Report to the United States Congress on Radon in Drinking Water: Multimedia Risk and Cost Assessment of Radon


       United States
       Environmental Protection
       Agency
Office of Water
4601
EPA811-R-94-001
March 1994
wEPA  REPORT TO THE
       UNITED STATES CONGRESS ON
       RADON IN DRINKING WATER

       MULTIMEDIA RISK AND COST
       ASSESSMENT OF RADON
                                      Recycled/Recyclable
                                      Printed with Soy/Canola Ink on paper that
                                      contains at least 50% recycled fiber

-------

-------
                      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 WATER SUPPLIES	   2-1

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 SAB COMMENTS	,	...,,,...,.,....	 5 .   6-1

  Radiation Advisory Committee (RAC) Comments
  Drinking Water Advisory Committee (DWAC) Comments
  Executive Committee Comments

7 EPA DISCUSSION OF ISSUES RAISED BY  SAB 	   7-1


BIBLIOGRAPHY
 ATTACHMENTS

-------

-------
EXECUTIVE SUMMARY
OVERVIEW

Radon, a naturally occurring gas, is colorless, odorless, tasteless, chemically inert and
radioactive. People are exposed to radon primarily in their homes from radon gas seeping up
from the soil. People can also be exposed to radon by drinking tap water or by inhaling radon
released into indoor air from tap water used for showering, washing, or other domestic uses, or
when the water is stirred, shaken, or heated before being ingested. Radon is second only to
cigarette smoking as a leading cause of lung cancer in the United States.

In 1992, Congress directed EPA to report on the risks from exposure to radon, the costs
to control this exposure and the risks from treating to remove radon. This report presents the
findings in response to that Congressional directive. The following table is a summary of
EPA's risk estimates, cancer cases avoided and costs for both air and water.
Summary of EPA's Estimates of Risk, Fatal Cancer Cases, Cancer Cases Avoided,
and Costs for Mitigating Radon in Water and Air

The voluntary air program is estimated to have avoided 100 fatal cancer cases per year based on 1992 data.
Annual lung cancer^cases averted by the voluntary air program are expected to increase each year as
additional mitigations occur and new construction is built to be radon resistant. The air program has the
potential to avert 2,200 cancer cases per year assuming 100% voluntary monitoring and mitigation.

-------
EPA estimates that approximately 19 million people are exposed to a radon level above
the proposed drinking water standard, or Maximum Contaminant Level (MCL), of 300
pCi/L^tcr. Approximately 2 of every 10,000 individuals exposed at 300 pCi/L would develop a
fatal case of cancer as a result of exposure to radon at this level. Approximately 15 million
people are exposed to airborne residential radon at levels above EPA's voluntary indoor air
action level of 4 pCi/L^. The 4 pCi/1 action level is equivalent to a water exposure level of
40,000 pCi/1. It is estimated that approximately 1 in every 100 individuals could suffer lung
cancer deaths as a result of exposure to radon at this level.

The total annual cost to treat radon in drinking water is $272 million. The regulated
industry has estimated higher costs than EPA. The American Water Works Association
(AWWA) estimated national costs at $2.5 billion per year. The Association of California Water
Agencies (ACWA) estimated annual costs of $ 520 million for California alone. The major
differences in Agency and industry cost estimates result from differences in the number of water
systems affected by the proposed standard, differences in the treatment costs that would be
incurred by a typical public water system to comply with the proposed standard, and the interest
rate charged for the purchase of treatment equipment. The biggest differences are in treatment
costs. EPA has modified its original cost estimates in response to industry comments. For
example, EPA's revised cost estimates recognize that many public water systems relying on
ground water will need to disinfect the water once they draw it out of the ground to aerate it.
The revised costs also recognize that most ground water systems rely on more than one well
and, therefore, need to install treatment at more than one location. Industry's estimates are mores
typical of additional costs likely to be incurred by large1 systems (e.g., higher labor rates in
urban areas and more engineering design work rather than purchase of off-the-shelf designs).
Although each party's estimate of the cost of a treatment technology may be reasonable in and
of itself, EPA believes its estimates better reflect likely industry practice for the small systems
which are most affected by the rule.

As required by Congress, the Science Advisory Board (SAB) reviewed EPA's study. The
SAB noted the cost differences discussed above and also noted that EPA had employed a
reasonable approach to the analysis of occurrence data, technologies, and costs as a function of
system size. In response to earlier SAB concerns about uncertainties in the radon risk
assessment, EPA conducted a quantitative uncertainty analysis pf the risks associated with
exposure to radon in drinking water. While noting some uncertainties continue in the risk
assessment, the SAB committee reviewing this analysis cited it as "state of the art". The
uncertainty analysis is more explicit for radon now than for any previous rule, not because of
any greater uncertainty in these estimates but in response to the SAB's request and EPA's
decision to present uncertainty more explicitly in future rulemaking.

The cancer risks from radon in both air and water are high. While radon risk in air
typically far exceeds that in water, the cancer risk from radon in water is higher than the cancer
risk estimated to result from any other drinking water contaminant. This report is the most
comprehensive assessment to date on radon exposure and risk and forms a sound foundation for
policy decision-making.

-------
SUMMARY OF FINDINGS

EPA prepared this report in response to the Congressional mandate in Public Law 102-389
(the Chafee-Lautenberg Amendment to EPA's Appropriation Bill, enacted October 6, 1992)
which 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 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. In EPA's Appropriation Bill, Congress
required EPA to (1) report on the risk of adverse 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 regula-
tion; (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 recom-
mendation 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.

Congress placed these requirements 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 inert and
radioactive. People can be exposed to waterborne radon either by ingestion or inhalation.
When ingested, radon is distributed throughout the body, which increases the cancer risk to
many organs. Radon also is released into indoor air from 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, or progeny, 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 drinking water in
perspective.

People are exposed to waterborne 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. EPA estimates that an individual's

-------
combined risk during a lifetime from constant use of drinking water with one picocurie1 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 are determined partly 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.

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
(pCi/LwaterXlOjOOO pCi/L^ter is equivalent to 1 pCi/L^). Figure 1, on page xi, depicts the
portion of individuals impacted at various levels. Radon in water exceeds 100 pCi/L^,. in 72
percent of the ground water sources surveyed, but these sources serve only 60 percent of the
population. The number of small systems impacted is out of proportion to their numbers
because they generally rely on ground water, and tend to have higher radon concentrations.
Because smaller systems tend to have higher radon concentrations the burden of the costs for
mitigating radon in drinking water would weigh more heavily on the small systems. The
distribution of systems and population impacted in community ground water systems is
illustrated in Figure 2 on page xii. :

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. When radon and its progeny
decay in the body, the surrounding tissues are irradiated by alpha particles. However, the doses
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.

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 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.
'A curie (Ci) is a standard measure of radioactivity, and a picocurie (pCi) is one trillionth (1 x 10'12) of a curie.

-------
     Assessing the risks from inhaling radon progeny from drinking water 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 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 exposure of radon progeny (expressed in working level months); and 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 from drinking water 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.

     Since the time the Proposed Rule was developed and published in July 1991, EPA has
used new data on radiation dosimetry and risk to improve the accuracy of the calculations.  In
terms of lung cancer death (LCD) inhalation risk of radon  progeny, EPA revised the dose
estimate per pCi/L of radon in air about 30 percent lower based on the dosimetry differences
between the mines and homes in the 1991 NAS report Comparative Dosimetry of Radon in
Mines and Homes.  Also, EPA modified the risk  model in estimating the LCD  inhalation risk
factor based on the recommendation of the Radiation Advisory Committee (RAC) of the Science
Advisory Board (SAB).  These modifications reduced the LCD risk from inhalation of radon
progeny by 38 percent.

     The risk for ingested radon has also been revised, based on revised organ-specific risk per
unit dose estimates, and additional modifications of intestinal and lung dosimetry treatment.
The result was to increase the ingestion risk by a factor of 2.3, mostly because  of increased risk
estimates of stomach and  colon cancers.  These new risk estimates are shown in the following
table.

-------
              Summary of Proposed and Revised Fatal Cancer Risk Estimates
                                   For Radon in Water
Lifetime Cancer Risk per pCi/L in Water
Exposure Pathway
Inhalation of Radon Progeny Derived
from Waterborae Radon Gas
Inhalation of Radon Gas Released from
Water to Indoor Air
Ingestion of Radon Gas in Direct Tap
Water
Sum of All Pathways
Proposed
4.9 x ID'7
(74%)
0.2 x IQr*
(3%)
l.SxlO'7
(23%)
6.6 x lO'7
(100%)
Revised
3.0 x lO'7
(45%)
2.0 x 10-"
(3%)
3,5 x lO'7
(52%)
6.7 x 10'7
(100%)
     Based on SAB's recommendation, EPA also conducted a quantitative uncertainty analysis
of the risks associated with exposure to radon in drinking water.  This analysis quantifies the
uncertainties in exposure and toxicology and estimates variation in exposure among individuals.
This analysis was reviewed by SAB and further expanded in Section 7.1 of the report) based on
SAB's recommendations.

     The combined lifetime fatal cancer risk per pCi/L 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 10~7 with a
credible range of 2.6 x 10~7 to 1.8 x I'Qf6. .EPA's nominal estimate for the individual lifetime
inhalation risk of lung cancer  deaths per pCi/L of radon in drinking water is 3.0 x 10'7 with a
median of 3.9 x  10'7 and a credible range of 1.4 x 10'7 to 1.4 x  10"6.  EPA's nominal estimate
for the individual lifetime ingestion risk of fatal cancers per pCi/L of water is 3.5 x 10"7 with a
median of 1.7 x  10'7 and a credible range of 3.7 x 10'8 to 7.4 x  10'7.  The combined  lifetime
fatal cancer risk per pCi/L in  water from inhalation of radon progeny, inhalation of radon gas
released from water to indoor air and ingestion of radon gas in direct tap water is 6.7 x 10'7,
with a median of 6.5 x lO'7 and a credible range of 2.7 x 10'7 and 1.8 x lO"6. These credible
ranges reflect an increase of approximately 17 percent over  the February 1993 analysis. Lastly,
EPA estimated that the individual lifetime fatal cancer risk for inhaling waterbome radon gas is
2 x 10*8per pCi/Lw,^.  Due to the small contribution (3%) to the overall risk, the uncertainty of
the risk from inhaling radon gas was not quantified.

     The following table shows the number of estimated cancer fatalities per year due to
various pathways of radon exposure, based on revised fatal cancer risk estimates and  occurrence
data.  EPA's nominal estimate of total lung cancer deaths caused by inhalation of radon progeny
in drinking water, for example,  is 86 (median of 113), with a credible range of 40 to 408 per
year.  EPA's nominal (or most likely) estimate of total fatal cancer cases caused by ingestion of
radon in drinking water is  100 (median of 46), with a credible range of 11 to 212 per year.

-------
                                            vu
The threat from radon in drinking water is about half (48 percent) due to inhalation and about
half (52 percent) due to ingestion of drinking water. These estimates are for the exposed
population served by community ground water systems. Due to the time constraint for
completing the report, risk estimates for waterborne radon via various exposure routes and their
credible range calculations were only performed for exposed populations served by community
ground water supplies and do not include non-transient, non-community (NTNC) ground water
supplies (such as schools or hospitals).  These calculations suggest that  the population risks from
radon in drinking water are similar to or higher than currently known risks from most chemical
pollutants in drinking water that are now subject to regulation.

                   Cancer Fatalities per Year due to Exposure to Radon


                                                       EPA's Nominal
Exposure Pathway
Inhalation due to Radon Treatment
Inhalation of Radon* Gas Released
from Water to Indoor Air
Inhalation of Radon* Progeny Derived
from Waterborne Radon Gas
Ingestion of Radon* hi Drinking Water
Inhalation from Outdoor Air
Inhalation from Indoor Air
Lower Estimate
—
—
40
11
280
6,740
Estimate
—
6
86
100
520
13,600
. Upper Estimate
<1
—
408
212
1,500
30,600
     * Estimates due to exposure through community water supplies only.
     The following table is a summary of EPA's risk estimates, cancer cases avoided and costs
for both air and water. EPA estimates that approximately 19 million people (17.2 million
people served by community ground water systems and 1.7 million people served by non-
transient, non-community ground water systems) are exposed to a radon level above the
proposed drinking water standard, or Maximum Contaminant Level (MCL), of 300
Figure 3 on page xiii depicts the nominal cases avoided at various radon levels in water
provided by community ground water systems only.  Approximately 2 of every 10,000
individuals exposed at 300 pCi/L would develop a fatal case of cancer as a result of exposure to
radon at this level.
     Approximately 15 million people are exposed to airborne residential radon at levels above
EPA's voluntary indoor air action level of 4 pCi/L^.  It is estimated that approximately 1 in
every 100 individuals could suffer lung cancer deaths as a result of exposure to radon at this
level.

-------
Summary of EPA's Estimates of Risk, Fatal Cancer Cases, Cancer Cases Avoided,
and Costs for Mitigating Radon in Water and Air

Number of Fatal Cancer Cases per Year*
Proposed Level or Target Level
Individual Lifetime Risk of Fatal Cancer
at Target Level
Total Number of People Above Target
Level
Number of Fatal Cancer Cases Avoided
Annually by Meeting the "Target" Level
Total Annual Cost for Mitigating Radon
Average Cost per Fatal Cancer Case
Avoided
- For largest size systems (serving
greater than one million persons)
- For smallest size systems (serving
between 25 and 100 persons)
Drinking Water
192*
300 pCi/L^,
2 in 10,000
19 million**
84**
$272 million
$3.2 million
$1.2 million
$7.9 million
Indoor Air
13,600
4 pCi/L^
1 in 100
15 million
100(2,200)"*
$1,504 million
$0.7 million
N/A
N/A
Includes those exposed above and below the target level in community ground water systems only.
ii
*" Includes cases in non-community systems as well as community systems.

*~ The voluntary air program is estimated to have avoided 100 fatal cancer cases per year based on 1992 data.
Annual lung cancer cases averted by the voluntary air program are expected to increase each year as
additional mitigations occur and new construction is built to be radon resistant. The air program has the
potential to avert 2,200 cancer cases per year assuming 100% voluntary monitoring and mitigation.

As noted above, the total annual cost is $272 million to treat radon in drinking water. The
following exhibit shows the national cost separated into total capital, annual amortized capital,
annual operation and maintenance (O&M), and total annual costs. It is important to note that
the regulated industry has estimated higher costs than EPA. The American Water Works
Association (AWWA) estimated national costs at $2.5 billion per year. The Association of
California Water Agencies (ACWA) estimated annual costs of $ 520 million for California
alone. The differences between the estimates are included in chapter 4 of the report. Figure 4
on page xiv depicts the total annual costs for different MCLs. 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 hi the largest water systems.

-------
                                             IX
                  National Costs for Controlling or Mitigating Radon in Water
                                    (millions of dollars*)
Type of Cost
Total Capital over 20 jrs
Annual Amortized Capital
Annual O&M
Total Annual
Proposed
Estimate
$1,579
106
74
180
Revised Best
Estimate
$1/02
151
121
272
          *Cost estimates include community and non-transient non-community systems.

     The following exhibit presents the estimates of annual costs for testing and mitigating
radon in indoor residential air.  EPA's best estimate of the national cost for addressing radon in
indoor residential air for a fully implemented voluntary program is $1,504 million annually
amortized (the period of time over which a loan will be paid off) over a 74 year period.  The
amortization period is based on the average life expectancy of the U.S. population and a time
period representative of at least the average life of a home. As the table indicates, most of the
costs (approximately 75 percent) are in the operation and maintenance of the system,  while only
25 percent of the costs are for the testing and initial installation of the mitigation systems.
           Annual Cost Estimates for Testing & Mitigating Radon in Indoor Air
                                (millions of dollars per year)
                        Type of Cost
  Best
Estimate
                        Annualized Testing Costs

                        Annualized Installation Costs

                        Annualized O&M

                        Total Annual Costs
   $90

   324

  1,090

  1,504
     EPA conducted an analysis of the cost per statistical cancer case avoided for different
hypothetical levels of regulatory control in community water systems.  Community water
systems are those which serve residences and comprise the bulk of the water systems which
would be affected by an EPA drinking water regulation. (Other systems which would be
affected by the rule but which are not included in this analysis are non-transient, non-community
water systems, or those which serve facilities such as schools, hospitals, and office buildings.)
This analysis is reflected in Figures 3 and 4. For each level of control,  Figure 3 shows the

-------
estimated number of cases avoided, and Figure 4 shows the estimated total national annual
costs. Using 300 pCi/1 as an example, the annual number of cases avoided on Figure 3 is 68,
while on Figure 4 the total national annual costs associated with 300 pCi/1 are estimated to be
$202 million. The cost per case avoided is 202 divided by 68, or approximately $3 million.
This analysis can be completed for various hypothetical levels of regulatory control.

When all water systems, including non-transient, non-community water systems are
considered, the total annual cost of mitigating radon exposure through drinking water is
calculated to be $272 million to prevent the deaths of approximately 84 people each year. This
is an average of $3.2 million dollars per life saved, as shown in the table on page viii. Based
on the total annual cost of mitigating radon exposure in indoor residential air, EPA calculates
that it would cost an average of $0.7 million per life saved (or $0.9 million at a 7 percent
discount rate) to prevent the deaths of approximately 2,200 people each year due to exposure to
radon in residential air.

The combined annual cost estimate is $1,776 million for controlling residential radon from
all sources. The component cost estimates for indoor air ($1,504 million dollars per year) and
drinking water (272 million dollars per year) are based on a 3 percent and a 7 percent discount
rate, respectively. If both the water and air cost estimates were calculated at a 3 percent
interest rate the costs would be $230 million per year for water and $1,504 million per year for
indoor air. If both the water and air cost estimates were calculated at a 7 percent
interest/discount rate the costs would be $270 million per year and $1,980 million per year
respectively. The drinking water cost estimate is based on a 20-year amortization period for
installation of treatment equipment at public water systems facilities, however, the indoor air
cost estimate is based on a 74-year amortization period. Comparing the national costs for air
and water is difficult because the air program costs are based on 100 percent compliance with a
voluntary program, whereas the cost for water are based on 100 percent compliance for public
water systems to meet the requirements of the Safe Drinking Water Act*.
* Most of the costs of mitigating indoor airborne radon are associated with operation and maintenance
(O&M) of the systems (75%). The estimates presented throughout the remainder of the document utilize
the 3 percent rate which yields an annual cost of $1,504 million for a fully implemented program and a
cost per cancer case avoided of $0.7 million. '

-------
XI
03
*-•
co
CO
c
Q
c

J2
0
>
0)
_J
Co
•a
Occ
*
03
DC
C
O
"•*-•
O
c
3
U.

as
CO
0)

o
QL
X
UJ
c
o
"•*-•
JO
a
Q.
O
Q_
v
CO
u_
0)
4-«
CO
•a
CD
o 0)
o «^

-------
xii

Figure 2A
w
Q
W
C!
O.S
o/A# Community Water Systems (CWS)
that exceed 300 pCi/l or greater.
15 is 6
10
1 0 _
25-100 01-600 601-1K 1K-3J3K .3Kr10K, 10K-2SK 28K-50K 60K-7SK 76Kr100K 100K< 500K-1M 1M+
Served fey Sy^ivt Size 600K
Pop. Exposed
tMHMMrJMMMMl *^
Systems Affected
2J?
o/A// Community Water Systems (CWS)
-a

-------
xiii
CD
+-•
CO

CD
C
Q
c
_w
CD
>
CD
o
"•«—•
o
c
ZJ
LL.
co
to
CO
-^
to
be
o
o
§
o
o
•o
o
Q.
C

CO
£
CD
••—•
to

CO
1_
0)
"CO
.-
C
E
£
o
•a
a>
o
Q.
\_
3
CO
JO
CD

CD
—I
C
o

co
DC
g
£
a
i
1
i

pspiOAy
-------
XIV
I
-------
PART ONE
INTRODUCTION
-------
-------
1. BACKGROUND AND INTRODUCTION
Public Law 102-389 (the Chafee-Lautenberg Amendment to EPA's Appropriation 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 costs 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.

1. 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 Two responds to Congress' direction to EPA to report on the human health
risks associated with various types of exposure to radon. Chapter Two explains
EPA's assessment of the risks from ingestion and inhalation of waterborne
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, and explains how
EPA determined the national and household cost estimates. Chapter Four also
discusses the different cost estimates from AWWA, ACWA, and EPA. Chapter
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, 199 la); National Primary
Drinking Water Regulations; Radionuclides: Notice of Proposed Rulemaking (USEPA, 1991e);
Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking Water (USEPA,
1993h); The Occurrence and Exposure Assessments for Radon, Radium-226, Radium-228,
-------
1-2
Uranium, and Gross Alpha Particle Activity in Public Drinking Water Supplies (Revised
Occurrence Estimates Based on Comments to the Proposed Radionuclides Regulations) (USEPA,
1992a); Technical Support Document for the 1992 Citizen's Guide to Radon (USEPA, 19921);
A memo from Marc Parrotta to Greg Helms Regarding An Analysis of Potential Radon
Emissions from Water Treatment Plants Using the Minedose Code (Nov. 22, 1989a); A Memo
from Christopher Nelson, Office of Radiation and Indoor Air, to Marc Parrotta regarding A
Review of Risk Assessments of Radon Emissions from Drinking Water Treatment Facilities (Jan.
28, 1993); and the Regulatory Impact Analysis of Proposed National Primary Drinking Water
Standards for 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, including 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. In the
fall of 1992, Congress passed the EPA's 1993 Appropriation Bill with a requirement that EPA
conduct risk and cost assessments of radon. This report is due to Congress 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
-------
1-3
Exhibit 1-1
Requirements of the Chafee-Lautenberg Amendment to EPA's Appropriation Act
In EPA's 1993 Appropriations Bill, Congress required EPA to:

(1) Report on the risk of adverse 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.
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
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 BEIR 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,
-------
1-4
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
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 soil and other sources. Although radon in
drinking water accounts for only 5 percent of the exposure, EPA is able to control that 5 percent
through reliable drinking water treatment technology. The 95 percent of radon in air can be
reduced through mitigation techniques, but the program to control it is currently a voluntary
program within EPA. Nonetheless, the drinking water risk from radon is significant and
controllable especially when viewed in the light of other risks already regulated under the Safe
Drinking Water Act.

Physical 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 enters the air, while some remains below the surface and dissolves in ground water
(water that collects and flows under the ground's surface). Due to their very long half-life (the
time required for half of a given amount of a radionuclide to decay), uranium and radium persist
in rock and soil.

Radon itself undergoes radioactive decay and has a radioactive half-life of about four days.
When radon atoms decay they emit radiation in the form of alpha particles, and transform into
decay products, or progeny, which also decay. Unlike radon, 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 its 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 from 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 can 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 Drinking Water

People can be exposed to waterborne radon by ingestion and inhalation (see Exhibit 1-2).
Radon dissolved hi 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 water 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
concentration of radon in this water varies widely. While high levels may occur in areas with
large amounts of uranium below ground, radon levels in surface water generally are very low.
Surface water usually lacks a source of radon from rocks, and radon in surface water escapes
quickly into air. Radon levels depend partly on the geologic formations that store and transport
ground water, but also are influenced by the proximity of radioactive precursors to radon.

Due to the time constraints for completing the revision of risk estimates for waterborne
radon and the quantitative uncertainty analysis associated with these risk estimates as requested
by the SAB, only risks associated with exposed populations served by community ground water
supplies were addressed. Based on the National Inorganics and Radionuclides Survey (NIRS)
(Longtin, 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 liter1 of water (pCi/L^), with a 90 percent confidence interval
of 205-306 pCi/L^ter- 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 pCiiL,,^ is
approximately 19 million people. Among these, 17.2 million are served by community ground
water supplies (USEPA, 1992a).

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
concentrations in indoor and outdoor air. To determine the annual average radon concentration
'A curie (Ci) is a standard measure of radioactivity, and a picocurie (pCi) is one trillionth (1 x 10'12) of a curie.
-------
1-6
Exhibit 1-2
Exposure Pathways for Radon
WATER

Radon gas escapes
from tap water
t
Radon in indoor air comes from
tap water, soil, and outdoor air
Routes of Exposure
From ground water source
Jb'rom groun water i
to Public Water Systi
tern
-------
1-7
in U.S. homes, EPA used the results of its National Residential Radon Survey (a survey
measuring annual radon concentrations in 6,000 homes statistically representative of all U.S.
residences) (USEPA 199 If). EPA estimates that the annual average radon activity levels in all
U.S. homes averages 1.25 pCi/Lair. 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/Lair.

Fatal Cancer Risks Associated with Exposure 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 pCi/L 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 radon exposure from water or air, EPA multiplied the unit risk by the
individual's exposure. 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.
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 air pollutants (NESHAP) for
radionuclides. That report led to the formation of the Radiation Advisory Committee (RAC) of
the SAB 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/SAB 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 published prior to or at the same time that the
-------
1-8
proposed rule was published in 1991 (USEPA, 1991a). 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 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 result hi occupational exposure. Prior to the review of the drinking
water documents, the SAB reviewed the Office 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 Chafee-Lauteriberg Radon Report requirements, it
initiated a series of teleconference calls to discuss with SAB members which of the items listed
above 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 quantitative uncertainty of the fatal cancer risk estimates.
Exhibit 1-3
Primary Documents Reviewed by the SAB in February 1993
EPA Document Reviewed
Completion
Date
Cost Modeling Update \
Packed Tower Aeration Cost Estimates for Radon Removal
Technical Support Document for the 1992 Citizen's Guide to Radon ',
Simplified Equations for Estimating Radon Removal Cost via Packed Tower Aeration
Technologies and Costs for the Removal of Radionudides from Potable Water Supplies
Addendum to the Occurrence and Exposure Assessments for Radon, RadiUm-226,
Radium-228, Uranium, and Gross Alpha Particle Activity in Public Drinking Water
Supplies .
Technical Memorandum: Problems with the Use of Granular Activated Carbon for
Radon Removal '••
Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking Water
Working Draft of the Regulatory Impact Analysis for Final NPDWR for Radionudides
February 21, 1992
March 11, 1992
May 1992
July 16, 1992
July 1992

September 1992

January 1993

January 29, 1993
to be published
* 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 analysis of the extensive data available from epidemio-
logic investigations conducted during decades of'Study, the application of mathematical models,
and carefully considered judgment, and with detailed oversight by the SAB, EPA prepared its
"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.

Risk From Unit Exposure

People are exposed to radon in drinking water in three ways: from ingesting radon
dissolved in water; from inhaling radon gas released from water during household use; and from
inhaling radon progeny derived from radon gas released from water. EPA estimated the lifetime
fatal cancer risk from exposure to one picocurie per liter (pCi/L) 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 uncertainties in exposure and cancer risk, as well as
variation in exposure among individuals. EPA's nominal estimate for the individual lifetime
inhalation risk of lung cancer deaths per pCi/L of radon in drinking water is 3,0 x 10"7 with a
median of 3.9 x 10"7 with a credible range of 1.4 x 10"7 to 1.4 x 10"6. EPA's best estimate for
the individual lifetime ingestion risk of fatal cancers per pCi/L of water is 3.5 x 10"7 with a
median of 1.7 x 10'7 with a credible range of 3.9 x 10'8 to 7.2 x 10'7. In,addition, EPA estimated
that the individual lifetime cancer risk for inhaling waterborne radon gas is 2 x 10"8 per
pCi/Lwater. 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.
Assuming a lifetime of 70 years, EPA's nominal estimates for individual lifetime risk associated
with exposure to radon progeny in indoor and outdoor air are 3.0 x 10"3 and 4.8 x 10"4 per
pCi/Lair, respectively.
-------
1-10
Exhibit 1-4
Summary of Proposed and Revised Fatal Cancer Risk Estimates
for Radon in Water
Exposure Pathway
Lifetime Cancer Risk per pCi/L in Water
Proposed
Revised
Inhalation of Radon Progeny Derived
from Waterborne Radon Gas
Inhalation of Radon Gas Released
from Water to Indoor Air
Ingestion of Radon Gas in Direct Tap
Water
Sum of All Pathways
4.9 x 10'7
(74%)
2 x 10'8 ?
(3%) ;
1.5 x 10'7
(23%) ;
6.6 xlO'7
(100%) ;
3.0 x 10'7
(45%)
2 x 10-"
(3%)
3.5 x lO'7
(52%)
6.7 x 10'7
(100%)
Individual Risk

The estimated lifetime fatal cancer risk from exposure to 1 pCi/L of radpn 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 pCi/Lwater. Thus, the average lifetime risk for people
served by community ground water systems containing radon is 165 chances in a million. By
comparison, the lifetime risk for lung cancer death from exposure to airborne radon in homes is
3,024 chances in a million for each pCi/L 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.

Population Risk

EPA has developed annual population risk 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 40 to 408 deaths per
-------
1-11
year arid 11 to 212 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
indoor air, with a credible range of 6,700 to 30,600 deaths. An estimated 520 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
Nominal
Estimate
Upper
Estimate
Inhalation due to Radon Treatment

Inhalation of Radon Gas Released
from Water to Indoor Air*

Inhalation of Radon Progeny Derived
from Waterborne Radon Gas*

Ingestion of Radon Gas in Drinking
Water*

Inhalation from Outdoor Air

Inhalation from Indoor Air
40

280

6,740
6

100

520

13,600
408

212

1,500

30,600
For populations served by community ground water supplies only.
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 pCi/L for drinking water was set as close to the MCLG of 0 pQ/L as at
the time technically feasible.1 The target level for air is EPA's action level of 4 pCi/Lair set in
light of current mitigation and measurement technology.
Since the proposal, newly evaluated data indicates the MCL in water could be technically feasible at well below 300
-------
1-12
As the exhibit shows, approximately 19 million people are exposed to a radon level above
the proposed MCL of 300 pCUL^^. EPA estimates that approximately 2 of every 10,000
individuals exposed would develop a fatal cancer as a result of exposure to radon in water at
300 pCi/L^t^. Approximately 15 million people are exposed above EPA's voluntary action
level of 4 pCi/L^. EPA estimates that approximately 1 in every 100 individuals would develop
a fatal cancer as a result of exposure to radon in indoor air at 4
The cost figures presented in Exhibit 1-6 are difficult to compare. In 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. Based 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 84 people
each year due to radon in water. Similarly, based on the total annual cost of mitigating radon
exposure in indoor residential air, EPA calculates that it would cost an average of $0.7 million
per life saved to prevent the deaths of approximately 2,200 people each year due to exposure to
radon in residential air.
Exhibit 1-6
Summary of EPA's Nominal Estimates of Risk, Fatal Cancer Cases,
Cancer Cases Avoided, and Costs for Mitigating Radon in Water and Air

Number of Fatal Cancer Cases per Year
Proposed or Target Level
Individual Lifetime Risk of Fatal Cancer at
Target Level
Number of People Above Target Level
Number of Fatal Cancer Cases Avoided
Annually by Meeting the "Target" Level
Total Annual Cost for Mitigating Radon
Cost per Fatal Cancer Case Avoided
Drinking Water
192*
300 pCi/L
2 in 10,000
19 million**
84*"
$272 million
$3.2 million
Indoor Air
13,600
4 pCi/L
1 in 100
15 million
100 (2,200)**
$1,504 million
$0.7 million
Includes those exposed above and below the target level in community ground water
systems only.
Includes both community and non-community water systems.
The voluntary air program is estimated to have avoided 100 fatal cancer cases per
year based on 1992 data. Annual lung cancer cases averted by the voluntary air
program are expected to increase each year as additional mitigations occur and new
construction is built to be radon resistant. The air program has the potential to avert
2,200 cancer cases per year assuming 100% voluntary monitoring and mitigation.
-------
1-13
As noted above, the total annual cost is $272 million to treat radon in drinking water. The
following exhibit shows the national cost separated into total capital, annual amortized capital,
annual operation and maintenance (O&M), and total annual costs. It is important to note that
the regulated industry has estimated higher costs than EPA. The American Water Works
Association (AWWA) estimated national costs at $2.5 billion per year. The Association of
California Water Agencies (ACWA) estimated annual costs of $ 520 million for California
alone. The differences between the estimates are included in chapter 4 of the report. Figure 4,
attached at the end of the executive summary, depicts the total annual costs are different MCLs.
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.
National Costs for Controlling or Mitigating Radon in Water
(millions of dollars*)
Type of Cost
Total Capital
Annual Amortized Capital
Annual O&M
Total Annual
Proposed
Estimate
$1,579
106
74
180
Revised Best
Estimate
$1,602
151
121
272
*Cost estimates include community and non-transient non-community systems.

