United States
Environmental Protection
Agency
Research and Development
Air and Energy Environmental
Research Laboratory
Research Triangle Park NC 27711
January 1990
The 1990 International
Symposium on Radon
and Radon Reduction
Technology:
Volume I. Preprints
Session I: Government
Programs, Policies, and
Public Information
Relating to Radon
Session A-l: Government
Programs, Policies, and
Public Information
Relating to Radon—
POSTERS
Session II: Radon Related
Health Studies
Session A-ll: Radon
Related Health Studies—
POSTERS
February 19-23,1990
Stouffer Waverly Hotel f
Atlanta, Georgia
-------
Session I:
Government Programs, Policies, and Public Information
Relating to Radon
-------
1-1
EPA'S RADON PROGRAM AND THE INDOOR RADON ABATEMENT ACT
by: Margo T. Oge
U.S. Environmental Protection Agency
Washington, DC
ABSTRACT
The Environmental Protection Agency established the Radon Action Program
in 1985 to reduce the health risks of radon. EPA's program includes
activities in four key area: problem assessment; mitigation and prevention;
capability development; and public information.
Congress enacted new legislation in October, 1988 which supports EPA's
approach to the radon problem and adds important new responsibilities. The
Indoor Radon Abatement Act contains the following key provisions:
• National Goal
• Citizen's Guide to Radon Revision
• State Grants Programs
• Technical Assistance Activities
• Model Building Standards for Radon-resistant New Construction
• Regional Radon Training Centers
• Radon Contractor Proficiency Program
• Radon in Federal Buildings Study
Implementing the new legislation requires the full participation and
commitment of States, health organizations, universities and the private
sector. Only through the combined efforts of these groups can we expect to
achieve meaningful reductions in the public health risks of indoor radon.
This paper will highlight major EPA Radon Program activities, and
provide an update on implementation of the Indoor Radon Abatement Act of 1988.
It will outline progress on and potential impacts of combined efforts to
reduce radon health risks.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
-------
1-2
REGULATION OF RADON IN DRINKING WATER
by: Gregory Helms
Office of Drinking Water
U. S. Environmental Protection Agency
Washington, D. C.
This paper has been reviewed in accordance with the XJ. S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
-------
2
Radon gas produced in the soil gradually seeps to the
surface and into the ambient air, or if it is under a building,
up through cracks in building foundations and into the indoor air
of the building. The average radon level in the air of homes is
about 1.5 pCi/1; the average outdoor background radon level is
about 0.2 pCi/1 (EPA/Ad Council 1989).
Radon gas produced in the soil can also dissolve in ground
water. Radon in ground water used for drinking water is a
problem because it contributes to radon in the indoor air of
homes. Because radon is volatile, water use such as showering
and bathing, dish or clothes washing, even toilet flushing, all
contribute radon to the indoor air if radon is present in the
water. (Radon is a problem only in ground or well water; it is
readily released from surface water to the ambient air before it
enters homes or other buildings).
Based on the results of a national survey of water in PWSs (the
National Inorganics and Radionuclides Survey, or NIRS), average
radon levels in groundwater are about 900 pCi/1, with a median
value of about 300 pCi/1. The highest level found in this survey
was 26,000 pCi/1, but wells in excess of 100,000 pCi/1 are known
to exist (EPA 1988, USGS 1989). In some parts of the country,
such as the west and north east, radon levels of 10,000 pCi/1 are
not uncommon (Connecticut Dept of Health Services, 1987; USGS
1989).
-------
3
Radon is carcinogenic to humans, and EPA estimates that 500-
20,000 lung cancer cases are caused by radon every year. It is
second only to smoking cigarettes, which causes more than 100,000
lung cancer deaths per year. Of the radon related lung cancer
cases, about 100-500 cases are attributable to radon in water
increasing radon air levels in buildings. EPA also believes that
radon in water may pose some increased risk of stomach cancer
when ingested. A quantitative estimate of the stomach cancer
risk is being developed but is not yet available.
The contribution of drinking water to radon in house air can
vary widely. On average, ten thousand pCi/1 of radon in water
will enrich the house air by about 1 pCi/1 (or 1:10,000).
However, this ratio can vary from house to house, due to
differences in house volume, ventilation rate, and water use
rates in the house. Air/water transfer rates for radon have been
reported as high as 1:3000 (EPA 1989; Region 4) or as low as
1:40,000 (Hess, 1989, preliminary report). Using a modeled
approach, Becker and Lazakjik (1984) estimated on average that
about 14,000 pCi/1 radon in water would enrich the air by 1
pci/i, with a range of 1:3300 to 1:59,000. Nazaroff et. al.
(1985) estimated the ratio at 1:15,000 using a statistical
modeled approach. He estimated the range for two standard
deviations at 1:2300 to 1:127,000. Relying on only one
significant figure, EPA has chosen the 1:10,000 ratio as a
-------
4
national average estimate for purposes of estimating risk from
radon in water.
Using the air/water transfer factor of 1:10,000, water from
PWSs would contribute about 0.09 pCi/1 to air on average or 0.03
pCi/1 as a median measure, or about 5% to overall indoor radon
levels, as an average.
In reviewing these estimates, it may appear that the radon
hazards due to water are unimportant except in those cases where
water radon levels are very high. Unfortunately, this is not
true. While it may require water containing 40,000 pCi/1 to
reach the current EPA action level of 4 pCi/1 in the air, it is
important to recognize that this action level is technology based
(and is only several times above background levels). That is, it
is set at the level where mitigation was judged, based on
available occurrence and mitigation effectiveness data, to be
both feasible and affordable for private home owners. The
estimated 2 cases occurring per 100 persons exposed is not
necessarily a "safe" level when measured against EPA regulation
of toxic chemicals discharged to the water or air, pesticides or
hazardous waste disposal sites. However, in a non-regulatory
program such as the radon program, where individual home owners
are being urged to test and mitigate, reduction of radon levels
in houses down to the 4 pCi/1 level, or lower, represents a
feasible and significant reduction in risk for many families.
-------
5
In setting environmental standards, EPA typically views
lifetime cancer risks of approximately 1/10,000 or lower (down to
one per million) as "acceptable" or "safe" levels. In the recent
regulation of benzene emissions to the ambient air, an
approximate 1/10,000 presumptive "safe" lifetime cancer risk
level was presented (54 FR 38044, September 14, 1989), and
reiterated in the final regulations controlling emissions of
radionuclides to the air (get cite). In regulating drinking
water contaminants under the Safe Drinking Water Act (SDWA), EPA
has sought to reduce individual upper bound lifetime cancer risks
from drinking water into the range of 1/10,000 to 1 per million
(52 FR 25700-25701, July 8, 1987).
Standards under the SDWA are set in two parts: the maximum
contaminant level goal (MCLG) and the maximum contaminant level
(MCL). The MCLG is a health based regulatory goal, which is set
solely on an assessment of the adverse health effects information
available on a contaminant. The MCLG is required to be set a the
level "at which no known or anticipated adverse health effect
occurs, with an ample margin of safety" (SDWA Section
1412(b)(4)). For carcinogens, EPA, as a matter of policy, sets
MCLGs at zero (50 FR 46881, November 13, 1985). This policy is
based on the legislative history of the SDWA, and the non-
threshold theory of carcinogenicity. According to this theory,
there exists some cancer risk (albeit small) at any exposure
-------
6
level.
The MCL, is an enforceable standard required by the SDWA to
be set as close to the MCLG as is feasible, taking both
technology and cost into account (SDWA Section 1412(b)(5)). In
evaluating the feasibility of different possible MCLs, EPA
considers what treatment for removal of drinking water
contaminants is available, the levels to which water can be
treated, and how affordable the treatment is for large
metropolitan water supplies (e.g., 50,000 to 75,000 population)
using relatively clean source water (Report No. 93-1185, 93rd
Congress, 2nd session, page 18). EPA also considers the
feasibility of reliably measuring contaminants in water in a
compliance monitoring setting (rather than a research setting).
The lowest level that can be reliably measured within defined
limits of precision and accuracy is called the practical
quantitation limit (PQL). Reviewing these factors together, EPA
identifies the lowest contaminant level to which water can be
treated and each contaminant measured, and which is affordable,
and sets the MCL at that level. EPA reviews the individual
risks remaining at this level, and uses the 1 in 10,000 risk
range (for probable or known carcinogens) as a target reference
risk range. EPA considers these to be safe levels and protective
of public health (52 PR 25700, July 8, 1987).
The treatment for radon in water at public water systems
-------
7
consist of various types of aeration. Very low-tech spray
aeration (into a storage tank, with ventilation of the tank) can
achieve 60-70 % removal of radon, and costs very little. Packed
tower aeration can achieve up to 99%, or better removals. At an
intake level of 20,000 pCi/1 radon, 99% removal would produce
water having 200 pCi/1. Studies indicate that granular activated
carbon (GAC) can also remove radon from drinking water. Kinner
et.al (1988) reported an overall average radon removal of 80%.
However, radon can require contact times that are prohibitive in
some cases (Dixon and Lee, 1988). Also, the carbon used in the
Kinner study became contaminated with radium 226 and was
classified as a low level radioactive waste in New Hampshire,
where the study was conducted, The investigators found disposal
of such low level waste to be problematic. Estimated disposal
costs were $13,000-$15,000 for the 47 cubic feet generated.
Gamma radiation was also reported to be emitted by the GAC units.
GAC has been used to treat private wells, where treatment
efficiencies of 90-95% have been reported (Lowry et.al., 1989).
In assessing the cost feasibility of different technologies
and possible standards, one of the key criteria that EPA
considers is the estimated increase in a residential water bill
that would result from installation and operation of the
treatment. An estimated annual increase of $100 or less in the
residential water bills for a community is generally considered
affordable. For water systems serving 50,000 to 75,000 people,
-------
8
and using aeration treatment to remove radon, the annual increase
in a family's water bill would be about $1.50 to $3.00, depending
on the treatment level. These household costs increase for
residents served by smaller public water systems, but become more
burdensome only in systems serving 100 or fewer persons, where
increases could range from $60.00 to $140.00 per year. For most
people served by public water systems, the cost of treating to
remove radon from their water will be small. For people relying
on private wells for water, radon treatment costs will be
substantially higher. They should review the radon contributed
to indoor air from soil as well as from water use in making
decisions about home radon mitigation.
EPA also reviews and evaluates the analytic methods
available for measuring contaminants in drinking water. For
radon, liquid scintillation counting and the Lucas Cell methods
are undergoing review. These methods are being evaluated for
their practicality and precision and accuracy at low radon levels
to determine which is appropriate for compliance monitoring in
regulating radon. While able to measure radon at lower levels,
the Lucas Cell method may be too cumbersome and labor intensive
to use of the large number of water samples that will need to be
analyzed to determine public water system compliance with the
radon regulations. While generally believed to be reliable to
relatively low levels, high quality interlaboratory testing of
these methods was lacking until recently. One report presented
-------
9
test results for the two methods for precision and accuracy down
to 1600 pCi/1 in water (EPA 1989). EPA is conducting additional
study of lab performance using these two methods down to 100
pci/l radon. The results of this additional work will be
presented in the proposed rule.
EPA expects to propose the radon standards in the fall of
1990. EPA is considering a range of options for proposal.
Some of the impacts and benefits of the options under
consideration are as follows:
Increment
MCL
cases/yr $/yr
$/case systems Ind. risk
200
94
$360M $4.8M 33,000 1/10,000
500
53
$170M $4.ON 18,000 3/10,000
1000
32
$ 82M $3.5M 9,400 6/10,000
2000
18
$ 35M $1.9M 4,300 1/1,000
EPA has not yet selected one of these alternatives for
proposal, but will most likely discuss all of them as
alternatives considered, with an explanation of why one was
-------
10
selected for proposal and the others not selected.
EPA is also working to develop additional public information
materials to explain the relative hazards from radon in water and
from basement soil gas. While it is very cost efficient for
public water supplies to centrally treat water for radon, private
well owners will not have the economies of scale that make
centralized radon treatment a bargain. Home water treatment
devices currently on the market include both GAC and aeration
devices. These devices cost about $1000.00 to $2500.00 to
install, plus require some maintenance. While efficient in
removing radon from water, often greater reductions in home radon
levels can be achieved through other, less costly mitigation,
such as sealing basement cracks or basement ventilation.
Homeowners contemplating radon mitigation need to understand all
the sources of radon in their homes, and choose the most cost
efficient method of mitigation, in order to reduce their risks as
much as possible while spending amounts that are affordable for
them.
The work described in this paper was not funded by the U. S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
-------
11
References
Dixon, K.L., and Lee, R.G., 1988, Occurrence of Radon in Well
Supplies, Journal AWWA, July, 1988, pp 65-70
EPA/Ad Council, 1989, Has Your Home Been Invaded by Radon?
EPA 1988. National Inorganics and Radionuclides Survey.
Memorandum Jon Longtin to Arthur Perler, February 23, 1988.
USGS, 1989. Personal communication. Letter from Richard Wanty,
USGS, Denver CO, December 20 1989.
Conn. Dept of Health Services, 1987. Connecticut RAdon Survey of
Private Well Water band indoor Air: Assessing Geologic,
Hydrologic and Household Parameters. Preventable Diseases
Division, December 1987.
EPA 1989. Radionuclide Concentrations from Waters of Selected
Aquifers in Georgia. EPA Region 4, Atlanta Georgia, August 1989.
Hess 1989.
Becker and Lazakjik, 1984. Evaluation of Waterborne Radon Impact
on Indoor Air Quality and Assessment of Control Options. EPA-
600/7-84-093.
Nazaroff, 1985
Kinner, N.E. Schell, G., Quern, P., and Lessard, C., 1988. Radon
Removal from Drinking Water Using Granular Activated Carbon,
Packed Tower Aeration, and Diffused Bubble Aeration Techniques.
Symposium on Radon and Radon Reduction Technology, Denver
Colorado, October 17-21, 1988.
Lowry et.al, 1989
EPA 1989. Two Test Procedures For Radon in Drinking Water.
Interlaboratory Collaborative Study. EPA/600/2-87/082
-------
Regulation of Radon in Drinking Water
Gregory Helms
U. S. EPA Office of Drinking Water
EPA's Office of Radiation Programs is presently making
considerable effort to address health risks associated with
exposure to radon. These efforts have sought to heighten public
awareness of radon as a public health hazard, induce the public
to test their homes for radon, and when the EPA action level is
exceeded, employ one of a number of mitigation techniques to
reduce household levels and exposure to radon. The Office of
Drinking water has also been working to develop regulatory
standards for radon in drinking water, as required by the 1986
amendments to the Safe Drinking Water Act (SDWA Section
1412(b)(1). EPA plans to propose standards for radon and other
radionuclides in drinking water in fall of 1990, take public
comment on the proposal, and publish final standards about a year
and a half after proposal. These standards will apply to all
public water systems (PWSs), which are supplies regularly serving
at least 15 service connections or 25 or more people. Private
wells are not covered by SDWA standards, but because they are
part of the radon in water problem, EPA plans to provide
information on radon in water to private well owners.
-------
1-3
"SOCIAL MARKETING" AND THE REDUCTION OF INDOOR RADON
by: Thomas J. Bierma, Ph.D.
Department of Health Sciences
Illinois State University
Normal, Illinois 61761
Daniel Swartzman, J.D.
School of Public Health
University of Illinois at Chicago
Chicago, Illinois 60680
ABSTRACT
Indoor radon monitoring and mitigation has apparently been
conducted by only a small percentage of private homeowners despite
extensive media coverage of the radon issue and public information
programs at the federal, state, and local levels. Whether public
education programs should be informative or persuasive is an ethical
decision. Arguments on each side of the issue are presented. A
framework for the development of a persuasive campaign is then
presented, using the concepts of social marketing and the limited
empirical evidence available on radon-related behavior. A four-step
process is proposed; 1) identify problem dimensions using focus groups
and other methods, 2) confirm dimensions for market segmentation using
probability samples and factor analysis, 3) implement program, and 4)
evaluate and revise, has not occurred spontaneously
BACKGROUND
The lung cancer risk from exposure to indoor radon is well known.
Testing for radon is easy and inexpensive for the homeowner. This
information has been carried by newspapers and the broadcast media
nationwide for several years. Many states have had public radon
-------
information programs in place since 1986 (1). One might expect, given
that the radon cancer risk is apparently greater than from all other
environmental exposures (2), and the relative ease with which elevated
radon concentrations can be detected and reduced, that public response
would be swift and extensive.
This does not appear to be the case. Though valid estimates do not
exist for the proportion of U.S. homes that have tested for radon, many
unofficial estimates have been offered by those in the field. The
highest of which we are aware does not exceed 2%. Even in New Jersey,
which has had an active state radon program for several years,
including mandatory reporting of home test results, the proportion of
homes tested does not exceed 10% (personal conversation with Mohammed
Rahman, New Jersey Department of Environmental Protection). In a
survey of public response to the radon issue in New Jersey, apathy,
rather than anger or panic, characterizes the public attitude (3).
It is time to critically evaluate why, in the face of such high risk,
and with having the tools at hand for reducing that risk, so few
homeowners choose to address the potential for radon problems in their
homes. In this paper, we review a number of models, borrowed from the
fields of communications and health behavior, that are relevant to
understanding radon-related behavior. Next, we highlight the findings
of a number of studies that have explored the motivational factors for
radon-related behaviors in selected populations. Finally, we present a
tool, Social Marketing, for translating the behavioral models into
effective intervention programs.
Before one can meaningfully discuss public education programs, it
is necessary to clarify the ethical basis for such programs. Indeed,
it may be that differences in our ethical positions account, in large
part, for our inability to motivate public response to the radon issue.
THE ETHICS OF PUBLIC MOTIVATION
One of the founding principles of government in the United States
is protection of human liberty, the freedom to act as one sees fit.
Protection of liberty is prominent in our Bill of Rights. It is the
foundation of the capitalist market system. This legal and economic
system is shared by most western nations.
Many have suggested this legal and economic model as a basis for
social policy as well (see reference 4 as a prominent example). This
view has been characterized as the "minimalist ethic": any behavior is
socially acceptable as long as it does not interfere with another's
liberty (5).
Guidelines for government intervention in the economy are seen as
an acceptable model for intervention in social issues. In particular,
-------
the government should intervene in the marketplace when the
distribution of information is insufficient to allow consumers to make
informed purchase decisions. The intervention is limited to those
actions necessary to assure distribution of factually correct
information.
As a model for the radon policy, the minimalist ethic dictates
that government intervention should strictly be informative.
Information should be made readily available to homeowners representing
the best factual information on the subject. This characterizes, to
date, most state and federal radon programs. Booklets presenting
factual information on indoor radon are provided to homeowners on
request. A hotline is often provided to minimize information cost to
the homeowner. Research is largely focused on risk communication and
the means by which educational materials can overcome the apparent
"irrationality" of public risk perceptions.
There are, however, challenges to the minimalist ethic as a basis
for social policy. One challenge focuses on the "irrationality" of the
public, arguing that government intervention in the form of persuasion
or coercion may be warranted if the individual does not recognize what
is in his or her own best interest (6). This Paternalistic view of
social policy would consider radon risks to be needless, and thus
intolerable.
Another challenge to the minimalist ethic is an appeal to the
greater good of the community as a whole in acting to protect the
public health (7, 8, 9) . This Communitarian ethic, as argued by
Beauchamp (8), is the foundation of the public health profession. To
base public health policy on the narrow interests of individual well-
being is to risk undermining the validity of public health.
The minimalist ethic, and its opposing ethics frame the central
issue in radon education: shall we persuade or shall we only inform?
Either of the challenges to the minimalist ethic support the use of
persuasion. It is this view that the authors adopt in the presentation
of this paper. Radon education campaigns should be informational, but
should be persuasive as well. The objective should be to reduce radon-
induced lung cancer risk, not simply to inform the public about it.
BEHAVIORS IN RADON REDUCTION
Reduction of indoor radon involves a series of decisions on the
part of the homeowner (10). First, the decision must be made to
perform initial testing of the home. This involves purchasing a
testing device from a retail store or through the mail. Secondary
testing, to confirm the results of initial testing, may then be
required, depending upon the conditions of initial monitoring and the
-------
anticipated expense of mitigation. Finally, the decision must be made
to mitigate if warranted by the results of testing. Radon reduction
will not take place if the homeowner fails to proceed at any point in
this series of steps.
To devise an effective education program, it is necessary to
understand the weak links in this series of steps and the factors that
motivate each decision.
MODELS OF HEALTH BEHAVIOR AND THEIR APPLICATION TO RADON
The health behavior literature is rich and voluminous. No attempt
is made here to present the full array of proposed models for health
behavior motivation. We will highlight some of the dominant factors in
many health behaviors and discuss their application to radon-related
behaviors. The intent here is to set the stage for the value of
social marketing in the construction of radon public education
programs.
The Health Belief Model (HBM) is a popular and insightful model in
the field of health behavior. Figure 1 summarizes the key components
of the HBM, adapted from Rosenstock (11). In their review of HBM
research, Janz and Becker (12) found "perceived barriers" to be the
construct most predictive of health-related behaviors in general.
"Perceived susceptibility" tended to be a better predictor than
"perceived benefits" in behaviors related to prevention. "Perceived
severity" was a relatively poor predictor of preventive behaviors.
Empirical evidence of the factors motivating radon-related
behavior is limited. Evidence is available primarily from a study of
New Jersey residents (3, 13, 14), a study of Maine residents (15), and
a series of focus groups conducted by USEPA in Maryland and
Pennsylvania (16).
Empirical evidence does not suggest that the cost of monitoring or
mitigation is a significant barrier to action for most homeowners.
However, a lack of adequate information about monitoring and mitigation
methods, the uncertainty of dealing with mail-order testing companies
and with mitigation contractors, and concern about property values, may
all be significant barriers to testing and mitigation. Interestingly,
concern over property values can be both motivating and inhibitory
(16).
The low level of perceived susceptibility to the risks from radon
appears to be a significant problem in motivating testing and
mitigation. Most residents, in the studies performed, believed the
risks from radon in their own homes were lower or much lower than in
others homes. Among those who have tested for radon, test results are
-------
generally only weakly correlated with perceived risks and intentions to
mitigate. Results from Maine, New Jersey, and the USEPA focus groups
suggest that the longer one has lived in the house or community, the
less one's perceived level of radon-related risk.
Perceived severity has been found to have a dual role in
motivating many health behaviors. At a moderate level, it appears to
be motivational, yet at a high level it may be inhibitory. The
inhibitory effect may either be due to denial, a feeling of
helplessness, or some other cause. Its role in motivating radon-
related behaviors is unclear. Generally, fear should be used for health
behavior motivation only with great caution. Fear is often inhibitory
unless it is immediately followed with clear and simple steps to be
taken to reduce risks (17). Radon messages stressing the risks from
exposure should include clearly presented information on the ease and
effectiveness of mitigation methods.
One of the limitations of the Health Belief Model is its focus on
the risks to the individual. A number of other models have identified
social influence and self-esteem variables as possible factors
influencing motivation of health behaviors. These variables come from
the work of Bandura (18, 19, 20), Rogers (21, 22), Cleary (23), Suchman
(24), Gottlieb and Baker (25), Langlie (26), Muhlencamp and Sayles
(27), and others. The greatest value in the use of these alternative
models is their focus away from the risk implications of radon,
recognizing the multidimensionality of the problem as perceived by the
public.
Peer pressure may be a significant motivational factor in radon-
related behaviors. USEPA (16) found that those who had monitored for
radon typically had talked about it with others before deciding to
test. Those who had not monitored generally had not discussed testing
with others. Among the best predictors of monitoring intentions in the
New Jersey study (3) was the belief that other adults in the home were
concerned or wanted to test, the belief that the people one knows are
concerned, what one has heard others say about testing, and whether a
friend has tested.
Though the "credibility" of radon messages has been evaluated, and
government sources have generally been considered "credible" (3, 16),
this may only be a small part of what makes an "effective" message
source. To motivate behavior, a message source should be a "model"
(19); that is, the audience should desire to emulate the actions of the
message source.
Many variables, such as the nature of peer and institutional
networks, control or freedom, self-esteem, denial mechanisms, and many
others, have yet to be explored in radon behavior research.
-------
SOCIAL MARKETING AND RADON BEHAVIOR
The marketing of commercial products has become a sophisticated
and successful process. Several concepts have evolved as the key to
effective marketing programs. The application of these key concepts to
the marketing of social issues has become known as social marketing
(28). While social marketing has been used in many areas of public
policy for some time, its use for marketing health-related issues is
fairly recent (see reference 29 for a recent review).
A central focus of social marketing is its use of the "marketing
mix" (traditionally - product, price, place, and promotion). This
focus enhances the use of the many attributes of a public health issue.
In the case of radon, this may be lung cancer risk, property values,
one's effectiveness in testing and reducing radon, peer perceptions,
etc. Social marketing also recognizes the importance of the message
channel and source, including the "model" value of the source.
Social marketing has a consumer orientation; that is, the bottom
line is the perceptions of the consumer and what the consumer considers
important in an issue. This does not imply "bending the truth" about
an issue to fit the needs of the consumer, but emphasizing those
aspects of an issue that the consumer wants to know about, particularly
to the end of clarifying misconceptions about the issue. Related to
this consumer orientation is the use of "market segmentation". This is
the recognition that different segments of the population may have
different views of what is important about an issue, and require
different marketing mixes as well as message channel/source
combinations.
Commercial marketing, as well as social marketing, relies heavily
on formative research in the development of marketing strategies. In-
depth cases studies, focus groups, community surveys, and pilot tests,
are among the tools used to initiate and refine a program.
Social marketing does not replace the health behavior models, but
provides a framework in which the models can be used to build a
comprehensive campaign systematically.
SOCIAL MARKETING AS A FRAMEWORK FOR RADON EDUCATION
Social marketing appears to hold much promise for increasing the
effectiveness of radon education programs. It is clear from the
limited empirical evidence available that the radon issue is
multidimensional, with many factors other than lung cancer risk playing
a role in determining testing and mitigation behavior.
-------
The framework of social marketing implies a series of steps in the
development of radon education campaigns.
1. IDENTIFY THE RELEVANT DIMENSIONS OF THE RADON PROBLEM
What does indoor radon actually mean to people? It is clear from
the limited data already available that to some people, it is a lung
cancer risk; to others, it is a threat to property values. For some it
is highly technical and difficult to understand; for others it may be
an opportunity to renew, if only in a small way, their bonds to the
community by taking action on an issue perceived as important by the
community. Their are undoubtedly many other dimensions of the radon
problem that are relevant to certain segments of the population.
Identifying such dimensions requires the use of qualitative, open-
ended, research methods such as the focus group and case study. The
focus group involves a small number of residents who meet as a group
with a non-technical "facilitator". The facilitator's job is to
encourage discussion of individual feelings and perceptions about the
issue. Focus groups are not considered representative samples of the
population in a statistical sense, but only serve to identify concerns
qualitatively. Case studies involve an in-depth interview of
individuals to explore the process by which they have made their
decisions. This can shed light on changes in perceptions and the
sequence of those changes.
In addition, an attempt is made to identify the psychological or
socio-demographic characteristics associated with the different
dimensions of the issue, as well as message source and channel
combinations likely to be relevant to different individuals.
Though focus group methods have recently begun to receive
attention in the risk communication literature (30), relatively few
researchers in the health behavior field have developed the research
skills needed for such qualitative studies. Collaborative work with
the fields of anthropology, sociology, or business marketing is likely
to be necessary.
The focus group work performed by USEPA in Maryland and
Pennsylvania represents an excellent start in this direction.
2. CONFIRM THE RELEVANT DIMENSIONS AND BASIS FOR MARKET SEGMENTATION
Based upon the results of the focus groups and cases studies a set
of relevant dimensions is assembled. With this are the psychological
and socio-demographic factors hypothesized to be important in
segmenting the market, as well as the message source and channel
combinations believed to be effective for each segment. A survey
instrument is developed with several questions used to define each
-------
dimension, psychological and demographic variable, and message
source/channel combination. Probability samples from the community are
then used to collect responses to the survey instrument.
Data analysis techniques such as factor analysis, discriminant
analysis, and logistic regression can be used to identify:
1. the hypothesized dimensions that best represent unique
concerns within the population,
2. the psychological and socio-demographic variables which can
best be used to segment the market into unique subpopulations with
common concerns, and
3. the message source and channel combinations most closely
associated with each population segment.
An excellent beginning in this direction is the work by Ferguson
and Valenti (31) examining dimensions of radon perceptions among
Florida residents. They examined attitudes about risk-taking behaviors
along with other characteristics, including education level and gender.
Using factor analysis, they identified three distinct population
segments ("adventurous", "impulsive", and "rebellious"). These were
then examined in relation to reactions to different radon messages
(risk to self versus risk to children, specific versus vague desired
action, and familiar versus abstract risk comparisons) and message
source (newspaper article versus government brochure). This approach
represents the type of work that will be required to develop radon
messages that can target specific segments of the population with
messages about those factors each segment considers important.
3. DEVELOP APPROPRIATE MESSAGES AND CONVEY THESE USING APPROPRIATE
SOURCES AND CHANNELS
Once the critical dimensions for each market segment have been
identified, messages can be developed addressing each dimension.
Probably of greatest importance is the selection of appropriate sources
for such communications. Effective models, identified through the
previous two steps, should be used as message sources. These models
may be local opinion leaders, nationally recognized figures
(celebrities, sports figures, political leaders, etc.), or stereotypes
(farmers, housewives, physicians, etc.). Government agencies are
unlikely to be effective models as message sources, although this is a
hypothesis to be tested in the above work.
4. EVALUATE THE EFFECTIVENESS OF THE MESSAGES AND REVISE PROGRAMS
Follow-up of the effectiveness of an education campaign and making
appropriate revisions to the program are critical parts of social
-------
marketing. Some messages are likely to be ineffective and alternative
approaches will need to be considered. Documentation of effective
messages will aid in deciding which messages to apply in other areas of
the country. Information on message effectiveness can be obtained
through focus groups, community surveys, and through data from the
testing and mitigation industry. Effective tracking of testing
activity may require mandatory reporting, through a confidential means,
of all radon tests.
CONCLUSIONS
Indoor radon is clearly a widespread issue. In most parts of the
country, a substantial percentage of the housing stock (exceeding 50%
in some cases) has elevated concentrations of radon. The problem is
unlikely to be short-lived. Radon testing and mitigation behavior has
proven exceedingly difficult to motivate. Even following mitigation,
evidence suggests that many mitigation systems lose effectiveness with
time. Periodic testing and mitigation system maintenance are likely to
be required in most homes.
Short-term or hastily assembled public education programs are
unlikely to be effective. A sustained and avowedly persuasive effort,
using a systematic framework to develop education programs based upon
current theory and empirical evidence, is greatly needed.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official endorsement
should be inferred.
REFERENCES
1. US Environmental Protection Agency. Summary of state radon programs,
EPA 520/ 1-87-19-1, 1987a.
2. US Environmental Protection Agency, Office of Policy Analysis.
Unfinished business: A comparative assessment of environmental
problems. US Environmental Protection Agency, Washington, D.C.,
1987.
3. Weinstein, N.D., Sandman, P.M., Klotz, M.L. Public response to the
risk from radon, 1986. Final Report to the New Jersey Department of
Environmental Protection, Trenton, New Jersey, 1987.
4. Friedman, M., Friedman, R. Free to Choose, Harcort, Brace, Janovich,
1980.
-------
5. Callahan, D. Minimalist Ethic. The Hastings Center Quarterly Report,
October, 1981.
6. Dworkin, G. Paternalism. The Monist, 56:64, 1972.
7. Beauchamp, D. Public health and individual liberty. Annual Review of
Public Health, 1:121, 1980.
8. Beauchamp, D. Community: The neglected tradition of public health.
Hastings Center Report, 1985.
9. Forster, J. A communitarian ethical model for public health
interventions: An alternative to individual behavior change
Strategies,", Journal of Public Health Policy, v2, pl50, 1982.
10. Svenson, 0., Fischhoff, B. Levels of environmental decisions.
Journal of Environmental Psychology, 5:55-67, 1985.
11. Rosenstock, I.M. Historical origins of the Health Belief Model.
Health Education Monographs. 2:328, 1974.
12. Janz, N.K,, Becker, M.H. The Health Belief Model: A decade later.
Health Education Quarterly, 11:1, 1984.
13. Weinstein, N.D., Sandman, P.M., Klotz, M.L. Optimistic biases in
public perceptions of the risk from radon. American Journal of
Public Health, 78:796-800, 1988.
14. Sandman, P.M., Weinstein, N.D., Klotz, M.L. Public response to
risk from geological radon. Journal of Communication, 37:93-108,
1987.
15. Johnson, F.R., Luken, R.A. Radon risk information and voluntary
protection: evidence from a natural experiment. Journal of Risk
Analysis, 7:97-107, 1987.
16. US Environmental Protection Agency, Office of Policy, Planning and
Evaluation. Radon risk communication project interim report. U.S.
Environmental Protection Agency, Washington, D.C., 1987.
17. Job, R.F.S. Effective and ineffective use of fear in health
promotion campaigns. American Journal of Public Health, 78:163-167,
1988.
18. Bandura, A. Principles of behavior modification. New York, Holt-
Re inhart-Winston, 1969.
19. Bandura, A. Agression: A social learning analysis. Englewood
Cliffs, NJ, Prentice-Hall, 1973.
-------
20. Bandura, A., Adams N.E., Beyer J. Cognitive processes mediating
behavioral change, Journal of Personality and Social Psychology,
35:125-139, 1977.
21. Rogers, E.M. Diffusion of Innovations. New York, Free Press, 1983.
22. Rogers, E.M. The Diffusion of Innovations Perspective. In:Taking
Care ed. N.D. Weinstein, pp79-94, New York, Cambridge University
Press, 1987.
23. Cleary, P.D. Why people take precautions against health risks,
In:Taking Care ed. N.D. Weinstein, New York, Cambridge University
Press, 1987.
24. Suchman, E.A. Preventive health behavior: A model for reseach on
community health campaigns. Journal of Health and Social Behavior,
8:197-209, 1967.
25. Gottlieb, N., Baker, J. The relative influence of health beliefs,
parental and peer behaviors and exercise program paricipation on
smoking alcohol use and physical activity. Social Sciences and
Medicine, 22:915-927, 1986.
26. Langlie, J.K. Social networks, health beliefs, and preventive
health behavior, Journal of Health and Social Behavior, 18:244-260,
1977.
27. Muhlenkamp, A.F., Sayles, J.A. Self-esteem, social support, and
positive health practices. Nursing Research, 35:334-338, 1986.
28. Manoff, R.K. Social Marketing. New York, Praeger, 1985.
29. Lefebvre, R.C., Flora, J.A. Social marketing and public Health
intervention. Health Education Quarterly, 15:299-315, 1988.
30. Ferguson, M.A., Valenti, J.M. Risk-taking tendencies and radon
messages: a field experiment testing an information processing model
for risk communication. Seventy-first Annual Meeting of the
Association for Education in Journalism and Mass Communication.
Portland, OR July 2-5, 1988.
31. Smith, V.K., Desvousges, W.H., Fisher, A., Johnson. F.R.
Communicating Radon Risk Effectively: A Mid-Course Evaluation. EPA-
230-07-87-029, US Environmental Protection Agency, 1987.
-------
Figure 1 : Health Belief Model (adapted from Rosenstock, 1974)
-------
1-4
PIIBI.TC POT-ICY CONSIDERATIONS AND THE DEVELOPMENT OF A CODE
FOR THE CONTROL OF RADON IN RESIDENCES
by: Mike Nuess and Stan Price
Washington Energy Extension Service
Washington State Energy Office
Spokane, WA 99201
ABSTRACT
Building codes that address radon control in residential buildings are a
relatively new development in the larger trend toward increased efforts to
understand and control indoor air quality. A residential radon construction
standard has been developed in the Pacific Northwest region of the United
States. The Northwest Residential Radon Standard (NRRS) seeks to provide a
measured public policy response that is commensurate with current knowledge of
both the health risk and the state of building science. This paper reviews the
range of potential public policy responses available to deal with radon as a
public health problem, describes the policy framework upon which the NRRS is
structured, and explains the development process.
Time and budget constraints limited the scope of the NRRS to identifying
that minimum set of measures necessary to reliably achieve radon reductions
wit--hoiif impa-lT-lny xt-rnnt-nrai integrity, capability to control other indoor air
pollutants, occupant comfort, or energy efficiency. Though it looks more
favorably at measures that enhance the linkages between durability, indoor air
quality, and comfort; it does not require them unless they are part of the
minimum set of requirements necessary for radon control. The NRRS, then,
serves to provide a useful interim step toward the larger goal of a systemic
approach.
INTRODUCTION
The NRRS was developed under the auspices of the Washington State Energy
Office with funding support from the Bonneville Power Administration(BPA), a
federal regional power-marketing agency.
Radon is an indoor pollutant requiring a different policy response than
some other indoor contaminants. As an external pollutant source, radon is
dependent on certain aspects of building science for control. There is a need
for governmental intervention to increase public awareness of the issue,
encourage voluntary action by individuals, and create the opportunity for
individuals to live free of high radon exposures.
Though the NRRS looks more favorably at measures that enhance the linkages
between durability, indoor air quality, and comfort; it does not require them
unless they are a part of the minimum set of requirements necessary for radon
control. As a result, the potential for optimizing net system performance and
cost is not impaired, but it is also not realized by the NRRS. Better control
of radon is possible, but it requires broader dispersion of already available
-------
information, further development of technical support, supportive changes in
other building codes, and the different emphasis of a whole systems approach.
THE REGIONAL CONTEXT
The Pacific Northwest region encompasses several states in the
Northwestern United States. The region is blessed with a large hydroelectric
resource which historically provided abundant low-cost electricity. In the
past decade the region has taken aggressive steps to preserve its hydroelectric
resource and avoid the cost of new electrical generation capacity. A major
component of this effort has been the acquisition of energy efficiency in
buildings — a conservation resource.
In 1980, the U.S Congress established the Northwest Power Planning
Council, a regional body mandated to develop a regional plan for ensuring
adequate supplies of electrical energy. The initial plan (and subsequent
revisions) have emphasized conservation as the most cost effective resource in
the region (1).
