use of various technologies for arsenic removal under different conditions, and has
incorporated that input into its estimates of technology performance and costs. The results are
documented in the Cost and Technology Document for the Arsenic Rule (US EPA, July 1999).
The technology cost functions and removal efficiencies presented in that document are used as
inputs for the cost analyses presented in this MA.
EPA reviewed fourteen treatment technologies. Of these, five are the most relevant for small
systems:
Ion exchange
Activated alumina
Reverse osmosis
Nanofiltration
Electrodialysis reversal
Chapter 6, Cost Analysis 6-1 Proposed Arsenic in Drinking Water Rule RIA
-------
Two technologies are used primarily in larger systems and are not expected to be installed solely
for arsenic removal:
Coagulation/filtration
Lime softening
Finally, seven additional alternative technologies were characterized as still emerging:
Iron-oxide coated sand
Granular ferric hydroxide
Iron filings
Sulfur-modified iron
Greensand filtration
Iron addition with microfiltration
Conventional iron/manganese removal
Some technologies generate wastes which require disposal, or require pre-treatment (e.g., pre-
oxidation or corrosion control)in order to be effective. These associated requirements were
identified for different technologies and system types, and their costs were included in the costs of
treatment where relevant.
In addition to these centralized treatment options, small systems may elect to use point-of-use
(POU) or point-of-entry (POE) devices to achieve compliance with the MCLs. POE involves
whole house treatment, whereas POU treats water at the tap. The available POE/POU
technologies for arsenic removal are essentially smaller versions of reverse osmosis, activated
alumina, and ion exchange. The technologies will have to be maintained by the water system,
involving some additional recordkeeping and maintenance costs.
Finally, some systems may elect to comply with the lower MCL by obtaining water from
alternative sources that meet the standard, or by interconnecting with another water system to
combine treatment or share the cost of treatment (referred to as "regionalization").
The result of the review of technologies that would effectively remove arsenic and bring a water
system into compliance is summarized in Exhibit 6-1. The list includes 25 treatment trains
available to systems, consisting of various combinations of compliance technologies, waste
disposal technologies, or pre-treatment technologies as required.
Chapter 6, Cost Analysis 6-2 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-1
Arsenic Rule Treatment Trains by Compliance Technologies Component, with Associated Removal Efficiencies
Treatment Technology
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Regionalization
Alternate Source
Modify Lime Softening
Modify Coagulation/Filtration
Anion Exchange (25 mg/L SO4)
Anion Exchange (150 mg/L SO4)
Anion Exchange (25 mg/L SO4)
Anion Exchange (150 mg/L SO4)
Activated Alumina (2,000 BV)
Activated Alumina (10,000 BV)
Reverse Osmosis
Reverse Osmosis
Reverse Osmosis
Coagulation Assisted Microfiltration
Coagulation Assisted Microfiltration
Oxidation Filtration (Greensand)
Anion Exchange (25 mg/L SO4)
Anion Exchange (150 mg/L SO4)
Activated Alumina (2,000 BV)
Activated Alumina (10,000 BV)
Anion Exchange (90 mg/L SO4)
Anion Exchange (90 mg/L SO4)
POE Activated Alumina
POU Reverse Osmosis
POU Activated Alumina
Waste Disposal Technology
POTW
/
/
/
s**
/
/
/
Evaporation
Pond
/
/
/
Non-
Hazardous
Landfill
/
/
£*
£*
/
/
/
/
/
/
/
/
Direct
Discharge
/
Chemical
Precipitation
/
/
/
Mechanical
De-
Watering
/
Non-
Mechanical
De-
Watering
/
Corrosion
Control
/
/
/
/
/
/
/
/
/
/
/
/
Pre-
Oxidation
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Removal
Efficiency
90%
90%
80%
95%
95%
95%
95%
95%
90%
90%
95%
95%
95%
90%
90%
50%
95%
95%
90%
90%
95%
95%
99%
99%
99%
non-hazardous landfill (for spent media)
* POTW for backw ash stream
-------
6.2.2 Unit Costs and Compliance Assumptions
Treatment
EPA estimated the costs of the various compliance technologies, including centralized treatment
technologies (with associated waste disposal and pre-treatment), POE/POU treatment, and
regionalization. Costs of each treatment train are estimated as functions of system size; design
flow is used to calculate capital costs and average flow is used to calculate operating and
maintenance (O&M) costs. Exhibit 6-2 presents a summary of compliance technology costs by
cost component for the treatment trains listed in Exhibit 6-1, annualized over 20 years at a seven
percent discount rate. Costs are in May 1999 dollars and are based on average and design flows
for median populations of each system size category, assuming one entry point per system. Note
that the capital and O&M cost components are listed separately for the treatment and waste
disposal components of the treatment train, and totaled to equal the annual cost, excluding
corrosion control or pre-oxidation costs. Detailed descriptions of the assumptions and
methodologies used to develop these cost estimates are available in the Cost and Technology
Document for the Arsenic Rule (US EPA, July 1999).
Chapter 6, Cost Analysis 6-4 Proposed Arsenic & Drinking Water Rule RIA
-------
Exhibit 6-2
Average Compliance Technology Costs (Treatment Train 1 through 8)
Size Category
<100
101-500
501-1000
1,001-3,300
Treatment Train No.
Treatment Capital Costs
Treatment O&M Costs
Waste Disposal Capital Costs
Waste Disposal O&M Costs
Annual Costs (7%)
Treatment Capital Costs
Treatment O&M Costs
Waste Disposal Capital Costs
Waste Disposal O&M Costs
Annual Costs (7%)
Treatment Capital Costs
Treatment O&M Costs
Waste Disposal Capital Costs
Waste Disposal O&M Costs
Annual Costs (7%)
Treatment Capital Costs
Treatment O&M Costs
Waste Disposal Capital Costs
Waste Disposal O&M Costs
Annual Costs (7%)
$
$
$
$
$
$
$
$
$
-------
Exhibit 6-2 continued
Average Compliance Technology Costs (Treatment Train 1 through 8)
Size Category
Treatment Train No.
1
3,300-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
Treatment Capital Costs $ 280,000
Treatment O&M Costs $
Waste Disposal Capital Costs $
Waste Disposal O&M Costs $
Annual Costs (7%) $ 26,430
Treatment Capital Costs $ 280,000
Treatment O&M Costs $
Waste Disposal Capital Costs $
Waste Disposal O&M Costs $
Annual Costs (7%) $ 26,430
Treatment Capital Costs $ 280,000
Treatment O&M Costs $
Waste Disposal Capital Costs $
Waste Disposal O&M Costs $
Annual Costs (7%) $ 26,430
Treatment Capital Costs $ 280,000
Treatment O&M Costs $
Waste Disposal Capital Costs $
Waste Disposal O&M Costs $
Annual Costs (7%) $ 26,430
$ 20,000 $
1,
225,058 $ 239,649
49,872 $ 113,459
9,334 $ 9,334
32,781 $ 32,781
104,778 $ 169,742
20,000 $1,237,501 $1,393,797
- $ 162,684 $ 736,014
- $ 10,898 $ 10,898
- $ 157,559 $ 157,559
1,888 $ 438,083 $1,026,166
20,000 $2,055,598 $1,853,803
- $ 368,722 $1,137,506
- $ 13,766 $ 13,766
- $ 413,912 $ 413,912
1,888 $ 977,967 $1,727,703
$ 2,495,535 $ 2,495,535
$ 8,922 $ 8,922
$ 5,085 $ 5,085
$ 8,312 $ 26,268
$ 253,275 $ 271,231
$ 9,160,742 $ 9,160,742
$ 33,090 $ 33,090
$ 5,085 $ 5,981
$ 36,512 $ 124,966
$ 934,791 $ 1,023,330
$ 19,339,675 $ 19,339,675
$ 70,123 $ 70,123
$ 5,735 $ 7,076
$ 94,447 $ 327,739
$ 1,990,640 $ 2,224,058
20,000 $6,936,516 $3,672,160 $139,484,072 $139,484,072
- $2,753,482 $2,931,378 $ 497,678 $ 497,678
- $ 48,804 $ 48,804 $ 9,572 $ 20,505
- $3,402,673 $3,402,673 $ 769,897 $ 2,691,813
1,888 $6,815,520 $6,685,285 $ 14,434,787 $ 16,357,736
$ 2,495,535 $ 2,495,535
$ 8,922 $ 8,922
$ 859,811 $ 3,312,119
$ 49,672 $ 127,510
$ 375,315 $ 684,633
$ 9,160,742 $ 9,160,742
$ 33,090 $ 33,090
$ 3,945,421 $ 15,570,260
$ 166,059 $ 448,436
$ 1,436,278 $ 2,815,958
$ 19,339,675 $ 19,339,675
$ 70,123 $ 70,123
$ 9,956,569 $ 33,925,583
$ 357,083 $ 993,137
$ 3,192,564 $ 6,091,123
$139,484,072 $ 139,484,072
$ 497,678 $ 497,678
$ 62,342,119 $140,291,110
$ 2,038,240 $ 5,985,719
$ 21,586,882 $ 32,892,195
NOTE:
1) Refer to Exhibit 6-1 fora description of treatment train technologies.
2) Average costs per size category are based on median population and associated flows, assuming one entry point per system.
-------
Exhibit 6-2 continued
Average Compliance Technology Costs (Treatment Train 9 through 16)
Size Category
<100
101-500
501-1000
1,001-3,300
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
9
49,780
2,816
-
21
7,536
210,637
9,693
-
137
29,712
485,414
23,047
-
360
69,227
1,302,755
67,590
-
1,099
191,659
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
10
49,780
2,816
-
21
7,536
210,637
9,693
-
137
29,712
485,414
23,047
-
360
69,227
1,302,755
67,590
-
1,099
191,659
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
11
156,629
25,439
3,520
1,368
41,924
433,067
51,709
3,457
11,724
104,639
710,071
92,892
3,552
31,207
191,460
1,623,013
222,259
4,571
74,606
450,497
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
12
156,629
25,439
3,520
1,368
41,924
433,067
51,709
3,457
11,724
104,639
710,071
92,892
3,552
31,207
191,460
1,623,013
222,259
4,571
74,606
450,497
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
13
156,629
25,439
40,587
11,267
55,322
433,067
51,709
70,251
17,439
116,658
710,071
92,892
109,999
27,080
197,381
1,623,013
222,259
152,058
40,523
430,336
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
14
65,210
13,108
23,260
3,701
25,160
145,362
17,159
26,273
4,941
38,301
304,855
28,132
31,420
7,330
67,205
726,146
37,152
51,820
21,160
131,746
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
15
65,210
13,108
27,122
2,088
23,911
145,362
17,159
37,421
2,299
36,711
304,855
28,132
52,696
2,707
64,590
726,146
37,152
122,455
5,696
122,950
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
16
24,983
7,747
3,955
464
10,943
104,869
9,495
3,955
694
20,462
218,393
12,948
3,955
1,137
35,073
507,481
24,350
3,955
2,598
75,224
NOTE:
1) Refer to Exhibit 6-1 fora description of treatment train technologies.
2) Average costs per size category are based on median population and associated flows, assuming one entry point per system.
-------
Exhibit 6-2 continued
Average Compliance Technology Costs (Treatment Train 9 through 16)
Size Category
3,300-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
9
3,803,627
219,658
-
3,630
582,324
15,997,473
1,075,859
-
17,881
2,603,788
38,371,441
2,834,902
-
47,159
6,504,053
312,631,164
23,343,166
-
388,505
53,241,841
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
10
3,803,627
219,658
-
3,630
582,324
15,997,473
1,075,859
-
17,881
2,603,788
38,371,441
2,834,902
-
47,159
6,504,053
312,631,164
23,343,166
-
388,505
53,241,841
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
11
3,935,731
561,953
5,382
108,199
1,042,166
15,030,714
2,429,780
7,898
526,046
4,375,365
35,105,717
5,457,968
12,515
1,376,897
10,149,778
268,100,910
40,762,901
69,107
10,542,021
76,618,275
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
12
3,935,731
561,953
5,382
108,199
1,042,166
15,030,714
2,429,780
7,898
526,046
4,375,365
35,105,717
5,457,968
12,515
1,376,897
10,149,778
268,100,910
40,762,901
69,107
10,542,021
76,618,275
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
13
3,935,731
561,953
197,836
65,423
1,017,556
15,030,714
2,429,780
476,817
153,221
4,046,803
35,105,717
5,457,968
988,706
333,600
9,198,627
268,100,910
40,762,901
7,263,436
2,436,597
69,191,945
14
$ 1,510,476
$ 73,586
$ 118,824
$ 21,276
$ 248,656
$ 5,348,970
$ 280,577
$ 177,968
$ 55,113
$ 857,394
$ 7,496,682
$ 580,569
$ 285,159
$ 124,631
$ 1,439,750
$16,835,071
$ 4,050,692
$ 1,459,396
$ 935,117
$ 6,712,677
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
15
1,510,476
73,586
410,913
22,801
277,753
5,348,970
280,577
907,819
61,314
932,488
7,496,682
580,569
1,816,207
139,960
1,599,600
16,835,071
4,050,692
12,597,315
1,009,510
7,838,411
$ 1
$
$
$
$
$ 4
fl-
ip
$
$
fl-
ip
$ 8
$
$
$
$ 1
$50
$ 6
fl-
ip
fl-
ip
$11
16
,253,212
63,448
5,085
8,312
190,535
,186,608
283,584
5,085
36,512
715,762
,718,795
735,847
5,735
94,447
,653,828
,584,105
,008,680
9,572
769,897
,554,262
NOTE:
1) Refer to Exhibit 6-1 fora description of treatment train technologies.
2) Average costs per size category are based on median population and associated flows, assuming one entry point per system.
-------
Exhibit 6-2 continued
Average Compliance Technology Costs (Treatment Train 17 through 23)
Size Category
3,300-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
17
2,495,535
8,922
120,582
43,485
299,350
9,160,742
33,090
271,621
67,770
991,208
19,339,675
70,123
548,757
117,662
2,065,112
139,484,072
497,678
3,945,878
699,342
14,735,792
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
18
2,495,535
43,567
120,582
43,485
333,995
9,160,742
165,678
271,621
67,770
1,123,796
19,339,675
350,577
548,757
117,662
2,345,566
139,484,072
2,488,884
3,945,878
699,342
16,726,998
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
19
3,803,627
219,658
4,624
36,409
615,539
15,997,473
1,075,859
6,097
175,430
2,761,913
38,371,441
2,834,902
8,799
461,046
6,918,771
312,631,164
23,343,166
41,919
3,790,974
56,648,268
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
20
3,803,627
219,658
4,624
36,409
615,539
15,997,473
1,075,859
6,097
175,430
2,761,913
38,371,441
2,834,902
8,799
461,046
6,918,771
312,631,164
23,343,166
41,919
3,790,974
56,648,268
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
21
3,173,443
822,852
-
-
1,274,679
13,884,129
3,866,644
-
-
5,843,432
34,069,251
9,909,392
-
-
14,760,087
239,916,151
76,696,105
-
-
110,854,767
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
22
1,644,085
586,793
-
-
820,874
7,198,773
2,643,298
-
-
3,668,241
17,673,110
6,602,335
-
-
9,118,589
124,585,586
48,323,042
-
-
66,061,227
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
23
562,676
861,662
-
-
941,775
2,462,155
3,842,689
-
-
4,193,244
6,042,288
9,539,669
-
-
10,399,955
42,558,851
68,900,160
-
-
74,959,583
NOTE:
1) Refer to Exhibit 6-1 fora description of treatment train technologies.
2) Average costs per size category are based on median population and associated flows, assuming one entry point per system.
-------
Exhibit 6-2 continued
Average Compliance Technology Costs (Treatment Train 17 through 23)
Size Category
<100
101-500
501-1000
1,001-3,300
$
$
$
$
$
$
$
$
$
fl-
ip
fl-
ip
$
$
$
$
$
$
fl-
ip
fl-
ip
fl-
ip
17
39,477
5,614
32,187
7,404
19,782
83,934
10,099
37,318
7,447
28,991
132,451
16,999
46,083
7,713
41,565
770,479
35,971
70,291
18,846
134,179
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
18
39,477
6,431
32,187
7,404
20,599
83,934
13,220
37,318
7,447
32,113
132,451
15,164
46,083
7,713
39,729
770,479
21,577
70,291
18,846
119,786
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
19
49,780
2,762
3,500
584
8,375
210,637
9,793
3,500
1,712
31,719
485,414
23,236
3,500
3,888
73,274
1,302,755
67,590
3,665
11,170
202,077
$
$
$
fl-
ip
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
20
49,780
2,762
3,500
584
8,375
210,637
9,793
3,500
1,712
31,719
485,414
23,236
3,500
3,888
73,274
1,302,755
67,590
3,665
11,170
202,077
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
21
26,357
5,420
-
-
9,172
152,478
34,134
-
-
55,844
374,156
87,479
-
-
140,750
1,049,843
258,026
-
-
407,500
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
22
13,619
4,433
-
-
6,372
78,866
26,552
-
-
37,781
193,617
66,321
-
-
93,888
543,574
189,924
-
-
267,317
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
23
4,671
6,725
-
-
7,390
27,027
39,804
-
-
43,652
66,325
98,815
-
-
108,258
186,123
280,998
-
-
307,497
NOTE:
1) Refer to Exhibit 6-1 fora description of treatment train technologies.
2) Average costs per size category are based on median population and associated flows, assuming one entry point per system.
-------
6.2.3 Monitoring and Administrative Costs
Monitoring Costs
Monitoring under the existing arsenic standard occurs annually for surface water systems, and
triennially for ground water systems. Currently, when triggered by a violation the system must
perform three additional tests within the month. Under the proposed rule to be promulgated in
January 2001, systems will still perform monitoring annually (for surface water systems) or every
three years (for ground water systems); however, when triggered by a violation, the system will
perform quarterly monitoring rather than three more samples in one month. All large systems
must comply no later than three years after promulgation (by December 31, 2004). Small systems
must comply within five years of promulgation. Specifically, small surface water systems must
comply by December 31, 2006, and small ground water systems must comply by December 31,
2007.
If quarterly monitoring is required it will continue until the State determines that the system is
"reliably and consistently" below the MCL. States are able to make this determination after
ground water systems have taken two quarterly samples and surface water systems have taken
four quarterly samples. Additionally, States may grant a nine year monitoring waiver to qualifying
systems, an option not previously available. To be eligible for a waiver, a system must meet the
following criteria:
1. Demonstrate adequate source water protection;
2. Demonstrate that the arsenic is naturally occurring;
3. Demonstrate that three previous samples were below the MCL.
The proposed requirements will impose new costs for some systems as follows:
NTNCs will incur the full costs of the monitoring requirements for the first time, unless
they are located in States that already require NTNCs to monitor for arsenic. For NTNCs
that are currently required to monitor for arsenic, the incremental monitoring costs will
depend on how the proposed national requirements compare with the current State
requirements. (It is assumed that states currently require NTNCs to monitor using the
ground water requirements. It is also assumed that NTNCs will continue to follow the
ground water requirements under the revised rule.)
CWSs may incur additional costs if they find exceedances more frequently at the proposed
MCL.
The cost of monitoring includes preparing and analyzing the sample. Collecting the sample,
arranging for delivery to the laboratory, and reviewing the results of the analysis is assumed to
require one hour of the system operator's time (at an estimated cost of $28 per hour). EPA has
assumed that all systems are equipped to collect samples. Therefore, no additional costs are
assumed for installing taps, re-piping of wells or other investments to permit sampling. EPA has
assumed that systems will utilize one of two laboratory methods: 1) stabilized temperature
platform graphite furnace atomic absorption (STP-GFAA) or 2) graphite furnace atomic
Chapter 6, Cost Analysis 6-11 Proposed Arsenic in Drinking Water Rule RIA
-------
absorption (GFAA). Both techniques cost $40 per sample.
Total net monitoring costs were estimated over a 20 year period at discount rates of three and
seven percent. The net costs are equal to the difference between the cost of the proposed
monitoring requirements and the cost of the current monitoring requirements. Cost and hour
burden, to the system and the State are listed below in Exhibit 6-4. The cost of routine
monitoring, triggered monitoring, waiver application and public notification are all included in the
total system costs. Miscellaneous costs related to sending samples to be analyzed and sending
public notification to customers are also included in the system cost.
During the first year of implementation all systems will incur costs related to routine monitoring.
In addition, systems in violation will incur cost related to triggered quarterly monitoring. Under
the revised rule, a percentage of the systems will have monitoring waivers in subsequent years
when monitoring is required Monitoring waivers are not granted under the existing rule,
therefore the number of systems required to conduct routine monitoring under the revised rule is
less than that under the existing rule. For this reason, the annual net cost of monitoring between
the revised rule and the existing rule may be negative, or less expensive, after the initial year of
implementation. The inputs and methodology associated with this analysis are presented in detail
in the Information Collection Request for the Proposed Arsenic in Drinking Water Rule (EPA
,1999)
Administrative Costs
States and systems will incur administrative costs to implement the revised arsenic program
proposed under the Arsenic Rule. States and systems will need to allocate time for their staff to
establish and maintain the programs necessary to comply with the revised arsenic standard and the
new monitoring requirements. Exhibit 6-3(a) lists the one-time state activities involved in starting
up the program following promulgation of the proposed rule. For example, start-up activities
may include developing and adopting state regulations that meet the new Federal arsenic
requirements. Resources are estimated in terms of full time equivalents (FTEs), which EPA has
assumed to cost $64,480 per FTE, including overhead and fringe. Systems also have start-up
costs for reviewing the regulation and training operators. Exhibit 6-3(b) lists the one-time system
start-up activities. The two primary activities that systems will perform to comply with the
revised arsenic rule are reading and understanding the rule and operator training. For all systems
the estimated time required to review the rule is eight hours. Systems serving fewer than 10,000
people have an estimated time of 16 hours to train operators; the estimated time for systems
serving more than 10,000 people is 32 hours. The rate for all start-up activities for systems
serving fewer than 10,000 people is $15.03 per hour and $29.03 per hour for systems serving
more than 10,000 people.
States will also be required to spend time responding to systems that report MCL exceedances or
systems that request a waiver (Exhibit 6-4). Hour burdens for States to review waiver
applications, record monitoring of a sample, and issue a violation letter are the same for small and
large systems. The number of hours required to review a single permit is twice as large for
systems serving more than 10,000 people than for systems serving less than 10,00 people. The
Chapter 6, Cost Analysis 6-12 Proposed Arsenic in Drinking Water Rule RIA
-------
unit cost for all activities is consistent across all activities and size categories ($41.47 per hour)
(EPA, 1997).
The number of hours required at the system level to perform the responsibilities related to
monitoring are the same for systems serving fewer than 3,300 people and systems serving more
than 3,300 people. However, the hourly rate for systems serving more than 3,300 people is
almost double ($29.03) the rate for systems serving fewer than 3,300 people ($15.03).
Exhibit 6-3 (a)
Estimated One-Time State Resources Required for Initiation of the Arsenic Rule
Administrative Activity
Estimated State
Resources (FTE)
One Time Start-up Activities
Regulation Adoption and Program Development (CWS)
System Training and Technical Assistance (CWS)
System Training and Technical Assistance (NTNC)
Staff Training (CWS)
Subtotal CWS Costs
Subtotal NTNCs Cost
National Total* CWS Costs
National Total* NTNCWS Costs
0.5
1.0
1.0
0.23
1.73
1.0
100.34
58.0
Estimated Cost
$32,200
$64,500
$64,500
$14,800
$111,500
$64,500
$6,467,000
$3,741,000
*National totals include estimates for all states, territories, and tribes.
Exhibit 6-3 (b)
Estimated One-Time System Resources Required for Initiation of the Arsenic Rule
System Size Category
One Time Start-up Activity
Reading and Understanding
Rule
Operator Training
< 10,000 people
Hours
8
16
Rate
$15.03
$15.03
> 10,000 people
Hours
8
32
Rate
$29.03
$29.03
Chapter 6, Cost Analysis
6-13
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-4
Unit Resources Required for Monitoring, Implementation, and Administration*
System Size Category
State Activity
Review a waiver application
Record monitoring of a
sample result
Issue a single violation letter
Review a single permit
System Activity
Apply for a waiver
Take a sample
Report a sample
Prepare and Send Public
Notification
< 10,000 people
Hours
8
1
4
16
Rate
$41.47
$41.47
$41.47
$41.47
<3,000 people
Hours
16
1
1
8
Rate
$15.03
$15.03
$15.03
$15.03
> 10,000 people
Hours
8
1
4
16
Rate
$41.47
$41.47
$41.47
$41.47
>3,300 people
Hours
16
1
1
8
Rate
$29.03
$29.03
$29.03
$29.03
*Estimates are provided in May
Source: Information Collection
1999 dollars, updated from 1997 dollars using the CPI-U for all items.
Request for the Public Water System Supervision Program.
6.2.4 Predicting Compliance Decisions (Compliance Decision Tree)
There is substantial variability in how systems will elect to comply with the Arsenic Rule. Choices
of compliance method will vary depending on baseline source water arsenic concentrations,
system size and location, types of treatment currently in place, and availability of alternative
sources. In addition, the source water pH, total dissolved solids, sulfides and other salts can
change the effectiveness of technologies in removing arsenic.
The RIA reflects this variability by predicting a range of compliance responses for different system
types and sizes. The compliance decision tree specifies the percentage of systems in different
categories that will choose specific compliance options, given the removal required by the MCL
option and the baseline occurrence of arsenic in source water. For example, for a target MCL of
10 ug/L, the decision tree specifies the probability of different compliance choices for systems
with different baseline influent concentrations (e.g., <10 ug/L, 10-20 ug/L, etc.), different sizes
(e.g., population < or > 1,000), different sources (ground water or surface water), and different
existing treatment facilities. The compliance choices are defined by a treatment technology and
(where relevant) a waste disposal option, and/or pre-treatment technology.
EPA reviewed a draft of the compliance decision tree at an American Water Works Association
(AWWA) technical workgroup meeting in February 1999, and made revisions based on the
comments received at that meeting. The final compliance decision tree, as well as a discussion of
Chapter 6, Cost Analysis
6-14
Proposed Arsenic in Drinking Water Rule RIA
-------
the assumptions made during its development, is provided in Appendix A ("Cost Analysis
Appendix") by system size and type.
6.2.5 Calculating Costs
Different methods were used to assess costs for three different categories of systems. A Monte-
Carlo simulation model (SafeWaterXL) was used to estimate costs for community water systems,
excluding the largest CWSs. A deterministic spreadsheet analysis was performed for NTNC
water systems, while a separate case-by-case analysis were performed for the very large systems
(serving more than one million people) that are expected to exceed one or more MCL options in
the baseline. The costs for the three system categories were then summed to calculate total
national costs. The methodology for calculating the costs for each of these system categories is
described separately below, beginning with a description of the SafeWaterXL model.
CIVS Cosfs
The national cost of compliance across CWSs (except those serving over one million people) was
estimated using SafeWaterXL, a Monte-Carlo simulation model developed in Microsoft Excel©
using the Crystal Ball© Monte-Carlo simulation add-in. SafeWaterXL forecasts a distribution of
costs around the mean compliance cost expected for each system size category. The Monte-Carlo
provides the flexibility to incorporate as much data as is available, while maintaining uncertainty
bounds to prevent any individual input from skewing the results. When sample data is not
available as single point estimates, this technique is an invaluable tool.
Historically most drinking water regulatory impact analyses used point-estimates to describe the
average system-level costs. By using SafeWaterXL, this analysis contains more detailed
descriptions of system-level cost. SafeWater XL describes system-level costs in terms of a
distribution. From the distribution, mean and median costs are available, as well as percentile
costs.
Model Structure
SafeWaterXL determines regulatory compliance costs for individual systems and subsequently
calculates a national average based on the mean value of these data points. To do so, each system
is assigned a random concentration from an occurrence distribution. This system concentration is
distributed across the number of sites of possible contamination for that system. The average
number of sites per system is determined based on the distribution of system intake sites for the
size category as estimated from the CWSS. However, SafeWaterXL does not assume that all
sites are equally likely to exceed the MCL standard. The likelihood of contamination is
determined on a site-by-site basis. The sum of the mean arsenic concentration of all sites within a
system must equal the mean arsenic concentration of the system. Given this upper bound, each
site is assigned a concentration based on the assumed relative standard deviation (RSD) around
the mean system occurrence.
The model then compares the concentration at each site to the proposed MCL standard; no costs
are incurred for those sites whose concentrations fall below the specified MCL. If the site is
Chapter 6, Cost Analysis 6-15 Proposed Arsenic in Drinking Water Rule RIA
-------
determined to be in violation of the MCL, then SafeWaterXL calculates the percent reduction in
arsenic concentration required to reduce the site concentration to 80 percent of the MCL standard
(this is a safety factor which includes a 20 percent excess removal to account for system over-
design). A treatment train is then assigned to the site based on a decision tree for the size and
type of the system. The decision tree and the selected treatment train reflect the removal
efficiencies of the chosen technology. For example, a technology is chosen based on matching the
removal efficiencies and the percentage removal required at the site (SafeWaterXL identifies three
categories of required removal: < 50 percent, 50-90 percent, > 90 percent).
In this manner, capital and O&M costs are calculated at the site level for the selected treatment
train. The system's cost of compliance is then determined by summing across the treating sites.