The following exhibit presents the estimates of annual costs for testing and mitigating
radon in indoor residential air. EPA's best estimate of the national cost for addressing radon in
indoor residential air for a fully implemented voluntary program is $1,504 million annually
amortized over a 74 year period. The amortization period is based on the average life
expectancy of the U.S. population and a time period representative of at least the average life of
a home. As the table indicates, most of the costs (approximately 75 percent) are in the
operation and maintenance of the system, while only 25 percent of the costs are for the testing
and initial installation of the mitigation systems.
-------
.! j.
-------
PART TWO
RISK
-------
-------
2. RISK ASSESSMENT OF EXPOSURE TO RADON FROM PUBLIC WATER
SUPPLIES
This 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 based
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
response 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 risk 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 community ground water supplies derived from ground water;

3. Section 2.3 explains EPA's assessment of the risks from inhalation of
radon in community ground water supplies relying on ground water; and

4. Section 2.4 explains 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, 1991a), National
Primary Drinking Water Regulations; Radionuclides: Notice of Proposed Rulemaking (USEPA,
1991e); 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
1992a).
-------
2-2
2.1 PROPOSED AND REVISED UNIT RISK ESTIMATES FOR WATERBORNE
RADON _

People are exposed to waterborne 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 used. It is estimated that an individual's
combined fatal cancer 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.

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 from 74 percent of the total risk in the proposed rule to 45 percent in the revised
estimates. The remainder of the risk (3 percent) from waterborne radon conies from inhaling
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 (107) people exposed per pCi/Lwater)
I
2L

0.2 0.2
Ingestion of Radon Inhalation of Radon Progeny
Inhalation of Radon Sum of All Pathways
Proposed H Revised
-------
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 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. 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 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 first four parameters listed below. To calculate the population risk, total
exposed population also needs to be taken into consideration. Due to the time constraint for
preparing the report, the risks for individual and population were only estimated for people
served by community ground water supplies.

(1) 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.
-------
2-4
Based on this survey and other data (Pennington, 1983; USEPA, 1984; Ershow and
Cantor, 1989), 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 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, 1984b; Ershow and Cantor, 1989). Because
different people ingest different amounts of water, there is considerable variability in the
amount of water ingested among individuals. The median of the direct tap water ingestion rate
per day is 0.526 liters per day, with a credible range of 0.518 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 the aforementioned drinking
water studies 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, 1991a).
• 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 number
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 of one rad (or organ-specific risk coefficients).
• Organ-Specific Dose (Rad per Pkocurie)

Several studies (in USEPA 1993h) 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 (Correia, et al., 1987)i The tissues receiving the highest
dose are the stomach and the intestine, but estimating the dose to these tissues is complicated by
uncertainty over: 1) the possible concentration gradient in the lining of the gastrointestinal tract
-------
2-5
and 2) the possible sweeping of radon's short-lived progeny from the tissue by the blood prior to
decay. In addition, there is uncertainty 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.

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)
Cancer Site
Stomach
Intestine
Small intestine
Ascending colon
Descending colon
Liver
Lung
General tissue

Proposed
1.0 x 10-8

5.2 x 10'9
7.4 x 10'9
4.1 x 10'9
1.5 x 10'9
1.8 x 10'9*
6.7 x 10'10
Dose
.Revised
1.0 x 10'8

3.1 x 10'9
1.5 x 10'9
8,2 x 10'10
1.5 x l
-------
2-6
the credible range between the upper and lower estimates is less than a factor of 10 to 30,
depending on the organ.

Exhibit 2-3
Estimated Ingestion Risk by Cancer Site
(Risk per Rad)
Cancer Site
Stomach
Intestine
Small intestine
Ascending colon
Descending colon
Colon
Liver
Lung
General tissue
Draft Criteria
Document
3.7

3.7
7.3
7.3

4.0
5.7
1.6
x ICT4

x W5
x 10'5
xlO'5
-
x ID"*
x lO'4
x 10'3
Revised
8.9 x 1C'4

.
'
-
2.2 x 10'3
13.0 x lO'4
:1.7x 10'3
14.2 x 10'3
• 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 radon ingested! Because the cancer risk for the total
body is additive, based on the risk to the individual tissues, the total risk of developing a fatal
cancer (risk factor) can be calculated using the equation in the box below.
Risk Factor=y\ dr.
*-**] J J
Where: dj = Dose (rad) per pCi ingested for the target cancer site j.
ij = 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
-------
2-7
cancer risk per pCi of radon ingested to 1.7 x 10~u, or 17 chances out of a trillion, which is
about 2.3 times the risk estimated in the proposed rule. This increase primarily reflects higher
estimates of the risk per unit dose for irradiation of the stomach and colon.

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 estimate for the liver.

Exhibit 2-4
Estimated Risk from Ingested Radon by Cancer Site
(risk x 10~12 per picocurie)
Estimated Risk Factor
Cancer Site
Proposed
Revised
Stomach
Intestine
Small intestine
Colon
Liver
Lung
General tissue
Total body
3.7

0.19
0.84*
0.6
1.0
1.1
7.4
8.9

0
2.6
0.45
2.2
2.8
17.0
Summation 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, 1990). In addition,
cancer statistics indicate that mortality from cancer of the small intestine is very low compared
to colon cancer (NCI, 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
-------
2-8
The risk factor for ingested radon (the summation of the total fatal cancer risk due to
ingestion of 1 pCi of radon), is estimated to be 1.7 x 10"". Monte Carlo simulation was used to
estimate the overall uncertainty in this factor (risk per pCi ingested), taking into account each of
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 radon activity is 246
pCi/Lwate, with a credible range of 205-306 pCi/Lwater Exhibit 2-5 demonstrates how the
population-weighted average radon activity was calculated.
-------
2-9
Exhibit 2-5
Summary Characteristics for Community Ground Water Supply Systems and for
Radon Ground Water Concentrations
System Size
Very, Very
Small
25-100 people
Very Small
101-500 people
Small
501-3,300 people
Medium
3,301-10,000
people
Large/Very
Large
> 10,001 people
Totals
Number of
Community
Ground Water
Supply Systems1
16,634
15,422
9,952
2,302
1,316
45,626
Population Served
by Community
Ground Water
Supply Systems1
(in thousands)
956
3,931
13,884
13,599
48,711
81,081
Average Radon
Concentration2
pCi/L
844
684
284
204
205
2463
Credible Range of
Mean
Concentration2
645-1090
522-876
208-402
147-271
137-295
205-306
1 September 30, 1992 Wade-Miller addendum prepared for USEPA. Data based on Federal Data Reporting
System (PROS)
2 Source: National Inorganics and Radionuclides Survey (Cothern, Rebers, 1990)
3 Population-weighted average and credible range of mean radon concentration for community ground water
supply systems.
• 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, 1992a & 1993h).
-------
2-10
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 uncertainty of
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 individual'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
Nominal
Estimate
Median
High
1.3 x 10'7
1.2 x 106
5.7 x 10
,-7
2.6 x 10
,-6
Population Risk Estimates

EPA estimated the total number of deaths for the fraction of the U.S. population exposed
to radon 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 11 to 212 deaths per year. These values exclude exposure of people
to waterborne radon from private wells and NTNC water supplies.
2.3 RISK ASSESSMENT OF INHALING RADON FROM PUBLIC WATER SUPPLIES

Public water supplies from ground water can be a significant controllable household source
of airborne radon and its decay products (i.e, radon progeny). This section explains 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
-------
2-11
risk estimates for inhalation of radon progeny and radon, respectively. This section also
summarizes EPA's combined inhalation risk estimates due to radon and progeny from drinking
water. Due to the time constraint for completing the report, individual and population risks were
only estimated for people served by community ground water supplies. The information was
drawn from the Drinking Water Criteria Document for Radon in Drinking Water (US EPA,
1991a), 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, Radium-228, Uranium, and Gross Alpha Particle Activity in Public Drinking Water
Supplies (Revised Occurrence Estimates Based on Comments to the Proposed Radionuclides
Regulations) (US EPA, 1992a).

EPA's inhalation risk assessment covers the risks to human health from both radon gas
and radon progeny. Radon progeny (or decay products) pose far greater risks than radon 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 supplies by community ground water systems 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 1991 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 BEIR IV model for calculating risk; in the proposal EPA used an average of the BEIR
IV 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 "nominal estimate" of each of these risks and quantitatively estimated the
credible range of values that could exist given the uncertainty in each of the key parameters in
the analysis.
-------
2-12
Parameters Influencing the Risk from Inhaling Radon Progeny from 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:
(1)
(2)

(3)
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;
(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 death 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. EPA has invested considerable effort gaining knowledge about each 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 veiy 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 on drinking water samples supplied by community
groundwater systems, the population-weighted average radon activity is 246 pCi/L^^,
with a credible range of 205-306 pCi/L,,^. Details of the occurrence data of radon in
drinking water are in Exhibit 2-5 on page 2-9.
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. Numerous studies have investigated the cross-media
-------
2-13
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 1:10,000. Existing studies suggest
that the uncertainty around this estimate gives a credible range from 0.7:10,000 to
1.9:10,000.
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 risks than radon itself, EPA focused on
radon progeny in estimating the inhalation risks of radon. In developing a 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 Comparative 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).

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
BEIR IV model. EPA had previously considered the results of an ICRP risk 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
-------
2-14
since the exposure occurred; and (2) the age at which the exposure occurred. The BEIR IV
model was more consistent with this evidence than was the ICRP model.

These changes led to a revised radon risk factor of 224 lung cancer deaths (LCDs) per
million persons exposed to a working level month (WLM1) of radon, a common unit for
measuring radon exposure. For the proposed rule EPA used a risk factor of 360 LCDs per
million people exposed to one WLM. This change represents a reduction of 38 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).

• 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
one-third of the population of the United States in 1990.
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 not 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
represents a house as a single compartment 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.

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 shower, a bathroom, and the remainder of the house.
WLM is defined as a working level month, which is a standard unit of measure of exposure to radon decay products.
-------
2-15
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 calculated from the resulting estimates of radon concentration in each
compartment and estimates 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-compartment model. EPA also conducted a sensitivity analysis of 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, 1993h).
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 presents the EPA's best estimates of the mean individual risk with a credible range.
EPA assumed that one lifetime is equal to 70 years.
-------
2-16
Exhibit 2-7
EPA's Estimates of Mean Individual Risk from Radon Inhalation at Occurrence Levels
Served by Community Ground Water Supply Systems
(lung cancer deaths per person per year)

Individual Risk
Low
4.9 x ID'7
Nominal Estimate
1.1 x 10 -6
Median
1.4 x 10'6
High
5.0 x 10'6
Population 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
water-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 credible range due to
uncertainty in this estimate of 40 to 408 LCDs per year.

The Risks from Inhaling Radon Gas
f! ' ' "
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 1 pCi of radon gas 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 pCi/L of radon in
water gives rise to 1.0 x 10"4 pCi/L 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 4.2 x 108 liters
of air in a 70-year lifetime. Thus, the amount of radon gas inhaled over a 70-year lifetime from
1 pCi/L of radon in water is 4.2 x 10"4 pCi. Using the risk factor of 4.7 x 10~13 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 xlO"8 fatal cancer cases per person per year.
The estimated annual cancer deaths due to inhalation in the population exposed to radon in water
(81 million) at an average concentration of 246 pCi/L is 6 deaths. This risk from inhalation of
radon gas contributes a small percentage (approximately 3 percent) of the total waterborne radon
risk. For this reason, EPA did not analyze the uncertainty in its estimates.
-------
2-17
Combined Inhalation Risks from Radon Progeny and Radon Gas

Based on its revised risk estimates as summarized above, EPA estimated the number of
fatal cancer cases from inhaling waterborne rad and its progeny for the entire population exposed
to radon through community water supplies using ground water on a long-term basis. Given that
these water suppliers expose an estimated 81 million people each year to increased risk of cancer
from the 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 46 to 414 deaths per year. These values exclude exposure of people to
waterborne 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 nominal 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
Comparison of EPA's Revised Individual and Population Risk Estimates for People
Served by Community Ground Water Supply Systems
Individual Risk
Population Risk
Inhalation Ingestion
Inhalation Ingestion
Statistic
Lower Credible Bound
Nominal Estimate
Upper Credible Bound
(fatal cancer cases per person per year)
4.9 x lO'7 1.3 x 10'7
LlxW* 1.2 xlO'6
5.0 x 10'6 2.6,x 10'6
(fatal cancer cases per year)
40 11
86 100
408 212
-------
-------
3. RISK ASSESSMENT OF EXPOSURE TO RADON IN AIR
The four sections outlined below present and explain how EPA estimated the risk due to
radon in air:

1. 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 originally presented in Technical Support Document for the
Citizen's Guide to Radon (USEPA, 1992i);

4. Section 3.4 presents the inhalation risks that result from radon emissions at water
treatment plants 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 summarizes 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 radon 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 National Residential Radon Survey (a survey measuring
annual average 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.
-------
3-2
homes of 1.25 picocuries1 per liter (pCi/Lair). EPA used the same estimates for the occupancy
factor and equilibrium factor described in Chapter 2,

Exhibit 3-1 summarizes EPA's best estimate of the results. EPA calculated both the
average individual risk and the population risk per year. For the entire U.S. population of 250
million, the total annual lung cancer deaths from inhalation of radon progeny in indoor
residential air were estimated to be 13,600. These risk estimates cover exposure to indoor
residential 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. . • ••
Exhibit 3-1
Best Estimate of the Annual Risk From Residential Inhalation of Indoor Radon
Progeny
=
20,000 -
1
10,000 -
5,000 -
0
30?400
,
-
- • -
-
•

6,700
•* •> ;
i )
Low

T3?600
i i
;••,->•. ••• •

Best Estimate High
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, occupancy
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
'A curie (Ci) is a standard measure of radioactivity, and a picocurie (pCi) is one trillionth (1 x 10"12) of a curie.
-------
3-3
sources: (1) 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 median
of 14,410 and a range of estimates of total population risk of 6,740 to 30,600 lung cancer deaths
per year (USEPA, 1992i).
3.2 RADON IN OUTDOOR AIR

For purposes of comparison, EPA estimated the expected risk from inhalation of radon
progeny in outdoor air (USEPA, 1993h). 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.

In outdoor air, radon decay products are nearly in equilibrium with radon gas, and the
equilibrium factor is higher than it is indoors. 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 risk 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
median of 657 with a credible range of 280 to 1,500 lung cancer deaths per year (USEPA,
1993h).
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
(GAG) to treat radon, GAG 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 tower 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 prior to proposing the radionuclides regulations. EPA based the risk assessment on two
prior studies: Preliminary Risk Assessment for Radon Emissions from Drinking Water Treatment
Facilities (USEPA, 1988) arid Analysis of Potential Radon Emissions from Water Treatment
Plants Using the MINEDOSE
-------
3-5
Revised Estimate of Risk From Treatment Plant Emissions
Commenters asserted that EPA failed to consider sufficiently in the proposed rule the risk
tradeoff of increasing air emissions from the use of PTA. In response, in 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 generally 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 BEIR IV
Committee (224 lung cancer deaths per million person-WLM), as reported in the Technical
Support Document. After recalculating based on the revised risk coefficient, ORIA reached two
conclusions: that the maximum 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 0.016 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 person-WLM)

Best Estimate of Individual Risk
(deaths per 100,000 people)

Incidence
(deaths per year)
460

0.016
224

0.004
ORIA's uncertainty analysis indicated that individual risks are likely to be overstated by
the generic assessments 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
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. As described in Chapter 2, 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 lung cancer deaths to 233 lung cancer deaths
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 Air"
(deaths per year)
Source of Radon in Air
AH Indoor Progeny0
Outdoor Progeny
Treatment Plant Emissions
Total for Radon in Air0
Low
7,000
280
—
7,280
Best
Estimate"
13,600
520
0.04
14,120
High
30,000
1,500
—
31,500
* For all indoor and outdoor progeny, values apply to the entire U.S. population of
250 million.
* Value calculated using best estimates of parameters and risk equation.
e Does not include structures other than residences.
-------
PART THREE
COST
-------
-------
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 of mitigating radon in indoor air. EPA developed
the cost estimates presented in the 1991 proposed rule for drinking water based on cost
modelling that incorporated available data and research 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 Drinking Water
Regulations for Radionudides (USEPA, 199 Ij) and revised data presented in Working Draft of
Regulatory Impact Analysis for Final National Primary Drinking Water Regulations for
Radionudides (USEPA, to be published). Sections six and seven present the costs for
controlling radon in air, originally presented in the Technical Support Document for the Citizen's
Guide to Radon (USEPA, 1992i), and the combined costs for controlling residential radon from
all sources. The last section discusses the cost estimates and differences between AWWA,
ACWA and EPA. The eight sections in this chapter are:
1.
4.
5.
6.
7.
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;

Section 4.2 describes qualitatively the key factors that account for
variations in costs per household for treating radon in public water
systems relying on ground water, with quantitative data on these factors
presented in graphs and other exhibits;

Section 4.3 provides a quantitative description of the cost estimates used
in the proposed rule for public water systems relying on ground water;

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;

Section 4.5 details the revised national and household cost estimates for
treating radon in public water systems relying on ground water;

Section 4.6 summarizes national cost estimates for controlling radon in
indoor air; and

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
8 Section 4.8 presents the cost estimates from American Water Works Association and
the Association of California Water Agencies and discusses the major differences
between their estimates and those of the United States Environmental Protection
Agency.

COSTS OF CONTROLLING RADON IN PUBLIC WATER SYSTEMS RELYING
ON GROUND WATER
Three primary factors determine the national cost of treating radon in public water systems
relying on ground water:

1. 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 per liter of
water (pCi/U, ) After further analysis, described below, EPA; raised its estimate to 27,294
affected systems. This revision 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 Inorganics and Radionuclides Survey (NIRS) data that
includes separate distribution parameters for each of five size categories, and
disaggregation of the smallest group into two separate groups; and

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
facUities. Therefore, EPA now estimates that at an MCL of 300 pCi/Lwater, 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 has increased slightly from the
18 million previously estimated. Exhibit 4-1 summarizes the 1991 estimates for radon
occurrence and the revised best estimates for radon occurrence. In the exhibit the PWSs include
community and non-transient non-community systems.
'A curie (Ci) is a standard measure of radioactivity, and a picocurie (pCi) is one trillionth of a curie.
-------
4-3
Exhibit 4-1
Occurrence Estimates for Treating Radon in
PWSs Relying on Ground Water

Systems Affected
Number of Sites
Population Affected
1991
Proposed Estimates
25,907
25,907
18 million
Best Estimates
27,294
41,136
19 million
The cost of treatment at a particular site depends both on the flow rate at the treatment site
(i.e., the amount of water that flows through the treatment site each day) and the concentration
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 water 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 GAG 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 longer
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 in the revised analysis that aeration was the most economical option for all
system sizes and derived final 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, are applicable. 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 several required removal efficiencies. The revised cost
estimates reflect an expected distribution (based upon the radon occurrence data) of PTA
-------
4-4
removal efficiencies required for treating radon in drinking water in order to meet the MCL.
Exhibit 4-2 illustrates the revised assumption that 32 percent of affected sites will require an 80
percent removal efficiency, 31 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 at a large number of sites that use aeration treatment.
Exhibit 4-2
Percent Radon Removal Required at Treatment Sites to
Meet the Proposed MCL for PWSs Relying on Ground Water
80% Radon Removal
(32%)
50% Radon Removal
(31%)
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.
-------
4-5
Exhibit 4-3
Unit Cost Derivation for PTA Treatment
Type of Cost Equipment
Cost Basis
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, and
distributor rings at six-foot intervals along the packing height. Delivered prices
for the three items, with diameters ranging from 1 to 10 feet, were obtained
from a major equipment supplier.

Packing Material Unit cost of 1-inch plastic saddle packing material, ranging from 100 to 2500
cubic feet, 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 obtained from a major supplier 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 transferring 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 (MGD). 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 requiring treatment. The air well was
assumed to be a below-grade concrete structure which functioned as an effluent
holding tank and a foundation for the packed columns. The 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, and motor control stations
was assumed to be 25% of blower capital cost.
Indirect 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)
Type of Cost Equipment
Cost Basis
Blower Electrical Electric power cost of the blower is based on air flow and pressure drop
Power through packing material plus 0.07 psi.

Operating Labor Operating labor cost estimate is based on an estimate of labor hours per year
and a labor rate of $14.70 per hour. Labor hours per year were estimated to be
the product of 0.25 hour per column per shift, number of PTA columns in
operation, number of shifts per day, and 365 days per year.
Maintenance Cost
Maintenance cost is based on 10% and 4% of mechanical and non-mechanical
process equipment cost, respectively.
Administrative Administrative cost is based on 20% and 25% of the labor and maintenance
Cost cost, respectively.
Total Annual Cost
Total annual cost is the sum of amortized capital and operating cost.
Total Production Cost
Total production cost is the total annual cost divided by the volume of water
treated per year.
4.2 HOUSEHOLD AND COMMUNITY COSTS OF TREATING RADON IN 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 systenis;

• Use a smaller percentage of their system capacity on average than 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 design flow rate of 0.024 million gallons per day (MOD);
The largest water systems (serving populations over 1 million people) have a design flow rate of
430 MOD. Exhibit 4-4 displays the 12 size categories used in this report.
-------
4-7
Exhibit 4-4
Public Water System Size Categories
System Size
1
2
3
,4
5
6
7
8
9
10
11
12
Average System Flow
Population Range (MGD)
25-100
101 - 500
501 - 1000
1,001 - 3,300
3,301 - 10,000
10,001 - 25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
100,001 -500,000
500,001 - 1,000,000
Over 1,000,000
0.0056
0.024
0.056
0.23
0.70
2.1
5.0
8.8
13
27
120
270
Design System
Capacity Flow
(MGD)
0.024
0.087
0.27
0.65
1.8
4.8
11
18
26
51
210
430 , ' : ; ;
Larger systems are likely to have more treatment sites that might require separate treatment
facilities, while smaller systems are more likely to have no more than one or two facilities that
treat the entire water flow. The flow rate at any single site in the larger systems may be several
orders of magnitude 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 .fate of 0,02 MGD,
and 20 percent are expected to have two sites with flow rates of 0.01 MGD 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 MGD, 30 percent have flow rates of 7.17 MGD, another 30
percent have flow rates of 5.38 MGD, 20 percent have flow rates of 10.75 MGD, and 10 percent
have flow rates of 21.5 MGD. 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. ,
-------
4-8
03
•S Ur
I
CO
O
I-.O

O
S
o
a
O
-------
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. Although treatment facilities must be designed to accommodate the maximum 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.
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 additional
disinfection costs, while less than 10 percent of the largest systems are expected to incur such
costs.

Exhibit 4-7 presents revised aeration treatment costs incurred by small and large systems.
The smallest systems incur costs of over $2.20 per thousand gallons, while the largest systems
-------
4-10
incur costs of almost $0.06 per thousand gallons. For systems adding disinfection, costs would
be an additional $1.00 to $0.04 per thousand gallons, approximately, for very small and large
systems, respectively. ;

Exhibit 4-7
Aeration Treatment Cost per Thousand Gallons
System Size
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
thousand gallons for the largest systems.2 Exhibit 4-8 shows that average costs per gallon are
1 The proposed rule, as published in the Federal Register on July 18, 1991, provided cost estimates only for radon
removal efficiency of 80 percent.
-------
4-11
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.

Exhibit 4-8
Average Cost per Thousand Gallons
Estimated in Proposed Rule
(Aeration Treatment)
System Size
In 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)
Population Served
Type of Cost
25 - 100 101 - 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)
0.2
0.6
1.4
3.1
7.6
3,400
-------
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 assumptions.

Exhibit 4-10
Cost Estimate Revisions
Cost Estimates
Type of Cost
Proposed
Revised
Labor Rate for Small Systems

Operating Labor

Mobilization and Bonding

Cost Indexes

Safety Factors

New Finished Water Pumps

Process Piping Labor Costs

Disinfection Treatment
$5.90/hr

Costs based on flat rate of 0.3 cents per
thousand gallons treated

Not included

1986-1989

20% Transfer Coefficient

Not included

$30 per hour labor costs; 1988 cost index

25% of systems add disinfection
$14.70/hr

0.25 labor hours per PTA column in
operation per shift

Contingencies Increased to 20% from 15%

Updated to 1991

• Increased to 40% to Cover Uncertainties
(Overdesign)

New Costs included

Increased to $50 per hour labor costs; 1991
cost index

Graded estimate: 50% of small systems
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 assumed 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 arid
communities.
-------
4-13
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 equal 0.25 hours per column per shift (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 during 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 Index

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.
-------
4-14
• 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
process. Based upon national survey data, approximately 50 percent of small systems and
10 percent of the largest systems would add disinfection treatment as a result of radon in
water regulation.

Also, in EPA's revised national cost estimates (see below) the Agency accounts for a
significant amount of treatment of water at the wellhead. EPA estimates that because of the
non-centralized nature of many ground water supplies there will be approximately 50 percent
more treatment sites than estimated in the proposed radon in drinking water regulation.
4.5 REVISED SUMMARY COST ESTIMATES FOR TREATING RADON IN
DRINKING WATER : ' ..
I
Exhibit 4-11 shows EPA's revised estimates for national costs. (The reader is referred to
the Attachments to this report for EPA's revised estimates of cost and other impacts at various
MCLs.) Exhibit 4-12 presents the cumulative effect of the occurrence and cost revisions for all
affected public systems by system size, in terms of average annual household cost. EPA's
revised best estimate of the national cost for treating radon at community drinking water supplies
is $272 million, versus the 1991 estimate of $180 million for an MCL of 300 pCi/L. (The reader-
is referred to the Attachments to this report which contain EPA's revised estimates of the
impacts on community water supplies only: cancer cases avoided, total annual costs, populations
exposed, and community water systems affected at various MCLs are presented.)
Exhibit 4-11
National Costs for Controlling or Mitigating Radon in Water
(millions of dollars*)
Type of Cost
Total Capital
Annual Amortized Capital
Annual O&M
Total Annual
Proposed
Estimate
$1,579
106
74
180
Revised Best
Estimate
$1,602
151
121
272
* Cost estimates include community :and non-
transient non-community systems.
p
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,
-------
4-15
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.

Exhibit 4-12
Average Annual Household Cost
of Treating Radon in PWSs Relying on Ground Water
300
5678
System Size
10 11 12
4.6 COSTS OF MITIGATING RADON IN INDOOR AIR
Exhibit 4-13 presents the estimates of annual costs for testing and mitigating radon in
indoor residential air. EPA's best estimate of the national cost for addressing radon in indoor
residential air is $1,504 million annually amortized over a 74 year period. The amortization
period is based on the average life expectancy of the U.S. population and a time period
representative of at least the average life of a home. As the table indicates, most of the costs
(approximately 75 percent) are in the operation and maintenance of the system, while only 25
percent of the costs are for the testing and initial installation of the mitigation systems.
-------
4-16
Exhibit 4-13
Annual Cost Estimates for Testing & Mitigating Radon in Indoor Air
(millions of dollars per year)
Type of Cost
Annualized Testing Costs

Annualized Installation Costs

Annualized O&M

Total Annual Costs
Best
Estimate
$90

324

1,090

1,504
4.7 COMBINED COSTS OF CONTROLLING SOURCES OF RESIDENTIAL RADON

4.8 COST ESTIMATES AND DISCUSSION OF DIFFERENCES BETWEEN AWWA,
ACWA AND EPA "

EPA estimates that the total annual cost of compliance with the proposed radon maximum
contaminant level (MCL) of 300 pCi/1 would be $272 M, whereas the American Water Works
Association (AWWA) estimates the cost at $2.5 B per year. The Association of California
Water Agencies (ACWA) estimates annualized costs, for California alone, at $520 M (low
estimate) to $710 M (high estimate). The major reasons for the differences in national estimates
between EPA and AWWA are summarized below in Exhibit 4-14. ACWA's numbers are not
included in this table because they apply only to California, whereas the numbers below are
national numbers.
-------
4-17
Exhibit 4-14
Comparison of Costs Estimates of AWWA, ACWA and EPA
Item
Public Systems Affected (No.)
Interest Rate
Annual Costs Per Household (3 smallest system
sizes)
25-100 population served1
101-500 population served
501-1000 population served
Annualized National Costs
AWWA
32,750
10%

$2,200
$700
$210
$2,500 M
EPA
Estimate
1991
25,907
3%

$170
$84
$20-45
$180 M
EPA Estimate
1993
27,294
7%

$240
$94
$25-46
$270 M
1 On average there are three people per household. The number of households in the three smallest system sizes
range between 8 homes in a system service population of 25 to 333 homes in a system service population of 1000.

As shown in the table, the major differences in Agency and industry cost estimates result
from differences in the number of water systems affected by the proposed standard of 300 pCi/1,
differences ,in the treatment costs that would be incurred by a typical public water system to
comply with the proposed standard, and the interest rate charged for the purchase of treatment
equipment
Occurrence
AWWA3 estimated that 26 percent more systems than predicted by EPA would be out of
compliance with the proposed MCL. EPA has revised its analysis in response to some of the
points raised by AWWA to more accurately reflect the current total number of drinking water
systems to be affected. The residual difference between EPA's and AWWA's,occurrence figures
is approximately 20 percent. The major factor in this variation is that AWWA utilized an
independent survey of State drinking water regulators to estimate approximately 3,500 more
groundwater-based non-transient, non-community water system (NT-NCWS) (e.g., schools and
hospitals) than are indicated in the national EPA data base, and then calculated a national
estimate. EPA's analysis extrapolated directly to a national estimate. The difference of
approximately 20 percent is in the range which would be expected from using these different
approaches to estimate national occurrence in more than 70,000 groundwater supplies from a
national survey of approximately 1,000 supplies.
3 Raucher, R.S. The Cost for Compliance With the Proposed Federal Drinking Water Standards for Radionuclides.
Prepared for American Water Works Association, Water Industry Technical Action Panel, October 10, 1991.
-------
4-18
Several close comparisons of individual State and regional radon data with EPA's national
data base tend to reinforce our confidence in the current EPA estimates. For example, two
sources, the EPA national survey and a State survey, have produced representations of radon
levels in public water supplies in Pennsylvania. They are within a few percentage points in
terms of indicating non-compliance with a proposed radon standard of 300 pCi/1 (66% versus
68%). In another example, ACWA analyzed radon data in California. These data (contained in
an ACWA survey and a Southern California Metropolitan Water District survey), were based on
approximately 500 and 200 samples, respectively, and indicated that 59 percent and 56 percent
of the samples, respectively, exceeded the proposed radon MCL. The California data in EPA's
national survey indicated 57 percent of the water systems would be above the proposed MCL.

