UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
'"" WASHINGTON. D C 20460
SAB-RftC-88-009
Jamiary 14 r 1988
THE AOMIN1STR A TQ!=>
Honorable Lee M, Thomas
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
Dear Mr. Thomas:
The Science Advisory Board's Radiation Advisory Committee has
completed its review of the Office of Environmental Engineering and
Technology Demonstration's (OEETD) rad'on mitigation research program.
In response to OEETD's request of May 5, 1987, The Radon Mitigation
Subcommittee met publically October 13-14, 1
-------
effectiveness. The Committee recommends that the differing needs of
mitigators, homeowners, and policy makers be addressed and that total
lifetime costs of each mitigation technique be estimated as accurately as
possible, and reported as concisely as possible.
The Committee appreciates the opportunity to review this important
research program and requests that the Agency formally respond to our
scientific advice.
Sincerelyj
Norton Nelson, Chairman
Executive Committee
Science Advisory Board
William J. Schull,
Radiation Advisory
Science Advisory Boa
Keith Sthiager, Chairman
Radon Hitigation Subcommittee
Radiation Advisory Committee
Science Advisory Board
cc: J. Skinner
S. Meyers
-------
SAB-RAC-88-009
Review of the
Radon Mitigation Research Program
A Report of the
Radiation Advisory Committee's
Radon Mitigation Subcommittee
U*S» Environmental Protection Agency
Science Advisory Board
Washington, D.C«
January 1988
-------
FPA NOTICE
This report has been written as part of the activities of the Science
Advisory Board, a public group providing extramural advice on scientific
information and advice to the Administrator and other officials of the
Environmental Protection Agency. The Board is structured to provide a
halanced expert assessment of scientific matters related to problems
facing the Agency. This report has not been reviewed for approval by the
Agency, and hence, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorsement or
recommendation for use.
-------
U.S. ENVIRONMENTAL PROTECTION AfiENCY
SCIENCE ADVISORY BOARD
RADIATION ADVISORY COMMITTEE
RADON MITIGATION SUBCOMMITTEE
ROSTER
CHAIRMAN
Dr. Keith Schiager, Orson-Spencer Hall, Room 100, University of Utah,
Salt Lake City, Utah fl4ll?
MEMBERS/CONSULTANTS
Dr. Michael R, Brainbley, Campus Box 1106, One Rrookings Drive
Washington University, St. Louis, Missouri 63130
Dr. Michael,Ginevan, Environ Corporation, 1000 Potomac Street, NW
Washington, DC 20007
Or. Thomas F, Gesell, U.S. DOE Idaho Operations Office, 785 DOE Place,
Idaho Falls, Idaho 83402
Dr. fierald, P. Nicholls, Radiation Protection Programs, CN411,
New Jersey Department of Environmental Protection,
Trenton, New Jersey 08625
Mr. William Turner, Harriman Associates, 292 Court Street, Auburn,
Maine 04210
EXECUTIVE SECRETARY
Kathleen W. Conway, Deputy Director, Science Advisory Board
U.S. Environmental Protection Agencyt 401 M Street, S.W.,
A101-F, Washington, D.C. 20460
STAFF SECRETARY
Dorothy M. Clark, Secretary, Science Advisory Board, U.S. Environmental
Protection Agency, 401 M Street, S.W., A101-F, Washington, D.C. 20460
-------
Table of Contents
PAGE
INTRODUCTION ....*....*...... 1
MITIGAT!ON MATRIX. ,,,..,........ I
DATA COLLECTION AND PRESENTATION..... ... 2
DATA ANALYSIS., » ........ 3
COST- EFFECTIVENESS. ....... 4
LITIGATORS, ............. **........ 5
HOMEOWNERS ... 5
POLICY MAKERS AND ANALYSTS.......... 6
-------
Introduction
The Radon Mitigation Subcommittee of the Radiation Advisory Committee
has reviewed the EPA Office of Environmental Engineering and Technology
Development's Program Description and Plans, March 9, 1987; the IA8
report, CollecTfon ojf Field Data, September 18, 1987; the technical
guidance manual Radon Reduction Techniques for Detached Houses (2nd
edition), September 11, 1987V and a report, Data Analysis In EPA1s Rad_gn_
Reduction, Technology/Demonstration Program, September 21, 1987.
