REPORT TO THE UNITED
STATES CONGRESS ON
RADIONUCLIDES IN DRINKING
WATER
MULTIMEDIA RISK AND COST
ASSESSMENT OF RADON IN
DRINKING WATER
-------
REPORT TO THE UNITED
STATES CONGRESS ON
RADiONUCLIDES IN DRINKING
WATER
DRAFT
MULTIMEDIA RISK AND CC3T
ASSESSMENT OF RADON IN
DRINKING WATER
PREPARED FOR PL 102-389
Office of Water
U.S. Environmental Protection Agency
Washington, D.C.
JULY 15, 1993
-------
TABLE OF CONTENTS
SECTION PAGE
EXECUTIVE SUMMARY
PART ONE: INTRODUCTION
1 BACKGROUND AND INTRODUCTION 1-1
PART TWO: RISK
2 RISK ASSESSMENT OF EXPOSURE TO RADON FROM
PUBLIC WATFD PPLIES ... 2-1
.*»• ' i
3 RISK ASSESSMENT OF EXPOSURE TO RADON IN AIR 3-1
PART THREE: COST
4 COST ESTIMATES FOR CONTROLLING RADON 4-1
5 COST OF RISK REDUCTION 5-1
PART FOUR: COMMENTS
6 SUMMARY OF SAB COMMENTS 6-1
7 EPA DISCUSSION OF ISSUES RAISED BY SAB COMMENTS 7-1
BIBLIOGRAPHY
-------
EXECUTIVE SUMMARY
EPA prepared this report in response to the Congressional mandate in Public Law 102-389
(the C ‘ utenberg Amendment to EPA’s i ppropriation Bill, enacted . 5er 6, 1992)
which directs the Auininistrator of the U.S. Environmental Protection Agency (EPA) to report to
Congress on EPA’s findings regarding the risks of human exposure to radon, the costs for
controlling or mitigating that exposure, and the risks posed by treating water to remove radon.
The Chafee-Lautenberg Amendment called for an explicit multimedia comparison of the risks
from radon in indoor air and drinking water. Congress placed this requu’ement on the Agency
because of the concern voiced in the United States over the costs to be incurred by public water
systems in the control of radon in drinking water while a larger threat from indoor air was not
being addressed except through voluntary measures. Amendments to the Safe Drinking Water
Act in 1986 called for the regulation of radon in drinking water.
Radon, a naturally occurring gas, is colorless, odorless, tasteless, chemically ineit and
radioactive. People c ‘ie exposed to waterborne radon either by ingestion or inhalr ‘n When
ingested, radon is distributed throughout the body, which increases the cancer risk to many
organs. Radon also is released into indoor air f om tap water used for showering, washing, or
other domestic uses, or when the water is stirred, shaken, or heated before being ingested.
Radon released to the air from water adds to the airborne radon from other sources, increasing
the risk of lung cancer.
Radon decay products pose far greater risks than radon gas itself. Therefore, EPA has
given them the greatest attention in its analysis of the inhalation risks of radon. The analyses for
outdoor radon and residential radon focus on the risks from radon progeny only. The results of
those analyses help place the inhalation risks from radon in dnnking water in perspective. EPA
recently completed an analysis that evaluates the uncertainty of key parameters in the risk
analysis for radon progeny.
People c :.posed to waterboine radon 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 released from water. An individual’s combined risk during a
lifetime from constant use of drinking water with one picocurie 1 of radon per liter is close to 7
chances in 10 million of contracting fatal cancer.
Many public water supplies use water from ground water wells containing radon, although
the concentration of radon in drinking water varies widely. While high radon levels may occur
in drinking water from ground water supplies in areas where there are large amounts of
underground natural radioactive materials such as radium and uranium, radon levels in surface
water typically are very low. Surface water generally lacks a source of radon from rocks, and
radon in surface water escapes quickly into the air. Radon levels aie determined partiy by the
geologic formations that store and transport ground water, but also are influenced by the
proximity of radioactive elements like uranium that are precursors to radon.
iA curie (Ci) is a standard measure of radioacuviiy, and a picocune (pCi) is one lnilionth (I x 1O 2) of a cune
-------
VI
EPA estimates that 81 million Americans obtain their water from community ground water
supplies. Based on EPA’s analysis of existing data, the population-weighted average radon
activity in ground water serving these 81 million people is 246 picocuries per liter of water
(PCiILwaier). Radon in water exceeds 100 PCi/Lwater in 72 percent of the ground water sources
surveyed. nia 1 . blic water systems generah rely on ground wateP, nd nd to have higher
radon concentrations Because smaller systems tend to have higher radon concentrations it
follows that the burden of the costs for mitigating radon in drinking water would be on the small
systems
After a person ingests radon in water, the radon passes from the gastrointestinal tract into
the blood, pri. cipally by way of the small intestine. The blood then circulate. 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 vanes from organ to organ The
tissues of the stomach, intestines, liver, and lungs appear to receive the greatest doses.
The hi man health risks from ingesting radon in water depend on the total quantity of radon
ingested and the risk factor for ingested radon. The quantity of radon people ingest depends on
the volume of water they ingest and the initial concentration of radon in the water. It also
depends on the fraction of the radon remaining in the water at the time of ingestion. That
amount varies because radon is a volatile gas; it begins to escape from water as soon as the
water is discharged from the tap. EPA’s estimates of the health risk associated with ingesting
radon in drinking water supplied by ground water have taken all of these factors into account.
Consequently, calculated estimates of the individual health risk from ingesting radon in water are
a product of the volume of water ingested that contains radon, the fraction of radon remaining in
water at the time of ingestion, the cancer risk factor (cancer fatality risk per picocurie (pCi) of
radon ingested) and the concentration of radon in water. To calculate the population risk, total
exposed population also needs to be taken into consideration.
Assessing the risks from inhaling radon progeny requires information on how much of the
radon released through household water use enters the air and is converted into piogeny that
individuals inhale. Given the amount of radon progeny individuals inhale, EPA uses a dose-
response factor that estimates the relationship between the radon dose received and the health
effects that result. EPA calculated radon risk as a product of the concentration of radon in
drinking water, a transfer factor, which is the relationship between the radon concentration in
indoor air derived from water and the initial concentration of radon in water; the equilibrium
factor, which is the fraction of the potential energy of radon progeny that actually exists in
indoor air compared to the maximum possible energy under true equilibrium; the occupancy
factor, which is the fraction of time individuals spend in their homes, exposed to indoor radon; a
risk factor, which estimates the risk of lung cancer death from exposure to a given amount of
radon; and the total exposed population, which is the number of pepple exposed to the airborne
radon progeny resulting from household use of water. The first four factors determine the
amount of exposure to radon progeny that occurs. The risk factor describes the exposure
response relationship between lung cancer deaths and exposure. This factor enables EPA to
estimate the risk that can result from a given level of exposure. To calculate the population risk,
the total exposed population is also taken into consideration.
-------
V II
Since the Proposed Rule for Radionuclides in Drinking Water was published in July 1991,
EPA has revised the risk estimates for radon and conducted a quantitative uncertainty analysis.
This analysis quantifies the uncertainties in exposure and toxicology and estimates variation in
exposure among individuals The revised analysis of the risks of inhaling radon r Jeased from
drinking produced lower overall est i. ate of risk by roughly ond thfrL --l1owever. the
revised analysis of risks óf ingestion of radon gas from drinking water has produced higher
overall estimates of risks by roughly one third
The combined lifetime cancer risk per pCiIL in water from all pathways (inhalation of
radon progeny due to radon released from water, inhalation of radon gas released from water to
indoor air, and ingestion of radon gas in direct tap water) was revised to 6.7 x l0 . EPA’s best
estimate for the individual lifetime inhalation risk of lung cancer deaths per pCiIL of radon in
drinking water is 3.0 x l0 with a credible range of 1.8 x l0 to 7.0 x l0 . EPA estimated
that the individual lifetime fatal cancer risk for inhaling waterborne radon gas is 2 x 10-8 per
P” 1 waier Lastly, EPA’s best estimate for the individual lifetime ingestion risk of fatal cancers
per pCi/L of water is 3.5 x l0 with a credible range of 6.9 x 10-8 to 6.4 x l0- .
Summary of Proposed and Revised Fatal Cancer Risk Estimai.c,
for Radon in Water
Exposure Pathway
Lifetime Cancer Risk
per pCiIL in Water
Proposed
Revised
Inhalation of Radon Progeny
Derived from Waterborne
Radon Gas
4.9 x l0-
74%
3.0 x l0
45%
Inhalation of Radon Gas
Release’ 1 from Water to
Indoor Air
0.2 x l0
3%
0.2 x l0
3%
Ingestion of Radon Gas in
Direct Tap Water
Sum of All Pathways
1.5 x
23%
-7
6.6x10
3.5 x l0
52%
-7
6.7x 10
The following exhibit shows the number of estimated cancer fatalities per year due to
various pathways of radon exposure, based on revised fatal cancer risk estimates and occuiTence
data. EPA’s best estimate of total lung cancer deaths caused by inhalation of radon in drinking
water, for example, is 86, with a credible range of 48 to 233 per ye ,ar. EPA’s best estimate of
total lung cancer deaths caused by ingestion of radon in drinking water is 100, with a credible
range of 19 to 166 per year. The threat from radon in drinking water is about half (45 percent)
due to inhalation and about half (52 percent) due to ingestion of drinking water. These
calculations also indicate that the population risks from radon in drinking water are similar to or
-------
Viii
higher than currently known risks from most chemical pollutants in drinking water that are now
subject to regulation.
( ncer Fatalities per Year due to Exposure to RadQn
Exposure Pathway
Lower
Estimate
EPA’s Best
Estimate
Upper Estimate
Inhalation due to Radon
Treatment
--
--
-------
I x
Summary of EPA’s Best Estimates of Risk, Fatal Cancer Cases, Cancer Cases Avoided,
and Costs for Mitigating Radon in Water and Air
Drinking Water Indoor Air
Target Level 300 PCi/Lwater 4 PC /L .
lndiv dual Lifetime Risk of Fatal
2 in 10,000 1 in 100
Cancer at Target Level
Number of People Above Target
19 million 15 million
Level
Number of Fatal Cancer Cases per 192 13,600
Year
Number of Fatal Cancer Cases Avoid-
ed Annually by Me 1 11 the Target 85** 2.2tiU
Level
Total Annual Cost for Mitigating
$272 million $l,9S0 million
Radon
Average Cost per Fatal Cancer Case .
$3.2 million $0.9 million
Avoided
: Includes those exposed above and below the target level.
Includes community and non-transient non-community water systems
Assumes 100% voluntary monitonng and mitigation.
The cost numbers presented in the above table for mitigating radon in drinking water and
indoor air are difficult to compare. In the case of water, the costs incurred by public water
systems whose ground water supplies have radon levels above the proposed Maximum
Contaminant Level would be passed on to consumers in their water bills. Because indoor air
mitigation is a voluntary program, homeowners bear the cost of testing their own radon levels
and taking corrective action to lower those levels if necessary.
Based on the total annual cost of mitigating radon exposure through drinking water, EPA
calculates that $272 million would be required to prevent the deaths of approximately 85 people
each year due to radon in water. This is an average of $3.2 million dollars per life saved.
Similarly, if the air in all homes with radon concentrations greater than the target level of 4
pCi1L jj for indoor air is reduced to a level of 2 PCiJL r (the average level achievable with
current mitigation technologies), approximately 2,200 lives would be saved each year ; a total
cost of $1,980 million. EPA calculates that the average cost borne by homeowners to mitigate
radon in indoor air would total approximately $900,000 per life saved.
-------
PART ONE
INTRODUCTION
-------
1. BACKGROUND AND Th TRODUCTION
Public Law 102-389 (the Chafee-Lautenberg Amendment to EPA’s Appropnation Bill,
enacted October 6, 1992) directs the Administrator of the U.S Environmental Protection Agency
(EPA) to report to Congress on EPA’s findings regarding the risks of human exposure to radon,
the cost for controlling or mitigating that exposure, and the risks posed by treating water to
remove radon. EPA has prepared this report in response to the Congressional mandate.
I. Part One reviews the requirements of the Safe Drinking Water Act (SDWA)
and Congress’ mandate to EPA regarding radon It also reviews the toxicity
and occurrence of radon in water and air, summarizes and compares EPA’s key
risk estimates for radon, and summarizes the total fatal cancer risks from radon.
2. Part Iwo responds to Congress direction to EPA to report on the human health
risks associated with various types of exposure to radon. Chapter T u explains
EPA’s assessment of the risks from ingestion and inhalation of radon (and its
progeny) from community public water supplies relying on ground water.
Chapter Three addresses EPA’s risk assessment of radon in indoor and outdoor
air. Chapter Three also includes an assessment of the risk from drinking water
treatment facilities.
3. Part Three addresses the costs for treating radon and compares the costs of
water treatment to the costs of reducing radon in indoor air. This section
responds to Congress’ direction that EPA consider the costs of mitigating
exposure to radon as well as the costs that households and communities --
including small communities -- would experience as the result of regulating
radon. Chapter Four discusses the unit costs of treatment, anc 1 xp1ains how
EPA determined the national and household cost estimates Chaptef Five
discusses the cost-effectiveness of treatment for radon in both water and air.
4. Part Four addresses the Science Advisory Board’s (SAB’s) review of EPA’s
risk and cost assessment. Chapter Six is the SAB’s review of EPA’s studies of
radon in drinking water. Chapter Seven is EPA’s responses to the SAB’s
comments.
Information presented in these sections are drawn from the following documents: Drinking
Water Criteria Document for Radon in Drinking Water (USEPA, l99la); National Pri,nary
Drinking Water Regulations; Radionuclides: Notice of Proposed Rulenwking (USEPA, 1991e);
Uncertainiy Analysis of Risk Associated with Exposure to Radon in Drinking Water (USEPA,
1 993h); The Occurrence and Exposure Assessments for Radon, Radium-226, Radium-228,
Uranium, and Cro. Alpha Particle Activity in Public Drinking Water Supplies (Revised
Occurrence Estimates Based on conzment to the Proposed Rudionuclides Regulations) (USEPA,
-------
1-2
1 992a); Technical Support Document for the 1992 Citizen’s Guide to Radon (USEPA, l992i);
A memo fi onz Marc Purrotta to Greg Helms Regarding An Analysis of Potential Radon
Emissions from Water Treatment Plants Using the Minedose C’ode (Nov. 22, 1 989a); A Memo
from Christopher Nelson, Office of Radiation anc Indoor Air, to Marc Parrotta regarding A
Review of Ri Asi ’Lenrs of Radon from Drinking Water Tr .itment Facilities (Jan
28, 1993); and the Regulatory Impact A ialysis of Proposed National Primary Drinking Water
Standai (Is f )r Radionuclides (USEPA, 1991k).
1.1 SAFE DRINKING WATER ACT REQUIREMENTS _____________
In 1974, the United States Congress enacted the Safe Drinking Water Act. In 1986,
Congress updated the program to set mandatory guidelines for regulating key contaminants,
require the monitoring of unregulated contaminants, establish benchmarks for treatment
technologies, bolster enforcement, and promote protection of ground water sources. The
amendments gave EPA three years to set standards for 83 contaminants, includin radon.
Section 1412 of The SDWA requires EPA to publish Maximum Contaminant Level Goals
(MCLGs) and promulgate National Primary Drinking Water Regulations (NPDWRs) at
enforceable Maximum Contaminant Levels (MCLs) for contaminants that may cause any adverse
effect on human health and that are known or anticipated to occur in public water systems.
1.2 THE CONGRESSIONAL MANDATE TO EPA
As required by the SDWA and in accordance with a court-ordered deadline, EPA issued
proposed MCLGs and MCLs for radon and other radionuclides in drinking water on July 18,
1991. EPA was under a court order to promulgate final regulations by April 15, 1993. En the
fall of 1992, Congress passed the EPA’s 1993 Appropriation Bill with a requirement that EPA
conduct risk cost assessments of radon. This report is due to Congr s by July 30, 1993.
The bill also authorized the Administrator to seek an extension of the deadline for the final radon
regulations to October 1, 1993, which was subsequently approved by the court. Exhibit 1-1
recounts the items to be considered in EPA’s report under P.L. 102-389.
Analysis of data from a variety of studies since 1986 provided the scientific basis for the
proposed rule on radon in drinking water and the subsequent revisions summarized in this report.
This report to Congress summarizes a detailed study of health risk due to radon exposure from
drinking water and the potential cost of mitigating these risks. Although the focus of this report
is on the risk and cost of radon from drinking water, it also compares drinking water risks and
costs with those of radon from air and with those from treatment of radon in drinking water as
called for by the Chafee/Lautenberg amendment. The scientific evidence amassed to support the
risk assessment of radon is among the strongest EPA has used to assess the health effects of an
environmental pollutant. EPA’s risk estimates are based on reports from the Biological Effects
of Ionizing Radiation (BEIR) committee of the National Academy of Sciences (NAS) and on the
work of the International Commission on Radiological Protection (ICRP). Throughout the radon
-------
1-3
Exhibit 1-1
Requirements of the Chafee-Lautenberg Amendment to EPA’s Appropriation Act
In “‘93 Appropriations Bill, Congre required EPA to:
(I) Report on the risk of adveise human health effects associated with exposure to various
pathways of radon;
(2) Report on the costs of controlling or mitigating exposure to radon;
(3) Report on the costs for radon control or mitigation experienced by households and
communities, including the costs experienced by small communities as the result of such
regulation;
(4) Consider the risks posed by the treatment or disposal of any waste produced by water
treatment;
(5) Have the Science Advisory Board review the EPA’s study and submit a recommendation
to the Administrator on its findings; and
(6) Report the Administrator’s findings and the Science Advisory Board’s recommendations to
the Senate Committee on Environment and Public Works and the House Committee on
Energy and Commerce.
risk assessment process, the independent SAB, an advisory panel of experts from the scientific
community, reviewed EPA documents for scientific accuracy and made specific comments and
recommendations.
1.3 DESCRIPTION OF THE PROBLEM
National and international health organizations have established that radon is a human
carcinogen. In 1988, the International Agency for Research on Cancer (IARC) convened a panel
of world experts who agreed unanimously that sufficient evidence exists to conclude that radon
causes cancer in humans and in experimental animals. The BEER Committees, the ICRP, and the
National Council on Radiation Protection and Measurement (NCRP) also have reviewed the
available data and agreed that radon exposure causes cancer in humans. EPA has concurred with
these determinations and classified radon in Group A, meaning that it is considered by EPA to
be a human carcinogen based on sufficient evidence of cancer in humans. After smoking, radon
is believed to be the second largest cause of lung cancer deaths in the United States.
There is a great deal of concern within Congress as to the extensive costs of the Safe
Drinking Water Act requirements especially as incurred by the small water systems. This
-------
1-4
concern has brought the radon regulation in drinking water under considerable scrutiny because
radon in drinking water supplied by small systems accounts for 5 percent of the overall exposure
to radon, the other 95 percent coming from radon in air. Although radon in drinking water
accounts for only 5 percent of the exposure, EPA is able to control that 5 percent through
reliable drinking . ier treatment technology. - e ”95 percent of rado l\ .n Jt an ber , educed
through mitigation techniques, but the program to control it is currently a voluntary program
within EPA. Nonetheless, the dnnking water risk from radon is significant and controllable
especially when viewed in the light of other risks already regulated under the Safe Drinking
W iter Act.
Physica. and Chemical Properties of Radon
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 nters 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 foui 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, these progeny easily attach to and
can be transported by dust and other particles in air. The decay of progeny continues until
stable, nonradioactive 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.
Nature of the Problem
The potential hazard of radon-222 and its progeny was first identified in Bohemian
underground miners in the 1920s. In the 1940’s, the increased lung cancer mortality in these
miners was shown to be associated with radon-222. By the 1950s, the hazard was attributed to
the short half-life progeny of radon-222. Since that time, epidemiological studies of various
underground miner groups have led to the development of a dose-response relationship for radon
and it.s progeny. At this time, radon-222 and its short half-lived progeny are designated as
known human carcinogens by national and international groups active in assessing carcinogens.
Because initial exposure data came fi ’m mines with very high radon levels, not much
consideration was given to environmental radon exposure until recently. As new miner groups at
lower radon exposure levels were added to the data base, it became evident that environmental
radon exposure might be an important source of risk for the U.S. population.
-------
1-5
Although the major hazard of environmental radon has been shown to be inhaled radon and
its progeny, ingested radon may also be hazardous. ingested radon is absorbed and distributed to
all body organs. Some of it will decay and emit radiation, which has the potential of inducing
cancer in any irradiated organ.
Radon Levels in inking Water
People can be exposed to waterborne radon by ingestion and inhalation (see Exhibit 1-2).
Radon dissolved in water, when ingested, is distributed throughout the body, which increases the
cancer risk to many organs. in addition, radon dissolved in tap water is released into indoor air
when it is used for showering, washing or other domestic uses, or when the .ater is stirred,
shaken, or heated before being ingested. This adds to the airborne radon from other sources,
increasing the risk of lung cancer.
Many public water supplies use water from ground water wells containing radon. The
concentratrnn of radon in this water varies widely. While high levels may occur in areas with
laige amounts of uranium helow ground, radon velc in surface water generally are very low.
Surface water usually lacks a source of radon from rocks, and radon in surface witer escapes
quickly into air. Radon levels depend partly on the geologic formations that store and tiansport
ground water, but also are influenced by the proximity of radioactive precursors to radon.
Based on the National Inorganics and Radionuclides Survey (NIRS) (Cothem, Rebers,
1990) and the Federal Reporting Data System (FRDS), EPA estimates that 81 million people -- a
majority of those consuming ground water -- use community ground water supplies. The risk
analysis completed for this report focuses only on these supplies and not on the non-transient,
non-community (NTNC) water supplies. Based on EPA’s analysis of existing data, the
population-weighted average radon activity in ground water supplied by community water
systems is 246 picocuries per liter’ of water (pCilLwater), with a 90 percent confidence interval
of 205-306 P Lwater. Radon exceeds 100 PC1tLwater in 72 pement of the ground water sources
surveyed. Small public water systems generally rely on ground water, and tend to have higher
radon concentration. The total population served by community and NTNC ground water
supplies with radon concentrations in excess of the proposed MCL of 300 PCi/Lwater is approxi-
mately 19 million people.
Radon Levels in Air
Radon exists in both indoor and outdoor air. Radon released from the soil may travel
indoors through cracks in foundations. To a lesser degree, building materials may also be a
source of radon. EPA reviewed the results of several investigations to determine radon
concent!-ations in indoor and outdoor air. To determine the annual average radon concentration
in U.S. homes, EPA used the results of its National Residential Radon Survey (a survey
measuring radon concentrations in 6,000 homes statistically represebtative of all U.S. residenc-
es) (USEPA 19910. EPA estimates that radon activity levels in all U.S. homes averages 1.25
curie (Ci) is a standard measure of radioactivity, and a picocurie (pCi) is one trillionth (I x 10 12) of a curie
-------
1-6
Exhibit 1-2
Exposure Pathways for Radon
WATER
MR
Radon in indoor air comes from
tap water, soil, and outdoor air
A From ground water source
to Public Water System
I
-------
1-7
P ILair Based on a review of The National Ambient Radon Study (Hopper, 1991) and The 1988
UNSCEAR Report (UNSCEAR, 1988; Gesell, 1983), EPA estimates the average outdoor air
radon concentration to be 0.3 PCi/Lwr.
Fatal Cancer ; ks Associated with Expusure to Radon
EPA defined the extent of the risk associated with exposure to radon (in water or air) in
terms of unit risk, individual risk, and population risk. The unit risk is the risk of fatal cancer
for an individual exposed to radon (in water or air) at 1 pCiIL for prolonged time intervals. The
individual risk is the fatal cancer risk for an individual exposed to radon (in water or air) at the
reported occurrence levels for prolonged time intervals. The population risk is the total number
of fatal cancer cases per year expected for a population exposed to radon (in water or air) at the
reported occurrence level.
The estimated risk per unit dose exposure is derived from human epidemiological studies
and from radiobiological dose estimation. To estimate the risk to an individual of getting fatal
cancer from a given r eAposure from water or air, EPA multiplied the unit ri:. ,
individual’s exposure. The population risk is calculated by summing all the indi Ju”’ ri. ks in
the population of interest, taking into account the distribution of exposure levels, as detemiined
from measured levels of radon in water or air.
1.4 SUMMARY OF DOCUMENTS PRESENTED TO THE SAB CONCERNING THE
RADON IN DRINKING WATER RULE
The EPA Science Advisory Board (SAB) was integrally involved in reviewing all phases of
the preparation of the proposed rule for radon in drinking water. This summary begins with the
SAB’s review of EPA’s scientific basis for proceeding with a risk assessment of radon in
drinking water and proceeds chronologically through June 1993.
In 1984, a specialized ad hoc subcommittee of the SAB reviewed the scientific basis for
EPA’s proposed national emissions standards for hazardous au pollutants (NESHAP) for
radionuclides. That report led to the formation of the Radiation Advisory Committee (RAC) to
“review risk assessments for radiation standards.” The RAC has reviewed subsequent EPA
studies of human health risks due to radionuclide exposure.
Since 1986, the RAC has been reviewing the adequacy of EPA’s ingestion and inhalation
risk assessment for waterborne radon. Prior to the Notice of Proposed Rulemaking, the SAB
reviewed EPA’s draft criteria documents and issued comments. The criteria documents were
revised and expanded by EPA to address SAB’s comments and were reissued with the proposed
rule. In letters sent to the EPA Administrator on January 9 and January 29, 1992, SAB
expressed its concerns about EPA’s documents. However, the documents to which SAB referred
were the earlier documents since their letters acknowledged they had not reviewed the revised
version. SAB’s concerns included: (a) uncertainties associated with the selection of particular
models, specific parameters used in the models, and the final risk estimates were not adequately
-------
1-8
addressed in any of the documents; (b) high exposure from water at the point of use (e.g.,
shower) had not been adequately addressed; (c) regulation of radon in drinking water introduces
risk from the disposal of treatment byproducts, tradeoffs which the EPA should consider more
explicitly in its regulatory decision-making, and (d) regulation and removal of radon in drinking
water may reslAll in occupational exposure Prior to the review of the drinkinj water documents,
the SAB reviewed the Oftice of Radiation Program’s risk and associated models for ionizing
radiation (1992e; 1992f) and concluded that the analysis was scientifically acceptable.
Once EPA began working on the radon report requirement, it initiated a series of
teleconference calls to discuss with SAB members what items they wanted EPA to address in its
presentation to them. These analyses were presented to the SAB in February 1993. Exhibit 1-3
summarizes the documents reviewed by the SAB in February 1993. This report is largely a
summary of those presentations. A major component of the presentations was a quantitative
uncertainty analysis. The document on uncertainty analysis briefly reviews the methods used to
derive estimated cancer risk levels from various exposure routes, quantifies the uncertainty of
each parameter used in the risk assessment, and provides the overall quantit t ve uncertainty of
the fat’il cancer r:.. .siirnates.
Exhibit 1-3
Primary Documents Reviewed by the SAB in February 1993
Completion
EPA Document Reviewed Date
Cost Modeling Update February 21, 1992
Packed Tower Aeration Cost Estimates for Radon Removal March II, 1992
Technical Support Document for the /992 Citizen’s Guide to Radon* May 1992
Simplified Equations for Estimating Radon Removal Cost via Packed Tower Jul 16 1992
Aerati ’n
Tec/titologies and Costs for the Removal of Radionuclides from Potable Water 1992
Supplies U Y
Addendum to the Occurrence and Exposure Assessments for Radon, Radium-
226, Radium-228, Uranium, and Gross Alpha Particle Activity in Public September 1992
Drinking Water Supplies
Technical Memorandum: Problems wit/i the Use of Granular Activated Carbon
January 1993
for Radon Removal
Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking 1 ’
.JdflUary ,
Water
Working Draft of the Regulatory I npact Analysis for Final NPDWR for published
Radionuclides
* The risk and measurement portions were reviewed by the SAB prior to February 1993.
-------
1-9
1.5 SUMMARY OF KEY RISK ESTIMATES FOR RADON EXPOSURE PATHWAYS
Based on a combination of rigorous inalysis of the extensive data available from epidemio-
logic i:tvestigations conducted during dec. Jes of study, the application mathernatical models,
and carefully considered judgment, and with detailed oversight by the SAB, EPA prepared its
best” or ‘nominal” point estimates regarding the number of fatal cancer cases induced by radon.
EPA also estimated the credible range of estimates that could exist given the uncertainty in each
of the key parameters. The credible range involves the use of data when it is available,
supplemented by expert judgement when necessary. It represents EPA’s best judgement of the
range expected to include the estimate with a high degree of confidence. To emphasize the
subjective dimension, uncertainty ranges are referred to as “credible ranges” rather than
confidence intervals. This report refers to EPA’s nominal point estimates as the “best estimates,”
and the upper and lower ends of the range as “high” and ‘low” estimates, respectively.
Risk From Unit Exposure
People are exposed to radon in three ways: from ingesting radon dissolve(i in water; from
inhaling radon gas released from water during household use; and from inhaling radon progeny
derived from radon gas released from water. EPA estimated the lifetime fatal cancer risk from
exposure to one picocurie per liter (pCiIL) of radon in water in public water supplies; from
exposure to one pCi/L of radon in indoor residential air; and from exposure to one pCi/L of
radon in outdoor air.
Since the proposed rule for radionuclides in drinking water was published, the EPA has
revised the risk estimates for radon and a quantitative uncertainty analysis has been conducted.
A person’s combined lifetime risk from constant use of drinking water with one picocurie of
radon per liter is close to 7 chances in 10 million of contracting a fatal case of cancer. The
uncertainty analysis incorporates quantifiable jncertainties in exposure and cancer risk, as well as
variation in exposure among individuals. EPA’s best estimate for the ind vidual lifetime
inhalation risk ot lung cancer deaths per pCifL of radon in drinking water is 3.O x l0 with a
credible range of 1.8 x l0- to 7.0 x l0- . EPA’s best estimate for the irdividual lifetime
ingestion risk of fatal cancers per pCiJL of water is 3.5 x l0 with a credible range of 6.9 x 108
to 6.4 x I 0 . in addition, EPA estimated that the individual lifetime cancer risk for inhaling
waterborne radon gas is 2 x 10-8 per PC water. Ingestion of waterborne radon accounts for 52
percent of the estimated risk associated with radon in water. The remainder of the risk comes
from inhaling radon progeny (45 percent) and radon gas (3 percent). Exhibit 1-4 presents both
the proposed and revised fatal cancer risk estimates for radon in water by exposure pathway.
The estimates of risk from exposure to residential radon levels and radon in outdoor air
only include the risk associated with radon progeny and do not consider the risk of inhaling
radon gas itself, since that risk is estimated to be a small portion of the total radon-related risk.
EPA’s best estimates for individual lifetime risk associated with exposure to radon progeny in
indoor and outdoc. air tre 3.0 x l0 and 4.8 x l0 per pCi/L r, respectively.
-------
i-jr
Exhibit 1-4
Summary of Proposed and Revised Fatal Cancer Risk Estimates
for Radon in Water
Exposure Pathway
Lifetime Cancer Risk
per pCiIL in Water
Proposed
Revised
inhalation of Radon Progeny Derived
from Waterhorne Radon Gac
49 x 10
74%
30 x
45%
inhalation of Radon Gas Released from
Water to Indoor Air
02 x l0
3%
0 2 x i0
3%
Ingestion of Radon Gas iii Direct Tap
Water
I 5 x i0
23%
3 5 x i0
52%
Sum of All Pathways
66 x l0
I00 %
67 x I0
i0O9
Individual Risk
The estimated lifetime fatal cancer risk from exposure to 1 pCiIL of radon in public water
supplies is less than one chance (0.67) in a million, but the average level of radon in public
water supplies containing radon is 246 PCdLwaier Thus, the average lifetime risk for people
served by public water systems containing radon is 165 chances in a million. By comparison,
the lifetime fatal cancer risk from exposure to airborne radon in homes is 3,024 chances in a
million for each pCiIL of radon in indoor air. Because the average level of radon in homes is
1.25 PCI/Lair, the estimated lifetime risk is 3,780 chances in a million.
Populat ii Risk
EPA has developed annual population nsk estimates for the three types of radon exposure
covered in this report. EPA estimates cancer deaths per year due to exposure to radon in public
water supplies for the 81 million people served by community ground water supplies; from
exposure to radon in indoor residential air for the total U.S population (250 million people); and
from exposure to radon in outdoor air for the total U.S population.
Based on the populations of people currently exposed to all levels of radon, EPA estimates
that the total number of fatal cancers that will occur as a result of exposure to radon supplied by
community ground water systems is 192 per year as shown in Exhibit 1-5. The total number of
fatal cancers is broken down into three categories: inhalation of radon gas, 6; inhalation of radon
progeny, 86; and ingestion of radon, 100. However, there is uncertainty in estimating the risk
from ingestion and inhalation of radon. The total number of deaths from inhalation of
waterborne radon progeny and ingestion of radon in water could range from 19 to 166 deaths per
-------
I—li
year and 48 to 233 deaths per year, respectively. Since the relative risk from inhalation of radon
gas is small (about 3 percent), no uncertainty analysis was performed.
An estimated 13,600 lung cancer cases every year will result from exposure to radon in
indou •th a credible range of 6,700 to 3( 500 deaths. An estim tted lung cancer deaths
every year will result from exposure to radon’in outdoor air, with a credible range of 280 to
1,500.
Exhibit 1-5
Cancer Fatalities per Year due to Exposure to Radon
Exposure Pathway
Lower
Estimate
EPA’s Best
Estimate
Upper
Estimate
Inhalation due to Radon Trealment
--
--
<1
inhalation or Radon Gas Released
from Water to indoor Air
--
6
--
Inhalation of Radon Progeny Derived
from Waterborne Radon Gas
48
86
233
ingestion of Radon Gas in Drinking
Water
19
100
166
Inhalation from Outdoor Air
280
520
1,500
Inhalation from Indoor Air
6,740
13,600
30,600
1.6 SUMMARY OF RISK AND COST OF MITIGATION
Exhibit 1-6 compares EPA’s estimates of individual lifetime risk of fatal cancer at the
target levels with the total annual fatal cancer cases caused by radon at all levels of exposure,
the number of cancer cases that could be avoided each year by reducing radon exposure to the
target levels, and the costs for reducing radon exposure through both water and air pathways to
the target levels. The target level for drinking water is the proposed MCL of 300 pCi/L The
proposed MCL of 300 pCiJL for drinking water was set as close to the MCLG of 0 pCi/L as at
the time technically feasible. 1 The target level for air is EPA’s action level of 4 pCUL r set in
light of current mitigation and measurement technology.