The number of systems affected by a regulation as well as the number of people served by
those systems are used to estimate the size of the population affected. EPA and the industry
differ on the number of systems needing to treat their water; thus, the two differ on the
population affected by the rule.
Treatment Costs
The second major category of cost difference is the typical cost of installing treatment.
Based on public comments on EPA's proposal, EPA has nearly doubled its unit cost estimates
for aeration treatment for the smallest public water systems. EPA incorporated several suggested
industry revisions. For example, cost estimates recognize that many public water systems relying
on ground water will need to disinfect the water once they draw it out of the ground to aerate it.
The revised costs also recognize that most ground water systems rely on more than one well and,
therefore, need to install treatment at more than one place. EPA has also increased labor and
equipment installation costs since proposal. SAB's Drinking Water Advisory Committee
determined that EPA had employed a reasonable approach4 to the analysis of occurrence data,
technologies, and costs as a function of system size.

Differences remain in technology cost estimates. Agency treatment cost projections are
based upon a validated model. Industry's estimates are more typical of additional costs likely to
be incurred by large systems (e.g., higher labor rates in urban areas and more engineering design
work rather than purchase of off-the-shelf designs). Although each party's estimate of the cost
of a treatment technology may be reasonable in and of itself, EPA believes its estimates better
reflect likely industry practice for the small systems which are most affected by the rule.
4 SAB letter dated July 29, 1993
-------
4-19
Interest Rate
Finally, some of the differences between EPA and the regulated industry estimates of costs
result from industry's use of a ten percent interest rate for borrowing money to purchase
treatment equipment versus the seven percent recommended by OMB and used by the Agency.
-------
-------
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 picocuries5 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 84 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 5-1
Low, Best, and High Estimated Cost Per Life Saved from
Controlling Radon in PWSs Relying on Ground Water
(in millions of dollars) '
$6
I
§
$5
'$3
I
2!
!
7
Low
Best
High
5A Curie (Ci) is a measure of radioactivity, and a picocurie (pCi) is one trillionth of a Curie. One pCi equals 0.037
atomic disintegrations per second, or one atomic disintegration every 30 seconds.
-------
5-2
The Science Advisory Board suggested that EPA compare cost per life 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.

The variation in cost per cancer case avoided between large and small systems can be
explained by both differences in costs and current exposure levels. Due to the number of
households to share the cost, the estimated annual household cost of treatment for radon is
much higher in the smallest systems than in the largest systems ($242 per household per year in
the smallest systems versus $5 per household per year in the largest systems, as discussed on
page 4-15). However, the variation in cost per cancer case avoided is mitigated because
households in some of the smallest systems are likely to be exposed to higher levels of radon
than larger systems. According to the national survey data, as analyzed by Longtin (1988,
1990), the occurrence of higher levels of radon is more concentrated within the smallest system
sizes. Longtin's (1988) analysis of the same data by State also indicates that population-
weighted average radon levels are generally higher for the srhailest systems. Trie higher levels
of radon in smaller ground water supplies may 'be due mainly to geologic factors, and to system
storage and distribution residence times which tend-to be larger at large supplies.

5.2 COST OF RISK REDUCTION FROM RADON IN AIR .

Exhibit 5-2 presents two estimates of the cost of risk reduction from testing and mitigating
radon in indoor air. Based on the estimate of 2,200 cancer cases avoided and the cost estimates
presented in Exhibit 4-13, the agency's best estimate for the cost per life saved is $0.7 million.
If a 7 percent discount rate were used, the total annual cost would be $1.98 billion and the cost
per life saved is $0.9 million6.
Exhibit 5-2

Annual Cost Estimates for Testing & Mitigating Radon in Indoor Air
(millions of dollars per year)
Type of Cost
Annualized Testing Costs

Annualized Installation Costs

Annualized O&M

Total Annual Costs
Best
Estimate
$90

324

1,090

1,504
6 Most of the costs of mitigating indoor airborne radon are associated with operation and maintenance (O&M) of the
systems (75%). The estimates presented throughout the remainder of the document utilize the 3 percent rate which
yields an annual cost of $1,504 million for a fully implemented program and a cost per cancer case avoided of $0.7
million.
-------
5-3
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 3 percent and a 7
percent discount rate, respectively. 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 $.8 million per life saved.
Exhibit 5-3
Cost of Risk Reduction from Controlling Selected Sources of Residential Radon
(annual lives saved and millions of dollars)
Residential Radon Source
Indoor Air
Drinking Water*
All Sources
Annual
Lives Saved
2200
84
2284
Annual
Cost
$1,504
272
1,776
Cost/Life
Saved
$0.7
3.2
•8
These estimates include community and non-transient non-
community drinking water supplies. •
-------
-------
PART FOUR
COMMENTS ON THE RADON
IN DRINKING WATER RULE
-------
-------
6. FINAL SAB COMMENTS RECEIVED JULY 1993
Radiation Advisory Committee (RAC) Comments

Drinking Water Advisory Committee (DWAC) Comments

Executive Committee Comments
-------
-------
Unfed State*
Environmental
Protection Ag«ncy
SCMHC* Advisory
Board (A-101)
EPA-SAB-RAC-M-OU
July. 19S3
vvEPA AN SAB REPORT:
MULTI-MEDIA RISK
ASSESSMENT FOR RADON
REVIEW OF UNCERTAINTY
ANALYSIS OF RISKS
ASSOCIATED WITH
EXPOSURE TO RADON
-------
-------
WASHINGTON, D.C. 20460
July 9, 1993
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
EPA-SAB-RAC-93-014
Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 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
commentary (EPA-SAB-EC-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 the Radiation Advisory Committee's
review of the EPA risk assessment study, Uncertainty Analysis of Risks Associated
with Exposure the Radon in Drinking Wafer (January 29, 1993), related documents
and public comment. The review was conducted at a public meeting February 17-19,
1993.

The Committee commends the EPA staff for having produced an excellent
document 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 alt 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 stated commitment 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
from underground sources 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 and
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 on xenon
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. Further, the changes in most cases
would not substantially change the document's estimates of central values for risks.

Policy Considerations

The comments below, to some extent, reach beyond the strictly technical issues
examined by the Committee. However, the Committee feels that it was important thaft
-------
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.

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 summary, within the limitations of the data currently available, the EPA has
now successfully prepared a scientifically credible multi-media risk 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,
)r. Raymond C. Loe
Chair, Executive Committee
Science Advisory Board
IU. Matanoski
Chair, Radiation Advisory Committee
Science Advisory Board
-------
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.
-------
ABSTRACT

The Radiation Advisory Committee has reviewed the EPA's, "Ur._jrtainty
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
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
RADIATION ADVISORY COMMITTEE

ROSTER
CHAIR
Dr. Genevieve M. Matanoski, 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-514, 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

*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 Adalbert Road, Cleveland, Ohio 44106

Dr. Richard G. Sextro, Indoor Environment Program, Lawrence Berkeley Laboratory
Building 90, Room 3058, Berkeley, California 94720

"Mr. Paul G. Voilleque, MJP 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
HI
-------
SCIENCE ADVISORY BOARD STAFF

Mrs. Kathleen W. Conway, Designated Federal Official, Science Advisory Board
(A-101F), U.S. Environmental Protection Agency, 401 M Street, S.W.,
Washington, DC 20460

Mrs. Dorothy M. Clark, Staff Secretary, Science Advisory Board (A-101F), U.S.
Environmental Protection Agency, 401 M Street, S.W., Washington, DC 20460
* 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.
iv
-------
TABLE OF CONTENTS

1. EXECUTIVE SUMMARY . .... . v. . . : 1
1.1 Background ...:............... 1
1.2 Technical Considerations 2
1.3 Policy Considerations ...... .....! -.:^'.".'.'•". . . ./....'• .'":. : ... . 6

2. INTRODUCTION 8
2.1 Relevant Prior SAB Reports . 8
2.2 Procedural History of this Review . . . .' ... . . . :: . .... . . . ... .... 8

3. FINDINGS AND DETAILED DISCUSSION 11
3.1 Adequacy of Revisions to Ingestion and Inhalation Risk Estimates ... 11
3.1.1 Are revisions of ingestion risk estimates for waterfaorne
radon and its progeny adequate? ...... .'". . /../....... 11
3.1.2 Are revisions of inhalation risk estimates for waterborne
radon and its progeny adequate? 12
3.1.3 Discrepancies in Numerical Values: Are EPA's choices for
risk parameters and the uncertainties adequately defended? . 13
3.1.3.1 Estimates of risk due to inhalation of indoor air . 13
3.1.3.2 Estimates of risk associated with inhalation of
outdoor air 13
3.1.3.3 Estimates of risks and uncertainties
associated with water ingestion 14
3.2 Adequacy of Quantitative Uncertainty Analyses Regarding Risk
Assessment 14
3.2.1 Are the basic methods used to propagate uncertainty
acceptable? 14
3.2.2 Are the probability density functions (PDFs) selected to
describe Type A and Type B uncertainty of each variable
reasonable? 15
3.2.3 Are there any important terms or assumptions that have not
been adequately evaluated? 16
3.3 Adequacy of Characterization of Risks from Water Treatment
Facilities 17
3.3.1 Has the EPA adequately characterized the risks introduced
by radon that would be released by aeration from water
treatment facilities? . . 17
3.3.2 Has the EPA adequately characterized the risks introduced
by radon that would be released from other types of water
treatment facilities? 19
3.3.3 Occupational Exposures 19
'"V
-------
3.4 Other Scientific Issues 20
3.4.1 Recommended extensions of the risk and uncertainty
analysis and publication of results in peer-reviewed journals . 20
3.4.1.1 Individual risks 20
3.4.1.2 Population risks 21
3.4.2 Estimate of Lives Saved 22
3.4.3 Peer Review and Publication . . . . : 23

4. POLICY CONSIDERATIONS , ...... 24
4.1 The Importance of Quantitative Uncertainty Analysis 24
4.2 The Relative Risk of Radon in Drinking Water 25
4.3 Harmonizing 25

5. REFERENCES .27
5.1 Documents Received by the Radiation Advisory Committee During
this Review 27
5.2 Science Advisory Board Reports of Potential Interest ....'.' 35
5.3 Literature cited 36
APPENDIX A:
APPENDIX B:
Brief Chronology of Relevant SAB Reports
Congressional Record-Senate, S15103, September 25, 1992
VI
-------
1. EXECUTIVE SUMMARY
1.1 Background
In EPA's 1993 appropriation1, Congress required EPA to, "conduct a risk
assessment of radon considering: . . . the risk of adverse human health effects
associated with exposure to various pathways of radon .... Such an evaluation shall
consider the risks posed by the treatment and disposal of any wastes produced by
water treatment." Congress also required that, "The Science Advisory Board shall
review the Agency's study and submit a recommendation to the Administrator on its
findings." This letter and the accompanying report set forth the Radiation Advisory
Committee's findings and recommendations based on its 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
EPA uncertainty analysis addressed four radon exposure pathways: inhalation indoors
of radon from non-water sources, inhalation of radon outdoors, ingestion of waterborne
radon, and inhalation of waterborne radon. 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. In considering adequacy in the review, the Committee was mindful
of concerns it had expressed in reports about earlier EPA documents on radon in
drinking water transmitted to the Administrator on January 9 and 29, 1992: (a) that
uncertainties associated with the selection of particular models, specific parameters
used in the models, and the final risk estimates were not adequately addressed in any
of the documents; (b) that high exposure to radon from water at the point of use (e.g.,
a shower) had not been adequately addressed; (c) that 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) that
regulation and removal of radon in drinking water may result in occupational
exposures.
1 Departments of Veterans Affairs and Housing and Urban Development, and Independent Agencies Appropriation Act 1993.
PUB. L 102-398. Section 519.106 STAT 1618 (1992)

1
-------
1.2 Technical Considerations

Regarding the Committee's charge and concerns (a) and (b) above, the
Committee commends the EPA staff for producing an excellent document 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 for a
regulatory agency and significantly enhances the scientific credibility of the EPA's ,
decision-making basis. The Committee assumes that this reflects the EPA's recently
stated commitment to a more rigorous approach to evaluating uncertainties in its risk
analyses of radiological .and other hazardous exposures in the future. With respect to
concerns (c) and (d) above, the Committee recommends that EPA re-examine its
assumptions about which water treatment technologies will be used for radon removal.
When EPA has determined the likely treatment options, then EPA should perform an
uncertainty analysis for occupational exposure based on that distribution (including the
uncertainty about how frequently the various options will be used). If granular
activated carbon is among those treatment options, then EPA should broaden the
uncertainty analysis to include the disposal of granular activated carbon.

With respect to the EPA's analysis, the risk assessment of radon in drinking
water has been revised and an uncertainty analysis has been conducted using Monte
Carlo simulation methods. The uncertainty analysis incorporates quantifiable
uncertainties in exposure and toxicology, as well as true variation in exposure among
individuals. EPA's mean estimate for the lifetime individual inhalation risk of lung
cancer deaths per pCi/L of radon in drinking water is 3.6 x 10"7, with a stated 90%
confidence interval around the mean of 1,8 x 10"7to 7.0 xlO'7. The Agency's mean
estimate for the lifetime individual ingestion risk of fatal cancers per pCi/L of radon in
drinking water is 1,8 x 1Q~7 with a stated confidence interval: around the mean of 6.9 x
10"* to 6.4 x 10"7. The Agency's nominal estimate for individual lifetime inhalation and
ingestion risk per pCi/L for radon in drinking water,are 3.0 x 1Q'7 and 3.5, xlO'7,
respectively. Therefore, for drinking water risks, the contributions of the inhalation and
ingestion are almost equal. , ,r

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 Radiation Advisory Committee believes that the stated uncertainty range
for the ingestion risk is too small in comparison with that for inhalation, because the
-------
ingestion risk estimate is based on two major factors: (a) an estimate of the
distribution of radon to organs in the gastrointestinal tract, based on an unpublished
study using xenun-133, and (b) the use of organ radiation risk factors that are based
on high dose and high-dose rate exposures to low-LET radiation extrapolated to low
dose and low-dose rates. These risk factors are then converted to high-LET radiation
risks for alpha particles associated with radon and its progeny. The Committee
recommends that EPA not only make this dear in its documents but also consider
keeping the estimates or risks from inhalation and ingestion separate in its discussion
of standards for radon in drinking water.
• - - L* ' , ' * i ' . r "«,,.'••
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. The
Committee makes the following scientific comments and recommendations for
additional improvements to the document, but notes that these issues can generally
be addressed by including clarifying statements and that the changes in most cases
would not substantially change the document's estimates of central values for risks.
(A more detailed discussion of each of the comments and recommendations can be
found in the report section identified in parentheses.)

a) Recommendation Organ-specific doses used in the document for
' assessment of ingestion risks are based, in part, upon a single study of
kinetics of xenon in humans, work that has not been published in the
peer-reviewed literature. The cited study also did not include a mass
balance determination: Consequently, the Committee recommends that
the EPA carefully review this study to evaluate whether the uncertainties
attributed to the results are adequately described. (3.1.1)
'-.,• • •' ^ " • '•':- ':::•": •/ '- - . v. •' '• '; •:*••• --•:••.••••-,:•: ;' '• - .
b) Comment With regard t& assessment of inhalation risks associated with
drinking water exposure (e.g., showering), the Committee believes that
the EPA's uncertainty analysis is satisfactory and that, given the nature
of the uncertainties, the transfer factor approach used in the document
adequately accounts for risks arising from episodic shower exposures.
(3.1.2)
3
-------
c) Recommendation The Committee noted some minor inconsistencies
between values in relevant documents and recommends that the EPA
review its selection of parameter values (including ranges and their
uncertainties) for each exposure pathway to ensure consistency with
original data sources. (3.1.3)

d) Comment The Committee believes that the basic methods used to
propagate uncertainty are acceptable. Proper consideration has been
given to the possibility of covariance, and the Monte Carlo simulation
methods are state-of-the-art. (3.2.1)

e) Recommendation The Committee recommends that particular attention be
given to more completely addressing uncertainty about the variance and
shape of the probability density functions (PDFs) that have been
assumed by the EPA to represent variability in exposures among
individuals. (3.2.2)

f) Recommendation The Committee recommends that the EPA include in its
uncertainty analysis a qualitative discussion of known uncertainty
variables which were not quantified in the uncertainty analysis. These
include the issue of a linear dose rate response extending to low doses,
the influence of smoking on increasing lung-cancer risks from radon, and
the effect of population mobility on the distribution of risks. (3.2.3)

g) Recommendation In order to increase the scientific credibility of the results,
the Committee recommends that EPA consider upgrading the uncertainty
analysis for the risks associated with aeration for radon removal;
however, the proposed revisions to the analysis will not change the
conclusion that the risk for a maximally exposed individual attributable to
radon released from a water treatment facility will.be less than or equal
to the average risk attributable to 300 pCi/L of radon in drinking water
used in the home. (3.3.1)

h) Recommendation If EPA determines that granular activated carbon will be
used for radon removal, the Committee urges EPA to thoroughly and
completely analyze any potential risk and/or disposal problems related to
-------
the use of granular activated carbon (GAG) for radon removal from
drinking water

i) Recommendation EPA did not provide an analysis of occupational exposures
as a result of water treatment for radon. The potential for such
exposures appears to depend heavily upon the choice of water treatment
technology, and the Committee recommends that such a comparative
analysis be conducted for different technologies, such as aeration or
granular activated carbon filtration, especially in view of waste disposal
problems that may result from use of the latter technology. (3.3.3)

i) Recommendation The Committee recommends that the document include a
summary of the results of the uncertainty analysis regarding the
contribution of the various exposure pathways to the overall radon risk to
individuals and to the general populatipn. This summary should alsp
highlight the major sources of uncertainty contributing to the total
uncertainty in the risk estimate for each pathway. Such a discussion
would provide the information necessary to factor uncertainties and
variabilities into the cost-benefit analysis for the proposed regulation and
to calculate a range for the estimates of cost/life saved. (3.4.1)

k) Recommendation The Committee recommends that the EPA extend its
population risk assessment and uncertainty analysis to obtain an
estimate of the lives that would be saved by the proposed maximum
contaminant level, using the same assumptions as were used to
calculate present-day risks but using for radon concentration a lognormal
probability density function truncated at the maximum contaminant level.
(3.4.2)

I) 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 EPA's science is of high-quality and that it
becomes part of the mainstream of scientific criticism, revision, and
acceptance (or rejection). Publication will also assist in raising
awareness within the scientific community to the risk issues associated
with radon. (3.4.3)
-------
1.3 Policy Considerations

The comments below, to some extent, reach beyond the strictly technical issues
examined by the Committee. However, the Committee felt 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. The Committee is
extremely pleased to see that the EPA has done such 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, npt only those
associated with radiation risks. This information should be a valuable aid in guiding
EPA in its consideration of possible regulatory strategies, .: -: fi

The Committee believes strongly that the explicit disclosure of uncertainty in
quantitative risk assessment is necessary. 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. :(*&«>/

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 ad the decision-maker, to evaluate the degree of confidence that one
should have in the risk assessment, (deleted sentence redundant with end of
previous paragraph)

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-90-021). Because radon in drinking water is a
very small contributor to radon risk except in rare cases, the Committee suggested
that the EPA focus its efforts on primary rather than secondary sources of risk. Within
the limitations of the data currently available, the EPA has now successfully prepared
a scientifically credible multi-media risk 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. The -nphasis on
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.

In its May 8, 1992 Commentary on Harmonizing Chemical and Radiation Risk
Reduction Strategies (EPA-SAB-RAC-COM-92-007), the Committee brought to the
EPA's attention, the need for a more coherent policy for making risk reduction
decisions with respect to radiation and chemical exposures. The control of radon in
drinking water presents a situation where a radiological contaminant being regulated
by a paradigm developed for chemicals, yet radon in drinking water represents only a
small fraction of radon exposure. The Committee appreciates the EPA's difficulty in
establishing a coherent risk reduction strategy under the variety of statutes governing
EPA and acknowledges that harmonization does not necessarily imply identical
treatment. However, the Committee urges the EPA to explain clearly why the risks
from radiation (in this case radon in indoor air) and chemicals (in this case radon in
drinking water) are treated differently under specified conditions and in specified
exposure settings. The Committee urges EPA, the Congress and the public to
carefully consider how chemical and radiation risks are being regulated in this case.
-------
2. INTRODUCTION
2.1 Relevant Prior SAB Reports
i.
For many years the Radiation Advisory Committee and other SAB committees
have urged the incorporation of quantitative uncertainty analysis into EPA
assessments to explicitly disclose the extent of confidence that one should have in the
results of these assessments and to identify areas where the acquisition of additional
information could lead to substantial improvements in the estimation of risks and
uncertainties. In its recent multi-media radon'risk assessment study entitled,
Uncertainty Analysis of Risks Associated with Exposure the Radon in Drinking Water
(January 29, 1993) the EPA has implemented most of the SAB's recommendations in
a scientifically credible manner. A brief chronology of relevant SAB reports can be
found in Appendix A.

2.2 Procedural History of this Review

This review resulted from the Chaffee-Lautenberg amendment. (A copy of the
complete language can be found in Appendix B.) More formally known as the
Departments of Veterans Affairs and Housing and Urban Development, and
Independent Agencies Appropriation Act 1993, PUB. L. 102-398, Section 519, 106
STAT 1618 (1992), the amendment was also published in the U.S. Congressional
Record and appears as Attachment 1 to this report. Regarding this review, Congress
required EPA to,

conduct a risk assessment of radon considering: (A) the risk of adverse human
health effects associated with exposure to various pathways of radon; (B) the
costs of controlling or mitigating exposure to radon; and (C) the costs for radon
control or mitigation experienced by households and communities, including the
costs experienced by small communities as the result of such regulations.
Such an evaluation shall consider the risks posed by the treatment or disposal
of any wastes produced by water treatment The Science Advisory Board shall
review the Agency's study and submit a recommendation to the Administrator
on its findings.

This report by the SAB's Radiation Advisory Committee is a review of EPA's '>
work in response to (A). The SAB's Drinking Water Committee is reviewing the

8
-------
Agency's work in response to (B) and (C) and is generating a separate SAB report. In
addition, a subcommittee of the SAB Executive Committee will generate a third SAB
report that reviews the Agency's "synthesis document" that is being geneiaied by EPA
for submission to the Congress.

At publicly announced conference call meetings November 30, December 2,
December 3, and December 17, 1992, the Radiation Advisory Committee together
with members of the Drinking Water Committee, Environmental Engineering
Committee, and Indoor Air Quality Committee provided a consultation to the EPA staff.
The consultation was on EPA's outline for a multi-media radon risk assessment and
on the parameters and uncertainty analysis for the assessment. The SAB has
developed the consultation as a mechanism to advise the EPA on technical issues
that should be considered in the development of regulations, guidelines, or technical
guidance before the EPA has taken a position. Consultations differ from other SAB
activities in that no report is generated by the SAB and no response from the EPA is
required.

The review of "Uncertainty Analysis of Risks Associated with Exposure to
Radon in Drinking Water " (January 29, 1993), related documents and public comment
was conducted at a February 17-19, 1993 publicly announced meeting of the
Radiation Advisory Committee. The first draft of this report was made available to the
EPA and the public on February 19. Written comments were received from the EPA
and the public subsequent to the meeting. The Committee held non-public writing
sessions by conference call to revise the draft prior to its submittal to the Executive
Committee.

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. In considering adequacy in the review, the Committee was mindful
of concerns it had expressed in reports about earlier EPA documents on radon in
drinking water transmitted to the Administrator on January 9 and 29, 1992: (a) that
uncertainties associated with the selection of particular models, specific parameters
used in the models, and the final risk estimates were not adequately addressed in any
of the documents; (b) that high exposure to radon from water at the point of use (e.g.,
a shower) had not been adequately addressed; (c) that 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; auu (Q) that
regulation and removal of radon in drinking water may result in occupational
exposures.
10
-------
3. FINDINGS AND DETAILED DISCUSSION

3.1 Adequacy of Revisions to Ingestion and Inhalation Risk Estimates

3.1.1 Are revisions of ingestion risk estimates for waterborne radon and
its progeny adequate?

Recommendation Organ-specific doses used in the document for assessment
of ingestion risks are based, in part, upon a single study of kinetics of xenon in
humans, work that has not been published in the peer-reviewed literature. The cited
study also did not include a mass balance determination. Consequently, the
Committee recommends that the EPA carefully review this study to evaluate whether
the uncertainties attributed to the results are adequately described.

Discussion. Revisions of ingestion risk resulted from modifications of
gastrointestinal (Gl) and lung dosimetry and from the use of revised organ-specific risk
coefficients, particularly that for the stomach. The revised ingestion risk is greater
than the previous estimate (EPA, 1991) by a factor of 2.3. The Committee has
reviewed these revised risk coefficients. The Committee's primary concern is that
radon retention times in organs are based upon a single study of kinetics of xenon in
humans (Correia et al., 1987), work that has not been published in the peer-reviewed
literature. The xenon study also did not include a mass balance determination.
Consequently, the Committee recommends that the EPA carefully review this study to
evaluate whether the uncertainties attributed to the results are adequately described.
Other factors in the' EPA's biological model that are difficult to verify are the
assumptions that a diffusion gradient exists in the G9 tract and that lead-214 and
subsequent decay products are removed from the Gl tract before decaying and do not
contribute to dose. The implications of these assumptions have been considered in
the uncertainty analysis, and in this case also the Committee recommends the EPA
carefully review these factors to evaluate whether the uncertainties are adequately
described. Many of these uncertainties are difficult to quantify because alternative
formulations and parameter values have not been proposed. EPA has adequately
captured the apparent quantifiable uncertainties in the ingestion risk estimates and has
propagated them property, in the opinion of the Committee. However, the quantitative
uncertainty bounds may give rise to a false sense of the overall reliability of the
ingestion risk estimates. Qualitative uncertainties about the formulation of the
exposure models and the applicability of high-dose, high-dose-rate, low-LET risk

11
-------
coefficients to the low-dose, low-dose-rate, high-LE T exposure conditions present with
ingestion of radon in drinking water are substantial. An expanded discussion of the
implication of these qualitative uncertainties is important to EPA's consideration of
regulations for radon in drinking water.

3.1.2 Are revisions of inhalation risk estimates for waterborne radon and
its progeny adequate?

Comment With regard to assessment of inhalation risks, the Committee
believes that the EPA's uncertainty analysis is satisfactory and that, given the nature
of the uncertainties, the transfer factor approach used in the document adequately
accounts for risks arising from episodic shower exposures.

Discussion. The analysis of inhalation risk from radon in water has two
components. The first considers exposures from radon released from general water
use within a house. The EPA applied a general transfer factor that describes radon
release from water indoors. The factor used had a value of 1 in 10,000 (i.e., 10,000
pCi/L in water yields an average indoor air concentration of 1 pCi/L), which is
consistent with values used and published by others. In order to investigate whether
exposures to radon from releases in showers represent a significant episodic peak
exposure not captured by an average transfer factor approach, the EPA used a
multicompartment model, based on one developed by McKone (1987). Because the
analysis of shower exposures required that radon progeny ingrowth and decay be
accounted for, the model specifically recognized the differences between radon and
radon progeny exposures. The multicompartment model yielded results that were
somewhat higher for radon but somewhat lower for radon progeny when compared
with the analysis based on use of an average transfer factor.

The Radiation Advisory Committee believes, first, that the EPA's analysis,
incorporating an uncertainty analysis, is satisfactory and, second, that given the nature
of the uncertainties, the EPA's conclusion that episodic shower exposures are
adequately accounted for by a transfer factor approach is also satisfactory.
12
-------
3.1.3 Discrepancies in Numerical Values: Are EPA's choices for risk
parameters and the uncertainties adequately defended?

Recommendation The Committee noted some minor inconsistencies between
values in relevant documents and recommends that the EPA review its selection of
parameter values (including ranges and their uncertainties) for each exposure pathway
to ensure consistency with original data sources.

Discussion. Some examples of discrepancies follow.
3.1.3.1
Estimates of risk due to inhalation of indoor air
In general, the estimated central value for the annual number of lung cancer
cases and the corresponding upper and lower bounds appear to be in the same range
in the present assessment as in the previous assessment. However, the lack of
consistency in the risk factor used is troubling. The summary information presented in
Table 6-2 of the EPA document (EPA, 1993) does not appear to be entirely consistent
with the parameter values used previously. The Committee recommends that the
previous values be used throughout or that clarification of the differences be made in
the document.
3.1.3.2
Estimates of risk associated with inhalation of outdoor air
Although the total risk associated with inhalation of radon and its progeny in
outdoor air is small compared with that attributable to inhalation of radon and its
progeny in indoor air, the estimated lung cancer risks due to outdoor radon/radon
progeny exposures are, in fact, larger than those estimated to arise from radon in
drinking water. Hence, it is important that the uncertainties in the risk assessment for
the outdoor pathway be assessed in a manner consistent with that used for the indoor
(drinking water) pathway. Examples of points of concern follow:

a) There are inconsistencies in the inhalation risk factors used and in their
uncertainties. For example, the text (at p. 6-2) states that one would
expect the unattached fraction to be lower outdoors than indoors, which
is consistent with the few measurements that have been made.
However, this reduction - which would reduce the dose conversion factor
— is not reflected in the geometric mean chosen for this value, nor is the

13
-------
geometric standard deviation (GSD) increased to capture this
uncertainty.

b) The average outdoor radon concentration used in the calculations
presented (0.3 pCi/L) does not appear to be consistent with the
UNSCEAR (1988) observation that a population-weighted average valu©
is about 0.14 pCi/L. In fact, the UNSCEAR value falls outside the stated
credibility interval of 0.19 to 4.6 pCi/L. A GSD of 1.3 is dearly much too
small for a concentration as uncertain as this.

c) Similarly, relatively few measurements are available to assess the
average equilibrium factor for outdoor exposure settings. Although the
observed values fall in a small range, the GSD of 1.05 implies greater
accuracy in the value chosen (0.8) than is warranted.

d) Time spent outdoors is estimated to be 7.5%, on average. The
variability in this factor is much larger than a GSD of 1.1 would imply.

3.1.3.3 Estimates of risks and uncertainties associated with water
ingestion

The variability assumed for the amount of direct tap water consumed appears
to be biased high, at least as reflected in the analyses presented on pp. 5-26+.