For each system in SDWIS in which a violation is expected, a cost is calculated with this method,
thereby creating an estimate of national compliance costs. Since household costs are also
calculated for each system, a similar distribution of the cost of compliance at the household level
are also created.
In order to develop more detailed results the compliance decision tree is employed at the site
level, so that only those sites requiring treatment would incur costs. The resulting total national
compliance cost is expected to be a truer representation of the impact of the Arsenic Rule on
systems. The sections below will describe the data needed to develop cost estimates for the entire
universe of systems affected by the Arsenic Rule. After the discussion of data requirements, the
SafeWaterXL model is described as it is used for this rule
Model Inputs
Number of Systems: The universe of public and private ground and surface water systems is
taken from Safe Drinking Water Information System (SDWIS), EPA's national regulatory
database for the drinking water program. Based on data extracted December 1998, a total of
54,352 CWSs and 20,255 NTNCs are subject to the new requirements proposed under the
Arsenic Rule. It is necessary to compile this data by system size, water source, and ownership, as
costs may vary by these characteristics. SafeWaterXL calculates costs for public and private
systems (the latter also includes "other" or "ancillary" systems), and surface and ground water
systems. A summary table of this breakdown is provided in Chapter 4, "Baseline Analysis."
Entry points per System: SafeWaterXL estimates each system's cost of compliance at the
treatment site level. This modeling approach is used because a system may include more than one
treatment site. Entry points are used as a proxy for potential or actual points of treatment. For
example, a given water system may have three entry points: one entry point that currently treats,
while two may not have treatment in place. Data on the distribution of the number of system
entry points for each size category and type was extracted from the Community Water Supply
Survey (CWSS). Linear interpolation was used to estimate values for the number of sites in cases
where there were no survey data (see Chapter 4, "Baseline Analysis").
Population Served by System: A system's size is determined by the number of people served by
that system. These numbers were extracted from the SDWIS database (see Chapter 4, "Baseline
Analysis"). Systems are grouped into eight categories to help identify systems with related
Chapter 6, Cost Analysis 6-16 Proposed Arsenic in Drinking Water Rule RIA
-------
characteristics so that any data or resources may be pooled during analysis. SafeWaterXL
recognizes the following size categories:
< 100
101-500
501-1000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001- 1,000,000
Flow Rate Parameters: System size is further defined by its flow, which is calculated as a power
law function of the population served. These functions were derived by EPA, and their derivation
can be found in the Model Systems report (Geometries and Characteristics of Public Water
Systems., May 1999, EPA pending). The equation form is shown below.
Average Flow = a, (Population)
(Equation 1)
Design Flow = max
2 Average Flow
aD (Population) D
(Equation 2)
Where: aA, bA, aD, bD =
Population =
the regression parameters derived for flow vs.
population
the population served by the appropriate system type and
primary source.
The regression parameters used in the cost model are provided in Exhibit 6-5. Values are
provided for design and average flow for public and private ground water and surface water
supplies. SafeWaterXL divides system design flow and average daily flow equally among all entry
points. Treatment costs are only assigned to the minimum portion of flow that must be treated in
order to achieve the new concentration standard, a process referred to as "blending".
Chapter 6, Cost Analysis
6-17
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-5
Flow Regression Parameters
by Water Source and System Ownership
Average Row
a b
Design Flow
a b
Ground Water
Pubic
Rivate
Publc-Rrch
Private-Purch
0.08558
0.06670
0.04692
0.05004
1.05840
1.06280
1.10190
1.08340
0.54992
0.41682
0.31910
0.32150
0.95538
0.96078
0.99460
0.97940
Surface Water
Pubic
Private
Pubic-Rrch
Pnvate-Purch
0.14004
0.09036
0.04692
0.05004
0.99703
1.03340
1.11020
1.08340
0.59028
0.35674
0.20920
0.20580
0.94573
0.96188
1.04520
1.00840
Average Consumption per Household: Household costs depend on the average annual
consumption per residential connection. These mean estimates are provided in Chapter 4,
"Baseline Analysis." Depending on the system's characteristics, SafeWaterXL multiplies the
appropriate mean consumption (kgal) with the system's computed cost per thousand gallons to
arrive at the average annual cost of compliance per household for a community water system.
Mean System Occurrence: Arsenic occurrence data are based on the EPA's Arsenic Occurrence
in Public Drinking Water Supplies report and are represented by a lognormal distribution. The
distribution is truncated at 50 ug/L, the current arsenic standard, because it is assumed that all
arsenic reductions attributable to the new standard start at the previous standard (i.e. all systems
are currently in compliance with the current standard). Baseline occurrence is distinguished
between ground and surface water systems and is provided in Chapter 4 ("Baseline Analysis") as a
lognormal distribution. Exhibit 6-6 shows the percent of ground and surface water systems with
arsenic concentrations greater than 3, 5, 10, and 20 parts per billion (ppb).
Exhibit 6-6
Arsenic Occurrence by Water Source
Source
GW
SW
% of systems greater than (ppb)
3
19.90
6.01
5 10 20
12.10 5.43 2.06
2.90 0.75 0.26
Chapter 6, Cost Analysis
6-18
Proposed Arsenic in Drinking Water Rule RIA
-------
Relative Intra-System Standard Deviation of Arsenic Concentrations: The relative intra-system
standard deviation of the site concentrations within a system is calculated using data from a 25
state arsenic occurrence study ("Arsenic Occurrence in Public Drinking Water Supplies," EPA
1999). SafeWaterXL uses a default value of 0.64. This standard deviation is applied to the mean
system concentration to generate individual entry points concentrations within the system.
Compliance Decision Trees: The decision trees represent EPA's best estimate of the treatment
train technologies system operators will choose to achieve a particular percentage reduction in
arsenic concentration. Decision trees are specific to the system's size categories and source
water. These are provided in Appendix A, "Cost Analysis Appendix."
Removal Efficiencies, Treatment Target, and Blending: Each treatment train is associated with an
arsenic removal efficiency that is assumed to be constant across system types. The removal
efficiencies for the 25 treatment trains available under the Arsenic Rule were presented in Exhibit
6-1. SafeWaterXL employs these efficiencies with the blending principle, to determine the
amount of flow that requires treatment in order for the entry point to meet the treatment target.
Blending uses the entry point concentration and treatment train removal efficiency to determine
the fraction of flow required to obtain the treatment target. The treatment target is set at 80
percent of the MCL and represents the level to which systems will be over-designed to ensure
compliance with the MCL.
SafeWaterXL employs the blending concept through the following equation at the entry point
level:
TreatmentTarget
-1) (% Site Flow}
Fraction of flow treated = min^ sueConcentration (Equation 4)
% Removal Efficiency
Where: TreatmentTarget = the target MCL with 80% safety factor
Site Concentration = arsenic concentration at the site
% Removal Efficiency = % removal efficiency of treatment train chosen
% Site Flow = % of total flow at that site
Note that the blending technique is used only for the those systems expected to require greater
than 90 percent removal in order to achieve compliance with the new MCL standard. In addition,
SafeWaterXL does not employ this technique for those systems that select treatment trains
involving POE or POU devices.
Equipment Life, Discount Rate and Capitalization Rates: System and State implementation costs
are tracked for a twenty-year period. This time frame was selected for two reasons:
1) technologies are estimated to have a twenty-year life (which includes replacement for some
technologies with lives shorter than twenty years);
Chapter 6, Cost Analysis 6-19 Proposed Arsenic in Drinking Water Rule RIA
-------
2) water systems often finance their capital improvements over a twenty year period.
Exhibit 6-7
Summary of Recommended Cost of Capital Estimates
(as of March 1998)
Ownership Type Size Category
Estimated After-Tax
Cost of Capital
NON-SMALL
Investor owned
Publicly owned
SMALL
Private
Public
10,001-50,000
>50,000
10,001-50,000
>50,000
1-500
501-10,000
1-500
501-10,000
5.26%
5.94%
5.26%
5.23%
4.17%
4.17%
5.10%
5.20%
Source: Development of Cost of Capital Estimates for Public
Water Systems (Draft Final Report). Prepared for U.S. EPA by
Apogee/Hagler Bailly, Inc. under subcontract to International
Consultants, Inc. June 1998.
Two different adjustments are made in this analysis in order to render future costs comparable
with current costs, reflecting the fact that a cost outlay today is a greater burden than an
equivalent cost outlay sometime in the future. The first adjustment is made when the cost
estimates that are derived are being used as an input in benefit-cost analysis. In this instance,
costs are annualized using a social discount rate so that the costs of each regulatory option can be
directly compared with the annual benefits of the corresponding regulatory option. Annualization
is the same process as calculating a mortgage payment; the result is a constant annual cost to
compare with constant annual benefits.
The choice of an appropriate social discount rate has been, and continues to be, a very complex
and controversial issue among economists and policy makers alike. Therefore, the Agency
compares costs and benefits using two alternative social discount rates, in part to determine the
effect the choice of social discount rate has on the analysis. The annualized costs of each
regulatory option are calculated and displayed using both a seven percent discount rate required
by the Office of Management and Budget (OMB) and a three percent discount rate which the
Agency believes more closely approximates the true social discount rate.
The second adjustment is made when the cost estimates that are derived are being used as an
input into an economic impact analysis, such as an affordability analysis or an analysis of system-
level costs or household-level costs. In these cases, rather than use a social discount rate when
determining the annualized costs, an actual cost-of-capital rate is used instead. This rate should
reflect the true after-tax cost of capital water systems face, net of any government grants or
subsidies. The cost of capital rates used in this analysis are shown in Exhibit 6-7 above.
Chapter 6, Cost Analysis
6-20
Proposed Arsenic in Drinking Water Rule RIA
-------
NTNC Costs
The cost for NTNCs is estimated using the mean values for system population for each system
service category, as shown in Chapter Four. As with the CWSs, cost is annualized over a twenty-
year period, at discount rates of three and seven percent. Assumptions regarding the monitoring
schedule correspond to the monitoring schedule for small ground water systems, including hour
burdens and hourly labor rates. The remaining assumptions required for determining cost are
described below.
Number of Systems, Sites per System and the Population Served: The non-transient non-
community water supply treatment decisions are modeled similarly to those for community water
supplies. The number of non-transient non-community water supplies is taken from EPA's
SDWIS, and include those systems as described in Geometries and Characteristics of Public
Water Supplies. For each service area type, the report lists the number of systems and the
average population served. The non-food manufacturing service area combines 16 categories that
were listed separately in the report. For this service area, the number of systems is the sum of the
16 categories and the average population served is the mean of the individual populations
weighted by the number of corresponding systems. Each of these systems has only a single site.
System Flows and Treatment Choices: For each service area, both design and average flows have
been derived by the Agency using literature values and best engineering judgment. There is no
primary survey data for non-community water systems that is equivalent to the CWSS that
provided data for the community water system flow calculations (Smith, personal
communication). The design flow is used to calculate the treatment capital costs while the
average flow is used in the operating and maintenance cost equations. For the non-transient non-
community water supplies, one of two treatment technologies was chosen based on the level of
the design flow. For service areas with design flows less than 2,000 gallons per day, POE
activated alumina is used; for all others, centralized activated alumina with 2,000 bed volume is
chosen (Kapadia, 1999a personal communication). Both treatment trains include pre-oxidation
and the centralized activated alumina also includes non-hazardous landfilling of the spent media
(Kapadia, 1999a personal communication).
Mean Arsenic Occurrence: The arsenic occurrence distribution used for ground water
community water supplies is also used for non-transient non-community water supplies. The
number of systems exceeding the MCL for each service area was calculated from the percent of
the distribution between the MCL and 100 jig/L. For this analysis, 100 jig/L was chosen as the
upper concentration limit because the non-transient non-community supplies have not been
previously regulated and occurrence values above the 50 jig/L regulatory level are possible.
Removal Efficiencies, Treatment Target, and Blending: The removal efficiency associated with
both POE activated alumina and centralized activated alumina is 95 percent. The NTNC model
uses this efficiency with the blending principle in the case of centralized activated alumina to
determine the amount of flow that requires treatment in order for the site to meet the treatment
target. The treatment target is set at 80 percent of the MCL and represents the level to which
systems will be over-designed to ensure compliance with the MCL. For POE activated alumina
Chapter 6, Cost Analysis 6-21 Proposed Arsenic in Drinking Water Rule RIA
-------
systems, all the flow is treated which may result in finished water below the treatment target
concentration.
Equipment Life, Discount Rate, and Capitalization Rates: As with the community water
supplies, the system implementation costs are tracked for a 20 year period. For the two service
areas using POE activated alumina, construction and forest service, the equipment is assumed to
last ten years with purchases in year zero and year ten. For the centralized activated alumina the
equipment is estimated to last 20 years. The cost estimates are annualized in the same manner as
those for the community water supplies.
Very Large CWS Costs
EPA evaluated the regulatory costs of compliance for very large systems that would be subject to
the new arsenic drinking water regulation. The nation's 25 largest drinking water systems (i.e.,
those serving a million people or more) supply approximately 38 million people and generally
account for about 15 to 20 percent of all compliance-related costs. Accurately determining these
costs for future regulations is critical. As a result, EPA has developed compliance cost estimates
for the arsenic and radon regulations for each individual system that serves greater than 1 million
persons. These cost estimates help EPA to more accurately assess the cost impacts and benefits of
the arsenic regulation. The estimates also help the Agency identify lower cost regulatory options
and better understand current water systems' capabilities and constraints.
The system costs were calculated for the 24 public water systems that serve a retail population
greater than 1 million persons and one public water system that serves a wholesale population of
16 million persons. The following are distinguishing characteristics of these very large systems:
(1) a large number of entry points from diverse sources;
(2) mixed (i.e. ground and surface) sources;
(3) occurrence not conducive to mathematical modeling;
(4) significant levels of wholesaling;
(5) sophisticated in-place treatment;
(6) retrofit costs dramatically influenced by site-specific factors; and
(7) large amounts of waste management and disposal which can
contribute substantial costs.
Generic models cannot incorporate all of these considerations; therefore, in-depth
characterizations and cost analyses were developed utilizing several existing databases and
surveys.
The profile for each system contains information such as design and average daily flows,
treatment facility diagrams, chemical feed processes, water quality parameters, system layouts,
and intake and aquifer locations. System and treatment data were obtained from the following
sources:
Chapter 6, Cost Analysis 6-22 Proposed Arsenic in Drinking Water Rule RIA
-------
(1) the Information Collection Rule (1997);
(2) the Community Water Supply Survey (1995);
(3) the Association of Metropolitan Water Agencies Survey (1998);
(4) the Safe Drinking Water Information System (SDWIS); and
(5) the American Water Works Association WATERSTATS Survey (1997)
While these sources contained much of the information necessary to perform cost analyses, the
Agency was still missing some of the detailed arsenic occurrence data in these large water
systems. Where major gaps existed, especially in ground water systems, occurrence data obtained
from the States of Texas, California, and Arizona, the Metropolitan Water District of Southern
California Arsenic Study (1993), the National Inorganic and Radionuclides Study (EPA, 1984),
and utility data was used. Based on data from the studies, detailed costs estimates were derived
for each of the very large water systems.
Cost estimates were generated for each system at several MCL options. The total capital costs
and operational and maintenance (O&M) costs were calculated using the profile information
gathered on each system, conceptual designs (i.e., vendor estimates and RS Means), and modified
EPA cost models (i.e., Water and WaterCost models). The models were modified based on the
general cost assumptions developed in the Phase I Water Treatment Cost Upgrades (EPA, 1998).
EPA consulted with the system operators to determine how each system would comply with
various MCL options and to assess the costs of their compliance responses. Preliminary cost
estimates were sent to all of the systems for their review. Approximately 30 percent of the
systems responded by submitting revised estimates and/or detailed arsenic occurrence data.
Based on the information received, EPA revised the cost estimates for those systems. EPA
developed cost estimates for three very large systems that are expected to have arsenic levels
below 50 |ig/L. These systems are located in Houston, TX, Phoenix, AZ, and Los Angeles, CA.
This analysis resulted in the estimated costs listed in Exhibit 6-8.
Exhibit 6-8
Annual Treatment Costs for Three Large CWSs Expected to
Undertake or Modify Treatment Practice to Comply with the Arsenic Rule
($ millions)
Large CWSs
Phoenix, AZ
Annual cost (3%)
Annual cost (7%)
Houston, TX
Annual cost (3%)
Annual cost (7%)
Los Angeles, CA
Annual cost (3%)
Annual cost (7%)
Population
Served
1,360,751
2,216,830
3,700,000
MCL(p.g/L)
3
$ 4.2
$ 4.8
$ 4.5
$ 5.3
$ 2.3
$ 2.3
5
$ 4.1
$ 4.7
$ 1.6
$ 1.9
$ 1.8
$ 1.8
10
$ 1.8
$ 2.1
$ 0.4
$ 0.4
$ 1.8
$ 1.8
20
$ 0.2
$ 0.2
$ 0.4
$ 0.4
$ 1.8
$ 1.8
Chapter 6, Cost Analysis
6-23
Proposed Arsenic in Drinking Water Rule RIA
-------
6.3 Results
This section presents the results of the national cost analysis. Unless otherwise specified, national
costs are presented in May 1999 dollars throughout this chapter.
6.3.1 National Costs
Exhibit 6-9 shows the total national cost breakdown across the four MCL options for the Arsenic
Rule. The system and state cost components of the total annual compliance costs are presented at
discount rates of three and seven percent. Expected system costs include treatment costs,
monitoring costs, and administrative costs of compliance. State costs include monitoring and
administrative costs of implementation. These cost components are also displayed. Exhibits 6
through 9 and 6 through 10 show the annual national costs of the Arsenic Rule under two
scenarios. Exhibits 6 through 9 shows the costs when both CWSs and NTNCs are required to
comply with an MCL. Exhibits 6 through 10 shows the costs when CWSs are required to comply
with an MCL, and NTNCs are only required to monitor for Arsenic.
Under the both scenarios, CWS costs are approximately $641 million at the 3 |ig/L MCL, $376
million at the 5 |ig/L MCL, $162 million at the 10 |ig/L MCL, and $61 million at the 20 |ig/L
MCL (at a three percent discount rate). State costs associated with CWS administration, at a 3
percent discount rate, are approximately $2.2 million at the 3 |ig/L MCL, $1.7 million at the 5
|ig/L MCL, $1.5 million at the 10 |ig/L MCL, and $1.3 million at the 20 |ig/L MCL.
When compliance with an MCL is required, the cost to NTNCs ranges from $26 million at the 3
|ig/L MCL, $16 million at the 5 |ig/L MCL, $7 million at the 10 |ig/L MCL, and $4 million at the
20 |ig/L MCL (at a three percent discount rate). State costs associated with NTNC
administration, at a three percent discount rate, are approximately $1.2 million at the 3 |ig/L
MCL, $1.1 million at the 5 |ig/L MCL, $1 million at the 10 |ig/L MCL, and $1 million at the 20
|ig/L MCL. When NTNCs are only required to monitor, their costs are approximately $1 million
across the MCLs (at a three percent discount rate). State costs associated with NTNC
administration under this scenario, at a three percent discount rate, also are approximately $1
million across the MCLs.
Chapter 6, Cost Analysis 6-24 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-9
Annual National System and State Compliance Costs
(CWSs and NTNCs Comply With MCL)
($ millions)
Discount Rate
cws
3% 7%
NTNC
3% 7%
TOTAL
3% 7%
MCL = 3 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$639.2 $746.4
$2.2 $3.0
$1.6 $1.9
$643.1 $751.4
$25.2 $30.5
$1.0 $1.2
$0.8 $0.9
$27.0 $32.7
$664.4 $777.0
$3.2 $4.2
$2.4 $2.9
$670.1 $784.0
MCL = 5 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$374.0 $436.0
$2.0 $2.8
$1.3 $1.6
$377.3 $440.4
$14.7 $17.8
$1.0 $1.2
$0.8 $0.9
$16.4 $19.8
$388.7 $453.8
$3.0 $3.9
$2.1 $2.5
$393.8 $460.2
MCL = 10 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$160.4 $186.7
$1.8 $2.5
$1.1 $1.3
$163.3 $190.5
$6.1 $7.4
$1.0 $1.2
$0.7 $0.8
$7.8 $9.3
$166.5 $194.1
$2.8 $3.7
$1.8 $2.1
$171.1 $199.9
MCL = 20 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$58.9 $68.3
$1.7 $2.4
$1.0 $1.2
$61.6 $71.8
$2.1 $2.6
$2.2 $2.4
$0.7 $0.8
$5.0 $5.8
$61.0 $70.8
$3.9 $4.9
$1.7 $2.0
$66.5 $77.6
-------
Exhibit 6-10
National Annual System and State Compliance Costs
(CWSs Comply With MCL/ NTNCs Monitor)
($ millions)
Discount Rate
cws
3% 7%
NTNC*
3% 7%
TOTAL
3% 7%
MCL = 3 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$639.2 $746.4
$2.2 $3.0
$1.6 $1.9
$643.1 $751.4
$0.9 $1.1
$0.6 $0.7
$1.5 $1.8
$639.2 $746.4
$3.1 $4.1
$2.2 $2.6
$644.6 $753.2
MCL =5 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$374.0 $436.0
$2.0 $2.8
$1.3 $1.6
$377.3 $440.4
$0.9 $1.1
$0.6 $0.7
$1.6 $1.8
$374.0 $436.0
$2.9 $3.9
$2.0 $2.3
$378.9 $442.2
MCL = 10 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$160.4 $186.7
$1.8 $2.5
$1.1 $1.3
$163.3 $190.5
$1.0 $1.1
$0.6 $0.7
$1.6 $1.9
$160.4 $186.7
$2.8 $3.7
$1.7 $2.1
$164.9 $192.4
MCL =20 mg/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$58.9 $68.3
$1.7 $2.4
$1.0 $1.2
$61.6 $71.8
$1.0 $1.1
$0.7 $0.7
$1.6 $1.9
$58.9 $68.3
$2.7 $3.5
$1.6 $1.9
$63.2 $73.7
-------
6.3.2 Costs by System Size and Type
This section presents the overall national compliance costs for water systems and for states at
three and seven percent discount rates. Exhibits 6-11 and 6-12 show a detailed breakout of
treatment and monitoring and administrative costs, respectively, by system type and size for the
various MCLs.
Exhibit 6-11
Total Annual CWS Treatment Costs Across MCL Options
by System Size and Type
($ millions)
System Size
MCL ((j.g/L)
3
5
3%
<100
101-500
501-1,000
1001-3300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
1,000,000 +
$
$
$
$
$
$
$
$
$
22.5
66.7
35.6
83.4
93.9
166.8
59.3
100.0
11.0
$
$
$
$
$
$
$
$
$
7%
<100
101-500
501-1,000
1001-3300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
>1, 000,000
$
$
$
$
$
$
$
$
$
24.8
73.5
40.2
96.7
113.7
199.6
70.6
114.9
12.4
$
$
$
$
$
$
$
$
$
Discount Rate
13.9
41.0
21.8
49.0
53.9
96.2
34.0
56.6
7.6
Discount Rate
15.3
45.1
24.7
56.7
65.1
115.2
40.5
65.0
8.5
10
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
6.3
17.9
9.5
21.2
23.2
40.6
14.2
23.5
4.0
6.9
19.7
10.7
24.5
28.0
48.6
16.9
27.0
4.4
20
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
2.4
6.9
3.6
7.9
8.3
14.3
5.0
8.0
2.4
2.6
7.6
4.1
9.1
10.0
17.1
6.0
9.2
2.5
Chapter 6, Cost Analysis
6-27
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-12
Total Annual CWS Monitoring and Administrative Costs Across MCL Options
by System Size and Type ($ millions)
System Size
MCL (^g/L)
3
5
3%
<100
101-500
501-1,000
1001-3300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
$
$
$
$
$
$
$
$
0.5
0.5
0.2
0.2
0.3
0.3
0.1
0.1
$
$
$
$
$
$
$
$
7%
<100
101-500
501-1,000
1001-3300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
$
$
$
$
$
$
$
$
0.6
0.7
0.2
0.3
0.4
0.4
0.1
0.1
$
$
$
$
$
$
$
$
Discount Rate
0.5
0.5
0.2
0.2
0.2
0.3
0.1
0.1
Discount Rate
0.6
0.7
0.2
0.3
0.3
0.4
0.1
0.1
10
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
0.4
0.5
0.1
0.2
0.2
0.2
0.04
0.04
0.6
0.7
0.2
0.3
0.3
0.3
0.1
0.1
20
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
0.4
0.5
0.1
0.2
0.2
0.2
0.03
0.03
0.6
0.7
0.2
0.3
0.3
0.3
0.05
0.04
Chapter 6, Cost Analysis
6-28
Proposed Arsenic in Drinking Water Rule RIA
-------
6.3.3 Costs per Household
Household level costs are considered a good proxy for the affordability of rule compliance with
regard to CWSs, since water systems recover costs at the household level through increased
water rates. This of course assumes that non-residential customers of water systems, such as
businesses, can pass along any increase in water costs to their customers through increased prices
on their goods or services. In order to calculate the number of households served by systems that
will treat, the expected number of treating systems is multiplied by the average number of
households per system (varies by system type and size). Exhibit 6-13 presents the total number of
households served by CWS that treat, by size category.
Exhibit 6-13
Number of Households in CWSs Expected to Treat
by Size Category and MCL (M9/L) Option
3
5
10
20
<100
93,886
58,654
26,856
10,525
101-500
366
229
103
41
,195
,192
,290
,357
501-1,000
355
223
102
40
,994
,048
,186
,811
1,001-
3,300
1,006,599
626,510
290,138
117,842
3,301-
10,000
1,623,698
1,012,724
477,938
195,671
10,001-
50,000
3,254,167
2,084,853
997,914
405,714
50,001-
100,000
1,452,752
905,886
465,003
188,798
100,001-
1,000,000
3
1
,084,826
,806,388
936,602
364,907
SafeWaterXL determines household costs separately for each affected CWS, by first dividing the
CWSs annual compliance cost by the CWS's average daily flow (1,000 gallons per day), and then
multiplied by 365 days to determine the CWS's cost of compliance per 1,000 gallons produced.
Finally, the CWS's cost of compliance per 1,000 gallons (kgal) is multiplied by the average annual
consumption per residential connection (kgal), to arrive at the average annual cost of compliance
per household for the CWS. The estimates of average annual consumption per residential
connection used in this analysis are provided in Chapter 4, "Baseline Analysis."
Given expected household costs for each individual system, the average is then calculated for each
size category. Exhibit 6-14 shows the average annual household costs by system size, across the
four regulatory options.
The range of household costs for the MCL of 5 |ig/L ranges from less than $2 to approximately
$373; the costs for the MCL of 3 |ig/L range from less than $3 to $374; the costs for the MCL of
10 |ig/L, range from less than $1 to $378; and the costs for the MCL of 20 |ig/L, range from less
than $1 to $398.
Exhibits 6-15 through 6-18 compare the distribution of annual household costs across public
water systems serving fewer than 10,000 people, for MCLs 3, 5, 10, and 20, respectively. The
exhibits demonstrate the maximum annual costs that different percentages of households in
treating systems face. Comparison of Exhibits 6-15 through 6-18 illustrates that regulatory
compliance costs decrease across MCLs. This observation is depicted by the consistent shift to
the left of cost curves across system size categories, when comparing incremental increases in the
MCL.
Chapter 6, Cost Analysis 6-29 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-14
Mean Annual Household Costs Across MCL Options by System Size
SIZE CATEGORIES
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
1,000,000 +
All categories
MCL (^g/L)
3
$368.13
$258.68
$106.07
$63.61
$44.28
$36.39
$29.52
$23.47
$2.70
$43.73
5
$363.65
$253.64
$103.74
$60.38
$40.77
$33.22
$26.64
$21.20
$1.73
$39.18
10
$357.17
$246.38
$98.35
$56.51
$37.04
$29.13
$22.80
$18.32
$0.89
$33.05
20
$348.72
$237.67
$93.25
$51.80
$32.52
$24.99
$19.44
$15.41
$0.55
$23.62
Chapter 6, Cost Analysis
6-30
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-15
Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving < 10,000 People
MCL 3 ug/L
O
.C
HI
to
3
0
I
'o
HI
1 00%
90%
80%
70%
60%
50%
o 40%
HI
Q.
30%
20%
10%
$100 $200 $300 $400 $500
Maximum Annual Household Treatment Costs
$600
Exhibit 6-16
Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving < 10,000 People
MCL 5 ug/L
1 00%
90%
80%
O 70%
.c
HI
3
0
I
'o
HI
HI
Q.
