Jump to main content or area navigation.

Contact Us

Water: Radon

Radon in Drinking Water Health Risk Reduction and Cost Analysis



[Federal Register: February 26, 1999 (Volume 64, Number 38)]

[Notices]

[Page 9559-9599]

From the Federal Register Online via GPO Access [wais.access.gpo.gov]

[DOCID:fr26fe99-119]


[[Page 9559]]


______________________________________________________________________


Part II

Environmental Protection Agency

_______________________________________________________________________



Radon in Drinking Water Health Risk Reduction and Cost Analysis; Notice



[[Page 9560]]



ENVIRONMENTAL PROTECTION AGENCY



[FRL-6304-3]



Radon in Drinking Water Health Risk Reduction and Cost Analysis



AGENCY: Environmental Protection Agency.



ACTION: Notice and request for public comments and announcement of

stakeholder meeting.



-----------------------------------------------------------------------



SUMMARY: The Safe Drinking Water Act (SDWA), as amended in 1996,

requires the U.S. Environmental Protection Agency (EPA) to publish a

health risk reduction and cost analysis (HRRCA) for radon in drinking

water for public comment. The purpose of this notice is to provide the

public with the HRRCA for radon and to request comments on the

document. As required by SDWA, EPA will publish a response to all

significant comments to the HRRCA in the preamble to the proposed

National Primary Drinking Water Regulation (NPDWR) for radon, due in

August, 1999.

    The goal of the HRRCA is to provide a neutral and factual analysis

of the costs, benefits, and other impacts of controlling radon levels

in drinking water. The HRRCA is intended to support future decision

making during development of the radon NPDWR. The HRRCA evaluates radon

levels in drinking water of 100, 300, 500, 700, 1000, 2000, and 4000

pCi/L. The HRRCA also presents information on the costs and benefits of

implementing multimedia mitigation (MMM) programs to reduce the risks

of radon exposure in indoor air. The SDWA, as amended, provides for

development of an Alternative Maximum Contaminant Level (AMCL), which

public systems may comply with if their State has an EPA approved MMM

program to reduce radon in indoor air. The concept behind the AMCL and

MMM option is to reduce radon health risks by addressing the larger

source of exposure (air levels in homes) compared to drinking water. If

a State chooses to employ a MMM program to reduce radon risk, it would

implement a State program to reduce indoor air levels and require

public water systems to control water radon levels to the AMCL. If a

State does not choose a MMM program option, a public water system may

propose a MMM program for EPA approval. Today's notice does not include

any decisions regarding the choice of a Maximum Contaminant Level (MCL)

for radon in drinking water. Today's notice also announces a

stakeholder meeting on the HRRCA and framework for the MMM program.



DATES: The Agency must receive comments on the HRRCA on or before April

12, 1999. EPA will hold a one day public meeting on Tuesday, March 16,

1999 from 9 a.m. to 5:30 p.m. EST.



ADDRESSES: Send written comments on HRRCA to the Comment Clerk, docket

number W-98-30, Water Docket (MC4101), USEPA, 401 M St., SW,

Washington, DC 20460. Please submit an original and three copies of

your comments and enclosures (including references).

    Commenters who want EPA to acknowledge receipt of their comments

should enclose a self-addressed, stamped envelope. No facsimiles

(faxes) will be accepted. Comments may also be submitted electronically

to ow-docket@epa.gov. Electronic comments must be submitted as an

ASCII, WP6.1, or WP8 file avoiding the use of special characters and

any form of encryption. Electronic comments must be identified by the

docket number W-98-30. Comments and data will also be accepted on disks

in WP6.1, WP8, or ASCII file format. Electronic comments on this notice

may be filed online at many Federal Depository Libraries.

    The record for this notice has been established under docket number

W-98-30, and includes supporting documentation as well as printed,

paper versions of electronic comments. The full record is available for

inspection from 9 a.m. to 4 p.m. EST Monday through Friday, excluding

legal holidays at the Water Docket, Room EB57, USEPA Headquarters, 401

M St., SW, Washington, DC 20460. For access to docket materials, please

call 202-260-3027 to schedule an appointment.

    The stakeholder meeting on the HRRCA and multimedia mitigation

framework will be held at the offices of at RESOLVE, Inc., 1255 23rd

Street, N.W,. Suite 275, Washington, DC 20037. Check-in will begin at

8:30 a.m.



FOR FURTHER INFORMATION CONTACT: For general information, please

contact the EPA Safe Drinking Water Hotline at 1-800-426-4791 or 703-

285-1093 between 9 a.m. and 5:30 p.m. EST. (For information on radon in

indoor air, contact the National Safety Council's National Radon

Hotline at 1-800-SOS-RADON.) The HRRCA, including the appendices, can

also be accessed on the internet at https://www.epa.gov/safewater/

standard/pp/radonpp/html. For specific information and technical

inquiries, contact Michael Osinski at 202-260-6252 or

osinski.michael@epa.gov.

    For general information on meeting logistics, please contact Sheri

Jobe at RESOLVE, Inc., at 202-965-6382 or Email: sjobe@resolv.org.



SUPPLEMENTARY INFORMATION: The purpose of the March 16, 1999

stakeholder meeting is to cover the following key issues, including:

(1) Discussion of the Health Risk Reduction and Cost Analysis published

in this notice; and (2) present information and discuss issues related

to status of development of a framework for multimedia mitigation

programs. This upcoming meeting is the fifth of a series of

stakeholders meetings on the NPDWR for radon, intended to seek input

from State and Tribal drinking water and radon programs, the regulated

community (public water systems), public health and safety

organizations, environmental and public interest groups, and other

stakeholders. EPA encourages the full participation of stakeholders

throughout this process.

    To register for the meeting, please contact Sheri Jobe at RESOLVE,

Inc., 1255 23rd Street, N.W,. Suite 275, Washington, DC 20037, Phone:

202-965-6382, Fax: 202-338-1264, Email: sjobe@resolv.org. Please

provide your name, affiliation/organization, address, phone, fax and

email if you would like to be on the mailing list to receive further

information about the meeting (including agenda and meeting summary). A

limited number of tele-conference lines will be available. Please

indicate whether you would like to participate by phone. Those

registered for the meeting by February 26, 1999 will receive an agenda,

logistics sheet, and other information prior to the meeting.



    Dated: January 5, 1999.

Dana D. Minerva,

Acting Assistant Administrator, Office of Water, Environmental

Protection Agency.



Radon in Drinking Water Health Risk Reduction and Cost Analysis



Table of Contents



1. Executive Summary

2. Introduction

    2.1 Background

    2.2 Regulatory History

    2.3 Safe Drinking Water Act Amendments of 1996

    2.4 Specific Requirements for the Health Risk Reduction and Cost

Analysis

    2.5 Radon Levels Evaluated

    2.6 Document Structure

3. Health Effects From Radon Exposure

    3.1 Radon Occurrence and Exposure Pathways

    3.1.1 Occurrence

    3.1.2 Exposure Pathways

    3.2 Nature of Health Impacts

    3.3 Impacts on Sensitive Subpopulations



[[Page 9561]]



    3.4 Risk Reduction Model for Radon in Drinking Water

    3.5 Risks from Existing Radon Exposures

    3.6 Potential for Risk Reductions Associated with Removal of Co-

Occurring Contaminants

    3.7 Potential for Risk Increases from Other Contaminants

Associated with Radon Removal

    3.8 Risk for Ever-Smokers and Never-Smokers

4. Benefits of Reduced Radon Exposure

    4.1 Nature of Regulatory Impacts

    4.1.1 Quantifiable Benefits

    4.1.2 Non-Quantifiable Benefits

    4.2 Monetization of Benefits

    4.2.1 Estimation of Fatal and Non-Fatal Cancer Risk Reduction

    4.2.2 Value of Statistical Life for Fatal Cancers Avoided

    4.2.3 Costs of Illness and Lost Time for Non-Fatal Cancers

    4.2.4 Willingness to Pay to Avoid Non-Fatal Cancers

    4.3 Treatment of Monetized Benefits Over Time

5. Costs of Radon Treatment Measures

    5.1 Drinking Water Treatment Technologies and Costs

    5.1.1 Aeration

    5.1.2 Granular Activated Carbon (GAC)

    5.1.3 Storage

    5.1.4 Regionalization

    5.1.5 Radon Removal Efficiencies

    5.1.6 Pre-Treatment to Reduce Iron and Manganese Levels

    5.1.7 Post-Treatment--Disinfection

    5.2 Monitoring Costs

    5.3 Water Treatment Technologies Currently In Use

    5.4 Cost of Technologies as a Function of Flow Rates and Radon

Removal Efficiency

    5.5 Choice of Treatment Responses

    5.6 Cost Estimation

    5.6.1 Site and System Costs

    5.6.2 Aggregate National Costs

    5.6.3 Costs to Community Water Systems

    5.6.4 Costs to Consumers/Households

    5.6.5 Costs to Non-Transient Non-Community Systems

    5.7 Application of Radon Related Costs to Other Rules

6. Results: Costs and Benefits of Reducing Radon in Drinking Water

    6.1 Overview of Analytical Approach

    6.2 Health Risk Reduction and Monetized Health Benefits

    6.3 Costs of Radon Mitigation

    6.4 Incremental Costs and Benefits of Radon Removal

    6.5 Costs to Community Water Systems

    6.6 Costs and Impacts to Households

    6.7 Summary of Cost and Benefit Analysis

    6.8 Sensitivities and Uncertainties

    6.8.1 Uncertainties in Risk Reduction and Health Benefits

Calculations

    6.8.2 Uncertainty in Cost and Impact Calculations

7. Implementation Scenarios--Multimedia Mitigation Programs

    7.1 Multimedia Mitigation Programs

    7.2 Implementation Scenarios Evaluated

    7.3 Multimedia Mitigation Cost and Benefit Assumptions

    7.4 Annual Costs and Benefits of Multimedia Mitigation Program

Implementation

    7.6 Sensitivities and Uncertainties



List of Tables and Figures



Table 3-1. Radon Distributions by Region

Table 3-2. Radon Distribution in Public Water Systems

Table 3-3. Population Exposed Above Various Radon Levels By System

Size

Table 3-4. Estimated Radon Unit Lifetime Fatal Cancer Risks in

Community Water Systems

Table 3-5. Radon Treatment Assumptions to Calculate Residual Fatal

Cancer Risks

Table 3-6. Annual Fatal Cancer Risks for Exposures to Radon from

Community Water Systems

Table 3-7. Radon Risk Reductions Across Various Effluent Levels and

Percent Removals

Table 3-8. Radon Risk Reduction from Treatment Compared to DBP Risks

Table 3-9. Annual Lung Cancer Death Risks Estimates from Radon

Progeny for Ever-Smokers, Never-Smokers, and the General Population

Table 4-1. Proportion of Fatal Cancers by Exposure Pathway and

Estimated Mortality

Table 4-2. Estimated Medical Care and Lost-Time Costs Per Case for

Survivors of Lung Cancer

Table 4-3. Estimated Medical Care and Lost-time Costs Per Case for

Survivors of Stomach Cancer

Table 5-1. Unit Treatment Costs by Removal Efficiency and System

Size

Table 5-2. Estimated Proportions of Ground Water Systems With Water

Treatment Technologies Already in Place

Table 5-3. Decision Matrix For Selection of Treatment Technology

Options

Table 5-4. Number of Sites per Ground Water System by System Size

Table 6-1. Risk Reduction and Residual Cancer Risk from Reducing

Radon in Drinking Water

Table 6-2. Estimated Monetized Health Benefits from Reducing Radon

in Drinking Water

Table 6-3. Risk Reduction and Monetized Benefits Estimates For Ever-

Smokers

Table 6-4. Risk Reduction and Monetized Benefits Estimates For

Never-Smokers

Table 6-5. Estimated Annualized National Costs of Reducing Radon

Exposures

Table 6-6. Capital and O&M Costs of Mitigating Radon in Drinking

Water

Table 6-7. Estimates of the Annual Incremental Costs and Benefits of

Reducing Radon in Drinking Water

Table 6-8. Number of Community Water Systems Exceeding Various Radon

Levels

Table 6-9. Average Annual Cost Per System

Table 6-10. Annual Costs per Household for Community Water Systems

Table 6-11. Per Household Impact by Community Water System as a

Percentage of Median Household Income

Table 6-12. Estimated National Annual Costs and Benefits of Reducing

Radon Exposures--Central Tendency Estimate

Table 6-13. Total Annual Costs and Fatal Cancers Avoided by System

Size

Table 6-14. Annual Monetized Health Benefits by System Size

Table 7-1. Central Tendency Estimates of Annualized Costs and

Benefits of Reducing Radon Exposures with 50% of States Selecting

the MMM/AMCL Option

Table 7-2. Central Tendency Estimates of Annualized Costs and

Benefits of Reducing Radon Exposures with 100% of States Selecting

the MMM/AMCL Option

Figure 3-1. General Patterns of Radon Occurrence in Ground Water

Figure 3-2. EPA Map of Radon Zones in Indoor Air

Figure 6-1. Sensitivity Analysis of Water Mitigation Costs

Figure 7-1. Sensitivity Analysis to Changes in MMM Cost Estimates



Abbreviations Used in This Document



AF: Average Flow

AMCL: Alternative Maximum Contaminant Level

AWWA: American Water Works Association

BAT: Best Available Technology

CWS: Community Water System

DA: Diffused-Bubble Aeration

DBP: Disinfection By-Products

DF: Design Flow

GAC: Granular Activated Carbon

EPA: US Environmental Protection Agency

FACA: Federal Advisory Committee Act

HRRCA: Health Risk Reduction and Cost Analysis

MCL: Maximum Contaminant Level

MCLG: Maximum Contaminant Level Goal

MMM: Multimedia Mitigation program

MSBA: Multi-Stage Diffused Bubble Aeration

NAS: National Academy of Sciences

NDWAC: National Drinking Water Advisory Council

NIRS: National Inorganics and Radionuclides Survey

NPDWR: National Primary Drinking Water Regulation

NTNCWS: Non-Transient Non-Community Water System

OGWDW: Office of Ground Water and Drinking Water

O&M: Operation and Maintenance

OMB: Office of Management and Budget

pCi/l: Picocurie Per Liter

POE GAC: Point-of-Entry Granular Activated Carbon

PTA: Packed Tower Aeration

RIA: Regulatory Impact Analysis

SAB: Science Advisory Board

SDWA: Safe Drinking Water Act, as amended in 1986 and 1996

SDWIS: Safe Drinking Water Inventory System

THM: Trihalomethane

VSL: Value of a Statistical Life

WTP: Willingness To Pay



1. Executive Summary



    This document constitutes the Health Risk Reduction and Cost

Analysis (HRRCA) in support of development of a National Primary

Drinking Water Regulation (NPDWR) for radon in drinking water, as

required by Section 1412(b)(13) of the 1996 Amendments to



[[Page 9562]]



the Safe Drinking Water Act (SDWA). The goal of the HRRCA is to provide

a neutral and fact-based analysis of the costs, benefits, and other

impacts of controlling radon levels in drinking water to support future

decision making during development of the radon NPDWR. The document

addresses the various requirements for the analysis of benefits, costs,

and other elements specified by Section 1412(b)(13) of the SDWA, as

amended.

    This is the first time the Environmental Protection Agency (EPA)

has prepared a HRRCA under the SDWA, as amended. As such, the EPA is

very interested in seeking comment on the techniques, assumptions, and

data inputs upon which the analysis is based. The Agency recognizes

that there may be other methods of conducting the analysis and

presenting the data required for this HRRCA, and encourages meaningful

input from all stakeholders during the public comment period.

Therefore, the specific analysis and findings presented here are

intended as an initial effort to frame an analysis that can support

development of the NPDWR. Since the HRRCA is a cost-benefit tool to

analyze an array of radon levels during development of the NPDWR, many

of the issues to be addressed in the regulatory development process

(e.g. the selection of a Maximum Contaminant Level (MCL), Best

Available Technology (BAT), and monitoring framework) are not analyzed

here, but will be presented in the proposed rule.

    The HRRCA evaluates radon levels in ground water supplies of 100,

300, 500, 700, 1000, 2000, and 4000 pCi/l. The HRRCA also presents

information on the costs and benefits of implementing multimedia

mitigation (MMM) programs. The scenarios evaluated are described in

detail in Section 2.5. This executive summary presents a background on

the radon in drinking water problem, followed by a summary of findings

arranged according to each provision for HRRCAs as specified by the

SDWA, as amended.



Background: Radon Health Risks, Occurrence, and Regulatory History





    Radon is a naturally occurring volatile gas formed from the normal

radioactive decay of uranium. It is colorless, odorless, tasteless,

chemically inert, and radioactive. Uranium is present in small amounts

in most rocks and soil, where it decays to other products including

radium, then to radon. Some of the radon moves through air or water-

filled pores in the soil to the soil surface and enters the air, and

can enter buildings through cracks and other holes in the foundation.

Some radon remains below the surface and dissolves in ground water

(water that collects and flows under the ground's surface). Due to

their very long half-life (the time required for half of a given amount

of a radionuclide to decay), uranium and radium persist in rock and

soil.

    Exposure to radon and its progeny is believed to be associated with

increased risks of several kinds of cancer. When radon or its progeny

are inhaled, lung cancer accounts for most of the total incremental

cancer risk. Ingestion of radon in water is suspected of being

associated with increased risk of tumors of several internal organs,

primarily the stomach. As required by the SDWA, EPA arranged for the

National Academy of Sciences (NAS) to assess the health risks of radon

in drinking water. The NAS released the ``Report on the Risks of Radon

in Drinking Water,''(NAS Report) in September 1998 (NAS 1998B). The NAS

Report represents a comprehensive assessment of scientific data

gathered to date on radon in drinking water. The report, in general,

confirms earlier EPA scientific conclusions and analyses of radon in

drinking water (US EPA,1994C).

    NAS recently estimated individual lifetime unit fatal cancer risks

associated with exposure to radon from domestic water use for ingestion

and inhalation pathways (Table 3-4). The results show that inhalation

of radon progeny accounts for most (approximately 89 percent) of the

individual risk associated with domestic water use, with almost all of

the remainder (11 percent) resulting from directly ingesting radon in

drinking water. Inhalation of radon progeny is associated primarily

with increased risk of lung cancer, while ingestion exposure is

associated primarily with elevated risk of stomach cancer.

    The NAS Report confirmed that indoor air contamination arising from

soil gas typically account for the bulk of total individual risk due to

radon exposure. Usually, most radon gas enters indoor air by diffusion

from soils through basement walls or foundation cracks or openings.

Radon in domestic water generally contributes a small proportion of the

total radon in indoor air.

    The NAS Report is one of the most important inputs used by EPA in

the HRRCA. EPA has used the NAS's assessment of the cancer risks from

radon in drinking water to estimate both the health risks posed by

existing levels of radon in drinking water and also the cancer deaths

prevented by reducing radon levels.

    In updating key analyses and developing the framework for the cost-

benefit analysis presented in the HRRCA, EPA has consulted with a broad

range of stakeholders and technical experts. Participants in a series

of stakeholder meetings held in 1997 and 1998 included representatives

of public water systems, State drinking water and indoor air programs,

Tribal water utilities and governments, environmental and public health

groups, and other federal agencies.

    The HRRCA builds on several technical components, including

estimates of radon occurrence in drinking water, analytical methods for

detecting and measuring radon levels, and treatment technologies.

Extensive analyses of these issues were undertaken by the Agency in the

course of previous rulemaking efforts for radon and other

radionuclides. Using data provided by stakeholders, and from published

literature, the EPA has updated these technical analyses to take into

account the best currently available information and to respond to

comments on the 1991 proposed NPDWR for radon. As required by the 1996

Safe Drinking Water Act (SDWA), EPA has withdrawn the proposed NPDWR

for radon (US EPA 1997B) and will propose a new regulation by August,

1999. The HRRCA does not include any decisions regarding the choice of

a Maximum Contaminant Level (MCL) for radon in drinking water.

    The analysis presented in this HRRCA uses updated estimates of the

number of active public drinking water systems obtained from EPA's Safe

Drinking Water Information System (SDWIS). Treatment costs for the

removal of radon from drinking water have also been updated. The HRRCA

follows current EPA policies with regard to the methods and assumptions

used in cost and benefit assessment.

    As part of the regulatory development process, EPA has updated and

refined its analysis of radon occurrence patterns in ground water

supplies in the United States (US EPA 1998L). This new analysis

incorporates information from the EPA's 1985 National Inorganic and

Radionuclides Survey (NIRS) of 1000 community ground water systems

throughout the United States, along with supplemental data provided by

the States, water utilities, and academic research. The new study also

addressed a number of issues raised by public comments in the previous

occurrence analysis that accompanied the 1991 proposed NPDWR, including

characterization of regional and temporal variability in radon levels,

and



[[Page 9563]]



the impact of sampling point for monitoring compliance.

    In general, radon levels in ground water in the United States have

been found to be the highest in New England and the Appalachian uplands

of the Middle Atlantic and Southeastern states (Figure 3-1). There are

also isolated areas in the Rocky Mountains, California, Texas, and the

upper Midwest where radon levels in ground water tend to be higher than

the United States average. The lowest ground water radon levels tend to

be found in the Mississippi Valley, lower Midwest, and Plains states.

When comparing radon levels in ground water to radon levels in indoor

air at the State level, the distribution of radon concentrations in

indoor air (Figure 3-2) do not always mirror distributions of radon in

ground water.

    In addition, the 1996 Amendments to the SDWA introduce two new

elements into the radon in drinking water rule: (1) an Alternative

Maximum Contaminant Level (AMCL) and (2) multimedia radon mitigation

(MMM) programs. The SDWA, as amended, provides for development of an

AMCL, which public water systems may comply with if their State has an

EPA approved MMM program to reduce radon in indoor air. The NAS Report

estimated that the AMCL would be about 4,000 pCi/L, based on SDWA

requirements. The concept behind the AMCL and MMM option is to reduce

radon health risks by addressing the larger source of exposure (air

levels in homes) compared to drinking water. If a State chooses to

employ a MMM program to reduce radon risk, it would implement a State

program to reduce indoor air levels and require public water systems to

control radon levels in drinking water to the AMCL. If a State does not

choose a MMM program option, a public water system may propose a MMM

program for EPA approval.