As the exhibit shows, approximately 19 million people are exposed to a radon level above
the proposed MCL of 300 PCiiLwater. Approximately 2 of every 10,000 individuals exposed
Since the proposal, newly evaluated data indicates the MCL in water could be technically feasible at weil below
300 P 1Lwater
-------
1-12
would develop a fatal cancer as a result of exposure to radon in water at 300 PCiiLwater
Approximately 15 million people are exposed above EPA’s voluntary action level of 4 pCi/L 1 .
Approximately I in every 100 individuals would develop a fatal cancer as a result of exposure to
radon in indoor air at 4 pCiIl
The cost figures presented in Exhibit 1-6 are difficult to compare. En comparing drinking
water to indoor air it is important to understand that EPA’s drinking water program is under the
jurisdiction of the Safe Drinking Water Act and is a regulated program, whereas the indoor air
program is voluntary and has no regulatory authority. Although both numbers assume 100
percent compliance, the actual costs for the air program will be lower because it is a voluntary
progiam. B ,ed on the total annual cost of mitigating radon exposure through both pathways,
EPA has calculated that it would cost an average of $3.2 million dollars per life saved to prevent
the deaths of approximately 85 people each year due to radon in water. Similarly, if the air in
all homes with radon concentrations greater than the target level of 4 pCi/L tir for indoor air is
reduced to a level of 2 pCiIL 1 (the average level achievable with current mitigation
technologies), approximately 2,200 lives would be saved each year. EPA has calculated that the
cost borne y homeowners to mitigate radon in indoor air would total approximately $900,000
per life saved. -
Exhibit 1-6
Summary of EPA’s Best Estimates of Risk, Fatal Cancer Cases,
Cancer Cases Avoided, and Costs for Mitigating Radon in Water and Air
Drinking Water Indoor Air
Target Level 300 pCiIL 4 pCi/L
Individual Lifctime Risk of Fatal Cancer at
Target Level 2 in 10,000 1 in 100
Number of People Above Target Level 19 million 15 nimlimnu
Number of Fatal Cancer Cases per Year 192 13,600
Number of Fatal Cancer Cases Avoided 2
Annually by Meeting the Target Level 85 ,200
Total Annual Cost for Mitigating Radon $272 million $1,980 million
Cost per Fatal Cancer Case Avoided $3.2 million $09 million
Includes those exposed above and below the target level.
Includes community and non-transient non-community water systems.
‘ ‘ Assumes 100 pen ent voluntary monitonng and mitigation to an average of 2
PCI/Lair
-------
PART TWO
RISK
-------
2. RISK ASSESSMENT OF EXPOSURE TO RADON FROM PUBLIC WATER
SUPPLIES
Thi’ chapter responds to the Congressional mandate to EPA to report on the risk of
adverse human health effects associated with various pathways of exposure to radon It
discusses the health risks from radon in drinking water from public water systems dependent on
ground water. It also examines how EPA developed risk estimates for both ingestion and
inhalation based on a large number of studies and publications. Much of the analysis is ba. ed
on findings from BEIR reports and the ICRP.
The chapter also discusses the risk estimates presented by EPA in the proposed rule, which
were based on initial risk calculations and EPA’s final estimates, and which were later revised in
iesponse to public comments and suggestions from EPA’s Science Advisory Board (SAB). EPA
performed a quantitative uncertainty analysis of the revised risk estimates by estimating the
uncertainty of each k and exposure parameter used in obtaining the nominal estimates: The
overall uncertainties in the risk estimates were then determined by integrating the uncertainties
of the individual parameters. The four sections outlined below explain how EPA reevaluated its
estimates of the human health risks associated with radon in water:
1. Section 2 1 briefly summarizes the proposed and revised unit risk
estimates for exposure to waterborne radon and progeny via various
pathways;
2. Section 2.2 explains EPA’s assessment of the risks from ingestion of
radon in public water supplies derived from ground water;
3. Section 2.3 explains EPA’s assessment of the risks from inhalation of
radon in public water supplies relying on ground water; and
4. Section 2.4 explams the combined fatal cancer risk.
The information used in these four sections was extracted from EPA documents including:
Drinking Water Criteria Document for Radon in Drinking Water (USEPA, 199 Ia), National
Primary Drinking Water Regulations; Radionuclides: Notice of Proposed Rulemaking (USEPA,
199 I e); Uncertainty Analysis of Risks Associated with Exposure to Radon in Drinking Water
(USEPA, 1993h); The Occurrence and Exposure Assessments for Radon, Radium-226, Radium-
228, Uranium, and Gross Alpha Particle Activity in Public Drinking Water Supplies (Revised
Occurrence Estimates Based on Comments to the Proposed Radionuclides Regulations) (USEPA,
1 992a).
-------
2-2
2.1 I’R(fl’OSED AIND REVISED UNiT RISK ESTIMATES FOR WAIERBORNE
RADON
Peop.e arc exposed to waterborne radon 1,1 thiee ways: from ingesting ri ‘ dissolved in
water; from inhaling radon gas released from water during household use; and from inhaling
radon progeny derived from radon released from water used. An indivi ual’s combined risk
during a lifetime from constant use of drinking water with one picocurie of radon per liter is
close to 7 chances in 10 million of contracting fatal cancer.
Exhibit 2- 1 summarizes the proposed and revised unit risk estimates of individual lifetime
fatal cancer risk for different pathways of exposure. The graph shows that although the
estimated risk for ingestion has increased, the overall risk has changed little from the risk
estimated in the proposed rule due to an offsetting decrease in inhalation risk. In the proposed
rule, the risk associated with ingestion accounted for only 23 percent of the lifetime risk, while it
accounts for 52 percent in the revised analysis. Risk associated with inhalation of radon progeny
decreased f om 74 ç ..t of the total risk in t.ie proposed rule to 45 percent in u c jevised
estimates, The remainder of the risk (3 percent) from waterborne radon comes iwn haling
radon gas. The estimates and methodologies used in the risk estimate revisions are described in
sections 2.2 and 2.3.
Exhibit 2-1
Estimated Individual Lifetime Fatal Cancer Risk by Exposure Pathway
(expected cases per 10 million (10 ) people exposed per PCi/Lwater)
8
LI Proposed Revised 1
=
E4
1.
‘I
U
0
Ingestion of Radon Inhalation of Radon Progeny
Inhalation of Radon Sum of All Pathways
-------
2-3
2.2 RISK ASSESSMENT OF INGESTING RADON IN WATER
After a person ingests radon in water, the radon passes from the gastrointestinal tract into
the blc’ud prin oally by way of the small intestine. The blood then circulates he 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. The
tissues of the stomach, intestines, liver, and lungs appear to receive the greatest doses.
Parameters Affecting Risk Estimates From Ingestion of Radon in Water
The human health risks from ingesting radon in water depend on the total quantity of
radon ingested and the risk factor for ingested radon The quantity of radon people ingest
depends on the volume of water they ingest and the initial concentration of ‘don in the water.
It also depends oii . traction of the radoi remaining in the water at the time .. estion. That
amount varies bcLause radon is a volatile gas; it begins to escape from water as soon as the
water is discharged from the tap. EPA’s estimates of the health risk associated with ingesting
radon in drinking water supplied by ground water have taken all of these factors into account.
Consequently, calculated estimates of the individual health risk from ingesting radon in water are
a product of the first four parameters listed below. To calculate the population risk, total
exposed population also needs to be taken into consideration.
(I) volume of water ingested that contains radon; and
(2) fraction of radon remaining in water at the time of ingestion; and
(3) cancer risk factor (cancer fatality risk per picocurie (pCi) of radon ingested); and
(4) concentration of radon in water; and
(5) total exposed population
Volume of Tap Water Ingested
The most complete survey of water ingestion patterns by people in the United States is the
Nationwide Food Consumption Survey conducted by the U.S. Department of Agriculture in 1977
and 1978. The three-day diary study sampled more than 30,000 people living in the continental
United States, who statistically represent the entire U.S. population living in households. Survey
questions specifically addressed the number of eight-ounce cups of water consumed by each
individual in the household each day. EPA used the findings from this short-term study to
estimate long-term drinking patterns of U.S. adults.
Based on this survey and other data, EPA estimates that U.S. residents ingest an average of
1.2 liters of tap water a day, which includes both “direct” and “indirect” tap water. Direct tap
water is ingested as soon as it is taken from the tap, so that most of the dissolved radon remains
in the water and is ingested. On the other hand, radon in indirect tap water (water used for
-------
REPORT TO THE UNITED STATES CONGESS
ON
RADIONUCLIDES IN DRINKING WATER
MULTIMEDIA RISK
AND
COST ASSESSMENT OF
RADON [ N DRINKING WATER
Office of Water
U.S. Environmental Protection Agency
Washington, D.C.
JULY 15, 1993
-------
2-4
making coffee, etc.) will escape before it is ingested. Therefore, EPA is concerned only about
radon in direct tap water. Slightly more than half the total tap water ingested (an average of
0.65 liters per day) is direct tap water (Pennington, 1983; USEPA, 1984; Ershow and Cantor,
l9 9). Because different people ingest different amounts of water, there is considerable
variability in the amount of water ingested among individuals. The ge . etric mean of the direct
tap water ingestion rate per day is 0.526 liters per day, with a credible range ot 0.5 18 to 0.534
liters per day.
EPA has used a value of two liters per day for total tap water consumption per person in
regulating other drinking water contaminants. EPA’s analysis of existing data shows that 90
percent of all people in the United States consume no more than two liters of tap water daily.
Consequently, based on existing data and for consistency in regulating radon in water, EPA
considers a protective value of one liter per day for direct tap water intake to be a reasonable.
protective value It is 50 percent of EPA’s commonly assumed total tap water consumption rate
(USEPA, 199la).
• Fraction of Radon Remaining During Water Transfer From the Tap
As water runs from a faucet into a glass, some of the radon escapes. Based on a numbei
of studies, EPA assumed that 20 percent of the radon escapes from direct tap water, with an
estimated credible range between 10 and 30 percent. Thus, the fraction of radon remaining is
typically 80 percent, with an estimated credible range from 70 to 90 percent.
• Ingestion Dose and Risk
EPA estimated the unit risk factor (i.e., fatal cancer risk to a person from ingesting one
pCi of radon) based on organ-specific dose estimates (i.e., organ radiation dose in rad per pCi
ingested) and organ-specific risk per unit dose (or organ-specific risk coefficients).
Organ-Specific Dose (Rad per Picocurie)
Several studies in humans have measured the amount of radon exhaled and the rate at
which it is exhaled after a person drank water containing high levels of radon. From these
studies the overall retention of ingested radon in the body can be determined, but not the
distribution among organs over time. EPA’s best estimate of radon levels in different organs
and how they change with time is derived from models that use data extrapolated from a study
of human ingestion of xenon. The tissues receiving the highest dose are the stomach and the
intestine, but estimating the dose to these tissues is complicated by uncertainty over: I) the
possible concentration gradient in the lining of the gastrointestinal tract and 2) the possible
sweeping of radon’s short-lived progeny from the tissue by the blood prior to decay. In
addition, there ‘s i’’.cel-Ldinty regarding how much the dose to tissues depends on the age of the
person. Overall, the credible range between the upper and lower estimates is less than a factor
of six, depending on the organ.
-------
2-5
Exhibit 2-2 compares the radon doses for individual organs used by EPA in the proposed
rule to the revised estimates As the table shows, EPA has decreased dose estimates for the
lung, small intestine, and both parts of the colon. Other estimates of doses have not changed.
Exhibit 2-2
Estimated Dose from Radon Ingested by Cancer Site
(Rad per Picocurie)
Dose
Cancer Site Proposed Revised
Stomach 1.0 x 10-8 1 0 x
Intestine
Small imesune 5 2 x l0 3.1 x i0
Ascending colon 7.4 x 10 1.5 x j 9
Descending colon 4 1 x l0 8 2 x 1010
Liver 1.5 x l0 1.5 x 10
Lung 1.8 x 109* 1.3 x l0
(eneral tissue 6.7 x 1010 6.7 x
* Alveoli only
Organ-Specific Risk per Unit Dose (Risk per Rad)
Following publication of the proposed rule in 1991, EPA revised its methodology for
estimating organ-specific risk coefficients (risk per unit dose) from specified doses of ionizing
radiation, (USEPA, l992e, l992f, 1993b). Exhibit 2-3 lists EPA’s proposed and revised
estimates of organ-specific risk coefficients (risk per unit dose). These estimates are based
mainly on updated studies of cancer risk in atomic bomb survivors, and contain a number of
sources of uncertainty. Specifically, uncertainty is contributed by: I) sampling variation, 2)
age/time dependence of risk; 3) extrapolation of data from the Japanese population to the U.S.
population; 4) errors in dosimetry; and 5) uncertainty in the relative biological effectiveness of
alpha particles. These five factors were part of the uncertainty ana ysis calculations. Overall,
the credible range between the upper and lower estimates is less than a factor of 10 to 30,
depending on the organ
-------
2-6
Exhibit 2-3
Estimated Ingestion Risk by Cancer Site
(Risk per Rad)
Cancer Site
Draft Criteria
Document
Revised
Stomach
3.7 x i0
8.9 x
i0
Intestine
Small intestine
3 7 x l0
--
Ascending colon
7.3 x l0
--
Descending colon
7.3 x i0
--
Colon
--
22 x
j .3
. er
4.0 X l0
30 x
10
Lung
57 x i0 ’
1.7 x
l0
General tissue
1.6 x 1O
4.2 x
lO
Ingestion Risk Factor (Risk per Picocurie)
The risk that an individual could develop a fatal cancer in any organ by ingesting one pCi
of radon can be calculated by multiplying an organ-specific risk per rad for that organ by the
organ-specific dose of radiation per pCi of waterborne radon ingested. Because the cancer risk
for the total hndy is additive, based on the risk to the individual tissues, the total risk of
developing a tULU! cancer (nsk factor) can be calculated using the equation in the box below.
Where: d = Dose (rad) per pCi ingested for the target cancer site j.
r = Risk (cancer fatality) per Rad of radiation for the target cancer site j.
Based on updated scientific data, EPA recalculated the ingestion risk factor using revised
organ-specific risk coefficients (risk per unit dose) and additional modifications of intestinal and
lung dosimetry. The end result was that EPA increased its estimate of the total body fatal
cancer risk per pCi of radon ingested to 1.7 x 10w, or 17 chances out of a trillion, which is
about 2.3 times the risk estimated in the proposed rule. This incmase primarily reflects higher
estimates of the risk per unit dose for irradiation of the stomach and colon.
-------
2-7
Exhibit 2-4 lists EPA’s cancer risk estimates for six cancer sites, plus the total body
estimate, for every pCi ingested. The table compares organ-specific risks used in the proposed
rule to EPA’s revised estimates. As the table shows, EPA more than doubled the risk estimates
for the stomach. general tissues, and the total body. but decreased by one-quarter the calculated
risk estimaie fo wC liver
Exhibit 2-4
Estimated Risk from inFested Radon by Cancer Site
(risk x 10 2 per picocurie)
Cancer Site
Estimated
Risk Factor
Proposed
Revised
Stomach
3.7
8.9
Intestine
Small intestine
0 19
0
Colon
0 84*
2.6
Liver
06
045
Lung
1.0
2.2
(‘,eneral tissue
LI
2.8
Total body
74
17.0
*
Suinmanon of ascending and descending colon.
In revising the risk estimates associated with the colon, EPA initially considered
calculating cancer risk to the intestines by weighing the doses as follows: small intestines (20
percent), ascending colon (40 percent), and descending colon (40 percent). However, no
evidence has been found for radiogenic risk of small intestine cancer (NAS, 199U) In addition,
cancer statistics indicate that mortality from cancer of the small intestine is very low compared
to colon cancer (Nd, 1981). Therefore, EPA decided to calculate the risk of cancer in the
intestine based simply on an average dose to the colon.
In revising risk estimates associated with the lung, EPA noted that most human lung
cancers occur in airways of the lung, not in the alveoli. Therefore, EPA’s revised estimates of
cancer risk due to exposure to waterborne radon consider the dose to both airways and alveoli
rather than just to the alveoli.
Uncertainty of the Risk Factor
The nsk factor for ingested radon (the summation of the total fatal cancer risk due to
ingestion of I pCi of radon). is estimated to be 1.7 x 10h1. Monte Carlo simulation was used to
estimate the overall uncertainty in this factor (risk per pCi ingested), taking into account each of
-------
2-8
the sources of uncertainty in dose (rad per pCi) and risk (risk per rad) discussed above. Overall,
the credible range between the upper and lower estimates of the risk factor is less than a factor
of 17.
• Concentration of radon in drinking water
The concentration of radon in drinking water is very low in surface water supplies. In
contrast, radon activity in ground water supplies is highly variable. It is typically highest in
areas where granite is near the surface of the ground. Based on EPA’s analysis of existing data
for all drinking water containing radon, the population-weighted average of iddon activity is 246
PC1ILW Iter, with a credible range of 205-306 PCIlLwqter. Exhibit 2-5 demonstrates how the
population-weighted average radon activity was calculated.
Exhibit 2-5
Summary Characteristics for Community and Non-Transient, Non-Community
(NTNC) Ground Water Supply Systems and for Radon Ground Water Concentrations
System Size
Number of Systems’
Population Served 1
(in thousands)
Average Radon
Concentration 2
pCi/L
Credible
Range of
Mean
Concentration
Community
NTNC
Community NTNC
Very, Very
Small
25-100 people
16,634
13,842
956 . 625
844
645-1090
Very Small
101-500 per’nle
15,422
7,512
3,931 1,950
684
522-876
Small
501-3,300 people
9,952
2,4-44
-
13,884 2,346
284
208-402
Medium
3,301-10,000
people
2,302
63
13,599 366
204
147-271
Large/Very
Large
>10,001 people
1,316
4
48,711 76
205
137-295
Totals
45,626
23,865
81,081 5,363
246
205-306
September 30, 1992 Wade-Miller addendum prepared for USEPA Data based on Federal Data Reporting System
(FRDS)
2 Source National li organics and Radionuclides Survey (Cothern, Rebers, 1990)
Population-weighted average radon concentration for afl systems
‘3D
-------
2-9
Total Exposed Population
EPA estimates that 81 million people, or about one-third of the population of the United
States in 1990, are served by public community water supply systems using ground water
(USEPA, -)931 i)
Estimates of Risk and Overall Uncertainty
Based on the information on risk and exposure parameters described above, EPA estimated
the mean individual risk and population risk due to ingested radon. The overall uncertaintyof
these estimates, expressed as the upper and lower estimates of the credible range, were derived
by Integrating the uncertainty of each risk and exposure parameter by Monte Carlo simulation
Individual Risk
To estimate an -d vidual’s fatal cancer risk for a given radon exposure, EPA multiplies
the estimate of the risk per unit exposure with the individual’s exposure. Exhibit 2-6 presents
the mean individual risk estimates in fatal cancer cases per person per year for the population
served by community ground water supplies.
Exhibit 2-6
EPA’s Estimates of Mean Individual Risk from Radon Ingestion
(fatal cancer cases per person per year)
Low
Best Estimate
High
2.4 x l0
1.2 x 106
2.1 x 10-6
Population Risk Estimates
EPA estimated the total number of deaths for the fraction of the U.S. population exposed
to iadon through community ground water supplies. The population risk (the probable total
number of fatal cancer cases per year) is calculated by summing all the individual risks in the
population of interest, taking into account the distribution of exposure levels as determined from
measured levels of radon in water or air. EPA’s best estimate of the number of fatal cancer
cases per year resulting from the ingestion of radon from drinking water is 100, given that 81
million people are exposed to radon annually. The total number of fatal cancer cases from
ingestion could range from 19 to 166 deaths per year. These values exclude exposure of people
to waterborne radon from private wells and NTNC water supplies.
‘ :31
-------
2-10
2.3 RISK ASSESSMENT OF INHALING RADON FROM PUBLIC WATER SUPPLIES
Public water supolies from ground water can be a significant controllable household source
of airborne radon is decay products (i.e, radon progeny). This section explaijis how EPA
estimated the risk associated with inhalation of waterborne radon. It gives an overview of the
inhalation risk of radon derived from drinking water, then presents the parameters that affect the
risk estimates for inhalation of radon progeny and radon, respectively. This section also
summanzes EPA’s combined inhalation risk estimates due to radon and progeny from drinking
watci- The infoi-mation was drawn from the Drinking Water Criteria Document for Radon in
Drinking Water (US EPA, 1991 a), the Uncertainty Analysis of Risks Associated With Exposure
to Radon in Drinking Water (US EPA, 1993h), and Addendum to: The Occurrence and
Exposure Assessments for Radon, Radium-226, Radiunz-228, Uranium, and Gross Alpha Particle
Activity in Public Drinking Water Supplies (Revi.sed Occurrence Estinwtes Based Ofl Co nments
to the Proposed Radionuclides Regulations) (US EPA, 1 992a).
EPA’s inha1ati’ k assessment covers i e risks to human health fiom bo :. . uun gas
and radon progeny. Radon progeny (or decay products) pose far greater risks th 1 r: o i gas
itself. Therefore, EPA has given them the greatest attention regardless of whether the radon is
from water or soil, or whether it is in outdoor or indoor air. The analysis for outdoor radon and
the analysis for residential indoor radon focus only on the risks from radon progeny.
Inhalation Risks of Radon Progeny from Drinking Water
When EPA developed the drinking water standards proposed in July 1991, it analyzed the
inhalation risks from radon gas and from radon progeny in public water supplies using ground
water. In that analysis EPA estimated individual risks as well as the total risk for the 81 million
people who use publicly supplied ground water for drinking.
Since the proposal, EPA has revised its analysis of the health effects caused by inhaled
radon progeny in two areas. First, the risk factor was changed to reflect newly available
scientific information. This factor derives from information on the effects of radon on
underground miners. In order to apply it to residential radon exposure, it is necessary to
understand the relative effects of radon in mines and in homes. A [ 991 National Academy of
Sciences report on radon dosimetry in mines and homes provided new information on this issue,
which EPA then used to revise the risk factor for residential radon exposure. Second, EPA used
only the BEI1R IV model for calculating risk; in the proposal EPA used an average of the BEIR
[ V and ICRP models. In addition, EPA expanded the analysis quantitatively to cover the
uncertainty that exists in many of the key parameters.
In analyzing the inhalation risks from radon and radon progeny, EPA focused on the unit
risk (i.e., the inhalation risk per pCi per liter of water (PCi!Lwater)), individual risk (risk per
unit exposure multiplied by the exposure), and the population risk (number of cancer cases per
year to all exposed households using ground water for drinking containing varying levels of
radon). EPA prepared its “best estimate” of each of these risks and quantitatively estimated the
-------
2-Il
credible range of values that could exist given the uncertainty in each of the key parameters in
the analysis.
ParameL. Influencing the Risk from Inhaling Radon Progeny froni Drinking Water
Assessing the risks from inhaling radon progeny requires information on how much of the
radon released through household water use enters the air and is converted into progeny that
individuals inhale. Given the amount of radon progeny individuals inhale, EPA uses a dose-
response factor that estimates the relationship between the radon dose received and the health
effects that result. EPA calculated radon risk as a product of the following six parameters:
(I) the concentration of radon in drinking water;
(2) a transfer facto,-, which is the relationship between the radon concentration in indooi
air denved from water and the initial concentration of radon in watei;
( ) the eq’ t ;un factor, which is the fraction of the potential enei y (‘ Iou progeny
that ‘::ually exists in indoor air compared to the maximum possible energy under
true equilibrium;
(4) the occupancy factor, which is the fraction of time individuals spend in their homes,
exposed to indoor radon;
(5) a risk factor, which estimates the risk of lung cancer from exposure to a given
amount of radon; and
(6) the total exposed population, which is the number of people exposed to the airborne
radon progeny resulting from household use of water.
The first four factors determine the amount of exposure to radon progeny that occurs. The
risk factor describes the exposure response relationship between lung cancer deaths and
exposure. This factor enables EPA to estimate the risk that can result from a given level of
exposure. EFA has invested considerable cffort gaining knowledge about eaLli of these six
factors, which is summarized below.
• Concentration of Radon in Drinking Water
The concentration of radon in drinking water from surface water is very low as compared
to radon activity in ground water supplies, which is highly variable. Radon concentrations
are typically highest in areas where granite is near the surface of the ground. Based on
EPA’s analysis of existing data for all drinking water containing radon, the population-
weighted average radon activity is 246 P ’Lwater , with a credible range of 205-306
PCI/Lwater Details of the occurrence data of radon in drinking water are in Exhibit 2-5 on
page 2-8.
-------
2-12
Transfer Factor
The concentration of radon released from water to indoor air is highly variable. Levels
vary from room to room and vary over time, depending on household water use patterns,
room sizes, and ventilation rates. N i tierous studies have investig. . d the cross-media
transfer of radon from water to air. The results from a series of studies expressed the
transfer factor as the ratio of the amount of radon transferred to the air given the amount
of radon initially contained in the drinking water. From its review of these studies, EPA
decided that the best estimate of the transfer factor was : 10,000. EPA believes that the
uncertainty around this estimate gives a credible range from 0.7:10,000 to I 9:10,000
• Equilibrium Factor
The equilibrium factor has been measured in a number of homes. A value of about 0.5 is
believed to be representative of U.S. homes, with a credible range of 0.35 to 0 55.
• Occupancy Factor
The occupancy factor, or fraction of time spent in the home, varies with people’s lifestyles.
EPA selected 75 percent as representative of the amount of time individuals spend inside
their homes. EPA based its selection on its review of data from eight studies conducted
between 1978 and 1990 (USEPA, 1992i). Estimates of the occupancy factor in these
studies ranged from 60 to 80 percent.
• Inhalation Risk Factor
Because radon decay products pose far greater nsks than radon itself EPA focused on
on progeny in estimating the inhalation risks of radon. In deve1opiii t risk factor for
radon progeny, EPA modified a risk projection model developed by the National Academy
of Science’s (NAS) Biological Effects of Ionizing Radiation (BEIR IV) Committee in
1988. The BEIR IV model used information from four major epidemiological studies of
underground miners exposed to radon. EPA adjusted the results of the BEIR IV model to
account for differences in physical and biological factors between mines and homes. For
example, breathing rates of physically active miners would be higher than those of
sedentary people at home. The NAS’s Gomparative Dosimetry of Radon in Mines and
Homes (1991) indicated that residents of homes are likely to receive a lower dose of
radiation to their lungs than miners when both are exposed to the same environmental level
of radon. Using information from this study, EPA reduced the estimated risk factor to
account for this difference (USEPA, 1992i).
-------
2- I ’
Total Exposed Population
EPA estimates that 81 million people who are served by public water supply systems using
ground water are covered by its risk assessments (USEPA, 1993h). This number is about
(L’ie-th” of the “opulation of the Li tates in 1990
The information from the NAS study was the principal reason EPA adjusted the risk factor
for radon following the proposed rule (EPA’s preamble to the proposal had noted that the risk
factor could change as a result of the NAS study.) Another reason for the change was that EPA
accepted an SAB recommendation to use a risk factor based solely on EPA’s version of the
BE1R IV model. EPA had previously considered the results of an ICRP i k model as well.
The recommendation to use only the BEIR IV model was based on a reassessment of the
available information, including evidence on the relationship between radon risk and (1) the time
since the exposure occurred, and (2) the age at which the exposure occuiied. The BEER IV
model was more consistent with this evidence than was the ICRP model
These changes 1 a revised radon risk factor of 224 lung cancer deaths i Ds) per
million persons exposed to a working level month (WLM’) of radon, a common unit for
measuring radon exposure. For the proposed rul , EPA used a risk factor of 360 LCDs per
million people exposed to one WLM. This change represents a reduction of nearly 40 percent in
the risk factor. EPA estimates that the uncertainty in the risk factor ranges from 140 LCDs per
million people to 570 LCDs per million people (USEPA, 1992i).
Estimates and Overall Uncertainty of Risks Associated with Inhalation Exposure to Radon
Progeny
EPA estimated the mean individual and population lung cancer death risks based on the
exposure and risk parameters described above. The overall uncertainty of these estimates was
derived by integrating the uncertainties of all the individual parameters used.
Special Analysis of Transfer Factor Approach and Uncertainty Due to Peak Exposure
For the transfer factor, EPA hot only looked at variations in that factor, but also considered
whether a more sophisticated modelling approach was warranted. EPA historically has used a
simple approach to assess the transfer of radon from water to air. The simple approach
iepresents a house as a single comparmient with uniform radon concentration, and assumes that
water use is constant over time. Exposure is calculated by determining the concentration of
radon in water, using the transfer factor to estimate the concentration of radon in indoor air, and
estimating occupancy.
WLM is defined as a working level month, which is a standard unit of measure of exposure to radon decay products
-------
2-14
In reality, water use is episodic, and walls divide houses into rooms, restricting the mixing
of radon. The result is higher levels in some rooms (e.g. bathrooms with showers) and the
possibility of episodic peak exposures to radon. Therefore, as an alternative, EPA evaluated the
use of a multi-compartment model to estimate average exposure. The ‘house” was partitioned
into a iu- . a bathroom, and the remainder c i the house.
The three-compartment model was used to predict radon concentration in each compart-
ment as a function of time, using estimates of input variables including.
• the volume of each compartment;
• the residence time of air in each compartment;
• the volume of water used and when it is used; and
• the fraction of radon released from water to air with each use. -
Exposure is cahiilated from the resulting estimates of radon concentration in each
compartment and esuiuates of the amount of time a person spends in each compartment
throughout the day.
EPA compared the results of the simple and multi-compartment models. The concentration
predicted by the simple model is lower than that predicted by the multi-compartment model
during peak exposure, but it is higher during the rest of the day. In other words, the simple
model underestimates exposure during showers, but overestimates exposure during the rest of the
day. Overall, EPA found that mean radon exposures predicted by the multi-compartment model
were only 1.3 times higher than the exposure level estimated by the simpler approach. However,
the cancer risk from peak exposures to radon is reduced by the time lag that exists due to the
buildup of radon and progeny. For example, in a typical shower scenario, the level of progeny
achieves only 2 to 4 percent of its maximum possible value. Taking this into account, it is
likely that the transfer factor approach may slightly overestimate exposure to progeny compared
to the multi inpartment model. EPA also conducted a sensitivity analysis cf the influence of
each input variable on the radon and progeny concentration and concluded that the use of the
more realistic multi-compartment house model would not significantly affect estimates of radon
exposure (USEPA, l993h).
Individual Risk
To estimate an individual’s risk of contracting a fatal cancer from a given radon exposure,
EPA multiplies the estimate of the risk per unit exposure by the individual’s exposure.
Exhibit 2-7 prese ts the EPA’s best estimates of the mean individual risk with a credible range.
EPA assumed that one lifetime is equal to 70 years.
-------
2-15
Exhibit 2-7
EPA’s Estimates of Mean Individual Risk from Radon Inhalation at Occurrence Levels
(lung cancer deaths per person per year)
Low Best Estimate High
Individual Risk 5.9 x lO 1.1 x 10.6 2.9 x 10-6
Populaton Risk
The population risk is calculated by summing all the individual risks in the population of
interest, taking into account the distribution of exposure levels as determined from measured
levels of radon in water or air. To derive annual population risk, EPA multiplied annual mean
individual risk by the number of people exposed. The population served by community ground
watei--based drinking water systems is estimated to be about 81 million, and the total risk for this
population is estimated to be 86 LCDs per year EPA estimates a credib e range due to
uncertainty in this estimate of 48 to 233 LCDs per year.
The Risks from Inhaling Radon Gas
The risk factor for fatal cancer per pCi of radon in inhaled air is estimated to be 4.7 x 10
13 (USEPA, 1989c). This cancer fatality unit risk for inhalation of I pCi of radon ga.s was
modeled by the RADRISK program using organ-specific radiation doses (rad/pCi) and cancer
risk coefficients (risk/rad). Organ-specific doses were estimated by absorption rate, distribution,
metabolism, and excretion of radon and its progeny (Dunning et al., 1980; Sullivan et al., 1981).
Organ-specific risk coefficients were derived by quantitative evaluation of epidemiological data
on human cancer risk following exposure to several types of ionizing radiation.
EPA also used an average transfer factor of 1:10,000 to evaluate inhalation exposure to
radon gas released from household use of water. With this transfer factor, 1 pCiJL of radon in
water gives rise to 1.0 x i0 pCIJL of radon in air. Assuming 75 percent occupancy and a
breathing rate of 22,000 L/day (USEPA, 1989b), an individual inhales a total of &2 x 108 liters
of air in a 70-year lifetime. Thus, the amount of radon gas inhaled over a 70-year lifetime from
I pCi/L of radon in water is 4.2 x l0 pCi. Using the risk factor of 4.7 x l0 for fatal cancers
per pCi of radon gas inhaled, the risk per pCi/L in water is 2.0 x 10-8 for total fatal cancers.
The estimated fatal cancer risk for an individual exposed to radon in water at the population-
weighted average concentration of 246 pCi’L is 7.0 x10 8 fatal cancer cases per person per year.
The estimated annual cancer deaths due to inhalation in the population exposed to radon in water
(8 I million) at an average concentration of 246 pCiJL is 6 deaths. This risk from inhalation of
radon gas contnbutes a small percentage (approximately 3 percent) of the total waterborne radon
risk. For this reason, EPA did not analyze the uncertainty in its estimates.
‘zl
-------
2-16
Combined Inhalation Risks from Radon Progeny and Radon Gas
Based on it,s revised risk estimates as summarized above, EPA estimated th number of
inhalation a’ii. wf the entire population expk d to radon through comn1 unit)’ Water supplies
on a long-term basis. Given that public water suppliers expose an estimated 8 I million people
each year to inc reased risk of cancer from th inhalation of radon and radon progeny, EPA
estimates that a total of 92 fatal cancer cases per year will result from the inhalation of radon
and its progeny by people relying on public water supplied by community ground water systems.