The Power Council's plan encourages the Bonneville Power Administration —
the agency which manages and distributes much of the region's electricity — to
pursue the conservation resource aggressively. This region has long been
recognized for the pioneering work of BPA, both in the transmission of
electricity and for the development of public power in the United States. More
recently, the BPA has pioneered the development of conservation resources. It
is now estimated that BPA has spent $1 billion (U.S.) on conservation programs,
purchasing electrical energy savings at an average cost of $.02-.03 per Kwh
saved (2).
Over the past decade, BPA has supported (through participating
publicly-owned electric utilities) the weatherization of homes that use
electricity for space heating. In about 1980, as a component of its
weatherization activities, BPA began to study indoor air quality in homes.
Initially, restrictions on available weatherization measures were imposed
pending an Environmental Impact Statement. Then, in 1984 the EIS was
completed, restrictions were softened, and indoor air quality information was
provided to all program participants (3).
Radon testing was initiated as part of BPA's indoor air quality effort.
Participating electric utilities tested residences throughout their service
territories for radon levels. Measurements were made with alpha track monitors
for a minimum of three months during the heating season. The result is one of
the largest data sets ever collected on radon levels in residential buildings.
Over 32,800 residential sites in approximately 400 townships were measured in
Oregon, Washington, Idaho, Montana, and Wyoming (see Figure 1). The average
measured radon concentration in roughly one-half of the 400 townships was
greater than 37 Bq/m3 (1 pCi/1) . None of the townships had an average measured
radon concentration at or below 7.4 Bq/m3 (.2 pCi/1), the new long-term
national goal enacted by the U.S. Congress. A few areas of the region show a
large number of homes with elevated radon test results. Notable, is the
Spokane River Valley region on the border of Washington State and Idaho. In
the City of Spokane, Washington nearly half the homes tested at levels above
150 Bq/m3 (4 pCi/1) .
As part of the Pacific Northwest's aggressive pursuit of energy efficiency
savings, the Northwest Power Planning Council developed the Model Conservation
Standards (MCS) for the construction of new buildings. These standards require
higher insulation levels in the building envelope, tighter building
construction to reduce air leakage, ventilation provided by mechanical systems,
and certain indoor air pollutant control measures. Roughly 25 building code
-------
jurisdictions in the region have adopted the MCS.
The MCS are the first adopted and enforced standards in the U.S. that
begin to address radon. In addition to specific requirements for sub-slab
gravel and
> 444
407-426 i
370-389::
333-352::
296-315::
259-278::
Bq/m3 222-241 :
185-204 : j
148-167 ' '
111-130
74-93
37-56
0-19
0 2000 4000 6000 8000 10000 12000
Number of Houses
Fjourft 1 platribm-lnn of 32.885 Radon Measurements in the PNW
crawlspace ventilation, the MCS contain an appendix which specifies technical
measures to be incorporated into certain residential buildings.
These standards implicitly recognize that more stringent energy codes do
not necessarily create elevated radon levels. In fact they may provide
opportunities to decrease the probability.
In 1987, BPA became interested in the development of a model radon code
for new residential construction. In the summer of 1988, BPA contracted with
the Washington State Energy Office's Energy Extension Service to research and
develop a model radon code.
The Washington Energy Extension Service(WEES) has had an active public
education program on indoor air quality for the past decade. When radon became
an issue of public concern, WEES had been able to respond quickly with
educational services. In a one year effort WEES developed the Northwest
Residential Radon Standard.
A REGULATORY APPROACH - LOOKING FOR PRECEDENT
The task of determining an appropriate public policy response to the
public health issue of radon presents interesting challenges. As a naturally
occurring indoor air pollutant that largely originates from outside the
building, radon is categorically different from many other indoor contaminants.
it is not generated by occupant activity and it is not responsive, in large
part, to behavioral adjustments by the occupant. In this light, radon appears
as a more appropriate pollutant for some level of regulation.
Alpha-track measurements from a minimum of
three winter months to one year. Data from
BPA. Distribution of test sites weighted toward
large participating utilities in certain areas.
January 1989.
-------
o
o
0.
(Q
O
X
£
a
Q.
c
o
5
O
<
3
e
4>
E
c
b
a>
>
o
O
®
c
c
s
(0
3
O
.2
«>!
1 9
ac qc
si
£
>, o
£
«
e
o
a
m
•
cc
3
>
€
e
I
s
3
E
i 8
O «
te .¦£* 3 S «
ll II1
£ cr £ « 5
c
§
CO
S>
©
X
Ul
J?
J
§
c
2
CL
.1
Flgura 2.
Public Health Pnllnv Options In t.h» tlfi
The range of policy options for addressing public health threats is quite
diverse (see Figure 2). At one end of the continuum, society does nothing.
-------
This, of course, maximizes individual freedom, but may do little to protect the
public good. At the other end of the continuum is the more draconian measure
of quarantine, obviously reserved for only the most severe of public health
threats. A number of potential policy responses exist along this continuum.
Some possible responses include funding research, public education, expanding
administrative efforts within existing regulations, moral suasion, and various
degrees of restrictive rules and regulations (4).
There are multiple levels of governmental response to health issues, and
policy responses vary in the United States for different public health issues.
It is illustrative to look at some health issues in light of the governmental
response. The response to saccharin use and tobacco smoking relies almost
exclusively on personal choice and public education (Though in the case of
tobacco smoke, local communities are becoming more aggressive in regulating
where the activity can take place). Childhood vaccinations and AIDS are two
well publicized health issues that have received a stronger regulatory
response: Though in both cases there remain elements of personal choice, the
public health response has relied heavily upon administrative regulations (e.g.
public schools require evidence of vaccination before enrolling children) and
moral suasion. A most notable recent example was a produce tampering case
(spring 1989) in which two Chilean grapes were found to be tainted with
cyanide. The public health response was swift and aggressive: all Chilean
produce was pulled from retail shelves throughout the United States.
THE SHARED RESPONSIBILITY VALUE
There has been very little regulatory control of radon in buildings in the
United States. As such, the project required at the outset many value choices
about both the technical structure and policy framework of the code. The fact
that a code would be developed at all assumed the problem warranted
intervention by government, but at what point in the regulatory continuum?
WEES assumed public health in the area of radon is a shared responsibility
between individuals and government. Unlike outdoor air quality (where the
costs and benefits of clean air cannot be rationally apportioned to an
individual, and attained through voluntary individual action); the benefits
that accrue to the individual from voluntary actions to maintain healthy indoor
¦it are clear (5) . WEES assumed it was the role of government to empower
individual choice by providing:
• education about radon health effects, measurement, control, etc.
• access to necessary resources by nurturing the development of necessary
technology.
• quality control through industry coordination and regulation.
• regulation necessary to provide the opportunity to live in healthier
indoor air (including construction standards).
WEES assumed the individual's freedom of choice should be preserved to the
extent possible, and that it was the individual's responsibility to:
• recognize the value of healthy indoor air.
• choose whether or not to live in it.
Therefore, the NRRS was structured as a governmental intervention that
enables voluntary action. It regulates the building in order to enable radon
control and preserve the individual's option to live in a healthier indoor
environment. It stops short of requiring an individual to test or mitigate in
order to continue to live in that environment. Because it is a construction
standard, its scope is very focused and it addresses but one of several
important regional and national issues with regard to radon and health.
-------
SCIENTIFIC UNCERTAINTIES AND POLICY MOMENTUM
The issue of radon as an indoor air contaminant has a relatively brief
history. In the U.S., an ever-increasing understanding of radon as a threat to
public health has generated governmental activity at federal, state, and local
levels.
At the federal level, the U.S.EPA issued action guidelines to the public
for mitigation activity based on radon test results. As research more firmly
established radon as a public health threat and as the public's awareness of
the radon issue increased, governmental response also increased. The U.S.
Environmental Protection Agency recommended the testing of all homes. The U.S.
Surgeon-General issued a report on the health threat presented by radon. He
encouraged all Americans to test their homes. The U.S. Congress recently
passed legislation that provided funds to the states for radon programs,
established a long-term national goal to lower radon levels in buildings to
outdoor levels (7.4 Bq/m3), and mandated the development of National Model
Construction Standards by June 1990.
Despite increased levels of governmental activity in the area of radon,
some uncertainty remains:
• estimates of the level of risk to human health at various exposure
levels still vary.
• measurement protocols need improvement.
• we do not have long-term experience with the techniques of radon control
and several technical questions remain unanswered (and probably
unasked).
It is within this environment of scientific uncertainty and governmental
desire to respond to the perceived threat, that the NRRS had to be developed.
WEES is confident that techniques required by the NRRS represent a
reasonable and appropriate "good practice" standard at this time. It is
redemonstrably evident that the radon control approaches required by the NRRS
are very effective. Several radon mitigators utilize these techniques in the
mitigation of existing residential buildings and guarantee their performance.
However, it should be clearly understood that new information will likely
emerge that results in the need for these measures to change.
THE VALUE OF A SYSTEMS APPROACH
The control measures required by the NRRS are intended to represent the
minimum set of measures necessary to reliably achieve radon reductions without-.
impairing structural integrity, capability to control other indoor air
pollutants, occupant comfort, or energy efficiency. These measures are
designed to mesh with current building practices, materials, and building
codes. Hence, the NRRS requires:
1. practical techniques that reduce the number of openings available for
soil gas transport to the indoor air.
2. a pressure difference control system designed to override other
house/soil pressure differences contributing to soil gas transport.
However, it does not require:
1. as-tight-as-possible building envelope construction.
2. mechanical ventilation in all residential occupancies.
3. decoupling of all combustion appliances from the indoor air.
4. attention to pressure difference control in design of HVAC systems.
-------
These additional measures serve multiple purposes and cannot always be
justified for one purpose alone. For example:
• Mechanical ventilation (properly installed) would contribute to further
reduction of indoor radon but its contribution is more than an order of
magnitude less than that of the sub slab depressurization system
capability required by the NRRS. Yet mechanical ventilation would
enhance the control of other indoor air pollutants and increase occupant
comfort, if installed in a tight house.
• Envelope tightness could reduce the volume rate of soil gas transport by
enhancing pressure difference control capability at minimal energy cost.
It would also enhance mechanical ventilation effectiveness, moisture
control, comfort, and energy efficiency.
These and other measures could contribute significantly to further radon
reductions. However, they would serve multiple purposes and the costs should be
appropriately proportioned. A reciprocal effect is that part of the cost of
the required radon control measures, such as substructure/crawlspace sealing
and sub slab depressurization, could be charged to comfort, control of other
indoor air pollutants, control of moisture(several tons/heating season from the
soil), and control of other soil gas pollutants. (Jim White, of Canada
Mortgage and Housing Corporation, reported that garbage gasses have been
measured several kilometers away from land fill sites. Also some bacteria,
fungi, and viruses found in soils can produce serious health problems (6)).
A whole systems approach which attempted to optimize residential buildings
for durability, health and safety, comfort, and energy efficiency would include
at least the additional measures listed above. Such an approach would further
rationalize the cost of radon control. The increased durability, safety,
comfort, and energy efficiency could increase the net value of residential
buildings.
Because of these limitations the proposed NRRS is not an optimal standard.
Better control of radon is possible, but it requires broader dispersion of
already available information, further development of technical support,
supportive changes in other building codes, and the different emphasis of a
whole systems approach.
WEES is encouraged to think that the NRRS serves to provide a useful step
toward the larger goal of a systemic approach.
INTRA-REGIONAL VARIABILITY
Radon exposures in some areas of the Pacific Northwest are relatively low:
in some areas relatively high: in some areas unknown. It was an original
intention that the NRRS would be offered to the region for optional adoption by
local jurisdictions. Jurisdictions that were sufficiently concerned could adopt
the NRRS.
FLEXIBILITY - THE ROLE OF A DUAL PATH STANDARD
The national model codes of the U.S., such as the Uniform Building Code,
are performance codes. Performance codes specify levels of performance rather
than specific materials or procedures. You must attain the end goal but are
free to choose the means of attainment. Performance codes allow flexibility,
cost optimization, and readily allow the development of hew and improved
materials and systems.
-------
On the other hand/ a prescriptive standard requires installation of
certain materials and systems. It specifies a path that must be taken.
Prescriptive standards/codes are simpler and easier to follow, but they lack
the flexibility of performance codes, as well as the potentials for innovation
and cost reduction.
The Council of American Building Officials (CABO), an umbrella
organization of the three national model code organizations, distributes the
CABO One and Two Family Dwelling Code. It is a prescriptive code. According to
Dick Kuchnicki, President of CABO, the One and Two Family Dwelling Code was
developed as a response to builders' requests for a prescriptive standard for
residential construction (7). Currently the National Association of
Homebuilders favors a prescriptive standard for radon control if and only if it
relieves builder liability. However, not all builders concur. Some would like
to see a performance standard, because it allows them the flexibility to
determine the most cost effective path.
Jim Gross, Deputy Director of the Center for Building Research, of the
National Institute of Standards and Technology, has encouraged a dual path
standard: a performance standard with the option of specified measures "deemed
to satisfy" the standard (8). This seems the most practical approach. The
proposed NRRS follows this dual path pattern.
The NRRS seeks to provide increased protection for all new and
significantly remodeled residential occupancies in any jurisdiction of the
Pacific Northwest that chooses to adopt it. Xt seeks to limit exposure to
indoor radon for occupants by requiring for every such occupancy either:
• demonstration of post-construction tested indoor radon levels at or
below 150 Bq/m3 (4 pCi/1), or
• installation of certain specified materials and systems during
construction that reduce the potential for elevated indoor radon and
establish the capability to further reduce radon levels should the
owner desire.
Option 1 (Chapter 3 of the NRRS) is a performance requirement. If the
building does not meet the performance specification it must be modified until
it does. There are no specified control requirements to be met during
construction. It allows both flexibility and the demonstration and use of new
and different approaches to controlling radon.
Option 2 (Chapter 4 of the NRRS) specifies certain preanripH
requirements, primarily substructure and crawlspace sealing, and the rough-in
of a sub slab depressurization system. If the prescribed measures are correctly
installed there are no future responsibilities for radon control.
NEED FOR A LONG-TERM MEASUREMENT TEST
The performance path of the NRRS requires verification that the
performance goal has been reached. The intent is to ensure, within a reasonable
level of certainty, that a building will perform as required.
The EPA's Interim Protocols for Screening and Follow-up Radon and Radon
Decay Product Measurements state that "The EPA does not recommend taking any
significant remedial action on the basis of a single screening measurement
(9)." The screening measurement is a short term test.
The short-term test can be a reasonably accurate measure of the radon
levels during the actual test period, but the range and period of variation is
-------
too great to enable a reasonably accurate measure of the long-term average
radon levels. Arthur Scott of American Atcon Inc. has suggested that the
decision level of a short-term (3 day charcoal for example) radon test is
really very different from that of a long-term test (6 to 12 month alpha
track), and that short-term tests are not being interpreted correctly.
Short-term tests cannot predict long-term averages. He indicated that if a
long-term average radon exposure is really 185 Bq/m3 (5 pCi/1), then the
probability of a short-term test result of 37 Bq/m3 (1 pCi/1) is the same as
the probability of a short-term test result of 750 Bq/m3 (20 pCi/1) (10) .
Currently the short-term test is being misinterpreted in many sectors.
William Ethier, an attorney for the National Association of Homebuilders,
recently suggested at the National Radon Conference (Cincinnati, March 1989),
that utilization of a short-term test to imply that radon levels are below 150
Bq/m3 and therefore acceptable, could provide reasonable grounds for a claim of
fraud or misrepresentation if a long-term test later showed levels over 150
Bq/m3. According to Ethier, NAHB takes the position that short-term tests
should not be part of a real estate transfer contract (11).
In its report to the U.S. House of Representatives, the U.S. Committee on
Energy and Commerce noted concern "about people making decisions not to
mitigate based on low readings from short-term radon tests. Accordingly, the
Committee expects EPA to evaluate the appropriate use of results from short-
and long-term tests by the public. In particular, the Committee expects EPA to
consider whether the Agency should recommend that only results from long-term
tests should be used (12)
The EPA screening protocol would be inappropriate for the NRRS, because of
its reliance on short-term measurements. One NRRS reviewer suggested that a
separate measurement protocol be developed, rather than rely on the EPA
screening protocol. Another reviewer cautioned that developing a protocol
outside that of the EPA might make it difficult to compare the results to
measurements elsewhere.
A longer term measurement is necessary in order to attain a reasonable
estimate of the building's actual performance and to avoid cheating by "smart"
testers, who could affect results by coordinating test periods with rainfall,
weather systems, and other factors. This requires addressing the additional
difficulty of testing after occupancy. However there is a positive side to
this: the occupant has the least incentive for fraud (unless he or she is
preparing for resale).
The NRRS requires a long-term test by specifying adherence to certain EPA
follow-up measurement protocols. According to the EPA's Interim Protocols for
Screening and Follow-up Radon and Radon Decay Product Measurements, "The
purpose of the follow-up measurement is to estimate the long-term average radon
or radon decay product concentrations in general living areas with sufficient
confidence to allow an informed decision to be made about risk and the need for
remedial action (13)."
SHOULD WE REQUIRE MONITORING FOR ALL RESIDENTIAL BUILDINGS?
Currently the NRRS requires monitoring only for the performance path
because it is the responsibility of the builder to meet the performance
standard. It does not require monitoring for the prescriptive path because the
builder completes his or her responsibility upon complying with specifications
which are "deemed to satisfy" the standard. Once the builder has met the
requirements of this standard his or her responsibilities have been completed.
At this point the responsibility for addressing indoor radon is passed to the
owner. The proposed NRRS stops short of governing the owner or occupant.
-------
Neither compliance path guarantees that, for any given residential
building, future indoor radon levels will be below 150 Bq/m3. If a building
has conformed to the prescriptive path, the owner or occupant will not know
radon exposures until he or she tests. If a building has conformed to the
performance path, there is no certainty that future events will not alter
long-term average radon levels. Periodic measurements over the course of the
useful life of any building, built to this standard, will be necessary if
knowledge of radon exposures is desired. For all governed buildings, the NRRS
requires measures to:
• Inform all future occupants of the radon control measures taken.
• Strongly encourage them to test for radon.
• Provide them access to further information about health effects,
testing, and mitigation.
Some NRRS reviewers recommended monitoring all new residential buildings.
Other policy approaches were offered. For example, a member of EPA's National
Radon Standards and Codes Work Group who has been involved in several
mitigation demonstration projects, expressed a concern that the only workable
way to reduce radon exposure in buildings is to have a standard that is at once
both a performance and a prescriptive standard. Buildings would be built to
specifications, tested, then mitigated if necessary. He felt that quality
control was so essential, yet so lacking, that this approach might be
necessary.
Testing following construction or remedial action, plus continued testing
for several years afterward, is warranted by the lack of knowledge of:
• the short-term effects of specific measures in specific houses.
• the longevity of the effects of those measures.
WEES concluded that such follow-up testing should be encouraged (perhaps funded
for research purposes), but not required.
THE NEED FOR EDUCATIONAL SERVICES
There is little system-wide coordination within the building industry.
Many builders receive training on the job and then must make do with what they
have learned from this rather local sphere of influence. There is significant
variation in construction methods by both geographical area and climate.
In addition, builders must survive in an economic milieu in which emphasis
on first costs forces builders to resist any increase in housing costs.
Builders face a forest of regulation and will, in many cases, be less than
eager to comply with additional regulations.
Educational and technical support services will be of significant value.
While no radically new construction techniques are required, many are new to
large portions of the residential sector.
An example is the Soil Gas Retarder Membrane required by the NRRS. Many
reviewers supported its inclusion, considering it feasible and reasonable.
Others were concerned about both the difficulty and cost incurred by having
this technique as a requirement, it has become clear from several discussions
that perceptions about this issue vary widely within the building trades.
Successful (and unsuccessful) experiences with the sub slab membrane are
closely linked to perceptions about correct concrete practices, and these
perceptions also seem to vary widely.
-------
More stringent aggregate specifications and sealing techniques may also
require educational services in the residential sector.
THE NRRS DEVELOPMENT PROCESS
Decisions about public health risks (in this case radon) can be extremely
complicated. They involve elements of risk assessment, risk management, and
risk communication. All too often, difficult decisions about risk assessment
and risk management are made remotely by experts, then poorly communicated to
the public. Often the result is conflict, with experts feeling misunderstood
and the public feeling misused. Often both are right. Conflicts about health
risk issues usually contain the underlying issues of equity and control.
"Public participation" is usually too late, and does not involve the kind of
information and power sharing necessary for the realization of enduring
policies. Risk communication, with the goal of an actively concerned public,
and within a context of real openness to public input, is vitally important.
It may be difficult but it is both possible and necessary.
A good process can serve to align public perceptions with the perceptions
of the scientific community. It can serve to eliminate the inappropriate
extremes of either panic or apathy. It can empower a community with the sense
that it can take charge and address the issues that confront it.
PARTICIPATORY PROCESS - SEQUENCED INPUT
The process for developing the NRRS was very participatory. Input was
solicited from a diversity of economic sectors including realtors; builders and
builder associations; technical specialists and generalists in the fields o£
building science, radon, and ventilation; consumer protection organizations;
energy utilities; state and federal agencies; code organizations; and research
organizations.
While broadly solicited, the input was sequenced: technical input was
solicited first and the range of known technical solutions identified.
Technical specifications had to meet criteria for control effectiveness, ease
of implementation by typical tradesmen, availability of materials, cost,
compatibility with comfort, and compatibility with other indoor air pollutant
control techniques. Legal and policy related input followed.
CHRONOLOGY - INITIAL SCOPING
The effort began in June 1988. A literature search was conducted.
Researchers, mitigators, and policymakers who were known to have radon-related
experience were contacted by telephone. By July, 1988 referrals to additional
resources had become very circular, and a sense of closure with regard to
available national resources had developed.
The initial effort was very broad. Persons contacted were asked to
identify and prioritize the radon issues that they perceived to be most
important. They were then asked more specifically to identify those issues they
thought were important to the development of construction standards for new
residential construction, if they were aware of any efforts to develop
construction standards, and if they knew of any standards already in place.
The U.S. Environmental Protection Agency is nationally recognized as the
lead Federal resource for addressing indoor radon, and WEES assumed that all
local, state, and regional efforts to address radon, particularly efforts to
develop construction standards, would include communication with EPA regional
offices. With assistance from EPA Region 10, all EPA regional radon
-------
representatives were contacted. They were all helpful, identifying issues of
concern, key people, and any developing construction standards in their areas.
Inquiries were also directed to the National Model Code organizations and the
National Association of Homebuilders.
In August 1988 WEES visited and interviewed several radon researchers,
mitigation contractors, and policymakers who were particularly knowledgeable
about the techniques, costs, and policy issues pertinent to new residential
construction. This included persons representing the National Association of
Homebuilder's National Research Foundation; U.S. Geological Survey; U.S. EPA
New Construction Division; New Jersey Department of Environmental Protection;
Princeton University Center for Energy and Environmental Studies; Fairfax
County, Virginia, a local jurisdiction actively addressing a known radon
problem; Camroden Associates, a major radon research contractor; Garnet
Homes, a large construction company voluntarily incorporating sophisticated
radon control measures in all new home projects; R.F. Simon Co. and Buffalo
Homes, two home construction contractors with significant radon mitigation
experience.
WEES deliberately avoided the formulation of any specific code structure
or provisions until the initial three months of research had been completed.
DEVELOPMENT CHRONOLOGY - TECHNICAL AND LEGAL REVIEW
On October 1, 1988 an initial draft of the NRRS was completed and
circulated for technical review. Circulation for legal review followed. More
than 35 technical reviewers contributed comments about the initial draft. They
included persons from the EPA; national research laboratories; university
researchers; private sector builders, contractors, radon mitigators,
tradesmen, engineers, architects and product suppliers; code officials and
code organizations; builder associations; state energy offices; BPA; the
Northwest Power Planning Council. The time allowed for the technical and legal
review comment period had to be extended considerably longer than originally
anticipated in order to obtain important and valuable review comments. The
need for a longer review period may be in part due to the unanticipated
intensity of activity in the radon industry in 1988, which included a national
symposium, and the passage by the U.S. Congress on October 28, 1988 of the
Indoor Radon Abatement Act which set a new national goal of indoor radon levels
no higher than outdoors.
In January 1989, The U.S. Environmental Protection Agency asked WEES to
contribute to EPA's effort to develop Model Construction Standards by June 1990
and partake in a National Radon Standards and Codes Work Group. The group
included persons representing the national Model Code Organizations (ICBO,
SBCCI, BOCA, and CABO), U.S HUD, National Institute of Building Sciences,
National Institute of Standards and Technology, National Association of Home
Builders, Canada Mortgage and Housing Corporation, members of an ASTM committee
on radon, and representatives from states actively working on radon codes. In
February WEES presented an introduction to the first draft of the NRRS to that
group and received several constructive comments.
DEVELOPMENT CHRONOLOGY - POLICY REVIEW
The second draft of the NRRS was distributed March 30, 1989. It was
circulated to a Policy Review Committee consisting of state and local officials
in the general government, building code, and public health areas; policy level
representatives from BPA, the Northwest Power Planning Council, utilities, and
the shelter industry; the EPA National Radon Standards and Codes Work Group;
the National Institute of Standards and Technology; the Canada Mortgage and
Housing Corporation. The second draft was also recirculated to technical and
legal reviewers as a courtesy copy.
-------
The first draft of a generic Implementation Plan was completed in May and
circulated for review by a local advisory committee. The Implementation Plan
is a guidance document intended to assist local jurisdictions with considering,
adopting and implementing the NRRS. The plan seeks to provide the conceptual
framework for a reasonable, equitable, and informed process for consideration
of the NRRS. It is not meant to encourage adoption of the NRRS. The intent is
to encourage and enable a good choice.
The final Implementation Plan and final draft of the NRRS were completed
in June 1989.
The work described in this paper was not funded by the U. S.
Environmemntal Protection Agency and therefore the contents do not necessarily
reflect the views of the Agency and no official endorsement should be inferred.
ACKNOWLEDGEMENT S
The authors deeply appreciate the willingness of the many capable (and
therefore very busy) people who generously provided time, effort, encouragement
and support to the development of the NRRS.
-------
REFERENCES
Northwest Power Planning Council, Portland, Oregon, USA. "Northwest
Conservation and Electric Power Plan," Volume 1, 1983.
Northwest Power Planning Council, Portland, Oregon, USA, "Northwest
Conservation and Electric Power Plan," Volume 1, 1986, p. 6-1.
Bonneville Power Administration, P.O. Box 3621, Portland, Oregon, USA.
"Final EIS: The Expanded Residential Weatherization Program," August,
1984. DOE/EIS 0095F.
Spengler, John D. and Sexton, Ken. "Indoor Air Pollution: A Public
Health Perspective," Science, Vol. 221, No. 4605, July 1983, p.14.
Spengler, John D. and Sexton, Ken. "Indoor Air Pollution: A Public
Health Perspective," Science, Vol. 221, No. 4605, July 1983.
White, Jim H. "Radon—Just Another Soil Gas Pollutant?" Presented at
81st Annual Meeting of the Air Pollution Control Authority, Dallas TX,
USA, June 1988, p.5.
Kuchnicki, Dick. Comments to National Radon Standards and Codes Work
Group of the U.S. EPA , Washington D.C., February 1, 1989.
Gross, James. National Institute for Standards and Technology,
Gaithersburg, Maryland, USA. Personal Communication.
U.S. Environmental Protection Agency, "Interim Protocols for Screening
and Followup Radon and Radon Decay Product Measurement," No.
520/1/86-014. January 1987.
Scott, Arthur. American Atcon Inc., Box 164, Mississuaga, Ontario,
L5L3A2, Canada. Personal communication.
Ethier, William. Comments at the March 1989 National Radon Conference,
Cincinnati, Ohio, USA.
Coranittee on Energy and Commerce, Report (to accompany H.R.2837) to the
U.S. House of Representatives, Report 100-1047, October 4, 1988.
U.S. Environmental Protection Agency, "Interim Protocols for Screening
and Followup Radon and Radon Decay Product Measurement," No.
520/1/86-014. January 1987.
-------
APPENDIX - ORGANIZATIONAL OUTLINE OF THE NORTHWEST RESIDENTIAL RADON STANDARD
Chapter 1
ADMIN.
&
ENFORCEMENT
Title Intent & Scope
Materials & Equipment
Alternate Materials & Methods
Plans & Specifications
Enforcement, Inspections, & Fees
Validity
Violations
Liability
Chapter 2
DEFINITIONS
Chapter 3
PERFORMANCE'
PATH
Chapter 4
PRESCRIPTIVE
PATH
Scope
Financial Assurance Requirement
Measurement
Mitigation
Verification of Mitigation
Notification
Written Description
Scope
Floors in Contact With the Ei
Masonry Block Walls
Concrete Walls
Crawlspaces
Combinations of Floor Assemblies
HVAC Systems
Sub-Slab Depreaaurization System
Chapter 8
REFERENCED |0ccuP#nt Notice Document
STANDARDS "l Rafwanced Documents
bulk water drainage
concrete floors
other floor assemblies
passive sub-slab
depressurization
placement
strength
workability
a| air entrapment
curing
control joints
sealing joints/cracks
soil gas retarder
DOCUMENTS
continuous solid course
mortar joints
penetrations
waterproofing
continuous air barrier
i
V
workability
sealing penetrations/
joints/cracks
damproofing
{continuously vented
intermittently vented
plenums
J Juctwork
^condensate drains
sealed pipe
perforated pipe
pipe characteristics
condensation/stack assist
tan location
fan wiring/circuit/label
pipe identification
system description
-------
1-5
RADON CONTRACTOR PROFICIENCY PROGRAM
by: Jed Harrison
Lee Salmon
John Hoornbeek
Dave Price
Lynne Gillette
Gene Fisher
Office of Radiation Programs
U.S. Environmental Protection Agency
Washington, DC
ABSTRACT
The Indoor Radon Abatement Act (IRAA) of 1988 requires the Environmental
Protection Agency (EPA) to develop a voluntary program to evaluate radon
reduction contractors and provide information to the public on proficient
contractors. The EPA established the Radon Contractor Proficiency Program
(RCPP) in response to this directive. Contractors who participate in the
program must pass a written examination, agree to use EPA mitigation protocols
in their work, and meet continuing education requirements. The Agency
includes contractors who meet these requirements in a list of proficient radon
reduction contractors. The list is made available to States and the public
for use in selecting capable radon mitigators. This paper outlines RCPP
requirements, describes how they ensure contractor proficiency, and reviews
the program's performance during its first three months of existence.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
-------
Session A-l:
Government Programs, Policies, and Public Information
Relating to Radon—POSTERS
-------
nmnoR RAnrrn t.euet.s m COLUMBUS AMD FRAMia.TW rnijyrY. OHIO
RESIDENCES- CO»#ffiRCI*T- RimnTMGS. AMD SCHOOLS
By Harry E. Grafton
Columbus, Ohio Health Department
Columbus, Ohio 43215
ABSTRACT
Data is presented for 2 residential radon surveys, one survey of city-
owned buildings, and survey of Columbus Public Schools. The first residential
survey used a volunteer participants and employed a 48 hour activated carbon
measurementi 4425 measurements in the data. The second survey consisted of
120 randomly selected residences in which alpha track detectors were placed
for from 60 to 120 days. A survey of 52 city-owned buildings in which
screening measurements were obtained using activated carbon, alpha track, and
E-PERM radon detectors is included in the data. Also a survey of 25 Columbus
Public Schools in which E-PERM radon monitors were used to obtain measurements
is detailed in the data. More than 72% of the volunteer survey residences
showed screening measurements of 4.0 pCi/L or greater while the random survey
revealed 92% of the residences with radon levels of 4.0 pCi/L or higher.
Schools tested in the survey also showed elevated radon levels with 20% of the
tested structures with an average radon level of 4.0 pCi/L. Work is still in
progress on city-owned building and Columbus Schools. Concludes that any
owner or lessor of occupied buildings in Franklin County, Ohio should perform
screening measurements and should be prepared to also perform follow-up
measurements.
-------
In response to encouragement by U.S. EPA (1) and cognizant of other Ohio
research from Cleveland (2) and Dayton (3) indicating elevated radon levels.
The Columbus Health Department developed a plan to investigate radon levels in
Columbus and Franklin County, Ohio. A three phase procedure was used survey
Franklin County. First, residences were investigated via a short-term
volunteer survey and also a long-term random survey. Second, city-owned
buildings were tested. Third, schools were investigated.
All measurements represented in the data in this paper are screening
measurements. The U.S. EPA recommends that follow-up measurements to
determine the annual average radon level before diagnostic or remedial action
is taken. Care should be exercised in interpreting the data.
RADON IN RESIDENCES
A two-fold approach was taken to determine the radon level in residences
in Franklin County. First, a volunteer survey conducted in cooperation with a
local television station and a large grocery retailer. A second survey,
initiated at the same time as the charcoal survey, used a randomly generated
sample of Franklin County residences.
VOLUNTEER SURVEY, METHOD AND PROCEDURE
The volunteer population was recruited by public service announcements
and news announcements provided by the participating television station.
Involvement by the public was enhanced by non-participating television
stations airing radon information before and during the recruitment period.
Volunteers were asked to purchase a radon detector at their own expense at the
participating retailer conduct the radon test observing the manufacturers
instructions, complete a questionnaire, and return the detector and
questionnaire to the laboratory.
A charcoal detector of the "tea bag" type was provided at a reduced cost
by Ryan Nuclear Laboratories who also provided analysis for the exposed
detectors. The appropriateness of the use of charcoal type detectors has been
established (4). Detectors were provided at the reduced cost for three weeks
during the promotion. A count of the total number of detectors sold was not
obtained, however 4425 detectors with completed questionnaires were returned
over a two month period.
A questionnaire was distributed with each detector sold and volunteers
were required to complete the questionnaire to the best of their knowledge. A
sample questionnaire is found in Figure 1.
Detectors were available for sale at 49 retail outlets within or near
Franklin County. Many detectors were received from outside of Franklin
County. All address were plotted on a map and sorted for their
appropriateness for inclusion in the data. Each address was plotted on a grid
system of half mile (160 acre) squares to enhance the site specific nature of
the data beyond the usual zip code territories.
-------
Figure 1 Sample Questionnaire
OCTECTOA NUMBER.
QUESDONNAIRS
1. Wheralntheliu»dlng*«»theteet«onducled?(Marfconlyone)
( ) CrnHp»o«ft—im»nt ( ) First floor en eoncteieilab
( JFtrit floor o¥erbeeetrwnt ( ) FM floor «Mrorwik«M
( ) Second floor or r
2. W«* in* teet «m etaMd prior to Martng the tatt (In order to etabiltt* the
oondMom)? (Math only on*)
( ) Lett than itt (•) hour* ( ) About twelwe (12) houra or rtwe
1 Do you own and occupy the teeleditruolute? (Mark on# on*)
( ) Own but do not occupy
( ) Occupy bul do not own
( )Omandooeupy
4. Mark tha langth o( th* tact parted. (Mark only on*)
( ) 1 k> 11/2 days, 24-38 houra ( ) 11Sto21«day*.3M#hou»
( ) 21/2 to 31/2 day*. 80-83 hour* ( ) 31/2 to 41/2 day*, 84-108 hour*
5. Doaa th* taatad itructure ban: (Mark only one)
( ) *ha« arrant
( ) A orwMepace under about on»haf or mora of th* MwcXura
( ) A crawlapeee under leee than one-quarter d the uructure end a itato
( )AM)
( ) Aaawtpaceundwloootiononoiiiartoraltieoltuatwoandaljaaiitw*
8. > tha laalad itouctwe haa a biamai*. doaa * haw: (Mark al that apply)
( ) Aumppunv ( ) 8c»d ooncrelo waHi
( ) Block or brick wale ( ) Al awathir wood awl
( )Aitafc floor ( )Adht floor
( ) Any laraa crack or dalacla In alab floor
( ) W«erln«looeely pooling tmough a wal or th* floor
7. Die loaieel le*er you oooupy In your IMng un* aHthln the teeled itruelure:
(Mark only on*)
( ) flamM ( (Second Floor
( )FM Floor ( )TWrt floor or *m*>
a What typia 01 heeling lyitarm wore In uaa during tha laat? (Mark al that
mbM
( ) Central Hydronlc (hot water) ( ) Central forced air
( ) Wood burning llrepleoe ( )Ga* or dl heater
( )B*cMoal ( ) Aettve or paaekw aolar
( )VHoodnowe(lndudliigltaplooalnieit»), unwanted (no oNrrna|f),taiaeeno
( ) No heating lyetem In um during Mating
9. If an air to air heel exchanger wae In uee: (CFM » Air flwr In oiMo teat per
minute) (Mark aH that apply)
( ) Wat bull In (ai part d tha oentraliytom)
( ) Wae In en area (through the wall or window) unit
( ) CFM ntlng beta* 100 CFM
( ) CFM rating 100-300 CFM
( ) CFM rating atxwe 300 CFM
( ) No air to Ht heat exchanger amused during Mdng
1a Cheek My o< the Maadng lerraAondUOM ttut om piMonl In the o*Mm
(between m* main living are* end th* attte) ol th* teated Wucnm. (Markal
ttwtwoM
( ) Flame aMMpar barrier
( ) Any ak regMere
( ) Loar permaaMty point (High qualtyolboae)
( ) Fold-down Main or eny (Inild*) eooaae door/panel
( ) Any oVwriifflllar "partwayi* tram Indoor* to »» ante
( ) Rio—id Ighdng fttturea
( ) Extra eeuMngaOoMi (abowe«andard technique)
( ) CMrmeya that pw through oalingaltaeted area
( ) Any plmitilng/electfoal mna (that can be aeon Indoora)
( ) Noneoltheabo**
11. How old la the teated (tructura? (Mark only one)
( )0to»Ye*ri ( ) 18 to 20 yean
( ) 5 to 10Yeera ( ) 21 Yaar* or older
( ) 10to1SYaare
12. What hoe been dona in tha tan ten years » tha teeted itruoture (or
weatherttattonpuipom? (Mark ril thai apply)
( ) Weather itr^iplng on wktdowe and door*
( )Clo*lngg*e unoer doora to outrtde by thing door i
( )CauMng,glailng around wlndowi
( ) No work done
The questionnaire asked the volunteer to designate the length of the
exposure period for the detector. Data from detectors placed for less than 48
hours or longer than 84 hours were excluded from the data. The questionnaire
also indicated whether the test was conducted in the lowest livable level were
excluded from the data bank. Similarly, respondents failing to keep the house
closed 12 hours prior to testing were excluded.