Summary of Findings



Quantifiable and Non-Quantifiable Costs



    The capital and operating and maintenance (O&M) costs of mitigating

radon in Community Water Systems (CWSs) were estimated for each of the

radon levels evaluated. The costs of reducing radon in ground water to

specific target levels were calculated using the cost curves discussed

in Section 5.4 and the matrix of treatment options presented in Section

5.5. For each radon level and system size stratum, the number of

systems that need to reduce radon levels by up to 50 percent, 80

percent and 99 percent were calculated. Then, the cost curves for the

distributions of technologies dictated by the treatment matrix were

applied to the appropriate proportions of the systems. Capital and O&M

costs were then calculated for each system, based on typical estimated

design and average flow rates. These flow rates were calculated on

spreadsheets using equations from EPA's Safe Drinking Water Suite Model

(US EPA 1998N). The equations and parameter values relating system size

to flow rates are presented in Appendix C. The technologies addressed

in the cost estimation included a number of aeration and granular

activated carbon (GAC) technologies described in Section 5.1, as well

as storage, regionalization, and disinfection as a post-treatment. To

estimate costs, water systems were assumed, with a few exceptions, to

select the technology that could reduce radon to the selected target

level at the lowest cost. CWSs were also assumed to treat separately at

every source from which water was obtained and delivered into the

distribution system.

    The costs of reducing radon to various levels are summarized in

Table 6-5, which shows that, as expected, aggregate radon mitigation

costs increase with decreasing radon levels. The cost ranges presented

in the table represent plausible upper and lower bounds of 50 percent

above to 50 percent below the central tendency estimates. For CWSs, the

costs per system do not vary substantially across the different radon

levels evaluated. This is because the menu of mitigation technologies

for systems with various influent radon levels remains relatively

constant.



Quantifiable and Non-Quantifiable Health Benefits



    The quantifiable health benefits of reducing radon exposures in

drinking water are attributable to the reduced incidence of fatal and

non-fatal cancers, primarily of the lung and stomach. Table 6-1 shows

the health risk reductions (number of fatal and non-fatal cancers

avoided) and the residual health risk (number of remaining cancer

cases) at various radon in water levels. Since preparing the

prepublication edition of the NAS Report, the NAS has reviewed and

slightly revised their unit risk estimates. EPA uses these updated unit

risk estimates in calculating the baseline risks, health risk

reductions, and residual risks. Under baseline assumptions (no control

of radon exposure), approximately 160 fatal cancers and 9.2 non-fatal

cancers per year are associated with radon exposures through CWSs. At a

radon level of 4,000 pCi/l, approximately 2.2 fatal cancers and 0.1

non-fatal cancers per year are prevented. At the lowest level evaluated

(100 pCi/l), approximately 115 fatal and 6.6 non-fatal cancers per year

would be prevented.

    The Agency has developed monetized estimates of the health benefits

associated with the risk reductions from radon exposures. The SDWA, as

amended, requires that a cost-benefit analysis be conducted for each

NPDWR, and places a high priority on better analysis to support

rulemaking. The Agency is interested in refining its approach to both

the cost and benefit analysis, and in particular recognizes that there

are different approaches to monetizing health benefits. In the past,

the Agency has presented benefits as cost per life saved, as in Table

6-5. An alternative approach presented here for consideration as one

measure of potential benefits is the monetary value of a statistical

life (VSL) applied to each fatal cancer avoided. Since this approach is

relatively new to the development of NPDWRs, EPA is interested in

comments on these alternative approaches to valuing benefits, and will

have to weigh the value of these approaches for future use.

    Estimating the VSL involves inferring individuals' implicit

tradeoffs between small changes in mortality risk and monetary

compensation. In the HRRCA, a central tendency estimate of $5.8 million

(1997$) is used in the monetary benefits calculations, with low- and

high-end values of $700,000 (1997$) and $16.3 million (1997$),

respectively, used for the purposes of sensitivity analysis. These

figures span the range of VSL estimates from 26 studies reviewed in

EPA's recent draft guidance on benefits assessment (US EPA 1998E),

which is currently under review by the Agency's Science Advisory Board

(SAB) and the Office of Management and Budget (OMB).

    It is important to recognize the limitations of existing VSL

estimates and to consider whether factors such as differences in the

demographic characteristics of the populations and differences in the

nature of the risks being valued have a significant impact on the value

of mortality risk reduction benefits. Also, medical care or lost-time

costs are not separately included in the benefits estimate for fatal

cancers, since it is assumed that these costs are captured in the VSL

for fatal cancers.

    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 lung and stomach cancers. The use of



[[Page 9564]]





such WTP estimates is supported in the SDWA, as amended, at Section

1412(b)(3)(C)(iii): ``The Administrator may identify valid approaches

for the measurement and valuation of benefits under this subparagraph,

including approaches to identify consumer willingness to pay for

reductions in health risks from drinking water contaminants.''

    A WTP central tendency estimate of $536,000 is used to monetize the

benefits of avoiding non-fatal cancers (Viscusi et al. 1991), with a

range between $169,000 and $1.05 million (1997$). The combined fatal

and non-fatal health benefits are summarized in Table 6-2. The annual

health benefits range from $13 million for a radon level of 4000 pCi/l

to $673 million at 100 pCi/l. The ranges in the last column of Table 6-

2 illustrate how benefits vary when the upper and lower bound estimates

of the VSL and WTP measures are used.

    Reductions in radon exposures might also be associated with non-

quantifiable benefits. EPA has identified several potential non-

quantifiable benefits associated with regulating radon in drinking

water. These benefits may include any peace of mind benefits specific

to reduction of radon risks that may not be adequately captured in the

VSL estimate. In addition, treating radon in drinking water with

aeration oxidizes arsenic into a less soluble form that is easier to

remove with conventional removal technologies. In terms of reducing

radon exposures in indoor air, it has also been suggested that

provision of information to households on the risks of radon in indoor

air and available options to reduce exposure is a non-quantifiable

benefit that can be attributed to some components of a MMM program.

Providing such information might allow households to make informed

choices about the appropriate level of risk reduction given their

specific circumstances and concerns. These potential benefits are

difficult to quantify because of the uncertainty surrounding their

estimation. However, they are likely to be somewhat less significant

relative to the monetized benefits estimates.



Incremental Costs and Benefits of Radon Removal



    Table 6-7 summarizes the central tendency and the upper and lower

bound estimates of the incremental costs and benefits of radon exposure

reduction. Both the annual incremental costs and benefits increase as

the radon level decreases from 4000 pCi/l down to 100 pCi/l.

Incremental costs and benefits are within 10 percent of each other at

radon levels of 1000, 700, and 500 pCi/l. The table also illustrates

the wide ranges of potential incremental costs and benefits due to the

uncertainty inherent in the estimates. There is substantial overlap

between the incremental costs and benefits at each radon level.



Impacts on Households



    The cost impact of reducing radon in drinking water at the

household level was also assessed. As expected, costs per household

increase as system size decreases (Table 6-10). Costs to households are

higher for households served by smaller systems than larger systems for

two reasons. First, smaller systems serve far fewer households than

larger systems and, consequently, each household must bear a greater

percentage share of the capital and O&M costs. Second, smaller systems

tend to have higher influent radon concentrations that, on a per-capita

or per-household basis, require more expensive treatment methods (e.g.,

one that has an 85 percent removal efficiency rather than 50 percent)

to achieve the applicable radon level.

    Another significant finding is that, like the per system costs,

costs per household (which are a function of per system costs) are

relatively constant across different radon levels within each system

size category. For example, there is less than one dollar per year

variation in household costs, regardless of the radon level being

considered for households served by large public or private systems

(between $6 and $7 annually), by medium public or private systems

(between $10 and $11), and by small public or private systems (between

$19 and $20 annually). Similarly, for very small systems (501-3300

people), the cost per household is consistently about $34 annually for

public systems and about $40 annually for private systems, varying

little with the target radon level. Only for very very small systems is

there a noticeable variation in household costs across radon levels.

The range for per household costs for public CWSs serving 25-500 people

is $87 per year (at 4,000 pCi/l) to $135 per year (at 100 pCi/l). The

corresponding range for private CWSs is $139 to $238 per year. For

households served by the smallest public systems (25-100 people) the

range of cost per household ranges from $292 per year at 4,000 pCi/l to

$398 per year at 100 pCi/l. For private systems, the range is $364 per

year to $489 per year, respectively.



Summary of Annual Costs and Benefits



    Table 6-12 reveals that at a radon level of 4000pCi/l (equivalent

to the AMCL estimated in the NAS Report), annual costs are

approximately twice the annual monetized benefits. For radon levels of

1000pCi/l to 300 pCi/l, the central tendency estimates of annual costs

are above the central tendency estimates of the monetized benefits,

although they are within 10 percent of each other. However, as shown in

Tables 6-2 and 6-5, due to the uncertainty in the cost and benefit

estimates, there is a very broad possible range of potential costs and

benefits that overlap across all of the radon levels evaluated.



Benefits From the Reduction of Co-Occurring Contaminants



    The occurrence patterns of other industrial pollutants are

difficult to clearly define at the national level relative to a

naturally occurring contaminant such as radon. Similarly, the Agency's

re-evaluation of radon occurrence has revealed that the geographic

patterns of radon occurrence are not significantly correlated with

other naturally occurring inorganic contaminants that may pose health

risks. Thus, it is not likely that a clear relationship exists between

the need to install radon treatment technologies and treatments to

remove other contaminants. On the other hand, technologies used to

reduce radon levels in drinking water have the potential to reduce

concentrations of other pollutants as well. Aeration technologies will

also remove volatile organic contaminants from contaminated ground

water. Similarly, granular activated carbon (GAC) treatment for radon

removal effectively reduces the concentrations of organic (both

volatile and nonvolatile) chemicals and some inorganic contaminants.

Aeration also tends to oxidize dissolved arsenic (a known carcinogen)

to a less soluble form that is more easily removed from water. The

frequency and extent that radon treatment would also reduce risks from

other contaminants has not been quantitatively evaluated.



Impacts on Sensitive Subpopulations



    The SDWA, as amended, includes specific provisions in Section

1412(b)(3)(C)(i)(V) to assess the effects of the contaminant on the

general population and on groups within the general population such as

children, pregnant women, the elderly, individuals with a history of

serious illness, or other subpopulations that are



[[Page 9565]]



identified as likely to be at greater risk of adverse health effects

due to exposure to contaminants in drinking water than the general

population. The NAS Report concluded that there is insufficient

scientific information to permit separate cancer risk estimates for

potential subpopulations such as pregnant women, the elderly, children,

and seriously ill persons. The NAS Report did note, however, that

according to the NAS model for the cancer risk from ingested radon,

which accounts for 11% of the total fatal cancer risk from radon in

drinking water, approximately 30% of the fatal lifetime cancer risk is

attributed to exposure between ages 0 to 10.

    The NAS Report identified smokers as the only group that is more

susceptible to inhalation exposure to radon progeny (NAS 1998A, 1998B).

Inhalation of cigarette smoke and radon progeny result in a greater

increased risk than if the two exposures act independently to induce

lung cancer. NAS estimates that ``ever smokers'' (more than 100

cigarettes over a lifetime) may be more than five times as sensitive to

radon progeny as ``never smokers'' (less than 100 cigarettes over a

lifetime). Using current smoking prevalence data, EPA's preliminary

estimate for the purposes of the HRRCA is that approximately 85 percent

of the cases of radon-induced cancer will occur among current and

former smokers. This population of current and former smokers, which

consists of 58 percent of the male and 42 percent of the female

population (US EPA 1999A), will also experience the bulk of the risk

reduction from radon exposure reduction in drinking water supplies.



Risk Increases From Other Contaminants Associated With Radon Exposure

Reduction



    As discussed in Section 5.1, the need to install radon treatment

technologies may require some systems that currently do not disinfect

to do so. Case studies (US EPA 199D) of twenty-nine small to medium

water systems that installed treatment (24 aeration, 5 GAC) to remove

radon from drinking water revealed only two systems that reported

adding disinfection (both aeration) with radon treatment (the systems

either had disinfection already in place or did not add it). In

practice, the tendency to add disinfection may be much more significant

than these case studies indicate. EPA also realizes that the addition

of chlorination for disinfection may result in risk-risk tradeoffs,

since, for example, the disinfection technology reduces potential for

infectious disease risk, but at the same time can result in increased

exposures to disinfection by-products (DBPs). This risk-risk trade-off

is addressed by the recently promulgated Disinfectants and Disinfection

By-Products NPDWR (US EPA 1998I). This rule identified MCLs for the

major DBPs, which all CWSs and NTNCWSs must comply. These MCLs set a

risk ceiling from DBPs that water systems adding disinfection in

conjunction with treatment for radon removal could face. The formation

of DBPs is proportional to the concentration of organic precursor

contaminants, which tend to be much lower in ground water than in

surface water.

    The NAS Report addressed several important potential risk-risk

tradeoffs associated with reducing radon levels in drinking water,

including the trade-off between risk reduction from radon treatment

that includes post-disinfection with the increased potential for DBP

formation (NAS 1998B). The report concluded that, based upon median and

average total trihalomethane (THM) levels taken from EPA's 1981

Community Water System Survey, a typical ground water CWS would face

incremental individual lifetime cancer risk due to chlorination

byproducts of 5 x 10<SUP>-5</SUP>. It should be emphasized that this

risk is based on average and median THM occurrence information that

does not segregate systems that disinfect from those that do. Further,

the NAS Report points out that this average DBP risk is smaller than

the average individual lifetime fatal cancer risk associated with

baseline radon exposures from ground water (untreated for radon), which

is estimated at 1.2 x 10<SUP>-4</SUP> using a mean radon concentration

of 213 pCi/l.

    A more meaningful comparison is to look at the trade-off between

risk reduction from radon treatment in cases where disinfection is

added with the added risks from DBP formation. This trade-off will

affect only a minority of systems since a majority of ground water

systems already have disinfection in place. For the smallest systems

size category, approximately half of all CWSs already have disinfection

in place. The proportion of systems having disinfection in place

increases as the size categories increase, up to >95% for large systems

(Table 5-2). In addition, although EPA is using the conservative

costing assumption that all systems adding aeration or GAC would

disinfect, not all systems adding aeration or GAC would have to add

post-disinfection or, if disinfecting, may use a disinfection

technology that does not forms DBPs. For those ground water systems

adding treatment with disinfection, this trade-off tends to be

favorable since the combined risk reduction from radon removal and

microbial risk reduction outweigh the added risk from DBP formation.

    An estimate of the risk reduction due to treatment of radon in

water for various removal percentages and finished water concentrations

is provided in Table 3.7. As noted by the NAS Report, these risk

reductions outweigh the increased risk from DBP exposure for those

systems that chlorinate as a result of adding radon treatment.

    The ratios between risk reduction from radon removal and the risks

from THMs at levels equal their MCLs (a conservative assumption) are

shown in Table 3.8. The data indicate that the risk ratios are

favorable for treatment with disinfection, ignoring microbial risk

reduction, even assuming the worst case scenario that ground water

systems have THM levels at the MCL. It is worth noting that there is

the possibility that accounting quantitatively for the increased risk

from DBP exposure for systems adding chlorination in conjunction with

treatment for radon may somewhat decrease the monetized benefits

estimates.

Other Factors: Uncertainty in Risk, Benefit, and Cost Estimates



    Estimates of health benefits from radon reduction are uncertain. A

few of the variables affecting the uncertainty in the benefit estimates

include the distribution of radon in ground water systems, the NAS's

risk models for ingestion and inhalation risks, and the transfer factor

used to estimate indoor air radon activity levels. EPA plans to include

an uncertainty analysis of radon in drinking water risks with the

proposed rule. Monetary benefit estimates are also strongly affected by

the VSL estimate that is used for fatal cancers. The WTP valuation for

non-fatal cancers has less impact on benefit estimates because it

contributes less than 1 percent to the total benefits estimates, due to

the fact that there are few non-fatal cancers relative to fatal

cancers.

    Estimates of the regulatory costs also have associated uncertainty.

The major factors affecting this uncertainty include assumptions

regarding the distribution of radon levels among ground water systems

and among treatment sites within systems, uncertainties in unit cost

models, the assumed prevalence of the various compliance decisions, and

the exclusion of NTNCWSs in the HRRCA's national cost estimates.

    To deal with a lack of information regarding the intra-system

variability of



[[Page 9566]]





radon levels between treatment sites (source wells), the national cost

estimates are based on the assumption that all CWSs above a target

radon level, as estimated by system-level average radon occurrence

predictions from the occurrence model, will install separate treatment

systems at each site. Ideally, occurrence information at each treatment

site will provide a better estimate of national costs, since the wells

within a water system would exhibit a range of radon occurrence levels,

some of which may be below the target radon level, others above this

level. Since it is not obvious whether the system-level approach will

lead to either a positive or negative bias in the national cost

estimates, EPA is in the process of performing an analysis of the

intra-system variability for radon occurrence and will include this

analysis in support of the upcoming proposed rule.

    There are also significant uncertainties in estimated treatment

unit costs and in the decision-trees that are used to model national

level compliance decisions that will by made by the system-size

stratified universe of drinking water systems in response to a range of

radon influent levels. It is possible to estimate the uncertainties in

both the unit costs and the decision-tree by performing sensitivity

analyses for the factors affecting costs. Regarding unit costs, this

analysis leads to a spread in costs that adequately resembles the

``real-world'' as shown by ranges in treatment cost case studies.

Regarding the uncertainty in the decision-tree, it is unfortunately not

possible to verify results in this way. However, since there are so few

technologies to mitigate radon in water, the decision-tree is fairly

robust.



Other Impacts: Costs and Benefits of Multimedia Mitigation Program

Implementation Scenarios



    In addition to evaluating the costs and benefits across a range of

radon levels, two scenarios were evaluated that reduce radon exposure

through the use of MMM programs. The two scenarios evaluated assume:

(1) 50 percent of States (all water systems in those States) select MMM

implementation; and (2) 100 percent of States select MMM. These two

scenarios are described in detail in Section 7. For the MMM

implementation analysis, systems were assumed to mitigate water to the

4,000 pCi/l Alternative Maximum Contaminant Level (AMCL), if necessary,

and that equivalent risk reduction between the AMCL and the radon level

under evaluation would be achieved through a MMM program. Therefore,

the actual number of cancer cases avoided is the same for the MMM

implementation scenarios as for the water mitigation only scenario.

    In calculating the cost of MMM programs, the cost per fatal cancer

case avoided was estimated at $700,000 (1997$). This value was

originally estimated by EPA in 1992 using 1991 data. The same nominal

value is used in the HRRCA based on anecdotal evidence from EPA's

Office of Radiation and Indoor Air (ORIA) that there has been an

equivalent offset between a decrease in testing and mitigation costs

since 1991 and the expected increase due to inflation in the years

1992-1997. This dollar amount reflects that real testing and mitigation

costs have decreased, while nominal costs have remained approximately

constant.

    Tables 7-2 and 7-3 illustrate that, as expected, the costs of

reducing radon exposures decrease with increasing numbers of States

(i.e. CWSs) selecting the MMM implementation scenario. Also, as would

be expected, the annual costs of implementing MMM are, on average,

lower compared to reducing radon exposures in drinking water alone.

Central tendency estimates of the total annualized benefits exceed the

annualized costs for both the 50 and 100 percent MMM participation

scenarios over all radon levels. The cost per fatal cancer case avoided

is also lower for both the 50 and 100 percent MMM implementation

scenarios compared to the scenario in which no States elect to develop

a MMM program. In addition, the cost per fatal cancer case avoided is

significantly lower for the MMM scenario with 100 percent of the States

electing the MMM program compared to when 50 percent of the States

choose the MMM scenario, especially at the lower radon levels. The

costs and benefits estimates are also broken out into their respective

MMM and water mitigation components. With the exception of 4000pCi/l

(the NAS estimated AMCL), annual monetized benefits are significantly

larger than annual costs for the MMM component of the total costs. For

the water mitigation component, the annual costs are larger than the

annual monetized benefits across all radon levels.



2. Introduction



2.1  Background



    This Health Risk Reduction and Cost Analysis (HRRCA) provides the

Environmental Protection Agency's (EPA) analysis of potential costs and

benefits of different target levels for radon in drinking water. The

HRRCA builds on several technical components, including estimates of

radon occurrence in drinking water supplies, analytical methods for

detecting and measuring radon levels, and treatment technologies.

Extensive analyses of these issues were undertaken by the Agency in the

course of previous rulemaking efforts for radon and other

radionuclides. Using data provided by stakeholders, and from published

literature, the EPA has updated these technical analyses to take into

account the best currently available information and to respond to

comments on the 1991 proposed regulation for radon in drinking water.

As required by the 1996 Safe Drinking Water Act (SDWA), EPA has

withdrawn the proposed regulation for radon in drinking water (US EPA

1997B) and will propose a new regulation by August, 1999.

    One of the most important inputs used by EPA in the HRRCA is the

National Academy of Sciences (NAS) September 1998 report ``Risk

Assessment of Radon in Drinking Water'' (NAS Report). EPA has used the

NAS assessment of the cancer risks from radon in drinking water to

estimate both the health risks posed by existing levels of radon in

drinking water and also the estimated cancer deaths potentially

prevented by reducing radon levels. The NAS Report is the most

comprehensive accumulation of scientific data gathered to date on radon

in drinking water. SDWA required the NAS assessment, which generally

affirms EPA's earlier scientific conclusions and analyses on the risks

of exposure to radon and progeny in drinking water.

    The analysis presented in this HRRCA uses updated estimates of the

number of active public drinking water systems obtained from EPA's Safe

Drinking Water Information System (SDWIS). Treatment costs for the

removal of radon from drinking water also have been updated. The HRRCA

follows EPA policies with regard to the methods and assumptions used in

cost and benefit assessment.

    In updating key analyses and developing the framework for the cost-

benefit analysis presented in the HRRCA, EPA has consulted with a broad

range of stakeholders and technical experts. Participants in a series

of stakeholder meetings held in 1997 and 1998 included representatives

of public water systems, State drinking water and indoor air programs,

tribal water utilities and governments, environmental and public health

groups, and other federal agencies. EPA convened an expert panel in

Denver in November of 1997 to review treatment technology costing

approaches. The panel made a number of



[[Page 9567]]



recommendations for modification to EPA cost estimating protocols that

have been incorporated into the radon cost estimates. EPA also

consulted with a subgroup of the National Drinking Water Advisory

Council (NDWAC) on evaluating the benefits of drinking water

regulations. The NDWAC was formed in accordance with the Federal

Advisory Committee Act (FACA) to assist and advise EPA. A variety of

stakeholders participated in the NDWAC benefits working group,

including utility company staff, environmentalists, health

professionals, State water program staff, a local elected official,

economists, and members of the general public.