60%
50%
40%
30%
20%
10%
$100 $200 $300 $400 $500
Maximum Annual Household Treatment Costs
$600
Chapter 6, Cost Analysis
6-31
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 6-17
Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving < 10,000 People
MCL 10 ug/L
-101-500
- - -501-1,000
1,001-3,300
-3,301-10,000
10%
$100 $200 $300 $400 $500
Maximum Annual Household Treatment Costs
$600
Exhibit 6-18
Annual Treatment Costs Per Household Across Public GW CWSs
Expected to Treat and Serving < 10,000 People
MCL 20 ug/L
100%
o
.c
0)
to
I
5
HI
10%
$100 $200 $300 $400 $500
Maximum Annual Household Treatment Costs
Chapter 6, Cost Analysis
6-32
Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 6, Cost Analysis 6-33 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapter 7: Comparison of Costs and Benefits
7.1 Introduction
In this RIA, EPA has analyzed the cost and benefits of regulating arsenic concentrations in
drinking water to four different MCL standards. The four options considered reflect increasing
levels of protection against exposure to arsenic in drinking water, employing a range of MCLs
from 20 |ig/L to 3 |ig/L. As the MCL provisions for the four options become increasingly strict,
the associated costs and benefits also increase incrementally. Chapter 5 ("Benefits Analysis")
describes in detail the estimated national health benefits of the Arsenic Rule options, while
Chapter 6 ("Cost Analysis") describes the projected national compliance cost estimates. This
chapter presents a summary and comparison of the national benefits and costs and a cost-
effectiveness analysis for each of the MCL options.
7.2 Summary of National Costs and Benefits
7.2.1 National Cost Estimates
National compliance costs to PWSs for treatment (both annualized capital and O&M), monitoring
and administrative activities, and costs to states, including any one-time start-up costs, for
regulatory implementation and enforcement, were estimated and described in Chapter 6. The
national costs for community water systems to comply with the four MCL options ranges from
$61.6 million (MCL=20 |ig/L) to $643.1 million (MCL=3 |ig/L) annually based on a discount rate
of three percent. Assuming a seven percent discount rate, the range of total national cost for
community water systems ranges from $71.8 million to $751.4 million annually. The national cost
of compliance with the Arsenic Rule options for NTNCs is approximately $1.5 million at a 3
percent discount rate, and $2 million at a 7 percent discount rate.
7.2.2 National Benefits Estimates
Chapter 5 contains a detailed summary of the methodology used to estimate a range of national
health benefits from avoided bladder cancer cases as a result of the four Arsenic Rule MCL
options. The dollar value of the estimated health benefits associated with each of the four rule
options were calculated based on lower and upper bound estimates of avoided bladder cancer
cases. Although the value of reducing the risk of lung cancer was estimated, these estimates
contain a great deal of uncertainty. Therefore, only the value of reducing the risk of bladder
cancer will be used in the comparison of benefits and costs. For community water systems that
must install treatment equipment or make other modifications to their treatment processes
resulting in reduced arsenic concentrations, the national benefits range from $7.9 million
(MCL=20 |ig/L) to $43.6 million (MCL=3 |ig/L) annually, based on the lower bound estimates of
bladder cancer cases avoided. Under the upper bound scenario, the health benefits from avoided
bladder cancer increase from $29.8 million at an MCL of 20 |ig/L to $104.2 million annually at an
MCL of 3 |ig/L.
Chapter 7, Comparison of Costs and Benefits 7-1 Proposed Arsenic in Drinking Water Rule RIA
-------
7.3 Comparison of Benefits and Costs
This section presents a comparison of total national benefits and costs for each of the Arsenic
Rule options considered. Three separate analyses are considered, including a summary of
benefit/cost ratios and net-benefits, a direct comparison of aggregate national cost and benefits,
and the results of a cost-effectiveness analysis of each regulatory option.
7.3.1 National Net Benefits and National Benefit-Cost Comparison
Exhibit 7-1 describes the net benefits and the benefit/cost ratios under various MCL options for
PWSs at 3 and 7 percent discount rates. Under both the lower and upper bound scenarios of
avoided bladder cancer cases in Exhibit 7-1, the net benefits are negative and decreasing as the
Arsenic Rule MCL options become increasingly more stringent. Similarly, the benefit/cost ratios
are less than one and decrease with each more stringent MCL option. Costs outweigh the
quantified benefits under the four MCL options, with benefit/cost ratios all below one, and range
from 0.07 (MCL=3 |ig/L) to 0.48 (MCL=20 |ig/L) at a 3 percent discount rate, and from 0.06
(MCL=3 |ig/L) to 0.42 (MCL=20 |ig/L) at a 7 percent discount rate. Of the MCL options
examined, the net benefits are maximized at an MCL of 20 jig/L. The benefit cost ratio is also
greatest at an MCL of 20 jig/L.
Chapter 7, Comparison of Costs and Benefits 7-2 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 7-1
Summary of Annual National Net Benefits and Benefit-Cost Ratios,
CWSs Comply With MCL and NTNCs Only Monitor*
(Bladder Cancer Cases Only, in $ millions)
MCL (ug/L)
3
5
10
20
3% Discount Rate
lower bound
upper bound
Net Benefits
Benefit/Cost Ratio
Net Benefits
Benefit/Cost Ratio
$ (599.5)
0.07
$ (538.9)
0.16
$ (345.6)
0.08
$ (287.4)
0.24
$ (145.4)
0.11
$ (111.2)
0.32
$ (53.7)
0.13
$ (31.8)
0.48
7% Discount Rate
T3
C
3
O
.0
i_
0
1
upper bound
Net Benefits
Benefit/Cost Ratio
Net Benefits
Benefit/Cost Ratio
$ (707.8)
0.06
$ (647.2)
0.14
$ (408.7)
0.07
$ (350.5)
0.20
$ (172.6)
0.09
$ (138.4)
0.27
$ (63.9)
0.11
$ (42.0)
0.42
"Costs include treatment, O&M, monitoring, and administrative costs to CWSs, monitoring and
administrative costs to NTNCWSs, and State costs for administration of water programs.
The lower-end estimate of bladder cancer cases avoided is calculated using the lower-end risk estimate
(see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group
was 100 percent. The upper-end estimate of bladder cancer cases is calculated using the upper-end risk
estimate (see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese
study group was 80 percent.
Chapter 7, Comparison of Costs and Benefits
7-3
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 7-2 graphically depicts the absolute difference between the total value of national costs
and benefits under each proposed MCL at a 7 percent discount rate.
Exhibit 7-2
Comparison of Costs and Benefits of Bladder Cancer Cases Avoided
(CWSs Comply with MCL / NTNCs Monitor, 7% Discount Rate)
ccynn nn
c
o
<&
cconn nn
CM nn nn
CC
DCoste
D Benefits (upper bound)
D Benefits (lower bound)
$
$
$
ug/L
753.1
04.2
43.6
HI
6
0
3
,
$
5
5
, 1 f ,
Jug/L 10 ug/L 20 ug/L
442.21 $192.40 $73.72
89.90 $52.10 $29.80
31.70 $17.90 $7.90
Finally, Exhibit 7-3 shows the results of an analysis in which the average national cost of
achieving each unit reduction in cases of bladder cancer avoided, was calculated. The average
annual cost per cancer case avoided was computed at each MCL option, for both 3 and 7 percent
discount rates. At a 3 percent discount rate, the cost per bladder cancer case ranges from $12.3
million to $29.4 million at an MCL of 3 ug/L, from $8.4 million to $23.8 million at an MCL of 5
Hg/L, from $6.3 million to $18.3 million at an MCL of 10 ug/L, and from $4.2 million to $15.9
million at an MCL of 20 |ig/L. At a 7 percent discount rate, the cost per bladder cancer case
ranges from $14.4 million to $34.4 million at an MCL of 3 ug/L, from $9.8 million to $27.7
million at an MCL of 5 |ig/L, from $7.3 million to $21.4 million at an MCL of 10 |ig/L, and from
$4.9 million to $18.6 million at an MCL of 20
Chapter 7, Comparison of Costs and Benefits
7-4
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 7-3
Cost per Bladder Cancer Case Avoided for the Proposed Arsenic Rule
(CWSs Comply with MCL / NTNCs Monitor, in $ millions)
Arsenic Level
(jiQ/L)
lower bound**
upper bound**
3% Discount Rate
3
5
10
20
$
$
$
$
29.4
23.8
18.3
15.9
$
$
$
$
12.3
8.4
6.3
4.2
7% Discount Rate
3
5
10
20
$
$
$
$
34.4
27.7
21.4
18.6
$
$
$
$
14.4
9.8
7.3
4.9
*Costs all treatment, O&M, monitoring, and administrative costs to CWSs,
monitoring and administrative costs to NTNCWSs, and State costs for
administration of water programs.
"Lower/upper bounds correspond to estimates of bladder cancer cases avoided.
Chapter 7, Comparison of Costs and Benefits
7-5
Proposed Arsenic in Drinking Water Rule RIA
-------
7.3.2 Cost-Effectiveness
Cost-effectiveness analysis is another commonly used measure of the economic efficiency with
which regulatory options are meeting the intended regulatory objectives. Exhibit 7-4 is a
comparison of annual national costs (computed at a seven percent discount rate) and annual cases
of bladder cancer avoided at each MCL option. The two lines represent the cost per cancer case
avoided under the lower and upper bound estimates of bladder cancer cases avoided. These
plotted lines depict the trend in marginal cost and benefits (expressed as health effects avoided)
between each point on these curves (corresponding to each MCL option). Points along these
lines represent each increment of cost that is incurred in order to achieve the next increment of
risk reduction, i.e. additional bladder cancer case avoided. The steepness of the curves under
both benefits scenarios suggests that additional increments of risk reduction and benefits are
achieved at increasingly greater cost to the nation.
Chapter 7, Comparison of Costs and Benefits 7-6 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 7-4
Comparison of Annual Costs to Cases of Bladder Cancer per Year
(7% Discount Rate)
$800
$700
20 30 40
Bladder Cancer Cases Avoided per Year
50
60
based on lower bound estimates of avoided cases
-based on upper bound estimates of avoided cases
Exhibit 7-5 further reinforces the fact that as the MCL becomes more stringent, the incremental
cost per cancer case avoided increases. For example, the additional cases of bladder cancer
avoided in moving from an MCL of 10 jig/L to 5 jig/L are achieved at a cost per case of $13
million annually under the high bound and seven percent discount rate scenario. Similarly, in
moving from an MCL of 5 |ig/L to a more stringent MCL of 3 |ig/L, the cost per case avoided
increases to $44 million per year under this same scenario.
Chapter 7, Comparison of Costs and Benefits
7-7
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 7-5
Incremental Cost per Incremental Bladder Cancer Case Avoided
(CWSs, 7% Discount Rate, in $ millions)
£ I
10 >-
o
-------
benefits of the rule to an appreciable degree, even if the assumption were made that the risk of
skin cancer were equivalent to that of bladder cancer, using EPA's 1988 risk assessment. Skin
cancer is highly treatable (at a cost of illness of less than $3,500 for basal and squamous cell
carcinomas vs. a cost-of-illness of $178,000 for non-fatal bronchitis) in the U.S., with few
fatalities (less than 1%).
In addition to potentially reducing the risk of skin cancer, there are also a large number of other
health effects associated with arsenic, as presented in Exhibit 5-1, which are not monetized in this
analysis, due to lack of appropriate data.
Other benefits not monetized in this analysis include customer peace of mind from knowing
drinking water has been treated for arsenic and reduced treatment costs for currently unregulated
contaminants that may be co-treated with arsenic. To the extent that reverse osmosis is used for
arsenic removal, these benefits could be substantial. Reverse osmosis is the primary point of use
treatment, and it is expected that very small systems will use this treatment to a significant extent.
(These benefits of avoided treatment cannot currently be monetized; however, they can be readily
monetized in the future, as decisions are made about which currently unregulated contaminants to
regulate.)
Chapter 7, Comparison of Costs and Benefits 7-9 Proposed Arsenic in Drinking Water Rule RIA
-------
Chapters: Economic Impact Analyses
8.1 Introduction
EPA is required to perform a series of analyses that addresses the distribution of regulatory
impacts associated with the proposed Arsenic Rule. This chapter presents analyses that support
EPA's compliance with the following federal mandates:
Executive Order 12886 (Regulatory Planning and Review);
Regulatory Flexibility Act (RFA) of 1980, as amended by the Small Business Regulatory
Enforcement Fairness Act (SBREFA) of 1996;
National Affordability determination required by the 1996 amendments to the Safe
Drinking Water Act (SOWA);
Unfunded Mandates Reform Act (UMRA) of 1995;
Technical, Financial, and Managerial Capacity Assessment required by Section 1420(d)(3)
of the 1996 amendments to the Safe Drinking Water Act (SDWA);
Executive Order 13045 (Protection of Children From Environmental Health Risks and
Safety Risks);
Executive Order 12989 (Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations);
Paperwork Reduction Act; and
Health Risk Reduction and Cost Analysis (HRRCA) as required by Section 1412(b)(3)(C)
of the 1996 SDWA Amendments;
Initial Regulatory Flexibility Analysis (IRFA)
These analyses draw on the cost analyses presented in Chapter 6 and an analysis of administrative
requirements presented in a separate document, Information Collection Request for the Arsenic
Rule. Throughout this chapter, it is assumed that CWSs will have to comply with the newly
proposed MCL, while NTNCs will only have to monitor for Arsenic.
Several of these federal mandates require an explanation of why the rule is necessary, the
statutory authority upon which it is based, and the primary objectives it is intended to achieve.
Background information on the problems addressed by the proposed rule, and EPA's statutory
authority for promulgating the rule are presented in Chapter 2. In this chapter, Section 8.2
presents the RFA and SBREFA analysis of impacts on small entities. Also described are the
economic impacts of the proposed rule on households. Section 8.3 discusses coordination of the
arsenic rule with other Federal rules. The minimization of economic burden, UMRA, system
capacity assessments, and the Paperwork Reduction Act are addressed in Sections 8.4, 8.5, 8.6
and 8.7, respectively. Section 8.8 discusses the rules' protection of children's health, Section 8.9
addresses environmental justice issues and Section 8.10 contains the HRRCA.
Chapters, Economic Impact Analyses 8-1 Proposed Arsenic in Drinking Water Rule RIA
-------
8.2 Regulatory Flexibility Act and Small Business Regulatory Enforcement
Fairness Act
8.2.1 Summary of EPA's Small Business Consultations
Under the Regulatory Flexibility Act, 5 U.S.C. 601 et seq.. as amended by the Small Business
Regulatory Enforcement Fairness Act of 1996, EPA is required to prepare a regulatory flexibility
analysis unless the Agency certifies that a rule will not have "a significant economic impact on a
substantial number of small entities." If it is determined that the rule will have a significant impact
on a substantial number of small entities the Agency must convene a Small Business Advocacy
Review (SBAR) Panel prior to publication of the proposed rule. The SBAR Panel has 60 days to
consult with small entity representatives (SERs) likely to be impacted by the rule and to make
recommendations designed to reduce the impact of the proposed rule on small entities. The
Agency must consider these recommendations when drafting the proposed rule.
As required by section 609(b) of the RFA, as amended by SBREFA, EPA also conducted
outreach to small entities and convened a Small Business Advocacy Review Panel to obtain
advice and recommendations of representatives of the small entities that potentially would be
subject to the rule's requirements.
EPA identified 22 representatives of small entities, in this situation small systems, that were most
likely to be subject to the proposal. In December, 1998, EPA prepared and distributed to the
small entity representatives (SERs) an outreach document on the arsenic rule titled "Information
for Small Entity Representatives Regarding the Arsenic in Drinking Water Rule" (EPA 1998 ).
On December 18, 1998, EPA held a SER conference call for small systems from Washington D.C.
to provide a forum for input on key issues related to the planned proposal of the arsenic in
drinking water rule. These issues included, but were not limited to issues related to the rule
development, such as arsenic health risks, treatment technologies, analytical methods, and
monitoring. Fifteen SERs from small water systems participated on the call from the following
States: Alabama, Arizona, California, Georgia, Massachusetts, Montana, Nebraska, New
Hampshire, New Jersey, Utah, Virginia, Washington, and Wisconsin.
Efforts to identify and incorporate small entity concerns into this rulemaking culminated with the
convening of a SBAR Panel on March 30, 1999, pursuant to section 609 of RFA/SBREFA. The
four person Panel was headed by EPA's Small Business Advocacy Chairperson and included the
Director of the Standards and Risk Management Division within EPA's Office of Ground Water
and Drinking Water, the Administrator of the Office of Information and Regulatory Affairs with
the Office of Management and Budget, and the Chief Counsel for Advocacy of the SB A. For a
60-day period starting on the convening date, the Panel reviewed technical background
information related to this rulemaking, reviewed comments provided by the SERs, and met on
several occasions. The Panel also conducted its own outreach to the SERs and held a conference
call on April 21, 1999 with the SERs to identify issues and explore alternative approaches for
accomplishing environmental protection goals while minimizing impacts to small entities.
Consistent with the RFA/SBREFA requirements, the Panel evaluated the assembled materials and
small-entity comments on issues related to the elements of the IRFA (See section 8.2.1). A copy
Chapters, Economic Impact Analyses 8-2 Proposed Arsenic in Drinking Water Rule RIA
-------
of the June 4, 1999 Panel report is included in the docket for this proposed rule (U.S. EPA,
1999).
The proposed rule addresses all of the recommendations on which the Panel reached consensus.
In addition, to help small systems comply with the arsenic rule, EPA is committed to addressing
several other Panel recommendations regarding guidance, which are discussed in detail in the
pages to follow.
Treatment Technologies. Waste Disposal, and Cost Estimates
The Panel recommended the following: further develop the preliminary treatment and waste
disposal cost estimates; fully consider these costs when identifying affordable compliance
technologies for all system size categories; and provide information to small water systems on
possible options for complying with the MCL, in addition to installing any listed compliance
technologies.
In response to these recommendations the Treatment and Cost document describes: development
of cost estimates for treatment and waste disposal; identification of affordable compliance
technologies, including the consideration of cost; and options for complying with the MCL other
than installing compliance technologies, such as selecting to regionalize.
Regarding POU devices, the Panel recommended the following: continue to promote the use of
POU devices as alternative treatment options for very small systems where appropriate; account
for all costs, including costs that may not routinely be explicitly calculated; consider liability issues
from POU/POE devices when evaluating their appropriateness as compliance technologies; and
investigate waste disposal issues with POE devices.
In response to these recommendations, EPA will include in the proposed rule's preamble: an
expanded description regarding available POU compliance treatment technologies and conditions
under which POU treatment may be appropriate for very small systems; a description of the
components which contribute to the POU cost estimates; a discussion that clarifies that water
systems will be responsible for POU operation and maintenance to prevent liability issues from
customers maintaining equipment themselves.
Relevance of Other Drinking Water Regulations
The Panel recommended the following: include discussion of the co-occurrence of arsenic and
radon in the proposed rule for arsenic; take possible interactions among treatments for different
contaminants into account in costing compliance technologies and determining whether they are
nationally affordable for small systems; and encourage systems to be forward-looking and test for
the multiple contaminants to determine if and how they would be affected by the upcoming rules.
In response, the proposed rule's preamble will include a discussion on the co-occurrence analysis
of radon and arsenic: the treatment section of the preamble will be expanded to describe the
relationship of treatment for arsenic with other drinking water rules and how this issue was taken
Chapters, Economic Impact Analyses 8-3 Proposed Arsenic in Drinking Water Rule RIA
-------
into account in cost estimates. In addition, the preamble will encourage systems to consider other
upcoming rules when making future plans on monitoring or treatment.
Small Systems Variance Technologies and National Affordability Criteria
The Panel recommended the following: include a discussion of the issues surrounding appropriate
adjustment of its national affordability criteria to account for new regulatory requirements;
consider revising its approach to national affordability criteria, to the extent allowed by statutory
and regulatory requirements, to address the concern that the current cumulative approach for
adjusting the baseline household water bills is based on chronological order rather than risk; and
examine the data in the 1995 Community Water Supply Survey to determine if in-place treatment
baselines can be linked with the current annual water bill baseline in each of the size categories for
the proposed rule.
In response to these recommendation, the treatment section of proposed rule's preamble will
include an expanded discussion about the national affordability criteria and how it may be adjusted
to account for new regulations. In addition, information regarding methodology and rationale will
be added to explain the national affordability approach.
Monitoring and Arsenic Species
The Panel recommended the following: EPA consider allowing States to use recent compliance
monitoring data to satisfy initial sampling requirements or to obtain a waiver; and that EPA
continue to explore whether or not to make a regulatory distinction between organic and
inorganic arsenic based on compliance costs and other considerations.
In response, the monitoring section of the rule's preamble and the proposed regulatory language
will describe the allowance of monitoring data that meet analytical requirements and have
reporting limits sufficiently below the revised MCL and collected after 1990.
Considerations in setting the MCL
The Panel recommended the following: in performing its obligations under SDWA EPA should
take cognizance of the scientific findings, the large scientific uncertainties, the large potential
costs (including treatment and waste disposal costs), and the fact that this standard is scheduled
for review in the future; give full consideration to the provisions of the Executive Order 12866
and to the option of exercising the new statutory authority under SDWA Sections 1412(b)(4)(C)
and 1412(b)(6)(A) in the development of the arsenic rule; and fully consider all of the "risk
management" components of its rulemaking effort to ensure that the financial and other impacts
on small systems are factored into its decision-making processes. The Panel also recommended
that EPA take into account both quantifiable and non-quantifiable costs and benefits of the
standard and the needs of sensitive sub-populations.
In response to all these recommendations, EPA describes in detail the factors that were
considered in setting in the MCL and provides the rationale for this selection.
Chapters, Economic Impact Analyses 8-4 Proposed Arsenic in Drinking Water Rule RIA
-------
Applicability of proposal
The Panel recommended that EPA carefully consider the appropriateness of extending the scope
of the rule to Non-Transient, Non-Community Water Systems (NTNCWSs).
In response, the arsenic proposal does not apply to NTNCWSs and the MCL section of the rule's
preamble will describe the basis for this decision, including the incremental costs and benefits
attributable to coverage of these water systems.
Other Issues
The Panel recommended that EPA encourage small systems to discuss their infrastructure needs
for complying with the arsenic rule with their primacy agency to determine their eligibility for
DWSRF loans, and if eligible, to ask for assistance in applying for the loans. In response, the
UMRA analysis has been expanded to discuss funding options for small systems and to encourage
systems to be proactive in communicating with their primacy agency.
Regarding health effects, the Panel recommended the following: further evaluate the Utah study
and its relationship to the studies on which the NRC report was based and give it appropriate
weight in the risk assessment for the proposed arsenic standard; and examine the NRC
recommendations in the light of the uncertainties associated with the report's recommendations,
and any new data that may not have been considered in the NRC report. In response to these
recommendations, the benefits analysis includes a discussion of the qualitative benefits evaluation
and use of research data.
8.2.2 Definition of Small Entity for the Arsenic Rule
The Agency has proposed, taken comment, and finalized its intent to define "small entity" as a
public water system that serves 10,000 or fewer persons for purposes of its regulatory flexibility
assessments under the RFA for all future drinking water regulations. (See the Consumer
Confidence Reports Final Rule, 63 FR 44511, Aug. 19, 1998 and Proposed Rule, 63 FR 7620
Feb. 13, 1998.) The Agency discussed at length, in the preamble to the proposed Consumer
Confidence rule, the basis for its decision to use this definition and to use a single definition of
small public water system whether the system was a "small business", "small nonprofit
organization", or "small governmental jurisdiction." EPA also consulted with the Small Business
Administration on the use of this definition as it relates to small businesses. The Agency has used
this definition in developing subsequent regulations under the Safe Drinking Water Act. In
defining small entities in this manner, EPA recognizes that baseline conditions in source water and
treatment and operational practices may differ for systems serving fewer than 10,000 people
versus systems serving 10,000 or more persons.
According to the latest estimates (December 1998) contained in the EPA's public water system
database, Safe Drinking Water Information System (SDWIS), there are 54,352 community water
systems and 20,255 non-transient noncommunity water supplies providing potable water to the
public. Of these, 90 percent of the CWSs (50,776) and nearly all of the NTNCWSs (20,235) are
Chapters, Economic Impact Analyses 8-5 Proposed Arsenic in Drinking Water Rule RIA
-------
classified by EPA as small entities. Exhibit 8-1 presents a breakdown of the universe of small
water systems by size category, type, and ownership.
Exhibit 8-1
Profile of the Universe of Small Water Systems
Regulated Under the Arsenic Rule
Type Water
System
Publicly-Owned:
cws
NCWS
Privately-Owned:
CWS
NCWS
Total Systems:
CWS
NCWS
TOTAL
System Size Category
<100
1,729
1,783
13,640
8,178
15,369
9,961
25,330
101-500
5,795
3,171
1 1 ,266
4,162
17,061
7,333
24,394
501-1,000
3,785
1,182
2,124
902
5,909
2,084
7,993
1,001-3,300
6,179
361
1,955
411
8,134
772
8,906
3,301-10,000
3,649
29
654
56
4,303
85
4,388
Source: Safe Drinking Water Information System (SDWIS), December 1998 freeze.
Of the total number of small systems governed by the proposed rule, 72 percent of the systems are
CWSs and 28 percent are NTNCWSs. A table of the total number of systems regulated by the
proposed rule is provided in Chapter 4, "Baseline Analysis."
8.2.3 Requirements for the Initial Regulatory Flexibility Analysis
The Regulatory Flexibility Act requires EPA to complete an Initial Regulatory Flexibility Analysis
(IRFA) addressing the following:
The need for the rule;
The objectives 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;
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; and
Chapter 8, Economic Impact Analyses
8-6
Proposed Arsenic in Drinking Water Rule RIA
-------
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. Significant
regulatory alternatives may include:
- 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; and
- Exempting small entities from coverage of the rule or any part of the rule.
8.2.4 Small Entity Impacts
The results of the economic impact analysis for small water systems under the MCL options of 3,
5, 10, and 20 jig/L are summarized below. Estimates of the number of small systems expected to
be affected and the cost of complying with each component of the regulatory approach are
presented.
As seen in Exhibit 8-2, at an MCL of 5 |ig/L, approximately thirteen percent of community water
systems serving less than 100 people are expected to be affected by the Arsenic Rule. In general,
for the small systems size categories, the Arsenic Rule is expected to affect approximately twelve
percent of the universe of CWSs.
Table 8-2
Number of CWSs Expected to Undertake or Modify Treatment Practice
MCL 3
GW
SW
MCL 5
GW
SW
MCL 10
GW
SW
MCL 20
GW
SW
<100
3,024
62
1,898
32
874
10
343
3
101-500
3,256
109
2,048
57
934
18
377
5
501-1,000
1,058
67
671
34
312
11
126
3
1,001-
3,300
1,406
137
893
70
424
23
177
5
3,301-
10,000
696
94
444
50
218
16
91
5
10,001-
50,000
434
76
286
41
144
13
61
3
50,001-
100,000
53
16
35
8
19
3
8
1
100,001-
1,000,000
33
16
21
8
11
4
5
1
It is useful to put the number of systems affected by the rule in to context by comparing the
number of systems expected to be affected by the promulgation of the Arsenic Rule to the number
of systems not expected to undertake or modify any of their existing treatment practices. Exhibits
8-3, 8-4, 8-5, and 8-6 show these comparisons forMCLs of 3, 5, 10, 20 |ig/L..
Chapter 8, Economic Impact Analyses
8-7
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-3
Number of CWSs Expected to Undertake or Modify Treatment Practice
MCL 3 ug/L
15,000
Exhibit 8-4
Number of CWSs Expected to Undertake or Modify Treatment Practice
MCL 5 ug/L
w 10,500 -
E
&
OT 7*nn
O
1
1,500 -
DTreating GW systems
D Not treating GW systems
DTreating SW systems
DNot treating SW systems
r
r
jn _L
<100
1,898
12,379
32
1,060
I 1
I 1 I I I 1 I 1
J r- _n H L ^ m
101-500 501-1,000 1,001-3,300 3,301-10,000
2,048 671 893 444
13,010 4,018 4,821 2,015
57 34 70 50
1,946 1,186 2,350 1,794
Chapter 8, Economic Impact Analyses
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-5
Number of CWSs Expected to Undertake or Modify Treatment Practice
MCL 10 ug/L
15,000
Exhibit 8-6
Number of CWSs Expected to Undertake or Modify Treatment Practice
MCL 20 ug/L
(/)
~ 9,000 -
CO
o
k.
E
D Treating GW systems
D Not treating GW systems
D Treating SW systems
D Not treating SW systems
n
<100
343
13,934
3
1,089
n
n
n 1 In 1-1
r n J n
101-500 501-1,000 1,001-3,300 3,301-10,000
377 126 177 91
14,681 4,563 5,537 2,368
5355
1,998 1,217 2,415 1,839
Chapter 8, Economic Impact Analyses
8-9
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-7 details the average system-level annual compliance costs for small systems that are
expected to undertake or modify treatment. As one would expect, the costs, per treating system,
are slightly higher under the most stringent MCL.
Exhibit 8-7
Average Annual System Compliance Costs for CWSs*
MCL (ng/L)
3
Treatm ent**
Monitoring/
Administrative
5
Treatm ent**
Monitoring/
Administrative
10
Treatm ent**
Monitoring/
Administrative
20
Treatm ent**
Monitoring/
Administrative
<100
$ 7,523
$ 37
$ 7,414
$ 35
$ 7,316
$ 34
$ 7,124
$ 34
cws
101-500
$ 20,551
$ 37
$ 20,163
$ 35
$ 19,460
$ 34
$ 18,705
$ 33
Size Category
501-1,000 1,
$ 33,440 $
$ 34 $
$ 32,745 $
$ 32 $
$ 31,018 $
$ 30 $
$ 29,347 $
$ 29 $
001-3,300
58,153
36
54,633
33
50,892
30
46,472
28
3,301-
10,000
$ 131,121
$ 76
$ 120,330
$ 69
$ 109,215
$ 63
$ 95,928
$ 55
*CWS system costs were calculated using a commercial discount rate.