Unfortunately, no quality control measurements were conducted.
Initially we felt that by far the greatest source of error was in the handling
of the detectors by volunteers. Since our original quality control procedures
checked only the reliability of the laboratory, we felt our quality control
measures would be invalid. In retrospect, we should have conducted duplicate
and blank measurements and introduced them into the sample stream at the point
of purchase.
VOLUNTEER SURVEY, FINDINGS
Certainly, we were not expecting the results we received form either the
volunteer survey or the random survey. A large proportion of Franklin County
residences showed elevated radon levels. The average radon level was 8.9 for
the entire County. More than 72% of the residences tested had radon levels of
4.0 pCi/L or above. In addition, 28.9% of the residences had radon levels
above 10.0 pCi/L and 1% of all homes participating had radon levels of 50.0
pCi/L or greater. Data is presented in Table 1 below.
-------
Table 1 - Frequency of Elevated Radon Levels
in Franklin County, Ohio
Survey Area
# of High Avg. %4.0 or %10.0 or %50.0 or
Readings Reading Level Greater Greater Greater
City of
Columbus
2207 320.2 8.6 71.5
24.6
1%
Franklin County
(excluding City
of Columbus)
2218 234.5
9.2
73.8
28.9
1%
Franklin County
Total
4425
320.2
8.9
72.7
26.8
1%
Questionnaire parameters for the volunteer survey were identical to the
response observed with random survey and will be discussed later.
Since residential radon data had been plotted on a grid system rather
than recycling on zip code territories for site specificity, a contour amp of
radon levels in Franklin county could be developed. The Ohio Geological
survey provided the contour map found in Figure 2. This paper will not
attempt to correlate geological features with elevated radon levels in homes.
RANDOM SURVEY, METHOD AND PROCEDURE
Participants were selected from a randomly generated list of Franklin
County telephone numbers provided by Ohio State University Department of
Polimetrics. Prospective participants were contacted by telephone and read a
prepared statement and their participation was strongly encouraged. Calls
were made 7 days a week and from 8 >00 am to 8i00 pm in an effort to reduce
sample bias. Approximately 40% of the prospective participants contacted
agreed inclusion in the survey. No information is available on those who
chose not to participate.
Alpha track detectors were selected for use in this survey because of
their modest cost and their longer deployment period. Detectors were provided
at no cost to participants. The appropriateness of this type of detector has
been established (5). Detectors marketed by Radon Environmental Monitoring,
Inc. were used in the survey. Manufacture's instructions for deployment were
at all times observed.
Detectors were placed and retrieved by trained health department
employees during the period from November 1987 through December 1987.
Department employees also retrieved detectors during the period from February
1988 through April 1988. Exposed detectors were sent to the manufacturer for
analysis.
When detectors were retrieved, a questionnaire was completed by health
department employees. The same questionnaire found in Figure 1 was used in
this survey.
One Hundred twenty detectors were deployed and accompanying
questionnaires, were received, detectors were exposed for a duration of from
60 to 120 days. All duplicates deployed, 4, were within 10% agreement and
both blanks introduced into the sample stream provided readings of 0.0 pCi/L.
-------
Figure 2 - Radon Levels in Franklin County (5)
RANDOM SURVEY FINDINGS
Table 2 Indicates that about 92% of the homes tested
levels ot 4.0 pCi/L or greater. In addition, 43% of the readings «re (ln
excess of 10.0 pCi/L and 11% exceeded 20.0 pCi/L. The average
all homes participating in the study was 11.3 pCi/L.
-------
Table 2 - Range and Distribution of Radon Levels
Number
Range of Responses % Total Sample
Less than 4.0 pCi/L 10 8%
4.0 to 9.9 pCi/L 59 49%
10.0 to 19.9 pCi/L 38 32%
20.0 pCi/L or Higher 13 11%
Total Sample 120
Placement of the detector within the structure is a significant
parameter to examine. Average indoor radon levels from detectors placed in a
basement, and in the first floor over a basement showed close correlation as
Table 2 indicates; However, readings taken in first floor on a concrete slab
and first floor over crawlspace also closely correlate with these readings.
Substantially lower readings were on the second of higher floor not with
standing radon levels obtained were in excess of the EPA action guideline,
including the detectors placed on the second floor or above.
Table 3 - Average Radon Level by Location of
Detector within the Structure
Location Number Average
of Detector of Responses Radon Level
Basement 11 11.90
First Floor over Basement 60 12.44
First Floor on Concrete Slab 25 10.6
First Floor over Crawlspace 20 9.1
Second Floor or Higher 3 6.8
Homes with a basement and homes with a crawlspace and a basement
demonstrated indoor radon levels nearly 50% higher than homes without a
basement. Structural features aside, radon levels exceeded the action
guidelines as demonstrated in Table 3. In each category radon levels were 2
to 3 times the guideline.
Table 5 - Soil Contact Characteristics of Structures
Number Average
Basement Has: of Responses Radon Level
Sump Pump
46
12.2
Block or Brick Walls
86
12.1
A Slab Floor
71
10.5
Any Large Cracks or Defects
in Slab Floor
16
10.1
Water Lines Loosely Passing
Through a Wall or the Floor
9
16.1
Solid Concrete Walls
2
11.6
A dirt Floor
1
8.3
No Response
23
9.0
-------
Type of heating system did not contriku^ °asor oil heater furnaces
radon levels. Even though central forced airJ^d °^vels than other
represented more highly elevated average ind°?5 itTore than twice
heating systems, each type of heating system exhibi e that no homes
the action guideline as shown in Table 6. It is n . . ^ random
using passive or active solar heating systems was s
sample.
Table 6 - Average Radon Level by Type of
Heating System in use During Test
Number Average
Type of Response Radon Level
Central Hydronic
' - • — * ¦« 21
8.5
9.4
Wood Burning Fireplace g.7
Electrical }JJ 8*. i
12.7
10.8
0.0
0.0
10
Woodstove _2
Central Forced Air
Gas or Oil Heater Ti
Active or Passive Solar
None in Use
Another structural characteristic tested Structure.
ceiling between the main living area and ttie att levels than others,
Some openings displayed more elevated indoor radon ^levels^ tnan^ ^
particularly recessed lighting fixtures, ext o£ ^ openings
plumbing or electrical runs, and structures h g respond to this
investigated as seen in Table 7. Participants choosing not to "sponot
question showed the highest average indoor radon level nearly twice me
observed by any other characteristics.
Table 7 - Average Radon Level by Openings in
the Ceiling of the Tested Structure
Number Average
Type nf Responses Radon Level
Plastic Air/Vapor Barrier *
Any Air Registers 21
Low Permeability Paint 3
Fold Down Stairs or Access Panel/Door 38 ^ ^
Any Similar Pathways to the Attic 24 ^'3
13.2
10.3
8.6
8.8
10.4
10.7
Recessed Lighting Fixtures 15
6
Extra Caulking at Joints
Chimneys that Pass through the
Ceiling of the Tested Structure 42
Any Plumbing or Electrical Runs 12 •
None of the Above 27
No Response
4
24.8
-------
Age of structure had little impact on average indoor radon levels albeit
elevated levels. Only structures from 10 to 15 years old varied substantially
and these structures tested at nearly half the level of structures of other
ages. See Table 8. However, even homes 10 to 15 years old tested above the
action guideline.
Table 8 - Average Radon Level by Age of Structure
Number Average
Age of Structure of Responses Radon Level
0-05 11 11.3
5-10 15 9.1
10 - 15 10 5.9
15 - 20 11 11.4
21 or Older 74 12.4
RADON IN COMMERCIAL BUILDINGS
As a result of discovering a high frequency of elevated radon levels in
residences, the Columbus Health Department felt it necessary to investigate
the radon levels in city-owned buildings in an effort to estimate worker risk
from exposure to radon. In the literature, the issue of radon in commercial
buildings had received much less attention than radon in residences. Early
studies suggested that few commercial buildings would exhibit elevated radon
levels (6). Later studies, however, suggested that several states had
discovered a number of commercial buildings with elevated radon levels (7).
METHOD AND PROCEDURE
Protocols for measurement of radon in commercial buildings had not been
firmly established at the time the survey was designed. Measurement devices
were deployed observing interim protocols (8) and interim radon in school
guidelines (9). A measurement device was placed in each space defined as any
room, closet, utility room, storage room, or area of greater than 2000 square
feet.
Four devices have been used to determine the radon level in city-owned
buildings. First, charcoal packets of the pro chek design provided by Air
Chek, Inc. were used in the first 12 buildings. The appropriateness of this
device has been cited previously. In another 6 buildings, alpha track
detectors provided by Radon Environmental Monitoring, Inc. were used. The
balance of the buildings were measured using Electret Passive Environmental
Radon monitors (E-PERM) provided by Rad Elec, Inc. The appropriateness of E-
PERMS for performing measurements of this type has been established (10). In
addition to the above devices, a Pylon AB-S continuous radon monitor was used
as a quality control measure.
Measurement devices were placed and retrieved by trained Health
Department personnel. Activated carbon detectors were deployed for 48 hours,
alpha track detectors were exposed for from 60 to 120 days, and E-PERMS were
placed for from 48 to 72 hours. Manufacturer's instructions for deployment
were at all times observed.
-------
Quality control was maintained by placing 10% duplicate monitors of the
type used in each building. At least one blank detector was introduced into
the sample stream in each building. A continuous radon monitor was used in
approximately 5% of the buildings tested. With the exception of one building,
all duplicate measurements fell within 15% or within 0.5 pCi/L. All control
measurements had readings of 0.2 pCi/L or less. Side by side measurements
using a continuous radon monitor fell within 15%.
Recommended protocol was not observed for 10 of the structures tested.
In 10 firestations, screening measurements were conducted during June and July
without attempting to maintain closed conditions. Since the data did not
differ from fall and winter measurements, they are included in the tables.
FINDINGS
Radon measurements were obtained in 52 city-owned buildings. The
building were selected by no other criteria than their availability to
measure. Table 9 provides pertinent data from the survey. A variety of
structural characteristics are represented with 32 buildings having at least
one level of basement and 20 buildings with a slab-on-grade construction. Of
~?e buildings with basements, 47% had at least one room of 4.0 pCi/L or
higher, while 40% of buildings with slab-on-grade construction had at least
one room of 4.0 pCi/L or greater. The characteristic of basement or slab-on-
grade construction is not a reliable predictor of elevate radon levels.
RADON IN SCHOOLS
The Indoor Radon Abatement Act of 1988 provided incentive to public
h? k°*S to forn,u^ate plans to test for radon in schoolrooms. Because of the
high percentage of Franklin County residences and commercial buildings with
elevated radon levels, there seemed little reason to delay initiating an
effective regimen. Since the Columbus Health Department had exhibit expertise
measuring residences and commercial buildings, a cooperative venture
between the Health Department and Columbus Public Schools was formulated to
efficiently investigate radon levels in school buildings.
METHOD AND PROCEDURE
At the time of this publication, data was available on 25 of the 150
buildings owned by Columbus Public Schools. Nearly all of the 25 buildings
are elementary schools although 2 middle schools and 1 high school are also
included.
Protocol presented in Radon Measurement in Schools (11) was observed in
each of the schools tested. Each space on every building level with ground
contact or below was measured. Measurements were conduct both while students
were present and over weekends while students were not present.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
Table 9 - Radon Measurements in Commercial Buildings
by Structure
# Measurement #>4 oCi/L Average Highest Basement/Slab
9
8
5.0
9.3
B
13
12
12.8
36.6
B
4
2
4.3
6.3
B
5
3
5.2
7.0
B
4
4
5.5
6.6
B
35
4
2.1
7.1
B
4
2
7.4
17.1
B
4
1
2.3
6.7
B
3
10
3.4
9.7
B
3
1
3.6
4.2
B
4
3
3.8
4.1
B
6
6
8.7
12.0
B
3
1
3.8
7.9
B
5
2
3.9
5.2
B
41
0
0.2
1.9
B
6
0
1.9
3.0
B
2
0
2.3
2.9
B
6
0
1.1
1.9
B
13
0
1.1
1.8
B
3
0
1.9
2.5
B
4
0
1.1
2.1
B
3
0
0.5
0.6
B
2
0
0.8
1.3
B
3
0
3.5
3.8
B
4
0
1.4
1.5
B
5
0
0.8
1.0
B
8
0
1.1
1.3
B
5
0
2.4
3.8
B
4
0
1.1
1.5
B
4
0
2.74
3.1
B
3
0
0.7
1.0
B
7
6
33.3
185.0
B
10
1
2.0
4.1
S
3
0
1.3
1.5
S
17
0
0.4
1.2
S
2
0
0.2
1.2
S
1
0
0.1
0.1
S
22
0
0.9
1.9
S
6
0
1.7
3.2
S
7
1
1.0
5.2
S
5
0
0.4
0.6
S
9
0
0.7
3.0
S
20
5
4.0
16.3
B
11
0
0.7
0.9
B
6
2
5.9
23.7
S
59
5
2.6
7.8
S
18
4
5.8
23.7
S
10
3
5.0
21.1
S
9
3
6.4
22.8
S
2
0
2.6
3.8
S
1
0
0.7
0.7
S
21
13
5.9
12.2
S
-------
An important consideration when performing screening measurements in
schools is the operation of the ventilation system. In each school, the
ventilation system was examined by Health Department Personnel and school
engineers to determine worst case conditions-negative pressure at ground
contact levels - and these conditions were maintained throughout the
measurement period. In most cases, the air handlers and bathroom exhaust fans
were kept on throughout the test period.
From Table 10, a significant proportion (25%) of the buildings had
average radon levels of 4.0 pCi/L of above and 10% the buildings with average
radon levels of 10.0 pCi/L or higher. Perhaps of more significance is the
number of buildings with at least one room with elevated radon levels.
Buildings with at least one room above 4.0 pCi/L comprised 48% of all
buildings, 19% had at least one room above 10.0 pCi/L, and 12% had at least
one room above 20.0 pCi/L.
Table 10 - City Owned Buildings with Elevated Radon Levels
Radon Characteristic
Average Radon Level
4/0 pCi/L or Greater
Average Radon Level
10.0 pCi/L or Greater
Highest Radon Measurement
10.0 pCi/L or Greater
Highest Radon Measurement
20.0 pCi/L or Greater
Number
Percent of all
Measurements
13
25%
10%
25
19%
12%
E-PERM radon monitors provided by Rad Elec, Inc. were used for testing
choolrooms. A Pylon AB-S continuous radon monitor was used for quality
ontrol measurements. The appropriateness of E-PERMS for screening
measurements has been cited.
Monitors were placed and retrieved by trained health department
®Ployees. Measurements were during the fall and winter months and continues
at the time of publication.
Quality control consisted of duplicates, blanks, and side-by-side
®easurements. in each tested structure, 10% duplicates and at least 1 blank
Placed side-by-side measurement were conducted in 10% of the buildings
tested.
school findings
No rooms with extremely high measurements were discovered such as that
round in city-owned buildings. The highest measurement recorded in any school
33.5 pCi/L. See Table 11 The number of rooms measure in each school
varied from 6 to 77 excluding quality control measurements.
-------
Screening measurements revealed that 20% of the schools had an average
radon level of 4.0 pCi/L or greater and 4% had an average radon level of 10.0
pCi/L or greater. Table 12 also shows that 72% of the schools had at least
one room with a radon level of 4.0 pCi/L or greater. In addition, 28% of the
schools had at least one room with a radon level of 10.0 pCi/L and 16% had at
least on measurement of 20.0 pCi/L or higher.
Table 11 - Radon Levels in Columbus Schools by Structure
Building High
Building
Number
Total Rooms
JL
Measurement
Averaae
Greater 4.0
Measured
l
33.5
4.5
10
29
2
19.5
3.4
7
27
3
9.3
2.8
7
38
4
17.0
3.6
17
60
5
32.4
10.1
26
27
6
8.5
3.6
16
47
7
5.6
2.8
6
26
8
6.7
3.0
14
62
9
7.3
1.3
6
69
10
29.4
6.3
6
21
11
1.9
0.7
0
47
12
5.2
1.3
2
65
13
1.1
0.7
0
44
14
2.3
1.1
0
45
15
2.3
1.2
0
44
16
3.2
2.0
0
18
17
3.2
1.6
0
28
18
3.6
2.2
0
33
19
31.0
5.6
17
41
20
8.4
3.0
7
47
21
6.9
3.5
7
20
22
11.2
5.9
5
6
23
7.2
1.2
2
47
24
6.2
2.0
11
77
25
4.6
2.2
1
56
With the exception of one school, all duplicate measurements varied no
more than 15% or 0.5 pCi/L. All blank measurement read 0.0 pCi/L. Side-by-
side measurements using a Pylon AB-S continuous radon monitor varied by less
than 15%.
Tampering with monitors has not proved to be a major obstacle. Host
monitors are placed on a teacher's desk when it is situated in an acceptable
location. Tampering is usually evident such as electrets loose from the
shell, closed monitors, moved monitor, and fully discharged monitors. Only
one monitor has been stolen to date.
-------
Table 12 - Columbus Schools with Elevated Radon Levels
Percent of all
Radon Characteristic Number Measurements
Average Radon Level of 5 2®%
4.0 pCi/L or Greater
1 4%
Average Radon Level of 1
10.0 pCi/L or Greater
Highest Radon Measurement 18
of 4.0 pCi/L or Greater
Highest Radon Measurement 7
of 10.0 pCi/L or Greater
1
Highest Radon Measurement 4
of 20.0 pCi/L or Greater
CONCLUSIONS
a large amount of data available for Columbus and Franklin County and
ith other researchers gathering similar data, a reliable picture of the radon
h h^ti0n k®g*ns t0 form. A significant proportion on residences, commercial
lldings, schools have exhibited elevated radon levels. Several
conclusions can be drawn«
1. From 72% to 92% of residences showed radon levels of 4.0 pCi/L or
greater. All residents of Franklin county, regardless of structural
characteristics of their residences, should perform screening
measurements and should be prepared to conduct follow-up measurements.
2. City-owned buildings, 48% of which had at least one room of 4.0 pCi/L
or higher, demonstrate the significant probability elevated radon
levels in commercial buildings. Owners and/or lessors of commercial
buildings should, as a part of occupational hazard assessment, perform
screening measurements to determine occupational exposure.
3. Columbus Public Schools had 72% of the buildings with at least one room
with a radon level of 4.0 pCi/L or greater. School officials in
Franklin County should begin to formulate plans to determine the radon
levels in school buildings.
A special note of appreciation is due to WSYX television and Kroger stores
r their participation in the volunteer survey, to Ohio State University for
assistance in the random survey, to Columbus Division of Facilities Management
th^ guidance in the city-owned building survey, to Columbus Public Schools for
necessary input on building and ventilation characteristics, and to the
hio Geological Survey for their guidance and interpretive insight throughout
Phases of this project.
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the
views of he agency and no official endorsement should be inferred.
REFERENCES
1. U.S. EPA, A Citizen's Guide to Radon. OPA-86-004. U.S. Environmental
Protection Agency Washington D.C., 1986
2. Ghahremani, D.T. Radon Occurrences in Northeast Ohio - An Environmental
Hazard. Ohio Journal of Science. 87>11, 1987
3. Regional Air Pollution control Agency. The Radon Sampling Project Final
Report. Regional Air Pollutant Control Agency, Dayton, Ohio 1988. 75pp
4. Cohen B.L. Comparison of Nuclear Track and Diffusion Barrier Charcoal
Adsorption Methods for Measurement of Radon 222 in Indoor Air. Health
Physics. 50»828, 1986
5. Alter H.W. Fleischer R.L. Passive Integrating Radon Monitor for
Environmental Monitoring. Health Physics. 40>693, 1981
6. Dziuban, J. A. et al.t Residential Radon survey of 14 States: in
Proceedings of the 1988 Sympossium on Radon and Radon Reduction
Technology; NTIS, Springfield, VA. 22161, 1988
7. Cohon, B.L. et al. Health Physics. 47j399, 1984.
8. U.S. EPA. Interim Protocols for Screening and Follow-up Radon and Radon
Decay Product Measurements. EPA 520/1-86-014-1, 1986
9. U.S. EPA Radon Measurements in Schools - An Interim Report. Unnumbered
EPA document for internal use. Date unknown
10. Kattrappa P. et al. An electret Passive Environmental Radon-222 Monitor
Based on Ionization Measurement. Heath Physics. 54>47, 1988
11. U.S. EPA Radon Measurements in Schools« An Interim Report. EPA 520/1-89-
010, 1989
-------
A-I-2
MODEL STANDARDS AND TECHNIQUES FOR CONTROLLING
RADON LEVELS WITHIN NEW BUILDINGS
by: David M. Murane
U.S. Environmental Protection Agency
Washington, DC 20460
ABSTRACT
This paper provides the current status of the Environmental Protection
Agency (EPA) program to develop, "by June 1, 1990¦ Model Standards and Tech-
niques for Control of Radon in New Buildings, as required by Section 30^ of
the Indoor Radon Abatement Act of 1988. The paper includes specific details
on the scope and draft recommendations for implementation of the various
construction standards and techniques contained in the document, and summar-
izes those standards and techniques.
This paper has been
Protection Agency's peer and administrative review policies, and. nas
approved for presentation and publication.
-------
MODEL STANDARDS AND TECHNIQUES FOR CONTROLLING
RADON LEVELS WITHIN NEW BUILDINGS
INTRODUCTION
The purpose of this paper is to provide a status report on the "Stand-
ards and Techniques for Control of Radon in New Buildings." This is the
document required by Section 304 of the Indoor Radon Abatement Act of 1988
and it is to be available for public use by June 1, 1990•
In early February 1989» EPA formed a Radon Standards and Techniques
Work Group to develop the scope and structure of the document. This Work
Group included a broad spectrum of people representing private, government-
al, and building industry organizations. At the same time, EPA was comply-
ing with the IRAA requirement to obtain assistance from organizations in-
volved in developing building standards. We were working with the National
Association of Home Builders/National Research Center to develop a draft
ASTM Standard Guide on Radon Reduction in New Construction, and we devel-
oped a Cooperative Agreement with the National Institute of Building Scien-
ces to review the Standards and Techniques drafts for consistency with
existing publications.
After the first meeting of the NIBS Radon Project Committee, it was
decided to merge the efforts of the EPA Work Group and the NIBS Project
Committee and use the existing NIBS consensus review mechanism as the pri-
mary method for developing and reviewing the Draft Standards and Techniques
for Radon Control in New Buildings. The Project Committee met in July,
August and October of 1989 and for a final draft review in January 1990.
Based on comments from NIBS Committee members and other interested parties
at these meetings, a draft of the Standards and Techniques was produced.
The draft is currently being processed for publication in the Federal Reg-
ister for public comment. A final version of the document should be avail-
able for public use by the June 1st deadline. In describing the basic
features of the document at this time, it should be noted that it is still
in draft status and that it may undergo some additional changes and refine-
ment as a result of public comment.
SCOPE
The Model Building Standards and Techniques will be applicable to all
new construction intended for human occupancy. The standards and tech-
niques are also applicable to additions made to the foundations of existing
occupied buildings or when modifications are made to their central air
handling systems. Builder of high-rise residential and non-residential
-------
buildings are cautioned that these standards and techniques should be con-
sidered only as guidelines pending further research and we do not recommend
their adoption in codes at this time. It is made clear that the document
is not intended to be adopted verbatim as a building code, but rather it is
to be used as a model by the code organizations, states and other jurisdic-
tions as they develop codes specifically applicable to their unique local or
regional radon control requirements. The document acknowledges the fact
that the proposed model building standards and techniques are based primari-
ly on experience in reducing radon in existing single-family homes and in a
limited number of new homes and schools. The assumption is made, however,
that techniques that work when retrofitted into existing homes will work
equally well in new homes and at a cost considerably less if installed dur-
in original construction.
GOAL
The goal of the document is portrayed as moving the building industry
toward the goal that Congress established in the Indoor Radon Abatement
Act, namely, to achieve radon levels indoors that are no higher than the
ambient radon levels outdoors. The draft acknowledges that technologically,
this may not be an achievable goal, but that the underlying objective is to
move towards the lowest possible levels of radon in new buildings.
CONSTRUCTION OPTIONS
The draft document contains a number of recommendations for implement-
ingthe specific construction standards and techniques. Two Construction
Options are proposed. Option 1 is a prescriptive option requiring instal-
lation of all the radon-resistant building standards and techniques listed
in the document to include a complete, operational sub-slab depressuriza-
tion system. Use of this option would relieve the builder of any responsi-
bility for post-construction radon control and no testing for radon would
be required either before or after occupancy. Verifying the installation
and performance of the various radon control features would be a normal
function of local building officials and inspectors.
Construction Option 2 is also a prescriptive option but it includes a
performance feature. Option 2 requires installation of only a limited
number of the building standards and techniques (primarily those designed
to resist radon entry) and includes a roughed-in sub-slab depressurization
capability. To satisfy the performance feature of this option, the builder
is required to conduct a long term (one year) radon test after the building
is occupied. If the results of the test are above 2 pCi/L, the builder is
required to complete and activate the sub-slab depressurization system. He
then has no further responsibility for radon control.
RECOMMENDATIONS
The current draft recommends that jurisdictions using these model
standards and techniques as the basis for developing their own codes, sel-
ect the construction option most appropriate to the radon potential that
exists in their areas. EPA will help in that regard by producing radon
-------
potential maps covering all States and many counties. It is anticipated
that these maps will "be available by this coming summer.
It is further recommended that Option 1 be used in any area identified
by EPA and designated by local authorities as having a high radon potential.
In all other areas, use of Option 2 is recommended to provide a basic level
of built-in radon resistance and a post-construction radon removal capability
if that capability is found to be needed.
STANDARDS AND TECHNIQUES
As for the specific construction standards and techniques included in
the draft, those applicable to Option 2 are listed first since they are more
limited in scope. They essentially include those techniques intended to
resist radon entry and prepare the building for post-construction radon re-
duction, if necessary. Many of these features are already standard construc-
tion practice throughout the building industry,
(1) A permeable layer of material should be placed under the slab to
facilitate installation of a sub-slab depressurization system.
(2) Polyethylene or equivalent sheeting material should be placed
immediately under the slab or other floor assembly to serve as a soil-gas
retarder.
(3) Slabs should be designed, mixed, placed, consolidated and cured
in accordance with American Concrete Institute (ACl) publications, to reduce
cracks.
(k) Openings around bathtub, shower or toilet drains should be sealed
to retard soil-gas entry.
(5) Openings around plumbing, wire or other objects that penetrate
slabs or floors should be sealed to retard soil-gas entry.
(6) Control joints, isolation joints and construction joints should
be sealed to retard soil-gas entry.
(?) Floor drains should be run through solid pipe to daylight or
mechanical or water traps should be installed.
(8) Wood foundations should be installed as prescribed in National
Forest Products Association publications.
(9) Pentrations of below-grade walls should be sealed.
(10) Exterior surfaces of below-grade walls should be sealed.
(11) If installed, sumps should be covered and sealed.
(12) Dirt floors in unvented crawlspaces should be covered with poly-
ethylene sheeting and a vent pipe installed to facilitate hookup of a sub-
poly depressurization system.
-------
(13) Joints in ductwork that passes through crawlspaces should "be
sealed.
(ik) Openings in floors above crawlspaces should he sealed.
(15) Access doors and other openings between crawlspaces and adjoin-
ing basements should be sealed.
(16) Non-closeable vents should be installed in ventilated crawlspaces
and located in accordance with local codes.
(l?) The components of a sub-slab depressurization system should be
roughed-in during construction.
To meet the more stringent requirements of Construction Option 1, the
draft adds the following features to those listed above. These features are
primarily designed to reduce radon entry by reducing the "stack" effect.
(1) All between-floor air passages (thermal bypasses) such as openings
+h°Un^ chimney flues, plumbing chases, ductwork, and wires routed through
the base and tops of walls, and around toilet, shower and bathtub traps
should be sealed.
(2) Attic access stairs or hatches should be sealed.
(3) Unsealable recessed ceiling lights should be avoided in top floor
ceilings.
(*0 Large capacity (whole-house) exhaust fans should be avoided.
(5) External air supply sources should be provided for all fireplaces,
wood stoves and other combustion appliances,
(6) External windows and doors in the superstructure should be weather
Btripped in conformance with model energy code or ASHRAE standards.
(7) A complete, operational sub-slab depressurization system should be
installed to include a warning system to alert occupants if the system fails.
Although the draft contains no cost projections, we believe these new
+KnS^IU°^i°n ^ea^ures can b® built-in to the average size home for less
than $500 added cost. It Bhould also be noted that many of these construc-
tion features play a secondary role in conserving energy and reducing water
infiltration problems in basements.
We hope to obtain industry-side acceptance of these construction stand-
ards and techniques when they are published in June, and that model building
codes will soon thereafter begin to include them. Subsequent adoption of
these new codeB by local jurisdictions will provide the needed mechanism to
ensure that new buildings will be as radon-resistant as possible.
-------
A-I-3
THE FLORIDA RADON RESEARCH PROGRAM: SYSTEMATIC
DEVELOPMENT OF A BASIS FOR STATEWIDE STANDARDS
by
David C. Sanchez
U. s. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
Richard Dixon
Florida Department of Community Affairs
Tallahassee, FL 32399
Ashley D. Williamson
Southern Research Institute
Birmingham, AL 35255-5305
ABSTRACT
The 1988 Florida Legislature has mandated development of
standards for the construction of radon-resistant buildings. By
February 1991, the Florida Department of Community Affairs (DCA)
* **ave adopted standards for building codes (new construction)
construction standards for mitigation of radon in existing
^ In suPP°rt these standards, a program of research
and development was initiated involving the State University
rPDt?ln' the DCA» and the U.S. Environmental Protection Agency
(EPA).in order to coordinate over 20 research projects involving
taculty members at four universities and five contractors, the
Florida Radon Research Program was subdivided into five topical
areas based on standards development needs. In each area, a
standards Development Committee was formed, chaired by the
principal investigator of a key project in the area. The committee
areas are: Improved Floor Barriers, Subslab Suction System Design
criteria, HVAC Specifications, Foundation Fill Material
Specifications, and Alternate Performance Standard. To assist in
interproject coordination, centralized reporting and data
Management systems have been implemented, and an online Radon
Literature Database has been developed to serve all projects within
the Florida Radon Research Program.
This paper has been reviewed in accordance with the U. S.
Environmental Protection Agency's peer and administrative review
policies, and has been approved for presentation and publication.
-------
INTRODUCTION AND LEGISLATIVE HISTORY1
The Florida Legislature on July 5, 1988, enacted legislation:
which provided comprehensive requirements for a statewide radon
protection program. Events leading to this legislation began when
the 1984 Legislature authorized the Florida Department of Healtft
and Rehabilitative Services (DHRS) to establish and enforce a rule
for an environmental radiation standard for land which emits
radiation. One of the key (and controversial) provisions of this
rule was to require the use of radon-resistant building techniques
for all new construction occurring in high radon risk areas.
Rule implementation was to follow the issuance of a map
defining the high radon risk areas. In March 1986 DHRS released
a map which identified these areas based on three criteria: (1)
areas previously mined for phosphate, (2) areas where elevated
radiation levels had been measured, and (3) areas surrounded by
lands characterized by the two previous criteria and underlain by
geologic deposits of phosphate ore. It was the intent to ado
additional areas to the map as they were identified over a 3-yea*
phase-in of the rule.
The 1986 Legislature intervened in the implementation of the
1984 legislation by requiring the Florida Institute of Phosphate
Research (FIPR) to conduct a comprehensive statewide study to
identify those areas of the state where the radon-resistant
building construction requirements would apply and prohibited the
DHRS from enforcing the new construction requirements until: (1)
the study was completed, (2) the study was evaluated by a Pee*
Review Committee, and (3) the DHRS had used the results of the
study to identify by regulations the affected areas.
The comprehensive statewide study2 conducted by FIPR was
completed in 1987 and peer reviewed. The Peer Review Committee/
in its final report (January 1988), made the following prograH
recommendations to the State:
(1) Development of radon-resistant building codes for
new construction,
(2) Certification of radon measurement and mitigation
companies,
(3) Provision of public information services about
radon,
(4) Notification on real estate sales contracts
regarding the health risks of radon,
-------
(5) Maintenance of the current radon standard (as low
as reasonably achievable not to exceed 0.02
working level (WL) or 4 picocuries per
liter (pCi/L).
The Amendments to Florida Statute Section 404.056 and the
addition of Florida Statute Section 553.98 by the 1988 Florida
Legislature addressed all of the above considerations and in
addition provided for:3
(1) Conducting a coordinated, cooperative program
of research, training, and service activities
related to the detection, control, and abatement
of radon, and
(2) Creating a trust fund, to ensure that
provisions of the statutes could be carried out.
It is these two provisions, especially the latter, that in
good measure facilitates and distinguishes the Florida radon
®£lslation from both the Federal Indoor Radon Abatement Act 4 and
other state radon acts5'6. The remainder of this paper will describe
tne coordinated and cooperative program of research that has been
developed under, and in response to, the Florida legislation.
THE FLORIDA RADON RESEARCH PROGRAM
STATUTORY BASIS
The Florida Radon Research Program (FRRP) is a direct
response, by the Florida DCA, to the intent and to certain specific
requirements of Florida Statutes Sections 404.056 (2) and 553.98
li)• These statutes' sections, respectively, call for (1) radon
t s®®rch (i.e., research, training, and service activities related
o the detection, control, and abatement of radon) and (2) the
hie^?i?pment' Publication, and adoption of standards for (a)
ouiiding codes for radon resistant buildings and (b) construction
onffards for roitigation of radon in existing buildings; training
imr? * use of such standards; and, within 2 years of adoption, an
update and adoption of standards based on the most current
fuSearcb* Florida Statutes Section 404.056 (4) ensures, through
„ f.**a(ion Trust Fund, financial support to the FRRP at a minimum
J-k building codes have been adopted and 1 year of training in
he use of the code has been provided.*
+ It is noted that the Florida Statutes provi
-------
PROGRAM DEVELOPMENT
The Florida radon protection legislation provided general
direction with regard to (1) areas of required activity, (2)
guidance and regulations to be developed, and (3) a timetable to
produce the guidance and regulation. Legislative hearings leading
to the Florida statute requirements attempted to define the
legislation in the context of the current state of knowledge with
regard to radon standards for new and existing structures. Much
of the experience with these topics had to be drawn from areas with
distinctly different environmental conditions, building practices,
and occupancy patterns and practices. It was with this as
background that the first effort by the DCA was to organize a Radon
Research Seminar and Workshop.
RADON RESEARCH SEMINAR AND WORKSHOP
On February 13-15, 1989, the DCA hosted a 1 day research
seminar followed by a 1-1/2 day research workshop. The primary
objectives of these two events were to clarify the current status
of radon research issues and to identify additional information
needed to develop the technical support necessary for the
development of a comprehensive, statewide code for radon-resistant
construction.
The 1 day research seminar consisted of invited presentations
by nationally recognized authorities in the field of radon problem
diagnosis and mitigation in large and small buildings. The
presentations addressed (1) radon availability and soil gas
permeability, (2) concrete slab integrity and reduction of slab
cracking potential, (3) house dynamics and radon entry, (4)
seasonal and spatial variability of radon levels in houses, (5)
radon in large scale buildings, (6) mitigation research in Florida
houses, and (7) integration of mitigation techniques in new
construction.
Participation in the 1-1/2 day workshop that followed the
research seminar included representatives from national research
groups, private sector practitioners, EPA research program
management, university researchers, industrial trade and
professional associations, and staff members of DCA and DHRS.
In order to facilitate a consensus evaluation of the status
of current research with regard to radon problem assessments, the
processes affecting radon entry, and the effectiveness of available
control techniques as well as to arrive at a consensus listing of
research issues in need of further research, the participants were
asked to take part in one of four working groups. The working
groups were asked to address: (1) radon availability, (2) barriers,
(3) building dynamics, and (4) large buildings. The findings of
the working groups are summarized in Table 1.
-------
TABLE 1. Summary of Research Needs Identified by the Florida
Radon Research Seminar and Workshop
Radon Availability
Measure and develop predictive techniques for
subbarrier radon concentrations based on soil
characteristics
Determine what volume of soil or fill effectively
contributes to radon at the subbarrier interface
Define the extremes of environmental conditions
affecting radon availability
Evaluate fill material construction criteria
impacts on radon availability
Determine the relationship of subbarrier and site
radon availabilities
Assess and interpret existing data with regard to radon
availability and the construction of radon-resistant
buildings
Barriers
o Identify and test effectiveness of physical
barriers to radon convection and diffusion
o Evaluate effect of ventilation and venting of
crawlspace as a barrier
o Develop innovative in-situ subfloor barriers
o Develop and evaluate dynamic pressure barriers
Building Dynamics
o Characterize the cause, magnitude, and
variability of building pressure differentials causing
radon entry
o Identify leakage pathways through the building
superstructure and assess their relevance to radon entry
Large-scale Buildings
o Assess the extent and magnitude of radon problems
in large buildings
o Assess the static and dynamic building conditions
affecting radon entry
-------
The results and findings of the Radon Research Seminar and
Workshop were used to define specific objectives, and possible
approaches for meeting these program objectives. The Workshop was
productive because it was carried out within a framework
established by the previous radon mitigation experience
illustrated in Figure 1. As shown in the figure, while some early
research was done around 1979 and 1983, it was not until late in
1986 that more extensive problem assessment and mitigation
activities were commenced by Florida agencies and EPA. It was this
early yet current research in Florida and national research
activities that permitted a consensus by the Workshop attendees as
to the areas of emphasis for the Florida Radon Research Program.
The Workshop findings with changes in emphasis attributable to
input from the Florida Coordinating Council on Radon Protection,
which met in December 1988, allowed the DCA to affirm the program
development scheme shown in Figure 2.
RADON RESEARCH PROGRAM DEVELOPMENT PLAN
The program development scheme (plan) outlined in Figure 2 is
consistent with the objectives defined by the Workshop and
emphasizes a need for: (1) studies of fundamentals, (2) applied
research, and (3) demonstration of solutions. The plan laid out
in Figure 2 shows the minimum time to demonstration and validation
for radon standards and codes may be 2 to 5 years depending on the
current state of development of possible radon resistant
construction approaches. There are thus near-term as well as
long-term products to be derived from the research plan which
relate to small (residential) and large (commercial) building radon
problems as well as to the continued development of alternative
mitigation measures for new construction. Table 2 presents some
examples of the kinds of tasks that were contemplated for the FRRP,
by research category, in April 1989.