However, uncertainty exists in estimating the inhalation risks. Consequently, the number of fatal
cancer cases from inhalation could range from 54 to 239 deaths per year. Tiiese values exclude
exposure of people to waterbome radon from private wells and NTNC ground water supplies.
2.4 COMBINED FATAL CANCER RISK
Combined Individual Lifetime and Population Risks
Based on the revised risk data summarized above, EPA’s best estimate of the total number
of deaths from radon in water is 192 per year. However, there is uncertainty in estimating the
risk involved from ingestion and inhalation of radon progeny. The best estimates of fatal cancer
risks from ingesting radon in water and inhalation of waterborne progeny with their credible
ranges are presented in Exhibit 2-8.
Exhibit 2-8 compares EPA’s current individual and population estimates of fatal cancer
risk. The table shows both the lower and upper credible bounds for inhalation and ingestion
risks.
Exhibit 2.8
Comparison of EPA’s Revised Individual and Population Risk Estimates
Individual Risk
Population Risk
Inhalation Ingestion
Inhalation Ingestion
Statistic
(fatal cancer cases per person per year)
(fatal cancer cases per year)
Lower Credible Bound
5.9 x iO 24 x lO
48 19
Best Estimate
1.1 x 10 1.2 x 10.6
S6 100
Upper Credible Bound
2.9 x 106 2.1 x l06
233 166
-------
3. RISK ASSESSMENT OF EXPOSURE TO RADON IN AIR
— ‘c--
The four sections outlined below present and explain how EPA estimated the risk due to
radon in :
I. Section 3.1 discusses the inhalation risks of indoor radon;
2. Section 3.2 discusses the inhalation risks of outdoor radon; -
3. Section 3 3 reviews EPA’s 1992 analysis of the inhalation risks of radon progeny in
residential air, which it onginally presented in Technical Support Document for the
Citizen’s Guide to Radon (USEPA, 1992i);
4. Section 3.4 pi . nts the inhalation risks that result from radon emissions at water
treatment wits that remove radon from ground water in order to comply with EPA’s
proposed drinking water standard for this contaminant; and
5. Section 3.5 summanzes risks of inhalation of radon and radon progeny from all
sources.
Radon decay products pose far greater risks than radon gas itself. Therefore, EPA has given
them the greatest attention in its analysis of the inhalation risks of radon. The analysis for
outdoor radon and residential iadon focus on the risks from radon progeny only. The results of
those analyses help place the inhalation risks from radon in drinking water in perspective. EPA
recently completed an analysis that evaluates the uncertainty of key parameters in the risk
analysis for radon progeny.
3.1 RADON IN INDOOR AIR
Radon released from drinking water is only one source of airborne radon progeny in
homes. Radon may also be present in soil, and travel indoors through cracks in foundations. To
a lesser degree, it may be released from building materials. Radon progeny from all sources
contribute to the inhalation risk of radon in homes. In a separate analysis, EPA assessed the
total residential risk from inhalation of radon progeny.
EPA used the same risk factor discussed in Chapter 2 for airborne radon progeny derived
from drinking water: 224 lung cancer deaths per million persons exposed to one WLM of
radon. EPA used the results of its Na onal Residential Radon Survey (a survey measuring
radon concentrations in 6,000 homes statistically representative of all U.S. residences) to provide
EPA’s best estimate of the annual average radon concentration in U.S. homes of 1.25
-------
EPA used the same estimates for the occupancy factor and
picocuries 1 per liter (pCi/L. 1 ).
equilil)riunl factor described in Chapter 2.
Exhibt 3 I suii& iarizes EPA ’s best estim ie of the results. EPA calculateJ both the
average individual risk and the population risk per year. For the entire U.S. population of 250
million, the total annual deaths from inhalation of radon progeny in indoor air were estimated to
he l .ñ()() with a range of 6.74() to 3(),(0ft These risk estimates cover exposure to indoor radon
from all sources, including release from water. The risk estimates shown in Chapter 2 for the
inhalation of radon progeny from drinking water are a subset of the total risks presented here.
Best Estimate of the Annual Risk
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
Inhalation of Indoor Radon
EPA also conducted an uncertainty analysis in which it considered the range of possible
values for the key input parameters (i.e., risk factor, the average radon concentration, oocupancy
rate, and equilibrium factor). For the risk factor, as described above, the analysis considered
uncertainty resulting from statistical variability in the epidemiological data used, from three
1 A iurie (Ci) is a standard measure of radioactivity, and a picocune (pCi) is one trillionth (1 x 1012) of a curie.
Exhibit 3-1
From Residential
Progeny
1
Low Best Estimate High
I
L . o
-------
3-3
sources: (I) predicting the effects of radon over long time periods; (2) using data for adult
miners to describe the effect on other age groups; and (3) differences between mines and homes
that may influence the effect of radon on health. EPA’s uncertainty analysis produced a range
of estimates of total population risk of 6,740 to 30,600 lung cancer deaths per yep! (USEPA,
1992i)
3.2 RADON iN OUTDOOR AIR
For purposes of comparison, EPA estimated the expected risk from inhalation of radon
piogeny in outdoor air (USEPA, l993h). The variables used in this calculation are similar to
those used to calculate the risk from inhalation of radon progeny released from drinking water.
For the concentration of radon in outdoor air in the U.S., EPA made use of its results of its
outdoor radon survey and the information in The 1988 UNSCEAR Report (UNSCEAR, 1988;
Gesell, 1983) to derive an estimate of 0.3 pCi/L.
Ii. outdoor a , ..idOn decay products re nearly in equilibrium with rauon a . .., .. nd the
equilibrium facto 1 is higher than it is indoois. EPA used a best estimate of 0.8 for this factor.
Using data on US population activity patterns, EPA estimated for the average time spent
outdoors (“occupancy factor’) is 7.5 percent. Finally, EPA used the same risk factor used above,
224 lung cancer deaths per million person per WLM The true risk factor may not be the same
outdoors as indoors; among other things, breathing rates may differ between indoor and outdoor
activities, which would affect the radiation dose that reaches lung tissues. However, the
difference between indoor and outdoor risk factors is not expected to be large, and adequate data
are not available to measure it.
Based on these values and a U.S. population estimate of 250 million, EPA’s best estimate
of the nsk to the population is 520 lung cancer deaths per year from outdoor radon exposure.
Based on estimates of the uncertainty distributions for the Input variables, EPA estimated a
credible range of 280 to 1,500 lung cancer deaths per year (USEPA, l993h).
3.3 EMISSIONS FROM PLANTS TREATING DRINKING WATER TO CONTROL
RADON
EPA identified aeration as the best available technology (BAT) for removing radon from
public water systems. While EPA also recognizes the capability of granular activated carbon
(GAC) to treat radon, GAC is not BAT because of the long contact time required for radon
removal, which makes it less efficient and more costly than aeration treatment, and an infeasible
technology for large municipal systems.
This section describes how EPA assessed the risks associated with emissions of radon
progeny from packed tewer aeration (PTA) treatment. It first examines the analyses performed
in support of the proposed rule, and then discusses the revised estimates.
-------
3-4
Proposed Estimate of Risk From Treatment Plant Emissions
EPA considered the potential risk associated with radon air emissions from water treatment
plants I i-ior to proposing the radionuchdes gulations. EPA based the: . a s ssment on two
prior studies: Preliinina,y Risk Assessment for Radon Emissions from Drinking Water Treatment
Facilities (USEPA, 1988) and Analysis of Potential Radon Emissions from Water Treatment
Plants Using the MINEDOSE Code (USEPA, 1989). Both studies concluded that the risks from
potential human exposure to PTA emissions appear very small in comparison to risks associated
with radon in drinking water.
The 1988 risk assessment of treatment plant emissions used the EPA Nationwide Radon
Survey to select 20 dnnktng water systems that contained the highest levels of radon in drinking
water and affected the largest populations. The study included parameters for:
• air dispersion of radioactive emissions, and
• estimates of fatal cancers to exposed persons within a 50 kilometer r dJus
of 20 water treatment facilities.
EPA estimated the individual risk and number of annual deaths based on urban, suburban, and
rural exposure settings for 20 treatment facilities. This approach is similar to the approach EPA
used in assessing risks associated with dispersion of coal and oil combustion products.
Using dispersion and exposure assumptions that were considered realistic, EPA’s best
estimate of the lifetime cancer risk to individuals from treatment plant emissions from the 20
treatment facilities considered was four cancer deaths per 100,000 people and the estimated
incidence of cancer was 0.016 deaths per year. EPA estimated that the total U.S. deaths per year
due to air emissions at all drinking water supplies requiring treatment for radon, given a
hypothetical MCL of 200 pCiiL, would be 0.1 deaths per year.
In 1989, EPA augmented the 1988 study with a “worst case” study of ldUOfl emissions
from four actual water treatment facilities to determine if such a worst ca.e facility might fail to
comply with a National Emissions Standard for Hazardous Air Pollutants (NESHAPs) limit of 10
millirem per year. EPA selected four facilities that had radon levels from 1,330 to 110.000
pCi/L. EPA concluded that all but possibly the very largest systems would be in compliance.
Because of EPA’s very conservative assumptions, however, it appears possible that the very
largest systems, those with very high flow rates, may pose a slight potential for appreciably
increasing ambient air radon exposure. EPA concluded that the risk to the population
nationwide attributable to radon released from source water during treatment for removal of
radon is two to four orders of magnitude smaller than the risks estimated for untreated radon-
containing drinking water.
Ljl
-------
3-s
Revised Estimate of Risk From Treatment Plant Emissions
Commenters asserted that EPA failed to cnrisider sufficiently in the proposed rule the risk
tradeofi of ii easin ir emissions from the u . of PTA. In response, early 1993, EPA’s
Office of Ground Water and Drinking Water requested EPA’s Office of Radiation and Indoor
Air (ORIA) to review their earlier analyses to ensure consistency, incorporate a radon risk
coefficient based on Technical Support Document for the 1992 Citizens Guide to Radon (the
Technical Support Document’), and provide a simple quantitative uncertainty analysis of the
individual risk estimates.
ORIA concluded that the 1989 study genetally corroborated the 1988 study and that the
assumptions and findings, while different, were complementary. ORIA noted that the risk
coefficient used in the 1988 study (460 lung cancer deaths per million person-WLM) was
approximately two times higher than the more recent risk coefficient proposed by the BEER IV
Committee (224 lung cancer deaths per million oerson-WLM), as reported in the Technical
Suppot Document. recalculating based oii the revised risk coefficient, ORIA iea hed two
conclusions: that th Thaximum individual risk was two cancer deaths per 100,000 people
instead of four cancer deaths per 100,000 people, as reported in the 1988 study; and that the
incidence (deaths per year) was 0.004 instead of 0.0 16 for the 20 water systems treating to
remove elevated levels of radon in water. Exhibit 3-2 summarizes ORIA’s findings.
Exhibit 3-2
EPA’s Proposed and Revised Risks from
Treatment Plant Emissions for 20 Sites Combined
Risk Measure
Proposed
Revised
Risk Coefficient
(deaths per million persori-WLM)
460
224
Best Estimate of Individual Risk
(deaths per iOO,000 people)
4
2
Incidence
(deaths per year)
0.016
0.004
ORIA’s uncertainty analysis indicated that individual risks are likely to be overstated by
the generic assessmcnts used in the 1988 and 1989 studies, particularly for the latter. This
overstatement, according to ORIA, is neither surprising nor inappropriate for a screening model,
the purpose of which is to provide a reasonable estimate of risk that will not be exceeded by a
more detailed analysis.
-------
3-6
3.3 SUMMARY OF INHALATION RISK RESULTS
EPA’s revised analysis of the risks of inhaling radon released from drinV ng water has
produL J .r overall estimates of risk from i a1ation of radon progeny tlhtnthe estimates in
the proposed rule. In addition, the analysis provides new information on the uncertainty in the
risk estimate. Exhibit 3-3 summarizes the annual population risk estimates from inhalation
exposure to radon and radon progeny in air. It also presents the results of EPA’s uncertainty
analysis for these risk estimates. Lung cancer deaths from exposure to radon progeny in indoor
air include those caused by waterborne radon and progeny in air Without treatment, inhalation
of radon progeny from drinking water is expected to result in 86 lung cancer deaths per year,
while inhalation of waterborne radon gas is expected to cause six deaths. EPA’s uncertainty
analysis shows a range from 48 LCD to 233 LCD per year from inhalation of progeny from
drinking water.
Exhibit 3-3
EPA’s Estimates of Population Lung Cancer Fatalities
by Source of Radon in Aira
(deaths per year)
Best
Source of Radon in Air
Low
Estimateb
High
All Indoor Progeny
7,000
13,600
30,000
Outdoor Progeny
280
520
1,500
Treatment Plant Emissions
--
0.04
--
Total for Radon in Air
7,280
14,120
31,500
a
For all indoor and outdoor progeny, values apply to the clitire U S population of
250 million
b Value calculated using best estimates of parameters and risk equation
L-t t -k
-------
PART THREE
COST
L1’ .
-------
4. COST ESTIMATES FOR CONTROLLING RADON
Many factors influence the costs of treating radon in public water systems that rely on
ground water and others influence the costs o mitigating radon in indoor air. EPA developed
the cost estimates presented in the 1991 proposed rule for drinking water based on cost
modelling th it incorporated available data and re ;earch on radon occurrence in drinking water
and the unit costs of treatment technologies. In response to public comments. EPA has revised
the radon occurrence and unit cost estimates. The first five sections of this chapter present and
explain these costs for treating radon in public water systems relying on ground water, based on
data presented in the Regulatory Impact Analysis of Proposed National Primary Dunking Water
Regulations for Radionuclides (USEPA, 1991j) and revised data presented in Working Diafi of
Regulatoty hnpact Analysis for Final National Puima,y Drinking Water Regulations for
Radionuclides (USEPA, to be published). The last two sections present the costs for controlling
radon in air, originally pi . , ited in the Technicw’ Support Document for the Citizen’s Guide to
Radon (USEPA, 19921), alid the combined costs for controlling residential radon jiufli all
sources. The seven sections in this chapter are:
I. Section 4. 1 presents the proposed and revised estimates of the number of
water systems and water treatment sites affected, as well as qualitative
descriptions of the unit cost factors that are the basis for the national cost
estimates for treating radon in public water systems relying on ground
water;
2. Section 4.2 describes qualitatively the key factors that account for
variations in costs per household for treating radon in public water
s ‘tems relying on ground water, with quantitative data on these factors
presented in graphs and other exhibits;
3. Section 4.3 provides a quantitative description of the cost estimates used
in the proposed rule for public water systems relying on ground water;
4. Section 4.4 describes the basis for the cost estimate revisions for public
water systems relying on ground water, made in response to public
comments;
5. Section 4.5 details the revised national and household cost estimates for
treating radon in public water systems relying on ground water;
6. Section 4.6 summarizes national cost estimates for controlling radon in
indoor air; and
7. Section 4.7 presents the combined cost estimates for controlling radon in
indoor air and in public water systems relying on ground water.
-------
4-2
4.1 COSTS OF CONTROLLING RADON Thl PUBLIC WATER SYSTEMS RELYING
ON GROUND WATER
Three pnm ry tctors determine the nauon.il cost of treating radon in public water systems
relying on ground water:
I. Number of sites (wells or well groups) requiring treatment;
2. Extent of treatment required at each site to achieve the Maximum Contaminant
Levels (MCLs); and
3. Unit costs associated with treatment requirements at each site.
EPA based cost estimates presented in the 1991 proposed rule on a national occurrence
estimate of 25,907 affected drinking water systems at an MCL of 300 picocuries 1 per liter of
water (PCi/Lwater). After f ther analysis, descriued below, EPA raised its estimate to 27,294
affected systems. Th. :e .ision resulted from a detailed analysis of the most recent available
data, including:
• An updated Federal Reporting Data System (FRDS) inventory of public water
systems (PWSs), including non-transient, non-community water systems;
• Re-stratified National Enorganics and Radionuclides Survey (NIRS) data that
includes separate distribution parametei s for each of five size categories, and
disaggregation of the smallest group into two separate groups; and
• “Radon decay sensitivity analysis.” Radon decays between when water enters
the public water system and when it reaches the tap, where samples are taken.
As a result, MRS samples underestimate, to a limited extent, the frequency of
occurrence of radon at the entry point (i.e., point of compliance).
EPA also considered and found merit in public comments on the proposed rule contending
that the number of systems affected understates the total number of sites affected, because a
single water system may have several wells or well groups that require separate treatment
facilities. Therefore, EPA now estimates that at an MCL of 300 PCi/Lwaier, 27,294 systems
(including community systems and non-transient non-community systems) contain a total of
41, 136 treatment sites that will incur costs in treating radon in water. This revised estimate of
the number of affected sites is a 58 percent increase over the 1991 estimate presented in the
proposed rule. However, the estimated number of people affected remains unchanged, at
approximately 18 million. Exhibit 4-1 summarizes the 1991 estimates for radon occurrence and
the revised low, best, and high estimates for radon occurrence. In the exhibit the PWSs include
community and non-transient non-community systems.
curie (Ci) is a standard measure of radioactivity, and a picocune (pCi) is one triflionth of a cune
-------
4-3
Exhibit 4-1
Occurrence Estimates for Treating Radon in
PWSs Relying o i Ground Water
1991
Proposed Estimates
Revised Estimates
Low
Best
Estimate
High
Systems Affected
25,907
24,715
27,294
30,364
Number of Site.s
25,907
38,068
41,136
45,624
Population Affected
IS million
17 million
18 million
20 million
The cost of treat- at a particular site d pends both on the flow rate at thc . - itICflt Site
(i.e., the amount of water that flows through the treatment site each day) and the ,nc it ation
of radon in the water. Larger, more expensive treatment facilities are required for sites with
higher flow rates, although the higher total cost at larger sites is offset by higher flow rates,
resulting in a lower cost per gallon due to economies of scale in treatment. Sites with high
radon concentrations will also incur higher treatment costs because more expensive treatment
facilities are needed for greater radon removal. For example, a site with lower radon
concentrations may need to remove only 80 percent of the radon in the watem to achieve the
MCL, while a site with high concentrations may need to remove 99 percent. A treatment facility
with 99 percent removal efficiency is more expensive than a facility with a 80 percent removal
efficiency. The difference in cost of the two removal efficiency treatments for small and large
systems ranges from 20 percent to 50 percent respectively. The difference is due to the height
of the tower required for the removal efficiencies.
EPA considered the costs of two treatment methods in developing the proposed rule:
Granular Activated Carbon (GAC) and Packed Tower Aeration (PTA). Although GAC has been
used to remove a variety of water contaminants and is capable of removing radon from drinking
water, EPA found this technology to be less effective and more expensive than PTA. Capital
and operating costs for removing radon with GAC are high because radon requires a lo ger
treatment contact time compared to other contaminants. Waste disposal costs may be very high
because the GAC process may produce a waste containing residual radioactivity. Therefore,
EPA determined that aeration was the most economical treatment option for all system sizes and
derived final regulatory cost estimates assuming that PTA would be used in all affected systems.
EPA recognizes that other aeration technologies, such as diffused bubble aeration and spray
aeration, may also be applied. Costs for these technologies are comparable to PTA.
The costs of PTA treatment also vary with the radon concentration at each site and the
corresponding radon removal efficiency required to reach the MCL. In developing the proposed
rule, EPA evaluated the costs of seveial required removal efficiencies. The revised cost
estimates reflect an expected distribution (based upon the radon occurrence data) of PTA
removal efficiencies required for treating radon in drinking water in order to meet the MCL.
-------
4-4
Exhibit -1-2 illustrates the revised assumption that 32 percent of affected sites will require an
percent removal efficiency. 3 1 percent will require a 50 percent removal efficiency, and 37
percent will require a 99 percent removal efficiency. EPA also included cost estimates in the
proposed rule and in the revised analysis (to a greater extent) for adding disinfection of drinking
water a’ a l trge dnber of sites that use aei tion treatment.
Exhibit 4-2
Percent Radon Removal Required at Treatment Sites to
Meet the Proposed M(L for PWSs Relying on Ground Water
M) X l tdou Reiiiuval
99 Radon Removal
(37’/ )
EPA derived cost estimates for treating radon in drinking water by estimating the capital
and operating cost components of PTA treatment facilities. Exhibit 4-3 provides details on the
derivation of each cost component.
5( )‘ Radon Removal
(3 1 )
‘ 11
-------
4-5
Exhibit 4-3
Unit Cost Derivation for PTA Treatment
l’ype of Cost
Equipment
Cost Rasis
Process
Equipment
Column Shell
Based on the mass of steel required to construct one column shell, the unit cost
of steel, fabricator mark-up, and number of column shells The mass of steel
required is based on the volume of stainless steel necessary to construct the
column shell using packing height plus one meter, column diameter, 1/4 inch
wall thickness, and a factor of two to account for access ports, flanges, ladders,
and other extra items
Internals
Internals per column require one support plate, one liquid distributor, arid
distributor rings at six-foot intervals along the packing height Delivered pricc
for the three items, with diameters ranging from I to 10 feet, were obtained
from a major equipment supplier
Packing Material
Uiiit cost of 1-inch plastic saddle packing material, ranging from 100 to 2500
cubic feel, was obtained from a major supplier
Blowers
Delivered cost was estimated as a function of total air flow and total air
pressure drop Purchase prices were obtaimied from a major sC j 1ier for air flow
rates from 200 standard cubic feet per minute (SCFM) to 6500 SCFM and
pressures from 0 2 to 8 0 inch water column
Pumps
Pumps were assumed to be capable of transfemng water from air well to
distribution system at a pressure of 100 pounds per square inch (psi) Pumps
were assumed to be 1.750 RPM vertical split-case cast iron pumps with totally
enclosed fan cooled motors for water volumes ranging from 10 gallons per
minute through 4 million gallons per day (MCD) Multiple pumps were
assumed necessary for volumes greater than 4 MGD
Support
Equipment
Installation
Installation represents the cost to install delivered equipment Installation costs
estimated at 50%, 25%, 100%, 25%, and 25% of the cost of the column shell,
internals, packing, blower, and pump, respectively
Air Well
One air well was used at each site requinng treatment The air well was
assumed to he a below-grade concrete structure whic- functioned as an effluent
holding tank and a foundation for the packed columns I ne installed cost of the
air well was estimated by the volume of excavation, select fill, and concrete
Piping
Piping cost was estimated by pipe length, diameter, number of connections
required, unit cost of pipe, and the labor hours required to install the piping
Air Duct
Cost assumed to be 20% of blower capital cost
Electrical
Installed cost of electrical equipment, transformers arid motor control stations
was assumed to be 25% of blower capital cost
Inuirect Cost
Includes all non-physical items required for PTA Indirect costs were assumed
to be a percentage of direct costs, and were made up of sitework (15% of direct
cost), design engineering (15%), contractor overhead and profit (12%), legal and
financial (2 5%), interest during construction (6%), and contingencies (20%)
Annual Cost
Annual cost consists of amortized capital and operating costs Amortized
capital cost is total capital cost amortized over a 20-year life cycle time period
of the PTA process at a 7 percent interest rate
Operating Cost Pump Electrical Electric power cost of pumping water is based on the average volume of water
Power treated per year and electrical power consumed by pump(s)
-------
4-6
Exhibit 4-3 (continued)
Administrative cost is based on 20% and 25% of the labor and maintenance
cost, respectively
4.2 HOUSEHOLD AND COMMUNITY COSTS OF TREATING RADON [ N PUBLIC
WATER SYSTEMS
Small public water systems incur higher per capita costs than do large water systems for
treating radon in drinking water derived from ground water because small systems:
Have a lower water flow capacity than large systems;
• Us a smaller percentage of their system capacity on average an large
systems; and
• Are more likely to require disinfection and pump replacements than large
systems.
Small systems by definition handle a lower water flow than larger systems. As illustrated
below, water systems can be grouped in 12 size categories, ranked by population served and
water flow rate. The smallest water systems (serving populations of 25 - 100) typically have
only one well or well group with a maximum flow late of 0.0056 million gallons per day
(MOD). The largest water systems (serving populations over I million people) have a maximum
flow rate of 430 MOD. Exhibit 4-4 displays the 12 size categories used in this report.
Type of Cost
Equipment Cost Basi
Blower Electrical
Power
Electric power cost of the blower is based on air flow and pressure drop
through packing .naterial plus 007 psi
Operating Labor
Operating labor cost estimate is based on an estimate of labor hours per year
and a labor rate of l4 70 per hour Labor hours per year were estimated to be
the product of 0 25 hour per column per shift, number of E’I’A columns in
operation, number of shifts per day, and 365 days per ear
Maintenance Cost
Maintenance cost is based on 10% and 4% of mechanical and non-mechanical
process equipment cost, respectively
Administrative
Cost
Total
Annual Cost
Total annual cost is
the sum of
amortized capital and operatin” list
Total
Production Cost
Total production Cost
treated per year
is the total
annual cost divided by the
volume of water
-------
4.7
Exhibit 4-4
Public Water System Size Categories
System Size
Population Range
erage System Flow
(MGI))
Desi •stem
Capacity Flow
(M CD)
I
25- 100
0.0056
0024
2
101 - 500
0024
0087
3
501 - 1000
0056
0.27
4
1,001 - 3,300
023
0.65
5
3,301 - 10,000
0.70
1.8
6
10,001 - 25,000
2 I
4 8
7
25,001 - 50,000
5.0
11
8
50,001 - 75,000
8 8
18
9
75,001 - 100,000
13
26
10
100,001 -500,000
27
51
II
500,001 - 1,000,000
120
210
12
Over 1,000,000
270
430
Larger systems are likely to have more treatment sites that might require separate treatment
facilities, w!Y smaller systems are more likely to have no more than one or ..vo facilities that
treat the entire water flow. The flow rate at any single site in the larger systems may be several
orders of magflitude greater than the flow rate at the smallest systems.
Exhibit 4-5 illustrates the relationship between system size and estimated flow rate at
affected PWS sites relying on ground water. Eighty percent of the smallest systems requiring
aeration treatment for radon are expected to have just one site with a flow rate of 0.02 MOD,
and 20 percent are expected to have two sites with flow rates of 0.01 MOD each. Of the largest
systems requiring aeration treatment for radon, EPA estimates that only 10 percent of affected
sites have flow rates as small as 4.3 MOD, 30 percent have flow rates of 7.17 MOD, another 30
percent have flow rates of 5.38 MOD, 20 percent have flow rates of 10.75 MOD, anu 10 percent
have flow rates of 21.5 MOD. EPA based this flow distribution on EPA analysis of data from
two national surveys. The much larger flow rates for affected sites at larger systems allow these
systems to realize substantial economies of scale in PTA treatment and reduce the treatment cost
per gallon treated.
-------
Exhibit 4-5
Estimated Water Flow Capacity
by System Size
Size Category: 1 Size Category: 6 Size Category: 12
Population Served: 25 - 100 Population Served: 10K - 25K Population Served: Over 1M
Total System Design Flow: 0.024 MGD Total System Design Flow: 4.8 MGD Total System Design. 431) MGD
H (ii 1 iI
- — — .——.— —
0.02 MG I ) — —
(80%)
- KEY 1
— z% of the sites in this system havt
a design flow capacity of x.xx M( .
1
717 MGI)
0.8 NH iD
(2O )
o G MGD
— — 50cc)
5.3 MGD
(30 )
x.xxMGD
(z%)
-------
4-9
In addition to having lower design flow rates, small water systems also tend to use a
smaller percentage of their flow capacity than do large systems. Exhibit 4-6 shows that while
the smallest systems have an average flow rate that is less than 25 percent of their design flow
rate, the largest systems have an average flow rate that is almost 65 percent of their design flow
rate. Alth h L 1 nent facilities must be des ’ ed to accommodate the m ixiI giai water flow
rate, the cost of these facilities must be spread over the amount of water actually used (i.e., the
average flow rate) Larger facilities effectively lower their cost per gallon by achieving a higher
utilization rate for their systems.
0
N
0..
65
60
55
50
45
40
35
30
25
20
0
Exhibit 4-6
Average Daily Flow as Percent of Design Flow
System Size
Aeration treatment of ground water for radon removal may also introduce the need for
disinfection and pump replacement. Larger systems are more likely than smaller systems to have
already incurred these costs. A greater percentage of small systems, however, will incur
disinfection and pump replacement costs in addition to basic aeration treatment costs. EPA
estimates that approximately 50 percent of the smallest systems are expected to incur addiUonal
disinfecuon costs, while less than 10 percent of the largest systems are expected to incur such
costs.
Exhibit 4-7 presents the cumulative effect of radon trea ient costs incurred by small and
large systems. The smallest systems incur costs of over $2.20 per thousand gallons, while the
2 3 4 5 6 7 8 9 iO 11 12
-------
4-10
largest systems incur Costs of almost $0.06 per thousand gallons. For systems adding
disinfection, costs would be additional $1.00 to $0.04 per thousand gallons, approximately, for
very small and large systems, respectively.
24
22
2
18
1.6
4 1.2
0
L I
0.8
0.6
04
0.2
0
ExhiUit 4.7
Average Treatment Cost per Thousand Gallons 2
4.3 SUMMARY OF COST ESTIMATES USED IN PROPOSED RULE FOR PWSs
The proposed rule provided costs as then estimated for radon treatment in PWSs relying on
ground water. EPA estimated that approximately 26,000 water systems require at least some
radon treatment, with a total cost of about $180 million to bring all water systems into
compliance with the MCL. Radon mitigation cost estimates ranged from approximately $170 per
household per year for the smallest systems to approximately four dollars per household per year
for the largest systems. The proposed aeration treatment cost estimates for 80 percent radon
removal efficiency ranged from $0.94 per thousand gallons for the smallest systems to $0.05 per
2 Exhibit 4-8 presents a weighted average of costs per thousand gallons, assuming (I) a revised flow distribution for
each system size, (2) 50% radon removal required for 31% of affected systems, 80% radon removal required for 32% of
affected systems, and 99% radon removal required for 37% of affected systems, and (3) additional disinfection treatment
and pump replacements required for some systems
2 3 4 5 6 7 8 9 10 11 12
System Size
-------
4-Il
thousand gallons for the largest systems. 3 Exhibit 4-8 shows that average costs per gallon are
highest for the four smallest system size categories (serving populations of less than 3,300),
while the fifth system size category (serving populations of 3,300 to 10,000) InCUrS costs per
gallon similar to larger systems.
09
ox
07
06
0
04
03
02
0l
0
Exhibit 4-8
Average Cost per Thousand Gallons
Estimated in Proposed Rule
(Aeration Treatment)
En addition to costs per gallon, the proposed rule presented estimated capital costs and
operation and maintenance (O&M) costs by system size. Exhibit 4-9 summarizes O&M costs
for systems requIring 80 percent radon removal
Exhibit 4-9
Costs From Proposed Rule for
80 Percent Radon Removal
(thousands of dollars)
Type of Cost 25 100
Po
pulation Served
iOI - 500
501
- 1,000
1,001 - 3,300 3,301 - 10,000 Over 1 Million
Capital Cost (per system) 15
33
58
78 100 13,000
O&M (per system per year) 02
06
1 4
3 1 76 3 4C0
The proposed rule, as published in the
removal efficiency of 80 percent
Federal Regist
er on July 18,
1991, provided cost estimates only for radon
2 3
System Size
4 5 12
-------
4-12
More recent analyses conclude that the proposed rule underestimated the actual costs of
radon treatment. The following section explains the changes in the assumptions used to derive
total CoSts.
4.4 SUMMARY OF REVISIONS TO COST ESTIMATES FOR TREATING RADON IN
PWSs
After reviewing and considering public comments on the proposed rule, EPA revised
several underlying assumptions used to calculate the unit costs of PTA treatment to remove
radon from public water systems relying on ground water. Exhibit 4-10 summarizes the original
and revised costs.
Exhibit 4-JO
Cost Estimate Revisions
Cost Estimates
Type of Cost Proposed Revised
Labor Rate for Small Systems $5 90/hr $14 70/hr
Costs based on flat rate of 0 3 cents per 0 25 labor hours per PTA i oluinn in
Operating Labor
thousand gallons treated operation per shift
Mobilization and Bonding Not included Contingencies Increased to 20% from 15%
Cost Indexes 1986-1989 Updated to 1991
Increased to 40% to Cover Uncertainties
Safety Factors 20% Transfer Coefficient
(Overdesign)
New Finished Water Pumps Not included New Costs included
Increased to $51’ - ‘iour labor costs, 1991
Process Piping i i” r Costs $30 per hour lai’ costs, 1988 cost index
cost index
Graded estimate 50% of small systems
[ )isinfcctioii Treatment 25% of systems add disinfection
and 10% of large systems add disinfection
EPA decided to revise the estimates presented in Exhibit 4-10 for the following reasons.
•
Labor Rate
The new estimated labor rate of $14.70 per hour, which includes additional costs such as
worker benefits, is substantially higher than the previously as umed rate of $5.90 per hour.
The previous rate of $5.90 only included small systems, whereas the revised rate of $14.70
includes both small and large systems The rate increase was based on a survey of rural
water associations. Because $14.70 is an average, it may overstate labor costs in some
small systems and communities and may understate labor costs for some larger systems
and communities.
ci
-------
4-Is
Operating Labor
Total labor costs depend on the hourly labor rate and the number of labor hours required.
EPA previously based labor costs on a flat rate of 0.3 cents per thousand gallons treated.
Operating labor is now estimated to ual 0.25 hours per column r i shill (i e., 0.25 hours
times the number of PTA columns in operation times the number of shifts per day times
365 days per year). EPA based the estimate of 0.25 hours per column per shift on an
operator checking and recording the water and air flows, air pressure drop, pump operation
and blower operation for each column dunng each shift
• Mobilization and Bonding
These two factors were not included in the proposed rule. EPA incorporated mobilization
and bonding (i.e., construction start-up and financing) by adding five percent of total direct
costs to the prior 15 percent contingency factor.
• Cost 1nde
EPA used the most current available cost index (1991) to update all cost estimates (from
1986-1989 basis).