**Treatment cost only applies if system undertakes or modifies treatment for arsenic.
Of course, system treatment costs range within size each size catagory. Also, it is important to
look at both absolute compliance costs, as well as compliance costs relative to current operating
expenditures. Using data on total current expenses from the CWSS (1995), EPA developed a
Monte-Carlo simulation model to estimate the effect of increased small system expenditures as a
result of Arsenic Rule compliance. Exhibits 8-8 through 8-17 show the distribution of system
compliance costs, as well as the distribution of the percentage increase in total operating expenses
that will result from the proposed arsenic rule. A separate exhibit is provided for each small size
catagory and each proposed MCL.
Chapter 8, Economic Impact Analyses
8-10
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-8
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving < 100 People (MCL 3 |jg/L)
$90,000
1000.0%
-- 100.0%
o
>
v>
0)
-------
Exhibit 8-10
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving < 100 People (MCL 10 |jg/L)
at tfl
tfl OJ
re «
-------
Exhibit 8-12
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 101-500 People (MCL 3 |jg/L)
$1,200,000
$1,000,000 --
1000.0%
0>
Q.
X
LJJ
000 --
S $600,000 --
E
X
$400,000 --
$200,000 --
10.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Exhibit 8-13
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 101-500 People (MCL 5 |jg/L)
$1,200,000
$1,000,000 --
c
0)
Q.
X
UJ
01
$800,000 --
~ $600,000 --
$400,000 --
$200,000 --
1000.0%
100.0% £ 5
O Q.
0)
c
01
Q.
10.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Chapter 8, Economic Impact Analyses
8-13
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-14
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 101-500 People (MCL 10 |jg/L)
$1,200,000
$1,000,000 --
0)
in
£
Q.
X
UJ
$800,000 --
S $600,000 --
E $400,000 --
$200,000 --
1000.0%
-- 100.0%
a>
to
>*
CO
c
at w>
8> 01
ro «
c
01
u
01
a.
^ 10.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Exhibit 8-15
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 101-500 People (MCL 20 |jg/L)
$1,200,000
$1,000,000 --
1000.0%
$800,000 --
CO
E $400,000 --
$200,000 --
10.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Chapter 8, Economic Impact Analyses
S-7₯
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-16
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 501-1,000 People (MCL 3 |jg/L)
1000.0%
V
to
c
aJ «
--100.0% s;
U Q.
X
O)
re
'c
ID
u
-------
Exhibit 8-18
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 501-1,000 People (MCL 10 |jg/L)
$2,000,000
1000.0%
-- 100.0% E
u
10.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses
Baseline Expenses - - -Percent Increase in Expenses
.
X
"J
0)
c
01
Q.
Exhibit 8-19
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 501-1,000 People (MCL 20 |jg/L)
1000.0%
-- 100.0%
-- 10.0%
1.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
v
to
0)
-------
Exhibit 8-20
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 1,001-3,000 People (MCL 3 |jg/L)
$3,000,000
$2,500,000 --
V)
0>
V)
I
X
LJJ
$2,000,000 --
E $1,000,000 --
'
$500,000 --
1000.0%
1.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Exhibit 8-21
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 1,001-3,300 People (MCL 5 |jg/L)
$3,000,000
$2,500,000 --
01
ifl
£ $2,000,000 --
Q.
X
UJ
$1,500,000 --
E $1,000,000 --
$500,000 --
1000.0%
-- 100.0%
-- 10.0%
0) W>
-------
Exhibit 8-22
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 1,001-3,300 People (MCL 10 |jg/L)
$3,000,000
$2,500,000 --
1000.0%
X
LJJ
$2,000,000 --
E $1,000,000 --
'
$500,000 --
1.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Exhibit 8-23
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 1,001-3,000 People (MCL 20 |jg/L)
$3,000,000
1000.0%
v
to
-- 100.0%
-- 10.0%
c
oj in
re <2
I Q)
U Q.
= X
O)
re
'c
ID
u
0)
Q.
1.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Chapter 8, Economic Impact Analyses
8-18
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-24
Comparison of CWS Baseline and Post-Compliance Total Expenses
for Systems Serving 3,301-10,000 People (MCL 3 |jg/L)
$6,000,000
$5,000,000 --
V)
0>
V)
I
X
LJJ
$4,000,000 --
E $2,000,000 --
'
$1,000,000 --
$0
-- 100.0%
1000.0%
01
to
>
w
0)
W> d)
ra «
o Q.
O)
S
01
Q.
1.0%
10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 95.0% 97.5% 99.9%
Percentage of Systems
-After Rule Expenses Baseline Expenses - - -Percent Increase in Expenses
Chapter 8, Economic Impact Analyses
8-19
Proposed Arsenic in Drinking Water Rule RIA
-------
Chapters, Economic Impact Analyses 8-20 Proposed Arsenic in Drinking Water Rule RIA
-------
8.2.5 Small System Affordability
Section 1415(e)(l) of SDWA allows States to grant variances to small water systems (i.e.,
systems having fewer than 10,000 customers) in lieu of complying with an MCL if EPA
determines that there are no nationally affordable compliance technologies for that system
size/water quality combination. The system must then install an EPA-listed variance treatment
technology (§1412(b)(15)) that makes progress toward the MCL, if not necessarily reaching it.
To list variance technologies, three showings must be made:
1) EPA must determine, on a national level, that there are no compliance technologies that
are affordable for the given small system size category/source water quality combination.
2) If there is no nationally affordable compliance technology, then EPA must identify a
variance technology the may not reach the MCL but that will allow small systems to make
progress toward the MCL (it must achieve the maximum reduction affordable). This
technology must also be listed as a small systems variance technology by EPA in order for
small systems to be able to rely on it for regulatory purposes.
3) EPA must make a finding on a national level, that use of the variance technology would be
protective of public health.
States must then make a site-specific determination for each system as to whether or not the
system can afford to meet the MCL based on State-developed affordability criteria. If the State
determines that compliance is not affordable for the system, it may grant a variance, but it must
establish terms and conditions, as necessary, to ensure that the variance is adequately protective
of human health.
In the Agency's draft national-level affordability criteria published in the August 6, 1998 Federal
Register, EPA discussed the affordable treatment technology determinations for the contaminants
regulated before 1996. The national-level affordability criteria were derived as follows. First an
"affordability threshold" was calculated. The affordability threshold was based on the total annual
household water bill as a percentage of household income. In developing this threshold value,
EPA considered the percentage of median household income spent by an average household on
comparable goods and services such items as housing (28%), transportation (16%), food (12%),
energy and fuels (3.3%), telephone (1.9%), water and other public services (0.7%), entertainment
(4.4%) and alcohol and tobacco (1.5%).
Another of the key factors that EPA used to select an affordability threshold was cost
comparisons with other risk reduction activities for drinking water. Section 1412(b)(4)(E)(ii) of
the SDWA identifies both Point-of-Entry and Point-of-Use devices as options for compliance
technologies. EPA examined the projected costs of these options. EPA also investigated the
costs associated with supplying bottled water for drinking and cooking purposes. The median
income percentages that were associated with these risk reduction activities were: Point-Of-Entry
(> 2.5%), Point-of-Use (2%) and bottled water (> 2.5%).
Chapters, Economic Impact Analyses 8-21 Proposed Arsenic in Drinking Water Rule RIA
-------
Based on the foregoing analysis, EPA developed an affordability criteria of 2.5% of median
household income, or about $750, for the affordability threshold (EPA, 1998). The median water
bill for households in each small system category was then subtracted from this threshold to
determine the additional expenditure per household that was considered affordable for new
treatment. This difference is referred to as the "available expenditure margin." Based on EPA's
1995 Community Water System Survey, median water bills were about $250 per year for small
system customers. Thus, an average available expenditure margin of up to $500 per year per
household was considered affordable for the contaminants regulated before 1996. EPA next
identified treatment technologies for all pre-1996 contaminants with average per household costs
below $500 per year. Therefore it was not necessary to list any small system variance
technologies for existing contaminant rules.
Applying this criterion to the case of arsenic in drinking water, EPA has determined that
affordable technologies exist for all system size categories and has therefore not identified a
variance technology for any system size or source water combination at the proposed MCL. (See
Exhibit 8-28, Mean Annual Household Costs Across MCL Options by System Size.) In other
words, annual household costs after installation of the compliance technology are projected to be
below the available affordability threshold for all system size categories for the proposed MCLs.
EPA solicits comment on its determination in this case as well as its affordability criteria more
generally.
Exhibit 8-28
Mean Annual Costs to Households Served by CWSs, by Size Category
SIZE CATEGORIES
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
1,000,000 +
All categories
MCL (^g/L)
3
$368.13
$258.68
$106.07
$63.61
$44.28
$36.39
$29.52
$23.47
$2.70
$43.73
5
$363.65
$253.64
$103.74
$60.38
$40.77
$33.22
$26.64
$21.20
$1.73
$39.18
10
$357.17
$246.38
$98.35
$56.51
$37.04
$29.13
$22.80
$18.32
$0.89
$33.05
20
$348.72
$237.67
$93.25
$51.80
$32.52
$24.99
$19.44
$15.41
$0.55
$23.62
Chapter 8, Economic Impact Analyses
8-22
Proposed Arsenic in Drinking Water Rule RIA
-------
EPA recognizes that individual water systems may have higher than average treatment costs,
fewer than average households to absorb these costs, or lower than average incomes, but believes
that the affordability criteria should be based on characteristics of typical systems and should not
address situations where costs might be extremely high or low or excessively burdensome. EPA
believes that there are other mechanisms that may address these situations to a certain extent. In
any case, EPA believes that small system variances should be the exception and not the rule.
EPA expects the available expenditure margin to be lower than $500 per household per year for
the Arsenic Rule because some sources of data, for example the Current Population Survey,
indicate that water rates are currently increasing faster than median household income. Thus, the
"baseline" for annual water bills will rise as treatment is installed for compliance with regulations
promulgated after 1996, but before the Arsenic Rule is promulgated.
EPA notes, however, that high water costs are often associated with systems that have already
installed treatment to comply with a NPDWR. Such in-place treatment facilities may facilitate
compliance with future standards. EPA's approach to establishing the national-level affordability
criteria did not incorporate a baseline for in-place treatment technology. Assuming that systems
with high baseline water costs would need to install a new treatment technology to comply with a
NPDWR may thus overestimate the actual costs for some systems.
To investigate this issue, during the derivation of the national-level affordability criteria, EPA
examined a group of five small surface water systems with annual water bills above $500 per
household per year. All of these systems had installed disinfection and filtration technologies to
comply with the Surface Water Treatment Rule. If these systems were required to install
treatment to comply with the revised arsenic standard, modification of the existing processes
would be much more cost-effective than adding a new technology. As a result, because these
systems have already made the investment in treatment technology, and the cost is incorporated
into current annual household water bills, costs to the household may not increase substantially.
Installing new technologies may interfere with treatment in-place or require additional treatment
to address side effects which will increase costs over the arsenic treatment technology base costs.
(An example is corrosion control for lead and copper, which may need to be adjusted to
accommodate other treatment). While EPA tries to account for such interference in its cost
estimates for each new compliance technology, it is not possible to anticipate all the site specific
issues which may arise.
EPA believes that there is another mechanism in the SDWA to address cost impacts on small
systems composed primarily of low-income households. Systems that meet criteria established by
the State could be classified disadvantaged communities under Section 1452(d) of the SDWA.
They can receive additional subsidization under the Drinking Water State Revolving Fund
(DWSRF) program, including forgiveness of principal. Under DWSRF, States must provide a
minimum of 15% of the available funds for loans to small communities and have the option of
providing up to 30% of the grant to provide additional loan subsidies to the disadvantaged
systems, as defined by the State.
Chapters, Economic Impact Analyses 8-23 Proposed Arsenic in Drinking Water Rule RIA
-------
8.3 Coordination With Other Federal Rules
Several Federal drinking water rules are under development involving treatment requirements that
may relate to the treatment of arsenic for this drinking water rule. Although it is very difficult to
determine how compliance with the proposed Arsenic Rule might effect compliance with other
drinking water regulations, the following briefly describes each rule, the impact the Arsenic Rule
may have on that rule, and/or how each rule may impact the arsenic standard. The Arsenic Rule is
expected to be promulgated in a similar time frame as the Ground Water Rule, the Radon Rule,
and the Microbial and Disinfection By-Product Rule.
Ground Water Rule (GWR^)
The goals of the GWR are to: (1) provide a consistent level of public health protection; (2)
prevent waterborne microbial disease outbreaks; (3) reduce endemic waterborne disease; and (4)
prevent fecal contamination from reaching consumers. To assure public health protection, EPA
has the responsibility to develop a GWR which not only specifies the appropriate use of
disinfection, but also addresses other components of ground water systems. This general
provision is supplemented with an additional requirement that EPA develop regulations specifying
the use of disinfectants for ground water systems as necessary. To meet these requirements, EPA
is working with stakeholders to develop a GWR proposal by Fall 1999 and a final rule by Fall
2000.
The GWR will result in more systems using disinfection. If a system does add a disinfection
technology, it may contribute to arsenic pre-oxidation. This largely depends on the type of
disinfection technology employed. For example, if a system chooses a technology such as
ultraviolet radiation, it may not affect arsenic pre-oxidation. However, if it chooses chlorination,
it will contribute to arsenic pre-oxidation. Arsenic pre-oxidation from arsenic (III) to arsenic (V)
will enhance the removal efficiencies of the technologies. Another option is that systems may use
membrane filtration for the GWR. In that case, depending on the size of the membrane, some
arsenic removal can be achieved. Thus, the GWR is expected to alleviate some of the burden of
the Arsenic Rule.
Radon
Like the Ground Water Rule, the Radon Rule will also be finalized before the Arsenic Rule. In
the 1996 Amendments to the SDWA, Congress [Section 1412(b)(13)] directed EPA to propose
an MCLG and NPDWR for radon by August 1999 and finalize the regulation by August 2000
(§1412 (b)(13)). One option for compliance with the Radon Rule is that systems may employ
aeration. Aeration alone, however, will not likely be sufficient to oxidize arsenic (III) to arsenic
(V). However, if systems do aerate, they may be required by State regulations to also disinfect.
The disinfection process may oxidize the arsenic, depending on the type of disinfection employed.
In particular, ultraviolet disinfection may not assist in arsenic oxidation (still under investigation
by US EPA), whereas chemical disinfection or oxidation is likely to. Thus, the Radon Rule is
expected to alleviate some of the burden of the Arsenic Rule.
Chapters, Economic Impact Analyses 8-24 Proposed Arsenic in Drinking Water Rule RIA
-------
Microbial and Disinfection By-product Regulations
To control disinfection and disinfection byproducts and to strengthen control of microbial
pathogens in drinking water, EPA is developing a group of interrelated regulations, as required by
the SDWA. These regulations, referred to collectively as the Microbial Disinfection By-product
(M/DBP) Rules, are intended to address risk trade-offs between the two different types of
contaminants.
EPA proposed a Stage 1 Disinfectants/Disinfection By-products Rule (DBPR) and Interim
Enhanced Surface Water Treatment Rule (TESWTR) in July 1994. EPA issued the final Stage 1
DBPR and IESWTR in November, 1998.
The Agency has finalized and is currently implementing a third rule, the Information Collection
Rule, that will provide data to support development of subsequent M/DBP regulations. These
subsequent rules include a Stage 2 DBPR and a companion Long-Term 2 Enhanced Surface
Water Treatment Rule (LT2ESWTR).
Stage 1 DBPR and IESWTR will primarily affect large surface water systems, so EPA does not
expect much overlap with small systems treating for arsenic. Stage 2 DBPR and possibly the
LT2ESWTR, however, would have significance as far as arsenic removal is concerned. For
systems removing DBF precursors, systems may use nanofiltration. The use of nanofiltration
would also be relevant for removing arsenic, and as a result, would ease some burden when
systems implement these later rules.
8.4 Minimization of Economic Burden
The proposed Arsenic Rule includes several provisions that will insure that the economic burden
to water systems is minimized, while still ensuring that the public health objectives of the rule are
met. First, the rule is developed around the concept of a performance target known as the
maximum contaminant level (MCL). Rather than prescribe a single treatment technique that must
be installed in all water systems, EPA is only requiring those systems that currently provide
finished water with an arsenic concentration above the target to undertake or modify treatment.
As seen above, this will exclude the vast majority of systems from having to undertake any
additional treatment under this proposed rule. In addition, if a system does have to undertake or
modify treatment, EPA is allowing systems to choose from a broad list of technologies, and is
encouraging systems to choose the treatment technique that minimizes their total costs.
Second, EPA is allowing states to grant nine year monitoring waivers to those systems that have a
history of arsenic monitoring results below the proposed MCL, and that do not show a substantial
risk of future arsenic contamination. This provision of the rule will further reduce the cost to
systems that currently provide finished water with low arsenic concentrations.
Finally, EPA is allowing small systems with finished water concentrations above the proposed
MCL to install POU or POE technologies. This option will further allow small systems to
minimize their total cost of compliance with the proposed rule.
Chapters, Economic Impact Analyses 8-25 Proposed Arsenic in Drinking Water Rule RIA
-------
8.5 Unfunded Mandates Reform Act
Title II of the Unfunded Mandates Reform Act of 1995 (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 one year.
Before promulgating an EPA rule for which a written statement is needed, Section 205 of the
UMRA generally requires EPA to identify and consider a reasonable number of regulatory
alternatives and adopt the least costly, most cost effective or least burdensome alternative that
achieves the objectives of the rule. The provisions of Section 205 do not apply when they are
inconsistent with applicable law. Moreover, Section 205 allows EPA to adopt an alternative
other than the least costly, most cost effective or least burdensome alternative if the Administrator
publishes an explanation why the more "costly" alternative was preferred for the final rule.
Prior to establishing any regulatory requirements that may significantly or uniquely affect small
governments, including Tribal governments, EPA must develop a small government agency plan
under Section 203 of the UMRA . 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.
EPA has determined that this rule contains a Federal mandate that may result in expenditures of
$100 million or more for State, local, and Tribal governments, in the aggregate and the private
sector in any one year. Accordingly, under Section 202 of the UMRA, EPA is obligated to
prepare a written statement addressing:
1. The authorizing legislation;
2. Cost-benefit analysis including an analysis of the extent to which the costs of State, local
and Tribal governments will be paid for by the Federal government;
3. Estimates of future compliance costs and disproportionate budgetary effects;
4. Macro-economic effects;
5. 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; and
6. 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.
Chapters, Economic Impact Analyses 8-26 Proposed Arsenic in Drinking Water Rule RIA
-------
The legislative authority for the arsenic rule is discussed in Chapter 2. Items two through five are
addressed below, with the exception of future compliance costs, which are discussed in Chapter 6.
Regulatory alternatives, the last item, are addressed in Chapters 3, 6 and 7.
8.5.1 Social Costs and Benefits
Chapters Five, Six and Seven contain a detailed cost-benefit analysis in support of the Arsenic
Rule. At a 7 percent discount rate, the proposed rule is expected to have a total annualized cost
of $645 million for a MCL of 3 |ig/L, $445 million for a MCL 5 |ig/L, $195 million for a MCL of
10 |ig/L, and $63.9 million for a MCL of 20 |ig/L.
EPA estimates the proposed arsenic rule will have total health benefits as a result of avoided
bladder cancer cases of approximately $43.6 to $104.2 million if the MCL were set at 3 |ig/L,
$31.7 to $89.9 million if the MCL were to be set at 5 |ig/L, $17.9 to $52.1 million if the MCL
were set at 10 |ig/L, and $7.9 to $29.8 million if the MCL were set at 20 |ig/L. These monetized
health benefits of reducing arsenic exposures in drinking water are attributable to the reduced
incidence of fatal and non-fatal bladder cancer. Currently under baseline assumptions (no control
of arsenic exposure), there are annual fatal cancers and non-fatal cancers associated with arsenic
exposures through CWSs. At an arsenic MCL level of 3 |ig/L, an estimated 6 to 14 fatal cancers
and 16 to 39 non-fatal cancers per year are prevented; at a arsenic level of 5 jig/L, an estimated 4
to 12 fatal cancers and 12 to 33 non-fatal cancers per year are prevented; at 10 |ig/L, 2 to 7 fatal
and 7 to 19 non-fatal cancers per year are prevented; and at 20 jig/L, 1 to 4 fatal and 3 to 11 non-
fatal cancers per year are prevented.
EPA estimates that should avoided lung cancer cases be monetized as well, the potential benefits
from reducing lung cancer cases would range from $47.2 to $448.0 million at an MCL of 3 |ig/L,
$35.0 to $384.0 million at an MCL of 5 |ig/L, $19.6 to $224.0 million at an MCL of 10 |ig/L,
and $8.8 to $128.0 million at an MCL of 20 |ig/L. EPA estimates that the potential number of
lung cancer cases avoided ranges from 8 to 80 at an MCL of 3 |ig/L, from 6 to 68 at an MCL of 5
|ig/L, from 3 to 40 at an MCL of 10 |ig/L, and from 2 to 23 at an MCL of 20 |ig/L. A more
detailed discussion of the lung cancer risk and benefits calculation may be found in Chapter 5,
"Benefits Analysis."
In addition to quantifiable benefits, in Chapter 5, EPA has identified many potential non-
quantifiable benefits associated with reducing arsenic exposures in drinking water. These
potential benefits are not able to be quantified at this time, but may include reduced risk skin
cancer, and numerous non-cancerous health effects. In addition, certain non-health related benefits
may exists, such as ecological improvements and an increase in consumers' perception of drinking
water.
8.5.2 State Administrative Costs
Chapters, Economic Impact Analyses 8-27 Proposed Arsenic in Drinking Water Rule RIA
-------
States will incur a range of administrative costs in complying with the arsenic rule. Administrative
costs can include program management, inspections, and enforcement activities. EPA estimates
the total annual costs of State administrative activities for compliance with the MCL at a 7
percent discount rate are approximately $5.5 million for an MCL of 3 |ig/L, $5.0 million for an
MCL of 5 |ig/L, $4.6 million at an MCL of lOpg/L, and $4.4 million for an MCL of 20|ig/L.
Various Federal programs exist to provide financial assistance to State, local, and Tribal
governments in complying with this rule. The Federal government provides funding to States that
have a primary enforcement responsibility for their drinking water programs through the Public
Water Systems Supervision (PWSS) Grants program. 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, the 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 eligible public water systems. The DWSRF also assists public water
systems with financing the costs of infrastructure needed to achieve or maintain compliance with
SDWA requirements. Each State will have considerable flexibility to determine the design of its
program and to direct funding toward its most pressing compliance and public health protection
needs. States may also, on a matching basis, use up to ten percent of their DWSRF allotments for
each fiscal year to assist in running the State drinking water program.
Under PWSS Program Assistance Grants, the Administrator may make grants to States to carry
out public water system supervision programs. One State use of these funds is to develop
primacy programs. States may "contract" with other State agencies to assist in the development
or implementation of their primacy program. However, States may not use program assistance
grant funds to contract with regulated entities (i.e., water systems). PWSS Grants may be used
by States to set-up and administer a State program which includes such activities as: public
education, testing, training, technical assistance, developing and administering a remediation grant
and loan or incentive program (excludes the actual grant or loan funds), or other regulatory or
non-regulatory measures.
8.5.3 Future Compliance Costs and Disproportionate Budgetary Effects
To meet the requirement in Section 202 of the UMRA, EPA analyzed future compliance costs
and possible disproportionate budgetary effects of the MCL options. The Agency believes that
the cost estimates, shown in Exhibit 8-7 and discussed in more detail in Chapter Six, accurately
characterize future compliance costs of the proposed rule.
With regard to the disproportionate impacts, EPA considered available data sources in analyzing
the disproportionate impacts upon geographic or social segments of the nation or industry. No
rationale for disproportionate impacts based on geographic area were identified. To the extent
that there may be disproportionate impacts to low-income or other segments of the population,
EPA will prepare a small entity compliance guide, a monitoring/analytical manual, and a small
systems technology manual that will assist the public and private sector. To fully consider the
potential disproportionate impacts of this proposed rule, this analysis also developed three other
measures:
Chapters, Economic Impact Analyses 8-28 Proposed Arsenic in Drinking Water Rule RIA
-------
(1) reviewing the impacts on small versus large systems;
(2) reviewing the costs to public versus private water systems; and
(3) reviewing the household costs for the proposed rule.
The first measure, the national impacts on small versus large systems, is shown in Exhibit 8-29.
Small systems are defined as those systems serving 10,000 people or less and large systems are
those systems that serve more than 10,000 people.
The second measure of disproportionate impacts evaluated is the relative total costs to public
versus private water systems, by size. Exhibit 8-29 also presents the annual system level costs for
public and private systems by system size category for 3 |ig/L, 5 |ig/L, 10 |ig/L, and 20 |ig/L.
The costs are comparable for public and private systems across system sizes for all options. For
example, for systems serving less than 100 people at the 5 jig/L MCL public system treatment
costs are $9 thousand and private system treatment costs are $7 thousand. This pattern may be
due in large part to the limited number of treatment options, resulting in similar treatment choices
by both public or private systems.
Chapters, Economic Impact Analyses 8-29 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-29
Average Annual Cost per CWS by Ownership
System Size
Treatment and Monitoring Costs
Public
Private
Total Cost
All Systems
MCL = 3 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 9,475
$ 25,228
$ 34,688
$ 60,929
$ 135,573
$ 578,591
$ 3,885,713
$ 7,354
$ 18,570
$ 31,646
$ 51,097
$ 111,396
$ 547,969
$ 7,559
$ 20,588
$ 33,474
$ 58,189
$ 131,197
$ 573,423
$ 3,885,713
MCL = 5 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 9,720
$ 24,560
$ 34,124
$ 57,277
$ 124,552
$ 518,647
$ 2,669,474
$ 7,212
$ 18,223
$ 30,697
$ 48,198
$ 102,005
$ 459,930
$ 7,450
$ 20,198
$ 32,778
$ 54,666
$ 120,399
$ 508,640
$ 2,669,474
MCL = 10 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 9,453
$ 23,584
$ 32,271
$ 53,357
$ 113,338
$ 458,340
$ 1,395,498
$ 7,135
$ 17,675
$ 29,160
$ 44,785
$ 91,244
$ 415,520
$ 7,350
$ 19,551
$ 31,048
$ 50,921
$ 109,278
$ 450,835
$ 1,395,498
MCL = 20 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 9,121
$ 22,778
$ 30,493
$ 48,399
$ 99,872
$ 394,742
$ 921,121
$ 6,950
$ 16,954
$ 27,668
$ 41,625
$ 79,128
$ 334,737
-
$ 7,157
$ 18,738
$ 29,376
$ 46,501
$ 95,983
$ 384,868
$ 921,121
"Costs were calculated at a commercial interest rate and include system
treatment, monitoring, and administrative costs; note that systems serving over
1 million people are public surface water systems.
Chapter 8, Economic Impact Analyses
8-30
Proposed Arsenic in Drinking Water Rule RIA
-------
The third measure, household costs, can also be used to gauge the impact of a regulation and to
determine whether there are disproportionately higher impacts in particular segments of the
population. A detailed analysis of household cost impacts by system size is presented in Chapter
6. The costs for households served by public and private water systems are presented in Exhibit
8-30. As expected, cost per household increases as system size decreases. Cost per household is
higher for households served by smaller systems than larger systems for two reasons. First,
smaller systems produce less water than large systems and are therefore unable to utilize
economies of scale. Consequently, each household must bear a greater percentage share of the
system's costs.
Table 8-30 presents the costs per household for systems exceeding the MCL. For each size
category there is a moderate difference in annual cost per household for 3 |ig/L, 5 |ig/L, 10 |ig/L
and 20 jig/L across source and ownership. In general, costs per household are higher for private
systems than for public systems. This difference could be attributable to a discrepancy in the cost
of capital for public versus private entities. For public systems, the cost per household ranges
from approximately $25 to $342 per year at 5 |ig/L and from approximately $22 to $329 per year
at 10 |ig/L (excluding systems serving greater than 1 million people). For private systems, the
ranges are $22 to $369 per year, and $19 to $363 per year, respectively. The range of costs for 3
|ig/L and 20 |ig/L is similar.
To further evaluate the impacts of these household costs, the average costs per household were
compared to median household income data for each system-size category. The result of this
calculation, presented in Exhibit 8-31 for public and private systems, indicate a household's likely
share of incremental costs in terms of its household income. For all system sizes and MCLs
average household costs as a percentage of median household income are less than one percent.