-------
TIMELINE
YR/MO
FLORIDA RADON
RESEARCH PROGRAM
FLORIDA COORDINATING
COUNCIL RADON
PROTECTION MEETINGS
STATE UNIVERSITY FLORIDA ft
SYSTEM'S RADON EPA RADON
RESEARCH RESEARCH
EXTRA-FLORIDA
RADON RESEARCH
EPA,
FIPR PZPR
FSRS NCS
EPA
EPA,DHRS 4.
PI
*WP
EPA
P2
NJ/NY,PL
VA,MD,FL
19««
1999
1990
XL, TV
on
CO, NM
-L
Lagand:
* ¦••tinqa
o raporta
Pi ph«M 1, Cant*al Florida Mitigation Study
P2 phta* 2, North Central Florida Mitigation Study
PR prograaa raport
Wp work plan
BOR - Stata Univaraity Syataa Board of Raganta
DCA - Florida Dapartaant of Coaaunlty Affairs
DBRS - Florida Dapartaant of Haalth and Rahabilitatlva Sarvicaa
EPA - U. S. Environaantal Protaction Agancy
FIPR - Florid* Inatituta of Phosphata xaaaarch
FSRS - Florida Statawlda Radiation Study
NCS - Naw Construction Study
Figure 1. Florida Radon Research Timelines.
-------
TIMELINE RESIDENTIAL (SMALL BUILDING) COMMERCIAL (LARGE BUILDING)
FY
PROBLEM FUNDAMENTAL FIELD PROBLEM FUNDAMENTAL FIELD
ASSESSMENT EVALUATION CRITERIA VALIDATION ASSESSMENT EVALUATION CRITERIA VALU^njgi
AND TRAINING JK>
•«-«7
FIPR
StAdy
BOK
I
cod*
DCA
Standard
BOR
Cod*
93-94
94-95
•S-M
DCA
Cod#
OCA
Coda
ftavialon
DCA
Standard
DCA
Coda
DCA
Coda
Ravialon
Lagands "
BOX - atata univaraity syttaa Board of Raqanta
ftnfi • piorida Dapartaant of Coaaunity Affairs
fXM - Florida Xnatituta of Phoaphata Saaaaroh
p.S. - 1999 Florida statutaa on Jtadon Protaotion
Figure 2. Long Range Research Timelines.
-------
TABLE 2. Examples of Research Tasks Considered as Part of
Fundamental Studies, Applied Research, and
Demonstration Projects in the FRRP, April 1989*
Fundamental Studies
o Characterize the temporal variations of radon and
progeny source strengths and concentrations in
buildings
o Characterize pressure differences between source and
building spaces
o Characterize soil and fill material gas permeabilities
for subslab mitigation system design
Applied Research
o Develop improved designs and construction systems for
barriers to radon entry
o Develop standards for evaluation of sealants
o Develop subslab depressurization designs and design
guidance for Florida construction
Demonstrations
o Demonstrate the practical use of design standards
and the effectiveness of the installed construction
features in near- and long-term applications. The first
designs or techniques suitable for demonstration
studies may be enhanced barriers to soil gas entry and
design standards for subslab radon soil-gas removal
systems
This Table does not differentiate between near- and long-term
tasks. Numerous near-term tasks were initiated under State
University System projects which began in January 1989 and
terminated in December 1989 in order to provide a core of technical
support to the draft building code on radon resistant construction
and mitigation techniques required of the State university system's
Board of Regents (BOR) by February 1, 1990.
-------
INTEGRATION OF ONGOING RESEARCH ACTIVITIES
Between January and June 1989, results and findings of
research activities by the State University System projects and EPA
projects further led to a definition of critically important
research areas, identification of organizations that could
undertake the work, and determination of the coordination and
management mechanisms that would be needed to ensure a timely
response to the requirements of the Florida radon protection
legislation.3
This management assessment effort culminated in the
organization and creation of the Florida Radon Research Program as
a coordinated, comprehensive research effort under the management
of the Florida DCA and the U.S. EPA.
The FRRP has as its purpose the provision of technically
supported and defensible standards for building codes for radon-
resistant buildings and construction standards for mitigation of
radon in existing buildings.
WORK PLAN MEETING OF THE FLORIDA RADON RESEARCH PROGRAM
Prior to the Work Plan Meeting of the FRRP, held on July
25-26, 1989, in Gainesville, a tentative organization of
management, reporting, and research task areas was defined by DCA
and EPA and communicated to lead research groups.
The primary objectives of the Work Plan Meeting were to
present an overview of the FRRP, present draft work plans for
technical task areas, form standards development committees,
develop consensus protocols for diagnosis and assessment
measurements, and define data flow and coordination between
projects. This initial meeting of the FRRP was attended by 13
research groups representing 19 ongoing research projects. All
projects were sponsored by DCA, EPA, or State University System's
Board of Regents (BOR).
On July 25, 1989, the Work Plan Meeting participants discussed
and outlined the objectives and approaches of the tentatively
defined FRRP task areas. Figure 3 is an organizational chart for
each task area listed by its current Standards Development
Committee name. Each Standards Development Committee chairman, in
addition to being the principal investigator on a project, is the
lead technical developer and synthesizer of the technical content
for a specific task area. Finally, as the name suggests, each
Committee chairman will be responsible for coordinating Committee
activities directed at turning research results into standards and
codes in the respective task area.
Table 3 summarizes the major technical task areas of the FRRP as
well as its management and research data base features.
-------
INTEGRATION OF ONGOING RESEARCH ACTIVITIES
Between January and June 1989, results and findings of
research activities by the State University System projects and EPA
projects further led to a definition of critically important
research areas, identification of organizations that could
undertake the work, and determination of the coordination and
management mechanisms that would be needed to ensure a timely
response to the requirements of the Florida radon protection
legislation.5
This management assessment effort culminated in the
organization and creation of the Florida Radon Research Program as
a coordinated, comprehensive research effort under the management
of the Florida DCA and the U.S. EPA.
The FRRP has as its purpose the provision of technically
supported and defensible standards for building codes for radon-
resistant buildings and construction standards for mitigation of
radon in existing buildings.
WORK PLAN MEETING OF THE FLORIDA RADON RESEARCH PROGRAM
Prior to the Work Plan Meeting of the FRRP, held on July
25-26, 1989, in Gainesville, a tentative organization of
management, reporting, and research task areas was defined by DCA
and EPA and communicated to lead research groups.
The primary objectives of the Work Plan Meeting were to
present an overview of the FRRP, present draft work plans for
technical task areas, form standards development committees,
develop consensus protocols for diagnosis and assessment
measurements, and define data flow and coordination between
projects. This initial meeting of the FRRP was attended by 13
research groups representing 19 ongoing research projects. All
projects were sponsored by DCA, EPA, or State University System's
Board of Regents (BOR).
On July 25, 1989, the Work Plan Meeting participants discussed
and outlined the objectives and approaches of the tentatively
defined FRRP task areas. Figure 3 is an organizational chart for
each task area listed by its current Standards Development
Committee name. Each Standards Development Committee chairman in
addition to being the principal investigator on a project, is the
lead technical developer and synthesizer of the technical content
for a specific task area. Finally, as the name suggests, Committee
chairman will be responsible for coordinating Committee activities
directed at turning research results into standards and codes in
the respective task area.
Table 3 summarizes the major technical task areas of the FRRP as
well as its management and research data base features.
-------
Figure 3. Standards Development-Committees and 1989 FRRP Projects.
-------
TABLE 3. Summary Description of Major Technical Task Area
Research Activities and Program Management and Data
Base Features
Management & Data Base Features
o Central Reporting and Database System - A computer
system consisting of a multi-user database and
electronic mail capabilities is maintained at the
University of Florida Geoplan Center. All house and
site specific research data are being placed on an
ARC/INFO geographic information system. FY 90 data
will include multiple inputs (10-25) for numerous study
sites (40-100-3000). Time series data are included,
o Radon Literature Data Base - All literature referenced
and developed as part of the FRRP is entered in
abstract to the library indexing program, XYINDEX, for
literature searching purposes. The data base currently
contains over 1000 screened references,
o Standard Measurement Protocols - A set of 25
quantitative measurements for use in the FRRP projects
have been identified and accepted by consensus for use
in characterizing radon source strengths, building site
soils, building static and dynamic conditions, and
indoor radon and radon decay product conditions.7
Improved Floor Barriers
o Research and development is ongoing in four areas:
(1) and (2) evaluation of the effectiveness of radon-
resistant construction in large- (schools) and small-
scale buildings (ca. 40 houses), (3) laboratory testing
of physical barriers to radon,* and (4) problem assessment
of a vented crawlspace as a barrier to radon entry.
Subslab Suction System Design Criteria
o A computer model of subslab communication (i.e.,
extension of pressure fields in subslab soils) has been
developed and used to design successful mitigation
systems.' The model is being extended to address
complex building footprint geometries, inhomogeneous
subslab permeabilities, and the use of perforated
drainage materials to extend depressurization fields.
Model predictions are being verified against measured
conditions in test plots and houses and will be tested
under new house conditions.14'11
-------
Table 3. Continued
HVAC Specifications
o This task area has as its near term objective the
assessment of the typical pressure distributions in
Florida houses. This assessment will be used to set
the model boundary conditions for the subslab pressure
field extension and fill material soil gas transport
model. The long-term goal is to establish standard
criteria which limit the pressure differentials induced
by HVAC systems.
Foundation Fill Material Specifications
o This task area encompasses the development of a model
for evaluating the radon potential and indoor radon
impacts of fill material used in construction." The
model will relate impacts to fundamental characteristics
of fill materials such as radium content, grain size,
moisture content, and thence to more easily obtained
site characteristics such as soil types and
permeabilities or degrees of compaction. The model
will define what depths of fill are significant as
radon sources. Fill materials at the source and as used
on building sites are being sampled extensively
statewide. This data base is being used to
determine the conditions to be modeled as well as to
establish needed interparameter correlations.
Alternate Performance Standard
o A performance standard may complement new construction
codes and standards for mitigation. The performance
standard must be compatible with state indoor radon
standards expressed as an annual average. Data from
about 40 houses in four regions of Florida are being
collected to establish the statistical relation
reliability and confidence of short- (1-30 day) vs
long-term averages." A model will be developed to
generalize the study results.
This task area research will also investigate the
feasibility of using a stress (house depressurization)
condition to test the adequacy of new construction
codes and mitigation standards (prescriptive
standards).
-------
STATUS OF THE FLORIDA RADON RESEARCH PROGRAM
The FRRP is now in its second year of funded research.
Initial (FY 88-89) efforts were directed at supplementing ongoing
BOR and EPA research. The content task areas (Standards
Development Committees) identified in Figure 3 were addressed, only
in part, by this initial supplement. These initial efforts were
generally fundamental research studies directed at producing
definition and characterization of sources, building static and
dynamic conditions, and currently available approaches for control
of radon entry. This work was used to define and support the
current recommendations of the draft building code provided to the
DCA by the BOR.
Current (FY 89-90) resources are directed at expanding the
task area data bases started in FY 88-89. Major new initiatives
are in the fundamental studies of slabs as barriers, investigation
of the effects of vented crawlspaces on radon entry, and
characterization of the short- and long-term variations of total
house dynamic conditions. Applied studies are being directed at
providing models for implementation and testing of subslab suction
system design criteria, foundation fill material specifications,
and alternate performance standards. These FY 89-90 initiatives
are directed to laying the foundation for the futher development
and revision of standards and to beginning the long-range research
and development stage of the program. (See Figure 2.)
FUTURE DIRECTION OF THE FRRP
The main objective which will guide the FRRP in future
research years is the need to ensure the availability of
technically defensible and cost effective standards for radon
protection, which balance the State's policies regarding affordable
housing, energy conservation, and structurally safe buildings as
implemented through building codes. Figure 2 shows the FRRP
shifting emphasis, after FY 90-91, from fundamental and applied
research to field evaluation or validation; i.e., demonstration
studies. Figure 2 indicates about an initial 2 year lag between
small (residential) and large (commercial) building research
developments with both building types being at about the same
development level by the FRRP's fifth year. It is expected that,
by that point in the program, new research efforts will be special
purpose efforts directed at revisions of code priority areas.
-------
REFERENCES
1. Committee Information Record. House of Representatives,
Committee on Community Affairs, Bill No. HB 1420. Staff
Analysis and Economic Impact Statement. April 7, 1988.
2. Florida Statewide Radiation Study. Contract Number 087-044,
Florida Institute of Phosphate Research. November 5, 1987.
3. Chapter 88-285. Committee Substitute for House Bill No. 1420.
An act amending s. 404.056, F.S. and creating s. 553.98, F.S.
Filed in Office Secretary of State. July 5, 1988
4. Indoor Radon Abatement Act of 1988. 15 U.S.C. 2601, Title
III. December 1988.
5. Radon Gas Demonstration Project of 1986 and Home Improvement
Loan Act. Act 62-1986. April 8, 1986. (Pennsylvania)
6. An Act Concerning Radon Gas and Radon Progeny Contamination.
Amendment to P.L. 1958. September 9, 1985. (New Jersey)
7. Florida Radon Research Program Standard Measurement Protocols
Part l Soil Measurements, Part 2 Building Measurements, Part
3 Radon Measurements. Florida Development of Community
Affairs, Tallahassee, FL, January 1990.
8. Pugh, T. D. "Test Barrier Facility for Radon Reduction
Technology." Paper Accepted for Presentation at The 1990
International Symposium on Radon and Radon Reduction
Technology, Atlanta, GA, February 19-23, 1990.
9. Hintenlang, D. "Sub-slab Suction System Design for Low
Permeability Soils." Paper accepted for presentation at the
1990 International Symposium on Radon and Radon Reduction
Technology, Atlanta, GA, February 19-23, 1990.
10. Furman, R. A. "Sub-slab Pressure Field Extension Studies in
Four Test Slabs Typical of Florida Construction." Paper
accepted for presentation at the 1990 International Symposium
on Radon and Radon Reduction Technology, Atlanta, GA, February
19-23, 1990.
11. Fowler, C. S. et al. "Engineering, Design Criteria for Sub-
slab Depressurization Systems in Low Permeability Soils."
Paper accepted for presentation at the 1990 International
Symposium on Radon and Radon Reduction Technology, Atlanta,
GA, February 19-23, 1990.
-------
12. Rogers, V. C. and K. K. Nielson. "Benchmark and Application
of the RAETRAD Model." Paper accepted for presentation at the
1990 International Symposium on Radon and Radon Reduction
Technology, Atlanta, GA, February 19-23, 1990.
13. Roessler, c. E. "Temporal Patterns of Indoor Radon in North
Central Florida and Comparison of Short-term Monitoring to
Long-term Averages." Paper accepted for presentation at the
1990 International Symposium on Radon and Radon Reduction
Technology, Atlanta, GA, February 19-23, 1990.
-------
A-I-4
TEN PRACTICAL LESSONS FOR AN EFFECTIVE RADON RISK COMMUNICATION PROGRAM
by: Ann Fisher
Office of Policy, Planning and Evaluation
U. S. Environmental Protection Agency
F. Reed Johnson
Economics Department, U. S. Naval Academy, and
Office of Policy, Planning and Evaluation
U. S. Environmental Protection Agency
ABSTRACT
Those responsible for state and local radon programs often express
frustration about the small share of homes that have been tested for radon,
and the small share of those with high readings that have been mitigated.
Several recent studies have examined how well alternative ways of
communicating about radon's risk have accomplished the goals of motivating
appropriate testing and mitigation. Unfortunately, the results of these
studies have not reached practitioners. This paper is for them. It
summarizes the practical implications that are most crucial for planning and
implementing an effective radon risk communication program--a program that
will motivate people to test for radon and mitigate when radon levels are
high, without unduly alarming those whose radon levels are low.
Our more complete paper, "Radon Risk Communication Research: Practical
Lessons" is forthcoming in the Journal of the Air and Waste Management
Association (formerly JAPCA1. It identifies six reasons why people do not
respond to radon as a serious threat, provides some remedies suggested by the
recent radon risk communicaiton research, and gives references to those
studies. The remedies are summarized here; the paper in Air and Waste has
more information on why these lessons will help make your radon risk
communicaiton program more effective.
1. Two points must be emphasized together: Radon truly is a serious
health threat, but one that is easy and cheap to detect and mitigate.
2. Stress the benefits of testing: either the peace of mind in
learning that there is no problem, or the opportunity to eliminate a serious
health risk inexpensively.
3. Use communication channels that reach women, who are more likely to
test than men.
-------
A-I-5
INTERACTIVE HOUSE INVESTIGATION AND RADON DIAGNOSTICS
COMPUTER PROGRAM
by: Lynne M. Gillette
U.S. Environmental Protection Agency
Office of Radiation Programs
401 H St. SW
Washington, DC 20460
Terry Brennan
Camroden Associates
RD # 1, Box 222
Oriskany, NY 13424
ABSTRACT
The interactive computer program called "Dungeons and Radon"
was developed as part of the Environmental Protection Agency's
(EPA's) Radon Contractor Proficiency (RCP) Program's "Radon
Technology for Mitigators" (RTM) course which is currently being
offered in the Regional Radon Training Centers (RRTCs). The
program was designed by Terry Brennan to be used in training radon
mitigation contractors. The Macintosh based program consists of
a series of animated, sound and voice enhanced house scenes. The
participants choose where and what to investigate and where to
perform diagnostic tests in order to gather enough information to
design a successful mitigation system.
This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
-------
4. Emphasize that testing for radon is inexpensive and requires no
special technical skill.
5. Preserving home values can be a powerful incentive. Stress that in
addition to protecting the family's health while in the home, testing (and
mitigation when needed) will prevent the loss of a potential buyer concerned
about radon.
6. Target home buyers for your messages about radon testing and
mitigation.
7. Get people--especially community leaders--to publicize their own
test results and encourage others to test (and mitigate when appropriate).
8. Use comparisons with other risks that have similar characteristics
(such as long latency and illness leading to death) to help people have a more
realistic perception of their risk from radon.
9. Acknowledge scientific uncertainty but concentrate on scientists'
agreement about important practical conclusions.
10. Design the message to attract attention, make it concrete, and
repeat it often. Use as many media channels as possible.
-------
A-I-5
INTERACTIVE HOUSE INVESTIGATION AND RADON DIAGNOSTICS
COMPUTER PROGRAM
by: Lynne M. Gillette
U.S. Environmental Protection Agency
Office of Radiation Programs
401 M St. SW
Washington, DC 20460
Terry Brennan
Camroden Associates
RD # 1, Box 222
Oriskany, NY 13424
ABSTRACT
The interactive computer program called "Dungeons and Radon"
was developed as part of the Environmental Protection Agency's
(EPA's) Radon Contractor Proficiency (RCP) Program's "Radon
Technology for Mitigators" (RTM) course which is currently being
offered in the Regional Radon Training Centers (RRTCs). The
program was designed by Terry Brennan to be used in training radon
mitigation contractors. The Macintosh based program consists of
a series of animated, sound and voice enhanced house scenes. The
participants choose where and what to investigate and where to
perform diagnostic tests in order to gather enough information to
design a successful mitigation system.
This paper has been reviewed in accordance with the U.S.
Environmental Protection Agency's peer and administrative review
policies and approved for presentation and publication.
-------
PURPOSE
The interactive computer program "Dungeons and Radon" was
developed for the Environmental Protection Agency's (EPA's) Radon
Technology for Mitigators (RTM) course which is part of the Radon
Contractor Proficiency (RCP) program. The RCP program was
implemented in response to the Indoor Radon Abatement Act of 1988
which mandates that EPA operate a proficiency program for
evaluating the gualifications of private firms offering radon
mitigation services. The RCP program includes the RTM course and
other radon mitigation contractor training programs which are
delivered by the Regional Radon Training Centers (RRTCs) and
others. Beginning in October 1989, the National RCP Exam has been
offered. In the spring of 1990 the RCP list will be published with
the names of persons who have met the proficiency requirements.
During the development of the RTM course it became
increasingly clear that the training courses could not all be
taught using the same training aids. Some training sites would
have training mock-ups available, others would be using houses or
other buildings during the training. The computer simulation of
a house with a radon problem was seen as a way to consolidate key
concepts from the course and provide a more consistent training
experience.
PROGRAM DESCRIPTION
"Dungeons and Radon" is a simple interactive computer program
designed for use with an Apple Macintosh computer equipped with at
least one (l) megabyte of memory (RAM) and a monochrome monitor.
The program is very user friendly and can be operated by a student
with little or no computer experience through the use of a "point
and click" mouse and a simple menu system. The program is designed
around a simple ranch house with a basement and attached
crawlspace. The program can be expanded or altered by those
interested in investing some time and money in the program
development software "Course Builder" which is sold by TeleRobotics
International Inc., 8410 Oak Ridge Highway, Knoxville, Tn 37931.
Their telephone number is (615) 690-5600.
Once the program begins, students move from the title screen
to a screen with simple operating instructions by executing a point
and click with the mouse and following the instructions on the
screen. All other operations are performed in the same manner. As
the program proceeds, students are presented with a view of the
house from the street. Students then move into the basement of the
house where they can choose to investigate parts of the basement
more closely, perform diagnostic radon "sniffer" tests, smoke tests
to determine air flow or sub-slab communication tests. As the
students move around the basement and attached crawl space they are
presented with sometimes animated visual information as well as
-------
audio and text information. The sub-slab pressure test results are
dependent on where the student chooses to drill the test holes.
The data collected during the student's interaction with the
computer is stored in a report which the student can then use in
mitigation system design exercises.
As the program presents much visual and audible information
it should be experienced to be fully appreciated. The program will
be demonstrated on February 19th during the first Poster Session
of the Symposium.
SYSTEM REQUIREMENTS AND SOFTWARE AVAILABILITY
In order to run the program in it's current form an Apple
Macintosh with at least one (1) megabyte of memory (RAM) is
required. "Dungeons and Radon" was developed under an EPA contract
and is public domain and has been distributed to the Radon Regional
Training Centers (RRTCs) and other interested trainers for use with
the Radon Technology for Mitigators (RTM) Course. Those that are
interested in having their own copy of the software may purchase
it directly from Camroden Associates for a nominal fee to cover the
cost of materials, duplication and postage. To alter or build onto
this program or to create your own program, you will need the
software program "Course Builder" mentioned above.
-------
STATE RADON PROGRAMS
John Paul Reese
New York State Energy Office
Albany, New York 12223
WITHDRAWN BY AUTHOR
-------
A-I-7
COMMUNITY-BASED RADON EDUCATION PROGRAMS
by: Joseph Laquatra
Department of Design and Environmental Analysis
Cornell University
Ithaca, NY 14853
ABSTRACT
In the United States, educational programs about radon gas have been developed and
implemented by federal and state government entities and other organizations, including the
Cooperative Extension Service and affiliated land grant universities. Approaches have included
the production of brochures, pamphlets, workshops for targeted audiences, and consumer
telephone hotlines. In a free market for radon mitigation products and services, these efforts
can be appropriate for their credibility, lack of bias, and individualized approaches. The
purpose of this paper is to report on an educational program about radon undertaken by Cornell
Cooperative Extension, including county-based workshops targeted to homeowners, housing
professionals, high school teachers, and others. An analysis of survey data from program
participants forms the basis for a discussion of the effectiveness of the Cooperative Extension
Service in reaching the public about this topic.
-------
A-I-7
THE COOPERATIVE EXTENSION SERVICE
A three-way partnership involving federal, state and county resources, the Cooperative
Extension Service is a national network which links communities with land-grant universities
in an informally structured educational system. The land-grant universities were authorized
through the Morrill Acts of 1862 and 1890. In 1914, the Smith-Lever Act established
Cooperative Extension as a partnership of the U.S. Department of Agriculture and these
universities. At the local level, this partnership includes professional staff in nearly all of the
3,150 counties of the United States. These professionals are educators, and include Extension
Agents, their staff associates, program assistants and volunteers, all of whom draw from
university resources to develop and implement educational programs for adults and young
people in their communities.
Cooperative Extension may be most well known for its agricultural programs in crop
development, pest management and others. Another well known component of Extension is
4-H, which consists of numerous programs for young people in a variety of areas, such as
animal husbandry and leadership training. Home economics is the traditional name for a
program area that in its early days focused on domestic skills for homemakers. It has grown
over the years to include programs in home construction and repairs, energy conservation,
financial management, nutrition, and human development. Audiences for these programs have
expanded from individuals and families to include groups and organizations, other educators,
and business professionals.
Cooperative Extension is organized at the county level as a membership association.
Volunteers serve on program committees that provide overall guidance for program focus and
objectives. Actual programs are delivered to the general public, regardless of membership
status or income levels.
A key resource for county Cooperative Extension associations is their link with the land-
grant university. University faculty members are assigned responsibilities to provide program
leadership in their areas of expertise for Extension field staff. This leadership consists of
conducting research, and using the results of that research to provide field staff with continuing
education as well as resources for local programs.
Faculty of the land-grant university in New York State, Cornell University, have primary
responsibility to build and maintain the resource base for programs implemented throughout the
statewide network of Cornell Cooperative Extension, which involves local associations in 57
counties and the five boroughs of New York City.
THE PROGRAM
Research and program development are activities that university faculty often undertake in
cooperation with state and federal agencies. "Radon in the Home Environment" is a program
of Cornell Cooperative Extension that resulted from such cooperation. A two-department
research project investigated radon levels in 250 randomly selected homes within a fourteen-
county area of New York State (1). Funded by the U.S. Department of Agriculture, issues
examined in this project included radon levels and relationships with housing conditions, soil
conditions, and residential energy efficiency (2).
During the course of the radon research effort, preliminary activities were undertaken to
plan community-based educational programs on the issue, for implementation through the
-------
Extension network. Data from the research project were collected during the early months of
1987. At the same time, awareness of the radon issue was rapidly increasing among the
general public, because of exposure in both print and broadcast media. One result of this
exposure was the desire among people for more information. Because county Extension
Associations are widely perceived to be reliable and unbiased sources of information, they
began to receive a large volume of requests for assistance in understanding the radon issue.
To help Extension field staff respond to the increasing volume of information requests
they were receiving, written materials were developed for their use in the form of newsletter
articles and fact sheets. These were supplemented with bulletins from the U.S. Environmental
Protection Agency (EPA). At the same time, assistance was provided to the author of this
report, who is the state Extension housing specialist, by the New York State Energy Office, in
the development of an inservice education course that was held in June of 1987 on the Cornell
campus. While the inservice was of a technical nature, it proved to be understandable to an
audience of Extension professionals with diverse educational backgrounds.
Following the inservice, plans were made to present the material in workshops for the
general public throughout the state. Selected materials from those used in the inservice were
phosen for this purpose. A committee of Extension agents working on plans for
implementation of the public workshop recommended that the New York State Department of
Health (DOH) be involved. At that time, the DOH was planning to undertake its Radon
Screening Program initiative, a component of which is public education. This identical goal of
the two organizations facilitated a great deal of cooperation in developing and implementing the
Workshop.
Decisions about the public information workshop were made by Extension agents; Mr.
Laurence Keefe, the coordinator of the DOH Radon Program; and the author of this report.
The workshop was to be two hours in length and would cover basic information about radon:
its origins and health effects; how it gets in houses; mitigation measures and their effectiveness;
state and federal programmatic responses in general; and specifically, resources from the EPA,
Cornell Cooperative Extension, the New York State Energy Office, and the New York State
department of Health. The Radon Screening Program of the Department of Health, in
Particular, was discussed, and applications for free radon detectors were distributed to
Workshop participants.
In the first year of the Extension Radon Education Program, workshops were conducted
individually and regionally for counties around the state. While most of these workshops were
held during evening hours on week nights, some were held during working hours for housing
Professionals, such as builders, developers, and real estate brokers. In conjunction with these
events, media appearances were made on television and radio talk shows by Mr. Keefe,
Extension agents, and the author of this report. These shows were used as opportunities to
Publicize educational materials on radon and the workshops, as well as to take call-in questions
from the viewing or listening public.
Most of the workshops were marked by high attendance levels. At one program, potential
Participants had to be turned away because of space limitations. Demand for follow-up
activities has been high in many counties since the initial workshop. Cooperative Extension
Agents have fielded call-in questions to consumer hotlines, written articles for local
newspapers, appeared on radio talk shows, and have conducted additional workshops for
small groups. Resources for the follow-up activities have included a slide/tape presentation
developed by the author of this report, fact sheets, newsletter articles, and video tapes
Produced by the EPA and other organizations.
-------
PROGRAM EVALUATION
Because of the variety of educational methods undertaken in this program, not all of those
reached could be identified. However, because pre-registration was required for most of the
workshops that were held, as a method for estimating space and materials requirements, names
and addresses were obtained for many of the participants. As part of a program evaluation
effort, a 20-item questionnaire was mailed to 443 of these participants in 18 counties, along
with a cover letter and return envelope that was stamped and addressed. Questions included
demographic characteristics of program participants, their ratings of the quality of the
workshop, actions they took as a result of the information received, numbers of people with
whom they shared information, whether they knew about Extension prior to the workshop,
and what types of information on radon they thought Cornell Cooperative Extension should
continue to provide, if any. In return for completing the survey, respondents received free
Extension publications of their choice, from the Housing and Home Environments Notes
series. 148 questionnaires were returned, and of these, 140 were useable.
A summary of demographic characteristics of the respondents is shown in Table 1. On
average the workshop participants can be described as middle-aged, having some college
education, and from small households with higher than average incomes. Fifty-three percent
described their occupations as being of a professional nature, 21 percent were retired, none
were unemployed.
Regarding how respondents learned of the workshop, 44 percent read of it in a
newspaper, 34 percent had received a direct mail announcement from Cornell Cooperative
Extension, seven percent learned of it from an Extension calendar of events, 7 percent from
radio or television, and 8 percent from word-of-mouth.
TABLE 1. DEMOGRAPHIC CHARACTERISTICS OF WORKSHOP PARTICIPANTS
Characteristic Mean Standard Error Range
Age 49.06 1.256 14-79
Number in
household 2.91 0.106 1-7
Household
income, 1988 $51,197.24 2,825.73 7,200 - 200,000
Years in school 15.25 0.221 9-21
Participants were asked to rate two aspects of the workshop, using a scale from 1 - 5, with
1 defined as "low," and 5 as "high." When asked to rate the workshop based on the amount
of information that they actually used, the mean rating was 3.42. Based on the value of the
information to them, the mean rating was 3.92. Respondents were asked to check all
applicable workshop benefits from a list. Eighty percent checked "answers to questions," 76
percent indicated "resource materials," 28 percent "help in decision making," 11 percent
"insight and support from others," 31 percent "new ideas," and 10 percent listed other benefits,
such as the free radon test kit applications and professional contacts.
-------
Fifty-five percent of survey respondents indicated that they had conducted radon screening
tests in their homes after attending the workshop. Seventy-one percent of those who tested
received their testing devices from the New York State Health Department; 29 percent obtained
fhem from private firms. Fifty-one people reported their results from charcoal canisters, 17
mdicated levels from alpha track detectors, and four reported grab sample readings. These test
results are summarized in Table 2.
Thirty-four percent of the respondents indicated that they had taken follow-up actions
based on high radon screening test results. This was broken down as 16 people conducting
follow-up tests, 12 increasing house ventilation, 2 having mitigation systems installed, and 6
taking other actions such as crack sealing.
IABLE 2. RADON TEST REST TT .TS REPORTED BY WORKSHOP PARTICIPANTS
Statistic Charcoal Canister Alpha Track Grab Sample
N 51 17 4
Mean(pCi/L) 6.45 2.33 25.10
Range (pCi/L) 0- 100 0-8 0- 100
Standard Error 2.10 0.57 24.97
When asked to indicate how else workshop information was used, 77 indicated teaching
someone else what was learned, 19 helped another person conduct a radon screening test, and
^4 specified other uses, such as incorporating some of the material in elementary and high
school science courses, and in one college course. Sixty-nine respondents reported sharing the
formation with one to five people; thirty-one percent with more. One respondent reported
^aching more than 400 people at over 20 presentations.
The radon workshop was the first exposure to Cornell Cooperative Extension for 34
Percent of survey respondents. Twenty-four percent had received additional information on
radon from Extension since the workshop, and 65 percent indicated that the educational
Program should continue. Some specific suggestions were made, including incorporating the
listing program into the public school system, and expanding the workshop focus to cover
other radon-related issues, such as procedures for real estate transactions and radon in schools,
specific audiences and delivery methods were also suggested; and clubs and service
°rganizations were named as possible collaborators.
£Valuation issues
« resources SdentifiedTsuch an effort will be undertaken. A field experiment would i
observations of differences between program participants and a control group, so that
Conclusions can be made about program impacts. Without a control group, conclusions
^arding the program's role as an influence in the participants' decision to test cannot be made
°ojectively. Workshop attendees may represent a specialized group. Their demographic
cf aracteristics indicated that they are highly educated. They may have been people who had
®*rcady decided to test their houses, but wanted verification of their decisions.
. Even with a field experiment design, it would be difficult to link program participation
^tth results, such as conducting a radon test. Nonprogram factors, such as voluntary
-------
participation in an Extension program, are difficult to control statistically, because
"...it is difficult to tell whether (a) program participants who are
more in contact with an Extension program therefore use more
recommended practices, or, (b) program participants who use
more recommended practices have sought more frequent contact
with Extension and other sources of similar information, or, (c) a
combination of (a) and (b) are actually the case over time in some
cyclical fashion." (5, p. 37)
The survey used in this evaluation can be described as a perceptual one, in that the
perceptions of program participants were solicited. They were asked for perceptions based on
reflections of knowledge levels before the program, and whether the change in their knowledge
influenced their decision making. Rivera et al. (5) discuss the validity of findings based on
perceptual surveys. On objectivist grounds, retrospective data do not produce adequate
evidence of program results because of questions regarding clientele perceptions, memory loss,
or distortion. A subjective view, on the other hand, holds that because of the links between
human experience and perception, it is important to allow program participants to identify
effects of program participation. At the very least, an evaluation based on perceptual data
allows an examination of opinions about Extension programming activities from clientele.
PARTICIPANT PERCEPTIONS
A close examination of program participant perceptions is useful as follow-up activities are
planned or undertaken. Efforts can be appropriately targeted based on insights gained from
such an analysis. To this end, a model was constructed to identify explanatory variables that
may have affected ratings assigned to the workshop by survey respondents, based on the
amount of information they actually used. Multiple regression analysis was applied to the
following model:
RAi=/(RTi, AGEi, NHi, HIi),
where RAj = the rating applied to the workshop based on the amount of information that the i1^
respondent actually used (1 = low, 5 = high); RTj = whether a radon screening test was
conducted in the i^1 household (0 = no, 1 = yes); AGEj = age of the i^1 respondent; NHj =
number of people in the i^1 household; HIi =1988 income of the i^1 household. Results of the
analysis are presented in Table 3.
As shown in Table 3,14 percent of the variation in the dependent variable is explained by
the model. The significant F-test indicates a good fit for the model. Significant independent
variables are whether a radon test was conducted and household income. The significance of
the testing variable is expected, as much of the workshop dealt with the mechanics of
conducting a screening test. The significance of the income variable may be related to financial
implications of radon mitigation measures.
/
-------
TABLE 3. REGRESSION RESULTS OF WORKSHOP RATINGS
Independent
-VariflftH
RT
AGE
NH
HI
R2 = 0.14
&
i-Yahrc
Probability
0.688
3.173
0.002
0.005
0.064
0.579
-0.061
0.631
0.529
0.00001
2.278
0.025
F =3.749; p = 0.007
Responses to the surveys were then divided into two groups: those who conducted radon
screening tests in their homes and those who did not. Paired t-tests were conducted to examine
potential differences in workshop ratings and demographic characteristics of the respondents.
Results are presented in Table 4. Significant differences in demographic characteristics
between the two groups were not observed. But workshop ratings did differ significantly,
with the group of those who tested assigning higher ratings to the workshop on the both the
value and amount of information they received,
IABLE 4,
DTFFERENCFfi IN TESTING AND NON-TESTING RESPONDENTS
Workshop Ratings:
Value of
Information
Amount of
Information
MsanX-r
0.344
0.574
Paired t-Valw
2.065
2.967
Probability
0.043
0.004
Age
Income
Education
Number in
household
1.117
-7770
0.783
0.306
0.420
-1.08
1.906
1.297
0.676
0.286
0.062
0.199
X = respondents who tested; Y = respondents who did not test
-------
CONCLUSIONS
In order to develop strategies to educate community residents about radon in the home
environment, a number of areas must be examined. These relate to the credibility of
information sources, characteristics of communities and community residents, and individual
motivations for seeking information. This paper presented findings from a preliminary
evaluation of a community-based radon education effort. A future study could examine
characteristics and behaviors of both program participants and non-participants. Even though
this study was limited to participants, some interesting observations were made which will be
useful in guiding continued program activities.
The study showed that participants in a radon education program developed by Cornell
Cooperative Extension rated the amount and value of information they received highly,
especially if they subsequently conducted radon screening tests in their houses. The delivery
of information was not limited to program participants, as many people shared what they had
learned with others, with some participants including the material in elementary, high school,
and college courses. A majority of participants requested that educational activities in this area
continue.
Some findings reported in this study indicate that additional efforts may be necessary to
reach people with lower income and education levels than those of program participants to date.
The high levels in both of these variables among the initial program participants may be an
indication that they are early adopters with high information seeking skills. More widespread
awareness could be achieved with targeted educational efforts, which are currently taking
place. A future evaluation effort could include a focus on this aspect of the program.
At a time when radon has been covered widely in both print and broadcast media, the
Cooperative Extension Service has been perceived as a credible source of unbiased information
on the issue. Successful partnerships in New York State involving Cornell Cooperative
Extension, thie New York State Department of Health, and the New York State Energy Office
have resulted in the delivery of educational materials to individuals, groups and organizations
throughout the state. On a national basis, the Cooperative Extension Service is a logical
organization for federal and state agencies to rely on for the development and implementation of
community-based radon education programs. In the past, this system has been shown to be
responsive to concerns of local communities and to make a positive impact on clientele who
participate in Extension programming (6). Its flexibility and adaptability to educational needs
of the general public, the resource base of research from universities and elsewhere, and
ongoing partnerships with federal, state, and local groups and organizations can ensure the
effectiveness of continued programs in radon education.