• Safety Factors
The original cost estimate included a 20 percent safety factor to cover lower than expected
mass transfer properties of packing materials. Upon further consideration, EPA decided to
add an additional 20 percent to cover other uncertainties, such as the possibility of
overdesign of aerators, because it will be easy and economical to design systems to meet
radon levels below the MCL.
• New Replacement Pumps
Some water systems, especially smaller ones, may require replacement pumps as part of
the PTA treatment. These costs, not included originally, are now included in revised cost
estimates.
• Process Piping
The American Water Works Association (AWWA) recommended that EPA increase the
amount of process piping reflected in the cost estimate. EPA has revised the labor rate for
installing process piping to $50 per hour-from $30 per hour in the proposed rule and
updated the cost index. To evaluate capital costs, EPA added the cost of 150 feet of piping
to its cost model. The additional piping affected piping costs but did not significantly
increase total capital costs
-------
4 -1
Disinfection Treatment
Similar to the AWWA recommended approach, EPA has upgraded the estimate of systems
that will add disinfection to protect against microbiological contamination in the aeration
poces Based pon national surve . approximately 50 percer. of small systems and
10 percent of the largest systems would add disinfection ti atment as a result of radon in
water regulation.
4.5 REViSED SUMMARY COST ESTIMATES FOR TREATING RADON EN
DRINKING WATER
Exhibit 4- 11 shows EPA’s revised “low’ t and “high” estimates for national costs, given the
uncertainty associated with occurience and cost estimates. Exhibit 4-12 presents the cumulative
effect of the occurrence and cost estimate revisions for community and non-transient non-
community systems EPA’s revised best estimate of the national cost for treating radon in
drinking water is $ - illion, versus the 1991 estimate of $180 million for an MCL of 300
pCi
Exhibit 4-11
National Cost Estimates for Controlling or Mitigating Radon in Water
(millions of dollars)
Type of Cost
Proposed
Estimate
Revised Best
Estimate
Total Capital
$1,579
$1,602
Annual AmortiLed
Capital
106
151
Annual O&M
74
21
Total Annual
180
272
As noted above, the total annual cost of over $272 million to treat radon in drinking water
will not be distributed evenly among large and small water systems. Exhibit 4-12 shows that,
based on the best revised cost estimates, EPA estimates that the average household cost of radon
treatment will range from $242 per household per year in the smallest water systems (i.e., those
serving fewer than 100 people) to about $5 per household per year in the largest water systems.
c i
-------
4-t5
Exhibit 4-12
Average Annual Household Cost
of Treating Radon in PWSs Relying on Ground Water
300
i s o
2oo
‘so
<100
50
0
4.6 COSTS OF’ MITIGATING RADON IN INDOOR AIR
Exhibit 4- 13 presents two estimates of the annual cost for testing and mitigating radon in
indoor air under different discount rate and amortization period assumptions. EPA based the
annual cost estimate of $1,504 million for testing and mitigating radon in indoor air on a 3
percent discount rate over a 74-year time period. This annual cost estimate is not directly
comparable to the annual cost estimate for reducing radon in drinking water, because the
drinking water cost estimate is based on a 7 percent discount rate over 20 years. The 7 percent
discount rate used in the drinking water cost estimate reflects updated guidance from the Office
of Management and Budget on appropriate discount factors, and the 20-year amortization period
reflects the estimated 20-year life of a PTA water treatment facility. The 74-year amortization
period used in the indoor air cost estimate is based on the average life expectancy of the U S.
population and housing demolition data indicating that the average home life is longer than 74
years However, EPA also estimated costs under shorter amortization periods to reflect the
lowest available estimate of the average home resale interval (i.e., five years) and the period of a
typical mortgage (i.e., 30 years). Under a 74-year amortization period assumption, the annual
cost for testing and mitigating radon in indoor air would be $1,980 million if costs were
annualized at a 7 percent discount rate.
1 2 3 4 5 6 7 8 9 iO 11 i2
System Size
-------
4-16
Exhibit 4-13
Annual Cost Estimates for Testing & Mitigating Radon in Indoor Air
(millions of dollars per year)
Cost
D
iscount Rate, Amortization Period Estimate
3
percent, 74
years
$1,504
7
percent, 74
years
1,980
4.7 COMBINED COSTS OF CONTROLLING SOURCES OF RESIDENTIAL RADON
The combined annual cost estimate is $2,252 million for controlling residential radon from
all sources. The component cost estimates for indoor air ($1,980 millions of dollars per year)
and drinking water (272 millions of dollars per year) are both based on a 7 percent discount rate.
The drinking water cost estimate is based on a 20-year amortization period for PTA treatment
facilities, however, whereas the indoor air cost estimate is based on a 74-year amortization
period. Comparing the costs for air and water is difficult because the air program costs are
based on 100 percent compliance of a voluntary program, whereas the cost for water are based
on compliance for public water systems to meet the requirements of the Safe Drinking Water
Act
-------
5. COST OF RISK REDUCTION
Section 5. 1 discusses the cost of risk reduction from radon in drinking water, based on the
revised cost estimates and the number of deaths per year that would be avoided by reducing all
ground water-based public water systems to a level below the MCL of 300 picncuries 4 per liter
of water (PCi/Lwater). Section 5.2 discusses the cost of risk reduction from indoor air treatment,
and Section 5.3 discusses the cost of risk reduction from all sources of residential radon.
5.1 COST OF RISK REDUCTION FROM RADON IN WATER
EPA’s best estimate indicates that about 86 deaths per year could be avoided by reducing
radon concentration in all ground water-based public water systems to 300 pCi per liter. Based
on this estimate of avoided deaths per year and the low, best, and high estimates for the annual
costs of treatment for radon in water, the estimated cost per life saved ranges from $3. 1 million
to $3.7 million for treating radon in PWSs relying on ground water. Exhibit 5-1 portrays these
estimates.
Exhibit S-I
Low, Best, and High Estimated Cost Per Life Saved from
Controlling Radon in PWSs Relying on Ground Water
(in millions of dollars)
$6
$37
/Ti .: 1
—
. $2
/ / /
Low -i;;;- H h
4 A Curie (C1 is a measure of radioactivity, and a picocuric (pCi) is one trillionth of a Curie. One pCi equals 0.037
atomic disintegrations per second, or one atomic disintegration every 3() seconds.
-------
5-2
The Science Advisory Board suggested that EPA compare cost per hfe saved for large
versus small water systems. Revised cost estimates indicate that the average cost per life saved
ranges from about $1.2 million per life saved in the largest water systems to $7.9 million per life
saved in the smallest water systems.
5.2 COST OF RISK REDUCTiON FROM RADON IN AIR
Exhibit 5-2 presents two estimates of the cost of risk reduc on from testing and mitigating
radon in indoor air. Based on the estimate of 2,200 cancer cases avoided and the two cost
estimates presented in Exhibit 4-13, the cost per life saved may range from $0.7 million to $0.9
million, depending on the discount rate and amortization period assumptions used in the annual
cost estimate.
Exhibit 5-2
Cost of Risk Reduction from Testing & Mitigating Radon in Indoor Air
(millions of dollars per life saved)
Discount
Ra
Cost per
te, Amortization Period Life Saved
3
percent,
74
years
$0.7
7
percent,
74
years
$0.9
5.3 COST OF RISK REDUCTION FROM ALL SOURCES OF RESIDENTIAL RADON
Exhibit 5-3 presents the combined cost of risk reduction from all sources of radon. The
component cost estimates for indoor air and drinking water are based on a 7 percent discount
rate. However, the drinking water cost estimate is based on a 20-year amortization period for
PTA treatment facilities, whereas the indoor air cost estimate is based on a 74-year life
expectancy for the U.S. population. Based on these assumptions, the combined cost of risk
reduction from controlling radon in drinking water and indoor air is approximately $1 million per
life saved
“3
-------
5-3
Exhibit 5-3
Cost of Risk Reduction from Controlling Selected Sources of Residential Radon
(annual lives saved and millions of dollars)
Annual Annual Cost/Life
Residential Radon Source Lives Saved Cost Saved
Indoor Air 2200 S 1,980 $0.9
Drinking Water 85 272 3.2
All Sources 2285 2252 1.0
-------
PART FOUR
COMMENTS ON THE RADON
IN DRINKING WATER RULE
V
-------
6. SAB COMMENTS RECEIVED JUNE 15, 1993
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D C. 20460
Re: Review of Uncertainty Analysis of Risks Associated with Exposure to Radon--
“Chafee-Lautenberg Multi-media Risk Study”
Dear Ms. Browner:
The Science Advisory Board (SAB) is working with the Agency to reply to the so-called
Chafee-Lautenberg amendment” which is a part of the Agency’s FY93 appropriation act.
The Act calls for Agency generation and SAB review of a Study that addresses a) a multi-
media risk assessment of radon gas and b) an assessment of the costs of mitigating those
risks. As described in our recent commentaiy (EPA-SAB-RAC-COM-93-003), the attached
report is the first of three SAB reports that you will receive in connection with the Chafee-
Lautenberg Study. This report addresses the risks posed by radon gas in various media (e g.,
basements of homes and drinking water), with a focus on the Agency’s quantitative
uncertainty analysis associated with these risk estimates.
Specifically, this report is based upon Radiation Advisory Committee’s review of the
EPA risk assessment study, “Uncertainty Analysis of Risks Associated with Exposure the
Radon in Drinking Water” (January 29, 1993), related documents and public comment. The
review was conducted at a public meeting February 17-19, 1993.
The Committee’s charge was to review the adequacy of revisions of inhalation and
ingestion risk from radon progeny and the adequacy of uncertainty analysis regarding risk
assessment of water-borne radon, including health risk analysis and exposure analysis. En
considenng adequacy in the review, the Committee was mindful of concerns it had expressed
in two earlier SAB reports about EPA documents on radon in drinking water which were
transmitted to the Administrator in January, 1992
Technical Observations
The Committee commends the EPA staff for having produced an excellent docwnent
that responds to previous SAB comments on uncertainty analysis and the exposure to radon
gas at the point of use (e.g., showering). This response is all the more impressive given the
constraint of tight deadlines imposed upon it by Congressional and Court mandates. Its
quantitative analysis of uncertainties in the radon risk assessment represents a methodology
that is essentially state-of-the-art and significantly enhances the scientific credibility of the
EPA’s decision-making basis. The Committee assumes that this reflects the EPA’s recently
-------
6-2
stated commimlent to a more rigorous approach to evaluating uncertainties in its risk analyses
of radiological and other hazardous exposures in the future. However, the Committee
continues to have concerns about the exposures and risks that could be associated with certain
treatment options (e.g., granular activated carbon), once those options are selected.
Based on the current analysis, the risks associated with radon gas in homes is
considerably greater than the risks associated with the risks posed by radon gas in the
drinking water supply. That smaller risk from radon gas in drinking water is composed of
nearly equal contributions of the inhalation arid ingestion pathways. The Committee notes,
however, that the quantitative uncertainty analysis for the drinking water case does not cover
some of the more important uncertainties. In particular, the Committee believes that the
overall uncertainty regarding the ingestion risk estimate is substantially greater than would be
inferred from the quantitative confidence interval
Overall, the Committee finds that the EPA has adequately addressed most of the issues
raised by the Committee in its earlier reports, either by incorporating the Committee’s
previously recommended changes into the new documents or by providing additional
background documentation supporting the EPA’s position. In the accompanying report the
Committee makes a number of specific scientific comments and recommendations for
additional improvements to the document. These deal with important issues such as
uncertainties associated with an unpublished study that contributes significantly to the
estimated internal doses from ingested radon-containing drinking water, the influence of
smoking on lung cancer risks from radon, and, again, unsettled question of treatment
technologies. These issues can generally be addressed by including clarifying statements.
Futher, the changes in most cases would not substantially change the document’s estimates of
central values for risks.
Policy Observations
The comments below, to some extent, reach beyond the strictly technical issues
examined by the Committee. However, the Committee feels that it was important that the
Agency have the benefits of these thoughts, also, as the decisionmaking process continues.
The Radiation Advisory Committee has long encouraged the use of integrated
quantitative uncertainty analysis in a variety of EPA assessments. As noted above, the
Committee is extremely pleased to see that the EPA has done such an analysis in this case
The Committee applauds EPA for its timely incorporation of a full quantitative uncertainty
analysis for each pathway in its assessment and hopes that the use of quantitative uncertainty
analysis will become a routine part of all EPA assessments, not only those associated with
radiation risks. This information should be a valuable aid in guiding EPA in its consideration
of possible regulatory strategies.
-------
6-3
The Committee agrees with the Agency’s Feb. 26, 1992 “risk characterization memo”
that articulates the EPA policy of explicitly disclosing uncertainty in quantitative risk
assessment Screening risk assessments involve only point estimate calculations, and
assumptions used to derive these estimates are generally biased on the conservative side and
can be misleading in terms of indicating the need for regulatory action In contrast,
regulatory action must be based on realistic estimates of risk and these require a full
disclosure of uncertainty. The disclosure of uncertainty enables the scientific reviewer, as
well as the decisionmaker, to evaluate the degree of confidence that one should have in the
risk assessment.
In its January 29, 1992, Commentary: Reducing Risks from Radon; Drinking Water
Criteria Documents (EPA-SAB-RAC-COM-92-003), the Committee noted that the radon risk
reduction situation reflects the fragmentation of environmental policy identified in Reducing
Risk (SAB-EC-90-02 I). Therefore, the Committee suggested that the EPA focus its efforts
on primary sources (e g., radon in some home basements) rather than on secondary sources of
risk, such as radon in drinking water, which is a very small conthbutor to radon risk, except
in rare cases.
In summary, within the limitations of the data currently available, the EPA has now
successfully prepared a scientifically credible multi-media nsk assessment for regulatory
decision-making on radon. The Committee’s agreement with the principle of radiation
protection optimization and in the concepts articulated in Reducing Risk lead it to note once
again that radon in drinking water represents only a small fraction of radon exposure and risk
compared to radon in indoor air from non-water sources. We acknowledge, however, that the
relative emphasis given to various radon exposure reduction methods--whether for radon from
water or non-water Sources--is a policy choice for which scientific analysis is only one of
many important inputs.
The Radiation Advisory Committee appreciates the opportunity to comment on the
EPA’s uncertainty analysis of risks associated with exposure to radon. We look forward to
receiving the EPA’s response to the this report, particularly as it relates to our explicit
recommendations.
Sincerely,
Dr. Raymond C. Loehr Dr. Genevieve M. Matanoski
Chair, Executive Committee Chair, Radiation Advisory Committee
Science Advisory Board Science Advisory Board
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
NOTICE
This report has been written as a part of the activities of the Science Advisory Board, a
public advisory group providing extramural scientific information and advice to the
Administrator and other officials of the Environmental Protection Agency. The Board is
structured to provide balanced, expert assessment of scientific matters related to problems
facing the Agency. This report has not been reviewed for approval by the Agency and,
hence, the contents of this report do not necessarily represent the views and policies of the
Environmental Protection Agency, nor of other agencies in the Executive Branch of the
Federal government, nor does mention of trade names or commercial products constitute a
recommendation for use.
-------
U
-------
III
ABSTRACT
The Radiation Advisory Committee has reviewed the EPA’s, “Uncertainty Analysis of
Risks Associated with Exposure the Radon in Drinking Water “(January 29, 1993), related
documents and public comment. The Committee reviewed the adequacy of the EPA’s
revisions of the risk assessment for both the ingestion and inhalation exposure pathways. and
the adequacy of the associated uncertainty analysis has been examined. The Committee also
considered the EPA’s estimates of risks associated with radon exposures due to releases at
drinking water treatment facilities. The Committee was mindful of its previously expressed
concerns regarding the Agency’s a) lack of quantitative uncertainty analyses, b) failure to
consider direct exposure to radon and its progeny released by showers, c) lack of an
assessment of risks associated with drinking water treatment, and d) lack of consideration of
potential occupational exposures and risk.
Overall the Committee finds that EPA has adequately addressed most of the issues
raised in earlier reports from the Committee. The quantitative uncertainty analysis developed
by the EPA represents a methodology that is state-of-the-art and significantly improves the
scientific basis for the EPA’s decision-making. The revised estimates for ingestion and
inhalation risks due to radon in drinking water are scientifically acceptable. There is concern,
however, that the uncertainties in the estimate of ingestion risk are larger than suggested by
the quantitative uncertainty analysis. The Committee recommends that the EPA incorporate a
qualitative discussion of known, but not quantified, uncertainties in its analyses and given the
larger uncertainty bounds associated with the ingestion risk, that consideration be given to
keeping the ingestion and inhalation risks separate in the EPA’s deliberations on standards for
radon in drinking water. The Committee also reiterated its previously stated concerns that the
overall risks associated with radon in drinking water are small compared with the average
radon exposures due to indoor air and that the drinking water risks be placed in context with
other radon risks in the summary documents developed by the EPA.
The Committee’s report also provides comments and recommendations regarding the
adequacy of the analysis and the approaches taken. Among these was the recommendation
that the EPA look at a range of water treatment technologies and include in the analyses risks
due to occupational radiation exposures and potential waste disposal issues. Finally, the
Committee also recommends that particular attention be given to the uncertainties associated
with the variance and shape of the probability density functions used by the EPA to represent
variability of exposures among individuals.
KEYWORDS: radon, drinking water, uncertainty, inhalation, ingestion
-ii
-------
Iv
-------
U S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
RADEATION ADVISORY COMMITI’EE
ROSTER
CHAIR
Dr. Genevieve M. Matanoski
Professor of Epidemiology
The Johns Hopkins University
School of Hygiene and Public Health
Department of Epidemiology
624 North Broadway, Room 280
Baltimore, Maryland 21205
MEMBERS
Dr. Stephen L. Brown
ENSR Consulting & Engineering
1320 Harbor Bay Parkway
Alameda, California 94501
Dr. June Fabryka-Martin
Los Alamos National Laboratory
Mail Stop J-5l4
Los Alamos, New Mexico 87545
Dr. Ricardo Gonzalez
U.P.R. School of Medicine
Post Office Box 365067
San Juan, Puerto Rico 00936
Dr. F. Owen Hoffman
SENES Oak Ridge, Inc.
Center for Risk Analysis
677 Emory Valley Road
Oak Ridge, Tennessee 37830
73
-------
VI
Dr. Arjun Makhijani*
Institute for Energy and Environmental Research
6935 Laurel Avenue
Takoma Park, Maryland 20912
Dr Oddvar F. Nygaard
Division of Radiation Biology
Case Western Reserve University
2199 Adelbert Road
Cleveland, Ohio 44106
Dr. Richard G Sextro
Lndoor Environment Program
Lawrence Berkeley Laboratory
Building 90, Room 3058
Berkeley, California 94720
Mr Paul G. Voilleque**
MW Risk Assessment, Inc.
Historic Federal Building
591 Park Avenue
Idaho Falls, Idaho 83405-0430
Dr. James E. Watson, Jr.
Department of Environmental Sciences
and Engineering
Campus Box 7400
University of North Carolina at Chapel Hill
Chapel Hill, North Carolina 27599-7400
* Although Dr Makhijani attended the February 17-19 meeting, his participation in this
review was limited.
**Mr. Voilleque was unable to attend the February 17-19, 1993 meeting where this review
was conducted and has subsequently resigned from the Radiation Advisory Committee
-------
V I I
DESIGNATED FEDERAL OFFiCIAL
Mrs. Kathleen W. Conway
Science Advisory Board (A-IOIF)
U.S. Environmental Protection Agency
401 M Street. S.W.
Washington, D.C. 20460
STAFF SECRETARY
Mrs. Dorothy M. Clark
Secretary, Science Advisory Board (A-IO1F)
U.S Environmental Protection Agency
401 M Street. S.W.,
Washington, D.C. 20460
-7 ’ ; ,
-------
vu’
-------
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY.
I I Background . .
1 .2. Technical Considerations
1.3 Policy Observations .
2. INTRODUCTION 6
2. I Relevant Prior SAB Reports . 6
2.2 Procedural History of this Review 6
3 FINDINGS AND DETAILED DISCUSSION RESPONDING
TO ThE QUESTIONS IN THE CHARGE
3 1 Adequacy of Revisions to Ingestion and
Inhalation Risk Estimates
3 2 Adequacy of Quantitative Uncertainty Analyses
Regarding Risk Assessment
3.3 Adequacy of Characterization of Risks from
Water Treatment Facilities
3.4 Other Scientific Issues
4 POLICY CONSIDERATIONS
4. 1 Quantitative Uncertainty Analysis
4.2 Relative Risk
4.2 Harmonizing
5. REFERENCES ....
5 I Documents Received m the Review Process
5 2 SAB Documents . . . . . . .
5.3 Literature Cited
APPENDIX A: Brief Chronology of Relevant SAB Reports
APPENDIX B: Congressional Record--Senate, S 15103, September 25, 1992
Distribution List
I
I
....
1
4
8
8
11
13
16
19
19
19
19
• 21
21
• . . 26
27
-77
-------
I 1. EXECUTIVE SUMMARY
2 1.1 Background
3 En EPA’s 1993 appropriatlo&, Congress required EPA to, conduct a risk assessment
4 of radon considering: . the risk of adverse human health effects associated with exposure to
5 various pathways of radon . . . Such an evaluation shall consider the risks posed by the
6 treatment and disposal of any wastes produced by water treatment. Congress also required
7 that. “The Science Advisory Board shall review the Agency’s study and submit a
8 recommendation to the Administrator on its findings. This letter and the accompanying
9 report set forth the Radiation Advisory Committee’s findings and recommendations based on
10 its review of the EPA risk assessment study, “Uncertainty Analysis of Risks Associated with
11 Exposure the Radon in Drinking Water “(January 29, 1993), related documents and public
12 comment. The EPA uncertainty analysis addressed four radon exposure pathways: inhalation
13 indoors of radon from non-water sources, inhalation of radon outdoors, ingestion of
14 waterborne radon, and inhalation of waterborne radon. The review was conducted at a public
15 meeting February 17-19,1993.
16 The Committee’s charge was to review the adequacy of revisions of inhalation and
17 ingestion risk from radon progeny and the adequacy of uncertainty analysis regarding risk
18 assessment of water-borne radon, including health risk analysis and exposure analysis. In
19 considering adequacy in the review, the Committee was mindful of concerns it had expressed
20 in reports about earlier EPA documents on radon in drinking water transmitted to the
21 Administrator on January 9 and 29, 1992: (a) that uncertainties associated with the selection
22 of’ particular models, specific parameters used in the models, and the final risk estimates were
23 not adequately addressed in any of the documents; (b) that high exposure to radon from water
24 at the point of use (e.g., a shower) had not been adequately addressed; (c) that regulation of
25 radon in drinking water introduces risk from the disposal of treatment byproducts, tradeoffs
26 which the EPA should consider more explicitly in its regulatory decision-making; and (U) that
27 regulation and removal of radon in drinking water may result in occupational exposures.
28 1.2 Technical Considerations
29 Regarding the Committee’s charge and concerns (a) and (b) above, the Committee
30 commends the EPA staff for producing an excellent document all the more impressive given
3 1 the constraint of tight deadlines imposed upon it by Congressional and Court mandates. Its
32 quantitative analysis of uncertainties in the radon risk assessment represents a methodology
33 that is essentially state-of-the-art for a regulatory agency and significantly enhances the
34 scientific credibility of the EPA’s decision-making basis. The Committee assumes that this
35 reflects the EPA’s recently stated commitment to a more rigorous approach to evaluating
36 uncertainties in its risk analyses of radiological and other hazardous exposures in the future.
37 With respect to concerns (c) and (d) above, the Committee recommends that EPA re-examine
38 its assumptions about which water treatment technologies will be used for radon removal.
39 ‘Departments of Veterans Affairs and Housing and Urban Development, and Independent Agencies Appropnation Act, 1993,
40 PUB L 102-398, Section 519, 106 STAT 1618 (1992)
-------
2
When EPA has determined the likely treatment options, then EPA should perform an
2 uncertainty analysis for occupational exposure based on that distribution (including the
3 uncertainty about how frequently the various options will be used). If granular activated
4 carbon is among those treatment options, then EPA should broaden the uncertainty analysis to
5 include the disposal of granular activated carbon.
6 With respect to the EPA’s analysis, the risk assessment of radon in drinking water has
7 been revised and an uncertainty analysis has been conducted using Monte Carlo simulation
8 methods. The uncertainty analysis incorporates quantifiable uncertainties in exposure and
9 toxicology, as well as true variation in exposure among individuals. EPA’s mean estimate for
10 the lifetime individual inhalation risk of lung cancer deaths per pUlL of radon in drinking
11 water is 3.6 x l0- , with a stated 90% confidence interval around the mean of 1.8 x l0 to
12 7.0 x l0- . The Agency’s mean estimate for the lifetime individual ingestion risk of fatal
13 cancers per pCilL of radon in drinking water is 1.8 x l0 with a stated confidence interval
14 around the mean of 6.9 x 10-8 to 6.4 x l0 . The Agency’s nominal estimate for individual
15 lifetime inhalation and ingestion risk per pCIJL for radon in drinking water are 3.0 x l0 and
16 3.5 x lO , respectively. Therefore, for drinking water risks, the contributions of the
17 inhalation and ingestion are almost equal.
18 The Committee notes, however, that the quantitative uncertainty analysis for the
19 drinking water case does not cover some of the more important uncertainties. In particular,
20 the Radiation Advisory Committee believes that the stated uncertainty range for the ingestion
21 nsk is too small in comparison with that for inhalation, because the ingestion risk estimate is
22 based on two major factors: (a) an estimate of the distribution of radon to organs in the
23 gastrointestinal tract, based on an unpublished study using xenon-133, and (b) the use of
24 organ radiation risk factors that are based on high dose and high-dose rate exposures to low-
25 LET radiation extrapolated to low dose and low-dose rates. These risk factors are then
26 converted to high-LET radiation risks for alpha particles associated with radon and its
27 progeny The Committee recommends that EPA not only make this clear in its documents
28 but also consider keeping the estimates or risks from inhalation and ingestion separate in its
29 discussion of standards for radon in drinking water.
30 Overall, the Committee finds that the EPA has adequately addressed most of the issues
31 raised by the Committee in its earlier reports, either by incorporating the Committee’s
32 previously recommended changes into the new documents or by providing additional
33 background documentation supporting the EPA’s position. The Committee makes the
34 following scientific comments and recommendations for additional improvements to the
35 document, but notes that these issues can generally be addressed by including clarifying
36 statements and that the changes in most cases would not substantially change the document’s
37 estimates of central values for risks. (A more detailed discussion of each of the comments
38 and recommendations can be found in the report Section identified in parentheses)
39 a. Recommendation Organ-specific doses used in the document for assessment of
40 ingestion risks are based, in part, upon a single study of kinetics of xenon in
41 humans, work that has not been published in the peer-reviewed literature. The
42 cited study also did not include a mass balance determination. Consequently, the
-------
I
I Committee recommends that the EPA carefully review this study to evaluate
2 whether the uncertainties attributed to the results are adequately described. (3. 1. I)
3 b Comment With regard to assessment of inhalation risks associated with drinking
4 water exposure (e.g., showering), the Committee believes that the EPA’s
5 uncertainty analysis is satisfactory and that, given the nature of the uncertainties,
6 the transfer factor approach used in the document adequately accounts for n.sks
7 arising from episodic shower exposures. (3. 1.2)
8 c. Recommendation The Committee noted some minor inconsistencies between values
9 in relevant documents and recommends that the EPA review its selection of
10 parameter values (including ranges and their uncertainties) for each exposure
11 pathway to ensure consistency with original data sources. (3.1.3)
12 d. Comment The Committee believes that the basic methods used to propagate
13 uncertainty are acceptable. Proper consideration has been given to the possibility
14 of covanance, and the Monte Carlo simulation methods are state-of-the-art. (3.2.1)
IS e Recommendation The Committee recommends that particular attention be given to
16 more completely addressing uncertainty about the variance and shape of the
17 probability density functions (PDFs) that have been assumed by the EPA to
18 represent variability in exposures among individuals. (3.2.2)
19 f. Recommendation The Committee recommends that the EPA include in its uncertainty
20 analysis a qualitative discussion of known uncertainty variables which were not
2 1 quantified in the uncertainty analysis. These include the issue of a linear dose rate
22 response extending to low doses, the influence of smoking on increasing lung-
23 cancer nsks from radon, and the effect of population mobility on the distribution
24 of risks. (3.2.3)
25 g. Recommendation In order to increase the scientific credibility of the results, the
26 Committee recommends that EPA consider upgrading the uncertainty analysis for
27 the risks associated with aeration for radon removal; however, the proposed
28 revisions to the analysis will not change the conclusion that the risk for a
29 maximally exposed individual attributable to radon released from a water
30 treatment facility will be less than or equal to the average risk attributable to 300
31 pCi/L of radon in drinking water used in the home. (3.3.1)
32
33 h Recommendation If EPA determines that granular activated carbon will be used for
34 radon removal, the Committee urges EPA to thoroughly and completely analyze
35 any potential risk and/or disposal problems related to the use of granular activated
36 carbon (GAC) for radon removal from drinking water
37 I. Recommendation EPA did not provide an analysis of occupational exposures as a
38 result of water treatment for radon The potential for such exposures appears to
39 depend heavily upon the choice of water treatment technology, and the Committee
40 recommends that such a comparative analysis be conducted for different
-------
4
technologies, such as aeration or granular activated carbon filtration, especially in
2 view of waste disposal problems that may result from use of the latter technology.
3 (3.3.3)
4 j Recommendation The Committee recommends that the document include a summary
5 of the results of the uncertainty analysis regarding the contribution of the various
6 exposure pathways to the overall radon risk to individuals and to the general
7 population. This summary should also highlight the major sources of uncertainty
8 contributing to the total uncertainty in the nsk estimate for each pathway. Such a
9 discussion would provide the information necessary to factor uncertainties and
10 variabilities into the cost-benefit analysis for the proposed regulation and to
11 calculate a range for the estimates of cosrllife saved. (3.4. 1)
12 k. Recommendation The Committee recommends that the EPA extend its population
13 risk assessment and uncertainty analysis to obtain an estimate of the lives that
14 would be saved by the proposed maximum contaminant level, using the same
15 assumptions as were used to calculate present-day nsks but using for radon
16 concentration a lognormal probability density function truncated at the maximum
17 contaminant level. (3.4.2)
18 1. Recommendation The Committee urges the EPA to submit its risk analyses for
19 publication in appropriate journals which would provide peer-review and
20 recognition that the EPA’s science is of high-quality and that it becomes part of
21 the mainstream of scientific criticism, revision, and acceptance (or rejection).
22 Publication will also assist in raising awareness within the scientific community to
23 the risk issues associated with radon. (3.4.3)
24 1.2 Policy Observations
25 The comments below, to some extent, reach beyond the strictly technical issues
26 examined by the Committee. However, the Committee felt that it was important that the
27 Agency have the benefits of these thoughts, also, as the decisionmaking process continues.
28 The Radiation Advisory Committee has long encouraged the use of integrated
29 quantitative uncertainty analysts in a variety of EPA assessments. The Committee is
30 extremely pleased to see that the EPA has done such analysis in this case. The Committee
3 1 applauds EPA for its timely incorporation of a full quantitative uncertainty analysis for each
32 pathway in its assessment and hopes that the use of quantitative uncertainty analysis will
33 become a routine part of all EPA assessments, not only those associated with radiation risks.
34 This information should be a valuable aid in guiding EPA in its consideration of possible
35 regulatory strategies.
36 The Committee believes strongly that the explicit disclosure of uncertainty in
37 quantitative risk assessment is necessary. Screenmg risk assessments involve only point
38 estimate calculations, and assumptions used to derive these estimates are generally biased on
39 the conservative side and can be misleading in terms of indicating the need for regulatory
40 action.
-------
5
Regulatory action must be based on realistic estimates of risk and these require a full
2 disclosure of uncertainty. The disclosure of uncertainty enables the scientific reviewer, as
3 well as the decision-maker, to evaluate the degree of confidence that one should have in the
4 risk assessment. (deleted sentence redundant with end of previous paragraph)
5 In its January 29, 1992, Commentary: Reducing Risks from Radon: Drinking Water
6 Criteria Documents (EPA-SAB-RAC-COM-92-003), the Committee noted that the radon risk
7 reduction situation reflects the fragmentation of environmental policy identthed in Reducing
8 Risk (SAB-EC-90-021). Because radon in drinking water is a very small contributor to radon
9 risk except in rare cases, the Committee suggested that the EPA focus its efforts on primary
10 rather than secondary sources of risk. Within the limitations of the data currently available,
11 the EPA has now successfully prepared a scientifically credible multi-media risk assessment
12 for regulatory decision-making on radon. The Committee’s agreement with the principle of
13 radiation protection optimization and in the concepts articulated in Reducing Risk lead it to
14 note once again that radon in drinking water represents only a small fraction of radon
15 exposure and risk compared to radon in indoor air from non-water sources. The emphasis on
16 various radon exposure reduction methods--whether for radon from water or non-water
17 sources--is a policy choice for which scientific analysis is only one of many important inputs.
18 In its May 8, 1992 Commentary on Harmonizing Chemical and Radiation Risk
19 Reduction Strategies (EPA-SkB-RAC-COM-92-007), the Committee brought to the EPA’s
20 attention the need for a more coherent policy for making risk reduction decisions with respect
21 to radiation and chemical exposures. The control of radon in drinking water presents a
22 situation where a radiological contaminant being regulated by a paradigm developed for
23 chemicals, yet radon in drinking water represents only a small fraction of radon exposure.
24 The Committee appreciates the EPA’s difficulty in establishing a coherent risk reduction
25 strategy under the variety of statutes governing EPA and acknowledges that harmonization
26 does not necessarily imply identical tream ent. However, the Committee urges the EPA to
27 explain clearly why the risks from radiation (in this case radon in indoor air) and chemicals
28 (in this case radon in drinking water) are treated differently under specified conditions and in
29 specified exposure settings. The Committee urges EPA, the Congress and the public to
30 carefully consider how chemical and radiation risks are being regulated in this case.