    The American Water Works Association (AWWA) convened a ``Radon

Technical Work Group,'' in 1998 that provided technical input on EPA's

update of technical analyses (occurrence, analytical methods, and

treatment technology), and discussed conceptual issues related to

developing guidelines for multimedia mitigation programs. Members of

the Radon Technical Work Group included representatives from State

drinking water and indoor air programs, public water systems, drinking

water testing laboratories, environmental groups and the U.S.

Geological Survey. EPA also held a series of conference calls with

State drinking water and indoor air programs, to discuss issues related

to developing guidelines for multimedia mitigation programs.



2.2  Regulatory History



    Section 1412 of the Safe Drinking Water Act (SDWA), as amended in

1986, requires the EPA to publish Maximum Contaminant Level Goals

(MCLGs) and to promulgate National Primary Drinking Water Regulations

(NPDWRs) for contaminants that may cause an adverse effect on human

health and that are known or anticipated to occur in public water

supplies. In response to this charge, the EPA proposed NPDWRs for

radionuclides, including radon, in 1991 (US EPA 1991). The proposed

rule included a maximum contaminant level (MCL) of 300 pCi/l for radon

in drinking water, applicable to both community water systems and non-

transient non-community water systems. A community water system (CWS)

is defined as a public water system with at least 15 or more service

connections or that regularly serves at least 25 year-round residents.

A non-transient non-community system (NTNCWS) is a public water system

that is not a CWS and that regularly serves at least 25 of the same

persons for at least six months per year. Examples of NTNCWSs include

those that serve schools, offices, and commercial buildings. Under the

proposed rule, all CWSs and NTNCWSs relying on ground water would have

been required to monitor radon levels quarterly at each point of entry

to the distribution system. Compliance monitoring requirements were

based on the arithmetic average of four quarterly samples. The 1991

proposed rule required systems with one or more points of entry out of

compliance to treat influent water to reduce radon levels below the MCL

or to secure water from another source below the MCL.

    The proposed rule was accompanied by an assessment of regulatory

costs and economic impacts, as well as an assessment of the risk

reduction associated with implementation of the MCL. The Agency

received substantial comments on the proposal and its supporting

analyses from States, water utilities, and other stakeholder groups.

Comments from the water industry questioned EPA's estimates of the

number of systems that would be out of compliance with the proposed

MCL, as well as the cost of radon mitigation. EPA's Science Advisory

Board (SAB) provided extensive comments on the risk assessment used by

the Agency to support the proposed MCL. The SAB recommended that EPA

expand the analysis of the uncertainty associated with the risk and

risk reduction estimates. In response to these comments, the assessment

was revised twice, once in 1993 and again in 1995 (US EPA 1995). Both

of the revised risk analyses provided detailed quantitative uncertainty

analysis.



2.3  Safe Drinking Water Act Amendments of 1996



    In the 1996 Amendments to the Safe Drinking Water Act, Congress

established a new charter for public water systems, States, and EPA to

protect the safety of drinking water supplies. Among other mandates,

amended Section 1412(b)(13) directed EPA to withdraw the drinking water

standards proposed for radon in 1991 and to propose a new MCLG and

NPDWR for radon by no later than August 6, 1999. As noted above, the

amendments require NAS to conduct a risk assessment for radon in

drinking water and an assessment of risk reduction benefits from

various mitigation measures to reduce radon in indoor air (Section

1412(b)(13)(B)). In addition, the amendments introduce two new elements

into the radon in drinking water rule: (1) An Alternative Maximum

Contaminant Level (AMCL) and (2) multimedia radon mitigation (MMM)

program.

    If the MCL established for radon in drinking water is more

stringent than necessary to reduce the contribution to radon in indoor

air from drinking water to a concentration that is equivalent to the

national average concentration of radon in outdoor air, EPA is required

to simultaneously establish an AMCL that would result in a contribution

of radon from drinking water to radon levels in indoor air equivalent

to the national average concentration of radon in outdoor air (Section

1412(b)(13)(F)). If an AMCL is established, EPA is to publish

guidelines for State programs, including criteria for multimedia

measures to mitigate radon levels in indoor air, to comply with the

AMCL.

    States may develop and submit to EPA for approval an MMM program to

decrease radon levels in indoor air (Section 1412(b)(13)(G)). These

programs may rely on a variety of mitigation measures, including public

education, testing, training, technical assistance, remediation grants

and loan or incentive programs, or other regulatory and non-regulatory

measures. EPA shall approve a State's program if it is expected to

achieve equal or greater health risk reduction benefits than would be

achieved by compliance with the more stringent MCL. If EPA does not

approve a State program, or a State does not propose a program, public

water supply systems may propose their own MMM programs to EPA,

following the same procedures outlined for States. Once the MMM

programs are established, EPA is required to re-evaluate them no less

than every five years.



2.4  Specific Requirements for the Health Risk Reduction and Cost

Analysis



    Section 1412(b)(13)(C) of the 1996 Amendments requires EPA to

prepare a Health Risk Reduction and Cost Analysis (HRRCA) to be used to

support the development of the radon NPDWR. SDWA requires the HRRCA be

published for public comment by February 6, 1999, six months before the

rule is to be proposed. In the preamble of the proposed rule, EPA must

include a response to all significant public comments on the HRRCA.

    The HRRCA must also satisfy the requirements established in Section

1412(b)(3)(C) of the amended SDWA. According to these requirements, EPA

must analyze each of the following when proposing an NPDWR that

includes a 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



[[Page 9568]]



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 subpopulations

that are identified as likely to be at greater risk of adverse health

effects due to exposure to contaminants in drinking water than the

general population; (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.

    To the extent possible, this HRRCA follows the new cost-benefit

framework being developed by the Office of Ground Water and Drinking

Water (OGWDW) . As provided in the SDWA, as amended, the HRRCA

discusses the costs and benefits associated with a variety of radon

levels. Summary tables and figures are presented that characterize

aggregate costs and benefits, impacts on affected entities, and

tradeoffs between risk reduction and compliance costs. More in-depth

discussions of input data and assumptions will be provided in a

companion ``Analytical Support Document'' and an in-depth presentation

and discussion of the results will appear in a separate ``Cost/Benefit

Document'' that will accompany the proposed rule. The HRRCA by itself

does not constitute the complete Regulatory Impact Analysis (RIA), but

serves as a foundation upon which the RIA can be developed for the

proposed rule.



2.5  Radon Levels Evaluated





    The HRRCA is intended to present preliminary estimates of the

potential costs and benefits of various levels of controlling radon in

drinking water. The HRRCA assumes that all systems drawing water from

sources above a defined radon level will employ treatment technologies

to meet the target level or ``regionalize'' to obtain water from

another source with lower radon levels. This analysis evaluates radon

levels of 100, 300, 500, 700, 1,000, 2,000, and 4,000 pCi/l. The

analysis did not include any provisions for exemptions or phased

compliance and assumed that a simple quarterly monitoring scheme would

be used to determine the need for mitigation and ongoing compliance.

    The HRRCA also evaluates national costs and benefits of MMM

implementation scenarios, with States choosing to reduce radon exposure

in drinking water through an Alternative Maximum Contaminant Level

(AMCL) and radon risks in indoor air through MMM programs. Based on NAS

recommendations, the AMCL level that is evaluated is 4,000 pCi/l. Under

the scenarios that include an AMCL, the HRRCA assumes that a portion of

the States would adopt an AMCL supplemented with MMM programs to

address indoor air radon risks. In the absence of information

concerning the number of States that would choose to implement radon

risk reduction through the use of AMCL plus multimedia programs, the

HRRCA assumes that either 50 or 100 percent of the systems in the

United States would choose to implement MMM programs and comply with

the AMCL. For the MMM implementation scenarios, a single multimedia

cost estimate is used, based on the cost-effectiveness of current

voluntary mitigation efforts. These issues are discussed in more detail

in Section 7.



2.6  Document Structure



    The HRRCA is organized into 7 sections and a number of appendices.

The appendices, while not included in this Federal Register Notice, are

available in the docket for review and can be downloaded from the web

at www.epa.gov/safewater/standard/pp/radonpp/html. Section 3 discusses

the health effects of exposure to radon. Section 4 describes the

assumptions and methods for estimating quantifiable benefits and

assessing non-quantifiable benefits. Section 5 discusses the water

treatment and MMM methods used to calculate the national costs of the

various radon levels examined. Section 6 presents the results of the

cost and benefit analysis of reducing radon levels in drinking water,

and evaluates economic impacts on households. In addition, the major

sources of uncertainty associated with the estimates of costs,

benefits, and economic impacts are identified. Section 7 estimates the

costs and benefits of two different implementation scenarios in which

States and water systems elect to develop and implement a MMM program

and comply with the AMCL. Appendices provide details of the risk

calculations, cost curves for treatment technologies, methods used to

calculate system flows, and detailed breakdown summaries of the cost,

benefit and impact calculations.



3. Health Effects of Radon Exposure



    This Section presents an overview of the major issues and

assumptions addressed in order to characterize the health impacts and

potential benefits of reductions in radon exposures. The methods that

have been used to characterize risk and benefits in the HRRCA are also

described. The assumptions and methods presented below are used in

Section 4 to derive detailed estimates of the health reduction benefits

of different radon levels in ground water supplies.



3.1  Radon Occurrence and Exposure Pathways



    As part of the regulatory development process, EPA has updated and

refined its analysis of radon occurrence patterns in ground water

supplies in the United States (US EPA 1998L). This new analysis

incorporates information from the EPA 1985 National Inorganic and

Radionuclides Survey (NIRS) of 1000 community ground water systems

throughout the United States, along with supplemental data provided by

the States, water utilities, and academic researchers.

    The new study also addressed a number of issues raised by public

comments on the previous occurrence analysis. These include

characterization of regional and temporal variability in radon levels,

variability in radon levels across different-sized water systems,

impact of sampling point, and the proper statistical techniques for

evaluating the data.

3.1.1  Occurrence

    Radon is a naturally occurring volatile gas formed from the normal

radioactive decay of uranium. It is colorless, odorless, tasteless,

chemically inert, and radioactive. Uranium is present in small amounts

in most rocks and soil, where it decays to other products including



[[Page 9569]]



radium, then to radon. Some of the radon moves through air or water-

filled pores in the soil to the soil surface and enters the air, while

some remains below the surface and dissolves in ground water (water

that collects and flows under the ground's surface). Due to their very

long half-life (the time required for half of a given amount of a

radionuclide to decay), uranium and radium persist in rock and soil.

    Radon itself undergoes radioactive decay and has a radioactive

half-life of about four days. When radon atoms decay they emit

radiation in the form of alpha particles, and transform into decay

products, or progeny, which also decay. Unlike radon gas, these progeny

easily attach to and can be transported by dust and other particles in

air. The decay of progeny continues until stable, non-radioactive

progeny are formed. At each step in the decay process, radiation is

released. The term radon, as commonly used, refers to radon-222 as well

as its radioactive decay products.

    In general, radon levels in ground water in the United States have

been found to be the highest in New England and the Appalachian uplands

of the Middle Atlantic and Southeastern States (Figure 3-1). There are

also isolated areas in the Rocky Mountains, California, Texas, and the

upper Midwest where radon levels in ground water tend to be higher than

the United States average. The lowest ground water radon levels tend to

be found in the Mississippi Valley, lower Midwest, and Plains States.

When comparing radon levels in ground water to radon levels in indoor

air at the State level, the distribution of radon concentrations in

indoor air (Figure 3-2) do not always mirror distributions of radon in

ground water.

    In addition to large-scale regional variation, radon levels in

ground water also vary significantly over smaller distance scales.

Local differences in geology tend to greatly influence the patterns of

radon levels observed at specific locations (e.g., not all radon levels

in New England are high; not all radon levels in the Gulf Coast region

are low). Over small distances, there is often no consistent

relationship between measured radon levels in ground water and radium

levels in the ground water or in the parent bedrock (Davis and Watson

1989). Similarly, no significant national correlation has been found

between radon levels in individual ground water systems and the levels

of other inorganic contaminants or conventional geochemical parameters.

Potential correlations between radon levels and levels of organic

contaminants in ground water have not been investigated, but there is

little reason to believe any would be found. Radon's volatility is

rather high compared to its solubility in water. Thus, radon

volatilizes rapidly from surface water, and measured radon levels in

surface water supplies are generally insignificant compared to those

found in ground water.



Figure 3-1. General Patterns of Radon Occurrence in Groundwater in

the United States



    Figure 3-1 is not printed in the Federal Register. It is available

in the Water Docket at the address listed in the ADDRESSES section.



Figure 3-2. EPA Map of Radon Zones in Indoor Air



    Figure 3-2 is not printed in the Federal Register. It is available

in the Water Docket at the address listed in the ADDRESSES section.

    Because of its short half life, there are relatively few man-made

sources of radon exposure in ground water. The most common man-made

sources of radon ground water contamination are phosphate or uranium

mining or milling operations and wastes from thorium or radium

processing. Releases from these sources can result in high ground water

exposures, but generally only to very limited populations; for

instance, to persons using a domestic well in a contaminated aquifer as

a source of potable water (US EPA 1994B).

    Table 3-1 summarizes the regional patterns of radon in drinking

water supplies as seen in the NIRS database. This survey of 1,000

ground water systems, undertaken by EPA in 1985, provides the most

representative national characterization of radon levels in drinking

water.

    However, the NIRS has the disadvantage that the samples were all

taken from within the water distribution systems, making estimation of

the naturally occurring influent radon levels difficult. In addition,

the NIRS data provide no information to allow analysis of the

variability of radon levels over time or within individual systems.



                          Table 3-1.--Radon Distributions by Region (All System Sizes)

----------------------------------------------------------------------------------------------------------------

                                                                                                     Geometric

                                                                    Arithmetic    Geometric Mean     standard

                             Region                                mean (pCi/l)     \1\ (pCi/l)    deviation \2\

                                                                                                      (pCi/l)

----------------------------------------------------------------------------------------------------------------

Appalachian.....................................................           1,127             333            4.76

California......................................................             629             333            3.09

Gulf Coast......................................................             263             125            3.38

Great Lakes.....................................................             278             151            3.01

New England.....................................................           2,933           1,214            3.77

Northwest.......................................................             222             161            2.23

Plains..........................................................             213             132            2.65

Rocky Mountains.................................................             607             361           2.77

----------------------------------------------------------------------------------------------------------------

\1\ The geometric mean is the anti-log of the average of the logarithms (log base e) of the observations.

\2\ The geometric standard deviation is the anti-log of the standard deviation of the logarithms (log base e) of

  the observations.



Source: US EPA 1998L. The values given are not population-weighted, but reflect averages across systems.



    The NIRS data illustrate the wide regional variations in radon

levels in ground water. The arithmetic mean and geometric mean radon

levels are substantially higher in New England and the Appalachian

region (in this analysis, all the States on the east coast between New

York and Florida) than in other regions of the United States. The large

differences between the geometric (anti-log of the average of the

logarithms (log base e) of the observations) and arithmetic means

indicate how ``skewed'' (i.e., ``stretched'' in a positive direction; a

bell-shaped curve with a tail out to the right) the radon distributions

are. The Agency selected a lognormal model as the best approach to

evaluating these data.

    EPA's current re-evaluation of radon occurrence in ground water

uses data from a number of additional sources to supplement the NIRS

information and to develop estimates of the national



[[Page 9570]]



distribution of radon in ground water systems of different sizes. Data

from 17 States were used to evaluate the differences between radon

levels in ground water and radon levels in distribution systems in the

same regions. The results of these comparisons were used to estimate

national distributions of radon occurrence in ground water. Table 3-2

summarizes EPA's latest characterization of the distributions of radon

levels in ground water supplies of different sizes and populations

exposed to radon through CWSs.

    In this table, radon levels and populations are presented for

systems serving various population ranges from 25 to greater than

100,000. For purpose of estimating costs and benefits, the CWSs are

aggregated to be consistent with the following system size categories

identified in the 1996 SDWA, as amended: very very small systems (25-

500 people), further subdivided into 25-100 and 101-500; very small

systems (501-3,300 people); small systems (3,301-10,000 people); medium

systems (10,001-100,000 people); and large systems (greater than

100,000 people).

    In the updated occurrence analysis, insufficient data were

available to accurately assess radon levels in the highest CWSs size

stratum. Thus, data from the two largest size strata were pooled to

develop exposure estimates for the risk and benefits assessments.

    The Agency estimates that approximately 89.7 million people are

served by community ground water systems in the United States based on

an EPA analysis of SDWIS data in 1998). The data in Table 3-2 show that

systems serving more than 500 people account for approximately 95

percent of the population served by ground water systems, even though

they represent only 40 percent the total active systems (USEPA 1997A).

The estimated system geometric mean radon levels range from

approximately 120 pCi/l for the largest systems to 312 pCi/l for the

smallest systems. Arithmetic mean values for the various size

categories range from 175 pCi/l to 578 pCi/l, and the population-

weighted arithmetic mean radon level across all the community ground

water supplies is 213 pCi/l.



                             Table 3-2.--Radon Distributions in Public Water Systems

----------------------------------------------------------------------------------------------------------------

                                                                 System size (population served)

                                                ----------------------------------------------------------------

                                                    25-100      101-500     501-3,300      3,301-      >10,000

-------------------------------------------------------------------------------------------10,000---------------

Total Systems..................................       14,651       14,896       10,286        2,538        1,536

Geometric Mean Radon Level, pCi/l..............          312          259          122          124          132

Geometric Standard Deviation...................          3.0          3.3          3.2          2.3          2.3

Population Served (Millions)...................         0.87         4.18         14.2         14.5         65.9

----------------------------------------------------------------------------------------------------------------

Radon Level, pCi/l.............................       Proportions of Systems Exceeding Radon Levels (percent)

----------------------------------------------------------------------------------------------------------------

100............................................         84.7         78.7         56.9         60.4         62.9

300............................................         51.4         45.1         22.1         14.3         16.2

500............................................         33.6         29.1         11.4          4.6          5.5

700............................................         23.4         20.3          6.8          1.8          2.3

1000...........................................         14.7         12.9          3.6          0.6          0.8

2000...........................................          4.7          4.4          0.8          0.0          0.1

4000...........................................          1.1          1.1          0.1          0.0          0.0

----------------------------------------------------------------------------------------------------------------

    Table 3-3 presents the total exposed population above each radon

level by system size category. Approximately 20% of the total

population for all system sizes are above the radon level of 300 pCi/l

and 63% are above a radon level of 100 pCi/l.



                                        Table 3-3.--Population Exposed Above Various Radon Levels By System Size

                                                                       [Thousands]

--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                           Very very     Very very

                   Radon level (pCi/l)                       small         small      Very small      Small        Medium         Large         Total

--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                               25-100     101-500      501-3,300   3,301-10K      10K-100K         >100K    ............

--------------------------------------------------------------------------------------------------------------------------------------------------------

4,000...................................................          9.4          46             20           0.2           0.9           0.4          77.2

2,000...................................................           41         183            119           5.7          21.7          11.0         381

1,000...................................................          128         541            513          85.5         289           147         1,695

700.....................................................          202         848            962         267           859           436         3,558

500.....................................................          290       1,210          1,620         672         2,070         1,050         6,893

300.....................................................          445       1,880          3,140       2,080         6,060         3,070        16,641

100.....................................................          733       3,290          8,080       8,760        23,400        11,900        56,054

--------------------------------------------------------------------------------------------------------------------------------------------------------



    Radon exposures also arise from NTNCWSs. The Agency estimates that

approximately 5.2 million people use water from NTNCWSs (US EPA 1998G).

An analysis of SDWIS data in 1998 shows there are approximately 19,500

active NTNCWSs in the United States. Over 96 percent of these systems

serve fewer than 1,000 people. EPA recently identified useful data on

radon levels in NTNCWSs from six States. A preliminary analysis of data

from these States suggested that geometric mean radon levels are

approximately 60 percent higher in NTNCWSs than in CWSs in the same

size category.

    There are currently no data which enable the agency to determine

the extent to which the populations exposed to radon from CWSs and

NTNCWSs overlap. Some portion of individuals exposed through a CWS at

home may be exposed to radon from a NTNCWS at school or at work.



[[Page 9571]]



Similarly, the same populations may be exposed to radon from two

different community systems in the course of their normal daily

activities. Further, in the case of NTNCWSs, it is possible that the

same individual could be exposed sequentially throughout their life to

radon from a series of different systems; at school, then at work, etc.

3.1.2  Exposure Pathways

    People are exposed to radon in drinking water in three ways: from

ingesting radon dissolved in water; from inhaling radon gas released

from water during household use; and from inhaling radon progeny

derived from radon gas released from water.

    Typically, indoor air contamination arising from soil gas accounts

for the bulk of total individual risk due to radon exposure (NAS

1998B). Nationally, levels of radon in household air average

approximately 1.25 pCi/l (US EPA 1992A). Usually, the bulk of the radon

enters indoor air by diffusion from soils through basement walls or

foundation cracks or openings. Radon in domestic water generally

contributes a small proportion of the total radon in indoor air. The

NAS recommends that EPA use the central estimate of a transfer factor

of 1.0 pCi/l for radon in domestic water contributing 1x10<SUP>-4</SUP>

pCi/l to indoor air. As an example, for a typical ground water CWS with

a radon level of 250 pCi/l, the increment in indoor air activity would

be 0.025 pCi/l. This is about 2 percent of the average indoor level,

which is derived mostly from soils.

    As noted, the bulk of radiation exposure through inhalation comes

from radon progeny, which tend to bind to airborne particulates. When

the particles are inhaled, they become deposited in the respiratory

tract, and further radioactive decay results in a radiation dose to the

respiratory epithelium. In contrast, when radon gas is inhaled, it is

absorbed through the lung, and much of this fraction remains in the

body only a short time before being exhaled.

    Direct ingestion of radon gas in water is the other important

exposure pathway associated with domestic water use. If water is not

agitated or heated prior to consumption, the bulk (80 to 100 percent)

of the radon remains in the water and is consequently ingested with it

(US EPA 1995). Heating, agitation (for example, by a faucet aerator),

and prolonged standing cause radon to be released and the proportion

consumed to be reduced. After a person ingests radon in water, the

radon passes from the gastrointestinal tract into the blood. The blood

then circulates the radon to all organs of the body before it is

eventually exhaled from the lungs. When radon and its progeny decay in

the body, the surrounding tissues are irradiated by alpha particles.