3.2 Adequacy of Quantitative Uncertainty Analyses Regarding Risk
Assessment

Are quantitative uncertainty analyses regarding risk assessment of water-borne
radon, including health risk analysis and exposure analysis, adequate? At the
suggestion of the EPA staff, this question has been broken down into three subparts:

3.2.1 Are the basic methods used to propagate uncertainty acceptable?

Comment The Committee believes that the basic methods used to propagate
uncertainty are acceptable. Proper consideration has been given to the possibility of
covariance, and the Monte Carlo simulation methods are state-of-the-art.
14
-------
Discussion. In making this determination, the Committee considered the
following:

a) The EPA acknowledged uncertainty in each step of the calculation.
b) The EPA identified the sources of that uncertainty.
c) The EPA examined uncertainty about best estimate values and about
best estimate distributions whereby the distributions represent variability
in exposures and risk among individuals.
d) This latter approach whereby uncertainty is expressed about a best
estimate distribution of exposures is the current state-of-the -art in
uncertainty analyses.
e) The EPA distinguished between variability and uncertainty, which past
analyses have not always done.
f) Perhaps most important, the EPA has also shown what the most
dominant sources of uncertainty are in the calculation. In the case of the
multi-media exposures to radon, the dominant source of uncertainty is
associated with the uncertainty of translating an exposure to radon to an
estimate of health risk. This risk conversion factor will probably be the
parameter which is most difficult to estimate accurately.
g) Nevertheless, the uncertainty associated with the dose to risk conversion
for radon, although it is the dominant contributor to overall uncertainty, is
still much less than the uncertainty associated with other carcinogens
that EPA regulates. » ,

3.2.2 Are the probability density functions (PDFs) selected to describe
Type A and Type B uncertainty of each variable reasonable?

Recommendation The Committee recommends that particular attention be
given to more completely addressing uncertainty about the variance and shape of the
probability density functions (PDFs) that have been assumed by the EPA to represent
variability in exposures among individuals.

Discussion. The Committee believes that the general treatment of the PDFs
used by the EPA in its uncertainty analysis is adequate, subject to the points made
below. The EPA analysis considers two types of uncertainty. First, it recognizes that
different individuals living in an area with the same level of radon in water will have
different exposures, and therefore risks, as a result of differences in household

15
-------
characteristics, water consumption rates, and other factors. The uncertainty due to
stochastic variability in the lifetime exposure per individual in the U.S. population (Type
A uncertainty) differs from uncertainty attributable to limitations in our knowledge about
the quantities (mean, variance and shape) that describe the true distribution of
individual lifetime exposures (Type B uncertainty). This latter uncertainty also reflects
limitations that influence the average risk per individual.

While the Committee notes that the EPA analysis has not completely
recognized these distinctions, it believes that the EPA has captured the most
important features of quantitative uncertainty analysis and has adequately documented
its choice of PDFs used in its analysis for describing uncertainty about the true value
of risk for the average individual.

3.2.3 Are there any important terms or assumptions that have not been
adequately evaluated?

Recommendation The Committee recommends that the EPA include in its
uncertainty analysis a qualitative discussion of known uncertainty variables which were
not quantified in the uncertainty analysis. These include the issue of a linear dose
rate response extending to low doses, the influence of smoking on increasing lung-
cancer risks from radon, and the effect of population mobility on the distribution of
risks.

Discussion. The EPA is well aware that other model and parameter
uncertainties may be important but are difficult to quantify given current state of
knowledge. Many of these are mentioned in its draft documents, such as the issue of
a linear dose response extending to low doses. Another issue that the Committee
would like to see discussed qualitatively in the document is the influence of smoking
on increasing lung-cancer risks from radon. The risk coefficient for airborne radon is
an average value that underestimates the risk to smokers and overestimates it for
nonsmokers. The average risk value thus depends implicitly upon assumptions about
the nature of the relationship between lung cancer risk factors of smoking and radon
exposure, and on the fraction of smokers in the population.

The EPA assessment of radon in water is designed to apply to people whose
water supplies have the same radon content for their entire 70-year lifetimes. The
Committee recognizes that this design assumption is consistent with EPA policy to

16
-------
promulgate an MCL for radon that is protective for those people who might live out
their lives in a water service area with radon at the maximum contaminant level. The
Committee notes, however, that the mobility of the population implies that not every
person currently living in an area with especially high or especially low radon levels in
water will remain there. The distribution of radon exposures and risks therefore will
not be the same as if every person remained in the same area for a lifetime. In
general, fewer people will have very high or very low exposures and risks and more
will have intermediate levels of risk than under the no-mobility assumption. The effect
of mobility on overall population risk (cancers per year in the United States arising
from radon in drinking water), in contrast, wilt likely be negligible because most people
moving from a high radon area to a low one will be replaced by people moving in the
other direction, except for any effect of net population migration within the country.

3.3 Adequacy of Characterization of Risks from Water Treatment Facilities

3.3.1 Has the EPA adequately characterized the risks introduced by radon
that would be released by aeration from water treatment facilities?

Recommendation In order to increase the scientific credibility of the results, the
Committee recommends that EPA consider upgrading the uncertainty analysis for the
risks associated with aeration for radon removal; however, the proposed revisions to
the analysis will not change the conclusion that the risk for a maximally exposed
individual attributable to radon released from a water treatment facility will be no more
than the average risk attributable to 300pCi/L of radon in drinking water used in the
home.

Discussion. The EPA has proposed air-stripping as Best Available Technology
(BAT) for achieving the proposed radon standard for drinking water where current
levels exceed the proposed standard. Recognizing that this technique would
discharge much of the waterborne radon to the atmosphere, the EPA analyzed the
risks of such discharges in terms of the risks to a maximally exposed individual (MEI)
living near the treatment facilities. The EPA also projected the population risk or
annual cancer incidence assuming that each water supplier exceeding the proposed
standard were to use air-stripping at a single location in order to bring itself into
compliance with the proposed standard (EPA, 1988; 1989).
17
-------
The EPA reasoned that if the individual and population risks from the treatment
facilities were small relative to the risks avoided by applying the proposed standard,
then a comparative risk tradeoff would favor implementation of the standard. To
ensure that this comparison would not favor the proposed standard solely through
differences in assessment methods, the EPA estimated the risks attributable to waiter
treatment by using two radiation risk models, AIRDOSE and MINEDOSE. Although
the Committee has reservations about the degree of validation of these models, the
MINEDOSE model is thought to provide conservative risk estimates. In the
assessment of risk from water treatment, the EPA also made assumptions that were
the same as or more conservative than those used for assessing the risks of radon in
water used in the home. Specifically, the individual risks were calculated for an MEI
who was defined as exposed to the highest concentrations for the longest possible
time from discharges under worst-case meteorological conditions. The Committee
concurs that the set of assumptions chosen was generally quite conservative.

The MEI risks presented to the Committee ranged up to 8 x 10"4, or about 4
times the nominal value for the risk of 300 pCi/L radon in drinking water. However,
this was a single value derived from largely unrealistic assumptions, and more typical
MEI risks appear to be much lower, generally falling at or below the risk due to
exposure to radon in drinking water at 300 pCi/L

The EPA also projected population risk using AIRDQSE and estimated total
cancer death rate of approximately 0.1/yr, a value that is considerably less than the
reduction of 80 cancer deaths/yr estimated to be achieved by implementing the
proposed standard.

The EPA conducted a semiquantitative uncertainty analysis of the MEI risk
calculation and concluded that upper bound risks would remain in the vicinity of 1 x
10"4, given the conservative nature of the nominal values. The uncertainty analysis
was less rigorous and more subjective than that for the risks of radon in drinking
water. Although more rigor is unlikely to change the conclusion, improvement of the
uncertainty analysis would improve the scientific credibility of the results.
18
-------
3.3,2 Has the EPA adequately characterized the risks introduced by radon
that would be released from other types of water treatment
facilities?

Recommendation If EPA determines that granular actived carbon wilt be used
for radon removal, the Committee urges EPA to thoroughly and completely analyze
any potential risk and/or disposal problems related to the use of granular activated
carbon (GAG) for radon removal from drinking water

Discussion Another technology for radon removal from drinking water is
Granular-Activated-Charcoal (GAG). Although GAC has not been designated a best
available treatment (BAT) for radon removal, in a draft technical memorandum from
the Office of Water (dated January 1993 and circulated to the RAG on February 18,
1993), EPA discussed various issues related to the use of this technology which
mentioned radioactivity accumulation in the GAC (mostly lead-210). However, while
the memorandum mentioned the issue of GAC building up levels of radioactivity such
that the residuals would require disposal at a low-level-radioactive-waste (or naturally
occurring radioactive material waste) repository, the memorandum was without
sufficient data or analysis for the Committee to evaluate this possibility and the
implications of this problem.

The Committe'e urges EPA to thoroughly and completely analyze any potential
risk and/or disposal problems related to the use of GAC for radon removal from
drinking water.

3.3.3 Occupational Exposures

Recommendation EPA did not provide an analysis of occupational exposures
as a result of water treatment for radon. The potential for such exposures appears to
depend heavily upon the choice of water treatment technology, and the Committee
recommends that such a comparative analysis be conducted for different technologies,
such as aeration or granular activated carbon filtration, especially in view of waste
disposal problems that may result from use of the latter technology.

Discussion. The EPA did not provide an analysis of potential radiation
exposures to workers in water treatment or.ancillary facilities. The RAC notes that in
the case of aeration techniques, proper ventilation of the water treatment facility

19
-------
should result in little increase in radon concentrations and exposures to personnel.
There should be no other significant sources of radiation due to such treatment.
However, the EPA has not ruled out treatment by other means, including granular
activated carbon filtration (GAG), in which case build-up of radon progeny in the bed
can result in an increased radiation field near the beds. Furthermore, the handling
and disposal of GAG beds containing radionuclides has not been analyzed nor, in fact,
have provisions been made for such disposal in the event it is necessary. In order to
provide a complete risk analysis, the Committee believes that the EPA needs to
consider the possibility of worker exposures either to radiation or to chemicals (such
as those used as biocides in aeration facilities) resulting from some water treatment
technologies.

3.4 Other Scientific Issues

3.4.1 Recommended extensions of the risk and uncertainty analysis and
publication of results in peer-reviewed journals

Recommendation The Committee recommends that the document include a
summary of the results of the uncertainty analysis regarding the contribution of the
various exposure pathways to the overall radon risk to individuals and to the general
population. This summary should also highlight the major sources of uncertainty
contributing to the total uncertainty in the risk estimate for each pathway. Such a
discussion would provide the information necessary to factor uncertainties and
variabilities into the cost-benefit analysis for the proposed regulation and to calculate a
range for the estimates of cost/life saved. (3.4.1)

Discussion. One aspect that was lacking in the reviewed document was a
summary and interpretation of the uncertainty analysis for radon in drinking water.
The Committee has studied the results presented by the EPA and offers the following
interpretation.
3.4.1.1
Individual risks
The following table lists the unit risks attributable to drinking water by inhalation
and ingestion pathways, including the 90% confidence interval around the median, the
upper-bound 95th percentile, and the lower-bound 5th percentile for risk.
20
-------
Table 1. Unit Risk Boundaries for Exposure to Radon in Drinking Water
(Fatal cancers/person/year per pCi/L)

Inhalation
Ingestion
5th
percentile
Lower Bound
1.6x 10'"1
1.2 x 10'"
5th percentile
Median
1.1 x 10"'
3.7 i! 10'18
Median
2.7 X 10''
1.7X10*
95th pereentile
Median
6.3 x 10-'
«.5 x 10"*
95tfi
Percentile
Upp«r Bound
4.2 X 10-*
2.0 X 10"1
The nominal unit risk in the proposed rule is 9.4 x 10'9 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 median inhalation 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 risk
presented in Chapter 3 of the reviewed document.
3.4.1.2
Population risks
The estimates of cancer fatalities due to exposure of radon in drinking water
are based upon 81 million people being exposed. This number was presented to the
Committee during a briefing on 2/17/93, and comes from a preliminary contractor
report on occurrence of radon in drinking water (Wade Miller, 1992). That report is
being reviewed by the EEC of the SAB. Any changes in that estimate will affect the
results presented below.

Table 2. Cancer Fatalities per Year due to Exposure to Radon
Exposure Pathway
Inhalation due to Water
Treatment
Inhalation from Drinking
Water
Ingeeiion from Drinking
Water
Inhalation from Outdoor Air
Inhalation from Indoor Air
Sth Percent!!*
Median
—
4«
18
280
6.790
Median
—
105
53
657
14.410
95th PeroentJto
Median
—
233
166
1.SOO
30.950
Upper
Bound
< 1
—
—
—
—
21
-------
The estimated lung cancer deaths attributable to inhalation exposure to radon in
drinking water range from 48 to 233 per year. The estimated fatal cancer cases ,
attributable to ingestion exposure to radon in drinking water range from 19 to 166 per
year. Therefore, estimated total fatal cancer cases attributable to waterborne radon
will be about a quarter of the risks associated with exposure to radon in outdoor air,
and about one percent of the risks associated with exposure to radon in indoor air and
of the total risks attributable to exposure to radon by all pathways. These calculations
also indicate that population risks from exposure to radon in drinking water are similar
to or higher than those normally addressed by regulation of chemical pollutants in
drinking water. Although the risk attributable to inhalation and ingestion of radon in
drinking water were apportioned equal weight in the calculation of the nominal value in
Chapter 3, the weight obtained as a result of the uncertainty analysis is approximately
two-thirds for inhalation and one-third for ingestion. This last set of values is similar to
those presented in the Proposed Rule (EPA, 1991).

3.4.2 Estimate of Lives Saved
Recommendation The Committee recommends that the EPA extend its
population risk assessment and uncertainty analysis to obtain an estimate of the lives
that would be saved by the proposed maximum contaminant level, using for radon
concentration the same assumptions as were used to calculate present-day risks but
using a lognormal probability density function truncated at the maximum contaminant
level.

Discussion. The Committee could not carry out an analysis of the estimated
number of lives that would be saved by the Proposed MCL of 300 pCi/L because no
uncertainty analysis was done on the number of cancer fatalities projected for the rule
in place. The Committee recommends that a population risk assessment and
uncertainty analysis be carried out, using the same assumptions as were used to
calculate present-day risks but using for radon concentration a lognormal PDF
truncated at the proposed MCL. An uncertainty for the tolerance in the measurement
of radon as described in the section regarding monitoring of the Proposed Rule should
also be factored into this uncertainty analysis. From these calculations, one would
obtain a 90% confidence interval for the cancer fatalities that would remain with
enforcement of the proposed MCL, and the difference between the values in Table 2
and those calculated with the truncated PDF would yield a range of lives saved. This
analysis would then allow the persons conducting the cost-benefit analysis to factor

22
-------
these uncertainties and variabilities into their calculations, leading to a range of costs
per life saved. The Committee believes that this extension to the EPA's uncertainty
analysis would enhance the usefulness of tne document reviewed.

3.4.3 Peer Review and Publication

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 EPA's science is of high-quality and that it becomes part of the mainstream of
scientific criticism, revision, and acceptance (or rejection). Publication will also assist
in raising awareness within the scientific community to the risk issues associated with
radon.

Discussion. The Committee believes that overall, the use of the peer-reviewed
literature as both a source of data and information and also as a method of
disseminating the EPA's own scientific work is an important means by which the EPA
and the public can be assured that the best science is being used or produced. In
this particular case, the estimate of the ingestion risk due to radon in drinking water
rests heavily upon data and analyses that have not been published and therefore have
not been broadly circulated within the scientific community. Reliance upon such
results should be done with considerable caution.

Although publication in peer-reviewed journals does not, by itself, assure
infallibility, it is the only generally recognized means by which scientific work gets
accepted by members of the scientific community. In seeking to improve the quality
and the scientific acceptability of its science, the EPA should encourage its scientists
to submit their work for peer-reviewed publication. The work and methodologies
presented here mark an important advance in the risk and uncertainty analyses
undertaken by the EPA and are certainly worthy of such publication.
23
-------
4. POLICY CONSIDERATIONS

4.1 The Importance of Quantitative Uncertainty Analysis

The Radiation Advisory Committee has long encouraged the use of integrated
quantitative uncertainty analysis in a variety of EPA assessments. The Committee is
extremely pleased to see that the EPA has done such 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.

The Committee believes strongly that the explicit disclosure of uncertainty in
quantitative risk assessment is necessary any time the assessment is taken beyond a
screening calculation. Screening risk assessments typically involve only point
estimate calculations. The assumptions used to derive these point estimates are
generally biased on the conservative side to ensure that the true risk to individuals will
not be underestimated. Screening calculations are thus useful for identifying situations
that are dearly below regulatory risk levels of concern. They can be grossly
misleading in terms of indicating the need for regulatory action.

The need for regulatory action must be based on more realistic estimates of
risk. Realistic risk estimating, however, requires a full disclosure of uncertainty. The
disclosure of uncertainty enables the scientific reviewer, as well as the decision-maker,
to evaluate the degree of confidence that one should have in the risk assessment.
The confidence in the risk assessment should be a major factor in determining
strategies for regulatory action.

Large uncertainty in the risk estimate, although undesirable, may not be critical
if the confidence intervals about the risk estimate indicate that risks are clearly below
regulatory levels of concern. On the other hand, when these confidence intervals
overlap the regulatory levels of concern, consideration should be given to acquiring
additional information to reduce the uncertainty in the risk estimate by focusing
research on the factors that dominate the uncertainty. The dominant factors
controlling the overall uncertainty are readily identified through a sensitivity analysis
conducted as an integral part of quantitative uncertainty analysis. Acquiring additional

24
-------
data to reduce the uncertainty in the risk estimate is especially important when the
cost of regulation is high. Ultimately, the explicit disclosure in the risk estimate should
be factored into analyses of the cost-effectiveness of risk reduction as weii as in
setting priorities for the allocation of regulatory resources for reducing risk.

4.2 The Relative Risk of Radon in Drinking Water

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-90-021). Because radon in drinking water is a
very small contributor to radon risk except in rare cases, the Committee suggested
that the EPA focus its efforts on primary rather than secondary sources of risk. Within
the limitations of the data currently available, the EPA has now successfully prepared
a scientifically credible multi-media risk 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. The emphasis on
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.

4.3 Harmonizing

In its May 8, 1992 Commentary on Harmonizing Chemical and Radiation Risk
Reduction Strategies (EPA-SAB-RAC-COM-92-007), the Committee brought to the
EPA's attention the need for a more coherent policy for making risk reduction
decisions with respect to radiation and chemical exposures. The control of radon in
drinking water presents a situation where a radiological contaminant being regulated
by a paradigm developed for chemicals yet radon in drinking water represents only a
small fraction of radon exposure. The Committee appreciates the EPA's difficulty in
establishing a coherent risk reduction strategy under the variety of statutes governing
EPA and acknowledges that harmonization does not necessarily imply identical
treatment. However, the Committee urges the EPA to explain dearly why the risks
from radiation (in this case radon in indoor air) and chemicals (in this case radon in
drinking water) are treated differently under specified conditions and in specified

25
-------
exposure.settings. The Committee urges EPA, the Congress and the public to
carefully consider how chemical and radiation risks are being treated in this case.
26
-------
S. REFERENCES

5.1 Documents Received by the Radiation Advisory Committee During this
Review

A. DOCUMENTS RECEIVED BEFORE THE FEBRUARY 17-19 PUBLIC
MEETING

Documents Provided by EPA

1. Departments of Veterans Affairs and Housing and Urban Development, and
Independent Agencies Appropriation Act, 1993, PUB. L. 102-398, Section 519,
106 STAT 1618 (1992)

2. Draft 2 "Uncertainty Analysis of Risk Associated with Exposure to Radon in
Drinking Water" prepared by U.S. EPA 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, January 29, 1993

3. Proposed Revisions in EPA Estimates of Radon Risks and Associated
Uncertainties
4. An Analysis of the Uncertainties in Estimates of Radon-Induced Lung Caner by
Jerome S. Puskin in Risk Analysis Volume 12, Number 2. 1992

5. Response to SAB Radon Comments

6. Preliminary Risk Assessment for Radon Emissions from Drinking Water
Treatment Facilities, a memorandum from Warren D. Peters and Christopher B.
Nelson to Stephen W. Clark, June 28, 1988

7. An Analysis of Potential Radon Emissions from Water Treatment Plants Using
the MINEDOSE Code, a memorandum from Pare. J. Parrotta to Greg Helms,
November 22, 1989

8. Proposed Methodology for Estimating Radiogenic Cancer Risks (no author or
date given)

27
-------
9. Cancer Fatalities from Waterborne Radon (Rn-222) by Douglas J. Crawford-
Brown in Risk Analysis, Volume 11, Number 1, 1991

Public Comment

10. Letter re: National Primary Drinking Water Regulations: Radionudides (Radon)
[WH-FRL 3956-4] from John H. Sullivan of the American Water Works
Association to Honorable Carol Browner, Administrator of the Environmental
Protection Agency, January 26, 1993. There were 27 Appendices to this letter.

1. EPA Technical Support Document for the 1992 Citizen's Guide
to Radon, EPA 400-R-92-011 (May 1992)
2. "Harmonizing Chemical and Radiation Risk-Reduction Strategies -
A Science Advisory Board Commentary,"(May 18,1992)
3. Letter from SAB Chairman Raymond Loehr to EPA
Administrator William Reilly Re: "Radionudides in
Drinking Water" (EPA-SAB-RAC-91-XXX) (September 1991)
4. "An SAB Report: Radionudides in Drinking Water
(EPA-SAB-RAC-91-009) (December 1991)
5. Letter from SAB Chairman Raymond Loehr to EPA
Administrator William Reilly Re: "Reducing Risks
from Radon; Drinking Water Criteria Documents,"
(EPA-SAB-RAC-COM-92-003) (January 29, 1992)
6. Letter from SAB Chairman Raymond Loehr to EPA
Administrator William Reilly Re: "Status of EPA
Radionudides Model" (EPA-SAB-RAC-COM-92-001)
(January 9, 1992)
7. SAB, "Review of the office of Drinking Water's
Assessment of Radionudides in Drinking Water and
Four Draft Criteria Documents" (July 1987)
8. Letter from SAB Chairman Raymond Loehr to EPA
Administrator William Reilly Re: "Review of Draft
Criteria Documents for Radionudides in Drinking
Water" (EPA-SAB-RAC-92-0009) (January 9, 1992)
9. Letter from SAB Chairman Raymond Loehr to EPA
Administrator William Reilly Re: "Revised Radon
Risk Estimates and Assodated Uncertainties" (EPA-

28
-------
SAB-RAC-LTR-92-003) (January 9, 1992)
10. Puskin, Jerome, "An Analysis of the Uncertainties in
Estimates of Radon-Induced Lung Cancer," Risk
Analysis, Vol. 12, No. 2, p. 277 (1992)
11. SENES Consultants Limited Memorandum Re: "Exposure
and Risk from Radon Released in Showers" (December 3, 1992)
12. Fensterheim, Robert, Stolwijk, Jan, "Critique of
Hess and Bernhardt Radon Shower Exposure Study,"(1992)
13. Testimony of Jonathan M. Samet before the
Subcommittee on Transportation and Hazardous
Materials, House Energy and Commerce Committee
(June 3, 1992)
14. Neuberger, John S., "Residential Radon Exposure and
Lung Cancer: An Overview of Published Studies,"
Cancer Detection and Prevention, Vol.15, Issue 6,
(1991) 435-443
15. Neuberger, John S., et al., "Residential Radon
Exposure and Lung Cancer: Evidence of an Inverse
Association in Washington State," Journal of
Environmental Health, Nov/Dec. 1992, 23-25
29
-------
16. "Proposed Guideline for Radon-222 in Drinking
Water," prepared by SENES Consultants Limited for
Health Protection Branch, Health and Welfare Panada
(March 1992)
17. Draft of SAB Radiation Advisory Committee Comments
on EPA's "Suggested Guidelines for the Disposal of
Drinking Water Treatment Wastes Containing
Naturally-Occurring Radionuclides" (July 6, 1992)
18. Testimony of Dr. Jill Lipoti on HR 3258, the "Radon
Awareness and Disclosure Act of 199111 before the
House Subcommittee on Transportation and Hazardous
Materials (June 3, 1992)
19. Factor Analysis for Differences Between EPA and RCG
Compliance Cost Estimates
20. Table Comparing Compliance Costs for A Radon MCL of
300 pci/1; Letter to Editor and Response in American
Water Works Association Journal
21. Comments of the State of Idaho Department of Water
Resources (May 13, 1992)
22. Letter from Dr. Alvin Young, Chairman of Committee
on Interagency Radiation Research and Policy
coordination, to Dr. Donald Henderson, Office of
Science and Technology Policy (May 21, 1992)
23. Testimony of Dr. Jan Stolwijk before the House
Subcommittee on Transportation and Hazardous
Materials (June 3, 1992)
24. Valentine, Richard, "Radon and Radium From
Distribution System and Filter Media Deposits/" AWWA
Water Quality Technology conference, Toronto (1992). 24
25. Comments of the State of New York Department Health
to EPA (February 12, 1992)
26. "Evaluation of the Impact of a Radon-222 MCL on
Small Water Systems," by John E. Reanier, Alabama
Rural Water Association (May 10, 1992)
27. Comments of the Association of State Drinking Water
Administrators (November 19, 1991)
30
-------
11. Letter to Administrator Browner and three SAB Chairs from Bill Mills, Steve
Hall, and Tom Levy of the Alliance for Radon Reduction, February 2, 1993
31
-------
B. DOCUMENTS RECEIVED AT THE FEBRUARY 17-19 PUBLIC MEETING

Documents Provided by EPA

1. Draft Summary (no date or author given, appears to be a draft summary for the
"Uncertainty Analysis of Risk Associated with Exposure to Radon in Drinking
Water"

2. Overheads: Briefing for SAB on Multimedia Risk Assessment of Human
Exposure to Radon, Office of Science and Technology, Office of Radiation and
Indoor Air, Office of Policy, Planning, and Evaluation, Office of Ground Water
and Drinking Water.

3. Overheads: Risk Assessment for Radon Emissions from Drinking Water
Treatment Facilities, EPA Office of Radiation and Indoor Air, February 17, 1993

4. Overheads: Cancer Risks Associated with Radon in Drinking Water-
Uncertainty and Variability Analysis

5. "Review of Risk Assessments of Radon Emissions from Drinking Water
Treatment Facilities" from Christopher Nelson ORIA to Mark Parrotta ODW

6. Radon Documents for SAB Review, a memorandum from Nancy Chiu of
OST/OW to William F. Raub, Science Advisor

7. Draft Technical Memorandum: Problems with the Use of GAC for Radon
Removal, printed date is January 1993 (handwritten date is 2/11)

Public Comment
8. Review of Technical Justification of Assumptions and Methods Used by the
Environmental Protection Agency for Estimating Risks Avoided by Implementing
MCLs for Radionuclides by S.C. Morris, M.D. Rosw, S. Holtzman, and A.F.
Meinhold and Brookhaven National Laboratory, November, 1992

9. Letter from Edward J. Schmidt to Comments Clerk-Radionuclides, Subject
Comments on National Primary Drinking Water Regulations: Radionuclides

32
-------
Proposed Rule, 40CFR Parts 141 & 142, Thursday, July 18, 1991, September,
30, 1991

10. Letter to James R. Elder from Raymond F. Pelletier, Office of
Environmental Guidance, U.S. Department of Energy, January 27, 1993

C. DOCUMENTS RECEIVED SUBSEQUENT TO THE FEBRUARY 17-19
PUBLIC MEETING

Documents Provided by EPA

1. One-page note to Kathleen Conway from Jan Auerbach, February 23, 1993

2. Note to Kathleen Conway, RAC DFO from Nancy Chui OGWDW, faxed to the
Radiation Advisory Committee, March 10, 1993

Public Comment

3. Letter to the SAB Radiation Advisory Committee from Frederick W. Pontius of
the American Water Works Association, February 24. This letter had seven
enclosures:
a. Lognormal Distributions for Water Intake by Children and Adults,
by Ann M. Roseberry and David. E. Burmaster in Risk Analysis,
Volume 12, Number 1, 1992
b. Distribution and Expected Time of R e sidence for U.S.
Households by Milton Israeli and Christopher B. Nelson in Risk
Analysis, Volume 12, Number 1, 1992
c. Review of Risk Estimates for Inhalation of Radon Progeny by
Miners: Presentation by the Atomic Energy Control Board of
Canada (ACB) before the ICRP Main Commission, printed date is
November 1992, there is also a stamped date of February 12,
1993
d. A Cohort Study in Southern China of Tin Miners Exposed to
Radon and Radon Decay Products by Xuan Xiang-Zhen, Jay. H.
Lubin, and others in Health Physics, Volume 62, Number 10,
pages 120-131, February 1993
33
-------
5.
6.
7.
8.
e.
f.
g.
Contribution of Waterborne Radon to Home Air Quality, prepared
by Arun K. Deb of Roy F., Eston, Inc. for the AWWA Research
Foundation, undated
Final Report: Risk and Uncertainty Analysis for Radon in Drinking
Water prepared by Douglas J. Crawford Brown for the American
Water Works Association.
Proposed Guideline for Radon-222 in Drinking Water prepared by
SENES Consultants Limited for the Health Protection Branch of
Health and Welfare Canada, March 1992
Letter to the Radiation Advisory Committee from Douglas Crawford-Brown of
the University of North Carolina, March 2, 1993

Letter to Dr. Genevieve Matanoski from Bill Mills, Steve Hall and Tom Levy of
the Alliance for Radon Reduction, March 11, 1993

Letter to Dr. Genevieve Matanoski from Robert J. Fensterheim, consultant to
the Alliance for Radon Reduction, March 16, 1993

Fax from Robert J. Fensterheim referencing Brown-Senate Letter and Naomi
Hariey Study, March 16, 1993. This fax included both a March 11, 1993 letter
from nine senators to Administrator Carol M. Browner and A Biokinetic Model
for the Distribution of Rn-22 Gas in the Body Following ingestion by Naomi H.
Hariey and Edith S. Robbins, March 12, 1993

Letter to Dr. Vem Ray, Chairman of the Radon Engineering Cost Subcommittee
from Stephen Hall of the Association of California Water Agencies, March 22
1993
34
-------
5.2 Science Advisory Board Reports of Potential Interest

1. Report o* 'he Scientific Basis cf EPA's Proposed National Emission Standards
for Hazardous Air Pollutants for Radionuclides: A report of the Subcommittee
on Risk Assessment for Radionuclides, August 1984 (There is no report
number because this report was produced before the SAB developed a report
numbering system.)

2. Radionuclides in Drinking Water (SAB-RAC-87-035)
3. Effective Dose Equivalent Concept (SAB-RAC-88-026)
4. Radon Risk Estimates (SAB-RAC-88-042)
5. Radionuclides NESHAP (SAB-RAC-89-003)
6. EEC Mathematical Models Resolution (SAB-EEC-89-01)
7. Radionuclides NESHAP (SAB-RAC-89-024)
8. Radon Risks (SAB-RAC-91-LTR-001)
9. Status of EPA Radionuclide Models (EPA-SAB-RAC-COM-92-00)
10. Revised Radon Risk Estimates and Associated Uncertainties
(EPA-SAB-RAC-LTR-92-003)
11. Criteria Documents for Radionuclides in Drinking Water
(EPA-SAB-RAC-92-009)
12. Reducing Risks from Radon/Drinking Water Criteria Documents
(EPA-SAB-RAC-COM-003)
13. Harmonizing Chemical and Radiation Risks (EPA-SAB-RAG-COM-92-007)
14. Drinking Water Treatment Wastes Containing NORM (EPA-SAB-RAC-LTR-92-
018)
15. Radon in Water: Consultation (EPA-SABrRAC-CON-92-002)
35
-------
5.3 Literature cited

Correia JA, Weise SB, Callahan RJ, Strauss HW, 1987. The kinetics of ingested Rn-
222 in humans determined from measurements with Xe-133. Massachusetts
General Hospital, Boston MA, unpublished report Prepared for Health Effects
Research Laboratory, U.S. EPA, Report No. EPA/60071-87/013.