In response to specific requests by the Agency, the Subcommittee is
presenting observations and suggestions on the following topics:
(1) the variables addressed by the mitigation test matrix,
(2) the collection and management of useful, high-quality data,
(3) the strategy and methodologies for data analysis for the two
separate data sets, e.g. the mitigation effectiveness data
(also referred to as the general data) and the time-series data
(also referred to as the detailed data), and
(4) the definition and application of cost*effectiveness as an
objective of the mitigation research effort.
The Subcommittee was very favorably impressed with both the quantity
and quality of the Agency's efforts in mitigation of indoor radon exposures,
The expanding public demands for guidance on needs for mitigation and on
techniques, efficacy and costs of meeting those needs, is a challenging
task and the Agency is pursuing its share of the task with vigor. The
Subcommittee's recommendations are directed toward strengthening the few
weaker areas in an otherwise very strong program.
litigation Matrix
In 1986 the Office of Environmental Engineering and Technology
Demonstration (OEETO) requested comments on the proposed matrix for
conducting radon mitigation test projects. The Radiation Advisory
Committee in its report (January 1987, SAB-RAC-87-016) recommended
that OEETD reduce the number of cells in the matrix on the basis of the
physical principles involved. The matrix subsequently described in
Program Description and Plans, March 9, 1987, has not been reduced
¥cc"ording to those principles. The Radon Mitigation Subcommittee
believes that although it may be desirable to collect data according to
the existing matrix, analyzing mitigation results within the context of
broad physical characteristics could help to generalize and extrapolate
the data to new situations. The Subcommittee recommends that the EPA
continue its efforts to consolidate the matrix by combining cells with
-------
- 2 -
common characteristics. Split-level houses are actually combinations of
basement a*^»»lab-on-grade construction. As a further example of possible
consolidatlW, it would appear that techniques such as increased ventilation,
participate removal, gas removal, and water treatment could be tested in
one or two of the most common substructure types rather than six to ten
as indicated in the matrix,
Data Collection and Presentation
EPA must impose certain quality standards for data collection and
presentation upon their contractors, or be prepared to reformat and verify
the quality of the data prior to public release. Use of a common computer
data-base-management system by the various contractors would also be
helpful.
With regard to data collection EPA should consider further
standardization of the data collected in the several non-data-lntensive
programs to include comparable before and after radon measurement methods
and relevant engineering details such as:
(1) pressure field developed in block wall venting installations
(value for each wall is suggested),
(2) pressure field developed In slab venting installations (four values
in four separate areas of the slab are suggested)t
(3) fan flow (measured),
(4) pressure differential at the fan,
(5) fan power consumption (measured),
(6) percent time fan is "on" if not continuous, and
(7) estimated annual operating cost of system based on measured
power consumption.
These seven items, currently not being collected, would require
minimal additional effort after the completion of the mitigation effort
and be valuable in addressing system performance criteria and cost-
effectiveness.
With regard to data presentation, as part of a quality control
program, EPA should consider further standardization of measurement,
display of uncertainties on graphics, and consistent labeling of graphs
and tables. Many of the preliminary graphics presented to the Subcommittee
could be improved by more descriptive labeling or by use of different
formats. For example, a scatter plot of radon concentration before
-------
- 3 -
"if.
mitigation (j^ixis) versus concentration after mitigation (y axis) would
provide a simple and direct indication of whether mitigation generally
reduced concentration to a uniformly low level or by a relatively consistent
fraction of the initial value,
Hata Analysis
The Subcommittee reviewed the objectives of the extensive time series
data being collected from the group of houses in the Piedmont area, as
well as the analytical and modeling methods employed by the Agency. The
data consist of 30-minute interval measurements of radon concentrations
and several environmental variables such as barometric temperature and
wind speed. The Subcommittee agrees that although studies of this type
would Nave little practical application in simply evaluating the effectiveness.
of mitigation efforts, they may be useful in developing a better understanding
of the dynamics of radon concentration in dwellings. For evaluation of
effectiveness, comparable pre- and postmitigation measurements integrated
over appropriate time intervals are adequate.