However, the dose of radiation resulting from exposure to radon gas by

ingestion varies from organ to organ. Stomach, followed by the tissues

of colon, liver, kidney, red marrow, and lung appear to receive the

greatest doses.

    Exposure patterns to radon vary with different exposure settings.

Depending on the relative radon levels in water and air, water use

patterns, and exposure frequency and duration, the relative

contribution of ingestion and inhalation exposure to total risks will

vary. In the case of domestic water use, inhalation of radon progeny

accounts for most of the total individual risk resulting from radon

exposure (Section 3.2). Inhalation exposure to radon from NTNCWSs is

expected to be less than for CWSs, however, because buildings served by

these systems tend to be larger, and ventilation rates higher, than the

corresponding values for domestic exposures. In addition, exposure at

these facilities tend to be less frequent and of shorter duration than

exposure from CWSs. Therefore, overall exposures at NTNCWSs will likely

be lower.



3.2  Nature of Health Impacts



    Exposure to radon and its progeny is believed to be associated with

increased risks of several kinds of cancer. When radon or its progeny

are inhaled, lung cancer accounts for most of the total incremental

cancer risk (NAS 1998A). Ingestion of radon in water is suspected of

being associated with increased risk of tumors of several internal

organs, primarily the stomach (NAS 1998B). As discussed previously, NAS

recently estimated the lifetime unit fatal cancer risks associated with

exposure to radon from domestic water use for ingestion and inhalation

pathways. EPA subsequently calculated the unit risk of inhalation of

radon gas to 0.06 percent of the total risk from radon in drinking

water, using radiation dosimetry data and risk coefficients provided by

the NAS (NAS 1998B). The lifetime unit fatal cancer risk is defined as

the lifetime risk associated with exposures to a unit concentration (1

pCi/l) of radon in drinking water. The findings are summarized in Table

3-4.



     Table 3-4.--Estimated Radon Unit Lifetime Fatal Cancer Risks in

                         Community Water Systems



------------------------------------------------------------------------

                                     Cancer unit risk    Proportion of

         Exposure pathway              per pCi/l in        total risk

                                          water            (percent)

------------------------------------------------------------------------

Inhalation of radon progeny<SUP>1</SUP>......        5.55 x 10<SUP>-7</SUP>                 89

Ingestion of radon<SUP>1</SUP>...............        7.00 x 10<SUP>-8</SUP>                 11

Inhalation of radon gas<SUP>2</SUP>..........       3.50 x 10<SUP>-10</SUP>               0.06

                                   -------------------------------------

Total.............................        6.25 x 10<SUP>-7</SUP>               100

------------------------------------------------------------------------

\1\ Source: NAS 1998B.

\2\ Source: Calculated by EPA from radiation dosimetry data and risk

  coefficients provided by NAS (NAS 1998B).



    These updated risk estimates indicate that inhalation of radon

progeny accounts for most (approximately 89 percent) of the individual

risk associated with domestic water use, with almost all of the

remainder (11 percent) resulting from ingestion of radon gas.

Inhalation of radon progeny is associated primarily with increased risk

of lung cancer, while ingestion exposure is associated primarily with

elevated risk of stomach cancer. Ingestion of radon also results in

slightly increased risk cancer of the colon, liver, and other tissues.

Inhalation of radon gas is estimated to account for approximately 0.06

percent of the total risk from household radon exposures, and the major

target organ is again believed to be the lung. In the following

sections, methods and parameter values developed by the NAS are applied

to the estimation of baseline population risks and the levels of risk

reduction associated with the different radon levels.

    Radon, a noble gas, exhibits no other known toxic effects besides

carcinogenesis. The 1998 NAS report indicates that there is no

scientific



[[Page 9572]]



evidence to show that exposure to radon is associated with reproductive

or genetic toxicity. Therefore, the endpoints characterized in the risk

assessment for radon exposure are primarily increased risk of lung and

stomach cancers.

    For the purposes of this Health Risk Reduction and Cost Analysis,

EPA is using the best estimates of radon inhalation and ingestion risks

provided by the NAS Report. In order to finalize the Agency's estimate

of lung cancer deaths arising from indoor air exposure, EPA's Office of

Radiation and Indoor Air is currently assessing various factors

integral to the approach for estimating the lung cancer risks of

inhaling radon progeny in indoor air provided in the NAS 1998 report

``The Health Effects of Exposure to Radon-BEIR VI'' (BEIR VI Report).

This assessment will be reviewed by the Agency's SAB and may result in

some adjustment to the estimated unit risk, and its associated

uncertainty, for inhalation of radon progeny used in this HRRCA



3.3  Impacts on Sensitive Subpopulations



    Populations that might experience disproportional risk as a result

of radon exposure fall into two general classes: those who might

receive higher exposures per unit radon in water supplies and those who

are more sensitive to the exposures they receive. The former group

includes persons whose domestic water supplies have high radon levels,

and whose physiological characteristics or behaviors (high metabolic

rate, high water consumption, large amounts of time spent indoors)

result in high exposures per unit of exposure concentration. As noted

above, a portion of the population could be exposed to radon from more

than one source. For example, a student or worker might be exposed to

radon from the CWS in the household setting and also from a NTNCWS (or

from the same or different CWS) at school or work.

    Different age and gender groups may also experience exposure

dosimetric differences. These differences in radiation dose per unit

exposure have been taken into account in the BEIR VI Report addressing

radon in indoor air (NAS 1998A), the NAS Report addressing radon in

drinking water (NAS 1998B), and the EPA Federal Guidance Report 13 (US

EPA 1998F).

    The NAS Report concluded that there is insufficient scientific

information to permit separate cancer risk estimates for subpopulations

such as pregnant women, the elderly, children, and seriously ill

persons. The report did note, however, that according to the NAS risk

model for the cancer risk from ingested radon, which accounts for 11%

of the total lifetime fatal cancer risk from radon in drinking water,

approximately 30% of this fatal lifetime cancer risk is attributed to

exposure between ages 0 to 10.

    The NAS did identify smokers as the only group that is more

susceptible to inhalation exposure to radon progeny. Inhalation to

cigarette smoke and radon progeny result in a greater increased risk

than if the two exposures act independently to induce lung cancer.



3.4  Risk Reduction Model for Radon in Drinking Water



    Risk and risk reduction were estimated using a Monte Carlo model

that simulated the initial and post-regulatory distributions of radon

activity levels and population cancer risks. Each iteration of the

model selected a size stratum of community water systems. The system

sizes were stratified according to the following populations served:


<100; 101-500; 501-3,300; 3,301-10,000; and > 10,000 served. For each

size category, a lognormal distribution of uncontrolled radon levels

had been defined based on the updated occurrence analysis (USEPA

1998L). The model sampled randomly from the radon distribution for the

selected CWS size category to determine if the radon level was above

the selected maximum exposure level. The proportion of iterations

choosing each size stratum were determined by the relative national

populations served by each size stratum of systems. Thus, over a large

number of iterations (generally, benefit calculations were carried out

using 20,000 to 50,000 iterations), the model produced a population-

weighted distribution of radon levels.

    In each iteration of the model, the simulated influent radon

activity level was compared to the maximum radon levels under

consideration (100, 300, 500, 700, 1000, 2000, and 4000 pCi/l). When

the simulated influent radon level was less than the target level, the

simulated level was passed directly to the risk calculation equations.

The equations calculated population fatal cancer risks from ingestion

of radon gas, inhalation of radon gas, and inhalation of radon progeny

using standard exposure factors and unit risk values derived by the

NAS.

    When the simulated influent radon level in a given iteration

exceeded a target radon level, the model reduced the value by a

proportion equivalent to the performance of selected mitigation

technologies. The degrees of reduction are presented in Table 3-5:



   Table 3-5.--Radon Treatment Assumptions to Calculate Residual Fatal

                              Cancer Risks

------------------------------------------------------------------------

           If the radon level is              Then the treated level is

------------------------------------------------------------------------

Less than the target level................  None; Influent = Effluent.

Above but less than two times the target    Influent = 0.5  x  Effluent.

 level.

Above two times but less than five times    Influent = 0.2  x  Effluent.

 the target level.

Greater than five times the target level..  Influent = 0.01 Effluent.

------------------------------------------------------------------------



    Using this approach implies that a greater level of control is

achieved than if all the systems were simply assumed to reduce

exposures to the maximum exposure level. For example, a system with an

initial uncontrolled concentration of 400 pCi/l would need to employ a

mitigation technology with a 50 percent removal efficiency to comply

with a maximum exposure limit of 300 pCi/l, resulting in a final radon

level of 200 pCi/l. Limited sensitivity analysis suggests that this

approach does not provide very much in the way of extra risk reduction.

The preponderance of population risk reduction is achieved by reducing

radon levels in the relatively few systems that have initial

uncontrolled values far above the maximum exposure limits, not by the

relatively small incremental reductions below the target radon levels.



3.5  Risks From Existing Radon Exposures



    In support of the regulatory development process for the revised

radon rule, EPA has updated its risk assessment for radon exposures in

drinking water. Previously, EPA developed estimates of risk from total

population exposure to radon in drinking water in support of the

proposed rule for radon in 1991 (US EPA 1991). In response to comments

from the SAB, EPA updated the risk assessment to include an analysis of

uncertainty in 1993 (US EPA 1993B). The assessment was further revised

to include revisions to risk factors and other variable values. The

latest uncertainty analysis was completed in 1995 (US EPA 1995).

    EPA's revised risk analysis in support of this HRRCA takes into

account new data on radon distributions and exposed populations

developed in the updated occurrence analysis, as well as new

information on dose-response relationships developed by the NAS (NAS

1998B). For the HRRCA,



[[Page 9573]]



population risks are estimated using single-value ``nominal'' estimates

of the various exposure factors which determine individual risk, and

Monte Carlo simulation techniques are used to estimate risks associated

with the distributions of radon exposures from the various size

categories of CWSs. The risk equations and parameter values used in the

revised risk assessment are summarized in Appendix A. EPA is currently

conducting a comprehensive uncertainty analysis of radon risks using

two-dimensional Monte Carlo methods to better judge the level of

uncertainty associated with the radon risk estimates.

    Table 3-6 summarizes the results of EPA's revised baseline risk

assessment. Because the NAS and EPA-derived dose-response and exposure

parameters factors discussed above were used in the risk assessment,

the proportions of risk associated with the various pathways were the

same as shown in Table 3-4. The total estimated population risks

associated with the current distribution of radon in CWSs was 160 fatal

cancers per year, 142 of which were associated with progeny inhalation.

Approximately 18 fatal cancers per year were associated with ingestion

of radon. These totals are similar to, but somewhat lower than, EPA's

1991 and 1993 baseline risk estimates (US EPA 1994C). In comparison,

there are an estimated 15,400 to 21,800 fatal lung cancers per year due

to inhalation of indoor air contaminated with radon emanating from soil

and bedrock (NAS 1998A).

    The risks summarized in Table 3-5 do not include any contribution

from NTNCWSs, Thus, the potential baseline risks and benefits of a

radon rule may be somewhat underestimated. The limited available data

concerning radon levels in NTNCWSs suggest that levels may be

considerably higher (perhaps by 60 percent, on average) than those in

CWSs of similar size (US EPA 1998L). However, it appears that the

average exposure per unit activity in NTNCWSs is likely to be lower

than that for CWSs. Because of the expected lower inhalation exposures,

water ingestion rates, and frequencies and durations of exposure, the

individual fatal cancer risk associated with a NTNCWS is expected to be

lower compared to a CWS with similar radon levels. EPA is currently

conducting additional analyses of NTNCWS exposures from radon in an

attempt to refine the current approximate risk estimates.



            Table 3-6.--Annual Fatal Cancer Risks for Exposures to Radon From Community Water Systems

----------------------------------------------------------------------------------------------------------------

                                                                    Annual unit

                                                                    risk (fatal       Annual

                                                                    cancers per     population     Proportion of

                             Pathway                                person per      risk (fatal    total annual

                                                                  year per pCi/l    cancers per   risk (percent)

                                                                    in water)<SUP>1</SUP>        year) <SUP>2</SUP>

----------------------------------------------------------------------------------------------------------------

Inhalation of progeny...........................................     7.44 x 10<SUP>-9</SUP>             142              89

Ingestion of radon gas..........................................    9.30 x 10<SUP>-10</SUP>            17.8              11

Inhalation of radon gas.........................................     4.7 x 10<SUP>-12</SUP>             0.1            0.06

      Total.....................................................     8.37 x 10<SUP>-9</SUP>             160            100

----------------------------------------------------------------------------------------------------------------

\1\ Derived using NAS lifetime unit fatal cancer risks.

\2\ Estimated through simulation analysis described in Section 3.4; the risk equations and parameter values used

  in the simulation analysis are summarized in Appendix A.



3.6  Potential for Risk Reductions Associated With Removal of Co-

Occurring Contaminants





    Because radon is a naturally occurring ground water contaminant,

its occurrence patterns are not highly correlated with those of

industrial pollutants. Similarly, the Agency's re-evaluation of radon

occurrence has revealed that the geographic patterns of radon

occurrence are not significantly correlated with naturally occurring

inorganic contaminants that may pose health risks. Thus, it is not

likely that a relationship exists between the need to install radon

treatment technologies and treatments to remove other contaminants.

    On the other hand, technologies used to reduce radon levels in

drinking water have the potential to reduce concentrations of other

pollutants as well. All of the aeration technologies discussed remove

volatile organic contaminants, as well as radon, from contaminated

ground water. Similarly, GAC treatment for radon removal effectively

reduces the concentrations of organic (both volatile and nonvolatile)

chemicals and some inorganic contaminants. Aeration also tends to

oxidize dissolved arsenic (a known carcinogen) to a less soluble form

that is more easily removed from water. The frequency with which radon

treatment would also reduce risks from other contaminants, and the

extent of risk reduction that would be achieved, has not been evaluated

quantitatively in the HRRCA.



3.7  Potential for Risk Increases From Other Contaminants Associated

With Radon Removal



    As discussed in Section 5.1, the need to install radon treatment

technologies may require some systems that currently do not disinfect

to do so. While case studies (US EPA 1998D) of twenty-nine small to

medium water systems that installed treatment (24 aeration, 5 GAC) to

remove radon from drinking water revealed only two systems that

reported adding disinfection (both aeration) with radon treatment (the

systems either had disinfection already in place or did not add it), in

practice the tendency to add disinfection may be much more significant

than these case studies indicate. EPA also realizes that the addition

of chlorination for disinfection may result in risk-risk tradeoffs,

since, for example, the disinfection technology reduces potential for

infectious disease risk, but at the same time can result in increased

exposures to disinfection by-products (DBPs). This risk-risk trade-off

is addressed by the recently promulgated Disinfectants and Disinfection

By-Products NPDWR (US EPA 1998I). This rule identified MCLs for the

major DBPs, with which all CWSs and NTNCWSs will have to comply. These

MCLs set a risk ceiling from DBPs that water systems adding

disinfection in conjunction with treatment for radon removal could

face. The formation of DBPs is proportional to the concentration of

organic precursor contaminants, which tend to be much lower in ground

water than in surface water.

    The NAS Report addressed several important potential risk-risk

tradeoffs associated with reducing radon levels in drinking water,

including the trade-off between risk reduction from radon treatment

that includes post-disinfection with the increased potential for DBP

formation (NAS 1998B). The



[[Page 9574]]



report concluded that, based upon median and average total

trihalomethane (THM) levels from EPA's 1981 Community Water System

Survey, a typical ground water CWS will face an incremental individual

lifetime cancer risk due to chlorination byproducts of

5x10<SUP>-5</SUP>. It should be emphasized that this risk is based on

average and median THM occurrence information that does not segregate

systems that disinfect from those that do. Further, the NAS Report

points out that this average DBP risk is smaller than the average

individual lifetime fatal cancer risk associated with baseline radon

exposures from ground water (untreated for radon), which is estimated

at 1.2 x 10<SUP>-4</SUP> using a mean radon concentration of 213 pCi/l.

    A more meaningful comparison is to look at the trade-off between

risk reduction from radon treatment in cases where disinfection is

added with the added risks from DBP formation. This trade-off will

affect only a minority of systems since a majority of ground water

systems already have disinfection in place. For the smallest systems

size category, approximately half of all CWSs already have disinfection

in place. The proportions of systems having disinfection in place

increases as the size categories increase, up to >95% for large systems

(Table 5-2). In addition, although EPA is using the conservative

costing assumption that all systems adding aeration or GAC would

disinfect, not all systems adding aeration or GAC would have to add

post-disinfection or, if disinfecting, may use a disinfection

technology that does not forms DBPs. For those ground water systems

adding treatment with disinfection, this trade-off tends to be

favorable since the combined risk reduction from radon removal and

microbial risk reduction outweigh the added risk from DBP formation.

    An estimate of the risk reduction due to treatment of radon in

water for various removal percentages and finished water concentrations

is provided in Table 3.7. As noted by the NAS Report, these risk

reductions outweigh the increased risk from DBP exposure for those

systems that chlorinate as a result of adding radon treatment.



              Table 3-7.--Radon Risk Reductions Across Various Effluent Levels and Percent Removals

----------------------------------------------------------------------------------------------------------------

                                                  Risk reduction  Risk reduction  Risk reduction  Risk reduction

                  % Removal \1\                     @ 50 pCi/L      @ 100 pCi/L     @ 200 pCi/L     @ 300 pCi/L

----------------------------------------------------------------------------------------------------------------

60..............................................          \2\ NA              NA         1.9E-04         2.8E-04

80..............................................              NA         2.5E-04         5.0E-04         7.6E-04

90..............................................         2.8E-04         5.7E-04         1.1E-03         1.7E-03

99..............................................         3.1E-03         6.2E-03         1.2E-02        1.9E-02

----------------------------------------------------------------------------------------------------------------

\1\ Influent levels used in risk reduction calculations are determined by the relationship, Effluent Level =

  Influent Level*(1--%Removal/100).

\2\ NA = Not applicable since associated influent level would be outside the range of realistic values.



    Comparing the risk reductions in Table 3.7 to the risks from THMs

at their MCL values (the maximum risk allowable under the DBP rule),

the ratios between risk reduction from radon removal and the

conservative assumption that DBPs are present at their MCL values are

shown in Table 3.8.



                      Table 3-8.--Radon Risk Reduction from Treatment Compared to DBP Risks

----------------------------------------------------------------------------------------------------------------

                                                   Estimated risk ratios (risk reduction from radon removal/risk

                                                                     from THMs at 0.080 mg/L)

                  % Removal \1\                  ---------------------------------------------------------------

                                                  Ratio @ 50 pCi/   Ratio @ 100     Ratio @ 200     Ratio @ 300

                                                         L             pCi/L           pCi/L           pCi/L

----------------------------------------------------------------------------------------------------------------

60..............................................          \2\ NA              NA             1.6             2.4

80..............................................              NA             2.1             4.2             6.3

90..............................................             2.4             4.7             9.5            14.2

99..............................................            26.0            52.0           104.0          155.9

----------------------------------------------------------------------------------------------------------------

Notes: \1\ Influent levels used in risk reduction calculations are determined by the relationship, Effluent

  Level = Influent Level*(1--%Removal/100).

\2\ NA = Not applicable since associated influent level would be outside the range of realistic values.



    As can be seen in Table 3.8, the risk ratios are favorable for

treatment with disinfection, ignoring microbial risk reduction, even

assuming the worst case scenario that ground water systems have THM

levels at the MCL. There is the possibility that accounting

quantitatively for the increased risk from DBP exposure for systems

adding chlorination in conjunction with treatment for radon may

somewhat decrease the monetized benefits estimates.



3.8  Risk for Ever-Smokers and Never-Smokers



    As noted previously, cancer risks from inhalation of radon progeny

are believed to be greater for current and former smokers than for

``never smokers''. The NAS defines a ``never smoker'' as someone who

has smoked less than 100 cigarettes in their lifetime. Therefore,

``ever smokers'' include current and former smokers. EPA and NAS have

developed estimates of unit risk values (estimates of cancer risks per

unit of exposure) for radon progeny for ``ever-smokers'' and ``never-

smokers'' as shown in Table 3-9 (US EPA 1999A). The estimated unit risk

values for inhalation of radon progeny for ever-smokers (and therefore

the individual and population risk) is approximately 5.5 times greater

than that for never smokers.

    Because of estimated higher individual risks for smokers, this

group accounts for a large proportion of the overall population risk

associated with radon progeny inhalation. The last two columns of the

table show that, given the current assumptions about smoking prevalence

and the relative impact of radon progeny on ever smokers and never

smokers, about 85 percent of the cancer cases from water exposures to



[[Page 9575]]



progeny will occur in the ever-smoker population.



 Table 3-9.--Annual Lung Cancer Death Risk Estimates From Radon Progeny for Ever-Smokers, Never-Smokers, and the

                                               General Population

----------------------------------------------------------------------------------------------------------------

                                                    Annual unit

                                                    risk (fatal   Average annual      Annual       Proportion of

                                                   cancer cases     individual      population     total annual

                 Smoking status                    per year per    risk per year    risk (fatal     population

                                                     pCi/l in       of exposure     cancers per        risk

                                                      water)                           year)

----------------------------------------------------------------------------------------------------------------

Ever............................................       1.31X10-8        2.8X10-6             120              85

Never...........................................       2.44X10-9        5.1X10-7              22              15

Combined........................................       7.44X10-9        1.6X10-6             142             100

----------------------------------------------------------------------------------------------------------------

Source: EPA analyses derived from NAS (1998) estimates.



Note: Ever-smoking prevalence was assumed to be 58 percent in males and 42 percent in females, and these rates

  were assumed to be age independent.



4. Benefits of Reduced Radon Exposures



4.1  Nature of Regulatory Benefits



4.1.1  Quantifiable Benefits

    The benefits of controlling exposures to radon in drinking water

take the form of avoided cancers resulting from reduced exposures.

Cancer risks (both fatal and non-fatal cancers per year) are calculated

using the risk model described in Section 3 for the baseline case

(current conditions) and each of the radon levels. The health benefits

of controls are estimated as the baseline risks minus the residual

risks associated with each radon level. The more stringent the radon

level, the lower the residual risks, and the higher the benefits.