Crawford-Brown DJ, 1991. Cancer fatalities from waterborne radon (Rn-222). Risk
Anal. 11:135-143.

EPA, 1988. "Preliminary Risk Assessment for Radon Emissions from Drinking Water
Facilities," memorandum from Warren Peters and Christopher Nelson to Stephen
Clark, June 28, 1988.

EPA, 1989. "An Analysis of Potential radon Emissions from Water Treatment Plants
using the MIMEDOSE Code," memorandum from Marc Parrotta to Greg Helms,
November 22, 1989.

EPA, 1991. "Notice of Proposed Rulemaking for Radibnuclides in Drinking Water"

EPA, 1989. "Draft 2 "Uncertainty Analysis of Risk Associated with Exposure to
Radon in Drinking Water" prepared by U.S. EPA 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, January 29, 1993

McKone, TE, 1987. Human exposure to volatile organic compounds in household tap
water the indoor inhalation pathway. Environ. Sci. Technol. 21:1194-1201

UNSCEAR, Sources, Effects and Risks of Ionizing Radiation, United Nations Scientific
Committee on the Effects of Atomic Radiation, United Nations: New York, 1988,
p.64.

Wade Miller Associates, 1992, Draft addendum to the occurence and exposure
assessments for radon, radium-226, radium-228, uranium, and gross alpha particle
activity in public drinking water supples. EPA contract No. 68-CO-0069 September
30, 1992.
36
-------
APPENDIX A: Brief Chronology of Relevant SAB Reports

In 1984 a specialized ad hoc Subcommittee of the Science Advisory
Board reviewed the scientific basis for EPA's proposed national emissions
standards for hazardous air pollutants for radionudides. That report led to the
formation of the Radiation Advisory Committee to "review risk assessments for
radiation standards". The report also stated,"A scientifically defensible risk
assessment for radionudides should address at least five major elements. These
indude 1) identification of the significant. . . sources; 2) a description of the
movement ••.'. from a source ... to people; 3) calculation of doses; 4) estimation of
. . . health effects, and 5) incorporation of estimates of uncertainty into elements
1 -4. . . ." The routine incorporation of uncertainty analysis into risk assessments /
has been a recurring theme in Radiation Advisory Committee reports.

In the summer of 1986, the Drinking Water Subcommittee of the
Radiation Advisory Committee reviewed the Office of Drinking Water's
Assessment of Radionudides jn Drinking Water and Four Draft Criteria
Documents, (SAB-RAC-87-035). This Subcommittee did not explicitly address
uncertainty analysis. While recommending spm'd improvements in science and
presentation, the Subcommittee concluded, "that the Office of Drinking Water has
developed scientifically comprehensive assessment documents." This report was
transmitted to the Administrator July 27, 1987.

In 1988 and 1989 reviews of revisions to the scientific basis for the
radionudides NESHAP, the Radiation Advisory Committee again raised concerns
about quantitative uncertainty analysis. The cover letter of the November 10,
1988 report (SAB-RAC-89-003) highlighted three findings for serious attention by
the EPA, induding, "To date, EPA's treatment of modeling uncertainties has been
qualitative rather than quantitative although state-of-the-art methods for estimating
uncertainty are available." The June 30, 1989 report (SAB-RAC-89-024) noted in
the cover letter (p.2), "... the Radiation Advisory Committee and the Science
Advisory Board has repeatedly urged the use of best estimates and ranges in the
specifications of risk, and a detailed explanation of the uncertainties in the
estimates themselves."

On January 13, 1989, the SAB transmitted to the Administrator the
Environmental Engineering Committee's Resolution on the Use of Mathematical

A-1
-------
Models by EPA for Regulatory Assessment and Decision-Making (EPA-SAB-EEC-
89-012). The Committee (p.1) had reviewed "a number of integrated
environmental modeling studies" and "noted a number of problems" including, "a
lack of studies quantifying the uncertainties associated with model predictions, and
concurrently, the potential misuse of particular uncertainty analysis techniques."
The resolution's fourth recommendation (p.3) was, "Sensitivity and uncertainty
analysis of environmental models and their predictions should be performed to
provide decision-makers with an understanding of the level of confidence in model
results, and to identify key areas for future study."

In the summer of 1990, the Radionudides in Drinking Water
Subcommittee of the Radiation Advisory Committee reviewed draft criteria
documents for radionuclides in drinking water, including those for uranium, radium,
radon, and a combined document on beta particles and gamma emitters.
The Subcommittee found that, "The overall quality of the four draft criteria
documents was not good.. . . recommendations from a 1987 Science Advisory
Board report on its review of the standards for radionuclides in drinking water
(SAB-RAC-87-035) had not been addressed. Nor did the new criteria documents
address recommendations from other available SAB reports that are directly
relevant (such as SAB-RAC-88-026 and SAB-EEC-89-012). . . . Uncertainties
associated with the selection of particular models, specific parameters used in the
models, and the final risk estimates are not adequately addressed in any of the
documents." Although the review was conducted in 1990 and draft reports
circulated at that time, this SAB report was not transmitted to the Administrator
until January 9, 1992. (EPA-SAB-RAC-92-009)

In the summer and fall of 1991, the Radiation Advisory Committee
received revised criteria documents and declined to review them. It did, however,
produce a commentary which noted (p.4) that, "Although each criteria document
now includes a chapter discussing uncertainty, the content of the chapters is very
qualitative and is not the rigorous technical analysis envisioned by the
Committee." In its section on policy considerations, the Committee also noted
(p.3) that, "radon in drinking water is a very small contributor to radon risk except
in rare cases and the Committee suggests the EPA focus its efforts on primary
rather than secondary sources of risk." This commentary was transmitted to the
Administrator January 29, 1992 (EPA-SAB-RAC-COM-92-003)
A-2
-------
The January 9 and 29, 1992 reports also contained other advice
relevant to the scientific assessment of the risk of radon in drinking water.
Additionally, the January 29, 1992 report provided policy-related comments on
radon in light of the SAB report, Reducing Risk. A May 8, 1992 Radiation
Advisory Committee commentary, "Harmonizing Chemical and Radiation Risk
Reduction Strategies," described chemical and radiation risk reductions
paradigms, discussed the difficulties of applying a paradigm developed for one
type of contaminant to the other, and recommended harmonization.

In the winter and spring of 1992, the Committee conducted a review
of the EPA's, "Suggested Guidelines for the Disposal of Drinking Water Treatment
wastes Containing Naturally-Occurring Radionudides" dated July 1990. The
Committee found that such guidelines were needed because of the potential
radiation doses to treatment plant workers and the public. However, the 1990
guidelines did not fully assess the magnitude of risk from exposure to treatment
wastes, nor did the document specify whether the radiation exposures to workers
should be considered as occupational exposures or viewed against dose limits for
the general public, a decision which will have considerable bearing on any final
guidelines. This letter report was transmitted to the Administrator September 30,
1992 (EPA-SAB-RAC-LTR-92-018).
A-3
-------
-------
APPENDIX B: Chaffee-Lautenberg Language from the Congressional Record
B-1
-------
Jptcmber25, 1992

ctptents to further science, technology develop-
• '.' „•„':•-••?
Rtsotved, Tlxac ca* Hoc** recede trora its
dl**er*«si*nt to Ui» amendment of th* Sen-
tu numbered >*S -to th* aforesaid bill, and
pooeor therein with aa amAOdmact as fol-
low*:
In lUo. of th* matter proposed by said
tneodmeac, insert:
Chapter X of tute XI ef the Dire Emergency
Supytsmemtal Aforvtriationt Act. 1892. Includ-
ing Disaster Attutonce ta iCeet the Pretent
EmtraenOet Arising from the Consequences of
Hurricane Andrew. Typhoon. Omar. Hurricane
IrJU. and Other MUvrel Disaster*, and Addi-
tional Assistance ta Distressed Conununiaet
(H.R. 5630) it amended by (1) striking the matter
Knder the headtne "Disaster ntHef' and insert
In Heu thereof: "Far necetaory expenset in osr-
ryOtf out the Robert T. Stafford Disaster Relief
and Emergency Assistance Act. at amended.
tlJS3.000.000. oftt+tch net to noted £50.000.000
may te transferred' to the "Disaster Atsittance
Drrtct Loan Procnm' aowwit for admtrJstra-
at* expense* and naOttdutJor direct leant pro-
vided under teeaen 411 ef tuch Act. and .of
uMch tHijDOOjQOO ***a be available only to the
extent an offlctat budget refuert. for o> specific
dollar am*unt,'thmt tnchtde* dtalfnatlOK of the
entire amount of the reqvett a* an enerpency
tejutrtsnent at defined in the Balanced Budget
and Emergency Deficit Control Act of 19IS. )(3)(D)(t) of the Balanced Budget and
Emergency Deftcu Control Aat of /MS. at
amended." and (II ttntang the natter under the.
heading "Disaster assistance forest loan pro-
gram account" e*d insert in Heu thereof: "The
Unitarian on datrt to*** for t\e 'Disaster at-
ratance direct lo*n yogrocn account' is in-
creased, within eraong fundt. by UX.OOOJOOO ta
not ta exceed UUOUOOOO: Provided further.
That net to exceed tajUOfM it .available for di-
rect loan obligations provided to eligible appli-
cants or to Stattt under section 311 of the Rob-
ert f. Stafford Dttatter At*ltta,nce and &ner-
sency KeUef Act, as attended: Provided further.
That not to exceed HUOJUOO.OOO It available for
eooununlty dt*aiter loans to local oocenunents
under section W of the Robert T. Stafford Dis-
aster Assistance and Emergency Relief Act. at
amended: Provtf»d further. That any unused
portion of the drrart loan ttmitatton ikaO be
dtxxUaote until Setxaxber x, 1S93: Provided fur-
ther. That the entste amount if designated by
Conpteu as en rm£*t*ncy reqfidrenent pursuant
(0 Uf&on ZSlttMHDM of th* Batanaid Budget
and Smeraencv Deftat Control Act of 13SS. at
amended.".
Resolved. That th* Boos* roced* from Its
dlsA(Te«m*at to the amendresmt of th*
au aambered M7 to th* aforstald bill.
concur th*r*la with aa aanecdment as fol-
lows:
In lies of the ana-proposed by said ameod-
mont, laaerb
tlfO.40S.000: P'ovtded further. That up to
11.000.000 ef the funat appropriate* under Mis
heading may be transferred ta and merged vith
turns appropriated for "Office of Inspector Gen-
eral"
Renlftd. That tb* HOBS* rtieede from Its
dlta«r»*ra*nt to t&a amendment of th* Sen-
ate Dumber*! Stfl to th* afomeald bill, aad
concur thereta with aa amendment aa fol-
low*-.
la lisa of the sure proposed by said amend-
ment, lawn: til3.2<3,000
Rfio'.vtd. That us House r««ode from its
dlMcrMment to the amendment of the Sen-
ate numbered 364 to the afoitioaid hill, and
concur therein wlta an am«ndmeat as fol-

Rei>tor* th* matter stricken by said
amendment, amended to read as follows:
CONGRESSIONAL RECORD—SENATE
S IS 103
(I) Hz
for the Office of the Diree-
requtrwneatK. with wucUl emnbaios oa
Couiuei.
ArMhwd. Th»t tb« Hou*o rtoed« tron Ita
dtca«r««m«Dt to th« aoMndoMat of Cb* S«n-
tt* nomb«r*d ZSfl so tb« tforuald bill, and
concur tJi«rela *1t& an ara«adm«a£ a* fol-
lows:
In llwi of tfce matter propood by a«Id
amendmsBt, tMtrt:
Wortoitfc«o«<««v «»» o*** proc«o« «/ tW»
or any o6k«r Act uttft rap«t t» any fleeal year.
the Hazardous Material* Branch of the Office ef
Technological Hazard*, and all fundt and ttaff
yean provide* to it by thlt Act. than be (rant-
fcrred from the State and local Program* and
Support Directorate to Ole United State* fire
Adntintttratum vlthtn SO dart of the enactment
of this Act.
Retotued. Tlifct tlM UOOM r*o*d« from la
distutreaoMnt to tho fttimntmtnt of tlM 8«B-
ac« aamb*r*d 3CO to th* *jTar«aald bill, and
concur th«r*la with an im*Ti'fr'*t't u tol-
Iow«:
In U«c of tho matter propo**d by
Th* Director of the federal Smergencs Uan-
agement Agency thai! undertake a reetM o/t*«
aoency't organisational structure and. within
ISO days of enactment of Out Act. tubmtt te the
appropriate commUtem ef the Congress a reor-
ganisation plan which reflectt changing mission
requirements and priorities. The review thaU tm-
clude an anecxMnt of the National Prepared-
nets Directorate and examine potential altar-
natrue* te meet that directorate's principal ob-
jectives while increasing oteraO Offency effl-
ctency.
Resolved. That the House recede from
Its diaacreeraent to the amendment of
the Senate numbered 277 to the afore-
said bill, and concur therein with an
amendment as follows:
la lien of the matter ieserted by said
amendment, inaart:
The Mission Simulator and Training FocWty.
Suildlne Number S. of the National Aeronoutict
and Space Administration, located at the John-.
ten Space Center in Houston. Texas, tt hereafter
named and designated the "Jake Corn Mission
Simulator and Training Facility". Any ref-
erence in a late. rule. map. regulation, docu-
ment, record, or other paper of the United States
to tuch'facUity thall be held to be a reference to
the "Jake Gam Mission Siirailator and Training
Facility".
Retoteed. That the hooes recede from Its
dlaacreareeot to the amudnieat.of th* Sea-
ate namband 903 to the aforesaid bill, and
concur therein with an amendment as fol-
lows;
IB liea of the matter proposed by said
ndroant. lagan:
sec. £91 SArt DMKKPW WATCH ACT IKPL»
(a) Safe Drtnklnr Water Act Raporc—The
Administrator of the Environmental Protec-
tion Afency shall report to tiie Conireea
within nlna months of the date of eaaeaaeat
of this section recommendations ooacerataf
tb* reauthorixation of th* Safe Drinking
Water Act. Bach report shall address—
a) th* adverse health effect* associated
with contaminants ta drtaklnc water aad the
public health aad other benefits that may b*
teallied by removta* each oontamtnaate;
(3) th* process for Idantlfylnt conteml-
naau la drinking, water sad isleetlac coa-
tamlnaatt for control:
(3) schedules for th* derslopment of reffU-
lattons and compUanc* with drlnklnc water
ataadsrdt:
(4) tho floanclal and technical capacity of
drinking water systems to Implement moa-
Itortnr reo.utrament« associated with reera-
lated aad acrerulated contamiaanu aad op-
tions to facilitate implementation of such
drinking viator nttaia u inttall treatment fa-
cilities needed to autire eoraplicnce with drink-
ing voter K&ndardi and options to faciHtate
compliance with tuch itandardt. with ipeclal
emphatit on tmaU cermnmtttet;
(6) the financial a • technical capacity of
Statet to Implement the drtnlctng voter program.
intruding option* for tncreattng funding of
State program* and
ff) innovaUve and alternative method* te fn-
ereaac the financial and technical capacity of
drinking water ntt&nt and the Statet U atture
effective Implementation of tuch Act.
(b) Mo/urajum* AHO Ktfour on RADIO.
mcUDtS Ot DnlMKOta WjkTM.—(l) The Admtn-
ittrator of the Environmental Protection Agency
thaU conduct a risk ataament af radon consid-
ering: (At the riafc of advene human health ef-
fect* associated wttA expoture to tartout path-
aavt of raden: (B) the cottt of controlling or
mitigating ecpemre to raden; and (O tha cottt
for ndan control or mltieation esperiencod by
houteholdi and cemmunsttet. including the
eoett experienced by tmatl cammuniite* at the
remit of tuch regulation. Such a* evaluation
ahatt conttdef tk» rukt pated by the tretttmcnt
or iftf»«* of omit taattei produced by u>ater
treatment. The Sctetae Advitary Board thall ie-
vieie the Ageacy'l study and tubmtt a rec-
ommendation ta the Administrator on itt flnd-
Inat. The Adminlttratar thall report the Admin-
ittrator't flntttnoe and the Science X<*uuo*y
Board miiii»in niTuffiin to the Senate Committee
on Environment and Public Work* and the
Houte Committee on Energy and Commerce. Mot
later than July 31. ISM. the Admtnittfotar thall
pubHth the Adinintttrator't study and rit* at-
Mttment and the Saence Advttory Beard rec-
ommendation.
(2) The Xdmhrfstrator it directed, if additional
ante it r«ndr«H to ettabttth the radon standard.
to seek an ertuntion of the deadline contained
In the fiuUctalfa-lmpMed content decree for pro-
mulgation of the rad«« ttandard to a dose not
later than October I, 3993.
(c) Small Syttem Monitoring Coat Reduc-
tion.—With respect to monitoring rfg/uiremfatt
for organic chaitlcalt. pettiddet. PCBt. or un-
regulated eonttaidnantt promulgated in January
l»l (knovn at the Phate tt rule), the Adminu-
trator or a primacy State may modify tuch re-
euiranentt to provide that any drinking water
tyttem teroing a population of Ittt than UOO
person* shall mot be required to conduct aadl-
ttonal quarterly monitoring for a tpecific oon-
tamtnant or eontandnOMti prior ta October t.
19ti. if monitoring far any one quarter con-
dticud after the date of enactment of thta tub-
tection and prior te October t, 1933 for any tuch
ctmtamtnant or contaminants fail* U detect the
pretence of tuch contaminant or contamnantt
in the water tupplted by the drinking water tyt-
tan.
The PRESIDIO OFFICER. The
clerk will report the amendment.
The l«fflal&tlv« clerk read aa followe:
Resolved. That the Hoes* recede from Its
dia*«re*m*at to the amendment of ta* Sen-
ate numbered M3 to the aforesaid btil. aad
concur therein, with an amendment ta fol-
lowe:
la lieu of the natter proposed by taid
amendment. Insert ": Provided, That th*
CooacU oa Savtroameiital Quality aad Office
of BnrironmcBtal Qaal(«y shall reimbures
other aceaeleji for not lees than oae-half of
th* personnel compenaatioa cosu of Indllrid-
uals detailed to it.".
Mr. METZENBAUM. Mr. President. I
rise on behalf of Senator WIRTB to ad-
dress myself to this amendment, which
I very atronffly support. Senator VIRTU
la unable to be with us at this late
-------
Distribution List
Administrator
Deputy Administrator
Assistant Administrators
Regional Administrators
Office of Policy, Planning and Evaluation
Office of Radiation Programs
Office of Water
DOE
NWTRB
NRC
-------
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
n r
Oft 1QQ9 OFFICE OF THE ADMINISTRATOR
tJU, 191M SCIENCE ADVISORY BOARD
EPA-SAB-EC-LTR-93-010

Hdnorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460

Re: SAB Review of Multimedia Risk and Cost Assessment of Radon in
Drinking Water

Dear Ms. Browner:

The EPA Science Advisory Board (SAB) is pleased to comment on the
multimedia risk of exposure to radon and the cost of mitigation as required by
Public Law 102-389 (the Chafee-Lautenberg Amendments to EPA's FY 1993
Appropriation Bill enacted October 6, 1992). The Chafee-Lautenberg Amendment
states that "The Science Advisory Board shall review the Agency's study and
submit its recommendation to the Administrator on its findings." The study
report made available to the SAB is entitled "Multimedia Risk and Cost
Assessment of Radon in Drinking Water".1 This SAB report on the Agency's
study, prepared by the Chafee-Lautenberg Study Review Committee of the SAB,
complements previous detailed SAB comments transmitted to you on the
uncertainty analysis of radon risks (July 9, 1993) and on costs of mitigation of
risks from radon in water (July 30, 1993).

The issues of major concern in assessing risks of radon exposure and costs
of mitigation may be grouped into four categories: a) population exposure profiles;
b) risk estimation procedures; c) mitigation costs; and d) integration of these for
regulatory decision making. The EPA study considered each of these issues and,
in turn, they have been addressed by the SAB.
*By way of background, the SAB early in 1993 b?xan interactions with EPA, including receipt of background material on
thit study. However, the specific report reviewed by the Cummittee mas not received until July 9, 1993, and thus. Itmtud tune
wot available to review and comment on the report bn-uuse of the July 31, 1993 deadline for submission to Congrttt.
Continuing to the present study report, there has been it steady improvement in the quality of the analyses conducted by EPA.
Qxr
-------
A Ponula1non.jg?T>osur.e profiles

The A^ncy report estimates th?t 81 million people use water originating
from community groundwater supplies with a population-weighted average radon
activity of 246 picocuries per liter of water (pCi/Lwater). The Agency report
estimates that approximately 19 million people are served by water supplies with
radon concentrations in excess of 300 pCi/Lwater, the Maximum Concentration
Level proposed by the Agency. It is the SAB's impression from information
provided by public commenters, that the Agency's estimates of population exposure
to radon in drinking water are rather uncertain and may seriously underestimate
the number of community water systems impacted by the proposed drinking water
standard. This uncertainty in exposure estimates ultimately impacts the costs of
mitigation. There is clearly a need for more information and a better presentation
of available data on the profile of population exposure to radon in drinking water,
including the distribution of radon in drinking water exposures for communities of
varying size.

B. Risk estimation procedures

The risk estimation procedures used by the Agency address both the risks
from radon inhaled in air and ingested in water. The risk estimates from airborne
radon with lung cancer as an endpoint are based on strong epidemiological
evidence from studies of uranium miners, augmented by data on other
underground miners, and supported by data from laboratory animal studies.
However, there continues to be debate about the extrapolated lung cancer risk at
lower levels of exposure. This issue may be clarified during the next several years
when the results of several major epidemiological studies focusing on exposure to
radon in homes become available. However, even though there .is a potential risk
at low levels of exposure to air borne radon, it must be recognized that the
populations available for epidemiological studies are relatively small, the majority
of residential exposures are not particularly high, and the postulated levels of risk
are sufficiently low that epidemiological studies might well be unable to identify
any increase in risk attributed to residential radon exposure if such a risk is
present.
-------
,<. r . S.O . I
aii.ua.uoii IS qUiC£ uiij.eifci.ifc io
drinking water. In this case, there is no direct epidemiological or laboratory
animal evidence of cancer being caused by ingestion of radon in drinLlxig water.
Thus, the approach to estimation of cancer risk from radon in drinking water is
more indirect than for radon in air. In the absence of direct evidence, it is not
possible to exclude the possibility of zero risk from ingested radon.

The indirect risk estimation approach involves several steps. First, the dose
to various tissues has been calculated from models for the distribution of radon in
the body following ingestion of radon. The model calculation is based, in part, on
organ distribution, information from an. unpublished study with radio-xenon (as a
surrogate for radon, since both are noble gases) using human subjects. The
meager data base results in uncertainty in estimating tissue doses from ingested
radon in drinking water. This uncertainty could be reduced through further
research. In the next step, the calculated doses have been used along with organ-
specific risk estimates per unit dose, derived from data on the Japanese atomic
bomb survivors, to calculate cancer risk to various organs. To a large extent, this
involves an extrapolation from the very acute, high dose rate, gamma (low Linear
Energy Transfer) exposure of the Atomic Bomb survivors to a very protracted,
very low dose rate, alpha particle (high Linear Energy Transfer) exposure with
ingested radon. The SAB is of the opinion that the estimates of risk from
ingested radon have additional uncertainty due to possible differences in the
distribution of dose, and resulting effects, from alpha particles from radon and
progeny. However, it should be noted that even at the upper bound of the
uncertainly analysis for ingested radon, for most situations the risk from radon
ingested in drinking water is still much lower than the risk from airborne radon
entering the house directly from the soil. Indeed, for many homes the risk from
the radon in water is even lower than that from radon in the outdoor air.

The available information on exposure and risk have been generally
integrated under a scientifically satisfactory framework by the Agency as evidenced
in the Agency's multimedia risk assessment for radon (EPA-SAB-RAC-93-014, July,
1993). However, the uncertainties noted earlier in this report are carried forward
into most of the integrated analyses. However, the differences of opinion,
especially with regard to the extent of the exposed population, with interested
parties are not reflected in the Agency report or in the integrated analyses.
-------
risk p"-*-im*tefl are illustrated in Figures 1 and 2. The population risk
estimates for airborne radon indoors are cne mosc cercaa*, v»x^ v«.i i»t,i^i;^
estimate of 13,600 lung cancer deaths per year (range of 6740 to 30,600 lung
cancer deaths) from exposure to indoor air . Less than one percent of this lung
cancer risk is attributable to radon reaching homes via water. In contrast,
exposure to ratdon in outdoor air is estimated to produce 520 lung cancer deaths
per year (range of 280 to 1500 lung cancer deaths)3. And finally it is estimated
that ingestion of radon in water is estimated to cause 46 cancers per year (range
of 11 to 212 cancers per year)4. This latter estimate is the most uncertain of all
the estimates made. Airborne radon arising from water is estimated to result in
113 lung cancers per year (range of 40 to 408 lung cancers per year)5 which are
included in the estimate presented above for indoor residential air. These risk
estimates for radon can be placed in perspective by comparison with an estimate
of approximately 30,000 cancer deaths per year from all exposures to naturally
occurring radiation, including approximately 13,600 deaths from inhaled radon and
approximately 2,500 cancers estimated for naturally occurring radio-potassium in
the human body.

C. Mitigation costs

The costs of mitigation of radon in the water and indoor air are also
uncertain. Part of the uncertainty for mitigation costs of radon in water relates to
differences of opinion between the Agency staff and interested parties over the
cost of mitigation systems. For example, the Agency staff estimates capital costs
for mitigation of radon in water at less than $2 billion, while interested parties
have estimates of capital costs in excess of $10 billion. Similar differences exist
for recurring maintenance and operating costs. The other part of the uncertainty
for mitigation costs of radon in water relates to the representativeness of the data
base on the occurrence of radon in groundwater used by the Agency. These data,
2
Report to (&• United State* Congreaa on RadionuelidM in Drinking Water: Multimedia Rbk and Ccwt Auecament
of R«don in Drinkinf Water. Praparad for PL 102-389. Offiai of Water. US Environmental Protection Agamy. July 9. 1993. pagd

3Ibid, p. 3-3.
A
Ibid, Table 7-3 beta model estimate*.
e
Tabfc 7-3.
-------
are the source for estimates of the number and size of communities that would
require radon mitigation depending on the level of the MCL finally selected for
regulation. In contrast to the potential mandated regulation of radon in water,
mitigation of radon in indoor air involves voluntary actions by homeowners. Total
cost estimates of the latter are highly uncertain because the extent and cost of
testing for radon in homes and the extent of voluntary participation in mitigation
action in affected homes are unknown.

The SAB is of the opinion that the mitigation cost uncertainties for radon
in drinking water could be reduced by the EPA working with interested parties to
resolve issues related to the occurrence of radon in community systems of various
sizes, the cost of the various process treatment operations and processes for
various system sizes, and the frequency of the need for disinfection after aeration.
This may require reopening the comment period for this rulemaking. The SAB
recommends that EPA, if necessary, request from the Court and Congress
sufficient time to do this work to reduce uncertainties in the cost estimates and
the cost per cancer avoided. The public interest will be served if the Agency
carries out activities over several years which provide a better basis for deciding
how to most effectively mitigate risks from radon exposure in drinking water.

D. Integration for regulatory decision-making

Because of uncertainties in both risk estimates and costs of mitigation there
is substantial uncertainty in ihe cost per cancer death avoided. This uncertainty
is especially large for mitigation of cancers related to ingestion of water. However,
even with this uncertainty, it is clear that the cost per lung cancer avoided from
mitigation of indoor air radon is substantially less than the cost per cancer death
avoided due to mitigation of exposure from radon in drinking water. This
difference appears to be at least a factor of 4 ($3.2 million per cancer death
related to drinking water and $0.7 million per cancer death related to airborne
radon) and may be substantially larger. The highest costs may be those associated
with mitigation of risks for radon in water for the smallest communities.

In summary, the SAB notes the extent of the uncertainties in the population
exposure profiles, the risk estimates for ingested radon in drinking water and the
costs of mitigation. In view of these large uncertainties for risk estimates for
-------
ingested rauoui m urijikiiig wduoi' «»vi. Li:,.'"1! .T*.^ •>!" ''V< "'j'brt.TTTtf.sHy gr^ate* risk*
associated witli airborne radon indoors and outdoors directly from soil, the SAB
advises that EPA consider various options for mitigating radon <;»i'-?r risks. The
options all include continuing the Agency's efforts to encourage voluntary actions
to reduce indoor air radon in view of the cost effectiveness of this approach for
reducing risks.

With regard to water, as one option the Agency could promulgate a
standard at 300 pCi/Lwater as has been proposed. However, in doing so it must be
recognized that this involves selecting a risk reduction strategy for radon that is
the most costly in terms of costs per cancer death avoided; i.e., more than four
times the cost of cancer risk avoidance for airborne radon indoors. Alternatively,
as another option a standard might be set at some higher level such as 1000 to
3000 pCi/L^^, to initiate mitigation of the highest potential risks. For example,
setting a water standard at 3000 pCi/Lwater would result in water contributing no
more radon to indoor air than is present in outdoor air. (Keep in mind that the
radon in outdoor air arises by natural processes from soil gas and there is no way
to alter the outdoor radon levels.) At the same time it would be appropriate to
intensify research on radon ingestion and radon mitigation, data gathering on
radon occurrence for all media, and dialogue with interested parties. These
actions would serve to reduce the uncertainties in the risk estimates, the costs of
mitigation, and, ultimately, the estimates of cost per cancer avoided. We cannot
emphasize too strongly the SAB view that a relative risk orientation should be
applied to the decision making process. Comparative analysis of uncertainties on
the risks of various exposure scenarios and mitigation approaches should be
developed and provided to the risk managers.