-------
6
1 2. INTRODUCTION
2 2.1 Relevant Prior SAB Reports
3 For many years the Radiation Advisory Committee and other SAB committees have
4 urged the incorporation of quantitative uncertainty analysis into EPA assessments to
5 explicitly disclose the extent of confidence that one should have in the results of these
6 assessments and to identify areas where the acquisition of additional information could lead to
7 substantial improvements in the estimation of risks and uncertainties. In its recent multi-
8 media radon risk assessment study entitled, “Uncertainty Analysis of Risks Associated with
9 Exposure the Radon in Dnnking Water” (January 29, 1993) the EPA has implemented most
10 of the SAB’s recommendations in a scientifically credible manner. A brief chronology of
I I relevant SAB reports can be found in Appendix A.
12 2.2 Procedural History of this Review
13 This review resulted from the Chafee-Lautenberg amendment. (A copy of the complete
14 language can be found in Appendix B.) More formally known as the Departments of
15 Veterans Affairs and Housing and Urban Development, and Independent Agencies
16 Appropriation Act 1993, PUB. L. 102-398, Section 519, 106 STAT 1618 (1992), the
17 amendment was also published in the U.S. Congressional Record and appears as Attachment
18 1 to this report. Regarding this review, Congress required EPA to, “conduct a risk assessment
19 of radon considering: (A) the risk of adverse human health effects associated with exposure to
20 various pathways of radon; (B) the costs of controlling or mitigating exposure to radon; and
2 I (C) the costs for radon control or mitigation experienced by households and communities,
22 including the costs experienced by small communities as the result of such regulations. Such
23 an evaluation shall consider the risks posed by the treamient or disposal of any wastes
24 produced by water treatment. The Science Advisory Board shall review the Agency’s study
25 and submit a recommendation to the Administrator on its findings.” This report by the SAB’s
26 Radiation Advisory Committee is a review of EPA’s work in response to (A). The SAB’s
27 Drinking Water Committee is reviewing the Agency’s work in response to (B) and (C) and is
28 generating a separate SAB report In addition, a subcommittee of the SAB Executive
29 Committee will generate a third SAB report that reviews the Agency’s “synthesis document”
30 that is being generated by EPA for submission to the Congress.
3 I At publicly announced conference call meetings November 30, December 2, December
32 3. and December 17, 1992, the Radiation Advisory Committee together with members of the
33 Drinking Water Committee, Environmental Engineering Committee, and Indoor Air Quality
34 Committee provided a consultation to the EPA staff. The consultation was on EPA’s outline
35 for a multi-media radon risk assessment and on the parameters and uncertainty analysis for
36 the assessment. The SAB has developed the consultation as a mechanism to advise the EPA
37 on technical issues that should be considered in the development of regulations, guidelines, or
38 technical guidance before the EPA has taken a position. Consultations differ from other SAB
39 activities in that no report is generated by the SAB and no response from the EPA is required.
-------
7
The review of ‘Uncertainty Analysis of Risks Associated with Exposure to Radon in
2 Drinking Water “ (January 29, 1993), related documents and public comment was conducted
3 at a February 17-19, 1993 publicly announced meeting of the Radiation Advisory Committee.
4 The first draft of this report was made available to the EPA and the public on February 19
5 Written comments were received from the EPA and the public subsequent to the meeting.
6 The Committee held non-public writmg sessions by conference call to revise the draft pnor to
7 its submittal to the Execuuve Committee.
8 The Committee’s charge was to review the adequacy of revisions of inhalation and
9 ingestion nsk from radon progeny and the adequacy of uncertainty analysis regarding risk
10 assessment of water-borne radon, including health risk analysis and exposure analysis. In
I I considering adequacy in the review, the Committee was mindful of concerns it had expressed
12 in reports about earlier EPA documents on radon in drinking water transmitted to the
13 Administrator on January 9 and 29, l992 (a) that uncertainties associated with the selection
14 of particular models, specific parameters used in the models, and the final risk estimates were
IS not adequately addressed in any of the documents, (b) that high exposure to radon from water
16 at the point of use (e.g., a shower) had not been adequately addressed; (c) that regulation of
17 radon in drinking water introduces risk from the disposal of treatment byproducts, tradeoffs
IS which the EPA should consider more explicitly in its regulatory decision-making; and (d) that
19 regulation and removal of radon in drinking water may result in occupational exposures.
-------
8
1 3. FINDINGS AND DETAILED DISCUSSION
2 3.1 Adequacy of Revisions to Ingestion and Inhalation Risk Estimates
3 3.1.1 Are revisions of ingestion risk estimates for waterborne radon and its progeny
4 adequate?
5 1. Recommendation Organ-specific doses used in the document for assessment of
6 Ingestion risks are based, in part, upon a single study of kinetics of xenon in humans, work
7 that has not been published in the peer-reviewed literature. The cited study also did not
8 include a mass balance determination. Consequently, the Committee recommends that the
9 EPA carefully review this study to evaluate whether the uncertamties attributed to the result.s
10 are adequately described.
11 Discussion . Revisions of ingestion risk resulted from modifications of gastrointestinal
12 (Gl) and lung dosimetry and from the use of revised organ-specific risk coefficients,
13 particularly that for the stomach. The revised ingestion risk is greater than the previous
14 estimate (EPA, 1991) by a factor of 2 3. The Committee has reviewed these revised risk
15 coefficients. The Committee’s primary concern is that radon retention times in organs are
16 based upon a single study of kinetics of xenon in humans (Correia et al., 1987), work that
17 has not been published in the peer-reviewed literature. The xenon study also did not include
18 a mass balance determination. Consequently, the Committee recommends that the EPA
19 carefully review this study to evaluate whether the uncertainties attributed to the results are
20 adequately described. Other factors in the EPA’s biological model that are difficult to verify
2 1 are the assumptions that a diffusion gradient exists in the GI tract and that lead-2 14 and
22 subsequent decay products are removed from the GI tract before decaying and do not
23 contribute to dose. The implications of these assumptions have been considered in the
24 uncertainty analysis, and in this case also the Committee recommends the EPA carefully
25 review these factors to evaluate whether the uncertainties are adequately described. Many of
26 these uncertainties are difficult to quantify because alternative formulations and parameter
27 values have not been proposed. EPA has adequately captured the apparent quantifiable
28 uncertainties in the ingestion risk estimates and has propagated them properly, in the opinion
29 of the Committee. However, the quantitative uncertainty bounds may give rise to a false
30 sense of the overall reliability of the ingestion risk estimates. Qualitative uncertainties about
31 the formulation of the exposure models and the applicability of high-dose, high-dose-rate,
32 low-LET risk coefficients to the low-dose, low-dose-rate, high-LET exposure conditions
33 present with ingestion of radon in drinking water are substantial. An expanded discussion of
34 the implication of these qualitative uncertainties is important to EPA’s consideration of
35 regulations for radon in drinking water.
36 3.1.2 Are revisions of inhalation risk estimates for waterborne radon and its
37 progeny adequate?
38 2. Comment With regard to assessment of inhalation risks, the Committee believes that
39 the EPA’s uncertainty analysis is satisfactory and that, given the nature of the uncertaintiec,
-------
9
1 the transfer factor approach used in the document adequately accounts for risks arising from
2 episodic shower exposures.
3 Discussion . The analysis of inhalation risk from radon in water has two components.
4 The first considers exposures from radon released from general water use within a house.
5 The EPA applied a general transfer factor that describes radon release from water indoors.
6 The factor used had a value of 1 m 10,000 (i.e., 10,000 pCiIL in water yields an average
7 indoor air concentration of 1 pCiIL), which is consistent with values used and published by
8 others. In order to investigate whether exposures to radon from releases in showers represent
9 a significant episodic peak exposure not captured by an average transfer factor approach, the
10 EPA used a multicompartment model, based on one developed by McKone (1987). Because
I 1 the analysis of shower exposures required that radon progeny ingrowth and decay be
12 accounted for, the model specifically recognized the differences between radon and radon
13 progeny exposures. The multicompartment model yielded results that were somewhat higher
14 for radon but somewhat lower for radon progeny when compared with the analysis based on
15 use of an average transfer factor
16 The Radiation Advisory Committee believes, first, that the EPA’s analysis, incorporating
17 an uncertainty analysis, is satisfactory and, second, that given the nature of the uncertainties,
18 the EPA’s conclusion that episodic shower exposures are adequately accounted for by a
19 transfer factor approach is also satisfactory.
20 3.1.3 Discrepancies in Numerical Values: Are EPA’s choices for risk parameters
21 and the uncertainties adequately defended?
22 3. Recommendation The Committee noted some minor mconsistencies between values
23 in relevant documents and recommends that the EPA review its selection of parameter values
24 (including ranges and their uncertainties) for each exposure pathway to ensure consistency
25 with original data sources.
26 Discussion . Some examples of discrepancies follow.
-------
i0
3.1.3.1 Estimates of risk due to inhalation of indoor air
2 En general, the estimated central value for the annual number of lung cancer cases and
3 the corresponding upper and lower bounds appear to be in the same range in the present
4 assessment as in the previous assessment. However, the lack of consistency in the risk factor
5 used is troubling. The summary information presented in Table 6-2 of the EPA document
6 (EPA, 1993) does not appear to be entirely consistent with the parameter values used
7 previously. The Committee recommends that the previous values be used throughout or that
8 clarification of the differences be made in the document.
9 3.1.3.2 Estimates of risk associated with inhalation of outdoor air
10 Although the total risk associated with inhalation of radon and its progeny in outdoor air
II is small compared with that attributable to inhalation of radon and its progeny in indoor air,
12 the estimated lung cancer risks due to outdoor radon/radon progeny exposures are, in fact,
13 larger than those estimated to arise from radon in dnnking water. Hence, it is important that
14 the uncertainties in the risk assessment for the outdoor pathway be assessed in a manner
15 consistent with that used for the indoor (drinking water) pathway. Examples of points of
16 concern follow:
17 a. There are inconsistencies in the inhalation risk factors used and in their uncertainties.
18 For example, the text (at p. 6-2) states that one would expect the unattached fraction to
19 be lower outdoors than indoors, which is consistent with the few measurements that
20 have been made. However, this reduction -- which would reduce the dose conversion
21 factor -- is not reflected in the geometric mean chosen for this value, nor is the
22 geometric standard deviation (GSD) increased to capture this uncertainty.
23 b The average outdoor radon concentration used in the calculations presented (0.3 pCiJL)
24 does not appear to be consistent with the UNSCEAR (1988) observation that a
25 population-weighted average value is about 0 14 pCiJL. In fact, the UNSCEAR value
26 falls outside the stated credibility interval of 0.19 to 4.6 pCIJL. A GSD of 1.3 is clearly
27 much too small for a concentration as uncertain as this.
28 c. Similarly, relatively few measurements are available to assess the average equilibrium
29 factor for outdoor exposure settings. Although the observed values fall in a small range,
30 the GSD of 1.05 implies greater accuracy in the value chosen (0 8) than is warranted.
3 1 d. Time spent outdoors is estimated to be 7.5%, on average. The variability in this factor
32 is much larger than a GSD of 1.1 would imply.
33 3. 1.3.3 Estimates of risks and uncertainties associated with water ingestion
34 The variability assumed for the amount of direct tap water consumed appears to be
35 biased high, at least as reflected in the analyses presented on pp. 5-26+.
-------
ii
I 3.2 Adequacy of Quantitative Uncertainty Analyses Regarding Risk Assessment
2 Are quantitative uncertainty analyses regarding risk assessment of water-borne radon.
3 including health risk analysis and exposure analysis, adequate’) At the suggestion of the EPA
4 staff, this question has been broken down into three subparts:
5 3.2.1 Are the basic methods used to propagate uncertainty acceptable?
6 4. Comment The Committee believes that the basic methods used to propagate
7 uncertainty are acceptable Proper consideration has been given to the possibility of
8 covanance, and the Monte Carlo simulation methods are state-of-the-art.
9 Discussion . In making this determination, the Committee considered the following.
10 1. The EPA acknowledged uncertainty in each step of the calculation.
I I 2. The EPA identified the sources of that uncertainty
12 3. The EPA examined uncertainty about best estimate values and about best estimate
13 distributions whereby the distributions represent variability in exposures and risk among
14 individuals.
IS 4 This latter approach whereby uncertainty is expressed about a best estimate distribution
16 of exposures is the current state-of-the -art in uncertainty analyses.
17 5. The EPA distinguished between variability and uncertainty, which past analyses have
18 not always done.
19 6. Perhaps most important, the EPA has also shown what the most dominant sources of
20 uncertainty are in the calculation. In the case of the multi-media exposures to radon,
21 the dominant source of uncertainty is associated with the uncertainty of translating an
22 exposure to radon to an estimate of health risk This risk conversion factor will
23 probably be the parameter which is most difficult to estimate accurately.
24 7. Nevertheless, the uncertainty associated with the dose to nsk conversion for radon,
25 although it is the dominant contributor to overall uncertainty, is still much less than the
26 uncertainty associated with other carcinogens that EPA regulates.
27 3.2.2 Are the probability density functions (PDFs) selected to describe Type A and Type
28 B uncertainty of each variable reasonable?
29 5. Recommendation The Committee recommends that particular attention be given to
30 more completely addressing uncertainty about the variance and shape of the probability
31 density functions (PDFs) that have been assumed by the EPA to represent variability in
32 exposures among individuals.
33 Discussion . The Committee believes that the general treatment of the PDFs used by the
34 EPA in its uncertainty analysis is adequate. subject to the points made below. The EPA
35 analysis considers two types of uncertainty First, it recognizes that different individuals
36 living in an area with the same level of radon in water will have different exposures, and
37 therefore risks, as a result of differences in household characteristics, water consumption
38 rates, and other factors. The uncertainty due to stochastic variability in the life me exposure
-------
12
per individual in the U.S. population (Type A uncertainty) differs from uncertainty
2 attributable to limitations in our knowledge about the quantities (mean, variance and shape)
3 that describe the true distribution of individual lifetime exposures (Type B uncertainty) This
4 latter uncertainty also reflects limitations that influence the average risk per individual.
5 While the Committee notes that the EPA analysis has not completely recognized these
6 distinctions, it believes that the EPA has captured the most important features of quantitative
7 uncertainty analysis and has adequately documented its choice of PDFs used in its analysis
8 for describing uncertainty about the true value of risk for the average individual.
9 3.2.3 Are there any important terms or assumptions that have not been adequately
10 evaluated?
11 6. Recommendation The Committee recommends that the EPA include in its uncertainty
12 analysis a qualitative discussion of known uncertainty variables which were not quantified in
13 the uncertainty analysis. These include the issue of a linear dose rate response extending to
14 low doses, the influence of smoking on increasing lung-cancer risks from radon, and the
15 effect of population mobility on the distribution of risks.
16 Discussion . The EPA is well aware that other model and parameter uncertainties may
17 be important but are difficult to quantify given current state of knowledge. Many of these are
I 8 mentioned in its draft documents, such as the issue of a Linear dose response extending to low
19 doses. Another issue that the Committee would like to see discussed qualitatively in the
20 document is the influence of smoking on increasing lung-cancer risks from radon. The risk
21 coefficient for airborne radon is an average value that underestimates the risk to smokers and
22 overestimates it for nonsmokers. The average risk value thus depends implicitly upon
23 assumptions about the nature of the relationship between lung cancer risk factors of smoking
24 and radon exposure, and on the fraction of smokers in the population.
25 The EPA assessment of radon in water is designed to apply to people whose water
26 supplies have the same radon content for their entire 70-year lifetimes. The Committee
27 recognizes that this design assumption is consistent with EPA policy to promulgate an MCL
28 for radon that is protective for those people who might live out their lives in a water service
29 area with radon at the maximum contaminant level. The Committee notes, however, that the
30 mobility of the population implies that not every person currently living in an area with
3 I especially high or especially low radon levels in water will remain there. The distribution of
32 radon exposures and risks therefore will not be the same as if every person remained in the
33 same area for a lifetime. In general, fewer people will have very high or very low exposures
34 and risks and more will have intermediate levels of risk than under the no-mobility
35 assumption. The effect of mobility on overall population risk (cancers per year in the United
36 States arising from radon in drinking water), in contrast, will likely be negligible because
37 most people moving from a high radon area to a low one will be replaced by people moving
38 in the other direction, except for any effect of net population migration within the country.
-------
13
1 3.3 Adequacy of Characterization of Risks from Water Treatment Facilities
2 3.3.1 1-las the EPA adequately characterized the risks introduced by radon that would be
3 released by aeration from water treatment facilities?
4 7. Recommendation In order to increase the scientific credibility of the results, the
5 Committee recommends that EPA consider upgrading the uncertainty analysis for the risks
6 associated with aeration for radon removal; however, the proposed revisions to the analysis
7 will not change the conclusion that the risk for a maximally exposed individual attnbutable to
S radon released from a water treatment facility will be no more than the average risk
9 attributable to 300pCiiL of radon in drinking water used in the home.
10
11 Discussion . The EPA has proposed air-stripping as Best Available Technology (BAT)
12 for achieving the proposed radon standard for drinking water where current levels exceed the
13 proposed standard. Recognizing that this technique would discharge much of the waterborne
14 radon to the a osphere, the EPA analyzed the risks of such discharges in terms of the risks
15 to a maximally exposed individual (MEL) living near the treatment facilities. The EPA also
16 projected the population risk or annual cancer incidence assuming that each water supplier
17 exceeding the proposed standard were to use air-stripping at a single location in order to bring
18 itself into compliance with the proposed standard (EPA, 1988, 1989).
19 The EPA reasoned that if the individual and population risks from the treatment
20 facilities were small relative to the risks avoided by applying the proposed standard, then a
21 comparative risk tradeoff would favor implementation of the standard To ensure that this
22 corn parison would not favor the proposed standard solely through differences in assessment
23 methods, the EPA estimated the risks attributable to water treatment by using two radiation
24 risk models, AJIRDOSE and MINEDOSE. Although the Committee has reservations about the
25 degree of validation of these models, the MINEDOSE model is thought to provide
26 conservative risk estimates. In the assessment of risk from water treamient, the EPA also
27 made assumptions that were the same as or more conservative than those used for assessing
28 the risks of radon in water used in the home. Specifically, the individual risks were
29 calculated for an MEl who was defined as exposed to the highest concentrations for the
30 longest possible time from discharges under worst-case meteorological conditions. The
31 Committee concurs that the set of assumptions chosen was generally quite conservative.
32 The MEl risks presented to the Committee ranged up to 8 x l0 , or about 4 times the
33 nominal value for the risk of 300 pC1IL radon in drinking water. However, this was a single
34 value derived from largely unrealistic assumptions, and more typical MEL risks appear to be
35 much lower, generally falling at or below the risk due to exposure to radon in drinking water
36 at 300 pCiIL.
37 The EPA also projected population risk using AIRDOSE and estimated total cancer
38 death rate of approximately 0. 1/yr. a value that is considerably less than the reduction of 80
39 cancer deaths/yr estimated to be achieved by implementing the proposed standard.
-------
14
The EPA conducted a semiquantitative uncertainty analysis of the MEl risk calculation
2 and concluded that upper bound risks would remain in the vicinity of 1 x i0 , given the
3 conservative nature of the nominal values. The uncertainty analysis was less rigorous and
4 more subjective than that for the risks of radon in drinking water. Although more rigor is
5 unlikely to change the conclusion, improvement of the uncertainty analysis would improve the
6 scientific credibility of the results.
7 3.3.2 Has the EPA adequately characterized the risks introduced by radon that would be
8 released from other types of water treatment facilities?
9 8.Recommendation If EPA determines that granular actived carbon will be used for
10 radon removal, the Committee urges EPA to thoroughly and completely analyze any potential
I I risk and/or disposal problems related to the use of granular activated carbon (GAC) for radon
12 removal from drinking water
13 Discussion Another technology for radon removal from drinking water is Granular-
14 Activated-Charcoal (GAC). Although GAC has not been designated a best available
15 treatment (BAT) for radon removal, in a draft technical memorandum from the Office of
16 Water (dated January 1993 and circulated to the RAC on February 18, 1993), EPA discussed
17 various issues related to the use of this technology which mentioned radioactivity
18 accumulation in the GAC (mostly lead-210). However, while the memorandum mentioned
19 the issue of GAC building up levels of radioactivity such that the residuals would require
20 disposal at a low-level-radioactive-waste (or naturally occurring radioactive material waste)
2 I repository, the memorandum was without sufficient data or analysis for the Committee to
22 evaluate this possibility and the implications of this problem
23 The Committee urges EPA to thoroughly and completely analyze any potential risk
24 and/or disposal problems related to the use of GAC for radon removal from drinking water
25 3.3.3 Occupational Exposures
26 9. Recommendation EPA did not provide an analysis of occupational exposures as a
27 result of water treatment for radon. The potential for such exposures appears to depend
28 heavily upon the choice of water treatment technology, and the Committee recommends that
29 such a comparative analysis be conducted for different technologies, such as aeration or
30 granular activated carbon filtration, especially in view of waste disposal problems that may
31 result from use of the latter technology.
32 Discussion . The EPA did not provide an analysis of potential radiation exposures to
33 workers in water treatment or ancillary facilities The RAC notes that in the case of aeration
34 techniques, proper ventilation of the water treatment facility should result in little increase in
35 radon concentrations and exposures to personnel. There should be no other significant
36 sources of radiation due to such treatment. However, the EPA has not ruled out treatment by
37 other means, including granular activated carbon filtration (GAC), in which case build-up of
38 radon progeny in the bed can result in an increased radiation field near the beds.
39 Furthermore, the handling and disposal of GAC beds containing radionudides has not been
-------
15
I analyzed nor, En fact, have provisions been made for such disposal in the event it is
2 necessary. In order to provide a complete risk analysis, the Committee believes that the EPA
3 needs to consider the possibility of worker exposures either to radiation or to chemicals (such
4 as those used as biocides in aeration facilities) resulting from some water treatment
5 technologies.
-------
16
I 3.4 Other Scientific Issues
2 3.4.1 Recommended extensions of the risk and uncertainty analysis and publication of
3 results in peer-reviewed journals
4 10. Recommendation The Committee recommends that the document include a
5 summary of the results of the uncertainty analysis regarding the contribution of the various
6 exposure pathways to the overall radon risk to individuals and to the general population
7 This summary should also highlight the major sources of uncertainty contributing to the total
8 uncertainty in the risk estimate for each pathway. Such a discussion would provide the
9 information necessary to factor uncertainties and variabilities into the cost-benefit analysis for
10 the proposed regulation and to calculate a range for the estimates of costilife saved. (3.4.1)
I 1 Discussion . One aspect that was lacking in the reviewed document was a summary and
12 interpretation of the uncertainty analysis for radon in drinking water. The Committee has
13 studied the results presented by the EPA and offers the following interpretation.
14
15 341.1 Individual risks
16 The following table lists the unit risks attributable to drinking water by inhalation and
17 ingestion pathways, including the 90% confidence interval around the median, the upper-
18 bound 95th percentile, and the lower-bound 5th percentile for risk.
19 Table 1. Unit Risk Boundaries for Exposure to Radon in Drinking Water
20 (Fatal cancers/person/year per pCi/L)
21
22
23
24
25
26
27
28
29
30
5th
perccit i ie
LOWeL Bound
5th
peLCent i
Median
Median
95th
peicent i I c
Medial)
95th
PeLcent i Ic
Uppei Bound
inha lat on
1 6 x 1O 0
1 1 x 1O
2 7 \ 1O
b 3 x 1O
4 2 x IO
Inciest ion
I 2 x 10 _ia
x i0i
i 7 X lO
b 5 i l0
2 0 X I0
The nominal unit risk in the proposed rule is 9.4 x l0 fatal cancers/person/year per
pCi/L. This nominal risk can be compared to the median inhalation and ingestion risks from
radon in drinking water shown in Table 1. The nominal risk is larger than the inhalation
median risk by a factor of 3.5 and is larger than the ingestion median risk by a factor of 5.5.
Therefore, the combined unit risk from inhalation and ingestion exposure will be <3.5, and
well within the range encompassed by the 90% confidence interval of risk about the median.
The same comment applies to the nominal unit nsk presented in Chapter 3 of the reviewed
document
31 34.1.2 Population risks
32 The estimates of cancer fatalities due to exposure of radon in drinking water are based
33 upon 81 million people being exposed. This number was presented to the Committee during
34 a briefing on 2/17/93, and comes from a preliminary contractor report on occurrence of radon
-------
17
in drinking water (Wade Miller, 1992). That report is being reviewed by the EEC of the
2 SAB. Any changes in that estimate will affect the results presented below.
3 Table 2 Cancer Fatalities per Year due to Exposure to Radon
4
( )
l0
E’ JUSULe Pdthwey
5th Petcent L le
Med L an
Medi ar
p5th PeLcent 1 I?
Med an
UppeL
Puuiicl
I nha let ion Uue to
Watci TLeatmeIit
— — —
— — —
— ——
I
Itijiat loin tLOm
i Iiii Watei
48
105
233
-—-
Ii j -nt ion ioin
Di ii mc ; Macen
1
53
1 6
-—-
Inhalat loin riom
‘ iitctooi Au
280
57
1500
m nhalatmon tiom
tililoot Alt
b79 0
14 ,410
30, )50
15 The estimated lung cancer deaths attributable to inhalation exposure to radon in drinking
16 water range from 48 to 233 per year. The estimated fatal cancer cases attributable to
17 ingestion exposure to radon in drinking water raiige from 19 to 166 per year. Therefore,
18 estimated total fatal cancer cases attributable to waterborne radon will be about a quarter of
19 the risks associated with exposure to radon in outdoor air, and about one percent of the risks
20 associated with exposure to radon in indoor air and of the total risks attributable to exposure
21 to radon by all pathways. These calculations also indicate that population risks from
22 exposure to radon in drinking water are similar to or higher than those normally addressed by
23 regulation of chemical pollutants in drinking water. Although the risk attributable to
24 inhalation and ingestion of radon in drinking water were apportioned equal weight in the
25 calculation of the nominal value in Chapter 3, the weight obtained as a result of the
26 uncertainty analysis is approximately two-thirds for inhalation and one-third for ingestion.
27 This last set of values is similar to those presented in the Proposed Rule (EPA, 1991).
28 3.4.2 Estimate of Lives Saved
29 I 1. Recommendation The Committee recommends that the EPA extend its population
30 risk assessment and uncertainty analysis to obtain an estimate of the lives that would be saved
3 1 by the proposed maximum contaminant level, using for radon concentration the same
32 assumptions as were used to calculate present-day risks but using a lognormal probability
33 density function truncated at the maximum contaminant level.
34 Discussion . The Committee could not carry Out an analysis of the estimated number of
35 lives that would be saved by the Proposed MCL of 300 pCi/L because no uncertainty analysis
36 was done on the number of cancer fatalities projected for the rule in place. The Committee
37 recommends that a population risk assessment and uncertainty analysis be carried out, using
38 the same assumptions as were used to calculate present-day risks but using for radon
39 concentration a lognormal PDF truncated at the proposed MCL. An uncertainty for the
40 tolerance in the measurement of radon as described in the section regarding monitoring of the
41 Proposed Rule should also be factored into this uncertainty analysis. From these calculations,
42 one would obtain a 90% confidence interval for the cancer fatalities that would remain with
43 enforcement of the proposed MCL, and the difference between the values in Table 2 and
-------
18
those calculated with the truncated PDF would yield a range of lives saved. This analysis
2 would then allow the persons conducting the cost-benefit analysis to factor these uncertainties
3 and variabilities into their calculauons, leading to a range of costs per life saved. The
4 Committee believes that this extension to the EPA’s uncertainty analysis would enhance the
5 usefulness of the document reviewed.
6 3.4.3 Peer Review and Publication
7 12. Recommendation The Committee urges the EPA to submit its risk analyses for
publication in appropriate journals which would provide peer-review and recognition that the
9 EPA’s science is of high-quality and that it becomes part of the mainstream of scientific
10 criticism, revision, and acceptance (or rejection). Publication will also assist in raising
I 1 awareness within the scientific community to the risk issues associated with radon.
12 Discussion . The Committee believes that overall, the use of the peer-reviewed literature
13 as both a source of data and information and also as a method of disseminating the EPA’s
14 own scientific work is an important means by which the EPA and the public can be assured
15 that the best science is being used or produced. In this particular case, the estimate of the
16 ingestion risk due to radon in drinking water rests heavily upon data and analyses that have
17 not been published and therefore have not been broadly circulated within the scientific
18 communIty Reliance upon such results should be done with considerable caution.
19 Although publication in peer-reviewed journals does not, by itself, assure infallibility, itis
20 the only generally recognized means by which scientific work gets accepted by members of
21 the scientific community. In seeking to improve the quality and the scientific acceptability of
22 its science, the EPA should encourage its scientists to submit their work for peer-reviewed
23 publication. The work and methodologies presented here mark an important advance in the
24 risk and uncertainty analyses undertaken by the EPA and are certainly worthy of such
25 publication.
qc
-------
‘9
4. POLICY CONSIDERATIONS
2 4.1 The Importance of Quantitative Uncertainty Analysis
3 The Radiation Advisory Committee has long encouraged the use of integrated
4 quantitative uncertainty analysis in a variety of EPA assessments. The Committee is
5 extremely pleased to see that the EPA has done such analysis in this case. The Committee
6 applauds EPA for its timely incorporation of a full quantitative uncertainty analysis for each
7 pathway in its assessment and hopes that the use of quantitative uncertainty analysis will
8 become a routine part of all EPA assessments, not only those associated with radiation risks.
9 This information should be a valuable aid in guiding EPA in its consideration of possible
10 regulatory strategies.
II The Committee believes strongly that the explicit disclosure of uncertainty in quantitative
12 risk assessment is necessary any time the assessment is taken beyond a screening calculation.
13 Screening risk assessments typically involve only point estimate calculations. The
14 assumptions used to derive these point estimates are generally biased on the conservative side
15 to ensure that the true risk to individuals will not be underestimated. Screening calculations
16 are thus useful for identifying situations that are clearly below regulatory risk levels of
17 concern. They can be grossly misleading in terms of indicating the need for regulatory
18 action.
19 The need for regulatory action must be based on more realistic estimates of risk.
20 Realistic risk estimating, however, requires a full disclosure of uncertainty. The disclosure of
2 1 uncertainty enables the scientific reviewer, as well as the decision-maker, to evaluate the
22 degree of confidence that one should have in the risk assessment The confidence in the risk
23 assessment should be a major factor in determining strategies for regulatory action.
24 Large uncertainty in the risk estimate, although undesirable, may not be critical if the
25 confidence intervals about the risk estimate indicate that risks are clearly below regulatory
26 levels of concern. On the other hand, when these confidence intervals overlap the regulatory
27 levels of concern, consideration should be given to acquiring additional information to reduce
28 the uncertainty in the risk estimate by focusing research on the factors that dominate the
29 uncertainty. The dominant factors controlling the overall uncertainty are readily identified
30 through a sensitivity analysis conducted as an integral part of quantitative uncertainty
3 1 analysis. Acquiring additional data to reduce the uncertainty in the risk estimate is especially
32 important when the cost of regulation is high. Ultimately, the explicit disclosure in the risk
33 estimate should be factored into analyses of the cost-effectiveness of risk reduction as well as
34 in setting priorities for the allocation of regulatory resources for reducing risk.
35 4.2 The Relative Risk of Radon in Drinking Water
36 En its January 29, 1992, Commentarv Reducing Risks from Radon: Drinking Water
37 Criteria Documents (EPA-SAB-RAC-COM-92-0O3), the Committee noted that the radon risk
38 reduction situation reflects the fragmentation of environmental policy identified in Reducing
39 Risk (SAB-EC-90-02l). Because radon in drinking water is a very small contributor to radon
40 risk except in rare cases, the Committee suggested that the EPA focus its efforts on primary
9k
-------
20
I rather than secondary sources of risk. Within the bmitations of the data currently available,
2 the EPA has now successfully prepared a scientifically credible multi-media risk assessment
3 for regulatory decision-making on radon. The Committee’s agreement with the principle of
4 radiation protection optimization and in the concepts articulated in Reducing Risk lead it to
5 note once again that radon in drinking water represents only a small fraction of radon
6 exposure and risk compared to radon in indoor air from non-water sources. The emphasis on
7 various radon exposure reduction methods--whether for radon from water or non-water
sources--is a policy choice for which scientific analysis is only one of many important inputs.
9 4.2 Harmonizing
10 In its May 8, 1992 Commentary on Harmonizing Chemicaj and Radiation Risk Reduction
I I Strategies (EPA-SAB-R CCOM.92.O07) the Committee brought to the EPA’s auention the
12 need for a more coherent policy for making risk reduction decisions with respect to radiation
13 and chemical exposures. The control of radon in drinking water presents a situation where a
14 radiological contaminant being regulated by a paradigm developed for chemicals yet radon in
IS drinking water represents only a small fraction of radon exposure. The Committee
16 appreciates the EPA’s difficulty in establishing a coherent risk reduction strategy under the
17 variety of statutes governing EPA and acknowledges that harmonization does not necessarily
18 imply identical treatment. However, the Committee urges the EPA to explain clearly why the
19 risks from radiation (in this case radon m indoor air) and chemicals (in this case radon in
20 drinking water) are treated differently under specified conditions and in specified exposure
2 I settings. The Committee urges EPA, the Congress and the public to carefully consider how
22 chemical and radiation risks are being treated in this case.