    The primary measures of regulatory benefits that are used in this

analysis are the annual numbers of fatal and non-fatal cancers

prevented by reduced exposures. Due to a lack of knowledge about how to

account for the latency period for radon-induced cancers, it has been

assumed that risk reduction begins to accrue immediately after the

reduction of exposures.

    Exposures to radon and its progeny are associated with increases in

lung cancer risks. Ingestion of radon in drinking water is suspected of

being associated primarily with increased risks of tumors of the

stomach, and with lesser risks to the colon, lung, and other organs.

The first column of Table 4-1 summarizes the estimates of the

distribution of cancers by organ system for inhalation and ingestion

exposures given. For purposes of the risk assessment, inhalation of

progeny and radon gas are assumed to be associated exclusively with

lung cancer risk. In the case of radon ingestion, stomach cancer

accounts for the bulk (approximately 87 percent) of the total risk by

this pathway. Cancers of several other organ systems account for far

smaller proportions of the cancer risk from radon ingestion, and are

not included in this analysis.



               Table 4-1.--Proportion of Fatal Cancers by Exposure Pathway and Estimated Mortality

----------------------------------------------------------------------------------------------------------------

                                                                                   Proportion of

                                                                                   fatal cancers

                                                                                   by organ and      Mortality

              Exposure pathway                          Organ affected               exposure      (percent) \2\

                                                                                      pathway

                                                                                   (percent) \1\

----------------------------------------------------------------------------------------------------------------

Inhalation of progeny, radon gas...........  Lung...............................              89              95

Ingestion of radon gas.....................  Stomach............................             9.5              90

                                             Colon..............................             0.4             550

                                             Liver..............................             0.3              95

                                             Lung...............................             0.2              95

                                             General Tissue.....................             0.5              --

----------------------------------------------------------------------------------------------------------------

\1\  Source: US EPA analysis of dosimetry data and organ-specific risk coefficients (NAS 1998).

\2\ Source: US EPA analysis of National Cancer Institute mortality data.



    The last column of Table 4-1 provides estimates of the mortality

rate associated with the various types of radon-associated cancers.

These values are used in this analysis to estimate the proportion of

fatal and non-fatal cancers by organ system and exposure pathway. Both

of the cancers that account for the bulk of the risk from radon and

progeny exposures (lung and stomach) have high mortality rates.

4.1.2  Non-Quantifiable Benefits

    Reductions in radon exposures might also be associated with non-

quantifiable benefits. EPA has identified several potential non-

quantifiable benefits associated with regulating radon in drinking

water. These include any peace of mind benefits specific to reduction

of radon exposure that may not be adequately captured in the VSL

estimate. In addition, treating radon in drinking water with aeration

oxidizes arsenic into a less soluble form that is easier to remove with

conventional arsenic removal technologies. In terms of reducing radon

exposures in indoor air, it has also been suggested that provision of

information to households on the risks of radon in indoor air and

available options to reduce exposure is a non-quantifiable benefit that

can be attributed to some components of a MMM program. Providing such

information might allow households to make informed choices about the

appropriate level of risk reduction given their specific circumstances

and concerns. These potential benefits are



[[Page 9576]]



difficult to quantify due to the uncertainty surrounding their

estimation. However, they are likely to be somewhat less in magnitude

relative to the monetized benefits estimates.



4.2  Monetization of Benefits



4.2.1  Estimation of Fatal and Non-Fatal Cancer Risk Reduction

    The ``direct'' health benefits of the regulation, as discussed

above, are the reduced streams of cancer cases associated with reduced

radon exposures. In this analysis, the data in Table 3-6 were used to

estimate the numbers of fatal cancers of each organ system associated

with inhalation and ingestion pathway from the risk model described in

Section 3.1. (These proportions, by the nature of the risk model that

is used, stay constant for all radon levels.) Subsequently, the total

number of cancers of each organ system was estimated. This is necessary

because the output of the risk model is fatal cancers, and the cost of

illness and willingness to pay for non-fatal cancers are only applied

to individuals who survive the disease. The total number of cancers per

year of exposure, and the number of non-fatal cancers were estimated

from the fatal cancer numbers using the mortality data in Table 4-1.

Thus, for example, a benefit of 100 cases of fatal lung cancer avoided

implies approximately 105 total lung cancers avoided, five of which are

non-fatal. This calculation omits rounding error, and the total number

of cases is equal to the fatal cases divided by the mortality rate.

    Fatal and non-fatal population cancer risks under baseline

conditions were estimated first. Then, the residual cancer risks were

estimated for each of the radon levels. Consistent with the assumptions

made in the cost analysis, residual water radon levels were calculated

using a similar range of technology efficiencies. Radon levels were

assumed to be reduced below baseline levels by either 50, 80, or 99

percent, using the least stringent reduction which could comply with

the radon level under evaluation. Benefits took the form of the

reductions in the numbers of fatal and non-fatal cancers associated

with each final level compared to the baseline risks.

4.2.2  Value of Statistical Life for Fatal Cancers Avoided

    As one measure of potential benefits, this analysis assigns the

monetary value of a statistical life saved to each fatal cancer

avoided. The estimation of the value of a statistical life involves

inferring individuals' implicit tradeoffs between small changes in

mortality risk and monetary compensation (US EPA 1998E). A central

tendency value of $5.8 million (1997$) is used in the monetary benefits

calculations, with low- and high-end values of $700,000 (1997$) and

$16.3 million (1997$), respectively, used for the purposes of

sensitivity analysis. These figures span the range of value of

statistical life (VSL) estimates from 26 studies reviewed in EPA's

recent guidance on benefits assessment (US EPA 1998E) which is

currently being reviewed by EPA's SAB and the Office of Management and

Budget (OMB). It is important to recognize the limitations of existing

VSL estimates and to consider whether factors such as differences in

the demographic characteristics of the populations and differences in

the nature of the risks being valued have a significant impact on the

value of mortality risk reduction benefits. As noted above, no separate

medical care or lost-time costs are included in the benefits estimate

for fatal cancers because it is assumed that these costs are captured

in the VSL for fatal cancers.

4.2.3  Costs of Illness and Lost Time for Non-Fatal Cancers

    Two important elements in the estimation of the economic impacts of

reduced cancer risks for non-fatal cancers are the reductions in

medical care costs and the costs of lost time. The costs of medical

care represent a net loss of resources to society (not considering the

economic hardship on the cancer patient and family). The cost of lost

time represents the value of activities that the individual must

abandon (e.g., productive employment or leisure) as a result of radon-

induced cancer. Together, these two elements are often referred to as

the costs of illness (COI).

    Medical care and lost-time costs have been estimated for lung and

stomach cancers, which are the two most common types of tumors

associated with radon exposures, and which account for 99 percent of

the total radon-associated cancers. Table 4-2 summarizes the Agency's

latest medical care and lost-time cost estimates for lung cancer (US

EPA 1998B, 1998C). Medical care costs have been estimated from survey

data for ten years after initial diagnosis. The medical costs in the

first year correspond to the costs of initial treatment, while medical

costs in subsequent years correspond to the average medical costs

associated with monitoring and treatment of recurrences among

individuals who survive to that year. These out-year costs are weighted

by the proportion of patients surviving to the given year.

    The lost time due to the radon-induced tumors is assumed to be

concentrated in the first year after diagnosis. This is why the out-

year estimates for the costs of lost time in Table 2-8 are all zero.

The dollar costs of lost time given in the table are derived by

assigning values lost productive (work) and leisure (non-productive)

hours. The costs given in the top row of Table 4-2 correspond to 776

lost productive hours and 1,493 lost leisure hours per patient. The

estimates of lost hours are relatively low for lung cancer primarily

because the average age at diagnosis is advanced (fewer than 34 percent

of lung cancer patients are diagnosed before age 65).

    Using a discount rate of seven percent, the estimated discounted

present value in 1997 dollars of combined medical care and lost-time

costs for a cancer survivor is approximately $108,000. The estimated

value varies with different discount rates. Using a discount rate of

three percent, combined costs are $121,600; at ten percent, combined

costs are approximately $100,200.

    Table 4-3 summarizes the estimation of medical and lost-time costs

for survivors of stomach cancer. The combined discounted costs for

stomach cancer are similar to those for lung cancer, but slightly

higher. At a seven percent discount rate, combined discounted costs for

stomach cancer are approximately $114,000 (1997$). At three percent,

they are about $126,300 (1997$). Discounted at ten percent, the average

combined cost is $106,400 (1997$).

          Table 4-2.--Estimated Medical Care and Lost-Time Costs Per Case for Survivors of Lung Cancer

----------------------------------------------------------------------------------------------------------------

                                                            Medical care       Cost of lost       Cost of lost

                                                               costs             leisure        productive time

                  Year after diagnosis                     (undiscounted      (undiscounted      (undiscounted

                                                         1997 dollars) \1\  1997 dollars) \2\  1997 dollars) \2\

----------------------------------------------------------------------------------------------------------------

1......................................................            $34,677             $9,886            $14,393



[[Page 9577]]





2......................................................              9,936                  0                  0

3......................................................              9,383                  0                  0

4......................................................              8,969                  0                  0

5......................................................              8,604                  0                  0

6......................................................              8,262                  0                  0

7......................................................              7,934                  0                  0

8......................................................              7,609                  0                  0

9......................................................              7,287                  0                  0

10.....................................................              6,974                  0                  0

Discounted Present Value at 7 Percent..................             85,225              9,390             13,671

Total Discounted Value (1997 dollars)..................           108,287

----------------------------------------------------------------------------------------------------------------

\1\ Medical care cost estimates derived from US EPA 1998B.

\2\ Lost productive and leisure hours estimates from US EPA 1998B; value of productive time estimated at $12.47/

  hr, value of leisure hour estimated at $9.64/hour (from US EPA 1998J).





         Table 4-3.--Estimated Medical Care and Lost-Time Costs Per Case for Survivors of Stomach Cancer

----------------------------------------------------------------------------------------------------------------

                                                                              Cost of lost        Cost of lost

                                                       Medical care costs        leisure        productive time

                 Year after diagnosis                  (Undiscounted 1997  (undiscounted 1997    (undiscounted

                                                          dollars) \1\        dollars) \2\     1997 dollars) \2\

----------------------------------------------------------------------------------------------------------------

1....................................................          $37,507.28          $19,337.84             13,288

2....................................................            9,328.23                0                     0

3....................................................            8,749.24                0                     0

4....................................................            8,265.39                0                     0

5....................................................            7,829.62                0                     0

6....................................................            7,423.51                0                     0

7....................................................            7,035.81                0                     0

8....................................................            6,663.46                0                     0

9....................................................            6,300.32                0                     0

10...................................................            5,946.38                0                     0

Discounted Present Value at 7 Percent................           82,997.35           18,368                12,621

Total Discounted Value (1997 dollars)................         113,987

----------------------------------------------------------------------------------------------------------------

\1\ Medical care cost estimates derived from US EPA 1998C.

\2\ Lost productive and leisure hours estimates from US EPA 1998C; value of productive time estimated at $12.47/

  hr, value of leisure hour estimated at $9.64/hour (from US EPA 1998J).



4.2.4  Willingness to Pay to Avoid Non-Fatal Cancers

    As was the case for fatal cancers, willingness to pay (WTP)

measures of the values of avoiding serious non-fatal illness have also

been developed. These WTP measures were developed because the cost of

illness estimates may be seen as understating total willingness to pay

to avoid non-fatal cancers. The main reason that the cost of illness

understates total WTP is the failure to account for many effects of

disease--it ignores pain and suffering, defensive expenditures, lost

leisure time, and any potential altruistic benefits (US EPA 1998E).

Recently, EPA applied one such study to evaluate the benefits of

avoiding non-fatal cancers in the Regulatory Impact Analysis for the

Stage I Disinfection By-Products Rule (US EPA 1998M). That study

estimated a range of WTP to avoid chronic bronchitis ranging from

168,600 to 1,050,000 with a central tendency (mean) estimate of 536,000

(Viscusi et al. 1991). In the benefits assessment, EPA uses the central

tendency measure as a surrogate for the cost of avoiding non-fatal

cancers and an alternative to the cost of illness measures discussed

above. The high and low ends of the range are used in sensitivity

analysis of the monetized benefit estimates.



4.3  Treatment of Monetized Benefits Over Time



    The primary measures of regulatory benefits that are used in this

analysis are the annual numbers of expected fatal and non-fatal cancers

prevented by reduced exposures to radon in drinking water. The monetary

valuation of fatal cancer risks used is a result of a benefits transfer

exercise from the risk of immediate accidental death to the risk of

fatal cancer. No adjustments to the benefits calculations have been

made to reflect the time between the reduction in exposure and the

diagnosis and illness or possible death from cancer. Also, no

adjustments have been made for any other factors which might affect the

valuation. Cancer valuations could be adjusted for how they differ from

accidental death valuations with respect to timing (latency) and with

respect to other factors that may affect individuals' willingness-to-

pay for cancer risk reduction, including dread, pain and suffering, the

degree to which the risk is voluntary or involuntary, and the amount by

which life spans are shortened. Such adjustments have been under debate

in the academic literature. In the absence of quantitative evidence on

the relative impact of each factor, EPA has not adjusted the benefits

estimates in this HRRCA to account for the factors discussed here. The

Agency is currently reviewing the various issues raised; at this time

no Agency policy regarding any such adjustments is in place.



[[Page 9578]]



5. Costs of Radon Treatment Measures



    This section describes how the costs and economic impacts of

reductions in radon exposures were estimated. The most commonly used

and cost-effective technologies for mitigating radon are described,

along with the degree of radon removal that can be achieved. Costs of

achieving specified radon removal levels for specific flow rates are

discussed, along with the need for pre-and post-treatment technologies.

The methods used to estimate treatment costs for single systems and

aggregate national costs are explained, and the approach for

translating the costs into economic impacts on affected entities is

also described.



5.1  Drinking Water Treatment Technologies and Costs



    The two most commonly employed methods for removing radon from

water supplies are aeration and granular activated carbon (GAC)

absorption. These treatment approaches can be technically feasible and

cost-effective over a wide range of removal efficiencies and flow

rates. In addition to the radon treatment technologies themselves,

specific pre-or post-treatment technologies may also be required. When

influent iron and manganese levels are above certain levels, pre-

treatment may be required to remove or sequester these metals and avoid

fouling the radon removal equipment. Also, aeration and GAC absorption

may introduce possible infectious particulates into the treated water.

Thus, disinfection is generally required as a post-treatment when radon

reduction technologies are installed.

    When only low removal efficiency is required, and sufficient

capacity is available, simple storage may in some cases be sufficient

to reduce radon levels in water below specified radon levels. Radon

levels rapidly decrease through natural radioactive decay, and if

storage is in contact with air, through volatilization. Therefore,

storage has also been included in the cost analysis.

    In some cases, water systems will choose to seek other sources of

water rather than employ expensive treatment technologies. Systems may

choose a number of strategies, such as shutting down sources with high

radon levels and pumping more from sources with low levels, or

converting from ground water to surface water. In the cost analysis,

however, it has been assumed that such options will not be available to

most systems, and they will need to obtain water from other systems.

This option is referred to as ``regionalization'' in the following

discussions.

    These general families of technologies, along with the specific

variants used in the cost analysis, are described.

5.1.1  Aeration

    Because of radon's volatility, when water containing radon comes

into contact with air, the radon rapidly diffuses into the gas phase.

Several aeration technologies are available. As will be discussed in

more detail below, the specific technology adopted in response to the

rule will depend on the system's influent radon level, size, and the

degree of radon removal that is required. The following common aeration

technologies have been included in this analysis. Other aeration

technologies are available (spray aeration, tray aeration, etc.) that

can potentially be used by water systems to remove radon. These

technologies have not been included in the analysis either because they

have technical characteristics that limit their use in public water

systems, or because their removal efficiencies are lower, and/or their

unit costs are higher than the three aeration technologies included in

the analysis.

    Packed Tower Aeration (PTA). During PTA treatment, the water flows

downward by gravity and air is forced upward through a packing material

that is designed to promote intimate air-water contact. The untreated

water is usually distributed on the top of the packing with sprays or

distribution trays and the air is blown up a column by forced or

induced draft. This design results in continuous and thorough contact

of the liquid with , air (US EPA 1998O). In terms of radon removal, PTA

is the most effective aeration technology. Radon removal efficiencies

of up to 99.9 percent are technically feasible and not prohibitively

expensive for most applications. In this analysis, two different PTA

treatments are used to estimate radon removal cost. The costs are

dependant on the degree of reduction required to achieve compliance

with the allowable radon level. The first design is capable of reducing

radon levels by 80 percent; the second and more costly version reduces

radon in drinking water by 99 percent.

    Diffused Bubble Aeration (DA). Aeration is accomplished in the

diffused-air type equipment by injecting bubbles of air into the water

by means of submerged diffusers or porous plates. The untreated water

enters the top of the basin and exits from the bottom [having been]

treated, while the fresh air is blown from the bottom and is exhausted

from the top (US EPA 1998O). Diffused bubble aeration can achieve radon

removal efficiencies greater than 90 percent. In this analysis, a DA

system with a removal efficiency of 80 percent is used as the basis for

estimating compliance costs.

    Multiple Stage Bubble Aeration (MSBA). MSBA is a variant of DA

developed for small to medium water supply systems (US EPA 1998O). MSBA

units consist of shallow, partitioned trays. Water passes through

multiple stages of bubble aeration of relatively shallow depth. In this

analysis, an MSBA radon removal efficiency of 80 percent is assumed.

    All of the aeration technologies discussed above are assumed to be

``central'' treatments in the cost analysis. That is, a single large

installation is used to treat water from a given source, prior to the

water entering the distribution system to serve many users. It is also

technically feasible to apply some of these technologies at the point

of entry (e.g. just before water from the distribution system enters

the household where it is to be used). However, most aeration

technologies are only cost-effective at minimum flows far above that

corresponding to the water usage rate of a typical household, and thus

would not likely be selected as the treatment of choice.

    Also, in all of the aeration systems just discussed, the radon

removed from water is released to ambient (outdoor) air. In this

analysis, it has been assumed that the air released from aeration

systems will not itself require treatment, result in appreciable risks

to public health, or result in increased permitting costs for water

systems. For the 1991 proposed rule, EPA conducted analyses on radon

emissions and potential risks associated with radon and its progeny as

they disperse from a water treatment facility (US EPA 1988, 1989). In

summary, these analyses concluded that the annual risk of fatal cancer

from radon and its progeny in off-gas emissions was 2,700 times smaller

(108 cases/0.04 cases) than the annual risk of fatal cancer from radon

and its progeny from tap water after all ground water systems were at

or below the 1991 target level of 300 pCi/L. Using the occurrence

estimates at that time, the off-gas risk was estimated to be 4800 times

smaller (192 cases/0.04 cases) than the radon in tap water risk if no

water mitigation was done (US EPA 1994C). The EPA's SAB reviewed the

Agency's report and concluded that: (1) while the uncertainty analysis

could be upgraded to lend greater scientific credibility, the results

of modeling would not likely change, i.e., the risk posed by release of

radon through treatment would be less



[[Page 9579]]



than that posed by drinking untreated water; and (2) it is likely that

the conservative assumptions adopted by EPA in its air emissions

modeling resulted in overestimates of risk (US EPA 1994C).

5.1.2  Granular Activated Carbon (GAC)

    The second major category of radon removal technology is treatment

with granular activated carbon. GAC adsorption removes contaminants

from water by the attraction and accumulation of the contaminant on the

surface of carbon. The magnitude of the available surface area for

adsorption to occur is of primary importance, while other chemical and

electrochemical forces are of secondary significance. Therefore, high

surface area is an important factor in the adsorption process (US EPA

1998O). GAC systems are commonly used in water supply systems to remove

pesticides or other low-volatility organic chemicals that cannot be

removed by aeration. Radon can also be captured by GAC filtration, but

the amounts of carbon and the contact times needed to produce a high

degree of radon removal are generally much greater than those required

to remove common organic contaminants. For most system sizes and design

configurations evaluated in this study, aeration can achieve the same

degree of radon reduction at lower cost than GAC. However, in the cost

analysis for the radon rule, it has been assumed that a small minority

of systems will nonetheless choose GAC technology over aeration

alternatives, due to system-specific needs (e.g., land availability).

Also, POE GAC (see below) may be cost-effective for systems serving

only a few households. Depending on the specific design and operating

characteristics, GAC can remove up to 99.9 percent of influent radon,

but high removal efficiencies require large amounts of carbon and long

contact times.

    Two types of GAC systems have been evaluated: Central GAC and Point

of Entry GAC (POE GAC). Central GAC refers to a design configuration in

which the activated carbon treatment takes place at a central treatment

facility, prior to entry into the distribution system. GAC may be

combined with other treatments and may be used to remove contaminants

other than radon in large, centralized facilities. In this analysis,

costs are estimated for central GAC systems with removal rates of 50,

80, and 99 percent. POE GAC generally refers to small- to medium-sized

carbon filtration units placed in the water distribution system just

before use occurs (e.g., before water enters a residence from the

distribution system.) System maintenance involves periodic replacement

of the filter units. As noted previously, POE GAC may be the most cost-

effective treatment for very small systems serving few households.

Costs are estimated for POE GAC with removal rates of 99%.

5.1.3  Storage

    Another technology that may be practical when only a relatively

slight reductions in radon levels are needed is the storage of water

for a period of time necessary for radioactive decay and volatilization

to reduce radon to acceptable levels. Depending on the configuration of

the vessel, storage for 24 to 48 hours may be sufficient to reduce

radon levels by 50 percent or more. The mode of removal is a

combination of radon decay and transfer of the radon from the water to

the storage tank headspace, which is refreshed through ventilation (US

EPA, 1998D). It has been assumed that a proportion of the smallest CWSs

(serving 500 people or fewer) with relatively low influent radon levels

and sufficient storage capacity may choose storage as the preferred

radon treatment technology. In estimating costs for the storage option,

it is assumed that the entire capital and O&M costs of the storage

system is attributable to the need to reduce radon levels. In fact, the

majority of CWSs choosing storage are likely to already have at least

some storage capacity available (ten percent of small systems have

atmospheric storage in place (US EPA 1997A)). These systems may be able

to add ventilation and/or other mechanisms to increase air/water

contact with a small capital investment, which supports the conclusion

that the present assumption of no storage in place is a conservative

assumption.