The Subcommittee suggests that the time-series data might be used
effectively to determine the optimum time interval for the pre- and
postmitigation measurements. The Agency staff presented a graph illustrating
the correlation between radon measurements integrated over 2 days and 4
days (not derived from the time-series data). The Subcommittee found
this correlation to he of limited interest, but felt that the real issue
is the relative variance of average radon concentrations determined by
measurements integrated over various time intervals, e.g. 2, 4, 8 or 16
days. Such averages could be obtained from the time-series data by an
appropriate random-sampling scheme. Identification of the shortest
interval that has an acceptably small variance would be an important
input to the mitigation testing protocol. These various estimates could
also be used to develop uncertainty bounds for reporting the effectiveness
of mitigation.
The Subcommittee questions the value of starting the analysis
by fitting a full blown auto-regressive integrated moving average (ARIMA)
time-series model. More immediately informative might be analyses of
serial correlations (correlations over time) among the various variables
and the lag time that maximizes the correlations. (A lagged correlation
coefficient 1s one calculated between variable 1 at time t and variable 2
at time t - n, where n is the number of intervals in the time difference.)
Autocorrelation of radon concentrations must also be considered in the
analysis.
Cross validation is also an analytical area of concern to the
Subcommittee. In a typical cross validation, the data are split into two
parts. Then the first part is used to fit a model, and the fitted model
-------
_ 4 -
is used to predict the second part. If the second part-prediction is a
success, confidence in the general liability of the model is increased.
In some time-series analyses, the model is cross validated iteratively,
i.e., it is fit to half,the data, then tested on the other half. If this
procedure is a failure, the model is refitted to the original data and
retested. After several such iterations, a model which "cross validates"
well is obtained. Such exercises are of questionable validity. A
discussion of exactly what was done in cross validating the radon time
series data would be helpful.
Cost- Effect 1 ve_ness
One of the issues referred to the Science Advisory Roard for comment
in Or. Skinner's memorandum of May 1987, was the Agency's definition of
the term "cost*effectiveness."
The Subcommittee believes that the goal adopted by OEETD of developing
cost-effective mitigation techniques is preferable to emphasis on development
of low cost techniques, since both cost and performance are important for
decisions concerning mitigation. In order to accomplish this goal, OEETD
should expand and refine its working definitions of cost-effectiveness
for radon mitigation as soon as possible. The two measures of mitigation
effectiveness currently used by the Agency, i.e. the percentage reduction
in the average indoor radon concentration and the final value of the
concentration, are appropriate for certain applications of cost-effectiveness,
as illustrated later. However, other applications of cost*effectiveness
evaluations would benefit from other measures of effectiveness, e.g. the
absolute reduction of the radon concentration, which might also be expressed
in terms of risk.
As applied by EPA in other situations, cost*effectiveness has been a
comparison of the costs of achieving equal results by use of different
engineering or equipment elements; or, conversely, a comparison of the
magnitude of results by use of different control elements of equal costs.
An appropriate measure for radon mitigation, for example, might be performance
per dollar of total cost of mitigation. However, the Subcommittee believes
that the factors that influence the use of cost-effectiveness data
may be at least as important (If not more important) as the specific
definition {or units) used for its presentation.
Because mitigation requires decisions by different groups that
participate 1n the mitigation process, three specific definitions may be
required for this program, one for the perspective of each of the three
primary use groups: (1) mitigators, (2) homeowners, (3) policy makers and
analysts (e.g., legislators, risk analysts and managers). Each of
these interested groups—mitigators, homeowners, and policy makers—need
realistic projections of total costs of mitigation, including diagnostic
-------
- 5 -
and verification measurements, equipment and labor for the initial
installatlflp., and lifetime operational and maintenance expenses (including
such costs-*!*- increased energy consumption), Therefore, the Subcommittee
recommends that the total lifetime costs of each mitigation technique be
estimated as accurately as possible and reported as concisely as possible.