The SAB strongly supports the use of a relative risk reduction orientation
as an important consideration in making risk reduction decisions on all sources of
risk, including those attributable to radon. Other important considerations include
legislative authorities, environmental equity, economics, and the like. In short, the
relative risk approach calls for giving the highest priority to mitigating the largest
sources of risks first* especially when the cost-effectiveness of risk reduction of
such sources is high. The SAB recognizes that the large number of laws under
which EPA operates makes it difficult to implement a relative risk reduction
strategy uniformly across the Agency. Radon is an excellent example of the
-------
with rador« irs ^.rMWflsr water governed under one statute (Safe Drinking
Water Act) while radon in indoor air is not currently suoject to regulation under a
specific statute. The SAB strongly encourages the Agency and the Congress to
work together to consider changes in existing statutes that would permit
implementation of relative risk reduction strategies in a more efficient and
effective manner.
The SAB appreciates this opportunity to advise you and the Congress on
this important matter, and we look forward to receiving a response on these
suggestions.
Sincerel
Dr. Raymond C. Loehr
Chair, Executive Committee
Science Advisory Board
toger (7. McCle
Chair, Chafee-Lautenberg Study
Review Committee
-------
-------
I!
>
If
s

I I

C«5
-------
Figure2. Bst/mareoAnnualue&u/6 rrcxr, L*
to Radon (in Cancer Deaths/Year)
Estimated Cancer Deaths/Year
35,000
30,000
25,000
20,000
15,000
10.000
5.000
High
Median
Low
13,600
520
Ingested DW Inhaled Outdoor Air Inhaled Indoor Air

Sources of Exposure

Source: 'Report to the U.S. Congress on Radtonucltdos In Drinking Water:
Multimedia Risk and Cost Assessment of Radon In Drinking Water*,
Office of Water, US EPA Jufy 9, 1993.
-------
-
SCIENCE ADVISORY BOARD
CHAFEE-LAUTENBERG STUDY REVIEW COMMITTEE
CHAIRMAN

Dr. Roger O. McClellan
President
Chemical Industry Institute of Toxicology
P.O. Box 12137
Research Triangle Park, NC 27709

MEMBERS

Mr. Richard Conway
Senior Corporate Fellow
Union Carbide Corporation 770/341
P.O. Box 8361
South Charleston, WV 25303-0361

Dr. Morton Lippmann
Professor
Institute of Environmental Medicine
New York University
Long Meadow Road
Tuxedo, NY 10987

Dr. Genevieve M. Matanoski
Professor of Epidemiology
School of Hygiene and Public Health
The Johns Hopkins University
601 Wolff Street, Room 6019
Baltimore, MD 21205

Dr. Verne Ray
Senior Technical Advisor
Medical Research Laboratory
Pfizer Inc.
Groton, CT 06340
-------
DESIGNATED FEDERAL OFFICIAL
L>r. EiU'waru o. iyenaer
Environmental Protection Agency
Science Advisory Board
401 M Street, S.W., A-101
Washington, DC 20460

STAFF SECRETARY
Mrs. Marcia K. Jolly
Environmental Protection Agency
Science Advisory Board
401 M Street, S.W., A-101
Washington, DC 20460
-------
Environmental
ProtoctjcnAgency
Board (A-lui)
EPA-SA8-DWC-93-015
July 19*"'
EPA AN SAB REPORT:
REVIEW OF ISSUES
RELATED TO THE COST
OF MITIGATING INDOOR
RADON RESULTING
FROM DRINKING
WATER
REVIEW OF THE OFFICE OF
GROUNDWATER AND DRINKING
WATER APPROACH TO THE COSTS
OF RADON CONTROL OR
MITIGATION EXPERIENCED BY
HOUSEHOLDS OR COMMUNITIES
-------
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
, u.o.
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
July 29, 1993

EPA-SAB-DWC-93-015

Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street SW
Washington, DC 20460

Subject: Review of issues related to the cost of mitigating indoor radon
resulting from drinking water.
Dear Ms. Browner:

The Science Advisory Board (SAB) has completed its review of the Agency's
approach to ascertaining the costs of radon control or mitigation experienced by
households or communities in response to Public Law 102-398, Section 519 (106
STAT 1618) pertaining to implementation of the Safe Drinking Water Act
(SDWA). This report is part of a larger study by the SAB of regulating drinking
water radon levels, cost, uncertainty of risk, and overarching issues.

On February 8 and 9, 1993, the Radon Engineering Cost Subcommittee
(REGS) of the SAB's Drinking Water Committee (DWC) conducted a review
focused on the following charge: to determine whether EPA offices are employing
a reasonable approach for estimating the costs of mitigating indoor radon from
drinking water in residences, and whether the technologies that have been judged
by EPA as being Best Available Technology (BAT) for central or well-head
treatment for each size water treatment-facility category are appropriate, and
whether the design, operation, installation and maintenance of these technologies
are reasonably estimated. Additionally, the SAB was asked to address the relative
cost-effectiveness of controlling radon exposure from drinking water in comparison
to controlling other sources of indoor radon. "Effective," in this context, means
the extent to which radon exposure is reduced by the treatment applied to produce
significant reductions in adverse health effects. These results can be normalized
Recyeted/Racyclable
Printed on pap«r thai contains
at toast 75% racydod fibor
-------
i lie ' imcungs ana conclusions ^. Cue
Subcommittee follow.
1. Exposure Issues

a) The Subcommittee determined that the EPA offices are employing a
reasonable framework for estimating the cost-effectiveness of
mitigating airborne indoor radon from soil and water sources in
residences. The cost factors for testing and mitigation of soil gases
are based on a substantial body of data from actual practice and
represent the consensus of a group of industry experts.

b) Based on one national sampling survey, the average concentration of
radon in potentially regulated U.S. water supplies at point of use (not
well head) is approximately 300 pCi/Lwater in groundwater systems
(100 pCi/L when considering a population-weighted average of ground
and surface water systems); certain state and regional survey data
were not included. EPA estimates that a 300 pCi/Lwater standard
would reduce total risk from radon by approximately 2.5%. However,
assuming an equilibrium ratio of 10,000 to 1, water to household air,
the average contribution to airborne radon from waterborne radon is
estimated to be 0.01 pCi/Lwater. This contrasts with an average
indoor airborne radon concentration of between 1 and 1.5 pCi/L^ for
all sources of airborne radon. Regulation of waterborne radon then
will reduce the total airborne radon risk (all sources of radon
considered) by less than 1%. This contribution to the total reduction
of risk (1%) is lessened by the fact that a regulatory limit on
waterborne radon would reduce, not remove radon from all water
supplies. Current estimates are that a regulatory limit of 300
nCL/L would reduce the average U.S. concentration of waterborne
" ^ water
radon to approximately 50% of the present value, indicating that
regulation of waterborne radon at 300 pCi/Lwater would reduce the
total risk of airborne radon by less than 0.5% from the currently
existing risk. By whatever route one arrives at the calculation of
total risk from radon, that is whether it is 0.5% or 2.5%, it most
assuredly is a small risk level compared to soil gas radon.
-------
c) The wide discrepancy between the cost-effectiveness or mitigating
water-borne radon versus soil gas radon underscores the minor role
that waterborne radon plays in the overall indoor health hazard. The
EPA estimates that approximately 80 deaths (range 81-89) could be
avoided per year by reducing all groundwater-based public systems to
300 pCi/Lwater with the maximum individual lifetime risk of fatal
cancer reduced to 2 x 10" .

The most recent cost estimates are about $400M per year, or about
$5M per life saved. On the other hand, the primary source of radon
in indoor air is soil gas which produces an ambient outdoor air
concentration of about 0.4 pCi/L^, and an average indoor
concentration of about 1.3 pCi/L^. EPA estimates that if all homes
with concentrations above 4 pCi/L^ were mitigated with present
technology, then about 3,000 of the 13,600 yearly deaths (range 6740
to 30,600 lung cancer deaths) attributed to indoor radon could be
eliminated, and under this scenario, the cost per life saved would be
about $700,000 and the maximum individual lifetime lung cancer risk
o
reduced to 10" .
2. Cost and Engineering Issues:

a) The Office of Groundwater and Drinking Water (OGW&DW) has
approached the development of the unit costs for the removal of
radon from drinking water by the Packed Tower Aeration (PTA)
method using a reasonable framework. Problems do arise in
calculating the total unit costs, however, because of the assumptions
made on the individual items that make up the total unit costs.
Other water treatment authorities have made their own estimates,
using nearly the same approach as OGW&DW, and have estimated
different total costs.

b) With regard to consideration of alternative aeration technologies (that
is, "engineered" versus modular systems) with systems of different
sizes: EPA's estimates are based on the use of a PTA for all system
sizes. Operation and Maintenance (O&M) costs are also based on a
-------
uniform approach for all sizes. Actual costs and practice will vary
with the size of the system installing the treatment. Very small
systems are likely to experiment with a variety of their own informal
designs as well as a variety of packaged systems and their style of
operation and interactions with the public and with regulatory
agencies are likely to be more informal as well. Larger systems are
likely to impose a more formal design, bid, and construction practice
and engender closer regulatory review and greater public input. If
EPA's purpose is to produce an estimate reflecting the most likely
cost, then these estimates should better reflect the impact of system
size on design practice.

c) Certainly PTA is an effective technique for removing radon from
groundwater and qualifies as Best Available Treatment (BAT) for
central treatment. However, there may also be a perceived problem
in using PTA in certain localities because of off-gas dispersal.
Granular Activated Carbon (GAG) was also discussed as a possible
BAT. EPA cited long contact times required and difficulties in
disposing of waste GAC as reasons for rejecting this technology. Yet,
it seems that GAC has been demonstrated to remove radon, and that
problems of waste disposal may be manageable where influent radon
levels are modest. Additionally, GAG may be a particularly well
suited technology for the smallest systems, since it could be installed
as an in-line pressure vessel not requiring repumping.

d) The cost of disinfection resulting from radon PTA treatment is a
significant factor in the cost of radon mitigation and should be
explicitly stated for different size systems. Groundwater can be
distributed without disinfection only if the system has appropriate
barriers to contamination by micro-organisms. Also, the cancer risks
associated with exposures to disinfection by-products were not
discussed.
-------
3. Recommendations:

a) We are pleased that the OGW&DW has recalculated their unit costs
for Packed Tower Aeration (PTA) in response to the comments
already received and recommend that they continue this iterative
process with the commenters and work cooperatively with other
responsible interested parties. We consider this necessary because we
find merit in some of the non-EPA data.

b) We recommend that EPA review its choices for, BAT and more
carefully state the reasons for choices made reducing the cost of the
GAG process.

c) Since EPA's purpose is to produce aii estimate reflecting the most
likely cost of units, these estimates should more accurately reflect the
impact of system size on design practice.

d) The Subcommittee suggests that summary tables be included in the
report that compare and contrast the impact of several levels of radon
exposure (e.g. 300 pCi/Lwater versus 1000 pCi/Lwater and 3000
pCi/Lwater) on system and national costs including cancer deaths
avoided at various confidence levels. This would be most helpful to
highlight the impact of various remediation efforts to members of
Congress, the states, various water treatment authorities and the
interested public.

e) The EPA analysis shows that mitigating radon from water as
required by the SDWA, is 10 times more expensive than mitigating
radon from soil gas. This regulatory requirement (policy) however,
should not negate logical and practical considerations related to
determining U.S. cost burdens, compared and contrasted to potential
health benefits.

f) One important part of the OGW&DW'S cost calculations on which the
SAB does want to comment specifically is that of the Interest rate
-------
assumptions used, interest rate assumptions markedly impact the
annualized capital costs for radon removal from drinking water. The
operation and maintenance (O&M) costs are insensitive to interest
rates. The SAB recommends that an Interest rate higher than the
3% currently employed by the Agency be used.

g) The cost of disinfection resulting from radon treatment apparently
has not been explicitly itemized In the cost of radon control and the
SAB recommends that this oversight be corrected.

h) The Subcommittee was provided with two thoughtful and detailed
analyses by the American Water Works Association (AWWA) and the
Association of California Water Agencies (ACWA). This commentary
was appreciated by the Subcommittee and provided insights and a
greater diversity of opinion that was useful in our deliberations and
should be considered by EPA in their reevaluation of the radon
issues.
i) The SAB recommends that the OGW&DW participate in the
upcoming "Radon Removal by Packed Tower Stripping" research
project of the AWWA so that they can have their impact on project
design and data collection.

j) Finally, the SAB realizes that it has recommended considerable work
to be done to make EPA's cost studies more creditable and therefore
recommends that the EPA, if necessary, request from the Courts and
the Congress sufficient time to do the work.

Subsequent to the February meeting of the Subcommittee and prior to the
publication of this SAB report, the OGW&DW provided revised cost estimates to
the Subcommittee. These estimates were not available at the time of the public
meeting, and have not been given the usual public scrutiny and discussion that is
such an integral part of all SAB meetings. Therefore, we have not addressed
them in this report. However, we do recognize that the issues contained in the
revised estimates are of great interest and warrant further public and SAB
interaction in the future.
-------
The SAB has offered a number of broad-rangir 3, as wen o» specu'ic
and recommendations on the Agency's radon engineering cost and treatment
technology issues. We are pleased to have had the opportunity to be of service to
the Agency. We trust that these comments will help in your guidance of this
important program, and look forward to your response.
Sincerely,
Dr. Itaymond C. Loehr, Chair
Executive Committee
Science Advisory Board
Dr. Verne A. Ray, Chain
Drinking Water Committee
Science Advisory Board
-------
-------
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 a balanced,
expert assessment of scientific matters related to problems facing the Agency.
This report has not been reviewed for approval by the Agency; hence, the
comments of this report do not necessarily represent the views and policies of the
Environmental Protection Agency or of other federal agencies. Any mention of
trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
ABSTRACT

The Radon Engineering Cost Subcommittee (REGS) of the Drinking Water
Committee (DWC) of the EPA Science Advisory Board (SAB) has reviewed the
Agency's approach to the costs of radon control or mitigation experienced by
households or communities. On February 8 and 9, 1993, the Radon Engineering
Cost Subcommittee (REGS) of the SAB's Drinking Water Committee (DWC)
conducted a focused review of the cost issues.

As part of its charge REGS evaluated EPA's approach for estimating the
costs of mitigating indoor radon from drinking water in residences, assessed EPA's
judgement on Best Available Technology (BAT) for central or well-head treatment
for each size water treatment-facility category are appropriate, and reviewed cost
estimates for design, operation, installation and maintenance of these technologies.
The SAB also compared the cost-effectiveness of controlling radon exposure from
drinking water with the costs of controlling other sources of indoor radon.
"Effective," in this context, means the extent to which radon exposure is reduced
by the treatment applied to produce significant improvements in health. These
results can be normalized using calculated dose-effect values.

The Subcommittee determined that the EPA offices are employing a
reasonable framework for estimating the cost-effectiveness of mitigating airborne
indoor radon in residences. The approach for soil gases embodies standard Agency
and industry methodology, and the cost data for testing and mitigation are based
on a substantial body of data from actual practice and represent the consensus of
industry experts.

The Subcommittee recommends that EPA invite more direct interaction with
various water industry commenters regarding radon removal from drinking water
in order to obtain better data on actual construction, operation, and cost
estimating practice before making its independent judgements. Of particular
concern were the representativeness of the data base on occurrence of radon in
groundwater, the elements used to calculate costs of treatment unit operations, the
effect of system size on unit costs, and the incidence and cost of disinfection after
air stripping.

Key Words: Radon, Radon Engineering Cost, Radon Treatment
-------
Science Advisory Board
Radon Engineering Cost Subcommittee
Drinking Water Committee

(Thair;
Dr. Verne A. Ray, Medical Research Laboratory, Pfizer, Inc.; Groton, Connecticut

Members and
Dr. Judy A. Bean, Universiiy of Miami, Department of Epidemiology, Miami,
Florida

Mr. Keith E. Cams, Carns, Perkins, Associates, Pinole, California

Mr. Richard A. Conway, Union Carbide Corporation, South Charleston, West
Virginia

Dr. Ben B. Ewing, Lummi Island, Washington

Dr. James H. Johnson, Department of Civil Engineering, Howard University,
Washington, DC

Mr. David W. Saum, Infiltec, Inc., Falls Church, Virginia

Dr. James M. Symons, Department of Civil and Environmental Engineering,
University of Houston, Houston, Texas

Dr. Vernon L. Snoeyink, Department of Civil Engineering, University of Illinois,
Urbana, Illinois

Dr. Rhodes Trussell, James M. Montgomery Consulting Engineers, Inc., Pasadena,
California

Dr. James E. Watson, Department of Environmental Sciences and Engineering,
University of North Carolina, Chapel Hill, North Carolina

Invited
Dr. Douglas Crawford Brown, University of North Carolina, Department of
Environmental Sciences and Engineering, Chapel Hill, North Carolina

Science Advisory Board Staff*

Dr. K. Jack Kooyoomjian, Designated Federal Official, U.S. EPA, Science Advisory
Board (A-101F), 401 M Street, SW, Washington, DC 20460

Mrs. Diana L. Pozun, Staff Secretary, U.S. EPA, Science Advisory Board
111
-------
TABLE OF CONTENTS

1. EXECUTIVE SUMMARY 1
1.1 Overview 1
1.2 Occurrence and Risk Estimates 2
1.3 Reasonableness of Cost Estimates For Mitigating Radon 3
1.4 The Technologies for Central or Well-Head Treatment and
Judgements on Best Available Technology 4
1.5 The Cost Estimates of Design, Operation Installation and
Maintenance of These Technologies for Each Size Range 5

2. INTRODUCTION 7

3. REGULATORY RATIONALE 8

4. OCCURRENCE AND RISK ESTIMATES 10

5. RESPONSES TO THE CHARGE 13
5.1 Response to Charge Question 1 13
5.2 Response to Charge Question 2 14
5.2.1 BAT Judgements '. '. 15
5.2.2 Appropriate Technologies For Each Size Range 15
5.3 Response to Charge Question 3 18

APPENDIX A REVIEW, BRIEFING AND BACKGROUND MATERIALS ... A-l

APPENDDC B - LITERATURE CITED B-l

APPENDED C - COST ESTIMATES AND UNCERTAINTY MEASURES .... C-l

APPENDDC D - GLOSSARY OF TERMS AND ACRONYMS . D-l
IV
-------
1. EXECUTIVE SUMMARY
This report presents the Science Advisory Board's (SAB) review of the
Agency's approach to the costs of radon control or mitigation experienced by
households or communities. Our findings and recommendations are aimed at
improving the Agency's overall approach.

1.1 Overview
Radon in drinking water is a two-fold concern in environmental health. It
is a drinking water contaminant which can impact health through the ingestion
route as can many other contaminants regulated under the 1988 SDWA. It also
can contribute to indoor air radon concentrations which expose occupants through
inhalation. Both concerns need to be addressed by the Agency, but it should not
ignore the issue of whether a regulatory focus on waterborne radon will
significantly reduce the health risk posed by radon in comparison with other viable
approaches.

It is recognized that current statutes mandate that EPA regulate radon in
drinking water to reduce exposure to radon in homes, even though the
contribution of drinking water to indoor air radon concentration is quite small
compared with radon from soil emission. But it is also recognized that radon
from water may yield potentially greater health impacts through the combined
inhalation and ingestion routes than other water contaminants which are
regulated by EPA.

The primary source of radon in indoor air is soil gas which produces an
ambient outdoor air concentration of about 0.4 pCi/L^, and an average indoor
concentration of about 1.3 pCi/L^. EPA estimates that if all homes with
concentrations above 4 pCi/L^ were mitigated with present technology, then about
3,000 of the 13,500 yearly deaths attributed to indoor radon could be eliminated.
Under this scenario, the cost per life saved would be about $700,000.

The contribution of waterborne indoor radon is much smaller, and it is
estimated by EPA that there is a ratio of about 10,000 to 1 (one) between the
-------
water concentration and the increase in the indoor air concentration, with typical
household water use. Therefore, 300 pCi/L in water contributes approximately
0.03 pCi/Lgjj, to the indoor air concentration. The EPA estimates that from 81 to
89 deaths (depending on the model) could be avoided per year by reducing all
ground-based public water systems to 300 pCi/Lwater> The most recent cost
estimates are about $400M ($400,000,000) per year, or about $3.2M per life saved.

This wide discrepancy between the cost-effectiveness of mitigating water-
borne radon versus soil gas radon underscores the minor role that waterborne
radon plays in the overall indoor health hazard. Still, its regulation is required
under the Safe Drinking Water Act (SDWA).

The question addressed is: To what degree will regulation of radon in water
bring about a reduction in exposure, and risk, to airborne radon in homes, and has
the U.S. EPA shown that a focus on waterborne radon is reasonable and cost-
effective in light of this goal? The inclusion of non-inhalation pathways of
exposure, especially direct ingestion, does not significantly alter the conclusions
here. (NOTE: Estimates of exposure from direct ingestion were included in EPA's
analysis.)
1.2 Occurrence and Risk Estimates

The occurrence data employed by the U.S. EPA in estimating exposures to
airborne radon are both the best available and a reasonable basis for making such
estimates. The affected population was determined based on a properly random
national survey of indoor radon in U.S. homes. The measurement technique
employed was alpha track detectors placed into homes for an entire year. While
new data continue to be produced, it is unlikely that these data will significantly
change the existing estimates of the distribution of airborne radon concentrations
in U.S. homes.
The contribution of waterborne radon to indoor air concentrations is less
well established. Current estimates are that a regulatory limit of 300 pCi/Lwater
would reduce the average U.S. concentration of waterborne radon to approximately
50% of the present value, indicating that regulation of waterborne radon at 300
-------
would reduce the total risk of airborne radon by less than 0.5% tiom
the currently existing risk.

It should be noted that homes with very high concentrations of waterborne
radon may contain a higher contribution of airborne radon from water. A
waterborne radon concentration of 10,000 pCi/L^^ would yield an average
contribution of 1 pCi/L^, which is significant relative to the national average.
Homes utilizing water with high radon concentrations, however tend also to have
high airborne concentrations from subsoil sources. The U.S. EPA should develop
estimates of ..the distribution of airborne radon contributions from waterborne
radon both in the presence and in the absence of potential regulatory limits on
waterborne radon.
The airborne risk estimates used by EPA are based on recommendations of
the National Academy of Sciences's (NAS) Committee on the Biological Effects of
Ionizing Radiation (BEIR). These estimates are extrapolated from data obtained
from uranium miners, and a modifying factor has been added to account for
differences between exposure conditions and physiological properties of individuals
in mines and homes. Both the BEIR Committee and EPA have noted the
uncertainty in these estimates. The SAB's Radiation Advisory Committee (RAC)
has reviewed this in the past and concurred with the EPA's risk estimates. These
uncertainties in the risk estimates are reflected in the EPA's range of values for
the cost per life saved.

1.3 Reasonableness of Cost Estimates For Mitigating Radon

The Subcommittee determined that the EPA offices are employing a
reasonable framework for estimating the cost and cost-effectiveness of mitigating
airborne indoor radon in residences. The approach embodies standard Agency and
industry methodology, focuses on existing homes, considers inhalation as the only
exposure pathway and does not differentiate by radon source (i.e., soil versus
water). This approach includes the determination of the affected population,
determination of the cost for testing and mitigation, analysis of the risk reduction
from mitigation and calculation of the cost per life saved. The cost data for
testing of air and mitigation of subsoil sources are based on a substantial body of
-------
data from actual practice and represent the consensus of a group of industry
experts.

In summary the Subcommittee considers the approach to be reasonable, and
the availability of the data from a national survey of indoor radon and actual cost
data strengthen the final result. The Subcommittee considers EPA's calculations
of cost per life saved to be based upon reasonable occurrence estimates, risk
estimates and cost estimates.

The Office of Groundwater and Drinking Water (OGW&DW) has
approached the development of the unit costs for the removal of radon from
drinking water by Packed Tower Stripping (PTS) in a reasonable manner.
Problems do arise in calculating the total unit costs, however, because of the
assumptions made on the individual items that make up the total unit costs.
Other water treatment authorities have made their own estimates, using nearly
the same approach as OGW&DW, and have estimated different total costs. The
SAB does not wish to comment on which is the "correct" assumption for each
component of the total, but does recommend that OGW&DW meet with these
other groups and their consultants to understand and resolve these differences.
The impact of significant differences can be severe in terms of national costs to
implement a radon rule.

1.4 The Technologies for Central or Well-Head Treatment and Judgements on
Best Available Technology

Certainty aeration is an effective technique for removing radon from
groundwater and qualifies as BAT for central treatment. However, there may be a
piecemeal problem in using Packed Tower Aeration (PTA) in certain localities
because of off-gas dispersal. Granular Activated Carbon (GAG) was also discussed
as a possible BAT. EPA cited long contact times required and difficulties in
disposing of waste GAG as reasons for rejecting this technology. Yet, it seems
that GAC has been demonstrated to remove radon, and that problems of waste
disposal may be manageable where influent radon levels are modest. Additionally,
GAC may be a particularly well-suited technology for small systems, since it could
be installed as an in-line pressure vessel not requiring repumping. We
recommend that EPA review its choices for BAT, more carefully state the reasons
-------
for its choices made and estimate the likely number of systems using GAG and the
attendant costs.
1.5 The Cost Estimates of Design, Operation Installation and Maintenance of
These Technologies for Each Size Range

The basic approach the EPA is taking reflects a standard framework to cost
estimation: compiling and analyzing data on occurrence, determining the likely
technology tp- be used, and estimating the cost of technology implementation as a
function of the water quality and system size. However, there are three concerns
that the SAB has about these estimates: a) the basic objective of the cost
estimation process, b) the consideration of the relationship of system size, style
of design and operation, and c) whether the costs are appropriately estimated?

It is unclear whether EPA's purpose is to estimate the costs industry will
most likely incur as a result of the radon regulations or to estimate the lowest
possible cost industry could incur. There was extensive discussion on this point by
the Subcommittee at its review meeting of February 8, 1993. In either case, EPA
would do well to invite more direct interaction with various commenters to obtain
better data on actual construction, operation, and cost estimating practice before
making its independent judgements.

With regard to consideration of alternative aeration technologies at different
sizes: EPA's estimates are based on the use of a PTA for all system sizes.
Operation and Maintenance (O&M) costs are also based on a uniform approach for
all sizes. Actual practice is likely to vary with the size of the system installing the
treatment. Some experience might be gained from the volatile organic carbon
(VOC) rule here. Very small systems are likely to experiment with a variety of
their own informal designs as well as a variety of packaged systems and their
style of operation and interface with the public and with regulatory agencies is
likely to be more informal as well. Larger systems are likely to impose a more
formal design, bid, and construction practice and experience closer regulatory
review and greater public input. If EPA's purpose is to produce an estimate
reflecting the most likely cost, then these estimates should better reflect the
impact of system size on design practice.
-------
Interest rate assumptions markedly impact the annualized capital costs for
radon removal from drinking water. O&M costs are insensitive to interest rates.
Capital improvements for many small systems require interest rates of 10% or
higher. Annual costs for radon removal by PTA were based on a three percent
interest rate. The impact of a 10 percent interest was also evaluated, but the
emphasis was on a three percent interest rate. The SAB recommends that an
interest rate higher than the 3% currently employed by the Agency be used. The
cost of disinfection resulting from radon PTA treatment is a significant factor in
the cost of radon mitigation and should be explicitly stated for different size
systems. Groundwater can be distributed without disinfection only if the system
has appropriate barriers to contamination by micro-organisms. Also, the cancer
risks associated with exposures to disinfection by-products were not discussed.

It would be most helpful to members of Congress, the states, various water
treatment authorities and the interested public to have the Agency's presentation
and summary of data, as well as the Agency's recommendations succinctly
presented in the report to Congress in a few well-planned and clearly labeled
summary tables, histograms or charts. This will serve to focus the many issues
onto the key recommendations of the Agency, and to obtain a summary of the
trade-offs involved with this issue.

Finally, the SAB recommends that the OGW&DW participate in the
upcoming "Radon Removal by Packed Tower Stripping" American Water Works
Association (AWWA) research project so that they can have their input on project
design and data collection. This will make the output of this important study as
useful to OGW&DW as possible.

In summary, the SAB is pleased that the OGW&DW has recalcxilated their
unit costs for PTA in response to the comments already received and the SAB
recommends that they continue this iterative process with the commenters and
work cooperatively with other responsible interested parties.
-------
2. INTRODUCTION
At the request of the Office of Drinking Water (ORD), the Radon
Engineering Cost Subcommittee (REGS) of the Science Advisory Board's (SAB)
Drinking Water Committee (DWC) met on February 8 and 9, 1993 to review
background reports and documents related to the cost of mitigating indoor radon.
(See Appendix A - References 1-6, 10-13, and 15). The Subcommittee was made
up of members of the DWC, the Radiation Advisory Committee (RAC), and the
Environmental Engineering Committee (EEC). Presentations by EPA staff (J.W.
Conlon, F. Marcinowski, M.J. Parrotta, M. Cummins, J.A. Auerbach, and others.
See Appendix A - Reference 11.) were also heard by the Subcommittee.

The statement of charge to the Subcommittee, as accepted by the
Subcommittee, was as follows:

a) To determine whether the EPA is employing a reasonable approach
for estimating the cost of mitigating indoor radon from drinking
water in residences;

b) To assess whether the EPA has made appropriate judgements of Best
Available Technology (BAT) for central of well-head treatment of each
size water treatment facility category, and whether the cost estimates
of design, operation installation and maintenance of these
technologies are accurately estimated; and

c) To address the relative cost-effectiveness of controlling radon
exposure from drinking water in comparison to controlling other
sources of indoor radon.

Each of the three elements of the charge are addressed below.
^'Effective" in this context means the extent to which radon exposure is reduced
by the treatment applied to produce significant improvements in health.
-------
3. REGULATORY RATIONALE
The history of concern over radon in water (See, for instance, Appendix B -
References 1-3, 5, and 7-9) provides the framework within which the following
discussion must be placed. That history is that (a) radon was found to be a source
of risk in mining populations exposed to airborne radon, (b) airborne radon was
found in indoor air of homes, (c) radon was found in water supplies in the U.S.,
and (d) radon was found to emanate from water to air in homes. It was
concluded that waterborne radon might pose a potential risk to human health
through its contribution to the concentration of airborne radon in homes (See, for
instance, Appendix B, Reference 5).

Subsequent risk analyses performed by the U.S. EPA and others indicated
that emanation from water to air is not the only route of exposure to waterborne
radon (See, for instance, Appendix B, References 7-9), direct ingestion also being of
potential significance based on calculation, the historical focus on airborne radon
remains. Even if ingestion exposures are considered, the conclusions of this report
are not altered. This raises the issue of whether a regulatory focus on waterborne
radon will significantly affect the health risk posed by radon in the general
environment. The present section examines the three goals of any potential radon
policy.

The first goal is to reduce the risk from waterborne pollutants. This goal
requires an answer to the question of whether waterborne radon produces a
significant additive risk and whether a focus on waterborne radon, rather than on
other pollutants, is a reasonable means for reaching a significant reduction in
health risk. The second goal might be to reduce the risk from environmental
radon. This requires an answer to the question of whether environmental radon
produces a significant risk and whether a focus on waterborne radon will
reasonably reduce the overall risk posed by environmental radon. The third goal
might be to reduce the overall environmental risk from all sources of risk. The
question to be addressed here is whether a focus on environmental radon will be
the most effective means to reach this goal (See, for instance, Appendix B,
References 4 & 6 pertaining to reducing risk).
8
-------
The following discussion focuses only on the second policy goal. The
question addressed is: To what degree will regulation of radon in water bring
about a reduction in exposure, and risk, to airborne radon in homes, and has the
U.S. EPA shown that a focus on waterborne radon is reasonable and cost-effective
in light of this goal? It is presumed that inclusion of non-inhalation pathways of
exposure, especially direct ingestion, does not significantly alter the conclusions
here.
-------
4. OCCURRENCE AND RISK ESTIMATES
The occurrence data employed by the U.S. EPA in estimating exposures to
airborne radon are both the best available and a reasonable basis for making sucli
estimates. The affected population was determined based on a random national
survey of indoor radon in U.S. homes. The measurement technique employed was
alpha track detectors placed into homes for a year. While new data continue to
be produced, it is unlikely that these data will significantly change the existing
estimates of the distribution of airborne radon concentrations in U.S. homes.