5.1.4  Regionalization

    The last technology whose costs are included in the HRRCA is

regionalization. In this analysis, regionalization is defined as the

construction of new mains to the nearest system with water below the

required radon level. This cost is estimated to be $280,000 per system

(1997$). The cost of actually purchasing water is not included in

regionalization costs, for several reasons. In the first case,

regionalization may involve the actual consolidation of water systems,

and thus there may be no charge to the system which is

``regionalized''. In addition, the system which supplies the water to

the regionalized system will still incur the same (or nearly the same)

costs for radon treatment as before regionalization and could be

expected to pass them on to the regionalized system. This assumes that

the water production cost ($/kgal) for the CWS before it regionalizes

is equal to the unit price ($/kgal) it will pay to the water system

from which it purchases water. In reality, this will over-estimate

costs in some cases and under-estimate in others. Including a water

purchase price in the cost estimate for regionalization without

correcting it for the removal of water production costs would lead to

an over-estimate in the costs of regionalization.

5.1.5  Radon Removal Efficiencies

    The amount of radon that the various technologies can remove from

water varies according to their specific design and operating

characteristics. At the most costly extreme, both aeration and GAC

technologies can remove 99 percent or more of the radon in water. Less

costly alternative designs remove less radon. In this analysis, one or

more cost estimates have been developed for the technologies discussed

above, corresponding to one or more radon removal levels. Approximate

cost ranges for achieving specified radon reduction efficiencies using

the various technologies are shown in Table 5-1. These costs are

estimated based on flow rates for a single installation, which may

treat water for an entire system or from a single source. For the

aeration and GAC technologies, costs have also been derived for

combined radon removal and post-treatment technologies, as discussed

below. The basis for the derivation of these cost estimates is

described in more detail in Section 5.4.

    The procedures used to decide what proportion of CWSs will adopt

the various radon removal technologies is described in more detail in

Section 5.5. In general, however, the large majority of the systems are

assumed to select the least-cost technology required to achieve a

target radon level. Other systems, for reasons of technical

feasibility, may need to choose more costly treatment technologies.

5.1.6  Pre-Treatment to Reduce Iron and Manganese Levels

    Pre-treatment technologies may also need to be part of radon

reduction systems. Aeration and GAC technologies can be fouled by high

concentrations of iron and manganese (Fe/Mn). EPA believes that Fe/Mn

concentrations greater than 0.3 mg/l would generally require

pretreatment to protect aeration/GAC systems from fouling. However,

since this level is near to the secondary MCL, it is believed that

essentially all systems with iron and manganese levels



[[Page 9580]]



above 0.3 are likely to already be treating to remove or sequester

these metals. Therefore, costs of adding Fe/Mn treatment to radon

removal systems are not included in the HRRCA. Preliminary EPA

estimates suggest that inclusion of Fe/Mn treatment costs will not

significantly effect overall cost estimates for radon removal. More

detailed analysis will be presented when the proposed NPDWR is

published.



BILLING CODE 6560-50-P



[[Page 9581]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.000







BILLING CODE 6560-50-C



[[Page 9582]]





5.1.7  Post-Treatment--Disinfection

    In addition to pre-treatment requirements, the installation of some

radon reduction technology may also require post-treatment, primarily

to reduce microbial contamination. Both aeration and GAC treatment may

introduce potentially infectious particulate contamination, which must

be addressed before the water can enter the distribution system. The

treatment of water for other contaminants may also introduce microbial

contamination. This is one reason why the majority of systems already

use disinfection technologies. As will be discussed in more detail

below, a substantial proportion of ground water systems (ranging from

50 percent in the smallest size category, to about 68 percent of the

largest systems) already disinfect. Costs of disinfection are only

attributed to the radon rule only for that proportion of systems not

already having disinfection systems in place. For systems that do not

already disinfect, chlorination is assumed to be the treatment of

choice. Alternative technologies are available, for example UV

disinfection, but chlorination is widely used in all size classes of

water supply systems, and the chlorination is considered to provide a

reasonable basis for estimating disinfection costs.



5.2  Monitoring Costs



    While not strictly speaking a water treatment technology, ground

water monitoring will play an important role in any strategy to reduce

radon exposures. Therefore, monitoring costs have been included as a

cost element in the cost analysis. Although EPA has not yet defined a

monitoring strategy for the proposed NPDWR, it is clear that systems

will, first, have to sample influent water to determine the need for

treatment, and second, continue to monitor after treatment (or after a

decision is made not to mitigate). For the purpose of developing

national cost estimates, it has been assumed that all systems will have

to conduct initial quarterly monitoring of all sources, and continue to

conduct radon monitoring and analysis indefinitely after the rule is

implemented. This is a conservative assumption (likely to overstate

monitoring costs) because in reality a large proportion of systems with

radon levels below the MCL will probably be allowed to monitor less

frequently after the initial monitoring period.

    Monitoring costs are simply the unit costs of radon analyses times

the number of samples analyzed. The number of intake sites per system

is estimated from SDWIS data, as discussed in Section 5.7. The cost of

analyzing each sample is estimated to be between $40 and $75, with an

representative cost of $50 per sample used for the national cost

estimate (US EPA 1998K).



5.3  Water Treatment Technologies Currently In Use



    EPA has conducted an extensive analysis of water treatment

technologies currently in use by ground water supply systems (Table 5-

2). This table shows the proportions of ground water systems with

specific technologies already in place broken down by system size

(population served). Many ground water systems currently employ

disinfection, aeration, or Fe/Mn removal technologies. This

distribution of pre-existing technologies serves as the baseline

against which water treatment costs are measured. For example, costs of

disinfection are attributed to the radon rule only for the estimated

proportion of systems that would have to install disinfection as a

post-treatment because they do not already disinfect.

    Within current EPA cost models, the estimate of the number of sites

(entry points into the distribution system) is ideally broken down into

three parts: estimates of the average national occurrence of the

contaminant in drinking water systems, the intra-system variability of

the contaminant concentration, and the typical number of sites within

system size categories. In prior RIAs, EPA modeled all drinking water

systems requiring treatment as installing centralized treatment, which

assumes that there is one point of treatment within a system. A more

accurate estimate of treatment would be to calculate costs according to

treatment installed at each well site that is predicted to be above the

target radon level within a water system. This intra-system variability

analysis accounts for the fact that, in reality, multi-site water

systems do not necessarily have the same radon level at each site.

However, because the analysis of intra-system variability for radon

occurrence is not yet complete, it is not possible to use this approach

to calculate treatment costs. For future rules, including the proposed

rule for radon, EPA will calculate national cost estimates based on the

number of sites rather than by the system as a whole. These estimates

will more accurately reflect the percentage of the population receiving

drinking water that has been treated in some way and will result in

more accurate national compliance cost estimates.

    The cost analysis assumes that any system affected by the rule will

continue to employ pre-existing radon treatment technology and pre-and

post-treatments in their efforts to comply with the rule. Where pre-or

post-treatments are already in place, but radon treatment is currently

not taking place, it is assumed that compliance with the radon rule

will not require any upgrade or change in the pre-or post-treatments.

Therefore, no incremental cost is attributed to pre-or post-treatment

technologies. This may underestimate costs if pre-or post-treatments

need to be changed (e.g., a need for additional chlorination after the

installation of packed tower aeration). The potential magnitude of this

cost underestimation is not known, but is likely to be a very small

fraction of total treatment costs.



  Table 5-2.--Estimated Proportions of Ground Water Systems With Water Treatment Technologies Already in Place

                                                  (Percent) \1\

----------------------------------------------------------------------------------------------------------------

                                                          System size (population served)

 Water treatment technologies in -------------------------------------------------------------------------------

              place                25-100    101-500   501-1K    1K-3.3K  3.3K-10K   10K-50K  50K-100K   100K-1M

----------------------------------------------------------------------------------------------------------------

Fe/Mn Removal & Aeration &             0.4       0.2       1.2       0.6       2.9       2.2       3.1       2.0

 Disinfection...................

Fe/Mn Removal & Aeration........       0.0       0.1       0.2       0.1       0.4       0.1       0.4       0.1

Fe/Mn Removal & Disinfection....       2.1       5.1       8.3       3.0       7.8       7.4       9.7       6.8

Fe/Mn Removal...................       1.9       1.5       1.5       1.0       1.1       0.4       1.1       0.2

Aeration & Disinfection Only....       0.9       3.2       9.8      13.7      20.9      19.7      18.6      19.9

Aeration Only...................       0.8       1.0       1.8       2.9       2.9       1.0       2.1       0.6

Disinfection Only...............      49.6      68.2      65.0      65.0      56.3      66.0      58.3      68.3



[[Page 9583]]





None............................      44.3      20.7      12.2      13.7       7.7       3.2       6.7       2.1

----------------------------------------------------------------------------------------------------------------

\1\ Source: EPA analysis of data from the Community Water System Survey (CWSS), 1997, and Safe Drinking Water

  Information System (SDWIS), 1998.



5.4  Cost of Technologies as a Function of Flow Rates and Radon Removal

Efficiency



    EPA has developed a set of cost curves that describe the

relationships between the capital and operating and maintenance costs

of the various treatment technologies, flow rates, and the degree of

radon removal that is required (US EPA 1998A, 1998O). Cost curves were

developed using the most recent available data and standard cost

estimation methodologies. Separate functions for capital and operation

and maintenance (O&M) costs have been developed for each technology and

radon removal rate. For all of the technologies except regionalization,

both the capital and O&M cost curves are functions of flow rates.

Capital costs are estimated as a function of the design flow (DF) of

the technology. The DF for a technology is equal to a technology's

maximum flow capacity, or the largest amount of water that can be

processed per unit time. The DF is typically two to three times greater

than the average amount of water treated by a given system. O&M costs

are functions of the average flow (AF) through the system. Labor,

treatment chemicals and materials, periodic structure maintenance, and

water stewardship expenses are estimated based on daily average flows.

The cost curves developed by OGWDW for the various radon removal

technologies are provided in Appendix B.



5.5  Choice of Treatment Responses



    The Agency has developed a set of assumptions regarding the choices

that CWSs will make in deciding how to mitigate water radon levels to

meet specific exposure reduction requirements. These assumptions have

been developed taking into account the expected influent radon levels,

the degree of radon removal needed to reach specified levels, the types

of technologies that would be technically feasible and cost-effective

for systems of a given size, and the distribution of pre-existing

technologies shown in Table 5-2. Generally, it is assumed that a system

will choose the least-cost alternative technology to achieve a given

radon level. For example, to achieve a radon level of 100 pCi/l, all

systems with average influent levels below 100 would not need to

mitigate, systems with influent radon levels between 100 and 200 pCi/l

would need to employ technologies that achieve 50 percent reduction,

systems with influent levels between 200 and 500 pCi/l would employ

technologies capable of 80 percent radon removal, and systems with

influent radon above 500 pCi would employ technologies with removal

efficiencies of 99 percent. In actuality, removal efficiencies would be

more variable; e.g., a removal efficiency of 90 percent, rather than 99

percent, could be employed for radon levels between 500 and 1,000 pCi/

l. However, this cost analysis has been limited to three removal

efficiencies to simplify the analysis. EPA does not believe that this

has introduced any significant bias into the assessment.

    Table 5-3 presents the estimated proportions of systems of given

sizes that are expected to choose specified radon reduction

technologies for given degrees of radon removal. Most systems in most

size classes are assumed to choose aeration as the preferred radon

reduction technology with or without disinfection, depending on the

proportion of systems in that size stratum already disinfecting. This

is because some form of aeration is generally the most cost-effective

option for a given degree of radon reduction. For small systems and low

required removal efficiencies, multistage fixed-bed (MSBA) and diffused

bubble aeration (DA) tend to be the most cost-effective. For large

systems and high removal efficiencies, packed tower aeration (PTA) is

the only feasible aeration technology.

    Small proportions of the smallest system size categories (less than

5 percent in all cases) are assumed to choose central GAC with or

without disinfection. A few percent of the smallest systems are also

assumed to choose POE GAC. Storage is assumed to be a viable option for

two percent of small systems where radon reduction of 50 percent or

less is required, and regionalization is assumed to be feasible for one

percent of the smallest systems. EPA has assumed in this HRRCA that no

systems would choose spray aeration or alternative source technologies.

It is believed that these technologies would be chosen only rarely, and

their omission has not biased the compliance cost estimates. This issue

will be addressed in more detail in the proposed NPDWR.



                           Table 5-3.--Decision Matrix for Selection of Treatment Technology Options: Up to 50 Percent Removal

--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                                 Percent of system size category (population served) choosing treatment technology

               Treatment technology option               -----------------------------------------------------------------------------------------------

                                                             <100       101-500    501-1000    1001-3.3K   3301-10K     10-50K      50-100K    100-1000K

--------------------------------------------------------------------------------------------------------------------------------------------------------

PTA (80)................................................         2.6         7.8        16.8        31.9        60.8        86.9        86.3        96.4

PTA (80) + disinfection.................................         2.4         2.2         3.2         8.1         9.2         3.2        13.7         3.6

MSBA/STA (80)...........................................        13.2        21.8        22.7        15.9         8.7         0.0         0.0         0.0

MSBA/STA (80) + disinfection............................        11.8         6.2         4.3         4.1         1.3         0.0         0.0         0.0

DA (80).................................................        31.7        43.4        42.7        31.9        17.4         9.7         0.0         0.0

DA (80) + disinfection..................................        28.3        12.6         8.3         8.1         2.6         0.4         0.0         0.0

Retrofit Spray..........................................         0.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

GAC (50)................................................         2.6         2.3         0.8         0.0         0.0         0.0         0.0         0.0



[[Page 9584]]





GAC (50) + disinfection.................................         2.4         0.7         0.2         0.0         0.0         0.0         0.0         0.0

POE GAC (99)............................................         2.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

Storage (50)............................................         2.0         2.0         1.0         0.0         0.0         0.0         0.0         0.0

Regionalization (99)....................................         1.0         1.0         0.0         0.0         0.0         0.0         0.0         0.0

Alternate source (99)...................................         0.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

All Systems.............................................       100         100         100         100         100         100         100         100

PTA (80)................................................         4.2        10.9        20.2        31.9        60.8        96.5        86.3        96.4

PTA (80) + disinfection.................................         3.8         3.1         3.8         8.1         9.2         3.5        13.7         3.6

MSBA/STA (80)...........................................        14.8        21.0        21.0        15.9         8.7         0.0         0.0         0.0

MSBA/STA (80) + disinfection............................        13.2         6.0         4.0         4.1         1.3         0.0         0.0         0.0

DA (80).................................................        29.6        42.8        42.0        31.9        17.4         0.0         0.0         0.0

DA (80) + disinfection..................................        26.4        12.2         8.0         8.1         2.6         0.0         0.0         0.0

Retrofit Spray..........................................         0.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

GAC (80)................................................         2.6         2.3         0.8         0.0         0.0         0.0         0.0         0.0

GAC (80) + disinfection.................................         2.4         0.7         0.2         0.0         0.0         0.0         0.0         0.0

POE GAC (99)............................................         2.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

Regionalization (99)....................................         1.0         1.0         0.0         0.0         0.0         0.0         0.0         0.0

Alternate source (99)...................................         0.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

All Systems.............................................       100         100         100         100         100         100         100         100

PTA (99)................................................        15.3        26.5        35.3        47.8        69.4        96.5        86.3        96.4

PTA (99) + disinfection.................................        13.7         7.5         6.7        12.2        10.6         3.5        13.7         3.6

MSBA/STA (99)...........................................        34.3        49.1        48.7        31.9        17.4         0.0         0.0         0.0

MSBA/STA (99) + disinfection............................        30.7        13.9         9.3         8.1         2.6         0.0         0.0         0.0

GAC (99)................................................         1.6         1.6         0.0         0.0         0.0         0.0         0.0         0.0

GAC (99) + disinfection.................................         1.4         0.4         0.0         0.0         0.0         0.0         0.0         0.0

POE GAC (99)............................................         2.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

Regionalization (99)....................................         1.0         1.0         0.0         0.0         0.0         0.0         0.0         0.0

Alternate source (99)...................................         0.0         0.0         0.0         0.0         0.0         0.0         0.0         0.0

      Totals............................................       100         100         100         100         100         100         100         100

--------------------------------------------------------------------------------------------------------------------------------------------------------

Notes:

1. Technology abbreviations: PTA = packed tower aeration, MSBA/STA = multi-stage bubble aeration, GAC = granular activated carbon, POE GAC = point of

  entry granular activated carbon. Numbers in parentheses indicate removal efficiencies.

2. Capital costs for small systems include land costs. For large systems, it is assumed that additional land is not required.

3. Sequestration costs are included in PTA and MSBA/STA capital costs.

4. Additional housing costs are included in PTA, MSBA/STA, and GAC capital costs and are weighted under the assumption that 50% of small systems will

  require additional housing, 100% of large systems will require additional housing.

5. Permitting costs are included and are assumed to be 3% of capital costs, with a minimum of $2500.

6. Pump and blower redundancies are included in capital costs.



5.6  Cost Estimation

5.6.1  Site and System Costs

    The costs of reducing radon in ground water to specific radon

levels was calculated using the cost curves discussed in Section 5.4

and the matrix of treatment options presented in Section 5.5. For each

radon level and system size stratum, the number of systems required to

reduce radon levels by up to 50 percent, 80 percent and 99 percent were

calculated. Then, the cost curves for the distributions of technologies

dictated by the treatment matrix were applied to the appropriate

proportions of the systems. Capital and O&M costs were then calculated

for each system, based on typical estimated design and average flow

rates. These flow rates were calculated on spreadsheets using equations

from EPA's Safe Drinking Water Suite Model (US EPA 1998N). The

equations and parameter values relating system size to flow rates are

presented in Appendix C.

    The distributions of influent radon levels in the various system

size categories were calculated using the results of EPA's updated

radon occurrence analysis (exceedance proportions calculated from data

in US EPA 1998L).

    Capital and O&M costs were estimated separately for each ``site''

(a separate water source, usually a well) within systems. Where systems

obtained water from only one site, costs are calculated by applying the

entire system flow rate to the appropriate cost curves. Where systems

consisted of more than one site, the total system flow rate was divided

by the number of sites, capital and O&M costs were then calculated for

the resulting flow rate, and the total system cost was obtained by

multiplying this result by the number of sites in the system. This

approach provides conservative cost estimates, because it assumes that

separate treatment systems would be built at each site. This approach

also obscures some of the effects of variability in system sizes on

costs, because each system in a given size category is assumed to have

the same flow rate.



    Table 5-4 summarizes the numbers of sites per system for the

various size categories of combined public and private community ground

water systems. The average ranges from 1.1 site per system serving less

than 100 people to almost nine sites per system serving greater than

100,000 people. The distributions of the numbers of sites per systems

are very skewed, with ninetieth-percentile values ranging from 2 to 20

sites per system for the smallest and largest size categories,

respectively. A large proportion of the systems serving 10,000 people

or less obtain water from only one site. Public and private water

systems differ with regard to system design and average flows. For



[[Page 9585]]



this reason, separate cost estimates have been developed for the public

and private community ground water systems.



   Table 5-4.--Numbers of Sites per Ground Water System by System Size

------------------------------------------------------------------------

                                                                 90th

                                                  Average     percentile

        System size (population served)          sites per    sites per

                                                   system       system

------------------------------------------------------------------------

25-100........................................          1.1            2

101-500.......................................          1.2            2

501-1,000.....................................          1.4            3

1,001-3,300...................................          1.7            4

3,301-10,000..................................          2.3            4

10,001-50,000.................................          3.9           10

50,000-100,000................................          8.7           20


>100,000......................................          8.8          20

------------------------------------------------------------------------

Source: EPA analysis of CWSS data, 1998.



    In addition to the costs of radon treatment and disinfection,

monitoring costs were also calculated for each system. As noted

previously, the average cost of monitoring was estimated to be $50 per

sample, and it was assumed that each site in a system would need to be

monitored quarterly. Monitoring costs were added as an ongoing cost

stream to the O&M costs.

5.6.2  Aggregated National Costs

    The estimated costs of reducing radon levels to meet different

radon levels were estimated by summing the costs for the individual

sites and systems in each size category and influent range. Separate

totals were compiled for capital and O&M costs. Capital costs were

annualized (over 20 years at a seven per cent discount rate) and added

to the annual O&M costs to provide single aggregate estimates of

national costs for each radon level. This approach implicitly assumes

that treatment devices have useful lives that are identical to the

period of financing. In reality, the useful life and period of

financing are not necessarily the same. The aggregate cost estimates

are presented in Section 6. As will be discussed in more detail below,

separate cost estimates were developed for implementation options

involving MMM programs and are presented in Section 7. Summary outputs

of the spreadsheet models used to estimate costs are provided in

Appendix D.

5.6.3  Costs to Community Water Supply Systems

    As noted above, costs were estimated separately for public and

private ground water systems. Costs per system were calculated by

dividing total costs for a given size category of public or private

system by the total number of systems needing to mitigate radon. The

results of these assessments are presented in Section 6.

5.6.4  Costs to Consumers/Households

    Costs to households have also been calculated for public and

private ground water systems. Costs are calculated by multiplying the

average annual treatment costs per thousand gallons by the estimated

average household consumption (83,000 gal/year). This approach assumes

that all water systems pass incremental costs attributable to the radon

rule on to system's residential customers and that the residential

customers will pay the same proportion of costs as other users. Average

household costs are calculated separately for public and private

community water systems across various system-size categories. Per

household costs are then compared to median household income data (US

EPA 1998H) for the same system-size categories. These impacts are

discussed in Section 6.

5.6.5  Costs of Radon Treatment by Non-Transient Non-Community Systems

    Very little data are available that will support the development of

detailed estimates of radon treatment costs for the NTNCWS that could

be affected by a radon NPDWR. EPA is currently conducting a more

detailed evaluation of the characteristics of NTNCWSs that will be

completed in time for the proposed rule.



5.7  Application of Radon Related Costs to Other Rules



    The baseline for the radon rule compliance cost estimates presented

in this draft HRRCA consists of the pre-existing treatment technology

distribution shown in Table 5-2. As the radon rule is implemented,

however, other rules may also require additional systems to install new

technologies (e.g., disinfection). Thus, attributing all costs of

increased use of disinfection at systems with high radon levels to the

radon rule would overstate its cost. At the present time, EPA has not

quantified the potential degree to which the costs of the radon rule

may be overstated.