The work described in this paper was not funded
by the U.S. Environmental Protection Agency and
therefore the contents do not necessarily reflect the
views of the Agency and no official endorsement
should be inferred.
-------
ACKNOWLEDGMENTS
For assistance in the development and implementation of the radon education program, the
author is grateful to Mr. John Paul Reese of the New York State Energy Office and Mr.
Laurence Keefe of the New York State Department of Health. For their impressive local
programming efforts and assistance with this evaluation, the author is grateful to the following
Cooperative Extension Agents: Doris I. Broten, Frances Carlson, Pamela W. Chiverton,
Dianne Cooper, Suzanne F. Doin, Eileen Donahoe, Maxine Duroe, Nancy Fink, Katherine
Giacomi, Patricia Gere, Margaret Howe, Eleanor Z. Hibben, Maijorie L.N. Keith, Nancy
Potter, Mary C. Raymond, Ann Robson, Elizabeth Shields, Barbara Smith, Judy Webster,
and Ann Verbeck.
REFERENCES
1. Laquatra, Joseph and Peter S.K. Chi. "Radon in the Home Environment." The Nelson
A. Rockefeller Institute of Government Rgpnnt SffrigS» Spring, 1989.
2. Chi, Peter S.K. and Joseph Laquatra. "Energy Efficiency and Radon Risks in Residential
Housing." Energy, in press, 1990.
3. Bennett, Claude F. Analyzing Impacts of Extension Programs, U.S. Department of
Agriculture (ESC-575), 1976.
4. Weiss, Carol H. Evaluation Research. Englewood Cliffs: Prentice-Hall, Inc., 1972.
5. Rivera, William M,; Claude F. Bennett; Sharon M, Walker. Designing Studies of
Extension Program Results: A Bsssairss Specialists (Volume I).
Cooperative Extension Service, University of Maryland, lyoi.
6. Futures Task Force to the Extension Committee on Organization and Policy. Extension ill
Transition. Virginia Cooperative Extension Service, 1987.
-------
Session II:
Radon Related Health Studies
-------
II-l
LUNG CANCER MORTALITY AMONG NONSMOKING
URANIUM MINERS EXPOSED TO RADON DAUGHTERS
by: Robert J. Roscoe
Kyle StenlaTid
National Institute for Occupational
Safety and Health
Cincinnati, OH
William E. Halperin
Harvard School of Public Health
Boston, MA
James J. Beaumont
University of California at Davis
Davis, CA
Richard J. Waxweiler
Centers for Disease Control
Atlanta, GA
ABSTRACT
Radon daughters, both in the workplace and in the household, are a
continuing cause of concern because of the well-documented association between
exposure to radon daughters and lung cancer. To estimate the risk of lung
cancer mortality among nonsmokers exposed to varying levels of radon
daughters, 516 white men who never smoked cigarettes, pipes, or cigars were
selected from the U.S. Public Health Service cohort of Colorado Plateau
uranium miners and followed up from 1950 through 1984. Age-specific mortality
fates for nonsmokers from a study of U.S. veterans were used for comparison.
Fourteen deaths from lung cancer were observed among the nonsmoking miners,
while 1.1 deaths were expected, yielding a standardized mortality radio of
12.7 with 95% confidence limits of 8.0 and 20.1. These results confirm that
exposure to radon daughters in the absence of cigarette smoking is a potent
carcinogen that should be strictly controlled.
Copies of this paper will be distributed separately.
-------
11 - 2
RADON AND LUNG CANCER AMONG NEW JERSEY WOMEN
by: Janet Schoenberg, Judith Klotz, Homer Wilcox
New Jersey State Department of Health
Trenton, NJ 08625-0369
Gerald Nicholls
New Jersey State Department of Environmental
Protection
Trenton, NJ 08625-0027
ABSTRACT
An epidemiologic study previously conducted in New Jersey women was extended
to examine the association of lung cancer with radon exposure. The substudy
included 433 cases and 402 controls who lived in a single "index" residence for
10+ years during the period 10-30 years prior to diagnosis or selection. Lung
cancer risks showed a significant trend (p->0.04) with increasing year-round
living area radon concentrations (based on alpha track measurements), and a
Weaker (p-0.09) trend with estimated cumulative radon exposure. The relative
risk coefficient of 3.4% per working level month (WLM) was consistent with the
range of 0.5-4%/WLM generally reported for underground miners. These study
Results must be interpreted cautiously due to the small number of subjects with
high radon exposures and the possibility of selection biases. Nevertheless, the
study suggests that findings of radon-related lung cancer in miners can be
Applied to the residential setting.
INTRODUCTION
Epidemiologic studies of miners have shown a strong and consistent dose-
r®sponse relationship between lung cancer mortality and cumulative radon exposure
(1*2). Extrapolations of the miners' data to levels of residential exposure
"uggest a substantial risk in some houses (2-7). However, there are uncertainties
these extrapolations. Also, many citizens continue to express doubt about
-------
the seriousness of elevated radon exposure in their homes. Therefore, studies
specifically in the residential setting are needed to determine whether and to
what extent radon in homes is associated with increased lung cancer risk.
Case-control studies are well suited to addressing this issue. Persons with
lung cancer and similar persons who do not have lung cancer are characterized
with respect to past exposure to radon and other factors such as smoking. The
relationship between radon and lung cancer is measured by the "relative risk,"
which is the rate of lung cancer among those exposed to higher levels of radon,
compared to the rate of lung cancer among those exposed to lower amounts. In
case-control studies, the relative risk is approximated by the "odds ratio"; it
is possible to take smoking and other factors into account by calculating an
"adjusted odds ratio."
Several case-control studies (8-13) conducted primarily in Sweden have
suggested an association between residential radon and lung cancer risk. However,
some of these studies had very few subjects, included actual radon measurements
in only a sample of houses, or did not account adequately for smoking and other
lung cancer risk factors.
In 1985, we recognized that we could address some of these questions on radon
by extending a female lung cancer case-control study which we had been
conducting. The collaborators in this study extension included the New Jersey
Departments of Health (NJDOH) and Environmental Protection (NJDEP) and the U.S.
National Cancer Institute.
METHODS
The original study cases included all female residents of New Jersey who
were newly diagnosed with lung cancer from August 1982 through September 1983
(14). The controls were population-based, i.e., selected at random from the New
Jersey population, through drivers' license, Medicare, and death certificate
files. In-person interviews, including lifetime smoking, occupational, and
residential (towns only) histories and detailed information on dietary vitamin
A consumption, had been completed for 994 (76%) of the 1,306 eligible cases and
for 995 (69%) of the 1,449 eligible controls.
In order to examine risk in relation to radon exposure, subjects or their
next of kin were recontacted to collect information on specific addresses where
each subject lived in the 30 years prior to case diagnosis or control selection.
Because the budget only allowed for measurements in one house per subject, a
residence criterion was established: subjects were included in the radon substudy
if they had lived in a single "index" residence for at least 10 years in the 10-
30 year period prior to diagnosis or selection. This residence criterion also
allowed for sufficient duration of exposure and assumed a minimum ten-year period
between relevant radon exposure and lung cancer diagnosis (15).
A year-long alpha track detector measurement of radon (Type SF, Terradex
Radon Detection Products, Glenwood, IL) was conducted in the living area of the
index residence and was assumed to provide the best estimate of the average radon
/
-------
levels to which a subject had been exposed when she was a resident of this house.
A second alpha track detector was also placed in each house, generally in the
basement. In about 15% of the houses, a third detector was paired with one of
the others as a quality control check.
At each index residence, 4-day charcoal canister measurements were also
conducted, usually under closed house conditions during the heating season.
These screening tests were included in case some houses had very high levels of
radon which would need immediate remediation, and as a back-up in case the alpha
track tests were not completed. This protocol also allowed for a comparison of
basement screening measurements with year-round living area measurements. House
construction and ventilation data were collected, including information on
changes which had occurred since the current occupants had moved into the house.
Cumulative radon exposure was estimated for cases and controls based on the
living area alpha track results and the number of years the subject lived in the
index residence during the period 5-30 years prior to diagnosis or selection.
This time period was used because data had been published (2,16) which indicated
a shorter period (five years) between relevant radon exposure and diagnosis of
lung cancer than had been assumed at the time of study design. For each year
during this 25-year period when the subject did not live at the index residence,
it was assumed that exposure was minimal, i.e., equivalent to the median
concentration among controls (0.6 pCi/L).
STATISTICAL METHODS
Odds ratios and 90% confidence intervals (CI) for the association of lung
cancer with radon or with cumulative radon were estimated by multiple logistic
regression analysis (17) using the microcomputer-based LOGRESS program (18).
The final logistic model included adjustment for lifetime average number of
cigarettes per day, time since smoking cessation, age, occupation, the type of
respondent (subject or next of kin) who provided the information in the original
interview, and interaction terms between respondent type and cigarettes per day.
Other variables such as race, education, county of residence, vegetable
consumption (a measure of dietary vitamin A), duration of smoking, and# cigarette
tar content were considered in the analysis but were not included in the final
model. Trends in risk with increasing radon exposure were calculated using
logistic regression with a weighted categorical exposure variable.
RESULTS: RISK ESTIMATES
Of the original 994 cases and 995 controls, 433 (44%) of the cases and
402 (40%) of the controls were included in the radon substudy. The remaining
subjects were not included because address-specific information could not be
collected (140 cases, 126 controls), because no address met the residence
criterion (253 cases, 256 controls), or because radon tests could not be
conducted at the index residence (168 cases, 211 controls).
Among the substudy subjects, year-long living area alpha track measurements
were completed for 346 cases and 318 controls. For the remainder, living area
-------
alpha track results were estimated from basement alpha tracks (27 cases, 28
controls) or from canisters (38 cases, 39 controls). For a small number of
subjects (22 cases, 17 controls) who lived in apartments on the third floor or
higher, year-round living area radon levels were assumed to be less than 1 pCi/1.
Table 1 shows the distributions of cases and controls by year-long living
area alpha-track measurements or estimates, as well as the adjusted odds ratios
at each exposure level. The highest living area alpha track measurement was 11.3
pCi/L. Because of the small number of subjects in the 2-3.9 and 4-11.3 pCi/L
groups, these were combined, giving an adjusted odds ratio of 1.8 (90% CI 0.89,
3.5). None of the individual odds ratios was statistically significant, but
there was a significant trend of increasing risk with increasing radon (1-sided
p value - 0.04).
For comparison purposes, Figure 1 shows odds ratios for the associations
of lung cancer with radon and with smoking. Lung cancer risk increased with each
factor, but much more with smoking. Among smokers of 15-24 cigarettes per day,
for example, the odds ratio was 11.2 (90% CI 7.8,16.1), relative to women who
had never smoked.
Table 2 shows the distributions of cases and controls by estimated
cumulative radon exposure, as well as the adjusted odds ratios at each exposure
level. The odds ratio of 7.2 for the 100-155 pCi/L-years group was statistically
significant, but was based on only 4 cases and 1 control. Because of these small
numbers, the subjects in the 50-99 and 100-155 pCi/L-years groups were combined,
giving an adjusted odds ratio of 1.4 (90% CI 0.65,3.0). There was a pattern of
increasing risk with increasing cumulative radon, but the trend was not
statistically significant (1-sided p value-0.09).
Figure 2 shows the odds ratios for cumulative exposure in units of working
level months (WLM) and compares these to the average risk in some of the miners
studies (2). The pattern of increasing risk with increasing residential exposure
is consistent with the pattern for occupational exposure.
Another way of comparing the residential data with the miners data is to
calculate a relative risk coefficient, or the increase in risk over background
risk for each working level month (WLM) of exposure. In this study, the relative
risk coefficient was 3.4% per WLM (90% CI 0,8.0%), which is again quite
consistent with the range of 0.5-4% reported for the miners studies (2,5).
RESULTS: COMPARISON OF SCREENING VS. YEAR-ROUND MEASUREMENTS
In addition to the findings in relation to lung cancer risk, the study
design allowed comparisons to be made between year-round living area alpha track
measurements and 4-day screening measurements using charcoal canisters. Table
3 gives the average value and range for living area alpha track measurements in
different categories according to the basement screening results. For the 474
houses which screened below 4 pCi/L, the year-round living area alpha track
measurement was always less than 4 pCi/L.
/
-------
Table 3 also gives, for each basement screening category, the ratios between
living area alpha tracks and basement canisters, between living area and basement
canisters, and between living area alpha tracks and living area canisters. For
houses with basement screening measurements less than 1 pCi/L, the year-long
living area alpha tracks averaged 64% of the basement canisters. However, as
the screening measurement increased, the year-long living area test represented
an increasingly smaller percentage of these canisters. For houses which screened
above 4 pCi/L, the year-long living area alpha track averaged only about 20% of
the screening value. This difference was related both to the difference between
basement and first floor results, and to the difference between 4-day heating
season and year-round results. Except at the lowest radon level, living area
canister results were about 40% of basement canister results. The ratio of
canisters to year-long alpha track results, both in the living area, decreased
with increasing screening concentrations and was around 50% for those houses
which screened above 4 pCi/L in the basement.
DISCUSSION
This is the first large-scale study of lung cancer and indoor radon based
on actual measurements for all subjects (except those who lived in apartments
above the second floor) as well as detailed smoking, occupational, and dietary
histories. Furthermore, year-long radon measurements in the living areas of
houses were completed for most subjects.
The study found a significant trend, with increasing lung cancer risk for
increasing radon in homes, even after taking smoking and other risk factors into
account. The increase in risk for each unit of cumulative exposure was
consistent with the results of the miners studies. This supports the
extrapolation of the miners data to the residential setting.
Questions might be asked about the generally low exposure levels which were
found in this study. In a radon surveillance study conducted for the NJDEP, 33%
of the 5,727 houses tested had basement screening values of 4 pCi/L or higher
(19). However, in this study, only 13% of houses had basement screening values
of 4 pCi/L or higher, and only 1% of the houses had year-long living area
measurements at this level.
There are several reasons for this difference. This case-control study was
statewide and population-based, with more houses in heavily populated counties
of New Jersey, which tend to have lower radon levels. Only 11% of the houses
were in five high radon counties. On the other hand, the NJDEP study was designed
to identify areas in the state with the most serious radon problems; therefore,
it used a geographic sampling scheme, and 58% of the houses were in five high
radon counties. A county-population weighted average for the NJDEP data suggests
that only 14% of houses in New Jersey have basement screening levels above 4
pCi/L.
In this study, even within high radon counties, most of the houses were in
more heavily populated urban or suburban areas, rather than rural areas. In the
NJDEP study, these urban and suburban houses tended to have lower radon levels
-------
than rural houses. Another difference involved the residence criterion; this
study included only houses where the subjects had lived for at least 10 years
in the 10-30 years prior to diagnosis. This meant that these houses had to be
at least 25 years old, and most were more than 40 years old. In the NJDEP study,
radon levels tended to be lowest for these older houses. Finally, as shown in
the results of Table 3, there was a substantial difference between year-round
living area and basement screening measurements, particularly at higher levels
of exposure.
The results of the exposure comparisons between charcoal canisters and year-
round living area alpha track detectors have several additional implications.
They suggest that basement screening results are usually exaggerations of typical
radon concentrations inhaled by people on a year-round basis. This supports the
use of follow-up tests, especially in the living area, rather than screening
tests, before making decisions on remediation. Such procedures are already
advised by the USEPA, the NJDEP, and the NJDOH, but are not necessarily
understood by the public. The finding that all houses which screened below 4
pCi/L had year-round living area measurements below 4 pCi/L means that making
decisions on the advisability of follow-up testing based on these screening
measurements would not have resulted in any houses with high exposures in the
living areas being missed.
The results of this study with respect to the radon-lung cancer association
need to be interpreted cautiously, not only because of the small numbers of
houses with high exposures, but also for several other reasons. The possibility
of selection biases is a weakness of the study. The residence criterion and the
need to have the cooperation of both the original interview respondents and the
current occupants of homes reduced the number of subjects originally eligible
to those who were included in the substudy. Comparison of subjects with
interviews who were and were not included in the radon substudy showed some
differences (20). It is unknown whether any of these differences resulted in
a substudy sample which was biased with respect to radon distribution.
Another weakness of any case-control study of this type is that exposure
data are collected in the present time, when the exposure of interest actually
occurred in the past. There is the possibility that changes in house
construction, heating, ventilation, and occupants' activity could cause major
inaccuracies in the exposure estimates. Although limited data on house
construction changes were collected (20), it is unclear what impact these
specific changes would have had on the radon exposures.
In this study, detailed information on smoking habits, occupation, diet,
and other demographic characteristics were available and were taken into account
in the analysis. However, it is still possible that the observed relationship
between lung cancer and the low levels of radon in this study is distorted due
to other factors which were not evaluated in the analysis.
Finally, the results of the analyses of risk in relation to cumulative radon
exposure were slightly weaker than those for radon concentration. This is
probably related to our use of a conservative exposure estimate for those years
when the subject did not live in the index residence. Measurements in additional
/
-------
houses which are now being conducted as part of a second phase of this study may
partially correct this problem.
The findings of this study need to be corroborated by other residential
radon studies currently underway in New Jersey and worldwide. Despite the
limitations discussed above, these results in combination with previous
occupational, residential and experimental data suggest that radon is a
carcinogen in the residential setting. Thus, the findings support the
comprehensive radon-related programs which various agencies have developed based
on extrapolations from the miner data, and the actions to reduce radon exposure
to the lowest feasible levels. Finally, the study reiterates the well-known
finding that smoking is the major risk factor for lung cancer. Radon-related
lung cancer risk at low, residential exposure levels is modest in comparison to
smoking-related risk. This again emphasizes the importance of avoiding both
smoking and high radon exposure, and the importance of including smoking
avoidance education along with radon reduction activities, in order to have the
greatest public health impact.
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the views
of the Agency and no official endorsement should be Inferred.
ACKNOWLEDGEMENTS
This study was funded in part by a special New Jersey Legislative
appropriation, and in part by Contract N01-CP-61031 and Grant R01-CA-37744 from
the National Cancer Institute. Other persons who made important contributions
to this study include Maria Gil-del-Real and Annette Stemhagen (NJDOH), Mary
Cahlll (NJDEP), and Zdenec Hrubec and Thomas Mason (National Cancer Institute).
Administrative support and manuscript reviews were provided by William Parkin,
Rebecca Zagraniski, and Thomas Burke (NJDOH). We also thank Victor Archer, Olav
Axelson, and Jonathan Samet for reviewing the initial technical report, and Karen
Tuccillo (NJDEP), Joanne Bill and Asora Carpenter (NJDOH), and other field staff
for assistance with data collection.
REFERENCES
1. National Council on Radiation Protection and Measurements (NCRP).
Evaluation of occupation and environmental exposures to radon and radon
daughters in the United States. NCRP Report 78, National Council on
Radiation Protection and Measurements, Bethesda, Maryland, 1984, 204 pp.
2. National Research Council, Committee on Biological Effects of Ionizing
Radiation. Health risks of radon and other internally-deposited
alpha-emitters: BEIR IV. National Academy Press, Washington, D.C. 1988.
602 pp.
-------
3. Harley, N. Radon and lung cancer in mines and homes. New Engl. J. Med.
310: 1525, 1984.
4. Radford, E.P. Potential health effects of indoor radon exposure. Environ.
Health Perspec. 62: 281, 1985.
5. Klotz, J.B. Estimating lung cancer risks of indoor radon: applications
for prevention. In: Indoor Radon - APCA Specialty Conference Proceedings.
Air Pollution Control Association, Pittsburgh, Pennsylvania, 1986. p. 37.
6. Jacobi, W. , Lafuma, J., and Land, C.E. Lung cancer risk from indoor
exposures to radon daughters. A report of a Task Group of The
International Commission on Radiological Protection. Annals ICRP 17: 1,
1987.
7. Samet, J.M. Radon and lung cancer. J. Natl. Cancer Inst. 81: 745, 1989.
8. Axelson, 0., Edling, C., and Kling, H. Lung cancer and residency - A case
referent study on the possible impact of exposure to radon and its
daughters in dwellings. Scand. J. Work Environ. Health 5: 10, 1979.
9. Edling, C., Kling, H., and Axelson, 0. Radon in homes - A possible cause
of lung cancer. Scand. J. Work Environ. Health 10:25, 1984.
10. Svensson, C., Eklund, G., and Pershagen, G. Indoor exposure to radon from
the ground and bronchial cancer in women. Int. Arch. Occup. Environ.
Health 59: 123, 1987.
11. Lees, R.E., Steele, R., and Roberts, J.H. A case-control study of lung
cancer relative to domestic radon exposure. Int. J. Epidemiol. 16: 7,
1987.
12. Axelson, 0., Anderson, K., Desai, G., et al. Indoor radon exposure and
active and passive smoking in relation to the occurrence of lung cancer.
Scand. J. Work Environ. Health 14: 286, 1988.
13. Svensson, C., Pershagen, G., and Klominek, J. Lung cancer in women and
type of dwelling in relation to radon exposure. Cancer Res. 49: 1861,
1989.
14. Schoenberg, J.B., Wilcox, H.B., Mason, T.J., et al. Variation in
smoking-related lung cancer risk among New Jersey women. Am. J. Epidemiol.
130: 688, 1989.
15. Whittemore, A.S. and McMillan, A. Lung cancer mortality among U.S. uranium
miners: A reappraisal. J. Natl. Cancer Inst. 71: 489, 1983.
16. Howe, G.R., Nair, R.C., andNewcomb, H.B. Lung cancer mortality (1950-80)
in relation to radon daughter exposure in a cohort of workers at the
Eldorado Beaverlodge uranium mine. J. Natl. Cancer Inst. 77: 357, 1986.
-------
17. Breslow, N.E. and Day, N.E. The Analysis of Case-Control Studies. IARC
Scientific Publication No. 32. International Agency for Research on
Cancer, Lyon, France, 1980. 338 pp.
18. McGee, D. A program for logistic regression on the IBM PC, Am. J,
Epidemiol. 124: 702, 1986.
19. New Jersey State Department of Environmental Protection. Statewide
scientific study of radon. New Jersey State Department of Environmental
Protection, Trenton, NJ, 1989.
20. New Jersey State Department of Health. A case-control study of radon and
lung cancer among New Jersey women: Technical report, Phase I. New Jersey
State Department of Health, Trenton, New Jersey, 1989. 85 pp.
-------
Luna cancer Relative Eisi
24
22
20
16
16
14
12
10
6
6
4
RADON
I—If3!
SMOKING
<1 1-1.9 2-11
pCl/I
0 <15 1S-2425-74
Clffarettes/Day
Figure 1. Association of lung cancer with year-round living area radon
concentrations or with lifetime average number of cigarettes smoked per day,
New Jersey radon-female lung cancer study, 1982-1988. Bars and error lines
represent odds ratios + 90% CI.
Lung Cancer Relative Risk
3,5
2.5
1.5
0.5
r
long cancer zlst
and KT residential
radon szposuibs
long cancer nsi
and occupational
radon exposures
_L
_l_
50 100 150
Cumulative Radon Exposure CWLM3
Figure 2. Association of lung cancer with cumulative radon exposure in
occupational studies of miners and in the New Jersey radon-female lung cancer
study, 1982-1988. Occupational data is averaged from studies summarized in
reference 2. Errors bars for residential data represent 90% CI.
/
-------
TABLE 1. DISTRIBUTION OF LUNG CANCER CASES AND CONTROLS BY YEAR-LONG LIVING AREA
ALPHA-TRACK RESULTS*, NEW JERSEY RADON-FEMALE LUNG CANCER STUDY, 1982-1988
Radon (pCi/L)
<1.0 1-1.9 2-3.9 4-11.3 Total
Cases 342 (79.0%) 67 (15.5%) 18 (4.2%) 6 (1.4%) 433
Controls 324 (80.6%) 66 (16.4%) 10 (2.5%) 2 (0.5%) 402
Adjusted OR| 1.0 1.1 1.3 4.2
(90% CI) (0.79,1.7) (0.62,2.9) (0.99,17.5)
* Living area alpha-track measurements (n-664); estimates derived from basement
alpha track or charcoal canister measurements, or for index residences which were
apartments above the second floor (n-171).
| Odds ratios (OR) and 90% CI, adjusted for lifetime average cigarettes/day,
years since smoking cessation, age, occupation, respondent type, and the
interaction terms between respondent type and cigarettes/day. Test for trend
in OR with increasing radon: p-0.04. OR for 2-11.3 pCi/L: 1.8 (0.89,3.5).
TABLE 2. DISTRIBUTION OF LUNG CANCER CASES AND CONTROLS BY ESTIMATED CUMULATIVE
RADON EXPOSURE*, NEW JERSEY RADON-FEMALE LUNG CANCER STUDY, 1982-1988
Cases
Controls
Adjusted ORf
(90% CI)
Cumulative radon (pCi/L-vears)
_<25_
361 (83.4%)
340 (84.6%)
1.0
25-49
56 (12.9%)
52 (12.9%)
1.2
(0.83,1.9)
50-99
12 (2.8%)
9 (2.2%)
0.94
(0.41,2.2)
100-155
4 (0.9%)
1 (0.2%)
7.2
(1.0,50.3)
Total
433
402
* Cumulative radon exposure during 25 years from 5-30 years prior to case
diagnosis or control selection; assumes exposure of 0.6 pCi/L (median for
controls) for any of the 25 years during which the person did not live in the
index residence where the measurements were made.
f Odds ratios (OR) and 90% CI, adjusted for lifetime average cigarettes/day,
years since smoking cessation, age, occupation, respondent type, and the
interaction terms between respondent type and cigarettes/day. Test for trend
in OR with increasing cumulative radon: p-0.09. OR for 50*155 pCi/L-years: 1.4
(0.65,3.0).
-------
TABLE 3. COMPARISON OF RADON MEASUREMENTS USING YEAR-ROUND LIVING AREA ALPHA TRACK DETECTORS (LivATD)
AND FOUR-DAY BASEMENT AND LIVING AREA CANISTERS* (BasCan, LivCan), BY BASEMENT CANISTER RADON LEVEL,
NEW JERSEY RADON-FEMALE LUNG CANCER STUDY, 1982-1988.
LivATD Ratios of radon measurement results (%)
BasCan
(uCi/L)
Number of
houses
Average*
roCi/L)
Range
roCi/D
LivATD/BasCan
LivCan/BasCan
LivATD/LivCan
<1
221
0.4
0.1-2.0
64%
100%
64%
1-1.9
82
0.6
0.1-2.0
36%
41%
88%
2-3.9
171
0.7
0.1-3.7
24%
33%
73%
4-7.9
69
1.0
0.2-3.4
18%
32%
56%
8-20
17
2.4
0.3-11.3
20%
40%
49%
* Canister measurements conducted under closed house conditions during the heating season,
t Geometric mean.
-------
II-3
Estimate of Annual Radon-Induced Lung Cancer Deaths -
EPA*s Approach
Anita Schmidt
Jerome s. Puskin
Christopher Nelson
Neal Nelson
Office of Radiation Programs
U.S. Environmental Protection Agency
ABSTRACT
EPA has revised its previous estimate of annual radon-induced lung
cancer deaths to the general population from its previous estimate
of 5000 to 20,000 given in the Citizen's Guide to Radon. The
revised central estimate is approximately 21,000 deaths per year.
The Agency is also using the range of uncertainty described in ICRP
50 report which gives a range of approximately 8000 to 40,000 lung
cancer deaths per year attributable to radon.
The revised lung cancer deaths are are based on BEIR IV and ICRP
50 risk estimates (lung cancer deaths per 10+6 WLM) derived from
risk coefficients from the miner studies. The calculation of the
number of annual deaths is also based on revised exposure
assumptions. A dose conversion factor was previously used to
reflect possible differences in the dose per unit exposure between
miners and people exposed indoors to radon. Based on the
recommendations of outside experts, EPA believes that risk
estimates for mining populations can be used directly to predict
risks to the general population without this factor.
-------
II-4
RADON DAUGHTER EXPOSURE IN DWELLINGS AND MULTIPLE MYELOMA
Inge Tell, Robyn Attewell, and Inger Bensryd
Department of Occupational and Environmental Medicine
University Hospital
S—221 85 Lund, Sweden
Sven Hertzman
Department of Radiation Physics
Sahlgren Hospital
S-413 45 Goteborg, Sweden
Gilbert Jonsson
Department of Physics
Lund Institute of Technology
Box 118, S—221 00 Lund, Sweden
ABSTRACT
The results of a case-referent study on a possible association between
myelomatosis and exposure to radon in dwellings are presented. Long term
residents, 10 years or more in the same houses, were studied in three muni-
cipalities in southern Sweden, where 64 cases of myelomatosis had been
diagnosed during the period 1971 to 1984. A case control study was designed
with a five to one matching scheme, selecting controls by means of strati-
fied random sampling from the population register, and matching controls to
cases on sex, age, death year, if dead, and on municipality. Radon daughter
levels were measured with SSNTD technique in each dwelling. Data on reci-
dence type, smoking habits, and other life style factors, were obtained by
a telephone interview, based on a questionnaire. Information from interviews
and radon levels were obtained for 49 cases and 186 matched referents. Ihe
arithmetic mean radon daughter level for the whole material was 46,Bq/m .
The geometricynean radon daughter levels was for referents 18 Bq/m , and for
cases 25 Bq/m , which is not a statistically significant difference.
-------
INTRODUCTION
Several studies have suggested that ionizing radiation may induce mye-
loma. Increased risk has been reported among radiologists (1,2), atomic bomb
survivors in Japan (3), and workers in nuclear plants (4). The exposure to
radiation of such populations is probably mainly external. However, exposure
to radon has been reported as a possible cause of leukaemia (5,6).
The lung is the main target organ of radiation from radon daughters,
and other anatomic sites are considered to receive comparatively little
radiation from radon exposure. However, reports showing internal contami-
nation of individuals by radon daughters have been presented (7). The possi-
bility of skeletal doses due to ^ Po has been discussed (8). Also, measure-
ments of Rb in lungs have led to discussion of recirculation of 210pb
from accumulated skeletal burdens to the lungs (9). Circulation of internal
radon daughters in the body can hardly be accounted for without assuming
the lymphatic fluid as a transportation medium.
The problem of a dose to B-lymphocytes, whose biological purpose has not
been finally determined, should not be disregarded, even if the dose to the
bone marrow is actually low. Knowledge is lacking of where in the human body
the development and determination of B-lymphocytes takes place. In birds,
this process occurs in the bursa of Fabricius, but in man no equivalent to
this tissue has been identified. It has been assumed that B-lymphocytes are
determined in what is called GALT, i.e. "gut associated lymphatic tissue".
If this is the case, B-lymphocytes may be assumed to occur more or less
diffusely along the gastrointestinal tract. There, they might be exposed to
various toxic agents and also to radon daughters. Also, even if most of the
alpha activity is released in the respiratory tract, there might be an ex-
posure to lymphatic tissue caused by transportation of radon progeny by
macrophages to lymph nodes located close to the respiratory tract.
Thus, there is an underlying assumption of a low dose exposure to
lymphatic tissue and to undetermined B-lymphocytes behind this study.
Moreover, there is an indication of excess risk, not only for lung cancer,
but also for lymphoma in radon-exposed underground miners (10). The possi-
bility of a relation, in the general population, between cancers other
than lung cancer and exposure to radon in ground water has also been
discussed (11).
STUDY DESIGN AND METHODS
The study was carried out in three municipalities in southern Sweden.
In two of these municipalities, U-rich black shale of cambrian age occurs
in the bedrock. The U content of the shales varies between 50 and 100 ppm.
The soils are mainly glacial tills, both gravel, sand, and clay, but also,
to some extent, eskers and other glaciofluvial deposits.
-------
The total population is approximately 42,000 persons. Compared to most
other parts of Sweden, a large fraction of the population is connected with
agriculture, living on farms or in detached houses in the countryside, or in
small villages within an area of 1,010 km . In each of the municipalities,
there is an urban center, and in each of these centers live approximately
5,000 persons.
During the five year period 1977 to 1982, the total cancer incidence
in the three municipalities did not differ compared with expected incidence
for Sweden.
From the local registers of two county hospitals, the names and birth
data were obtained for persons, who were diagnosed during the period from
1971 to 1984 with multiple myeloma. During the 14 year period, 64 cases of
multiple myeloma were diagnosed. The number of diagnosed cases was checked
against data from the Swedish National Cancer Register. In some cases, per-
sons now living in the actual dwellings were unwilling to have radon mea-
surements made in their homes. Also, only cases who had lived at the same
address (and in the same dwelling) during 10 of the 15 years prior to
diagnosis were included in the study. The study group thus consists of
49 cases of multiple myeloma.
The study was designed with a five to one matching scheme, in which
controls were selected by means of stratified random sampling from the
Population register. The controls were matched to the cases on sex, birth
and death year, if dead, and on domicile. The matching criteria, however,
reduced the sampling frame and in some domiciles less than 5 controls could
by matched to the case. In total 186 controls entered the study. Primarily
selected controls, who had a diagnosed neoplastic disease, were excluded.
Radon daughter levels were measured with solid-state nuclear track
detector technique (12) and expressed as Becquerel/nr (Bq/m-*). The detec-
tors were exposed in each house for three months during the period January-
May, which is mainly during the heating season in Sweden.
Data on previous illnesses, medication, number of radiographics, smo-
king habits (of case/control and spouses), and a number of other life style
factors, believed to be of possible importance, were obtained by a tele-
Phone interview, based on a questionnaire, and undertaken by a nurse. As
the majority of the participants were not alive, these interviews had to
be made with next of kin of the subjects. This was possible with only a few
Exceptions.
Data on building material, basement construction, ventilation, changes
ventilation, heating, major renovations, use of private well, and other
technical or hygienic factors were obtained in telephone interviews with
the persons living in the dwellings at the period of measurement.
The Mann-Whitney U-test was used to compare the distribution of radon
levels for cases and referents ignoring the matching. Conditional logistic
Agression (as implemented in PECAN) was used for estimating relative risks
for different levels of radon (taking into account the matching). The radon
-------
levels were categorized into 3 groups with approximately equal numbers of
each group; the cutpoints being 15 and 35 Bq/m-5. A possible confounding by
smoking in the household was checked by including a smoking indicator
variable in the model (yes versus no). Also, possible effect modification
by smoking was checked by making separate relative risk estimates for the
3 radon levels in the presence and absence of smoking.
RESULTS
The radon daughter concentrations in the material are low, with few
exceptions. The geometric mean was 19 Bq/rrr5, and the range 0-406. When the
results were plotted geographically, it was obvious that low values prevail
in urban areas. In rural parts of the municipalities, with ll-rich bedrock,
higher values seem to cluster in areas characterized by porous glaciofluvial
ground in proximity to occurrences of black, U-rich shale.
In Table 1, radon daughter levels are presented as related to case-con-
trol status. Arithmetic mean radon daughter level for the whole material wa^
46 Bq/m . The geometric mean radon daughter level was for referents 17 Bq/m
and for cases 25 Bq/m . Ignoring the matching, the radon levels for the cases
were not significantly higher than for the controls (Mann-Whitney U test
p=0.3).
Table 2 shows the results of the conditional logistic regression model
involving radon daughter levels (in 3 groups) with adjustment for smoking
status.^A two-fold increase in risk is observed even for homes with at least
15 Bq/m . The relative risk estimate smoking versus non-smoking was 1.89
(1.00, 3.58). There was no apparent confounding effect by smoking since the
risk estimates in the 2 higher exposed groups were essentially unchanged
(2.02, 1.81) when smoking was excluded from the model.
In Table 3, the relative risks are shown by radon daughter level and
smoking status. The highest risks are observed for homes of smokers with
the highest radon daughter content. Despite the clearer trend in increasing
risk with radon levels among the smokers (1.0, 1.5, 1.8), compared with non-
smokers (1.0, 2.7, 2.0), the confidence intervals are broad, so there appears
to be major effect modification due to smoking. Similarly the higher risk
associated with smoking is observed at all 3 radon levels.
DISCUSSION
This study does not provide conclusive evidence for an association be-
tween exposure to radon in dwellings and a risk of developing multiple mye-
loma. However, the population studied is fairly small. Also, the average ra-
don exposure is low. In parts of the investigated areas, U-rich soils give
very high indoor radon daughter levels (up to 10.000 Bq/m have been found).
No such high levels we^e observed in this study. On the contrary, the arith-
metic mean was 46 Bq/m , which is below the mean value for Sweden. This has
been estimated as 50 Bq/m5 for the period 1980 to 1982 in buildings existing
in 1975. In 1956, the mean radon daughter level in Swedish homes was estima-
ted to have been 15 Bq/m .
-------
One of the results of this study is the increased relative risk
of developing multiple myeloma due to smoking. This is interesting be-
cause tobacco smoking, although generally known as a cancer hazard, is not
a well known factor as a cause of multiple myeloma. A larger study group
would enable a better investigation of a possible synergistic effect be-
tween smoking and radon exposure and the risk for developing myeloma.
The work described in this paper was not
funded by the U.S. Environmental Protection
Agency and therefore the contents do not
necessarily reflect the views of the Agency
and no official endorsement should be inferred.
REFERENCES
1. Lewis EB, Leukemia, multiple myeloma, and aplastic anemia in American
radiologists. Science 142:1492-1494, 1963.
2. Matanoski GM, Seltser R, Sartwell PE, et al, The current mortality
rates of radiologists and other physician specialists: specific causes
of death. Am J Epidemiol 101(3):199—210, 1975.
3. Ichimaru M, Ishimaru T, Mikami M, et al, Multiple myeloma among atomic
bomb survivors in Hiroshima and Nagasaki by dose, 1950-1976. Radiation
Effects Research Report, Technical Report 9-79, 1979.
4. Dolphin GW, A comparison of the observed and the expected cancers of
the haematopoietic and lymphatic systems among workers at Windscale:
a first report. National Radiological Protection Board, NRPB-R54,
Harwell, 1976.
5. Lucie NP, Radon exposure and leukaemia. The Lancet, ii 99-100, July 8,
1989.
6. Cook-Mozaffari P, Darby S, and Doll R, Cancer near potential sites of
nuclear installations. The Lancet, ii 1145-1147, November 11, 1989.
7. Stebbings JH, and Dignam OJ, Contamination of individuals by radon
daughters: a preliminary study. Arch Environ Health, 43(2): 149-154, 1988.