The contribution of waterborne radon to indoor air concentrations is less
well established for two reasons. First, the occurrence data on waterborne radon
continue to be weakened by considerations of sample size and conflicting sets of
data. Second, the equilibrium ratio between airborne radon concentration (as
produced by only waterborne radon) and the waterborne radon concentration is
not well established. At present, the estimate of 1 per 10,000 for this ratio as
adopted by the U.S. EPA is reasonable in light of the existing data but must be
viewed as preliminary. An equilibrium ratio of 1 per 10,000 is assumed in this
discussion.

The average concentration of radon in potentially regulated U.S. water
supplies is estimated by EPA to be approximately 300 pCi/Lwater in groundwater
systems (100 pCi/Lvr&ter when considering a population-weighted average of ground
and surface water systems). EPA estimates that a 300 pCi/Lwater standard would
reduce total risk from radon by approximately 2.5% (See Appendix B - references
11 and 12, as well as memo dated 4/20/93 from Douglas Crawford-Brown, which
includes relevant citations.) Assuming an equilibrium ratio of 10,000 to 1, water
to household air, the average contribution to airborne radon from waterborne
radon is estimated to be 0.01 pCi/L^. This contrasts with an average indoor
airborne radon concentration of between 1 and 1.5 pCi/L^ for all sources of
airborne radon. Regulation of waterborne radon then will reduce the total
airborne radon risk (all sources of radon considered) by less than 1%. By
whatever route one arrives at the calculation of total risk from radon, that is
whether it is 0.5% or 2.5%, it most assuredly is a small risk level compared to soil
gas radon.

10
-------
This contribution to the total reduction of risk (1%) is lessened by the fact
that a regulatory limit on waterborne radon would reduce, not remove radon from
all water supplies. Current estimates are that a regulatory limit of 300 pCi/Lwater
would reduce the average U.S. concentration of waterborne radon to approximately
50% of the present value, indicating that regulation of waterborne radon at 300
pCi/Lwater would reduce the total risk of airborne radon by less than 0.5% from
the currently existing risk.

It should be noted that homes with very high concentrations of waterborne
radon may contain a higher contribution of airborne radon from radon emanated
by water. A waterborne radon concentration of 10,000 pCi/Lwater would yield an
average contribution of 1 pCi/L^, which is significant relative to the national
average. Homes utilizing water with high radon concentrations, however tend also
to have high airborne concentrations from other sources. The U.S. EPA should
develop estimates of the distribution of airborne radon contributions from
waterborne radon both in the presence and in the absence of potential regulatory
limits on waterborne radon.

The risk estimates used by EPA are based on recommendations of the
NAS's Committee on the Biological Effects of Ionizing Radiation (BEIR). These
estimates are extrapolated from data obtained from uranium miners, and a
modifying factor has been added to account for differences between exposure
conditions and physiological properties of individuals in mines and homes. Both
the BEIR Committee and EPA have noted the uncertainty in these estimates. The
SAB's Radiation Advisory Committee (RAG) has reviewed and concurred with the
EPA's risk estimates. These uncertainties are in the risk estimates are reflected
in the EPA's range of values for the cost per life saved.

The Subcommittee notes that two major points related to epidemiology that
were not covered in the meeting. First, the miner risk data is for exposures
considerably higher than the 4 pCi/L^ action level recommended for homes, and
the linear extrapolation to 4 pCi/L^ has been the subject of controversy in the
past. In the case of the very small incremental change in radon levels from water
contributions (0.03 from 300 pCi/Lwater), there is some question as to its effect if
the initial house levels are also very low (e.g., house at 0.5 pCi/L^ and water
contribution of 0.03 pCi/L results in a net 0.53 pCi/L. The linear
11
-------
extrapolation becomes even more questionable at the low levels. Second, the risk
to non-smokers is at least a factor of 10 lower than the risk to smokers.

The Subcommittee determined that the EPA offices are employing a
reasonable framework for estimating the cost and cost-effectiveness of mitigating
airborne indoor radon in residences. The approach embodies standard Agency and
industry methodology, focuses on existing homes, considers inhalation as the only
exposure pathway and does not differentiate by radon source (i.e., soil versus
water). The cost data for testing and mitigation of radon in indoor air are based
on a substantial body of data from actual practice and represent the consensus of
industry experts.

The national costs for testing and mitigation are based on an action level of
4 pCi/Lgjj. and a mitigation reduction level of 2 pCi/L^. (The action level
corresponds to 224 Lung Cancer Deaths (LCDs) per million). The national radon
mitigation coste are based on the summed weighted costs for installation, O&M for
various mitigation methods and foundation types.

In summary the Subcommittee considers EPA's calculations of cost per life
saved appear to be based upon reasonable occurrence estimates, risk estimates and
cost estimates; however, much of these data are in a continuing state of evolution
and refinement.
12
-------
5. RESPONSES TO THE CHARGE
5.1 Response to Charge Question 1

Charge 1: To determine whether EPA is employing a reasonable approach
for estimating the cost of mitigating indoor radon from ambient and
drinking water sources in residences.

The Subcommittee determined that the EPA offices are employing a
reasonable approach for estimating the cost-effectiveness of mitigating airborne
indoor radon in residences. The approach embodies standard Agency and industry
methodology, focuses on existing homes, considers inhalation as the only exposure
pathway and does not differentiate by radon source (i.e., soil versus water) in its
occurrence estimates. This approach includes the determination of the affected
population, determination of the cost for testing and mitigation, analysis of the
risk reduction from mitigation and calculation of the cost per life saved.

The affected population was determined based on a national survey of
indoor radon in U.S. homes. The survey was conducted using alpha track
detectors which were placed in homes for a full year. Cost data for testing and
mitigation are based on a substantial body of data from actual practice. The risk
estimates used by EPA are based on recommendations of the National Academy of
Science's Committee on the Biological Effects of Ionizing Radiation (BEIR). These
estimates are extrapolated from data obtained from uranium miners, and a
modifying factor has been added to account for differences between exposure
conditions and physiological properties of individuals in mines and homes. Both
the BEIR Committee and the EPA have noted the uncertainty of these estimates.
The SAB's Radiation Advisory Committee has reviewed and concurred with the
EPA's risk estimates. In addition to the uncertainty in the risk estimates, there
are other uncertainties that are reflected in the EPA's range of values for the cost
per life saved. In summary, the Subcommittee considers the approach to be
reasonable, and the availability of the data from a national survey of indoor radon
and actual cost data support the Agency's final result.

The contribution of waterborne radon to indoor air concentrations is less
well established for two reasons. First, the occurrence data on waterborne radon

13
-------
continue to be weakened by considerations of sample size and conflicting sets of
data. Second, the equilibrium ratio between airborne radon concentration (as
produced by only waterborne radon) and the waterborne radon concentration is
not well established. At present, the estimate of 1 per 10,000 for this ratio as
adopted by the U.S. EPA is reasonable in light of the existing data but must be
viewed as preliminary. An equilibrium ratio of 1 per 10,000 is assumed in this
discussion.

The average concentration of radon in potentially regulated U.S. water
supplies is approximately 300 pCi/L^^ in groundwater systems (100 pCi/Lwater
when considering a population-weighted average of ground and surface water
systems). EPA estimates that a 300 pCi/Lvrater standard would reduce total risk
from radon by approximately 2.5%. (See Appendix B - references 11 and 12, as
well as memo dated 4/20/93 from Douglas Crawford-Brown, which includes
relevant citations.) Assuming an equilibrium ratio of 10,000 to 1, water to
household air, the average contribution to airborne radon from waterborne radon
is estimated to be 0.01 pCi/L. This contrasts with an average indoor airborne
radon concentration of between 1 and 1.5
air
for all sources of airborne
radon. Regulation of waterborne radon then will reduce the total airborne radon
risk (all sources of radon considered) by less than 1%. By whatever route one
arrives at the calculation of total risk from radon, that is whether it is 0.5% or
2.5%, it most assuredly is a small risk level compared to soil gas radon.

5.2 Response to Charge Question 2

Charge 2: To assess whether the EPA has made appropriate judgements of
Best Available Technology [BAT] for central or well-head treatment of each
size water treatment-facility category, and whether the cost estimates of
design, operation installation and maintenance of these technologies, are
accurately estimated.

The question of BAT will be discussed in two parts: 1) Are EPA's BAT
Judgments appropriate?, and 2) Are appropriate technologies selected for each size
range?
14
-------
5.2.1 BAT Judgements

Certainly aeration is an effective technique for removing radon from
groundwater and qualifies as BAT. Granular Activated Carbon (GAG) was also
discussed as a possible BAT. EPA cited long contact times required and
difficulties in disposing of waste GAC as reasons for rejecting this technology. Yet
is seems that GAC has been demonstrated to remove radon, and that problems of
waste disposal may be manageable where influent radon levels are modest.
Moreover, GAC may be a particularly important technology for small systems,
because the units would be small, regardless of the longer contact time, and more
importantly, can be applied as a pressure vessel not requiring repumping.
Moreover there may also be a problem in using Packed Tower Aeration (PTA) in
certain localities because of off-gas dispersal and the cost of repumping.
Additional community aesthetic concerns deal with the unsightly character of air
towers. We recommend that EPA reconsider its choices for BAT and more
carefully state the reasons for its choices made, as well as estimate the likely
number of systems using GAC and the attendant costs.

5.2.2 Appropriate Technologies For Each Size Range

There are three concerns that the SAB has about these estimates: a) the
basic objective of the cost estimation process, b) the consideration of the
relationship of system size style of design and operation, and c) are the costs
appropriately estimated?

a) Basic objectives: It is unclear whether EPA's purpose is to estimate
the costs industry will most likely incur as a result of the radon
regulations or to estimate the lowest possible cost industry could
incur. There was extensive discussion on this point by the
Subcommittee at its review meeting of February 8, 1993. In either
case, EPA would do well to invite more direct interaction with
various commenters to obtain better data on actual construction,
operation, and cost estimating practice before making its independent
judgements.
15
-------
b) Consideration of alternative aeration technologies at different sizes:
EPA's estimates are based on the use of PTA for all system sizes.
O&M costs are also based on a uniform approach for all sizes. Actual
practice is likely to vary with the size of the system installing the
treatment. Some experience might be gained from the VOC rule here.
Very small systems are likely to experiment with a variety of their
own informal designs as well as a variety of packaged systems and
their style of operation and interface with the public and with
regulatory agencies is likely to be more informal as well. Again, GAG
may be a particularly well-suited technology for small systems, since
it could be installed as an in-line pressure vessel not requiring
repumping. Larger systems are likely to impose a more formal
design, bid, and construction practice and experience closer regulatory
review and greater .public input. Larger systems also often have
wells in residential areas and are required by local planning boards to
design the facility to blend in with the surrounding homes. These
requirements significantly increases costs. If EPA's purpose is to
produce an estimate reflecting the most likely cost, then these
estimates should better reflect the impact of system size on design
practice.

c) Costs Appropriately Estimated: The basic approach the EPA is
taking to cost estimation, that is, compiling and analyzing data on
occurrence, determining the likely technology to be used, and
estimating the cost of technology implementation as a function of the
water quality and system size are appropriate.

d) Additional Considerations: (Refer to Appendix C - Tables of Cost
Estimates and Uncertainty Measures.)

(1) The EPA analysis was of drinking water mitigation, and only
considered the cost per life saved for an aggregation of all sizes of
central water systems, but this data can be used to determine the cost
per life saved for each system size category. When this is done, the
largest systems show a cost of less than $500,000 per life saved, and
the smallest systems show a cost of over $50M per life saved. This
16
-------
disaggregatiori suggests that central system mitigation may not be
economical for the smaller systems. Continuing this line of analysis,
systems with very high radon concentrations might be analyzed
separately, and PTA might be shown to be cost-effective (e.g., a small
system with 30,000 pCi/L^^ rather than the assumed 300
pCi/Lwater would have a cost-effectiveness of about $500,000 per life
saved). The Subcommittee recommends that the EPA use this type of
disaggregated analysis.

(2) If central water system mitigation is not cost-effective for some
systems, then other non-central radon mitigation technologies might
be investigated so that mitigation advice or assistance can be provided
to the public at risk. From the EPA analysis of air and water
mitigation, the effectiveness of installing a standard soil gas radon
mitigation system in-house with water radon problems can be
estimated. Assuming a background of 0.4 pCi/L^, an average house
level of 1.3 pCi/L^, and a mitigation system effectiveness of 50%,
radon levels could be lowered O^SpCi/L^ for a cost of about $185
per year. The cost per life saved is about $1.2M per year. The
Subcommittee recommends that the EPA use this type of analysis to
investigate the cost of standard soil gas mitigation as an alternative
to central system treatment.

(3) In the small central water systems, where central radon
mitigation might not be cost-effective, there may be very simple non-
central system mitigation systems that would provide even more cost-
effective mitigation than the standard Active Sub-Slab
Depressurization (ASD) systems (for soil gas mitigation). For
instance, entry from washing clothes or showering might be mitigated
with exhaust fans, and drinking water might be filtered with a very
small GAG system. The Subcommittee recommends that the EPA
study new types of low cost mitigation alternatives and perform
research if necessary.
17
-------
5.3 Response TO Charge Question 3

Charge 3!: To address the relative cost-effectiveness2 of controlling radon
exposure from drinking water in comparison to controlling other sources of
indoor radon.

Radon in drinking water is a two-fold concern in environmental health. It
is a drinking water contaminant which can impact health through the ingestion
route as can many other contaminants regulated under the 1988 SDWA. It also
can contribute to indoor air radon concentrations which expose occupants through
inhalation. Both concerns need to be addressed by the Agency.

The SAB review of the EPA approach to evaluation of the overall indoor
airborne radon exposure and its mitigation cost-effectiveness has been discussed
above under Charge 1. The cost-effectiveness of mitigation of radon in water in
comparison to controlling other sources of indoor radon is discussed here.

It is recognized that current statutes mandate that EPA regulate radon in
drinking water to reduce exposure to radon in homes, even though the
contribution of drinking water to indoor air concentration is quite small compared
with radon from soil emission. But it is also recognized that radon from water
yields potentially greater health impacts through the combined inhalation and
ingestion routes than do the concentration of many other water contaminants
which are regulated by EPA.

The primary source of radon in indoor air is soil gas which produces an
ambient outdoor ah- concentration of about 0.4 pCi/L^, and an average indoor
concentration of about 1.3 pCi/L^. EPA estimates that if all homes with
concentrations above 4 pCi/L^ were mitigated with present technology, then about
3,000 of the 13,500 yearly deaths attributed to indoor radon could be eliminated.
Under this scenario, the cost per life saved would be about $700k.
2"Effective" in this context means the extent to which radon exposure is reduced
by the treatment applied to produce significant improvements in health.

18
-------
The contribution of waterborne indoor radon is much smaller, and it is
estimated that there is a ratio of about 10,000 to (1) one between the water
concentration and the increase in the indoor air concentration, with typical
household water use. Therefore, 300 pCi/Lwater in water contributes
approximately 0.03 pCi/L^ to the indoor air concentration. The EPA estimates
that approximately 80 deaths could be avoided per year by reducing all ground-
based public water systems to 300 pCi/Lwater The most recent cost estimates are
about $400M per year, or about $3.2M per life saved.

This wide discrepancy between the cost-effectiveness of mitigating
waterborne radon versus soil gas radon underscores the minor role that
waterborne radon plays in the overall indoor health hazard. Still, its regulation is
required under the SDWA. This regulatory policy, however, should not negate the
logic and practical considerations that relate to determining U.S. cost burdens.

The Office of Groundwater and Drinking Water (OGW&DW) has
approached the development of the unit costs for the removal of radon from
drinking water by PTS in a reasonable manner. Problems do arise in calculating
the total unit costs, however, because of the assumptions made on the individual
items that make up the total unit costs. Other thoughtful groups have made their
own estimates, using nearly the same approach as OGW&DW, and have estimated
different total costs.

The SAB does not wish to comment on which is the "correct" assumption
for each component of the total, but does recommend that OGW&DW meet with
these other groups and their consultants to understand these differences. The
result of these meetings probably will be a range of costs, the low end being a
"bare bones" system that smaller systems might install, the high end being a
"engineered" system that a larger system might install. This result would have
two advantages; one, the assumptions supporting the cost for each system would
be clearly delineated and two, OGW&DW would have a better understanding of
the expected range of costs around their estimate of "best", rather than the
assumed 20 percent lower and 30 percent higher that is now being used.

One important part of the OGW&DW's calculation on which the SAB does
want to comment specifically is that of the interest rate assumptions used.

19
-------
Interest rate assumptions markedly impact the annualized capital costs for radon
removal from drinking water. O&M costs are insensitive to interest rates.
Capital improvements for many small systems require interest rates of 10% or
higher. Annual costs for radon removal by Packed Tower Aeration (PTA) were
based on a three percent interest rate. The impact of a 10 percent interest was
also evaluated, but the emphasis was on a three percent interest rate. The SAB
recommends that an Interest rate higher than the 3% currently employed by the
Agency be used.

The cost of disinfection resulting from radon PTA treatment is a
significant factor in radon cost mitigation and should be explicitly stated for
different size systems. Systems that require PTA, which are not now disinfected,
will require disinfection. Groundwater can be distributed without disinfection only
if the system has appropriate barriers to contamination by microorganisms. PTA
introduces the possibility of such contamination, and, thus, disinfection is required.
The cost of such disinfection has not been explicitly itemized in the cost of radon
control, and the SAB recommends that this oversight be corrected. The
Subcommittee understands, based on subsequent discussions with OGW&DW staff,
that the cost of disinfection varies considerably based on system size. For
instance, predominantly small systems have costs for disinfection ranging typically
from $100 to $200/year/household, while large systems may only cost a few dollars
per household per year. It is the Subcommittee's view that costs of disinfection,
especially in small systems, needs to be reviewed thoroughly.

In summary, the SAB is pleased that the OGW&DW has recalculated then*
unit costs for PTA in response to the comments already received and the SAB
recommends that they continue this iterative process with the cornmenters and
works cooperatively with other responsible interested parties.

Finally, the SAB recommends that the OGW&DW participate in the
upcoming "Radon Removal by Packed Tower Stripping" American Water Works
Association research project so that they can have their input on project design
and data collection and consider the new AWWA and ACWA documents in their
future review. This will make the output of this important study as useful to
OGW&DW as possible.
20
-------
Currently, the EPA has considered PTA as the only feasible BAT. However,
the Subcommittee considers that treatment with GAG should be revisited,
especially since it would enable a system to use it as an in-line pressure vessel, not
requiring repumping. It could also eliminate the need for disinfection.
21
-------
-------
APPENDIX A
REVIEW, BRIEFING AND BACKGROUND MATERIALS
1) Cummins, Michael D., Memorandum to Marc. Parrotta, entitled "Removal
from Contaminated Ground Water by Packed Column Air Stripping, " U.S.
EPA, Water Supply Technology Branch, Cincinnati, Ohio, April 26, 1988

2) Cummins, Michael D., Memorandum to Marc Parrotta entitled "Packed
Tower Aeration Cost Estimates for Radon Removal," U.S. EPA, Office of
Ground Water and Drinking Water, Technical Support Division, March 11,
1992

3) Longtin, Jon (Project Engineer, TSD) and Denning, George (Economist,
OPD&E), "Draft for Discussion with QAMS of the Data Quality Objectives
for the National Inorganics and Radionuclides Survey," Technical Support
Division, Office of Drinking Water, Office of Water, U.S. EPA, Cincinnati,
Ohio, March 8, 1985

4) Mills, William R., Stephen K. Hall and Thomas E. Levy, Letter to Carol M.
Browner, Raymond C. Loehr, Genevieve Matanoski, and Verne Ray from the
Alliance for Radon Reduction, February 2, 1993

5) Cummins, Michael D. Memorandum to Marc Parrotta entitled "Simplified
Equations for Estimating Radon Removal Cost via Packed Tower Aeration,"
U.S. EPA, Office of Ground Water and Drinking Water, Technical Support
Division, July 16, 1992

6) Parrotta, Marc, Memorandum to Addressees entitled "Cost Modeling
Update," U.S. EPA, Office of Water, February 21, 1992

7) Saum, David, Memorandum to Members of the SAB Radon Engineering
Cost Subcommittee, dated February 3, 1993

8) Sullivan, John H., Letter to Carol Browner Pertaining to National Primary
Drinking Water Regulations: Radionuclides (Radon) [WH-FRL 3956-4], from
the Government Affairs Office of the American Water Works Association,
January 26, 1993

9) U.S. Congressional Record - Senate, S15103, Sec. 591 SAFE DRINKING
WATER ACT IMPLEMENTATION, September 25, 1992

10) U.S. EPA, "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
A-l
-------
Based on Comments to the Proposed Radionuclides regulation), A Draft
Document Prepared by Wade Miller Associates, Inc. under EPA Contract
No. 68-CO-0069, Work Assignment 1-32, for the Office of Ground Water and
Drinking Water, September 30, 1992

11) U.S. EPA, Briefing Materials by Dr. Janet A. Auerbach, Mr. James M.
Conlon, Mr. Michael Cummins, Mr. Frank Marcinowski, and Mr. Marc J.
Parrotfca, February 8, 1993

12) U.S. EPA, "National Primary Drinking Water Regulations; Radionuclides;
Proposed Rule, 40 CFR Parts 141 and 142," Federal Register. Vol. 56, No.
138, pages 33050 to 33127, July 18, 1991 (Attention to the cost components
in the Table of Contents, Section V, where the mitigation technologies and
the costs are discussed.)

13) U.S. EPA, "Regulatory Impact Analysis of Proposed National Primary
Drinking Water Regulations for Radionuclides," Prepared by Wade Miller
Associates, Inc. under EPA Contract No. 68-CO-0069, Work Assignment No.
0-1 for the Office of Drinking Water, Washington, D.C., July 17, 1991

14) U.S. EPA, "Technical Support Document for the 1992 Citizen's Guide to
Radon," Office of Air and Radiation (ANR-464), EPA 400-R-92-011, May 20,
1992

15) U.S. EPA, "Technologies and Costs for the Removal of radionuclides from
Potable Water Supplies," Prepared by Malcom Pirnie, Inc. for the Drinking
Water Technology Branch, Office of Ground Water and Drinking Water,
July 1992 (NOTE: This is the primary review document.)
A-2
-------
3)

4)
APPENDIX B - LITERATURE CITED

1) Safe Drinking Water Act Amendments of 1986, Public Law 99-339, 100
STAT642

2) Departments of Veterans Affairs and Housing and Urban Development, and
Independent Agencies Appropriation Act, 1993, PUB. L. 102-398, Section
519, 106 STAT 1618 (1992) (This is the citation adopted from the
Congressional Record (See Reference #3, below) that requires the EPA
Study of Radon.)

U.S. Congressional Record - Senate, S15103, Sec. 591 SAFE DRINKING
WATER ACT IMPLEMENTATION, September 25, 1992

U.S. EPA, "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], EPA/600/9-91/050, January 8, 1992

5) U.S. EPA, "Technical Support Document for the 1992 Citizen's Guide to
Radon," Office of Air and Radiation (ANR-464), EPA 400-R-92-011, May 20,
1992

6) U.S. EPA/SAB, "Reducing Risk: Setting Priorities and Strategies for
Environmental Protection," SAB-EC-90-021, September 25, 1990

7) U.S. EPA/SAB, ''Review of the Office of Drinking Water's Assessment of
Radionuclides in Drinking Water and Four Draft Criteria Documents: Man-
Made Radionuclide Occurrence, Uranium, Radium, Radon," Prepared by the
Drinking Water Subcommittee of the Radiation Advisory Committee of the
Science Advisory Board, EPA-SAB-RAC-87-035, July 27, 1987

8) U.S. EPA/SAB, "Status of Radionuclide Models," Prepared by the Radiation
Advisory Committee of the Science Advisory Board, EPA-SAB-RAC-92-001,
January 9, 1992

9) U.S. EPA/SAB, "Revised Radon Risk Estimates and Associated
Uncertainties," Prepared by the Radiation Advisory Committee of the
Science Advisory Board, EPA-SAB-RAC-LTR-92-003, January 9, 1992
B-l
-------
10; U.S. EPA/SAB, "Review of Drail Criteria Documents for Radionuclides in
Drinking Water," [Drinking Water Criteria Document for Uranium,
November 1989; External Review Draft for the Quantification of
Toxicological Effects Document on Radium, (TR-1242-67), 10 July 1990;
Quantitative Risk Assessment for radon in Drinking Water, May 1990; and
Quantitative Risk Assessment for Beta Particle and Gamma Emitters in
Drinking Water, May 1990], Prepared by the Radionuclides in Drinking
Water Subcommittee of the Radiation Advisory Committee of the Science
Advisory Board, EPA-SAB-RAC-92-009, January 9, 1992

11) Longtin, Jon "Occurrence of Radionuclides in Drinking Water: A National
Study," p. 97 Radon, Radium and Uranium in Drinking Waterf edited by C.
Cothern and P. Rebers, Lewis Publishers, Cheslea, Michigan, 1990

12) Milvy, P. and C. Cothern, "Scientific Background for the Development of
regulations for Radionuclides in Drinking Water," p. 1 Radon. Radium and
Uranium in Drinking Water, edited by C. Cothern and P. Rebers, Lewis
Publishers, Cheslea, Michigan, 1990
B-2
-------
APPENDIX C - COST ESTIMATES AND UNCERTAINTY
MEASURES
NOTE: The following tables of cost estimates and uncertainty measures, which
dissaggregate totals to include soil gas mitigation, have been provided by an
SAB/RECS consultant for illustration, comparison and discussion purposes only
and are not quality-checked or peer-reviewed for accuracy.
-------
-------
8
50K-75K
15
125
880,330
4.79%
$1,013
$23,637
$7
$23,637
$1,013
$2,602
0.67%
$3,789
$€79
3.83
$7
382,752
9
75K-100K
6
97
459,868
2.50%
$626
$14,554
$8
$14,554
$626
$1,605
0.41%
$2,336
$802
2.00
$8
199,943
10
100K-500K
10
177
1,981,030
10.78%
$1,872
$41,341
$5
$41,341
$1,872
$4,651
1.20%
$6,728
$53i
8.62
$5
861,317
11
SOOK-1M
1
59
721,366
3.92%
$732
$15,343
$6
$15,343
$732
$1,764
0.46%
$2,535
$562
3.14
$6
313,637
12
>1M
0
12
326,155
1.77%
$291
$5,210
$5
$5,210
$291
$641
0.17%
$903
$452
1.42
$5
141,807
ROW SUM
27,294
41,135
18,381,967
100.00%
$263,864
$1,834,136

$1,834,136
$263,864
$387,150
100.00%
$479,303
$4,839
80.00
$48
7,992,160
SOURCE
EPA
EPA
EPA
Saum
EPX
EPA
EPA
EPA
EPA
EPA
Saum
EPA
Saum
Saum
EPA
Saum
****************
$1,500
$80
$574
$30,620
$70,809
$185
1.3
0.4
50%
0.45
57.5
$1,232
0.55
$1,500
$80
$300
$15,995
$36,989
$185
1.3
0.4
50%
0.45
30.0
$1,232
0.65
$1,500
$80
$1,292
$68,905
$159,344
$185
1.3
0.4
50%
0.45
129.3
$1,232
0.44
$1,500
$80
$470
$25,091
$58,023
$185
1.3
0.4
50%
0.45
47.1
$1,232
0.46
$1,500
$80
$213
$11,345
$26,234
$185
1.3
0.4
50%
0.45
21.3
$1,232
0.37
$1,500
$80
$11,988
$639,373
$1,478,550
$185

1200.0

EPA
EPA
Saum
Saum
Saum
Saum
EPA
EPA
Saum
Saum
Saum
Saum
Saum
C-2
-------

^
c
o
• 1
•3
1
JJ
_ q
e *j
u S
JT3
3^

*l
• S
||
w o
ac
— *
n
**T?
•s,
ft •
S3
s°
**** 9

8 5P «[
°i s
M S U
i5 5
So S
»***
!5 If
S- *2S
a| 1"

A 2
Is «•
0 > •
-« « >
•JQ «
5>. "
a - •
«« »* •
OTJ e 5
.: s 3
• > i „
>•• 3 «
•3- ttjf
S 8X1
|
5
g
•G

25
A

23
1

^
*

«B

O
|
S
N
-s
c*
1
0

"5
I
n

n
I
*»tt
«
I
o
«•

"§
n
i
9
O
M
JJ

X
1?
2i
3:
:*
S2
t
^»rtr» o
**m<4>o
r* >H *4 o
<•*
*
M

«»
*4 •
M
n
Msssi
ct
r*

•»
•»
**^
"SSS
•»**
<»

«»
tfino**
»^
o
4»
4»
•P
•n^tnCi
^ *o
M»4
4»
<4»0«n

* »4»
«"*
«
o222
M"*?*
«0
M
s5s^
M««^M
S
VtMOMl

v»no •
M**«»
'M
*•
«»
» » **•
-5S**
rf.
tf>*4«k«»
1*1*1
S2S
p«
t9 M.
o2r) j
5^5^*
*?5 *
3«4 O
u|*-
• ll"*
a&
• M

N 4»
«» %
4V

MO
MM
£

ss
*4

• ss
•S

Ntfl
•» «»
J.