6. Results: Costs and Benefits of Reducing Radon in Drinking Water



    This section presents benefit, cost, and impact estimates for the

various radon levels. Section 6.1 provides an overview of the

analytical approach. Sections 6.2 and 6.3 present the monetized benefit

and cost estimates for the various radon levels evaluated. Section 6.3

summarizes the economic impacts on the various affected entities.

Section 6.5 compares the costs and benefits of the radon levels

evaluated. Section 6.6 presents a brief summary of the major

uncertainties in the cost, benefit, and impact estimates.

    The presentation of costs and benefits in this Section is based on

analysis of radon levels of 100, 300, 500, 700, 1,000, 2,000, and 4,000

pCi/l in CWSs served by ground water.



6.1  Overview of Analytical Approach



    The analysis of benefits quantifies the reduction in health risks/

impacts to the general population and considers the risks to

potentially sensitive subpopulations (qualitatively). The evaluated

health benefits of the rule consist of reduced fatal and non-fatal

cancer risks, and the monetary surrogates for these benefits have been

estimated, as described in Section 4.0. The national cost estimates

developed include the capital and O&M costs to reduce radon, along with

pre- and post-treatment costs where appropriate, as well as monitoring

costs. Record keeping and reporting costs and implementation costs to

States and government entities will be addressed in the RIA prepared

for the proposed rule.

    The costs and benefits of a radon NPDWR will result in economic

impacts on affected individuals, corporate entities, and government

entities. In this analysis, the impacts on water systems and households

have been evaluated. These include: (1) the cost to systems of

different sizes and ownership types, and (2) changes in water costs to

households as a proportion of income. Public systems include those

owned by government entities. Private systems consist of investor-owned

entities that provide drinking water as their primary line of business.

Ancillary systems include drinking water systems that are operated

incidentally to another business. The vast majority of ancillary

systems are mobile home parks, but some are schools, hospitals, and

other entities. The economic impacts of the MMM programs on systems or

households have not been calculated, because there is no information at

present as to how these programs would be funded or upon whom the costs

would fall.



6.2  Health Risk Reduction and Monetized Health Benefits



    The probabilistic risk model was used to calculate the cancer risk

reduction benefits of the various levels. Risk reduction benefits were

calculated by subtracting the estimated population risk (number of

fatal cancers per year at a particular radon level) from the



[[Page 9586]]



baseline (pre-regulation) population cancer risk due to radon exposure.

Estimates of the number of non-fatal cancers avoided were developed as

described in Section 4.2.1. The results of this analysis are summarized

in Table 6-1. Under the baseline scenario, the estimated number of

fatal cancers per year caused by radon exposures in domestic water

supplies is 160, and the number of non-fatal cancers is 9.2. As radon

levels decrease, residual risks decrease, and the risk reduction

benefits increase. Since very few people are exposed at levels above

2,000 pCi/l, the benefit of controls in this range is relatively small

(fewer than 7 cancers prevented per year). The health risk reduction

benefits then increase rapidly as radon levels decrease because

progressively larger populations are affected as more and more systems

are required to mitigate exposures.



            Table 6-1.--Residual Cancer Risk and Risk Reduction From Reducing Radon in Drinking Water

----------------------------------------------------------------------------------------------------------------

                                                Residual fatal   Residual non-   Risk reduction   Risk reduction

                                                 cancer risk     fatal cancer    (fatal cancers     (non-fatal

         Radon level (pCi/l in water)             (cases per      risk (cases     avoided per    cancers avoided

                                                    year)          per year)       year) \1\      per year) \1\

----------------------------------------------------------------------------------------------------------------

(Baseline)...................................            160               9.2              0                0

4,000 \2\....................................            158               9.1              2.2              0.1

2,000........................................            153               8.8              6.5              0.4

1,000........................................            143               8.2             16                0.9

700..........................................            135               7.8             25                1.4

500..........................................            124               7.1             36                2.1

300..........................................            101               5.8             58                3.4

100..........................................             44.8             2.6            115                6.6

----------------------------------------------------------------------------------------------------------------

\1\ Risk reductions and residual risk estimates are slightly inconsistent due to rounding.

\2\ 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA provisions of Section 1412(b)(13).



    At the lowest level (100 pCi/l) analyzed, the residual cancer risk

(the cancer risk occurring after controls are installed) is

approximately 45 fatal cancers per year. The risk reduction from this

radon level is 115 fatalities per year, a reduction of approximately 72

percent from the baseline of 160 per year. A similar proportional

reduction in non-fatal cancers is seen with decreasing radon levels.

    The monetary valuation methods discussed in Section 4 were applied

to these risk reductions, as shown in Table 6-2. The central tendency

benefits estimates are based on a VSL of $5.8 million (1997$) and a WTP

to avoid fatal cancers of $536,00 (1997$). The ranges of benefits

estimated using the upper and lower bound estimates of the VSL and WTP

to avoid non-fatal cancers are also provided in the table.



 Table 6-2.--Estimated Monetized Health Benefits From Reducing Radon in

                             Drinking Water

------------------------------------------------------------------------

                                                Monetized

                                                 health       Range of

                                                benefits,     monetized

                                                 central       health

             Radon Level (pCi/l)                tendency      benefits

                                              (annualized,  (annualized,

                                               $millions,    $millions,

                                                1997) \1\     1997) \2\

------------------------------------------------------------------------

4,000 \3\...................................            13          2-35

2,000.......................................            38         5-106

1,000.......................................            96        12-268

700.........................................           145        18-403

500.........................................           212        26-591

300.........................................           343        43-955

100.........................................           673      84-1875

------------------------------------------------------------------------

\1\ Includes contributions from fatal and non-fatal cancers, estimated

  using central tendency estimates of the VSL of $5.8 million (1997$),

  and a WTP to avoid non-fatal cancers of $536,000 (1997$).

\2\ Estimates the range of VSL between $0.7 and $16.3 million (1997$),

  and a range of WTP to avoid non-fatal cancers between $169,000 (1997$)

  and $1.05 million (1997$).

\3\ 4,000 pCi/l is equivalent to the AMCL estimated by the NAS based on

  SDWA provisions of Section 1412(b)(13).



    Using central tendency estimates for each of the monetary

equivalents, the baseline health costs of fatal and non-fatal cancers

associated with household radon exposures from CWSs are estimated to be

$933 million per year. Central tendency estimates of monetized benefits

range from $13 million per year for a level of 4,000 pCi/l up to $673

million for the most stringent level of 100 pCi/l. When different

values for the VSL are used, the benefits estimates change

significantly. Using a lower bound VSL of $0.7 million, the benefits

estimates are reduced approximately 9-fold compared to the central

tendency estimates. Using an upper bound VSL of 16.3 million increases

the benefits estimates by approximately 3-fold relative to the central

tendency estimate. Variations in the estimated WTP to avoid non-fatal

cancers affect benefit total estimates only slightly (i.e., less than 1

percent), since non-fatal cancers represent a very small proportion of

estimated radon cancer cases.

    A more detailed breakout of the risk reduction, monetized benefits

estimates, and the total cost per fatal cancer case avoided for ever-

smokers and never-smokers is provided in Tables 6-3 and 6-4.



                  Table 6-3.--Risk Reduction and Monetized Benefits Estimates for Ever-Smokers<SUP>1</SUP>

----------------------------------------------------------------------------------------------------------------

                                                                     Radon level, pCi/l

                                           ---------------------------------------------------------------------

                                              4000<SUP>3</SUP>     2000      1000       700       500       300       100

----------------------------------------------------------------------------------------------------------------

Fatal Cancers Avoided Per Year............       1.7       5.2      13.2      19.9      29.2      47.1      92.5

Non-Fatal Cancers Avoided Per Year........       0.1       0.3       0.8       1.1       1.7       2.7       5.2

Annual Monetized Health Benefits                10.2      30.6      77.1     115.8     170.0     274.7     539.3

 ($Millions, 1997)--Central Tendency......



[[Page 9587]]





Annual Incremental Health Benefits              10.2      20.4      46.5      38.7      54.2     104.7     264.6

 ($Millions/year)--Central Tendency.......

Annual Cost Per Fatal Cancer Avoided             7.0       4.4       3.7       3.7       3.7       4.0      4.3

 ($Millions, 1997) <SUP>2</SUP>......................

----------------------------------------------------------------------------------------------------------------

\1\ Risk reductions for ever- and never-smokers were estimated using the NAS unit risk estimates summarized in

  Table 3-4, an ever-smoking prevalence of 58% males and 42% females, a central VSL estimate of $5.8 million

  (1997$), and central WTP estimate to avoid non-fatal cancer of $536,000 (1997$).

\2\ Total cost estimates come from Table 6-5. The cost per fatal cancer case avoided is calculated by dividing

  the estimates of fatal cancers avoided per year by the annualized mitigation costs for each population. For

  purposes of this analysis, it was assumed that the mitigation costs (for both water and MMM programs) would be

  allocated equally to smoking and non-smoking populations.

\3\ 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on the SDWA provisions of Section

  1412(b)(13).







                  Table 6-4.--Risk Reduction and Monetized Benefits Estimates for Never-Smokers

----------------------------------------------------------------------------------------------------------------

                                                                  Radon Level, pCi/l

                                    ----------------------------------------------------------------------------

                                       4000 *      2000       1000       700        500        300        100

----------------------------------------------------------------------------------------------------------------

Fatal Cancers Avoided Per Year.....       0.4        1.3        3.2        4.8        7.0       11.4       22.3

Non-Fatal Cancers Avoided Per Year.       0.03       0.09       0.22       0.33       0.48       0.78       1.54

Annual Monetized Health Benefits          2.4        7.4       18.6       27.9       41.0       66.3      130.2

 ($Millions, 1997)--Central

 Tendency..........................

Annual Incremental Health Benefits        2.4        5         11.2        9.3       13.1       25.3       63.9

 ($Millions/year)--Central Tendency

Annual Cost Per Fatal Cancer             29.2       18.3       15.3       15.4       15.5       16.4       17.8

 Avoided ($Millions, 1997).........

----------------------------------------------------------------------------------------------------------------

*4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



6.3  Costs of Radon Mitigation



    This section describes the incremental costs associated with each

of the radon levels. Discussion of the cost results includes: the total

nationally aggregated cost to all water systems that must comply with

the target radon levels. These include capital and O&M costs; the

average annualized cost per system exceeding the applicable radon

level; the average annualized costs per system and incremental costs

per household, broken out by public and private water system; and costs

and impacts to households under each radon level. All costs are

incremental costs stated in 1997 dollars. Capital costs were annualized

using a seven percent discount rate and a 20-year amortization period.

6.3.1  Aggregate Costs of Water Treatment

    The total annual nationally aggregated cost varies significantly by

the specific radon level. Total national cost estimates for CWSs are

presented in Table 6-5. As demonstrated by the exhibit, water

mitigation costs increase substantially from the highest radon level

analyzed ($24 million at 4000 pCi/l) to the lowest level analyzed ($795

million at 100 pCi/l).



                   Table 6-5.--Estimated Annualized National Costs of Reducing Radon Exposures

                                                [$Million, 1997]

----------------------------------------------------------------------------------------------------------------

                                                                      Central

                                                                     tendency        Range of     Cost per fatal

                       Radon level (pCi/l)                          estimate of     annualized      cancer case

                                                                    annualized    costs (+/-50%)      avoided

                                                                       costs

----------------------------------------------------------------------------------------------------------------

4000*...........................................................              24      12-36                 11.3

2000............................................................              46      23-70                  7.1

1000............................................................              98     49-146                  5.9

700.............................................................             148     75-223                  6.0

500.............................................................             218    109-327                  6.0

300.............................................................             373    187-560                  6.4

100.............................................................             795   398-1193                  6.9

----------------------------------------------------------------------------------------------------------------

*4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



    The costs borne by water systems are made up of annualized capital,

O&M, and monitoring costs. The contributions of these cost elements are

broken out in Table 6-6. As the radon level increases (i.e., is made

less stringent), the proportion of costs due to monitoring increases

relative to capital and O&M costs.



[[Page 9588]]







                     Table 6-6.--Capital and O&M Costs of Mitigating Radon in Drinking Water

                                                [$Million, 1997]

----------------------------------------------------------------------------------------------------------------

                                                                                      Annual

             Radon levels (pCi/l)                Annual capital  Annual O&M cost    monitoring      Total costs

                                                      cost                             costs

----------------------------------------------------------------------------------------------------------------

4000 *........................................              8.0              5.2            11.4              25

2000..........................................             19.8             15.3            11.4              46

1000..........................................             48.9             37.4            11.4              98

700...........................................             77.9             58.5            11.4             148

500...........................................            119               87.7            11.4             218

300...........................................            210              124              11.4             373

100...........................................            460.             324              11.4            795

----------------------------------------------------------------------------------------------------------------

* 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



6.4  Incremental Costs and Benefits of Radon Removal



    Table 6-7 summarizes the central tendency and the upper and lower

bound estimates of the incremental costs and benefits of radon exposure

reduction. Both the annual incremental costs and benefits increase as

the radon level is incrementally decreased from 2000 pCi/l down to 100

pCi/l. The exhibit also illustrates the wide ranges of potential

incremental costs and benefits due to the uncertainty inherent in the

estimates. Incremental costs and benefits are within 10 percent of each

other at radon levels of 1000, 700, and 500 pCi/l. There is substantial

overlap between the incremental costs and benefits at each radon level.



                         Table 6-7.--Estimates of the Annual Incremental Costs and Benefits of Reducing Radon in Drinking Water

                                                                    [$Millions, 1997]

--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                                                                   Radon Level, pCi/l

                                                              ------------------------------------------------------------------------------------------

                                                                  4000 *       2,000        1,000         700          500          300          100

--------------------------------------------------------------------------------------------------------------------------------------------------------

Annual Incremental Cost......................................           24           46           52           50           70          156          422

Range of Annual Incremental Costs............................        12-36        11-34        26-76        26-77       34-104       78-233      211-633

Annual Incremental Monetized Benefits........................           13           25           58           48           67          130          329

Range of Incremental Monetized Benefits......................         2-35         3-71        7-162        6-135        8-188       17-364       41-920

Incremental Cost Per Fatal Cancer Case Avoided...............         11.3          5.0          5.2          6.1          6.1          7.0          7.5

--------------------------------------------------------------------------------------------------------------------------------------------------------

* 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



6.5  Costs to Community Water Systems



    This section examines the regulatory costs that will be incurred by

individual CWSs at the various radon levels analyzed. Systems above the

target radon level will incur monitoring costs and treatment costs.

Systems below the target radon level will incur only monitoring costs.

    The number of CWSs exceeding the applicable radon level increases

considerably with each decrease in the radon level analyzed as shown

Table 6-8. The table also shows that the vast majority (90 percent or

more) of affected systems, regardless of radon level, are very, very

small (serving 25-500 people) or very small (serving 501-3,300 people).



               Table 6-8.--Number of Community Ground Water Systems Exceeding Various Radon Levels

----------------------------------------------------------------------------------------------------------------

                                     VVSVS

   Exposure level (pCi/l)    --------------------  VS (501-     S (3,301-       M (10,000-        L       Total

                              (25-100)    (101-     3,000)       10,000)         100,000)      (>100K)

------------------------------------------500)------------------------------------------------------------------

4000 \1\....................       364       759         60             5               1            0     1,190

2000........................       949      1448        205            19               8            0     2,630

1000........................      2149      2613        668            75              44            2     5,552

700.........................      3090      3459      1,153           151              94            5     7,951

500.........................      4201      4434      1,796           287             177            9    10,904

300.........................      6302      6233      3,059           657             387           19    16,657

100.........................    10,922    10,349      6,077         1,707             995           48    30,098

----------------------------------------------------------------------------------------------------------------

\1\ 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



Source: (USEPA 19989L).



    For CWSs that have radon in excess of a given level within each

size category, the average cost per system to reach the target level

varies little as the radon levels decrease. This is shown in Table 6-9,

which presents the average annualized cost per public and private CWS

by system size category. This pattern is due in large part to the

limited number of treatment options assumed to be available to systems

that may (in aggregate) be encountering a relatively wide range of

radon levels. In some cases (e.g., for very very small systems), the

average cost per system for a given



[[Page 9589]]



system size increases as the radon level decreases. In other cases, the

average cost per system remains virtually constant as the radon level

decreases. These inconsistent patterns are due to two competing

effects: (1) The average cost will tend to increase because some

systems must select a more costly treatment option; yet (2) the average

cost will also tend to decrease with the inclusion of previously

unaffected systems (those with lower radon levels) that are most likely

to use lower-cost treatments. The cases where average costs decrease

with decreasing radon levels are due to the latter effect.

    These results show that changing the radon level affects the number

of CWSs that must treat for radon, but generally does not significantly

alter the cost per system for those systems above the target level.

Moreover, while large systems bear the greatest burden in terms of cost

per system, there are relatively few large systems with radon levels

above the exposure scenarios analyzed. The cost per system for CWSs

with a radon concentration below a target radon level will be the same

because monitoring costs are dependent on system size and not on

concentration. Monitoring costs range from less than $250 for the very

very small systems to almost $2,000 for large systems, again due to the

larger number of sites requiring monitoring.



BILLING CODE 6560-50-P

[GRAPHIC] [TIFF OMITTED] TN26FE99.001



BILLING CODE 6560-50-C



6.6  Costs and Impacts to Households



    This section reports incremental household costs and impacts

associated with each radon level, assuming that costs incurred by

systems above the target radon levels are passed on to the systems'

customers (i.e., households). Costs per household reflects only

monitoring and treatment costs to CWSs above the target level. In

addition, households served by CWSs falling under the target radon

level also will incur monitoring costs, but no treatment costs. Costs

for these CWSs are relatively low, however, and are not evaluated at

the household level. As with per system costs, the results are

presented separately for public and for private CWSs. This is important

in considering impacts on households not only because the costs per

system are different for public versus private systems, but also

because the smallest private systems tend to serve fewer households

than do the smallest public systems. Therefore, the average household

served by a private system must bear a greater percentage of the CWS's

cost than does the average household served by a public CWS. This is

particularly important where capital costs make up a large portion of

total radon mitigation costs.

    The annual cost per household is presented in Table 6-10 for

households served by public and private CWSs. As expected, costs per

household increase as system size decreases. Costs per household is

higher for households served by smaller systems than larger systems for

two reasons. First, smaller systems serve far fewer households than

larger systems and, consequently, each household must bear a greater

percentage share of the CWS's costs. Second, smaller systems tend to

have higher influent radon concentrations that, on a per-capita or per-

household basis, require more expensive treatment methods (e.g., one

that has an 85 percent removal efficiency rather than



[[Page 9590]]



50 percent) to achieve the target radon level.

    Another significant finding regarding annual cost per household is

that, like the per-system costs, household costs (which are a function

of per system costs) are relatively constant across different radon

levels within each system size category. For example, there is less

than $1 dollar per year variation in cost per household, regardless of

the radon level being considered for households served by large public

or private systems (between $6 and $7 per year), by medium public or

private systems (between $10 and $11 per year, and by small public or

private systems (between $19 and $20 per year). Similarly, for very

small systems, the costs per household is consistently about $34 per

year for public systems and consistently about $40 per year for private

systems, varying little across radon level. Only for very very small

systems is there a modest variation in household costs. The range for

per household costs for public systems serving 25-500 people is $87 per

year (at 4000 pCi/l) to $135 per year (at 100 pCi/l). The corresponding

range for private systems is $139 to $238 per year. For households

served by the smallest public system (25-100 people), the range of cost

per household ranges from $292 per year at 4000 pCi/l to $398 per year

at 100 pCi/l. For private systems, the range is $364 to $489 per year,

respectively. Costs per household for very very small systems differ

more than do household costs for other system size categories because

very very small systems serve only between 25 and 500 people and,

consequently, serve fewer households. Therefore, even though per system

costs show little difference for any system size category, all system

size categories (other than for very very small systems) spread the

small difference out among many more households such that the

difference is indistinguishable.



BILLING CODE 6560-50-P

[GRAPHIC] [TIFF OMITTED] TN26FE99.002



BILLING CODE 6560-50-C



    To further evaluate the impacts of these household costs on the

households that must bear them, the costs per household were compared

to median household income data for households in each system-size

category. The result of this calculation indicates a household's likely

share of incremental costs in terms of its household income. The

analysis considers only households served by CWSs with influent radon

levels that are above the target radon level. Households served by CWSs

with lower radon levels may incur incremental costs due to new

monitoring requirements, but these costs are not significant at the

household level.

    Results are presented in Table 6-11 for public and private CWSs,

respectively. For all system sizes but one (very very small private

systems), household costs as a percentage of median household income

are less than one percent. Impacts exceed one percent only for

households served by very very small private systems, which are

expected to face impacts of just under 1.1 percent. Similar to the cost

per household results on which they are based, household impacts

exhibit little variability across radon levels.



BILLING CODE 6560-50-P



[[Page 9591]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.003





BILLING CODE 6560-50-C



6.7  Summary of Costs and Benefits



    Table 6-12 summarizes the central tendency estimates of annual

monetized benefits and annualized costs of the various regulatory

alternatives. The central tendency national cost estimates are greater

than the monetized benefits estimates for all radon levels evaluated,

although they are within 10 percent at levels of 1000, 700, 500, and

300 pCi/l. Mitigation costs increase more rapidly than the monetized

benefits as radon levels decrease. However, it is important to

recognize that due to the uncertainty in the costs and benefits

estimates, there is a very broad possible range of potential costs and

benefits that overlap across all of the radon levels evaluated.



Table 6-12.--Estimated National Annual Costs and Benefits of Reducing Radon Exposures--Central Tendency Estimate

                                                [$Millions, 1997]

----------------------------------------------------------------------------------------------------------------

                                                                                                      Annual

                       Radon level (pCi/l)                          Annualized    Cost per fatal     monetized

                                                                       costs      cancer avoided     benefits

----------------------------------------------------------------------------------------------------------------

4000 \3\........................................................              25            11.3              13

2000............................................................              46             7.1              38

1000............................................................              98             5.9              96

700.............................................................             148             6.0             145

500.............................................................             218             6.0             212

300.............................................................             373             6.4             343

100.............................................................             795             6.9             673

----------------------------------------------------------------------------------------------------------------

Notes: 1. Benefits are calculated for stomach and lung cancer assuming that risk reduction begins immediately.

  Estimates assume a $5.8 million value of a statistical life and willingness to pay of $536,000 for non-fatal

  cancers.