-------
2I
1 5. REFERENCES
2 5.1 Documents Received by the Radiation Advisory Committee During this Review
3 A. DOCUMENTS RECEIVED BEFORE THE FEBRUARY 17.19 PUBLIC MEETING
4 Documents Provided by EPA
5 I. Departments of Veterans Affairs and Housing and Urban Development, and Independent
6 Agencies Appropnation Act, 1993, PUB. L. 102-398, Section 519, 106 STAT 1618
7 (1992)
8 2. Draft 2 “Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking
9 Water’ prepared by U S. EPA Office of Science and Technology, Office of Radiation and
10 Indoor Air, Office of Ground Water and Drinking Water, and Office of Policy Planning
Ii and Evaluation, January 29, 1993
12 3. Proposed Revisions in EPA Estimates of Radon Risks and Associated Uncertainties
13 4. An Analysis of the Uncertainties in Estimates of Radon-Induced Lung Caner by Jerome
14 S Puskin in Risk Analysis Volume 12, Number 2. 1992
15 5. Response to SAB Radon Comments
16 6. Preliminary Risk Assessment for Radon Emissions from Dnnking Water Treatment
17 Facilities, a memorandum from Warren D. Peters and Christopher B. Nelson to Stephen
18 W. Clark, June 28, 1988
19 7. An Analysis of Potential Radon Emissions from Water Treatment Plants Using the
20 MINEDOSE Code, a memorandum from Parc. J. Parrotta to Greg Helms, November 22,
21 1989
22 8. Proposed Methodology for Estimating Radiogenic Cancer Risks (no author or date given)
23 9. Cancer Fatalities from Waterborne Radon (Rn-222) by Douglas J. Crawford-Brown in
24 Risk Analysis, Volume 11, Number 1, 1991
25 Public Comment
26 10. Letter re National Primary Drinking Water Regulations: Radionuclides (Radon) [ WH-FRL
27 3956-4] from John H. Sullivan of the American Water Works Association to Honorable
28 Carol Browner, Administrator of the Environmental Protection Agency, January 26, 1993.
29 There were 27 Appendices to this letter
30 1. EPA Technical Support Document for the 1992 Citizen’s Guide to Radon,
31 EPA 400-R-92-0ll (May 1992)
32 2. “Harmonizing Chemical and Radiation Risk-Reduction Strategies - A Science
33 Advisory Board Commentary,’(May 18,1992)
34 3. Letter from SAB Chairman Raymond Loehr to EPA
35 Administrator William Reilly Re ‘Radionuclides in
36 Drinking Water” (EPA-SAB-RAC-91-X)O() (September 1991)
37 4 ‘An SAB Report: Radionuclides in Drinking Water”
38 (EPA-SAB-RAC-91-009) (December 1991)
39 5. Letter from SAE Chairman Raymond Loehr to EPA
40 Administrator William Reilly Re “Reducing Risks
-------
22
1 from Radon; Drinking Water Criteria Documents,”
2 (EPA-SAB-RAC-COM-92-003) (January 29, 1992)
3 6 Letter from SAB Chairman Raymond Loehr to EPA
4 Administrator William Reilly Re: ‘Status of EPA
5 Radionuclides Model” (EPA-SAB-RAC-COM-92-001)
6 (January 9, 1992)
7 7. SAB, “Review of the office of Drinking Water’s
8 Assessment of Radionuclides in Drinking Water and
9 Four Draft Criteria Documents” (July 1987)
10 8. Letter from SAB Chairman Raymond Loehr to EPA
11 Administrator William Reilly Re “Review of Draft
12 Criteria Documents for Radionuclides in Drinking
13 Water” (EPA-SAB-RAC-92-0009) (January 9, 1992)
14 9. Letter from SAB Chairman Raymond Loehr to EPA
15 Administrator William Reilly Re: “Revised Radon
16 Risk Estimates and Associated Uncertainties” (EPA-
17 SAB-RAC-LTR-92-003) (January 9, 1992)
18 10. Puskin, Jerome, “An Analysis of the Uncertainties in
19 Estimates of Radon-Induced Lung Cancer,” Risk
20 Analysis, Vol. 12, No. 2, P. 277 (1992)
21 11. SENES Consultants Limited Memorandum Re: “Exposure and Risk
22 from Radon Released in Showers” (December 3, 1992)
23 12. Fensterheim, Robert, Stolwijk, Jan, “Critique of
24 Hess and Bemhardt Radon Shower Exposure Study,”(l992)
25 13. Testimony of Jonathan M. Sa.met before the
26 Subcommittee on Transportation and Hazardous
27 Materials, House Energy and Commerce Committee
28 (June 3, 1992)
29 14. Neuberger, John S., “Residential Radon Exposure and
30 Lung Cancer: An Overview of Published Studies,”
31 Cancer Detection and Prevention, Vol.15, Issue 6,
32 (1991)435-443
33 15. Neuberger, John S., et at, “Residential Radon
34 Exposure and Lung Carcer: Evidence of an Inverse
35 Association in Washington State,” Journal of
36 Environmental Health, Nov/Dec. 1992, 23-25
-------
23
I 16. “Proposed Guideline for Radon-222 in Drinking
2 Water,” prepared by SENES Consultants Limited for
3 Health Protection Branch, Health and Welfare Canada
4 (March 1992)
5 17. Draft of SAB Radiation Advisory Committee Comments
6 on EPA’s ‘Suggested Guidelines for the Disposal of
7 Drinking Water Treatment Wastes Containing
8 Naturally-Occurring Radionuclides” (July 6, 1992)
9 18. Testimony of Dr. Jill Lipoti on HR 3258, the “Radon
10 Awareness and Disclosure Act of 199111 before the
I I House Subcommittee on Transportation and Hazardous
12 Materials (June 3, 1992)
13 19. Factor Analysis for Differences Between EPA and RCG
14 Compliance Cost Estimates
15 20 Table Comparing Compliance Costs for A Radon MCL of
16 300 pcill; Letter to Editor and Response in American
17 Water Works Association Journal
18 21. Comments of the State of Idaho Department of Water
19 Resources (May 18, 1992)
20 22 Letter from Dr. Alvin Young, Chairman of Committee
21 on Interagency Radiation Research and Policy
22 coordination, to Dr. Donald Henderson, Office of
23 Science and Technology Policy (May 21, 1992)
24 23. Testimony of Dr. Jan Stolwijk before the House
25 Subcommittee on Transportation and Hazardous
26 Materials (June 3, 1992)
27 24 Valentine, Richard, “Radon and Radium From
28 Distribution System and Filter Media Deposits/” AWWA
29 Water Quality Technology conference, Toronto (1992)
30 25 Comments of the State of New York Department Health
31 to EPA (February 12, 1992)
32 26. “Evaluation of the Impact of a Radon-222 MCL on
33 Small Water Systems,” by John E. Reanier, Alabama
34 Rural Water Association (May 10, 1992)
35 27. Comments of the Association of State Drinking Water
36 Administrators (November 19, 1991)
37 1 I. Letter to Administrator Browner and three SAB Chairs from Bill Mills, Steve Hall, and
38 Tom Levy of the Alliance for Radon Reduction, February 2, 1993
/00
-------
24
I B. DOCUMENTS RECEIVED AT THE FEBRUARY 17-19 PUBLIC MEETING
2 Documents Provided by EPA
3 1 Draft Summary (no date or author given, appears to be a draft summary for the
4 ‘Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking Water’
5 2 Overheads: Briefing for SAB on Multimedia Risk Assessment of Human Exposure to
6 Radon, Office of Science and Technology, Office of Radiation and Indoor Air, Office of
7 Policy, Planning, and Evaluation, Office of Ground Water and Drinking Water.
8 3 Overheads: Risk Assessment for Radon Emissions from Drinking Water Treatment
9 Facilities, EPA Office of Radiation and Indoor Air, February 17, 1993
10 4. Overheads. Cancer Risks Associated with Radon in Drinking Water--Uncertainty and
11 Vanability Analysis
12 5. ‘Review of Risk Assessments of Radon Emissions from Drinking Water Treatment
13 Facilities” from Christopher Nelson ORIA to Mark Parrotta ODW
14 6. Radon Documents for SAB Review, a memorandum from Nancy Chiu of OST/OW to
15 William F Raub, Science Advisor
16 7. Draft Technical Memorandum: Problems with the Use of GAC for Radon Removal,
17 printed date is January 1993 (handwritten date is 2/11)
18 Public Comment
19 8 Review of Technical Justification of Assumptions and Methods Used by the
20 Environmental Protection Agency for Estimating Risks Avoided by Implementing MCLs
21 for Radionuclides by S.C. Morris, M.D. Rosw, S. Holtzman, and A.F. Meinhold and
22 Brookhaven National Laboratory, November, 1992
23 9 Letter from Edward J. Schmidt to Comments Clerk-Radionuclides, Subject Comments on
24 National Primary Drinking Water Regulations: Radionuclides Proposed Rule, 4OCFR
25 Parts 141 & 142, Thursday, July 18, 1991, September 30, 1991
26 ID Letter to James R. Elder from Raymond F. Pelletier, Office of Environmental
27 Guidance, U.S. Department of Energy, January 27, 1993
28 C. DOCUMENTS RECEIVED SUBSEQUENT TO THE FEBRUARY 17-19 PUBLIC
29 MEETING
30 Documents Provided by EPA
31 I. One-page note to Kathleen Conway from Jan Auerbach, February 23, 1993
32 2 Note to Kathleen Conway, RAC DFO from Nancy Chui OGWDW, faxed to the
33 Radiation Advisory Committee, March 10, 1993
34 Public Comment
35 3. Letter to the SAB Radiation Advisory Committee from Frederick W. Pontius of the
36 American Water Works Association, February 24. This letter had seven enclosures:
13I
-------
25
a. Lognormal Distributions for Water Intake by Children and Adults, by Ann M.
2 Roseberry and David. E. Burmaster in Risk Analysis, Volume 12, Number 1,
3 1992
4 b Distribution and Expected Time of R e sidence for U.S. Households by Milton
5 Israeli and Christopher B Nelson in Risk Analysis, Volume 12, Number 1, 1992
6 c. Review of Risk Estimates for Inhalation of Radon Progeny by Miners
7 Presentation by the Atomic Energy Control Board of Canada (ACB) before the
8 ICRP Main Commission, pnnted date is November 1992, there is also a stamped
9 date of February 12, 1993
10 d. A Cohort Study in Southern China of Tin Miners Exposed to Radon and Radon
11 Decay Products by Xuan Xiang-Zhen, Jay. H. Lubin, and others in Health
12 Physics, Volume 62. Number 10, pages 120-131, February 1993
13 e. Contribution of Waterborne Radon to Home Air Quality, prepared by Arun K.
14 Deb of Roy F., Eston, Inc. for the AWWA Research Foundation, undated
15 f Final Report Risk and Uncertainty Analysis for Radon in Drinking Water
16 prepared by Douglas J. Crawford Brown for the American Water Works
17 Association.
18 g. Proposed Guideline for Radon-222 in Drinking Water prepared by SENES
19 Consultants Limited for the Health Protection Branch of Health and Welfare
20 Canada, March 1992
2 1 4. Letter to the Radiation Advisory Committee from Douglas Crawford-Brown of the
22 University of North Carolina, March 2, 1993
23 5. Letter to Dr. Genevieve Matanoski from Bill Mills, Steve Hall and Tom Levy of the
24 Alliance for Radon Reduction, March 11, 1993
25 6. Letter to Dr. Genevieve Matanoski from Robert J. Fensterheim, consultant to the Alliance
26 for Radon Reduction, March 16, 1993
27 7. Fax from Robert J. Fensterheim referencing Brown-Senate Letter and Naomi Harley
28 Study, March 16, 1993. This fax included both a March 11, 1993 letter from nine
29 senators to Administrator Carol M. Browner and A Biokinetic Model for the Distribution
30 of Rn-22 Gas in the Body Following Ingestion by Naomi H. Harley and Edith S
31 Robbms, March 12, 1993
32 8. Letter to Dr. Vern Ray, Chairman of the Radon Engineering Cost Subcommittee from
33 Stephen Hall of the Association of California Water Agencies, March 22, 1993
-------
26
1 5.2 Science Advisory Board Reports of Potential Interest
2 1. Report of the Scientific Basis of EPA’s Proposed National Emission Standards for
3 Hazardous Air Pollutants for Radionuclides: A report of the Subcommittee on Risk
4 Assessment for Radionuclides, August 1984 (There is no report number because this
5 report was produced before the SAB developed a report numbering system.)
6 2. Radionuclides in Drinking Water (SAB-RAC-87-035)
7 3. Effective Dose Equivalent Concept (SAB-RAC-88-026)
8 4 Radon Risk Estimates (SAB-RAC-88-042)
9 5. Radionuclides NESHAP (SAB-RAC-89-003)
10 6. EEC Mathematical Models Resolution (SAB-EEC-89-O1)
Il 7. Radionuclides NESHAP (SAB-RAC-89-024)
12 8. Radon Risks (SAB-RAC-91-LTR-OOl)
13 9. Status of EPA Radionuclide Models (EPA-SAB-RAC-COM-92-OO)
14 10 Revised Radon Risk Estimates and Associated Uncertainties
15 (EPA-SAB-RAC-LTR-92- 0 03)
16 II Criteria Documents for Radionuclides in Drinking Water
17 (EPA-SAB-RAC-92-0 09)
18 12 Reducing Risks from Radon/Drinking Water Criteria Documents
19 (EPA-SAB .RAC-COM- 0 03)
20 13. Harmonizing Chemical and Radiation Risks (EPA-SAB-RAC-COM-92-007)
21 14. Drinking Water Treatment Wastes Containing NORM (EPA-SAB-RAC-LTR-92-0 18)
22 15. Radon in Water: Consultation (EPA-SAB-RAC-CON-92-002)
-------
27
5.3 Literature cited:
2 Correia JA. Weise SB, Callahan Ri, Strauss HW, 1987. The kinetics of ingested Rn-222 in
3 humans determined from measurements with Xe-133. Massachusetts General Hospital,
4 Boston MA. unpublished report Prepared for Health Effects Research Laboratory, U S
5 EPA. Report No. EPA/600/l-87/013.
6 Crawford-Brown DJ, 1991. Cancer fatalities from waterborne radon (Rn-222) Risk Anal.
7 11135-143.
8 EPA. 1988. Preliminary Risk Assessment for Radon Emissions from Drinking Water
9 Facilities,’ memorandum from Warren Peters and Christopher Nelson to Stephen Clark,
10 June 28, 1988.
I I EPA, 1989. “An Analysis of Potential radon Emissions from Water Treatment Plants using
12 the MINEDOSE Code,” memorandum from Marc Parrotta to Greg Helms, November 22,
13 l9 9
14 EPA, 1991 “Notice of Proposed Rulemaking for Radionuclides in Drinking Water”
IS EPA, 1989. “Draft 2 ‘Uncertainty Analysis of Risk Associated with Exposure to Radon in
16 Drinking Water” prepared by U.S EPA Office of Science and Technology, Office of
17 Radiation and Indoor Air, Office of Ground Water and Drinking Water, and Office of
18 Policy Planning and Evaluation, January 29, 1993
19 McKone, TE, 1987. Human exposure to volatile organic compounds in household tap water.
20 the indoor inhalation pathway. Environ. Sci. Technol. 21:1194-1201
2 I UNSCEAR, Sources, Effects and Risks of Ionizing Radiation, United Nations Scientific
22 Committee on the Effects of Atomic Radiation, United Nations: New York, 1988, p. 64
23 Wade Miller Associates, 1992, Draft addendum to the occurence and exposure assessments
24 for radon, radium-226, radium-228, uranium, and gross alpha particle activity in public
25 drinking water supples. EPA contract No. 68-CO-0069 September 30, 1992.
/o’ ,1
-------
28
APPENDIX A. Brief Chronology of Relevant SAB Reports
2 In 1984 a specialized ad hoc Subcommittee of the Science Advisory Board
3 reviewed the scientific basis for EPA’s proposed national emissions standards for
4 hazardous air pollutants for radionuclides That report led to the formation of the
5 Radiation Advisory Committee to “review risk assessments for radiation standards’ . The
6 report also stated,”A scientifically defensible risk assessment for radionuclides should
7 address at least five major elements. These include 1) identification of the significant
8 . sources, 2) a description of the movement. . from a source . . . to people; 3)
9 calculation of doses; 4) estimation of. . . health effects, and 5) incorporation of estimates
10 of uncertainty into elements 1-4. . . .“ The routine incorporation of uncertainty analysis
II into risk assessments has been a recurring theme in Radiation Advisory Committee
12 reports.
13 In the summer of 1986, the Drinking Water Subcommittee of the Radiation
14 Advisory Committee reviewed the Office of Drinking Water’s Assessment of
15 Radionuclides in Drinking Water and Four Draft Criteria Documents, (SAB-RAC-87-
16 035). This Subcommittee did not explicitly address uncertainty analysis. While
17 recommending some improvements in science and presentation, the Subcommittee
18 concluded, “that the Office of Drinking Water has developed scientifically comprehensive
19 assessment documents.” This report was transmitted to the Administrator July 27, 1987
20 In 1988 and 1989 reviews of revisions to the scientific basis for the
21 radionuclides NESHAP, the Radiation Advisory Committee again raised concerns about
22 quantitative uncertainty analysis The cover letter of the November 10, 1988 report
23 (SAB-RAC-89-003) highlighted three findings for serious attention by the EPA,
24 including, “To date, EPA’s treatment of modeling uncertainties has been qualitative rather
25 than quantitative although state-of-the-art methods for estimating uncertainty are
26 available.” The June 30, 1989 report (SAB-RAC-89-024) noted in the cover letter (p.2),
27 “ . . . the Radiation Advisory Committee and the Science Advisory Board has repeatedly
28 urged the use of best estimates and ranges in the specifications of risk, and a detailed
29 explanation of the uncertainties in the estimates themselves.”
30 On January 13, 1989, the SAB transmitted to the Administrator the
31 Environmental Engineering Committee’s Resolution on the Use of Mathematical Models
32 by EPA for Regulatory Assessment and Decision-Making (EPA-SAB-EEC-89-0 12). The
33 Committee (p.1) had reviewed “a number of integrated environmental modeling studies”
34 and “noted a number of problems” including, “a lack of studies quantifying the
35 uncertainties associated with model predictions, and concurrently, the potential misuse of
36 particular uncertainty analysis techniques.” The resolution’s fourth recommendation (p 3)
37 was, “Sensitivity and uncertainty analysis of environmental models and their predictions
38 should be performed to provide decision-makers with an understanding of the level of
39 confidence in model results, and to identify key areas for future study.”
40 In the summer of 1990, the Radionuclides in Drinking Water Subcommittee
41 of the Radiation Advisory Committee reviewed draft criteria documents for radionuclides
-------
29
I in drinking water, including those for uranium, radium, radon, and a combined document
2 on beta particles and gamma emitters.
3 The Subcommittee found that, “The overall quality of the four draft criteria documents
4 was not good. . . recommendations from a 1987 Science Advisory Board report on its
5 review of the standards for radionuclides in drinking water (SAB-RAC-87-035) had not
6 been addressed. Nor did the new criteria documents address recommendations from other
7 available SAB reports that are directly relevant (such as SAB-RAC-88-026 and SAB-
8 EEC-89-0 12). . . Uncertainties associated with the selection of particular models,
9 specific parameters used in the models, and the final risk estimates are not adequately
10 addressed in any of the documents. Although the review was conducted in 1990 and
II draft reports circulated at that time, this SAB report was not transmitted to the
12 Administrator until January 9, 1992 (EPA-SAB-RAC-92-009)
13 In the summer and fall of 1991, the Radiation Advisory Committee
14 received revised criteria documents and declined to review them. It did, however,
15 produce a commentary which noted (p.4) that, “Although each criteria document now
16 includes a chapter discussing uncertainty, the content of the chapters is very qualitative
17 and is not the rigorous technical analysis envisioned by the Committee.’ In its section on
18 policy considerations, the Committee also noted (p.3) that, “radon in drinking water is a
19 very small contributor to radon risk except in rare cases and the Committee suggests the
20 EPA focus its efforts on primary rather than secondary sources of risk” This
21 commentary was transmitted to the Administrator January 29, 1992 (EPA-SAB-RAC-
22 COM-92-003)
23 The January 9 and 29, 1992 reports also contained other advice relevant to
24 the scientific assessment of the risk of radon in drinking water. Additionally, the January
25 29, 1992 report provided policy-related comments on radon in light of the SAB report,
26 Reducing Risk . A May 8, 1992 Radiation Advisory Committee commentary,
27 “Harmonizing Chemical and Radiation Risk Reduction Strategies,” described chemical
28 and radiation risk reductions paradigms, discussed the difficulties of applying a paradigm
29 developed for one type of contaminant to the other, and recommended harmonization.
30 In the winter and spring of 1992, the Committee conducted a review of the
3 I EPA’s, “Suggested Guidelines for the Disposal of Drinking Water Treatment wastes
32 Containmg Naturally-Occurring Radionuclides” dated July 1990. The Committee found
33 that such guidelines were needed because of the potential radiation doses to treatment
34 plant workers and the public. However, the 1990 guidelines did not fully assess the
35 magnitude of nsk from exposure to treatment wastes, nor did the document specify
36 whether the radiation exposures to workers should be considered as occupational
37 exposures or viewed against dose limits for the general public, a decision which will have
38 considerable bearing on any final guidelines. This letter report was transmitted to the
39 Administrator September 30, 1992 (EPA-SAB-RAC-LTR-92-0l8).
-------
30
I APPENDIX B. Insert Chafee-Lautenberg
107
-------
• mher 25, 1992
-d
( C I-
a Id
—s
n a1
, be
the
eq)
0, - t o
ho:-
nce
Act
vital
Set-
1993
ient
re-
-on-
On-
5 7O-
Re-
a,,
trot
trot
3 of
Its
‘en-
a.nd
(01-
aid
it n l
li ly
‘ia n
f of
Its
en-’
(0 1-
4I(
2CC
‘5
ted
ep-
Va-
C, ’-
nfl
it
its
‘C-
od
31-
it-,
3’-
-4
r ..
Cl
‘-7
C-
S 1510’
re(luli ’emeatS. W’Lt SL’OC’ .I
ssna:l com.r ’ iunitie
(5) JZ ,f fitonct J and cca CO4YIdIV
drmidvw WCLCT 7Y5te,Tu to n.stci! Ire tnl .-nr fc
ctlt:tcs n.eeded to a suie Co wiz dri,’,
ing water :tan4ards and opt1on to
CCl ’ ipItanCE With SUCh vidCrd3 with sp
empha.rti on nnizll con —,,un(ctt,
(6) the financial ar4 tech71Ca1
States to Im pleT nent the drinking waLCr prorcr
lncludIr.g options for Lncraulng funding
State programs, and
(7) in ’wvat.Lu and citernadve mnethod to
crno . .se the financial and t hni 1 oapc tty
drlnklr ,g wJ .tcr sistans and the States to assu
effective isnp.’enentatlo ’t oj such Act
(b) M0R. ropJuM AND RrPoRr ON RADII
NUCLIO&S IN DRINKING WATSit —(1) The Admi,
totrator of the Environ m ntaJ Protection Agent
shall conduct a risk a.uessment of radon Cor t
ering (A) the rule of o4verse human h Zzh e
fects associated wit / i crposure to various pau
ways of radon. (8) (lie costs of COntrCt4fflg I
llltIlQatlrlQ erposure to radon, and (C) the Ca,
for radon control or rnhttgaeiovi e-perienced
households and communities, tiwluding Li
costs experienced by email cornmuni&s a.s ii
result of such regulation. Such an evalv.a,tic
shall c n,ldev the risks posed by the tr .bnc ’
or disposal of any bastes produced by tout.
ireabnent. The Science Ad ,t -tsorij Board 3halZ r.
view the Agency’s study and subirdt a no
oznnwitdatton to the Adjnintrtrntor on its fl’u
ings. The Administrator shaLl report the AdriL,
(stralor’s findings and tile Science Adviao’
Board recer ,ursendatlon to the Senate Co,nndct.
on Envirorunent and Public Works and Li
House Comini lice on nergi’ and Conunevce N
later than July 31. 1993. the Adednlsrator s/ia
publish the Adminl tratora at dy and risk a
sessmeni and the Science Advisory Board ‘a
onunendation.
(2) The Adnu’ijstsatrj, is directed, If additlcns
time is reqtdre4 to establish the radon standan
to seek an extension of the deadline COn taint
In the judiciaily-(mposed consenc decree far pr
mulgatlon of the rado, standard to a aute ,u
later than October 1, 1993.
(c) Small System MonUorlr.c Cost Redui
tlo ,t .—Wtth respect to onitoring re4,sfrer .en
for oroanic chenicaL ,, pesticides PCBr. or a,
reg-ulated contnnnnonts promulgated in Januar
1991 (known as the Phase 11 rule), the Admiru,
trator or a primacy State may modify such ri
qulrernentz to provide that any drinking watt
system serving a population of lass than 13 (
persons shall not be required to condu ct nods
tto ,uii qua rterly nwnf taring for a specific on
tgmlnant or con lamtn,ints prior to October 1
1993, If monitoring for airy or,e quavt .er con-
ducted after the dale of enactment of this rub-
section and prior to October 1, 1993 for any such
cc,ntandnant or con taininants faLls to detect the
presence of ch conian ’ .inant or Con(asrlnanLr
In the water supplied by tine drthi.anq water ry,-
-
The PRESIDING OFFICER. Th
clerk will report the amendment
The legislative clerk read as foilowe
Resolved, That the House re-cede from it.
disagreement to the arnendnusnt of toe Sen
ate nambered y42 to the aforesaid t,iij. an
concur thereto with an amenoment as Cal
1ow
In lieu of the matter proposed by ai
amendment, insert - Provided, Than tb
Coancil on Envtronmeotal Quality and Offlc-
of Environmental Quality shall retrnours.
other agencies for not less the.n one-half 0
the personnel compensation coats of t .ndivld
uala detailed to it.’.
Mr. METZEL4BAUM Mr President.
rise on behalf of Senator Wi ru to ad
dress myself to this amend.mer. ’. hk
I very strongly support Senator WIRTI
is unable to be with tie at thia lats
I ’aD— E:’4ATE
clpv vits (a f irte ,e ,, hr ’&gy iC , e A ’ ”1’ : 0 .1Lf 1 j’r A hf! ..‘ ::
rner t ed stWn.ardotherpurposss t.’,r 21 ? ,- rai
Resolved. That t e Rouse recedo fr
d1 ’wgree ent to tile amendment ci i2e .izi- R,jt’!res ‘ sv P€iu,jt, ft’jt(to r its
ate naznfr,red 246 to the aforesaid 5C . tc l n. een ’ i- t .. ‘ s endzzthn ris San-
concur therein with an amecdnloc . as c.i a,.e na t. ,r ’ , . - t.bs , L ’ icri a.sJ ti - s.nd
ou -.w ‘ . ,t tr- z :t ittoai ur (01-
In lieu of tbe matter proposed by sole
F amendment. Lneert. .tou , -iy , ’ ‘rj csrs4 \‘ iald
C’uiptex I Of aUC Xl of the Dire Dner tnc
Suppternev ital Appropriations Act, 1992, lr.tl’j4- li?tJ, no .otitrc yuidnan ‘i this
mg Disaster Assistance to Meet the Prczer.: e’- sep ocnc ici - rts t to’sr ,tl u’ ,ear,
5mergcncles Artiing from the Ccnaequcncez oj ip i. Haxc ae- h wrss .9iru,a-L v i ,’ usc 3 ce of
Hurricane Andrew, Typhoon Oir.ar, Hurrica.ne jap , ,r j , sz(irop ,a 4dJ curizin wic. staff
jnlk i, and Other Not,jrai Disasters, and AddI- years rwiec to it l-j W.s .aiV rhol! b’ -lans-
tlono.l Assistance to Distressed Cornru-uties jtrre4 J ’,ce line stale end .(acr,s ?roc vt and
(H.R 5631) Is wnended by (1) strikIng the ,na er Supr vnt D ’ -c ‘a e Unisot. $t ct Ptre
under 1)2 heading “Disaster relief’ and tr, ’ -’ A mlt-atstr -cLs tau 2 ida’js a,” ).‘ia I W3nent
In lieu thereof “For n nsary expenses in C - of this Ac..
nJtng Out the Robert T. Stafford Ditattey Reltet RosoLi , ,ed , ‘ t a ?rL wa , : r ’uije l unn it.,
and Emergency Assistance Act, as amended lsogreeusaon to tha actoudmnut vi t ha Sea-
l2.B93 ,CY .s3, of waich not to exceed $50 &iC C ’ ate ‘iber,b Ic , - the e(au-laid u’ ,ll, and
may be transferred to the ‘Disaster Assistance coccitr IZ%L5 cic c a-s (01-
Direct Loan Pro pum’ aecaunt for ud,nurdstra. lowx
Live expenses and suosidses for direct loans pro- In ‘don f tha cr . ,Qten o’or , ’nemf - said
vided ur4ev ao , ’ 417 of sited Act, and of amendment. meert
whIch 5143,000 t l shall be availablg only to the The D l rectae of the Federal Re,ercencs Man-
extent as official budget re-.p .iesl, for a. specific merit 4ge ’wy UseD snifertaksgu-l ’empw of tile
dollar ct-riount, that includes designatlor. of the ngency: orponLsstio,iol structus, and.. ijiukin
enbre amount of the re3uext as an emergency 2 days of er.acoamit of this rI, subra z 0 the
reguirement as deflnezi in the B&ar .ced Budget appvcpriale comnntieea of the Congnsss .z reor-
and Emergency Def ctL Control Act of 1955. U ganlsation pica which reflects chanutty .‘nisslon
transmitted by the President to the Congress, to requiranents cr4 prioiltlex, Tke na’i e chaD In-
remain avaliaale wial expended. Provided, That elude an assessment of the Nattanal Prepared-
th entire ainowit ‘a den qma ,Ced by Congress as ne : Directorate mul wne -pates-.iat alter-
ve emergency reqwrsseent pursuant to section nathia to meet that dU Orate s pritiir .al ob-
2S1(b)(2)(DXi) of the Balanced Budget and )ecttves while 1ncrstiss overall ip wncsJ effi.
Emergency Deficit Control Act of 1955, as
amended “and (7) w ’u ing tine matter under the p 0 j That’ thh Boone ri eade from
heading “Disaster assistance direct loon pro- Its disagreement to the asr. zdmiint of
p-am account and ln,sev’t in lieu thereof The , , , , , ., , , ,,.
1’.mU4tlOn on dtre-t loans for ‘Dls j a, e nate nttm. . - - . . e ore-
sirtnr ce direct loan p’oqram cccount’ . said bill, and concat’ therieisi w ’ th an
creased, within erucing funds, by u cs occt7 to amendment as foUo &
not to exceed ZOSS Ott) O ) Provided further, In lieu of the znsLter inaee’tod ‘by said
That not to errs,4 Z* )Ot ) Cs available for d l- amendment. tasert
red loan obhoanans provided to eligible appli- The Mission Simulator and S cssy Facility,
Canto or to States under section 319 of the Rob- Building NumberS, of the tVstid a2 Acronautla
en T. Stafford Duo,wcr Ajxtstan e and Emcr- and Space Adrtinlj’trt sU ,on , located at the John-
pency Relief Act, as amended Provided further, son Space Center in Houston, Ter,rs. to hereafter
That not to exceed $st,t),CCt),000 Is available for named arid designated the “Jiske Gore Mission
community disaster loans to local goverur ,ie,its Simulator and Training .Fhclh.’y ’. Any ref-
under stctmon ill of the Robert T. Stafford Dis- erence in a law, rule, map. r v ,latIo’*. docu-
aster A si tance and Emergency Relief Act, as inent, record, or other paper of the L’nlled States
amended, Provided P ,srnhor. That any unused to such facility shall be held to e a reference to
portion of the direct loan tin-iCc lion shall be the “Joke Gum Mission Slmr ,ulator and Training
available until September 30, 1993 Frovld d fur- Facility”
ther, That the entire amount is designated by Rtsolved, That the house recode from It
Cor.gress as an e ne.’vcwicy requirement pursuant disagreement to the amendment O S the San-
tosertionlSl(b)(2 ,lD,(t,Qft,- ,eBaZaflcp,dB.,4 0 ct ate cumb rr4 3 to the aforesaid bill, and
arid Err,ergency Deficit Control Act of 19.55, as concur therein with an amendment as f bI-
amended,”, lows
P.esolt’ed, That the House recede from its In lien of the matter prop , med by said
disagreement to the amendment of the Sen- amendment, ths rt
ace numbered 247 to the aforesaid bIll. ant ’ S 5c 001 S tJx DleeecxNo WArcis ACT IMP 5 -
concur Userein with in ameodrrtenc as Cal- jes’r*m —
lows. ‘ (a) Safe Dr-tcking Water Act Report —The
En lien of the sum proposed by ald amend- Adn ’ iiniats-ator of the Envtronn ’ ienuel Protec-
meat, insert tioc Agency shall report to the Congress
$150409000 Provided further, That up to wltaln nine months of the datei 01 nsactzueat
12,0( 1) of the Juiias appropriated under tins of this sectioc recomnnsendatlons n onovrning
heading ‘nay be era rs.sfrrrc4 to and merge4 with the reauthorization of the Sale r,rlnking
appropriated for “Office of !nspector Gen- Water Act Such report shall address—
era!’ (I) the adrerse htaith eifevte an cciated
Resolved, That the House recede from its with contaminants L a dztrski g’water and the
disagreement to the amendment of the Sea- public health and Other b vilta ttz .t may be
ate numbered 249 to the aforesaid bill, end realized by removing such eoat.amthaote,
concur therein with an amendment as fbI- (2) the ç ,rixew far ldsoztiytng Conr,aml-
l owe naots in ‘Jrrsktng water t ,ncj .sslecz tag con-
La lieu of the sum proposed by saId amend- tarrtthant.s (Or coatrol,
meat, insert $253 2 ,3 ( 0) (3) sctneli{es for the deveso imern of reg i s-
!u’so!ved, That toe House recode from ita latlons ant c.smp aace wi,”b dr1 u , water
dl’.ag’reemant to the amendment of the Sea- standards
ate numbered 254 to the aforesaId bill, and (4) the f’cnocial and technicrl capacity of
concur therein witn an amendment as (01- drtnkusg waar r ’Bterns to ltoplee-ient man-
tows itorl.ag reQutrrr-rate associated with regu-
R.epr ,ore the matter st,rlc en by said lated a d ucr gutst,ed contaci,iaai t., and op-
amendment, amended to read as follows Uocs to facilitate ijnpiemsois : oa ‘of such
-------
11
I Distribution List
2 Administrator
3 Deputy Administrator
4 Assistant Administrators
5 Regional Administrators
6 Office of Policy, Planning and Evaluation
7 Office of Radiation Programs
Office of Water
9 DOE
10 NWTRB
ii RC
I o
-------
7.1
7. ANALYSIS OF RISK AND COST ISSUES RAISED BY THE SCIENCE
ADVISORY BOARD
Chapter 7 discusses a number of specific risk and cost issues raised by the Science Advisory
Board (SAB) during its February 1993 review of EPA’s risk and cost assessments for radon
in dnnking water. Sections 7. 1 and 7.2 address issues relating to the nsk assessment; Section
7.3 addresses issues relating to costs and treatment.