*>r*
°»*^
sa

2C
»* ^
r?rj
*•
«»43
n M

rfi*
4»
**r*
*)«•»
« *
»M

SM
O4tt
MO
•free
*•>«*-
*4 Vft
»s
2S
*
2S
* ^
*4>**
••>
t»«
»0

^
4J
8"ff—
ci-—
123
HOU
aj&aip
i «
i ^ « r* o 4%
t •» M n ^
1 ««»«» 4».
!S
1
1
1 *»
A i o MMr* 4M>
i MM"*****?
i »** w
1 49

1 » * *M *
i vtfm o
t
t 4»
s i ssIS*!**
! ^^-r****
i w«r» o
I M 4*"^ M
J* „
2 i sssi*
lies s
i ' •

•• i ^*>^ •«•
1325*5
1
1 «»
o i »nMm«*o
M 1 W*«««*0
«*• 1 * * »4B*^
] ssl"!
0 1 »OMWI>
S 1 5!Sn*!S
i S ^2 * S
I SS» *
| 3^.2^5
2 ! ^•H'3 "3 S
a j-SSSaS
S ! Jr * * *
o^ [ oo e~ e
1 M M •• M 2
* M 1 • 4 O
?S ! S o «'**^

«»«v»
«*4 .4!
• M n a
* ««J
a* '"rf
«
- * * r
• «••«
et »
f^r4
M *» 4»«t •» O **
:s irss'
O O »4 *^ «k rf»
.0 m M p* « •
•««» ••«*
3 ^33— 53!
3 "'
C-3
-------
APPENDIX D - GLOSSARY OF TERMS AND ACRONYMS

ACWA ASSOCIATION OF CALIFORNIA WATER AGENCIES
ASD ACTIVE SUB-SLAB DEPRESSURIZATION
AWWA AMERICAN WATER WORKS ASSOCIATION
BAT BEST AVAILABLE TECHNOLOGY
BEIR BIOLOGICAL EFFECTS OF IONIZING RADIATION
DWG DRINKING WATER COMMITTEE (U.S. EPA/SAB)
EEC ENVIRONMENTAL ENGINEERING COMMITTEE (U.S. EPA/SAB)
EHC ENVIRONMENTAL HEALTH COMMITTEE (U.S. EPA/SAB)
EPA U.S. ENVIRONMENTAL PROTECTION AGENCY (U.S. EPA, or
"The Agency")
GAG GRANULAR ACTIVATED CARBON
k THOUSAND (DOLLARS)
NAS NATIONAL ACADEMY OF SCIENCE
O&M OPERATION AND MAINTENANCE
OGW&DW OFFICE OF GROUNDWATER AND DRINKING WATER
ORD OFFICE OF RESEARCH AND DEVELOPMENT, U.S. EPA
PTA PACKED TOWER AERATION
PTS PACKED TOWER STRIPPING
L LITER
LCD LUNG CANCER DEATHS
M MILLION (DOLLARS)
pCi PICO CURIE
pCi/Lwate Concentration in water
pCi/L • Concentration in air
RAC RADIATION ADVISORY COMMITTEE (U.S. EPA/SAB)
SAB SCIENCE ADVISORY BOARD (U.S. EPA)
SDWA SAFE DRINKING WATER ACT OF 1988
U.S. UNITED STATES
VOC VOLATILE ORGANIC CARBON
D-l
-------
DISTRIBUTION LIST

Deputy Administrator
Assistant Administrators
EPA Regional Administrators
EPA Laboratory Directors
Deputy Assistant Administrator for Office of Research and Development (ORD):
Director, Center for Environmental Research Information (CERI)
Director, Office of Environmental Engineering and Technology
Demonstration (OEETD)
Director, Office of Environmental Processes and Effects Research (OEPER)
Director, Office of Modeling, Monitoring Systems, and Quality Assurance
(OMMSQA)
Director, Office of Technology Transfer and Regulatory Support (OTTRS)
Deputy Assistant Administrator for Water:
Director, Office of Ground Water and Drinking Water (OGW&DW)
Director, Office of Science and Technology (OST)
Deputy Director, OST
Deputy Assistant Administrator for Air and Radiation:
Director, Office of Radiation and Indoor Air (ORIA)
Director, Office of Air Quality Planning and Standards (OAQPS)
Director, Office of Radiation Programs (ORP), Las Vegas, Nevada
Deputy Assistant Administrator for Office of Prevention, Pesticides and Toxic
Substances (OPPTS):
Director, Office of Pollution prevention and Toxics (OPPT)
Deputy Assistant Administrator for Office of Solid Waste and Emergency Response
(OSWER):
Director, Office of Emergency and Remedial Response (OERR)
Deputy Director, OERR
Director, Office of Solid Waste (OSW)
Deputy Director, OSW
Director, Technology Innovation office (TIO)
EPA Headquarters Library
EPA Regional Libraries
EPA Laboratory Libraries
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
July 30, 1993 OFFCE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD
EPA-SAB-EC-LTR-93-010

Honorable Carol M. Browner
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460

Re: SAB Review of Multimedia Risk and Cost Assessment of Radon in
Drinking Water

Dear Ms. Browner:

The EPA Science Advisory Board (SAB) is pleased to comment on the
multimedia risk of exposure to radon and the cost of mitigation as required by
Public Law 102-389 (the Chafee-Lautenberg Amendments to EPA's FY 1993
Appropriation Bill enacted October 6, 1992). The Chafee-Lautenberg Amendment
states that "The Science Advisory Board shall review the Agency's study and
submit its recommendation to the Administrator on its findings," The study
report made available to the SAB is entitled "Multimedia Risk and Cost
Assessment of Radon in Drinking Water".1 This SAB report on the Agency's
study, prepared by the Chafee-Lautenberg Study Review Committee of the SAB,
complements previous detailed SAB comments transmitted to you on the
uncertainty analysis of radon ri.sks (July 9, 1993) and on costs of mitigation of
risks from radon in water (July 30, 1993).

The issues of major concern in assessing risks of radon exposure and costs
of mitigation may be grouped into four categories: a) population exposure profiles;
b) risk estimation procedures; c) mitigation costs; and d) integration of these for
regulatory decision making. The EPA study considered each of these issues and,
in turn, they have been addressed by the SAB.
By way of background, the SAB early in 1993 bejtan interactions with EPA, including receipt of background maurtal on
Out study However, the specific report reuiewed by the Committee was not received until July 9. 1993, and thus, tuntud tune
was available to review and comment on the report bn-uuse of the July 31. 1993 deadline for submission to Congnss.
Continuing to the present study report, there has been a steady improvement in the quality of the analyses conducted by EPA.
-------
A. Population exposure profiles

The Agency report estimates th?t 81 million people use water originating
from community groundwater supplies with a population-weighted average radon
activity of 246 picocuries per liter of water (pCi/Lwater). The Agency report
estimates that approximately 19 million people are served by water supplies with
radon concentrations in excess of 300 pCi/Lwater, the Maximum Concentration
Level proposed by the Agency. It is the SAB's impression from information
provided by public commenters, that the Agency's estimates of population exposure
to radon in drinking water are rather uncertain and may seriously underestimate
the number of community water systems impacted by the proposed drinking water
standard. This uncertainty in exposure estimates ultimately impacts the costs of
mitigation. There is clearly a need for more information and a better presentation
of available data on the profile of population exposure to radon in drinking water,
including the distribution of radon in drinking water exposures for communities of
varying size.

B. Risk estimation procedures

The risk estimation procedures used by the Agency address both the risks
from radon inhaled in air and ingested in water. The risk estimates from airborne
radon with lung cancer as an endpoint are based on strong epidemiological
evidence from studies of uranium miners, augmented by data on other
underground miners, and supported by data from laboratory animal studies.
However, there continues to be debate about the extrapolated lung cancer risk at
lower levels of exposure. This issue may be clarified during the next several years
when the results of several major epidemiological studies focusing on exposure to
radon in homes become available. However, even though there is a potential risk
at low levels of exposure to air borne radon, it must be recognized that the
populations available for epidemiological studies are relatively small, the majority
of residential exposures are not particularly high, and the postulated levels of risk
are sufficiently low that epidemiological studies might well be unable to identify
any increase in risk attributed to residential radon exposure if such a risk is
present.
-------
The situation is quite different for estimating the risks of ingested radon in
drinking water. In this case, there is no direct epidemiological or laboratory
animal evidence of cancer being caused by ingestron of radon in drinLlng water.
Thus, the approach to estimation of cancer risk from radon in drinking water is
more indirect than for radon in air. In the absence of direct evidence, it is not
possible to exclude the possibility of zero risk from ingested radon.

The indirect risk estimation approach involves several steps. First, the dose
to various tissues has been calculated from models for the distribution of radon in
the body following ingestion of radon. The model calculation is based, in part, on
organ distribution information from an. unpublished, study with radio-xenon (as a
surrogate for radon, since both are noble gases) using human subjects. The
meager data base results in uncertainty in estimating tissue doses from ingested
radon in drinking water. This uncertainly could be reduced through further
research. In the next step, the calculated doses have been used along with organ-
specific risk estimates per unit dose, derived from data on the Japanese atomic
bomb survivors, to calculate cancer risk to various organs. To a large extent, this
involves an extrapolation from the very acute, high dose rate, gamma (low Linear
Energy Transfer) exposure of the Atomic Bomb survivors to a very protracted,
very low dose rate, alpha particle (high Linear Energy Transfer) exposure with
ingested radon. The SAB is of the opinion that the estimates of risk from
ingested radon have additional uncertainty due to possible differences in the
distribution of dose, and resulting effects, from alpha particles from radon and
progeny. However, it should be noted that even at the upper bound of the
uncertainty analysis for ingested radon, for most situations the risk from radon
ingested in drinking water is still much lower than the risk from airborne radon
entering the house directly from the soil. Indeed, for many homes the risk from
the radon in water is even lower than that from radon in the outdoor air.

The available information on exposure and risk have been generally
integrated under a scientifically satisfactory framework by the Agency as evidenced
in the Agency's multimedia risk assessment for radon (EPA-SAB-RAC-93-014, July,
1993). However, the uncertainties noted earlier in this report are carried forward
into most of the integrated analyses. However, the differences of opinion,
especially with regard to the extent of the exposed population, with interested
parties are not reflected in the Agency report or in the integrated analyses.
-------
The risk estimates are illustrated in Figures 1 and 2. The population risk
estimates for airborne radon indoors are the most certain, with the nominal
estimate of 13,600 lung cancer deaths per year (range of 6740 to 30,600 lung
cancer deaths) from exposure to indoor air2. Less than one percent of this lung
cancer risk is attributable to radon reaching homes via water. In contrast,
exposure to radon in outdoor air is estimated to produce 520 lung cancer deaths
per year (range of 280 to 1500 lung cancer deaths)3. And finally it is estimated
that ingestion of radon hi water is estimated to cause 46 cancers per year (range
of 11 to 212 cancers per year)4. This latter estimate is the most uncertain of all
the estimates made. Airborne radon arising from water is estimated to result in
113 lung cancers per year (range of 40 to 408 lung cancers per year)5 which are
included in the estimate presented above for indoor residential air,, These risk
estimates for radon can be placed in perspective by comparison with an estimate
of approximately 30,000 cancer deaths per year from all exposures to naturally
occurring radiation, including approximately 13,600 deaths from inhaled radon and
approximately 2,500 cancers estimated for naturally occurring radio-potassium in
the human body.

C. Mitigation costs

The costs of mitigation of radon in the water and indoor air are also
uncertain. Part of the uncertainty for mitigation costs of radon in water relates to
differences of opinion between the Agency staff and interested parties over the
cost of mitigation systems. For example, the Agency staff estimates capital costs
for mitigation of radon in water at less than $2 billion, while interested parties
have estimates of capital costs in excess of $10 billion. Similar differences exist
for recurring maintenance and operating costs. The other part of the uncertainty
for mitigation costs of radon in water relates to the representativeness of the data
base on the occurrence of radon in groundwater used by the Agency. These data
2IUport to tlu UniUd State* Congrew on Radionuclidee in Drinking Water: Multimedia Risk and Cart A*»e««ment
of Rmdon in Drinking Water. Prepared for PL 102-389. Ofltco of Water. US Environmental Protection Agency- Jwly 9, 1993. pagn
3-2.

Ibid, Table 7-3 beta model estimate*.
!Wd, Table 7-3.
-------
are the source for estimates of the number and size of communities that would
require radon mitigation depending on the level of the MCL finally selected for
regulation. In contrast to the potential mandated regulation of radon in water,
mitigation of radon in indoor air involves voluntary actions by homeowners. Total
cost estimates of the latter are highly uncertain because the extent and cost of
testing for radon in homes and the extent of voluntary participation in mitigation
action in affected homes are unknown.

The SAB is of the opinion that the mitigation cost uncertainties for radon
in drinking water could be reduced by the EPA working with interested parties to
resolve issues related to the occurrence of radon in community systems of various
sizes, the cost of the various process treatment operations and processed for
various system sizes, and the frequency of the need for disinfection after aeration.
This may require reopening the comment period for this rulemaking. The SAB
recommends that EPA, if necessary, request from the Court and Congress
sufficient time to do this work to reduce uncertainties in the cost estimates and
the cost per cancer avoided. The public interest will be served if the Agency
carries out activities over several years which provide a better basis for deciding
how to most effectively mitigate risks from radon exposure in drinking water.

D. Integration for regulatory decision-making

Because of uncertainties in both risk estimates and costs of mitigation there
is substantial uncertainty in Ihe cost per cancer death avoided. This uncertainty
is especially large for mitigation of cancers related to ingestion of water. However,
even with this uncertainty, it is clear that the cost per lung cancer avoided from
mitigation of indoor air radon is substantially less than the cost per cancer death
avoided due to mitigation of exposure from radon in drinking water. This
difference appears to be at least a factor of 4 ($3.2 million per cancer death
related to drinking water and $0.7 million per cancer death related to airborne
radon) and may be substantially larger. The highest costs may be those associated
with mitijgation of risks for radon in water for the smallest communities.

In summary, the SAB notes the extent of the uncertainties in the population
exposure profiles, the risk estimates for ingested radon in drinking water and the
costs of mitigation. In view of these large uncertainties for risk estimates for
-------
ingested radon in drinking water and knowledge of the substantially greater risks
associated with airborne radon indoors and outdoors directly from soil, the SAB
advises that EPA consider various options for mitigating radon COM-M- risks. The
options all include continuing the Agency's efforts to encourage voluntary actions
to reduce indoor air radon in view of the cost effectiveness of this approach for
reducing risks.

With regard to water, as one option the Agency could promulgate a
standard at 300 pCi/Lwater as has been proposed. However, in doing so it must be
recognized that this involves selecting a risk reduction strategy for radon that is
the most costly in terms of costs per cancer death avoided; i.e., more than four
times the cost of cancer risk avoidance for airborne radon indoors. Alternatively,
as another option a standard might be set at some higher level such as 1000 to
3000 pCi/L^^,., to initiate mitigation of the highest potential risks. For example,
setting a water standard at 3000 pCi/Lwater would result in water contributing no
more radon to indoor air than is present in outdoor air. (Keep in mind that the
radon in outdoor air arises by natural processes from soil gas and there is no way
to alter the outdoor radon levels.) At the same time it would be appropriate to
intensify research on radon ingestion and radon mitigation, data gathering on
radon occurrence for all media, and dialogue with interested parties. These
actions would serve to reduce the uncertainties in the risk estimates, the costs of
mitigation, and, ultimately, the estimates of cost per cancer avoided. We cannot
emphasize too strongly the SAB view that a relative risk orientation should be
applied to the decision making process. Comparative analysis of uncertainties on
the risks of various exposure scenarios and mitigation approaches should be
developed and provided to the risk managers.

The SAB strongly supports the use of a relative risk reduction orientation
as an important consideration in making risk reduction decisions on all sources of
risk, including those attributable to radon. Other important considerations include
legislative authorities, environmental equity, economics, and the like. In short, the
relative risk approach calls for giving the highest priority to mitigating the largest
sources of risks firstj especially when the cost-effectiveness of .risk reduction of
such sources is high. The SAB recognizes that the large number of laws under
which EPA operates makes it difficult to implement a relative risk reduction
strategy uniformly across the Agency. Radon is an excellent example of the
-------
problem with radon in drinking water governed under one statute (Safe Drinking
Water Act) while radon in indoor air is not currently subject to regulation under a
specific statute. The SAB strongly encourages the Agency and the Congress to
work together to consider changes in existing statutes that would permit
implementation of relative risk reduction strategies in a more efficient and
effective manner.

The SAB appreciates this opportunity to advise you and the Congress on
this important matter, and we look forward to receiving a response on these
suggestions.
Sincere!
Dr. Raymond C. Loehr
Chair, Executive Committee
Science Advisory Board
>ger 17. McCle
Chair, Chafee-Lautenberg Study
Review Committee
-------
-------
8 -s
l!

M
l

i
*
cr §
<•> <<5
• >•»
a
&
•s
$ -I
•s
s *
3
5

5
-------
Figure 2. Estimated Annual Deaths From Exposure
to Radon (in Cancer Deaths/Year)
Estimated Cancer Deaths/Year
35,000
30.000
25.000
20,000
15,000
10,000
5,000
High
Median
Low
13,600
\46
520
Ingested DW Inhaled Outdoor Air Inhaled Indoor Air

Sources of Exposure

Source: 'Report to ft* US. Congress on Radtonucitdes in Drinking Water:
Multtmedfa ff/s* and Cost Assessment of Radon in Drinking Watef",
Office of Water, US EPA July 9, 1993.
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
CHAFEE-LAUTENBERG STUDY REVIEW COMMITTEE
CHAIRMAN

Dr. Roger O. McClellan
President
Chemical Industry Institute of Toxicology
P.O. Box 12137
Research Triangle Park, NC 27709

MEMBERS

Mr. Richard Conway
Senior Corporate Fellow
Union Carbide Corporation 770/341
P.O. Box 8361
South Charleston, WV 25303-0361

Dr. Morton Lippmanh
Professor
Institute of Environmental Medicine
New York University
Long Meadow Road
Tuxedo, NY 10987

Dr. Genevieve M. Matanoski
Professor of Epidemiology
School of Hygiene and Public Health
The Johns Hopkins University
601 Wolff Street, Room 6019
Baltimore, MD 21205

Dr. Verne Ray
Senior Technical Advisor
Medical Research Laboratory
Pfizer Inc.
Groton, CT 06340
-------
DESIGNATED FEDERAL OFFICIAL
Dr. Edward S. Bender
Environmental Protection Agency
Science Advisory Board
401 M Street, S.W., A-101
Washington, DC 20460
STAFF SECRETARY
Mrs. Marcia K. Jolly
Environmental P.rotection Agency
Science Advisory Board
401 M Street, S.W., A-101
Washington, DC 20460
-------
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 drinking water. Sections 7.1 and 7.2 address issues relating to the risk 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 Uncertainty Analysis of Risks Associated with Exposure to Radon —
Chafee-Lautenberg Multi-media Risk Study (June 1993).

The risk analyses 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 changes provide a more robust analysis of radon risks attributable
to drinking water.

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
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 including the use of
linear dose-response relationships extending to low dose, the effect of population mobility on
-------
7-2

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 Document 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 al. 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.
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
-------
7-3 ^^^

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 a 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-133 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
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
-------
7-4
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 "slow" 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 hi 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 close, 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
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. 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 Junctions (pdf). Probability
density functions are mathematical expressions which provide numerical statements of the
probability that a model variable, e.g.s radon concentration in drinking water, is within some
-------
7-6
arbitrarily small interval. Each pdf is characterized by its parameters1 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., lognorrnal, 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 lognormal 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 termparameter is often used to refer to model variables, i.e., model "parameter" is used
interchangeably with model "variable". 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
-------
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 Estimates

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.
-------
-------
V
I
1
is
s?
1
W
Ov
tu
U4
W
t»j

VI
O\
Ul
U]
VI
w
f>

00
SO
w
v
VI
0\

ci
s s s
-------
ts
i
«B
JS Z-4
o\
o\
VC
vo
00
22
-------

1E-5-
b
i
o
~g 1E-6-
u

•
•*•
(

Figure 7.1 Credible Range for
Mean Individual Risks [1,2]
inhalation
•

»•• •••
I *

•»

^•1
) (
•
ingestion
__
M*>
<
t
«••
•
| |
•••

» •
i 4
•M*
MB*
•> «
<
•W
|
•

t 4
_ ""•
a •

) —<

HB>

combined
pathways

"""" Feb/93 beta Feb/93 beta Feb/93 beta
triangular triangular triangular
Figure 7.2 Credible Range for
Population Risk [1,2]
600 •
500 •
£
Annual Fatal Can<
h N> c*> £
> o o c
! O Q C
1UU '
fl .
combined
pathways
inhalation

i
•

» <
•

_ —
i <
* -

ingestion

t
•Ml

» •"
(
*
•••

» 1
• •

> «J
!•

L»
•
Feb/93 beta
triangular
Feb/93 beta
triangular
Feb/93 beta
triangular
[1] • represents nominal value; — represents median values
[2] Credible range for inhalation risk includes risk from inhaling radon progeny
bat 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
sense of underestimating risk reduction benefits.
Box 7.1
Variable Control Scenario: Concentration-
Dependent Efficiencies. This scenario uses the
removal efficiencies assumed by EPA in its cost
analysis as shown in Box 7.1.
Table 7.4 and Figure 7.3a summarizes the results of
this analysis for a control level of 300 pCi/L.
Figure 7.3b extends this analysis to other control
levels between 100 and 1500 pCi/L for the beta
model.
Radon Removal
Efficiencies for the
Variable Control Scenario
\. for system concentrations between
300 - 600 pCi/L
" , -' removal efficiency = 0.50

for system concentrations between
600 - 1,500 pCi/L ,./•:'
..- '••'" V' -. removal efficiency = 0180
i.. J-, ' * -' " '
s " ,''•.
" '" ,':: " " '' ••
" for system concentrations greater
than 1,500 pCi/L
' , removal efficiency = 0.99
-------
* s
1^.
t"
•
'•I
E &*
<""fe
^•3
•s
SI
•«
I
I
1
l
1
^

«
\o
yj
to
?
ui
OS <7\
*
i
O
i
§ s
5?
w
vq
>n

-------
Figure 7.3a Annual Population Risks Attributable
to Reductions to 300 pCi/L
600
500

I400
= 300
•§200
£
100
triangular
model
"iiieflian"
No Controls Variable*
Truncation*
No Controls Variable*
Truncation*
.
£

550
500
450
400
350
300
250
200
150
100
50
0
Figure 7.3b Reduction in Annual Fatal
Cancer Cases for Different MCLs
beta model
variable control
100 200 300 400 500 600 700 800 900 1000 1200 1500
MCL, pCi/L

[—* Median Estimate Credible Bound • Nominal Estimate I
* Truncation refers to reductions in groundwater concentrations to the control! level and no
more; variable refers to concentration-dependent removal efficiencies assumed by EPA
in its benefit/cost analysis.
-------
7-16

Issue: The Effects of Population Mobility on Radon Risk
Estimates

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 paltems 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
-------
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
Residency Periods with
Radon Exposures

0
1
2
3
4
5
6
U.S. Average
Years Exposed to
Radon-containing
Groundwater

0
11.7
23.3
35.0
46.7
58.3
70.0
Lifetime Exposure

Average
Probability of
Exposure, (%)
9.52
27.40
32.89
21.05
7.58
1.45
0.12
22.7 years

Population
Exposed
(million)
23.8
68.5
82.2
62.6
19.0
3.6
0.3
E = 226.2

Relative Risk
Proportion

0
0.141
0.338
0.324
0.156
0.037
0.004
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 risk will Increase, exactly off-setting decrease in
exposure.
-------
Probability of indicated Years of Exposure Based on 6
Moves per Lifetime, Constant Total Exposure
70 years (0.1%)~|
58.3 years (1.5%)
46.7 years (7.6%)
No Exposure (9.5%)
35 years (21.0%)
11.7 years (27.4%)
23.3 years (32.9%)
Exposed Population (millions) and Years of Exposure
Based on 6 Moves per Lifetime,
Constant Total Exposure
70 years (0.25 )-i
58.3 years (3.75)
46.7 years (19 )
No Exposure (23.75)
35 years (52.5 )
11.7 years (68.5)
23.3 years (82.25)
Figures 7.4 and 7.5 Distribution of Exposures and Population Exposed for the
Constant Total Exposure Model
-------
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 21.7% 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: (pt, P2 ): p, is the conditional
probability of moving from a ground water-supplied system to another ground water-supplied
system (no change in status), and pa 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 pz= 18.5%. On a
regional level, pj 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 p? 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 alt risk from 81 million to 201 - 242 million.
-------

o
ii
oo,
Al 2

00 •* fN

2 a 2
O O 00

gi s a
tn in O
o\
VO
vn
o t- •<*•
• • *
oo vo oo
**
«- v
— ON cs
vo in vo
^ o oo

\o •*' g
•-< »•« w
Tf
-------
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
Number of Lifetime Years Exposed to
Residency Periods Radon-containing
with Radon Exposures Groundwater
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
U.S.
0
4.7
9.3
14.0
18.7
23.3
28.0
32.7
37.3
42.0
46.7
51.3
56.0
60.7
65.3
70.0
Average Lifetime Exposure
Average
Probability of
Exposure (%)
3.1
10.6
16.9
16.7
11.5
5.8
2.3
1.1
1.4
3.0
5.4
7.3
7.3
5.0
2.1
0.4
26.6 years
Population
Exposed
(million)
7.7
26.4
42.3
41.9
28.7
14.5
5.8
2.7
3.6
7.6
13.5
18.3
18.2
12.5
5.3
1.1
£ = 242.3
Relative Risk
Proportion
0
0.0211
0.0673
0.1000
0.0914
0.0576
0.0277
0.0157
0.0251
0.0599
0.1183
0.1763
0.1913
0.1426
0.0655
0.0140
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
Residency Periods
with Radon
Exposures
0
1
2
3
4
5
6
U.S. Average
Years Exposed
to Radon-
containing
Groundwater
0
11.7
23.3
35.0
46.7
58.3
70.0
Lifetime Exposure
Average
Probability of
Exposure (%)>
19.7
27.1
16.5
9.0
10.5
11.6
5.7
32.1 years
Population
Exposed
(million)
49.2
67.7
41.2
22.5
26.1
29.0
14.3
E = 200.8
Relative
Risk
Proportion
0
0.1349
0.1650
0.1408
0.2274
0.3172
0.1882
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)

58.3 years (29)
46.7 years (26.1 )
35 years (22.5)
No Exposure (49.2)
11.7 years (67.7)
23.3 years (41.2)
Figures 7.6 and 7.7 Distribution of Exposures and Population Exposed for the
Mobility-based Exposure Model
-------
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 bexthe major problems.
Detailed description of the relationship between radon and smoking in causing lung cancer
may be found in EPA's Technical Support Document 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 BEIR 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 $1.90/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, 1993X), 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, presenting 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 as
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-treatment 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 risk
due to radon in aeration off-gas would play a significant role in local decision-making. 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 REGS 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
appropriate 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.
-------
7-30
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 are reasonable, but cited some differences between the
aeration cost model and the models put forth by water industry groups.

In 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 1 MOD 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 during 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 (American 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 axe 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.
-------
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 improvements, 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 cm
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.
-------
-------
BIBLIOGRAPHY
Bair, W.J. 1991. Overview of ICR Respiratory Tract Model. Radiat. Protect. Dosim.
38:147-152.

Chameaud, J., 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., Weise SB, Callahan RJ, Strauss HN. 1988. The Kinetics of Ingested Radon 222
in Humans Determined from Measurements with Xe-133. Massachusetts General Hospital,
Boston, MA. Report to U.S. Environmental Protection Agency under cooperative agreement
CR810942.

Crawford-Brown, D.J., 1990. A Calculation of Organ Burdens, Dose and Health Risks from
Rn-222 Ingested in Water. (Draft). Report to the Office of Drinking 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-7105.

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-810264.

Gesell, T.F., and H.M. Prichard. 1980. The Contribution of Radon in Tap Water to Indoor
Radon Concentrations. In: Gesell T.F., LowderW.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).

Gesell, T.F. 1983. Background atmospheric Rn-222 concentration outdoors and indoors. In:
Health Physics. Vol. 45. p.289.

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.
-------
IARC. 1988. Monograph on the Evaluation of Carcinogenic Risks to Humans, Volume 43:
Man-made Mineral Fibers and Radon. IARC, World Health Organization.

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.

Longtin, J. Occurrence of Radon, Radium, and Uranium in Groundwater. Journal of
American Water Works Association. July, 1988

Longtin, J. Occurrence of Radionuclides in Drinking Water, A National Study. Radon,
Radium, and Uranium in Drinking Water (C.R. Cothern, Ed.) Lewis Publishers, Chelsea, MI
(1990).

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.

NCI. 1981. Surveillance, Epidemiology, and End Results: Incidence and Mortality Data
1973-1977. 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 (NIOSH) 88-101.

Partridge, I.E., T,R. Horton, and E.L. Sensintaffar. 1979. A Study ofRadon-222 Released
from 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.
Rebers. Lewis Publishers, Chelsea, MI.
1990. eds. J. Longtin, C.R. Cothern, and P.
Sullivan, R.E., N.S. Nelson, W.H. Ellet, D.E. Dunning Jr., R.W. Leggett, M.C. Yalcintas,
and K.F. Eckerman. 1981. Estimates of Health Risk from Exposure to Radioactive Pollutants.
Oak Ridge National Laboratory. TM-7745.

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.

UNSCEAR. 1988. United Nations Scientific Committee on the Effects of Atomic Radiation.
Sources, Effects and Risks of Ionizing Radiation. New York: United Nations.

UNSCEAR. 1986. United Nations Scientific Committee 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 I.
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/l-84-02r.

USEPA. 1987. The Risk Assessment Guidelines of 1986. Office of Health and Environmental
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. 1989a. An Analysis of Potential Radon Emissions From Water Treatment Plants Using
the MINEDOSE 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. I.
Background information document. Washington, D.C.: United States Environmental
-------
Protection Agency. EPA 520/1-80-005.

USEPA. 1991a. Drinking Water Criteria Document for Radon in Drinking Water. (Draft).
Prepared by Life Systems, Inc. June 1991.

USEPA. 1991b. 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.

USEPA. 1991c. Health Effects Assessment Summary Tables, Annual FY-1991. OERR 9200.6-
303 (91-1). NTIS PB91-921199.

USEPA. 1991d. Memorandum to G. Helms from I. Deloatch Regarding Estimated Cost of
Analyses for Radionuclides. March 27, 1991.

USEPA. 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
Programs, 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-210 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.Sk Environmental
Protection Agency. May 20, 1991.

USEPA. 199 Ih. Occurrence and Exposure Assessment for Gross Alpha Particle Activity
(Draft). Prepared by Wade Miller Associates. May 13, 1991.

USEPA. 1991L Proposed Revisions in EPA Estimates of Radon Risks and Associated
Uncertainties. April 1991.

USEPA. 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. 1991k. 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. 1992a. 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.

USEPA. 1992d. Packed Tower Aeration Cost Estimates for Radon Removal Memorandum
from Michael Cummins to Marc Parrotta. March 11, 1992.

USEPA. 1992e. Proposed Methodology for Estimating Radiogenic Cancer Risks. Office of
Radiation Programs, U. S. Environmental Protection Agency. Submitted to Radiation Advisory
Committee. May 1, 1992.

USEPA. 1992f. Reevaluation of EPA's Methodology for 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]. EPA/600/9-91/050
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. 1992L Technical Support Document for the 1992 Citizen's Guide to Radon. Office
of Air and Radiation, U.S. Environmental Protection Agency . EPA 400-R-92-011 Mav
1992. ' J

USEPA. 1992J. Technologies and Costs for the Removal of Radionuclides from Potable Water
Supplies. (Draft Document). Prepared by Malcom Pirnie, Inc. for the Drinking Water
Technology Branch, Office of Ground Water and Drinking Water, U.S. Environmental
Protection Agency. July 1992.

USEPA. 1993a. 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. 3

USEPA. 1993b. Revised EPA Methodology for Estimating Radiogenic Cancer Risks. U.S.
Environmental Protection Agency. Draft, July 12, 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 Christopher 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.

USEPA. 1993e. Revised EPA Methodology for Estimating Radiogenic Cancer Risks. Office of
Radiation and Indoor Air. In preparation.

USEPA. 1993f. Risk Assessment for Radon Emissions from Drinking Water Treatment
Facilities. Office of Radiation and Indoor Air, U.S. Environmental Protection Agency.
February 17, 1993.

USEPA. 1993g. Technical Memorandum: Problems with the Use ofGACfor 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. 1993L U.S.Environmental Protection Agency. Office of Ground Water and Drinking
Water. Suggested Guidelines for Disposal of Drinking Water Treatment Wastes Containing
Radioactivity. (Draft Document, July 1993)

USEPA. Working Draft of the Regulatory Impact Analysis for Final NPDWRfor Radionuclides.
Prepared by Wade Miller Associates for Office of Drinking Water, U.S. Environmental
Protection Agency. To be published.
-------
-------
<»-a o
in
«• DO
O c
-• to
-05"

i'S
a»>

-------