Chapters, Economic Impact Analyses 8-31 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-30
Annual Compliance Costs per Household for
CWSs Exceeding MCLs
System Size
Groundwater
Public
Private
Surface Water
Public
Private
MCL = 3 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 338.44
$ 218.59
$ 108.63
$ 62.17
$ 44.67
$ 31 .29
$ 374.86
$ 285.61
$ 112.60
$ 83.24
$ 62.96
$ 31 .29
$ 328.94
$ 135.98
$ 45.44
$ 21.13
$ 18.34
$ 26.49
$ 2.70
$ 335.61
$ 183.96
$ 46.72
$ 27.91
$ 22.94
$ 22.81
MCL = 5 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 341 .78
$ 213.11
$ 106.00
$ 58.31
$ 40.60
$ 28.12
$ 369.21
$ 280.76
$ 108.40
$ 77.54
$ 57.25
$ 28.63
$ 323.48
$ 135.22
$ 44.86
$ 20.07
$ 16.89
$ 24.73
$ 1.73
$ 330.05
$ 182.65
$ 46.35
$ 26.57
$ 21 .54
$ 21.91
MCL = 10ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 329.17
$ 203.40
$ 99.45
$ 53.70
$ 36.30
$ 24.09
$ 363.09
$ 273.04
$ 102.19
$ 71.97
$ 50.41
$ 24.47
$ 317.80
$ 132.74
$ 42.98
$ 18.62
$ 14.68
$ 22.03
$ 0.89
$ 325.64
$ 180.88
$ 44.48
$ 25.49
$ 18.55
$ 19.06
MCL = 20 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
$ 320.13
$ 195.99
$ 93.27
$ 48.03
$ 31.38
$ 20.27
-
$ 352.42
$ 262.01
$ 96.63
$ 66.12
$ 44.14
$ 20.39
-
$ 310.11
$ 132.68
$ 42.26
$ 18.20
$ 13.35
$ 19.96
$ 0.55
$ 324.84
$ 179.93
$ 44.04
$ 24.87
$ 17.53
$
-
"Costs to households were calculated at a commercial interest rate and include system
treatment, monitoring, and administrative costs; note that systems serving over 1 million
people are public surface water systems.
Chapter 8, Economic Impact Analyses
8-32
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-31
Annual Compliance Costs per Household for CWSs Exceeding MCLs,
as a Percent of Median Household Income
System Size
Groundwater
Public
Private
Surface Water
Public
Private
MCL = 3 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
0.85%
0.55%
0.27%
0.16%
0.11%
0.08%
0.95%
0.72%
0.28%
0.21%
0.16%
0.08%
0.83%
0.34%
0.11%
0.05%
0.05%
0.07%
0.01%
0.85%
0.46%
0.12%
0.07%
0.06%
0.06%
MCL = 5 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
0.86%
0.54%
0.27%
0.15%
0.10%
0.07%
0.93%
0.71%
0.27%
0.20%
0.14%
0.07%
0.82%
0.34%
0.11%
0.05%
0.04%
0.06%
0.00%
0.83%
0.46%
0.12%
0.07%
0.05%
0.06%
MCL= 10 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
0.83%
0.51%
0.25%
0.14%
0.09%
0.06%
0.92%
0.69%
0.26%
0.18%
0.13%
0.06%
0.80%
0.33%
0.11%
0.05%
0.04%
0.06%
0.00%
0.82%
0.46%
0.11%
0.06%
0.05%
0.05%
MCL = 20 ng/L
<100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-1,000,000
1,000,000 +
0.81%
0.49%
0.24%
0.12%
0.08%
0.05%
-
0.89%
0.66%
0.24%
0.17%
0.11%
0.05%
-
0.78%
0.33%
0.11%
0.05%
0.03%
0.05%
0.00%
0.82%
0.45%
0.11%
0.06%
0.04%
0.00%
-
"Costs to household were calculated at a commercial interest rate and include system
treatment, monitoring, and administrative costs; median household income in May 1999
was $39,648 updated from the 1998 annual median household income from the Census
Chapter 8, Economic Impact Analyses
8-33
Proposed Arsenic in Drinking Water Rule RIA
-------
8.5.4 Macroeconomic Effects
As required under UMRA Section 202, EPA is required to estimate the potential macro-economic
effects of the regulation. These include effects on productivity, economic growth, full
employment, creation of productive jobs, and international competitiveness. Macro-economic
effects tend to be measurable in nationwide econometric models only if the economic impact of
the regulation reaches 0.25 percent to 0.5 percent of Gross Domestic Product (GDP). In 1998,
real GDP was $7,552 billion so a rule would have to cost at least $18 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
macro-economic effects on the national economy from the arsenic rule should be negligible based
on the fact that, assuming 100 percent compliance with an MCL, the total annual costs are
approximately $750 million at the 3 |ig/L level, $440 million at the 5 |ig/L level, $190 million at
the 10 |ig/L level, and $73 million at the 20 jig/L level (at a 7 percent discount rate). In addition,
the costs are not expected to be highly focused on a particular geographic region or industry
sector.
8.5.5 Consultation with State, Local, and Tribal Government
Under UMRA section 204, 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 rulemaking through various means. This included five stakeholder
meetings announced in the Federal Register and open to any one interested in attending in person
or by phone, and presentations at meetings of the American Water Works Association (AWWA),
the Association of State Drinking Water Administrators (ASDWA), the Association of California
Water Agencies (ACWA), and the Association of Metropolitan Water Agencies (AMWA).
Participants in EPA's stakeholder meetings also included representatives from the National Rural
Water Association, AMWA, ASDWA, AWWA, ACWA, Rural Community Assistance Program,
State departments of environmental protection, State health departments, State drinking water
programs, and a Tribe. EPA also made presentations at Tribal meetings in Nevada, Alaska, and
California.
To address the proposed rule's impact on small entities, the Agency consulted with
representatives of small water systems and convened a Small Business Advocacy Review Panel in
accordance with the Regulatory Flexibility Act (RFA) as amended by the Small Business
Regulatory Enforcement Fairness Act (SBREFA). Two of the small entity representatives were
elected officials from local governments. EPA also invited State drinking water program
representatives to participate in a number of workgroup meetings. In addition to these
consultations, EPA participated in and gave presentations at AWWA's Technical Workgroup for
Arsenic. State public health department and drinking water program representatives, drinking
water districts, and ASDWA participated in the Technical Workgroup meetings. A summary of
State, local, and Tribal government concerns on this proposed rulemaking is shown in the next
section.
Chapters, Economic Impact Analyses 8-34 Proposed Arsenic in Drinking Water Rule RIA
-------
In order to inform and involve Tribal governments in the rulemaking process, EPA staff attended
the 16th Annual Consumer Conference of the National Indian Health Board on October 6-8, 1998
in Anchorage, Alaska. Over nine hundred attendees representing Tribes from across the country
were in attendance. During the conference, EPA conducted two workshops for meeting
participants. The objectives of the workshops were to present an overview of EPA's drinking
water program, solicit comments on key issues of potential interest in upcoming drinking water
regulations, and to solicit advice in identifying an effective consultative process with Tribes for the
future.
EPA, in conjunction with the Inter Tribal Council of Arizona (ITCA), also convened a Tribal
consultation meeting on February 24-25, 1999, in Las Vegas, Nevada to discuss ways to involve
Tribal representatives, both Tribal council members and tribal water utility operators, in the
stakeholder process. Approximately twenty-five representatives from a diverse group of Tribes
attended the two-day meeting. Meeting participants included representatives from the following
Tribes: Cherokee Nation, Nezperce Tribe, Jicarilla Apache Tribe, Blackfeet Tribe, Seminole
Tribe of Florida, Hopi Tribe, Cheyenne River Sioux Tribe, Menominee Indian Tribe, Tulalip
Tribes, Mississippi Band of Choctaw Indians, Narragansett Indian Tribe, and Yakama Nation.
The major meeting objectives were to:
(1) identify key issues of concern to Tribal representatives;
(2) solicit input on issues concerning current OGWDW regulatory efforts;
(3) solicit input and information that should be included in support of future drinking
water regulations; and
(4) provide an effective format for Tribal involvement in EPA's regulatory
development process.
EPA staff also provided an overview on the forthcoming arsenic rule at the meeting. The
presentation included the health concerns associated with arsenic, EPA's current position on
arsenic in drinking water, the definition of an MCL, an explanation of the difference between
point-of-use and point-of-entry treatment devices, and specific issues for Tribes. The following
questions were posed to the Tribal representatives to begin discussion on arsenic in drinking
water:
(1) What are the current arsenic levels in your water systems?
(2) What are Tribal water systems affordability issues in regard to arsenic?
(3) Does your Tribe use well water, river water or lake water?
(4) Does your Tribe purchase water from another drinking water utility?
The summary for the February 24-25, 1999 meeting was sent to all 565 Federally recognized
Tribes in the United States.
EPA also conducted a series of workshops at the Annual Conference of the National Tribal
Environmental Council which was held on May 18-20, 1999 in Eureka, California.
Representatives from over 50 Tribes attended all, or part, of these sessions. The objectives of the
workshops were to provide an overview of forthcoming EPA regulations affecting water systems;
discuss changes to operator certification requirements; discuss funding for Tribal water systems;
Chapters, Economic Impact Analyses 8-35 Proposed Arsenic in Drinking Water Rule RIA
-------
and to discuss innovative approaches to regulatory cost reduction. Meeting summaries for EPA's
Tribal consultations are available in the public docket for this proposed rulemaking.
8.5.6 State, Local, and Tribal Government Concerns
State and local governments raised several concerns, including the high costs of the rule to small
systems; the burden of revising the State primacy program; the high degree of uncertainty
associated with the benefits; and the high costs of including Non-Transient Non-Community
Water Systems (NTNCWSs). EPA modified the revision of State primacy in order to decrease
the burden of the new arsenic regulation in response to State concerns, to minimize paperwork
and documentation of existing programs that would manage the arsenic regulation.
Tribal representatives were generally supportive of regulations which would ensure a high level of
water quality, but raised concerns over funding for regulations. With regard to the forthcoming
proposed arsenic rule, many Tribal representatives saw the health benefits as highly desirable, but
felt that unless additional funds were made available, implementing the regulation would be
difficult for many Tribes.
EPA understands the State, local, and Tribal government concerns with the above issues. The
Agency believes the options for small systems, proposed for public comment in this rulemaking,
will address stakeholder concerns pertaining to small systems and will help to reduce the financial
burden to these systems.
8.5.7 Regulatory Alternatives Considered
As required under Section 205 of the UMRA, EPA considered several regulatory alternatives in
developing an MCL for arsenic in drinking water. In preparation for this consideration, this
Regulatory Impact Analysis evaluated arsenic levels of 3 |ig/L, 5 |ig/L, 10 |ig/L, and 20 |ig/L.
Also evaluated were two scenarios for NTNCWSs: NTNCWSs treat and monitor, NTNCWSs
only monitor and do not treat.
This analysis also evaluated national costs and benefits of States choosing to reduce arsenic
exposure in drinking water. EPA believes that the regulatory approaches proposed for arsenic in
today's notice are the most cost-effective options that achieve the objectives of the rule and
provide the highest degree of public health protection.
8.5.8 Impacts on Small Governments
In developing this rule, EPA consulted with small governments pursuant to section 203 of the
UMRA to address impacts of regulatory requirements in the rule that might significantly or
uniquely affect small governments. In preparation for the proposed Arsenic Rule, EPA conducted
analysis on small government impacts and included small government officials or their designated
representatives in the rule making process. EPA conducted stakeholder meetings on the
development of the arsenic rule which gave a variety of stakeholders, including small
governments, the opportunity for timely and meaningful participation in the regulatory
development process. Groups such as the National Association of Towns and Townships, the
Chapters, Economic Impact Analyses 8-36 Proposed Arsenic in Drinking Water Rule RIA
-------
National League of Cities, and the National Association of Counties participated in the proposed
rulemaking process. Through such participation and exchange, EPA notified potentially affected
small governments of requirements under consideration and provided officials of affected small
governments with an opportunity to have meaningful and timely input into the development of the
regulatory proposal.
In addition, EPA will educate, inform, and advise small systems, including those run by small
governments, about the arsenic rule requirements. One of the most important components of this
process is the Small Entity Compliance Guide, required by the Small Business Regulatory
Enforcement Fairness Act of 1996 after the rule is promulgated. This plain-English guide will
explain what actions a small entity must take to comply with the rule. Also, the Agency is
developing fact sheets that concisely describe various aspects and requirements of the Arsenic
Rule.
8.6 Effect of Compliance With the Arsenic Rule on the Technical, Financial,
and Managerial Capacity of Public Water Systems
Section 1420(d)(3) of the SDWA as amended requires that, in promulgating a 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 has
been performed to fulfill this statutory obligation.
Overall water system capacity is defined in EPA guidance (EPA 816-R-98-006) (EPA 1998) 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. A water system's technical
capacity can be determined by examining key issues and questions, 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?
Chapters, Economic Impact Analyses 8-37 Proposed Arsenic in Drinking Water Rule RIA
-------
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 refers
to the system's institutional and administrative capabilities. 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 operator(s) 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?
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?
Generally, water systems are not expected to require significantly increased technical, financial, or
managerial capacity to comply with these new requirements.
8.7 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 three 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.
EPA is required to estimate the burden on PWS for complying with the final rule. Burden means
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
Chapters, Economic Impact Analyses 8-38 Proposed Arsenic in Drinking Water Rule RIA
-------
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. The Information Collection Rule for the Proposed Arsenic
Rule estimated a total burden of 3.33 million hours for 3 |ig/L, 3.23 million hours for 5 |ig/L, 3.15
million hours for 10 |ig/L, and 3.11 for 20 |ig/L.
8.8 Protecting Children From Environmental Health Risks and Safety Risks
Executive Order (EO) 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 E.O. 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.
As described in Chapter 5 ("Benefits Analysis"), there is insufficient toxicological data to
distinguish morbidity and mortality differences by age groups. No studies were located by
ATSDR (1998) that focused exclusively on evaluating unusual susceptibility to arsenic. However,
some members of the population are likely to be especially susceptible. For example, Chapter 5
describes several non-carcinogenic effects that may be of greater concern to children than adults,
such as cardiovascular or reproductive effects. Similarly, arsenic has been suggested to pose
significant problems in fetal development. This increased susceptibility may be due to a variety of
factors. These factors include increased dose (intake per unit of body weight) in children, genetic
predispositions, and dietary insufficiency (ATSDR, 1998), as well as pre-existing health
conditions.
Because children have increased fluid and food intake in relation to their body weight (NAS,
1995), their dose (milligrams per kilogram of body weight per day - mg/kg/day) of arsenic will be,
on average, greater than that of adults. For example, an intake of 1.2 liters per day in a 70 kg
adult yields an overall water intake of 0.017 liters per kg of body weight. An infant who
consumes 1 liter per day and weighs 10 kg is consuming 0.1 liter per kg of body weight, which is
more than 5 times the water intake per kg of an adult. Any contaminant which is present in the
water will be delivered at a correspondingly higher level, on a daily basis. Foy et al. noted that in
studies of chronic exposure, children appear to be more severely affected, probably due to a
higher exposure per body weight (1992 citation, reported in ATSDR, 1998). ATSDR (1998)
identified a need for additional data on the exposure of children in their arsenic analysis.
8.9 Environmental Justice
Executive Order 12898 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
Chapters, Economic Impact Analyses 8-39 Proposed Arsenic in Drinking Water Rule RIA
-------
and low-income populations. The Executive Order requires the Agency to consider environmental
justice issues in the rulemaking and to consult with Environmental Justice (EJ) stakeholders.
The Agency has considered environmental justice related issues concerning the potential impacts
of this regulation and has determined that there are no substantial disproportionate effects.
Because the arsenic rule applies to all community water systems, the majority of the population,
including minority and low-income populations will benefit from the additional health protection.
8.10 Health Risk Reduction and Cost Analysis
Section 1412(b)(3)(C) of the 1996 Amendments requires EPA to prepare a Health Risk
Reduction and Cost Analysis (HRRCA) in support of any NPDWR that includes an MCL.
According to these requirements, EPA must analyze each of the following when proposing a
NPDWR that includes an MCL:
1. quantifiable and non-quantifiable health risk reduction benefits for which there is a factual
basis in the rulemaking record to conclude that such benefits are likely to occur as the
result of treatment to comply with each level;
2. quantifiable and non-quantifiable health risk reduction benefits for which there is a factual
basis in the rulemaking record to conclude that such benefits are likely to occur from
reductions in co-occurring contaminants that may be attributed solely to compliance with
the MCL, excluding benefits resulting from compliance with other proposed or
promulgated regulations;
3. quantifiable and non-quantifiable costs for which there is a factual basis in the rulemaking
record to conclude that such costs are likely to occur solely as a result of compliance with
the MCL, including monitoring, treatment, and other costs, and excluding costs resulting
from compliance with other proposed or promulgated regulations;
4. the incremental costs and benefits associated with each alternative MCL considered;
5. the effects of the contaminant on the general population and on groups within the general
population, such as infants, children, pregnant women, the elderly, individuals with a
history of serious illness, or other sub-populations that are identified as likely to be at
greater risk of adverse health effects due to exposure to contaminants in drinking water
than the general population;
6. any increased health risk that may occur as the result of compliance, including risks
associated with co-occurring contaminants; and
7. other relevant factors, including the quality and extent of the information, the uncertainties
in the analysis, and factors with respect to the degree and nature of the risk.
This analysis summarizes EPA's estimates of the costs and benefits associated with various
arsenic levels. The summary tables below characterize aggregate costs and benefits, impacts on
Chapters, Economic Impact Analyses 8-40 Proposed Arsenic in Drinking Water Rule RIA
-------
affected entities, and tradeoffs between risk reduction and compliance costs. This analysis also
summarizes the effects of arsenic on the general population as well as any sensitive sub-
populations and provides a discussion on the uncertainties in the analysis.
8.10.1 Quantifiable and Non-Quantifiable Health Risk Reduction Benefits
Arsenic ingestion has been linked to a multitude of health effects, both cancerous and non-
cancerous. These health effects include cancer of the bladder, lungs, skin, kidney, nasal passages,
liver, and prostate. Arsenic ingestion has also been attributed to cardiovascular, pulmonary,
immunological, neurological, endocrine, and reproductive and developmental effects. A complete
list of the arsenic related health effects reported in humans is shown in Chapter 5. Of all the
health effects noted above, current research on arsenic exposure has only been able to define
scientifically defensible risks for bladder cancer. That is, EPA has adequate data to perform a risk
assessment on bladder cancer. Because there is currently a lack of strong evidence on the risks of
other arsenic-related health effects, the Agency has based its assessment of the quantifiable health
risk reduction benefits on the risks of arsenic induced bladder cancers. Avoided cases of lung
cancer presented in Exhibit 8-34 were estimated based on a comparison of lung cancer incidence
to bladder cancer indicidence.
The quantifiable health benefits of reducing arsenic exposures in drinking water are attributable to
the reduced number of fatal and non-fatal bladder cancers. Exhibit 8-32 shows a range of mean
bladder cancer risks for exposed populations at or above arsenic levels of 3, 5, 10, and 20 jig/L in
CWSs. Exhibit 8-33 shows the corresponding health risk reductions (number of total bladder
cancers avoided and the proportions of fatal and non-fatal bladder cancers avoided) at 3, 5, 10,
and 20 ug/L. These ranges of total, fatal, and non-fatal bladder cancer cases are based on the
range of risks shown in Exhibit 8-32.
Exhibit 8-32
Mean Bladder Cancer Risks, Exposed Population,
and Annual Cancer Cases Avoided in CWSs1
Arsenic Level
(M9/L)
3
5
10
20
Mean Exposed
Population Risk
2.1 -4.5x10-5
3.6-7.5x10-5
5.5-11.4x10-5
6.9- 13.9x10'5
Population
Exposed
34,599,915
21,347,435
8,530,370
3,579,085
Total Bladder Cancer
Cases Avoided per Year
22-52
16-45
9-26
4- 15
1The bladder cancer risks presented in this table provide our "best" estimates at this time. Actual risks could be
lower, given the various uncertainties discussed, or higher, as these estimates assume that the probability of illness
from arsenic exposure in the U.S. is equal to the probability of death from arsenic exposure among the Taiwanese
study group.
Chapter 8, Economic Impact Analyses
8-41
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-33
Annual Bladder Cancer Cases Avoided from Reducing Arsenic in CWSs1
Arsenic Level
(ug/L)
3
5
10
20
Total Bladder Cancer
Cases Avoided
22-52
16-45
9-26
4-15
Reduced Mortality
Cases*
6- 14
4-12
2-7
1 -4
Reduced Morbidity
Cases*
16-39
12-33
7- 19
3-11
* The lower-end estimate of bladder cancer cases avoided is calculated using the lower-end risk estimate (see
Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group was 100
percent. The upper-end estimate of bladder cancer cases avoided is calculated using the upper-end risk estimate
(see Exhibit 5-9) and assumes that the conditional probability of mortality among the Taiwanese study group was
80 percent.
**Assuming 20-year mortality rate in the U.S. of 26 percent.
Health benefit estimates for lung cancer were also quantified based on the "what if?" scenario,
where the risks of a fatal lung cancer case associated with arsenic are assumed to be 2-5 times
that of a fatal bladder cancer case. More detail on lung cancer health benefits is provided in
Chapter 5, "Benefits Analysis." Exhibit 8-34 shows the resulting annual lung cancer cases
avoided as a result of the "what if?" scenario analysis.
Exhibit 8-34
Potential Annual Lung Cancer Cases Avoided from Reducing Arsenic in CWSs
Arsenic Level
(ug/L)
3
5
10
20
Total Lung Cancer
Cases Avoided
8-80
6-68
3-40
2-23
Reduced Mortality
Cases*
7-70
5-60
3-35
1 -20
Reduced Morbidity
Cases*
1 - 10
1 -8
0-5
0-3
The Agency has developed monetized estimates of the health benefits associated with the risk
reductions from arsenic exposures. The approach used in this analysis for the measurement of
health risk reduction benefits is the monetary value of a statistical life (VSL) applied to each fatal
cancer avoided. For non-fatal cancers, willingness to pay (WTP) data to avoid chronic bronchitis
is used as a surrogate to estimate the WTP to avoid non-fatal bladder cancers. A WTP central
tendency estimate of $607,162 (May 1999$) is used to monetize the benefits of avoiding non-
fatal cancers (this value was updated from the $536,000 value EPA updated to 1997$ from the
Viscusi et al. 1991 study). The bladder cancer, lung cancer, and non-quantifiable health benefits
are summarized in Exhibit 8-35. Total annual health benefits resulting from bladder cancer cases
avoided range from $43.6 to $104.2 million at an MCL of 3 ug/L, $31.7 to $89.9 million at an
MCL of 5 ug/L, $17.9 to $52.1 million at an MCL of 10 ug/L, and $7.9 to $29.8 million at an
MCL of 20 ug/L. Potential annual health benefits resulting from lung cancer cases avoided range
from $47.2 to $448.0 million at an MCL of 3 ug/L, $35.0 to $384.0 million at an MCL of 5
Chapter 8, Economic Impact Analyses
8-42
Proposed Arsenic in Drinking Water Rule RIA
-------
ug/L, $19.6 to $224.0 million at an MCL of 10 ug/L, and $8.8 to $128.0 million at an MCL of
20 ug/L.
Exhibit 8-35
Estimated Monetized Total Cancer Health Benefits and
Non-Quantifiable Health Benefits from Reducing Arsenic in CWSs
Arsenic
Level
(ug/L)
3
5
10
20
Annual
Bladder Cancer
Health Benefits
($millions)12
$43.6 -$104.2
$31 .7 -$89.9
$17.9 -$52.1
$7.9 - $29.8
"What-if" Scenario and Potential Non-Quantifiable Health
Benefits
"What-if" Scenario
Annual
Lung Cancer
Health Benefits
($millions)13
$47.2 - $448.0
$35.0 - $384.0
$19.6 -$224.0
$8.8 -$128.0
Potential Non-Quantifiable
Health Benefits
Skin Cancer
Kidney Cancer
Cancer of the Nasal Passages
Liver Cancer
Prostate Cancer
Cardiovascular Effects
Pulmonary Effects
Immunological Effects
Neurological Effects
Endocrine Effects
Reproductive and Developmental
Effects
1. May 1999 dollars.
2. The lower-end estimate is calculated using the lower-end number of bladder cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was
100 percent. The upper-end estimate is calculated using the upper-end number of cancer cases avoided (see
Exhibit 5-12) and assumes that the conditional probability of mortality among the Taiwanese study group was 80
percent.
3. These estimates are based on the "what if" scenario for lung cancer, where the risks of a fatal lung cancer
case associated with arsenic are assumed to be 2-5 times that of a fatal bladder cancer case.
Reductions in arsenic exposures may also be associated with non-quantifiable benefits. EPA has
identified several potential non-health non-quantifiable benefits associated with regulating arsenic
in drinking water. These benefits may include any customer peace of mind from knowing that
their drinking water has been treated for arsenic. To the extent that the Arsenic Rule can reduce
households' perception of the health risks associated with arsenic in drinking water, household
averting actions and costs to avoid these risks, such as buying bottled water or installing home
treatment systems, could also be reduced.
8.10.2 Quantifiable and Non-Quantifiable Costs
The costs of reducing arsenic to various levels are summarized in Exhibit 8-36, which shows that,
as expected, aggregate arsenic mitigation costs increase with decreasing arsenic levels. Total
national costs at a 7 percent discount rate range are: $753.2 million per year at 3 ug/L; $442.2
million per year at 5 |ig/L; $192.4 million per year at 10 ug/L; $73.7 million per year at 20 ug/L.
Chapter 8, Economic Impact Analyses
8-43
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-36
Summary of the Total Annual National Costs of Compliance
($ millions)
Discount Rate
cws
3>/o 7%
INTTNC*
3P/o 7%
TOTAL
3P/o 7%
MCL = 3ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$639.2 $746.4
$2.2 $3.0
$1.6 $1.9
$643.1 $751.4
$0.9 $1.1
$0.6 $0.7
$1.5 $1.8
$639.2 $746.4
$3.1 $4.1
$2.2 $2.6
$644.6 $753.2
MCL=5ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$374.0 $436.0
$2.0 $2.8
$1.3 $1.6
$377.3 $440.4
$0.9 $1.1
$0.6 $0.7
$1.6 $1.8
$374.0 $436.0
$2.9 $3.9
$2.0 $2.3
$378.9 $442.2
MCL=10ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$160.4 $186.7
$1.8 $2.5
$1.1 $1.3
$163.3 $190.5
$1.0 $1.1
$0.6 $0.7
$1.6 $1.9
$160.4 $186.7
$2.8 $3.7
$1.7 $2.1
$164.9 $192.4
MCL=20ng/L
System Costs
Treatment
Monitoring/
Administrative
State Costs
TOTAL COST
$58.9 $68.3
$1.7 $2.4
$1.0 $1.2
$61.6 $71.8
$1.0 $1.1
$0.7 $0.7
$1.6 $1.9
$58.9 $68.3
$2.7 $3.5
$1.6 $1.9
$63.2 $73.7
*Costs include treatment, O&M, monitoring, and administrative costs to CWSs, monitoring and administrative costs to
NTNCWSs, and State costs for administration of water programs.
Chapter 8, Economic Impact Analyses
8-44
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-37
Mean Annual Costs per Household in CWSs
SIZE CATEGORIES
<100
101-500
501-1,000
1,001 ^3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
1,000,000 +
All categories
MCL (|ag/L)
3
$368.13
$258.68
$106.07
$63.61
$44.28
$36.39
$29.52
$23.47
$2.70
$43.73
5
$363.65
$253.64
$103.74
$60.38
$40.77
$33.22
$26.64
$21.20
$1.73
$39.18
10
$357.17
$246.38
$98.35
$56.51
$37.04
$29.13
$22.80
$18.32
$0.89
$33.05
20
$348.72
$237.67
$93.25
$51.80
$32.52
$24.99
$19.44
$15.41
$0.55
$23.62
The cost impact of reducing arsenic in drinking water at the household level was also assessed.
Exhibit 8-37 examines the cost per household for each system size category. As shown in the
table, costs per household decrease as system size increases. However, costs per household do
not vary significantly across arsenic levels. This is because costs do not vary significantly with
removal efficiency; once a system installs a treatment technology to meet an MCL, costs based
upon the removal efficiency that the treatment technology will be operated under remain relatively
flat. Per household costs are, however, somewhat lower at less stringent arsenic levels. This is
due to the assumption that some systems would blend water at these levels and treat only a
portion of the flow in order to meet the target MCL.
Chapter 8, Economic Impact Analyses
8-45
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit 8-38
Cost per Bladder Cancer Case Avoided for the Proposed Arsenic Rule
(CWSs Comply with MCL / NTNCs Monitor, in $ millions)
Arsenic Level
(H9/L)
lower bound**
upper bound**
3% Discount Rate
3
5
10
20
$
$
$
$
29.4 $
23.8 $
18.3 $
15.9 $
12.3
8.4
6.3
4.2
7% Discount Rate
3
5
10
20
$
$
$
$
34.4 $
27.7 $
21.4 $
18.6 $
14.4
9.8
7.3
4.9
*Costs all treatment, O&M, monitoring, and administrative costs to CWSs,
monitoring and administrative costs to NTNCWSs, and State costs for
administration of water programs.
"Lower/upper bounds correspond to estimates of bladder cancer cases avoided.
Exhibit 8-38 illustrates the cost per bladder cancer case avoided, based on national cost estimates
which include all the costs of treatment, O&M, monitoring and administrative costs to CWSs,
monitoring and administrative costs to NTNCWSs, and all State costs for administration of water
programs. At a 3 percent discount rate, cost per case ranges from approximately $29.4 million at
an arsenic level of 3 jig/L (lower bound estimate of avoided bladder cancer cases) to $4.2 million
at an MCL of 20 jig/L (upper bound of avoided bladder cancer cases). Similarly the range at a 7
percent discount rate is $34.3 million to $4.9 million.