A brief discussion of some other specific suggestions follows,
MltigatOrs, The mitigator must be able to predict which techniques
or .combination of techniques are most likely to achieve the reduction
desired by the homeowner (client) at minimal or nominal cost. The
nitigator also must know the probability of success (or failure) to
allow for liability or warranties that may be demanded by the homeowners.
As a result, an appropriate measure of cost-effectiveness for the mitigator
might be the ratio of expected reduction in radon level to total cost for
each radon mitigation technique. However* because the reduction achieved
depends not only on the mitigation technique, but also on the initial
radon concentration, OEETD should develop a. method that somehow accounts
for the initial concentration. The simplest approach may be to estimate,
the cost effectiveness for categories of pre-mitlgation concentration
(e.g., for 4-20 pCi/L, 20-100 pd/L, and more than 100 pCi/L).
The use of an expert system, or similar decision tree techniques, to
guide the mitigation efforts to those most likely to be cost-effective
should be considered. For example, diagnostic costs could be minimized
by mitigation guidance based on observed substructure features, and
expected performance results. Separate guidance could be provided for
hones with, say, more than 20 pCi/L, and those with less than 20 pCi/L.
This guidance can be generated from Che results of current demonstration
programs, and the strategies should then be prioritized.
Homeowners, The homeowner, as the consumer of a service, may be
interested in one, or both of two different concepts of effectiveness. In
connection with real estate transactions, homeowners are concerned with
maintaining the market value and marketability of their homes when elevated
radon levels are found. They are most likely to be interested in the
effectiveness of mitigation in achieving the EPA-recommended guideline of
4 pCi/L or less. In spite of EPA efforts to emphasize the flexibility of
the guideline, it is unfortunately being treated wore and more as a
boundary between safe and unsafe* Insufficient data at present preclude
evaluation of the effect of radon mitigation Installations on market
value of homes, but a measure of cost-effectiveness could guide selection
of mitigation techniques*
Owners of homes, faced with high radon concentrations but not intending
to sell a home may be more interested in the effectiveness of mitigation
for reducing health risks to family members* These homeowners may be
more receptive to concepts of risk than to EPA numerical recommendations.
For such homeowners, cost-effectiveness cannot be separated from the
concept of cost-benefit, i.e., what health benefit or risk avoidance can
-------
- 6 -
be obtained.ffcr a given expenditure on mitigation.. For such a homeowner
to make t'OtiftmmtL choices, the expected reduction in radon exposure should
be expressed In absolute rather than relative values. A useful definition
of cost-effectiveness for these homeowners might be the same as that
suggested for nitigatots* However, to facilitate interpretation by
homeowners, this form of cost-effectiveness might be converted to the
decrease in risk or expressed in terns of the statistical increase in
life expectancy per total dollar invested over the life cycle. In any
case, the definition adopted should help homeowners decide what level of
mitigation is desirable and how likely various mitigation techniques are
apt to be helpful. Allowance needs to be made for differences among
homeowners in willingness to take risks and recognizing the difficulty in
reducing extremely high concentrations to a prescribed low level (e.g., 4
pCi/L).
Pol_i_cy_Jiakers and Analysts* Policy makers analyze the total national
benefit, or risk reduction, expected for a given level of national (or
state) effort toward mitigation. The following are examples of the types
of questions related to cost-effectiveness that would involve national
policy. If the current distribution of indoor radon concentrations
causes 5000 - 20,000 cases of lung cancer per year* how much will this
number be reduced by a national effort based on voluntary participation
and a 4 pCi/L guideline? How much different would the results be if the
guideline was higher or lower? How different would the results be if the
current guideline was applied only to new construction and a different
guideline was applied to existing housing stock?
To help answer these questions* OEETD should consider developing a
method for determining Che decrease in total population risk per total
dollars spent on mitigation, including costs of research, development and
demonstration, diagnostics, installation, and operation of mitigation
measures. This definition might be applied to the estimated distribution
of radon concentrations in U.S. homes to determine which levels are cost-
effective from a societal perspective.
------- |