8. Clemente GF, Renzetti A, and Santori G, Assessment and significance of
the skeletal doses due to z Po in radioactive Spa workers. Environ
Research 18:120-126, 1979.
2io
9. Singh NP, Bennett DB, and Wreun ME. Concentrations of Pb and its
states of equilibrium with U. U, and Th in U miners' lungs.
Health Phys 51(4):501-507, 1986.
10. Samet JM, Radon and lung cancer. Review. I Nat Cancer Inst, 81(10):
745-757, 1989.
-------
222
11. Collman GW, Loomis DP, and Sandler DP, Radon concentration in
groundwater and cancer mortality in North Carolina. Int Arch Occup
Environ Health, 61(1/2):13-18, 1988.
222
12. Jonsson, G, Indoor Rn Measurements in Sweden with the Solid-State
Nuclear Trach Detector Technique. Health Phys 54(3):271-281, 1988.
-------
TABLE 1. RADON DAUGHTER CONCENTRATIONS (BECQUEREL/M^) FOR CASES
AND REFERENTS
Referent
Case
Total
Median
23.5
27.0
24.0
Mean (arithmetic)
45.8
46.7
46.0
Mean (geometric)
17.8
24.5
18.9
Range
0-406
0-330
0-406
Valid N
186
49
235
TABLE 2. RELATIVE RISK ESTIMATES AND 95 % CONFIDENCE INTERVALS
FOR MULTIPLE MYELOMA ACCORDING TO RADON LEVELS MEASURED AT HOME;
WITH ADJUSTMENT FOR SMOKING (CONDITIONAL LOGISTIC REGRESSION).
Radon (Bq/m^)
Referents
Cases
RR
(95 % CI)
0-14
64
11
1.00
15-34
56
19
2.16
(0.87, 5.36)
35+
66
19
1.89
(0.79, 4.54)
TABLE 3. DISTRIBUTION OF REFERENTS AND CASES ACCORDING TO RADON LEVELS
AND SMOKING STATUS. RELATIVE RISKS (WITH LOWEST RADON LEVELS IN NON-
SMOKING HOUSEHOLDS AS THE REFERENT CATEGORY) AND 95 % CONFIDENCE INTER-
VALS CALCULATED WITH CONDITIONAL LOGISTIC REGRESSION.
Referent Case
N
%
N
%
RR
95 % CI
SMOKING
Rn (Bq/m^)
No;
0-14
44
23.7
5
10.2
1.00
15-34
40
21.5
12
24.5
2.68
(0.84, 8.57)
35+
47
25.3
10
20.4
1.96
(0.59, 6.47)
Yes:
0-14
20
10.8
6
12.2
2.41
(0.68, 8.56)
15-34
16
8.6
7
14.3
3.60
(0.94, 13.88)
35+
19
10.2
9
18.4
4.32
(1.26, 14.88)
Total
186
100
49
100
-------
WHOLE BODY COUNTING OF RADON DAUGHTERS
R. A. Schlenker, M. A. Essling, J. H. Stebbings, and H. F. Lucas
Biological and Medical Research Division
Argonne National Laboratory
Argonne, Illinois 60521
ABSTRACT
Five adult males were exposed for one hour to radon and
radon daughter products in an exposure chamber and were
subsequently measured for radon daughter product activity in
the chest region by whole body counting methods. The
gamma-ray detection rate appeared to diminish as a single
exponential with 35 minute half period, a form that is
consistent with the physical decay of a mixture of RaB and
RaC. About half of the deposited activity was associated with
internal deposition and half with external deposition on
clothing, skin and hair. The average counting rate from radon
daughters on clothing was 10 times the average from skin and
hair. These two components together contributed 51.2% of
the total counting rate. A substantial contribution was made
to the counting rate from deposits on under as well as outer
clothing. A strong but statistically non significant correlation
was found between internal and external deposition indicating
that total activity provides an index of internal deposition.
-------
INTRODUCTION
The quantitative uptake of radon daughter products by the lung is normally estimated
from mathematical models of aerosol and unattached daughter product deposition on the
mucous membranes of the bronchial tree and alveolar sacs. However, the technology and
methods for experimental observation of radon daughter uptake have long existed, due
to interest in the health effects of radium deposition in the body. Both 22^ln and Ra
give rise to gamma-ray emitting decay chains of radon daughter products whose emissions
can be easily detected by standard radiation detectors. The earliest applications of this
method for radon daughter product detection employed electroscopes, and later Geiger-
Muller tubes, without spectral resolution. Current applications are based on sodium iodide
and semiconductor detectors to achieve moderate or high spectral resolution, respectively,
of the emitted gamma-rays, and permit the separate measurement of RaB and RaC. An
excellent review of the history of radium, i.e. radon daughter product, detection in the
body and of the current use of sodium iodide systems has been published by Toohey et
al. (1).
Our laboratory has employed whole body counting methods for radon daughter
detection for several decades while investigating radium and radon gas retention in
humans. The systems in use by us (1) are stationary and mobile whole body counters
that employ thallium activated sodium-iodide crystals (designated Nal(Tl)) of large size,
in the shape of right circular cylinders. The stationary whole body counters are
subterranean and shielded by concrete, steel, and lead to reduce terrestrial gamma-ray
and cosmic-ray background, and are located in rooms supplied with radon-free air to
reduce background from airborne radon daughter products. Our mobile detector is
contained in a trailer which can be hauled to field locations. Gamma-ray backgrounds
and background variability are higher in the mobile counter than in the stationary
counters due to the use of thinner shielding and untreated air in the mobile counter.
During the course of field investigations of gamma-ray emission from former radium
workers in Bloomsburg, Pennsylvania, we observed easily measurable levels of radon
daughter products in the bodies of some subjects (2). The radon daughters were first
attributed to radium retained from occupational exposure but later identified as radon
daughters directly deposited from indoor air or as the daughters of radon dissolved in
body tissues from breathing indoor air. An immediate question for the interpretation
of these data, many of which were collected with the subjects in their street clothes, was
whether or not the detected gamma rays were due primarily to radon daughters on the
clothing, body surfaces and hair, or to radon daughters in the body. A corollary was
whether or not total daughter product activity was correlated with internally deposited
activity. If so, total activity would be an index of internally deposited activity and could
be used as an index of lung deposition. This question has been investigated by
experimental exposures to radon and radon daughters in an exposure chamber. The
initial work in this series of experiments is reported in this paper.
-------
EXPERIMENTAL DESIGN AND METHODS
The subjects in this study were exposed to radon and radon daughter products and
measured for daughter product activity in our stationary whole body counter in order to
determine the time course of daughter product retention. The exposure chamber was
a defunct walk-in freezer that had been converted to use for the calibration of radon
daughter and radon gas measurement instruments. The chamber's interior dimensions
are large enough to comfortably accommodate a chair, measurement equipment, and a
radon source, with space to spare. Ventilation is by air seepage through the joints of
the structure. Chilled water flow through the heat exchanger of the old refrigeration unit
is used for temperature control. The bones of people with elevated occupational exposure
to radium were used as the source of radon gas for the chamber. The radon emanation
percentage from bone in vitro is generally several percent to several tens of percent.
There have been three phases to this work. In the first phase we determined the
distribution of radon daughter products between the exterior and interior of the body
during and after exposure in street clothes. In the second phase we established a
calibration factor for the quantitative determination of radon daughter deposition in the
chest. In the third phase we compared the results of combined internal and external
exposure, pure internal exposure, and pure external exposure in males and females.
The first phase, which is the subject of this paper, involved the exposures of five
males, all associated through their employment at Argonne, with this or related projects.
The subjects were middle-aged non smokers of average weight and height (Table 1).
Following a measurement in the whole body counter to establish baseline whole body
radioactivity, the subjects sat for one hour in the exposure chamber, dressed in street
clothes, and breathed exposure room air. The chamber air temperatures were in the mid
to high 20's, centigrade, and relative humidities were 40 to 60%. Working level values,
measured by an EDA Instruments model RDA-400 continuous monitor, set for 10 minute
data collection intervals, are presented in Table 1. Working level ratios, based on data
taken earlier, were probably in the range 0.1 - 0.3 during the exposure periods.
After exposure, the subjects were measured in the whole body counter while lying
supine with a Nal(Tl) crystal 10 cm high and 29 cm diameter positioned over their chests
to measure the gamma-ray emission from the upper body. Subjects then changed to
Pajamas, had two more whole body measurements, took a shower and shampoo and had
another two whole body measurements. In total, six whole body measurements were
made on each subject, one for baseline, one in street clothes after exposure, two in
clean clothes before shower and shampoo, and two in clean clothes after shower and
shampoo. The protocol with timing is given in Table 2. These measurements permitted
the estimation of the distribution of radioactivity between clothes, unwashed skin and hair,
and body interior, and washed skin and hair, and body interior. Throughout this work,
We have assumed that washing removes all residual radon daughter products from the
skin and hair.
-------
Table 1. Age, weight, height, and working level exposure of subjects.
ID
Age
Weight Height
RnD
(y)
(kg) (cm)
(WL)
50-032
45
90 184
0.18
50-040
61
91 184
0.22
50-109
46
89 173
0.14
50-128
60
76 170
0.17
50-166
53
83 183
0.20
Table 2. Exposure and measurement protocol.
Elapsed
Allotted
Activity
Time
Time
(min)
(min)
0
10
Don clean pajamas
10
30
First (baseline) whole body measurement
40
10
Don street clothes
50
60
Exposure to radon and radon daughters
110
10
Set up whole body measurement
120
5
Second whole body measurement (in radon
daughter contaminated street clothing)
125
10
Don clean pajamas
135
15
Third whole body measurement
150
2
Reset instruments
152
15
Fourth whole body measurement
167
20
Shower, shampoo, and don clean pajamas
187
15
Fifth whole body measurement
202
2
Reset instruments
204
15
Sixth whole body measurement
209
10
Don street clothes
-------
GAMMA-RAY DATA
For illustration, the gamma-ray spectrum of radon daughter products is shown in
Figure 1. Some gamma-ray spectrum peaks originating from the decay of RaB or RaC
are identified. In general, the data between the left-hand border of the graph and the
RaB marker are mostly from RaB, with contributions from the Compton scattering of
RaC gamma rays.
Channel Number
Figure 1. Gamma-ray spectrum of radon daughter products with the
identification of some peaks associated with RaB and RaC decay.
The sum of all data in the gamma-ray spectrum for energies above 200 keV provides
a simple measure of radon daughter product activity and was used, after conversion to
counting rate and adjustment for baseline radioactivity, as the variable for analysis of the
data from this experiment. The data for three subjects are presented in Table 3 and the
data for the remaining two subjects appear in Figure 2.
Hie decline of radon daughter counting rate in Table 3 and Figure 2 is due to
clothes change, shower and radioactive decay. The lines through the data in the figure
represent a single exponential with 35 min half life, and help when judging how much
of the decline is due to each source.
-------
Table 3. Net radon daughter counting rates over the chests of three subjects.
Subject
Condition
Time Since
Rate
Exposure
(min)
(min"1)
50-040
Contaminated clothing
12.5
2650
After clothes change
32.5
1060
After clothes change
49.5
768
After shower
84.5
335
After shower
101.5
266
50-109
Contaminated clothing
12.5
3750
After clothes change
32.5
1200
After clothes change
49.5
919
After shower
84.5
410
After shower
101.5
288
50-166
Contaminated clothing
12.5
3560
After clothes change
32.5
1391
After clothes change
49.5
1000
After shower
84.5
429
After shower
101.5
316
ANALYSIS AND RESULTS
The difference between the counting rate with contaminated clothing and the
counting rate represented by the exponential with 35 minute half life provides an estimate
of the portion of the rate due to radon daughters on the clothing. The balance of the
radon daughters for the counting period centered at 12.5 min post exposure are deposited
on the skin surface or in the body. Similarly, the difference between the counting rates
after shower and shampoo and the rates represented by the exponential are due to the
removal of skin and hair contamination. The radon daughter counting rates after shower
and shampoo are due solely to radon daughters in the body.
The difference between the exponential line and the observed counting rate is an
estimate of the radon daughters associated with clothing. As an example, the observed
counting rate, with subject 50-032 dressed in contaminated clothing, was 1920 min1 and
the counting rate represented by the exponential at 12.5 min was 985 min"1. Therefore,
-------
100%(1920 - 985)/1920 = 48.6% of the observed activity was due to radon daughters
deposited on the clothing, and 51.4% was due to radon daughters in the body and on
skin and hair. The observed counting rates after shower and shampoo were 210 and
171 min"1 at 84.5 and 101.5 minutes respectively. The counting rates represented by the
exponential line were 237 and 169 min1 at those same times respectively. The percentage
difference at 84.5 minutes is 100%(237 - 210)/237 = 11.4%. Using the same formula,
the difference at 101.5 minutes is -1.2%, and the average for the two times is 5.1%. Re-
expressed as a percentage of the total counting rate at 12.5 minute, this average becomes.
S.l%(51.4%)/100% = 2.6%.
10000
.£ 1000
£
L.
8.
100
D
-------
all are positive lends support to the notion that there is a removable radon daughter
deposit associated with skin and hair. Percentages for the body were obtained as the
difference, e.g. the percentage in the body of subject 50-032 is 100% - (48.6% + 2.6%)
= 48.8%.
Table 4. Percentage of radon daughter counting rate due to deposition on street clothing,
on skin and hair, and in the body.
Subject
Clothing Skin/Hair
Body
50-032
50-040
50-109
50-128
50-166
48.6
40.1
50.6
51.7
41.6
2.6
4.3
4.2
5.1
7.4
48.8
55.6
45.2
43.2
51.0
2500 i
500
1000
2000
External counts per min
3000
Figure 3. Counting rates from internal and external deposition of radon
daughter products. The plot and equation of a regression line
fitted to the data are shown.
-------
A scatter diagram of the internal (body) and external (clothing plus skin) counting
rates, estimated with the aid of percentages given in Table 4, is presented in Figure 3.
The slope of the regression line is not significant at the 5% level (p = 0.089) but the
percentages in Table 4 may be biased estimates that render tests of significance invalid.
DISCUSSION
Street clothing accounts for a substantial fraction of the radon daughter activity
associated with the body, based on the estimation procedure outlined above, which relies
on a single exponential with 35 min half life for the back extrapolation of data. To test
the plausibility of the extrapolation model, we fitted a sum of two exponentials,
representing contributions from RaB and RaC, to a single exponential with 35 minute
half life, i.e. the equation,
Aft) = b0ekB' + c0ekC (1)
was fitted, with nonlinear regression techniques, to:
Oft) = e't(ln2^35 (2)
for times between 12.5 and 101.5 minutes; kg and kg are the decay constants for RaB
and RaC, 0.0259 min'1 and 0.0352 min"1, respectively. The fitted equation (A(t)) with
bQ - 2.141 and cQ * -1.223 and the single exponential (O(t)) deviate by no more than
11% with an average absolute deviation of 2.9% and an average deviation of 1.8%
Thus, a single exponential provides a plausible empirical representation of the gamma-
ray emission from a RaB - RaC mixture whose activity is not significantly diminished by
biological processes. The negative sign of c0 is consistent with an initial RaC activity
that is less than the value at secular equilibrium. For clarification, the following equation
gives the gamma-ray counting rate, R(t), as a function of time after the RaA activity
(3.05 min half life) has fallen to negligible levels. This most likely occurred before the
subjects entered the whole body counter for the first time. The factors, fg and fg, are
calibration constants that relate the counting rates for RaB and RaC to the activities
associated with the subject. The counting rate has been normalized to BQ, the initial
activity of RaB.
R(t)/B0 - (fB + 3.77}c)e + (C0/BQ - 3.77)fceu'UJJ" (3)
Secular equilibrium occurs when the ratio of initial activities, is 3.77. A smaller
-------
ratio causes the coefficient of the second exponential term to be negative, verifying our
interpretation of the negative sign of c0.
During the course of this work we made direct measurements of clothing from the
exposure chamber to verify radon daughter uptake and to supplement our inferences of
substantial deposition on clothing based on observations of a reduction in counting rate
after a clothes change. One of the most interesting aspects of these exploratory
measurements was the observation of significant daughter product deposition in
underclothing, following an exposure in loose fitting garments. The counting rates for
gamma rays in the 609 keV peak of RaC were analyzed using the extrapolation techniques
described above. We found that two-thirds of the 609 keV gamma rays associated with
clothing originated in the underclothing. Although these experiments were not repeated,
this one result suggests an important role for underclothing in radon daughter uptake.
During field studies in which subjects are asked to change their clothes it is insufficient
to remove only the outer clothing. Underclothing must also be removed to reduce
external contamination to low levels.
The correlation apparent in Figure 3 provides another observation important for
field studies. The linear trend in the data, emphasized by the regression line, suggests
a strong correlation between externally and internally deposited radon daughter products
even though the present data set is too small to establish a statistically significant
correlation. The spread of the data in Table 4 is not large, suggesting that 50% internal
deposition, 50% external deposition is a good working hypothesis. The small spread also
suggests that total counting rate above the chest is a good index of radon daughter
deposition in the lungs. This simplifies protocols for field measurements by eliminating
the requirement for a clothes change, although data collected with a clothes change are
preferable to those collected without one. It also permits a confident analysis of data
collected from people dressed in street clothes, when information about internal deposition
is desired.
ACKNOWLEDGEMENTS
We would like to acknowledge the contributions, at various stages of this study, of
R. E. Toohey, A. T. Keane, F. Markun, B. G. Oltman, E. G. Thompson, M. Hall, and
L. Michaels. This work was supported by National Institutes of Health grant number 5-
R01-CA40071. The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement should be inferred.
Ttw tubmitud manuttrlpl hai toon authored
by ¦ contractor of th* U. S. Govemmant
undar contract No. W-31-10B-ENG-38.
Accordingly, ttw U. 8. Government rtuim •
non«xclutlv«, roydtytfM Hum* to puWMi
or raproduca the publidiad form of thli
contribution, or allow other, to do to, for
U. S. Government purpom.
-------
REFERENCES
1. Toohey, R. E., Keane, A. T., and Rundo, J. Measurement techniques for radium and
the actinides in man at the Center for Human Radiobiology. Health Phys. 44 (Suppl.
#1): 323, 1983.
2. Stebbings, J. H., Kardatzke, D. R., Toohey, R. E., Essling, M. E., and Pagnamenta
A. Domestic and personal determinants of the contamination of individuals by
household radon daughters. In: D. D. Hemphill (ed.), Trace Substances in
Environmental Health - XX. University of Missouri Press, Columbia, Missouri, 1987,
p.392.
-------
II-6
HIGH RADON HOUSES: QUESTIONS ABOUT LOG NORMAL DISTRIBUTIONS
AND IMPLICATIONS FOR F.PTDEMTOLOOY AND RISK ASSESSMENT
R. Goble
Center for Technology, Environment, and Development
Clark University, Worcester, MA 01610
R, Socolow
Center for Energy and Environmental Studies
Princeton University, Princeton, N.J. 08544
ABSTRACT
There is a very real possibility that, for the United States
as a whole, the number of houses with indoor radon concentrations
tens or hundreds of times greater than the national median is
Neatly underestimated by the traditional log-normal distributions
^sed to characterize radon levels in houses. A fat tail could
imply that a significant fraction of the total national health
impact of indoor radon is being experienced by those living in
v®ry high radon houses, contrary to the conclusions drawn from
log-normal distributions. There is also a very real possibility
that geographically delimited distributions (across, say, on the
°*der of a million houses in a region) have fat-tailed
distributions, relative to log-normal. Fat-tailed distributions
Would permit case-control epidemiological studies to have much
Steater power, for a given level of effort, than is now presumed
0r* the basis of log-normal distributions.
-------
INTRODUCTION
By now many measurements of radon concentrations in houses
have been made. They have been taken by many different groups for
different purposes, using different methods, different protocols,
in different geographical areas, under different conditions. Many
of them have been taken for "screening" purposes with procedures
intended to maximize the observed radon levels. To what extent is
it feasible to formulate a picture of how radon concentrations are
actually distributed across a sample of houses?
At present the standard approach is based on the use of log-
normal distribution functions. For many data sets they provide,
as we shall see, a good fit covering the bulk of the data. They
do not, however, provide very secure guidance for estimating
numbers of houses with high concentrations of radon. In this paper
we discuss evidence that indicates that the numbers of high radon
houses may be underestimated by log-normal fits: we suggest an
alternative approach to analyzing the data, and describe some of
the reasons it would be desirable to make more realistic estimates
for the number of houses in various samples with high
concentrations. High radon houses are important in two contexts:
in the area of public policy, they constitute an environment in
which the public may be exposed to risks that are considered
unacceptable, or strongly undesirable, and they may well represent
the source of a significant fraction of the population exposure to
inhaled radon decay products; in the broader public health area,
they may provide the best opportunity to learn about radiation-
induced carcinogenesis in the general public. A lengthier
discussion by us of fitting high radon distributions and their
implications is available (1).
THE LOG-NORMAL FITS TO RADON MEASUREMENTS IN HOUSES
Because of the differences In measurement methods and
protocols, and, especially, because of differences in measurement
goals, identification of a representative national sample for
determining a national distribution of radon levels has not been
easy. Results from two national surveys specifically addressing
these difficulties have been published. They follow two different
approaches: one, by Cohen (2), is an attempt to use a uniform
protocol and consists of a measurement of about 450 homes of
physics professors; the second, by Nero et al. (3) is a
compilation of 800 measurements taken from nearly 40 data sets,
where the compilers made a substantial analysis of the differences
in measurement specifications. The two surveys are remarkably
consistent: they find median concentrations of approximately 1
pCi/1, and the distributions fit surprisingly well a log-normal
form with a geometric standard deviation of 2.4-2.8.
-------
In Figure 1, we show Nero's fit to the data. The log-normal
fit appears excellent, especially when one remembers that it is
determined by only two parameters, the geometric mean and standard
deviation. The excellence of the fit can, however, distract one
from worrying about how well the tail of the log-normal
distribution will match the numbers of high radon houses. Because
the distribution is determined by only two parameters and because
high radon houses constitute a small portion of the data, the log-
normal fit essentially predicts the number of high radon houses
based on the distribution of radon concentrations relatively near
the median. There is no particular reason to expect such a
prediction to hold. Neither the Cohen nor the Nero study serves to
test the prediction because of the relatively small numbers of
houses included.
222Rn Concentration (Bq/m3)
Figure 1. Log-normal fit to Nero's compilation of national
radon data.
EVIDENCE FOR FATTER-TAILED DISTRIBUTIONS
Before discussing the evidence that log-normal distributions
are likely to underestimate seriously the numbers of houses with
high concentrations, it is worth stressing the problematic quality
of the data. The radon concentrations of interest are not simple
to define. The connection between house measurements and exposure
to radon decay products is, unfortunately, neither direct nor
certain. Most reported measurements are of radon concentrations
taken in one or two locations of the house averaged over days,
weeks, or months, since such long-term averages are obtained with
relatively low cost methods. Most have been made for "screening"
purposes with procedures intended to maximize the measured level
(basement measurements, for instance, with windows kept closed for
the measurement period) . Research performed with more elaborate
-------
instrumentation has revealed that radon concentrations in a house
can change substantially within a few hours (associated with rain
or with diurnal changes in the difference between the indoor and
outdoor temperature) and that there are seasonal variations caused
by changes in soil conditions. Perhaps even more important are
the large differences found in radon levels in different parts of
a house at any particular time. Figure 2A shows some
characteristic fluctuations in radon levels in a basement. Figure
2B illustrates the sort of variation which may be found in radon
levels at different places in a house. Clearly where and when
people spend their time at home will strongly influence the
exposure they receive.
Radon House 22 basement:
March/June
Weekly radon averages, House 22,
March/June -•.-.jipstairs: staircase
«—basement upstairs tfireplace
160
140-
120'
100-
80.
•¦4
60-
•H
U
40-
o o
CM
tl 13 15 '1* 13' 15 17 19
March June
90 110" * 130 150
Julian Date
Figure 2. Examples of spatial and temporal variation in
radon concentrations in a house.
Despite such problems with using general radon surveys for
quantitative assessments, some recently available large data sets
taken under non-representative circumstances provide a useful
perspective on the distribution of radon exposures. One is a set
of nearly 20,000 screening measurements conducted by the
Pennsylvania Bureau of Radiation Protection (4) in five counties
which include the Reading prong, the large geological feature on
which many high radon houses have been found. A second is summary
data of measurements made nationwide by the Terradex corporation
(5). The particular subset of their data which we use is a
compilation of about 10,000 measurements made throughout the U.S.
from which 6 states (ME, NJ, NY, OR, PA, WA) were excluded. We
chose this subset because it avoids areas where there have been
large numbers of measurements within a small region and because
the total Terradex compilation very substantially overlaps the
Pennsylvania study.
We present in Figure 3 a reanalysis of the Nero et al.
compilation, the Reading prong area data, and the Terradex subset.
-------
On a log-log scale, the percent of houses having greater than the
indicated concentration is plotted against that concentration/ the
latter made dimensionless by dividing by the median concentration.
Since the median in the Nero data set is close to 1.0 pCi/liter,
the x variable may be regarded as the concentration in pCi/liter.
The median for the Reading Prong data set is roughly 5 pCi/1,
while it is approximately 1.7 pCi/1 for the Terradex data. Because
we are interested in the high radon tail/ we plot the data only
for concentrations greater than four times the median
concentration of that data set. This starting point is arbitrary,
but we have found it useful for a large number of data sets. When
the median value is 1 pCi/liter, the starting point is the EPA
action level.
Together with each data set we show four theoretical
distributions on each plot of Figure 3. Two are logarithmic
normal distributions, named LnNrm 2.8 and LnNrm 3.5, which have
geometric standard deviation 2.8 (Nero's fit (3)) and 3.5 (Alter*s
fit (5)) respectively. Two are power law distributions, named PWR
1.75 and PWR 1.25. They are specified by an exponent, 1.75 and
1.25, such that the percent of houses with concentration greater
Figure 3. Comparison of four theoretical high radon
distributions with three data sets.
-------
The three high radon data sets in Figure 3 are reasonably
well fit by a power law with an exponent bracketted by 1.75 and
1.25 and with the fractional number of houses with concentrations
greater than four times the median roughly between 10 and 15%.
The high radon tails of the Reading prong and Terradex data
clearly appear fatter than log normal. In fact, log normal
distributions fall off sufficiently at high levels that more
houses with concentrations greater than 100 pCi/1 have already
been found in these studies than are predicted by the Nero et al.
log normal fit for the entire nation, and the discrepancy at
higher levels is even worse. Cohen's most recent survey results
(6) support a power law distribution as well. He finds deviations
from log normal; the deviations are greater the higher the
observed levels, and his reported distributions for concentrations
from 100 to above 1000 pCi/1 are consistent with the range of
power law exponents considered here. We also have found similar
results in fitting a number of other non-representative survey
data sets.
To summarize: in the absence of large surveys with protocol
designed to assure the representativeness of the data set, these
fits must be considered suggestive rather than conclusive evidence
that houses with high radon concentrations will appear more
frequently than expected from log-normal fits. For the remainder
of the paper we argue that if the tails of the distribution are
indeed fatter, there will be important applications. Thus careful
study of high radon distributions is called for.
DISTRIBUTIONS OF RISK
Three separate questions about the distribution of radon
concentrations in houses across the nation continually appear in
policy analyses of the indoor radon problem. One question is how
many high radon houses are there? What is the size of the
population for which we have various levels of concern, or for
which protective measures might be addressed? A second question
is how are radiation exposures distributed across the population?
What fraction of the overall dose to the public is accumulated in
houses at various levels of exposure? The third question is how
well known are the average levels of exposure?
In Table 1, we show how the answers to these questions differ
for the four theoretical distributions shown in Figure 3. For
each distribution we display representative values of the % of
houses exceeding benchmark radon levels, the fraction of total
dose contributed by houses in each category, and the ratio of the
arithmetic mean to the median dose.
-------
TABLE 1: FRACTION OF HOUSES WITH HIGH RADON CONCENTRATIONS AND
THE FRACTION OF THE AGGREGATE RADIATION DOSE ASSOCIATED WITH THOSE
HOUSES FOR FOUR THEORETICAL MODELS.
Concentrations
of radon/median
% houses with
conc. >
% of dose from
these houses
% houses with
conc. >
% of dose from
these houses
Fit: PWR(1.25)
Fit: LnNrm (gsd = 3.5)
4
13.
72.
13.4
56.
10
4.1
57
3.3
28
20
1.7
48
.84
12.8
60
.44
37
.054
2.2
100
.23
32
.012
.77
200
.098
27
.0012
.15
arithmetic mean/median
3.6
2.2
Fit: PWR(1.75)
Fit: LnNrm (gsd = 2.8)
4
9.
44.
8.9
38.
10
2.7
22
1.3
11.
20
.54
13.
.18
3.
60
.079
5.8
.0035
.2
100
.032
3.9
.00029
.03
200
.096
2.4
.00001
.002
arithmetic mean/median
1.9
1.7
Table 1 shows that if a distribution of radon concentrations
across high radon houses is similar to one of the power law
distributions:
1) the 10% to 12.5% (6-7 million) of houses with
concentrations greater than 4 times the median may contribute
half or more (50% to 70%) of the expected aggregate exposure,
2) the roughly 1% (600,000) of houses with concentrations
greater than 20 times the median could be expected to
contribute 15% to 45% of the aggregate exposure, and,
3) as many as 20,000-120,000 houses may have levels greater
than 100 times the median and contribute 5% to 30% to the
aggregate exposure.
-------
These values for the power law fits differ strikingly from
those for log normal distributions, where only a few tenths of a
percent of houses have concentrations greater than 20 times the
median and these contribute only 2% to 6% to the aggregate
exposure. Only the values for log normal distributions are
considered in EPA's present perspective (7). The fatter tail also
increases the ratio of mean to median dose, but the difference is
substantial only for the slowly falling (exponent = 1.25) power
law.
IMPLICATIONS FOR EPIDEMIOLOGY
The presence of substantial numbers of houses with radon
concentrations that present high risks suggests that there should
be a significant number of lung cancers associated with those
houses and that such an association should be observable. Thus
far, attempts to observe such associations between environmental
radon exposure and lung cancer have been ambiguous at best (8,9).
There are, however, several reasons why it is difficult to observe
the association.
• It is difficult to estimate exposures:
— people move from house to house
— the most recent 5-10 years of exposure may be
irrelevant
— patterns of occupancy in a house strongly affect
exposures
— radon concentrations vary widely in houses in a single
neighborhood
— the characteristics of houses may change as a result
of changes in heating systems or other modifications
• Radon causes only a fraction of lung cancers
• There is great regional and local variability in other
causes of lung cancer, particularly smoking.
Fortunately, in our view, these complications have not
prevented some case-control studies from getting under way (10).
The public health community is pessimistic that radon-induced lung
cancer will be detectable but it has been willing to look.
Successful studies could be very valuable. A determination of the
rate at which environmental radon causes cancer should clarify
significantly the debate on public policy. Environmental
exposures will be the only source of reasonable data on effects of
exposures on women, and on exposures of children and the elderly;
they are the most likely source of information to clear up the
controversies about the interactive effect of smoking.
-------
Suppose that we postulate a geographical region where high
radon houses follow the sort of power-law distribution considered
above, and further postulate that cancer incidence is proportional
to radon dose (the "linear" hypothesis which underlies so much of
radiation risk assessment). In such a region the capability of
epidemiological studies to determine the effects of environmental
radon exposures is greatly enhanced, relative to the capability of
such studies in a region where high radon houses obey a log-normal
distribution.
The argument is a natural extension of the analysis presented
in Table 1: the larger the fraction of the total radon dose to
the population absorbed by those living in the houses with the
very highest radon levels (say, in the houses with radon levels 20
or more times the median, which may be roughly one percent of the
houses), the more frequently research seeking to link lung cancer
to radon in houses will find that the trail leads to high radon
houses.
The case-control experiments we consider here would begin
with a patient ("case") who has been diagnosed with lung cancer
and a control who is matched with appropriate characteristics.
The key step in the research is to establish the radon exposures
in the houses in which the case and the control lived. Greatly
complicating these studies are the difficulties involved in
documenting these exposures, as discussed above.
We present an example to show numerically how much more
productive such an investigation will be whenever the power-law
rather than the log-normal distribution more closely represents
reality. We find that confirmation of today's expectations about
the dose-response relationship is obtainable only from studies
involving tens of thousands of matched cases and controls if there
is a log-normal distribution of high-radon houses, but will be
achievable from studies involving roughly nno thousand matched
cases and controls if there is a power-law distribution of high-
radon houses.
The example assumes that 10% of cancers in the sample are
radon induced and that the dose response relation is linear and
independent of age at exposure except for a latency period. These
assumptions are consistent with EPA and other risk estimates. To
obtain representative numbers, we invent a simple and arbitrary
model of exposure histories.
• Every case and control moves precisely every ten years,
and dies at age 70.
A ten-year latency period is assumed,
-------
• Accurate exposures could be established for only the
most recent three of seven houses: consequently, the
most recent house is not of interest, precisely two of
the pertinent six houses are investigated, and thus one
third of the integrated effective radon exposure is
documented.
• There is no correlation in house radon level across
moves.
From this model, distributions of radon levels in housing
"data" were derived for both cases and controls using each of the
four theoretical distributions for radon levels in houses. The
predicted fraction of cases and of controls whose measured
exposures would exceed a value x is plotted against the exposure
x, for two of the distributions, PWR (1.25) and LnNrm (3.5) in
Figures 4A,4B. The numerical details of this exercise with
several further examples are summarized in our larger paper (1).
Figure 4. Calculated distributions of exposure for two
hypothetical case control studies.
The main feature of each Figure 4 is the crossover, the
location at which the slope of the graph becomes less steep. This
crossover appears in the lung cancer cases, as measured radon
exposures become more important; it is absent in the controls. It
is also much less apparent for log-normal distributions; although
there is a shift in slope for the cases, it does not look like a
break since the graph is already falling off rapidly.
-------
One good identifier of the crossover is the value of x, call
it XO/ above which half of the cases are associated with measured
radon. For each theoretical distribution, Table 2 tabulates xo
and gives the fraction of the houses above xo; the larger this
fraction, the smaller the sample size necessary to estimate the
dose response coefficient to any given level of confidence.
To calculate actual, rather than relative, sample sizes, it
is necessary to declare the size of some subsample desired. We
choose the following power-of-experiment condition to establish
sample sizes:
At the radon concentration in houses above which half of
the cases are radon-associated, there shall be 20 cases
{10 radon induced).
There will then be roughly 10 controls. The corresponding
sample sizes for each distribution are also tabulated in Table 2.
TABLE 2: THE DEPENDENCE OF THE NUMBERS OF CASES RECOMMENDED FOR A
HYPOTHETICAL CASE CONTROL STUDY ON THE DISTRIBUTION OF HIGH RADON
HOUSES FOR FOUR THEORETICAL DISTRIBUTIONS: RESULTS FOR BASE CASE
ASSUMPTIONS USED IN FIGURES 5, ASSUMING 10% OF CANCERS ARE RADON-
INDUCED AND ONE-THIRD OF LIFETIME EXPOSURE IS MEASURED IN THE STUDY.
THEORETICAL
DISTRIBUTION
CROSS OVER POINT
POINT Xo
(EXPOSURE/MEDIAN)
% OF POPULATION
WITH
EXPOSURE > Xo
NUMBER OF
CASES
PWR(1.25)
10.75
3.4
300
PWR(1.75)
12.25
0.8
1,300
LnNrm(3.5)
21.0
0.3
3,500
LnNrm(2.8)
18.25
0.05
20,000
Table 2 shows an order of magnitude difference in sample
sizes for experiments of equivalent statistical power when
experiments are done in a world with an underlying power-law
distribution, as compared with a world with an underlying log-
normal distribution. For the power-of-experiment condition just
stated, the sample sizes range from 3,500 to 20,000 for log-normal
distributions, but from 300 to 1,250 for power-law distributions.
-------
This order of magnitude increase in the power of the
epidemiological experiments currently under way is the good news
latent in the possibility, not otherwise a cheerful one, that
there are many very high radon houses. It opens up the prospect
of seeing a house-radon signal in lung cancer data many years
earlier than would be expected on the basis of log-normal
distributions.
Tables 3 and 4 further explore the sensitivity of the
estimate of sample size: in Table 3, to differences in the % of
cancer radon-induced (the objective of a study), and in Table 4,
to different fractions of exposure history obtained. The
dependence in each case is straightforward: the existence of fewer
radon-induced cancers requires larger samples, so that a 5%
induced cancer incidence means that even power law distributions
could require up to several thousand cases, while 20% induced
cancer incidence could be detectable even with lognormal radon
distributions; better radon histories can compensate for all but
the narrowest log normal distributions, while poor radon histories
may make any study too difficult.
TABLE 3: THE DEPENDENCE OF THE NUMBERS OF CASES RECOMMENDED FOR A
HYPOTHETICAL CASE CONTROL STUDY ON THE DISTRIBUTION OF HIGH RADON
HOUSES FOR FOUR THEORETICAL DISTRIBUTIONS: RESULTS SHOWING THE
DEPENDENCE ON THE FRACTION OF CASES OF LUNG CANCER WHICH ARE RADON-
INDUCED
(% of cancers radon-induced / # of houses measured)
20%/ 10%/ 5%/
2 houses 2 houses 2 houses
THEORETICAL NUMBER OF CASES
DISTRIBUTION
PWR (1.25)
PWR (1.75)
LnNrm (3.5)
LnNrm (2.8)
130 300 750
350 1,300 4,400
500 3,500 30,000
1,600 20,000 450,000
-------
TABLE 4: THE DEPENDENCE OF THE NUMBERS OF CASES RECOMMENDED FOR A
HYPOTHETICAL CASE CONTROL STUDY ON THE DISTRIBUTION OF HIGH RADON
HOUSES FOR FOUR THEORETICAL DISTRIBUTIONS: RESULTS SHOWING THE
DEPENDENCE ON THE NUMBER OF HOUSES FOR WHICH THERE ARE
MEASUREMENTS.