Chapter 8, Economic Impact Analyses
8-46
Proposed Arsenic in Drinking Water Rule RIA
-------
References
Agency for Toxic Substances and Disease Registry. 1998. Toxicological Profile for Arsenic
(Draft). U.S. Department of Health and Human Services.
Baker, Mary S., et al. 1989. Site specific treatment costs for cancer: an analysis of the medicare
continuous history sample file. Cancer Care and Cost. DRGs and Beyond. Richard M.
Scheffler and Neil C. Andrews, eds. Ann Arbor, MI: Health Administration Press
Perspectives.
Brouwer, OF, et al. 1992. Increased neurotoxicity of arsenic in methylenetetrahydrofolate
reductase deficiency. Clin Neurol Neurosurg. 94(4):307-10.
Buchet, JP and Lauwerys, R. 1987. Study of factors influencing the in vio methylation of
inorganic arsenic in rats. Toxicol Appl Pharmacol. 91: 65-74.
Clifford, D., G. Ghurye and A. Tripp (1998). "Arsenic Removal by Ion Exchange With and
Without Brine Reuse," AWWA Inorganic Contaminants Workshop, San Antonio, TX,
February 23-24, 1998.
Dallaire, L, and Beliveau, R. 1992. Phosphate transport by capillaries of the blood-brain barrier. J
Biol Chem. 267 (31): 22323-7.
Dong, JT and Luo, XM. 1993. Arsenic-induced DNA-strand breaks associated with DNA-protein
crosslinks in human fetal lung fibroblasts. Mutat Res 302(2):97-102.
Edelman, GM. 1992. Morphoregulation. Dev Dynamics. 193:2-10.
EGA, IRIS web site. Extracted 8199.
EPA. 1995. Community Water System Survey.
EPA, 1997a. Information Collection Request for the Public Water System Supervision Program.
EPA. 1997b. Community Water System Survey, Volume II: Detailed Survey Result Tables and
Methodology Report. Office of Water, Washington, D.C. EPA/815-R-97-001a and
-OOlb.
EPA. 1997c. The Benefits and Costs of the Clean Air Act, 1970-1990. Clean Air Act §812 report
prepared for U.S. Congress by U.S. EPA Office of Air and Radiation. October. EPA
410-R-97-002.
References R-l Proposed Arsenic in Drinking Water Rule RIA.
-------
EPA. 1998a. An Assessment of the Vulnerability of Non-Community Water Systems to SDWA
Cost Increases. Washington, D.C.
EPA. 1998b. Cost of Illness Handbook. Washington, D.C.
EPA. 1998c. National-Level Affordability Criteria Under the 1996 Amendments to the Safe
Drinking Water Act, Draft. Washington, D.C.
EPA. 1999a. Drinking Water Baseline Handbook, Draft 1st Edition. Washington, D.C.
EPA, 1999b Geometries and Characteristics of Public Water Systems. Prepared for Office of
Ground Water and Drinking Water by Science Applications International
Corporation(SAIC) EPA Contract No. 68-C6-0059.
EPA. 1999c. Arsenic Occurrence in Public Drinking Water Supplies. Washington, D.C.
Federal Register. 1998. Aug. 6.
Freeman, HC, et al. 1979. Clearance of arsenic ingested by man from arsenic contaminated fish.
Bull Environ Contam Toxicol. 22(l-2):224-9.
Geubel, AP, et al. 1988. Abnormal methylation capacity in human liver cirrhosis. Int J Clin
Pharmacol Res. 8:117-22.
Gunn, TM, et al. 1992. Further genetic studies of the cause of exancephaly in SELH mice.
Teratology. 45:679-86.
International Consultants, Inc. (ICI) and Malcolm Pirnie, Inc. (MPI) (1999). Technologies and
Costs for the Removal of Arsenic from Drinking Water, Prepared for the United States
Environmental Protection Agency, Office of Ground Water and Drinking Water,
Standards and Risk Management Division, Targeting and Analysis Branch, April 1999.
Jha, AN, et al. 1992. Genotoxic effects of sodium arsenite on human cells. Mutat Res. 1992
284:(2):215-21.
Kapadia, Amit 1999a. E-mail from Amit Kapadia, USEPA to Michael Muller, Abt Associates,
July 1.
Kapadia, Amit 1999b. Telephone conversation between Amit Kapadia, USEPA and Michael
Muller, Abt Associates, July.
Leonard, A and Lauwerys, RR. 1980. Carcinogenicity, teratogenicity and mutagenicity of arsenic.
Mutat Res. 75(l):49-62.
References R-2 Proposed Arsenic in Drinking Water Rule RIA.
-------
Li, W and Chou, IN. 1992. Effects of sodium arsenite on the cytoskeleton and cellular glutathione
levels in cultured cells. Toxicol Appl Pharmacol 114(1): 132-9.
Morris-Kay, GM, et al. 1994. Normal neurulation in mammals. In Ciba Foundation Symposium,
neural tube defects. New York: John Wiley & Sons, 51-69.
Mottet, NK and Perm, VH. 1983. The congenital teratogenicity and perinatal toxicity of metals.
In reproduction and development toxicity of metals. Clarkson TW, Nordberg GF, and
Sager, PR, eds. New York: Plenum Press, 93-125.
National Academy of Sciences. 1995. Risk Assessment in the Federal Government.
National Cancer Institute. 1999.
National Research Council. 1999. Arsenic in Drinking Water.
Petres, J, et al. 1977. Effects of arsenic cell metabolism and cell proliferation: cytogenetic and
biochemical studies. Environ. Health Perspect. 19:223-7.
Scheonwolf, GC and Smith, JL. 1990. Mechanisms of neurulation: traditional viewpoint and
recent advances. Development. 109:243-270.
SDWIS, 1998. December 1998 freeze.
Smith, Ben 1999a. Facsimile from Ben Smith, USEPA to Michael Muller, Abt Associates, June
30.
Somogyi, A and Beck, H. 1991. Nurturing and breast-feeding exposure to chemicals in breast
milk. International Workshop on the Impact of the Environment on Reproductive Health,
Copenhagen, Denmark, September 30-October 4, 1991. Environ Health Perspect 101
(suppl.2): 45-52.
Subramanian, K.S., T. Viraraghavan, T. Phommavong and S. Tanjore (1997). "Manganese
Greensand for Removal of Arsenic in Drinking Water," Water Quality Research Journal
Canada, 32:3:551-561.
Taubeneck, MW, et al. 1994. Altered material zinc metabolism following exposure to diverse
developmental toxicants. Reprod Toxicol 8(1):25-40.
Tsai, SM, et al. 1999. Mortality for certain diseases in areas with high levels of arsenic in drinking
water. Arch Environ Health. 54(3): 186-93.
References R-3 Proposed Arsenic in Drinking Water Rule RIA.
-------
Vahter, M. and Marafante, E. 1987. Effects of low dietary intake methionine, choline, or proteins
on the biotransformation of arsenite in the rabbit. Toxiol Lett. 37: 41-46.
Variance Technology Findings for Contaminants Regulated before 1996.
Viscusi, K., et al. 1991. Pricing Environmental Health Risks: Survey Assessments of Risk-Risk
and Risk-dollar Tradeoffs. Journal of Environmental Economics and Management.
201:32-57.
References R-4 Proposed Arsenic in Drinking Water Rule RIA.
-------
Appendix A: Decision Tree and Decision Matrix
A.1 Introduction
The purpose of this appendix is to present the rationale behind the development of the decision
tree and associated decision matrix. It includes an overview of the decision tree structure and
major factors impacting the decision- making process. The following list outlines the contents of
this appendix:
A.2 BACKGROUND - Presents a brief history of the arsenic regulation and the
statutory requirements impacting EPA and the decision-making process.
A.3 DECISION TREE OVERVIEW - Outlines the decision tree and presents major
factors affecting the decision-making process.
A.4 MAJOR FACTORS AFFECTING THE DECISION TREE - Presents the
rationale for selecting parameters which impact the decision tree, including MCL,
population, water type, region, and co-occurrence of solutes.
A.5 ADDITIONAL FACTORS AFFECTING THE DECISION TREE - Presents
other parameters in the process which impact the decision tree, including: corrosion
control, pre-oxidation, regionalization, and alternative technologies.
A.6. ARSENIC RULE-MAKING DECISION MATRIX - Complete copy of the
decision matrix developed for the arsenic rule-making process. These are provided in
Exhibits A-2 through A-17.
A.7. REFERENCES.
A.2 Background
In 1998 and 1999, EPA conducted technology and cost evaluations for the removal of arsenic
from drinking water. These evaluations looked into the effectiveness of various removal
technologies and the capital and operations and maintenance (O&M) costs associated with each
process. The following were evaluated and determined effective to varying degrees:
Coagulation/Filtration (C/F);
Modified Coagulation/Filtration (modifications to existing C/F plants);
Direct Filtration (DF);
Coagulation Assisted Microfiltration (CMF);
Lime Softening (LS);
Modified Lime Softening (modifications to existing LS plants);
Appendix A, Decision Tree and Decision Matrix A-l Proposed Arsenic in Drinking Water Rule RIA
-------
Activated Alumina (AA);
Ion Exchange (IX);
Ultrafiltration (UF);
Nanofiltration (NF);
Reverse Osmosis (RO);
Greensand Filtration (GF); and
Point-of-Entry (POE) and Point-of-Use (POU) Treatment Options.
The technology and cost evaluation yielded a document entitled Technologies and Costs for the
Removal of Arsenic From Drinking Water (ICI and MPI, 1999). The document includes detailed
evaluations of the above technologies, capital and O&M cost estimates for each of the listed
technologies, as well as other technologies that were considered ineffective or unproven.
EPA used the information contained in the technologies and costs (T&C) document to develop a
regulatory decision tree. The decision tree was then used to fashion a decision matrix which
contains the probability that a given system will choose a treatment technology based on the
percent removal required to meet the proposed MCL. The decision matrix, unit cost curves for
treatment and waste disposal (illustrated in the T&C), treatment-in-place data and occurrence
estimates will be used to develop national cost of compliance estimates.
A.3 Decision Tree Overview
The decision tree is a flow chart that details the thought process involved in estimating which
treatments will be installed to comply with the four MCL options. The decision tree was
developed under the overriding assumption that systems would attempt to comply with the
proposed MCL by choosing the lowest cost approach. However, it was also recognized that
some systems may not be able to choose the lowest cost technology, because they face certain
constraints. Therefore, the "branches" of the decision tree represent constraining factors that
must be taken into consideration when estimating which treatment technology a system will
select. The following questions were used to define the branches of the decision tree. A detailed
explanation of the rationale behind each question is presented in the next section.
1. What i s the target MCL?
2. What is the influent arsenic concentration of the system?
3. What is the population of the system?
4. In what region of the country is the system located?
5. Does the system utilize ground or surface water?
6. Does the system currently have some type of treatment in place?
Appendix A, Decision Tree and Decision Matrix A-2 Proposed Arsenic in Drinking Water Rule RIA
-------
7. Are there any co-occurrence issues that preclude the use of a particular treatment
technology?
8. What disposal options are available for the treatment technology selected?
Even though these eight questions represent major decisions within the decision tree structure and
each question must be answered to complete a branch of the tree, the tree was structured to limit
the number of branches that must be considered during the development of the decision matrix.
By first selecting the target MCL and influent arsenic concentration, some branches of the tree are
eliminated early in the process. For example, if the target MCL is 20 //g/L and the influent arsenic
levels are between 10 and 20 //g/L, no removal is necessary, and therefore the probability of
choosing any treatment technology is zero. Similar logic was applied to other decisions factors.
A.4 Major Factors Affecting the Decision Tree
This section explains the rationale behind selecting each particular decision factor. Specifically,
this section will discuss the following:
the MCL target,
influent arsenic concentration,
population,
region where the system is located,
source water,
whether a system has existing treatment in place,
co-occurrence of solutes, and
waste disposal issues.
A.4.1 MCL Target
The MCL target is the single most important factor in development of the decision tree because it
is essential for determining all other branches of the tree. The decision tree is structured such
that selection of the target MCL is the first step in the decision process. Four MCL scenarios
will be analyzed as part of the regulatory development process (3 //g/L, 5 //g/L, 10 //g/L, and 20
A. 4.2 Influent Arsenic Concentration
When coupled with the MCL target, the influent arsenic concentration was of major importance in
developing the decision tree. In developing the decision tree, influent arsenic concentrations were
grouped into four categories: less than 10 //g/L, 10 to 20 //g/L, 20 to 30 //g/L, and 30 to 50 //g/L.
Given the MCL under consideration, the influent arsenic concentration determines what percent
removal of arsenic is needed, if any, and lays the groundwork for remaining decisions in the tree.
Percent removal is critical for determining what additional technologies may be feasible. For
example, if a surface water plant has an influent arsenic level of 50 //g/L, and the target MCL is 2
Appendix A, Decision Tree and Decision Matrix A-3 Proposed Arsenic in Drinking Water Rule RIA
-------
Mg/L, then 96 percent removal is required.1 Research indicates that lime softening is only capable
of achieving approximately 80 percent removal; thus, lime softening would not be a viable
treatment option in this branch of the decision tree. Likewise, in the decision matrix, the
probability of choosing lime softening as a treatment technology would be zero whenever the
percent removal is over 80 percent.
A.4.3 System Size
System size, or population, also plays a significant role in determining the treatment options
available to a system, as well as the affordability of a particular technology for a system. EPA
established nine size categories to be used in the decision tree and RIA process:
1. 25 to 100;
2. 101-500;
3. 501-1,000;
4. 1,001-3,300;
5. 3,301-10,000;
6. 10,001-50,000;
7. 50,001-100,000;
8. 100,001-1,000,000; and
9. greater than 1,000,000.
In developing the decision tree EPA grouped size categories one through three (25 to 1,000
people) and four though eight (1,001 to 1,000,000 people) because POE and POU treatment
options are considered viable treatment alternatives only for the three smallest categories. With
these groupings, the available technologies are constant for each of the size categories within each
group. However, the probability of choosing a given technology is still assigned dependent on
system size category. Therefore, in the development of the decision matrix, the probabilities that
certain technologies are chosen change for each of the size categories. Hence, the decision matrix
accommodates each of the eight size categories (one through eight) individually. Systems within
the ninth size category (greater than 1,000,000) will be addressed on a case-by-case basis by EPA,
and will fall outside the scope of the decision tree process.
A.4.4 Region
The region of the nation that a system resides in does not effect the treatment options available.
Therefore, the decision tree is structured in such a way that, regardless of the region, the branches
are identical, and in fact refer to the same pages within the decision tree. However, the number of
systems that may select a particular option as defined in the decision matrix, is region-specific.
EPA has decided that the nation can be divided into three regions for the purpose of the decision
Required removal percentages in the decision tree are based on worst cast scenarios and therefore
correspond with the upper bound of the arsenic concentration range for each category.
Appendix A, Decision Tree and Decision Matrix A-4 Proposed Arsenic in Drinking Water Rule RIA
-------
making process: 1) Southwest Region; 2) Northwest Region; and 3) East Region. The regions
were selected based upon availability of water (i.e., scarcity of water) and availability of land. In
the Southwest Region, for example, water may be scarce and treatment technologies that generate
large volumes of reject water, such as RO, may not be appropriate. In the East Region, water
scarcity is much less a concern than the availability of land. Technologies or disposal options that
require significant amounts of land are less likely to be utilized in the East Region. The
Northwest Region, by comparison, is less affected by the scarcity of water or land availability than
either of the other two regions.
A.4.5 Source Water
The source of the system's raw water, either ground water or surface water, plays a major role in
determining the technologies that may already be in use by a system and what treatment options
are available if a system needs to install a new treatment facility.
For example, greensand filtration is affected by the level of iron in the raw water. Influent levels
greater than 300 mg/L (ppm) are conducive to removal of arsenic by greensand filtration. Surface
waters typically have low iron content, whereas ground waters often have levels in excess of 300
mg/L (Subramanian, et al., 1997). Accordingly, greensand filtration was not considered a viable
removal technology for surface water systems.
To determine the types of treatment that are currently being utilized throughout the country, EPA
reviewed the Community Water Systems Survey (CWSS). EPA determined there are few
surface water systems utilizing RO, IX or AA. As a result, when approximating the treatment in
place options, RO, IX, and AA were omitted for surface water systems.
Arsenic removal is significantly more efficient when arsenic is present as arsenate (As5+).
Research has demonstrated many of the technologies considered perform poorly when arsenite
(As3+) is the predominant form (ICI and MPI, 1999). Arsenite can be easily oxidized to arsenate
using conventional oxidation methods, such as chlorination and potassium permanganate addition.
Ground waters typically contain higher levels of As3+, whereas As5+ is the dominant species in
surface waters. As a result, ground waters are more likely to install pre-oxidation and use higher
oxidant doses, whereas surface waters may be able to get by with little or no pre-oxidation
capacity.
A.4.6 Systems with Treatment In-Place
If a system currently has treatment in-place it will significantly impact the decision tree. Many
existing treatment facilities will be able to achieve the necessary removal with little or no
modification, particularly at high MCLs. At the lower MCLs (2 and 5 //g/L), existing facilities
may be able to add polishing steps or make some modifications to the existing system to achieve
the required removal level. Table 1 outlines the treatment technologies included in the decision
tree, the percent removal assumed capable without modification or polishing, and the maximum
percent removal.
Appendix A, Decision Tree and Decision Matrix A-5 Proposed Arsenic in Drinking Water Rule RIA
-------
A.4.7 Systems without Treatment In-Place
Many factors affect the decision tree when considering the addition of a treatment option to
systems with no current treatment in place. Source water type and quality, system size, required
arsenic removal, and removal achievable by a particular technology are all major considerations.
Many of these considerations have been discussed earlier in Section 4, however, source water
quality, i.e., co-occurrence of solutes, is discussed in Section 4.8.
For ground water systems (any size) without treatment in-place, the most suitable treatment
technologies are IX, AA and RO, though CF and LS may be used. CF and LS are best suited
for large surface water systems without treatment in-place while IX, AA and RO are best suited
for small surface water systems without treatment in-place. In either case, RO is not a suitable
treatment technology in regions where water is scarce. Modified CF and LS are for those surface
water systems that already have CF or LS in-place.
The SDWA identifies POE and POU treatment units as potentially affordable technologies, but
stipulates that POE and POU treatment systems "shall be owned, controlled and maintained by the
public water system, or by a person under contract with the public water system to ensure proper
operation and compliance with the maximum contaminant level or treatment technique and
equipped with mechanical warnings to ensure that customers are automatically notified of
operational problems."
Preliminary affordability determinations have shown that POE and POU technologies will only be
considered viable for small systems. These determinations have shown the cost breakpoint to be
in the area of 200 persons served. This estimate does not account for waste disposal costs, which
would make central treatment estimates more expensive, thus increasing the breakpoint. As a
result, POE and POU units are only included in the decision tree for systems with populations
fewer than 1,000 individuals.
Appendix A, Decision Tree and Decision Matrix A-6 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit A-1
Treatment Technologies for Systems with Treatment In-Place and Percent Removals
Assumed and Achievable
Treatment Technology
Coagulation/Filtration2
Lime Softening2
Coagulation Assisted Microfiltration
Ion Exchange
Activated Alumina
Reverse Osmosis
Greensand Filtration3
POE Activated Alumina
POD Ion Exchange
Percent Removal of
In-Place System
50
50
NA
95
95
95
90
NA
NA
Maximum Percent
Removal1
95
80
90
>95
>95
>95
90
>95
>95
1 - For Percent Removals of In-Place Systems that are very close to Maximum Percent Removals (e.g., 95% and > 95%) polishing steps
may be required.
2 - Maximum Percent Removal involves modification to existing system in the form of additional chemical feed systems, pumping, piping,
etc.
3 - EPA is currently evaluating the removals achievable by this process. The final decision matrix may reflect significantly lower removals,
e.g., 50%.
NA - Not Applicable
A.4.8 Co-Occurrence of Solutes
There are a number of solutes and water quality parameters that may effect the viability of a
particular treatment option. Total dissolved solids (TDS), silica, sulfate and iron can all be major
detractors/benefactors for the use of a particular technology. The decision tree simply cannot
account for each individual situation where the influent water quality plays a role in selecting the
treatment option. Utilities are encouraged to read the T&C document (ICI and MPI, 1999) to
gather additional information on parameters which impact the performance of a particular
technology.
The decision tree uses influent sulfate and iron levels as decision factors in selecting treatment
technologies. Sulfate has been shown to decrease the effectiveness of ion exchange processes for
arsenic removal; therefore, these processes are not generally recommended where influent sulfate
levels exceed 120 mg/L (Clifford, et al., 1998). Iron, on the other hand, significantly improves the
effectiveness of greensand filtration (Subramanian, et al., 1997). Greensand filtration is best
suited for ground waters (which typically contain higher levels of iron than surface waters) with
high influent levels of iron (300 mg/L).
The decision tree has been structured to accommodate the impact of sulfate and iron on treatment
Appendix A, Decision Tree and Decision Matrix
A-7 Proposed Arsenic in Drinking Water Rule RIA
-------
effectiveness. For ground water sources, both sulfate and iron levels are considered. Ion
exchange is not considered a feasible treatment option when sulfate levels exceed 120 mg/L, and
greensand filtration is not considered viable when the iron level falls below 300 mg/L. For surface
waters in which high iron levels are rare, only sulfate has been considered. For purposes of
approximating national cost, greensand filtration is not considered a treatment option for surface
water systems. Ion exchange is not considered a feasible treatment option when the sulfate level
exceeds 120 mg/L. Therefore, in the development of the decision matrix, the probability that a
system will choose a particular technology is assumed with an eye on the regional levels of iron
and sulfate.
A.4.9 Waste Disposal
Waste handling and disposal options are specific to the treatment technology selected and the
availability of disposal options does not vary by system size in the decision tree. However, the
probability that a system will utilize a particular option does vary with system size. For example,
evaporation ponds may not be suitable for large systems in the Northwest Region where climatic
conditions do not facilitate evaporation. Mechanical dewatering devices can be expensive and
may require significant operator attention and may not be suitable for small systems.
A.4.9.1 Mechanical Dewatering
Mechanical dewatering processes include centrifuges, vacuum-assisted dewatering beds, belt filter
presses, and plate and frame filter presses. Such processes generally have high capital, as well as
high O&M costs, compared to similar capacity non-mechanical dewatering processes (e.g.,
storage lagoons). Due to the high costs, such processes are generally not suitable for application
with very small water systems.
Filter presses have been used in industrial processes for years and have been increasing in the
water treatment industry over the past several years. The devices have been successfully applied
to both lime softening process sludge and coagulation/filtration process sludge. Filter presses
require little land, have high capital costs, and are labor intensive.
Centrifuges have also been used in the water industry for years. Centrifugation is a continuous
process requiring minimal time to achieve the optimal coagulation/filtration. Centrifuges have low
land requirements and high capital costs. They are more labor intensive than non-mechanical
alternatives, but less intensive than filter presses. Again, due to the capital and O&M
requirements, centrifuges are more suitable for larger water systems.
A.4.9.2 Evaporation Ponds and Drying Beds
Evaporation ponds and drying beds are non-mechanical dewatering technologies wherein
favorable climatic conditions are used to dewater waste brines generated by treatment processes
such as reverse osmosis and ion exchange. Ponds and drying beds are not generally suitable for
lime softening or coagulation/filtration. Evaporation is an extremely land intensive waste handling
option requiring shallow basins with large surface areas which can be an important consideration
in densely populated regions. Evaporation ponds and drying beds have few operation and
Appendix A, Decision Tree and Decision Matrix A-8 Proposed Arsenic in Drinking Water Rule RIA
-------
maintenance requirements but are only feasible in regions with high temperatures, low humidity,
and low precipitation.
A.4.9.3 Storage Lagoons
Lagoons are the most common, and often least expensive, method to thicken or dewater
treatment sludge, however, they are land intensive. Storage lagoons are best suited for
dewatering lime softening process sludge, though they have been applied with some success to
coagulation/filtration process sludge. Coagulation/filtration process sludge do not typically
dewater well in storage lagoons. Thickened coagulation/filtration process sludge can be difficult
to remove from lagoons and often require dredging or vacuum pumping by knowledgeable
experts.
Since lagooning is a land intensive process, it has limited applicability in densely populated
regions, or regions with limited land availability. Lagoons are best suited for areas with favorable
climatic conditions, i.e., high temperatures, low humidity, and low precipitation. In fact, in
northern climates, winter freezing can dehydrate coagulation/filtration sludge.
A.4.9.4 Direct Discharge
Direct discharge to a surface water body is a common method of disposal for water treatment
byproducts. No pretreatment or concentration of the byproduct stream is necessary prior to
discharge, and the receiving water dilutes the waste concentration and gradually incorporates the
sludge or brine. The primary cost associated with direct discharge is that of the piping. Direct
discharge requires little oversight and operator experience and maintenance are minimal. This
method has been used to successfully dispose of lime softening and coagulation/filtration process
sludge materials, as well as brine streams generated at reverse osmosis and ion exchange water
plants.
A.4.9.5 POTW Discharge
Indirect discharge (POTW discharge) is a commonly used method of disposal for filter backwash
and brine waste streams. Coagulation/filtration and lime softening sludge materials have also been
successfully disposed of in this manner. The primary cost associated with POTW discharge is that
of the piping. Additional costs associated with POTW discharge may include lift stations,
additional piping for access to the sewer system, and any cost incurred by the POTW in
accommodating the increased demands on the POTW.
A.4.9.6 Dewatered Sludge Land Application
Dewatered sludge can be disposed of by spreading the material over an approved land surface.
Land application is limited by the availability of land. In areas where grassland, farmland, or
forested land is unavailable, transportation can significantly affect the cost effectiveness of this
disposal option. Land application can be a means of final disposing of lime softening sludge, and
to a lesser degree, coagulation/filtration sludge. Lime softening sludge can be used to neutralize
Appendix A, Decision Tree and Decision Matrix A-9 Proposed Arsenic in Drinking Water Rule RIA
-------
soil pH while coagulation/filtration sludge offer no benefit to soil chemistry and are generally used
as fill material.
A.4.9.7 Sanitary Landfill Disposal
Two forms of sanitary landfill are commonly used for disposal of water treatment byproducts:
monofills and commercial nonhazardous waste landfills. In some parts of the country, decreasing
landfill availability, rising costs, and increasing regulations are making landfill disposal more
expensive. Costs associated with the development of monofills are generally less than those
associated with commercial nonhazardous water landfill.
A.5 Additional Factors Affecting the Decision Tree
A.5.1 Pre-Oxidation
As mentioned above, inorganic arsenic occurs in two primary valence states, arsenite (As III) and
arsenate (As V). As(III) is dominant in ground waters while surface waters more typically
contain As(V). As(III) is easily oxidized to As(V) by conventional oxidation technologies such as
chlorination and potassium permanganate addition. Each of the treatment technologies
considered in the decision tree remove As(V) more readily than As(III) and as a result, pre-
oxidation may be necessary depending upon source water conditions.
Pre-oxidation is included in the decision tree for all treatment technologies. Systems without
treatment in-place may already be chlorinating which may meet pre-oxidation requirements. For
those systems, pre-oxidation may or may not need to be installed. Similarly, systems with
treatment in-place may have pre-oxidation in-place which could meet the pre-oxidation
requirements. For single-house (POE) or single tap (POU) treatment options, centralized pre-
oxidation is required.
A.5.2 Corrosion Control
Many of the treatment technologies considered in the decision tree (e.g., LS, AA, IX, and RO)
remove hardness and alkalinity. Removal of hardness and alkalinity can reduce the pH of finished
water and lead to corrosion problems within the system. Hardness and alkalinity, at the
appropriate levels, act as buffers against corrosion in the treatment plant and distribution system.
At these levels, alkalinity and hardness form protective coatings (metal hydroxides), control pH
and enhance the buffer effect against corrosion. Corrosion control is included in the decision tree
for all new constructions, that is, systems without existing treatment plants. It was assumed that
existing plants had adequate corrosion control in-place.
A.5.3 Regionalization
The term regionalization is used to define the process of purchasing water or transporting water
from one community to another. Numerous operational and other factors including
Appendix A, Decision Tree and Decision Matrix A-10 Proposed Arsenic in Drinking Water Rule RIA
-------
1) the availability of water,
2) water quality,
3) geography, and
4) economic factors influence the decision to implement regionalization.
Water quality also plays a role in the decision-making process. For example, if a community's
source water is contaminated, it may be less expensive for the community to purchase water from
another community than to treat its own water source. Regionalization is considered an option
only for small systems (
-------
Currently, GFH media costs approximately $4,000 per ton. In addition to the high cost of the
media, the direct deposition of spent GFH as hazardous waste is favored
A.5.4.3 Iron Filings
The Iron Filings process is essentially a filter technology, much like greensand filtration, wherein
the source water is filtered through a bed of sand and iron filings. Unlike some technologies (i.e.
ion exchange), sulfate is actually introduced in this process to encourage arsenopyrite
precipitation.