7.1 ISSUES RELATED TO RISK ATTRIBUTABLE TO DRINKING WATER
SAB comments on the Agency’s radon risk analysis were submitted to the Administrator in a
report entitled Review of Uncertainly Analysis of Risks Associated with &posure to Radon --
Chafee-Lautenberg Multi-media Risk Study (June 1993).
The risk anaiyses summarized in Section 7.1 represent revisions and expansions to the initial
analysis of radon risks attributable to drinking water which was prepared by EPA in
February 1993. As a result of these revisions, the specific numerical results presented in this
section differ from those presented in earlier EPA documents and presentations. Generally,
the analyses presented in Chapter 7 affect the width of the credibility intervals surrounding
the Agency’s earlier estimates of exposures and risks, with relatively minor influences on the
central estimates. These differences are judged not to be critical to the risk management
decisions that need to be made regarding protection against radon risks. Rather, these
changes represent “scientific fine-tuning” of the earlier findings and are felt to provide a
more robust analysis of radon risks attributable to drinking water. This conclusion is
consistent with the SAB’s position that these “changes in most cases would not substantially
change the document’s estimates of central values for risks.”
In its report, SAB stated that, in its judgment, EPA had adequately addressed the significant
radon risk assessment issues and that the quantitative uncertainty analysis developed by the
EPA represents a methodology that is state-of-art and significantly improves the scientific
basis for the EPA’s decision-making. SAB stated that the revised estimates for ingestion and
inhalation risk due to radon in drinking water are scientifically acceptable. In addition, the
SAB made several specific recommendations to EPA to improve its radon risk analysis.
SAB recommended that EPA (1) consider the adequacy of the uncertainties associated with
estimated internal doses from ingested radon from drinking water; (2) include uncertainty in
-------
7-2
the variance and shape of risk distributions used to represent variability, (3) extend the
uncertainty analysis to include risk reductions attributable to implementation of target levels
of interest, and (4) provide a qualitative discussion of several issues includihg the use of
linear dose-response relationships extending to low dose, the effect of population mobility on
radon risk estimates, and the influence of smoking on lung cancer risks from radon.
The following sections summarize EPA’s analysis of these issues. The focus of the
following sections is limited to estimates of mean exposures and risks and does not include
discussion of other exposure or risk statistics. These and other details as well as a full risk
characterization can be found in the EPA’s revised Uncertainty Analysis of Risks Associated
with Exposures to Radon in Drinking Water (U.S. EPA, 1993h).
Issue: Uncertainties Associated with Ingestion Risk and Dose
The SAB commented that the revised estimates for ingestion and inhalation risk due to radon
in drinking water are scientifically acceptable. However, the SAB stated that the organ-
specific doses used for assessment of ingestion risks are based, in part, upon an unpublished
study of kinetics of an analog, xenon, in humans. The cited study did not include a mass
balance determination. The SAB recommended that EPA carefully review this study to
evaluate whether the uncertainties attributed to the results are adequately described. Given
the larger uncertainty bounds associated with the ingestion risk, the SAB recommends that
the EPA consider keeping ingestion and inhalation risk separate in the EPA’s deliberations on
standards for radon in drinking water.
The ingestion risk estimates and its associated uncertainties are discussed in the Chapter 2 of
this report and described in detail in Uncertainty Analysis of Risks Associated with Exposure
to Radon in Drinking Water (USEPA, 1993h) and Drinking Water Criteria Docwnenz for
Radon in Drinking Water (USEPA, 1991a). After a person ingests radon in water, the radon
passes from the gastrointestinal tract into the blood, principally by way of the small intestine.
The blood then circulates the radon to all organs of the body before it is eventually exhaled
from the lungs. The biological half-life of radon in the body ranges from 30 to 50 minutes
(Hursh et a!. 1965; Suomela and Kahlon, 1972; Hess and Brown, 1991). When radon and
its progeny in the body decay, the surrounding tissues are irradiated by alpha particles.
However, the radiation dose and risk per radiation dose varies from organ to organ.
I U
-------
7.3
The Agency has expended considerable effort in estimating the organ radiation doses
following radon ingestion. There are several studies in which a small number of subjects
ingested radon laden water and were followed over time, either by whole body counting of
the penetrating emissions from the radon progeny or by measuring radon in expired air. All
of these studies suffer from the limitations that direct measurements of organ concentrations
could not be carried out with the experimental procedure used, and depended on inferring
radon and progeny concentrations from an equilibrated mixture of the parent and progeny.
Although these studies were taken into consideration, EPA tried to improve the dosimetry
estimates of ingested radon using an alternative procedure that would overcome the
limitations of these studies. Rather than using radon itself, the metabolism of xenon-l33 in
thirty-five human subjects was studied (Correia, 1987). Xenon, a photon emitter, is
chemically inert and behaves in the same manner as radon in tissue but differs slightly in
tissue solubility. After ingesting water laden with millicurie levels of xenon, the xenon
levels in human subjects were followed for periods of up to ten hours with a gamma camera.
Organ radioactivity concentration vs time curves were generated, quantified in absolute
concentration units and converted to radon kinetic curves using the measured tissue/blood
partition coefficients of radon and xenon.
The project was carried out at Massachusetts General Hospital which has the capability for
handling the imaging technology and radioisotopes for use in human subjects, and expertise
in mathematical modeling and data handling. EPA is confident in using the xenon study to
estimate radon concentrations by organ since: (1) the measured transfer coefficients and
rates in this xenon study are comparable to those reported in other human studies of xenon
and (2) the derived organ radon burdens and removal rates are comparable to human radon
studies for which such values have been determined. To further fine-tune the dosimetry
estimates, EPA later funded a project to generate a biokinetic model for ingested radon
fitting the empirical retention functions for eight organs from the xenon study into the model
(Crawford-Brown, 1990).
The mass balance determination suggested by SAB is probably not crucial for the organ dose
determinations since the actual radioactivity in various organs at different time intervals were
measured directly. However, for quality assurance and completeness, the Agency is
currently working with the principal investigator of the xenon study at Massachusetts General
Hospital to provide a mass balance calculation for the ingested xenon and radon.
As described in Chapter 2, the quantitative uncertainties associated with each risk, dose and
exposure parameter used in obtaining the ingestion risk estimates were determined. There is
-------
7-4
uncertainty associated with the dose to target epithelium cells in the wall of the stomach and
colon from ingested radon since the emitted alpha particles have a maximum range of only
70 microns. One public comment argued that any alpha particles emitted through radioactive
decay of radon atoms in the stomach will be unable to reach the epithelium target cells. To
support this argument, it was pointed out that very little ingested water passes directly from
stomach wall into the blood stream, suggesting that small molecules cannot penetrate the
stomach wall. EPA believes that this argument is misleading. Given its small size, chemical
inertness, solubility in both lipid and water, and the results of extensive studies of diffusion
of inert gases in water and tissue, radon would be expected to diffuse readily into the
stomach wall, including extracellular and intracellular spaces. The lack of significant bulk
uptake of water in the stomach does not imply any diffusion barrier for water either. Thus,
the movement of either radon or water into the wall of the stomach is expected to be
relatively rapid from a diffusion standpoint. Nevertheless, diffusion of material from the
lumen of the stomach to the blood is still a slowN process. As a result, most of the uptake
of ingested radon or water will occur in the small intestine, which because of its extremely
large surface area and extensive mixing of the contents that occur there, is a much more
efficient organ for absorption than the stomach.
Although it is expected that radon will readily diffuse into the epithelial cells of the stomach
and the intestines, there is a uncertainty with respect to the distribution of radon across the
wall of these organs and, consequently, with respect to the dose delivered to target cells.
EPA assumes that the ratio of doses at target cells to that in the gastrointestinal lumen is 1:3
with a credible range of 0.2 to 0.8, and that decay products after Po-218 are swept away and
reducing energy deposition to 60% with a credible range of 0.6 to 1.0. It is important to
keep in mind that only 67% of the radon ingestion risk is associated with irradiation of the
gastrointestinal tract. The remainder arises from a fairly uniform whole-body dose, the
estimate of which is more directly based on measured data. In estimating the concentration
in organ doses, EPA considered uncertainties in : (1) the xenon measurement, (2) the
dosimetry assumptions described above, and (3) the dependence of dose on age. Overall, the
credible range between the upper and lower dose estimates is less than a factor of six,
depending on the organ.
EPA also quantified the uncertainty associated with the organ-specific risks per unit dose of
radiation. The sources of uncertainty in these estimates are contributed by (1) sampling
variation, (2) age/time dependence of risk; (3) extrapolation of the data from the Japanese
population to the U.S. population; (4) errors in dosimetry, and (5) uncertainty in the relative
biological effectiveness of alpha particles. Overall, the credible range between the upper and
lower estimates is less than a factor of 10-30 depending on the organ. Taking each of the
-------
7-5
sources of uncertainty in dose (rad per pCi) and risk (risk per rad) discussed above, the
credible range of fatal cancer risk due to ingestion of 1 pCi of radon between upper and
lower estimates is less than a factor of 17.
EPA’s quantitative uncertainty analysis agrees with the SAB comment that the uncertainty in
radon ingestion risk is greater than for the inhalation risk. For the latter, the risk per unit
exposure is estimated directly from observed excess lung cancer rates in uranium and other
miners exposed to airborne radon and its progeny. For the former, the estimate of risk is
more indirect, depending on dosimetric experiments/assumptions and on a relative biological
effectiveness of alpha particles estimated from laboratory studies. As described above, EPA
tried carefully to identify the important sources of uncertainty in estimating radon ingestion
risk and to quantify them. The Radiation Advisory Committee of SAB reviewed EPA’s
analysis of radon ingestion risks and agreed that the uncertainty bounds placed on the various
parameters and the method used to combine the various components of uncertainty are
reasonable.
It should be noted in this context, however, that the uncertainty range for the ingestion risk
reflects a degree of professional judgment. Professional judgments were necessary to capture
the uncertainty in the underlying parameters used in estimating ingestion risk. For example,
while it is expected that radon will readily diffuse into the epithelial cells of the stomach and
the intestines, there is uncertainty with respect to the distribution of radon across the wall of
these organs and, consequently, with respect to the dose to target cells. The distribution will
depend on a number of factors, including the diffusion rate of radon, the detailed
microanatomy of the wall, and blood flow rate through the walls. The resulting confidence
bounds (or credible range) represent EPA’s best interpretation of the available dosimetry and
risk data. It is the Agency’s belief that, given the available evidence, ingestion risks are
within the credible range. As part of the radon risk assessment and characterization, EPA
presents separate numerical estimates of the ingestion and inhalation risks and their associated
uncertainty.
Issue: Uncertainty in Variability
In its February 1993 presentation to the Science Advisory Board, EPA discussed the basic
methodological approach and findings from its analysis of risks attributable to radon in
drinking water. In that analysis, EPA modeled the natural variability in each of the terms of
the radon risk models through the use of probability density functions (pdJ). Probability
f/LI
-------
7-6
density functions are mathematical expressions which provide numerical statements of the
probability that a model variable, e.g., radon concentration in drinking water, is within some
arbitranly small interval. Each pdf is characterized by its para neters’ which are ideally
estimated from data but often be supplemented by professional judgment when sufficient data
are lacking or not available. Pdfs can differ from each other in two ways: (1) in their basic
mathematical form (e.g., lognormal, normal, uniform, etc.) and (2) in the specific numerical
values of their parameters (e.g. the log-mean of one lognormal variable in contrast to a
different log-mean for another lognornial variable).
The exact roles of the parameters of a pdf depend on the mathematical form of the pdf itself.
The parameters of some key pdfs that are important in environmental risk analysis act to
characterize the location of the pdf and are called location parameters. For example, the
arithmetic mean is the location parameter for a normal distribution; different normal
distributions may have different means. Other parameters often describe the shape or
overall spread of the pdf (i.e, broad or narrow) and are called shape parameters. For the
normal distribution, the shape parameter is the standard deviation. Different normal
distributions may have different means and/or different standard deviations.
Specific numerical values for the parameters of the pdfs used in the radon uncertainty
analysis were estimated from a variety of data sources. Because data is always limited, there
is always uncertainty associated with each parameter estimate. In its February 1993
presentations to the SAB, EPA limited its uncertainty analysis to include only uncertainty in
the location parameters of each of the pdfs. In its review, SAB asked EPA to include
uncertainty in the shape parameters as well. The following is a summary of EPA’s analysis,
expanded to include uncertainty in the shape parameters.
‘The term parameter is often used to refer to model variables, i.e., model parameter is used
interchangeably with model uvariableu. This is common usage within the technical community. This
practice can lead to confusion when the model variable (model parameter) is a random variable and is
itself characterized by its own defining statistical parameters. Thus, parameters can refer to both
model variables as well as their statistical parameters. In this Chapter, an effort has been made to be
clear about the quantity being used.
-------
7.7
Radon Risk Parameter Uncertainties
Table 7. 1 summarizes the variables used in the revised risk assessment for radon in drinking
water. The primary changes over the analysis reviewed by the SAB in February, 1993 are
the inclusions of uncertainties in the shape parameters. For the radon concentration pdfs, the
uncertainty in shape parameters was estimated from data contained in the National Inorganics
and Radionuclides Survey (NIRS). For the water to air transfer factor, the uncertainty in
shape was estimated from the effective sample size inferred in the study by Nazaroff and his
co-workers (1987). For the occupancy factor, equilibrium factor, and fraction not
volatilized, the credible ranges for the minima and maxima represent professional judgment
of EPA staff and other experts.
Triangular Distributions and Beta Distributions. EPA developed two approaches for
assigning probability density functions to the uncertainties in the radon risk models. In
developing these models, EPA was guided by the principle of maximum entropy, which
seeks to make maximum use of the available knowledge. Maximum entropy distributions are
distributions chosen to make maximize use of existing information, minimizing the use of
assumptions that cannot be supported. In this sense maximum entropy distributions represent
conservative or bounding distributions. For example, the triangular distribution is the
maximum entropy distribution for a random variable when the only pieces of information
available are the most likely value (i.e., its mode) and its range. Similarly, when the only
information known about a finitely-bounded variable is the arithmetic mean, the maximum
entropy distribution is the truncated exponential distribution.
In its previous presentations, EPA used triangular pdfs to represent the variability in the
occupancy factor, equilibrium factor, and fraction not volatilized. The highest mean that can
be derived from a triangular distribution over the range (0,1) is 0.67. In modeling the
variability of the occupancy factor and fraction not volatilized as triangular, it was necessary
to treat the means for both factors as if they were the most likely values. Strictly, this is not
correct. However, it was judged not to be a critical assumption given the relatively smaller
variation and uncertainty in these factors compared to the wider variation and uncertainties in
the concentration, water to air transfer factor, and risk factors.
As part of its expanded analysis, EPA developed a more flexible approach to assigning pdfs
for the occupancy factor and the fraction not volatilized. This alternative approach is based
on use of the beta distribution. The standard beta distribution is a very flexible, 2-parameter
distribution, taking on a wide range of shapes over the interval (0,1). The beta distribution
is often used to model data when data are scarce. In EPA’s application, the means for each
11
-------
7-8
of the three model variables were modeled as uniform over the credible ranges listed in
Table 7.1. The beta pdfs were then allowed to take on all shapes consistent with these
ranges of means. In essence, the beta modeling approach doses not depend on any single
shape but considers all shapes consistent with a defined range of the mean.
Expanded Estimation Procedure for Radon Concentrations in Ground Water.
In the February 1993 radon risk analysis, EPA used a single pdf to represent the variability
of radon concentrations in ground water systems. The criteria used to select this single pdf
was based on finding an optimal match between central and upper percentiles derived from a
single pdf and the central and upper end percentiles derived from the population-weighted
sum of the concentrations from each of the five ground water system strata. Thus, the single
pdf was an approximation intended to minimize any central and high end approximation
differences. A validation exercise was conducted and it was judged that the discrepancies
introduced by the approximation were reasonably small.
For fine-tuning its analysis to address the issues raised by the SAB, EPA did not collapse the
concentration data into a single, representative pdf. Rather, the Agency used the full
population-weighting procedure which incorporates and maintains a unique pdf for each
ground water system strata. It was felt that this modification, although more computationaly
intensive, was better for aggregating exposures and risks across system strata.
Radon Risk Esthnates
Tables 7.1, 7.2 and 7.3 summarize the findings from the expanded radon risk analysis. In
the Figures, “triangular model” refers to the analysis in which the pdfs for the occupancy
factor, equilibrium factor and fraction not volatilized were assumed to be triangular; “beta
model refers to the more general beta model in which the shape of the distributions was
allowed to randomly vary, consistent with the means and ranges listed in Table 7.1.
The effects of including uncertainty in the shape parameters generally widened the credible
range, with the greatest effect on the high end estimate. This result was anticipated from
strictly analytical considerations. The credible range increased over the February, 1993
analysis by approximately 13% for the triangular model and 17% for the beta model.
‘(7
-------
Table 7.1 Radon Risk Variables and Their Parameters:
Credible Ranges Used In the Expanded Radon Risk Analysis
Risk Model Variable
pdf
expressing
variability
Location Parameters
Shape Parameters
Concentration (pCi/Liter)
very, very n Il
very naI1
aniall
medium
large & very large
lognormal
lognorinal
lognormal
lognornial
lognormal
Geometnc Mean Credible Range
267 236 - 301
216 194-240
126 109 - 145
133 107- 166
136 102 - 180
GSD Credible Range
4.5 18 4.003 - 5.098
4.540 3.951-5.217
3.550 2.934 - 4.295
2.467 2.118-2.874
2.406 1.928 - 3.003
Water to Air Transfer
Factor (unitless)
lognormal
Geometric Mean Credible Range
6.57E-05 4. OlE-05 - 1.08E-04
GSD Credible Range
2.880 1.429 - 5.804
Daily Drinking Water
Ingestion Rate (liters/day)
lognormal
Geometric Mean Credible Range
0.526 0.517 -0. 535
GSD Credible Range
1.92 1.889 - 1.952
Mean Inhalation Risk Factor
(LCD/pCl/L per yr)
uncertain
constant
Geometric Mean Credible Range-
2.83E-04 1 .41E-04 - 5.70E-04
- an - - an -
Mean Ingestion Risk Factor
(LCD/pCi/L per yr)
uncertain
constant
Geometnc Mean Credible Range
l.24E-11 2.90E-l2 - 5.31E-l1
- na - - an -
Occupancy Factor (imitless)
triangular
beta
Mode = 0.75
Mean = 0.75
Credible Range
0.65 - 0.80
Credible Range of
Minimum
0.17 - 0.33
Credible Range of
Maximum
0.85 - 0.95
Equilibriwn Factor
(iwitless)
triangular
beta
Mode = 0.45
Mean = 0.45
Credible Range
0.35 - 0.55
Credible Range of
Minimum
0.1 - 0.2
Credible Range of
Maximum
0 8 - 0.9
Fraction Not Volatilized
(unidess)
triangular
beta
Mode = 0.80
Mean = 0.80
Credible Range
0.7 - 0.9
Credible Range of
Minimum
0 4-0.6
Credible Range of
Maximum
0.9- 1.0
-------
Table 7.2 tndividnnl Risks (risk of fatal per person per year)
Comparison of Expanded F.sthnntes With February 1993 F .diznates
Statistic
Executive Summary February, 1993
TrIangular Model Estimates
Beta Model Estimates
Inhalation
Ingestion
.
Combined
(11
Inhalation
.
Ingestion
Combined
(21
Inhalation
Ingestion
Combined
(21
Credible
Bound
5.9E-O7
2 .4E-O7
8.3E-07
4.5E-07
l.3E-O7
8.2E-07
4.9E-07
I.3E-O7
9.2E-O7
Median
131
l.3E-06
(l.SE-06)
6.6E-07
(8.3E-07)
2.OE-06
(2.3E .06)
l.3E-06
5.2E-O7
2. IE-06
l.4E-06
5.7E-07
2.3E-06
Uppom ’
Credible
Bound
2.9E-06
2.IE-06
5.OE-06
4.5E-06
2.5E-06
5.6E-06
5.OE-06
2.6E-06
6.5E-06
This column was determined by anthmetic addition and not by statistical addition. Thus, tius column as an upper bound; the “true estimate will
be slightly smaller than this.
These columns were determined by proper statistical addition.
Numbers in parentheses represent average maan nsk estimates and are shown here for compansons with earlier presentations.
[ 1].
[ 2].
(3].
-------
Table 7.3 Population Risks (fatal cancers per year)
Comparison of Expanded Estimates With February 1993 Estimates
Statistic
Executive Swnmary February, 1993
Triangular Model Estimates
Beta Model Estimates
Inh alation
Ingestion
Con3bined
111
Inhalation
Ingestion
Combined
121
inhalation
Ingestion
Combined
[ 21
w&
Credible
Bound
48
19
67
37
11
66
40
II
75
Median
[ 31
105
(118)
53
(67)
159
(186)
108
42
170
113
46
186
Upper
CredIble
Bound
233
166
399
367
203
454
408
212
527
[ 11. This column was determined by arithmetic addition and not by statistical addition. Thus, this colunm is an upper bound; the true” estimate will
be slightly smaller than this.
[ 2]. These columns were determined by proper statistical addition.
[ 3J. Numbers in parentheses represent average mean population risk estimates and are shown here for comparison with earlier presentations..
-------
Figure 7.1 Credible Range for
Mean Individual Risks [ 1,2]
IE-5
inhalation
ingestion
tlE-8 f
U combined
pathways
1€-i’
Feb193 beta Feb/93 beta Feb/93 beta
triangular triangular triangular
Figure 7.2 Credible Range for
Population Risk (1,2]
600
combined
pathways
500
Inhalation
400
300
ingestion
200
100
0
Feb 193 beta Feb/93 beta Feb/93 beta
triangular triangular triangular
[ 11 I represents nominal value; — represents median values
[ 21 Credible range for inhalation risk includes risk from inhaling radon progeny
but not from inhaling radon gas since its contribution to the overall risk is small.
-------
7-13
Issue: Uncertainty in Risk Reduction Attributed to Implementation
of Different Target Levels
The Science Advisory Board asked EPA to conduct an uncertainty analysis of the risk
reduction effects of reducing radon concentration levels in ground water to the proposed
MCL. This section presents the results of EPA’s analysis of this issue. It should be noted
that the following considers community ground water systems risks only. This is in contrast
to the broader effort for the Agency’s benefit/cost analysis which included community ground
water systems and non-transient, non-community ground water systems. In addition
nominal or best point estimates were used in the Agency’s benefit/cost analysis rather than
medians and means.
It should also be noted that neither Agency effort includes drinking water supplied by private
wells.
Radon Reduction Assumptions. Every pollution control technology has an associated
removal efficiency. The issue relevant for assessing risks is: given some regulatory target,
what levels will actually be achieved in practice. Ignoring compliance questions, it is
generally believed that radon reduction measures are very effective and could reasonably be
expected to achieve levels well below any contemplated regulatory limit. The following
analysis discusses two radon reduction scenarios. The variable control scenario described
below was used by EPA in its benefit/cost analysis.
Truncation Scenario: Reduction to the Target Level and No More. In this scenario,
each system over the target level will reduce radon
concentrations to the target level and no more. This
should be a very conservative assumption in the BOX 7.1 Radon RanovaJ
Efficiencies for the
sense of underestimating risk reduction benefits. Vai Me Control Scenario
Variable Control Scenario: Concentration-
Dependent Efficiencies. This scenario uses the
removal efficiencies assumed by EPA in its cost 300 - 600 pCIJL
analysis as shown in Box 7.1. removal efficiency = 0.50
for system concentradom between
600 - 1,500 pCi/L
removal efficiency = 0.80
Table 7.4 and Figure 7.3 summarizes the results of
this analysis. than 1,500 pCi/L
removal efficiency = 0.99
-------
Table 7.4 Mean Individual Risk and Population Risk Attributable to
a Target Level of 300 pCi/L under Two Radon Reduction Assumptions
Tar d Lcvd
ln vMk l R
(t z1riru p p on p year)
l pulation Risk
(riinr rs pa year)
VMhw ios in
Am I C.iiri’v
(diff K In
ç ij )
Low d
Megan
1Iigb. id
Low-end
fe
MA i
Hlgfrmd
-
e Cua No Co
triai iiirmodd
beta modd
8.2E-07
9.2E-07
2.IE-06
2.3E-06
5.6E-06
6.5E-06
66
75
170
186
454
527
-
-
Truncation Scenario
(tnmtion to 300 pCiIL)
triangularmodd
beta modd
5.6E-07
5.7E-07
l.4E-06
1.4E-06
4.OE- .06
3.9E-06
45
47
114
114
324
316
56
72
Variable Control Scenario
(variable rmnoval elTiciency)
triangular model
beta model
4.6E-07
4.7E-07
l.1E-06
l.2E-06
3.2E-06
3.SE-06
37
38
89
97
259
284
81
89
-------
Figure 7.3 Annual Population Risks Attributable to Reductions to
300 pCi/L for Two Control Scenarios (1,2]
600
500 Beta
Model
Triangular
400 Model
300
U
200
100- ——
0
No Controls Variable No Controls Variable
Truncation Truncation
— represents median values
[ 2J Truncation refers to reductions in groundwater concentrations to 300 pCi/L and no more;
variable refers to concentration-dependent removal efficiencies assumed by EPA
in its benefit/cost analysis.
-------
7-16
Issue: The Effects of Popubtion Mobility on Radon Risk
E a
The Science Advisory Board asked EPA to consider the effects of population mobility on
exposures and risks attributable to radon in drinking water. The following discussion is a
summary of a more detailed analysis of the population mobility and its effects on radon risk
estimates presented in Uncertainty Analysis of Risks Associated with Exposures to Radon in
Drinking Water (USEPA, 1993h).
A thorough analysis of the effects of population mobility on risks attributable to radon in
drinking water would need three key pieces of information: (1) how often people move, (2)
where they move from and where they move to, and (3) the radon concentrations in their
drinking water at both locations. Unfortunately, these three key pieces of information are
not available in the detail necessary to conduct a comprehensive analysis. As a
simplification, the following analyses implicitly assume exposures derived from population-
weighted average radon concentration and ignore the finer details of geographical exposure
differences and differences associated with different ground water delivery systems.
There are two primary considerations for analyzing the effects of population mobility on
radon risks associated with waterborne radon. The first consideration is total residency time
(or total residency period). Total residency period is the length of time that the same
individuals occupy the same residence. The second consideration is moving patterns.
Moving patterns refer to geographic moving patterns, that is, where people move to and
where they move from. Since radon risk is directly proportional to exposure and, since
mobility patterns affect the probability of exposure and exposure duration, residency times
and moving patterns may be important factors affecting radon risks from waterborne radon.
Considering the highly mobile U.S. population, it is therefore likely that some portions of the
U.S. population will have wide swings in their exposures to waterborne radon over the
course of their lifetime. In addition to population mobility, radon concentrations in ground
water show strong local and regional variations from below 100 pCi/L to over 25,000 pCi/L.
Current Residency Periods. The statistic needed for assessing population mobility effects is
total residency period. Total residency period should be distinguished from current residency
period. Current residency period refers to the number of years that the current occupants
have lived at that same location - not the total time the occupants will live at that residence.
EPA, in its Exposure Factors Handbook (USEPA, 1989b) using data from a 1983 survey by
-------
7-17
the Bureau of the Census, estimated the percentage of owner-occupied houses for specified
periods of time. Based on these data, the Agency estimates the 50th percentile, the mean,
and the 90th percentile (reasonable worst case current resident period) to be 9 years, 13
years, and 30 years respectively.
Residency periods may be expected to vary by geographic region, as well as vary by rural,
suburban, owner versus renters, and urban residence patterns within a region. Residency
period data stratified by these variables has not been obtained by the Agency. Using the
Bureau of Census data, the average number of lifetime moves based on current residency
times for owner-occupied houses is estimated to be 5.4 (70 years/13 years-per household =
5.4 moves per lifetime).
Total Residency Period. Using 1985 and 1987 U.S. housing survey data, Israeli and Nelson
(1992) estimated the average total residence time for all U.S. household to be 4.6 years, 2.4
years for renters, and 11.4 years for owners. These estimates are in contrast to estimates of
the average current residency period as 10.6 years for all U.S. households, 4.6 years for
renters, and 14.0 years for owners. Using these estimates, 15.2 moves per lifetime (70/4.6
15.2) are expected, averaged across all households.
For this analysis, two estimates of the average number of lifetime moves will be used to
bound the analysis: 6 and 15 moves per lifetime.
Constant Total Exposure Model (CTE)
The most simple and most direct approach for modeling population mobility would be to
assume constant total exposure (CTE), that is, every person moving out of a ground water
system containing radon is immediately replaced by someone moving in, thereby conserving
total exposure. For any given move, the probability of moving into a radon-containing
groundwater system is approximately 32.4% , which is simply an estimate of the current
proportion of people exposed to radon via community ground water systems. In
mathematical terms, the average person’s exposure is then represented by a binomial pdf
with p 0.324 and q 0.676. Table 7.5 shows the resulting distribution of exposures and
relative risk based on six lifetime moves.
The relative riSk measure in Table 7.5 expresses the ratio of population risk under the
assumption of mobility relative to the total population risk under the assumption of no
) Z l
-------
7-18
mobility. The CTE model leads to three conclusions. First, there is no net change in
population risk (relative risk = 1.0). Second, there is a decrease in average exposure
duration from 70.0 years under the no mobility assumption down to 22.7 years in the CTE
mobility model. This decrease in average exposure years is exactly offset by the increase in
the size of the population at risk, that is, from 81 million in the no mobility model to 250
million in the simple mobility model (including the 23.8 million with no exposures over their
lifetime).
Table 7.5
Constant Total Exposure Model
The Effects of National Population Mobility on the Current Radon Risk Estimate
Based on Average of 6 Moves per Lifetime
Number of Lifetime
Years Exposed to
Residency Penods with
Radon-containing
Average
Population
Relative Risk
Radon Exposures
Groundwater
Probability of
Exposure, (%)
Exposed
(million)
Proportion
0
0
9.52
23.8
0
1
11.7
27.40
68.5
0.141
2
23.3
32.89
82.2
0.338
3
35.0
21.05
62.6
0.324
4
46.7
7.58
19.0
0.156
5
58.3
1.45
3.6
0.037
6
70.0
0.12
0.3
0.004
U.S. Average
Lifetime Exposure
22.7 years
E
= 226.2
E = 1.00
The following observations are independent of the number of lifetime moves assumed per
individual as well as independent of the population currently exposed to radon and depend on
the assumption that total exposure is constant:
1. population risk will remain unchanged,
2. average individual exposures will decrease, and
3. the total population at nsk will increase, exactly off-setting decrease in
exposure.
-------
Probability of Indicated Years of Exposure Based on 6
Moves per Lifetime, Constant Total Exposure
46.7
35 years (21.0%)
70 years (0.1%)—
23.3 years (32.9%)
years (27.4%)
Exposed Population (millions) and Years of Exposure
Based on 6 Moves per Lifetime,
Constant Total Exposure
35 years (52.5 )
23.3 years (82.25 ) —
1.7 years (68.5)
Figures 7.4 and 7.5
Distribution of Exposures and Population Exposed for the
Constant Total Exposure Model
iii No Exposure (9.5%)
70 years (0.25)
58.3 years (3,75)
46.7 years (19)—
No Exposure (23.75)
1 2
-------
7-20
Mobility Based Approach: A Model Taking Mobility Patterns into Account
The prior analysis assumed that the probability of moving to a community ground water
system was independent of where the move started and where the move ended. This
assumption ignores the fact that the majority of moves occur within the same county.
Data describing regional variability in moving patterns of resident populations by state and
geographic region were collected by the Bureau of the Census and are summarized in the
regional statistics presented in Table 7.6 (USEPA, 1989b). For the United States as an
average, when moves are undertaken, 56.3% of the moving population remain within their
same county; 22.0% remain within the state; and 2 1.7% leave the state. When a move does
occur and the new location remains within the county of origin, the individual’s radon
exposure status is assumed not to change; when a move does occur and it is out of the
county, it is initially assumed that there is a 50% likelihood that the individual’s ground
water radon exposure status will change; if a move occurs and it is out of the state, it is
assumed that the individual’s radon exposure status will change on a random basis defined
by the relative proportions of the national ground water supplied population and the non-
ground water supplied national population.
Because most moves take place within the same county, there is a probability greater than
0.563 that an individual’s radon status will not change after the move. The probability that a
person over his/her lifetime will experience a specific number of residency periods exposed
to radon in ground water can be modeled as a compound binomial probability density
function which is dependent on two conditional probabilities: (p 1 , p2): Pi is the conditional
probability of moving from a ground water-supplied system to another ground water-supplied
system (no change in status), and P2 is the conditional probability moving to a ground water-
supplied system from a surface water-supplied (change in status). Estimates of these
probabilities are shown in Table 7.6. For the U.S, P = 74.8% and P2 = 18.5%. On a
regional level, Pi ranges from 66.8% in the Mountain States region and 79.7% in the East
North Central States, with a population-weighted national average of 74.8% for the entire
United States; P2 ranges from a low of 15.5% in the Middle Atlantic region to a high of
21.4% in the Mountain States region, with a national population-weighted average of 18.5%.
Aggregating the effects of Pi and P2 requires that they be weighted according to the
population ratios, e.g., to reflect the ground water population/total population, and non-
ground water population/total population ratios.
Tables 7.7 and 7.8 show the estimated relative effect of population mobility on the current
radon risk estimate. Total population risk is predicted to be higher by about 14% . The
-------
7-21
relative risks are seen to be the same for 15 moves per lifetime and 6 moves per lifetime
(relative risk = 1.136), while the average U.S. lifetime exposures (26.6 years compared to
32. 1 years) and total population exposed (242.million and 202.4 million) are seen to be
different.
A sensitivity analysis of these findings was conducted by varying the parameters over a wide
range of values. The mean exposure period was found to range between 24.4 years and
35.4 years. However, this reduction is offset by an increase of between 0.2% and 40.6% in
the estimated population risk. The national population exposed also increased: 183 million to
248 million people exposed. Between 2 and 67 million people have no radon exposures via
drinking water; between 0.4 million and 27.8 million people experience a full 70 years of
radon exposures via the drinking water route.