(% of cancers radon-induced / # of houses measured)
10%/ 10%/ 10%/
3 houses 2 houses 1 houses
THEORETICAL
DISTRIBUTION NUMBER OF CASES
PWR (1.25)
120
300
1,500
PWR (1.75)
400
1,300
9,000
LnNrm (3.5)
700
3,500
62,000
LnNrm (2.8)
2,900
20,000
900,000
The details of any given research experiment will require
modification of the analysis, particularly if additional
information such as age, age at exposure, and smoking history is
used to enhance the power of the study. The modifications should
not change the overall message that a fatter tail at high dose
makes the prospect of seeing a signal much more likely. There
will always be a tradeoff between taking a larger sample and going
further back in time in finding the houses occupied. There are
also alternative ways of estimating actual past exposure,
including using physiological radon indicators instead of house
measurements.
CONCLUSIONS
Information about numbers of high radon houses is only a
small fraction of the information that can be obtained about the
distribution of radon concentrations. It is, however, information
that is of direct significance in formulating public policy
regarding radon exposures. Even more significantly, perhaps, it
is information that is crucial to research planning that could
lead to a better understanding of cancer induction by radiation.
Data about relative numbers of high radon houses is best sought
explicitly: the indications are that current practice
underestimates the numbers, based on inappropriate extrapolations
from the distributions of radon concentrations at lower levels.
-------
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCES
1. Goble, R. and Socolow, R. High radon houses: implications
for epidemiology and risk assessment, CENTED Research Report,
Clark University, Worcester, MA. 1990.
2. Cohen, B. L. A national survey of 222Rn in U.S. homes and
correlating factors, Health Phys. 51: 175. 1986.
3. Nero A. V. Schwehr M.B., Nazaroff W. W. and Revzan K. L.
Distribution of airborne radon-222 concentrations in U.S.
homes, Science 234: 992, 1986.
4. Gerulsky T. M. Pennsylvania: protecting the homefront,
Environment 29, No. 1,12.
5. Alter, H.W. and Oswald, R.A. Nationwide distribution of
indoor radon measurements: a preliminary data base JAPCA,
37 (3) : 227, 1987.
6. Cohen, B.L., and Gromicko, N. Variation of radon levels in
U.S. homes with various factors, JAPCA 38(2): 129, 1988.
7. Puskin, J.S. and Nelson, C. B. EPA's Perspective on Risks
from Residential Radon Exposure, JAPCA. 39(7): 918, 1989.
8. Ballew, M.A. Health risks from radon progeny. In 18th annual
national conference on radiation control- Frankfort, KY:
Conference of Radiation Control Program Directors, Inc.
Conf. Publication-87-2, pp.157-171. 1987.
9. Bodansky D. et al. Indoor Radon and Its Hazards, chapt.8,
University of Washington Press, 1987.
10. Schoenberg, J., and Klotz, J. A case control study of radon
and lung cancer among New Jersey women. Technical Report -
Phase I, New Jersey State Department of Health, Trenton, NJ
1989.
-------
Session A-ll:
Radon Related Health Studies—POSTERS
-------
A—11—1
RADON-INDUCED LUNG CANCER RISK ESTIMATES FOR THE NEW JERSEY POPULATION
by: Mary Cahill and Robert Stern
New Jersey Department of Environmental Protection
Princeton, New Jersey 07540
ABSTRACT
New Jersey excess population risks for radon-induced lung cancer mortality
were estimated for counties and the state using results of radon testing in 5,700
homes and risk factors extracted from the recent report of the National Academy
of Sciences' Committee on the Biological Effects of Ionizing Radiation (BEIR IV).
Current New Jersey specific female and male smoker/nonsmoker proportions and
residential occupancy rates were considered. Findings indicate a statewide
average excess lifetime population lung cancer risk of 5 x 10"3 and geometric
mean excess risk of 3 x 10"*. The estimated annual number of radon-induced lung
cancer deaths in New Jersey is 500. These risks are 50 percent higher than
earlier estimates as a result of updating baseline lung cancer mortality rate
information. Considering that 3 percent of the nation's population resides 1n
New Jersey, our current estimate of 500 radon induced lung cancer deaths each
year is 25 percent lower than the 650 that would be inferred from the current
national central estimate of 21,600 reported by the U.S. Environmental Protection
Agency (EPA). This difference is attributed primarily to lower New Jersey
Population exposures as compared to EPA estimates for the nation.
-------
INTRODUCTION
The most recent estimate by the U.S. Environmental Protection Agency (EPA)
of the annual number of radon-induced lung cancer deaths for the nation is given
as a central estimate of 21,600 (1). New Jersey, with 3 percent of the nation's
population would therefore be expected to contribute 650 radon-induced lung
cancer deaths each year to the national total, assuming that New Jersey specific
population exposure patterns, baseline lung cancer mortality rates, and age
distribution are similar to those for the nation. Based on the large number of
occurrences of very high levels of radon discovered in the Reading Prong region
and in other northern portions of the state, it has been suggested that the radon
problem in New Jersey may be of greater magnitude than that estimated for other
states and the nation as a whole.
To better characterize the New Jersey radon problem, a statewide survey
of indoor radon levels in approximately 5,700 homes was conducted by Camp Dresser
and McKee Inc. for the New Jersey Department of Environmental Protection. The
primary objective of the survey was to determine the geographical distribution
of indoor radon levels in the state. To best achieve this objective, a spatially
uniform, rather than population based sampling design was selected to ensure
adequate character!zation of less populated, potentially high radon areas. The
current trend of increasing residential development in areas of high radon
potential indicated the need for this approach. Radon measurements were made
with charcoal canisters on the lowest level of each home (85 percent of
measurements were made in basements) during the heating season. A second
objective of the study, which is the subject of this paper, was to estimate radon
progeny exposures and associated lung cancer risks for the New Jersey population.
While the radon measurement data obtained from the original survey did not
directly lend itself to reliable estimation of population exposures and risks,
a methodology was developed to enable a reasonable approximation of county and
statewide average population exposures. Appropriate risk factors, as reported
by the National Academy of Sciences' BEIR IV Committee (2), combined with New
Jersey specific exposure estimates and information obtained from survey
participants regarding smoking habits and house and floor occupancy rates were
used to estimate radon-induced lung cancer mortality risks. The resultant risk
estimates permit a more refined estimate of New Jersey specific radon risk than
can be inferred from national estimates reported by EPA.
METHODS
To estimate radon risks to a population, both representative exposure data
and a risk model for radon induced lung cancer are required. For the New Jersey
specific risk assessment, exposure information was inferred from shcrt term
radon measurements and average occupancy information collected for 5,700 homes.m
Risk factors derived from the BEIR IV modified relative risk model were applied
to exposure data to estimate excess population risks for counties and the state.
EXPOSURE ASSESSMENT
Estimation of exposure requires information on annual average radon levels
to which a person is exposed and the amount of time the person is exposed. The
-------
seasonal variability of radon levels, as well as variation of radon levels within
structures must be considered. Short term "screening" measurements made under
conditions and in locations in homes where radon levels tend to be highest do
not take into account this variability and therefore cannot directly be used to
infer exposures. Year long measurements of radon in frequently occupied areas
is the most appropriate method for estimating exposures.
Since only lowest floor radon screening measurements had been performed
in this study, it was necessary to develop an alternate method for estimating
annual average exposures. To project an annual average radon progeny
concentration from a single short term radon measurement required information
on equilibrium ratios and seasonal and floor to floor variability of indoor radon
and radon progeny levels. This information was obtained from shcrt term radon
and radon progeny measurements made during the heating and non-heating seasons
and on different floors in a 200 home subset of the 5,700 homes screened in the
initial survey. Average basement and first floor equilibrium ratios, ratios of
first floor to basement radon levels, and heating to non-heating season radon
levels were thus computed for use as correction factors in estimating annual
average radon progeny levels for the 5,700 radon screening results.
Correction factors were developed separately for three house substructure
categories and three heat distribution system categories. Substructure
categories included basement, slab-on-grade/crawl space and "other" homes. Homes
with forced air, hot water/electric and "other" systems made up three heat
distribution categories. Correction factors developed for basement homes were
then applied, as appropriate, to screening results for basement homes of the
larger 5,700 home data set to estimate annual average radon progeny levels in
basements and on upper floors of each home. A similar approach was used for
slab-on-grade/crawl space and "other" homes which represent about 15 percent of
the data set and New Jersey housing stock.
The second step was to estimate the average amount of time spent in
different floors of homes obtained from occupancy information provided by survey
Participants. Data from upper floors were combined and designated as first floor
occupancy hours.
Next, the annual average first floor and basenent rador progeny
concentrations for each sampled house were multiplied by the average number of
resident hours per day spent on the first floor and basement, respectively, of
each house. Hours not accounted for by in-house occupancy were assumed to occur
at a location characterized by a background radon progeny concentration of 0.001
WL.
Finally, exposures were totaled for each house and averaged over all and
averaged for all counties. A statewide average exposure estimate was determined
from population weighted average county exposures.
RISK ASSESSMENT
The BEIR IV Committee developed a modified relative risk model for
projecting radon risks in which the risk coefficient effectively varies from
0.5* per WLM to 3.0% per WLM, depending on age at risk and time since exposure.
-------
For a continuous lifetime exposure to radon progeny the Committee estimated, for
various annual exposures and female and male smoker and nonsmoker subgroups,
ratios of lifetime risks in exposed to nonexposed persons. It is these values,
combined with exposure estimates developed as described above and New Jersey
smoking rate information, that were used to estimate excess radon-induced lung
cancer risks.
Male and female baseline lung cancer mortality rates of 0.067 and 0.025,
respectively were adjusted to reflect New Jersey smokers and nonsmokers. A
relative risk for smoking of 12 for males and 10 for females was used.
Proportions of male and female smokers and nonsmokers for each county
were determined from questionnaires completed by study participants. Only
current smokers over age 18 were included in the smoker category. The proportion
of smokers estimated in the study may be biased low due to undersampling of urban
areas where smoking rates are generally higher, unwillingness of study
participants to admit to smoking, and higher participation rates among more
health conscious (less likely to smoke) individuals. Risks were therefore also
estimated using higher 1970 U.S. smoker proportions given in the BEIR IV report.
Once baseline lung cancer rates were determined, appropriate relative
risk coefficients given in the BEIR IV report for average county exposures were
applied to estimate excess lifetime risks for each subgroup.
The weighted sum of average excess risks for each subgroup was then used
to determine average excess population risks for counties. Subgroup proportions
used for counties were those estimated for the state from population weighted
county averages. County specific subgroup proportions, which were shown to be
highly variable, were not used in estimating county risks. Statewide excess
lifetime risks were projected from population weighted average county excess
risks.
RESULTS AND DISCUSSION
STATE/COUNTY EXPOSURE ESTIMATES
Average values of inter-floor radon ratios, basement and first floor
equilibrium ratios, and inter-season radon ratios for the basement homes of the
200 home subset used as correction factors for estimating annual average basement
and first floor radon progeny levels are compiled in Table 1. They are generally
consistent with findings of other studies. First floor to basement radon ratios
ranged from 0.27 to 0.51, with higher ratios found during the heating season and
for homes with forced air heat distribution systems. Average equilibrium ratios
were somewhat lower than the 0.5 value assumed by EPA in estimating radon progeny
exposures, ranging from a low of 0.34 during the heating season to a high of 0.46
1n warmer periods in basement homes with forced air systems. Equilibrium ratios
were slightly higher on first floors than 1n basements. Heating to non-heating
season radon ratios ranged from 0.6 to 0.97, higher than ratios of between 0.4
and 0.6 reported by others (3,4,5). The lack of sufficient radon measurement
-------
data in this study from summer months of the nonheating season when radon levels
are generally lower may partially explain this difference. A more complete
discussion of this phase of the exposure assessment is available in the
proceedings of the 1988 EPA Symposium on Indoor Radon (6).
Application of appropriate correction factors to each of the 5,700 home
screening results yielded annual average radon progeny levels in basements and
on upper floors in Table 2. County average screening levels ranged from 0.8 to
12.5 pCi/1. Corresponding estimates annual average basement radon progeny
concentrations 0.003 WL and 0.047 WL. Annual average radon progeny levels
ranging from 0.001 WL to 0.019 WL on upper floors were a factor of 3 to 4 lower
than basement radon progeny levels. For the statewide population-weighted
average screening level of 2.9 pCi/1 in homes with basements, annual average
radon progeny levels of 0.011 WL and 0.005 WL in basements and upper floors,
respectively, were estimated.
House occupancy Information used 1n computing exposures, also shown 1n
Table 2, Indicates that the average amount of time spent in basements is from
0.5 to 2.2 hours per day with a statewide average of 1 hour per day. The major
portion of time was spent on upper floors of homes, ranging from 15.2 to 17.2
hours per day with a statewide estimate of 16 hours per day. Participants of
the study, on average, were found to spend 71 percent of their time at home as
compared to the slightly higher 75 percent occupancy factor typically assumed
by the EPA.
Statewide exposures estimated using occupancy data were 0.028 WLM/yr 1n
basements and 0.160 WLM/year on upper floors. The contribution from out-of-home
exposures using an assumed radon progeny concentration of 0.001 WL was about
0.017 WLM/yr. The total exposure projected for the New Jersey population
residing in basement homes, assuming that the survey data are representative,
1s 0.21 WLM/yr. A combined estimate for basement and non-basement homes 1s about
10 percent lower, at 0.19 WLM/yr for an average screening level of 2.5 pC1/l.
The New Jersey exposure estimate 1s 25 percent lower than the current EPA
estimate of 0.25 WLM/yr for the U.S. population. The high concentration of the
population in low radon areas of the northeastern portion of the state and sparse
Population of high radon areas 1s the primary reason for the lower New Jersey
exposure estimate.
It 1s of Interest to relate our radon screening measurement results to
the EPA annual average remedial action level of 4 pC1/l. Based on this study,
a basement screening level of 11 pC1/l is estimated to result 1n an annual
average exposure 0.8 WLM/yr exposure (4 pC1/l EPA action level).
STATE/COUNTY RISK ESTIMATES
Statewide smoker proportions estimated for this study, 0.25 for males and
0.26 for females, were considerably lower than the 0.48 (males) and 0.36
(females) for the 1970 U.S. population reported by the BEIR IV Committee. County
smoker proportions ranged from a low of 0.13 1n males and 0.19 1n females to a
high of 0.55 1n males and 0.47 1n females. As discussed above 1n the section
on methods above, smoker proportions are likely to be underestimated 1n this
-------
study while those used by the BEIR IV Committee are likely to overestimate
current rates. Baseline lung cancer rates in male and female smokers and
nonsmokers shown in Table 3 were computed using both BEIR IV assumptions and
smoker proportions estimated for this study. Baseline rates in male smokers and
nonsmokers computed for this study were 70 percent higher than those estimated
by the BEIR IV Committee based on 1970 U.S. estimates. Female smoker and
nonsmoker baseline rates for the NJ study were 30 percent higher than BEIR IV
estimates. Population baseline lung cancer rates from which subgroup rates were
projected were the same for both the BEIR IV and NJ studies at 0.067 for males
and 0.025 for females. The rates are unadjusted to reflect baseline rates in
persons not exposed to radon since only a 10 percent reduction would be
indicated. The baseline population lung cancer risk of 0.046 currently used for
the New Jersey population is 50 percent higher than the rate used for previous
NJ estimates of radon induced lung cancer mortality rates (7). Baseline lung
cancer rates used in this analysis were revised upward to be more in line with
present rates.
Excess radon-induced lifetime risk estimates for male and female smokers
and non-smokers based on application of ratios of lifetime risks extracted from
the BEIR IV report for estimated annual average radon progeny exposures are shown
in Table 4 for each county and the state. The estimates shown were computed from
baseline rates estimated for the NJ study. The lowest excess risk was 0.001 in
non-smoking females and the highest, 0.072, in smoking males cf the county with
the highest average exposure. Had BEIR IV baseline risks been used, excess risks
for male smokers/nonsmokers would be 70 percent and for female smokers/nonsmokers
30 percent lower than those shown. The statewide average excess risk of 0.005
would be expected to result in about 500 radon induced lung cancer deaths (LCDs)
each year in New Jersey. The geometric mean excess risk was estimated at 0.003.
Considering that 3 percent of the nation's population resides in New Jersey, our
estimate of 500 LCDs is 25 percent lower than the 650 that would be inferred from
the current central EPA central estimate of 21,600 for the nation or directly
from screening measurement results.
Radon induced LCDs in smokers would be expected to contribute 76 percent
to the total deaths and 70 percent of the deaths would be expected to occur in
males. These percentages and the annual number of radon related LCDs would
remain unchanged whether either NJ study or BEIR IV baseline lung cancer rates
were used to project radon-induced lung cancer rates.
Figure 1 shows the number of radon Induced LCDs projected for each county
and the average screening level for that county. Predicted LCDs range from 3
in Salem County to 47 in Morris County. Counties with the highest average radon
screening levels, Warren and Sussex, are also those less sparsely populated and
and therefore annual LCD's from these counties, 22 and 19, respectively are about
mid-range for all counties. What these findings indicate is that efforts to
reduce population risks, in order to have a measurable effect, must be addressed
in both high and low radon areas.
Implicit in the methodology for estimating state/county radon risks used
in this study are numerous assumptions which lead to significant uncertainties
in the estimated risks. All uncertainties related to the form and use of the
-------
risk model for estimating residential risks have been discussed in the BEIR IV
report and elsewhere and will not be repeated here. Estimates of New Jersey
population exposures contain additional uncertainties which may bias high or low
the associated risks. Among these are: 1) overrepresentation of rural, high
radon areas; 2) the adequacy of correction factors developed for estimating
annual average exposures, and; 3) estimates of occupancy factors may be biased
high since residents had to be at home during daytime hours in order to be
included in the survey. Excess risks for subgroups of smokers and nonsmokers
are of higher uncertainty than combined estimates for the population.
The estimates of radon risks for the New Jersey population presented
here, uncertainties recognized, are based on the best information available at
this time. As new data become available for New Jersey, further refinements of
risk estimates will be made.
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
-------
REFERENCES
1. August 17, 1989 Memorandum from U.S. Environmental Protection Agency Division
Director to EPA Regional Offices.
2. National Academy of Sciences, National Research Council. Report of the
Committee on the Biological Effects of Ionizing Radiations. Health risks
of radon and other Internally deposited alpha-emitters. Washington, DC:
National Research Council, National Academy of Sciences, National Academy
Press; 1988.
3. Abu-Jarad, Falah, and J.H. Fremlin. Seasonal variation of radon
concentrations in dwellings. Health Physics. 46(5):1126-1129, 1985.
4. Nero, A.V., M.B. Schwehr, W.W. Nazaroff, and K.L. Revzan. Distribution of
airborne radon-222 concentrations 1n U.S. homes. Science. 234:992-997, 1986.
5. Wilkening, Marvin, and Andreas Wicke. Seasonal variation of indoor Rn at
a location in the southwestern United States. Health Physics.
51(4):427-436, 1986.
6. Ranney, C., K.E. Miller, R. Machever, and M. Cahill. Seasonal variability
and bilevel distribution of radon and radon progeny concentrations in 200
New Jersey homes, in: Proceedings of the 1988 Symposium on Radon and Radon
Reduction Technology. U.S. Environmental Protection Agency, Washington,
D.C., 1988.
7. New Jersey Department of Environmental Protection. Task 5 Final Report-Risk
assessment, statewide scientific study of radon prepared by Camp Dresser &
McKee Inc. New Jersey Department of Environmental Protection, Trenton, New
Jersey, 1989.
-------
Table 1. Average Inter-floor and inter-season radon ratios and equilibrium
ratios for basement homes (values in parentheses are standard
deviations).
heating
NON-HEATING SEASON
HEATING SEASON
1st fl:
base Rn
base 1st fl
1st fl:
base Rn
base
equi 1
1st fl
equi 1
NON-HEATING:
HEATINQ SEASON
RADON
base 1st fl
forced
air
0.45
(0.28)
0.46
(0.24)
0.48
(0.27)
0.51
(0.12)
0.34
(0.18)
0.41
(0.18)
0.97
(0.82)
0.80
(0.98)
hot water/
electric
0.27
(0.22)
0.40
(0.15)
0.50
(0.20)
0.33
(0.24)
0.37
(0.23)
0.38
(0.17)
0.89
(0.43)
0.75
(0.54)
Table 2. Average radon, radon progeny, and exposure data by county for
basement homes.
County
Radon
Screening
Level
(pCi/1)
Annual
Average Rn
Prog Cone ~
(WL)
Base Upstr
Exposure
(WLM/yr)
Base Upstr
Occupancy
(hrs/day)
Base Upstr
Atlantic
Bergen
Burlington
Camden
Cape May
Cumberland
Essex
Gloucester
Hudson
Hunterdon
Mercer
Middlesex
Monmouth
Morris
Ocean
Passaic
Salem
Somerset
Sussex
Union
Warren
Statewide
1.2
1.9
2.1
2.3
0.8
2.2
1.3
3.7
2.7
7.3
5.4
2.3
3.6
6.0
1.5
3.2
2.8
5.1
6.8
1.6
12.5
2.9
0.005 0.002 0.007 0.069
0.007 0.003 0.016 0.103
0.008 0.004 0.014 0.127
0.009 0.005 0.054 0.146
0.003 0.001 0.006 0.041
0.009 0.004 0.011 0.127
0.005 0.002 0.013 0.074
0.015 0.007 0.052 0.186
0.010 0.004 0.040 0.116
0.028 0.012 0.048 0.418
0.020 0.009 0.079 0.269
0.009 0.004 0.017 0.134
0.014 0.006 0.022 0.218
0.023 0.010 0.051 0.335
0.006 0.003 0.006 0.103
0.012 0.005 0.034 0.165
0.011 0.005 0.012 0.167
0.020 0.009 0.029 0.298
0.026 0.010 0.049 0.348
0.006 0.003 0.010 0.097
0.047 0.019 0.127 0.633
0.5
1.1
0.8
2.2
0.9
0.7
1.2
1.3
1.2
0.8
1.0
0.9
0.7
1.0
1.0
0.7
1.6
0.5
0.6
0.8
1.0
15.2
15.7
16.1
15.2
17.5
15.3
15.2
16.4
16.4
16.4
15.6
16.5
16.5
15.8
15.6
17.8
.2
.2
16.1
16.9
15.9
15.
17.
0.011 0.005 0.028 0.160 1.0
* Estimated using correction factors shown 1n Table 1 and radon screening
results.
Table 3. Baseline lung c*nc»r risks estimated in smokers and nonswokers based on NJ study
and BEIR IV smoking proportion assumptions.
ana ocih iv soioking prafm> w ¦ w¦ ——¦-
Males
Smoker Nonsmoker Smoker
Proportion Baseline Risk Baseline Risk
Smoker
Proportion
Females
Nonsmoker
Baseline Risk
Smoker
Baseline Risk
NJ
Study
BEIR IV
0.25 0.019 0.203
0.48 0.011 0.123
0.27
0.36
0.008
0.000
0.075
0.0S8
-------
Table 4. Average excess lifetime radon-induced lung cancer mortality risks
Average Total
Exposure * Males Females Population
County (WLM/yr) Non-smokers Smokers Non-smokers Smokers Risk
Atlantic
0.08
(0.
9)
0.002
0.008
0.001
0.004
0.002
Bergen
0.13
(1.
3)
0.002
0.013
0.001
0.005
0.003
Burlington
0. 14
(1.
3)
0.002
0.013
0.001
0.007
0.003
Camden
0.21
(2.
5)
0.002
0.021
0.001
0.009
0.005
Cape May
0. 11
(1.
1)
0.002
0.011
0.001
0.005
0.005
Cumberland
0. 14
(1.
9)
0.002
0.014
0.001
0.007
0.003
Essex
0. 10
(1.
3)
0.002
0.010
0.001
0.004
0.002
Gloucester
0.23
(3.
5)
0.002
0.022
0.001
0.010
0.006
Hudeon
0. 17
(2.
5)
0.002
0.016
0.001
O.OOB
0.005
Hunterdon
0.46
(7.
0)
0.005
0.043
0.003
0.020
0.01 t
Mercer
0.35
(5.
0)
0.003
0.034
0.001
0.014
0.008
Middlesex
0. 15
(2.
0)
0.002
0.014
0.001
0.007
0.005
Monmouth
0.22
(3.
0)
0.002
0.021
0.001
0.009
0.006
Morris
0.39
(5.
7)
0.003
0.019
0.001
0.016
0.008
Ocean
0.30
(1.
0)
0.002
0.011
0.001
0.005
0.003
Passaic
0.22
<3.
1)
0.002
0.021
0.001
0.009
0.005
Salem
0. 18
(2.
4)
0.002
0.018
0.001
o.ooe
0.003
Somerset
0.33
(4.
8)
0.003
0.032
0.001
0.014
o.ooa
Sussex
0.40
(6.
4)
0.005
0.038
0.001
0.017
0.011
Union
0.12
(1.
6)
0.002
0.011
0.001
0.005
0.003
Warren
0.76
(11
.6)
O.OOB
0.072
0.005
0.052
0.018
Statewide **
0. 19
(2.
5)
0.002
0.019
0.001
O.OOB
0.005
* Values in parentheses are average lowest floor radon screening levels in pCi/1
for basement and slab-on-grade/crawl space homes.
*~ Population weighted average excess risk.
-------
Figure 1. Average radon screening levels and annual radon Induced
lung cancer death estimates for New Jersey counties.
-------
A-II-2
INDOOR RADON EXPOSURE IN NORWAY AND LUNG CANCER RISK
Tore Sanner1 and Erik Dybing2
'institute for Cancer Research, The Norwegian Radium Hospital, Oslo, Norway,
National Institute for Public Health, Oslo, Norway.
ABSTRACT
The risk for lung cancer due to indoor radon in Norway was estimated. The
risk factor recommended by the World Health Organization was used. Corrections
were made for time not spent at home and type of activity. On the basis of
measurements by the Norwegian National Institute for Radiation Hygiene in
about 1.500 homes, it was estimated that the average concentration of radon
daughters was 55 Bq/m3 EER in dwellings on the ground floor or on both the
ground floor and first floor and 18 Bq/m3 EER-in dwellings on the first floor.
The level of exposure during tim? spent outside the home was assumed to be 10%
of that at home. It was calculated that indoor radon exposure may cause 110 -
340 lung cancer deaths per year. This corresponds to about 8 - 25% of all
lung cancer deaths in Norway.
-------
INTRODUCTION
Indoor exposure to radon may induce lung cancer. The induction of lung
cancer is caused by the deposition of radioactive decay products from radon
in the lung. The indoor level of radon and its radioactive decay products are
usually higher than ambient levels. Influx of radon gas from the ground into
basement of dwellings, is the most important source of indoor radon in Norway.
The indoor levels of radon and its decay products are highest in the basement
and decrease upwards in the dwellings. The effective dose equivalent received
will depend not only on the level of radioactive radon decay products and
their physical state, but also on the time spent in the dwelling and level
of activity.
RISK OF LUNG CANCER DEATH
Several international groups of experts have estimated the risk of lung
cancer death in association with exposure to radon and its radioactive decay
products. The data are mainly based on studies of workers in underground
mining. An equilibrium factor of 0.5 was used in calculation of indoor levels
of radioactive radon decay products from levels of radon (1). This factor
depends on the rate of air exchange, humidity and concentration of particles
in the air. Table 1 shows that large variations exist between the proposed
risk factors for lung cancer per Bq year/m3 EER (equilibrium equivalent
radon).
WHO (11) has for indoor radon exposure recently estimated a lung cancer
risk of 1-3 times the product (C x F0/106), where C is the indoor level of
radon decay products in Bq/m3 EER and FQ is the occupancy factor. In the
calculation of risk for lung cancer death in Norway, the risk estimate of WHO
has been used.
CALCULATION OF DOSE EQUIVALENT
The absorbed dose of radioactivity will depend on the amount of
radioactive radon decay products inhaled into the lungs. The calculation of
dose equivalent on the basis of the concentration of radon daughters is
uncertain. Such calculations have, however, been carried out in the present
paper in order to use the risk factor recommended by WHO and at the same time
taking into account specific factors such as type of activity at home and
housing conditions for the average Norwegians. The factor 0.076 mSv per Bq
year/ar EER for calculation of dose equivalent has been recommended by an OECD
expert group in 1983 (12). This factor is based on 100X indoor occupancy and
16 hours of light activity and 8 hours of resting.
-------
TABLE 1. RISK FOR LUNG CANCER DEATH FROM INDOOR EXPOSURE TO RADIOACTIVE RADON
DECAY PRODUCTS. THE RISK FACTORS GIVE THE RISK FOR LUNG CANCER DEATH DUE TO
RADON EXPOSURE IN A TYPICAL POPULATION WITH A LIFETIME OF 70-75 YEARS*
Report Lifetime risk for lung cancer death per 10® persons
per Bq year/m3 EER
AECB (2) 8.3
BEIR III (3) 10.1
BEIR IV (4) 4.9
EPA (5) at 74 Bq/m3 EER 3. 1 - 12.0
at 370 Bq/m3 EER 2.9 - 10.1
at 3.700 Bq/m3 EER 2.1 - 3.7
UNSCEAR (6) 2.8 - 6.3
NCRP (7) exposure 1 year 1.4
exposure the whole life 0.8
Swedish radon report (8) 3.1
ICRP (9) 2.1 - 6.3
ICRP (10) 1.3 - 2.9
WHO (11) 1 - 3
*A11 numbers have been calculated to an occupancy of 100Z.
As a first step in determination of lung cancer risk, attempts were made
to determine the average dose equivalent as a function of indoor levels of
radioactive radon decay products and the activities of the persons in the
dwellings.
NCRP (7) has calculated that during light activity at the same level of
radon daughters, men and women will receive nearly the same dose equivalents.
During resting, men will receive a dose corresponding to 2/3 of the dose at
light activity, while women will receive a dose corresponding to 1/2 of that
at light activity. Moreover, with 16 hours of light activity and 8 hours of
resting, the dose equivalent received by women is approximately 90% of the
dose equivalent received by men (12). The data in Table 2 were obtained on
the basis of calculations by NCRP (7) using the factor 0.076 mSv per Bq
year/m3 EER (12).
-------
TABLE 2. CALCULATION OF DOSE EQUIVALENT PER HOUR AS A FUNCTION OF THE LEVEL
OF RADIOACTIVE RADON DECAY PRODUCTS FOR PERSONS WITH DIFFERENT ACTIVITIES
Sex
Activity
4
nSv per Bq/m EER per hour
Male
light activity
resting
10.3
6.8
Female
light activity
resting
9.8
5. 1
HOUSING CONDITIONS
The population and housing census of the Central Bureau of Statistics of
Norway (13, 14) shows that among people living in sparsely populated areas,
about 98% live in dwellings on the ground floor or on both the ground floor
and the first floor, while among people living in cities with more than
100.000 inhabitants, approximately 50X have homes on ground floor or ground
floor and first floor. On the average 2.7 persons are living in each home.
Based on the housing statistics it is assumed that about 75% of the Norwegians
or 3.000.000 people have homes on the ground floor or ground floor and first
floor, while 10% or 400.000 are living in dwellings on the first floor. It
is assumed that persons living in the first floor will receive a dose which
is approximately 1/3 of the dose to persons living on the ground floor or
ground floor and first floor.
TINE BUDGET
A time budget survey has been carried out by the Central Bureau of
Statistics of Norway (15, 16). They found that retired persons and home
working women spent more than 80% of the day in their own homes, while in the
working population, men spent about 58-60% at home and women 65-72%. As an
average, men spent 14.6 hours per day at home. Of this 6.4 hours were used for
light activity and 8.2 hours for resting. Women spent 17.5 hours per day at
home. Of this 9.2 hours were used for light activity and 8.3 hours for
resting. The time spent outside the homes were used for light activities.
Based on the data in Table 2, it follows that a Norwegian man living in a home
with an average level of radon decay products of 100 Bq/m3 EER will at home
receive a dose equivalent of 4.4 mSv per year while a woman will receive 4.8
mSv per year. In addition the person will also be exposed to radon decay
products at work and when visiting other people. As an average the level of
exposure to radioactive radon products received at the time the person is
not at home is assumed to be 10% of that at home. This implies that the total
dose equivalent for a man and a woman respectively, living in a dwelling with
100 Bq/m3 will be 4.8 mSv and 5.0 mSv per year.
-------
LEVELS OF RADON IN NORWEGIAN DWELLINGS
The Norwegian National Institute for Radiation Hygiene has performed
radon measurements in about 1,500 homes (blocks of flats excluded) using an
integrating dosimeter. In each building, measurements were made for equal time
periods in the living room and in one of the bedrooms. All measurements were
made during the heating season (17). The indoor level of radon varied with the
time of the year, and it was found on the basis of measurements in 47 houses
that the average concentration during the summer was about 50% of the average
concentration during the winter (17). This is in agreement with other
measurements (18, 19). It was concluded that the average level during the year
of radioactive radon decay products was 55 Bq/m3 EER in dwellings located on
the ground floor or ground floor and first floor; and 18 Bq/m3 EER in
dwellings on the first floor.
RISK OF LUNG CANCER DEATH FROM INDOOR RADON IN NORWAY
Table 3 shows the calculated number of lung cancer deaths based on the
WHO risk factors. This corresponds to a life time risk of lung cancer
mortality of 13.2 - 40 per 10® persons per mSv using the factor of 0.076 mSv
per Bq year/m3 EER (12). It was calculated that in Norway approximately 110-
340 cases of lung cancer deaths per year may be due primarily to indoor
exposure to radon decay products. This corresponds to about 8 - 25% of all
lung cancer deaths in Norway. Since the indoor levels of radon and its decay
products probably have increased during the last years, these numbers do not
necessarily reflect the role of radon and its decay products in inducing lung
cancer today, but they represent the annual number of lung cancer deaths which
may be expected should the present level of radon and its radioactive de'cay
products in the future be at the present level.
Epidemiological studies as well as experimental data, indicate that
smoking increases the risk of lung cancer in a synergistic manner (multi-
plicatively or submultiplicatively) (4, 21). Moreover, as pointed out above
the main data available for estimating risk factors for lung cancer death in
connection with radon are based on studies of male workers in underground
mining. Lung cancer in connection with radon exposure will thus, preferen-
tially occur among people who smoke and since men have been smoking for a
longer time than women and the prevalence of daily smokers as well as the
number of cigarettes smoked per day is lower among women than among men, the
risk factor used for women is probably to high.
-------
TABLE 3. LUNG CANCER DEATHS IN CONNECTION WITH EXPOSURE TO RADON DECAY
PRODUCTS IN NORWAY*.
Location of
Number of
Radon decay
Dose
Number of lung cancer
dwelling
persons
products
equivalent
death
Men Women
Men Women
(Bq/m3 EER)
(mSv)
Ground floor
or ground
and first floor
3.000.000
55
2.6 2.8
51 - 156 55 - 168
First floor
400.000
18
0.9 1.0
2-7 3-8
Total
53 - 163 58 - 176
"it is assumed in the calculations that the numbers of men and women are
equal. The calculation of dose equivalents were based on the numbers given in
the paragraph "TIME BUDGET", The exposure during time spent outside the
dwellings is assumed to be 5.5 Bq/m3 EER both for persons living in dwellings
on the ground floor or ground floor and first floor and for persons living in
dwellings on the first floor. No estimates has been made for persons living
in dwellings above the first floor.
ESTIMATION OF LUNG CANCER DEATHS FROM INDOOR RADON BASED ON ESTIMATES
FROM DIFFERENT REPORTS
Due to difference in the proposed risk factors (see Table 1), it follows
that the estimated number of lung cancer deaths caused by a specific indoor
level of radon will differ. It was therefore of interest to calculate the risk
for lung cancer death per 1.000 deaths at a radon decay product level of 100
Bq/m3 EER based on estimates from different reports. The results are shown in
Table 4.
The present calculations are based on the risk factor proposed by WHO
(11) which is very close to that recently proposed by ICRP (10). Our
calculations agree well with those from NCRP (7), but are considerably lower
than the rest of the estimates. It should be noted the factor (12) used for
calculation of the dose equivalent in the calculation of the number of lung
cancer deaths in Norway does not affect the comparison of the numbers in
Table 4. The estimates by EPA (5) cover a wide range and their lower estimate
corresponds to the Swedish estimate (8) and to the upper limit of our
estimates. Part of the differences in the numbers are probably due to differ-
ences in the time spent at home and level of activity used for the calcula-
-------
cions. However, it is apparent from Table 1 that the risk factor proposed for
calculation of lung cancer deaths after exposure to radon has decreased
considerably during the last years. Thus, the risk factor proposed by BEIR III
(3) in 1980 was two times higher than the factor proposed by BEIR IV (4) in
1988. This is also true when the factor proposed by ICRP in 1979 and 1986 is
compared (9, 10).
TABLE 4. RISK FOR LUNG CANCER DEATH PER 1.000 DEATHS AT A RADON
DECAY PRODUCT LEVEL OF 100 Bq/m3 EER BASED ON ESTIMATES FROM
DIFFERENT REPORTS*
Report Number of lung cancer deaths per 1.000 deaths
This report 5-15
Swedish radon report (8) 19
EPA (5) 18 - 68
NCRP (7) 10
UNSCEAR (6) 21 - 48
BEIR III (3) 76
BEIR IV (4) 37
The numbers have been calculated on the basis of data given in
the report.
The work described in this paper was not funded by Che U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the
views of the Agency and no official endorsement should be inferred.
REFERENCES
1. United Nations Scientific Committee on the Effects of Atomic Radiation
(UNSCEAR). Ionizing radiation: Sources and biological effects. United
Nations, New York, 1982.
2. Atomic Energy Control Board (AECB). Risk estimate for exposure to Alpha
Emitters (ACRP-1) (INFO-0090), Advisory Committee on Radiological
Protection. Ottawa, Canada, 1982.
3. National Research Council. The effects to populations of exposures to
low levels of ionizing radiations. Committee on the Biological Effects
-------
of Ionizing Radiation (BEIR III), National Academy Press, Washington
D.C., 1980.
4. National Research Council. Health risks of radon and other internally
deposited alpha-emitters. Committee on the Biological Effects of
Ionizing Radiation (BEIR IV), National Academy Press, Washington D.C.,
1988.
5. EPA. A citizen's guide to radon. What it is and what to do about it.
EPA OPA-86-004. US Environmental Protection Agency, Washington D.C.,
1986.
6. United Nations Scientific Committee on the Effects of Atomic Radiation
(UNSCEAR), sources and effects of ionizing radiation. Report to the
General Assembly, with Annexes, UN publication, E. 77 IS 1., United
Nations, New York, 1977.