While this process seems to be quite effective, its use as a drinking water treatment technology
appears to be limited. There is no indication that this technology can reduce arsenic levels below
approximately 25 ppb. This technology also suffers from a study design which failed to test its
effectiveness at influent levels of concern in drinking water. Since the study design called for such
high influent levels - 470 to 20,000 ppb - there is no data to indicate how the technology performs
at normal source water arsenic levels, which most certainly are below the 470 ppb level used in
experimentation. This technology needs to be further evaluated before it should be recommended
as an approved arsenic removal technology for drinking water.
A.5.4.4 Iron Oxide Coated Sand
Iron oxide coated sand (IOCS) is a rare process that has shown some tendency for arsenic
removal. IOCS consists of sand grains coated with ferric hydroxide which are used in fixed bed
reactors to remove various dissolved metal species. Factors such as pH, arsenic oxidation state,
competing ions, EBCT, and regeneration time have significant effects on the removals achieved
with IOCS. Like other processes, the media must be regenerated upon exhaustion. IOCS has
only limited experience having only been tested at bench-scale. High levels of arsenite could
reduce IOCS effectiveness because the bonding is strong and may permanently damage the media.
Natural organic matter may also be problematic for arsenic removal. IOCS also takes a
considerable amount of time to produce in a laboratory setting. At full-scale this would likely
result in high capital cost.
A.6 Arsenic Rule Making Decision Matrix
The actual decision tree is illustrated as a flow chart and occupies over 300 written pages. Rather
than include the decision tree as part of this appendix, the decision matrix is presented. The
decision matrix contains the actual assumptions regarding compliance decisions that were used to
estimate system compliance costs. The decision tree is presented in Exhibits A-2 to A-16. Each
exhibit provides the decision matrix for a particular source water and size category pair (e.g.
Exhibit A-2 contains the decision matrix for ground water systems serving fewer than 100
people). The first column described the treatment technology train associated with the row. The
next three columns present three different decision probabilities for each technology treatment
train, depending on the required percentage reduction in influent arsenic (e.g. in the cost analysis,
a system which needed to remove between 50 and 90 percent of the influent arsenic would have a
seven percent chance of choosing treatment technology train 24, Reverse Osmosis POU with pre-
corrosion control).
Appendix A, Decision Tree and Decision Matrix A-12 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit A-2
Probability Decision Tree: Ground Water Systems Serving 100 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidafon
4 Modify Coagulation/Filtration and pre-oxidafon
5 Anion Exchange (25 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidafon
6 Anion Exchange (1 50 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidafon
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidafon
8 Anion Exchange (1 50 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidafon
9 Activated Alumna (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidafon
10 Activated Alumna (10,000 BV) and non-hazardous landfill (forspent media) and pre-oxidafon
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTW waste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Mcrofiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Mcrofiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTWfor backwash stream
17 Anion Exchange (25 mg/l SO4) and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumna (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumna (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidafon
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
0.0
0.0
3.0
2.0
2.0
0.0
2.0
0.0
60.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
4.0
4.0
5.0
100.00
0.0
0.0
3.0
2.0
2.0
0.0
2.0
0.0
70.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
7.0
7.0
7.0
100.00
0.0
0.0
1.0
1.0
17.0
0.0
17.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
18.0
8.0
10.0
10.0
100.00
-------
Exhibit A-3
Probability Compliance Decision Tree: Ground Water Systems Serving 101-500 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (1 50 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (forspent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTW waste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alurrina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
0.0
0.0
3.0
4.0
24.0
0.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
0.0
0.0
0.0
0.0
7.0
6.0
4.0
5.0
5.0
100.00
0.0
0.0
3.0
4.0
18.0
0.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
19.0
19.0
6.0
6.0
7.0
100.00
0.0
0.0
1.0
2.0
17.0
0.0
17.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
18.0
9.0
9.0
9.0
100.00
-------
Exhibit A-4
Probability Compliance Decision Tree: Ground Water Systems Serving 501-1,000 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and rrechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumna and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-80% >90%
0.0
0.0
2.0
2.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
24.0
0.0
0.0
0.0
15.0
15.0
0.0
0.0
0.0
100.00
0.0
0.0
2.0
2.0
18.0
3.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
18.0
19.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
0.0
0.0
1.0
1.0
17.0
13.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
13.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
-------
Exhibit A-5
Probability Compliance Decision Tree: Ground Water Systems Serving 1,001-3,300 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and rrechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumna and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-80% >90%
0.0
0.0
3.0
3.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
8.0
18.0
24.0
0.0
0.0
0.0
6.0
6.0
0.0
0.0
0.0
100.00
0.0
0.0
3.0
3.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
10.0
0.0
18.0
0.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
0.0
0.0
1.0
1.0
17.0
13.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
13.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
-------
Exhibit A-6
Probability Compliance Decision Tree: Ground Water Systems Serving 3,301-10,000 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and rrechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumna and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-80% >90%
0.0
0.0
3.0
8.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
3.0
18.0
24.0
0.0
0.0
0.0
9.0
9.0
0.0
0.0
0.0
100.00
0.0
0.0
3.0
8.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
7.0
0.0
18.0
0.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
0.0
0.0
1.0
4.0
17.0
12.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
11.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
-------
Exhibit A-7
Probability Compliance Decision Tree: Ground Water Systems Serving 10,001-50,000 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and rrechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alurrina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Percent of Treatment Required to
Achieve MCL
<50% 50-80% >90%
0.0
0.0
5.0
4.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
22.0
21.0
0.0
24.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
4.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
28.0
27.0
0.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
2.0
17.0
12.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
12.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-8
Probability Compliance Decision Tree: Ground Water Systems Serving 50,001-100,000 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and cherrical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and rrechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
17 Anion Exchange (25 rrg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 rrg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 rrg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumna and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-80% >90%
0.0
0.0
3.0
4.0
23.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
24.0
23.0
0.0
23.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
100.00
0.0
0.0
3.0
4.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
29.0
28.0
0.0
18.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
100.00
0.0
0.0
1.0
2.0
17.0
12.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
13.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
-------
Exhibit A-9
Probability Compliance Decision Tree: Ground Water Systems Serving 100,001-1,000,000 People
No. Treatment Technology Train
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTW waste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (forspent media) and pre-oxidation
1 1 Reverse Osmosis and direct discharge w aste disposal and corrosion control and pre-oxidation
1 2 Reverse Osmosis and POTW w aste disposal and corrosion control and pre-oxidation
1 3 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
1 4 Coagulation Assisted Microfiltration and mechanical de-w atering/non-hazardous landfill w aste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-w atering/non-hazardous landfill waste disposal and pre-oxidation
1 6 Oxidation Filtration (Greensand) and POTW for backw ash stream
1 7 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
1 8 Anion Exchange (1 50 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
1 9 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (1 0,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
Sum of Probabilities:
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
0.0
0.0
10.0
5.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
43.0
42.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
100.00
0.0
0.0
10.0
5.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
43.0
42.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
100.00
0.0
0.0
5.0
2.0
17.0
10.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
17.0
11.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
100.00
-------
Exhibit A-10
Probability Decision Tree: Surface Water Systems Serving 100 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-w atering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
4.0
22.0
2.0
0.0
2.0
0.0
56.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
5.0
4.0
0.0
0.0
4.0
22.0
2.0
0.0
2.0
0.0
46.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
8.0
8.0
0.0
0.0
2.0
11.0
17.0
0.0
17.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.0
3.0
15.0
16.0
16.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-11
Probability Decision Tree: Surface Water Systems Serving 101-500 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
9.0
53.0
15.0
0.0
14.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.0
3.0
3.0
0.0
0.0
9.0
53.0
17.0
0.0
17.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
2.0
0.0
0.0
0.0
0.0
0.0
4.0
26.0
16.0
0.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
19.0
19.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-12
Probability Decision Tree: Surface Water Systems Serving 501-1,000 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (forspent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
19.0
73.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
19.0
73.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
36.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
12.0
12.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-13
Probability Decision Tree: Surface Water Systems Serving 1,001-3,300 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
16.0
76.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
76.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
35.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
13.0
12.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-14
Probability Decision Tree: Surface Water Systems Serving 3,301-10,000 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
7.0
85.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
7.0
85.0
8.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
3.0
42.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
12.0
11.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-15
Probability Decision Tree: Surface Water Systems Serving 10,001-50,000 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
8.0
92.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.0
92.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
4.0
48.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
8.0
8.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-16
Probability Decision Tree: Surface Water Systems Serving 50,001-100,000 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4) and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
5.0
85.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
5.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
85.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
5.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
42.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
12.0
12.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Exhibit A-17
Probability Decision Tree: Surface Water Systems Serving 100,001-1,000,000 People
No.
Treatment Technology Train
Percent of Treatment Required to
Achieve MCL
<50% 50-90% >90%
1 Regionalization
2 Alternate Source
3 Modify Lime Softening and pre-oxidation
4 Modify Coagulation/Filtration and pre-oxidation
5 Anion Exchange (25 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
6 Anion Exchange (150 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
7 Anion Exchange (25 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
8 Anion Exchange (150 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
9 Activated Alumina (2,000 BV) and non-hazardous landfill (forspent media) and pre-oxidation
10 Activated Alumina (10,000 BV) and non-hazardous landfill (for spent media) and pre-oxidation
11 Reverse Osmosis and direct discharge waste disposal and corrosion control and pre-oxidation
12 Reverse Osmosis and POTWwaste disposal and corrosion control and pre-oxidation
13 Reverse Osmosis and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
14 Coagulation Assisted Microfiltration and mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
15 Coagulation Assisted Microfiltration and non-mechanical de-watering/non-hazardous landfill waste disposal and pre-oxidation
16 Oxidation Filtration (Greensand) and POTW for backwash stream
17 Anion Exchange (25 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
18 Anion Exchange (150 mg/l SO4) and chemical precipitation/non-hazardous landfill and corrosion control and pre-oxidation
19 Activated Alumina (2,000 BV) and POTW/non-hazardous landfill and pre-oxidation
20 Activated Alumina (10,000 BV) and POTW/non-hazardous landfill and pre-oxidation
21 Anion Exchange (90 mg/l SO4)and POTWwaste disposal and corrosion control and pre-oxidation
22 Anion Exchange (90 mg/l SO4) and evaporation pond/non-hazardous landfill and corrosion control and pre-oxidation
23 POE Activated Alumina and corrosion control and pre-oxidation
24 POU Reverse Osmosis and pre-oxidation
25 POU Activated Alumina and pre-oxidation
0.0
0.0
5.0
94.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
5.0
94.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
47.0
16.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
16.0
0.0
0.0
0.0
10.0
9.0
0.0
0.0
0.0
Sum of Probabilities:
100.00
100.00
100.00
-------
Appendix B: Bladder Cancer Risk Analysis
B.1 Community Water Systems
In order to calculate the number of bladder cancer cases avoided due to each regulatory option,
EPA developed a Monte-Carlo based risk model. This model is summarized in Chapter 5. This
appendix provides a more detailed description of the risk analysis, including the assumptions and
calculations used in the analysis.
The following sections explain how we calculated risk reductions for populations exposed to
arsenic concentrations of 3 //g/L and above in CWSs. First, the data used in the analysis will be
presented. Second, the calculations used in the analysis will be explained. Finally, the results of
the analysis will be explained.
B.1.1 Data Inputs
The inputs into a Monte-Carlo analysis can be separated into two categories, those data that
describe variation across a population, and those data that describe uncertainty in the underlying
population values. In the CWS risk analysis, two data inputs have elements of uncertainty, water
consumption and the risk distributions (lifetime risk at 50 //g/L, assuming consumption of 2 L per
day, and 70 kg body weight). In order to best understand the results of the analysis, it is
important to separate out the effects of variation and uncertainty.
This is accomplished in the following manner. Two different distributions for water consumption
are used. As described in the next section, one includes "total water consumption" and one
includes "community water consumption." The former captures the upper bound of water
consumption, and the latter captures the lower bound. Likewise, two different distributions for
lifetime risk are used. As described below, the first captures the lower bound of lifetime risk
(mean 0.731/1,000 people, 95% upper confidence limit 0.807/1,000 people) and the second
captures the upper bound of lifetime risk (mean 1.237/1,000 people, 95% upper confidence limit
1.548/1,000 people). Therefore, throughout the analysis, two scenarios are carried out. The
"Lower Bound" scenario uses the lower bound water consumption and lifetime risk assumptions.
The "Upper Bound" scenario uses upper bound water consumption and lifetime risk assumptions.
The range of results within one scenario represent the variation within the population. The
difference in results between the two scenarios captures the uncertainty in the risk analysis.
B.1.1.1 Water Consumption
EPA recently updated its estimates of personal (per capita) daily average estimates of water
consumption (Estimatedper Capita Water Consumption in the United States, external review
draft, EPA 1999). The estimates used data from the combined 1994, 1995, and 1996 Continuing
Survey of Food Intakes by Individuals (CSFII), conducted by the U.S. Department of Agriculture
Appendix B, Bladder Cancer Risk Analysis B-l Proposed Arsenic in Drinking Water Rule RIA
-------
(USD A). The CSFII is a complex, multistage area probability sample of the entire U.S. and is
conducted to survey the food and beverage intake of the U.S. Estimates of water consumed
include "Community Water" and "Total Water." Community Water includes water consumed
directly as a beverage as well as water added to foods and beverages during final preparation at
home or by food service establishments such as school cafeterias and restaurants. "Total"
includes Community Water plus bottled water (see Exhibit B-l).
Exhibit B-1
Source of Water Consumed
Source
Community
Water
Total
Water
Direct
(drinking)
X
X
Indirect
(from food and beverages)
X
X
Bottled water
X
Water consumption estimates for selected subpopulations in the U.S. are described in the analysis,
including per capita water consumption by source for gender, region, age categories, economic
status, race, and residential status and separately for pregnant women, lactating women, and
women in childbearing years. The water consumption estimates by age and sex were used in the
computation of bladder cancer cases avoided, and are shown in Exhibits B-2 through B-21.
Exhibit B-2
Water Consumption (ml) Male Less Than 1 Year Old (Lower Bound)
Consumption Male <1 LB
u
^
u
1.00
321.00
641.00
00
1,281.00
Appendix B, Bladder Cancer Risk Analysis
B-2
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-3
Water Consumption (ml) Male 1-10 Years Old (Lower Bound)
Consumption Male 1-10 LB
ra
.0
O
u
u
GC
Exhibit B-4
Water Consumption (ml) Male 11-19 Years Old (Lower Bound)
Consumption Male 11-19 LB
-a
n
.o
a
u
u
GC
7.00
1,001.25
1,995.50
2,989.75
3,984.00
Appendix B, Bladder Cancer Risk Analysis
B-3
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-5
Water Consumption (ml) Male 20-64 Years Old (Lower Bound)
ra
.0
O
u
u
GC
Consumption Male 20-64 LB
1,302.00
2,592.00
3,882.00
5,172.00
Exhibit B-6
Water Consumption (ml) Male Over 64 Years Old (Lower Bound)
Consumption Male 65+ LB
ra
.a
o
u
u
919.50
1,809.00
2,698.50
3,588.00
Appendix B, Bladder Cancer Risk Analysis
B-4
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-7
Water Consumption (ml) Male Under 1 Year Old (Upper Bound)
Consumption Male <1 UB
-a
n
u
^
u
DC
7.00
397.00
787.00
1,177.00
1,567.00
Exhibit B-8
Water Consumption (ml) Male 1-10 Years Old (Upper Bound)
Consumption Male 1-10 UB
ra
.a
o
u
u
Appendix B, Bladder Cancer Risk Analysis
B-5
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-11
Water Consumption (ml) Male Over 64 Years Old (Upper Bound)
Consumption Male 65+ UB
ra
.0
O
u
u
GC
1,117.50
2,011.00
2,904.50
3,798.00
Exhibit B-10
Water Consumption (ml) Male 20-64 Years Old (Upper Bound)
Consumption Male 20-64 UB
-a
n
.o
a
CL
u
u
GC
65.00
Appendix B, Bladder Cancer Risk Analysis
B-6
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-12
Water Consumption (ml) Female Under 1 Year Old (Lower Bound)
ra
.a
o
u
u
Consumption Female <1 LB
Appendix B, Bladder Cancer Risk Analysis
B-7
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-13
Water Consumption (ml) Female 1-10 Years Old (Lower Bound)
u
^
u
DC
6.00
Consumption Female 1-10 LB
508.00
1,010.00
1,512.00
2,014.00
Exhibit B-14
Water Consumption (ml) Female 11-19 Years Old (Lower Bound)
Consumption Female 11-19 LB
-a
n
u
^
u
7.00
654.75
1,302.50
1,950.25
2,598.00
Appendix B, Bladder Cancer Risk Analysis
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-15
Water Consumption (ml) Female 20-64 Years Old (Lower Bound)
Consumption Female 20-64 LB
-a
n
u
^
u
DC
11.00
Exhibit B-16
Water Consumption (ml) Female Over 64 Years Old (Lower Bound)
Consumption Female 65+ LB
ra
.a
o
u
u
936.25
1,856.50
2,776.75
3,697.00
Appendix B, Bladder Cancer Risk Analysis
B-9
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-17
Water Consumption (ml) Female Under 1 Year Old (Upper Bound)
ra
.0
O
u
u
GC
Consumption Female <1 UB
Exhibit B-18
Water Consumption (ml) Female 1-10 Years Old (Upper Bound)
Consumption Female 1-10 UB
ra
.a
o
u
u
Appendix B, Bladder Cancer Risk Analysis
B-10
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-19
Water Consumption (ml) Female 11-19 Years Old (Upper Bound)
Consumption Female 11-19 UB
us
j=
o
CL
u
u
GC
23.00
788.00
1,553.00
2,318.00
3,083.00
Appendix B, Bladder Cancer Risk Analysis
B-ll
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-20
Water Consumption (ml) Female 20-64 Years Old (Upper Bound)
I
u
Consumption Female 20-64 UB
Appendix B, Bladder Cancer Risk Analysis
B-12
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-21
Water Consumption (ml) Female Over 64 Years Old (Upper Bound)
Consumption Female 65+ UB
ra
.0
O
u
u
GC
B.1.1.2 Body Weight
In this analysis, EPA used body weight data from the U.S. census (DOC, 1999). The body weight
data included a mean and a distribution of weight for male and females on a year-to-year basis
throughout a lifetime (70 years). These data were used to develop statistical distributions of body
weight for each sex and age category included in the water consumption study discussed above.
The body weight distributions are provided in Exhibits B-22 through B-31.
Appendix B, Bladder Cancer Risk Analysis
B-13
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-22
Body Weight (kg) Male Less Than 1 Year Old
Normal distribution with parameters:
Mean 9.4
Standard Dev. 1.3
male wt<1
Exhibit B-23
Body Weight (kg) Male 1-10 Years Old
Normal distribution with parameters:
Mean 22.1
Standard Dev. 1.4
00
1 to 10
17.90 20.00 22.10 24.20
Appendix B, Bladder Cancer Risk Analysis B-14 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-24
Body Weight (kg) Male 11-19 Years Old
Normal distribution with parameters:
Mean 58.5
Standard Dev. 3.8
11 to 19
52.80 58.50 64.20 69.90
Exhibit B-25
Body Weight (kg) Male 20-64 Years Old
Normal distribution with parameters:
Mean 79.1
Standard Dev. 6.3
00
J3
O
CL
60.20
20 to 64
79.10 88.55
98.00
Appendix B, Bladder Cancer Risk Analysis B-15 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-26
Body Weight (kg) Male Over 64 Years Old
Normal distribution with parameters:
Mean 74.8
Standard Dev. 12.8
oo
J3
O
CL
36.40
65 & over
55.60
74.80
94.00
113.20
Exhibit B-27
Body Weight (kg) Female Under 1 Year Old
Normal distribution with parameters:
Mean 8.8
Standard Dev. 1.2
<1
Appendix B, Bladder Cancer Risk Analysis
B-16
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-28
Body Weight (kg) Female 1-10 Years Old
Normal distribution with parameters:
Mean 21.6
Standard Dev. 1.5
oo
J3
O
CL
1to10
17.10 19.35
23.85 26.10
Exhibit B-29
Body Weight (kg) Female 11-19 Years Old
Normal distribution with parameters:
Mean 53.8
Standard Dev. 3.6
11 to 19
43.00 48.40 53.80 59.20
Appendix B, Bladder Cancer Risk Analysis B-l 7 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-30
Body Weight (kg) Female 20-64 Years Old
Normal distribution with parameters:
Mean 65.8
Standard Dev. 7
co
J3
O
CL
44.80
20 to 64
55.30
65.80
76.30
Exhibit B-31
Body Weight (kg) Female Over 64 Years Old
Normal distribution with parameters:
Mean 66.6
Standard Dev. 13.8
CO
-a
o
it
65 & over
Appendix B, Bladder Cancer Risk Analysis
B-18
Proposed Arsenic in Drinking Water Rule RIA
-------
B.1.1.3 Lifetime Risk of Bladder Cancer
In its 1999 report, Arsenic in Drinking Water, the National Research Council (NRC) analyzed
bladder cancer risks using data from Taiwan. In addition, NRC examined evidence from human
epidemiological studies in Chile and Argentina, and concluded that risks of bladder and lung
cancer were comparable to those "in Taiwan at comparable levels of exposure (NRC 1999, page
7)." The NRC also examined the implications of applying different mathematical procedures to
the newly available Taiwanese data for the purpose of characterizing bladder cancer risk.
These risk distributions are based on bladder cancer mortality data in Taiwan, in a section of
Taiwan where arsenic concentrations in the water are very high by comparison to those in the
U.S. It is also an area of very low incomes and poor diets, and the availability and quality of
medical care is not of high quality, by U.S. standards. In its estimate of bladder cancer risk, the
Agency assumed, that within the Taiwanese study area, the risk of contracting bladder cancer was
relatively close to the risk of dying from bladder cancer (that is, that the bladder cancer morbidity
rate was equal to the bladder cancer mortality rate).1
In the NRC report, two tables are provided which provide six lifetime risk distributions at the
current MCL of 50 ug/L, which EPA's senior scientist feel are appropriate to use in this risk
analysis. Table 10-11 of the report shows three risk estimates based on the Taiwanese male
bladder cancer data, using a Poisson regression model (the models differ based on assumptions
regarding the availability of baseline data about the exposed population). The means of the risk
distributions are 0.731, 0.911, and 1.049 per 1,000 people. Table 10-12 of the report presents
excess lifetime risk estimates at the current MCL of 50 ug/L for bladder cancer in males
calculated using EPA's 1996 proposed revisions to the cancer guidelines (EPA 1996) (the
models differ based on assumptions regarding the availability of baseline data about the exposed
population). The means of these distributions are: 1.237, 1.129, and 1.111 cases per 1,000
people.
EPA has no means by which to choose one of these lifetime risk estimates over the other,
therefore, as mentioned above, the Agency has chosen to estimate the range of uncertainly
concerning the lifetime risk estimates. This was done by using the lowest lifetime risk estimate in
the "Lower Bound" scenario, and using the highest lifetime risk estimate in the "Upper Bound"
scenario. These risk estimates are shown in Exhibits B-32 And B-33 respectively.
1 We do not have data on the rates of survival for bladder cancer in the Taiwanese villages in the study
and at the time of data collection. We do know that the relative survival rates for bladder cancer in developing
countries overall ranged from 23.5% to 66.1 % in 1982-1992 ("Cancer Survival in Developing Countries,"
International Agency for Research on Cancer, World Health Organization, Publication No. 145, 1998). The age-
adjusted annual incidence rates of bladder cancer in Taiwan in 1996 for males and females, respectively, were 7.36
and 3.09 per 100,000, with corresponding annual mortality rates of 3.21 and 1.44 per 100,000 (correspondence
from Chen to Herman Gibb, January 3, 2000).
Appendix B, Bladder Cancer Risk Analysis B-19 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-32
Lifetime Bladder Cancer Risk at 50 /^g/L per 1,000 people (Lower Bound)
Normal distribution with parameters:
Mean 0.731
Standard Dev. 0.0388
o
it
0.61
Lifetime Risk/1000 LB
0.67
0.73
0.79
0.85
Exhibit B-33
Lifetime Bladder Cancer Risk at 50 /^g/L per 1,000 People (Upper Bound)
Normal distribution with parameters:
Mean 1.237
Standard Dev. 0.159
O
CL
0.76
Lifetime Risk/1000 UB
1.00
1.24
1.71
Appendix B, Bladder Cancer Risk Analysis
B-20
Proposed Arsenic in Drinking Water Rule RIA
-------
B.1.1.4 Arsenic Occurrence Estimates
EPA used statistical techniques to assess the national distribution of mean arsenic concentrations
in water systems. A detailed explanation of the occurrence data used in the risk analysis can be
found in the Occurrence and Exposure Document for the Arsenic in Drinking Water Rule (EPA,
1999). For each MCL under consideration, the Post-Compliance Exposure Distribution at
different concentrations (|ig/L) is provided in Exhibits B-34 to B-38 for ground water and
Exhibits B-39 to B-43 for surface water. Two important caveats should be mentioned. First, the
risk analysis only applies to persons exposed to arsenic concentrations at or above 3 |ig/L. This
represents the feasible range (at concentrations below this level, arsenic concentrations can not be
accurately measured). Second, for each MCL, it is assumed that all systems will provide finished
water with a maximum concentration equal to 80 percent of the MCL. This assumption is
consistent with industry practice, and the same assumption was made in the calculation of the
proposed rule's costs.
Exhibit B-34
Post-Compliance Exposure Distribution (/^g/L) at MCL= 50
(Ground Water)
GW Occurrence No Treatment
ra
.o
o
u
u
Appendix B, Bladder Cancer Risk Analysis
B-21
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-35
Post-Compliance Exposure Distribution (/^g/L) at MCL= 20
(Ground Water)
GW Occurance MCL=20
e
0_
u
3.00
7.25
11.50
15.75
20.00
Exhibit B-36
Post-Compliance Exposure Distribution (/^g/L) at MCL= 10
(Ground Water)
GW Occurence MCL=10
ra
.a
o
u
u
DC
Appendix B, Bladder Cancer Risk Analysis
B-22
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-37
Post-Compliance Exposure Distribution (/^g/L) at MCL= 5
(Ground Water)
GW occurence MCL=5
£
la
ra
0
u
.>
"u
DC
3.000 3.500 4.000 4.500 5.000
Exhibit B-38
Post-Compliance Exposure Distribution (/^g/L) at MCL= 3
(Ground Water)
GW Occurence MCL=3
u
u
2.399
2.400
2.400
2.401
2.401
Appendix B, Bladder Cancer Risk Analysis
B-23
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-39
Post-Compliance Exposure Distribution (/^g/L) at MCL= 50
(Surface Water)
E37
ra
.0
O
u
u
GC
14.75 26.50 38.25 50.00
Exhibit B-40
Post-Compliance Exposure Distribution (/^g/L) at MCL= 20
(Surface Water)
SW Post treatment occ @ 20
u
"
u
GC
Appendix B, Bladder Cancer Risk Analysis B-24 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-41
Post-Compliance Exposure Distribution (/^g/L) at MCL= 10
(Surface Water)
SW Post treatment occ @ 10(feasible)
ra
.0
O
u
u
GC
4.75
6.50
8.25
10.00
Exhibit B-42
Post-Compliance Exposure Distribution (/^g/L) at MCL= 5
(Surface Water)
SW Post treatment occ @ S(feasible)
-a
n
u
u
3.00
Appendix B, Bladder Cancer Risk Analysis
B-25
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-43
Post-Compliance Exposure Distribution (/^g/L) at MCL= 3
(Surface Water)
SW Post treatment occ @ 3
u
^
u
DC
2.40
2.40
2.40
2.40
2.40
B.1.1.5 CWS Population Affected
As mentioned above, the target population for the CWS risk analysis is people exposed to arsenic
at levels at or above 3 //g/L, as this is considered the feasible level. Exhibit B-44 provides the
population served by both surface water and ground water CWSs. In addition, the percentage of
people served that are exposed at arsenic levels above 3 //g/L is provided. Ground water systems
with finished water concentration above the minimum feasible level serve 16.5 million people,
while surface water systems with finished water above 3 //g/L serve 10.1 million people, for a
total exposed population of 26.7 million people. Systems serving overl million people were
identified from a system-by-system data collection.
Appendix B, Bladder Cancer Risk Analysis
B-26
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-44
Total Population Served by Water Systems
By Source Water, System Type and Service Population Category
Service
Population
Category
< 100
101-500
501-1,000
1,001-3,300
3,301-10,000
10,001-50,000
50,001-100,000
100,001-1,000,000
> 1,000,000
Total Pooulation Served
Percentage of Population Over 3^g/L
Population Served Over 3^g/L By Source Type
Total Population Served Over 3^g/L
SOURCE PERCENTAGE (Percentage of Total
Population Served Over 3wQ/U
Community
Groundwater
859,777
3,741,017
3,457,163
10,631,422
14,095,015
25,004,779
8,609,455
14,575,556
2,855,494
83.829.678
19.70%
16,514,447
Surface Water
61,450
570,448
921,449
4,797,855
10,995,980
36,819,575
20,500,370
65,375,183
28,658,586
168.700.896
6.01%
10,138,923
26,653,370
62%
38%
Source: Safe Drinking Water Information System (SDWIS), December 1998 freeze.