2. Costs are annualized over twenty years using a discount rate of seven percent.

3. 4000 pCi/l is equivalent to the AMCL estimated by the NAS based on SDWA requirements of Section 1412(b)(13).



    The total annualized cost per fatal cancer case avoided is $11.3

million at a radon level of 4,000 pCi/l, drops to around $6.0 million

for radon levels in the range of 1,000 to 500 pCi/l, and increase again

back to $6.9 million per life saved at the lowest level of 100 pCi/l.



BILLING CODE 6560-50-P



[[Page 9592]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.004







[[Page 9593]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.005





BILLING CODE 6560-50-C



[[Page 9594]]



6.8  Sensitivities and Uncertainties



6.8.1  Uncertainties in Risk Reduction and Health Benefits Calculations

    The estimates of risk and risk reduction are derived based on

models which incorporate a number of parameters whose values are both

uncertain and highly variable. Thus, the estimates of health risks and

risk reduction are uncertain. In addition, to the extent that age-

specific smoking prevalence rates change, the risk from radon in

drinking water will change.

    The cost of fatal cancers tend to dominate the monetized benefits

estimates. Approximately 94 percent of the cancers associated with

radon exposure and prevented by exposure reduction are fatal cancers of

the lung and stomach. In addition, the estimated value of statistical

life ($0.7 to 16.3 million dollars, with a central tendency estimate of

$5.8 million, 1997$) is much greater than the estimated willingness-to-

pay to avoid non-fatal cancers ($169,000 to $1.05 million, with a

central tendency estimate of $536,000, 1997$). If the COI measures are

used, non-fatal cancers account for an even smaller proportion of the

total monetized costs of cancers, since the medical care and lost-times

costs for lung and stomach cancer are on the order of $108,000 and

$114,000, respectively (1997$).



    Unless the VSL is assumed to be near the lower end of its range,

the assumptions made regarding the monetary value of non-fatal cancers

are not a major source of uncertainty in the estimates of total

monetary benefits. For most reasonable combinations of values, the VSL

is the major contributor to the overall uncertainty in monetized values

of health benefits. As shown in Table

6-2, the upper and lower estimates of the monetary benefits for a given

radon level vary by a factor of approximately 23, corresponding to the

ratios of the lower- and upper-bound estimates of the VSL.

6.8.2  Uncertainty in Cost and Impact Calculations

    The results of the cost and impact analysis are subject to a

variety of qualifications. As discussed in Section 5, the analysis is

subject to a variety of uncertainties in the models and assumptions

made in developing cost estimates. One important assumption is that for

all CWSs for which the estimated average radon level exceeds a given

level, treatment will be necessary at all sites. This is a very

important assumption, because if systems in reality have only a portion

of sites above the target level, then mitigation costs could be much

lower. EPA is currently evaluating intra-system variability in radon

levels, and will address this issue in more detail in the proposal.

    In addition, CWSs are assumed to select from only a relatively

small number of treatment methods, and to do so in known, constant,

proportions. In actuality, systems could select technologies that best

fit their needs and optimize operating conditions to reduce costs. The

analysis also relies on various cost-related input data that are both

uncertain and variable. Some of these variables are entered as

constants, others as deterministic functions. For example: treatment

technology cost functions are based on EPA cost curves derived for

generic systems; households are assumed to use a uniform quantity of

83,000 gallons/year of drinking water, regardless of geographical

location, system size, or other factors; MMM program costs are assumed

to cost $700,000 per fatal cancer case avoided, regardless of the

specific types or efficiencies of activities undertaken by the

mitigation programs. One factor that may contribute significantly to

the overall uncertainty in cost estimates is the set of the nonlinear

equations (Appendix C) used to convert population served data to

estimates of average and design flow rates for ground water systems.

Relatively small errors in the specification of this model could result

in disproportionately large impacts on the cost estimates. Similarly,

the cost curves for some of the technologies are highly nonlinear

function of flow, adding another level of uncertainty to the cost

estimates.

    Because of the complexity of the various cost models, EPA has not

conducted a detailed analysis of the uncertainty associated with the

various models and parameter values. Limited uncertainty analyses have

been performed, however, to estimate the impact of a few major

assumptions and models on the overall estimates of mitigation costs.

First, EPA has analyzed the impacts of errors of plus or minus 50

percent in the cost curves for the various radon treatment

technologies. The results of this analysis are shown in Figure 6-1.

Since water mitigation costs make up the bulk of the total costs of

meeting radon levels in the absence of MMM programs, the effect of

these changes is generally to increase or decrease the costs of

achieving the various levels by slightly less than 50 percent. It can

be seen from these results that the assumptions regarding costs can

affect the relationship between costs and monetized benefits. A

relatively small systematic change in water mitigation costs could

result in benefit estimates that either exceed, or are less than, a

wide range of radon levels.

    In addition to assuming across-the board changes in radon

mitigation costs, EPA also examined the extreme situation in which none

of the water systems would adopt GAC treatment. Since the GAC

technologies are the most expensive treatments evaluated, the costs of

meeting the various radon levels are reduced if GAC is eliminated and

systems are assumed to employ aeration instead (Figure 6-1). Since,

however, so few systems are assumed to elect GAC in the first place

(five percent or less of the smallest systems) the cost decrease of

eliminating GAC is quite small.



BILLING CODE 6560-50-P



[[Page 9595]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.006





BILLING CODE 6560-50-C



7. Implementation Scenarios--Multimedia Mitigation Programs Option



    This Section presents a preliminary analysis of the likely costs

and benefits under two different implementation scenarios in which

States choose to develop and implement multimedia mitigation (MMM)

programs to comply with the radon NPDWR.



7.1  Multimedia Mitigation Programs



    The SDWA, as amended, provides for development of an Alternative

Maximum Contaminant Level (AMCL), which public water systems may comply

with if their State has an EPA approved MMM program to reduce radon in

indoor air. The idea behind the AMCL and MMM option is to reduce radon

health risks by addressing the larger source of exposure (air levels in

homes) compared to drinking water. If a State chooses to employ a MMM

program to reduce radon risk, it would implement a State program to

reduce indoor air levels and require public water systems to control

water radon levels to the AMCL, which is anticipated to be set at 4000

pCi/l based on NAS's re-evaluation of the radon water to air transfer

factor. If a State does not choose a MMM program option, a public water

system may propose a MMM program for EPA approval.

    The Agency is currently developing guidelines for MMM programs,

which will be published for public comment along with the proposed

NPDWR for radon in August 1999. For the purpose of this analysis, the

MMM implementation scenarios are assumed to generate the same degree of

risk reduction as achieved by mitigating water alone. For example, a

MMM scenario which includes the AMCL of 4,000 pCi/l and a target water

level of 100 pCi/l is assumed to generate the same degree of risk

reduction as the 100 pCi/l level alone. Thus, the HRRCA estimates the

health risk reduction benefits of MMM implementation options to be the

same as the benefit that would be achieved reducing radon in drinking

water supplies alone.



7.2  Implementation Scenarios Evaluated



    EPA has evaluated the annual costs and benefits of two MMM

implementation assuming (1) all States (and all water systems) would

adopt MMM programs and comply with the AMCL, and (2) half of the States

(and half of the water systems) adopt the MMM/AMCL option. These

scenarios were analyzed in the absence of specific data on States'

intentions to develop MMM programs. The two scenarios, along with the

case where the MMM option is not selected by any States or water

systems (presented in Section 6), span the range of participation in

MMM programs that might occur when a radon NPDWR is implemented. At

this point, however, it is not possible to estimate the actual degree

of State participation. The economic impacts of the MMM programs at the

system or household level have not been calculated, because there is no

information at present as to how these programs would be funded or upon

who the costs would fall.

    The presentation of costs and benefits is based on analysis of

radon levels of 100, 300, 500, 700, 1,000, 2,000, and 4,000 pCi/l in

public domestic water supplies, supplemented by States (50 or 100

percent participation) implementing MMM programs and complying with an

AMCL of 4,000 pCi/l.

    For the scenario evaluated in which one-half of the States

(estimated to include 50 percent of all CWSs) were assumed to implement

a MMM program and comply with an AMCL of 4000 pCi/l option, while the

other half mitigated



[[Page 9596]]



radon in water to the target radon levels without MMM programs. In the

other scenario, all of the States (and 100 percent of the CWSs) were

assumed to adopt MMM programs and comply with the AMCL.



7.3  Multimedia Mitigation Cost and Benefit Assumptions



    For the HRRCA, a simplified approach to estimating the costs of

mitigating indoor air radon risks was used. Based on analyses conducted

by EPA (US EPA 1992B, 1994C) a point estimate of the average cost per

life saved of the current national voluntary radon mitigation program

was used as the basis for the cost estimate of risk reduction for the

MMM option. In the previous analysis, the Agency estimated that the

average cost per fatal lung cancer avoided from testing all existing

homes in the United States and mitigating all those homes at or above

EPA's voluntary action level of 4 pCi/l is approximately $700,000 (US

EPA 1992B). This value was originally estimated by EPA in 1991. The

same nominal value is used in the HRRCA based on to anecdotal evidence

from EPA's Office of Radiation and Indoor Air that there has been an

equivalent offset between a decrease in testing and mitigation costs

since 1992 and the expected increase due to inflation in the years

1992-1997. This dollar amount reflects that real testing and mitigation

costs have decreased, while nominal costs have remained relatively

constant. The estimated cost per fatal cancer case avoided by building

new homes radon-resistant is far lower (Marcinowski 1993). For the

purposes of this analysis, only the cost per fatal cancer case avoided

from mitigation of existing homes is used.

    To estimate the national cost of the MMM program's air mitigation

component, MMM costs were estimated by multiplying the cost per fatal

cancer case avoided by the number of fatal cases avoided in going from

a water radon level equal to the AMCL (4,000 pCi/l) to a water level

equal to various radon levels analyzed in the HRRCA. The number of

fatal cancer cases avoided was estimated using the risk reduction model

described in Section 3.



7.4  Annual Costs and Benefits of Multimedia Mitigation Program

Implementation



    The total annual cost of the radon levels analyzed varies

significantly depending on assumptions regarding the number of States

implementing MMM programs. This variation can be seen in Tables 7-1 and

7-2. Under an assumption that 50 percent of States choose to implement

MMM programs, the cost of the rule varies from about $38 million per

year to achieve a radon level in water of 2,000 pCi/l to about $450

million per year to achieve an level of 100 pCi/l. Assuming that 100

percent of States implement MMM programs, the cost of the rule varies

from about $29 million per year to achieve an radon level of 2,000 pCi/

l to about $106 million per year to achieve an level of 100 pCi/l.

    The monetized benefits of both MMM implementation scenarios exceed

the estimated mitigation costs across all radon levels. When the 50

percent MMM participation scenario is evaluated, the mitigation costs

at 2,000 pCi/l are just less than the estimated benefits ($38 million

versus $39.6 million, respectively). In the case of 100 percent

multimedia participation, mitigation costs begin at about 65 percent of

the benefits at a radon level of 2,000 pCi/l, and decrease rapidly so

that at 100 pCi/l the monetized benefits of radon reduction exceed the

mitigation costs by almost 7-fold.

    Assuming 50 percent MMM participation, the total cost per fatal

cancer case avoided is $5.8 million at a radon level of 2,000 pCi/l,

dropping to around $3.7 million at a level of 500 pCi/l, and increasing

slightly to about $3.9 at 100, pCi/l (Table 7-1). As expected, the cost

per fatal cancer case avoided is lowest for the 100 percent MMM

participation option, ranging from from $4.5 at a radon level of 2,000

pCi/l to about $900,000 at a level of 100 pCi/l.

    For the 50 percent MMM participation, the incremental cost per

fatal cancer case avoided decreases from 2000 pCi/l to 500 pCi/l ($8.7

million to $3.4 million, respectively), then increases to $4.1 million

at 100 pCi/l. In the case of the 100 percent MMM participation, the

incremental cost per life saved starts at about $4.3 million for the

maximum target levels of 2,000 pCi/l, and then drops sharply to about

700,000 per life saved for the other radon.



 Table 7-1.--Central Tendency Estimates of Annualized Costs and Benefits of Reducing Radon Exposures With 50% of

                                      States Selecting the MMM/AMCL Option

                                                [$million, 1997]

----------------------------------------------------------------------------------------------------------------

                                     Water mitigation component             Multimedia mitigation component

                             -----------------------------------------------------------------------------------

                                                             Cost per                                   Cost per

     Radon level (pCi/l)                            Fatal      fatal                          Fatal      fatal

                               Annual    Annual     cancer    cancer     Annual    Annual     cancer     cancer

                                costs   benefits    cases      case      costs    benefits    cases       case

                                 \2\               avoided    avoided                        avoided    avoided

----------------------------------------------------------------------------------------------------------------

Baseline....................         0         0        0    ........        0           0        0         0

4000........................        25        13        2.2      11.3        0           0        0         0

2000........................        35        25        4.3       8.2        2.3        13        2.2       1.1

1000........................        61        54        9.0       6.6        5.8        42        7.1       0.81

700.........................        86        78       13         6.4        8.6        66       11         0.77

500.........................       121       112       19         6.3       12.7        99       17         0.74

300.........................       199       177       30         6.6       20         164       28         0.73

100.........................       410       341       58         7.0       40         328       56         0.71

----------------------------------------------------------------------------------------------------------------

\1\ Equivalent to the cost of complying with an AMCL of 4000 pCi/l.





[[Page 9597]]





  Table 7-2.--Central Tendency Estimates of Annualized Costs and Benefits of Reducing Radon Exposures With 100% of States Selecting the MMM/AMCL Option

                                                                    [$million, 1997]

--------------------------------------------------------------------------------------------------------------------------------------------------------

                                                                          Water mitigation component               Multimedia mitigation component

                                                                   -------------------------------------------------------------------------------------

                                                                                                  Cost per

                        Radon level (pCi/l)                                               Fatal     fatal                           Fatal      Cost per

                                                                     Annual    Annual    cancer    cancer     Annual     Annual     cancer      fatal

                                                                    costs\1\  benefits    cases     case      costs     benefits    cases    cancer case

                                                                                         avoided   avoided                         avoided     avoided

--------------------------------------------------------------------------------------------------------------------------------------------------------

Baseline..........................................................         0         0       0.0  ........        0.0        0.0        0.0        0.0

4000..............................................................        25        13       2.2      11.3        0.0        0.0        0.0        0.0

2000..............................................................        25        13       2.2      11.3        4.6       25          4.4        1.1

1000..............................................................        25        13       2.2      11.3       12         83         14          0.81

700...............................................................        25        13       2.2      11.3       17        131         23          0.77

500...............................................................        25        13       2.2      11.3       25        198         34          0.74

    300...........................................................        25        13       2.2      11.3       41        328         56          0.73

100...............................................................        25        13       2.2      11.3       80        654        112          0.71

--------------------------------------------------------------------------------------------------------------------------------------------------------

\1\ Equivalent to the cost of complying with an AMCL of 4000 pCi/l.



7.6  Sensitivities and Uncertainties





    EPA conducted a sensitivity analysis associated with potential

uncertainty in the cost-effectiveness of MMM programs. Since the value

used is a point estimate ($700,000 per life saved), and since the

ability to employ MMM programs results in substantial decreases in

estimated costs, it might be expected that changes in the cost-

effectiveness value would affect the cost estimates for these options

substantially. Figure 7-1 summarizes the impact of different estimates

of the cost of MMM programs on the total cost of radon mitigation.

Costs are graphed for the 50 percent and 100 percent participation

options for radon level. Costs were estimated for a high-end case

(assuming a MMM cost 50 percent above the central tendency value), a

low-end case (50 percent below the central tendency), and for a central

tendency case that assumes the current $700,000 per life saved as the

MMM cost.

    The relative impacts of changing MMM costs on the total costs of

reducing radon exposure can also be seen in Figure 7-1. The figure

illustrates that the central tendency estimate of monetized benefits is

e well above the estimated costs for all ranges except for the high-end

estimate of the 50 percent MMM participation scenario. This is due to

the greater impact of water mitigation costs relative to the MMM cost

component to total costs compared to the 100 MMM scenario, where the

MMM component contributes the largest share to total costs.



BILLING CODE 6560-50-P



[[Page 9598]]



[GRAPHIC] [TIFF OMITTED] TN26FE99.007





BILLING CODE 6560-50-C



References



Davis, RMS and JE Watson Jr. ``The Influence of Radium Concentration

in Surrounding Rock on Radon Concentration in Ground Water,''

University of North Carolina, Chapel Hill: March 13, 1989.

Longtin, J.P. ``Occurrence of Radon, Radium, and Uranium in

Groundwater.'' Journal of the American Water Works Association.

July, 1987.

Marcinowski, F. and S. Napolitano. ``Reducing the Risks from

Radon.'' Air and Waste, Vol. 43, 955-962, 1993.

NAS. 1998A. ``Health Effects of Exposure to Radon--BEIR VI (Pre-

Publication Copy),'' National Academy Press: Washington, DC.

NAS. 1998B. ``Risk Assessment of Radon in Drinking Water (Pre-

Publication Copy),'' National Academy Press: Washington, DC,

September 15.

US EPA. 1999A. ``Point Estimate of Radon Unit Risks,'' Memo to Mike

Osinski from Nancy Chiu, Office of Science and Technology, January

22.

US EPA. 1999B. ``Radon in Drinking Water Health Risk Reduction and

Cost Analysis: Appendices,'' Office of Ground Water and Drinking

Water, February.

US EPA. 1998A. ``Cost Evaluation of Small System Compliance Options:

Point-of-Use and Point-of-Entry Treatment Units,'' Office of Ground

Water and Drinking Water.

US EPA. 1998B. ``Cost of Lung Cancer (Draft),'' Office of Ground

Water and Drinking Water, October.

US EPA. 1998C. ``Cost of Stomach Cancer (Draft),'' Office of Ground

Water and Drinking Water, October.

US EPA. 1998D. ``Evaluation of Full-Scale Treatment Technologies at

Small Drinking Water Systems: Summary of Available Cost and

Performance Data,'' Office of Ground Water and Drinking Water,

December.

US EPA. 1998E. ``Guidelines for Preparing Economic Analyses--Review

Draft,'' Office of Policy, November.

US EPA. 1998F. ``Health Risks from Low-Level Environmental Exposure

to Radionuclides,'' Office of Radiation and Indoor Air, Federal

Guidance Report No. 13, Part I--Interim Version, EPA 401/R-97-014.

US EPA. 1998G. ``Model Systems Report (Draft),'' Office of Ground

Water and Drinking Water, March.

US EPA. 1998H. ``National-Level Affordability Criteria Under the

1996 Amendments to the Safe Drinking Water Act,'' Office of Ground

Water and Drinking Water, August 19.

US EPA. 1998I. ``National Primary Drinking Water Regulations:

Disinfectants and Disinfection Byproducts; Final Rule,'' 63 FR No.

241, 69390-69476, December 16.

US EPA. 1998J. ``Potential Benefits of the Ground Water Rule--Draft

Final Report,'' Office of Ground Water and Drinking Water, February.

US EPA. 1998K. ``Radon Cost Estimate,'' Memo to Bill Labiosa from H.

McCarty, Office of Ground Water and Drinking Water, December 4.

US EPA. 1998L. ``Re-Evaluation of Radon Occurrence in Ground Water

Supplies in the United States--External Review Draft,'' Office of

Ground Water and Drinking Water, September 30.

US EPA. 1998M. ``Regulatory Impact Analysis of the Stage I

Disinfectants/Disinfection By-Products Rule,'' Office of Ground

Water and Drinking Water, November 12.

US EPA. 1998N. Safe Drinking Water Suite Model. Inputs from Version

3.4 of the Cost Library and Version 4.0 of the What If Module.

US EPA. 1998O. ``Technologies and Costs for the Removal of Radon

From Drinking Water,'' Office of Ground Water and Drinking Water,

September.



[[Page 9599]]



US EPA 1997A. ``Community Water System Survey. Volume II: Detailed

Survey Result Table and Methodology Report.'' Office of Ground Water

and Drinking Water. EPA 815-R-97-001b, January.

US EPA. 1997B. ``Withdrawal of the Proposed NPDWR for Radon-222,''

62 FR No. 151, 42221-42222, August 6.

US EPA. 1995. ``Uncertainty Analysis of Risks Associated with Radon

Exposures in Drinking Water,'' Office of Science and Technology,

Office of Policy, March.

US EPA. 1994A. ``A Citizen's Guide to Radon (Second Edition): The

Guide To Protecting Yourself and Your Family from Radon,'' Office of

Radiation and Indoor Air. EPA-402-K92-001, September.

US EPA. 1994B. ``Final Draft for the Drinking Water Criteria

Document on Radon,'' Office of Water, September 30.

US EPA. 1994C. ``Report to the United States Congress on Radon in

Drinking Water, Multimedia Risk and Cost Assessment of Radon,''

Office of Water. EPA-811-R-94-001.

US EPA. 1993A. ``EPA's Map of Radon Zones: National Summary,''

Office of Air and Radiation, EPA 402-R-93-071, September.

US EPA. 1993B. ``Uncertainty Analysis of Risk Associated with

Exposures to Radon in Drinking Water,'' Office of Science and

Technology, April 30.

US EPA. 1992A. ``National Residential Radon Survey: Summary

Report,'' Office of Air and Radiation, EPA 402-R-92-011, October.

US EPA. 1992B. ``Technical Support Document for the 1992 Citizen's

Guide to Radon,'' Office of Air and Radiation. EPA 400-R-92-011.

US EPA. 1991. ``National Primary Drinking Water Regulations:

Radionuclides: Notice of Proposed Rulemaking,'' 56 FR No. 138,

33050-33127, July 18.

US EPA. 1989. ``Analysis of Potential Radon Emissions from Water

Treatment Plants Using the MINEDOSE Code,'' Memo to Greg Helms from

Marc Parotta, Office of Ground Water and Drinking Water.

US EPA. 1988. ``Preliminary Risk Assessment for Radon Emissions from

Drinking Water Treatment Facilities,'' Memo to Stephen Clark from

Warren Peters and Chris Nelson, Office of Water.

Viscusi, WK, WA Magat, and J. Huber. ``Pricing Environmental Health

Risks: Survey Assessments of Risk-Risk and Risk-Dollar Trade-Offs

for Chronic Bronchitis.'' Journal of Environmental Economics and

Management, 21:32-51, 1991.



[FR Doc. 99-4416 Filed 2-25-99; 3:08 pm]

BILLING CODE 6560-50-P



Jump to main content.