7. National Council on Radiation Protection and Measurements (NCRP).
Evaluation of occupational and environmental exposures to radon and
radon daughters in the United States. NCRP Report No. 78, Bethesda,
MD, 1984.
8. Sveriges Radonutredning - Preliminart forslag til Atgarder mot
strAlrisker i byggnader. PM fr&n radonutredningen. Jordbruksdeparte-
mentet. DsJol979:9, Stockholm, 1979 (in Swedish).
9. International Commission on Radiological Protection (ICRP). Limits for
intake of radionuclides by workers. Publication 30. Pergamon Press,
Oxford, 1979.
10. International Commission on Radiological Protection (ICRP). Lung cancer
risk from indoor exposures to radon daughters. Publication 50. Pergamon
Press, Oxford, 1986.
11. World Health Organization (WHO). Indoor air quality: Radon and
formaldehyde. Environmental health. 13. World Health Organization,
Copenhagen, 1986.
12. Nuclear Energy Agency (NEA). Dosimetry aspects of exposure to radon
and thoron daughter products. Report by a NEA/OECD, Paris, 1983.
13. Central Bureau of Statistics of Norway. Population and housing census
1980. Oslo - Kongsvinger, 1982.
14. Central Bureau of Statistics of Norway. Survey of housing conditions
1981. Oslo - Kongsvinger, 1983.
15. Central Bureau of Statistics of Norway. The time budget survey 1971-72
Vol. I. Oslo, 1975.
-------
16. Central Bureau of Statistics of Norway. The time budget survey 1980-
81. Oslo - Kongsvinger, 1983.
17. Stranden, E. Radon-222 in Norwegian dwellings. In: Radon and its decay
products. Occurrence, properties and health effects (eds PK Hopke) ACS
Symposium Series 331; 70-83, American Chemical Society, Washington DC,
1987.
18. George, A.C. and Hinchliffe, L.E. Measurements or radon concentrations
in residential buildings in the eastern United States. In: Radon and its
decay products. Occurrence, properties and health effects (ed PK Hopke)
ACS Symposium Series 331: 42-62, American Chemical Society, Washington
DC, 1987.
19. Dudney, C.S., Hawthorne A.R., Bull, L.A., Cohen, M.A,, Daffron, C.R.,
Orebaugh, C.T. Radon levels in 300 houses in Roane County, Tennessee.
In: Proceedings of the 4th International Conference on Indoor Air
Quality and Climate. Vol 2. Berlin 1987. p. 393-397.
20. Sveriges Cancer Kommitte - Cancerkoomitten, cancer Arsaker, forebyggande
m.m. Betankande av Cancerkommitten, SOU, 1984:67, Sosialdepartementet,
Stockholm, 1984 (in Swedish).
21. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Man-
made Mineral Fibres, and Radon. Vol 43. Lyon 1988. p. 173-259.
-------
A-II-3
VALIDITY OF VARIOUS EPIDEMIOLOGICAL
APPROACHES TO ASSESSING RADON HEALTH RISK
by: Susan M. Conrath
U.S. Environmental Protection Agency
Washington, DC
abstract
Various study designs (i.e., ecologic, cohort, case control) will be
defined and evaluated for their utility in assessing radon health risk. The
strengths and limitations of these approaches will be addressed.
Common pitfalls and errors of epidemiologic method will be described
with examples and explanation of resultant consequences.
This paper has been reviewed in accordance with the U.S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
-------
A-II-4
ASSESSMENT OF HEALTH IMPACTS OF RADON EXPOSURES IN FLORIDA
by: W.T.Vonstille, Professor and Director,
Institute of Environmental Health and Toxicology
College of Health, University of Central Florida
Orlando, FL 32816
Hildegarde L.A.Sacarello, Director
Environmental and Toxicological Research, Hy-Tox
6944 Crown Gate Drive
Miami Lakes, FL 33014
ABSTRACT
Residential radon levels, from a statewide Florida survey, were used in an
analysis of over 150,000 medically treated episodes of malignancies and other
serious illnesses and conditions in whites, blacks and hispanics from all counties
in the state. No evidence of an increased percentage of cancer was found in
any sex or ethnic group from the areas with the highest radon exposure levels.
Age adjustment of data did not affect the results. The highest radon exposures
were associated with some of the lowest cancer rates and contradict the risk
assessment hypothesis based on extrapolation from exposures in mining. Points
for DOE and EPA errors in risk assessment methods are reviewed; predictions
from risk assessment should be empirically tested as in the case of any other
scientific hypothesis before being used as a basis for public policy. Thus, we
find that cancer risks of residential radon have been vastly overstated.
INTRODUCTION
Our previous paper (1) reported on research to find direct health hazard
evidence in Florida to support the claim that "20,000 annual lung cancer deaths
may be due to U.S. exposures to radon" (2). This often repeated environmental
hazard claim indicates that up to 20% of U.S. lung cancers may be caused by
radon. This claim remained untested even though cigarette smoking, asbestos
exposure (in buildings, automotive products, etc.), and urban air pollution have
each been supported by evidence from health studies as major or as the main
causes of lung and other cancers. Our report noted that the recent NIH atlas
of cancer rates by U.S. counties raised further etiological doubts about radon as
a carcinogen by showing that lung cancer death rates are highest in and near
the major industrial centers along the eastern and gulf coasts, the lower
-------
Mississippi River, and other urban areas. Elevated rates of lung cancer do not
occur in the radon areas of the Appalachian and Rocky Mountains, nor in the
radon belts of Central Florida (3). In general terms, this suggests that the role
of non-mining radon exposures in causing cancer may be small. Hence, we
began specific health studies of human cancers and radon levels to verify risk
assessments made using extrapolations based on lung cancer risks of high
exposure mining occupations.
The percentages of cancer versus other diseases and conditions requiring
medical treatment, analyzed by levels of radon exposure in all parts of Florida
failed to provide evidence of a hazard due to radon. Since the latency period
of lung cancer averages about 37 years, lung cancers of recent emigrants to
Florida would reflect exposures elsewhere rather than radon levels in Florida.
However, much of the recent emigration has been into the coastal locations
with the lowest radon levels or with essentially no excess radon exposure. The
highest radon levels are found in some of the phosphate and agriculture areas
of Central Florida with fewer recent emigrants, especially among blacks.
Hence, the cancers in whites and blacks, particularly in the areas of highest
radon levels, are expected to be the most sensitive indicator of a health hazard
of radon, if such exits.
METHODS
In 1985 the Florida Legislature initiated on a statewide survey of radon
exposures, and a summary of the final report was obtained from the Florida
Institute of Phosphate Research, Bartow, Florida--the organization commissioned
by the Legislature to make the radon measurements (4). Radon risks were
categorized from indoor and outside measurements, and the findings were
summarized by counties and by quadrangles of the U.S. Geological Survey maps.
About 6,500 homes were sampled and one or two school buildings per county
were also selected for inside radon measurements. Outdoor gamma radiation and
aerial radiation measurements from the National Uranium Resource Evaluation
were also used in the study to produce an index of possible radon exposure.
Definite evidence of elevated radon potential was found in 18 counties; limited
evidence in 14 counties; and no evidence of an elevated radon potential was
found in 35 counties. Figure 1, from the study (4), indicates the Florida
counties with elevated radon levels. Four of these counties, Alachua,
Hillsborough, Marion and Polk, were further described as having the highest
radon levels. A range of 5% to 22% of the homes sampled within each of
these four counties were found to have radon levels above the EPA action
standard of 4 pico Curies of radon per liter of air (5). Within these counties,
the radon Burvey noted the U.S. Geological Survey quadrangles with the highest
radon levels. The quadrangles with the highest radon levels in the four highest
radon level counties were indicated by shading in Figure 1. As covered below,
people living in the shaded quadrangles in the four highest radon counties could
be expected to have the highest radon exposures in the state and are designa-
ted as the "Highest Radon" exposure group.
Health information was obtained from the proprietary data base of serious
illnesses and conditions from all localities throughout the Southeast operated
jointly by Health Accounting/Assessment, Inc. and the College of Health,
-------
FIGURE 1. Radon gas levels of in-door environments of Florida from the
Statewide survey (4). The zip codes in the shaded quadrangles, shown above,
with the highest radon levels of the four counties with the highest radon levels
are designated as the "Highest Radon Levels" for this analysis. The remaining
zip codes of the shaded counties, shown above, are designated as "Low Radon
Levels" for this analysis. The non-shaded counties have no added radon
exposures in excess of ambient levels in air and are designated an "None".
University of Central Florida. All diseases and conditions were medically
diagnosed according to the International Classification of Disease, I CD, coding
recommendations and an episode of serious illness was defined as requiring one
or more days of bed rest, hospitalization or both. Additional information for
each illness episode included data on age, sex, race-ethnicity and zip code of
the persons involved. Florida data were extracted for analysis according to the
level of radon exposure. For 1987-88, over 200,000 privacy secure abstracts of
episodes were available. Illness data for whites, blacks and hispanics were well
represented. Oriental, American Indians and other minority groups were found
to have less than 100 records in the present data base at one or more of the
radon levels and were excluded from analysis because of small numbers.
-------
Episodes were also excluded from analysis if there were missing data items or if
the 1CD code indicated normal deliveries and child births (in the absence of
pathology, no effect of radon was expected). A total of 68,344 female and
29,645 male episodes of serious illness conditions were available for analysis.
To estimate the radon exposure, the zip code of each illness episode was
matched to the zip codes of the statewide radon survey maps and the radon
level determined. U.S. Post Office zip codes for Florida were located in U.S.
Geological Survey quadrangles and classified as high, low or having no radon
hazard potential. Of the 3,000 zip codes assigned to Florida, 99 were located in
the highest radon exposure quadrangles of the four counties with the highest
radon levels; 918 zip codes were located in counties with no radon hazard
potential (under .2 pico Curries per liter or .001 VTL); and 1,983 zip codes were
found in the low radon level quadrangles. A few zip codes that could not be
found were also included with this group after determining they were definitely
not in the highest radon exposure areas. The records were classified by zip
code by computer into the radon exposure categories and summarized by major
disease groups as shown in Table 1. Since studies of atom bomb and other
radiation survivors reported higher rates of a variety of cancers and leukemia
in various age groups, lung cancers and all malignant neoplasms were grouped
together for analysis. Since census data by radon exposure or by sex, age and
zip codes is not available, disease rate analysis is not possible. Our analysis
uses comparisons of the percentages of malignant neoplasms by radon exposure
levels. If higher radon exposure causes more lung or other cancers, as alleged,
then the percentages of lung cancer and other malignant neoplasms must be
greatest in people with high radon exposure; the lowest percentages must be in
those unexposed; and the percentages must be intermediate in people with low
radon exposures.
RESULTS
Table 1 shows that white, black and hiBpanic males from the localities with
the highest radon levels, have percentages of malignant neoplasms very similar
to males from essentially unexposed areas. Each ethnic group of males from
localities with the highest radon levels have lower percentages of cancers than
males from the low radon areas. Age adjustment of the white and black cancer
data, using the 1985 Florida populations as a standard, did not change the
relative differences among the highest, low or essentially no radon exposure
localities. Small numbers for some age groups and exposures prevented
meaningful age adjusted analysis for hispanics.
Table 2 shows that white, black and hispanic females from the localities
with the highest radon levels, have percentages of malignant neoplasms not only
very similar to, but actually lower than, females from essentially unexposed
areas. As in the case of males, each ethnic group of females from localities
with the highest radon levels have lower percentages of cancers than females
from the low radon areas. Age adjustment of the white and black cancer data,
using the 1985 Florida populations as a standard, did hot change the relative
differences among the highest, low or essentially no radon exposure localities.
-------
TABLE 1. FLORIDA MALES, SERIOUS ILLNESSES OF ALL AGES, 1987-1988
Radon Level: Highest Low or ??? None
Males-White
Infections
79
5.2%*
428
4.5%
311
5.0%
Malignant Neoplasms
71
4.7 %
631
6.6%*
266
4.3%
Age Adjusted MN
6.07%
6.16%*
4.94%
Benign Neoplasms
10
0.7%*
34
0.4%
41
0.7%
Metabolic-Blood
61
4.0%*
356
3.7%
237
3.8%
Mental Conditions
204
13.5%
1,473
15.4%*
946
15.1%
Nervous System
65
4.3%*
305
3.2%
260
4.2%
Circulatory System
111
7.3%
1,365
14.3%*
563
9.0%
Respiratory System
260
17.2%*
1,118
11.7%
861
13.8%
Digestive System
161
10.7%*
874
9.2%
585
9.4%
Genitourinary System
54
3.6%
600
6.3%*
253
4.0%
Skin Conditions
31
2.1%*
154
1.6%
123
2.0%
Musculoskeletal
31
2.1%
226
2.4%
191
3.1%*
Congenital-birth
117
7.7%
570
6.0%
509
8.1%*
Symptoms, etc.
104
6.9%*
543
5.7%
405
6.5%
Injuries
102
6.8%
549
5.8%
476
7.6%*
Poisoning, etc.
50
3.3%
308
3.2%
222
3.6%*
Total all conditions
1,511
100.0%
9,534
100.0%
6,249
100.0%
FL Males-
-Black
Infections
50
7.0%
345
6.9%
331
7.3%*
Malignant Neoplasms
23
3.2%
187
3.7%*
127
2.8%
Age Adjusted MN
5.66%
6.66%*
5.16%
Benign Neoplasms
2
0.3%
17
0.3%
24
0.5%*
Metabolic-Blood
48
6.7%
426
8.5%*
313
6.9%
Mental Conditions
57
7.9%
432
8.6%*
338
7.4%
Nervous System
38
5.3%*
197
3.9%
183
4.0%
Circulatory System
30
4.2%
262
5.2%*
235
5.2%
Respiratory System
137
19.1%
727
14.5%
974
21.5%*
Digestive System
82
11.4%*
352
7.0%
371
8.2%
Genitourinary System
30
4.2%*
189
3.8%
177
3.9%
Skin Conditions
14
2.0%
114
2.3%
124
2.7%*
Musculoskeletal
12
1.7%
103
2.1%
114
2.5%*
Congenital-birth
63
8.8%
776
15.5%*
420
9.3%
Symptoms, etc.
43
6.0%*
257
5.1%
263
5.8%
Injuries
65
9.1%
468
9.4%*
421
9.3%
Poisoning, etc.
23
3.2%*
151
3.0%
122
2.7%
Total all conditions
717
100.0%
5,003
100.0%
4,537
100.0%
(Continued)
-------
TABLE 1. FLORIDA MALES, SERIOUS ILLNESSES OF ALL AGES, Continued
Radon Level: Highest Low or ??? None
Males-Hispanic
Infections
7
9.1%*
147
8.2%
19
8.4%
Malignant Neoplasms
0
0.0%
66
3.7%*
3
1.3%
Benign Neoplasms
2
2.6%*
13
0.7%
2
0.9%
Metabolic-Blood
1
1.3%
43
2.4%
17
7.5%*
Mental Conditions
2
2.6 %
68
3.8%*
7
3.1%
Nervous System
2
2.6%
54
3.0%
12
5.3%*
Circulatory System
1
1.3%
93
5.2%
14
6.2%*
Respiratory System
15
19.5%
250
14.0%
57
25.1%*
Digestive System
14
18.2%*
179
10.0%
34
15.0%
Genitourinary System
2
2.6%
71
4.0%
11
4.8%*
Skin Conditions
2
2.6%*
38
2.1%
1
0.4%
Musculoskeletal
0
0.0%
37
2.1%*
3
1.3%
Congenita1-birth
13
16.9%
413
23.1%*
11
4.8%
Symptoms, etc.
8
10.4%*
98
5.5%
11
4.8%
Injuries
6
7.8 %
164
9.2%*
20
8.8%
Poisoning, etc.
2
2.6%
56
3.1%*
5
2.2%
Total all conditions
77
100.0%
1,790
100.0%
227
100.0%
* * Maximum for disease condition
TABLE 2. FLORIDA FEMALES, SERIOUS ILLNESSES OF ALL AGES, 1987-1988
Radon Level: Highest Low or ??? None
Females-
•White
Infections
100
2.9%
737
2.5%
477
3.1%*
Malignant Neoplasias
98
2.9%
,1,493
5.1%*
477
3.1%
Age Adjusted MN
3.19%
4.35%*
3.32%
Benign Neoplasms
34
1.0%
298
1.0%*
120
0.8%
Metabolic-Blood
125
3.7 %
1,254
4.3%
672
4.3%*
Mental Conditions
390
11.5%
3,462
11.8%
1,907
12.2%*
Nervous System
99
2.9%
709
2.4%
478
3.1%*
Circulatory System
258
7.6%
4,930
16.8%*
1,357
8.7%
Respiratory System
438
12.9%*
2,638
9.0%
1,664
10.6%
Digestive System
276
8.1%
3,049
10.4%*
1,382
8.8%
Genitourinary System
881
25.9%*
4,720
16.1%
3,689
23.6%
Skin Conditions
45
1.3%
393
1.3%
243
1.6%*
Musculoskeletal
86
2.5%
945
3.2%*
488
3.1%
Congenital-birth
140
4.1%
841
2.9%
665
4.3%*
Symptoms, etc.
181
5.3%
1,675
5.7%*
829
5.3%
Injuries
144
4.2%
1,431
4.9%*
743
4.8%
Poisoning, etc.
111
3.3%*
743
2.5%
438
2.8%
Total all conditions
3,406
100.0%
29,318
100.0%
15,629
100.0%
-------
TABLE 2. FLORIDA FEMALES, SERIOUS ILLNESSES OF ALL AGES, Continued
Radon Level: Highest Low or ??? None
Females-Black
Infections
72
5.5%*
385
4.2%
276
4.3%
Malignant Neoplasms
16
1.2%
218
2.4%*
132
2.1%
Age Adjusted MN
2.60%
3.11%*
3.02%
Benign Neoplasms
18
1.4%*
75
0.8%
45
0.7%
Metabolic-Blood
88
6.7%
572
6.2%
428
6.7%*
Mental Conditions
117
8.9%
1,046
11.3%*
498
7.8%
Nervous System
A2
3.2%*
216
2.3%
169
2.7%
Circulatory System
87
6.6%
553
6.0%
423
6.6%*
Respiratory System
152
11.6%
826
8.9%
763
12.0%*
Digestive System
88
6.7%
540
5.8%
459
7.2%*
Genitourinary System
370
28.1%
2,644
28.6%
1,899
29.8%*
Skin Conditions
26
2.0%*
178
1.9%
118
1.9%
Musculoskeletal
26
2.0%
179
1.9%
141
2.2%*
Congenital-birth
46
3.5%
790
8.6%*
245
3.8%
Symptoms, etc.
66
5.0%
383
4.1%
336
5.3%*
Injuries
60
4.6%
421
4.6%
291
4.6%*
Poisoning, etc.
41
3.1%*
206
2.2%
151
2.4%
Total all conditions
1,315
100.0%
9,232
100.0%
6,374
100.0%
Females-Hispanic
Infections
9
9.4%*
117
4.4%
11
3.7%
Malignant Neoplasms
1
1.0%
75
2.8%*
4
1.4%
Benign Neoplasms
1
1.0%
27
1.0%
7
2.4%*
Metabolic-Blood
2
2.1%
69
2.6%
13
4.4%*
Mental Conditions
2
2.1%
159
5.9%*
12
4.1%
Nervous System
1
1.0%
67
2.5%
13
4.4%*
Circulatory System
1
1.0%
119
4.4%*
6
2.0%
Respiratory System
15
15.6%*
191
7.1%
39
13.2%
Digestive System
11
11.5%
194
7.2%
35
11.9%*
Genitourinary System
28
29.2%
881
32.9%
98
33.2%*
Skin Conditions
2
2.1%*
34
1.3%
2
0.7%
Musculoskeletal
0
0.0%
43
1.6%*
4
1.4%
Congenital-birth
13
13.5%
394
14.7%*
18
6.1%
Symptoms, etc.
3
3.1%
120
4.5%*
12
4.1%
Injuries
4
4.2%*
111
4.1%
9
3.1%
Poisoning, etc.
3
3.1%
78
2.9%
12
4.1%*
Total all conditions
96
100.0%
2,679
100.0%
295
100.0%
* = Maximum for disease condition
-------
Figures 2 and 3 show the percentages of malignant neoplasms for white and
black males and females by ten year age groups. In Figure 2 (white males) the
solid line for high radon exposure repeatedly rises above and falls below the
dotted line for the no radon exposure group. In blacks, the solid line for high
radon exposures is below the other groups until age 60 and over. The dashed
line for low radon exposures tends to average slightly above the other groups.
In Figure 3 (white females) the solid line for high radon exposure levels shows
that higher percentages of cancers were found in age groups from 50 to 79, and
lower percentages at other ages. In black females, a similar pattern was found
with the solid line for the highest radon exposures associated with lower
percentages of cancers at all ages other than the 60 to 79 groups. From age
group to age group, the highest radon exposed females, shown by the solid line,
tended to have lower percentages of malignant neoplasms than the no radon
exposure group.
Clearly, the relative proportion of malignant neoplasms found in the areas
of highest radon exposure in Florida, compared to either the low or the
unexposed areas, give no indication of an increased cancer risk from radon
exposure.
DISCUSSION
There is a major disparity between our results and those of the NIH atlas
of cancers (3) versus the DOE and EPA risk assessment determination that non-
mining radon exposures are a major hazard to health. The recent book
reviewing risk assessment (6) contains a chapter devoted to the non-mining
health hazards of Radon-222 and its daughters. In the absence of direct health
studies of non-miners exposed to radon levels in the range found in homes,
work places, and other in-door environments, lung cancer data from iron and
uranium miners has been used. Errors of vast overstatement of the radon risks
appear to have been introduced into the risk assessment at many points--one or
more of the following examples may account for the disparity. Lung cancer in
iron and uranium miners may be caused by, or contributed to, by many other
job, work or community related exposures increasing the apparent mortality rate
that was attributed to radon exposure. Studies of prior work histories of
miners can, and have been found to account for apparent excess cancers (7)
instead of the carcinogenic exposures thought to be present in the current
employment under investigation. Where prior jobs involve exposure to lung
carcinogens, known or unknown, lung cancer latency periods will be found to
average less than the 37 year average reported in studies of smoking, asbestos,
and other lung carcinogens. Any inadvertent exaggeration of the magnitude of
radon risk is further compounded by deliberate choices of the most conservative
estimates used when more relevant facts are at hand. Unfortunately, an effort
to be sure that a possible hazard is identified carries the possibility of
overstatement of the radon hazard and needless economic burdens without
health benefits. It is regrettable that DOE and EPA risk assessments are not
immediately tested, as any other important scientific hypotheses might be, to
verify whether empirical observations support or will contradict risk assessment
predictions before having them enacting them into laws of the land.
-------
FIGURE 2. PERCENTAGES OF MALIGNANT NEOPLASMS, FLORIDA,
BY AGE AND RADON EXPOSURE LEVELS
WHITE MALES
0-10 20 30 40 50 60 70 80+
Age Group
BLACK MALES
0-10 20 30 40 50 60 70 80+
Age Group
Legend: Highest Radon Levels - - - Low Radon Levels . . . None
-------
FIGURE 3. PERCENTAGES OF MALIGNANT NEOPLASMS, FLORIDA,
BY AGE AND RADON EXPOSURE LEVELS
WHITE FEMALES
0-10 20 30 40 50 60 70 80f
Age Group
BLACK FEMALES
0-10 20 30 40 50 60 70 80+
Age Group
Legend: Highest Radon Levels ... Low Radon Levels . . . None
-------
CONCLUSIONS
Since 1986 the DOE and EPA have publicized the claim that up to 20,000
U.S. lung cancer deaths each year may be due to exposure to radon; our
analysis of illnesses by radon levels throughout Florida fails to find support for
this claim. We previously reviewed over 150,000 malignancies and other serious
illness conditions from all parts of the state. The current more detailed review
by ethnic group, sex and age finds that people in localities with the highest
radon levels tend to have a lower percentage of malignant neoplasms compared
to people from areas with no or lowest radon exposures. Our findings are
consistent with the NIH cancer atlas showing that U.S. lung and related cancer
death rates are highest in industrial areas and not in high radon areas of the
U.S. We conclude that the cancer risks of radon and its daughters are vastly
over stated. Further, we find it regrettable that risk assessments that may
form a basis for public policy are not treated as any other scientific hypothesis
and subjected to immediate testing for empirical verification.
This research was supported by Health Accounting/Assessment, Inc., Winter
Park, FL 32793-5051, using their proprietary Community Health Data Base and
data base methods and their assistance is gratefully acknowledged.
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the
views of the Agency and no official endorsement should be inferred.
REFERENCES
1. Vonstille, W.T. and Sacarello, H.L.A. Radon Exposures and Cancer in
Florida. J. Environmental Health. In Press, 1990.
2. U.S. Department of Energy. Indoor Air Quality Environmental Information
Handbook: Radon, Washington, D.C. 1986,
3. Pickle, L.W., Mason, T.J., Howard, N., Hoover, R., and Fraumeni, J.F.
Atlas of U.S. Cancer Mortality Among Whites: 1950-1980, U.S. DHHS
Publication No. (NIH) 87-2900. 1987.
4. Florida Institute of Phosphate Research. Florida Statewide Radiation
Study, Summary of Final Report, Florida Institute of Phosphate
Research, Bartow, FL. 1987.
5. U.S. Environmental Protection Agency. A Citizen's Guide to Radon, Report
OPA-86-004, Washington, D.C. 1986,
6. Hallenbeck, W.H. and Cunningham, K.M. Quantitative Risk Assessment For
Environmental and Occupational Health. Lewis Publishers, Inc., Chelsea,
Michigan. 1986.
7. Vonstille, W.T. and Tabershaw, I.R. The Mortality Experience of Upstate
New York Talc Workers. J. Occupational Medicine 24:480-484, 1982.
-------
A-II-5
REALISTIC EVALUATION OF TESTER EXPOSURE
BASED ON FLORIDA TESTING EXPERIENCE
by: Richard A. Schreiber
Razman Associates
1002 Fellsridge Court
Stone Mountain, Georgia 30083
ABSTRACT
A radon decay product exposure model for Florida Certified Radon
Measurement Technicians has been formulated based on the guidance
of 10CFR20. This model was used for to estimate the exposure of
44 Florida measurement technicians from January through November
of 1989. Comparing estimated testing and home exposure shows that
100% of the technicians observed received more exposure in the home
than during testing activities. Exposure during normal office
hours also exceed testing exposure in 86% of the technicians
observed. Health and safety exposure data for radon measurement
technicians does not follow the standard concepts of occupational
radiation exposure normally accepted in 10CFR20.
-------
BACKGROUND
The State of Florida requires that certified radon testing
personnel be monitored to track their exposure to radon decay
products and that a record of that exposure be maintained by the
testing company. The development of the testing company Worker
Health and Safety Plan is not specifically defined in the affecting
state rules and regulations.(1) The testing company is responsible
for the development of this plan which is executed by a certified
measurement specialists. Records of exposure are maintained by the
company for review by the regulating authority but are not reported
to that authority.
A Worker Health and Safety Plan was developed by Thomas R.
Stephenson of Certus Laboratories, Inc. and the author for Florida
testing companies using Certus radon detector analysis and
specialist consulting services.(2) This plan was predicated on
the principles of accepted radiation industry practice.(3) The
assumption is that the worker leaves an uncontrolled clean
environment and enters a controlled environment where there is
always a potential for unsafe radiation exposure. The object is
to record both controlled and uncontrolled (accidental) exposure
to the workers. Environmental and health related radiation
exposure is specifically excluded from consideration in calculating
the exposure record of each worker.
The exposure model developed by Certus also excluded
environmental exposures in calculating radon decay product exposure
to radon measurement technicians. In addition, exposure during
hours in the testing company office and in other structures other
than those tested was also included. Data for the exposure of 44
Florida Certified Radon Measurement technicians was recorded using
this model from January through November of 1989, The records were
reviewed to develop a realistic evaluation of this method of
exposure record keeping.
DEVELOPMENT OF THE EXPOSURE MODEL
Exposure of the radon tester is assumed to be from three
sources: the tested structures, the testing company office, and
other structures entered during the work day. Each tester was
required to provide an estimate of the number of hours per month
spent in the testing company office. An estimate of hours per
month spent in structures other than the office or structures
during radon testing was also made. For comparison, the tester
-------
provided an estimate of the percentage of time during a typical
month that was spent inside the home. An EPA protocol screening
test using 4" charcoal canisters (4) was conducted in the testing
company office and in the testers home. This data was recorded on
an Annual Exposure Opportunity Profile form.
TESTING EXPOSURE
Testing exposure values for each tester were derived from the
tests conducted. The total time for placement and retrieval of
test devices was assumed to average 0.5 hours per test. The time
estimate includes time spent with the costumer, data recording, and
selecting a test location. This time of exposure was applied to
each test even if there were multiple tests being conducted in the
structure. The example is assumed to over predict the exposure
time since less residence time in the structure is required to
place additional tests or quality assurance duplicates.
The pCi/1 values for the tests conducted by each tester over
the months of testing activities were summed to a variable T(i) for
each i" tester. The working level months (WLM) of exposure was
calculated assuming a 0.5 equilibrium value, 0.5 hours of exposure
per test, and 170 hours per month. The value for testing exposure,
Etbjt, for each tester, i, may be stated as
ETOM(i) - 1.47E-5 X T(i) (Eq. 1)
where:
E,„aT(i) is in WLM,
1.47E-5 is 0.5 hrs/(200 pCi/l/WL x 170 hrs/mo), and
T(i) is the sum of test values for tester i in
pCi/1.
OFFICE EXPOSURE
Exposure to radon decay products in the office environment is
calculated from the estimated hours in the office per month, the
office screening test value, and an assumed 0.5 equilibrium ratio.
Therefore, the value for monthly office exposure, EofPICE(i), for each
tester, i, may be calculated as
^omc« ( 1 ) = 2.94E-5 X TopriCi(i) X Rorrict(i) x N(i)
(Eq. 2)
where:
Eomci(i) is in WLM,
2.94E-5 is WLM/(200 pCi/1 X 170 hrs),
-------
Tofpice(i) is the test value for the office of tester
i in pCi/1,
Ron-Tcsfi) is the residence time in the office per
month for tester i in hours, and
N(i) is the number of testing months for tester i.
EXPOSURE IN OTHER STRUCTURES
Occupational exposure of radon testers occurs through out the
work day. This includes exposure while in structures other than
the office and structures being tested. This does not include non-
work time in the home or time not in structures, i.e. in transit.
The exposure from these other structures is based on the estimated
hours of exposure and an assumed maximum of 4 pCi/1 in these other
structures. The assumption of 4 pCi/1 as an average indoor level
in these structures is chosen for convenience due to lack of
sufficient data.
Exposure to radon decay products in other structures is
calculated from the estimated hours in the those structures per
month, the assumed value of 4 pCi/1, and an assumed 0.5 equilibrium
ratio. Therefore, the value for monthly other structure exposure,
EwHwCi), for each tester, i, may be calculated as
Eo—U) = 1 • 18E-4 x Rothi®( i) x N(i) (Eq. 3)
where:
EoWi) is in WLM,
1.18E-4 is 4 pCi/l/(200 pCi/l/WL x 170 hrs/mo),
Ronng.fi) is the residence time in other structures per
month in hours for tester i, and
N(i) is the number of testing months for tester i.
TOTAL OCCUPATIONAL EXPOSURE
The total occupational exposure, ETOrU(i), for each tester, i,
is the sum of the individual exposures and may be expressed as
~ E,n,OT(i) + E0PFICI( 1) + E0TKSR(i) (Eq. 4)
where:
Ew^i) is in WLM,
E
-------
COMPARISON HOME EXPOSURE
For purposes of comparison, the exposure model data included
an estimate of the percentage of time the tester spent in the home
each month and a screening test value for the home. Non-
occupational exposure is specifically excluded from consideration
in 10CFR20 and is so excluded in this model. However, a comparison
to an estimated home exposure is important in placing occupational
exposure in perspective.
The residence time in the home for tester i, RHOw(i), is
calculated as a percentage of 720 hours per month (30 days x 24
hours/day). The estimated home exposure for tester i, EHwls(i), may
be expressed as
®Ham(i) = 2.94E-5 x THaHI(i) x Rhomb(i-) x N(i) (Eg. 5)
where:
ehoi« (i) is in WLM,
2.94E-5 is WLM/(200 pCi/1 x 170 hrs),
Thow is the test value for the home of tester i in
pci/l,
R«o« is the residence time in the home of tester i
each month (720 hrs x percent in home/100) in hours,
and
N(i) is the number of testing months for tester i.
RESULTS OF MODEL CALCULATIONS
The results of calculating the various exposures according to
the Certus model are listed in Table 1. - Results of Exposure
Model. The number of months of testing and the total number of
tests conducted during that period are included in the table as an
indication of testing frequency. Each tester is identified by a
two letter, one number code. Those testers with the same two
letter code work for the same testing company. In some cases, the
testers shared placement and retrieval activities so that each
received one-half of the testing exposure.
For this sample of 44 testers, 100% of the testers had higher
estimated exposures at home than during testing. Home exposures
exceeded office exposures in 86% of the cases. There are no
testing companies in this stu4y whose sole business is radon
testing. The office exposure contains both radon testing office
time and other business office time.
-------
The results of this study indicate that the circumstances of
exposure to the sample group do not follow the typical exposure to
radiation workers. Exposures during testing activities are well
below that of non-occupational exposures. In the majority of cases
the exposure under normal working conditions in the office can
exceed testing exposure. Therefore, occupational radon decay
product exposure modeling should not be bound to the example of
occupational radiation exposure modeling for the nuclear industry.
TABLE 1. RESULTS OF EXPOSURE MODEL
ID
Etest
Eofpicb
Mother
E total
Ehqm
Mo's
NO.
AA1
0.0009
0.0334
0.1224
0.1567
0.6668
10
12
AA2
<10E-4
0.0006
0.2446
0.2452
0.0468
10
1
AB1
<10E-4
0.0070
0.1646
0.1716
0.0702
10
4
AC1
0.0002
0.0114
0.1164
0.1280
0.1106
11
9
AD1
0.0002
0.0090
0.2018
0.2110
0.0596
11
27
AD2
0.0002
0.0356
0.0672
0.1030
0.0442
11
15
AE1
0.0002
0.0106
0.0840
0.0948
0.0372
11
14
AF1
0.0003
0.0084
0.0300
0.0387
0.1652
10
39
AG1
0.0157
0.1010
0.1682
0.2849
0.7114
11
52
AH1
C10E-4
0.0052
0.1034
0.1086
0.0466
11
1
AK1
0.0002
0.0084
0.0306
0.0392
0.0476
5
18
AL1
0.0023
0.1362
0.3364
0.4749
0.9430
11
29
AMI
0.0007
0.0158
0.0706
0.0871
0.0858
10
47
AN1
0.0001
0.0238
0.0396
0.0635
0.0338
4
12.5*
AN 2
0.0005
0.0652
0.1086
0.1743
0.0396
11
43.5*
AQ1
0.0011
0.0264
0.0978
0.1253
0.1080
8
39
AS1
C10E-4
0.0072
0.1764
0.1836
0.0434
10
6
ATI
0.0009
0.0142
0.1364
0.1515
0.0710
10
72.5*
AT2
0.0015
0.0284
0
0.0299
0.0712
8
125
AT 3
0.0006
0.0156
0.1500
0.1662
0.0548
11
52.5*
AU1
<10E-4
0.0024
0.0396
0.0420
0.0254
8
2
AW1
0.0001
0.0024
0.0236
0.0261
0.0412
8
4
AM 2
<10E~4
0.0098
0.0470
0.0568
0.0214
8
1
AX1
0.0001
0.0137
0.0846
0.0979
0.0234
8
7
AY1
0.0004
0.0438
0.0672
0.1114
0.0294
11
22
AZ1
0.0005
0.0588
0.0303
0.0896
0.0318
5
16
BA1
<10E-4
0.0404
0
0.0404
0.0674
11
2
BB1
0.0125
0.1804
0.0044
0.1973
0.2224
11
235
BB2
0.0005
0.0592
0.0110
0.0707
0.0606
3
26
BC1
0.0001
0.0104
0.0988
0.1093
0.1048
10
8
BC2
<10E-4
0.0084
0.0790
0.0874
0.0168
8
2
BD1
0.0005
0.0024
0.1086
0.1115
0.0364
11
33
BE1
0.0006
0.0154
0.0564
0.0724
0.0704
8
19
BE2
0.0004
0.0308
0.0276
0.0586
0.3024
9
19
BE 3
0.0004
0.0078
0.1412
0.1494
0.2994
8
20
BF1
<10E—4
0.0014
0.0198
0.0212
0.0152
2
5.5*
BF2
0.0001
0.0012
0.0592
0.0605
0.0458
3
9.5*
-------
TABLE 1. RESULTS OF EXPOSURE MODEL (CONT.)
ID
EteST ^office
Mother
®*TOTAL
MO'S
No.
BG1
0.0002 0.0046
0.0494
0.0542
0.0562
10
9
BH1
<10E-4 0.0452
0.0988
0.1440
0.0402
10
l
BH2
0.0001 0.0302
0.0988
0.1291
0.1157
10
8
BJ1
0.0003 0.0244
0.0246
0.0493
0.0234
10
32
BJ2
<10E-4 0.0244
0.0246
0.0490
0.0234
10
2
BK1
0.0001 0.0062
0.2470
0.2533
0.0266
10
5
BK2
<10E-4 0.0122
0.2752
0.2874
0.0698
10
3
* Tests
i placed by one
tester
and retrieved by
another.
ACKNOWLEDGEMENTS
The author wishes to thank Certus Laboratories, Inc. and it's
president, Thomas R. Stephenson, in allowing the author to examine
the health and safety records of the testing technicians in that
program to be able to prepare this paper.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
REFERENCES
1. Rules of the State of Florida Department of Health and
Rehabilitative Services Chapter 10D-91, Florida Administrative
Code, Control of Radiation Hazards.
2. Certus Laboratories, Inc., "Health and Safety Plan", Atlanta,
Georgia, January l, 1989, Rev. 1.
3. Chapter 10 of the Code of Federal Regulations Part 20, Standards
for Protection Against Radiation.
4. "Indoor Radon and Radon Decay Product Measurement Protocols,"
USEPA, EPA 520/1-89-006, March, 1989.
*U.S. MMKNHENT PRINTING OFFICE: 1990 748'OU/ZSOQl
------- |