In addition, since the daily water consumption and body weight associated with each age category
varies by sex, it is necessary to know the percentage of the population made up by each sex. In
this risk analysis, it is assumed that 51.9 percent of the population is female, and 48.1 percent is
male (DOC, 1999).
B.1.2 CWS Risk Model
The CWS risk analysis is a Monte-Carlo based simulation model. This section will explain each
step is the simulation. The Monte-Carlo simulation is conducted at each MCL option (50, 20, 10,
5 and 3 //g/L). In addition, for each MCL option, the simulation is carried out for both the
"Lower Bound" and "Upper Bound" scenarios discussed above. Therefore, the simulation model
is carried out ten times. Each of these ten "runs" of the model is independent of the other, and can
be discussed in isolation. Therefore, this section will include a generalized discussion of the
model. The inputs that are used will depend on the MCL option and scenario being evaluated. It
is important not to confuse a "run" of the model" as just described, with a model iteration. Each
Appendix B, Bladder Cancer Risk Analysis
B-27
Proposed Arsenic in Drinking Water Rule RIA
-------
run of the model consists of 10,000 iterations. Within a single iteration, the model pulls a value
for each variable from its input distribution (e.g. body weight) and calculates a value for each
output variable (e.g. lifetime risk). This is done 10,000 times for each model run. The results of
the model run is the distribution of the 10,000 values for each output variable.
The first step of each iteration is to calculate the relative exposure factor for each sex and age
category. As shown in the following equation, the relative exposure factor is a function of daily
water consumption and body weight.
where;
REF = relative exposure factor
C = daily water consumption (L)
W = body weight (kg)
i = model iteration number
a = age category
m = male
f = female
Next, lifetime relative exposure factors are calculated for each sex using the relative exposure
factors and information regarding the number of years in each age catagory (e.g. a person spends
nine years in the 1-10 year old age category).
i= XREFmai*Na 770
LREFfi= £REFfai*Na 770
^ a '
where;
LREF = lifetime relative exposure factor
N = number of years in age category
The model then determines, for this iteration, if the modeled individual is a male or a female. This
Appendix B, Bladder Cancer Risk Analysis B-28 Proposed Arsenic in Drinking Water Rule RIA
-------
is done using a random number generator and information concerning the sex distribution of the
population.
fLREF , if RNi < MP
LREF =
1 I LREFfl otherwise
where;
RN., = random number from 0 to 1
MP = percentage of the population that is male (48.1%)
For each iteration, the model assumes that the individual is consuming water from either a surface
water system or a ground water system. If the individual is assumed to be associated with a
surface water system, then the arsenic concentration is chosen from the surface water
concentration distribution; likewise, if the person is assumed to be associated with a ground water
CWS, the arsenic concentration is chosen from the ground water concentration distribution.
ASai if RN2 < GP
AS = '
AS otherwise
where;
AS= concentration of arsenic
RN2= random number from 0 to 1
GP= percentage of the population served by ground water (62%)
Finally, the lifetime risk of bladder cancer associated with this iteration is calculated.
LR; = (RF; * LREF, * AS; / 50) * 10
where;
LR= lifetime risk of bladder cancer per 10,000 people
RF= lifetime risk of bladder cancer per 1,000 people, assuming
consumption of 2 liters per day, a body weight of 70 kg, and an
arsenic concentration of50^g/L.
In order to calculate the expected number of cancer cases associated with the model run, the
mean lifetime risk is multiplied by the exposed population, as follows:
Appendix B, Bladder Cancer Risk Analysis B-29 Proposed Arsenic in Drinking Water Rule RIA
-------
CA =
N
. r
U0,000y
where;
CA = expected number of bladder cancer cases
P = population
N = number of iterations
B.1.3 CWS Bladder Cancer Risk Estimates
Exhibits B-45 through B-54 provide the results of the CWS risk analysis. These exhibits provide
the mean and standard deviation of expected cases of bladder cancer per 10,000 people, as well as
the full distribution of expected risk values. The results for each MCL option are provided for
both the Lower and the Upper Bound Scenarios.
As mentioned above, these results can be used to calculate the expected number of bladder
cancers nationwide by multiplying the mean expected risk associated with an MCL option by the
exposed population (in 10,000s). Likewise, the population exposed to a certain risk threshold
(e.g. 10"4), at a given an MCL, can be obtained by using the distribution of expected risk values.
These results are provided in Chapter 5 of this RIA.
Appendix B, Bladder Cancer Risk Analysis
B-30
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-45
CWS Number of Cancer Cases per 10,000 people
Baseline Under Lower Bound Scenario
Stat
is tics:
Mean
S ta n d a rd D e v ia tio n
1 0,000 Tria
.056 -
£
a
a
e
Q. -014
Value
0.85
1.06
Forecast: Ind. Lifetime Risk/10000 Untreated LB
Is Frequency Chart 2
I
0.00
Illlllllllllhl .,
4
1.00 2.00 3.00 4.00
X10-4
09 Outliers
- 556
Tl
n
n
3
139 <5
- 0
Exhibit B-46
CWS Number of Cancer Cases per 10,000 people
Baseline Under Upper Bound Scenario
Stat
sties:
Mean
Standard Deviation
10,000 Tria
.050 -
49
o
Value
1.74
1.97
Forecast: Ind. Lifetime Risk/10000 Untreated UB
Is Frequency Chart 2
I
||
Illllllllllllllllll h,l. ,,...
85 Outliers
- 504
"*
n
- 252 -D
C
n
3
126 5
0.00 1.75 3.50 5.25 7.00
X10-4
Appendix B, Bladder Cancer Risk Analysis
B-31
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-47
CWS Number of Cancer Cases per 10,000 people
MCL = 20 yug/L Under Lower Bound Scenario
Stat
is tics:
Mean
S ta n d a rd D e v ia tio n
1 0,000 Tria
.035 -
£
-------
Exhibit B-49
CWS Number of Cancer Cases per 10,000 people
MCL = 10 Mg/L Under Lower Bound Scenario
Stat
is tics:
Mean
S ta n d a rd D e v ia tio n
1 0,000 Tria
.030 -
£
a
a
e
Qi_ -008
Is
Value
0.57
0.37
Forecast: Ind. Lifetime Risk/10000 MCL=10 LB
Frequency Chart 1
0.00
I
0.44
I
Illlllilllll,!!!
4
0.88 1.31 1.75
X10-4
35 Outliers
- 303
Tl
n
n
3
7575 <5
- 0
Exhibit B-50
CWS Number of Cancer Cases per 10,000 people
MCL = 10 yug/L Under Upper Bound Scenario
Stat
sties:
Mean
S ta n d a rd D e v ia tio n
10,000 Tria
.031 -
£
a
.a
e
Qi_ -008 -
Value
1.19
0.72
Forecast: Ind. Lifetime Risk/10000 MCL=10 UP
Is Frequency Chart 1
,lll
ilii
Illllli
0.00 0.88 1.75
X10-4
111
Illlllllllllll Mil.... .
4
2.63 3.50
28 Outliers
- 314
- 235.5
Tl
n
157 -g
n
3
- 78.5 ^
- 0
Appendix B, Bladder Cancer Risk Analysis
B-33
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-51
CWS Number of Cancer Cases per 10,000 people
MCL = 5 fjgIL Under Lower Bound Scenario
Stat
is tics:
Mean
S ta n d a rd D e v ia tio n
1 0,000 Tria
.029 -
-------
Exhibit B-53
CWS Number of Cancer Cases per 10,000 people
MCL = 3 jug/L Under Lower Bound Scenario
Stat
sties:
Mean
Standard Deviation
1 0,000 Tria
.026 -
TO
.O
e
Value
0.22
0.12
Forecast: Ind. Lifetime Risk/10000 MCL=3 LB
Is Frequency Chart 2
0.00
i
III
Dlhiiiiiii
Illllllllllllllhil
4
0.14 0.28 0.41 0.55
X10-4
53 Outliers
- 255
- 191.2
ft
n
- 63.75 ^
- 0
Exhibit B-54
CWS Number of Cancer Cases per 10,000 people
MCL = 3 jug/L Under Upper Bound Scenario
Stat
sties:
Mean
S tandard Dev ia tio n
10,000 Tria
.027 -
a
0
Va lue
0.47
0.22
Forecast: Ind. Lifetime Risk/10000 MCL=3 UB
s Frequency Chart
I
nil
,
0.00 0.28 0.55
L
Illlllllll 1
Illllllllllllllllllllllllllll
0.83 1.10
203 Outliers
- 267
- 200.2
Tl
n
- 133.5 J=
e
n>
- 66.75 Q
- 0
Appendix B, Bladder Cancer Risk Analysis
B-35
Proposed Arsenic in Drinking Water Rule RIA
-------
B.2 Non-Transient Non-Community Water Systems
B.2.1 Data Inputs
Most of the data described above under the CWS risk model is also used in the NTNC risk model.
This includes water consumption, body weight, and lifetime risk estimates (at 50 //g/L, 2 L
consumption, and 70 kg body weight). Also, the ground water arsenic concentrations at each
MCL used in the CWS risk model are used in the NTNC risk model.
B.2.1.1 NTNC Service Categories, Population and Exposure Time
The main differences between the CWS and NTNC risk models are how population is distributed
among the different types of establishments that make up the NTNC category of systems, and the
extent to which the worker and customer populations within a service category are exposed to
arsenic (both in terms of length of exposure and drinking water consumed).
In addition to the CWS data already discussed, Exhibits B-55 and B-56 provide all of the data
inputs necessary to model the bladder cancer risk associated with NTNC systems. First, note that
in Exhibit B-55, the NTNC universe has been divided into 35 service categories. This was
accomplished using the system descriptions in SDWIS (EPA, 1999b). For each service category,
the total number of NTNCs and the population served by these NTNCs is taken from SDWIS.
The population served by each NTNC often varies daily; the SDWIS population numbers are
interpreted to mean the peak population served (both workers and customers).
The next data field in Exhibit B-55 is the number of customer cycles per year, or the number of
times each year the customer base turns over. For example, if this parameter equals one, then the
same customer's are served each day. If the value is seven, then seven sets of customers use the
facility. The next field is the number of workers per person per day. For example, if the value is
0.1, as in the case of summer camps, then 10 percent of the peak population served (from
SDWIS) is assumed to be workers. Both the number of customer cycles per year assumptions
and workers per person per day data assumptions were made after investigating numerous data
sources, including trade-journals and trade association information.
The next set of data fields in Exhibit B-55 are assumptions about the characteristics of the
workers in each service type. The percent of workers' daily consumption is the percentage of
drinking water consumed on a work day that is consumed at work. This value is assumed to be
either 50 percent or 100 percent, depending on the service category. The number of days a person
works is assumed to be 250 for all service categories. The number of years a person works at the
NTNC establishment is assumed to be either 40 or 10, depending on the service category.
Information regarding customer behavior is provided in the next set of data fields in Exhibit B-55.
The percent of customers' daily consumption is the percentage of total drinking water consumed
on a day that the customer visits the NTNC, that is consumed at the NTNC. This value is
assumed to be either 25 percent, 50 percent or 100 percent, depending on the service category.
Appendix B, Bladder Cancer Risk Analysis B-36 Proposed Arsenic in Drinking Water Rule RIA
-------
The number of days a customer visits the NTNC is provided for each service category. For
example, the value for nursing homes of 365 indicates that nursing home customers are served by
the nursing home year round, while the value for churches of 52 indicates that churches are
assumed to serve their customers once per week. The number of years a person is assumed to
visits each service category is also provided.
Finally, the total exposed worker and customer populations for each service category are
provided in Exhibit B-55. These numbers are calculated as follows:
= (PC*CCC)*(1-WPC)
TW = P*WP
where:
TC= total number of customers
TW= total number of workers
P= SDWIS population
WP= workers per person per day
CC= number of customer cycles per year
c= NTNC service category
Exhibit B-56 provides the final set of data required to estimate bladder cancer risk from NTNCs.
The percent of worker lifetime exposure is the percent of lifetime water consumption which is
consumed at the NTNC by a worker. The percent of customer lifetime exposure is the percent of
lifetime water consumption consumed at the NTNC by a customer. These numbers are calculated
as follows:
PWLE- 365*70
PCDC *DC *YC
PCLE,, =
365*70
where;
PWLE =percent of worker lifetime exposure
PCLE =percent of customer lifetime exposure
PWDC =percentage of workers daily consumption
PCDC =percentage of customers daily consumption
DW =worker days per year
DC =customer days per year
YW =worker years
YC =customer years
Appendix B, Bladder Cancer Risk Analysis B-37 Proposed Arsenic in Drinking Water Rule RIA
-------
Returning to Exhibit B-56, the worker age bracket is the age range (corresponding to the age
ranges used in the CWS risk analysis) that a NTNC worker is assumed to fall in. For all service
categories, the worker age bracket is assumed to be 20-64 years of age. The customer age
bracket is the age range (corresponding to the age ranges used in the CWS risk analysis) that a
NTNC customer is assumed to be in. For most service categories, the customer age bracket is
assumed to be 0-70 years of age (all ages). However, certain service categories only serve certain
age groups (e.g. nursing homes and schools), therefore more specific age ranges are assumed.
Appendix B, Bladder Cancer Risk Analysis B-38 Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-55
NTNC Population and Exposure Time
Data
Water Wholesalers
Mobile Home Parks
Nursing Homes
Churches
Golf and Country Clubs
Retailers (Food related)
Retailers (Non-food related)
Restaurants
Hotels/Motels
Prisons/Jails
Service Stations
Agricultural Products/Services
Daycare Centers
Schools
State Parks
Medical Facilities
Campgrounds/RV Parks
Federal Parks
Highway Rest Areas
Misc. Recreation Services
Forest Service
Interstate Carriers
Amusement Parks
Summer Camps
Airports
Military Bases
Non-Water Utilities
Office Parks
Manufacturing: Food
Manufacturing: Non-Food
Landfills
Fire Departments
Construction
Mining
Migrant Labor Camps
Number
of
Systems
266
104
130
230
116
142
695
418
351
67
53
368
809
8,414
83
367
123
20
15
259
107
287
159
46
101
95
497
950
768
3,356
78
41
99
119
33
Total
SDWIS
Population
66,018
19,240
13,910
1 1 ,500
11,716
45,724
120,930
154,660
46,683
121,940
12,190
27,968
61 ,484
3,086,012
106,895
163,631
19,680
780
6,105
22,533
4,494
35,301
76,462
6,716
326,860
67,525
84,490
181,600
285,696
588,792
3,432
4,018
5,247
13,447
2,079
Number of
Customer
Cycles/Year
1.00
1.33
1.00
1.00
4.50
2.00
4.50
2.00
86.00
1.33
7.00
7.00
1.00
1.00
26.00
16.40
22.50
26.00
50.70
26.00
26.00
93.00
90.00
8.50
36.50
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Worker
Per
Person
Per Day
0
0.046
0.23
0.01
0.11
0.07
0.09
0.07
0.27
0.1
0.06
0.125
0.145
0.073
0.016
0.022
0.041
0.016
0.01
0.016
0.016
0.304
0.18
0.1
0.308
1
1
1
1
1
1
1
1
1
1
Percent of
Worker's
Daily
Consumption
n/a
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
100.0%
50.0%
50.0%
100.0%
50.0%
50.0%
50.0%
50.0%
50.0%
50.0%
100.0%
100.0%
100.0%
100.0%
100.0%
Worker
Days Per
Year
n/a
250
250
250
250
250
250
250
250
250
250
250
250
200
250
250
180
250
250
250
250
250
250
180
250
250
250
250
250
250
250
250
250
250
250
Worker
Years
n/a
40
40
40
40
40
40
40
40
40
40
40
10
40
40
40
40
40
40
40
40
40
10
10
40
40
40
40
40
40
40
40
40
40
40
Percent of
Customer's
Daily
Consumption
25.0%
100.0%
100.0%
50.0%
50.0%
25.0%
25.0%
25.0%
100.0%
100.0%
25.0%
25.0%
50.0%
50.0%
50.0%
100.0%
100.0%
50.0%
50.0%
100.0%
100.0%
50.0%
50.0%
100.0%
25.0%
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Customer
Days Per
Year
270.00
270.00
365.00
52.00
52.00
185.00
52.00
185.00
3.40
270.00
52.00
52.00
250.00
200.00
14.00
6.70
5.00
14.00
7.20
14.00
14.00
2.00
1.00
7.00
10.00
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Customer
Years
70.00
35.00
10.00
70.00
70.00
70.00
70.00
70.00
40.00
3.00
54.00
50.00
5.00
12.00
70.00
10.30
50.00
70.00
70.00
70.00
50.00
70.00
70.00
10.00
70.00
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Total
Worker
Population
0
885
3,199
115
1,289
3,201
10,884
10,826
12,604
12,194
731
3,496
8,915
225,279
1,710
3,600
807
12
61
361
72
10,732
13,763
672
100,673
67,525
84,490
181,600
285,696
588,792
3,432
4,018
5,247
13,447
2,079
Total
Customer
Population
66,018
24,412
10,711
1 1 ,385
46,923
85,047
495,208
287,668
2,930,759
145,962
80,210
171,304
52,569
2,860,733
2,734,802
2,624,510
424,645
19,956
306,428
576,484
114,974
2,284,963
5,642,896
51 ,377
8,255,830
0
0
0
0
0
0
0
0
0
0
Subtotal = 1,662,407 30,305,774
TOTAL = 31,968,181
-------
Exhibit B-56
NTNC Percent of Lifetime Exposure and Age at Exposure
Water Wholesalers
Mobile Home Parks
Nursing Homes
Churches
Golf and Country Clubs
Retailers (Food related)
Retailers (Non-food related)
Restaurants
Hotels/Motels
Prisons/Jails
Service Stations
Agricultural Products/Services
Daycare Centers
Schools
State Parks
Medical Facilities
Campgrounds/RV Parks
Federal Parks
Highway Rest Areas
Misc. Recreation Services
Forest Service
Interstate Carriers
Amusement Parks
Summer Camps
Airports
Military Bases
Non- Water Utilities
Office Parks
Manufacturing: Food
Manufacturing: Non-Food
Landfills
Fire Departments
Construction
Mining
Migrant Labor Camps
Percent
of Worker
Lifetime
Exposure
0.00%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
4.89%
15.66%
19.57%
19.57%
14.09%
19.57%
19.57%
19.57%
39.14%
19.57%
4.89%
7.05%
19.57%
19.57%
19.57%
19.57%
19.57%
19.57%
39.14%
39.14%
39.14%
39.14%
39.14%
Percent of
Customer
Lifetime
Exposure
18.49%
36.99%
14.29%
7.12%
7.12%
12.67%
3.56%
12.67%
0.53%
3.17%
2.75%
2.54%
2.45%
4.70%
1 .92%
0.27%
0.98%
1 .92%
0.99%
3.84%
2.74%
0.27%
0.14%
0.27%
0.68%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
0.00%
Worker
Age
Bracket
n/a
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
20 to 64
all
Customer
Age
Bracket
all
all
65+
all
all
all
all
all
all
20 to 64
16 to 70
all
<5
6 to 18
all
all
all
all
all
all
all
all
all
11 to 19
all
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
Appendix B, Bladder Cancer Risk Analysis
B-40
Proposed Arsenic in Drinking Water Rule RIA
-------
B.2.2 The NTNC Risk Model
Just like the CWS risk analysis, the NTNC risk analysis is a Monte-Carlo based simulation model.
This section will explain each step is the simulation. The Monte-Carlo simulation is conducted at
each MCL option (50, 20, 10, 5 and 3 //g/L). In addition, for each MCL option, the simulation is
carried out for both the "Lower Bound" and "Upper Bound" scenarios just like in the CWS case.
Therefore, the simulation model is carried out ten times. Each of these ten "runs" of the model is
independent of the other, and can be discussed in isolation. Therefore, this section will include a
generalized discussion of the model. The inputs that are used will depend on the MCL option and
scenario being evaluated at the time. It is important not to confuse a "run" of the model" as just
described, and a model iteration. Each run of the model consists of 10,000 iterations. Within a
single iteration, the model pulls a value for each variable from its input distribution (e.g. body
weight) and calculates a value for each output variable (e.g. lifetime risk). This is done for 10,000
times for each model run. The results of the model run is the distribution of the 10,000 values for
each output variable.
The first step of each iteration is to calculate the relative exposure factor for each sex and age
category. This is done exactly as it was done in the CWS risk analysis. As shown in the
following equations, the relative exposure factor is a function of daily water consumption and
body weight.
TJ-RT7 _ * mal
mai"^J lwmj
^
where;
REF = relative exposure factor
C = daily water consumption (L)
W = body weight (kg)
i = model iteration number
a = age category
m = male
f = female
Next, the lifetime risk of bladder cancer (1/100,000 people) is calculated for workers and
customers of each sex for each service category. The next four equations, therefore are:
Appendix B, Bladder Cancer Risk Analysis B-41 Proposed Arsenic in Drinking Water Rule RIA
-------
WLRfci = PWLEci * AS ; * (RF; / 50) *
*100
WLRmd = PWLEci * AS . * (RF; / 50) *
Z(REFmai*Zac)'
*100
CLRmci = PCLEC1 * ASgl *(RFj 750)'
:100
= PCLEci*ASgl*(RV50):
*100
where;
WLR = worker lifetime risk (per 100,000 people)
CLR = customer lifetime risk (per 100,000 people)
AS = arsenic concentration Cwg/L)
RF = risk of bladder cancer at 50 ^g/L, 2 liters consumption per day,
and 70 kg body weight
2. = years spent in age category
g = ground water
The sex of the worker and customer is then chosen for the iteration to determine the worker and
customer risk for each service category:
WLRmci if RNj < MP
ci [WLRfci otherwise
CLR.,, =
CLRmci if RNj < MP
CLRfci otherwise
Appendix B, Bladder Cancer Risk Analysis B-42 Proposed Arsenic in Drinking Water Rule RIA
-------
where;
RN., = random number between 0 and 1
MP = percentage of the population that is male
Finally, the lifetime risk for the model iteration is determined by choosing among the 70
combinations of worker and customer risk over of the 35 service categories. This is accomplished
using a population weighted probability distribution. First, the total worker and customer
populations served are computed.
TC =
TW = TW
c
Next, the probability that the lifetime risk for the model iteration will be equal to the worker
lifetime risk associated with a service category is calculated:
TW
WPR =
(TW + TC)
where;
WPR =probability of choosing lifetime risk estimate for any iteration to
be equal to the lifetime risk estimate of a worker in a given
service category
Likewise, the probability that the lifetime risk for the model iteration will be equal to the customer
lifetime risk associated with a service category is calculated:
TC
CPR =
(TW + TC)
where;
CPR =probability of choosing lifetime risk estimate for any iteration to
be equal to the lifetime risk estimate of a customer in a given
service category
Appendix B, Bladder Cancer Risk Analysis B-43 Proposed Arsenic in Drinking Water Rule RIA
-------
Given these probabilities, the lifetime risk estimate for each model iteration is chosen as follows:
LR; =
WLRci with Probability WPRC
CLRci with Probability CPRC
where;
LR =l_ifetime risk (1/100,000)
In order to calculate the expected number of cancer cases associated with the model run, the
mean lifetime risk is multiplied by the exposed population as follows:
CA =
N
(TC + TW)
100,000
where;
CA = expected number of bladder cancer cases
N = number of iterations
B.2.3 NTNC Risk Results
Exhibits B-57 through B-66 provide the results of the NTNC risk analysis. These exhibits
provide the mean and standard deviation of expected cases of bladder cancer per 100,000 people,
as well as the full distribution of expected risk values. The results for each MCL option are
provided for both the Lower Bound Scenario and the Upper Bound Scenario.
As mentioned above, these results can be used to calculate the expected number of bladder cancer
cases nationwide by multiplying the mean expected risk associated with an MCL option by the
exposed population (in 100,000s). Likewise, the population exposed to a certain risk threshold
(e.g. 10"4), at a given an MCL, can be obtained by using the distribution of expected risk values.
These results are provided in Chapter 5 of this RIA.
Appendix B, Bladder Cancer Risk Analysis
B-44
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-57
NTNC Expected Number of Cancer Cases per 100,000 people
Baseline Under Lower Bound Scenario
Statistics:
Value
Mean 1.9E-01
Standard Deviation 6.6E-01
1 0,000 Tria
.342 -
robability
D ^- K
) -si (.
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC
Is Frequency Chart 1
III
> 4
O.OE+O 5.0E-1 1.0E+0 1.5E+0 2.0E+0
X10-5
51 Outliers
- 3421
T|
5
JS
c
ft
a
- 855.2 ,$3
- 0
Exhibit B-58
NTNC Expected Number of Cancer Cases per 100,000 people
Baseline Under Upper Bound Scenario
Statistics: Value
Mean 4.1E-01
Standard Deviation 1.3E + 00
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC U
10, 000 Trials Frequency Chart 162 Outliers
.321 -
t -
ro
JS
o
||
- 3214
Mnq
CD
.Q
C
CD
O.OE+O 1.0E+0 2.0E+0 3.0E+0 4.0E+0
Appendix B, Bladder Cancer Risk Analysis
B-45
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-59
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 20 jug/L Under Lower Bound Scenario
Statistics:
Value
Mean 1.6E-01
Standard Deviation 4.4E-01
10,000 Trial
.281 -
robability
i £ '
.000 -
Forecast: Ind. Lifetime Composite Risk/100K NTNC
s Frequency Chart
::::::::::::::::
174 Outliers
- 2813
m
n
.0
c
CD
- 703.2 ^
4 ~
.OE+0 3.8E-1 7.5E-1 1 .1E+0 1 .5E+0
X10-5
Exhibit B-60
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 20 jug/L Under Upper Bound Scenario
Statistics:
Mean
Standard Deviation
Value
3.2E-01
8.4E-01
Forecast: Ind. Lifetime Composite Risk/100K NTNC U
10,000 Trials Frequency Chart
199 Outliers
4
2.8E+0
Appendix B, Bladder Cancer Risk Analysis
B-46
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-61
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 10 jug/L Under Lower Bound Scenario
Stat
istics:
Mean
Standard
10,000 Tria
.207 -
Probability
o o ^ -
Value
1.2E-01
Deviation 3.3E-01
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC
Is Frequency Chart 2
"I
lllllllllllllllllMM... .
> <
D.OE+0 2.5E-1 5.0E-1 7.5E-1 1.0E+0
X10-5
22 Outliers
- 2073
Tl
n
^
c
3
- 518.2 5
- 0
Exhibit B-62
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 10 jug/L Under Upper Bound Scenario
Stat
istics:
Mean
Standard
10,000 Tria
.184 -
n
A
o
.000 -
Value
2.6E-01
Deviation 6.4E-01
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC U
Is Frequency Chart 2
\
V r i ^
5.0E+0 5.0E-1 1.0E+0 1.5E+0 2.0E+0
66 Outliers
- 1843
r!
- 921.5 .D
c
n
3
- 460.7 >5
- 0
Appendix B, Bladder Cancer Risk Analysis
B-47
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-63
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 5 jug/L Under Lower Bound Scenario
Statistics:
Value
Mean 7.6E-02
Standard Deviation 2.0E-01
1 0,000 Tria
.184 -
£
«
JD
o
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC
Is Frequency Chart 1
1
Ill
°°° 4
O.OE+O 1.5E-1 3.0E-1 4.5E-1 6.0E-1
X10-5
50 Outliers
- 1844
Tl
n
- 922 -O
C
a
a
- 0
Exhibit B-64
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 5 jug/L Under Upper Bound Scenario
Statistics:
Value
Mean 1.6E-01
Standard Deviation 3.9E-01
1 0,000 Tri:
.175 -
Probability
Bo o -
*. 03 C
D -£ CO -
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC U
ils Frequency Chart 261 Outliers
Illllll
t 4
O.OE+O 3.1E-1 6.3E-1 9.4E-1 1.3E+0
Frequency
~M
f) CD CO
^ r^ co
i- co -=r o
Appendix B, Bladder Cancer Risk Analysis
B-48
Proposed Arsenic in Drinking Water Rule RIA
-------
Exhibit B-65
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 3 jug/L Under Lower Bound Scenario
Statistics:
Value
Mean 4.7E-02
Standard Deviation 1.2E-01
10,000 Trie
.207 -
.155 -
£
B
j=
o
CL -052 -
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC
Is Frequency Chart 2
Illlllll
> <
O.OE+0 1.0E-1 2.0E-1 3.0E-1 4.0E-1
X10-5
44 Outliers
- 2067
Tl
5
JO
c
ft
3
- 0
Exhibit B-66
NTNC Expected Number of Cancer Cases per 100,000 people
MCL = 3 jug/L Under Upper Bound Scenario
Statistics:
Value
Mean 1.0E-01
Standard Deviation 2.4E-01
1 0,000 Tria
.180 -
a
JD
e
Forecast: Ind. Lifetime Composite Risk/1 OOK NTNC U
Is Frequency Chart 2
Illlllll
O.OE+0 2.0E-1 4.0E-1 6.0E-1 8.0E-1
48 Outliers
- 1801
T1
3
- 900.5 J3
e
n
3
- 450.2 ,$5
- 0
Appendix B, Bladder Cancer Risk Analysis
B-49
Proposed Arsenic in Drinking Water Rule RIA
------- |