Conclusions on the Effects of Mobility
Under two different sets of modeling assumptions and two different moving rates, best
estimates of population risks were found to range between no change and a 14% increase.
Average individual risks were found to decrease by more than 50%; this decrease was offset
by the increase in the population at risk from 81 million to 201 - 242 million.
-------
TABLE 7.6
Mobility of Resident Pbpulatloas by Region: 1980
••. Sou : Fado Han ook Uu , 1989)
Region,
Division
1 rsoas S
years old
(thousands)
Percentage Distribution -
residence in 1975
Normalized Percentage, Given that a Move
Has Occurred
Same
Residence in
1980 as 1975
Different
House, Same
County
Different
County,
Same State
Different
State
Different
House,
Same
County
Different
County,
Same State
Different
State
United States 210,323
53.6 25.1 9.8 9.7
56.3 22.0 21.7
Northeast 46,052
New England 11,594
Middle Atlantic 34,458
61.7 22.3 8.0 6.1
59.1 23.4 6.7 9.2
62.6 21.9 8.4 5.0
61.3 22.0 16.8
59.5 17.0 23.4
62.0 23.8 14.2
Midwest 54,513
East North Central 38,623
West North Central 15,890
55.4 26.4 10.2 7.0
56.0 27.4 9.6 6.0
53.9 24.0 11.8 9.4
60.6 23.4 16.1
63.7 22 3 14.0
53.1 26.1 20.8
South 69,880
South Atlantic 34,498
East South Central 13,556
West South Central 21,826
52.4 24.1 10.0 12.0
52.7 22.4 9.7 13.6
56.0 25.9 7.9 9.5
49.6 25.6 11.8 11.0
52.3 21.7 26.0
49.0 21.2 29.8
59.8 18.2 21 9
52.9 24.4 22.7
West 39,879
Mountain 10,386
Padfic 29,493
43.8 28.3 11.0 13.4
42.7 25.1 9.1 21.1
44.2 29.4 11.6 10.7
53.7 20.9 25 4
45.4 16 5 38.2
56.9 22.4 20.7
-------
7-23
Table 7.7
The Effects of National Population Mobility on the Current Radon Risk Estimate
Based on Average of 15 Moves per Lifetime
Nwnber of Lifetime
Years Exposed to
Residency Penods
Radon-containing
Average
Population
Relative Risk
with Radon Exposures
Groundwater
Probability of
Exposure (%)
Exposed
(million)
Proportion
0
0
3.1
7.7
0
1
4.7
10.6
26.4
0.0211
2
9.3
16.9
42.3
0 0673
3
140
16.7
41.9
0.1000
4
18.7
11.5
28.7
0.0914
5
23.3
5.8
14.5
0.0576
6
28.0
2.3
5.8
0.0277
7
32.7
1.1
2.7
0.0157
8
37.3
1.4
3.6
0.0251
9
42.0
3.0
7.6
0.0599
10
46.7
5.4
13.5
0.1183
11
51.3
7.3
18.3
0.1763
12
56.0
7.3
18.2
0.1913
13
60.7
5.0
12.5
0. 1426
14
65.3
2.1
5.3
0.0655
15
70.0
0.4
1.1
0.0140
U.S. Average
Lifetime Exposure
26.6 years
E = 242.3
1.1360
Table 7.8
The Effects of National Population Mobility on the Current Radon Risk Estimate
Based on Average of 6 Moves per Lifetime
Number of Lifetime
Years Exposed
Residency Periods
to Radon-
Average
Population
Relative
with Radon
containing
Probability of
Exposed
Risk
Exposures
Groundwater
Exposure (%)
(million)
Proportion
0
0
19.7
49.2
0
1
11.7
27.1
67.7
0.1349
2
23.3
16.5
41.2
0.1650
3
35.0
9.0
22.5
0.1408
4
46.7
10.5
26.1
0.2274
5
58.3
11.6
29.0
0.3172
6
70.0
5.7
14.3
0.1882
U.S. Average
Lifetime Exposure
32.1 years
E = 200.8
E = 1.136
-------
Probability of Indicated Years of Exposure Based
on 6 Moves per Lifetime, Mobility-based Model
70 years (5.7%) -.
58.3 years (11.6%)
46.7 years (10.5%)—/
35 years (9.0%)
23.3 years (16.5%)—
No Exposure (19.7%)
11.7 years (27,1%)
Exposed Population (millions) and Years of
Exposure, 6 Moves per Lifetime
Mobility-based Model
70 years (14.3 )m_
58.3 years (29 )—“ .,.
46.7 years (26.1 )_(
35 years (22.5)-
No Exposure (49.2)
r—117 years (67.7)
Figures 7.6 and 7.7 Distribution of Exposures and Population Exposed for the
Mobility-based Exposure Model
23.3 years (41.2 ) J
-------
7-25
Issue: Non-Threshold, Linear Dose-Response Relationship
SAB recommended that EPA provide a qualitative discussion on the issue of using linear
dose-response relationship extending to low dose. EPA considers the linear non-threshold
dose-response model to be most appropriate for estimating radiocarcinogenic risk of radon.
There is scientific consensus that ionizing radiation is a non-threshold carcinogen (NAS,
1980; NAS, 1988; NAS, 1990; UNSCEAR, 1986; UNSCEAR, 1988). There is additional
evidence that a threshold is not applicable for inhalation exposure to radon progeny. The
range of lifetime residential exposure includes cumulative exposures exceeding those in
some mines. Statistically significant increases in lung cancer have been reported in miners
with mean cumulative excess exposure of about 40 WLM.
While there is evidence in some experimental systems that the risk of low-LET
radiation are reduced at low doses and dose rates due to the operation of cellular repair
processes, the risks (per unit dose) of high-LET radiation appear to be maximal at low doses
and dose rates. The assumption of linearity of dose-response function of radon and its decay
products are based on current scientific consensus that the dose-response relationship for
high-LET radiation (e.g. alpha particles) is linear in the range of environmental exposures
(NAS, 1980; NAS, 1988; NAS, 1990; ICRP, 1987). The assumption of non-threshold,
linear dose-response relationship was endorsed by the Radiation Advisory Committee of the
SAB.
Issue: Influence of Smoking on Lung Cancer Risk from Radon
SAB recommended that EPA provide a qualitative discussion of the influence of smoking on
lung cancer risks from radon. Issues surrounding the possible form of radon-smoking
interaction and its possible effect on radon risk are complex and presently intractable given
the limited information available. No attempt has been made by EPA to quantify the
uncertainty due to this source. Rather, EPA outlined what appear to be the major problems.
Detailed description of the relationship between radon and smoking in causing lung cancer
may be found in EPA’s Technical Support Docwne,u for the 1992 Citizen’s Guide to Radon
(USEPA, 1992i)
Information on the interaction between smoking and ionizing radiation in causing lung cancer
is somewhat conflicting. Human data are rather limited, including support for everything
from a multiplicative to a protective effect of smoking to radon risk. Animal studies, too,
are conflicting: one study showing a synergism with tobacco smoke (Chameaud et al., 1980);
another, a protective effect (Cross et aL, 1982).
-------
7-26
The EPA model of radon risk is based on the BEER IV assumption that radon and smoking
act multiplicatively in causing lung cancer. According to this model, the radon risk is
proportional to the baseline lung cancer rate, so that as the baseline various over time (due
largely to changes in smoking habits), the projected rate of radon-induced lung cancers varies
in parallel. Therefore, neglecting possible changes in average exposure rates, it is, according
to the model, the proportion of all lung cancers attributable to radon that remains constant
over time, not the absolute rate of radon-induced lung cancers.
It should also be noted that risks calculated are based on 1980 lung cancer mortality rates.
These rates are continually changing in response to changes in smoking patterns. Those
limits the ability of the current risk estimates to predict future radon-induced lung cancer.
It is important to determine the risk in smokers and non-smokers since there is substantial
difference in the projected population risk and its interpretation, depending on the nature of
smoking-radon interaction. The evidence available today indicates that reducing radon decay
products will reduce lung cancer mortality in both smokers and nonsmokers. The Agency is
continuing to review data on radon decay products exposure and lung cancer as it is
published.
7.2 RISK ISSUES RELATED TO TREATMENT AND REMOVAL
ISSUE: Risks Associated with Aeration Treatment
for Radon Removal
EPA agrees with SAB’s comment that a more vigorous analysis of the risks associated with
the use of air-stripping to the remove radon from drinking water could improve the scientific
credibility of its assessment. However, the Agency also agrees with the SAB conclusion that
such an assessment would not change the Agency’s conclusion that the risk for a maximally
exposed individual would not exceed that attributable to a radon concentration of 300 pCi/L
in drinking water used in a home. Therefore, the Agency does not consider such an
assessment critical to adequately characterize the risks from this method of treatment.
-------
7-27
Issue: Risks Related to the Use of Granular Activated Carbon
For Radon Removal
SAB recommended that EPA analyze the potential risks and potential disposal problems
related to the use of granular activated carbon (GAC) for removing radon from drinking
water. EPA has not determined that GAC is an appropriate technology for efficient radon
removal at public water supplies. In addition, the Agency maintains that cost, operational
and waste management concerns will likely undermine application of GAC for radon
treatment.
The Chafee Amendment Report provides some basis for a comparison of cost for small
systems that install aeration and GAC treatment. An approximate comparison may be found
in the Chafee Report, 6/16/93 draft: GAC at $6.60/kgal and PTA at $l.90/kgal for a very
small system. The large carbon filter bed required for efficient radon removal would make
GAC very uneconomical to apply. Other concerns that the Agency cites as justification for
precluding GAC as a BAT in this case include: (1) the potential for elevated exposure to
gamma emissions from the GAC treatment unit and radioactivity in wasted GAC, and (2) the
uncertainty of future regulations (State and Federal) on wastes containing elevated natural
radioactivity. A revision of an EPA guidance that was reviewed by the SAB, currently in
draft form (EPA, 1993i), provides an analytical tool for estimating radioactivity in water
treatment wastes, including GAC media. This is meant to provide insight prior to selection
of treatment. The draft document also discusses disposal problems and regulatory issues. A
complete analysis of potential risks due to use of GAC treatment would not be feasible
because there would not be sufficient data to support such an analysis.
Issue: Risks Associated with Occupational Exposures to Radon
from Water Treatment
SAB recommended that EPA provide an analysis of occupational exposures as a result of
water treatment for radon. The predominant treatment for radon involves aeration
technology. EPA believes that aeration technologies typically discharge radon to the
atmosphere where it is quickly dispersed, preseI t1ng minimal risks to workers in the
treatment facilities. Occupational protection for treatment facility workers is discussed in the
draft guidelines Suggested Guidelines for Disposal of Drinking Water Treatment Wastes
Containing Radioactivity (USEPA, 1993i). These guidelines seeks to assure safe handling,
transport and disposal of wastes containing radioactivity.
-------
7-28
7.3 Issues Related to Cost
BAT for Radon in Water
EPA explained in the proposed regulation the choice of aeration treatment as BAT for
removal of radon in drinking water. This decision is largely due to the volatility of radon in
water, a physical characteristic that makes radon amenable to aeration treatment Aeration us
BAT may include packed tower aeration (PTA), diffused bubble aeration, spray aeration, and
other treatment technologies that enhance air-liquid contact, allowing for effective transfer of
radon out of water. Aeration is the most efficient and economical treatment for removal of
radon and it is commonly employed for treatment or pre-treatrnent of water containing other
undesirable constituents.
In regard to implementation of a radon in drinking water standard, one issue cited by
SAB has been a perceived risk associated with the emissions of radon gas from treatment
sites. The risks of vented radon from sites treating for high levels of radon in water has been
estimated by EPA. This was done by application of various dispersion and risk assessment
models, as explained in the proposed regulation, and further discussed with the Science
Advisory Board’s Radiation Advisory Committee (SAB/RAC), The SAB in 1993 (RAC
Committee Report) determined that it generally agreed with EPA’s analysis and conclusion
that the actual risk to human health would be quite small in magnitude. EPA’s models
predicted maximum risks to individuals that were 100 to 10,000 times lower than the risks
posed by delivery of untreated water to homes. The SAB, while agreeing with EPA’s
analysis, recommended that EPA scientifically upgrade its uncertainty analysis of the risks
related to aeration treatment. However, EPA does not believe that this would add any
significant value to the results of the analysis or to the regulation, or contribute toward
protection of public health.
It is not evident, in EPA’s view of the likely circumstances, that the perception of nsk
due to radon in aeration off-gas would play a significant role in local decision-malung. Some
commenters on the proposed radon regulation (for example, California water industry
representatives) suggested that emissions control requirements may limit local applications of
aeration technology in congested neighborhoods, if radon release is deemed excessive. First,
EPA finds that the off-gas risks would not be significant at a typical treatment site or even at
a site removing moderately high levels of radon in water. EPA is not aware of any local air
quality board that actually does control or otherwise restricts radon emissions from water
treatment plants. Analyses by EPA and others using available atmospheric dispersion and
associated risk models shows that local regulators would not be compelled to restrict the use
of properly designed drinking water aeration plants emitting radon gas.
-------
7-29
Another perceived problem which was addressed by the SAB (draft Radon Engineering
Cost Subcommittee Report) was the aesthetics of the packed column treatment unit. EPA
notes that if a community is not comfortable with the visual impact of a PTA treatment
system. other low-profile aerators are available. The other type aerators which were
mentioned above are quite effective, i.e., can generally remove 90 to 99 percent radon from
water, and are cost-competitive with PTA.
The most serious concern regarding BAT for radon in water involves the implication
that some small (low-flow) treatment systems may employ granular activated carbon (GAC)
as an alternate to aeration treatment. The SAB (draft RECS and RAC Reports) advises EPA
that it should review its BAT decisions under the radon regulation, and produce costs related
to GAC treatment. The SAB advises EPA that GAC may be used for radon reduction to a
larger extent than anticipated by EPA. The SAB (RAC report) also expresses a concern that
EPA has not provided a risk analysis to quantify the hazards posed to treatment workers and
to others during the disposal of GAC wastes. Given this, the Agency does not believe it is
appropnate to expend the limited time and resources to develop GAC treatment costs and a
corresponding risk analysis. A risk analysis in particular would be quite speculative and not
sufficiently quantitative, due to a lack of scientific and engineering data and analytical
methods that would be required to lend credibility to such an analysis.
The Agency Continues to believe that other technical and cost considerations will
preclude the wide scale application of GAC for radon treatment. EPA presented to SAB a
preliminary analysis of the problems related to GAC treatment for radon (draft Technical
Memorandum, 2/11/93). EPA has not specified GAC as BAT because it is not a cost
effective treatment for radon in water. The large carbon filter bed required for efficient radon
removal would make GAC uneconomical to apply, particularly for larger flows. Even at
smaller installations in EPA’s judgement GAC would not be the technology of choice due to
cost, primarily capital cost. Other concerns that the Agency cites as justification for
precluding GAC as a BAT in this case include: (1) the potential for elevated exposure to
gamma emissions from the GAC treatment unit and radioactivity in wasted GAC, and (2) the
uncertainty of future regulations (state and federal) on wastes containing elevated natural
radioactivity.
Also pertinent to this discussion is the 1992 review by SAB of EPA’s 1990 guidance
for disposal of radioactive water treatment wastes (Letter to W.K. Reilly from R.C. Loehr and
O.F. Nygaard, September 30, 1992). That review produced recommendations on the 1990
disposal guidance, i.e., that EPA perform risk assessments on treatment wastes, collect data
on waste volumes and radioactivity, clarify the occupational guidance, and reevaluate the
criteria for disposal of Lead-210, the long-lived radioactive progeny of radon. EPA is
addressing the SAB recommendations on the waste guidance to the maximum extent feasible,
and will update and reissue the guidance document at the time of the final radionuclides in
drinking water standards. The Agency is considering adding to the reissued guidance and to
the preamble to the final regulation advice that water suppliers avoid use of technologies that
have the ability to concentrate to an extreme degree radionuclides in treatment media, thus
advising that GAC should not generally be applied for radon removal at public water supplies.
-------
7 -10
Treatment Costs
EPA has utilized packed column or packed tower aerator (PTA) costs as a basis for
determining the feasibility of implementing a radon standard. Relative to other water
treatments used, e.g., water softening technologies, PTA is less expensive and appears
affordable for the full range of very small to large water systems. The SAB generally agreed
that EPA’s cost estimating processes ase reasonable, but cited some differences between the
aeration cost model and the models put forth by water industry groups.
En addition, other aeration treatments are quite cost competitive and will likely be used
instead of PTA, particularly at small installations (i.e., less than approximately I MGD design
flow). EPA’s draft technologies and costs document reviews the technical bases for the PTA
cost estimates and draws a close cost comparison with one of the modular, low-profile
diffused bubble aerators currently on the market.
The Agency has recognized, in the proposed regulation and in subsequent analyses, that
aeration would be the most applicable treatment, and that some percentage of systems would
be required to add disinfection treatment to ensure the safety of water delivered that has been
opened (through aeration) to the atmosphere. The SAB (RECS Report) advises EPA in this
regard, i.e., that the Agency should include adequate allowances and costs for this additional
treatment.
While some states (approximately half) may already require disinfection of ground
water supplies, others may require some or all systems that aerate water to add disinfection
treatment as a precaution. That decision may be based upon a judgement of the likely degree
of exposure of the water to microbiological contaminants dunng treatment, and may be
specific to the treatment site and design. It may also be related to the Recommended
Standards for Water Works (1992), which suggests some degree of disinfection for ground
water supplies (depending upon protection and type of treatment employed). EPA estimated
in the impact analysis for the proposed rule that since approximately half of the states already
require disinfection of ground water supplies, and since some aerator designs afford some
protection from particulate and microbiological contamination through protection of air
intakes, approximately 25 percent of those affected by a radon standard would be assumed to
require additional disinfection treatment with the aeration treatment regime. National costs
estimates in 1991 were conducted under that assumption. However, commenters on the
proposed radon regulation (Amencan Water Works Association, Ocotber 1991) pointed to
data from the 1986 Community Water Survey that may be used to more accurately estimate
current disinfection practices. Application of that data, plus EPA’s judgement regarding the
potential need to disinfect aerated water supplies has produced a revised profile: it is now
estimated that the large water supplies would require less disinfection than previously
estimated, i.e, 10 percent (versus 25 percent), and that the smallest systems would add more
disinfection in comparison to the earlier estimate, i.e., 50 percent (versus 25 percent), since
small systems are less likely to be currently disinfecting water. In addition, EPA costs reflect
the assumption that minimal contact times would be required for the purpose of disinfecting
water that has been treated by aeration.
131
-------
7-31
The EPA cost model for estimating system level costs to aerate water for radon
reduction came under considerable scrutiny by the public and by the SAB (RECS
Committee). Some commenters have said that EPA cost estimates are too high, while the
majority of industry comments asserted that EPA costs are too low. There is a wide margin
in the cost estimating procedure for insertion of features that reflect considerable professional
judgement and factors and physical requirements that may or may not be related to radon
treatment per se EPA’s cost estimating goal is generally to reproduce engineered costs for
systems that will efficiently remove radon from water. While site work, excavation,
contractor and engineering fees are adequately represented, EPA does not include such site-
specific factors as acquisition of land, which would be very small or negligible in this case,
purely aesthetic site Lmprovements, local permitting, and replacement of parts that may be
replaced as a secondary result of the radon regulation.
The SAB recommended that EPA meet with industry groups to resolve cost estimating
differences and participate in a water works sponsored research project to collect field data on
radon treatment costs. EPA assisted in that regard by reviewing the industry research
proposal and questionnaire developed for acquiring costs in the field. In regard to meeting
with industry groups, the Agency has given the water works industry ample opportunity on
several occasions since the time of the proposal to meet with Agency officials and discuss
differences. EPA has reviewed all data submitted by the major industry groups and has
adjusted its final cost estimates to reflect new data and public comments.
i q
-------
BIBLIOGRAPHY
Bair, Wi 1991. Overview of ICR Respiratory Tract Model. Radiat. Protect. Dosim.
38 147-152
Chameaud. I., R. Masse, M. Morin, J. Laufuma. 1980. Lung Cancer Induction by Radon
Daughters in Rats. In Occupational Radiation Safety in Mining. Volume 1 . H. Stokes, ed.
Canadian Nuclear Association, Toronto, Canada.
Correia, J. A., et al. 1988. The Kinetics of Ingested Radon 222 in Humans Determined from
Measuiements with Xe-133. Project Summary. PB 88-145297/AS.
Crawford-Brown, D.J., 1990. A Calculation of Organ Burdens, Dose and Health Results from
Rn-222 Ingested in Water (Draft). Report to the Office of Dnnking Water, U.S. Environmental
Protection Agency.
Cross, F.T.. et al. 1982. Carcinogenic Effects of Radon Daughters in Uranium Ore Dust and
Cigarette Smoke in Beagle Dogs. Health Physics 42:35-42.
Dunning DE, Leggett RW, Yalcintas MG. 1980. A Combined Methodology for Estimating Dose
Rates and Health Effects from Exposures to Radioactive Pollutants. Oak Ridge National
Laboratory. TM-7 105.
Ershow, A.G., and K.P Cantor. 1989. Total Water and Tapwater Intake in the United States:
Population-Based Estimates of Quantities and Sources. Report prepared under National Cancer
Institute Order 0263-ND-8 10264.
Gesell, T.F., and H.M. Prichard. 1980. The Contribution of Radon in Tap Water to Indoor
Radon Concentrations. In: Gesell T.F., Lowder W.M., eds. Natural Radiation Environment HI ,
Vol 2. Washington, DC: U.S. Department of Energy, Technical Information Center, pp.
1347-1363 CONF-780422 (Vol. 2).
Hess, C T., and W L. Brown. 1991. The Measurement of the Biotransfer and Time Constant of
Radon from ingested Water by Human Breath Analysis. Final report to U.S. Environmental
Protection Agency under cooperative agreement CR 815156-01.
Hopper, R. 1991. National Ambient Radon Study. Proceedings of the 1991 International
Symposium on Radon and Radon Reduction Technology. Office of Radiation Programs, U.S.
Environmental Protection Agency.
Hursh, J.B., D.A. Morken, T.P. Davis, and A. Lovaas. 1965. The Fate of Radon Ingested by
Man. Health Phys. 11.465-476.
[ ARC. 1988. Monograph on the Evaluation of Carcinogenic Risks to Humans, Volume 43:
Man-made Mineral Fibers and Radon. [ ARC, World Health Organization.
-------
2
ICRP. 1987 Lung Cancer Risk from Indoor Exposures to Radon Daughters. Pergamon Press,
Oxford, U.K ICRP Publication 50
Israeli. M. and C. Nelson. 1992. Distribution and Expected Time of Residence for U.S.
Households. Risk Analysis , Vol. 12, No. 1. pgs. 65-72.
NAS. 1980. National Academy of Sciences - National Research Council. The Effects on
Populations of Exposure to Low Levels of Ionizing Radiation, Report of the Committee on
Biological Effects of Ionizing Radiations (BEIR III). Washington, D.C.: National Academy of
Sciences.
NAS. 1988. Health Risk of Radon and Other Internally Deposited Alpha-Emitters: BEIR IV
National Academy Press, Washington, DC.
NAS. 1990. Health Effects of Exposure to Low Levels of Ionizing Radiation (BEIR V).
National Academy of Sciences - National Research Council. Washington, DC, National
Academy Press.
NAS. 1991 National Academy of Sciences - National Research Council. Comparative
Dosimetry of Radon in Mines and Homes. Washington, D.C: National Academy of Sciences.
Nazaroff, W.W., S.M. Doyle, A.V. Nero, and R.G. Sextro. 1987. Potable Water as a Source of
Airborne Rn-222 in U.S. Dwellings: A review and assessment. Health Physics. 52:281-289
Nd. 1981. Surveillance, Epidemiology, and End Results: Incidence and Mortality Data
/973-1977. National Cancer Institute. Monograph No. 57. NIH Publication No. 81-2330.
NCRP. 1984. Evaluation of Occupational and Environmental Exposures to Radon and Radon
Daughters in the United States. NCRP Report 78.
NIOSH. 1987. Radon Progeny in Underground Mines. A Recommended Standard for
Occupational Exposure. DHHS (N1OSH) 88-101.
Partridge, J E., T.R. Horton, and EL. Sensintaffar. 1979. A Study of Radon-222 Released front
Water During Typical Household Activities, Montgomery, AL. U.S. Environmental Protection
Agency ORP/EEFR-79- 1.
Pennington, J.A.T. 1983. Revision of the Total Diet Study Food List and Diets J. Am. Dietetic
Assoc. 82:166-173.
Radon, Radium and Uranium in Drinking Water. 1990. eds. J. Longtin, C.R. Cothern, and P.
Rebers. Lewis Publishers, Chelsea, MI.
Sullivan, R.E., N.S. Nelson, W.F1. Ellet, D E. Dunning Jr., R.W. Leggett, M.C. Yalcintas, and
K.F Eckem-ian. 1981. Estimates of Health Risk from Exposure to Radioactive Pollutants. Oak
Ridge National Laboratory. TM-7745.
-------
I
Suomela. M., and H. Kahlon. 1972. Studies on the Elimination Rate and the Radiation
Exposure Following Ingestion of 222-Rn-Rich Water. Health Phys. 23:641-652.
UNSC EAR. 1988 United Nations Scientific Committee Ofl the Effects of Atomic Radiation
Sources, Effects and Risks of Ionizing Radiation. New York: United Nations.
UNSCEAR. 1986. United Nations Scientific comnuttee on the Effects of Atomic Radiation.
Genetic and Somatic Effects of ionizing Radiation. Report to General Assembly with Annexes
New York: United Nations.
USEPA. 1984a. Radionuclides: Background Information Document for Final Rules. Volume 1.
EPA 520/1-84-0221.
USEPA. 1984b. An Estimation of the Daily Average Food Intake by Age and Sex for Use in
Assessing the Radionuclide intake of individuals in the General Population. Office of Radiation
Programs, U S Environmental Protection Agency. EPA 520/1-84-02r.
USEPA. 1987. The Risk Assessment Guidelines of 1986. Office of Health and Envuonmental
Assessment, U.S. Environmental Protection Agency. EPA/600/8-87/045.
USEPA. 1988. Preliminary Risk Assessment for Radon Emissions from Drinking Water
Treatment Facilities. Memorandum from Warren D. Peters, U.S. Environmental Protection
Agency Office of Air Quality Planning and Standards to Stephen W. Clark, U.S. Environmental
Protection Agency Office of Ground Water and Drinking Water. June 28, 1988.
USEPA. 1 989a. An Analysis of Potential Radon Emissions From Water Treatment Plants Using
the MJNEDOSE Code. Memorandum from Marc J. Parrotta, U.S. Environmental Protection
Agency Office of Ground Water and Drinking Water to Greg Helms, Office of Ground Water
and Drinking Water, U.S. Environmental Protection Agency. November 22, 1989.
USEPA. 1989b. U.S Environmental Protection Agency. Office of Health and Environmental
Assessment. Exposure Factors Handbook . Washington, DC: U.S. Environmental Protection
Agency. EPA/600/8-89/043.
USEPA. 1989c. U.S. Environmental Protection Agency. Office of Radiation Programs. Risk
Assessment Methodology, Environmental Impact Statement for NESHAPS Radionuclides. Vol 1
Background information document. Washington, D.C. United States Environmental Protection
Agency. EPA 520/1-80-005.
US EPA. 1991 a. Drinking Water Criteria Docunient for Radon in Drinking Water. (Draft).
Prepared by Life Systems, Inc. June 1991
USEPA. 199 lb. Estimate of Average Equilibrium Fraction (F). Office of Radiation Programs,
U.S. Environmental Protection Agency Presented to the Radiation Advisory Subcommittee of
the Science Advisory Board. May 20, 1991
-------
4
USEPA. 199 Ic. I-Iealth Effects Assessment Summary Tables, Annual FY-1991. OERR 92006-
303 (91-1). NTIS PB9I-921199
USEPA. 199 Id. Memorandum to G Helms from I. Deloatch Regarding Estimated Gost of
Analyses for Radionuclides. March 27, 1991.
US EPA. 199 le National Primary Drinking Water Regulations; Radionuclides: Notice of
Proposed Rulemaking. FR Vol. 56, No 138. July 18, 1991.
USEPA. 199 If. National Residential Radon Survey Preliminary Results. Office of Radiation
Pro2rams, U.S. Environmental Protection Agency. Presented to the Radiation Advisory
Subcommittee of the Science Advisory Board May 20, 1991.
USEPA. 1991g. Occurrence and Exposure Assessment of Pb-2]O in Drinking Water, Food, and
Air Addendum to Occurrence of Man-Made Radionuclides in Public Drinking Water Supplies
(Draft). Prepared by Wade Miller Associates, Inc for U.S. Environmental Protection Agency.
May 20, 1991.
USEPA. 199 [ h. Occurrence and Exposure Assessment for Gross Alpha Particle Activity
(Draft). Prepared by Wade Miller Associates. May 13, 1991.
USEPA. 1991 i. Proposed Revisions in EPA Estimates of Radon Risks and Associated
Uncertainties. Api-il 1991.
US EPA. 199 Ij. Radon in Drinking Water: Assessment of Exposure Pathways. Prepared by
Life Systems Inc., for Office of Water and Office of Radiation Programs, U.S. Environmental
Protection Agency. TR-1242-87. June 14, 1991.
USEPA 199 1k. Regulatory Impact Analysis of Proposed National Primary Drinking Water
Regulations for Radionuclides. Prepared by Wade Miller Associates, Inc. for the Office of
Drinking Water, U.S. Environmental Protection Agency. July 17, 1991.
USEPA. 19911 Technologies and Costs for the Removal of Alpha Emitters from Potable Water
Supplies. Prepared by Malcolm Pirnie, Inc. for U.S. Environmental Protection Agency.
February 1991.
USEPA. I 992a. Addendum to: The Occurrence and Exposure Assessments for Radon, Radium-
226, Radium-228, Uranium, and Gross Alpha Particle Activity in Public Drinking Water
Supplies (Revised Occurrence Estimates Based on Comments to the Proposed Radionuclides
Regulations). (Draft Document). Prepared by Wade Miller Associates, Inc. September 30,
1992
USEPA 1992b. Cost Modeling Update. Memorandum from Marc Parrotta to EPA drinking
water engineers. February 21, 1992.
USEPA. 1992c. National Residential Radon Survey: Summary Report. October 1992.
-------
5
USEPA 1992d. Packed Tower Aeration Cost Estimates for Radon Renuwal Memorandum
from Michael Cummins to Marc Parrotta. March 11, 1992.
USEPA I 992e. Proposed Methodology fr Estimating Radio genic Cancer Risks. Office of
Radiation Programs, U. S. Environmental Protection Agency. Submitted to Radiation Advisory
Committee. May I, 1992.
USEPA. 1 992f. Reevaluation of EPA ‘s Methodology fir Estimating Radiogenic Cancer Risks.
Transmittal from M. Oge to D.G. Barnes. Office of Radiation Programs, U.S. Environmental
Protection Agency. January 10, 1992.
USEPA. 1992g. Safeguarding the Future; Credible Science, Credible Decisions. The Report of
the Expert Panel on the Role of Science at EPA, [ Panel members are Raymond C. Loehr,
Chairman, Bernard D. Goldstein, Anil Nerode and Paul G. Risser] EPAJ600/9-91/050 January
8, 1992.
USEPA. 1992h. Simplified Equations for Estimating Radon Removal Cost via Packed Tower
Aeration. Memorandum from Michael Cummins to Marc Parrotta. July 16, 1992.
USEPA. 1992i. Technical Support Document for the /992 Citizen’s Guide to Radon. Office of
Air and Radiation, U.S Environmental Protection Agency . EPA 400-R-92-0l 1. May 1992.
USEPA. 1 992j. Technologies and Costs for the Removal of Radionuclides from Potable Water
Supplies. (Draft Document) Prepared by Malcorn Pirnie, Inc. for the Drinking Water
Technology Branch, Office of Ground Water and Drinking Water, U.S. Environmental Protection
Agency. July 1992.
USEPA. l993a. Briefing MateriaLs. Prepared by Dr. Janet A. Auerbach, Mr. James M. Conlon.
Mr. Michael Cummins, Mr. Frank Marcinowski, and Mr. Marc J. Parrotta. February 8, 1993.
USEPA 1993b. Estimation of Risk Uncertainry. Memo from Henry D. Kahn and Helen Jacobs
to Nancy Chiu. January 21, 1993.
USEPA. 1993c. Radon Report to Congress Update and Conclusions. Rad Report Talk Draft
U.S. Environmental Protection Agency. April 15, 1993.
USEPA. 1993d. Review of Risk Assessments of Radon Emissions from Drinking Water
Treatment Facilities. Memorandum from Chnstopher B. Nelson, Office of Radiation and
Indoor Air, U.S. Environmental Protection Agency, to Marc J. Parrotta, Office of Ground Water
Drinking Water, U.S. Environmental Protection Agency. January 28, 1993.
US EPA 1 993e. Revised EPA Methodology for Estimating Radiogenic Cancer Risks. Office of
Radiation and Indoor Air. In preparation.
USEPA 1 993f. Risk Assessment for Radon Emissions from Drinking Water Treatment
Facilities Office of Radiation and Indoor Air, U S. Environmental Protection Agency
February 17, 1993.
!LK
-------
6
USEPA. 1993g. Technical Memorandum: Problems with the Use of GAC for Radon Removal.
(Draft). Marc Parrotta, Drinking Water Standards Division, U.S. Environmental Protection
Agency. January, 1993.
USEPA. 1993h. Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking
Water (Draft 2) Prepared by Office of Science and Technology, Office of Radiation and
Indoor Air, Office of Ground Water and Drinking Water, and Office of Policy Planning and
Evaluation, U S. Environmental Protection Agency. January 29, 1993.
USEPA. 1993i. U.S.Environmental Protection Agency. Office of Ground Water and Drinking
Water. Suggested Guidelines for Disposal of Dnnking Water Treatment Wastes Containing
Radioactivity. (Draft Document, July 1993)
USEPA. Working Draft of the Regulatory Impact Analysis for Final NPDWR fcr Radionuclide.s.
Prepared by Wade Miller Associates for Office of Drinking Water, U.S. Environmental
Protection Agency. To be published.
------- |